Student-Faculty Programs 2011 Abstract Book - Summer ...
Student-Faculty Programs 2011 Abstract Book - Summer ...
Student-Faculty Programs 2011 Abstract Book - Summer ...
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
<strong>Student</strong> <strong>Student</strong>-<strong>Faculty</strong> <strong>Student</strong> <strong>Faculty</strong> <strong>Programs</strong><br />
<strong>2011</strong> <strong>Abstract</strong> <strong>Book</strong>
STUDENT-FACULTY PROGRAMS<br />
<strong>2011</strong> <strong>Abstract</strong> <strong>Book</strong><br />
This document contains the abstracts of the research projects<br />
conducted by students in all programs coordinated by Caltech’s<br />
<strong>Student</strong>-<strong>Faculty</strong> <strong>Programs</strong> Office for the summer of <strong>2011</strong>.<br />
Table of Contents<br />
<strong>Summer</strong> Undergraduate Research Fellowships (SURF) 1<br />
MURF Undergraduate Research Fellowships 113<br />
Amgen Scholars Program 121<br />
Laser Interferometer Gravitational-Wave Observatory (LIGO) 129<br />
NASA/JPL <strong>Programs</strong> 139<br />
� Planetary Geology and Geophysics<br />
Undergraduate Research Program (PGGURP)<br />
� The National Space Grant College<br />
and Fellowship Program (Space Grant)<br />
� Undergraduate <strong>Student</strong> Research Program (USRP)<br />
� JPL <strong>Student</strong> Internship Program (JPLSIP)
SURF<br />
SUMMER UNDERGRADUATE<br />
RESEARCH FELLOWSHIPS<br />
S<br />
U<br />
R<br />
F
In-Plane Thin Film Lattices: Thermally Stable Mirrors<br />
Deeksha Agrawal<br />
Mentors: Chiara Daraio and Eleftherios Gdoutos<br />
In devices that experience large temperature changes such as high speed aerospace vehicles, substantial thermal<br />
stresses develop when high temperature components are connected to lower temperature structures. This leads to<br />
failure by yielding, fracture or fatigue, as well as the formation of gaps that require sealing and extreme forces on<br />
attachments to other structures. To suppress such failures, we fabricate and study expansion thin film bimaterial<br />
lattices with tunable thermal expansion. The lattice configuration is composed of Aluminium and Titanium which<br />
have widely different individual coefficients of thermal expansion (CTEs) and intervening spaces. According to<br />
theoretical predictions the higher CTE internal Aluminium expands more rapidly and pushes outward against the<br />
lower CTE exterior lattice of Titanium. This causes the latter to rotate, thereby counteracting the lengthwise<br />
elongation of the material. We fabricate our lattice structure on a silicon wafer, coat the wafer with photoresist,<br />
expose with UV and create the pattern, coat with Aluminium by E-beam evaporation, lift off the photoresist and<br />
then iterate the procedure with Titanium. We then etch the sacrificial gold layer to separate the Al-Ti lattice. We<br />
use techniques like Finite Element Analysis using ABAQUS 6.9-2and X-ray diffraction to test the various properties<br />
of this lattice.<br />
Thermal Analysis of the Solar Decathlon House<br />
Sara Ahmed<br />
Mentor: Melany Hunt<br />
A group of Caltech and SCI-Arc students have teamed together to build an energy efficient, innovative and most<br />
importantly a Net-Zero home. To achieve this goal, we are using solar energy for generating electricity and day<br />
lighting as well as reducing energy consumption throughout the home. The thermal load is a major factor of energy<br />
use, including the heat emitted by lights, sunlight through windows, heat transfer from the environment to the<br />
house, domestic hot water needs and heat absorbed by PV (photovoltaic) panels.<br />
During my SURF I have been working on the thermal analysis and building of our solar house, CHIP. We have<br />
integrated our HVAC (heating, ventilation, air conditioning) unit and hot water system which are responsible for all<br />
the heating and cooling needs of the house. We have accomplished this by utilizing the waste heat of the outdoor<br />
unit of the HVAC unit to preheat the domestic hot water. For optimizing the operation of the integrated heat<br />
exchange system, we have done comprehensive temperature control and dehumidifying tests in the house.<br />
Through iterative tests and analysis, we have significantly improved its performance. I have also researched on<br />
improving the energy efficiency of the house using heat sinks to the PV Solar panels and light fixtures.<br />
C-H Bond Activation Using Organometallic Iridium Metal Complexes<br />
Tonia Ahmed<br />
Mentors: John Bercaw and Ian Tonks<br />
A large part of petroleum used in today’s society consists mainly of methane. However, few methods have been<br />
found to efficiently and selectively convert methane to other valuable products. Moreover, because of the cost and<br />
difficulty of transporting gases, oil companies generally resort to burning methane gas instead of utilizing it. As an<br />
alternative, methane could be converted to liquid methanol through selective C-H bond activation. C-H activation<br />
was studied in this project using a variety of iridium hydroxyl complexes and observed using methods of nuclear<br />
magnetic resonance (NMR) spectroscopy and X-ray crystallography. From the data provided by these methods, the<br />
C-H activation of a variety of substrates was confirmed. To determine the mechanism of the reactions, several<br />
kinetics experiments were performed also using NMR spectroscopy. The rate determining step was determined to<br />
be substrate coordination based on kinetic and Eyring analysis.<br />
The Impact of High Frequency Trading on Correlations<br />
Emanuelle Alm<br />
Mentor: Benjamin Gillen<br />
From the onset of the financial crisis through late 2010, soaring correlations were observed in the financial<br />
markets. Since 2010, correlation levels have gone down to some extent. However, these high correlations have<br />
persisted longer than expected given the current macroeconomic environment and incipient recovery. The<br />
expansion of high-frequency trading (HFT) activity is often mentioned as one possible cause of this persistence,<br />
likely due to significantly changes in the underlying market microstructure. HFT has experienced strong growth<br />
over the past several years and makes up a large amount of the trading volume in the Unites States. This paper<br />
empirically identifies and evaluates causal implications unique to the hypothesis that HFT is driving increased<br />
correlations. We present a mathematical model that illustrates the mechanism by which HFT would drive increase<br />
correlations and empirically test the model. Our analysis uses data on 23 Exchange Traded Funds to compare the<br />
ratio between short-term and long-term correlation for a period with low incidence of High Frequency Trading and a<br />
1
period from latter years when High Frequency Trading is considered to make up a notable amount of the trading<br />
volume. While most of the preliminary data preparation and analysis is complete, much formal statistical<br />
estimation and testing remains to be performed over the following month.<br />
Mechanical Properties of Nanocrystalline Ni Nanopillars<br />
Eli E. Alster<br />
Mentors: Julia Greer and Jie Lian<br />
Materials sized on the order of nanometers possess mechanical properties far divergent from their bulk analogs. In<br />
order to quantify the effects of microstructural dimension and external scale on materials’ mechanical properties,<br />
both grain size and sample size were varied across the nanometer range and plotted against yield strength to help<br />
form a three dimensional plot. Fabrication of the nanopillars was accomplished through electrodeposition of Ni in<br />
template microarrays and mechanical properties tested through uniaxial compressions by a Hysitron nanoindenter<br />
with a flat punch tip. Original and post-deformation images were taken with either the Nova-200 or Nova-600<br />
scanning electron microscope (SEM). Grain size was determined through the lift-out procedure and transmission<br />
electron microscopy (TEM). Molecular dynamics simulation work was also undertaken using the LAMMPS software<br />
package to investigate the nucleation and movement of dislocations, which are difficult to capture from<br />
experimental observations. These efforts were designed to gain insight into the not yet understood mechanisms<br />
which drive deformation at nanoscale.<br />
Optimization of Dreiding Hydrogen Bond Parameters<br />
Anthony Alvarez<br />
Mentor: William A. Goddard III<br />
In chemistry, accurately modeling small systems on the order of a few dozen atoms can be done using quantum<br />
mechanics. For large systems, such as proteins, these simulations take an incredibly long time. To be able to model<br />
these systems quickly and accurately a method called force fields is used rather than quantum calculations to<br />
model these structures. However, these force fields require optimized parameters to be able to accurately<br />
reproduce the atomic level interactions being described by quantum mechanics. One force field, Dreiding, has<br />
parameters optimized for high dielectrics, like water, but does poorly in low dielectric environments, such as a<br />
protein. Hydrogen bond parameters are particularly affected by the dielectric change. Since hydrogen bonds play a<br />
large role in the folding of proteins those parameters are particularly important. This project is aimed at obtaining<br />
high level quantum mechanics data on a set of molecules that represent the interactions found in a protein and use<br />
that data to fit the Dreiding hydrogen bond parameters. Once the parameters are optimized, Dreiding will be able<br />
to more accurately describe proteins and their interactions with ligands.<br />
Hyperbolic Versus Euclidean Embedding of Real-World Networks<br />
Sadaf Amouzegar<br />
Mentors: Babak Hassibi and Elizabeth Bodine-Baron<br />
Recent research has suggested that social networks can be described in terms of a hidden metric space, which<br />
often can be used to improve applications requiring navigation, such as routing or search. Different algorithms<br />
exist to embed networks in either Euclidean or hyperbolic space – in this work we evaluate two representative<br />
algorithms to determine the optimal embedding for various types of networks. We use Landmark Multidimensional<br />
Scaling as representative of Euclidean embedding. This method measures the length of the shortest path between<br />
any two nodes in the network and assigns a coordinate to every node in Euclidean space based on this distance.<br />
However, since Euclidean space has a geometric growth rate and most real-world networks exhibit low diameters,<br />
implying exponential growth, we also employ Crovella’s “online greedy graph algorithm” as representative of<br />
hyperbolic embedding. This method uses a spanning tree algorithm to calculate the appropriate coordinates for<br />
each node in hyperbolic space. To determine which type of hidden space is optimal for different types of networks,<br />
we evaluate these algorithms on both real world and online social network data.<br />
Supersonic Relative Velocities of Baryonic Fluids and Dark Matter<br />
Nikhil Anand<br />
Mentor: Christopher Hirata<br />
At the epoch of recombination, the sound speed of the Universe dropped precipitously from a relativistic limit<br />
($\sim c/\sqrt{3}$) to that of the thermal RMS velocities of hydrogen atoms ($\sim 10^{-5} c$). Because the<br />
dark matter does not suffer Thomson scattering, the relative velocities between the baryonic fluids and the dark<br />
matter becomes supersonic (coherent over several Mpc, comoving), effectively raising the Jeans mass necessary to<br />
form the earliest structures. This delays the growth and population of these structures by an amount that depends<br />
on the magnitude of the relative velocity. We examine the effect of supersonic relative velocities, particularly in<br />
regards to the galaxy power spectrum and its consequences for future galaxy surveys and BAO measurements.<br />
2
Contribution of Organosulfur Compounds to Atmospheric Aerosol in the United States<br />
Avin Andrade<br />
Mentors: John Seinfeld and ManNin Chan<br />
Atmospheric aerosol—tiny particles suspended in the air—forms from a variety of sources such as vegetation,<br />
transportation, industry, agriculture, and urban land use. It not only poses serious problems to human health but<br />
also has an effect on weather and climate. By understanding its properties and composition, we can determine the<br />
most effective means of minimizing its production. Organosulfates (R’OSO4) and nitrooxyorganosulfates<br />
(O2NOR’OSO4) compose a new and important class of organosulfur and organic compounds that has recently been<br />
detected in laboratory generated aerosol and ambient aerosol. Its importance can be estimated by looking at the<br />
difference in concentration of total sulfur and inorganic sulfur. Using data from the the Environmental Protection<br />
Agency's Interagency Monitoring of Protected Visual Environments (IMPROVE) database it has been shown that<br />
organosulfur makes up a significant contribution to aerosol composition across the United States. The<br />
concentrations are on the order of 0.1 ug/m 3 or around 10 percent of the total organic carbon concentration. These<br />
numbers vary significantly by season and year with the highest concentrations occurring during summer and the<br />
years 2005 and 2006.<br />
The Firefighter Problem on Planar Graphs<br />
Zhi Yuan Ang<br />
Mentor: Niranjan Balachandran<br />
Let G be a connected graph. A fire starts out at a vertex v. At each discrete time step, k firemen are placed on k<br />
vertices that are not on fire and these vertices are being protected permanently. The fire then spreads from a burnt<br />
vertex to all its unprotected neighbours and the process repeats itself. Denote snk(v) as the maximum number of<br />
vertices that can be saved if the fire starts out at v. The k-surviving rate ρk(G) of G is defined to be<br />
k (v) / n 2 , which is the average proportion of saved vertices. Let C be the class of all planar graphs. In this<br />
report, we investigate the infimum of ρ3(G) taken over all G C. In other words, we want to know if 3 firemen<br />
suffice to save a good proportion of the graph. The ultimate goal is to solve or refute the conjecture that 2 firemen<br />
would suffice.<br />
Condensation of Water on Superhydrophobic Carbon Nanotube Arrays<br />
Patrick Arran<br />
Mentors: Mory Gharib and Indrat Adrianos<br />
It was thought that superhydrophobic vertically aligned carbon nanotube arrays would, uniquely among known<br />
superhydrophobic materials, not be subject to condensation of water. This hypothesis was made on the basis of the<br />
surface feature scale being orders of magnitude smaller than that for previously investigated materials. The study<br />
did not attempt to characterise the mechanisms of condensation, merely to observe the macro-level results. CNT<br />
array samples that were treated to be either superhydrophobic or hydrophilic along with traditional control samples<br />
such as polymers, glass and metals, were cooled with liquid nitrogen and exposed to humid air. The subsequent<br />
phenomena were recorded as video and results compared. The details of the methodology have evolved over the<br />
course of the study. Preliminary results indicate much smaller amounts of condensed water and much faster<br />
evaporation times for the nanotubes than for the control materials. It is hoped that more precisely controlled<br />
experiments will provide a more quantifiable, consistent result and an understanding of the effect of varying<br />
relative humidity. It is also hoped to more accurately determine the effect of the CNT treatments, which have so<br />
far made negligible differences.<br />
Data Reduction and Analysis Software for Robo-AO LGS Adaptive Optics System<br />
Ankit Arya<br />
Mentors: Christoph Baranec and Reed Riddle<br />
I developed and coded a library of reduction and analysis software for the telemetry system of a robotic adaptive<br />
optics telescope system (Robo-AO). In addition, I created a status page for the system, which helps the engineers<br />
and astronomers to monitor the overall functioning of the system instruments. The status page represents live<br />
information about the telescope instruments in the form of live charts and 2D heatmap images, making it a very<br />
useful tool for debugging and for representing errors quickly. First, the analysis software finds the minimum,<br />
maximum and median values for the deformable mirror actuator positions as well as the focus shape of the mirror.<br />
It also calculates the root mean square and leaky average of the input data. The results of calculations are written<br />
to data files, which are read by the status page asynchronously to display it live. The system code is in C++, so I<br />
used the same to integrate my library within the software. The status page makes use of scripting languages like<br />
JavaScript, jQuery, JSON, AJAX and other plotting API to represent information. All API used to design the page are<br />
open source to help the distribution of the software publically at a later stage.<br />
3
Analysis of Interpolation Schemes in the Quasicontinuum Method<br />
Hyunji Bae<br />
Mentors: Michael Ortiz and Malena I. Español<br />
The quasicontinuum method is used to effectively and accurately approximate the mechanics of solid materials at<br />
the atomistic level. The interpolation scheme is an important part of the method. The purpose of this research is to<br />
investigate the effect of the interpolation schemes in the quasicontinuum method. Specifically, we compare<br />
different interpolation schemes and the effect it has on approximating energies and displacement fields. We<br />
present numerical examples in the one-dimensional case.<br />
High Performance DSP on FPGAs Using Python<br />
Amit Bansod<br />
Mentors: John Carpenter and Glenn Jones<br />
The CASPER tools provide a design flow to rapidly develop complex digital signal processing (DSP) systems for the<br />
radio astronomy instrumentation using Field Programmable Gate Arrays (FPGAs). CASPER libraries provide high<br />
performance DSP functionality for current radio astronomy applications but require further optimizations to<br />
efficiently implement future wide bandwidth instrumentation like the CARMA (A combined array of 23 antennas<br />
with a maximum bandwidth of 8 GHz) correlator. In this project, it is attempted to develop a better toolflow using<br />
a python library called MyHDL which can use the existing library of optimized functions and an architectural<br />
paradigm that works well for radio astronomy applications. To test the functionality of MyHDL, a Biplex pipelined<br />
Fast Fourier Transform (FFT) function block was designed using MyHDL. Performance of the designed block was<br />
then compared with the CASPER FFT block. The results showed a better performance by the MyHDL block in terms<br />
of the maximum operating clock frequency and a similar resource usage as that of CASPER design. By making a<br />
user friendly design toolflow, MyHDL can be a promising tool for future radio astronomy instrumentation.<br />
Clamping Losses in Nanomechanical Resonators<br />
Aniruddha Bapat<br />
Mentors: Michael Roukes and Luis G. Villanueva<br />
The field of nanotechnology is an active area of research today. One important application of nanotechnology is the<br />
field of Nanoelectromechanical systems, or NEMS. The focus of this project is to study a particular loss mechanism,<br />
the clamping loss mechanism, in NEMS resonators by simulation using the software COMSOL and to try to minimize<br />
this loss by varying the clamping geometry. The simulation was used to calculate the ratio of strain energies in the<br />
resonator and in the clamps, for the desired eigenmodes. This parameter is proportional to the clamping-based<br />
Quality factor Q of the system. Different resonator geometries, such as square, disk and ring bulk and rectangular<br />
flexural resonators are analyzed versus changes in dimensions and clamp geometry.<br />
Analysis of a Crater Detection Algorithm Applied to the Moon<br />
Allison Haley Barnes<br />
Mentors: Yang Cheng, Matthew Golombek, and Fred Calef<br />
This project aims to analyze and optimize the result of an algorithm that fully automates crater detection on the<br />
Moon. The algorithm uses images taken by two Narrow Angle Cameras (NACs) on the Lunar Reconnaissance<br />
Orbiter (LRO) to detect a large number of craters (~200,000 craters per image). The Moon has the potential to aid<br />
further space exploration, reveal information about Earth, and provide resources. Impact cratering produces the<br />
majority of observed surface morphologies, reveals the age and composition of surfaces, serves as a mechanism<br />
for deposition and removal of material, and provides information about the solar system. The objective of this<br />
project is to develop an automatic evaluation procedure for the algorithm’s data product and apply this procedure<br />
to all of the NAC images of the Apollo 15 landing site. Accomplishing this objective is significant, as validating an<br />
automated procedure that detects craters will save researchers the arduous effort of manually measuring craters<br />
while also providing insight into the relationship between craters and their surrounding terrain.<br />
Synthesis of an Asymmetric Bimetallic Complex for Carbon Dioxide Reduction<br />
Marissa Barrientos<br />
Mentor: Jonas Peters<br />
The goal is to synthesize a symmetric bimetallic complex that can electrocatalytically reduce carbon dioxide. The<br />
Peters group has envisioned an asymmetric heptadentate, monoanionic dinucleating ligand based on a μ-phenolate<br />
scaffold. A soft site is created using a tris-chelating diphosphino amine “arm” which may be able to stabilize a lowvalent,<br />
nucleophilic metal center that can potentially bind to the carbon atom of carbon dioxide. A dipyridal amine<br />
“arm” serves as a hard site supporting a Lewis acidic metal center that can facilitate the binding of carbon dioxide<br />
through coordination of an oxygen atom. The ligand precusor HL py, Cl was synthesized from a known synthetic<br />
procedure within the group. After attaching the BH3 protected diphosphino amine “arm,” several deprotection<br />
routes were explored. After achieving a reaction mixture of HL py,PNP and organic byproducts, we were still able to<br />
metallate the compound. We believe we have synthesized the desired [Ni2(HL py,PNP )(OAc)2][BPh4].<br />
4
Cryogenic Noise Measurements of Amplifiers for the Q/U Imaging ExperimenT (QUIET) Phase II<br />
Nathaniel Baskin<br />
Mentors: Anthony Readhead and Kieran Cleary<br />
The principle objective of the QUIET Phase II project is to detect the B-mode polarization of the Cosmic Microwave<br />
Background, a characteristic signature of universal inflation. Detecting these B-modes requires the development of<br />
lower noise amplifiers. Caltech and JPL are currently collaborating on a program to improve the fundamental<br />
performance of coherent amplifiers. The goal of this SURF project is to perform a full characterization on packaged<br />
35 nm amplifiers developed for this program. In particular, this involves measuring the DC bias characteristics and<br />
noise temperatures of these amplifiers. By doing so, we can provide feedback to the designers and the fabrication<br />
process; this will facilitate the optimization of the amplifiers’ cryogenic performance.<br />
Iron Glyoximes as Catalysts in Artificial Photosynthesis Systems<br />
Geoffrey Beck<br />
Mentor: Harry Gray<br />
Developing a viable artificial photosynthesis system that could generate fuel from sunlight and water remains one<br />
of the main challenges of 21 st century chemistry. While many catalysts based on different metals have been<br />
proposed, the use of iron for the catalytic center could allow for more economically viable devices to be developed.<br />
This project used density functional calculations to investigate the viability of iron glyoximes for the catalytic<br />
production of molecular hydrogen and the activation of carbon dioxide. Properties of intermediate structures for<br />
several potential reaction mechanisms were determined such that the feasibility of these pathways could be<br />
established and the utility of this and other iron based catalytic systems could be ascertained.<br />
Spectra Simulation for ULXs NGC 1313 X-1 and NGC 1313 X-2<br />
Juliette Becker<br />
Mentors: Fiona Harrison and Kristin Kruse Madsen<br />
ULXs (ultra-luminous X-ray sources) are accreting point-sources with unexpectedly high rates of luminosity. Not<br />
much is known about ULXs and their internal mechanisms. NGC 1313 X-1 and NGC 1313 X-2 are two such ULXs<br />
and thus both sources of interest as targets for observation by the NuSTAR telescope (a high energy x-ray<br />
telescope to be launched in February of 2012). Whether or not they are worth observing depends on if the spectral<br />
interference from nearby point sources is too high and masks the spectra of the ULXs. A NuSIM simulation will<br />
determine whether NGC 1313 X-1 and NGC 1313 X-2 are too obscured for an observation by NuSTAR to be useful.<br />
Feynman Graphs and the Field With One Element<br />
Dori Bejleri<br />
Mentor: Matilde Marcolli<br />
In quantum field theory, Feynman graphs parameterize the perturbative expansion of the Green’s function.<br />
To each Feynman graph G, there is an associated integral U(G) that gives a term in this expansion. Let<br />
be the Kerchoff polynomial, where the sum runs over all spanning trees T of G and each te is an indeterminate<br />
corresponding to the edge e. This polynomial is homogenous and it defines a projective variety XG. It is well known<br />
that U(G) can be can be written as an integral of an algebraic differential form on the projective complement of XG.<br />
Often U(G) is a mixed zeta value of a mixed motive suggesting XG is a mixed motive. This is not true in general,<br />
but considering when this holds motivates an interesting problem to determine when XG corresponds to a scheme<br />
over the field with one element. There are many different but related approaches to geometry over the field with<br />
one element. In this paper we investigate necessary and sufficient conditions on the graph G for XG to be defined<br />
over the field with one element under several of these approaches. Additionally, we explore what XG being defined<br />
over the field with one element and G satisfying these conditions implies for the physical interpretation of the<br />
Feynman graph. Further implications of this research lead to deeper studies and understanding of the link between<br />
the physical and algebro-geometric interpretations of Feynman diagrams.<br />
5
Bound States of Fourth Generation Quarks and Antiquarks<br />
Tanvir Ahamed Bhuyain<br />
Mentor: Mark B. Wise<br />
Recently it has been pointed out that fourth generation quarks and antiquarks can form long lived bound states<br />
that do not decay to a lower generation, if they have very small mixing with the lower generations. In this paper<br />
we further investigate some properties of such possible bound states. We focus on getting better estimates for the<br />
spectrum of these states using variational techniques. We study the effects of relativistic and perturbative<br />
corrections to the Higgs couplings in all spin, color and flavor channels and several trial wave functions.<br />
Identification of Endogenous Inflammatory Cells in the Mammalian Hypothalamus<br />
Evan R. Biggs<br />
Mentors: Daniel Marks and Ellen Rothenberg<br />
Cachexia, characterized by involuntary weight loss and loss of muscle mass, is a comorbidity commonly observed<br />
with numerous chronic inflammatory diseases. Cachexia has been repeatedly linked to chronic inflammation, and<br />
central inflammation of the hypothalamus has been repeatedly implicated as a critical component in its pathology.<br />
Accordingly, characterization of the molecular mechanisms underlying inflammation in the hypothalamus<br />
represents an essential step towards comprehending and treating this condition. It is well-established that<br />
peripheral immunologic challenge with lipopolysaccharide produces a cachexia-like phenotype in rodents, and<br />
hypothalamic expression of IL-1β has been implicated as an integral part of this pathophysiology. However, the<br />
molecular and neuroanatomic mechanisms underlying this induction of IL-1β signaling remain largely undescribed.<br />
In this study, we determine the cell-type responsible for the observed IL-1β signaling in the arcuate nucleus of the<br />
hypothalamus in response to peripheral immunologic challenge. Double-label in situ hybridization was utilized to<br />
observe colocalization of IL-1β mRNA with either an astrocyte or microglial marker in order to determine the<br />
generative cell-type. Results will be discussed as data become available.<br />
Deep UV Microbial Detection On Weathered Basalts<br />
Alexa Bilek<br />
Mentors: Luther Beegle, Rohit Bhartia, and William Abbey<br />
The Raman laser system is a sophisticated machine used to detect mineral biology on the surface of rock samples.<br />
My partner and I, through this summer internship program, have been working with this laser for the past 4 weeks<br />
in order to familiarize ourselves with the common graph spectra associated with the minerals found within Basalts<br />
and Gabbros. Developing a good understanding of this technique will ultimately provide a stable background of<br />
knowledge, which will be helpful in the field when we collect live samples from the Mojave Desert in two weeks.<br />
Since organics are not fond of hot surfaces, it will be important for us to scan the entire rock, including the<br />
shadowed areas, throughout the day and night in order to come up with a large range of data. We will then be able<br />
to hopefully make a correlation between anthropogenic sources (pollution containing CO2 and other organic<br />
compounds released from semi-trucks and cars) and the sustainability of microbiotic life (microbes without<br />
chlorophylls vs. microbes using photosynthesis).<br />
Insight Into the Growth of Garnet Porphyroblasts Through Isotopic Analysis of Iron<br />
Suzanne Birner<br />
Mentors: Nicolas Dauphas and Paul Asimow<br />
Large chemically zoned garnet porphyroblasts are commonly found in high-temperature/pressure metamorphic<br />
systems. While this concentric zonation-- specifically of Mg 2+ and Fe 2+ -- is well-documented, the processes<br />
governing crystal growth that lead to such gradients are less well understood. Recently, it has been shown that an<br />
isotopic analysis can help constrain the timing and conditions of crystal growth in igneous and metamorphic rocks.<br />
The growth of garnet porphyroblasts may be limited by grain-boundary diffusion in the country rock. The extent to<br />
which Fe and Mg isotopes will be fractionated during such growth will depend on the growth rate, thus providing a<br />
geospeedometer of garnet growth in metamorphic rocks. In this study, a large garnet from the Zillertal region of<br />
the Austrian Alps is analyzed compositionally using a scanning electron microscope. The resulting chemical<br />
zonation data are plotted as a contour map overlying a back-scattered electron image to show the concentric<br />
nature of the gradient. An isotopic analysis of iron, carried out using a Neptune inductively coupled plasma mass<br />
spectrometer, then produces a profile of the isotopic fractionation with respect to radius. These two results are<br />
combined to gain unique insight into the chemical and diffusive processes that govern crystal growth in<br />
metamorphic rocks.<br />
6
Improving Automated Segmentation for Statistical Analysis of Bacterial Filaments<br />
John Blackwood<br />
Mentors: Grant Jensen and Songye Chen<br />
The Jensen Laboratory at Caltech has proven the value of using electron cryotomography (ECT) for gaining key<br />
insights into cellular structures and cytoskeletal support systems of bacteria. While significant discoveries have<br />
already been made, there is still a need for substantial advancements in processing and analysis of ECT data sets.<br />
Very low signal-to-noise ratios resulting from limitations of the ECT experimental setup create challenges in noise<br />
reduction and automated segmentation that are unique to the field. The Jensen lab has recently developed a<br />
program to automatically segment bacterial filaments in three-dimensional reconstructed tomograms that come<br />
from ECT data collection and processing. Segmentation programs such as this are generally computationally<br />
expensive and have long execution times. This project explores roughly segmenting out the cell membrane in<br />
bacterial tomograms with a fast algorithm, so that more computationally intensive segmentation programs search<br />
in a reduced volume. Volume reduction can be as great as 70 percent, significantly reducing execution time and<br />
the necessary computational resources. In addition, preliminary results of the Jensen lab’s filament segmentation<br />
program were analyzed. The lab’s long term goal is to efficiently search for filaments in a large database of<br />
tomograms for quantitative, statistical results.<br />
Computational Study of Solution Accuracy in Finite Volume Methods Using OpenFOAM<br />
Daniel E. Blado<br />
Mentor: Dan Meiron<br />
OpenFOAM is an open source C++ library primarily used solving specific problems in continuum mechanics via the<br />
finite volume method. Prior to the full implementation and integration of OpenFOAM into computational fluid<br />
dynamics research, however, the accuracy of the library’s algorithms must be acceptable. The accuracy of<br />
OpenFOAM’s solver methods was investigated via a convergence study on a driven cavity problem, in which a<br />
square cavity of incompressible fluid has its top lid moving at a constant velocity. Using OpenFOAM’s calculated<br />
velocity data for this problem, velocities were extrapolated at the boundary. The spatial derivatives for each<br />
velocity component and pressure at the boundary were computed as well, which are needed to solve for important<br />
quantities such as fluid stresses. So far the solutions for a Reynolds number of 10 have been worked out with mesh<br />
resolutions 20x20x1 and 40x40x1, and such numerical data is being compared with expected results. Future work<br />
entails analyzing boundary solutions for both higher mesh resolutions and higher Reynolds numbers. Should the<br />
convergence rate of error in OpenFOAM be adequate, we can later utilize OpenFOAM for the study of separated<br />
flows over airfoils.<br />
Gene Expression Levels as Markers of Irradiation Damage to the Brain<br />
Michelle E. Bobrow<br />
Mentor: Robert J. Brown and Marianne Bronner<br />
There are currently no available means of treating or preventing the long-term effects of therapeutic brain<br />
irradiation, and the molecular mechanisms of such damage are not well characterized. This study determined the<br />
effects of radiation on gene expression levels in the brain, in order to identify markers of radiation injury. Slices of<br />
brain tissue containing the hippocampus were harvested from irradiated and non-irradiated mice for extraction of<br />
RNA. Reverse transcription followed by real-time PCR at several post-irradiation time points was used to analyze<br />
the changes in expression of various inflammation-related genes. A database of genes known to be expressed in<br />
the brain was created and annotated to allow for future real-time PCR analysis of genes involved in neurological<br />
disease, apoptosis, vascular formation and damage, hypoxia, oxidative stress, radiation response, and other<br />
processes relevant to radiation injury.<br />
Using Riboswitches to Understand Stem Cell Maintenance<br />
Stephanie Bohaczuk<br />
Mentors: Elliot Meyerowitz and Paul Tarr<br />
Synthetic biology aims to engineer biological systems that elicit preprogrammed outputs. An application is<br />
riboswitches, which are RNA molecules that control translation of an mRNA. A riboswitch consists of a ligand<br />
binding aptamer, which is coupled to a hammerhead ribozyme that mediates mRNA degradation. Binding of the<br />
ligand induces a conformational change in the riboswitch, inactivating the ribozyme and allowing translation. We<br />
will use riboswitches to study how dynamic changes in the CLAVATA/WUSCHEL (CLV/WUS) genetic regulatory<br />
network and auxin-signaling pathway regulate stem cell differentiation and specification in the shoot apical<br />
meristem (SAM). Utilizing live imaging, we will follow stem cell proliferation and differentiation over several days<br />
after induction of a tetracycline or theophylline riboswitch to recover CLV or WUS expression in clv3 and wus<br />
mutants, respectively. The role of auxin in plant organ initiation in the SAM has been difficult to study due to the<br />
lack of a direct reporter. We will implement a riboswitch to act as a direct reporter of auxin localization. We use the<br />
SELEX protocol to develop de novo apatamers responsive to auxin, which will be coupled to a fluorescent protein.<br />
Preliminary results reveal similar motifs in the aptamer candidates, which require improvements in their affinities<br />
for auxin.<br />
7
Synthesis of New Fluorophores for Super Resolution Microscopy<br />
Umed Boltaev<br />
Mentors: Long Cai and Timur Zhiyentayev<br />
Super resolution microscopy techniques overcome the diffraction limit of traditional microscopy. However, there is<br />
a demand for photostable fluorophores with high contrast ratio in order to achieve the desired resolution and high<br />
labeling density. In order to use different fluorophores simultaneously one need to activate only one kind during<br />
imaging. The goal of this project was to synthesize new photoswitchable fluorophores with desired spectral<br />
properties. We have obtained 2 coumarins that are able to cage rosamine derivative with excitation maximum at<br />
550 nm. My final goal is to prepare rosamine with excitation maximum at 595 nm, cage it with spectrally distinct<br />
coumarins and test it as super resolution probes.<br />
The Characterization of Carbon Dots, Gold Nanorods, and Graphene as Components of Photothermal<br />
Tumor Treatment<br />
Carly Bond<br />
Mentor: Giyoong Tae<br />
Previous studies have shown that it is possible to use the photothermal effect of various materials as a method of<br />
killing tumor cells. Current objectives of researching this method include maximizing the accumulation of the<br />
photothermal material in the tumor, minimizing the non-specific cytotoxicity of the materials, and finding<br />
photothermal materials that can be heated using light (preferably NIR light). Several nanomaterials including<br />
carbon dots (CDs), gold nanorods (GNRs), and graphene will be compared as possible photothermal materials for<br />
tumor treatment. Loading materials into functional nanocarriers that can concentrate photothermal materials and<br />
deliver selectively to the target tissue before injecting them into the bloodstreamis a way to increase the efficacy of<br />
the photothermal therapy Chitosan-conjugated, pluronic-based nanocarriers have demonstrated the selective<br />
tumor targeting of loaded GNRs and thus highly effective tumor ablation upon laser irradiation. Therefore, the<br />
encapsulation efficiency and release of other materials from the nanocarriers and the photothermal heating of<br />
nanocarriers loaded with. CDs, GNRs, or graphene will be characterized for potential in vivo applications.<br />
Network for Collection of Seismic Data<br />
Prashant Borde and Prathamesh Juvatkar<br />
Mentors: Julian J. Bunn and K. Mani Chandy<br />
Earthquakes are natural disasters responsible for destruction, injuries and loss of life and property. However,<br />
earthquakes are not directly responsible for loss of life. The secondary events triggered by earthquakes such as<br />
structure collapse, fires, tsunamis and volcanoes cause actual human disasters. The losses due to earthquakes can<br />
be greatly reduced if we can give early warnings of earthquake or at least detect an earthquake’s early seismic<br />
activity. For building a system capable of early detection of seismic activity, we need high density arrays of<br />
spatially distributed motion sensors. Our project is to develop a private network of portable low cost seismometers<br />
that consume minimal resources. Each seismometer consists of a Single Board Computer (SBC), “Phidget”<br />
accelerometer and a Home-plug device for power line communication. Each seismometer sends the data to next<br />
device in linear topology using power-line communication. A base-station in the network sends the collected data to<br />
a central server via the internet. The operation of the seismometer network is completely automated after<br />
installation. Also the network is robust against intermediate node failure. Large numbers of such private networks<br />
will ensure sufficient amounts of data available for the detection of earthquakes with higher accuracy.<br />
A Criterion for Arithmeticity of Subgroups of PGL(2,C)<br />
Zarathustra E. Brady<br />
Mentor: Dinakar Ramakrishnan<br />
We attempt to generalize a result of Piatetski-Shapiro and Jiang (Geom. Funct. Anal., 8(3):586–605, 1998) from<br />
PGL(2,R) to the group PGL(2,C), giving a criterion for the arithmeticity of a discrete subgroup with finite covolume.<br />
We show that under the right assumptions on a subgroup Γ of PGL(2,C), Γ is arithmetic if it has an<br />
automorphic cusp form f such that f is a Hecke eigenform with respect to a prime p, and that this is equivalent to<br />
L(s,f) admitting an Euler factor at a prime p.<br />
Multiple Mirror Space Telescope Arrays: A Study of Imaging Applications and the Determination of<br />
Deformable Mirror Shape From Optical Wave Measurements<br />
Liana Braun<br />
Mentors: Sergio Pellegrino and Keith Patterson<br />
It has been proposed that a multi-mirror, self-attuning space telescope may be more cost-effective than a<br />
telescope with a large monolithic, high-mass mirror if the various smaller mirrors can produce results that rival the<br />
accuracy of currently used space telescope technologies. This project aims to optimize the focusing abilities of each<br />
individual deformable mirror in the telescope’s array in order to successfully image stellar objects. Using various<br />
references, an assortment of ideal stellar imaging targets was collected and plotted to determine accessibility by<br />
8
the spacecraft. Once this collection was completed, work was geared toward focusing light reflected from a<br />
deformable mirror in the lab. This was done through manipulation of voltages passed across the PVDF membrane<br />
of the mirror, coupled with an imaging process to detect changes in the image size. After a successful image<br />
focusing process is determined, it will be possible to redirect the image to a Shack-Hartmann sensor, which will<br />
provide more detailed image information, allowing for better shape sensing.<br />
Water on the Moon? A Quantitative Study of Water Solubility<br />
Alex Brett<br />
Mentors: E. Stolper and M. Newcombe<br />
Lunar water is a controversial subject: ongoing debate concerns its origin and quantity. Lunar glasses containing<br />
up to 40ppm water and melt inclusions containing up to 1500ppm water give broad constraints, but correct<br />
interpretation of these samples requires a thorough understanding of the diffusive behaviour of water at low<br />
concentrations in appropriate melt compositions. Here I present an experimental study on the solubility of H2O in<br />
anorthite-diopside and lunar-composition glasses at 1350°C and 1 atm (CO2-H2-H2O-CO). The results show that OH<br />
and not H2 is likely to be the dominant diffusive species, and establish empirical relationships between equilibrium<br />
concentrations and partial pressure of atmospheric water.<br />
Characterizing the Properties of Copper in Silicon Microwire Arrays<br />
Reuben Britto<br />
Mentor: Nathan Lewis<br />
Silicon microwires, grown from SiCl4 through the vapor-liquid-solid (VLS) process, have been shown to possess<br />
high material quality enabling solar cells to reach efficiencies that rival many wafer-based crystalline Si<br />
technologies. The VLS growth process requires annealing copper-patterned silicon wafers at 1000� C before<br />
introducing the Si precursor. Copper, due to its high diffusion coefficient, readily mixes with silicon at this high<br />
temperature, forming many different compounds in different phases. Selective etching, EDS-SEM, and XRD were<br />
used to determine what compounds and phases are forming, where they’re forming, and in what amounts. Copper<br />
oxide was detected and located on planar samples through SEM-EDS, and its cause of formation – sample<br />
exposure to atmosphere at >100� C – was established. Copper silicide (Cu3Si) was also detected through X-Ray<br />
diffraction on annealed planar samples.<br />
After identifying two phases of Cu impurity on planar samples, wire arrays were analyzed in the same manner, and<br />
several Cu-containing regions were identified. These results motivate further investigation of how changing growth<br />
and cooling conditions can minimize precipitation of Cu-rich phases in Si microwire arrays.<br />
Bioremediation of Endocrine Disruptors: Isolation of an Endocrine-Disruptor-Degrading Bacterium<br />
Julia Brown<br />
Mentor: Richard Murray<br />
Endocrine disruptors are compounds, such as estrogen or compounds that mimic estrogen, that interfere with the<br />
normal action of an organism’s endocrine system. The effects of this are most significant when exposure occurs<br />
during the organism’s development, as endocrine-disrupting compounds can disrupt the signaling pathways that<br />
direct the various stages of development. The Caltech iGEM team is developing a system, housed in E. coli, that is<br />
capable of detecting and degrading these compounds (specifically, bisphenol A, DDT, nonylphenol, or 17-αethinylestradiol).<br />
Using soil samples from the L.A. river as a source, we attempt to isolate a bacterium capable of<br />
degrading our compounds and determine the gene(s) responsible for degradation. We hope to express this gene in<br />
E. coli in conjunction with a modified human estrogen receptor to form a biological circuit capable of removing and<br />
degrading an endocrine-disrupting compound from water.<br />
Chemical Effects of Ice Crystal Growth<br />
Nina Budaeva<br />
Mentor: Kenneth Libbrecht<br />
The prediction of ice crystal growth rates and morphologies is a difficult problem which depends on nucleation<br />
dynamics, temperature history, chemical makeup of the sample, and many other parameters. This project<br />
investigates the effects of solute chemistry on the growth of ice crystals. By varying the physical parameter space<br />
of nucleation temperature and solution concentration, I am able to observe changes in the ice crystal growth of<br />
various water solutions.<br />
The ice crystals are grown in a wire loop and are recorded on camera so that their growth rates and morphologies<br />
can be observed from the recordings. Solutions of particular interest are salt and sucrose solutions as well as<br />
nanoparticle suspensions and antifreeze protein solutions. The results of these experiments, besides providing<br />
scientific insight into an insufficiently-explored field, may also prove useful for direct applications such as in freezedrying<br />
of pharmaceuticals.<br />
9
Meter-Scale Evolution of the Martian Residual South Polar Cap<br />
Peter B. Buhler<br />
Mentors: James W. Head III and Andrew P. Ingersoll<br />
We analyze high resolution imagery of the Residual South Polar Cap (RSPC) on Mars to document seasonal and<br />
interannual ablation and accumulation processes of CO2 ice. We observe vertical and horizontal changes in the<br />
RSPC and find that ablation and accumulation at meter-scales vary both spatially and temporally.<br />
This investigation details the mesoscopic (decimeter to decameter scale) processes by which the RSPC evolves on<br />
multi-day to annual timescales. We find that seasonal CO2 frost persists longer into the summer during Martian<br />
Year (MY) 30 than it persists in MY 29 in some locations. We postulate that this gradual accumulation offsets the<br />
sublimation of the RSPC, acting to stabilize the RSPC on timescales shorter than solar or orbital climate cycles.<br />
Solar Cell Enhancement by Whispering Gallery Modes From Dielectric Nanosphere Array Coatings<br />
Colton Bukowsky<br />
Mentors: Harry Atwater and Dennis Callahan<br />
Thin-film photovoltaic performance can be improved by increasing the absorption capability of the active layer.<br />
An innovate method utilizing dielectric-resonant nanospheres placed on top of the active layer has been shown by<br />
simulations to excite whispering gallery modes (WGMs) in the nanospheres, which then couple incident light into<br />
the active layer. This coupling increases the absorption at these wavelengths, and it is desired to experimentally<br />
show this phenomenon occurs in real solar thin-film solar cells. Efforts to create hexagonally close packed 2D<br />
colloidal crystals of silica nanospheres have been successful in creating long range order on various substrates with<br />
greater than 90% area coverage using the Langmuir-Blodgett technique. These arrays were coated onto semiconducting<br />
surfaces and functioning PV cells. Absorption measurements will be used to demonstrate absorption<br />
peaks at certain wavelengths and matched to those in the models in order to show that the WGMs are excited.<br />
The Receptor mGluR1 Is Reduced in Postsynaptic Densities of Mice With a Deletion of Densin<br />
Sarah R. Cantor<br />
Mentor: Mary B. Kennedy<br />
Densin is a scaffold protein that is abundant in the postsynaptic density. Mice with a homozygous knockout of<br />
densin exhibit abnormal behaviors that are characteristic endophenotypes of schizophrenia and autism spectrum<br />
disorders. Past research has shown that densin knockout mice have a 30% reduction of mGluR5, a metabotropic<br />
glutamate receptor, in the postsynaptic density compared to wildtype mice. Type I metabotropic glutamate<br />
receptors comprise two individual receptors, mGluR1 and mGluR5. Currently, little is known about the binding<br />
partners of mGluR1 and whether this protein associates with densin. Through the use of quantitative immunoblots,<br />
I showed that densin knockout mice have a 30% reduction in levels of mGluR1 in the postsynaptic density fraction,<br />
whereas levels in the total homogenate are unchanged. This indicates that densin plays an important role in the<br />
localization of both mGluR5 and mGluR1 to the postsynaptic density. It may be part of a protein complex that<br />
associates with both mGluR1 and mGluR5 at the postsynaptic membrane.<br />
Cell Phone Earthquake Detection: Small Memory Variations on Expectation Maximization for Gaussian<br />
Mixture Models<br />
Alejandro Carbonara<br />
Mentor: Andreas Krause<br />
Most smartphones come equipped with accelerometers, providing a wealth of untapped data. The Caltech Seismic<br />
Network project aims to use this data in order to detect earthquakes as they are happening and warn cell phone<br />
users as early as possible. Our part of this project is the client side abnormality detection. Since we have a limit on<br />
how much data can be transmitted, each cell phone needs to detect which of its recordings are statistically<br />
significant in order to only transmit only the most important information. We decided early on in our project that<br />
we wanted to use Gaussian Mixture models to model our data and that we wanted to use Expectation Maximization<br />
to fit these models. Due to the memory limitations of a cell phone, we examined various ways to compress our<br />
data, including randomization, coresets, and a streaming model. Using cell phone data collected to represent an<br />
average day, we coded up and ran several experiments to see how well these various small memory<br />
implementations worked.<br />
Forming a Porous Scaffold With Clover-Leaf Shaped Microgels for Tissue Engineering Formulations<br />
Jacqueline Chan<br />
Mentors: David A. Tirrell, Julie A. Kornfield, Amy Fu, Larry Dooling, and Loddie Foose<br />
Three-dimensional protein polymer scaffolds have been engineered to promote specific cell behaviors such as<br />
migration, differentiation, and proliferation. This work describes a method of fabricating clover-leaf shaped<br />
polyethylene glycol diacrylate microgels that can be modularly assembled into a macroscopic scaffold material with<br />
high porosity, permeability, and pore interconnectivity. A secondary crosslinking reaction is used to stabilize the<br />
10
assembled microgels. This bottom-up method allows for the rapid and scalable fabrication of microgels with defined<br />
geometrical and biological features, which can potentially be used for microscale tissue engineering applications.<br />
Future work entails fabricating these microgels with photocrosslinkable artificial extracellular matrix protein.<br />
Fluorescent Double Labeling of the Signal Recognition Particle to Study Interdomain Interaction<br />
Sandhya Chandrasekaran<br />
Mentors: Shu-ou Shan and Ishu Saraogi<br />
Proper cellular localization of proteins is essential for survival and errors in this process could lead to cell<br />
malfunction or cell death. About one-third of the proteins in the cell are delivered to their correct cellular<br />
destination by the Signal Receptor Particle (SRP). During targeting, SRP forms a heterodimeric complex with its<br />
receptor (SR), and the SRP-SR complex goes through a series of conformational changes that provides the spatial<br />
and temporal cues for the targeting reaction. The protein component of SRP, called Ffh, is a multi-domain protein<br />
comprising a substrate-binding domain connected to a GTPase domain via a flexible linker. The mechanism by<br />
which the movement of these two domains with respect to each other coordinates downstream signaling events is<br />
poorly understood. Here, we developed a double fluorescence-labeling scheme for Ffh to study these interdomain<br />
movements with Forster Resonance Energy Transfer (FRET). By optimizing labeling conditions, Ffh double-cysteine<br />
mutants were differentially labeled using cysteine-maleimide chemistry with complementary donor and acceptor<br />
fluorophores. Subsequent FRET experiments with these labeled proteins can provide important kinetic and<br />
thermodynamic information about events in the protein targeting process that involve inter-domain movements of<br />
Ffh.<br />
Simulac: A Stochastic Simulator for Modeling Biomolecular Circuits<br />
Arjun Chandrasekhar<br />
Mentor: Richard Murray<br />
The primary purpose of this research project is to develop and modify Simulac, a stochastic simulator for creating<br />
and simulating models of various biomolecular circuits. We want to add and modify different features of the<br />
simulator, and then use it to simulate environments expressing bimodal and graded gene expression to determine<br />
the factors that affect noise in gene expression. The first goal is to make the program easier to use by<br />
incorporating the graphical user interface from TinkerCell. Next we want to implement tau leaping algorithms to<br />
make the program run faster and more efficiently. Finally we want to expand the set of reactions that the simulator<br />
can account for, including anti-sense RNA binding and various polymerase interactions. In the future, once all of<br />
these features are in place, the next step will be to set up various environments and modify features that we think<br />
are correlated to noisiness in gene expression (combinatorial promotes, transcriptional efficiency, chromatin<br />
remodeling, etc.) in order to determine what causes bimodal and graded responses.<br />
Thioester-Functionalized Silica Microspheres as Substrates for Covalent Immobilization<br />
Christine H. Chang<br />
Mentors: Harry B. Gray and Paul J. Bracher<br />
This project focuses on the development of a binding system based on the thiol–thioester exchange reaction that<br />
uses thioester-functionalized silica microspheres as a substrate to bind thiols. The covalent immobilization of<br />
molecules to a solid phase is a key step in a wide range of applications, such as solid-phase organic synthesis,<br />
affinity chromatography, and assays for biological activity. While several binding systems are currently in use—<br />
each with its advantages and disadvantages—the system under investigation in this project is particularly<br />
advantageous due to its use of small and noninvasive molecular tags as binding partners, its ability to function in<br />
biologically compatible conditions, and the availability of several methods to cleave the tightly bound partners<br />
without the use of harsh reagents. The silica microspheres were fabricated by the Stöber process, the binding<br />
interactions were measured by fluorimetry, and the susceptibility of the microspheres to aminolysis and hydrolysis<br />
was also studied. Specific areas of interest included the effects of phosphorus(V) versus EDC coupling reagents in<br />
the preparation of the functionalized microspheres, metal species (e.g., Cu 2+ , Pb 2+ , and NaAuCl4) on the rates of<br />
aminolysis, and increased thiol functionalization on binding.<br />
11
Identification of Cis-Regulatory Regions in L. variegates DNA<br />
James Chang<br />
Mentor: Andrew Cameron<br />
Gene regulatory networks consist of elements that control gene expression and the connections between them.<br />
Computational genomics is a useful tool for finding non-coding but functional cis-regulatory regions which affect<br />
gene regulation in the Strongylocentrotus purpuratus (Sp) sea urchin genome. To find similarities, Sp sequence is<br />
juxtaposed with that of Lytechinus variegates (Lv), a species 50 million years diverged. Divergence of<br />
nonfunctional areas increases with evolutionary distance, but functional areas will be conserved. Thus, possible<br />
regulatory regions are defined as conserved areas upstream and downstream from genes. The search for<br />
regulatory regions begins by pairing similar genes as initial references. Then, the surrounding sequence from both<br />
genomes can be compared. Although similarities between known Sp genes and Lv sequences have been found,<br />
work on corresponding upstream and downstream sequences has not been completed. Once scripts have been<br />
adapted to large-scale sequence files, the process will be streamlined and potential cis-regulatory regions will be<br />
identified.<br />
An Exact Solution for the Time-Dependent Behavior of the Kitaev Toric Code Under a Rapid Quench<br />
Adrian Chapman<br />
Mentors: John Preskill and Steve Flammia<br />
Topologically ordered quantum systems have recently been of interest as potential mechanisms for the<br />
implementation of fault-tolerant quantum computation. We examine the time-dependent behavior of one such<br />
system, the Kitaev toric code, following the sudden application of a local transverse magnetic field -- a quantum<br />
quench. By mapping the transverse toric code to a series of decoupled one-dimensional Ising models, we exactly<br />
solve for the energy spectrum and eigenstates of the original model and use the result to calculate the time<br />
evolution of the expectation values of contractible and non-contractible loop operators, along with the topological<br />
entanglement entropy. We expect these quantities to remain constant over a timescale short compared to the<br />
system size divided by the Lieb-Robinson velocity. This study is useful for the understanding of topological phases<br />
and the construction of reliable quantum memories.<br />
Modeling Nonlinear Response in a Dispersion-Engineered Traveling-Wave Kinetic Inductance<br />
Parametric Amplifier<br />
Saptarshi Chaudhuri<br />
Mentor: Jonas Zmuidzinas<br />
Recently, Zmuidzinas and Day proposed a new concept for a low-noise cryogenic amplifier known as a Dispersion-<br />
Engineered Traveling-Wave Kinetic Inductance (DTWKI) amplifier. This superconducting transmission line device<br />
uses nonlinear kinetic inductance to convert pump wave photons into signal wave photons, producing significant<br />
signal power gain. In theory, the DTWKI amplifier should operate close to the noise limit imposed by quantum<br />
mechanics. However, the nonlinearity leads to the production of pump harmonics, and unless countermeasures are<br />
taken, shock waves form, preventing signal amplification. In the DTWKI device, shock front formation is prevented<br />
by periodically perturbing the width, and hence the impedance, of the transmission line to create a stop band at<br />
three times the pump frequency. Using a harmonic-balance computation scheme, a model of the amplifier was<br />
developed. Wave deformation and the growth of pump harmonics in uniform transmission lines were studied<br />
extensively. The effects of nonlinear kinetic inductance on traveling waveforms in periodically-loaded transmission<br />
lines were compared to the effects on waveforms in uniform lines.<br />
Sampling Tool Concepts for Comet Surface Sample Return<br />
Allen Chen<br />
Mentor: Paul Backes<br />
Three sampling tools for sampling of comet surfaces in a touch-and-go (TAG) mission architecture were developed<br />
and tested and compared with test results using a Brush Wheel Sampler (BWS) sampling tool. The BWS was<br />
developed previously and tested across a range of stimulants so it was used as a reference to compare the new<br />
tool concepts with. The three tools were a drive tube (DT), comet corer (CC), and comet auger (CA). The primary<br />
anticipated benefits of the DT, CC, and CA compared with the BWS were that they could acquire subsurface<br />
samples to 10cm depth and potentially enable a simpler sample acquisition, measurement, and transfer (SAMeT)<br />
capability compared to the BWS. The BWS has been shown to reliably acquire a large surface sample but it is of<br />
interest to determine whether an alternative tool could acquire a subsurface sample to at least 10cm depth with<br />
similar reliability and the same or reduced system complexity.<br />
12
Patterns of Nanog Sub-State Switching in Single Mouse Embryonic Stem Cells<br />
Chih-Ping Chen<br />
Mentors: Michael Elowitz and John Yong<br />
Nanog is a transcription factor thought to safeguard embryonic stem cells from differentiating. However, it is found<br />
to be expressed heterogeneously even in undifferentiated populations between a high-Nanog state and a low-<br />
Nanog state. Of particular interest for this project, it has been observed that individual cells can switch between<br />
high-Nanog and low-Nanog states while remaining undifferentiated. Using time-lapse fluorescent imaging, it is<br />
possible to track Nanog transcription in single cells through multiple cell cycles. From analyzing these tracks, we<br />
seek to characterize the frequency, time-scale and pattern of inheritance of the switch. These patterns are being<br />
compared to establish the uniqueness of the way Nanog transcription is being regulated. This analysis will serve as<br />
the basis for further study of the phenomenon under various culture conditions.<br />
Doped Polycrystalline Silicon Templates for Epitaxial Crystalline Thin-Film Solar Cells<br />
Clare Chen<br />
Mentors: Harry A. Atwater and Michael G. Deceglie<br />
We demonstrate the process of metal induced lateral crystallization (MILC) of phosphorus and boron doped<br />
amorphous silicon films to produce templates for epitaxial growth of thin-film crystalline silicon solar cells. In<br />
developing these thin-film crystalline templates, a low-temperature, low cost process enables the use of low-cost<br />
substrates such as glass. Upon annealing with nickel nanoparticles, amorphous silicon was crystallized by MILC.<br />
These thin films were characterized by optical microscopy, atomic force microscopy, and scanning electron<br />
microscopy to analyze MILC growth. Then surface nickel deposits were etched off the crystallized template before<br />
electrical characterization by Hall measurements and compositional analysis by secondary ion mass spectrometry<br />
(SIMS). These measurements were used to investigate the carrier concentration in this doped template layer to see<br />
if this polycrystalline template can provide back surface field for the solar cell, preventing flow of minority carriers<br />
to the rear of the cell thus minimizing surface recombination losses.<br />
A Modified Cellular Automata Model for Dendritic Computation<br />
Daniel Chen<br />
Mentors: Christof Koch and Adam Shai<br />
There exist many proposed models for post-synaptic integration in a typical pyramidal neuron. Many, for the sake<br />
of simplicity, do little, if anything, to account for the nonlinearities that exist throughout the neuron. Other more<br />
sophisticated models abstract the tree into a large set of complicated circuits, which require thousands of<br />
parameters. Consequently, it may be difficult and computationally intensive to use these models to perform<br />
analysis beyond small networks. In this paper, we propose a model that accounts for all of the nonlinearites while<br />
remaining relatively simple and easily adjustable to account for additional parameters. We have made several<br />
simplifications; particularly, abstracting the voltage transients induced by single synapses to discrete signals which<br />
interact with each other according to simple rules. The time and space evolution of these are governed according to<br />
known parameters. With this model, we hope to gain an understanding of the fundamental components of the<br />
single cell mechanism and thus be able to develop more refined experiments that focus on the more important<br />
factors.<br />
Establishing Ground Truths for Automated Monitoring and Analysis of Aggressive Behavior in Female<br />
Drosophila<br />
John Chen<br />
Mentors: David Anderson and Rod Lim<br />
Aggression is an innate social behavior observed in many species. Recently, Drosophila has emerged as a model<br />
system for studying aggression. Previous studies demonstrated that fruit flies, as in many other species, exhibit<br />
gender-specific patterns of aggressive behaviors. Although machine vision algorithms have been adapted for<br />
automatic detection of male-male social behaviors, no such tool exists for female-female behaviors. In order to<br />
characterize female behaviors in a systematic way, we seek to develop a machine vision system that can quantify<br />
and analyze various social behaviors in Drosophila. Currently these female behaviors are measured manually,<br />
which is slow and laborious. A machine algorithm that can detect and analyze these behaviors will be highly useful<br />
in studying the genetic and neural bases behind female aggression. By carefully annotating female-female<br />
aggression using high resolution cameras, we will establish definitive behavioral ground truths that, through<br />
collaboration with the Perona Lab, can help establish a machine algorithm for evaluating aggression in female<br />
Drosophila.<br />
13
Identifying Genes Involved in FGF-Dependent Cell Migrations in Drosophila<br />
Rebecca Chen<br />
Mentors: Angelike Stathopoulos and Young Bae<br />
Embryonic development of multicellular organisms is a complicated process of cell movement and cell<br />
differentiation. Drosophila serves as a model organism in which intricate gene regulatory networks that orchestrate<br />
the complex process of gastrulation can be studied. This project focuses on the study of two dynamic events that<br />
are regulated by fibroblast growth factor (FGF) signaling: migration of early mesoderm and caudal visceral<br />
mesoderm (CVM). To identify genes that play a role in mesoderm migration and differentiation, the gene<br />
expression profile of sorted mesodermal cells was obtained by RNA-seq. We are verifying gene expression in the<br />
mesoderm by in situ hybridization and testing if expression of any target genes is dependent on FGF signaling.<br />
Phenotypic analyses of those with specific expression in the early mesoderm will be performed. Similarly, for CVM<br />
cell migration, we confirmed expression of candidate genes through in situ hybridization. We are examining what<br />
role these candidate genes play in CVM cell migration using RNAi hairpins. In addition, a technique for cell<br />
transplantation has been developed so that cells can be transplanted and observed in an embryo of different<br />
genetic background.<br />
Parylene-C Characterization<br />
Rujian Chen<br />
Mentors: Yu-Chong Tai and Chun-Hui Ling<br />
Parylene-C is widely used as a biocompatible material for many kinds of micro analysis systems. As-deposit<br />
parylene coatings feature low thickness, high uniformity, good mechanical strength, barrier properties and<br />
adherence to substrates. However, parylene used in practical devices will go through many treatment processes<br />
which may alter its microstructure and hence its mechanical and chemical properties. During the fabrication of<br />
MEMS devices, it is observed that parylene films become more brittle after heat treatment, which shortens the<br />
film's lifetime and degrades its performance. This project investigates the cause and mechanism of the<br />
deterioration of mechanical properties. Parylene samples annealed under different conditions(air/vacuum, different<br />
temperatures and annealing times) are prepared, and structural changes are studied, by means of fluorescence,<br />
IR, XPS, XRD methods. Experimental results have shown evidence of the onset of densification, oxydization, and<br />
crystallization under different annealing conditions.<br />
Microfluidic Polymerase Chain Reaction (PCR) Platform for Pathogen Detection<br />
Samson Chen<br />
Mentor: Axel Scherer<br />
Microfluidic medical diagnostic systems, due to their ability to integrate incredibly complicated procedures into a<br />
single device, have the potential to drastically reduce the cost and increase the availability of modern medicine<br />
worldwide. However, there are still many significant barriers to achieving this goal with current microfluidic<br />
fabrication techniques. We present the first working application of a novel microfluidics platform which overcomes<br />
many of these barriers. Specifically, we fabricated a fully integrated Polymerase Chain Reaction (PCR) thermal<br />
cycler, a device which permits the amplification of very small concentrations of specific sequences of DNA, often<br />
found as part of many pathogen detection schemes. Additionally, we designed and optimized a low cost circuit on<br />
the same device to thermally regulate the PCR chamber, demonstrating the ability of this platform to tightly<br />
integrate electronics and fluidics in the same package at very low cost.<br />
Stereological Sampling of von Economo Neurons (VENs) in Aging Human Brain Tissue<br />
Victor Chen<br />
Mentor: John Allman<br />
The principal objective and scope of the project was to investigate the role of VENs in the human brain. It has been<br />
hypothesized that these neurons serve as homeostatic mechanisms to ensure proper functioning of core bodily<br />
functions, and therefore the project focused on determining whether there was a correlation between the number<br />
of VENs and longevity. The methodology of the project was based on the use of a stereoinvestigator to make<br />
accurate estimates of the number of VENs within the fronto-insular cortex. Only very preliminary results have been<br />
obtained so far, as very few aging brains have been completely counted with stereology for VENs. The project will<br />
be definitely continued with subsequent work focusing on enlarging the data pool so that reliable conclusions can<br />
be made about the importance of VENs in the brain.<br />
Identifying a Protein Lattice on the Outer Spore Membrane in Acetonema longum<br />
Wesley Chen<br />
Mentors: Grant Jensen and Elitza Tocheva<br />
During sporulation, an elaborate protein coat essential for spore viability assembles around the outer spore<br />
membrane. One protein of interest, SpoIVA forms the foundation of the protein coat to which all other proteins<br />
bind to. However, the exact structure of SpoIVA is unknown. While the protein coat has been studied extensively in<br />
14
Bacillus subtilis, there are no high-resolution images of SpoIVA in vivo during early stages of sporulation. One<br />
possible technique, electron cryo-tomography (ECT), is a powerful new tool that is capable of producing threedimensional<br />
reconstructions of microorganisms and proteins in their near-native state in close to atomic resolution.<br />
Acetonema longum is a sporulator suitable for ECT studies. Previously, a series of concentric protein rings has been<br />
imaged on the outer spore membrane of A. longum that we hypothesize to be SpoIVA. Here, we use a variety of<br />
biochemical assays, imaging techniques, and genetic manipulations to identify the protein lattice found on the<br />
protein surface of the endospore. We anticipate that these approaches will provide opportunities for identifying the<br />
protein lattice formed during early stages of sporulation in A. longum.<br />
Light Curve Classification of Time-Variable Astronomical Phenomena<br />
Yutong Chen<br />
Mentors: George Djorgovski, Baback Moghaddam, and Ashish Mahabal<br />
Recent development in digital synoptic sky surveys presents researchers with a more in-depth examination of<br />
known and newly discovered time-variable astronomical phenomena, or transients. This opportunity comes in the<br />
form of an overwhelming amount of data, and at the same time requiring identification and prioritization for followup<br />
observations. We employ a Decision Tree (DT) classifier for automatic classification of transients based on the<br />
transient’s light curve. Given a transient, we convert the transient’s light curve to its corresponding Δm vs. Δt<br />
histogram. We can generalize a particular class of transients’ Δm vs. Δt histogram; and thus building a probability<br />
density function (prototype histogram) for that class. Classification of an unknown transient is accomplished by<br />
sequentially cascading the transient’s Δm vs. Δt histogram against the DT’s prototype histograms. We assess the<br />
reliability of our DT classifier using a 5-fold cross-validation on a comprehensive confusion matrix. Our DT classifier<br />
performs well in correctly identifying CV, RR, and SN1a; but the reliability of our DT classifier significantly<br />
decreases for Blazar, Mira, SN1bc, and SN2np. A correlation with the reliability of our DT classifier is that wellperforming<br />
classes contains at least 350 examples in their training sets while poor-performing classes contains less<br />
than 100 examples in their training sets.<br />
Reorganization of Nanoparticles on Origami Surfaces by DNA Random Walkers<br />
Zibo Chen<br />
Mentors: Erik Winfree, Lulu Qian, Niranjan Srinivas, and Damien Woods<br />
The specificity of Watson-Crick base pairing has been shown to be useful when constructing nanostructures using<br />
DNA. Since the development of DNA nanotechnology, DNA structures with various functions have been<br />
constructed, such as DNA tiles that can process information, DNA origami that can serve as molecular assembly<br />
lines, etc. This study, conducted together with Shayan Doroudi, Gregory Izatt, Yae Lim Lee and Sarah Wittman, is<br />
part of the upcoming BIOMOD competition. In this study, we developed a technique to reorganize randomly placed<br />
nanoparticles using random walk. Using the algorithm NUPACK, we designed single-stranded DNA walkers that<br />
were capable of walking on DNA origami and picking up cargoes with DNA strands attached; they would continue<br />
the random walk until encountering goal strands and transferring the cargoes to the goals by strand displacement.<br />
The mechanisms were verified using polyacrylamide gel electrophoresis, fluorescent spectroscopy and atomic force<br />
microscopy. This system is useful when being integrated into a DNA assembly line to deliver DNA labeled<br />
nanoparticles, and it opens up the possibility of sorting different types of cargoes at the same time.<br />
Synthesis of Nickel Complexes of Redox-Active Protic Ligands for CO2 Reduction<br />
Christine Cheng<br />
Mentors: Theodor Agapie and Emily Tsui<br />
Carbon dioxide is a major cause of global warming and results from many common processes. Since CO2 is known<br />
to bind to dinuclear nickel complexes, a dinickel complex supported by a diphosphine p-terphenyl ligand was<br />
designed as a potential electrocatalyst for the reduction of CO2. In this project, variants of the ligand containing<br />
electron-donating methoxy and hydroxyl groups on the central arene were synthesized. Monometallic nickel(0)<br />
complexes and bimetallic nickel(I) complexes were synthesized using these diphosphine p-terphenyl ligand<br />
frameworks. These complexes were characterized by NMR and other methods and studied for CO2 reactivity.<br />
Using E. coli to Bioremediate Endocrine Disrupting Chemicals<br />
Puikei Cheng<br />
Mentors: Richard Murray, Nathaniel Glasser, Toni Lee, Joseph Meyerowitz, and Emmanuel Lorenzo de los Santos<br />
Endocrine disrupting chemicals, or EDC’s, disrupt the reproductive abilities of aquatic wildlife by imitating hormones<br />
present in their bodies. Traditional efforts to remove these chemicals have faced problems such as cost, disposal of<br />
products, and unforeseen consequences. This year, the Caltech iGEM team has decided to try a better method of<br />
eliminating these pollutants using bioengineering. Evolution has already selected for bacteria which metabolize<br />
EDC’s in polluted rivers, so a simple search for genes which can degrade toxic chemicals in the environment has<br />
proven fruitful. Two main approaches used were to either first select for these unusual bacteria by plating collect<br />
15
mud samples on media which only contain our chosen EDC’s as a carbon source, then sequence the gene, or the<br />
extraction DNA from the soil first and insertion into E. coli using a special type of plasmid afterwards. In addition,<br />
we explored the ability of a class of enzymes called cytochrome p450’s to degrade these chemicals.<br />
Towards the Use of Formal Methods for Dexterous Robotic Manipulation<br />
Sandeep Chinchali<br />
Mentors: Joel Burdick, Scott Livingston, and Ufuk Topcu<br />
We apply advances in formal methods and symbolic motion planning to synthesize robust, generalized control<br />
strategies for the canonical problem of finger gaiting in dexterous manipulation. In this scheme, fingers on a hand<br />
move continuously to re-orient an object until they encounter joint limitations, at which point the constrained<br />
fingers must be repositioned while always maintaining a force closure (FC) grasp.<br />
The problem of gaiting an object in the midst of external perturbations is modeled as a two person Generalized<br />
Reactivity (1) game between a control system (the robotic hand) and adversarial environment. Desired behavior of<br />
the system is specified using linear temporal logic (LTL) and a correct-by-construction control strategy in the form<br />
of an automaton is synthesized. Finally, results from a kinematics simulation of a planar robotic hand manipulating<br />
an object are presented to illustrate the efficacy of this approach.<br />
Purification of a Photon State Using Linear Optics<br />
Soonwon Choi<br />
Mentors: John Preskill and Netanel Lindner<br />
Many optical implementations of quantum information processing protocols require high quality on-demand singlephoton<br />
sources. Such photon sources must exhibit high collection efficiency and high purity for output photons. A<br />
promising route to such sources is based on solid state structures. A solid state based photon source exhibits good<br />
single-photon collection efficiency, but yields photonic states which are mixed in their spectral degrees of freedom.<br />
In this research, we demonstrated a simple method to improve the purity of a photon state using only linear optical<br />
devices, measurements, and post-selection conditioned on measurement outcomes. This restricted set of<br />
operations considered was chosen for simplicity and experimental relevance. The method we study utilizes the<br />
bosonic statistics of photons. The simplest example in which an improvement is achieved relies on interfering two<br />
photons in a 3 port interferometer. We shall further study possible optimizations and generalizations of this<br />
method.<br />
Numerical Simulation and Analysis of Darcy's Law in Fuel Cells<br />
David Choy<br />
Mentors: Guillaume Blanquart and Gerry Della Rocca<br />
Fuel cells are one of the most promising renewable energy technologies. Through liquid and gas transport,<br />
electricity is generated by a reduction reaction. The dynamics behind the gaseous molecular transport in microscale<br />
porous media are not well-understood in that an often-used law, Darcy’s law, is only an empiricism derived<br />
from simple channels and pipes. The validity of Darcy’s law in all structures, including complex geometries are the<br />
focus of this project. By adding a mask to a grid of cells, thereby creating solid and porous sections in a structure,<br />
various porous media were generated in which Darcy’s law was tested. With the aid of Fick’s law, Darcy’s law was<br />
compared to results from Monte Carlo simulations in the same structures. The flux through the structure provided a<br />
diffusive constant and a random walk Monte Carlo algorithm found the tortuosity used in the comparison. Data<br />
shows that Darcy’s law is valid for all structures at all logical porosities.<br />
Investigation of Different Isotope and Chemical Patterns of Pyrite Oxidation From the Río Tinto SW,<br />
Spain<br />
Justin Christensen<br />
Mentor: Max Coleman<br />
The Río Tinto, a river in southwest Spain, has become a premier site for Mars-analog research due to its highly<br />
acidic waters and sulfate rich nature. Influences from extremophile organisms on the isotope composition of sulfur<br />
and oxygen provide a unique opportunity for biomarker characterization. Our objective was to use oxygen isotope<br />
data collected from various sources on the Río Tinto system to determine the source of sulfate oxygen (either<br />
water or atmospheric), and the biological influence on the oxidation pathway and resulting isotopic signature.<br />
Preliminary data show ∆ 17 O ranging in values from +0.1‰ to +0.13‰. This is unexpected, since it indicates<br />
neither oxygen incorporation in sulfate from atmospheric (-0.35‰) nor water oxygen (0‰), the only two possible<br />
sources. Future δ 18 OSO4 measurements are planned and the results used as an additional indicator of source<br />
oxygen. As further data become available the relationship between δ 18 O and δ 17 O can be compared against the<br />
existing literature, δ 17 O = 0.52 δ 18 O + ∆ 17 O, and a refined slope specific to sulfate processes defined. In addition<br />
these data suggest the presence of sulfite intermediaries and isotope exchange processes during oxidation. Further<br />
data will help identify exact oxidation pathways and microbial influences.<br />
16
Constructing a Flexible Phage Display System for Characterizing the Peptide Substrate Specificity of the<br />
Bacterial Oligosaccharyltransferase<br />
Tinyi Chu<br />
Mentor: William Clemons<br />
The N-linked glycosylation is a post-translational modification of proteins containing the conserved sequon Asn-X-<br />
Ser/Thr. It is implicated in protein folding, trafficking and quality control, and ultimately exerts an impact on<br />
diverse physiological processes. The oligosaccharyltransferase (OST) is a membrane protein involved in<br />
transferring the oligosaccharide from the lipid-anchored oligosaccharide to the sequon on target protein. In<br />
general, All OSTs have their glycan donor specificity and acceptor peptide specificity. Recently, several bacteria<br />
species have been found to contain the N-linked glycosylation system, and their OST homologs, PglB, have been<br />
shown to have higher selectivity at the acceptor peptide sequon in addition to the requirement of Ser/Thr at the +2<br />
site. By constructing a flexible glycophage display system with high efficiency and controllability, we aim at<br />
characterizing the acceptor peptide sequon specificity of these various PglBs, including the pglBs with their saltbridge<br />
artificially mutated. The components of the phage display system are under construction and the efficacy of<br />
the system will be verified by in vivo glycosylation. The data provided by this phage display system will underlie<br />
the downstream characterization of the N-glycan structures from the corresponding species and their roles in<br />
virulence and cellular physiology. Furthermore, it may hopefully provide insight for developing the bacterial<br />
glycoprotein synthesizing system.<br />
AMMOS Cost Model Development<br />
Akash Chudasama<br />
Mentor: Eleanor Basilio<br />
AMMOS Cost Model Environment (ACME) is a model development project being undertaken by the GST (Ground<br />
Segment Team). The main purpose of ACME is to provide models that allow for the costing of the AMMOS<br />
(Advanced-Multi-Mission Operations System) tools and services suite to systematically implement mission<br />
operations to strategically enhance mission efficiency, reliability, and cost effectiveness. The ACME models will be<br />
used primarily during the Formulation phase to provide general cost estimates to allow for the appropriate design<br />
of a mission.<br />
This project focuses on the development of the Data Management element of Mission Operations, which is<br />
responsible for the end-to-end data flow during the life-cycle of the mission. Through detailed research and<br />
customer interviews, an algorithm is developed focusing on key mission elements and cost drivers. Having<br />
formulated an algorithm, a cost model is constructed and implemented into the larger ACME suite to yield cost<br />
estimates in the form of Full Time Employees (FTEs). This tool is relevant in providing insight to determine the<br />
required resources and feasibility of a proposed mission concept before the mission is implemented.<br />
Determination of Simple 3-Dimensional Abelian Surfaces With Complex Multiplication With Field of<br />
Moduli as Q<br />
Daniel Chun<br />
Mentor: Matthias Flach<br />
In this paper, I will be generalizing the results of Murabayashi's paper "The field of moduli of abelian surfaces with<br />
complex multiplication" to the case of 3-dimensional abelian varieties with complex multiplication. Furthermore, I<br />
will be showing some computational results involving computations of Hecke chracters for the 3-dimensional<br />
abelian varieties with complex multiplication which satisfy the condition of having its field of moduli coincide with<br />
the field of rational numbers. Basic results from number theory and paraphrasing of Murabayashi's arguments have<br />
been enough in achieving these aims. How- ever, for n-dimensional abelian varieties with complex multiplication<br />
where n is greater than or equal to 4, our approach here might not apply so well as there are far more caseworks<br />
and complications to be accounted for. Thus, for arbitrary n, a new method of attack seems well-advised.<br />
Predicting Lift-Off Height in Turbulent Lifted Flames Using Stochastic Flamelets<br />
Emmet M. Cleary<br />
Mentor: Guillaume Blanquart<br />
Autoignition is often considered the most crucial mechanism in predicting lift-off height. In this work, lift-off height<br />
of a turbulent ethylene jet flame in hot coflow is modeled numerically using a stochastic flamelet approach.<br />
Chemistry is modeled using a reduced mechanism for ethylene combustion, and is used to track ignition with a<br />
progress variable measuring CO2, CO, H2O, and H2 species. This progress variable is shown to be sufficient to track<br />
ignition even in the diffusion combustion regime, contributing only small errors to the ignition delay time. Modeling<br />
with stochastic flamelets takes temperature and the progress variable as deterministic parameters with a stochastic<br />
scalar dissipation rate. The results from the Direct Numerical Simulations (DNS) of a turbulent lifted ethylene jet<br />
17
flame reported by Yoo et al. (Proc. Comb. Inst., <strong>2011</strong>) are used to find the decay rate of the mean dissipation rate<br />
and the normalized variance. The stochastic flamelet approach is found to predict the mean lift-off height and its<br />
fluctuations reasonably well.<br />
An Examination of the Spatial Distribution of Carbon Dioxide and Systematic Errors Within<br />
Algorithmic Retrievals<br />
Brennan Coffey<br />
Mentor: Mike Gunson<br />
With the advent of technologies such as vehicles compounded with general population increases, the<br />
concentrations of CO2 have steadily risen. This increase, which is being closely studied, has ramifications<br />
throughout the biological world. Through applications of spectroscopy, and radiative transfer amongst others, it is<br />
possible to calculate how many molecules of CO2 lie in a defined column of air. However, other molecules are<br />
present in the atmosphere, such as aerosols and water, which diffract the light. Understanding the diffraction<br />
geometry and path length is vital to computing the concentration of CO2. Comparing these satellite readings with<br />
ground-truth data (TCCON) the systematic error contingent on the presence of other particles can be assessed.<br />
Once the error is understood, it can be scaled for in the data retrieval algorithms to acquire a set of data, which is<br />
closer to the TCCON readings. Using this process, the algorithms are being developed to be perfect, within .1%<br />
worldwide of the true value. At this stage, the accuracy is within 1%, but through amending small issues contained<br />
in the algorithms, such as accounting for the nonlinear path of sunlight, the desired accuracy can be achieved.<br />
Using Rubredoxin From Pyrococcus furiosus as a Model for Determining Oxidation States of Individual<br />
Fe Atoms in Nitrogenase as a Means for Probing the Mechanism of Biological Nitrogen Reduction<br />
Connor W. Coley<br />
Mentors: Douglas C. Rees and Limei Zhang<br />
Nitrogenase is the enzyme responsible for biological nitrogen reduction. The oxidation states of individual Fe in<br />
nitrogenase can be assigned using the multi-wavelength anomalous diffraction technique, which is critical to<br />
understand the functional mechanism of nitrogenase. The single Fe-bound Rubredoxin from Pyrococcus furiosus is<br />
used as a model for optimizing data collection and data analysis for such a goal. In this project, the wild type<br />
Rubredoxin from Pyrococcus furiosus has been successfully overexpressed, and purified with high homogeneity<br />
using anion exchange and size-exclusion chromatography. Crystallization methods have also been refined, yielding<br />
single crystals up to 1 mm in diameter through a combination of microseeding and macroseeding. Any high quality<br />
crystals produced in this study will be applied to refining data collection strategy in the future. Analysis of multiwavelength<br />
anomalous diffraction data of PfRd collected previously will also help to improve the algorithms for data<br />
analysis in the subsequent studies.<br />
Structure Determination of Two Commensal Colonization Factors in Bacteroides fragilis<br />
Skylar Cook<br />
Mentors: Sarkis Mazmanian and Sung-Eun Lee<br />
The human gut is home to an immense group of symbiotic bacterial species. Bacteroides fragilis is among the most<br />
numerically prevalent colonizers of the human gastrointestinal tract, however the mechanism of bacterial<br />
colonization remains unknown. Recent studies by the Mazmanian group have identified a highly conserved genetic<br />
locus required for colonization in B. fragilis, comprised of 5 “commensal colonization factor” (ccf) genes. The<br />
Mazmanian group has shown that B. fragilis ccfC and ccfD gene knockouts are easily outcompeted when challenged<br />
by wild-type B. fragilis; they are unable to maintain saturation of their anatomic niche. In this study we attempt to<br />
determine the structure of these two proteins.<br />
Optimization of Vertical Axis Wind Turbine Arrays Using Numerical Simulations<br />
Anna Craig<br />
Mentor: John O. Dabiri<br />
Vertical axis wind turbines (VAWTs) are currently under investigation as a potential means by which to make wind<br />
energy a more economically, socially, and environmentally viable option. Although individual VAWTs are less<br />
efficient than other types of turbines, VAWTs can be placed closely together due to their small aerodynamic<br />
footprint. Additionally, the disturbances generated by the VAWTs may be able to be beneficially employed via<br />
proper turbine arrangement in order to increase the individual power coefficients of the VAWTs in the array. The<br />
resulting high power output per unit land area of a VAWT wind farm eliminates the need to build very tall turbines<br />
in order to reach the greater wind resources at higher altitudes. A VAWT wind farm therefore has less aesthetic,<br />
acoustic, radar, and environmental impacts than wind farms with other types of turbines.<br />
Our work has focused on creating a computationally simple potential flow model of the time-averaged flow around<br />
a single turbine. Using this model, we are able to rapidly optimize different arrangements of turbines in order to<br />
find one that maximizes the power output per unit land area of the total array.<br />
18
Enantioselective Friedel-Crafts Alkylation Reaction Substrate Scope<br />
Nadine P. Currie<br />
Mentors: Sarah Reisman and Lindsay Repka<br />
The Reisman laboratory is interested in the development of enantioselective Friedel-Crafts alkylation reactions to<br />
prepare unnatural tryptophan derivatives. Recently, Reisman and coworkers found that C(3)-substituted indoles<br />
and 2-amidoacrylates react in the presence of Tin(IV)Chloride and +(R)-BINOL to give pyrroloindolines with high<br />
enantioselectivity in a single step. The proposed mechanism for the formal (3+2) cycloaddition includes an<br />
important enantioselective protonation which increases the enantioselectivity of the pyrroloindolines formed. Based<br />
upon this hypothesis of the pyrroloindoline mechanism, it was anticipated that Friedel-Crafts alkylation reactions<br />
could be studied using other kinds of substrates. Since the Friedel-Crafts products only have one stereocenter, the<br />
protonation must be enantioselective. Thus, using a variety of C(2)-substituted indoles, the enantioselectivity, and<br />
therefore the reactivity, of the Friedel-Crafts alkylation reactions can be better understood. In order to carry out<br />
these enantioselective Friedel-Crafts reactions, certain C(2)-substituted indoles must be prepared first. Many of<br />
these substrates are difficult to prepare, for instance, some substrates form dimers, or decompose on silica gel.<br />
Many of the C(2)-substituted indoles have been shown to react with the acrylate to give Friedel-Crafts alkylation<br />
products with high enantioselectivity.<br />
List-Coloring Graphs Fixing a Particular Union<br />
Megan Cutrofello<br />
Mentor: Richard Wilson<br />
There are many known results regarding list-colorings of graphs. These results, however, only consider the size of<br />
each list used in the assignment; list-chromatic number does not take into account the total number of colors used<br />
in the assignment. This project fixes the union and asks what the consequences are when each list must have m<br />
elements and also there must be exactly n colors used in total. I call this the list-chromatic number of a graph G<br />
fixing n. Many questions arise from this definition. For example, is there always a graph G that has list-chromatic<br />
number m fixing n for any m
Small Crater Analysis of the Mars Science Laboratory Landing Site<br />
Mackenzie Day<br />
Mentor: John Grotzinger and Fred Calef<br />
Gale Crater was chosen as the landing site for the Mars Science Laboratory (MSL) Curiosity rover, due to arrive at<br />
Mars in early August, 2012. The landing ellipse contains a number of interesting features for exploration,<br />
documented by Anderson and Bell (2010). These include an alluvial fan marked by low thermal inertia, a more<br />
distal alluvial fan marked by high thermal inertia, and less well defined materials of the surrounding plains. We<br />
explored the stratigraphic relationships between the high and low thermal inertia alluvial fans and surrounding<br />
geologic units within the 20x25 km landing ellipse at Gale Crater by analyzing over 50,000 small craters (1-500 m<br />
diameter). With this data, we created a crater retention age map of the landing site based solely on crater<br />
observations to help calibrate the geologic map units delineated by Anderson and Bell (2010). By calculating crater<br />
retention ages, we estimated formation time of the various units to constrain the geologic history of the region.<br />
Comparing calculated ages with the morphology and stratigraphy, we inferred which surfaces are more indurated<br />
and the time of their formation. Information gathered in this project about the terrain of the landing site could help<br />
in planning the MSL Curiosity rover traverses and targeting strategies.<br />
Examining the Neurobiology of Facial Processing: A Pattern-Classification Based Approach<br />
Archy de Berker<br />
Mentors: Doris Tsao and Sebastian Moeller<br />
Primates have a dedicated neural system for facial recognition. This system contains a series of 'face patches' in<br />
the temporal and frontal lobes, identified with functional MRI (fMRI) and elucidated by microelectrode recordings<br />
from single cells. Single-cell analyses have shown neurons responsive only to faces and tuned to a variety of<br />
dimensions, allowing a population response to encode information such as the identity and orientation of the<br />
viewed face. These recordings also reveal qualitative differences in the representation of faces in the different<br />
patches; in some, the representation of identity appears to be view-specific, and others, view-invariant. This<br />
project applies classification algorithms to extant single-unit recordings and demonstrates that the identity and<br />
orientation of faces viewed by a Macaque monkey can be decoded from the population response of neurons in each<br />
face patch. We also confirmed that the patch AM (anterior-medial) encodes identity independent of view- a<br />
classifier trained on one view can decode identity from a response to a different view, a property not seen in the<br />
other patches. We then conducted an fMRI experiment, imaging the neural response to pictures of four individuals<br />
over five orientations, and applied similar classification techniques (multi-voxel pattern analysis, MVPA) to the<br />
scans. We again achieved decoding of orientation and identity, and demonstrated differences in decoding potential<br />
across the face patches, albeit with significantly lower classifier performance. Our results highlight the potential<br />
benefits and drawbacks of lower-resolution, non-invasive imaging techniques when attempting to functionally<br />
characterize interesting regions of the brain.<br />
Scanning Tunneling Spectroscopic Studies of Impurity Resonances in Epitaxial Films of Topological<br />
Insulator Bi2Se3 on Si(111)<br />
Charles de Bourcy<br />
Mentor: Nai-Chang Yeh<br />
Bi2Se3 is a three-dimensional topological insulator whose electronic properties are characterized by a large bulk<br />
bandgap and a single gapless Dirac cone (linear energy versus momentum dispersion relation for all directions of<br />
momentum) on the surface. Our group’s scanning tunneling spectroscopic studies of Bi2Se3 samples grown by<br />
molecular beam epitaxy on Si(111) substrate have revealed resonances in the local density of electronic states<br />
versus energy at isolated impurities. Prompt vanishing of the resonance with recovery of the Dirac cone as we<br />
move away from an impurity site illustrates the topological protection of the electronic surface state. Consistent<br />
with theoretical expectations for Dirac fermion tunneling spectra, the intensity of the impurity resonances<br />
decreases rapidly with decreasing sample thickness because of the increasing difference between Fermi energy and<br />
Dirac-point energy. Consistent with the expected breakdown of the Dirac-type dispersion relation in the thinsample<br />
limit due to coupling of boundary modes from the opposite surfaces, no impurity resonances are observed<br />
for samples thinner than 6 quintuple layers. To complete our study, we measure the precise Dirac energy in our<br />
samples by placing them in magnetic fields of different strengths and identifying the field-independent zeroth<br />
Landau level, and we fit theoretical simulations to the resonance peaks.<br />
Multi-Phase Coexistence in Fatty Acid Membrane Systems<br />
Anik Debnath<br />
Mentors: Jack Szostak, Itay Budin, and Rob Phillips<br />
Membranes composed of monoacyl lipids such as fatty acids, are putative models for primitive cell membranes.<br />
Compared to phospholipids, fatty acids feature significant aqueous solubility, which allow them to rapidly exchange<br />
between discrete membranes. These exchange processes have been proposed as a mechanism by which early cells<br />
could have competed with each other and evolved in a Darwinian manner. Here we ask if fatty acid exchange<br />
occurs purely via a solution phase of monomers, or also involves a coexisting micellar phase. We introduce a novel<br />
20
fluorescent assay based on the spectral shift of the fatty acid analogue Laurdan, which can distinguish between<br />
bilayer and micellar aggregates based on these structures’ head group solvation. We use this assay, along with<br />
surface tension measurements, to characterize the aggregate phase behavior of fatty acids as functions of<br />
concentration and pH. We find that fatty acid membranes do feature a coexisting micellar phase and that the<br />
equilibrium between these two aggregates is dependent both on enthalpic (head group protonation state) and<br />
entropic (the total fatty acid concentration) factors. We also use this assay to characterize the kinetics of the rapid<br />
pH-driven transition between micelles and vesicles. Our results indicate that fatty acid membranes exist as<br />
complex solutions of heterogenous aggregates in rapid equilibrium.<br />
Determining the Link Between Tortuosity and Recoverability in Vertically Aligned Carbon Nanotubes via<br />
Image Analysis<br />
Elizabeth R. Decolvenaere<br />
Mentors: Julia Greer and Shelby Hutchens<br />
Vertically aligned carbon nanotubes (VACNTs) are nominally vertically aligned arrays of carbon nanotubes (CNTs)<br />
grown perpendicular to a substrate on the order of tens of microns tall. Multiple different techniques exist for<br />
growing such structures, and many of these produce materials with markedly different microstructures. One aspect<br />
of this microstructure is the tortuosity or ‘curviness’ of the individual CNTs. This tortuosity is expected to play a<br />
role in both the intertube interactions and load-bearing capabilities, thus affecting the resulting mechanical<br />
properties of the structure. The mechanical properties of VACNTs are of great interest in applications related to<br />
energy absorption and dissipation (such as drop-shock cushioning for electronics). Therefore determining a link<br />
between observable physical geometry and mechanical properties, especially recoverability, is of great importance<br />
in the ultimate goal of tuning mechanical properties toward match specific design constraints. In this study,<br />
VACNTs grown by three different methods in three different labs were obtained and tested mechanically via<br />
nanoindentation using a flat-punch, and the results show highly variant properties across these three growth<br />
methods. Visual inspection reveals qualitative differences in the tortuosity of each type of sample. Using highmagnification<br />
scanning electron microscope (SEM) images, a semi-quantitative figure of merit for the tortuosity of<br />
the nanotubes in these structures is determined both by human analysis of the images and by an automated<br />
program. The results of each method are compared.<br />
Electrochemical Characterization of Model Thin-Film Metal Electrodes for Solid Acid Fuel Cells<br />
Rachel Deghuee and Jeffrey Kowalski<br />
Mentors: Sossina M. Haile and Aron Varga<br />
To date, platinum is the catalyst of choice in intermediate temperature fuel cells including solid acid fuel cells<br />
(SAFCs), due to its compatibility with the solid acid electrolyte (CsH2PO4) and high catalytic activity. We can infer<br />
mechanistic pathways and extract fundamental rate parameters for hydrogen electro-oxidation and oxygen electroreduction<br />
via AC impedance spectroscopy measurements of electrode performance of model thin film electrodes.<br />
Thin films were deposited via sputtering and metal organic chemical vapor deposition (MOCVD) with varying<br />
thicknesses. In addition, asymmetric cells were fabricated with high performance counter electrodes and small<br />
working electrodes by masking the electrolyte with Teflon tape. A good isolation between the electrodes allows for<br />
the assumption that all measured electrode resistance is due to the reaction at the working electrode. Both the<br />
forward and backward electrochemical reactions were sampled via biased measurements. The thin film morphology<br />
and microstructure were characterized via SEM, EDS, and XRD in order to correlate with the specific catalytic<br />
activity.<br />
The Hopf Algebra of Rooted Trees With Natural Growth by Fan Graphs<br />
Colleen R. Delaney<br />
Mentor: Susama Agarwala<br />
In Connes and Moscovici (1998), the authors define a Hopf algebra of formal vector fields on the frame bundle of<br />
an n-dimensional manifold. In the dimension one case, this Hopf algebra on formal vector fields is closely related to<br />
the Hopf algebra of rooted trees used in the renormalization calculations of quantum field theory. We present a<br />
new Hopf algebra that generalizes the Hopf algebra of rooted trees under growth by one vertex to allow growth by<br />
families of fan graphs of any number of vertices.<br />
We identify the Hopf algebra from both the perspective of the rooted trees and the formal vector fields. This new<br />
Hopf algebra can be studied using noncommutative geometry.<br />
21
Characterizing and Investigating Spatio-Temporal Expression of Genes Critical to Strongylocentrotus<br />
purpuratus Coelomic Pouch Development in Eucidaris tribuloides, a Distantly Related Cidaroid Sea<br />
Urchin<br />
Neal Desai<br />
Mentors: Eric H. Davidson and Eric Erkenbrack<br />
The Hedgehog signaling pathway plays an important role in coelomic pouch development during embryogenesis of<br />
the purple sea urchin Strongylocentrotus purpuratus. To determine its role within the echinoderm class Echinoidea,<br />
we aim to study this pathway in a distantly related sea urchin, the pencil urchin Eucidaris tribuloides. We are<br />
investigating where and when genes in this pathway are expressed and how those expression patterns change<br />
when cultured in the presence of cyclopamine, a chemical inhibitor that binds to and stabilizes the ligand of the<br />
Hedgehog signaling pathway. Spatio-temporal expression profiles were obtained for the genes of interest using the<br />
quantitative polymerase chain reaction (qPCR) and whole-mount in situ hybridization (WMISH).<br />
Eclipsing Binary White Dwarfs in Archival Palomar Transient Factory Data<br />
Suhail Dhawan<br />
Mentors: Shrinivas Kulkarni and Shriharsh Tendulkar<br />
This project aims to find binary systems of white dwarfs (WDs) in which the primary and the companion eclipse<br />
each other. These are of great interest to astronomers since they offer accurate mass-radius relations and can be<br />
used as chronometers for galaxies. Such systems are exciting because their population distribution can help us<br />
determine whether or not they are progenitors of type Ia supernovae. Binaries with short orbital periods can be<br />
evidence for gravitational radiation and plausible candidates for space based gravity wave detectors. In this project<br />
we used Palomar Transient factory (PTF) photometry cross-matched with SDSS white dwarfs and looked for<br />
candidates of eclipsing binary systems. We found three white dwarf M-Dwarf binary systems that are possibly<br />
eclipsing and three variable WD-WD binaries. We also found candidates for pulsating WDs with DQ, DZ and DC<br />
spectral classification.<br />
Performance Optimization of Large-Area Bulk Heterojunction Solar Cells<br />
Marino Di Franco<br />
Mentors: Kwanghee Lee and Sooncheol Kwon<br />
Bulk heterojunction solar cells are easily fabricated layer-by-layer from solution, with the commonly available<br />
polymer poly(3-hexylthiophene) (P3HT) and fullerene [6,6]-phenyl-C61 butyric acid methyl ester (PC60BM) used as<br />
electron donor and acceptor in the device’s active layer. However, scaling the dimensions of these devices becomes<br />
problematic as the large sheet resistance of the indium-tin-oxide (ITO) electrode is rather high, leading to drops in<br />
the power conversion efficiency of the solar cell. Thus, a module based approach is taken to minimize resistive<br />
loses in each substrate, and fabrication conditions were optimized such that the solvent for deposition of the active<br />
layer, the thickness of the active layer and the time and temperature of thermal annealing led to the highest<br />
attainable device performance.<br />
Magnetostratigraphy of the Late Cretaceous Rocks in the James Ross Basin, Antarctica, to Investigate<br />
the K/Pg Mass Extinction<br />
Matthew Ross Diamond<br />
Mentors: Joseph Kirschvink, Tim Raub, and Sarah Slotznick<br />
During the end of the Cretaceous, Antarctica was a place of rapid sediment collection, giving rise to high resolution<br />
stratigraphy. Taking advantage of this opportunity to study the Cretaceous-Paleogene Mass Extinction that<br />
occurred globally at the end of the Mesozoic, a scientific team with researchers from Caltech, University of<br />
Washington, University of Buenos Aires, and Laboratorio de Geología Andina has worked since 2008 to recover<br />
fossils and rock samples of this carbonate sequence. Both block and core samples were collected to be analyzed in<br />
the Paleomagnetics lab at Caltech. Using the documented Maastrichtian magnetic reversals known from other parts<br />
of the world, we are using a 2G SQUID magnetometer to study the paleomagnetic directions of the cores to<br />
correlate them with specific magnetic chrons, helping pin down the stratigraphic carbonate sequence to a global<br />
timeline. The steps for uncovering the magnetic direction during lithification including liquid nitrogen cooling,<br />
running alternating field (AF) demagnetization, and baking samples by 20 degree intervals up to 260 degrees<br />
Celsius. This analysis is important for understanding the biostratigraphic and geochemical data that the researchers<br />
at the University of Washington are studying. Geologists are currently interested in knowing when individual<br />
species went extinct at this time and understanding the global environmental conditions leading up to the Cenozoic<br />
in order to investigate the factor leading up to the K/Pg mass extinction. While many scientists link the volcanic<br />
eruptions of the Deccan traps or the Chicxulub meteor impact to this mass extinction, we still do not understand<br />
the kill mechanism that caused 75% of the species on Earth to die. Knowing the paleomagnetic directions of these<br />
carbonates will allow us to piece together information in a relative time frame to better interpret ongoing and<br />
future findings.<br />
22
Localization of Lipid Biosynthesis Enzymes in the Asymmetrically Dividing Bacterium<br />
Rhodopseudomonas palustris Utilizing Fluorescent-Tagged Proteins<br />
Michael Dieterle<br />
Mentors: Dianne K. Newman, David M. Doughty, and Gargi Kulkarni<br />
Hopanoids are triterpenoid lipids that are structurally very similar to sterols, their proposed eukaryotic<br />
counterparts. Fossils of triterpenoid lipids called hopanes are preserved in ancient rocks, some of which are as old<br />
as 2.7 billion years. Hopanoids are produced by both aerobic and anaerobic bacteria, such as Rhodopseudomonas<br />
palustris. While the role of these lipids is not understood, there are indications that they may play a role in<br />
asymmetrical cell division, membrane permeability and stability, and protection of the cells under stress. Based on<br />
earlier work, the Newman Lab has proposed a hopanoid biosynthetic pathway containing several genes associated<br />
with the production and transportation of hopanoids in Rhodopseudomonas palustris. Fluorescent proteins such as<br />
mCherry will be used to tag these biosynthetic proteins to determine their intracellular localization, shedding light<br />
on where hopanoids are produced and where they are integrated into the cell membranes (outer, cytoplasmic or<br />
inner-cytoplasmic) during the cell cycle. The localization patterns of the different biosynthetic proteins range from<br />
uniform throughout the cell to dense points near the poles of the wild type cell. If these patterns are affected for<br />
each protein by the absence of other proteins in the pathway, this suggests that the proteins may be forming<br />
complexes.<br />
Expression of Neuromedin B in the Anterior Cingulate Cortex and Fronto-Insular Cortex<br />
Monisha Dilip<br />
Mentors: John Allman and Nicole Tetreault<br />
Neuromedin B (NMB) is a peptide found in the gastrointestinal tract as well as in the central nervous system in<br />
humans. It works to regulate feeding, the release of thyroid stimulating hormone, as well as internal body<br />
temperatures. Both the Anterior Cingulate Cortex (ACC) and Fronto-insular Cortex (FI) are regions of the brain that<br />
have previously been found to have an effect on appetite regulation pathways. More specifically, Von Economo<br />
Neurons (VENs) are found in both the FI and the ACC and appear to carry out these pathways. Thus, NMB<br />
expression was monitored in the VENs in the FI of multiple human brains from Mount Sinai. No concrete results<br />
were observed in the SURF time period, but the VENs themselves were counted in brains from Chicago, so that<br />
VENs could be both mapped out within the brains as well as quantified.<br />
Identification and Isolation of Novel Long Noncoding RNAs Involved in Myogenesis<br />
Race DiLoreto<br />
Mentors: Barbara Wold and Georgi Marinov<br />
Advances in functional genomics have allowed the discovery of a number of interesting classes of noncoding RNAs.<br />
Long noncoding RNAs (lncRNAs) have become a focus of study because several lncRNAs have been identified as<br />
playing a role in biological processes, such as X chromosome inactivation and cell cycle regulation. I am using<br />
computational methods in to identify potential novel lncRNAs involved in muscle differentiation by evaluating<br />
conservation and protein coding capacity in novel transcripts sequenced from polyA selected myogenic mouse cell<br />
lines. I am also working to isolate several lncRNAs adjacent to MyoD, the master regulator of myogenesis, which<br />
have a putative role in regulation of myogenesis. Once isolated, I can clone these transcripts and use the<br />
constructs to assay biological function of the lncRNAs.<br />
Test Setup to Measure Properties of the Crystalline Lens<br />
Hyung Wan Do<br />
Mentors: Yu-Chong Tai and Charles DeBoer<br />
Presbyopia is a condition in which the eye’s ability to focus on near objects is weakened due to the natural course<br />
of aging. Accommodation is the process by which the eye changes its optical power in order to focus on objects of<br />
different distances. Change in the optical power is accomplished by adjusting the curvature of the lens with a ring<br />
shaped ciliary muscle which surrounds the lens. It is known that presbyopia is most affected by the accommodation<br />
ability of the lens. In order to measure accommodation in an enucleated eye, a test setup with high repeatability<br />
and accuracy was built. An intraocular lens was tested to calibrate the system and response of the natural lens to<br />
stretching inside the capsule was observed. In particular, emphasis was placed on in vitro measurement of<br />
accommodation of porcine lenses as the ciliary muscle is stretched. Eventually, this could lead to better<br />
understanding of accommodation of the eye in relation to presbyopia.<br />
Saturn Ring Data Analysis and Thermal Modeling<br />
Coleman Dobson<br />
Mentors: Linda Spilker and Estelle Deau<br />
CIRS, VIMS, and ISS, (Cassini’s Composite Infrared Spectrometer, Visual and Infrared Mapping Spectrometer, and<br />
Imaging Science Subsystem, respectively), have each operated in a multidimensional observation space and have<br />
acquired scans of the lit and unlit rings at multiple phase angles. To better understand physical and dynamical ring<br />
23
particle parametric dependence, we co-register profiles from these three instruments, taken at a wide range of<br />
wavelengths, from ultraviolet through the thermal infrared, to associate changes in ring particle temperature with<br />
changes in observed brightness, specifically with albedos inferred by ISS and VIMS. From this compilation of<br />
multiple wavelength data, we construct and fit phase curves using independent dynamical thermal models for ring<br />
structure. From these fits we test our dynamical thermal models; and from the phase curves we derive ring<br />
albedos, and derive properties of the ring particle surfaces.<br />
Analysis of Plant Wax δD to Evaluate the Conservation of Rainfall δD in Higher Plant Fatty Acids in the<br />
Indonesian Archipelago<br />
Sabrina Dodgin<br />
Mentors: Alex Sessions and Eva Niedermeyer<br />
A key part of any paleoclimate study is understanding the regional hydrologic. This cycle is reflected in the<br />
hydrogen isotopic composition (δD) of environmental water and this δD value is recorded in higher plant fatty<br />
acids. However, how the δD values are recorded in the plant waxes vary. My research focuses on comparing the<br />
δD values of marine-surface sediments’ fatty acids extracts taken from the coasts of Sumatra and Java with the<br />
δD values of the rainfall in the same area. Leaf waxes are extracted and separated; their fatty acids are stabilized,<br />
identified using mass spectrometry and analyzed using isotope-ratio mass spectrometry. The δD values are then<br />
compared to the δD values of rainfall in their respective areas. The results may be used in a paleoclimate study<br />
and will further enable a study of terrestrial plant waxes in the area. A study of terrestrial plant waxes will give<br />
information about the preceding conditions of the hydrologic cycle and past climate changes of the Indonesian<br />
Archipelago.<br />
Performance of Linear Error Correcting Codes With Anytime Reliability<br />
Yishun Dong<br />
Mentors: Babak Hassibi and Ravi Teja Sukhavasi<br />
New upper bounds on the error probability of casual linear codes under maximum-likelihood (ML) decoding are<br />
proposed. Three types of channels considered here are the additive white Gaussian noise (AWGN) channel, the<br />
binary symmetric channel (BSC), and the binary erasure channel (BEC). The new bounds have shown to widen the<br />
ranges of rates for which the union bound analysis applies for all three channels. For the AWGN channel, the<br />
Divsalar bound for block codes is adapted to the casual coding scheme using the distance spectrum of code<br />
obtained from the Toeplitz ensemble. It is shown that using the Divsalar bound, the error exponent can be<br />
improved over the union bound exponent at all rates and all channel signal-to-noise ratios (SNR). For the BSC and<br />
BEC, similar techniques are used to give tighter upper bounds on the error probability, and it is shown that the<br />
error exponents can be improved at rates that are higher than some critical rate.<br />
Due to the complexity of ML decoder over BSC, practical decoder needs to be considered. In particular, linear<br />
programming (LP) decoder by Feldman is explored. It is known that LP decoder gives error exponent for block<br />
codes. We show that Toeplitz parity check matrix with banded or log-log sparsity cannot be anytime reliable.<br />
Furthermore, extensive simulations have been done with log sparse Toeplitz parity check matrix. Results have<br />
shown that this particular matrix structure gives relatively poor performance. Hence, we search for alternative<br />
parity check matrix structures that are non-Toeplitz.<br />
Deterministically Reorganizing DNA on a Nanoscale DNA Surface Using Random Walking<br />
Shayan Doroudi<br />
Mentors: Erik Winfree, Lulu Qian, Niranjan Srinivas, and Damien Woods<br />
Aside from its essential role as the carrier of genetic information in all forms of life, DNA has proven itself as a<br />
powerful tool for creating nanoscale structures and systems in the past few decades. We (Zibo Chen, Gregory Izatt,<br />
Yae Lim Lee, Sarah Wittman, and I) propose one such system built entirely out of DNA where a single strand of<br />
DNA takes an unbiased random walk on a DNA origami surface, picks up cargo molecules it encounters, and<br />
deposits those molecules at predetermined locations on the surface. In this fashion, we can add order to a<br />
seemingly random system, fueled only by the addition and exchange of base pairs between strands. To verify that<br />
the nanoscale mechanisms work as planned, we will primarily rely on (a) atomic force microscopy, which can<br />
indicate the position of certain strands at different points in time, and (b) fluorescence measurements using a<br />
spectrofluorometer, which shows the progression of reactions over time. While our system only uses one type of<br />
cargo, it can theoretically be scaled to reorganizing a surface with many different types of cargoes, moving each<br />
cargo to its own destination.<br />
24
Ensuring Socially Beneficial Equilibration in Competitive Loan and Insurance Markets: Theory and<br />
Experiment<br />
Kathrin Eichinger<br />
Mentor: Peter Bossaerts<br />
Asymmetric information, where one side of the market knows more than the other, presents a significant issue in<br />
economics, as for a competitive market to generate optimal resource allocations for sure, it must not exist.<br />
Economic theory has sharp predictions for equilibria in these markets where imperfect information is present, such<br />
as insurance and bank lending, but they may not necessarily hold true. This project puts the theory to the test by<br />
recreating a loan market environment in the laboratory, where lenders offer loan contracts (price and interest rate)<br />
and borrowers may accept them in order to fund their projects. As there are borrowers with higher and lower<br />
chances of payoff and without any influence over which type of borrower will accept their contracts, lenders need to<br />
rely solely on the terms of the contracts to distinguish between them in order to make profit. Observing public<br />
behaviour in these markets provides information about equilibration dynamics, where nothing is known about the<br />
dynamics in a setting where convergence involves both prices and contract specifications.<br />
Graph Theoretic Approaches to the Static Membership Problem<br />
Sam Elder<br />
Mentor: Niranjan Balachandran<br />
The static membership problem is concerned with the following question: How many bits of memory are required<br />
to store a set S from a finite universe U such that membership queries of the form, “Is x in S?” can be answered<br />
with a small number of probes to individual bits in the memory? For example, one may wish to encode the<br />
correctly spelled words of a given language from the universe of letter strings of a fixed length, and one wishes to<br />
easily ask whether a given string is a word. We hope to understand the asymptotic dependence of the size of<br />
required memory on the size of U. In this work, we investigate the static membership problem for particular cases<br />
when the size of S and the number of probes allowed are both small, hoping to extend the results of Alon and Feige<br />
(2008) and Radhakrishnan, Shah, and Shannigrahi (2010). These cases lend themselves to a graph theoretic<br />
interpretation, corresponding in particular to questions about colored bipartite graphs and hypergraphs.<br />
Synthesizing a Functional Pseudomonas aeruginosa Methionyl-tRNA Synthetase Mutant Recognizing<br />
Azidonorleucine<br />
Clara Eng<br />
Mentors: David Tirrell and Brett Babin<br />
BONCAT (bio-orthogonal non-canonical amino acid tagging) is a cell-specific protein labeling technique that utilizes<br />
functionalized non-canonical amino acids to enrich proteins produced by a subpopulation of interest in a mixed cell<br />
population. This project studies the feasibility of extending the functionality of the mutations L13N, Y260L, and<br />
H301L, used to engineer an Escherichia coli methionyl-tRNA mutant (NLL-MetRS) that selectively incorporates the<br />
methionine analog azidonorleucine (Anl) instead of methionine, to other biological systems. Anl contains an azide<br />
group that reacts with alkynyl probes, facilitating isolation of tagged proteins via affinity purification, analysis of<br />
cell lysates using a protein gel, and visualization of whole cells with fluorescence microscopy. The corresponding P.<br />
aeruginosa NLL-MetRS mutant was engineered and was demonstrated to selectively incorporate Anl, though with a<br />
lower efficiency than that of the E. coli NLL-MetRS. This is an important first step toward broadening the range of<br />
NLL-MetRS mutants as a tool for cell-specific labeling.<br />
The Neutron Electric Dipole Moment Experiment<br />
Joshua Escribano-Fontanet<br />
Mentors: Brad Filippone and Adrián Pérez Galván<br />
Studies of magnetic fluctuations of the synchrotron laboratory at the California Institute of Technology with and<br />
without magnetic shielding from high permeability materials show that to increase the magnetic shielding it is<br />
necessary to use a superconducting shield. For this purpose we started work on the optimization design of a large<br />
superconducting shield made out of lead. The superconducting shield belongs to a half scale magnet package that<br />
is a prototype for a new search of the electric dipole moment of the neutron to be carried out at Oak Ridge National<br />
Laboratory. The magnet package will provide the magnetic fields for the experiment. We present studies of the<br />
cooling rates of the half scale apparatus in temperatures between 300K to 5K. We show that by reducing the<br />
masses and improving the thermal connections of the superconducting shield we increase the cooling rates and<br />
decrease the amount of cryogenic liquid needed to cool down the lead shield as compared to a previous version of<br />
the apparatus.<br />
25
Low-Energy Limits of Noncommutative Calabi-Yau Manifolds<br />
Chris Estrada<br />
Mentor: Matilde Marcolli<br />
In this project, we consider algebraic formulations of Calabi-Yau manifolds that allow us to move into<br />
noncommutative territory, with the goal of constructing a spectral triple for a Calabi-Yau algebra, and interpreting<br />
the geometric model of Chamseddine, Connes, and Marcolli as a low-energy limit of a noncommutative Calabi-Yau<br />
manifold. In particular, we focus on a similar construction for an associative algebra obtained as a quotient of a<br />
path algebra for a quiver, and use it together with an embedding of the path algebra into the graph C*-algebra in<br />
order to obtain an algebra representation on a Hilbert space, and thus a spectral triple.<br />
Development and Testing of a Multifunctional Computer Fan Wind Tunnel for Wind Energy Applications<br />
Yuyang Fan<br />
Mentors: Morteza Gharib and Daegyoum Kim<br />
It is important to study non-uniform wind velocity profiles such as gust and wind shear to better understand flow<br />
physics and to develop new wind energy technologies. However, lacking of appropriate facilities that can generate<br />
realistic wind profiles in laboratory environment makes it hard to study the energy extraction from the wind. In<br />
order to overcome this problem, an inexpensive multifunctional wind tunnel with a test section area of 16 ft 2<br />
(4ft x 4ft) has been developed. This tunnel uses a 10 x 10 array of powerful computer fans to achieve different<br />
wind velocity profiles by controlling the speed of different rows of fans using pulse width modulation (PWM)<br />
through a computer.<br />
Telemedicine and the 10-Cent Checkup<br />
Alex Fandrianto<br />
Mentors: K. Mani Chandy and Julian J. Bunn<br />
We construct the framework for a remote medical checkup by developing an otoscope, stethoscope, and<br />
electrocardiogram (ECG) device. The otoscope and stethoscope devices are adapted to Nexus One Android phones.<br />
The otoscope casing covers the phone’s camera and contains a small lens and fiberoptic cable to shine the phone’s<br />
camera light out of its aperture. Acoustic signals sensed by the stethoscope head are converted to electrical signals<br />
by a microphone in a casing attached to the stethoscope; the electrical signals are fed to the microphone input of a<br />
phone or computer. ECG recordings from the 3-electrode ECG are taken by a laptop. Recordings are saved locally<br />
with associated medical metadata and can be uploaded to the Amazon Web Server (AWS) where they can be<br />
viewed securely from a browser. The ECG data is contaminated with 60 Hz power line noise; various filters are<br />
applied to improve the signal-to-noise ratio (SNR). The filtered data can then be used to detect pulse via<br />
triggering. The system performance of each device is determined by comparing its recordings against the medical<br />
ideal.<br />
Enhancing Performance of GaAs Solar Cells at High Incident Angles With Plasmonic Surface Structures<br />
Alta Fang<br />
Mentors: Harry Atwater and Ragip Palas<br />
Improving the performance of solar cells at large incident angles is valuable because it would allow solar cells to<br />
produce the same amount of electricity throughout the day as the sun travels across the sky, rendering mechanical<br />
solar-tracking mechanisms unnecessary. By adding metal or dielectric nanostructures to the surface of thin film<br />
GaAs solar cells, the solar cells' performance can be improved due to near-field absorption enhancements,<br />
waveguiding effects, and anti-reflection-coating effects. Finite-difference time-domain (FDTD) simulations are run<br />
to find the surface structure parameters that maximize light absorption in GaAs films at large angles. The optimal<br />
structures are then fabricated and tested in order to experimentally verify the simulation results and demonstrate<br />
the absorption enhancements.<br />
Thermal Bubble Generation for Flexible Microfluidics Control<br />
Kelvin C. Fang<br />
Mentors: Changhuei Yang, S. Pang, and C. Han<br />
Although the incredible accuracy of modern silicon fabrication processes allows for much flexibility in microfluidic<br />
channel design, the difficulty in post-fabrication manipulation of the channel currently limits microfluidic chips to<br />
relatively static operations. To provide runtime control over microfluidic chips, the Yang Biophotonics group is<br />
exploring the utilization of laser-generated channel bubbles as flexible fluid actuators. By selectively heating<br />
resistive metal pads underlying the channel, we produce adjustable microvalves and micropumps during chip<br />
operation. With the development of an integrated laser sensor and focusing system, we provide complete discretion<br />
in bubble placement and growth. This technology enhances the capabilities of present and future chip designs and<br />
creates numerous possibilities in microfluidics research and applications.<br />
26
Study of Hypervelocity Impacts Using Large-Scale, Long-Term Molecular Dynamics<br />
Chi Feng<br />
Mentors: William A Goddard III and Andres Jaramillo-Botero<br />
In the hypervelocity impact regime, where the impact velocity is tens of kilometers per second, the inertial stresses<br />
of the projectile and target are very large compared to the yield stress of both target and projectile materials which<br />
causes parts of the configuration to undergo solid-liquid-vapor phase transitions, ionization, plasma formation, and<br />
hydrodynamic instabilities. We use the electron force field (eFF) molecular dynamics simulation method developed<br />
at the Goddard Group at Caltech to model hypervelocity impact. This approach uses a mixed quantum-classical<br />
approach that enables us to track the long-term dynamics of large-scale electron excitations which overcomes the<br />
scale limitations of QM methods and the adiabatic nature of conventional force field methods. These simulations<br />
should reveal the effect of impact energy on the physics of hypervelocity impact with the ultimate goal of providing<br />
experimental observables such as plasma conductivity measurements for hypervelocity impact.<br />
Laser Desorption Infrared Spectroscopy for Icy Moon Surfaces<br />
Nate Figlewski<br />
Mentor: Luther W. Beegle<br />
The mechanics of Laser Desorption are constantly being researched and updated. Mass Spectroscopy is closely tied<br />
with Laser Desorption. Higher energy desorption techniques, such as Laser Induced Breakdown Spectroscopy<br />
(LIBS), and Matrix Assisted Laser Desorption/Ionization (MALDI) have been used to analyze the elemental and<br />
chemical makeup of extraterrestrial surfaces. However, processes like LIBS restrict the spectroscopic process to<br />
reviewing elements, as more complex compounds are broken apart during the desorption process. Laser<br />
Desorption Infrared Spectroscopy (LD-IR) takes advantage of water’s high absorption levels in the mid IR range,<br />
requiring less energy, thus simultaneously reducing the energy cost on a mission as well as allowing larger<br />
compounds to remain intact. First, this experiment will find the peak absorption wavelength of water using an<br />
Opotek Opolette 3034 tunable IR OPO. Then, using the wavelength found, we will find the relationship between<br />
pulse energy and quantity of material desorbed.<br />
Search for Blazars by Detection of Optical Transients Near Gamma Ray Sources<br />
Sveinbjörn Finnsson<br />
Mentors: S. George Djorgovski and Ashish Mahabal<br />
Blazars are active galactic nuclei (AGN), powered by material that falls onto a supermassive black hole. This<br />
material forms an accretion disc close to the black hole which generates energy. An AGN is defined as a blazar<br />
when the accretion disc produces a pair of relativistic jets and one of them is pointed towards the Earth. The jets<br />
are fast outflows of energetic plasma, perpendicular to the accretion disc. The energy ranges in form, from radio to<br />
gamma rays. We searched for blazars by looking for objects that show optical variability near known gamma ray<br />
sources with no known optical counterparts. The gamma ray sources were detected by the Fermi satellite and fields<br />
near them explored. The light curves (plots of magnitude of brightness vs. time) and images of objects in these<br />
fields were analyzed to identify variable objects. This data came from the Palomar Quest (PQ) and the Catalina<br />
Real-Time Transient Survey (CRTS). Interesting objects were then further inspected by referring to the NASA/IPAC<br />
Extragalactic Database (NED) to see if there was a known radio or x-ray source near any of them. The result is a<br />
list of blazar candidates that hopefully will later be identified as such.<br />
Microstructures and Liquidus Projection of Ternary Ag-Pb-Te System<br />
Wei Jian Foo<br />
Mentors: G. Jeffrey Snyder and Hsin-Jay Wu<br />
Thirty Ag-Pb-Te alloys were prepared, and their primary solidification phases were analyzed. The projection of<br />
liquidus troughs of the ternary Ag-Pb-Te system were determined using phase diagrams of the three constituent<br />
binary systems and experimental results. There are six primary solidification phase regions. In total, there exist<br />
three terminal solid solution phases and two binary compounds, with no ternary compound. Unique microstructures<br />
were observed during precipitation of PbTe in Ag2Te primary solidification phase. High Seebeck coefficients were<br />
also observed along the primary solidification phase boundaries between PbTe and Te.<br />
Neural Representations of Value and Loss: Prediction Error, Contrast Effects, and Loss Aversion<br />
Suzannah Fraker<br />
Mentor: John O’Doherty<br />
How people learn causal relationships is a central question of behavioral psychology. When an outcome is<br />
unexpected, the contrast between what was expected and reality is called a prediction error. It is believed that<br />
associations between predictive cues and their results are strengthened with each prediction error. We have<br />
studied the effects of contrasts between expectations and outcomes through multiple avenues. First, we have<br />
looked at the hypothesis that dopamine release encodes prediction error in the brain and is responsible for forming<br />
associations. This would imply that inducing extrinsic dopamine release using drugs of abuse, for example, would<br />
27
cause learning to continue even after prediction error has been reduced to zero and thus blocking would be<br />
inhibited. We test this in nicotine addicted subjects learning to predict receipt of nicotine and natural rewards<br />
(juice). Second, we looked at the effect of reference frames on people’s willingness to take risks or exert effort for<br />
rewards of various magnitudes. We hypothesize that decreases in reward magnitude are perceived as losses, and<br />
thus that subjects who are more averse to losing money will also show greater drops in effort level when faced with<br />
a negative reward contrast.<br />
Multiphase Flow in the Presence of Hydrophilic and Hydrophobic Surfaces<br />
Joshua W. Fromm<br />
Mentors: Guillaume Blanquart and Gerry Della Rocca<br />
Effective simulation of water droplets in multiphase environments is important to understand the countless<br />
applications of multiphase flow; this study, inspired by the hope of improving reactant transfer in fuel cells,<br />
examines the methods used in multiphase simulations and the implications of those simulations. Direct numerical<br />
simulations using the NGA code are used in a wide range of environments with varying parameters. Due to errors<br />
in NGA, multiphase flow initially was unable to be properly simulated. This led to the implementation and<br />
modification of different fluid front tracking methods to optimize multiphase flow simulation. Comparisons of<br />
simulated dynamic contact angles to the theoretical mathematical value that should be obtained using a given set<br />
of parameters are also carried out. Once multiphase flow is properly simulated, the movement of water through<br />
channels with hydrophilic and hydrophobic surfaces is compared to the movement of water through non polar<br />
channels and the results are considered with respect to fuel cell applications.<br />
Effect of Surface Chemistry on Trapped Air Phenomenon in Flat Plate Impacts<br />
Matthew K. Fu<br />
Mentor: Morteza Gharib<br />
A study was conducted to better understand the dynamics of hull slamming on the water free surface. A novel<br />
Slingshot Impact Testing System (SITS) was developed to simulate hull impacts by slamming flat plates at various<br />
impact angles and velocities. The loads on the plates was observed and used to determine the maximum force of<br />
the impact. Previous studies have shown that air can become trapped between the plates and water at small<br />
deadrise angles (deadrise angles less than 5 O ) resulting in a cushioning effect. This study sought to characterize<br />
the overall effect of surface chemistry on the presence of the trapped air, specifically, if hydrophobic plate surfaces<br />
can help stabilize the trapped air beneath the plate and reduce the overall impact force. Various hydrophobic and<br />
hydrophilic surfaces were attached to composite plates and slammed into the free surface at impact velocities<br />
between 4m/s and 5 m/s and deadrise angles between 0 o and 5 o . While preliminary results show some differences<br />
in the size of cushioning effect experienced by the hydrophilic and hydrophobic plates, further analysis and testing<br />
is needed.<br />
Building Vectors to Test the Roles of miR-146a Targets Traf6, Irak1, and Stat1 in HSC Biology<br />
Prakriti Gaba<br />
Mentors: David Baltimore, Ryan M. O’Connell, and Jimmy Zhao<br />
The immune system is an intricate system that aims to destroy any harmful invaders that enter an organism. An<br />
effective immune response typically involves a process called inflammation, which coordinates the recruitment of<br />
the proper immune cell types to eradicate infection. MicroRNAs (miRNAs) are small, non-coding RNAs that have<br />
recently been shown to be important players in the immune system. miRNAs function by targeting mRNAs leading<br />
to repression of their expression. Regulation of miRNA expression and function is controlled by transcription factors<br />
that regulate the production of miRNA containing primary transcripts in specific cell types during development or in<br />
response to various environmental signals. Specifically, we will be studying the transcription of a particular miRNA,<br />
miR-146a, which is present in immune cells and upregulated further in response to inflammatory stimuli such as<br />
Toll-like receptor ligands or pro-inflammatory cytokines. Hematopoietic stem cells (HSCs) give rise to all blood cell<br />
types, including those found in the immune system. They are located primarily in the bone marrow. In prior<br />
studies, mice that are lacking in ARS2, which contributes to pri-miRNA processing, have bone marrow failure that is<br />
thought to be due to defective HSC function. This experiment suggested that the miRNA pathway is probably<br />
important in HSC function. However, the identity of the responsive miRNAs remains elusive. Now, members of the<br />
Baltimore lab have found that miR-146a functions to promote proper HSC function and may be involved in how<br />
HSCs respond to inflammatory stress conditions. However, the mechanistic basis for miR-146a’s function in HSCs is<br />
presently unclear. Here, we focus on studying whether specific targets of miR-146a, including IRAK1, TRAF6, and<br />
STAT1, contribute to its biology in HSCs before and after inflammation, and assessing whether this occurs in a<br />
combinatorial manner. Our aim is to create retroviral vectors that produce siRNAs against one, two, or all three<br />
targets of miR-146a and determine which combination can rescue the miR-146a HSC phenotype. Our current<br />
results show that we are able to successfully knock down STAT1, TRAF6, and IRAK1 using the newly designed<br />
siRNAs.<br />
28
Implicit and Explicit Time-Stepping Fourier Continuation Solvers for the Burgers’ Equation<br />
Emmanuel Garza Gonzalez<br />
Mentors: Oscar Bruno and Edwin Jimenez<br />
Burgers’ equation is sometimes used as a simplified model of the incompressible Navier-Stokes equations to test<br />
numerical schemes. Since there is no general numerical stability theory for nonlinear problems, the quadratic<br />
convective term that appears in Burgers’ equation presents a unique challenge to formulate a robust numerical<br />
method. Solvers for the Burgers’ equation were implemented by using several time step schemes and a novel<br />
methodology referred as Fourier Continuation to approximate the spatial derivatives with high order accuracy. The<br />
diffusive term was treated implicitly since it would otherwise impose severe stability constraints. On the other<br />
hand, the convective term was treated with a variety of explicit and implicit approaches. For the 1D case a solver<br />
that is second order accurate in time was implemented which does not have an implicit quadratic term as in the<br />
case of a direct application of the Crank-Nicolson scheme, leading to a variable coefficients ODE instead of a<br />
nonlinear ODE. A 2D system of Burgers’ equations was solved by using Fourier Continuation along with the<br />
Alternating Direction Implicit methodology with different formulations for the convective term. The stability<br />
properties of these schemes are analyzed and the results suggest numerical formulations for the nonlinear<br />
convective term that yield robust solvers for similar nonlinear PDEs.<br />
Micro Measurements of 3-D Deformations Using Digital Image Correlation<br />
Alexandra Gdoutou<br />
Mentors: G. Ravichandran and Shuman Xia<br />
Digital Image Correlation (DIC) provides a full-field non-contact optical method for the accurate measurement of<br />
displacements and, therefore, strains during deformation of materials, devices and structures. The method is based<br />
on the comparison of speckle images taken before and after deformation. In the present work, we have developed<br />
an innovative 3D-DIC technique, which utilizes a transmission grating and a single camera, and is capable of 3D<br />
displacement measurement with a 2D-DIC algorithm. The accuracy and validity of the method has been<br />
demonstrated through measurement of deformations of a thin inflated membrane. The pressure chamber for<br />
performing the inflation of the membrane was designed and fabricated and the experimental setup was assembled.<br />
Calibration and translation experiments were performed to establish the resolution of the technique and effect of<br />
various speckle patterns on the measurements were established. Membrane inflation experiments were performed<br />
and the digital images were acquired. The acquired digital images were analyzed using a 2D-DIC software (VIC-2D,<br />
Correlated Solutions, Inc.) to obtain the 3-D displacement data that was compared favorably with the results from<br />
finite element simulations.<br />
Symmetries of Unrooted Polygons and Multiple Logarithms<br />
Meng Ge<br />
Mentor: Susama Agarwala<br />
Multiple polylogarithms are important algebraic objects. They have applications in number theory (mixed –Tate<br />
motives) and they also apply in the calculations of Feynman diagrams in physics. Previous papers have calculated<br />
dihedral group actions on rooted polygons, namely, σ which rotates the index of the sides of the polygon by 1, and<br />
which is the action of reflection. In this paper, we study the effect of transposition, denoted as on rooted<br />
polygons. We know that for a three-gon Q, Q= Q. Using induction, we can relate the action of<br />
polygon, which is defined in previous published papers, to that of its subpolygons.<br />
of a<br />
Conditional RNA Interference in a Mammalian Cell-Free System<br />
Tianjia J. Ge<br />
Mentors: Niles Pierce, Lisa Hochrein, and Ma’ayan Schwarzkopf<br />
RNA interference (RNAi) allows for specific downregulation of almost any endogenous gene. However, gene<br />
silencing by traditional, constitutive RNAi can lead to undesired effects when studying essential genes or cellspecific<br />
silencing. Conditional RNAi would regulate gene silencing, having applications ranging from studying gene<br />
function to therapies for human disease. Here we report a system of small engineered RNA molecules that execute<br />
conditional RNAi, in which gene A is silenced only in the presence of mRNA from gene B. The mRNA from gene B<br />
triggers a conformational change in the RNA molecules, leading to RNAi-mediated silencing of gene A. The system<br />
is designed for complete sequence independence between genes A and B, allowing its application to any<br />
combination of genes. We demonstrate the conditional RNAi system in HEK293 cell lysate, which should reflect in<br />
vivo conditions, with the intent of observing conditional siRNA production in the cell-free extract. Insights into the<br />
system from this study will hopefully help with an in vivo implementation of conditional RNAi.<br />
29
Exploring Optimal Electromagnetic Follow-Up Campaigns for Inspiralling Neutron Star Binaries<br />
Detected Using Ground-Based Gravitational-Wave Detectors<br />
Alexandra Georgieva<br />
Mentors: Tom Prince and Samaya Nissanke<br />
Merging neutron star binary systems are amongst the most promising sources for the network of gravitationalwave<br />
(GW) ground-based interferometric observatories. Observing an electromagnetic (EM) counterpart will<br />
provide complementary information which will help constrain further the physics of these sources. Possible EM<br />
counterparts are short hard bursts as well as optically faint short-lived transients (powered by radioactive decay<br />
ejecta) or longer-lived radio transients. Therefore, sky location and distance measurements (including their<br />
associated errors) from the GW signal are crucial when planning any EM follow-up campaign. We assume that the<br />
underlying distribution of neutron star binaries is such that: i) their density is uniform in constant comoving volume<br />
or ii) they are associated with a galaxy and their number is proportional to the galaxy luminosity. Using the 3.5PN<br />
approximation to model the binaries’ inspiral GW signature, we produce sets of detected events for different<br />
networks comprising combinations of possible detectors (including LIGO Australia and a proposed Indian detector<br />
INDIGO). We explore how the orientation of INDIGO influences the prior distribution of detected binaries. Then,<br />
using MCMC techniques, we compute error cubes for the sky positions of the detected systems. We then construct<br />
subsamples of neutron star binaries whose EM counterparts should be detectable as GRBs and/or transients in the<br />
optical and radio. This will allow us to consider various strategies for coordinated EM follow-up campaigns.<br />
Effect of Globally Imposed Reactant Gradients on Osmotic Propulsion<br />
Anirban Ghosh<br />
Mentors: John F. Brady and Nicholas Hoh<br />
One area of interest in the growing field of nanodevices and nanotechnology is the autonomous motion of<br />
molecular motors. The osmotic motor is one such model which uses a surface chemical reaction as its method of<br />
propulsion. By creating a concentration gradient in a bath of reactive particles, the motor is able to propel itself<br />
using the induced osmotic pressure difference. Computer simulations are designed to study the motion of osmotic<br />
motors in a half reactive, half nonreactive bath at different volume fractions of bath particles and initial orientations<br />
of the motor. The simulations will implement the Brownian dynamics method to model particle interactions in the<br />
colloidal dispersion. The average displacement and change in orientation of the motor are studied over time. The<br />
results will provide an understanding for how motor particles will behave in heterogeneous, non-uniform<br />
dispersions, and may allow for more control over osmotic motor movement through the use of reactant gradients<br />
in a mixture.<br />
Barium Tagging for the Enriched Xenon Observatory<br />
Lauren Gilbert<br />
Mentors: Martin Briedenbach, Liang Yang, and Maria Spiropulu<br />
One key component of the full Enriched Xenon Observatory (EXO) will be the in-situ tagging of xenon double beta<br />
decay daughter nucleus, barium. Such a tagging technique can significantly reduce radioactivity-induced<br />
background and increase the experiment sensitivity. Current R&D focuses on firstly, removing the barium from a<br />
platinum surface with a 750 mW Nd:YAG laser, and secondly, correctly identifying the barium ions with a time of<br />
flight spectrometer. While there is not enough data to draw any substantive conclusions, the early results look<br />
promising.<br />
Coherence of Majorana Qubit in a Topological Superconducting Wire Undergoing a Phase Slip Event<br />
Samuel Goldberg<br />
Mentors: Gil Refael, Doron Bergman, and David Pekker<br />
Recently, it has been suggested that topological states of quantum matter, those determined by the global<br />
properties of the system, may provide a decoherence-free realization of quantum computing. We are investigating<br />
the properties of one possible candidate: the spin-polarized p-wave superconductor wire. The quantum information<br />
is stored as a pair of Majorana modes located at the ends of the wire that together form a delocalized fermion state<br />
of zero energy. The goal of this project is to calculate the effect of superconductor order parameter fluctuations on<br />
the state of the Majorana qubit. In particular, we investigated the effect of a phase slip event, a fluctuation where<br />
the order parameter vanishes at a point, allowing the superconducting phase to unwind. We have found that a<br />
phase slip changes the delocalized fermion number and creates a bulk quasiparticle excitation with the energy of<br />
the superconducting gap. A second phase slip returns the delocalized fermion number to its initial value and<br />
destroys the quasiparticle. However, the unoccupied component of the Majorana qubit gains a Berry phase of<br />
relative to the occupied component. Phase slip events, therefore, will compromise the coherence of the qubit.<br />
30
Thermometry of He 3 Fridge<br />
Chen (Chris) Gong<br />
Mentors: Keith Schwab and Emma Wollman<br />
The Schwab group is in the process of building and testing a He 3 fridge to measure the properties a microwave<br />
amplifier and graphene bolometer. The He 3 fridge is designed to reach 300mK in its coldest stage. My work consists<br />
of the electronics of the fridge, which are the thermometry and the cryogen level detecting circuit. The precise<br />
measurement of the temperatures at various places in the fridge is a very important diagnostic tool for<br />
understanding the heat conduction circuit in the fridge. Since the measurements are done at very cool<br />
temperature, the design principle of the thermometry circuit is to limit the heat dissipation from the sensor as well<br />
as the heat conducted through the wires. Using four-wire sensing and lock-in amplification in this circuit we are<br />
able to maintain a high sensitivity while limit the current which might generate heat. The two circuits are currently<br />
still in testing stage.<br />
Temperature Variations of Jupiter’s Atmosphere From Mid-Infrared Imaging<br />
Jennifer J. Greco<br />
Mentor: Glenn Orton<br />
Over the last several years, Jupiter has been undergoing massive atmospheric changes, in what meteorologists<br />
classify as a period of ‘global upheaval.’ One aspect of these ongoing meteorological changes involves Jupiter’s<br />
North Temperate Belt (NTB), which turned a much lighter color in December 2002. In February and March 2007,<br />
the NTB ejected 2 plumes and, in the wake of the second, returned to a darker color, distinct from its typical<br />
shade. The normally dark brown South Equatorial Belt lightened in early 2007 and returned to its normal color later<br />
that year. It lightened again in 2009 – 2010, and began reviving in late 2010. Very little is understood about these<br />
atmospheric changes. In order to better understand them, images of Jupiter taken over the past decade at midinfrared<br />
wavelengths between 8.5 and 24.8 microns are being analyzed to examine the temperature in both the<br />
upper and deeper layers of Jupiter’s troposphere, together with changes in cloud properties and the distribution of<br />
ammonia gas and the fraction of molecular hydrogen in the para- vs. ortho states. The cloud and trace-gas<br />
distributions are useful as indirect indicators of vertical motions. Results have revealed some interesting<br />
temperature and gas profile changes including a decrease in temperature of the Equatorial Zone in 2000-2002,<br />
followed by a slow temperature rise, which correlates to an increase in the temperature of the North Equatorial<br />
Belt, followed by a slow decrease. Both of these are correlated to an increase in the para-H2 fraction in the<br />
Equatorial Zone.<br />
Metal-Decorated, Nanostructured Sm0.2Ce0.8O2-� Thin Films by PLD as High-Performance Anodes<br />
Kevin L. Gu<br />
Mentors: Sossina M. Haile and WooChul Jung<br />
The primary long-term goal for solid-oxide fuel cells (SOFCs) in general is to reduce operating temperatures to<br />
intermediate ranges (600–700°C) for reasons of material stability, cost, and wider applicability. Preliminary results<br />
involving platinum catalysis on vertically-aligned columns of Sm0.2Ce0.8O2-��(SDC) have shown significantly reduced<br />
electrode impedance when compared with other morphologies such as Pt on dense SDC for intermediate<br />
temperatures. The two primary goals of this project are to explore the suitability of Ni, Ti, or Cu as an inexpensive<br />
replacement for Pt, and investigate the dependence of performance on electrode morphology. Columnar thin film<br />
electrodes of SDC are grown on Y0.16Zr0.84O1.92 (YSZ) single crystals by pulsed-laser deposition. Metal catalyst<br />
particles are then deposited by means of sputtering. Results of electrochemical and morphological analyses carried<br />
out via AC impedance spectroscopy and scanning electron microscopy are presented.<br />
Proton Exchange Membranes With Through-Plane Aligned Ion Conducting Channels: A Study of Ionic<br />
and Electric Conductivity<br />
Mengyu (Kelly) Guan<br />
Mentors: Seung-Hyeon Moon and Sung-Hyun Yun<br />
Proton exchange, or polymer electrolyte, membranes (PEM) are semi-permeable membranes used in proton<br />
exchange membrane fuel cells (PEMFC). PEMs function to separate the reactants and transport protons across the<br />
membrane while blocking electrons. Current research focuses on improving membrane properties by (i) using<br />
different ionomers and (ii) modifying the structural properties of the membranes. Sulfonated poly(phenylene oxide)<br />
(SPPO) has been shown to be a promising PEM candidate, while applying an electric field perpendicular to the<br />
membrane surface may improve proton conductivity and PEMFC performance. In order to study the effects of<br />
applying an electric field along the membrane thickness direction, a series of SPPO membranes, with a target<br />
thickness of ~50 μm, was fabricated under an applied electric field and characterized. Since electrical insulation<br />
becomes an issue with thin membranes, this study will also attempt to measure the effects of electron transport<br />
across PEMs by electrical impedance spectroscopy. Finally, the optimal membrane will be tested for PEMFC<br />
performance against a commercially available Nafion membrane to compare and further optimize membrane<br />
properties.<br />
31
Studying the ABCB7 Transport Protein From Bacteria<br />
Trisha Guchait<br />
Mentors: Doug Rees and Jens Kaiser<br />
The ABCB7 trans-membrane protein is an ATP- binding cassette transporter thought to be responsible for<br />
transporting Fe-S compounds from the mitochondria to the cytosol. To learn more about the structure and function<br />
of this transporter in order to better understand iron metabolism in humans, homologues to the ABCB7 protein<br />
were studied in a wide variety of bacteria. First large colonies of the bacteria were grown and DNA was extracted<br />
using a phenol/chloroform extraction. Once the genomes were properly identified using the 16s RNA sequence, a<br />
PCR was performed on the ABCB7 sequence. The gene is then cloned and expressed using E-coli.<br />
Amplification of DNA-Mediated Redox Signals Using Enzyme-Mediated Electrocatalysis<br />
Luis F. Guerra<br />
Mentors: Jacqueline K. Barton and Catrina G. Pheeney<br />
The ability to rapidly and accurately detect the abundance of DNA sequences and lesions is necessary for medically<br />
relevant procedures such as genomic sequencing and the detection of genetic mutations that could lead to disease.<br />
Electrochemical biosensors achieve many of these criteria at low cost, and in particular, electrochemical DNA<br />
detection via DNA-mediated charge transfer (CT) has been shown by the Barton group here at the California<br />
Institute of Technology to have unique, robust advantages over other detection technologies. However, the signals<br />
obtained by DNA-mediated CT based biosensors—in which a redox reporter is intercalated into the distal end of<br />
gold-bound DNA and is then reduced via DNA-mediated CT—can be too weak to accurately detect low abundance<br />
oligonucleotides, such as miRNA. The feasibility of using an enzyme to electrocatalyze the redox reaction and thus<br />
enhance the sensitivity of these assays was investigated; in particular, hemoglobin—an important redox-active<br />
oxygen transporter in biological systems—was utilized due to its efficiency and lack of reactive side products. It<br />
was found that hemoglobin avoided many of the pitfalls of other electrocatalytic systems (such as direct reaction<br />
with the electrode) and catalyzed DNA-mediated CT redox reactions—important prerequisites for a successful<br />
electrocatalytic system.<br />
Design of a Program for Shear Induced Polymer Crystallization Control<br />
Ashley Guo<br />
Mentors: Richard Flagan, Ruoyu Zhang, and Xerxes Lopez-Yglesias<br />
By applying an external force to melted polymer, unique properties of the resulting crystallized polymer can be<br />
observed. We present a program developed in the software platform LabVIEW® for the control of a shear induced<br />
polymer crystallization experiment. In this system, the shear force is applied by the control of pressurized gas, and<br />
temperature is controlled by a proportional–integral–derivative controller (PID controller). All features of the<br />
program are accessible from a single interface which contains controls for the entire shear experiment. The<br />
program adjusts the temperature and pressure applied to the system based on user-defined values and both<br />
collects and saves temperature, pressure, and optical data. Additional features in the program include calibration<br />
functionality, flexibility in choice of hardware, live graphing of data, and the ability to perform a shear multiple<br />
times. A report file can be saved at any point while the program is running, containing the basic details of the<br />
experiment and custom comments by the user.<br />
Cost Model Comparison: A Study of Internally and Commercially Developed Cost Models in Use by NASA<br />
Garima Gupta<br />
Mentors: Eleanor Basilio and Greg Welz<br />
NASA makes use of numerous cost models to accurately estimate the cost of various components of a<br />
mission—hardware, software, mission/ground operations—during the different stages of a mission’s lifecycle. The<br />
purpose of this project was to survey these models and determine in which respects they are similar and in which<br />
they are different. The initial survey included a study of the cost drivers for each model, the form of each model<br />
(linear/exponential/other CER, range/point output, capable of risk/sensitivity analysis), and for what types of<br />
missions and for what phases of a mission lifecycle each model is capable of estimating cost. The models taken into<br />
consideration consisted of both those that were developed by NASA and those that were commercially developed:<br />
GSECT, NAFCOM, SCAT, QuickCost, PRICE, and SEER. Once the initial survey was completed, the next step in the<br />
project was to compare the cost models’ capabilities in terms of Work Breakdown Structure (WBS) elements. This<br />
final comparison was then portrayed in a visual manner with Venn diagrams. All of the materials produced in the<br />
process of this study were then posted on the Ground Segment Team (GST) Wiki.<br />
32
Iron Based Bulk Amorphous Alloy Sample Preparation for Annealing in Magnetic Field<br />
Karan Gupta<br />
Mentor: Marios D. Demetriou<br />
This project involved the preparation of samples of inductor and transformers cores of Iron based bulk amorphous<br />
alloys for annealing in a magnetic field. Two methods of preparation were employed. The first involved casting<br />
these alloys in an RF induction melter to make disks and the other used the RDF method (Rapid Discharge forming)<br />
to prepare toroidal samples. Amorphous alloy samples were obtained from both these methods. By annealing these<br />
samples in a magnetic field we hope to further reduce energy losses from already effecient bulk iron based<br />
amorphous alloy cores.<br />
Bootstrapping Robotic Systems: Implementation and Analysis<br />
Magnus Hakansson<br />
Mentors: Richard M. Murray and Andrea Censi<br />
This project covers a very extreme form of machine learning. An agent should start without having any information<br />
at all about neither its actuators nor its sensors. The agent should then learn about its dynamics by generating<br />
uninterpreted commands and receiving uninterpreted observations. Importantly, the learning algorithm does not<br />
depend on the actual robot configuration. Following the learning phase is the active phase where the learned<br />
dynamics are used for the agent to perform a specific task. The first part of the project has been spent on creating<br />
a set of programs that was used for the experiments. The convergence of the learning has been shown from<br />
loggings. One feasible task for the agent is the detection of sensors that fails. A strategy for accomplishing this is<br />
implemented and tested. Future work also includes examining whether it is possible to create an adaptive<br />
command generator for the agent.<br />
Electrical-Mechanical Properties of Graphene Cantilever Structures<br />
Jeff N. Han<br />
Mentors: Julia Greer and Mingyuan Huang<br />
Graphene, a 2-dimensional material composed of a single layer of graphite, has shown great promise in future<br />
electronic devices and Nano electromechanical systems. Due to its ultra-high electron mobility, low mass density,<br />
and exemplary mechanical properties, graphene has the potential to become the ideal material for future<br />
developments in transistor technology, sensors, and resonators. In mass, pressure and position sensing, graphene<br />
has proven to be accurate and robust in operation. Due to high strength, stiffness, and conductivity along the basal<br />
plane, a graphene cantilever can support itself with no built-in tension and exhibits a high resonance frequency.<br />
The fabrication of the cantilever structure involves mechanical exfoliation of graphite onto a SiO2 wafer The<br />
graphene is cut into a nanoribbon and is plated with electrodes before the SiO2 beneath the graphene is etched<br />
away. This suspended graphene is cut in half with a focus ion beam subsequently creating the cantilever<br />
structures. A voltage bias between the cantilever and the Si wafer below will be modulated and corresponding<br />
frequencies will be measured in a scanning electron microscope.<br />
Ab initio Quantum Computation of THz Vibrational Modes<br />
Matthew Harrigan<br />
Mentors: Geoffrey Blake and Matthew Kelley<br />
Terahertz time-domain spectroscopy (THz-TDS) uses light in the terahertz frequency region (3-300 cm -1 ) to excite<br />
low energy modes of molecules and materials. It promises to offer a new tool to investigate molecules with<br />
applications in pharmaceuticals, security and defense, and cosmochemistry. Whereas infrared (IR) vibrational<br />
modes can be assigned empirically or with simple free-molecule quantum mechanical computations, the observed<br />
THz absorptions cannot be linked to purely vibrational modes associated with individual species. Instead, the<br />
difficulty in assignment arises from the intermolecular nature of THz vibrations. Quantum mechanical codes which<br />
employ periodic boundary conditions were used to calculate intermolecular phonon modes. Both Gaussian basis<br />
sets and plane-wave basis sets with pseudopotentials were investigated.<br />
Optimizing Forging of Bulk Metallic Glasses Utilizing Rapid Discharge Capacitive Heating<br />
Thomas N. Harris<br />
Mentors: Marios Demetriou and Georg Kaltenboeck<br />
Bulk metallic glasses (BMG’s) possess equal or superior characteristics compared to modern engineering materials.<br />
However, BMG’s can be thermoplastically formed, enabling manufacturing processes to occur at much lower<br />
temperatures and stresses. In this study, we aim to forge amorphous GHDT into different shapes using a novel<br />
device, the Rapid Discharge Forming machine. After forming, we will analyze the quality of forging and use such<br />
analysis to refine future BMG manufacturing techniques.<br />
33
Cold Friends of Hot Jupiters<br />
Monica He<br />
Mentor: Sasha Hinkley<br />
With further Doppler monitoring and a better-characterized radial velocity curve of transiting exoplanets, we may<br />
establish the existence of additional planets on ~28% of our sample. By either confirming or placing strong limits<br />
on the presence of outer companions in these systems, we can directly test migration mechanisms involving fewbody<br />
interactions, such as planet-planet scattering, Kozai migration, as well as generic secular perturbations.<br />
Indeed, even the absence of these longer-period, eccentric objects would render migration processes, driven by<br />
secular instabilities, unlikely. A dynamical analysis of such a multiplanetary system, combined with models of the<br />
inner planet’s interior density structure, can also constrain the core mass and tidal quality factor, Q, of the<br />
transiting planet. Our objective of optimizing the observing strategy to be carried out by the California Planet<br />
Search (CPS) was approached by a Monte Carlo analysis, drawing for the period of the orbit, the planet mass,<br />
eccentricity, and a time baseline. A scaling relationship for the sensitivity for the planet’s mass was found,<br />
dependent on the dispersion of the planet’s linear radial velocities. With new observations taken over the next few<br />
years collated with the radial velocity measurement database of CPS, we hope to constrain the existence of<br />
companion planets.<br />
THz Emission From Photoinduced Charge Transfer in Re(I) Carbonyl Diimines<br />
Thomas Heavey<br />
Mentors: Geoff Blake and Marco Allodi<br />
Rhenium(I) carbonyl diimines absorb light in the near ultraviolet to generate a photoexcited state with a different<br />
dipole than the ground state of the molecule. The dipole change associated with this excitation causes the<br />
surrounding polar solvent molecules to rearrange. This rearrangement perturbs the electric field on the timescale of<br />
a few picoseconds, which corresponds to electromagnetic radiation in the terahertz region of the spectrum. Using<br />
ultrafast pulsed lasers, a solution containing these molecules can be photoexcited, and the corresponding terahertz<br />
radiation can be measured. The emitted electric field can be detected with a high signal-to-noise ratio using<br />
electro-optic sampling. Additionally, the emitted radiation can be simulated and consequently analyzed using<br />
computer models. The simulation model will be improved by being made more efficient, general, and user friendly.<br />
Using Phospolipid Bilayer Nanodiscs as Platforms for Small-Angle Scattering Investigations of a<br />
Membrane Protein<br />
Adriana Hertel-Wulff<br />
Mentors: Lise Arleth and William M. Clemons, Jr.<br />
Nanodiscs provide unique opportunities for structural and functional characterization of membrane proteins in<br />
solution. These systems consist of a form-maintaining membrane scaffold protein (MSP) surrounding a<br />
phospholipid bilayer, into which a membrane protein, such as bacteriorhodopsin (bR), can be inserted. Using this<br />
approach and verifying the results with spectrophotometry, circular dichroism, as well as small-angle x-ray<br />
scattering, we investigate the integration of trimeric bacteriorhodopsin into nanodiscs, which we then use to study<br />
the structural and functional characteristics of the proton pump. We thus find that the ideal conditions for<br />
successfully creating monodisperse, well-defined nanodiscs with bacteriorhodopsin include additional, non-native<br />
lipids, such as POPC; an appropriate concentration of a mild detergent, such as octyl glucoside; and the correct<br />
stoichiometry between the MSP, lipids, and bR.<br />
Spectral Gap Scaling of One Dimensional Quantum Spin Chains<br />
Matthew T. E. Heydeman<br />
Mentor: John Preskill<br />
Frustration-free Hamiltonians which are the sum of two-qubit projection operators have been considered as models<br />
for adiabatic quantum computation. The run time of such computation is determined by the scaling of the spectral<br />
gap with system size; in general, this property is not well understood. We show that the number of free<br />
parameters in any such Hamiltonian may be reduced to three real parameters and consider the decay of the gap<br />
with respect to the number of qubits. We present numerical results from a DMRG algorithm. For certain values of<br />
the real parameters, the scaling is shown to be an inverse polynomial of the system size.<br />
Investigation of Chemotrophic Sulfur and Iron Metabolisms Under Simulated Europa Ocean Conditions<br />
Samuel H. Ho<br />
Mentor: Robert T. Pappalardo<br />
Galileo data points to the probable existence of a subsurface ocean beneath the icy crust of the Jovian moon<br />
Europa. The water ocean provides a potential model for low-temperature redox cycling involving carbon, sulfur,<br />
and iron as energy sources for chemotrophic microorganisms. The presence of a solvent (water), a molecular<br />
scaffold (organic carbon) and an energy source (geochemical disequilibria) may provide the necessary conditions<br />
for life at Europa. In this study, the terrestrial organisms Shewanella putrefaciens and Shewanella oneidensis strain<br />
34
MR-1 (facultative anaerobic bacteria isolated from high-pressure, low-temperature, and nutrient-poor<br />
environments) are tested for growth and survival through sulfur and iron reduction under simulated Europa ocean<br />
conditions. Capillary electrophoresis can be used to monitor the increasing concentrations of reduced sulfur and<br />
iron during the course of cultivation experiments and compared with negative controls. Biomolecules generated<br />
from such reactions can be characterized using gas chromatography-mass spectrometry. Scanning electron<br />
microscopy can evaluate morphologies and distributions of biominerals and extracellular components that may be<br />
generated through the activities of these microorganisms. In the interests of studying reaction rates within feasible<br />
time-scales, free energies of Europa ocean geochemical reactions were calculated using the computer simulation<br />
program REACT within Geochemist’s Workbench (GWB) 6.0. Results from this project may provide valuable<br />
information about potential habitability and target biosignatures relevant to further in situ exploration of Europa.<br />
Using RGD Peptide Polymers to Promote the Attachment of Stem Cells to Parylene Surfaces<br />
Wilson Ho<br />
Mentors: Robert H. Grubbs and Paresma Patel<br />
Stem cells have the ability to differentiate into a variety of fetal or adult cell types and thus hold great potential as<br />
a medical treatment for tissues that do not regenerate when damaged. Stem cells need to attach and differentiate<br />
on a surface, and then be held in close contact with the damaged tissue for this treatment to work. A thin sheet of<br />
parylene, a bioinert material that is FDA-approved for use in medical implants, could serve as the solid support to<br />
hold the cells in place, but stem cells will not attach to and proliferate on this surface without modification or<br />
treatment. The RGD peptide sequence has been shown to bind integrin proteins expressed on the outer<br />
membranes of cells, and coating the parylene sheet with these peptides could promote stem cells to attach to its<br />
surface. Our approach is to embed the RGD peptide in a hydrogel polymer matrix, and then coat this matrix onto<br />
parylene prior to cell seeding. A number of methods for this coating process will be tested, and the RGDfunctionalized<br />
parylene surfaces will be sent to our collaborators for biological evaluation.<br />
Instrument Design and Construction for Photoacoustic Spectroscopy and Frequency Stabilized Cavity<br />
Ringdown Spectroscopy<br />
Daniel Hogan<br />
Mentor: Mitchio Okumura<br />
Based on earlier designs of photoacousic spectroscopy (PAS) and frequency stabilized cavity ringdown spectroscopy<br />
(FS-CRDS), highly accurate PAS and FS-CRDS instruments are constructed. The PAS apparatus was designed to<br />
operate near 770 nm, which is near the A-band of molecular oxygen. PAS will enable measurements at high<br />
pressures, where line mixing can be observed. The FS-CRDS apparatus is design to produce spectra around 2 um,<br />
near important carbon dioxide transitions. The spectra produced by the FS-CRDS instrument will be sufficiently<br />
detailed to precisely observe line shapes, and may be used to find deviations from revised theories. The spectra<br />
collected will be used as reference data for remote sensing projects, such as the Orbiting Carbon Observatory,<br />
which require extremely precise line parameters to achieve their desire precision.<br />
The Heroic Leaders of Renaissance Epics<br />
Nerissa Hoglen<br />
Mentor: Kristine Haugen<br />
Renaissance writers, particularly Ariosto, Tasso, Spenser, and Milton borrowed from Virgil and each other to write<br />
heroic epics, which practice allows us to analyze patterns emerging among their works. One trend involves<br />
portraying the hero of the epic in a leadership role. Though heroes are usually considered extraordinary figures,<br />
emphasizing their leadership qualities draws attention to the essential role other people, particularly the hero’s<br />
community and the audience of readers, play in defining the character as a hero. The Aeneid makes clear the<br />
importance of the leader’s community of followers, particularly his family. Experimentation with the role of the<br />
hero, particularly Tasso’s three-way division of personalities and responsibilities, pave the way for Milton to create<br />
two very different heroic figures with Adam and Satan in Paradise Lost. Milton explores their emotions and their<br />
relationships to their communities. The contrast between the two raises philosophical questions about what it<br />
means to be a hero.<br />
Rhesus Macaque Training for Neuroprosthetic Project<br />
Erin Hoops<br />
Mentor: Richard Andersen<br />
This neuroprosthetic project has two main objectives. First, to control a robotic arm using movement-planning<br />
signals. Second, to investigate simulation of tactile feedback using somatosensory stimulation. Our experiments<br />
involve three Rhesus Macaques (macaca mulatta) performing reaching tasks in 3D virtual reality. We compare<br />
control of a robotic arm using magnetic sensor information and actual arm movements with control directly from<br />
posterior parietal signals. Obtaining similar results from both tasks will verify the functionality of our signal<br />
recording hardware and software processing. To test the plausibility of simulating tactile feedback, we will repeat<br />
the aforementioned brain-control task while stimulating the somatosensory cortex upon completion of a reach. If<br />
35
the trials with stimulation are more accurate or lead to faster learning, more research will need to be done in order<br />
to determine the most effective way to incorporate this feedback into neuroprosthetic systems. Completion of this<br />
project will lead to a more effective neuroprosthetic device, which will help many people, including amputees,<br />
paraplegics and people with amyotrophic lateral sclerosis (ALS).<br />
Developing a Microfluidic Delivery System for Silicon Nanopillar Ion Sensors<br />
Maxwell Horton<br />
Mentors: Axel Scherer and Andrew Homyk<br />
Silicon nanopillars are nanoscale cylindrical features that can be fabricated on a silicon chip. Typically, they are<br />
about 10nm in diameter and over 100nm tall. Their high surface area to volume ratio makes them extremely<br />
sensitive to their environment. For example, if nanopillars are placed in a solution, the presence of ions in the<br />
solution can affect the current flow through the pillars when a voltage is applied across the pillars. By fabricating a<br />
sensor on a chip and delivering a test fluid to it, and applying a voltage across the sensor, the ion concentration of<br />
the test fluid can be deduced. Delivering test fluid is not easy. A simple PDMS mold cannot be used, because it is<br />
difficult to align the mold with the small features on the ion sensor. A microfluidic delivery system was designed to<br />
deliver test fluid to the nanopillars. SU8 was spun onto a chip and used to define channels, which were accurately<br />
aligned with the sensor using a mask aligner. A PDMS lid was placed on top to cover the channels.<br />
Faces in Motion: The Effect of Motion on the Saliency of Faces<br />
Gladia Hotan<br />
Mentors: Christof Koch and Blythe Towal<br />
The study of human visual saliency is key to the understanding of human visual attention. It is also important for<br />
the development of biologically feasible computer vision algorithms. Although much work has been done on the<br />
saliency of both low-level features such as color and intensity, and high-level features such as faces and text, most<br />
current saliency algorithms only apply to still images. Since a biologically feasible saliency algorithm must be<br />
applicable to real-life stimuli, we investigated how the motion of faces, which most people see on a daily basis,<br />
affects their saliency. Subjects performed a free-viewing experiment in which they viewed videos of people walking<br />
in natural and choreographed settings. Analysis of eye-tracking data from the experiment indicated which factors<br />
attracted visual attention the most. The data will be used to extend the current Itti-Koch saliency algorithm for use<br />
in natural video stimuli.<br />
Investigating the Sufficiency of “Leaky Gut” to Induce Autism-Like Symptoms<br />
Sophia Hsien<br />
Mentors: Paul H. Patterson and Elaine Y. Hsiao<br />
In the past few decades, there has been a dramatic increase in the number of autism cases diagnosed. The recent<br />
California Autism Twin Study estimates that 58% of autism development risk can be attributed to non-genetic<br />
factors; within these, there is accumulating evidence for a gastrointestinal (GI) disorder. Common GI complaints of<br />
autistic patients include chronic constipation and abdominal pain, and studies of some autistic individuals show<br />
eroded GI epithelium and increased intestinal permeability. Currently, one of the most widely used mouse models<br />
for autism is the maternal immune activation (MIA) model, in which the injection of viral mimic, poly(I:C), in a<br />
pregnant female yields offspring with the core symptoms of autism: deficits in communication, increased repetitive<br />
behavior, and impaired social interactions. Our laboratory has shown that these poly(I:C) offspring have increased<br />
intestinal permeability, commonly known as “leaky gut”. I am investigating whether leaky gut is sufficient for the<br />
development of autism-like behaviors. Postnatal day 21 mice were orally administered 0%, 1%, or 2% dextran<br />
sulfate sodium (DSS), which is commonly used to induce experimental colitis in mice. The resulting leaky gut was<br />
verified by monitoring diminished weight gain, scoring stool sample consistency, assessing colon pathology in H&Estained<br />
sections, and performing a leaky gut assay (to assay leakage of fluorescently-labeled dextran, from the GI<br />
tract into the serum). Using these assays, 2% DSS was determined to be the optimal dosage, and a second cohort<br />
of mice was started on this regimen. To characterize the behavior of DSS-treated mice, five assays with relevance<br />
to autism are being used: open field test (OFT), pre-pulse inhibition (PPI), marble burying, social preference, and<br />
sociability. Preliminary results show no difference in PPI.<br />
Characterization and Structural Study of a Zinc Exporting P1B-Type ATPase<br />
Jessica Hsu<br />
Mentors: Douglas Rees and Gabriele Meloni<br />
P1B-type ATPases actively transport transition metals across the cell membrane to regulate the concentrations of<br />
essential and toxic metal ions. However, the 3D atomic structure of this class of proteins and the mechanism by<br />
which the energy generated by ATP hydrolysis is coupled to metal transport and induces conformational changes in<br />
protein structure remain to be elucidated. Deletion mutants of ZntA from P. aeruginosa have been expressed,<br />
purified, and prepared for crystallization trials in a lipidic environment to identify the conditions under which the<br />
protein can be crystallized for X-ray diffraction structural determination. Nucleotide analogs and inhibitors have<br />
also been introduced to lock the ZntA into its intermediate states to map out the catalytic cycle. Its hydrolytic<br />
36
activity has been studied in various chemical environments, including its insertion back into proteoliposomes to<br />
recover the coupling of ATP hydrolysis to the transport of its substrate. By discerning the structural properties of<br />
transition metal transporters, we will add to our knowledge of the regulation of trace element homeostasis and the<br />
origins of diseases that arise from perturbations of such activity. Furthermore, the understanding of how metal<br />
selectivity is achieved will allow for the possibility to design modulators of P1B-type ATPase pumps.<br />
Dinosaur Body Temperatures From Clumped Isotope Paleothermometry in Carbonate From Eggshells<br />
David Hu<br />
Mentors: John Eiler and Rob Eagle<br />
The nature of the physiology and thermal regulation of non-avian dinosaurs is the subject of debate and arguments<br />
have been made for both endothermic and ectothermic metabolisms. Here clumped isotope thermometry of<br />
carbonate is used to show modern bird and reptile eggshells record the body temperatures the female experienced<br />
prior to egg laying. The method is applied to well preserved cretaceous oviraptor eggshells. The body temperatures<br />
of oviraptors are determined to be 29-33C, similar to modern reptiles. Soil nodules found near the cretaceous<br />
shells estimate paleoclimate temperatures at 21-25C. Evolutionarily, oviraptors are close relatives of modern day<br />
birds suggesting avian endothermy developed late on the evolutionary timeline of birds. It also suggests most<br />
dinosaurs were ectothermic and previous endothermic-like results in larger dinosaurs are the product of<br />
gigantothermy.<br />
Engineered Underdominance: A Gene Drive System for Drosophila melanogaster<br />
Jennifer Hu<br />
Mentors: Bruce A. Hay and Kelly J. Dusinberre<br />
A proposed strategy to control insect borne diseases such as malaria is replacement of insect vectors with<br />
transgenic counterparts that are incapable of disease transmission. In general, however, transgenic insects have no<br />
fitness advantage over wild type, and hence it is naïve to suggest that transgenes conferring disease refractoriness<br />
will be efficiently spread through natural selection. This project proposes the utilization of engineered<br />
underdominance to drive transgenes into populations. In engineered underdominance, we construct a pair of<br />
unlinked lethal genes, each of which is associated with a repressor of the other lethal gene. This effectively means<br />
that only if both engineered chromosomes are present will the individual survive. The head involution defective<br />
gene (Hid) of D. melanogaster functions in programmed cell death by triggering apoptosis. We have designed<br />
underdominance constructs in which the toxin is expression of engineered Hid, and the rescue consists of<br />
specifically targeted microRNA complementary to engineered Hid. Currently, efforts are being made to build each<br />
chromosome and generate transgenics with blocking sequences such that the killing component is off, so that both<br />
chromosomes can be brought together in a single individual to test drive.<br />
Sensing Module for CMOS Based Magneto Bio-Sensor<br />
Yunqing (Alexander) Hu<br />
Mentors: Ali Hajimiri and Alex Pai<br />
Current molecular level diagnostics are primarily carried out using stationary lab-based facilities, which require<br />
large equipment, high cost and extreme complexity. The CHIC Lab has designed a frequency-shift based CMOS<br />
based magnetic biosensor chip that can detect 1-μm diameter magnetic labels. By detecting the magnetic label,<br />
biological molecules can be detected and quantified in the biosensor’s proximity. However, a supporting system has<br />
yet to be designed to make the device handheld, easy to use, and cost effective. Thus, we have developed a<br />
solution consisting of an FPGA-PCB based design to give the test setup a handheld form factor and a Graphic User<br />
Interface (GUI) to ease the user’s experience. In our design, the PCB board is directly socketed on top of the FPGA,<br />
giving the final product the size of 2.1 inches by 3 inches, with a single USB wire providing both power and data<br />
channel. The software interface is made so that non-technical individuals can conduct assays for point of care<br />
applications.<br />
Modeling and Characterizing a Novel Nitric Oxide Donor Used as a Treatment for Dry Eye<br />
Tiffany A. Huang<br />
Mentors: Julia A. Kornfield and Dennis Ko<br />
Kertoconjunctivitis sicca, or dry eye, is a common eye condition affecting millions of people worldwide that arises<br />
from a malfunction of the tear film, resulting in low levels of tear production or tears that evaporate too quickly.<br />
Research has shown that nitric oxide (NO) can stimulate blood vessels in the eye to produce their own autologous<br />
serum, which increases tear production. To improve upon current treatments for dry eye, which are often<br />
inconvenient and uncomfortable, a protein that can periodically release NO and can be administered via contact<br />
lenses would be desirable. NO was conjugated to bovine serum albumin (BSA) to produce reduced nitrosylated<br />
bovine serum albumin (r-NO-BSA). Fluorimeter analysis and gel simulations showed that NO was released for a<br />
duration of up to 3 hours in vitro. The effects of pH, thiolene ratios, and amounts of BSA on r-NO-BSA’s<br />
biodegradability were shown by SDS-PAGE analysis. S-nitrosoglutathione (GSNO) was also synthesized and its<br />
release kinetics were studied as a standard for the NO release assays.<br />
37
Comparison of Lattice Boltzmann and Direct Numerical Simulations for Single Phase Flow<br />
Yifei Huang<br />
Mentors: Guillaume Blanquart and Gerry Della Rocca<br />
Due to their benign emissions and applicability to a broad range of applications, fuel cells serve as an alternative to<br />
combustion engines for energy conversion. However, slow removal of water from the cathode has been a limiting<br />
factor to their performance. Currently, the lattice Boltzmann method (LBM), based on statistical mechanics, and<br />
direct numerical simulation (DNS), based on Navier-Stokes continuum, are two methods used to simulate liquid<br />
flow in porous fuel cells. However, we desire a thorough comparison of the two methods. We studied both<br />
simulation methodologies in two cases with existing analytical solutions: flow through a channel (Pouseuille flow,<br />
analogous to flow through a pore), and flow around a cylinder (analogous to flow around an obstacle). Various<br />
Reynolds and Knudsen numbers were used to determine the range of applicability of each simulation. Specifically,<br />
we compared the entry length and velocity profiles of the two simulations with each other and with analytical<br />
solutions. This comparison provides a better understanding of the limitations of each simulation to improve flow<br />
simulations in fuel cells.<br />
Effect of Aging on the Foam Fractionation of Lactoferrin<br />
Yuehan Huang<br />
Mentors: Julia Kornfield and Robert Tanner<br />
Foam fractionation is an inexpensive and simple technique for concentrating proteins. The foamability of a protein<br />
can drastically change with the age of the protein. The foamability of solutions created from ten year old bovine<br />
lactoferrin (bLF) protein was investigated with varying concentration protein, air flow velocity, and the pH of the<br />
solution. The results suggest the foamability of the aged protein decreased to an insignificant level except at high<br />
pH with a protein concentration of 0.1 mg/mL.<br />
Linear Programming and Clustering<br />
Aditya Huddedar<br />
Mentor: Leonard Schulman<br />
We study an affine invariant approach to Strongly polynomial algorithm for linear programming by Vempala-<br />
Barasz. We prove that the method of combination of the improving rays (with non-negative coefficients) in a<br />
particular step does not matter as far as the analysis of the algorithm for Klee-Minty and Goldfarb-Sit cubes is<br />
considered. We present a counter-example in which a weaker version of their algorithm is forced to visit some<br />
facets of the polytope more than once, unlike in Klee-Minty or Goldfarb-Sit cases. We implement and analyse the<br />
efficiency of the seeding algorithm described in “The effectiveness of Lloyd-type Methods for the k-Means Problem”<br />
by Ostrovsky, Rabani, Schulman and Swamy.<br />
Optimizations of the Quasicontinuum Method on Lattice Structure Computation<br />
Ka Kin Kenneth Hung<br />
Mentors: Michael Ortiz and Malena Español<br />
In simulation of impacting materials with nanoindentors, it is crucial to compute the resultant displacements of<br />
atoms. These displacement of atoms in a lattice structure in response to an externally applied force field can be<br />
either computed atomistically by solving a system of non-linear equations or continuously by assuming even<br />
distribution of mass. However, fully atomistic calculation requires solving millions of non-linear equations which<br />
incur a huge computational cost while continuous calculations are not accurate. In this project, I have analyzed<br />
several possible methods based on the quasicontiuum method to approximate the total energy of the system more<br />
efficiently.<br />
Determining the Neutron-Star Equation of State From Mass and Radius Observations<br />
Nathaniel Indik<br />
Mentor: Lee Lindblom<br />
Neutrons stars are former main sequence stars that have undergone gravitational collapse to a state where the<br />
density is so high that the star is primarily composed of neutrons. Since this density is impossible to recreate in a<br />
lab, the high density equation of state is poorly understood. Many models have been proposed to describe Neutron<br />
star matter, but they disagree with each other by an order of magnitude. The goal of this SURF is to develop and<br />
improve theoretical tools that can be used to infer information about the high density equation of state governing<br />
the cores of neutron stars from the current and future Neutron star observations.<br />
The interior structures of these stars are governed by the Oppenheimer -Volkoff equation. By integrating this<br />
system of differential equations we can find the total mass and radius of the star. By assuming that the underlying<br />
equation of state is determined by a few parameters, we can use it together with the Oppenheimer –Volkoff<br />
38
equations to produce a mass-radii curve to compare with the observational mass-radii curve. When there are more<br />
measurements of neutron star mass and radii, the equation of state parameters that best fit the observational<br />
mass-radii curve will tell us which model for the neutron star equation of state best fits the data.<br />
Tracking and Analyzing Avalanche-Like Flows in Granular Media<br />
Robert (Rivers) Ingersoll and Wesley Swank<br />
Mentors: José Andrade and Carlos Avila<br />
Although phenomenological models for steady flows in granular media exist, the underlying mechanics of less<br />
predictable situations, such as avalanches, are difficult to simulate and predict with existing methods. By creating a<br />
link between the precision of the discrete element method (DEM) at the grain level and the flexibility of continuum<br />
models, the mentor is developing algorithms that can simulate sudden deformations with a high degree of<br />
accuracy. The goal of this project is to provide essential physical evidence to quantitatively assess the accuracy of<br />
the mentor’s theoretical framework and of simulations developed by the co-mentor. To generate this data, a highspeed<br />
camera captures thousands of images per second of the progression of short, recurring avalanches of steel<br />
beads in a steadily rotating plexiglass drum. These images are then reduced to an array of coordinates<br />
corresponding to the centroids of each bead in the frame of view, providing velocity and displacement values in a<br />
two-dimensional plane. With analysis, this data will be used to refine and eventually confirm the mentor’s model for<br />
the early progression of an avalanche.<br />
Neural Correlate to Attractiveness Leakage From Face to Hair<br />
Janis Intoy<br />
Mentors: sShinsuke Shimojo and Chihiro Saegusa<br />
The attractiveness of the face plays a major role in social interactions and can influence economic mobility and<br />
mate choice. However, few studies have delved into the interactions between face and its natural surrounding,<br />
hair. In this study, we used functional magnetic resonance imaging (fMRI) to determine brain activity in the event<br />
that attractiveness leaks from the face to the hair. The brain regions that may be active when leakage is occurring<br />
are related to the areas of the brain that represent facial attractiveness, including the orbitofrontal cortex, fusiform<br />
face area, amygdala, ventral striatum, and superior temporal sulcus. The prefrontal cortex is also expected to be<br />
involved in leakage because it plays a role in inhibition and suppression.<br />
Modification of SAMMS for Increased Hydrophilicity and Host-Guest Interactions<br />
Anna Ivanova<br />
Mentors: Darren Johnson, Sean Fontenot, and Robert Grubbs<br />
A previously-synthesized self-assembled monolayer on mesoporous supports (SAMMS) material was modified,<br />
replacing phenyl scaffolding with beta-cyclodextrin groups in hopes of increasing the material’s hydrophilic<br />
properties. The cyclodextrin groups were also intended to improve the interaction between the scaffolding and<br />
active layer by utilizing host-guest chemistry. After successful assembly of the cyclodextrin scaffolding, thiol guests<br />
were tested with varying degrees of success. The stability of the final material in water was evaluated by<br />
monitoring thiol leaching using Ellman’s reagent.<br />
A Mechanism for Random-Walk-Based Deterministic Object Sorting on DNA Origami<br />
Gregory Izatt<br />
Mentors: Erik Winfree, Lulu Qian, Niranjan Srinivas, and Damien Wood<br />
The specific base-pairing behavior of DNA, when combined with relatively cheap and easy-to-access base-specific<br />
commercial oligonucleotide synthesis, is a powerful tool for precisely designing and constructing nanoscale<br />
structures. Through careful design, these structures can be programmed to behave in productive ways. The Caltech<br />
BIOMOD team, comprised of Zibo Chen, Shayan Doroudi, Yae Lim Lee, Sarah Wittman, and I, presents work done<br />
to prove the functionality of such a mechanism for the spatial sorting of DNA-tagged cargo on the surface of a<br />
100nm by 70nm folded DNA rectangle -- hereby referred to as "DNA origami" -- through the parallel work of many<br />
random-walking single-stranded oligonucleotides cooperating with cargo-specific goal strands. We used a<br />
combination of gel analysis, fluorescence experiments, and atomic force microscopy to consecutively verify that the<br />
many steps and components of our mechanism function correctly, independently and together.<br />
Standing Before Stepping: A Home Stand Frame for Spinal Cord Injury Patients<br />
Anum Jang Sher<br />
Mentor: Joel Burdick<br />
About 250,000 people in US today suffer from severe spinal cord injury and approximately 11,000 new injuries<br />
occur each year. Hence, a tremendous amount of effort is been put in the research of reestablishing the connection<br />
between the brain and the body or to mimic the signals that the brain sends to the body to trigger motion after a<br />
spinal cord injury. A team of researchers from Caltech, UCLA and University of Louisville is investigating the use of<br />
high density electrical epidural stimulation to facilitate locomotion recovery after severe spinal cord injury.<br />
39
However, this technique has to be coupled with rigorous physical training to be effective. In this paper, we will<br />
focus on the key design features of a stand frame that people implanted with epidural stimulating arrays can use at<br />
home to extend their physical training outside of a clinic. As this stand frame is used at home, it is designed to be<br />
safely used without a trained therapist and with minimum assistance. The main structure is stable but light for<br />
convenience. There is a foot plate, various straps for support and adjustability features for people with different<br />
physique. Sensors are installed on this stand frame to collect data to analyze the force exerted by the user and the<br />
support provided by the fame.<br />
Development of Genetic Tools in Microtubule Possessing Bacteria<br />
Pierre M. Jean<br />
Mentors: Grant J. Jensen and Martin Pilhofer<br />
The genus Prosthecobacter is a bacterial group with increasing interest since the discovery of tubulin genes<br />
encoded in its genome. These tubulins were shown to form bacterial microtubules; however, their function remains<br />
unknown. Furthermore, their study is difficult due to the lack of genetic tools available in Prosthecobacter. A<br />
random transposon, noted for its high efficiency on a broad spectrum of bacteria, was selected for transformation<br />
of Prosthecobacter vanneervenii through electroporation. An electroporation protocol was designed and optimized<br />
for the transformation, and a transposon mutant library consisting of 63 P. vanneervenii clones was generated. The<br />
transposon was amplified using specific primers to confirm the insertion in the genome of the mutants. Using the<br />
two-step gene walking method, the sequences of the regions flanking the transposon were determined. These<br />
flanking sequences were BLASTed against the P. vanneervenii genome to determine the location of the insert and<br />
whether a gene was knocked-out during the process. It has been demonstrated that it is possible to introduce<br />
exogenous DNA in Prosthecobacter, providing opportunities to further manipulate these organisms and elucidate<br />
the function of genes of interests in these peculiar and poorly understood bacteria.<br />
Growth and Characterization of Silicon Germanium Microwires for Application in Multijunction Wire-<br />
Array Solar Cells<br />
Julie D. Jester<br />
Mentors: Harry Atwater and Dan Turner-Evans<br />
Microwire arrays have the potential to revolutionize the solar energy industry by reducing solar cell production cost.<br />
Additionally, multijunction arrays can increase the efficiency of non-concentrator solar cells. SiGe alloys are lattice<br />
matched to a variety of semiconductor materials of different band-gaps, allowing them to be used in conjunction<br />
with those materials in multijuntion devices. Thus, we have set out to grow SiGe wire arrays to serve as the base<br />
of multijunction wire solar cells. As a starting point, Si microwire arrays have been grown through Vapor-Liquid-<br />
Solid (VLS) growth with some success using copper catalyst substrates. The wire array growth fidelity degrades<br />
with water or oxygen contamination and thus great pains have been taken to remove these elements from the<br />
growth reactor. The reactor has been entirely rebuilt to remove leaks, and a Residual Gas Analysis (RGA) has been<br />
used to determine the amount of contaminants present. Furthermore, the copper catalyst may form a stable<br />
silicide rather than initiating growth, resulting in degradation of array fidelity. By looking at the phase diagrams for<br />
Cu-Si and Au-Si, it is apparent that Au catalysts should not experience the same silicide producing effect. Thus, we<br />
have begun to grow with a Au catalyst. In the future, the growth parameters, including temperature and gas flow,<br />
will be modified in order to produce high fidelity (111) oriented microwire arrays. These Si microwire arrays should<br />
then be able to be used to produce SiGe alloy microwires by introducing GeCl4 gas during the growth process.<br />
These microwires can then be doped with boron to give them appropriate electrical characteristics. Producing and<br />
characterizing these boron doped SiGe microwire arrays should increase understanding so that they can be used in<br />
multijunction solar cells.<br />
Engineering cTPR-Linked Antibody-Like Reagents to Improve Neutralization of HIV<br />
Siduo (Stone) Jiang<br />
Mentors: Pamela Bjorkman, Rachel Galimidi, and Anthony West<br />
The Human Immunodeficiency Virus (HIV) infects CD4+ T-helper cells via the binding of the viral surface<br />
glycoprotein gp120 with CD4. Clinical screens have identified a class of broadly-neutralizing antibodies that are<br />
capable of binding to and neutralizing many HIV strains. The natural design of these antibodies is in the IgG form,<br />
composed of two binding sites with a separation distance of 15nm. Due to the low surface density of gp120 on HIV<br />
however, natural IgGs may not be large enough to achieve bivalent binding, or binding of both antigen recognition<br />
sites at the same time to the virus. To improve the structure of these antibodies, we have engineered with a vast<br />
new library of all-protein reagents using cTPR linkers. cTPR proteins are all-helical structural motifs consisting of<br />
34 amino acid residues per repeating unit that demonstrate high stability. Our antibody-like reagents employ noncanonical<br />
combinations of Fabs and single-chain variable fragments of some of the most potent antibodies to date,<br />
and at the same time, are designed to have increased separation distance between their dual binding sites. As a<br />
result, we expect them to have increased breadth and improved potency of HIV neutralization. When put into an<br />
adeno-associated virus, an eternal supply of these reagents can be generated and may provide an alternative to<br />
costly conventional drug therapy.<br />
40
Design and Fabrication of Optically-Linked CMOS Neural Probes<br />
Raymond Jimenez<br />
Mentors: Axel Scherer and Aditya Rajagopal<br />
We present a fabricated, robust CMOS neural probe with a radical new level of integration. Our new design contains<br />
an optical power source (PiN diode), an optical transmitter (VCSEL), and a glass pipette electrode fully integrated<br />
into one die, with a total die area of 600 microns by 600 microns square. This is a breakthrough in standalone<br />
neural probes, as prior designs have never incorporated their own power source or sensing electrodes in their<br />
design area. We also present a new, stable oscillator design for sensing the action potential while still minimizing<br />
transistor count (and area). The resulting neural probe has wide applications in neurobiology, enabling scientists to<br />
sense the state of tens, if not hundreds of neurons simultaneously, several orders of magnitude more than<br />
currently possible. Further work could easily extend our design to triggering action potentials, which could enable a<br />
new realm of control in neuroscience.<br />
Post-WWI Scientific Communication Networks as Seen Through Albert Einstein’s Correspondence<br />
Xiao (Dawn) Jin<br />
Mentor: Diana Kormos-Buchwald<br />
Einstein’s correspondence during the post WWI period grew exponentially. His correspondents also diversified,<br />
ranging from Zionist leader Chaim Weizmann to celebrated author Stefan Zweig. In my project I utilize Einstein’s<br />
immense amount of letters as my primary sources, and examine both the reach and depth of his connections with<br />
colleagues, family, politicians et al. The most important method I use is statistics. I record the identities of<br />
Einstein’s correspondents, including nationality, occupation, family backgrounds, and relations to Einstein, in<br />
addition to frequency of correspondence. Then I carry out some typical statistical analyses such as sorting and<br />
distribution analysis. My project will expectedly reveal something new about the process of the science center<br />
shifting from Germany to the outside (possibly America), and the post-WWI scientific community in general.<br />
Regulation of Attraction by Small Signaling Molecules in Caenorhabditis elegans<br />
Yea-ra Jo<br />
Mentors: Paul Sternberg and Jagan Srinivasan<br />
Small molecule metabolites are known to play important roles in biology of many organisms including<br />
Caenorhabditis elegans. In Caenorhabditis elegans, ascarosides in varying amounts represent a structurally diverse<br />
set of signaling molecules that regulate its major life history traits by urging it to undergo specific responses. At<br />
very low concentrations, this chemical acts as a potent male attractant which allows the male nematode to search<br />
for potential hermaphrodites as mates. However, the specific genes that control response to ascarosides are still<br />
unidentified. Therefore, by running behavior assays on the male worms with various neuropeptide mutations and<br />
comparing mutant nematodes’ responses to previously identified ascarosides with those of wild-type male worms,<br />
we would like to identify the genes and the proteins encoded by the genes involved in the responses to sex<br />
pheromone in Caenorhabditis elegans and gain further insight into genetic pathways governing male recognition,<br />
and furthermore, general animal communication, and self-identification among other species.<br />
Interactive Image Segmentation Toolset: An Application for Crowdsourcing a New Benchmark<br />
Segmentation Dataset<br />
Alex Jose<br />
Mentor: Pietro Perona<br />
Image segmentation is an area of study in computer vision, usually referring to computer-generated partitioning of<br />
an image into regions. The goal of segmentation is typically to simplify the information in an image from a large set<br />
of unorganized pixels to a set of more manageable and meaningful areas. In creating algorithms for image<br />
segmentation, it is useful to have a base set of reference images such that the performance of multiple algorithms<br />
can be compared. This project aims to create a segmentation toolset that will enable the Vision Lab to generate<br />
such a dataset. This toolset will consist of programs enabling humans to quickly segment images, and will be<br />
distributed through Amazon’s Mechanical Turk crowdsourcing program. The tasks will be divided among<br />
participants who will complete small portions of the overall segmentation - such as outlining part of a single image<br />
- for a small amount of money. Ideally, from this process we will not only generate a new dataset but will elucidate<br />
some of the intricacies of human approaches to segmentation.<br />
Understanding Tumor Vascular Heterogeneity Using Whole Mount Optical Microscopy and MRI<br />
Devashish Joshi<br />
Mentors: Scott Fraser, David Koos, and Thomas Ng<br />
Tumors are heterogeneous in structure and a simple biopsy is often not sufficient to capture the complex<br />
relationship between the vasculature and tumor cells. Since many drugs are administered via the blood stream, it<br />
is essential to understand the tumor's vasculature to predict drug delivery. We are developing methods to correlate<br />
mobility maps identified via MRI with the 3D spatial organization of cells revealed by high resolution optical<br />
41
imaging. MRI imaging of intravenously delivered contrast agents demonstrates the fluid dynamics within the tumor,<br />
potentially identifying regions of vasculature leakage. Optical imaging of fluorescently labeled vasculature allows us<br />
to visualize the cellular architecture of the tumor. Correlation of the images obtained from the two modalities will<br />
allow us to identify the cellular underpinnings of the fluid mobility as seen through MRI. This study will lay forth the<br />
foundation for future studies aimed at quantifying and enhancing the uptake of drugs within tumors.<br />
Rugged Encapsulation Structures and Acoustic Instrument Development for Inexpensive Medical<br />
Sensors<br />
Theresa Juarez<br />
Mentors: K. Mani Chandy and Julian Bunn<br />
Public health in developing countries is limited by insufficient and inaccessible medical care. An innovative way to<br />
address this issue is by means of a health system capable of operating without direct interaction with a physician.<br />
The method proposed here would consist of portable cellphone based medical sensing devices that can be used by<br />
members of the local society in need. The instruments collect medical data and upload them into the Cloud where<br />
they are evaluated and can be assessed by physicians. Two such devices are heart auscultation and<br />
electrocardiograph (ECG) tools, both of which exist as prototypes. Problems with these devices include structural<br />
issues. The ECG and stethoscope lack robust casings and can break easily. The key elements of this study include<br />
creating an encapsulation for the current auscultation unit and developing a combined stethoscope and ECG device.<br />
The project will also outline an investigation to find acoustic shapes better suited to microphone recording in the<br />
stethoscope.<br />
Network for Collection of Seismic Data<br />
Prathamesh Juvatkar and Prashant Borde<br />
Mentors: Julian J. Bunn and K. Mani Chandy<br />
Earthquakes are natural disasters responsible for destruction, injuries and loss of life and property. However,<br />
earthquakes are not directly responsible for loss of life. The secondary events triggered by earthquakes such as<br />
structure collapse, fires, tsunamis and volcanoes cause actual human disasters. The losses due to earthquakes can<br />
be greatly reduced if we can give early warnings of earthquake or at least detect an earthquake’s early seismic<br />
activity. For building a system capable of early detection of seismic activity, we need high density arrays of<br />
spatially distributed motion sensors. Our project is to develop a private network of portable low cost seismometers<br />
that consume minimal resources. Each seismometer consists of a Single Board Computer (SBC), “Phidget”<br />
accelerometer and a Home-plug device for power line communication. Each seismometer sends the data to next<br />
device in linear topology using power-line communication. A base-station in the network sends the collected data to<br />
a central server via the internet. The operation of the seismometer network is completely automated after<br />
installation. Also the network is robust against intermediate node failure. Large numbers of such private networks<br />
will ensure sufficient amounts of data available for the detection of earthquakes with higher accuracy.<br />
Directed Evolution of Cytochrome P450 for Enantioselective Meso-Desymmetrization<br />
Karolina Kalbarczyk<br />
Mentors: Frances H. Arnold and Ryan Lauchli<br />
Cytochrome P450 monoxygenases (P450s) are a superfamily of enzymes responsible for a range of vital functions,<br />
from the biosynthesis of secondary metabolites to the degradation of toxic substances in the liver. Engineering<br />
cytochrome P450 enzymes is a budding area of research, with the goal of optimizing the critical functions<br />
performed by P450s, in this case with a focus on the potential for enantioselective meso-desymmetrization. Testing<br />
a variety of enzymes demonstrated that several had high activity and enantioselectivity, with substrates ranging<br />
from cyclopentanes to cyclooctanes. Particularly high enantioselectivity was present with the P450 9-10A F87I,<br />
which engineered for thermostability provides an enzyme with a backbone robust to mutation. Through the process<br />
of directed evolution with high-throughput screening assays, random mutations were generated in the 9-10ATS<br />
F87I to make a library which can subsequently be screened for higher enantioselectivity with cis-1,2dimethoxycyclooctane.<br />
Analysis by gas chromatography confirms the enantioselectivity of various mutants.<br />
Identifying, Analyzing, and Creating a Minimalistic in vitro System for the piRNA Pathway<br />
Amol Kamat<br />
Mentor: Alexei Aravin<br />
The Piwi-piRNA pathway is a unique, lesser understood process for silencing of selfish, repetitive transposable<br />
elements (TEs) of the genome via de novo DNA methylation in the germline. While it is generally thought that Piwi<br />
protein is involved in the biogenesis of piRNAs, the other components are not well known, especially in mammals.<br />
We sought to determine the basic components needed for piRNA biogenesis to occur, create a heterologous in vitro<br />
system for expressing the piRNA pathway, and study the characteristics of RNAs in the pathway. We expressed<br />
three proteins (Tdrd12, MOV10L1, and mitoPLD) paired with Mili (Piwi from mice) in stable cells. Their interactions<br />
with Mili (Piwi from mice) and associated RNAs were to be analyzed via immunoprecipitation, western blotting, RNA<br />
extraction, and creating and sequencing libraries of that RNA. Western blots of the Tdrd12/Mili IP indicated some<br />
42
interaction, and RNA gel analysis suggested the presence of small RNAs in the correct size range (23-30 nt) in this<br />
interacting complex. If sequencing confirms this result, we will be on an encouraging route towards identifying a<br />
minimalistic approach to creating an in vitro system for piRNA biogenesis, which will be an incredible tool for<br />
further study. We will continue the project with the remaining two proteins in order to form a more complete<br />
result.<br />
Computational Studies of Active Site Hydration in Cytochrome P450BM3 for the Development and<br />
Understanding of P450 Catalyzed C-H Amination<br />
Arvind Kannan<br />
Mentor: Frances Arnold<br />
The design of new enzymes capable of catalyzing novel reactions not yet observed in nature is a challenging<br />
multidisciplinary task. This project seeks to demonstrate that a combination of directed evolution and<br />
computational methods can be used to convert a fatty acid hydroxylase, cytochrome P450BM3, into an efficient<br />
catalyst for C-H nitrene insertion – an important synthetic reaction that lacks an enzymatic counterpart. We have<br />
thus far identified P450BM3 mutants that bind and activate aryl sulfonyl azides (one example of nitrene precursors),<br />
but analysis of the reaction products suggests that the desired amination reaction is outcompeted by hydrolysis of<br />
the key reactive intermediate, termed aza compound I. Here we have used computational molecular dynamics<br />
(MD) simulations in order to study active site hydration in P450BM3. Our simulations have identified a long hydrogen<br />
bonding network of waters in a channel that links the active site to the exterior solvent. A set of polar amino acids<br />
that stabilize this network of waters have been identified as candidates for future targeted mutagenesis<br />
experiments, with the ultimate goal of minimizing active site hydration in the presence of the bound nitrene<br />
intermediate.<br />
Entropy-Based Systematic Clustering for Constrained Molecular Dynamics<br />
Jonas T. Karlsen<br />
Mentors: William A. Goddard and Andres Jaramillo-Botero<br />
We propose using local entropy to investigate the dynamics of proteins, and we explore a simple way of utilizing<br />
this approach to develop systematic molecular clustering for constrained molecular dynamics. We interpret the<br />
notion of local entropy in terms of the physics of the quantum harmonic oscillator approximation, which is used to<br />
calculate entropies from the density of states that is the Fourier transform of the velocity autocorrelation function,<br />
which is obtained from a 20 ps molecular dynamics run. Based on our studies we propose a working hypothesis,<br />
stating that the local entropy of the backbone of a protein is a relative measure for how dynamic that part of the<br />
protein is, and that high entropy areas are the most likely to undergo large conformational changes in long term<br />
dynamics. This hypothesis implies that a short simulation from which entropies are calculated, contain information<br />
about the system on timescales several orders of magnitude larger.<br />
Object Recognition and Localization Using Audio and Tactile Sensors<br />
Robert Karol<br />
Mentor: Joel Burdick<br />
Due to the rapid increase in the number of robotic devices used in society, there is a strong desire to design<br />
hardware and software capable of being used in a variety of different situations. Reusable hardware and software<br />
would allow for robots to be more independent, and allow them to be integrated in a variety of situations. The<br />
Defense Advanced Research Projects Agency (DARPA) is sponsoring a two part competition to advance the state of<br />
research into the autonomous sensing, grasping and manipulation of objects. JPL/Caltech is one of six teams<br />
competing in this event. Prior work by the JPL/Caltech team has been done on using vision systems to perceive<br />
objects, estimate the position of each object, and plan paths to grasp the object, as well as controls for moving the<br />
arm along the planned path and manipulating the object. My current work focuses on the feasibility of identifying<br />
different sounds through the use of a Fast Fourier Transform, Principal Component Analysis, and Support Vector<br />
Machines. This approach should allow the robot to recognize the successful completion of a task. Additionally these<br />
techniques have been applied to tactile sensor data to determine the type of contact being made with an object as<br />
well as localize the object’s position within the hand.<br />
Gravitational Lensing of the Cosmic Microwave Background<br />
Ryan Keeley<br />
Mentors: Marc Kamionkowski and Fabian Schmidt<br />
Massive astronomical bodies deflect light causing a discrepancy between the actual position of the source and the<br />
observed position of the source. This gravitational lensing can be used to study the mass distribution of these<br />
astronomical bodies. The Cosmic Microwave Background (CMB) is useful as a source in such an analysis as its<br />
position (the last scattering surface) is known. This project will produce profiles of a lensed CMB distribution from a<br />
given unlensed distribution. This information will then be used to reconstruct the mass distribution of the lensing<br />
43
astronomical body. In the past, such analysis has been done with the assumption that the CMB can be modeled as<br />
a simple gradient on small scales. This project will do the full calculation for some general distribution in the<br />
unlensed CMB.<br />
Experimental Design for Two-Plasmon Quantum Interference Measurements<br />
Yousif A. Kelaita<br />
Mentors: Harry A. Atwater and Jim Fakonas<br />
In the last several decades, experiments in the field of quantum optics have opened new windows into the<br />
quantum nature of light. With a relatively simple setup, one can perform a suite of experiments that collectively<br />
demonstrate various quantum effects including quantum interference and entanglement. It is not yet known,<br />
however, whether surface plasmons—quanta of light that are sustained by the collective oscillation of a metal’s free<br />
electrons—exhibit all the same quantum effects as photons in free space do. This project proposes a set of<br />
experiments aimed at filling the gaps in our current understanding of plasmons at the quantum limit. To this end,<br />
the project requires work in precise laser alignment and calibration of collection optics, the design and<br />
implementation of a full fabrication process for surface plasmon waveguides on silicon nitride, and the design and<br />
construction of various custom parts. As a first goal, we aim to demonstrate two-plasmon quantum interference, an<br />
effect that has been predicted by analogy to two-photon interference but not yet observed. In the long term, we<br />
hope that this experiment and others will answer some of the many questions regarding the quantum nature of the<br />
plasmon.<br />
15<br />
N NMR Study of Intramolecular Hydrogen Bonding in Imidazole-4-Propanoic Acid<br />
Michael J. Kenney<br />
Mentors: John D. Roberts, William Carroll, and Mrinmoy Nag<br />
The presence of an intramolecular hydrogen bond in imidazole-4-propanoic acid in acetonitrile has been verified<br />
using 15 N NMR and 1 H NMR. Hydrogen bonding (deprotonated propanoic acid group) and non-hydrogen bonding<br />
(protonated propanoic acid group) cases were studied in acetonitrile-d3 using a 15 N-labeled 2-mercaptoimidazole-<br />
4-propanoic acid compound. Studies showed a downfield shift difference for N3 of 9 ppm, likely the result of an<br />
intramolecular hydrogen bond between H3 and the carboxylate anion. A CD3CN/H2O mixed solvent study, where<br />
the fraction gauche was measured as water was added to a solution of imidazole-4-propanoic acid in CD3CN in 5 μL<br />
increments, revealed the amount of water necessary to break the intramolecular hydrogen bond in imidazole-4propanoic<br />
acid. At 12 equivalents of water to one equivalent of imidazole-4-propanoic acid, the fraction of gauche<br />
conformers becomes statistical (0.66) indicating that the hydrogen bond is broken.<br />
Integration of Vertical-Cavity Surface-Emitting Lasers on CMOS Chips<br />
Aroutin Khachaturian<br />
Mentor: Axel Scherer<br />
Integration of vertical-cavity surface-emitting lasers (VCSEL) with the CMOS chip is an important step toward<br />
communicating with implantable biological sensors with high data transfer rates. VCSEL and CMOS chips have been<br />
previously bonded together through various fabrication techniques and an alternative approach is examined during<br />
this project. The initial goal of the project was to bond a 200µm thick VCSEL to a CMOS chip. This was achieved<br />
through a combination of photolithography and electroplating techniques so that the CMOS chip’s contact can be<br />
raised to the top contact of the VCSEL in order to make a connection between the two. Afterwards, they were<br />
tested in contact with skin, blood, and neural tissue using different types of data modulation techniques (e.g.<br />
amplitude and frequency modulation) and their efficiency is examined in terms of power and chip area consumed.<br />
Understanding the Role of Cilia in Hedgehog Signaling Pathway<br />
Sohini Khan<br />
Mentors: Rajat Rohatgi and David Tirrell<br />
The Hedgehog signaling pathway is known for its importance in embryonic development and cancer from<br />
medulloblastoma to basal cell carcinoma. Additionally, it requires primary cilia for full activation; however, in this<br />
pathway, it is still unknown which specific parts of the pathway require cilia. This project focuses on whether the<br />
interaction between Protein Kinase A (PKA) and Gli transcription factors depends on cilia. For this study, stable cell<br />
lines were generated containing a dominant negative form of Kif3A, a kinesin motor protein. These cells, in which<br />
primary cilia are absent, will then be used to explore the interaction between PKA and Gli proteins. Specifically,<br />
dominant negative PKA was tested for its ability to induce the Hedgehog pathway in the absence of cilia which<br />
proved to be ineffective. These cells were, however, able to activate the Hedgehog pathway after overexpression of<br />
Gli2. Also, PKA-insensitive mutants of Gli2 and Gli3 will be tested in the future for their ability to translocate into<br />
the nucleus and induce the expression of target genes.<br />
44
Design and Analysis of an in vitro Exact Adapter<br />
Ishan Khetarpal<br />
Mentors: Richard Murray and Jongmin Kim<br />
In the strict biological sense, by manipulating increasingly complex circuits, we gain much insight into how cells<br />
function. In terms of Computer Science, it has been shown that these biological processes are Turing complete,<br />
meaning that in theory, these systems can be used for computation. Several circuits have already been created,<br />
including oscillators, switches, and logical gates. The purpose of this research was to create and analyze an exact<br />
adapter.<br />
An exact adapter is a system whose outputs vary only transiently on its inputs. To implement the exact adapter in<br />
vitro, a gene was used that coded both for the creation of a strand fluorescent in the presence of malachite green<br />
(rMG), and its inhibitor (iMG). Effectively, the inhibitor degrades the fluorescent molecule after a transient period of<br />
visibility. One of the enzymes required for this process is ribonuclease R (RNase R). This enzyme is not fully<br />
understood, as it fails to operate on select fragments that are seemingly unrelated. Tests of degradation of rMG<br />
and iMG by RNase R were performed and analyzed via spectroflurometer and gel electrophoresis, to ensure the<br />
components would work correctly in the circuit. The circuit was then engineered with these components.<br />
Synthesizing a Pyrroloindoline Derivative Using a Formal (3+2) Cycloaddition Reaction<br />
Brian Kim<br />
Mentors: Sarah Reisman and Jane Ni<br />
A formal (3+2) cycloaddition reaction catalyzed by an (R)-BINOL · SnCl4 complex can be used to make different<br />
pyrroloindoline derivatives, the core structure of several biologically active products. 1,3-dimethyl-5-methoxyindole<br />
and methyl 2-trifluoroacetamidoacrylate were good substrates under the given reaction conditions, while<br />
1,3-dimethyl-5-bromoindole did not react due to the electron withdrawing effects of the bromine substituent.<br />
Thermodynamic Properties of Modified LiFePO4 for Li-ion Batteries<br />
Laura Kim<br />
Mentors: Brent Fultz, Hongjin Tan, and Hillary Smith<br />
Rechargeable lithium-ion batteries are one of the most popular batteries in consumer electronics owing to their<br />
high specific energy, long life cycles and reduced weight and volume. LiFePO4 has attracted significant interest<br />
because it is inexpensive, nontoxic, and made from naturally occurring minerals. Improvements to the intrinsic low<br />
electronic conductivity of LiFePO4 can be made by ball-milling to reduce the particle size and coating the material<br />
with carbon. We seek to understand the effect of these modifications on the thermodynamic properties of these<br />
materials. First, nano-sized LiFePO4 was prepared by ball milling. Then, the bulk and the nano-sized materials were<br />
characterized by x-ray diffraction. Next, coin cells were assembled in a glove box and their rechargeability and<br />
functionality were tested by a cell cycler. Finally, the coin cells were placed on an Electrochemical Thermodynamic<br />
Measurement System (ETMS) to compare the thermodynamics properties of the bulk, nano-sized, and carbon<br />
coated materials. Increased electrical conductivity in nano-sized particles and carbon-coated particles versus bulk<br />
and uncoated material has been observed, and ETMS measurements indicate that the configurational entropy of<br />
these materials has changed as a result of these modifications.<br />
Solid-State NMR Probe and Rotor Design at High Pressure and Corrosive Environment for Extended<br />
in situ Experiments<br />
Sangkyu Kim<br />
Mentors: Sonjong Hwang and S. I. Zones<br />
Zeolites make up an enormous class of three-dimensional aluminosilicate materials. The acid form of most zeolites<br />
is catalytic due to its ability to confine molecules in small spaces. The proposed mechanism under the condition of<br />
hydrofluoric acid is the following:<br />
SiO2 + SDAF + HF + H2O -> SDAF-MS + SDAF + HF + H2O<br />
where SDAF stands for the structure directing agent with fluorine, and MS for the molecular sieve.<br />
The in-situ method, synthesizing zeolite inside the NMR spectrometer, can offer important clues to understand<br />
zeolite frameworks and crystallization conditions. In this paper, the construction and performance (in zeolite<br />
syntheses) of high pressure magic angle spinning (MAS) inserts and rotor bodies made from the polymers PEEK<br />
and Teflon are described. The inserts are designed to fit inside standard commercial Bruker MAS rotors, and rotor<br />
bodies are designed as substitutes for the Bruker rotors.<br />
45
Progress Towards the Total Synthesis of Acutumine<br />
Seohyun (Chris) Kim<br />
Mentors: Sarah Reisman and Raul Navarro<br />
Acutumine is a propellane alkaloid natural product that possesses interesting biological properties, including<br />
selective T-cell cytotoxicity and antiamnesic activity. Structurally, acutumine comprises a highly fuctionalized azapropellane<br />
core, making it a challenging target for total synthesis. The goal of our studies is to achieve an efficient<br />
enantioselective synthesis of acutumine. Our lab recently reported a highly diastereoselective 1,2- addition of<br />
organometallic reagents to benzoquinone-derived sulfinyl imines, a strategy that allows for the preparation of<br />
enantioenriched dihydroindolones. These dihydroindolones serve as versatile intermediates in the preparation of<br />
propellane alkaloids; therefore, their gram-scale synthesis will facilitate synthetic studies en route to acutumine.<br />
We have achieved a multi-gram synthesis of the benzoquinone imine and grignard reagent required for our<br />
synthesis of acutumine. Current focus lies in the preparation of multi-gram quantities of sulfinamide addition<br />
product which can be readily converted to the dihydroindolone. By preparing larger amounts of addition product,<br />
we can thoroughly study the key steps in our proposed synthesis.<br />
Efficient Monte Carlo Methods for Radiative Transfer Modeling of Protoplanetary Disks<br />
Stacy Kim<br />
Mentor: Neal Turner<br />
When a protostar is born from a swirling, contracting cloud of dense gas, it leaves behind a rotating dusty and<br />
gaseous disk, produced by accretion of the surrounding interstellar material and the demands of angular<br />
momentum conservation. Translating observations of the disks into information about the conditions for planet<br />
formation and migration requires model temperature profiles and spectral energy distributions, which can be<br />
obtained relatively straightforwardly by applying the Monte Carlo method due to the stochastic nature of radiative<br />
transfer processes and the method’s adaptability to complex disk shapes. In practice, certain regions of the disk<br />
are under-sampled, such as the optically thick interior, or are of particular interest, such as the snow line (where<br />
water vapor condenses into ice) and the area surrounding a protoplanet. To improve the sampling, photon packets<br />
can be preferentially scattered and reemitted toward the preferred locations. This comes at the cost of deweighting<br />
packet energies by the same factor that the probability a packet travels in a particular direction was<br />
augmented. I present an analysis on the effectiveness of a variety of weighting schemes in terms of improved<br />
sampling, accuracy, and simulation speed.<br />
Electroporation of Engineered Bacterial Artificial Chromosomes (BACs) in Early Chick Embryos<br />
WoongBin Kim<br />
Mentors: Marianne Bronner and Marcos Simoes Costa<br />
Chicken embryos are important in embryology due to its similarity with human embryos. Making transient<br />
transgenic chicken embryos is the object of this project and it is an important tool for finding regulatory sequences<br />
of chicken embryos. Electroporation which increases the permeability and the electrical conductivity of a cell<br />
membrane by external electric field is the key technique to make transgenic chicken embryos. Electroporation of<br />
chicken embryos with different plasmid constructs has been established in high efficiency while electroporating<br />
BACs is much more difficult due to its big size. A [pCAGGS-eGFP (enhanced green fluorescent protein)-Kanamycin<br />
resistance gene] was made by ligation of C8 vector and PCR (polymerase chain reaction) product which were<br />
modified by XhoI and NotI restriction enzymes. The cassette was amplified with PCR. Purified PCR product of the<br />
cassette and BAC were electroporated into a strain of bacteria that expresses a recombinase. Bacterial cells which<br />
contain the BAC were selected with Kanamycin. Engineered BACs were extracted and purified with a BAC Maxiprep<br />
Kit. Electroporation of chicken embryos with BACs were processed under various conditions (different voltage and<br />
number of pulses). Survival of chicken embryos and expression of eGFP were observed after culturing chicken<br />
embryos.<br />
Limits on the Baryonic Contribution of Dark Matter via Astrometric Microlensing of Quasars<br />
Christina M. King<br />
Mentors: Charles J. Naudet and Christopher S. Jacobs<br />
Dark matter makes up 83% of the universe, though its composition remains a mystery. It is likely that dark matter<br />
partly consists of Massive Compact Halo Objects (MACHOs). The preferred methods for detecting MACHOs are<br />
photometric and astrometric microlensing. Existing MACHO surveys use nearby stars to detect photometric lensing,<br />
but it is more likely to observe an event when using extragalactic sources to search for astrometric lensing.<br />
Decades of Very Long Baseline Interferometry (VLBI) measurements have provided an extensive dataset of<br />
extremely fine measurements of positions of extragalactic radio sources. An upper limit for the precision required<br />
to detect microlensing events is established by comparing a trendline of a source’s motion with various Gaussian<br />
curves, which simulate expected microlensing signatures. Using VLBI data, is possible to set upper limits for<br />
detection of microlensing events on time scales of 1–30 years. The dominant systematic error is expected to arise<br />
46
from source structure, and can be isolated by rotating the coordinate system to analyze data parallel and<br />
perpendicular to the jet. Simulated astrometric events can distinguish smooth, predictable positional changes<br />
associated with gravitational lensing from erratic signatures produced by structural changes.<br />
Self Anchoring Drilling/Coring/Sampling System for in situ Planetary Applications<br />
Jonathan King<br />
Mentor: Aaron Parness<br />
The objective of future NASA exploration missions is in situ sampling and analysis of various astronomical bodies in<br />
the solar system (including Mars, Europa, Titan, comets and asteroids), and the return of samples to Earth for<br />
further study. Conventional drills require a relatively high axial preload or weight-on-bit, which limits their use in<br />
microgravity and on inverted surfaces. To resist these large reaction forces during drilling, an anchoring<br />
mechanism that effectively clings to rough and rocky surfaces has been integrated with a coring drill. The system<br />
consists of a compact rotary-percussive drill with a custom coring bit, a linear translation mechanism and spring<br />
system that will maintain preload and weight-on-bit, and a compliant anchoring mechanism that employs an omnidirectional<br />
array of micro-spines that cling to a surface by opportunistically catching on surface asperities. A<br />
prototype based on this design work is currently in fabrication, to be followed by testing and data collection.<br />
Additional efforts to develop a comprehensive model of the system are recommended so that further developments<br />
and predictions can be corroborated both theoretically and experimentally.<br />
The Phase and Proton Conduction Behavior of Solid Acids of the Form CsIx(H2PO4)1-x and<br />
Ba(3-x)MxHx(PO4)2 for M = Na, K, Rb, Cs<br />
Daniil Kitchaev<br />
Mentors: Sossina M. Haile and Ayako Ikeda<br />
The solid acid electrolyte that is currently believed to be optimal for fuel cell applications, CsH2PO4 (CDP), suffers<br />
from a low dehydration temperature, which limits its operating range, and water solubility, which causes problems<br />
during fuel cell startup and shutdown. We investigated the feasibility of two alternative materials: a doped variant<br />
of CDP that was expected to have a higher dehydration temperature, and water insoluble electrolytes based on<br />
Ba3(PO4)2. To investigate these samples, two atmosphere controlled furnaces were constructed. Then, electrolytes<br />
of various compositions were prepared by precipitation and high-temperature annealing, and examined via x-ray<br />
diffractometry, SEM/EDS, AC impedance spectroscopy and thermogravity. The solubility limits of CsIx(H2PO4)1-x,<br />
Ba(3-x)NaxHx(PO4)2, Ba(3-x)KxHx(PO4)2, Ba(3-x)RbxHx(PO4)2, and Ba(3-x)CsxHx(PO4)2, were measured and the structure of<br />
these compounds was confirmed. Furthermore, the preferred dopant site occupancy in the latter four compounds<br />
and its relation to the materials’ conductivity was determined. Finally, it was found that the effect of the type of<br />
dopant ion on the conductivity of the barium phosphate compounds was much less than could be expected.<br />
Characterizing and Investigating Spatio-Temporal Expression Patterns of Gene Expression in the Veg1<br />
and Veg2 Regions of E. tribuloides<br />
Rebekah Z. Kitto<br />
Mentors: Eric H Davidson and Eric Erkenbrack<br />
The goal of this project is to temporally and spatially profile early endomesodermal patterning events in the slate<br />
pencil urchin Eucidaris tribuloides. These data will be compared to those collected for other echinoderms in order to<br />
cast light on the conservation of early endomesodermal patterning events within the Echinodermata.<br />
The aim of this project is to produce qPCR primers and WMISH probes for otx, gataE, foxA, hox 11/13b, eve,<br />
blimp1, and wnt8, and then employ qPCR and whole-mount in situ hybridization (WMISH) to characterize the<br />
expression of these genes during E. tribuloides development. The data collected can be compared to S. purpuratus<br />
and other echinoderms to determine the similarities and differences in development between these two and other<br />
species despite their evolutionary divergence. Additionally, we aim to determine the role of Notch signaling in early<br />
endomesodermal patterning. To do this we will knockdown Notch signaling by treating developing embryos with the<br />
GSK-3 inhibitor DAPT. Ideally, the results obtained will provide information linking S. purpuratus and E. tribuloides<br />
that can shed light on the evolutionary divergence of these two species.<br />
Asbestos-Induced Alveolar Epithelial Cell Death: An Investigation Into the Role of the ASK1/JNK<br />
Pathway in Apoptosis<br />
Kathryn B. Knister<br />
Mentors: David Kamp, Paul Cheresh, and David Chan<br />
Asbestos fibers cause pulmonary fibrosis and cancer by mechanisms not fully established. Asbestos induces<br />
alveolar epithelial cell (AEC) death (apoptosis) via the p53- and mitochondria-regulated (intrinsic) death pathway.<br />
Mitochondrial and endoplasmic reticulum (ER) interactions regulate intrinsic apoptosis by activating ASK1/JNK<br />
signaling, but it is unknown whether asbestos induces this response. We examined whether a JNK inhibitor<br />
attenuates asbestos- and ER stress (using thapsigargin)-induced AEC DNA fragmentation (signifying apoptosis) and<br />
ASK1/JNK signaling (gauged by measuring protein levels with Western blotting). Our preliminary results were<br />
47
inconclusive because the JNK inhibitor alone was slightly toxic. This should be overcome by adjusting culture<br />
conditions and JNK inhibitor dose/diluent. Using transient transfection of an ASK1 dominant/negative plasmid (to<br />
inhibit the ASK1 gene), we will address similar endpoints. Future experiments will examine whether a p53 inhibitor,<br />
pifithrin, blocks asbestos-induced ASK1/JNK signaling and ER stress response. If the inhibitors block asbestos- and<br />
ER stress-induced AEC apoptosis and ASK1/JNK signaling, this will implicate a critical role for the ASK1/JNK<br />
pathway in mediating mitochondrial-ER-p53 crosstalk that can lead to intrinsic apoptosis. Our studies may also<br />
show that ER stress is an adaptive response to mitochondrial free radicals but independent of downstream<br />
ASK1/JNK and p53 activation.<br />
The Role of Elastic Interactions in Bacterial Chemoreceptor Organization and Function<br />
Peter D. Koch<br />
Mentors: Rob Phillips and Christoph A. Haselwandter<br />
We investigate the role of membrane-mediated elastic interactions in the spatial organization and function of<br />
bacterial chemoreceptors. It is believed that the smallest functional units of chemoreceptors are trimers of dimers<br />
which, upon binding to periplasmic ligands, initiate the chemotaxis signaling cascade. Many types of membrane<br />
proteins, including chemoreceptors, exhibit a hydrophobic mismatch with the surrounding lipid bilayer, thereby<br />
inducing thickness and curvature deformations of the lipid membrane. The energies of the membrane deformations<br />
induced by chemoreceptor trimers can be calculated using elasticity theory. We numerically compute these<br />
energies using a finite difference scheme. We first study the energetics of two interacting chemoreceptor trimers<br />
and find that our numerical results are in good agreement with previous analytical calculations, as well as previous<br />
numerical studies of related problems. Next, we study the energetics of chemoreceptors in lattice structures, as<br />
recent electron cryo-tomography studies have shown that chemoreceptor trimers form regular lattices which<br />
appear to be conserved across the bacterial kingdom. In particular, we examine how the total membrane<br />
deformation energy depends on the type of lattice symmetry and on the orientations of the chemoreceptor trimers<br />
within the lattice. Lastly, recent in vivo FRET studies have shown that bacterial chemoreceptors engage in<br />
cooperative signaling, and various theoretical models have suggested that this cooperative behavior is crucial for<br />
the high sensitivity of the chemotaxis signaling pathway. We study whether elastic interactions may promote such<br />
cooperative behavior among chemoreceptors, and whether lattice symmetry and trimer orientation may provide a<br />
mechanism for the observed size of cooperative signaling teams of chemoreceptors.<br />
Investigation of the Reported Separation of Nuclear Spin Isomers of Water Through Selective<br />
Adsorption<br />
Theodore Koenig<br />
Mentors: Geoff Blake and Daniel Holland<br />
Water exists as two nuclear spin isomers, ortho- and para-water, having different magnetic properties. At ambient<br />
temperatures, the isomers remain in thermodynamic equilibrium with a 3:1 ortho/para ratio. Various experiments<br />
claim to have shown significant enrichment of one isotope over the other through selective adsorption of water<br />
vapor over various sorbent materials. However, few if any of these experiments have been independently verified,<br />
and the existence of the claimed effect is still contested. The purpose of this project is to rigorously test the<br />
claimed effect, using various spectroscopic techniques upon which the effect was originally claimed to have been<br />
observed. To this end, a vacuum apparatus with sorbent and sounding cells was constructed, heeding claimed<br />
material incompatibilities with the claimed effect (i.e. no magnetic parts). A tunable diode laser absorption<br />
spectrometer was built. Further, an FT-IR was setup to complement the near-IR diode measurements and a<br />
procedure for data analysis of these results was chosen. Laser absorption specific to each of ortho- and para-water<br />
has been observed. The integrated areas of these absorption peaks will be used to measure the relative ratios of<br />
ortho- and para-water over the course of an adsorption experiment, as was performed in the original literature<br />
claims. The results of these experiments should inform the debate about these claims, with potential impact in the<br />
astronomical and earth sciences.<br />
Damage Mitigation and Performance Improvement of Vertical Axle Wind Turbines Using Compliant<br />
Wind Blades<br />
Chan-Hee Koh<br />
Mentors: Morteza Gharib and Julia Cossé<br />
We investigate the effect of using compliant blades utilizing unsteady aerodynamics on drag based vertical axle<br />
wind turbines. A model wind turbine was built using various compliant materials for the blades, potentially<br />
increasing the efficiency, performance, and durability beyond what conventional drag based wind turbines could<br />
achieve. The design is a simple vertical axle wind turbine with two flat plates as blades. The blades are designed so<br />
that their angle can be changed relative to the frame. The blade angles were changed to see the various effects the<br />
wind had on the drag. These angles would show how the flexible materials deform in the wind, resulting in different<br />
areas that face the wind. After measuring the angular velocity and the torque of the turbine in a wind tunnel, we<br />
were able to calculate the power output and the power coefficient, also known as the efficiency of the turbine, to<br />
48
get an idea of how our model compared with conventional turbines. We used two materials of different flexibility,<br />
aquaplast plastic and rubber, as well as a rigid control, aluminum sheet. We compared the obtained values with<br />
one another as well as the control case, which would not rotate on its own for any configuration.<br />
The turbine showed a periodic inconsistency in the angular velocity as the power output oscillated through each<br />
rotation due to flapping of the flexible blades. The unsteady effects caused by the flexible flapping significantly<br />
improved the performance of the wind turbine. This is evident in the fact that the flexible materials were able to<br />
rotate in configurations that the rigid material was not. However, the design had a number of problems that<br />
prevented the rigid material from turning. One aspect was that the design made the blades symmetric, which<br />
meant that there would be an equal amount of drag force on both sides of the blades in the opposing direction,<br />
resulting in stall. Conventional savonius blades would be curved to ensure that there is a difference in drag forces<br />
so that the turbine would rotate. Another problem was that the alignment of the axle was not optimum, and we<br />
undoubtedly lost efficiency to friction. Nevertheless, the results delineate the potential improvements in<br />
performance of current turbines.<br />
Neuroeconomic Foundations of Social Learning in Economic Choice<br />
Swadhruth Komanduri<br />
Mentors: Christof Koch and Michael Hill<br />
Modern neuroeconomic experiments primarily consist of studying reward-related systems by having subjects make<br />
decisions that will directly alter their future reward. These studies have shown that the orbitofrontal cortex, the<br />
amygdala, and the anterior cingulate have neurons which encode for relevant variables, such as the prediction<br />
error, the value of a choice, or the winnings of a certain trial. Such studies uncover underlying neural pathways for<br />
learning from one’s own mistakes and triumphs. These studies, however, do not take into account the impact of<br />
others’ choices and outcomes which help in the decision making process. To understand this impact, I have been<br />
working on an experiment to discover the neural basis of social learning. In this experiment, each subject plays a<br />
multiplayer game with two computer controlled characters. Each player must choose from one of two decks,<br />
knowing that one deck has a 70% chance of winning money and the other a 30% chance. The outcomes are<br />
revealed publicly, allowing the subject to gain information and learn while not actively participating. The goal of the<br />
project is to understand the neural network underlying this type of social learning and to compare this network<br />
with those pertaining to personal reward-related learning. Preliminary analysis shows that there are responsive<br />
neurons in the orbitofrontal cortex and amygdala, as expected. Moreover, there are also responsive neurons<br />
present in the entorhinal cortex, parahippocampal gyrus, and anterior hippocampus, areas potentially involved in a<br />
distinct neural pathway for social learning.<br />
Elucidating NF- κB Oscillation at a Single Cell Level<br />
Philip Kong<br />
Mentors: David Baltimore and Devdoot Majumdar<br />
Nuclear factor kappa-B (NF-κB) is a transcription factor that orchestrates the activation of several immune<br />
response genes in response to inflammation. It is found in virtually all animal cells and plays a key role in<br />
responding to infection. One hallmark of NF-κB is its oscillation from the cytoplasm to the nucleus on an hourly<br />
timescale. The dynamics of this periodic nuclear entry of NF-κB and its resulting gene expression pattern have<br />
been studied in populations of cells. However, there exists considerable asynchrony in the timing of NF-κB<br />
oscillations among members of a population of cells; therefore, the study of the NF-κB transcriptional response is<br />
complicated by cell-to-cell heterogeneity. Here, we ask: what is the NF-κB induced gene expression profile as a<br />
function of time at the single cell level? By using real-time quantitative PCR (RT-qPCR) and time-lapse microscopy,<br />
we connect NF-κB oscillatory movement and NF-κB mediated temporal gene expression profiles in response to<br />
inflammatory stimuli. Our results indicate that we are able to develop a qPCR system that can detect genes at a<br />
single cell level, which has only few picograms of RNA. With the methods developed herein, we explore the<br />
functional relevance of NF-κB’s oscillatory behavior.<br />
Search for Highly Dispersed Transient Radio Pulses<br />
Michael Kossin<br />
Mentor: Walid Majid<br />
This project is part of an effort to develop and test an accurate and efficient dispersed transient radio source<br />
detection system. Certain phenomena eject radiation that may only be observable for a fraction of a second. As<br />
this radiation travels through space, it may undergo dispersion before it arrives at our detectors. Dispersion is an<br />
effect experienced by waves as they travel through an ionized medium: their group velocities become altered by an<br />
amount related to their frequencies. If a burst of radiation across a range of frequencies undergoes dispersion, the<br />
higher frequency radiation will arrive at an observer sooner than lower frequency radiation does.<br />
49
To detect radio signals received after experiencing an unknown amount of dispersion, we use a program utilizing a<br />
trial and error process to search for dispersed radio signals in a data set by varying a guessed dispersion measure.<br />
The program translates signals in the dataset in time by a function of their frequencies and searches for instances<br />
of pulses existing at the same time across these adjusted frequency bands. We test the efficiency of this program’s<br />
ability to detect pulses with different dispersion and noise levels with artificial datasets.<br />
Sputter Deposition and Electrochemical Characterization of Model Thin-Film Metal Electrodes for Solid<br />
Acid Fuel Cells<br />
Jeffrey Kowalski<br />
Mentors: Sossina Haile and Aron Varga<br />
Solid Acid Fuel Cells (SAFCs) are promising alternative energy conversion devices since their high efficiency helps<br />
to address the dwindling supplies of fossil fuels and their effects on the environment. The specific problem<br />
examined in this project is the performance of several metal electrocatalysts (platinum, palladium and nickel) in<br />
SAFCs constructed with CsH2PO4 as the electrolyte. Differences in electrochemical activity under conditions of<br />
varying electrocatalyst film thickness were investigated. The thickness of the film was controlled by varying<br />
deposition times. The physical attributes of the films were then characterized using scanning electron microscopy<br />
and x-ray diffraction. A.C. impedance spectroscopy was used to elucidate the performance for hydrogen oxidation.<br />
Cyclic voltammetry was also used to determine the voltage required to overcome the formation of Pt oxide when<br />
oxygen is present in the atmosphere of measurement. A comparison of XRD patterns before and after keeping an<br />
intimate mixture of NiO and CsH2PO4 powder at fuel cell operating conditions for 18 hours suggests a reaction<br />
occurring between the components. These findings suggest that Ni cannot be used as a catalyst for SAFCs that use<br />
CsH2PO4 as the electrolyte.<br />
Characterization of High Speed Analog-to-Digital Converters for Deep Space Network Downlink<br />
Array IF Sampling<br />
Kara Kundert<br />
Mentors: Robert Navarro and Melissa Soriano<br />
The NASA Deep Space Network provides services for navigation and telemetry of spacecraft beyond Earth orbit.<br />
Currently, NASA is starting work on a new downlink array—DSN Downlink Array (DDA)—which will be focused on<br />
bringing more of the processing from the analog to the digital domain. This focus will take advantage of modern<br />
signal processing techniques and improve the accuracy of the DSN data. This project is primarily focused on the<br />
evaluation of several candidate analog-to-digital converters to determine which provides the best overall<br />
performance for the needs of the DDA. Of particular concern is the flatness of the gain and group delay of the<br />
converter over the IF bandwidth, as excessive variation will interfere with the beam forming that occurs when<br />
combining the signals from many antennas. In addition, converter nonlinearity and noise will be evaluated as these<br />
could limit the DDA’s ability to resolve weak signals from our satellites in deep space, particularly in the presence<br />
of large interferers.<br />
d-Wave Superfluidity in Two Dimensional Optical Lattices of Ultracold Polar Molecules<br />
Kevin Kuns<br />
Mentors: John Preskill and Alexey Gorshkov<br />
It has long been proposed that simulations of the t-J model by ultracold atoms in two dimensional optical lattices<br />
could be used to shed light on the physics of high-temperature superconductors and the underlying d-wave<br />
superfluidity. Recent work on ultracold polar molecules, governed by the t-J-V-W model, suggests that molecules<br />
may be better suited for studying d-wave superfluidity due to stronger interactions and larger tunability of the<br />
system. We compute the phase diagram for polar molecules in a two dimensional square lattice in a certain<br />
perturbative limit. In the simplest experimentally realizable case where there is only tunneling and an XX type<br />
spin-spin interaction, we identify the parameter regime where d-wave superfluidity occurs. We also study the<br />
effects of density-density, spin-density, and Ising type spin-spin interaction terms. Our findings can be used as a<br />
guideline for tuning the experimental system of polar molecules into the d-wave superfluid phase.<br />
Effect of Nanostructured Topologies on Air Flow Properties Over Vertical Axis Wind Turbine Blade<br />
Surface<br />
Adela Kuzmiakova<br />
Mentors: Beverley McKeon and Julia Greer<br />
During its lifetime, a vertical axis wind turbine (VAWT) encounters a variety of meteorological conditions – be they<br />
icing, thunderstorms, or heavy rains. Yet still, little remains known about the aerodynamic influence on VAWT’s<br />
performance due to heavy rain. One of the key mechanisms responsible for degradation of VAWT’s efficiency is the<br />
“effective” time-dependent roughness on the airfoil's surface due to uneven distribution of water film. Here, we<br />
propose nano-structuring of thin layers on the top of the turbine’s surface in order to improve its energy efficiency.<br />
The thin films are exposed to a simulated precipitation design and the change in their surface distribution is<br />
50
measured. Then, the layers are applied on the airfoil’s surface and tested under comparable conditions. The<br />
preliminary study of air flow properties over the nano-structured topologies is related to their wetting capabilities in<br />
order to gage the material selection and processing for optimal VAWT’s energy efficiency.<br />
Asthma and Thunderstorms: Observing Pollen Grain Rupture in Laboratory Simulated Conditions<br />
Sung Min Kwon<br />
Mentors: Richard Flagan and James House<br />
Thunderstorms have been known to be associated with widespread asthma attacks in recent years. Whole pollen<br />
grains are too large (20 to 100μm in diameter) to reach the lower airways, but at high relative humidity (~96%),<br />
they can rupture through osmotic shock and release cytoplasmic material small enough (0.2 to 3.5μm in diameter)<br />
to invade smaller bronchial tubes of the lung. With intermolecular van der Waals forces sufficiently lowered after<br />
evaporation, winds deposit these fragments in the human respiratory tract, likely causing asthma. Studies have<br />
found that a synergic combination of electricity and humidity leads to more pollen rupture than humidity alone.<br />
The purpose of this study is to determine the optimal conditions for pollen grains to rupture. An apparatus was<br />
designed from acrylic resin to observe the behavior of pollen grains subjected to humid air and varying electric<br />
fields under a light microscope. The humidity will be measured by a chip hygrometer/thermistor inserted into a<br />
controlled airflow. The electric field will be calculated from input voltage and distance between two copper wire<br />
leads. The pollen samples will be taken from the Chinese elm (Ulmus parvifolia), a major source of airborne<br />
allergens in Pasadena that blooms in August.<br />
Building Mechanical Actuators From C2C12 Skeletal Muscle Cells<br />
Thomas Kwong<br />
Mentors: Mory Gharib and Hesham Azizgolshani<br />
Muscle cells are linear contractile actuators that cause mechanical motion in biological systems. The directional<br />
action of muscles derives from coordination on all levels of muscle structure, from the sarcomeres in each cell to<br />
the myofibers in the tissue. That directionality is crucial for allowing muscles to efficiently deliver forces. However,<br />
C2C12 cells grown in culture lack the organization of sarcomeres. In order to design mechanical actuators powered<br />
by muscle cells, a method must be developed in order to coax cells grown in culture to adopt the directionality<br />
found in the muscle tissue of organisms. To this end microcontact printing (µCP) enables linear patterning of<br />
fibronectin onto a culture surface, thereby promoting selectively favorable cell attachment in specific linear<br />
patterns. Muscle cells cultured on fibronectin µCP surfaces display linear directionality when differentiated into<br />
myofibers. µCP is the primary method for patterning cell growth, but its limitations motivate study of other<br />
approaches to expand patterning capabilities. In specific, µCP is limited to planar surfaces. One promising approach<br />
utilizes the fact that cells placed in contact with a favorable surface can transfer to it from their culture surface.<br />
This method invites the possibility of transferring cell patterns achieved through µCP to untreated surfaces by<br />
contact. Further studies are being performed with the goal of patterning cells on cylindrical surfaces.<br />
Determining the Stellar Properties of Herbig-Haro Energy Sources<br />
Stephanie Laga<br />
Mentor: Lynne Hillenbrand<br />
Optical spectroscopy of T Tauri stars, or very young stars, continues to reveal evidence used in developing theories<br />
about the early stages of low mass star formation. In particular, this project focuses on those Class I objects that<br />
power Herbig-Haro flows, spatially resolved jets or jet shocked materials that signify strong outflow from the young<br />
star. In a study of a small set of such objects in the Taurus Molecular Cloud, White and Hillenbrand (2004) found<br />
that the stars that power these Herbig-Haro flows have stellar and circumstellar properties generally<br />
indistinguishable from similar mass stars that do not power these flows. For this project, the protostars of interest<br />
were observed from 6300 to 8800 Angstroms in wavelength using the Keck telescope. The emission and absorption<br />
lines in the spectra produced by these objects will reveal specific properties of the enveloped protostars. Stellar<br />
properties related to the central star like spectral type the rotational velocity and circumstellar properties like<br />
spectral veiling tell us about accretion rates and various emission line strengths, important to determining theories<br />
about star formation.<br />
Elastostatic Solutions for Realistic Slip and Stress Around Mode II and III Cracks<br />
Valère R. Lambert<br />
Mentors: Jean-Philippe Avouac and Sylvain Barbot<br />
A common practice for analyzing the displacement and stress caused by slip on buried faults is to discretize the<br />
fault area into finite size patches of assumed uniform slip and use the corresponding Green’s functions to predict<br />
strain at the surface or within the bulk of the medium. This formalism is commonly used to invert observation of<br />
co-seismic deformation for fault slip, in afterslip analyses, stress transfer studies and in numerical models of the<br />
seismic cycle, such as the Uniform California Earthquake Rupture Forecast (UCERF). The discretization of the fault<br />
into areas of uniform slip introduces stress singularities at the edges of each patch, which might make the results<br />
51
quite sensitive to the choice of the discretization scheme. Yet, little attention has been paid to this issue so far.<br />
Here, we derive analytic expressions of the Green’s functions, for which the Okada’s solution is a special case. We<br />
derive a full-space formulation for linearly interpolated fault slip distributions, which allows discretization of any<br />
fault slip distributions without introducing stress singularities. We investigate the case of a finite crack with a<br />
uniform distribution of stress on the fault. We show that the Okada solution provides an inadequate representation<br />
of the stress distribution for mode II and III cracks, even in the limit of infinitely small discretization. We introduce<br />
a critical size of fault discretization below which stress singularities dominate and bias the distribution. Our<br />
tapered-slip solution does not suffer from these shortcomings, suppresses all singularities and converges to a<br />
uniform stress with smaller discretization. Our results demonstrate the need to resolve the areas of stress<br />
concentration in our models of fault slip evolution.<br />
Quantitative Analysis of the Phototaxic Response of Artemia<br />
Megan C. Larisch<br />
Mentor: John O. Dabiri<br />
Biomixing- ocean mixing brought about by the movement of animals- has been the source of much debate over the<br />
past several years. Recent work has shown that biomixing via vertical migration may be a significant contributor to<br />
overall ocean mixing. The implications of such mixing are far-reaching. If biomixing were found to be significant, it<br />
would affect climate models and our understanding of global circulation. Additionally, vertical mixing by small<br />
crustaceans, such as krill, could potentially bring deeper, more nutrient-rich water to the surface, feeding the<br />
plankton the krill eventually eat.<br />
It has been shown that certain models of vertical migration result in significant ocean mixing. However, the work<br />
has been mainly theoretical, and has yet to be replicated in a quantitative study. As a result, the viability of<br />
biomixing as a significant contributor to overall ocean mixing is still undetermined. Since vertical migration holds<br />
the most promise for biomixing, we have set up an experiment to simulate the migration of bottom-dwelling<br />
animals to the surface. A 1-meter tall graduate cylinder will be filled with different salinities of salt water in a<br />
stratified manner, so that the densest, saltiest water makes up the bottom layer and the lightest, least salty water<br />
makes up the top layer. Artemia (brine shrimp) will be made to swim from the bottom to the top of the column by<br />
taking advantage of their phototaxic behavior.<br />
While some studies have addressed the phototaxic behavior of brine shrimp little has been done to quantitatively<br />
study this attraction. To best understand any mixing by the shrimp in the above experiment, we must first<br />
understand the phototaxic behavior of the shrimp.<br />
Patterned Quantum Dots for Nanotechnology<br />
Pawel Latawiec<br />
Mentors: Axel Scherer and Andrew Homyk<br />
Quantum dots in solid state systems have been an active area of research for years now, showing tremendous<br />
potential in applications to next-generation communication and computing systems. Traditional methods of<br />
fabrication of quantum dots on semiconductor substrates have largely been unable to place them selectively,<br />
resulting in poor device performance. This project explores novel methods for the design and construction of<br />
patterned quantum dot systems. Of particular interest are applications to transistors and photonic crystals. Novel<br />
vertical transistor devices were fabricated which operated via quantum mechanical resonant tunneling. This was<br />
achieved by a self-terminating oxidation of a scalloped silicon nanopillar followed by vertical gate construction.<br />
Measurements of preliminary structures indicate results consistent with theoretical predictions. Patterned quantum<br />
dots constructed for photonic crystals seek to maximize crystal-dot coupling and bring interaction into the cavity<br />
quantum electrodynamic regime. This is achieved by wet etching of quantum wells. Devices which take advantage<br />
of the positioning control are proposed.<br />
Investigating How the Brain’s Sensitivity to the Eyes in a Face Associates With Social Gaze During Real<br />
Conversations in Autism<br />
Valerie Latimore<br />
Mentors: Ralph Adolphs and Michael Spezio<br />
The ability to gather and process social information such as emotional expression is an important skill for all social<br />
animals to possess. Facial signaling is a particularly important medium of transmittance of social information for<br />
humans during conversation. However, people with autism exhibit deficits in many aspects of social behavior,<br />
including face gaze and the use of information in that face, such as from the eyes and mouth, in order to make<br />
judgments about a person’s facial expression. Because it is known that high functioning adults and children with<br />
autism show difficulties in social judgment and interaction, but retain normal perceptual and cognitive systems,<br />
research on autism has been focused on understanding how people with autism process salient social cues, notably<br />
from faces. The novelty of this project is that it seeks to establish a link between how the brains of people with<br />
52
autism process facial features, on the one hand, and the social gaze of those same people as they engage in real<br />
conversations, on the other. The project combines fMRI data with high-speed eyetracking in naturalistic contexts to<br />
yield new insight into the biology of autism.<br />
Determining Cloud Base and Thickness From Spaceborne Spectroscopic Imaging and Lidar Profiling<br />
Techniques<br />
Wingyee (Marie) Lau<br />
Mentors: Yuk L. Yung and Dong L. Wu<br />
Spaceborne stereoscopic imaging from the Multi-angle Imaging SpectroRadiometer enables global wide cloud base<br />
height and thickness determination, which cannot be achieved by in situ ground based Lidar height profiling<br />
instruments. This new technique enables more accurate determination of cloud base and thickness of boundary<br />
layer clouds, which in turn provides an important observational constraint to climate and weather models.<br />
To effectively compare height profiles determined from spaceborne images and Lidar profiling, we focus on an<br />
island far from the equator, i.e. Graciosa of the Azores, and we do not study clouds that are too thick for<br />
spaceborne cameras to see to the cloud base. We have found sufficient cases where height profiling from<br />
spaceborne imaging agrees with that from Lidar instruments.<br />
Analysis of Lunar Roughness and Thermal Emission<br />
Michael Lauria<br />
Mentors: Oded Aharonson and Antoine Lucas<br />
Lunar roughness, or variance in slope as a function of scale, reveals information concerning impacts and other<br />
processes which shaped the surface of the Moon. Roughness of a surface also affects temperature because it<br />
modifies the amount of incident solar illumination as well as reradiation from nearby surfaces. Lunar temperature is<br />
an important criterion for volatile ices stablity, and hence, relevant for future exploration. In this study, the<br />
relationship between lunar roughness on the one meter scale and the bolometric temperature of the lunar surface<br />
is evaluated through detailed elevation data in the form of Digital Elevation Models (DEMs). These models are<br />
constructed from high resolution stereo images acquired by Lunar Reconnaissance Orbiter Camera, and analyzed<br />
for their statistical roughness properties. Initial results indicate correlations between the roughness and<br />
temperature fields, dependent on the azimuth and elevation of the sun during the temperature measurement, as<br />
well as on large-scale features such as craters. This study provides a targeted analysis of specific areas of the lunar<br />
surface and establishes a method that can be applied more generally.<br />
Further Consequences of the Gross-Zagier Formulae<br />
Brian Lawrence<br />
Mentor: Dinakar Ramakrishnan<br />
A quadratic twist of the L-function associated with a modular form is known to satisfy a functional equation, which<br />
may be even or odd. A result due to Gross and Zagier explicitly computes the central value of the L-function or its<br />
derivative. For modular forms in prime level, in cases when the functional equation is even, Michel and<br />
Ramakrishnan have used an averaging method, averaging over new forms while keeping the quadratic character<br />
fixed, to prove several consequences of the Gross-Zagier formulae, including a non-vanishing result. The present<br />
research generalizes this result to forms of prime-squared level, for which the functional equation is always odd.<br />
Such an L-function necessarily has a central value of zero; the Gross-Zagier formulae compute the central value of<br />
its derivative. We compute the average value of these derivatives over different L-functions and investigate the<br />
consequences of this formula. In particular, we show that for sufficiently large level there exists an L-function with<br />
only a simple zero at the center of symmetry. Some of our results are conditional upon a conjectured nonnegativity<br />
for the central value of the derivative of a twisted L-function.<br />
Social Judgments at Varying Distances<br />
Dennis Lazar<br />
Mentors: Ralph Adolphs and Laura Loesch<br />
This experiment looks at how our social judgments about other people are influenced by our distance from them.<br />
People can make rapid judgments from faces, but it is still unclear what information in faces allows for these<br />
Spontaneous Trait Inferences. This experiment is performed by having viewers judge pictures of actors taken at<br />
different distances: far (20 meters), mid (10 meters), near (1 meter). For each stimulus, a viewer is asked to rate<br />
that stimulus in terms of trust, threat, and dominance. Additionally, a subset of subjects’ behavioral data is<br />
supplemented by eye tracking data. The data obtained shows good test-retest reliability within subjects, but<br />
further analysis still needs to be done. To conclude the project, I will analyze the effect of distance on the<br />
judgments made, as well as on where the subjects were looking while making their judgments.<br />
53
Change-Point Detection for the Community Seismic Network<br />
Tuan Anh Le<br />
Mentors: Andreas Krause and Matthew Faulkner<br />
Change-point detection is a technique used to finding anomalies in time-dependent data series, which can be<br />
applied to detecting natural phenomena. One of the examples in which change-point detection could be used in is<br />
earthquake data for the Community Seismic Network. Currently, the Community Seismic Network applies a simple<br />
binary picking algorithm for earthquake detection. Here, we consider the problem of applying multivariate changepoint<br />
detection on a dense, noisy sensor network for earthquake detection in the Community Seismic Network. We<br />
give an estimate of the performance of the method in general, and especially its feasibility for the Community<br />
Seismic Network. Performance is demonstrated on several sets, including synthetic and real world data series.<br />
Applied Statistics and Digit Analysis to Identify Fraud in Financial Data Sets<br />
Jetson Leder-Luis<br />
Mentor: Jean Ensminger<br />
In investigating fraudulent expenditure, it is useful to be able to determine the probability of the presence of fraud<br />
within a set of expenditure data simply by analyzing the numbers. Digit analysis is a tool in which the digits of a<br />
data set are compared to expected digit frequencies to indicate the probability that human tampering was present.<br />
Data produced by real-world expenditures follow different digit frequency patterns than those made up by humans;<br />
namely, human-produced data shows preference for certain numbers and fails to be significantly random or to<br />
follow Benford’s law, which gives an expectation for certain digits in certain circumstances. For this SURF project, I<br />
produced tools to analyze a financial data set of development expenditures for evidence of human tampering, then<br />
applied these tools to show the significance of the fraudulent patterns. The results indicate that the data set with<br />
which I have been working has a statistically significant probability of human tampering within certain categories,<br />
which signals fraud.<br />
Application of High-Throughput Screening to the Generation of a Thermodynamic Stability Database of<br />
Protein Mutant Libraries<br />
Ernest Lee<br />
Mentors: Stephen L. Mayo and Alex Nisthal<br />
The application of computational protein design to protein engineering is a burgeoning field. The potential of<br />
sophisticated protein-based drugs with specialized structures and unique chemical properties is at a premium.<br />
Therefore, it is essential to understand the basis of protein thermodynamics in the context of force-field<br />
development for computational models. In order to do this, the generation of a high quality, self-consistent protein<br />
stability database is required. Protein stability can be measured quantitatively by calculating the difference in free<br />
energy between a mutant and the wild type. Studying single-mutant proteins allows us to understand how amino<br />
acid mutations directly affect protein stability. Stability data for a variety of proteins was generated by utilizing a<br />
high-throughput screening method to assay the complete mutagenesis of a protein domain. From this data, the<br />
quality of this database was compared to those of ones derived from existing computationally generated models. In<br />
addition, a quantitative analysis of the physical effects of amino acid hydrophobicity, polarity, and volume change<br />
on protein stability was carried out. Ultimately, this analysis allows for a clearer understanding of the correlation<br />
between the structure and stability of proteins.<br />
Three Lead EKG Glove With Stethoscopes<br />
EunJee Lee<br />
Mentor: Mani Chandy<br />
According to India Today, in both urban and rural areas of India, heart diseases have emerged as the number one<br />
cause of death. For this reason, it is urgent to introduce heart monitoring device especially to the areas without<br />
strong health facility. EKG (electrocardiography) is useful source to identify abnormal heart function or damaged<br />
area of heart. EKG detects electrical activity due to the contraction and expansion of heart muscle.<br />
This project is for inexpensive yet complete device to operate as a medical device in the third world countries. To<br />
meet affordable price, the device is simply built with two quadruple operational amplifiers, resistors, diodes, and<br />
capacitors. To test and reduce noise, hardware band pass filter and software filter within the recording program are<br />
used. Also, the device is combined with stethoscopes for detecting heart sound simultaneously with the EKG signal.<br />
The device can be made with less than 10 dollars and can function as complete EKG and stethoscopes for heart<br />
sound.<br />
54
Increasing Epithelial Permeability to Enhance Topical Drug Delivery to the Cornea<br />
Sinwook Lee<br />
Mentors: Julie Kornfield and Dennis Ko<br />
The drug administration to the cornea is required for the treatment of certain eye diseases such as keratoconus<br />
and post-LASIK ectasia. Keratoconus is the corneal thinning disorder that affects approximately 1 in 2000. But the<br />
corneal epithelium blocks the drug penetration because of the tight junctions in the epithelial layer. The most<br />
common method to improve it is to remove the epithelium, which has side effects such as tissue scarring, infection,<br />
and blurred vision. We believe some chemicals can make the tight junctions open wide enough temporarily for the<br />
drugs to penetrate the epithelium. And in order to determine how much the epithelium is permeabilized by those<br />
chemicals, we used eosin Y as a model drug, which is used for the treatment of keratoconus. If our project is<br />
successful, we will be able to administer drugs using eye drops instead of removing the epithelium.<br />
Demonstration of a DNA-Based Random Walking Molecular Robot That Reorganizes<br />
DNA-Tagged Objects<br />
Yae Lim Lee<br />
Mentors: Erik Winfree, Lulu Qian, Niranjan Srinivas, and Damien Wood<br />
Biomolecular robotics is a relatively recent research field with visible and intuitive mechanisms using DNA<br />
molecules. Many walkers have been demonstrated to walk on 1-dimensional track, but just a few walk on<br />
2-dimensional track and even fewer perform specific functions such as transferring gold nanoparticle species as<br />
cargos while traversing the pathway. This SURF project aims to incorporate both 2-dimensional walking and a<br />
specialized function into a DNA-based robot. More specifically, a molecular-scale DNA-based robot will reorganize<br />
cargos on 2dimensional fields. The scheme will include DNA- tagged cargos scattered on 2-dimensional origami,<br />
and a DNA-based robot randomly walking to sort them into different piles. The 2-dimensional track, cargo, and<br />
cargo goals will be planted on rectangular DNA origami [P. W. K. Rothemnd, Nature, 2006]. The motion of the<br />
walker and cargos will be examined by atomic force microscopy imaging. Bulk behavior of the system, kinetics of<br />
walking, and mechanisms of cargo picking up, and cargo dropping off will be analyzed by SPEX experiment.<br />
Quantifying Uncertainty in a Spacecraft Deployable Structure Latch<br />
James Leet<br />
Mentor: Lee Peterson<br />
A spacecraft must often deploy an additional structure, such as a solar array or a scientific instrument, in order to<br />
fulfill its mission objectives. While such tasks are mechanically simple, one must ensure that the devices<br />
responsible for deploying such structures can operate within mission parameters while subjected to the harsh<br />
demands of space. In this case, we employed computational modeling techniques to evaluate a latch for the<br />
planned SWOT (Surface Water Ocean Topography) mission. Tasked with deploying an array to focus returning<br />
microwave signals that have been reflected off of surface water on Earth, this latch has the potential to undermine<br />
the accuracy of measurements obtained if it is not deployed within certain boundary conditions. Utilizing a highfidelity,<br />
finite element model of the latch prepared with the Sierra software suite, we quantified the uncertainty<br />
present in the design using a DAKOTA (Design Analysis Kit for Optimization and Terascale Applications) interface.<br />
By quantifying the uncertainty present in the latch, mission planners can validate the current latch as a viable<br />
design or attempt to redesign it in order to provide less uncertainty.<br />
Improve Syntheses for Tracking the Decomposition of Silica-Supported Olefin Metathesis Catalysts<br />
Jomya Lei<br />
Mentors: Robert H. Grubbs and Matthew Van Wingerden<br />
Synthesizing the second generation Grubbs catalyst with a triethoxysilyl-functionalized NHC ligand grafted onto<br />
silica gel creates an efficient heterogeneous solid catalyst that can be recycled many times and does not leach<br />
ruthenium during olefin metathesis. Unfortunately, this catalyst can only be studied by ICP-MS which destroys the<br />
catalyst. A photolabile linker was synthesized to attach the catalyst to the silica gel which is easily cleaved off with<br />
55
light – taking the silica gel with it. Thus, the decomposition of the catalyst could now be studied with<br />
nondestructive techniques like NMR. New techniques were also implemented to synthesize the catalyst with higher<br />
purity and yield.<br />
Cellular Delivery of an Inhibitory Branched Capture Agent Against Akt1 via Cell-Penetrating Peptides<br />
Erica Leung<br />
Mentors: James Heath and Ryan Henning<br />
The non-specific serine/threonine-protein kinase Akt1 inhibits cell apoptosis; therefore, it is an important factor in<br />
the study of cancerous cells’ survival abilities. We further developed a branched capture agent against Akt1 for<br />
delivery within a cell. To study its effect within a living cell, known cell-penetrating peptide sequences were added<br />
to the capture agent. Two sequences were used with spacers of GSG: polyarginine consisting of nine arginines and<br />
the HIV-1-TAT sequence consisting of RKKRRQRRRPPQQ. The cell-penetrating peptides were synthesized on the<br />
carboxy-terminus of the anchor ligand using solid phase peptide synthesis, and the branched secondary and<br />
tertiary ligands were appended using the copper catalyzed Huisgen 1,3-dipolar cycloaddition reaction. The cell<br />
permeable capture agent will be delivered into cells to assess AKT inhibition and cellular toxicity.<br />
Mechanical Properties of Nanopillars Composed of Two Metals: A Simple Model for Metal Matrix<br />
Composites<br />
Jarvis Li<br />
Mentors: Julia Greer and Qiang Guo<br />
With growing interest in the properties of nano-scaled structures, we attempt to formulate a more detailed model<br />
of metal matrix composites to better understand the properties that govern the microscopic interactions among<br />
different materials. In this project, using electrodeposition on templates made from electron beam lithography- .<br />
we fabricated bi-material Cu-Fe and Cu-Ni nano-pillars containing a single heterogeneous interface, with diameters<br />
ranging from 50nm to 500nm. We then performed both compression and tension tests on the pillars to determine<br />
their mechanical properties, such as yield strengths, flow stresses, failure strain, work-hardening behaviors and<br />
etc. Also, we characterized the pillars using transmission electron microscope (TEM) to study the their<br />
microstructures, and scanning electron microscopy (SEM) to study the morphology before and after deformation.<br />
These bi-material pillars represent a prototype metal-matrix composite with highly controllable microstructure, high<br />
strengths, as well as strong-metal-metal interfaces. The isolation and subsequent engineering of a single interface<br />
in a pillar allows for fundamental insights gained into their deformation mechanisms, leading to more intelligent<br />
design of interfaces.<br />
Analyzing the Effects of Ice on Stratospheric Water<br />
Jingyuan Li<br />
Mentors: Geoffrey A. Blake and Ke Zhang<br />
How water enters the stratosphere is a process that is largely unknown. Despite the low temperatures at the<br />
tropopause which lead to small mixing ratios, stratospheric water levels show no correlation with temperature.<br />
Known methods of transport across the tropopause point to ice as an important factor. This project concentrates on<br />
any possible correlation between the amount of ice and the water vapor by using isotopic ratios. Due to the lower<br />
vapor pressure of HDO, the H/D ratio decreases as the water vapor condenses. The data is taken from the Keck<br />
telescopes, which can be used to extract both emission spectra and atmospheric transmission. Data reduction was<br />
done using a pipeline and then fitted to a lineshape model. Different line profiles (namely, the Voigt and the<br />
Galatry) were compared using CO fits to see if a simpler model is enough for the HDO/H2O fits. An atmospheric<br />
model gives us the values of HDO and H2O with altitude, and the results are tested with water vapor levels above<br />
Mauna Kea.<br />
Back-Action Evasion Measurement and Squeezed State of a Mechanical Oscillator<br />
John Li<br />
Mentors: Keith Schwab and Junho Suh<br />
Back-action evasion measurement avoids introducing noise when measuring a system. Kalman filter is a statistical<br />
algorithm that produces a better estimation from the measurement. I investigated Kalman filter in order to perform<br />
the optimal control of cantilever in back-action evasion measurement. I derived and verified the technique to<br />
implement Kalman filter for this specific experiment into discrete algorithm that DSP or FPGA can run, as well as<br />
the possible hardware means, including both Digital signal processor and Field programmable gate array, to<br />
perform digital signal processing.<br />
56
Improving the Toughness of Responsive Double Physical Network Injectable Hydrogels<br />
Shuaili Li<br />
Mentors: Bradley D. Olsen, Matthew Glassman, and David A. Tirrell<br />
Hydrogels show significant promise in minimally invasive tissue engineering and drug delivery because of their high<br />
water content and tunable mechanical properties. Triblock copolymers with a gel-forming protein midblock and a<br />
thermoresponsive poly(N-isopropylacrylamide) (PNIPAM) endblock form a double physical network hydrogel that is<br />
both injectable and responsively stiffens post-injection, but control of endblock relaxation for enhanced toughness<br />
remains a challenge.To engineer the thermodynamic and dynamic behavior of the endblocks, variants of PNIPAM<br />
polymers have been synthesized via Reversible Addition Fragmentation Chain Transfer (RAFT) with variable alkyl<br />
chain ends and with hydrophobic comonomers. Copolymers with N-tert-butyl acrylamide (NtBAM) are expected to<br />
have particularly promising characteristics due to depressed lower critical solution temperature (LCST) and higher<br />
glass transition temperature than the homopolymer. PNIPAM copolymerized with photocrosslinkable benzophenone<br />
monomer has also been explored for biomedical testing purposes. Fundamental investigations have been<br />
performed in purely synthetic triblock copolymer models as well as in bioconjugates with gel-forming protein.<br />
Linear oscillatory shear rheology was used to probe the relaxation behavior of the double-network gels via<br />
frequency sweeps as a function of temperature. Nonlinear rheology was used to look at the yielding behavior.<br />
Progress towards demonstration of the responsively tough, injectable biomaterial will involve compression testing<br />
for fracture toughness, assessments of injectability, and cell viability and adhesion studies.<br />
Identification of Piwi-Binding Loci in the Drosophila Germline Genome<br />
Susan E. Liao<br />
Mentor: Katalin Fejes Toth<br />
Suppressing mobile DNA elements in the germline is important for protecting genome integrity. One protein, Piwi,<br />
has been shown to partner with small RNA segments to silence effects of these mobile DNA elements in the<br />
cytoplasm. Since Piwi is localized to the nucleus in all metazoans, we suspect that it may also regulate expression<br />
of mobile DNA elements in the nucleus. Using DNA adenine methyltransferase identification (DamID), we<br />
investigate whether Piwi localizes to to specific chromatin loci. Then, through reverse transcription polymerase<br />
chain reaction (RT-PCR) and chromatin immunoprecipitation (ChIP), we test Piwi's effects on gene expression<br />
levels and epigenetic histone marks.<br />
Directed, in situ Growth of Conductive Nanowires Between the Photoelectrodes in a Tandem Microwire<br />
Array Fuel-Forming Device<br />
Ben Lieber<br />
Mentors: Nathan S. Lewis and Shane Ardo<br />
The primary focus of my research was to fabricate a highly conductive electrical conduit between two silicon<br />
microwire arrays through electric-field directed growth of 7,7,8,8-tetracyanoquinodimethane (TCNQ) anion chargetransfer<br />
crystals. This is important, because tandem microwire array solar photoelectrosynthetic cells require highly<br />
conductive interconnects to maximize efficiency. [Cu(TCNQ)] nanowires were grown in situ via physical vapor<br />
deposition from solid TCNQ precursors. Successful nanowire formation was demonstrated on photolithographically<br />
patterned 3 µm diameter copper islands, and on the copper vapor-liquid-solid (VLS) growth catalyst found at the<br />
top of silicon microwires. Scanning electron microscopy (SEM) and powder X-ray diffraction (XRD) confirmed the<br />
structure and conductive phase of the [Cu(TCNQ)] nanowires. Sandwich-cell devices, employing two planar<br />
patterned wafers, were constructed to model the back-to-back arrangement of polymer-embedded silicon<br />
microwire arrays envisioned for the ultimate device; [Cu(TCNQ)] nanowires were grown to bridge the halves of<br />
these sandwich-cell devices. Ongoing research is directed toward integrating these results to a tandem silicon<br />
microwire-array device and directing the [Cu(TCNQ)] growth with an applied electric field.<br />
Faster Compression Leads to Higher Recovery and Better Energy Absorption in Carbon Nanotube<br />
Micro-Pillars: An in situ Study<br />
Ee Jane Lim<br />
Mentors: Julia R. Greer and Siddhartha Pathak<br />
While individual carbon nanotubes (CNTs) have been announced as the strongest material known, layers of<br />
vertically aligned CNT (VACNT) bundles are more likely to find use in applications requiring large compliance and<br />
deformability, such as in microelectromechanical systems (MEMS) and impact mitigation. In this study we report on<br />
the mechanical behavior of such a VACNT bundle, grown in the form of 30/30 (height/diameter) micron-sized<br />
pillars using the chemical vapor deposition method at 750 mbar pressure and ~750 o C. Compression tests on these<br />
CNT pillars revealed 3 distinct regimes in their mechanical deformation—a short initial elastic section followed by a<br />
sloped plateau regime corresponding to buckle formation in the pillars, and a final densification regime. Each<br />
regime is characterized by its own distinct mechanical properties, for example a 6 fold increase in modulus and<br />
drastic decrease in recoverability were observed in the densification stage as compared to the elastic regime. Loadunload<br />
cycles at different strain rates reveal the gradual accumulation of damage in the CNT pillars at lower rates,<br />
57
which leads to lower recovery (
molds and extracted as free-standing particles in solution. Ultimately, clover-leaf microgels will be fabricated using<br />
artificial extracellular matrix proteins and assembled into porous constructs to determine the impact of porosity on<br />
corneal epithelial cell migration.<br />
Self-Assembling Artificial Extracellular Matrix Proteins for Cell Encapsulation<br />
Erik Liu<br />
Mentors: David A. Tirrell and Maren E. Buck<br />
The development of new biomaterials is important for tissue engineering and cell-based therapies. In this project,<br />
we use recombinant DNA technology to generate artificial extracellular matrix (aECMs) proteins. We have targeted<br />
two multidomain protein polymers composed of domains derived from elastin and fibronectin sequences that<br />
alternate with heterodimeric leucine zipper domains. Mixing of multiblock protein polymers containing<br />
complementary leucine zippers should permit the spontaneous formation of hydrogels under mild conditions. The<br />
DNA sequences encoding these proteins have been cloned and sequence verified. Initial expressions of the proteins<br />
resulted in low yields. Currently, we are pursuing a dual plasmid approach for co-expressing multidomain proteins<br />
with the complementary leucine zipper as a potential route to improving protein yields.<br />
Determination of the Dihedral Angles in Predominantly Trans-Ethanes Using NMR Spectrometry<br />
Tianxiang Albert Liu<br />
Mentor: John D. Roberts<br />
Two predominantly trans-ethane systems (2-(N,N-dimethylamino)-N-1-ethylammonium chloride and 3,3dimethylbutyl-N,N,N-trimethylammonium<br />
iodide) were synthesized in relatively high yield. Variable temperature<br />
and solvent studies suggested their strong trans conformational preference. Two ordered media (aqueous lyotropic<br />
and 7% PMA gel) were used to obtain 1 H, 13 C and HSQC spectra of both trans-ethane systems, which were then<br />
used to extract scalar, dipolar and residual dipolar coupling constants. This shed light on trans dihedral angles of<br />
both systems, which were previously undetermined.<br />
Cell ECM Network Study for Application in Tissue Engineering<br />
Ashley Lo<br />
Mentors: Chin-Lin Guo and Yuhwa Lo<br />
Cells have an extracellular matrix (ECM) on their exterior surface, which regulates processes such as growth,<br />
reparation, intercellular communication, support, and anchorage. When the ECM’s of each cell “detects” the<br />
presence of other cells within their vicinity (600 μm), these ECM’s will extensively branch out and connect to those<br />
of neighboring cells.<br />
MDCK and MCF-10A cells were grown in a cell culture, and injected into microfluidic channels and polygonal wells<br />
of different shapes, containing a medium solution and dissolved collagen, in attempt to stimulate the in vitro<br />
formation of ECM and epithelial (cellular) structures.<br />
In living organisms, the ECM plays a critical role in the ability of cells to self-organize and form functional<br />
structures. This process is evidenced in the formation of connective tissues, such as organs, joints, ligaments,<br />
fibrils, and tendons. Because cellular organization is such a fundamental aspect of structure and survival, research<br />
on the factors that influence and enable it would provide valuable insight into medical field. Ideally, the ability to<br />
induce cellular self-organization and differentiation in vitro would provide potential solutions to the current<br />
shortages in regenerative medicines and organ transplants.<br />
Atmospheric Features at the Jupiter North Pole From Cassini Images<br />
Daniel Y. Lo<br />
Mentors: Andrew P. Ingersoll and Shawn P. Ewald<br />
Newly reprojected mosaics of the Jupiter North Pole from Cassini images taken with the CB2 (756 nm),<br />
MT2 (727 nm) and MT3 (889 nm) filters has allowed the first detailed observations of the dynamical features and<br />
processes that occur at the high latitudes (>60 deg N). This study found that spots and filamentary regions (FRs)<br />
also occur at the high latitudes as in the lower latitudes. However, interactions between spots leading to merger<br />
take place in a very different manner, and this, together with the splitting behavior of spots, suggests that<br />
dynamics at the high latitudes are very different. In addition, this study also discovered that FRs originate from<br />
cyclonic spots that migrate into shear zones, and found evidence of the asymmetry in behavior between cyclonic<br />
and anticylonic features in these shear zones. Furthermore, this study also verified that the spot budget at the high<br />
latitudes was balanced over the observational period, and characterized the size of the spots at various latitudes.<br />
These findings will pave the way and serve to inform the first direct observations of the Jupiter North Pole by Juno<br />
when it enters Jovian orbit in 2016.<br />
59
N-Terminal Functionalization of Engineered Calmodulin Constructs for Surface Immobilization<br />
Megan Lo<br />
Mentors: David A. Tirrell and Chethana Kulkarni<br />
Calmodulin is a regulatory protein that plays an important role in learning and memory. The goal of this work is to<br />
immobilize engineered calmodulin constructs onto surfaces to create microarrays for high-throughput studies of<br />
calmodulin and the proteins which it regulates. We are working towards this goal by taking advantage of<br />
myristoylation, a well-studied, natural labeling process in eukaryotes. Myristoylation is the transfer of myristic acid<br />
onto the N-terminus of protein substrates via the eukaryotic enzyme N-myristoyl transferase (NMT). This process<br />
can be utilized as part of a versatile site-specific labeling technique due to NMT’s tolerance towards reactive<br />
myristic acid analogs. By engineering calmodulin to display a short sequence that NMT recognizes, we can achieve<br />
N-terminal functionalization of the protein which enables it to be conjugated to other materials such as beads and<br />
glass slides; specifically, calmodulin can be coupled to surfaces to create microarrays. After the construction of<br />
these microarrays, high-throughput studies can be performed to study the interaction of calmodulin with its binding<br />
partners to allow for a deeper understanding of the protein.<br />
An Analysis of Properties of Galaxies With Close Pairs<br />
Cassandra Lochhaas<br />
Mentor: Nick Scoville<br />
Merging galaxy systems are very important in the creation of new galaxies. Eventually, a system of merging<br />
galaxies can become a single, new galaxy. During the early stages of merging, galaxies can be at a variety of<br />
separation distances, and these distances give clues to how the merging process evolves. Due to the difficulties of<br />
identifying merging galaxies, this study considers close pairs of galaxies (instead of attempting to classify actual<br />
merging systems). The data was taken from the COSMOS study, which contains hundreds of thousands of galaxies<br />
in a small portion of the sky out to distances that correspond to half the age of the universe. The number of<br />
galaxies with close pairs as a function of separation distance and cosmic age of the universe (equivalently redshift)<br />
was also studied, with implications to how galaxies merged together in the far past to form galaxies in the more<br />
recent past. Differences in properties such as star formation rate, mass, and absolute magnitude of galaxies were<br />
analyzed. A. Petric also helped in analyzing spectra of a sample of nearby luminous infrared galaxies to better<br />
understand the effects of merging.<br />
Core Formation in SuperEarths<br />
Simon J. Lock<br />
Mentor: David J. Stevenson<br />
It is usually assumed in the study of SuperEarths (rocky planets with a mass of a few Earth masses) that they are<br />
essentially Earth-like, i.e. they have a differentiated (mostly liquid) core with a solid mantle and crust. However<br />
SuperEarths are significantly more massive than the Earth and so this may not always be the case. When<br />
SuperEarths are first formed it is believed that they are hot enough that their constituents (mainly silicates and<br />
iron) are fully mixed, down to the atomic level. As the planet cools the rock and iron demix, with the iron rich<br />
components sinking to the centre of the planet. This releases a lot of gravitational potential energy, providing a<br />
significant buffer against cooling. This process is being studied by constructing a model of the energy balance for a<br />
SuperEarth as it cools, in order to determine the effect on planet formation. A basic model indicates that significant<br />
core formation does not occur for planets of a few Earth masses. A more advanced model is being developed to<br />
further explore the problem. This work has significant implications for modelling the internal structures of<br />
SuperEarths and its results will need to be considered in future attempts.<br />
Developing an Algorithm to Pinpoint Deep Sea Corals<br />
Paige Logan<br />
Mentor: Jess Adkins<br />
Studies of deep-ocean sediments have revealed a clear connection between deep ocean circulation patterns and<br />
global climate change. A replacement for these sediments has been found in the deep-sea coral species<br />
Desmophyllum dianthus. These corals can provide more accurate oceanic circulation rate data for past climates<br />
using radiocarbon and U-series dating. This data may help us further understand current global climate change.<br />
Since these corals grow thousands of meters underwater and are only a few centimeters tall, they are difficult to<br />
find. My goal is to find an easy way to locate the corals. Using Amazon’s Mechanical Turk service, we quantified<br />
thousands of seafloor pictures to resolve the presence or absence of corals at various depths. By combining this<br />
information with bathymetry data from the autonomous rover that took the pictures, we can study the relationship<br />
between coral presence and the depth, slope, and curvature of the seafloor. This allows us to create an algorithm<br />
to estimate the likelihood of finding corals in a given area based on the shape of the seafloor there.<br />
60
The Role of Pavlovian Processes and Payment Methods on Decision Making and Reinforcement Learning<br />
Tong Lu<br />
Mentors: Antonio Rangel and Vanessa Janowski<br />
In contrast to the economic assumption that consumers are rational in valuation of goods, consumer choice is<br />
affected by product presentation, payment method, and other situational variables, as explored in two related<br />
experiments. The first simulated a seller choosing prices for various items, with the hypothesis being that the<br />
physical, proximal presentation of money in cash form significantly increases its value, as opposed to usage of nonmonetary<br />
equivalents or the presence of a physical barrier, resulting primarily from Pavlovian consummatory<br />
mechanisms. The second involved subjects given either cash or tokens to bet on the win or lose outcome of a<br />
series of colored cards in a gambling game. Their objective was to maximize their winnings by tracking the<br />
stochastically varying probabilities associated with the different card colors, generated by a Gaussian walk<br />
simulation. Preliminary results are consistent with the hypothesis that subjects are more willing to take risks with<br />
tokens versus cash. Identification of the variables that affect consumer decision making have immediate<br />
applications to marketing strategy.<br />
Design, Fabrication, and Characterization of Low Threshold Photonic Quasicrystal Lasers<br />
Tsung-Ju J. Lu<br />
Mentors: Axel Scherer and Se-Heon Kim<br />
Photonic crystal lasers have a spectral gap in the electromagnetic wave spectrum in which electromagnetic wave<br />
propagation is forbidden in all directions, which makes them attractive optical materials for controlling and<br />
manipulating the flow of light. However, there are some fabrication challenges in creating two-dimension periodic<br />
photonic crystals, specifically two-dimensional slab structures. In the present time, only thin slab periodic photonic<br />
crystals are useful since they can support a high Q factor. However, thin slabs have many disadvantages when<br />
used for real-life applications. Thus, the goal of this project is to fabricate a thick slab Quasicrystal cavity laser that<br />
has a quasi-periodic arrangement of air cylinders, which would potentially have a higher Q factor compared to<br />
photonic lasers with periodic arrangements of air cylinders. The design and simulation of the Quasicrystal structure<br />
were done by using Finite Difference Time Domain method and a program called “harminv”. The fabrication of the<br />
photonic Quasicrystal lasers was done by using a scanning electron microscope and etching machine. The lasers<br />
were then tested to make sure that they do indeed have spectral band-gaps and operate with a low threshold.<br />
Conformational Analysis of Succinate Dianion in Liquid Crystal Media<br />
Yichen Lu<br />
Mentors: John D. Roberts, Mrinmoy Nag, and William Carroll<br />
Anisotropic phases of liquid crystals restrict the freedom of motion of molecules in normal isotropic phase of<br />
solution. Nuclear Magnetic Resonance Spectroscopy (NMR) can be used to obtain geometrical information and other<br />
solution order parameters. Most importantly, dipolar couplings that contain information about the relative angles of<br />
the nuclei in a molecule can be obtained by fitting the oriented NMR spectrum. The average dihedral angle rotating<br />
around the central C-C bond of a specific molecule can then be calculated using dipolar coupling constants. In this<br />
project, both succinic acid and succinate dianion have been studied in an environment of oriented aprotic media<br />
such as liquid crystals. Both H-NMR and 13 C-NMR spectrum have been taken at nematic phase temperatures to find<br />
dipolar coupling constants. Mixtures of gauche and trans conformers have been considered. A MATLAB program has<br />
been constructed to perform the calculation.<br />
The cpSRP54-cpFtsY Interaction in the Chloroplast SRP Pathway<br />
Zeyu (Mike) Lu<br />
Mentors: Shu-ou Shan and Sowmya Chandrasekar<br />
The universally conserved signal recognition particle (SRP) and SRP receptor (SR) mediate the cotranslational<br />
targeting of proteins to cellular membranes. In contrast, a recently discovered unique chloroplast SRP in green<br />
plants is primarily dedicated to the post-translational targeting of light harvesting chlorophyll a/b binding (LHC)<br />
proteins to the thylakoid membranes. The pathway involves cpSRP54 and its receptor cpFtsY, two GTPases that are<br />
closely homologous to the cytosolic SRP54 and SR GTPases. However, intriguingly enough, the otherwise<br />
universally conserved SRP RNA which accelerates complex formation between the two GTPases has not been found<br />
in the chloroplast pathway. Instead, a new 43-kDa protein, cpSRP43, forms the chloroplast SRP together with<br />
cpSRP54. Despite the early hypothesis that cpSRP43 might serve the same purpose as the SRP RNA, kinetic<br />
analyses have shown that it does not significantly affect the rate of the cpSRP54·cpFtsY complex formation or the<br />
GTP hydrolysis rate of the complex. Instead, cpSRP43 facilitates the substrate recognition through interactions with<br />
the cargo proteins, LHCPs (1-3). My overall goal for the <strong>2011</strong> SURF is to understand how cpSRP54 and cpFtsY are<br />
able to associate so much more efficiently than their bacterial homologues in the absence of the SRP RNA by<br />
studying their complex structure using crystallography and lysine footprinting mass spec.<br />
61
The Impact of Driving Force on Rates of Electron Transfer in Ruthenium-Labeled Cytochrome P450<br />
Katja E. Luxem<br />
Mentors: Harry Gray, Jeff Warren, and Maraia Ener<br />
This project aimed to examine the high oxidation state intermediates in the catalytic cycle of P450 by studying the<br />
impact of driving force on the rate of electron transfer in our protein model system. By covalently labeling P450BM3<br />
with four different ruthenium-based photosensitizers and photo-exciting them, we were able to generate the<br />
species of interest. This research sheds light on redox events involved in the reactions of the high oxidation state<br />
intermediates in the catalytic cycle of cytochrome P450, knowledge which can be applied to modifying the protein<br />
for human use and studying the health impact of the protein. We have finished synthesis and characterization<br />
(NMR, mass spectrometry, cyclic voltammetry, transient and steady state fluorometry) of the photosensitizers<br />
attached to P450 and free in solution.<br />
Development of a Physical Model of E. coli Carbon Metabolism for Predicting Cell Growth Parameters<br />
Brian Ma<br />
Mentors: Kristala J. Prather, Kevin Solomon, and Richard Murray<br />
Understanding how cells use nutrients is the first step in formulating a systematic practice of metabolic<br />
engineering. In order to optimize the activity of a recombinant pathway and optimize the yield of a desired product,<br />
it must be decided how much carbon a cell can direct towards its various tasks, and therefore how much should be<br />
optimally allocated towards an artificial pathway. Using glucose as the sole carbon source, a simple physical model<br />
has been developed that dynamically links the expression of glucokinase (a glucose-intake, and therefore growthlimiting<br />
protein) and glucokinase mRNA, and the overall growth rate of E. coli KTS strains developed by the Prather<br />
Group. The model incorporates parameters such as promoter strength, anti-sense RNA and induced repressor<br />
expression to provide insight into how glucokinase activity can be controlled through different means and predict<br />
the results. A further expansion of this model calculates from first principles the use of carbon and ATP by the cells<br />
to pursue growth, providing both greater accuracy and a possible physical explanation for the observed growth<br />
trends of E. coli strains.<br />
Heterogeneous Nucleation of Organic Compounds on Crystalline Substrates<br />
Xiaoya Ma<br />
Mentors: Bernhardt L. Trout and David A. Tirrell<br />
Crystallization enjoys diverse applications, including the purification of pharmaceutical ingredients, food<br />
engineering, construction design, and the manufacture of semiconductors. One of the key challenges that arise in<br />
any industry, however, is controlling the properties of the crystalline products. To do so requires a better<br />
understanding of nucleation, the critical step in the crystallization process. In reality, all crystallizations occur by<br />
heterogeneous means. In this project, the heterogeneous nucleation of various organic crystalline materials onto<br />
crystalline surfaces was studied. Epitaxial studies of these systems helped shed light on the fundamental<br />
mechanisms behind heterogeneous nucleation, which are not well understood at present. A long-term consequence<br />
of greater control over nucleation will be greater control over properties of the crystalline product, such as<br />
polymorphism.<br />
Study of Z boson + Higgs � Z boson W boson W* boson � leptons + neutrinos at the International<br />
Linear Collider<br />
James Macdonald<br />
Mentors: Maria Spiropulu and Barry Barish<br />
This project studies the Higgs decay channel with either two or four lepton plus neutrino final states at a<br />
proposed International Linear Collider (ILC), an electron positron collider with center of mass energy 250 GeV and<br />
an integrated luminosity of as assumed in the ILC Letters of Intent. To accomplish this we produced Monte<br />
Carlo data for Higgs production events as well as standard model background. The objective is to extract the Higgs<br />
signal from the dominant standard model background of ZZ and WW dibosons and possibly from “Higgs lookalikes.”<br />
Ultimately we would like to develop handles to distinguish a standard model Higgs from its look-alike states<br />
and improve on the selection criteria by using angular and other new kinematic variables.<br />
Argonaute Protein From Rhodobacter sphaeroides and Its Nucleic Acid Interactions<br />
Vishnu Manoranjan<br />
Mentors: Alexei Aravin and Ivan Olovnikov<br />
RNA interference plays an important role in controlling gene activity. RNA interference is also known to be<br />
mediated by small-interfering RNAs in eukaryotes. These small-interfering RNAs are derived from double-stranded<br />
precursors, and Argonaute proteins aid in the actual target cleavage. Here, we examine Argonaute proteins from<br />
bacteria and the nucleic acids that they interact with. So far, bacterial and archaea Argonaute x-ray structures<br />
have been studied. In vivo functions of these proteins, however, remain a mystery. Here, we observe the<br />
Argonaute protein in Rhodobacter sphaeroides and find that this protein binds not only to small RNAs but also to a<br />
62
number of small DNAs. Preliminary deep sequencing results show that the small RNAs observed have a strong 1-<br />
Uracil bias and a size distribution of 17-18 nucleotides. This research sets the tone for further study of the role of<br />
Argonaute proteins in processing of small nucleic acids<br />
Investigating Mechanisms of Isotopic Fractionation in Microbial Sulfate Reduction<br />
Taylor Martin<br />
Mentors: Alex Sessions and Jacob Waldbauer<br />
Changes in S isotope fractionation can indicate changes in bacterial evolution, the productivity of sulfate reducing<br />
bacteria, and even the oxidation state of the ocean throughout the geologic record. Though the biochemical<br />
mechanism of reduction of sulfate to sulfite is well understood, the physiological basis of S isotope fractionation is<br />
unknown. Desulfobacterium autotrophicum, a marine sulfate-reducing bacterium, was grown under several<br />
different conditions expected to impact fractionation. One culture was grown autotrophically using H2 and CO2,<br />
whereas the other cultures were grown heterotrophically using 10 different carbon sources. The cultures were<br />
sampled over the course of a month, and each sample was analyzed for sulfate isotope composition, sulfide isotope<br />
composition, sulfate concentration, sulfide concentration, and cell density. The isotopic composition of sulfate and<br />
sulfide was compared for each sample to provide a value for the isotopic fractionation that occurs via sulfate<br />
reduction. The different fractionation values from the cultures grown on different substrates, in conjunction with<br />
how each of the substrates is processed by the bacteria, provide insight into the physiological basis of S isotope<br />
fractionation.<br />
Implementation of Motion Simulation Software and Visual-Auditory Electronics for Use in a Low Gravity<br />
Robotic Testbed<br />
William C. Martin<br />
Mentor: John Leichty<br />
The Jet Propulsion Laboratory (JPL) is developing the All-Terrain Hex-Limbed Extra-Terrestrial Explorer (ATHLETE)<br />
to assist in manned space missions. One of the proposed targets for this robotic vehicle is a Near-Earth Asteroid<br />
(NEA), which typically exhibit a surface gravity of only a few micro-g. In order to properly test ATHLETE in such an<br />
environment, the development team has constructed an inverted Stewart platform testbed that acts as a robotic<br />
motion simulator. This project focused on creating physical simulation software that is able to predict how ATHLETE<br />
will function on and around a NEA. The corresponding platform configurations are calculated and then passed to<br />
the testbed to control ATHLETE’s motion. In addition, imitation attitude control thrusters were designed and<br />
fabricated for use on ATHLETE. These utilize a combination of high power LEDs and audio amplifiers to provide<br />
visual and auditory cues that correspond to the physics simulation.<br />
Microenvironments for Studying Extracellular Cues to Cell Behavior<br />
Jordan Maslov<br />
Mentor: Chin-Lin Guo<br />
Cell behavior, such as branching morphogenesis and structure formation, depends on mechanical and geometrical<br />
cues from the extracellular environment. Here, we develop multiple techniques and microenvironments to examine<br />
how these processes take place based on the cells’ external factors. One type of environment uses PDMS stamps to<br />
imprint traps into an agarose and gelatin mixed gel. Cells can be seeded and washed into the traps, and collagen<br />
can be added to induce either branching for MCF-10A cells or structure formation for MDCK cells. We also develop<br />
the use of channels of PDMS microfluidic devices as entrapments to shape and control the formation of cellular<br />
structures.<br />
Semi-Annual Oscillations and Seasonal Variations in Saturn’s Atmosphere<br />
Charlotte A. Mason<br />
Mentor: Glenn Orton<br />
Ground-based telescopes can provide us with images of thermal emission of Saturn in the mid-infrared spectral<br />
range that are useful in tracking various atmospheric properties. Using such data from observations carried out by<br />
my mentor and his colleagues with different mid-infrared wavelength filters, it is possible to study the variability of<br />
emission from different gases in Saturn’s atmosphere. I present the most recent study of low-latitude stratospheric<br />
variability of methane and ethane emission and discuss how the recent data fit with predictions for a semi-annual<br />
oscillation structure. Variability of strong emission from methane, being well mixed in Saturn’s atmosphere, tracks<br />
the variability of stratospheric temperatures. Variability of strong emission from ethane can, with the temperature<br />
information from the methane emission, be used to detect variability of the abundance of ethane itself. Besides the<br />
low-latitude oscillation, I also examine the high-latitude stratospheric variability of methane and ethane, and the<br />
variability of thermal emission in the troposphere as a function of time and compare these data with models for<br />
seasonal forcing of the atmosphere.<br />
63
Electromyographic Control of Tactile-Sensing Enabled Prosthetic Hand<br />
Blaine Matulevich<br />
Mentors: Gerald Loeb, Jeremy Fishel, and Joel Burdick<br />
Myoelectric prosthetic hands generally provide poor, if any, tactile feedback to their users; as a result, gripping<br />
objects is often both strenuous and difficult. Attaching multimodal tactile sensors to a currently available hand can<br />
provide tactile feedback for transmission to the user as well as enabling automated slip-detecting grip-control<br />
algorithms. While reflex-like grip control would improve the ability to grasp objects, fully autonomous grip control<br />
may be undesirable; for this project, we sought to integrate the user’s intent to scale the magnitude of the<br />
response. A circuit was designed and constructed that filtered raw electromyographic (EMG) signals from electrodes<br />
attached to the user’s forearm, outputting a smooth envelope indicating the approximate level of effort exerted by<br />
the user. A servo-amplifier was connected that enabled the user to drive the hand directly with EMG signals. A<br />
force gauge was designed, built, and utilized to create curves that correlate the voltage of the user’s signal with the<br />
force output by the hand. In future work, the user’s EMG signal will provide a measure of the intentions of the user<br />
and will be combined with tactile information about object loads and slip to determine the actual force generated<br />
by the actuator. This should result in a user-based control system that feels more natural and provides significantly<br />
better tactile feedback than current designs.<br />
Understanding the Effects of the Pulse Charging Method on Dendrite Growth in Lithium Metal Batteries<br />
Matthew Mayers<br />
Mentors: Thomas Miller and Jakub Kaminski<br />
The availability of powerful, lightweight, and long-lasting battery technologies is an essential prerequisite for the<br />
design and development of new technologies and devices. Lithium metal is an ideal anodal component for these<br />
battery technologies because of its high redox potential and low molecular weight. However, the realization of<br />
secondary, or rechargeable, lithium metal batteries has been frustrated by the formation of metal dendrites at the<br />
anode during recharging, leading to reduced lifetimes and catastrophic failures of the cells. Using coarse-grained<br />
molecular simulation methods, we develop a three-dimensional model to directly simulate the non-equilibrium<br />
processes of lithium reduction, plating, and dendrite formation on a nanoparticle. By observing the conditions<br />
under which dendrites nucleate and subsequently elongate, we hypothesize that the pulse charging method could<br />
aid in preventing the onset of dendrite formation. We show that the pulse charging method can be used with high<br />
overpotentials to cause plating densities identical to those seen in the case of low overpotentials. Finally, we see<br />
that although battery materials give rise to an inherent, relatively low upper limit on the pulsing frequencies that<br />
may be used, that upper limit still allows denser plating in the presence of a healthy SEI, a result that reinforces<br />
the importance of the SEI in helping to prevent dendrite formation.<br />
Online Real-Time Prediction of Motor Decisions Using Intracranial Brain Recordings in Humans<br />
Naomi McArthur<br />
Mentors: Christof Koch and Uri Maoz<br />
Neuroscientific research of free will research over the last 30 years has revealed that binary motor decisions might<br />
be predicted well in advance of a person’s conscious intention to move. However, such research was never<br />
conducted in a live setting – online and in realtime – with a patient before. The goal of this project was to gather<br />
data from patients implanted with intracortical electrodes as they press a button to advance through a series of<br />
images, and then incorporate the data gathered into a learning model that would predict the timing of the button<br />
press. This learning model will be used in real time to predict the movement of the patient before movement<br />
occurs. During the course of the project, new data was collected, the parameters of the patient’s task were<br />
changed, and the learning model was tested against available data. Future directions would involve refining the<br />
learning model further, testing in real time with new patients, and observing the effects on the patient’s agency<br />
and decision-making when they realize that their movements are being predicted.<br />
Seismocloud: An Investigation Into Efficient SCEDC Earthquake Data Manipulation<br />
Stephen Meisenhelter<br />
Mentors: Robert Clayton and Ellen Yu<br />
The Southern California Seismic Network currently stores all collected data to a server bank on Caltech campus and<br />
a redundant system in another on campus building. While this redundancy does provide some protection from<br />
failure, the entire system is still vulnerable to data loss from earthquakes in Pasadena. The system also suffers<br />
from bottlenecking leading to server slow-downs and denial of service during large events.<br />
Seismocloud is an experimental cloud based data server that utilizes the Google App Engine as a replacement for<br />
the current SCEDC server system. This system will solve the data redundancy and bottlenecking problems<br />
associated with maintaining physical servers while also reducing costs. Due to design changes in the Google App<br />
Engine, it became possible this year to drastically increase the efficiency of Seismocloud by precompiling a single<br />
unified record for each seismometer’s submitted data. This simplifies storage and allows for a more consistent data<br />
set. These changes make Seismocloud feasible for large-scale deployment.<br />
64
Improving Redox Couple Mass Transport in Silicon Wire-Array Liquid Junction Solar Cells<br />
Andrew C. Meng<br />
Mentor: Nathan S. Lewis<br />
Silicon wire-array liquid junctions show promise in solar to fuel conversion applications. However, high aspect ratio<br />
wire array geometries present additional mass transport complications relative to their planar analogs. Redox<br />
couple depletion at the bottom of the wire arrays was observed and accounted for carrier collection loss. To<br />
facilitate redox couple mass transport, a novel architecture is proposed to achieve higher efficiencies in liquid<br />
junction solar cells. Metal films were deposited onto the bottom of Si wire arrays by electron-beam evaporation. To<br />
prevent direct contacts to the p-Si wire array from the metal films, electrically insulating layers of SiO2 and Al2O3<br />
were prepared by thermal oxidation and electron-beam evaporation, respectively. During cell operation, the metal<br />
films served as additional counter electrodes, which regenerate redox couple species where it is most likely to<br />
deplete. Computational modeling suggested that the supply of redox species from the metal films minimizes<br />
concentration overpotential and solution resistance. In this study, the energy-conversion properties of Si wire-array<br />
photocathodes with metal film counter electrodes were characterized in contact with cobaltocene (CoCp2 +/0 ) redox<br />
couple. Significant improvements of current-voltage characteristics were observed compared to the conventional<br />
cell architecture. Under high illumination intensities or low stirring conditions, the new electrode design<br />
demonstrated superior performance in short-circuit current and fill factor. Spectral response also confirmed near<br />
unity collection of photogenerated carriers at the bottom of the wire array. The proposed structure, which<br />
essentially minimizes the effective ion transport path length to the photoelectrode, also provides design guidance<br />
for future tandem cells for water splitting reactions.<br />
Syntheses of Fluorinated Cobalt Diglyoxime Complexes for Electrocatalytic Hydrogen Evolution<br />
Rocio Mercado<br />
Mentors: Harry B. Gray and Michael J. Rose<br />
Work on various cobaloximes has shown them to be promising catalysts for electrocatalytic hydrogen evolution at<br />
low overpotentials. However, while cobaloximes of the forms Co(dmgBF2)2L2 and Co(dpgBF2)2L2 exhibit Co II/I<br />
reduction potentials in organic solvents close to NHE (-2.4 V vs SCE), placing them in water increases the<br />
magnitude of this potential enough to render the catalyst ineffective. Because of this, we are studying various<br />
cobalt complexes containing the fluorinated diphenylglyoxime ligand (“dAr F gH2”). We are studying these complexes<br />
because the electron withdrawing pentafluorophenyl groups will shift the Co II/I reduction potential in the positive<br />
direction. As a consequence, under aqueous conditions the potential of the catalyst will fall within an adequate<br />
range. So far, we have synthesized Co(dAr F gH)2(py)2, as was confirmed by X-ray crystallography and infrared<br />
spectroscopy. We have studied this compound by cyclic voltammetry, and found the Co II/I reduction potential to be<br />
-0.88 V vs. Ag/AgCl in methylene chloride. After -1 V, the CV is not well defined and suggests rearrangement of<br />
the Co I complex. These possible rearrangements could be due to the weak proton bridge linking the two ligands.<br />
Because of this, the proton bridges in the complex have been substituted by more stable difluoroboryl and<br />
diphenylboryl bridges, allowing for simple comparison of the BF2/BPh2-bridged cobaloxime to the proton-bridged<br />
cobaloxime. The syntheses, X-ray structures, and electrochemical characterizations of Co(dAr F gBF2)2(py)2, and<br />
Co(dAr F gBPh2)2(py)2 are underway at this time.<br />
Committee Karate: Divide and Conquer<br />
Brian Merlob<br />
Mentor: Charles Plott<br />
The outcome of a group decision making process is inherently dependent on the voting system employed, the<br />
availability of information, and the presence of veto players. This dependence potentially provides key members of<br />
the committee—particularly agenda-setters—with near-dictatorial power over the outcome. We explore through<br />
experimentation and simulation what additional influence over the outcome is gained by an agenda-setter that can<br />
divide committees into subcommittees.<br />
Novel Computational Methods for High Speed Radio Transient Searches<br />
Omar Mezenner<br />
Mentor: Walid Majid<br />
Radio transients are celestial objects that emit radiation in short time windows. One example is a pulsar, a rotating<br />
neutron star that emits a beam of electromagnetic radiation. Detection and classification of radio transients<br />
requires us to boost signal-to-noise on large datasets acquired from radio telescopes observing the radio sky.<br />
Errors in the datasets are due to dispersion, an effect of electromagnetic propagation through plasma in the<br />
interstellar medium that causes the signal to be dispersed across multiple frequency channels unevenly, with lower<br />
frequency components lagging behind higher frequency components. This causes a widened pulse width in the<br />
time-domain, and therefore a lower signal-to-noise ratio. To correct dispersive effects, dedispersion techniques<br />
may be applied incoherently (after detection) or coherently (before detection). This process involves applying a<br />
correctional time shift to lower frequency channels to align the signal with that of higher frequency channels, and<br />
65
then summing across all channels. The dedispersed time-series then possesses a narrower pulse width and higher<br />
signal-to-noise. We are exploring novel techniques to implement dedispersion strategies on GPU platforms because<br />
of the highly parallelizable nature of vector computations, as high-end GPUs hold two to three orders of maginitude<br />
more cores than their contemporary CPU counterparts.<br />
Effects of DNA Mismatches on DNA-Protein Crosslinking via Guanine Oxidation<br />
Kelsey Miller<br />
Mentor: Eric Stemp<br />
8-oxoguanine is a common oxidation product in DNA and can lead to mutation. The metallointercalator<br />
Ru(phen)2dppz 2+ is a useful luminescent probe for DNA that has also found use as a guanine-selective oxidizing<br />
agent via the flash-quench technique. Here, we use its osmium analogue as a selective way to oxidize 8oxoguanine<br />
in double-stranded DNA, as visualized by oxidative DNA-protein crosslinking. With the 3+/2+ couple at<br />
1.15 V for Os (phen)2dppz 2+ , this metallointercalator should be able to oxidize 8-oxo-G (~0.7 V) without oxidizing<br />
guanine (~1.3V). In plasmid DNA where 8-oxo-G has been incorporated, flash-quench treatment with Os<br />
(phen)2dppz 2+ and Co(NH3)5Cl 2+ leads to crosslinking with histone. Furthermore, in gel shift experiments with a<br />
duplex of the oligonucleotide 5’-ATATGATAT8GATATGATAT-3’ (8 = 8-oxo-G) and its complement, flash-quench<br />
treatment with Ru(phen)2dppz 2+ in the presence of histone leads to a band of intermediate mobility (presumably<br />
1:1 crosslink) and to a band of well-shifted material. In contrast, analogous treatment with Os(phen)2dppz 2+<br />
produces only the band of intermediate mobility, consistent with the presence of only site that is oxidizable by the<br />
osmium complex.<br />
The Solar Decathlon: Improving House Performance Through Appliance Optimization<br />
Zeke Millikan<br />
Mentor: Melany Hunt<br />
The Solar Decathlon is a biennial competition sponsored by the Department of Energy in which college teams from<br />
around the world design and build solar powered homes. The houses are brought to Washington D.C. where they<br />
are judged and opened to the public, showcasing new strategies of energy efficiency in residential living. The<br />
California Institute of Technology and the Southern California Institute of Architecture have joined together to<br />
make an innovative and energy conscious house for the competition. The home, CH:IP (Compact Hyper-Insulated<br />
Prototype), makes energy efficiency a priority through interconnected systems and careful energy budgeting. More<br />
than forty percent of the energy budget is devoted to appliances; thus their performance has a large impact on the<br />
overall home performance. The appliances were tested extensively for energy and water use in different settings<br />
and configurations. The resultant numbers were used to choose optimal operational strategies and to verify overall<br />
energy performance of the home.<br />
Exploiting the Weak Temperature Gradient Approximation for Climate Theory on Slowly Rotating,<br />
Tidally Locked Planets<br />
Sean M. Mills<br />
Mentors: Dorian Abbot, Raymond Pierrehumbert, and Yuk Ling Yung<br />
Tidally locked planets are subject to extreme temperature variations due to the stellar flux directly warming only<br />
one side of the planet. This is important because planets in the habitable zone around M dwarf stars are likely to be<br />
tidally locked. Such planets are unlikely to be habitable if their antistellar temperatures are low enough that CO2<br />
will condense. This problem has previously been investigated using GCMs, which explicitly solve for atmospheric<br />
dynamics. In order to gain a greater understanding of the effect of different mechanisms on the temperature<br />
profile, we use a lower-order energy balance model here. We consider tidally locked planets that rotate slowly<br />
enough that we can neglect the Coriolis force, which allows us to assume that atmospheric temperature at any<br />
given height is independent of horizontal position (the weak temperature gradient approximation). This allows us to<br />
easily isolate and contrast the effects of different physical mechanisms, such as greenhouse gas level and surface<br />
turbulent exchange, on the resulting temperature profile. We find that the effect of turbulent exchange on climate<br />
saturates at fairly low values (very smooth planets), whereas CO2 has a consistently strong effect on climate.<br />
Investigating the Role of miR125b in Hematopoiesis and Development of a Myeloid Leukemia<br />
Aarathi Minisandram<br />
Mentors: David Baltimore and Alex So<br />
miR125b has been shown to be an important player in hematopoiesis. When miR125b is overexpressed in mice,<br />
the mice develop severe leukemia. We identified this leukemia to be primarily myeloid in nature, although in<br />
certain cases, it can be lymphoid. The primary cause of the myeloid cancer was a rearrangement of the<br />
hematopoietic system favoring myeloid progenitors and decreasing the number of lymphoid progenitors. We<br />
developed a loss-of function system for miR125b expression in mice and confirmed that miR125b has a regulatory<br />
effect on hematopoiesis. Finally, we investigated the primary target of miR125b, lin28, and show that miR125b<br />
causes a rearrangement of the hematopoietic system by decreasing the expression of lin28.<br />
66
Proteomic Analysis of Histone Post-Translational Modifications and Their Cellular Targets<br />
Neeli Mishra<br />
Mentors: Benjamin A. Garcia and Jack L. Beauchamp<br />
Histones are small proteins that function to package genomic DNA into the basic unit of chromatin, the<br />
nucleosome. Post-translational modifications (PTMs) of histones are called epigenetic (heritable changes in gene<br />
expression that occur without alterations in gene sequence). Histone PTMs occur on multiple but specific sites,<br />
suggesting that histones can act as a signaling platform for nuclear events. A “histone code” hypothesis has been<br />
put forward to explain how these PTMs are responsible for the regulation of transcriptional activity through the<br />
alteration of chromatin states. In the first study, we seek to investigate the impact of epigenetic gene expression<br />
on both histone modifications and the nuclear proteome. The project focuses on the knockdown cell line of EZH2,<br />
a clinically significant histone methyltransferase associated with developing tumors. Using proteomics to directly<br />
investigate the cellular function, we combine SILAC labeling and mass spectrometry (MS) to characterize the<br />
alterations in abundances for nuclear proteins in these epigenetically modified cells. Histones are typically studied<br />
using electron capture/transfer dissociation or collision-induced dissociation MS. The second study focuses on<br />
testing an alternate, recently developed free-radical initiated peptide sequencing (FRIPS) technique to selectively<br />
cleave peptides at specific amino acid residues and potentially improve histone PTM analysis.<br />
Progress Toward the Synthesis of Gagunin E, a Polyoxygenated Terpene Isolated From the Sponge<br />
Phorbas sp.<br />
Austin Moehle<br />
Mentor: Brian Stoltz<br />
Gagunin E, one of a large set of polyoxygenated terpenes isolated from the sponge Phorbas sp., exhibits high<br />
cytotoxicity toward the human leukemia cell-line K562 with an LC50 of 0.03 μg/mL (50 nM). Despite its potent<br />
biological activity, a synthesis of gagunin E has yet to be reported. The initial goal of the project was to construct<br />
the tricyclic core of gagunin E, which contains fused 5-, 6-, and 7-membered rings. Although cyanthiwigin F, a<br />
diterpene recently synthesized by the Stoltz group, possesses a similar tricyclic core, gagunin E is more complex<br />
due to oxygenated stereocenters bearing various ester groups. An array of ruthenium catalysts were screened for<br />
the 7-membered ring-closing metathesis and several approaches to the construction of the three rings were<br />
attempted, ultimately resulting in the desired synthesis of the tricyclic core of gagunin E from succinic acid, a<br />
commercially available precursor. Future work on this project will involve performing C-H oxidations on this core to<br />
attach the appropriate hydroxyl groups and ester moieties, followed by an investigation of the effects of various<br />
hydroxyl functionalization patterns on this compound’s biological activity.<br />
Characterization of the Partial Duplication of the �7 Nicotinic Acetylcholine Gene (CHRFAM7A) and Its<br />
2-Base Pair Deletion Polymorphism<br />
Mahati Mokkarala<br />
Mentors: Henry A. Lester and Ying Wang<br />
Previous research indicates that the �7 nicotinic acetylcholine receptor mediates P50 auditory gating; a defect in<br />
the P50 auditory response is a common symptom for schizophrenia. Of particular interest is the partial duplication<br />
of exons 5-10 of the �7 nicotine acetylcholine receptor gene (CHRFAM7A) prevalent only in humans. A recent study<br />
found that CHRFAM7A has a 2bp deletion (CHRFAM7A2bpdel) polymorphism linked to schizophrenia. Unfortunately,<br />
the structure and function of the gene products of CHRFAM7A (dup�7) and CHRFAM7A2bpdel (dup��7) are<br />
unknown. We are studying the subcellular localization of dup�7 and dup��7 in neuronal culture by overexpressing<br />
fluorescently tagged �7, dup�7 and dup��7 in rat hippocampus neuron cultures and comparing the distribution of<br />
receptor subunit with stained cellular structures such as the endoplasmic reticulum. Receptor subunit localization<br />
patterns may help us understand the function of dup�7 and dup��7 in mammalian neurons. We are also<br />
investigating the N-terminal sequence of dup�7 and dup��7 overexpressed in IMR-32 human neuroblastoma cells<br />
by applying the Edman degradation technique. Optimization of transfection efficiency for rat hippocampus neuron<br />
and human IMR-32 neuroblastoma cell cultures are ongoing.<br />
Conformational Equilibria of 1,4-Butanedioic Acid and Its Anions as a Function of the Deuterium Oxide<br />
Mole Fraction in Mixed Solvent Environments of Deuterium Oxide and Dimethyl Sulfoxide as<br />
Determined Through NMR Spectroscopy<br />
Greg Moore<br />
Mentor: John D. Roberts<br />
Scalar J13/J14 vicinal proton couplings of 2,3- 13 C labeled 1,4-butanedioic acid and its anions were used to evaluate<br />
conformational equilibria as the mole fraction of deuterium oxide in mixed solvents of deuterium oxide and<br />
dimethyl sulfoxide was progressed from 0 to1. Results for all compounds considered showed a surprisingly gradual<br />
transition from the conformational preference exhibited in pure dimethyl sulfoxide to that seen in pure deuterium<br />
oxide. This gradual change suggests that deuterium oxide is not as effective at disrupting the intramolecular<br />
hydrogen bonding of monohydrogen 1,4-butanedioate or protonating the carboxylate groups of 1,4-butanedioate in<br />
67
dimethyl sulfoxide solutions as was initially thought. Additionally, the calculated fraction of gauche conformers for<br />
1,4-butanedioate in pure anhydrous dimethyl sulfoxide was ~.20, a lower figure than the ~.40 previously reported<br />
and more consistent with computational predictions.<br />
Activation Responses to Hallucinogen Administration in a Mouse Model of Maternal Immune Activation<br />
Marlyn J. Moore<br />
Mentors: Paul H. Patterson and Natalia Malkova<br />
Hallucinations are a hallmark symptom of schizophrenia. Moreover, schizophrenic patients show a greater response<br />
to hallucinogenic drugs than control subjects. Epidemiologic studies have found that maternal immune activation<br />
(MIA) increases the risk of schizophrenia in the offspring. In a mouse model of this risk factor we analyzed the<br />
behavioral and molecular response of MIA offspring to an injection of the hallucinogenic drug DOI. Compared to<br />
controls, these mice display greater behavioral responses and higher levels of the immediate early gene (IEG)<br />
egr-1, a marker of neuronal activity, in prefrontal cortex. Thus, as in human schizophrenia, MIA offspring appear to<br />
be more sensitive to a hallucinogen.<br />
Mist Control Due to the Complementary Interactions of Polymer End Groups<br />
Rachel S. Moore<br />
Mentors: Julia A. Kornfield and Jeremy Wei<br />
The reason airplane crashes are so deadly is that the, during the crash, the fuel forms a highly combustible, fine<br />
mist. Adding a certain polymer to the fuel can control this fine mist by making the droplet size bigger, rendering<br />
the mist non-explosive. Past polymers have been able control misting, but they were not usable due to either shear<br />
degradation or insolubility in fuel. My polymer is designed to associate via reversible, complementary associating<br />
groups on the polymer ends. This reversibility is what is unique about this polymer, and the complementary nature<br />
of the association allows for efficacy at low concentrations without drastically altering the rheological properties of<br />
the fuel. In my studies, I utilized different methods for characterizing the polymers, including rheology, a spray<br />
test, a combustion test, and a shear degradation test. All tests so far indicate that the polymer associates in the<br />
predicted way and that the interaction between polymers is very strong. This strong interaction means that only<br />
four associating groups on each end need to be used. In turn, this means that the synthesis and mass production<br />
of the polymers will be feasible.<br />
Low Cost, High-Density Protein Arrays for Sensors and Fuel Cells<br />
Melody A. Morris<br />
Mentors: Bradley Olsen, Carla Thomas, Matthew Glassman, and David Tirrell<br />
Compared to traditional chemical synthesis, the use of enzymes has dramatically expanded the diversity of<br />
chemical conversions possible. Enzymes have many advantages: substrate specificity, high catalytic rate, mild<br />
operating conditions, biodegradability and renewability, and the ability to operate on non-traditional substrates.<br />
However, enzymatic catalysts and sensors face three key limitations that this project seeks to remedy: low density<br />
of active sites, short enzyme lifetimes or limited operating conditions, and high costs of enzyme purification and<br />
device construction. Recent work has shown that conjugates of globular protein and thermoresponsive polymer will<br />
self assemble into lamellar nanostructures with high protein density and preservation of folding. Here carbonic<br />
anhydrase (hCAII), an enzyme that converts carbon dioxide and water into bicarbonate and protons, was chosen<br />
as a model catalytic globular protein due to its potential application in emission reduction and environmental<br />
remediation. A thermoresponsive elastin-like protein (ELP) was fused to hCAII as a tool for low-cost purification<br />
and to direct self-assembly of the enzyme into nanostructured plastics and hydrogels. This project aims to develop<br />
methods for self-assembly of hydrated gels and solid-state nanostructures of ELP-enzyme fusion proteins, to<br />
explore the effects of ELP secondary structure on the formation of the nanostructures, and to enhance stability and<br />
activity of the enzyme.<br />
Multiple Laser Frequency Modulated Ranging System<br />
Alexander T. Mouschovias<br />
Mentors: Amnon Yariv and Arseney Vasilyev<br />
We investigate the combination of swept frequency optical waveforms for resolution enhancement in a diode laser<br />
frequency-modulated ranging experiment. To this end we build an optoelectronic feedback system capable of<br />
stitching four swept-frequency laser sources. We use this system to concatenate chirped vertical cavity surfaceemitting<br />
lasers (VCSELs) at wavelengths of 1550 and 1310 nm, and demonstrate a four-fold improvement in range<br />
resolution over a single source ranging system.<br />
68
The Incidence of Active Galactic Nuclei in Galaxy Groups<br />
Eric Mukherjee<br />
Mentors: Christine Jones, Andrew Goulding, and Fiona Harrison<br />
Properties of Active Galactic Nuclei (AGNs) are linked to properties of their host galaxies. Reseach suggests AGNs<br />
are the result of accreting black holes at centers of galaxies, and affect properties like star formation rate, stellar<br />
mass, and galaxy color. Knowing how the large scale environment influences AGN evolution may further our<br />
understanding of this relationship. We seek to constrain AGN environments in the 9 square degree Chandra Bootes<br />
survey. We used ~15,000 optically observed galaxies in the AGN and Galaxy Evolution Survey (AGES) catalog to<br />
determine whether AGN form in groups. We developed a friends-of-friends algorithm to locate galaxy groups<br />
associated by distance and velocity, and a matching program to associate groups with possible X-ray detections.<br />
We found 4,758 (~30%) galaxies in 1,554 groups in the AGES catalog with 0.01
Density of States of Nematic Emulsion<br />
John C. Napp<br />
Mentors: Arjun Yodh, Matthew Lohr, and Fiona Harrison<br />
Knowledge of the vibrational density of states of a material is a fundamental key to understanding its physical<br />
properties. In this project, the vibrational density of states of an emulsion of glycerol in the nematic liquid crystal<br />
(NLC) 4'-pentyl-4-biphenylcarbonitrile (5CB) is investigated. A thin layer of NLC is placed on top of a thick layer of<br />
glycerol by dissolving the 5CB in hexane, which spreads evenly over the glycerol and quickly evaporates. The<br />
system is then heated, allowing the 5CB to enter the isotropic phase and the glycerol to diffuse through the liquid<br />
crystal. When cooled, the 5CB reenters the nematic phase, and the glycerol condenses into droplets in a hexagonal<br />
lattice pattern. Some of these droplets are trapped at the NLC-air interface, and their vibrational motion can be<br />
tracked via video microscopy and standard particle tracking algorithms. Measurement of the displacement<br />
correlations between particles yields information on the vibrational properties of the system which allows<br />
calculation of the vibrational density of states.<br />
Phosphino–Borane Transition Metal Complexes for Energy Storage Reactions<br />
Luis Navarro<br />
Mentors: Jonas C. Peters and William H. Harman<br />
This project’s goals have been the synthesis and characterization of first row transition metal complexes using<br />
diphosphino-borane ligands and studying their reactivity as they relate to activation of hydrogen and carbon<br />
dioxide. Altering the sterics of these systems has been shown to dramatically alter their reactivity and particularly<br />
the nature of the boron-metal bonds.<br />
Investigating the Protein and RNA Components of the Chromatoid Body<br />
Pushpa Neppala<br />
Mentors: Alexei Aravin and Evelyn Stuwe<br />
Small non-coding RNAs are known to associate with the Argonaute (Ago) family of proteins, and their presence<br />
usually results in the silencing of the target gene expression. A just recently discovered class of small noncoding<br />
RNAs, the piRNAs (Piwi-interacting RNA), interacts with Piwi proteins (a clade of the Ago family) and is involved in<br />
repression of transposable elements in germ line cells. In mammalian male germ cells, a cloudlike cytoplasmic<br />
structure called the chromatoid body has been thought be composed of proteins and small RNAs involved in the<br />
piRNA pathway. Elucidating the nature of the chromatoid body could help to improve our understanding of<br />
translation and the highly complex process of spermatogenesis. The objective of this research is to further study<br />
and identify the protein and RNA components of the chromatoid body, through developing a novel isolation protocol<br />
that couples cross-linking techniques with immunoprecipitation in mouse testes. After formaldehyde treatment and<br />
immunoprecipitation of a known protein component of the chromatoid bodies, MILI, proteins were identified<br />
through Western blot analysis. RNA molecules were also isolated and identified through quantitative PCR. In the<br />
future, the RNA will be sequenced and we hope to perform mass spectrophotometry on the chromatoid bodies. This<br />
involves also developing an isolation protocol in which the chromatoid bodies are not cross-linked but can still be<br />
immunoprecipitated while kept intact.<br />
Social Judgments of Other People: Comparing 2D, 3D, and Live Formats<br />
HongAn Nguyen<br />
Mentors: Ralph Adolphs and Laura Loesch<br />
We make inferences about people every day based on how they present themselves. People are free to post 2D<br />
videos of themselves as well as make video calls via the internet. With the recent introduction of consumer-level<br />
3D technology, they have yet another format for interacting with the greater community. We were interested in<br />
whether the inferences people make about each other in terms of attributes such as trustworthiness, social<br />
dominance, and physical threat, as well as the facial information they used to make those judgments, were<br />
influenced by the format of the interaction. Participants’ eyes were tracked as they looked at facial stimuli and gave<br />
explicit social judgment ratings. The facial stimuli varied across these different formats of interaction: 2D and 3D<br />
video, live one-way video, live two-way video, actors through a one-way mirror, and actors interacting face-toface.<br />
We plan on looking at the ratings using ANOVA and at eye tracking information through the SHINE toolbox in<br />
MATLAB to see if there are statistically significant differences among these formats of presentation.<br />
Programmable Genetic Regulation by sRNA-mRNA Hybridization<br />
Paul D. Nguyen<br />
Mentors: Richard M. Murray and Joseph Meyerowitz<br />
Small non-coding ribonucleic acids (sRNAs) play key roles in regulatory pathways in an extensive range of<br />
prokaryotes and eukaryotes. Drawing inspiration from this natural device, we have begun an investigation to<br />
imitate the mechanism of regulation in vivo with Escherichia coli, and develop engineered and programmable<br />
biological circuitry that utilizes sRNA-mRNA interaction for regulation. We constructed a simple genetic circuit that<br />
70
is designed to give a delayed response of a reporter gene for green fluorescent protein (GFP). This circuit can be<br />
used in a programmable way to give a desired regulation of the GFP gene by some intended temporal delay in<br />
expression, as shown by our computer simulations of the system. To confirm the intended delayed regulatory<br />
capabilities of our genetic circuit, we characterized this system by taking various fluorescence measurements of the<br />
cells with this system against other cells.<br />
Non-Covalent, Self-Assembled Polyads for Multistep Electron and Energy Transfer<br />
Eva Nichols<br />
Mentor: Daniel T. Gryko and Harry Gray<br />
Systems for multistep electron and energy transfer are important to the field of solar energy because they provide<br />
access to long-lived charge-separated states, allowing electrochemical potential to be harnessed for catalysis.<br />
Covalently bonded donor-acceptor systems have been documented, but a polyad system capable of self-assembly<br />
via intermolecular interactions still remains an elusive target. The design of three- and four-component noncovalent<br />
supramolecular systems for electron and energy transfer is described. Individual donor and acceptor<br />
moieties have been designed and synthesized with groups that allow them to self-assemble in a complementary<br />
and site-specific fashion (via hydrogen bonding and ionic interactions). Preliminary data concerning the selfassembly<br />
and electron or energy transfer of each supramolecular complex will be presented.<br />
State Space Partitioning Algorithm in TuLiP<br />
Lars Petter Nilsson<br />
Mentors: Richard Murray, Necmiye Özay, and Ufuk Topcu<br />
The Temporal Logic Planning (TuLiP) Toolbox is a Python software package implementing the correct-byconstruction<br />
controller synthesizer method Receding Horizon Temporal Logic Planning. One step in this method is<br />
to partition the continuous state space into discrete states and establish reachability relations between the discrete<br />
states. Cells in the partition are described by convex sets with flat sides, called polytopes. State space<br />
discretization requires intersections, set differences and projections to be performed on polytopes, the optimal<br />
choice of method for these operations depends on the characteristics of the polytopes. The algorithm used for<br />
discretization can be tuned with respect to cell requirements and termination criteria in order to produce a good<br />
partition.<br />
We also investigate if using overlapping instead of disjoint polytopes can decrease computational complexity. The<br />
discretization algorithm is evaluated by letting TuLiP synthesize a controller for a fuel tank problem.<br />
Experimental Investigation Into Shock Wave Boundary Layer Interactions of Reflected Detonation<br />
Waves<br />
John J. Odell<br />
Mentors: Joseph E. Shepherd and Jason Damazo<br />
Accurate prediction and modeling of the damage caused by gaseous detonation necessitates understanding the<br />
physics of shock wave reflection as it pertains to supersonic combustion waves. Previous experimental evidence<br />
suggests that detonation reflection deviates from one-dimensional inviscid wave reflection theory thus<br />
inhibiting accurate prediction of material deformation. We propose that this deviation is caused by interaction<br />
between the boundary layer induced by the incident detonation with the shock wave created when a detonation<br />
undergoes normal reflection analogous to the classic phenomenon of shock wave bifurcation studied in shock<br />
tubes. This shock wave boundary layer (SWBL) interaction is experimentally investigated using optical techniques<br />
alongside pressure and thermocouple measurements for the case of nitrous oxide--hydrogen detonations of varying<br />
stoichiometry. These measurements allow us to examine the boundary layer development behind gaseous<br />
detonations and the boundary layer effects on both the fluid dynamics of the reflected shock wave and the<br />
destructive potential of these waves: It is shown that viscosity plays a key role in the behavior of reflected<br />
detonation waves. Analytical theory is developed corresponding to non-reacting shock wave bifurcation as<br />
pioneered by Mark to explain the observed phenomena and highlight the differences between shock and detonation<br />
waves.<br />
Microfluidic Sensors With Integrated Germanium Photodetectors<br />
Nicholas Ogden<br />
Mentors: Axel Scherer and Jinqing Wang<br />
Microfluidic devices have great potential to supplement or replace current biochemical analysis machines such as<br />
those used for blood tests. For this to happen, microfluidic and electronic technology must be compatible. This<br />
requires accurate and reliable sensors to be developed for use in microfluidic devices. In this case, an optical<br />
sensor is considered due to the small size, good sensitivity, and lack of required ancillary equipment. Our sensor<br />
uses a silicon waveguide to channel 1.55um light (commonly used in telecommunications) into a germanium<br />
photodetector. Because, the size and shape of the germanium is critical in the performance of the sensor, Finite-<br />
Difference Time-Domain software is employed to test potential designs. Construction starts with a silicon-on-<br />
71
insulator substrate, on which a silicon waveguide is deposited. A germanium strip is then deposited and the device<br />
undergoes thermal annealing to improve its crystallinity. Lastly, a layer of gold is deposited on both sides of the<br />
germanium to serve as electrical contacts for analysis circuitry.<br />
In situ Amplification for Imaging miRNAs and mRNA Splice Variants<br />
Kanenori (Kane) Okamoto<br />
Mentors: Niles Pierce, Aneesh Acharya, and Harry Choi<br />
RNA plays a key role in gene regulation; it is involved in almost every DNA-related cellular process. In addition to<br />
providing transcripts for protein synthesis, RNA can promote gene expression, silence genes, and catalyze various<br />
cellular reactions 1 , among many other functions. Imaging mRNA expression patterns in intact cells or organisms<br />
with in situ hybridization technology has helped bring about numerous discoveries in the field of embryonic<br />
development and gene regulation. Recent developments such as locked nucleic acid (LNA) probes have enabled<br />
in situ imaging of mRNA splice variants 2 and microRNAs (miRNAs) in single mammalian cells 3 and whole-mount<br />
embryos 4-6 . Despite these advances, multiplexed in situ imaging of mRNA splice variants and miRNA within a single<br />
sample has not yet been demonstrated. Multiplexing minimizes sample variations and allows us to study the<br />
relationships among genes simultaneously in the same sample. The Pierce group has developed a multiplexed<br />
fluorescent in situ amplification using Hybridization Chain Reactions (HCR) in intact zebrafish embryos under the<br />
microscope 7 . In our studies, we have improved and adapted this technology for use with multiple initiator probes<br />
and with flow cytometry for quantitative analysis and propose a multiplexed in situ procedure to image miRNAs and<br />
splice variants in mammalian cells.<br />
Redesigning IgG Antibody 3BN60 to Improve Neutralization of HIV<br />
Alejandra I. Olvera<br />
Mentors: Pamela Bjorkman and Jennifer Keeffe<br />
Human Immunodeficiency Virus (HIV) has several properties that make it successful at evading the immune<br />
system, thus preventing antibody-mediated neutralization. One property is HIV’s ability to rapidly mutate and<br />
conceal epitopes on its trimeric gp120-gp41 envelope spikes. Recently, Pamela Bjorkman and colleagues suggested<br />
a novel evasion strategy by HIV: that the low density of HIV spikes impedes antibodies from binding with avidity,<br />
the ability to bind two epitopes simultaneously. Avidity enhances the affinity of an antibody for its antigen and has<br />
been shown to be important in the neutralization of other viruses. To address this problem, the Bjorkman Lab<br />
seeks to reengineer the structure of antibodies to enable multivalent binding. As part of this goal, we have<br />
designed novel antibody-like reagents with two Fabs joined by a variable length linker. We expressed the Fabs<br />
separately and ligated them with Sortase A, an enzyme that catalyzes the creation of a novel peptide bond. By<br />
changing the length of the linker we can optimize the distance between the antigen binding sites and potentially<br />
achieve inter- or intra-spike cross-linking, creating an avidity effect that may enhance the antibody’s potency<br />
against HIV. These type of antibody-like reagents could be administered via passive immunization or gene therapy<br />
to HIV infected individuals and inform efforts to develop a vaccine.<br />
Lewis Acid Centers in Zeolites: On the Route to Mechanistic Understandings of Carbohydrate<br />
Conversions in Aqueous Media<br />
Arna Pálsdóttir<br />
Mentor: Mark E. Davis<br />
There is a growing emphasis on conversion of biomass to fuels and chemicals. Thus, reactions of carbohydrates in<br />
aqueous media are of importance. The Davis lab has discovered that sugars can be isomerized by zeolites with<br />
Lewis acid centers. The Lewis acid centers in the zeolite involve Sn or Ti atoms that are four coordinated with<br />
oxygen. The oxygens are either all coordinated to silicon (denoted the closed site) or three silicon atoms and a<br />
proton (denoted the open site). The open site has an adjacent silanol group. I investigated whether this silianol<br />
group is required for the reaction mechanism. The reason is that previously, the Davis group showed that this<br />
silianol group participates in oxidation reactions with Ti containing zeolites. Here, I analyzed Ti and Sn zeolites by<br />
IR methods to observe the presence of the silanol group and when the proton was substituted by a sodium ion.<br />
Upon sodium ion substitution, the Ti zeolites are not active for oxidation reactions. However, when erythrose is<br />
used as a model sugar, the erythrose is converted to erythrulose (Lewis acid mediated isomerization). Thus, the<br />
silanol group is not required for the isomerization reaction. Based upon these results, we are not able to distinguish<br />
whether the active center is the open or closed site or both.<br />
Monitoring Searching Behavior and Arousal in Drosophila melanogaster<br />
Ketaki Panse<br />
Mentor: David Anderson<br />
Arousal, a state characterized by increase in activity and sensitivity, is fundamental to many animal behaviors, but<br />
its neuronal mechanism is poorly understood. The purpose of this project is to study the neuronal mechanism of<br />
arousal in Drosophila melanogaster. To do so, we are using sugar to induce arousal by quantifying the effects of<br />
sucrose on the searching behavior of starved Drosophila melanogaster. Here, we show that when starved flies<br />
72
come in contact with water, they do not show local searching behavior. However, when the flies come in contact<br />
with sucrose solution, searching behavior increases, and their path consists of several turns and loops near the<br />
original food source. Similar behavior has been documented with C. elegans and blow flies, suggesting that there is<br />
a fundamental behavior common for several species. To ensure that this behavior is caused by activity of sugar<br />
neurons and not feeding, we will repeat the experiments by expressing Channelrhodopsin 2, a light sensitive cation<br />
channel, in sugar neurons to induce the behavior with light. To understand the neuronal mechanism and detect<br />
which neuromodulator is responsible for this behavior, we will perform screening for mutants and use calcium<br />
imaging.<br />
Intelligent Queue Scheduling for Robo-AO<br />
Athanasios Papadopoulos<br />
Mentors: Christoph Baranec and Reed Riddle<br />
This presentation discusses the design and implementation of the intelligent queue scheduling software for the<br />
Robo-AO adaptive optics system. Robo-AO will be the first robotic laser guide star adaptive optics system and will<br />
be able to deliver diffraction limited resolution in both the visible and the near-infrared spectrum. Due to the fact<br />
that Robo-AO is completely autonomous, a queue scheduler will be used to prioritize a set of scientific targets,<br />
making the choice of which target to observe after each observation is completed. The queue scheduling software<br />
will make the choice of its next observation by taking into account all the important parameters and input data for<br />
each target. These include the scientific importance, the current observing conditions and feedback about the<br />
current performance of the adaptive optics system. The logic of the optimization is, in essence, balancing the<br />
quality of the observations, the slewing time overhead and the scientific priority of a target.<br />
Atomic Force Microscopy Studies of the Effect of ATP on XPD Helicase Interactions With DNA<br />
Alison Parisian<br />
Mentors: Jacqueline Barton and Pamela Sontz<br />
DNA-mediated charge transport (DNA CT) is a proposed mechanism used by repair proteins to cooperatively search<br />
for lesions on DNA. Since DNA CT is significantly attenuated over a strand containing mismatches, proteins<br />
containing a redox-active [4Fe4S] cluster may use the strength of signal sent across a strand by another protein to<br />
determine the presence of a mismatch. XPD is a 5'-3' DNA helicase involved in the nucleotide excision repair<br />
pathway which contains a [4Fe4S] cluster and binds to ATP. It has been shown to redistribute onto mismatched<br />
strands, but it was unclear as to the effect of ATP on the protein's activity and redistribution. In our experiments,<br />
atomic force microscopy was used to visualize and quantify the binding of XPD to strands of varying lengths, some<br />
containing a mismatch, in the presence of ATP. It was found that over a 24 hour incubation with ATP the helicase<br />
activity of the protein was activated, overriding the CT mechanism, but without the incubation time, ATP has little<br />
effect on the redistribution of XPD. Future studies will focus on visualizing the helicase activity of XPD and its<br />
activation by ATP.<br />
Introduction of the Azide Chemical Reporter to Newly Synthesized Fatty Acids for Lipid Detection and<br />
Antibiotic Screening<br />
Sungjin Park<br />
Mentors: David A. Tirrell and Janek Szychowski<br />
Once thought to be simple in nature, lipids are now known to be incredibly diverse, and to play key roles in many<br />
biological processes. However, the lack of non-radioactive tools and methods make the study of lipids difficult. The<br />
modern “chemical reporter strategy,” which involves the labeling of biomolecules by chemical reporter groups, was<br />
implemented to effectively label newly synthesized lipids within cells. Vibrio harveyi was fed with N3(CH2)5CO2H, a<br />
short chain azido fatty acid, and analysis of click-reacted cellular lipid extract by HPLC-UV-MS confirmed the<br />
successful elongation and incorporation of N3(CH2)5CO2H by the cellular lipid biosynthetic machinery. In order to<br />
improve incorporation of the azide chemical reporter in lipids, the effects of concentration, pulse length, and<br />
introduction time point of N3(CH2)5CO2H on azido fatty acid incorporation were studied. The level of incorporation of<br />
the azide group was determined by measuring the fluorescence of cells treated with a coumarin-cyclooctyne<br />
conjugate. It was determined that cells introduced to 5 mM of N3(CH2)5CO2H at OD 0.3 and grown for 5-7 hours are<br />
able to incorporate the azide chemical reporter into their newly synthesized lipids most efficiently. Using these<br />
optimized conditions for incorporation of the azide group in lipids, fatty acylated proteins were detected by SDS-<br />
PAGE in-gel fluorescence using a TAMRA-cyclooctyne conjugate. In addition, the observation of the significant<br />
disruptive effect of cerulenin, a fatty acid biosynthesis inhibitor, on N3(CH2)5CO2H incorporation in cellular lipids was<br />
used to take steps towards developing a high-throughput cell-based antibiotic screen for identifying antibiotics<br />
which specifically interfere with fatty acid synthesis.<br />
73
Transcription Factor Sox10 May Be Required for Microvillous Neuron Formation in the Zebrafish<br />
Olfactory System<br />
Brian Peng<br />
Mentors: Marianne Bronner and Ankur Saxena<br />
The zebrafish olfactory system has three distinct types of sensory neurons: ciliated, crypt, and microvillous.<br />
Microvillous neurons are an important component of olfaction across species, but their origin has remained unclear.<br />
In the Bronner lab, previous observations have shown transgenic Sox10:eGFP-labeled cells, thought to be neural<br />
crest, migrating into the olfactory region and differentiating into microvillous olfactory neurons. To determine the<br />
relevance of Sox10 in microvillous neuron formation, the expression profile of Sox10 was analyzed. Embryos were<br />
examined at different stages of development for Sox10 protein expression and assayed for colocalization with a<br />
variety of proteins including TRPC2, which is a marker for microvillous neurons. Results indicate that Sox10 protein<br />
is indeed highly expressed in neural crest cells that differentiate into microvillous neurons. In addition, loss-offunction<br />
experiments were performed using morpholino injections to determine Sox10’s role in microvillous neuron<br />
formation. Analysis was done primarily with confocal microscopy; preliminary results indicate a requirement for<br />
Sox10 in the differentiation of neural crest into microvillous neurons. Future experiments will investigate Sox10<br />
mRNA expression in order to fully determine the spatiotemporal span of Sox10 induction during neural crest<br />
migration into the olfactory region.<br />
Siegel Modular Varieties: Arithmetic Invariants, Cusps, and Real Points<br />
Christopher Perez<br />
Mentor: Andrei Jorza<br />
Siegel modular forms are a generalization of elliptic modular forms to several complex variables. As in the case of<br />
elliptic modular forms, there is an associated modular variety known as a Siegel modular variety. We attempt to<br />
generalize the notion of modular symbols to Siegel modular varieties and use these to generate a basis for Siegel<br />
modular forms.<br />
Hot Dust in Bright Galaxies: Using Mid-Infrared Photometry to Determine the Sources of Extragalactic<br />
Radiation and Spectroscopy<br />
John Pharo<br />
Mentor: Lee Armus<br />
Ultraluminous Infrared Galaxies (ULIRGs) have some of the highest energy output in the known universe. This is<br />
usually the result of some combination of stellar light, widespread star formation, and the accretion of matter by a<br />
central, supermassive black hole. These galaxies, however, contain vast amounts of dust, which absorb much of<br />
the energy produced and re-emit it as infrared radiation. By measuring the flux through the galaxy at different<br />
wavelengths of the infrared spectrum, and then by fitting and decomposing the spectrum, one can determine the<br />
respective energy contributions of star formation and accreting black holes. Many known ULIRGs are missing data<br />
between 2 and 5 microns, which can have significant effects on the shape and composition of the fit, as emissions<br />
from hot dust occur in this region. The Spitzer Space Telescope recently took IRAC images of 86 such objects,<br />
allowing us to measure fluxes at 3.6 and 4.5 microns. We are currently using using these data together with<br />
archived spectra and photometry from other telescopes to fit the spectra and estimate the energy contribution of<br />
an active, central black hole.<br />
Elk3 Is Required for Neural Crest Specification<br />
Jacquelyn Phillips<br />
Mentors: Marianne Bronner and Crystal Rogers<br />
Elk3/Net/Sap2 (here referred to as Elk3) is involved in tumor angiogenesis in mice, but is expressed in various<br />
tissues throughout embryonic development. Elk3 expression was characterized using whole mount in situ<br />
hybridization, and we found that Elk3 expression is very dynamic. During development, Elk3 is expressed in the<br />
developing head folds, the head mesenchyme, the developing somites, and in the migratory cranial neural crest.<br />
Because Elk3 is expressed in premigratory and migratory neural crest cells, we believe that it is involved in cranial<br />
crest specification and migration and may be required for the formation of many neural crest derivatives. Neural<br />
crest cells migrate throughout embryos and become various types of cells, including neurons, enteric ganglia, and<br />
melanocytes. Based on the Elk3 expression pattern, we wanted to explore the effect of Elk3 knockdown on neural<br />
crest cell development. Here we show that loss of Elk3 protein decreased the number of neural crest cells marked<br />
by Sox10 expression, which suggests a role for Elk3 in neural crest specification.<br />
74
Elucidation of the Architecture of a Protein Aggregate: The Role of Binding Interactions and<br />
Intermolecular Packing Interactions Within the Aggregate<br />
Samantha Piszkiewicz<br />
Mentors: Shu-ou Shan and Thang Nguyen<br />
Protein homeostasis is essential for all cells and requires the proper control of the folding, localization, and<br />
interactions of all proteins. The misfolding and aggregation of proteins are detrimental to cells and have been found<br />
to be the root cause of numerous age-related diseases. To study the general mechanism of disaggregase of a<br />
typical hydrophobic membrane protein, we will use LHCP as a model system to define the nature of such<br />
aggregates. The Shan Lab has recently shown the robust disaggregase activity of plant protein, the 43 kDa<br />
chloroplast signal recognition particle (cpSRP43), for the disassembly of aggregates of the family of the multitransmembrane<br />
light harvesting chlrophylla/b –binding (LHC) protein. We have shown through alkylation<br />
experiments of single-cysteine constructs of the LHC protein that L18 is a dominant sequence within the LHC<br />
aggregate, meaning that it is always presented on the surface of the aggregate, poised for disaggregase<br />
recognition. It appears that the third transmembrane domain of LHC is most involved in intermolecular packing<br />
interactions in the aggregate.<br />
Developing and Optimizing Polymer-Actuated Microfluidic Valves for Portable Applications<br />
Ahalya Prabhakar<br />
Mentors: Michael Roukes and Blake Axelrod<br />
The field of microfluidics has become a focus of nanobiotechnology due to the potential for lab-bench based<br />
biological experiments on a microscale. From large-scale arrays to single-cell assays, microfluidics allows<br />
researchers to conduct the same experiments in the lab for significantly less reagent cost, space, and time. Many<br />
groups have been working on portable microfluidic systems, but have focused on passive, valveless control. We are<br />
working on developing a portable, active-control microfluidic system through the use of electroactive polymers. I<br />
have focused on optimizing the polymer growth, chip design, and efficient and consistent voltage delivery for the<br />
control of the valve actuation. We are working towards successful long-term actuation of independent valves, which<br />
will work reliably for practical use.<br />
Corneal Fibroblast Migration on Cross-Linked Gelatin Hydrogels With Nanofibrous Scaffolds<br />
Amy C. Proctor<br />
Mentors: Julie Kornfield and Amy Fu<br />
The cornea is the clear protective surface that covers the front of the eye and aids in focusing and refracting light.<br />
Corneal endothelial cells cannot be regenerated if they are destroyed by disease or trauma, and blindness can<br />
result from the loss of too many endothelial cells. Synthetic corneal prostheses have successfully been used to<br />
treat corneal injuries; however, epithelialization of the implant surface remains a challenge. Although<br />
epithelialization is not necessary to restore vision, epithelial cell coverage can increase retention rates and reduce<br />
the risk of infection after implantation. Here we investigate the in vitro cellular migration of rabbit corneal<br />
fibroblasts on cross-linked gelatin hydrogels covered with cross-linked gelatin nanofibers of various directionalities<br />
as a first step toward developing a protocol to promote epithelization in a clinical setting.<br />
Identification of Regulatory Genes in the Endodermal Gene Regulation Network in Strongylocentrotus<br />
purpuratus During Gastrulation<br />
Ralph Edward Pursifull<br />
Mentors: Eric H. Davidson and Isabelle Peter<br />
Hnf1aLike and Hnf4 are transcription factor genes that are expressed in the later stages of development, from 30<br />
hours post fertilization (hpf) to 72 hpf and are potential components of the endoderm Gene Regulatory network.<br />
This project aims to detect the temporal and spatial localization of the expression of the targeted genes, Hnf1aLike<br />
and Hnf4 through in-situ hybridization. Expression patterns indicated by the in-situ hybridization will indicate<br />
whether the changes in expression of the genes within the endoderm cause changes to the regulatory state of the<br />
endoderm region. If this is the case, the expression patterns can be analyzed alongside other transcription factors,<br />
indicating potential factors that alter the expression of the target genes, as well as genes that experience altered<br />
expression based on the presence of Hnf1aLike and Hnf4. This will serve as a basis for further work to establish<br />
relations between the target genes and other genes in the late endoderm Gene Regulation Network of sea urchins.<br />
Surface Imaging of Mixed Methyl/Thienyl Monolayers on Si(111)<br />
Suyeon Pyo<br />
Mentors: Nathan S. Lewis and Leslie O’Leary<br />
Methyl monolayers and mixed methyl/thienyl monolayers on Si(111) were synthesized through a two step<br />
chlorination/alkylation procedure. The terrace size and surface roughness were investigated using transmission<br />
fourier transform infrared spectroscopy (FTIR) and atomic force microscopy (AFM). The surface composition was<br />
characterized via X-ray photoelectron spectroscopy (XPS) and the distribution of functional groups on mixed<br />
75
monolayers was determined using scanning tunneling microscopy (STM). The mixed methyl/thienyl monolayers<br />
were composed of 5-10% of thienyl groups, 85-90% of methyl groups and few hydrogen groups. AFM images<br />
showed straight terraces of 100-200nm width which were parallel to each other. The statistical distribution of<br />
functional groups will give deeper insight into monolayer formation via halogenation/alkylation, reactivity of<br />
monolayers, efficiency of secondary functionalization and catalyst attachment.<br />
Regulation of Huntingtin and REST by the IκB Kinase Complex in Huntington’s Disease<br />
Mike Qian<br />
Mentors: Paul H. Patterson and Ali Khoshnan<br />
Expansion of a polyglutamine repeat at the N-terminus of the huntingtin (Htt) protein causes Huntington’s disease<br />
(HD), a neurodegenerative disorder characterized by progressive cognitive decline, chorea, and psychiatric<br />
disturbances. Although the molecular pathways that regulate the disease remain poorly understood, many of its<br />
deleterious effects in the brain have been attributed to aberrant nuclear accumulation of neurotoxic N-terminal<br />
fragments of mutant Htt. The zinc-finger protein RE1 silencing transcription factor (REST), a master regulator of<br />
neuronal differentiation, is normally sequestered in the cytoplasm by Htt but becomes activated in HD, causing<br />
neuronal genes to be incorrectly silenced. REST levels are elevated in HD animal and cellular models. Many studies<br />
in these models have also implicated the IκB kinase (IKK) complex in HD pathogenesis. Interestingly, the REST<br />
promoter contains several binding sites for the transcription factor NF-κB. Since IKKβ plays a key role in activating<br />
NF-κB, we began testing a panel of REST promoter constructs in a cellular model to determine if blocking NF-κB<br />
activity could abolish abnormal REST expression. Preliminary results indicate that mutating NF-κB binding sites in<br />
the REST promoter decreases promoter activity, thus suggesting a possible link between IKKβ and REST<br />
expression. Future experiments will focus on direct inhibitors of IKKβ, which should help to establish whether this<br />
represents a novel pathway in the neuropathology of HD that could pave the road toward a promising new<br />
therapeutic strategy.<br />
Testing Properties of Parylene-C With X-Ray Photoelectron Spectroscopy<br />
Chenxi Qiu<br />
Mentors: Yu-Chong Tai and Jeffrey Lin<br />
Parylene-C is a class of poly(p-xylylene) polymers widely used for many biomedical implantable devices due to its<br />
high biocompatibility. The main use of a parylene is through a vapor deposition polymerization (VDP) process<br />
developed by Gorham, where a thin film of parylene is developed around usually sensitive materials to prevent<br />
them from corroding in harsh environments like the human body. This project focuses on understanding what<br />
causes parylene to become brittle after being processed at 100C to 200C and discovering methods to prevent it.<br />
X-Ray Photoelectron Spectroscopy (XPS) is a machine that irradiates the surface of a material with X-rays while<br />
simultaneously measuring the kinetic energy and number of electrons that escape from the material. With the use<br />
of XPS, it has been determined that oxidation might not be one of the causes of the brittleness for parylene<br />
processed at 100C in air because no significant amount of oxygen was discovered. Further work will involve finding<br />
a curve over temperature and time for the densification of parylene and verifying the possibility of crystallization as<br />
a cause for the brittleness.<br />
Stochastic Frequency Modulation of Bacillus subtilis Stress Response<br />
Sofia Quinodoz<br />
Mentors: Michael Elowitz and James Locke<br />
Genetic networks employ dynamic strategies to respond to different conditions. For example, it was recently shown<br />
that cells can activate gene expression in “pulses” under stress conditions. In particular, the sigB alternative sigma<br />
factor in B. subtilis exhibits discrete pulses of activation whose frequency is modulated by external stresses. To<br />
follow up on this discovery, I examined the dynamics of twelve other B. subtilis alternative sigma factors using<br />
quantitative single cell time-lapse microscopy. Alternative sigma factors control key regulatory pathways in<br />
bacteria, including chemotaxis response and antibiotic resistance. My investigations are providing a<br />
characterization of the dynamics of induction of these diverse alternative sigma factors and the dependence of<br />
these dynamics on the magnitude of activating inputs for each factor.<br />
A Study of the Binding Affinities of the Family I Carbohydrate-Binding Modules (CBM) and the<br />
Characterization of the CBM Chimeras Constructed Using SCHEMA Recombination<br />
Misha Raffiee<br />
Mentors: Frances H. Arnold and Indira Wu<br />
The skyrocketing cost of fossil fuels has given rise to the need to reduce America’s dependence on petroleum. By<br />
converting cellulosic biomass to glucose, and in turn, converting the glucose to biofuels, it is possible to reduce our<br />
dependence on petroleum-based fuels. Cellulose is known to be degraded by certain enzymes produced by specific<br />
bacterial and fungal organisms. Trichoderma reesi, a filamentous fungus with this ability, contains an array of<br />
cellulases which act as catalysts to the hydrolysis of the insoluble crystalline cellulose, which yields glucose as a<br />
product. Cellobiohydrolase, one such cellulase, has a unique structure consisting of a cellulose binding module<br />
76
(CBM) and a catalytic domain coupled by a linker region. Using overlap extension PCR, seven different CBMs were<br />
successfully fused with a previously engineered catalytic domain. The binding curves of the CBMs were examined in<br />
an effort to engineer and improve the binding affinities of the cellobiohydrolases. This research also utilized<br />
SCHEMA recombination to identify specific beneficial amino acid blocks in the CBMs. The goal of the study is to<br />
improve the binding affinity of the Family I CBMs, and allow for the successful and efficient conversion of biomass<br />
to biofuel.<br />
Automated Information Retrieval System (AIRS): An Economic Solution to Data Recovery From High<br />
Altitude Balloons<br />
Anthony Ransford<br />
Mentors: Jason Rhodes and Jeffery Booth<br />
Large amounts of data from high altitude balloons can be accrued during astronomic research such as the High<br />
Altitude Lensing Observatory (HALO). By confining all data to a balloon until mission termination, total data loss is<br />
risked in the event of undesirable landings. Conventional data transmissions are not feasible for large data rate<br />
missions due to low-data rate transmission capabilities. AIRS resolves these issues by deploying low-cost gliding<br />
aircraft, containing solid state drives with mission data, from the balloon. The AIRS vehicles autonomously guide<br />
themselves to within 300 ft. of designated GPS coordinates, automatically compensating for winds during descent.<br />
Initially velocity and lift predictions were made to determine the flight characteristics of different designs. The AIRS<br />
design was chosen and constructed from glass-fiber materials. A previously developed hobby autopilot was<br />
modified and used for the project needs. Low and high altitude, impact and velocity testing were performed with<br />
the vehicle. This project has showed it is feasible to build a safe, reliable and low cost data retrieval system that<br />
can be easily located.<br />
Accelerating Neuronal Model Simulations Using GPGPU Techniques<br />
Ben Razon<br />
Mentors: Alon Korngreen, Thomas Miller, and Roy Ben-Shalom<br />
Neuronal models are a critical tool that empower scientists to translate experimental results into knowledge of the<br />
underlying biophysical mechanisms. A common method involves optimizing kinetic parameters of the ion channels<br />
to match currents found in voltage-clamp protocol experiments. Both the genetic algorithm and model used in this<br />
technique are highly compute intensive, with execution times typically lasting several days even on large CPU<br />
clusters. By parallelizing these processes on graphics cards using the CUDA programming language, large<br />
performance gains can be realized. To fully accelerate these solutions a variety of optimization techniques such as<br />
utilizing shared memory, coalescing global memory accesses, and reducing memory transfers were implemented.<br />
Furthermore, the affordability of modern GPUs makes these solutions widely accessible, bringing super-computing<br />
capabilities to a new audience.<br />
Learning Representations of Astronomical Data With Missing Observations and Uneven Sampling<br />
Colorado Reed<br />
Mentors: Umaa Rebbapragada, David Thompson, and Kiri Wagstaff<br />
Radio telescopes are progressively observing candidate astronomical events at rates that are too high for human<br />
experts to visually inspect and analyze. As a result, automated statistical techniques for event detection and<br />
analysis are becoming commonplace for modern observatories. These methods often require the temporal data<br />
(i.e. light curves) to be robust and evenly sampled, while actual observational data usually contains missing or<br />
unevenly sampled data. Typical solutions include interpolating missing data (usually via linear interpolation) and<br />
marginalization (removing sections of data that are not robust). Missing values can be physically meaningful in<br />
astronomical data; however, and simply ignoring or interpolating this data may discard valuable information. In<br />
this work we present a novel technique for using contextual information to reconstruct probability distributions of<br />
missing values. The benefits of this technique are twofold: (1) using contextual observational data to impute<br />
missing data can outperform standard interpolation techniques, (2) the inferred probability distributions of the<br />
missing data provide an estimate of the uncertainty in the data representation that can be used in further analysis<br />
of the data. We demonstrate the effectiveness and generality of our technique with time-series data from the<br />
Green Bank Interferometer and the Very Long Baseline Array.<br />
Calibration of Mechanical/Electronic Platforms for Deep UV Imaging of Microbes on Natural Samples<br />
David Reid<br />
Mentors: Rohit Bhartia, William Abbey, and Luther Beegle<br />
The objective of my project is to help prepare the deep UV team for two major field excursions where the goal is to<br />
detect microbes and their distribution in the natural environment. One field sample will be basalt cores from the<br />
deep biosphere. For this I am reassembling and calibrating the current ultraviolet fluorescence scanner to reduce<br />
the need to refocus and to prevent sample contamination. The other field trip is to map microbial distributions on<br />
rocks in the Mojave. For this I will be calibrating a stand-off ultraviolet fluorescent scanners mounted on motorized<br />
tip/tilt. The tip/tilt will enable scans over 0.5x0.5m. This requires calibrations that enable assessment of the<br />
77
distance to the rock and determining the resolution capability of the instrument as well as the reproducibility to<br />
retarget a spot. With calibration of these instruments in the, the technology that has been demonstrated in the lab<br />
to non-invasively detect microbes in-situ can be successfully tested in the field.<br />
Algebraic Hecke Characters Which Are Rational Over Q<br />
Stephanie Reyes<br />
Mentor: Matthias Flach<br />
The motivation for this project comes from papers by Schütt (CM newforms with rational coefficients), Murabayashi<br />
(Determination of simple CM abelian surfaces defined over Q) and Anderson (Cyclotomy and a covering of the<br />
Taniyama group). The purpose of this project is to determine for which in Murabayashi’s list of degree four abelian<br />
fields there exists weight one or two Jacobi sum Hecke characters, to possibly find new degree four abelian fields<br />
for which there exists weight two Jacobi sum Hecke characters and to check if certain easily constructed weight two<br />
Hecke characters of imaginary quadratic fields are Jacobi sum Hecke characters.<br />
Precision Measurement of Growth Rates and Attachment Kinetics in Ice Crystals<br />
Mark E. Rickerby<br />
Mentor: Kenneth Libbrecht<br />
The basal growth of ice crystals between -2 o C and -40 o C was examined by using a purposely built convection<br />
chamber equipped with optical instruments. The obtained data was found to be in agreement with the model of<br />
nucleation limited growth. From this, the critical supersaturation σ0 was determined for a large number of crystals.<br />
This showed a very small nucleation barrier for the basal facet in the region -6 o C to -10 o C. An explanation of<br />
these findings by the effects of ice surface melting at temperatures below freezing point is presented. These results<br />
are consistent with the snow morphology diagram observed in nature that has remained unaccounted for since it<br />
was first proposed. Ideas on how to further develop the apparatus in order to also measure the radial growth are<br />
suggested.<br />
Paleoclimate Insights From the U-Th Isotope Geochemistry of the Buda Cave System<br />
Alexander D. Rider<br />
Mentors: Jess Adkins and Adam Subas<br />
The chemical and physical information preserved within speleothems (stalagmites and stalactites) can be an<br />
important tool for understanding past climates. However, this information is only useful if there is an understanding<br />
of the relationship between climate, speleothems, and the cave systems in which speleothems grow. Our goal is to<br />
incorporate the isotopes of the uranium decay series, often used for dating speleothems, into this understanding<br />
for the Buda Cave System. Using Caltech’s “Neptune” MC-ICP-MS we have made highly precise uranium series<br />
measurements of karst collected from the surface of the cave system, and drip waters collected from within. To<br />
complement these uranium series measurements we have also performed a broad-spectrum elemental analysis of<br />
the same samples using an Agilent 7500 series ICP-MS. This approach allows us to understand how the chemistry<br />
of drip water-rock interaction relates to the isotopes of the uranium series. We can then use our understanding of<br />
the uranium series geochemistry of the cave system to shed light on the record preserved within previously<br />
collected and analyzed stalagmite samples.<br />
Hot Flares From Cool Stars<br />
Joanna Robaszewski<br />
Mentors: George Djorgovski and Andrew Drake<br />
The Catalina Real-Time Transient Survey is a synoptic sky survey that observes variable objects across the entire<br />
sky and publishes information on them directly after observation. Many of these objects are flare stars, young stars<br />
that produce larges stellar flares. These flares can increase a star’s brightness by an order of magnitude and often<br />
only last for a few minutes. By examining the light cures of known flare stars, criteria to classify a light curve as<br />
belonging to a flare star or not can be determined. Once these parameters have been decided, they can then be<br />
applied to other transient objects that the Catalina Real-Time Transient Survey has discovered. After the objects<br />
have been classified, the frequency with which flare stars appear in our galaxy can be measured. Examining<br />
common characteristics in flare star light curves and looking correlations with spectral type can help us learn about<br />
the nature of stellar flares, as well as help us classify flare stars for future synoptic sky surveys.<br />
Error Quantification in Simulations of Variable Density Low Mach Number Turbulent Flows<br />
Nicholas Robertson<br />
Mentor: Guillaume Blanquart<br />
In numerical simulations of low Mach number flows, the density is often expressed as a function of one or several<br />
transported scalar quantities through the use of an equation of state (EOS). As a result, these scalars play a critical<br />
role in controlling the accuracy of the overall simulation. This study aims at understanding and quantifying the<br />
sources of errors introduced in performing simulations of variable density low Mach number laminar and turbulent<br />
78
flows. As a first step, the relative importance of the order of accuracy of the scalar transport scheme and the<br />
velocity scheme is analyzed in canonical variable density test cases. Then, various implementations of the EOS are<br />
investigated in laminar flows. Particular importance is placed on the robustness, the convergence, and the accuracy<br />
of these implementations. Finally, the different scalar transport schemes and implementations of the EOS are<br />
combined and evaluated in simulations of turbulent planar jets. Following this analysis, guidelines for performing<br />
accurate and robust simulations of variable density low Mach number laminar/turbulent flows are proposed.<br />
Using Uncertainty Quantification to Maximize the Efficiency of Parallel Resources<br />
Andrew C. Rodriguez<br />
Mentor: Mike McKerns<br />
When dealing with any batch computation across parallel resources, a common problem is encountered: the<br />
computation time for the batch is extended by calculations that take longer than the rest. More so, we find that<br />
with parallel computation for the process to continue, we must allow for the final computations to finish. Though<br />
each calculation within the batch may not take much time individually, compounded, we find that, overall, a lot of<br />
time is taken waiting for the final few computations. We can reduce overall computation time through use of<br />
completion criteria. By deeming a batch “complete” when certain criteria are met, we find that we both make a<br />
better use of resources and save time. And so, we are creating a “Completion” data structure that shall manage<br />
such features for use within mapping functions. Criteria shall include, but not be limited to, a simple “percent<br />
completion” condition and use of statistical methods to terminate based on the confidence of the collected results.<br />
This would increase efficiency and save overall computation time. This is important since even a simple 10% time<br />
reduction allows for us to save 2.4 hours on a computation that would normally take 24 hours.<br />
Creating a Peptide Capture Agent Against IgG Fc<br />
Errika C. Romero<br />
Mentors: Jim Heath and Jessica Pfeilsticker<br />
Inexpensive and robust HIV tests are needed for effective point-of-care diagnostics in many developing countries.<br />
Current HIV diagnostic assays employ monoclonal antibodies, which are expensive and inconsistent. Instead,<br />
peptides, which are low-cost and whose chemical properties are easily tuned, can be designed to bind with<br />
comparable affinity and selectivity to anti-HIV antibodies similar to monoclonal antibodies that were raised against<br />
the same target. Previous research has identified a 13 residue cyclic peptide, monikered Wells’ peptide, which<br />
binds to the IgG Fc region of HIV antibodies. There are two binding pockets on every Fc molecule, so we created a<br />
dimer of Wells’ peptide using standard amide coupling to interact with both pockets simultaneously. Affinity and<br />
selectivity assays are performed to determine the ideal dimer linker length and to measure the dimer’s efficacy in<br />
binding to IgG Fc.<br />
Investigating the Expression Patterns of Potential Regulators and Homologues of FoxD3<br />
Shourya Sonkar Roy Burman<br />
Mentors: Marianne Bronner and Marcos Simões-Costa<br />
Neural crest cells are a population of transient migratory cells unique to the vertebrate embryos, which give rise to<br />
a wide variety of derivatives like the peripheral nervous system and the facial skeleton. They serve as a model to<br />
study multipotency, epithelial to mesenchymal transition, cell migration, embryonic induction and cell-fate<br />
determination. Of specific interest to this project is a transcriptional repressor in the gene regulatory network<br />
governing these processes, FoxD3. FoxD3 is expressed early in the presumptive neural crest region, and changes<br />
the adhesion properties of these cells leading to their epithelial to mesenchymal transition and migration. The first<br />
part of this project involves identification of potential regulators of FoxD3 in the chicken embryo using a candidate<br />
gene approach. A set of transcription factors was identified based on binding site studies on an enhancer element<br />
of FoxD3. The expression patterns of these genes were compared to the region where the enhancer is activated,<br />
and a refined list of candidates was prepared. In the second part of the project, differences in the formation of<br />
neural crest between an early vertebrate, the lamprey and other vertebrates was explored by studying the<br />
expression pattern of FoxD3 and a related homologue, FoxD1.<br />
Keck Spectroscopy of X-Ray Sources in the Ssa22 Field<br />
Iva Rreza<br />
Mentors: Fiona Harrison and Dan Stern<br />
We present the results of a deep optical spectroscopic survey of X-ray sources in the SSA22 field. After the Great<br />
Observatories Origins Deep Survey (GOODS), SSA22 is the next deepest extragalactic field observed by the<br />
Chandra X-ray Observatory, with a total integration time of XXX~ks. We observed six slitmasks with the Keck<br />
telescopes, obtaining 90 new high quality redshifts. We report on the properties of these sources, including their<br />
redshift distribution, absorbing column distribution, and the range of optical properties for the X-ray host galaxies.<br />
We compare our results to the published literature, and discuss several interesting sources in more detail.<br />
79
Sunk Costs and Reference-Dependent Preferences<br />
Alexander Manek Karl Runkel<br />
Mentors: Shinsuke Shimojo and Vikram Chib<br />
Human decision-making is influenced by expectations and aversions of loss, ambiguity and risk. A reference point<br />
represents a person’s rational expectations about past events, determined in an endogenous equilibrium. In a<br />
deterministic setting humans tend to maximize the utility derived from their actions – but in case of uncertainty,<br />
decisions are influenced by aversions as proposed by the Prospect Theory. In order to understand how these<br />
expectations are induced and how they influence human behavior, we performed several fMRI experiments to<br />
determine, firstly, how real effort is encoded as value; secondly, how this value is transcribed into a point of<br />
reference and thirdly how an established reference point shifts if uncertainties or risks are introduced. We trained<br />
subjects with visual cues associated with different levels of effort and a separate visual stimulus upon successful<br />
completion of the specific effort task. We found that subjects prefer the low-effort cues but paradoxically are more<br />
partial to the success-stimuli associated with the higher effort and a lower average payoff yield – thus indicating<br />
that humans are emotionally invested in incurred costs that cannot be recovered. In economic terms we can state<br />
that humans are attached to sunk costs and that the classical consumption utility function has to be refined by<br />
introducing a model of gain-loss-utility.<br />
C. elegans Lethargus and Molting Behavior<br />
Elizabeth Ryan<br />
Mentors: Paul Sternberg and Julie Cho<br />
Lethargus is a sleep-like condition exhibited by the nematode C. elegans between successive developmental<br />
stages. During lethargus, the nematode enters a state of quiescence in which locomotive activity and feeding<br />
behavior decrease. Subsequently, the nematode molts its exterior cuticle, signaling the end of lethargus. To<br />
understand the relationship between availability of food and its effect on behavioral states such as lethargus and<br />
molting, we placed worms in their fourth larval stage, known as L4, inside a microfluidic tracking chip with and<br />
without OP50 E. coli available for sustenance. Although the time taken for C. elegans nematodes to shed their<br />
cuticle in the liquid environment of a microfluidic chip is comparable to the time taken on an agar plate, results<br />
indicate that molting lasts longer when the worms are starving. Additionally, we want to use this sleep-like<br />
behavior to test the effects of sleep-altering drugs. Therefore, we are exploring the effect on lethargus when L4<br />
worms were exposed to the stimulant dipropylcyclopentylxanthine (DPCPX). Using tracking software, worms placed<br />
in a rectangular field of view are monitored during their transition into lethargus. DPCPX works as an antagonist for<br />
the adenosine A1 receptor and has been shown to cause wakefulness in Danio rerio (zebrafish). However,<br />
because C. elegans pores are more resistant to chemical penetration, dimethyl sulfoxide and Triton X-100 are used<br />
to carry compounds across the nematode membrane, rendering it more permeable. Findings may provide a<br />
connection between C. elegans sleep-like state and that of other species in the Animalia kingdom due to similarities<br />
in biological systems.<br />
Carbon Nanotube Arrays for Drug Delivery and Transport Studies<br />
Stephanie Rae P. Samson<br />
Mentors: Morteza Gharib and Adrianus Aria<br />
Carbon nanotubes, rolled seamless graphene sheets, can be functionalized non-covalently or covalently with<br />
different chemical groups to modify or add to their already unique intrinsic properties. Pristine carbon nanotubes<br />
are generally unreactive, but by taking advantage of existing defects or by introducing defects without destroying<br />
the nanotubes’ structures via oxidation with UV/Ozone or Oxygen/Plasma treatments, it is possible to add carboxyl,<br />
carbonyl, and hydroxyl groups on the carbon nanotube surface. The resulting oxygen-containing groups can be<br />
utilized to covalently add functionality to the carbon nanotubes. The ultimate goal is to create an ester linkage<br />
between vertically aligned carbon nanotube arrays (nanocarpets) and large molecules, such as proteins, drugs, and<br />
nanoparticles. To test the reactivity of oxidized nanocarpets, they are subjected to esterifications with nitrogencontaining<br />
compounds using Fischer esterification, Steglich esterification, and esterification after conversion of<br />
carboxyl groups to acid chlorides. The reacting nanocarpets will then be analyzed using Energy-Dispersive X-ray<br />
Spectroscopy (EDS), Raman Spectroscopy, and Fourier Transform Infrared Spectroscopy (FTIR) to determine the<br />
resulting nitrogen content and the integrity of the carbon nanotubes. If those results show promise, then those<br />
esterifications will be repeated using larger molecules.<br />
Producing a Lab-on-a-Chip Microfluidics Device for Speedy Cell Diagnostics<br />
Troy Sandberg<br />
Mentor: James Heath<br />
The main advantage a microfluidics device has over its larger counterparts is the small volume of materials it<br />
requires and the speed at which it is able to analyze these small samples. A hospital patient could supply a single<br />
drop of blood from which the lab-on-a-chip could determine the various cells therein, checking for metabolic signs<br />
indicative of metastasizing cancer cells. This project focuses on fine-tuning the production of such a chip. Once a<br />
particular design is decided on and laser-etched into a silicon wafer, a silicone polymer is poured over the wafer<br />
80
and baked so that it hardens in the desired shape and surface geometry of the chip. This chip has multiple<br />
separated chambers which can be coated with protein substrates to which cells will adhere. Collaborating with a lab<br />
at UCLA, many different coatings and cell lines have been paired together on the chip to find the combination that<br />
provides optimal cell adherence. With the cells adhered to the chip, medium containing radiolabeled probes can be<br />
flowed through the chip’s chambers and their uptake by the cells quantified.<br />
An in vitro Study of the Hemodynamics of Valvulogenesis<br />
Laura F. Santoso<br />
Mentors: Morteza Gharib and Derek Rinderknecht<br />
Heart disease is a significant problem worldwide, and people often undergo valve replacement surgery to treat<br />
their condition. Valvulogenesis is being studied to improve upon existing replacement valves, which are currently<br />
produced from synthetic materials or animal tissue, have limited lifetimes, and require additional medication.<br />
Ample research has shown that shear forces from blood flow affect how cells grow. The shear stress forces caused<br />
by various flow rates have been measured in this research by particle image velocimetry. Regulation of shear<br />
forces allowed emulation of the flow conditions present during valvulogenesis, under which endothelial cell<br />
proliferation was observed. Additional studies from previous research have provided strong evidence that extra<br />
expression of Krueppel-like factor 2 (KLF2) is indicative of valvulogenesis. Studying expression of KLF2 ensured<br />
that the prolonged fluid shear stress applied was recognized by the observed cells. Subsequent studies analyzing<br />
proliferation characteristics under pulsatile flows would further this research.<br />
Design and Modeling of an Autoinhibiting Biomolecular Circuit Incorporating Scaffolding<br />
Keshav Sapatnekar<br />
Mentor: Richard Murray<br />
In nature, protein scaffolds allow processes to proceed more efficiently by localizing signaling proteins and their<br />
downstream elements, thus decreasing the time for signals to propagate. This could be used to create protein-level<br />
feedback systems utilizing scaffolds. In this project, a circuit was designed incorporating a histidine kinase, a<br />
response regulator and two scaffold elements bound to target proteins. One scaffold binds the kinase to the<br />
regulator, thereby activating production of the second scaffold. This second protein would bind the first, preventing<br />
its further production, ideally giving an equal concentration of the target proteins at steady state. An ordinary<br />
differential equation model was used to describe the behavior of the system, incorporating transcription,<br />
translation, and binding of proteins and scaffolds. Future investigation would include implementation of the system<br />
in E. coli using fluorescent target proteins and a plate reader to determine concentrations.<br />
3D Scientific Data Visualization in an Immersive Virtual Reality<br />
Franz Sauer<br />
Mentor: George Djorgovski<br />
As society moves into an age where scientific data becomes more complex, humans need more powerful<br />
visualization tools that would not only need to revolutionize the way we perceive data, but also provide effective<br />
ways for scientists to collaborate with one another. The Meta-Institute for Computational Astrophysics (MICA) was<br />
therefore created to explore, develop, and promote virtual worlds as a means of professional research. This specific<br />
project focused on developing and testing new data visualization tools in immersive virtual reality environments in<br />
order to gain insight into what methods work better than others in terms of understanding and collaboration.<br />
Currently set up in ScienceSim, a virtual world run by Intel, this project is developing a number of tools such as a<br />
multi-dimensional plotting device and viewing area, an in world method to format and scale data files, a<br />
collaborative tool that links data points to more information located on web databases, and an in world method to<br />
display online images to a group of researchers. As this is still and ongoing project many more tools will be<br />
developed with the eventual goal of developing an efficient virtual world as a means of scientific research<br />
collaboration and visualization.<br />
New Design for a Vertical Wind Turbine<br />
Brad Saund<br />
Mentor: John Dabiri<br />
Recent analysis and experiments demonstrate that carefully placed vertical axis wind turbines (VAWTs) have the<br />
potential for an order of magnitude increase in the power produced per land area compared to horizontal axis wind<br />
turbines. By placing a VAWT adjacent to a counter-rotating VAWT the efficiency of each turbine can be improved to<br />
above the efficiency of an isolated turbine. Currently commercially available vertical wind turbines are used at the<br />
field site for these experiments. Interest in designing and constructing a new wind turbine for this site stems from<br />
the desire for greater control over the experimental setup, increasing the beneficial effects of the turbine<br />
placement, and lowering the cost the turbines. A proposed design was analyzed, models of the design and key<br />
components were tested, and the turbine was constructed.<br />
81
Spectral Gap Scaling of One Dimensional Quantum Spin Chains<br />
Travis Scholten<br />
Mentors: John Preskill and Spiros Michalakis<br />
Frustration-free Hamiltonians which are the sum of two-qubit projection operators have been considered as models<br />
for adiabatic quantum computation. The run time of such computation is determined by the scaling of the spectral<br />
gap with system size; in general, this property is not well understood. We show that the number of free<br />
parameters in any such Hamiltonian may be reduced to three real parameters and consider the decay of the gap<br />
with respect to the number of qubits. We present numerical results from a DMRG algorithm. For certain values of<br />
the real parameters, the scaling is shown to be an inverse polynomial of the system size.<br />
Boy Meets Boy: A Study of Male-Male Aggression in Drosophila melanogaster<br />
Jonathan Schor<br />
Mentors: David Anderson and Liming Wang<br />
In Drosophila melanogaster, the fruit fly, the nonvolatile cuticular hydrocarbon (CH) (Z)-7-tricosene (7-T) regulates<br />
aggressive behavior in an independent and inverse fashion to courtship. Given that 7-T is a gustatory pheromone,<br />
it seems likely that its primary mode of transmission is through physical contact with another fly’s gustatory<br />
sensilla, though the kinetics of 7-T are yet unknown. Through use of a gal4/gal80-UAS system, pairs of oenocyteminus<br />
(oe-) flies lacking in all CHs, including 7-T, were observed in parallel with wild-type Canton S (CS) flies for<br />
their aggressive behavior. Films of the interactions were annotated by hand, though in the future annotation will be<br />
performed through use of a machine-learning algorithm currently under development. Preliminary results indicate<br />
that physical contact precedes aggressive behavior, before which male-male courtship is observed. Future<br />
experimentation will examine why CS flies raised in groups (group-housed flies) are less aggressive than those that<br />
are single-housed, with a focus on the frequency at which they touch and their sensitivity to 7-T.<br />
Improving Temperature Stability of Himar1 Transposase<br />
Stanford Schor<br />
Mentor: Frances H. Arnold<br />
Thermophilic microbes and the heat stable enzymes they produce have many potential applications in<br />
biotechnology. However, the lack of available genetic tools for thermophiles has hindered their widespread use in<br />
industry. High-density mutagenesis is an efficient method for probing the biology of such poorly characterized<br />
organisms. Himar1 is a mariner-family transposase isolated from the horn fly Haematobia irritans. It inserts a short<br />
inverted terminal repeat (ITR) between thymine and adenine nucleotides randomly throughout the genome of its<br />
host and requires no external factors, making it a versatile tool for disrupting gene function. Wild-type Himar1 does<br />
not function at temperatures above 40�C, and we propose to use directed evolution to shift the optimal<br />
temperature range of Himar1 and generate a set of mutants for use in thermophilic organisms. We will randomly<br />
introduce point mutations into Himar1 using error-prone PCR, and then test mutants for activity at higher<br />
temperatures using a novel in vitro transposition assay. Stabilized, functional mutants will be subjected to iterative<br />
rounds of error-prone PCR and activity testing at higher temperatures in order to produce a family of mutants with<br />
functionality well above that of the wild type enzyme.<br />
Separating Planck Bolometers and Beams via Simulated Planet Observations<br />
Kasey W. Schultz<br />
Mentors: Sunil Golwala and Brendan Crill<br />
For any Cosmic Microwave Background (CMB) observation, in order to extract the full cosmological information the<br />
telescope beam must be accurately modeled, and the data must be corrected for the beam’s filtering. The CMB<br />
satellite Planck scans the full sky by rotating on its spin axis at a constant rate. This causes a degeneracy between<br />
the detector time response and the beam, and necessitates detailed modeling to identify relationships between<br />
their respective parameters. This modeling allows their effects to be separated, and produces data that has been<br />
properly corrected. Utilizing and updating the Planck team’s in-place software pipeline, I have run planet<br />
observation simulations and detector and beam model reconstruction. Through identifying possible bias in the<br />
fitting routines and any degeneracy between parameters, my work provides results that will help separate these<br />
effects as much as possible. By analyzing best fit parameters of over 3000 simulations and fits, I have shown that<br />
the pipeline reconstructs the input parameters without bias, that proper planet motion introduces no bias, and<br />
hope to show that prior knowledge of the beam does not bias the fitting routines or parameters; an important<br />
result for analyzing real Planck data.<br />
82
Analysis of Energy Estimation From NOvA Detector Data<br />
James Scott<br />
Mentor: Ryan Patterson<br />
Determining the energy of a neutrino accurately is important for studying neutrino oscillations. The detector for<br />
NOvA, which is made of many cells of scintillator, is able to detect the outgoing particles from a neutrino<br />
interaction, which can then be used to reconstruct the event and thus the neutrino energy. Any reconstruction will<br />
have uncertainties introduced by limits in the resolution of the detector, as well as random variations and noise.<br />
Knowing the magnitude of the uncertainties is important not only for assigning uncertainty in measurements, but<br />
also in guiding detector calibration approaches and in determining the theoretical and practical limitations of the<br />
oscillation measurements. By analyzing simulated detector data the uncertainties can be determined for the<br />
various interactions.<br />
Fourier Continuation Method for Eigenvalue Problems on Arbitrary Two-Dimensional Domains<br />
Ayon Sen<br />
Mentors: Oscar Bruno and Timothy Elling<br />
The eigenvalues of the 2D Laplacian correspond to the fundamental modes of vibration of a membrane with given<br />
geometry. These eigenvalues can be determined by using an iterative eigenvalue solver on the numerical<br />
differential operator. To evaluate the operator, we propose a method for calculating numerical derivatives on<br />
arbitrary domains by decomposing the geometry into overlapping patches that can be mapped to rectangles and<br />
then applying Fourier series techniques to compute derivatives under the change of variables. The Fourier<br />
continuation method extends non-periodic data so that the periodic extension is smooth, meaning Fourier series<br />
can be used to accurately represent the extension. Thus, derivatives can be computed accurately and quickly (i.e.<br />
using fast Fourier transforms). Extending this methodology to differential operators with variable coefficients<br />
(corresponding to inhomogeneous membranes) is simple, and test cases suggest that eigenvalues can be<br />
computed accurately for various geometries.<br />
Fabrication of Silicon-Based Nanopore Membranes for DNA Nucleotide Characterization<br />
Bhargav Setlur<br />
Mentors: Axel Scherer and Aditya Rajagopal<br />
In this ongoing project, we focus on the design and fabrication of a silicon-based nanopore for use in single<br />
stranded DNA sequencing. We start by defining the nanopore geometry using a multiphysics modeling program,<br />
which automatically calculates various electrical and capacitive parameters relating to the structure. Then, based<br />
on the idealized design, we demonstrate fabrication of the silicon nanopore using nanofabrication equipment<br />
located in the Kavli Nanoscience Institute in the sub-basement of Steele Building. Fabrication processes involve<br />
boring nanoscale holes in silicon using a transmission electron microscope to bore holes in silicon, as well as an<br />
SEM to beam-write patterns. Finally we fit the nanopore structures with electrical contacts and integrate them into<br />
the structure.<br />
Functionalization and Biological Application of Carbon Nanotubes<br />
Juying Shang<br />
Mentor: Mory Gharib<br />
Carbon nanotubes (CNTs) are allotropes of carbon with a cylindrical nanostructure. Besides exhibiting extraordinary<br />
strength and unique electrical and thermal properties, functionalized CNTs are useful for a multitude of biological<br />
and chemical applications. For example, functionalized CNTs can be fashioned into new scaffolds for cells,<br />
supporting three-dimensional tissue formation. In this project, CNTs are grown using the chemical vapor deposition<br />
process and functionalized using the UV/O3 treatment. SEM/EDS and FTIR analysis were conducted before and<br />
after treatment to determine the effects of UV/O3 exposure. It was found that the UV/O3 treatment temporarily<br />
adds carboxyl groups to the surface of CNTs; the effects disappear after 4 days. In order to understand cell<br />
behavior on CNTs, proliferation and differentiation assays will be conducted with human mesenchymal stem cells<br />
(hMSCs) grown on CNTs. In addition, CNTs’ ability to absorb proteins will be determined. The same experiments<br />
will be performed using functionalized CNTs. From the combined results of cell study and functionalization<br />
characterization, cell adhesion to and affinity for CNTs can be evaluated, illustrating the possibilities of biological<br />
applications of carbon nanotubes.<br />
Classifying Things That Go *BANG!* in the Night<br />
Nihar Sharma<br />
Mentors: S.G. Djorgovski and Ciro Donalek<br />
Digital synoptic sky surveys promise to shed a whole new light on existing as well as previously unknown timevariable<br />
astronomical phenomena. The challenge is to automate the classification and prioritization of transients for<br />
follow-up observations within the data that we receive in such a way that minimizes the need to use expensive<br />
spectroscopic resources. Due to an inevitably large amount of missing input data, a Bayesian Network (BN) was<br />
83
initially used to generate probabilistically inferred results that placed transients into different classes by computing<br />
the distribution of a given subset of variables where other variables were observed. Effort is being directed to<br />
incorporate continuous nodes into the existing BN, so that the need for artificial binning of continuous data can be<br />
eliminated, avoiding loss of information that is currently inherent in the classifier. An ensemble of learners was<br />
then constructed by prototyping Decision Trees as well as Multi-Layer Perceptrons and integrating it into the overall<br />
classification framework. This framework was rigorously tested on parameters extracted from time-series data of<br />
light curves so that it could potentially be employed in the real-time variable classification pipeline.<br />
Branching Messaging for Anonymous Communication<br />
Isaac C. Sheff<br />
Mentor: Tracey Ho<br />
We consider anonymous communication between pairs of nodes in the presence of an adversary who can observe<br />
all network traffic. Existing schemes involve partially trusted central servers, a large amount of cover traffic, or<br />
high latency. We propose a scheme which improves upon the trade-off, in which messages require only logarithmic<br />
time to deliver, as well as computational time to send. In this scheme, any node receiving a message applies a<br />
private decryption key to discover content, or instructions to forward the message to zero or more other nodes.<br />
Furthermore, when a node applies its decryption key to a message not encrypted with that node's encryption key,<br />
the result is indistinguishable from a message with forwarding instructions. A sender wraps a message in an onion<br />
route of logarithmic length, branching (forwarding to multiple nodes) at random points, resulting in a tree of<br />
forwarded messages. Nodes wait to initiate messages of their own until they have received a message, so it is<br />
indistinguishable whether they are sending or forwarding. Through analysis and simulation, we show this system<br />
preserves a high degree of sender and receiver anonymity, as well as unlinkability between communicating pairs.<br />
Bioremediation of Endocrine Disrupting Compounds Using the Synthetic Biology BioBrick Standard<br />
Amanda Shelton<br />
Mentors: Richard Murray, Nate Glasser, Toni Lee, Joe Meyerowitz, and Emzo de los Santos<br />
Endocrine disruptors are organic pollutants that cause reproductive harm to fish, birds and possibly humans. The<br />
<strong>2011</strong> Caltech iGEM Team aims to use synthetic biology to create a biosensing and bioremediation system in<br />
Escherichia coli for any of four compounds: bisphenol A (BPA), dichlorodiphenyltrichloroethane (DDT), nonylphenol<br />
and 17α-ethynylestradiol. To detect the presence of these chemicals we characterize the human estrogen receptor<br />
in E. coli and then modify the protein for reduced toxicity to the chassis and to improve its ligand-binding specificity<br />
to some of the endocrine disruptors. We also have identified genes that degrade these compounds and attempt to<br />
characterize their function in E. coli. Using both Gibson and standard assembly to create plasmids, we put the<br />
modified estrogen receptor and the degradation genes into the BioBrick Standard, a way of formatting genes,<br />
promoters, plasmids and other parts for easy retrieval by iGEM teams and others to use in their own synthetic<br />
biology projects.<br />
A Catalytic Enantioselective Approach to the Synthesis of Physovenine<br />
Jeff Shen<br />
Mentors: Brian M. Stoltz and Boram D. Hong<br />
A convenient method for the construction of enantioenriched oxindoles bearing all carbon quarternary<br />
stereocenters is employed toward the synthesis of (�)-physovenine, a natural product containing a furoindoline<br />
moiety. The strategy involves a single step copper-catalyzed enantioselective stereoablative alkylation of the<br />
racemic halooxindole, synthesized from the commercially available p-anisidine, with up to 87% enantiomeric<br />
excess. Subsequent decarboxylation, N-methylation and reductive cyclization of the oxindole allow access to the<br />
desired furoindoline structure. The successful application of the enantioselective alkylation allows for future<br />
asymmetric syntheses of other biologically significant alkaloids such as physostigmine.<br />
3-D Visualization and Modeling of Building Shaking During Earthquakes<br />
Hong Sheng<br />
Mentor: Monica Kohler<br />
The goal of this project is to find a new way to visualize and model the shaking of buildings during earthquakes.<br />
Visualization provides a rapid portrayal of the extent of potentially damaging shaking during an earthquake and can<br />
be used for emergency response, loss estimation, and for public information through the media. For this project I<br />
used Millikan Library as our prototype building. I integrated the accelerations recorded by a 36-channel seismic<br />
network inside Millikan Library during four earthquakes (Mar 16, 2010, ML=4.4 Pico Rivera; Aug. 9, 2007, ML=4.6<br />
Chatsworth; Apr. 4, 2010 Mw=7.2 Baja California and July 29, 2008, ML=5.5 Chino Hills) and applied filters and<br />
numerical integration to get the displacements of each floor during the earthquakes. I then made movies from the<br />
images that resulted from applying the displacements to the initial building model. I also plotted the peak<br />
accelerations and velocities on each floor using a color scale that indicates the extent of shaking during the<br />
earthquake. The peak accelerations we compute from the data are an indirect measure of potential damage.<br />
84
Construction of a Polarization Sensitive Optical Coherence Tomography System<br />
Jeffrey D. Sherman<br />
Mentors: Scott Fraser and Reza Motaghian<br />
Neural degenerative diseases result in tissue damage and cause discomfort, disease, and death. Preemptively<br />
diagnosing these degenerative diseases and visualizing them will help in diagnostics as well as the search for cures<br />
and treatments. Recent advances in optical techniques have led to Optical Coherence Tomography (OCT), a<br />
noninvasive imaging technique without the use of ionizing radiation, particularly in the eye. The birefringence of<br />
tissue can act as a surrogate biomarker, allowing for an analysis of tissue structure and health. Polarization<br />
Sensitive OCT (PS-OCT) has emerged as an extension of OCT for capturing birefringence, particularly in the eye.<br />
Using PS-OCT, it is possible to detect the initiation of neural degeneration. Since the method is noninvasive, a<br />
longitudal analysis can be performed to monitor the progression of these degenerations and adjust therapies to<br />
better help the patient. A PS-OCT system was constructed in free space on optical breadboards for modularity.<br />
To verify the effectiveness of the system, eye scans taken by a Swept Source OCT system at 1060nm will be<br />
compared with scans that will be taken by the newly constructed PS-OCT system at 1060nm. If the PS-OCT system<br />
reveals tissue organization better than standard OCT, then the construction of the system will be considered a<br />
success.<br />
Analysis of Soot Evolution Using a Combined Fluid Mechanics and Chemical Model<br />
Kevin Shi<br />
Mentor: Guillaume Blanquart<br />
The importance of hydrocarbon fuels necessitates the study of their emissions, and in particular their sooting<br />
tendencies. This work relies on two programs, FlameMaster and NGA, to simulate the overall combustion process.<br />
The first uses a detailed chemical mechanism with 185 species and 1903 reactions to generate counterflow<br />
diffusion flames using a flamelet approach. These data are tabulated and integrated into a full 2D fluid dynamics<br />
solver to simulate an axisymmetric coflow diffusion flame. Special emphasis is placed on evaluating the relative<br />
importance of radiation and incorporating a radiation model to account for heat losses. A soot model, built on top<br />
of the chemistry and fluid dynamics, calculates the soot production rate. Sooting tendency is studied for aliphatic<br />
and aromatic dopant species, including polycyclic aromatic hydrocarbons. These results are validated with existing<br />
experimental and computational studies.<br />
Electron Transfer Studies of the Double Tryptophan Azurin Mutant H107W108W110 (All Phe)<br />
Dong Woo Shin<br />
Mentors: Harry Gray and Heather Williamson<br />
Simplified protein structure can be studied to help understand the complex electron transfer system present in<br />
nature. In this study, Azurin mutant with hydrophobic and hydrophilic tryptophan at sites 108 and 110 was used to<br />
look for communication and electron transfer in the system. Circular dichroism and steady-state fluorescence was<br />
used to probe the environment and the communication line of the two tryptophans. Luminescence and transient<br />
absorption of the protein was studied with the nanosecond lasers to determine the nature of the pathway of<br />
electron transfer system. No clear communication between the tryptophans has been isolated. The laser studies did<br />
indicate weak interaction between the rhenium photo sensitizer and the tryptophans, but no communication<br />
between the tryptophans and the native copper.<br />
A 2.4 GHz Printed Circuit Board Proof-of-Concept of a High Efficiency mm-Wave CMOS Active Radiator<br />
Gunnar Atli Sigurdsson<br />
Mentors: Ali Hajimiri and Steven Bowers<br />
The immense popularity of wireless communications has left the common frequency bands crowded, prompting<br />
researchers to utilize available spectrum at ever higher frequencies. At mm-wave frequencies there is pronounced<br />
need for novel antenna designs that are tightly integrated with their driving circuitry in order to reduce power<br />
losses. A radiator concept for 94 GHz CMOS-technology was reviewed, scaled up, and redesigned to work at 2.4<br />
GHz on a FR-4 printed circuit board, in the interest of testing the concept. The radiator works in similar manner to<br />
an array of dipoles, and can connect directly to the last amplifier stage without impedance matching, due to loadpull<br />
matched input impedances, accomplishing all of its power combining in the air. 3D full-wave electromagnetic<br />
field simulations were performed on all transmission line structures and furthermore, various ways to achieve<br />
symmetric power splitting and shorted transmission line stubs with coupled lines were designed and experimented<br />
with, in order to achieve acceptable efficiency and radiation pattern of the radiating array. Simulated results predict<br />
the radiating section having 99.5% efficiency, 4.62 dBi maximum antenna gain and the power amplifier section<br />
delivering 2.28 W of power to the radiating section, making the whole structure 76% efficient with 22.6 dB power<br />
gain.<br />
85
Cosmic Sparklers: A Systematic Search for Cataclysmic Variables Using the Palomar Transient Factory<br />
Gregory Simonian<br />
Mentors: Thomas Prince and David Levitan<br />
I identified potential Catalysmic Variable (CV) sources in the Palomar Transient Factory (PTF) photometric<br />
database. Accomplishing this involved utilizing basic cuts based on magnitude and duration of outbursts to<br />
eliminate spurious sources, and cross-matching remaining candidates using SDSS and SIMBAD catalogs. This<br />
method yielded 44 new potential CVs, and out of these, four were followed up with classification spectra. In order<br />
to study one of these systems in detail, we undertook a case-study of a known special CV: CR Boo, an AM CVn<br />
system. We expect to confirm the connection between quiescent variability and the orbital period of such AM CVn<br />
systems which would be very useful for determining properties of these systems with photometric surveys of the<br />
sky.<br />
Reduction of Thermal Conductivity in Graphene<br />
Prastuti Singh<br />
Mentors: Jim Heath and Slobodan Mitrovic<br />
One of the greatest challenges in thermoelectric is increasing the effectiveness of heat-to-electricity conversions.<br />
One solution is to fabricate materials whose thermal conductivity, ���can be reduced while their electrical<br />
conductivity, �� is kept intact. Recent research suggests the answer may lie with nanostructures. High surface-tovolume<br />
ratio of nanostructures allows for greater scattering of phonons (heat carrying particles). Further research,<br />
led by Prof. Jim Heath’s lab, led to the development of a phononic silicon nanomesh. The nanomesh is patterned<br />
with spacings shorter than the mean free path of phonons, results in greater scattering and therefore, a lower<br />
���The purpose of this project was to examine this process using graphene, a material known to possess extremely<br />
high values of ��and ���Graphene was exfoliated onto SiO2 wafers, and then its location mapped using the SEM.<br />
Four electrodes were written on the graphene, and then a 200 nm mesh was written and etched upon the graphene<br />
using E-Beam Lithography. This process allows for the electrical conductivity of the graphene device to be<br />
measured. Further research is needed to develop a process for measuring the thermal conductivity of the graphene<br />
mesh.<br />
The Regulatory Role of wnt16 in Strongylocentrotus purpuratus Endoderm Specification<br />
Natnaree Siriwon<br />
Mentors: Eric H. Davidson and Isabelle Peter<br />
Cell fate is determined by the gene regulatory network. The endoderm precursor cells in sea urchin embryos derive<br />
from two cell lineages: veg1 and veg2. Endoderm specification in both cell lineages requires the expression of<br />
hox11/13b. The expression of Hox11/13b in veg2 derived cells was predicted to be required for the activation of<br />
itself in veg1 derived cells, according to a Hox11/13b perturbation experiment. Based on that observation, it was<br />
proposed that a wnt signaling ligand expressed downstream of hox11/13b in veg2 derived cells might be involved.<br />
QPCR experiment showed that the expression of wnt16 decreased when Hox11/13b was perturbed. We showed by<br />
whole mount in situ hybridization that wnt16 is expressed in veg2 derived cells at 24h. Moreover, when the<br />
expression of wnt16 was analyzed in 24h Hox11/13b perturbed embryos, a decrease of expression in veg2 derived<br />
cells was observed. These results show that hox11/13b activates wnt16 in veg2 derived cells. When the expression<br />
of wnt16 was perturbed, the expression of bra, blimp and foxA was lowered at 18h and 24h post-fertilization. The<br />
expression of eve was lowered only at 18h. However, the expression of hox11/13b increased at 18h, 21h and 24h.<br />
A QPCR analysis of 15h, 28h, 21h, 27h and 30h Wnt16MASO treated embryos showed that the rise of hox11/13b<br />
and foxA expression delayed by 6 hours due to the perturbation of wnt16. This result shows that Wnt16 is<br />
responsible for the activation of hox11/13b in veg2 endoderm precursor cells. However, it is still inconclusive<br />
whether Wnt16 is the activator of hox11/13b in veg1 endoderm precursor cells.<br />
Risk Mitigated Optimal Power Flow With High Wind Penetration<br />
Emma Sjödin<br />
Mentors: Steven Low and Dennice Gayme<br />
Increased penetration of renewable energy sources into the power grid poses new challenges, mostly due to their<br />
intermittency. To ensure a safe and steady supply of power, a substantial amount of ancillary services will need to<br />
be introduced to the grid. We consider a case where intermittent wind generation is taken as the primary energy<br />
source and where energy storage and spinning reserves from conventional sources are provided. The problem is<br />
formulated as a risk mitigating optimal power flow (OPF) problem, where the objective is to schedule the spinning<br />
reserves to minimize the cost of ancillary services while compensating for uncertainties in generation and load.<br />
Using a modified version of the IEEE 14 bus benchmark system, the OPF is solved both as a finite and receding<br />
horizon control problem, where the optimal strategy of pre-allocation of spinning reserves is studied with both<br />
time-invariant and demand based cost functions. We also investigate the optimal distribution of storage capacity<br />
across different network topologies to minimize transmission losses and unnecessary charge/discharge cycles. The<br />
results quantify the need for ancillary services and suggest a strategy for their scheduling and placement.<br />
86
Automated Information Retrieval System (AIRS): An Economical Safeguard Against Data Loss for High<br />
Altitude Balloon Missions<br />
Russell Smith<br />
Mentors: Jeff Booth and Jason Rhodes<br />
Long duration balloon missions such as the High Altitude Lensing Observatory (HALO) acquire such high quantities<br />
of data that they exceed the downlink bandwidth which can be guaranteed. AIRS provides a backup data transfer<br />
mechanism so that flights need not be limited by fear of total data loss. An autonomous glider with GPS and<br />
autopilot carries a solid state hard drive to an approved location near the balloon’s position. For maximum safety,<br />
mass and final kinetic energy have been minimized. Accurate waypoint targeting both simplifies recovery and<br />
further mitigates risk. A low drag, low aspect ratio wing provides rapid penetration through high crosswinds, yet<br />
supports slow landings at a high angle of attack. The 800g laminated thin wall fiberglass and foam capsule includes<br />
a solid state drive rated to withstand 1500G, more than double the capsule's impact force in an unlikely control<br />
system failure. A radio signal is broadcast after landing to allow AIRS to be easily located with a directional<br />
receiver. AIRS has been designed, fabricated, fitted with an autopilot, and tested for functionality, accuracy and<br />
safety both in simulation and from a weather balloon drop, demonstrating the feasibility of safe and easy data<br />
retrieval at low cost.<br />
Construction and Testing of Low-Energy α- and β-Particle-Tracking Detector Prototype<br />
Daniel Sotolongo<br />
Mentors: Sunil Golwala and Robert Nelson<br />
The Cryogenic Dark Matter Search is searching for evidence of Weakly-Interacting Massive Particles as a candidate<br />
for dark matter. The dominant background for this type of experiment is low-energy electrons emitted by<br />
radioactive contaminants on detectors’ surfaces. Therefore, a detector capable of screening these crystal detectors<br />
for contamination at each step in their manufacturing process was needed. The proposed detector, dubbed the<br />
Betacage, is a drift chamber with energy resolution of a few keV. The Betacage is composed of several wire grids<br />
charged to ~3kV, a voltage that creates an electric field strong enough to cause an electron avalanche around a<br />
given wire from only a few ionized particles. These avalanches are read as pulses on the grids, which are<br />
reconstructed into a 3D path. The construction of the prototype, composed of only 1 grid and smaller than the final<br />
design, has revealed and resolved problems that would have been considerably more problematic at full scale. The<br />
prototype has confirmed the feasibility of the design, and work will soon begin on a larger-scale fully-functional<br />
prototype.<br />
Formation of Membrane Protein Nanocrystals Through Laminar Flow<br />
Kelsey M. Spaur<br />
Mentors: Michael Stowell, Rob Phillips, and Ray (Hsin-Jui) Wu<br />
Membrane proteins play a critical role in transport across membranes. The ability to easily determine membrane<br />
protein structures can greatly benefit the pharmaceutical world as well as enhancing our knowledge of the human<br />
body in general. The project focuses on using a microfluidic device to mix protein and lipid and then induce the<br />
diffusion of detergent through laminar flow to allow the membrane protein to form. To make this methodology<br />
operational, tests were performed to prove that the protein and lipid mix at the designated ratio and to determine<br />
the mixing ratios and flow rates that best formed membrane proteins. Results indicate that the device reliably<br />
mixes protein and lipid at the desired range of ratios. At this point, the flow rate ratios of 1 to 100 and 1 to 500,<br />
protein/lipid combination to buffer respectively, have produced membrane proteins. Determining the diffusion rate<br />
of the detergent would be valuable supporting information about the process. Tests should be run with ratios near<br />
these values, 1 to 90 and 1 to 120 for example, to narrow in on the best ratios. Tests should also be performed<br />
with different protein/lipid mixing ratios.<br />
Neural Predictors of Bargaining Decision Making Using Hyperscanning EEG<br />
Stasja Stanisic<br />
Mentors: Shinsuke Shimojo and Kyongsik Yun<br />
Striking is the act of not participating in beneficial circumstances with the intention of achieving a long-term goal.<br />
In this project we wanted to induce a possible strike situation using a bargaining game. In each experiment a<br />
subject pair would go through 120 trials of the game where they could reach a no deal (no payoff for either) or a<br />
deal (payoff for both) situation. Their brain activity was recorded using electroencephalography (EEG), EGI 128<br />
channel device. Their finger movement, with which they would coordinate their negotiations, was tracked using a<br />
Vicon infra-red camera. EEG data was processed with independent component analysis and standardized low<br />
resolution brain electromagnetic tomography for source localization. Brain activity and finger movement data could<br />
allow discovery of neural process that leads to a strike which can be used for prediction of strikes moments before<br />
the thought is fully formed in the brain. Comparing the brain activity of the two subjects and between the pairs, we<br />
aim to discover what leads to a strike.<br />
87
Do Tutte Polynomials Satisfy the Kontsevich Conjecture?<br />
Jessica Su<br />
Mentor: Matilde Marcolli<br />
The Kontsevich conjecture holds for a polynomial if its zeros are polynomially countable, i.e., if the number of zeros<br />
of that polynomial has polynomial dependence on the size of the field. Our aim is to determine which graphs yield<br />
Tutte polynomials that satisfy the Kontsevich conjecture. With computer software and logical reasoning we have<br />
indicated that very few polynomials satisfy the conjecture directly, but some satisfy a modified version of the<br />
conjecture. In those cases, the dependence is polynomial except at points where a certain parameter (and thus,<br />
the entire polynomial) is 0. Further work should focus on identifying qualities of graphs that yield deeper failures.<br />
The Mechanism Behind I-SceI Injection Method for Strongylocentrotus purpuratus<br />
Julia Y. Su<br />
Mentors: Eric Davidson and Miao Cui<br />
Understanding gene regulatory networks is crucial to comprehending the relationship between genes and their<br />
effects on embryonic development and evolution. Current research on gene regulatory networks has been impeded<br />
by the mosaic pattern that emerges from injecting reporter constructs. Although mosaic pattern does not affect the<br />
data of young embryos, it does impact the data gathered from older animals. Research has observed that this new<br />
I-SceI injection method has decreased the mosaic pattern in embryos. There is little if anything known about the<br />
mechanisms behind this injection method. If the mechanism of how the injected DNA is integrated is understood,<br />
then the knowledge can be used to further decrease the mosaic patterns and help scientists gather better data in<br />
their analysis of gene regulatory networks. In this study, TAIL-PCR is used to recover the adjacent sequence<br />
flanking the injected DNA which directly answer the question of where it is integrated into genome. In addition, It<br />
also indicates the form of exogenous insertion by obtaining injected DNA itself as the adjacent sequence.<br />
Gesture Based Interface for Solar Decathlon House<br />
Ka-Chuan Suen<br />
Mentor: Richard Murray<br />
Energy use in homes and residence accounts for nearly 25% of the total US energy output per year. While<br />
appliances such as lights and media devices are not the largest energy consumers in a house, they are easier to<br />
control and turn off when not in use. Using the recent advances in 3D cameras and pairing it with a home<br />
controller system, Control4, we provide a gesture based interface to control lights and appliances. The Xbox kinect<br />
with its 3D video stream is used to determine where a person is pointing and compares it with a virtual 3D map of<br />
appliances to find which appliance is selected. Commands are then sent using Control4. The location of the user is<br />
also used as a separate system for controlling devices. With greater ease in controlling household appliances that<br />
use energy, it is hoped that overall energy use in a house can be decreased.<br />
C-H Activation of Methane Using Homogeneous Rh-Based Catalysts: A Computational Study<br />
Vincentius Jeremy Suhardi<br />
Mentors: William Goddard III and Robert Nielsen<br />
Despite the fact that methane is one of the main components of natural gas, it has not been used to its full<br />
potential due to difficulties associated with its selective functionalization. Current technologies for the conversion of<br />
natural gas to liquid products proceed by the energy intensive Fischer-Tropsch process (Temperature = 850 o C). In<br />
this project, we explored different Rhodium-based catalysts for direct catalytic partial oxidation and<br />
functionalization of methane’s C-H bond. The advantage of direct functionalization of methane is the increased<br />
efficiency due to the ability for the process to proceed at much lower temperature (T=150-200 o C). Using Density<br />
Functional Theory (DFT) and the polarizable continuum model of solvation, we screened various Rh-bound pincer<br />
ligands (L), and various combinations of nucleophiles (X)/solvents combinations (HX) (X=TFA - , Cl - , HSO4 - , SO4 2- ,<br />
and OH2). Screening of free energy surfaces for CH activation and SN2 functionalization barriers predicted that<br />
when L is a CNC-derived ligand with X=Cl - (~ 1 mM) and water at pH~8 as solvent, a full catalytic cycle of<br />
conversion of CH4 to CH3Cl will be observed with turnover frequencies ~ 1 hr -1 at T�200 o C.<br />
88
Optical Feedback Cooling of a “Zipper” Optomechanical Cavity<br />
Aiden Sullivan<br />
Mentors: Oskar Painter and Alexander Krause<br />
In an optomechanical cavity, such as the “zipper” cavities used in this investigation, the circulating optical field<br />
gives rise to a radiation pressure force that can drive the mechanical modes of the cavity. In turn, the mechanical<br />
oscillations deform and detune the optical cavity and thus modulate the optical field. Such coupling of the<br />
electromagnetic and mechanical degrees of freedom allows energy and information transfer between the optical<br />
and mechanical modes. Cavity optomechanics allows for unprecedented tests of quantum mechanics at the<br />
mesoscale. However, the mechanical mode must be close to its quantum mechanical ground state for quantum<br />
effects to be apparent.<br />
Cryogenic cooling alone cannot sufficiently cool the mechanics. In order to cool further, we create a damping force<br />
with feedback. We measure the position of the mechanics with the light modulated by the cavity, and modulate the<br />
light going into the cavity to drive the mechanics with a force proportional to the velocity of the mechanics. We are<br />
working towards an implementation of such a feedback loop, and have developed a design that uses analog<br />
electronics to accomplish this. Further work includes implementation and tests of this design, as well investigating<br />
the possibility of using DSP technology to accomplish feedback.<br />
NO2 Mixing Ratio Determination by Topographic Target Light Scattering-Differential Optical Absorption<br />
Spectroscopy (ToTaL-DOAS) in Gwangju, South Korea<br />
Sylvia Sullivan<br />
Mentors: Young Joon Kim and Jihyo Chong<br />
Nitrogen dioxide is a primary pollutant from fossil fuel combustion, which acts as an anthropogenic fingerprint in<br />
the troposphere. The free-radical cycle of its chemical family (NOx) also has a direct influence on stratospheric<br />
formation and tropospheric removal of ozone. These roles make quantification of atmospheric NO2 presence<br />
especially important. In this study, differential optical absorption spectroscopy is coupled with the newly developed<br />
topographic target light scattering method (ToTaL-DOAS) to obtain NO2 mixing ratios in Gwangju. The sensitivity of<br />
ToTaL-DOAS to temporal and spatial variations in urban boundary layer NO2 levels makes it an especially promising<br />
technique. Three sun-illuminated targets of similar albedo at increasing distance (560 m, 920 m, and 1490 m) are<br />
scanned both horizontally and vertically using Multi-Axis DOAS instruments. NO2 fluctuations are considered in<br />
relation to different meteorological parameters, and the accuracy of the ToTaL-DOAS method is assessed by<br />
comparison with in-situ data from three nearby air quality monitoring stations.<br />
Fabrication and Studies of Overdoped High-Temperature Superconducting Cuprate Thin Films<br />
(Y1-xCax)Ba2Cu3O7-�<br />
Qunchao Sun<br />
Mentors: Nai-Chang Yeh and Zhenjie Feng<br />
Vortex liquid state is a state when the H field goes up and close to Hc2, the vortices in Type II superconductor will<br />
be mobilized, which gives rise to a tiny current. The curve that marks the transition from vortex solid state to<br />
vortex liquid state is in between the curve of Hc1 vs. T and that of Hc2 vs. T. In my project, my objective is using<br />
Pulsed Laser Deposition (PLD) to make four samples of (Y1-xCax)Ba2Cu3O7-� with different doping levels of Calcium,<br />
and get the phase boundary between the vortex solid state and the vortex liquid state in the magnetic field vs.<br />
temperature plot for each of them to find the correlation between the vortex phase boundary and the doping<br />
levels.<br />
Directed Evolution of Alcohol Dehydrogenases for Aldehyde Resistance<br />
Sabrina I. Sun<br />
Mentors: Frances Arnold and Devin Trudeau<br />
Detoxification of lignocellulosic degradation products is an important challenge in optimizing viability and sugar-toproduct<br />
yields of engineered biofuel pathways. Among these compounds, aldehydes interfere most with microbial<br />
functionality whereas alcohols are notably less toxic. Alcohol dehydrogenases reduce aldehydes to alcohols using<br />
NADPH as a cofactor, but existing enzymes lack catalytic activity on many industrially relevant toxins. We therefore<br />
aim to create a novel alcohol dehydrogenase with broad substrate specificity and high activity, to improve aldehyde<br />
resistance in a range of fermenting microorgansisms. We have cloned and overexpressed the Saccharomyces<br />
cerevisiae ADH6 (alcohol dehydrogenase 6) gene, and verified that this can improve resistance of the wildtype<br />
BY4741 to cinnamaldehyde and 5-hydroxymethylfurfural. Random mutagenesis was used to construct an ADH6<br />
mutant library of approximately 10 6 clones. To achieve both high enzymatic activity and promiscuity, variants are<br />
currently being screened on combinations of aldehydes at lethal concentrations.<br />
89
Detecting Dust Devil Tracks and Wind Streaks in the North Polar Region of Mars<br />
Vivian Sun<br />
Mentor: Leslie Tamppari<br />
This study utilizes images from the Mars Reconnaissance Orbiter’s (MRO’s) Context Camera (CTX) to detect dust<br />
devil tracks (DDT), windstreaks (WS), and active dust devils (DD) in the 65-75N latitude band of Mars’s North Polar<br />
Region. Such features are formed through aeolian processes and can be indicative of local wind directions. The CTX<br />
images were processed and map-projected using ISIS, a planetary image processing software. The dataset yielded<br />
roughly 1500 usable images, of which 75% contained at least one of DDT, WS, or DD, with over 60 occurrences of<br />
active dust devils. DDT and WS will be analyzed for their frequency of occurrence and orientation and distribution<br />
according to location, and overlapping images will be compared for temporal analysis to determine seasonal,<br />
annual, and/or interannual changes and to assess the lifetimes of DDTs and WS.<br />
CubeSat Radiometer Development<br />
Yichuan Sun<br />
Mentor: William R. Johnson<br />
NASA/JPL has developed two automated validation sites at Lake Tahoe CA/NV and Salton Sea CA to help calibrate<br />
and validate thermal infrared measurements from airborne and satellite platforms. Radiometers deployed at the<br />
sites are used to measure the surface temperature of the water and a variety of meteorological equipment is used<br />
to provide ancillary measurements including relative humidity, air temperature, wind speed and direction and<br />
atmospheric pressure. In order to provide National Institute of Standards and Technology (NIST) traceability<br />
between the radiometers and the remote measurements the field radiometers must first be calibrated in the<br />
laboratory in a controlled environment. This process involves changing multiple variables to simulate a range of<br />
conditions, and requires operating four instruments simultaneously: environmental chamber, water bath,<br />
thermistor, and a radiometer. In order to undertake the laboratory calibration I have developed a multithreaded<br />
application coded in Python to test all possible permutations of environmental conditions. The code developed<br />
allows this by controlling all the necessary hardware. Additionally the program will also be used to simulate real<br />
weather data obtained from existing setups.<br />
Investigating Bismuth Oxide as a Photocatalyst for Water-Splitting<br />
Dylan Sures<br />
Mentor: Jay Winkler<br />
The photocatalytic process of water-splitting has the potential to produce a “clean” energy source in the form of<br />
hydrogen gas. In order to accomplish this water-splitting, a semiconductor photocatalyst is necessary; bismuth<br />
oxide has been investigated as a possible semiconductor material. Through cyclic voltammetry and open circuit<br />
potential measurements, it was determined that bismuth oxide produces a moderate, albeit unstable photovoltage<br />
in neutral and acidic electrolytic solutions. In higher-pH electrolyte solutions, bismuth oxide exhibited larger<br />
photovoltages and better, but still suboptimal stability as a photocatalyst. By combining bismuth oxide with other<br />
transition metal oxides, there is great promise for the preparation of a strong, durable photocatalyst.<br />
Terahertz Time-Domain Spectroscopy and Far Infrared Spectroscopy of Minerals<br />
Kevin M. Sutherland<br />
Mentors: George R. Rossman and Geoffrey Blake<br />
Far infrared spectroscopy and terahertz time-domain spectroscopy are used to investigate the low-frequency<br />
vibrational modes and torsional modes of crystalline solids. This study presents far infrared spectra in the range of<br />
600-50 cm -1 and terahertz time-domain spectra (THz-TDS) in the range of 150-3 cm -1 of 100 minerals. The<br />
minerals used in the study represent a wide cross section of common and rare minerals including oxides,<br />
carbonates, phosphates, pyroxenes, sulfides, sulfates, native elements, clay minerals, tungstates, and various<br />
other minerals. Samples are prepared for terahertz and far infrared analysis by pressing a mixture of polyethylene<br />
powder and mineral sample into pellets. Far infrared absorption spectra were collected with a Fourier transform<br />
infrared spectrometer and terahertz absorption spectra were collected with a femtosecond laser-driven terahertz<br />
time-domain spectrometer; in each case, spectra were collected in a dry gas-purged environment. Terahertz<br />
spectra and far infrared spectra were collected for 100 different minerals phases, compiled into a continuous<br />
spectra from 600 cm -1 to 3 cm -1 for each mineral, and entered into our terahertz/far infrared spectral library. All<br />
minerals investigated in the study produced a unique spectral fingerprint in the 600-50 cm -1 range.<br />
Thermoelectric Optimization in Sr3AlSb3 and Ca5In2Sb6 Through Control of Carrier Concentration<br />
Jessica G. Swallow<br />
Mentors: G. Jeffrey Snyder and Alex Zevalkink<br />
Sr3AlSb3 and Ca5In2Sb6 are two attractive possibilities in the search for new, efficient, low-cost, nontoxic<br />
thermoelectric materials. With controlled doping, the conflicting and interdependent properties that determine the<br />
thermoelectric figure of merit of these materials, and therefore their efficiency, can be optimized. Both Zintl<br />
90
compounds were synthesized with several dopant concentrations using powder metallurgy and their transport<br />
properties, including Seebeck coefficient, thermal conductivity, and resistivity, were investigated to high<br />
temperature. Modeling based on these measurements allows prediction of the optimal carrier concentration at<br />
which the maximum figure of merit for both compounds can be achieved.<br />
Tracking and Analyzing Avalanche-Like Flows in Granular Media<br />
Wesley Swank and Rivers Ingersoll<br />
Mentors: José Andrade and Carlos Avila<br />
Although phenomenological models for steady flows in granular media exist, the underlying mechanics of less<br />
predictable situations, such as avalanches, are difficult to simulate and predict with existing methods. By creating a<br />
link between the precision of the discrete element method (DEM) at the grain level and the flexibility of continuum<br />
models, the mentor is developing algorithms that can simulate sudden deformations with a high degree of<br />
accuracy. The goal of this project is to provide essential physical evidence to quantitatively assess the accuracy of<br />
the mentor’s theoretical framework and of simulations developed by the co-mentor. To generate this data, a highspeed<br />
camera captures thousands of images per second of the progression of short, recurring avalanches of steel<br />
beads in a steadily rotating plexiglass drum. These images are then reduced to an array of coordinates<br />
corresponding to the centroids of each bead in the frame of view, providing velocity and displacement values in a<br />
two-dimensional plane. With analysis, this data will be used to refine and eventually confirm the mentor’s model for<br />
the early progression of an avalanche.<br />
HCAL Noise Rate Stability Using Fit-Based Filters<br />
Kelly Swanson<br />
Mentor: Maria Spiropulu<br />
Robust noise-filtering algorithms are crucial for analyzing new physics phenomena, particularly those characterized<br />
by missing transverse energy. This paper presents an overview of fit-based noise filters and their long-term<br />
behavior. Using <strong>2011</strong> data, we examined isolated, flat, spike, and triangle pulse-shape discriminants for HCAL<br />
barrel and endcap anomalous signals. From this, we determined the channels that were persistently noisy and<br />
examined each filter's performance in high pile-up runs.<br />
Cellphone Medicine: Medical Devices for a Mobile Platform<br />
Albert Tan<br />
Mentors: K. Mani Chandy and Julian Bunn<br />
Medical technologies and devices have advanced significantly in the past decades, but the cost of the equipment<br />
remains high. Many third world countries are unable to afford such quality devices and thus have a dire lack of<br />
medical resources. Cellphone technologies, widespread across the globe, offer a potentially inexpensive solution to<br />
this problem. This project developed third generation low cost electronic stethoscopes and electrocardiograms in<br />
addition to hardware and software required for the devices to interact with cellular phones. The systems developed<br />
allow heart auscultations and electrical activity to be recorded and analyzed concurrently on the cellphone<br />
platform. The data can subsequently be uploaded to the Cloud and downloaded by a medical professional for<br />
further review. By building upon existing cellphone technologies, the project aims to create an affordable 10 cent<br />
medical checkup and diagnostic system that can be used in countries with limited resources.<br />
Viral Targeting of Dopaminergic Neurons<br />
Alison Tan<br />
Mentors: Thanos Siapas and Maria Papadopoulou<br />
The objective of this project is to use viral tools and target gene expression selectively to dopaminergic neurons of<br />
the rat midbrain. These neurons play a critical role in reward processing, addiction and movement initiation. Their<br />
loss, as well as misregulation of pathways in which these neurons are involved, have been implicated in many<br />
neurological diseases and disorders; as such, targeting gene expression specifically to these neurons is of great<br />
clinical significance. To accomplish specific targeting, we are testing injections of three different adeno-associated<br />
viruses in the midbrain dopamine-rich areas of rats. In these viruses, GFP expression is placed under the influence<br />
of an aminergic promoter; we therefore expect infected cells in the midbrain expressing GFP to be dopaminergic.<br />
Success of viral targeting of dopaminergic neurons will be assessed by the co-localization of GFP expression and<br />
the labeling of dopaminergic cells through antibody staining. By comparing the three viruses, we aim to identify the<br />
virus that leads to the highest spread of infection while maintaining targeting specificity of gene expression to<br />
dopaminergic cells.<br />
91
Construction and Characterization of Artificial Vessels for in vitro Arterial Wave Dynamic Studies<br />
Michelle Tang<br />
Mentors: Morteza Gharib and Niema Pahlevan<br />
The study of the wave dynamics and hemodynamics of the arterial system will lead the medical community to a<br />
better understanding of the pathogenesis and prognosis of cardiovascular complications, some of which include<br />
arteriosclerosis, ventricular hypertrophy, and heart failure. These two cardiovascular biomechanical concepts are<br />
affected by such factors as heart rate, reflection sites, pumping characteristics, and aortic rigidity. Wave<br />
dynamics—a synthesis of the effects on both the fluid flowing through the vessels and the vessel walls—are<br />
specifically influenced by the different parameters associated with aortic rigidity. For the aorta, such dynamics<br />
affect the load on and perfusion of the heart; for other arterial vessels, such dynamics affect the fluid mechanics<br />
within the vessels. In this study, we are investigating the relationship between aortic rigidity and wave dynamics—<br />
more specifically, how aortic rigidity influences the wave dynamics within the cardiovascular system. We designed<br />
several aortic and other arterial vessels of various compliances with either latex or silicone and subsequently<br />
measured their volume compliance, local compliance, elastic modulus, arterial distensibility, pulse wave velocity,<br />
and fluid particle velocity. The calculated results were then compared to the corresponding clinical results<br />
generated from previous studies of males and females within the range of 20 to 80 years of age. From these<br />
comparisons, we were able to experimentally simulate the physiological wave conditions associated with the<br />
arterial systems of males and females of various ages.<br />
Acquiring the High Frequency Terahertz Spectra of Amino Acids and Sugars Using Terahertz Pulses<br />
Generated via Plasma Filamentation<br />
Justine Tawel<br />
Mentors: Geoffrey Blake and Marco Allodi<br />
This work examined various hydrated and dehydrated amino acids and sugars in the 0.1 to ~7 THz range of the<br />
electromagnetic spectrum, corresponding to the range used by astronomical observatories to collect data. The time<br />
domain THz spectrometer contained a plasma filament to generate the THz radiation and a GaP crystal was used<br />
for detection. We acquired spectra of D-arabinose, D-fructose, D-glucose, D-mannose, D-xylose, glycoaldehyde,<br />
sucrose, L-aspartic acid, L-glutamic acid, 2-aminoisobutyric acid, and γ-aminobutyric acid over a wide range of<br />
temperatures from 10 to 310 kelvin. This study will enable astronomers to identify the presence of amino acids and<br />
sugars in regions of the cosmos, as these molecules have unique spectral fingerprints in the THz range. Prebiotic<br />
compounds, like amino acids and sugars, are thought to have played a role in the origin of life on Earth. In order to<br />
better understand the process of how life originated on Earth, and to search for the precursors of life in other parts<br />
of the cosmos, it is necessary to be able to find and identify regions where these prebiotic molecules are present.<br />
FU Orionis Stars<br />
Jamie Tayar<br />
Mentor: Lynne Hillenbrand<br />
FU Orionis is the prototype of a class of young stellar objects most notable for their dramatic outbursts. The<br />
spectra of stars have several distinctive features including P Cygni profiles in lines such as Hydrogen, Helium,<br />
Sodium D, Calcium II H and K as well as optical spectra similar to F or G supergiants and near infrared spectra<br />
characteristic of K or M supergiants. The leading theory is that an instability in the circumstellar disk causes a<br />
dramatic jump in accretion. The purpose of this project was to examine high resolution optical spectra taken of<br />
several FU Orionis objects and compared them to the spectra of stars ranging from F supergiants to M dwarfs in<br />
order to better understand the physical state of these objects and compare them to the predictions of leading<br />
theoretical explanations.<br />
Correlation of Gamma-Ray Photons With Radio Giant Pulses From the Crab Pulsar<br />
Thanchanok Teeraratkul<br />
Mentors: Anthony Readhead and Glenn Jones<br />
To gain better understanding of giant pulse (GP) emission mechanism and test models of giant pulse emission, we<br />
have carried out observations of the Crab pulsar at radio wavelengths using the GAVRT telescope which can then<br />
be compared to simultaneous gamma-ray observations made by the Fermi space telescope. In 60.37 hours of<br />
simultaneous observations, we obtained 8950 giant pulses observed at radio frequency of 2.6 GHz and 847 Fermi<br />
photons with energies between 100 MeV and 153 GeV. One photon with energy 1872.87 MeV was found within 0.4<br />
ms window of the interpulse (IP) GP, and one with energy 134.56 MeV was found within 1.8 ms window of the<br />
main pulse GP. Higher energy photons are more likely to be found near GP window, indicating a weak correlation<br />
between GP and gamma-ray.<br />
92
Erasure Correction for Nested Networks<br />
Omer F. Tekin<br />
Mentor: Tracey Ho<br />
We consider the problem of finding erasure and error correction capacity regions of networks having a nested<br />
structure, where each sink receives a subset of the packets received by its successor. Such nested networks can be<br />
used to model streaming applications, where each sink represents a deadline by which a certain set of information<br />
is decoded. We first characterize the erasure correction capacity region of a nested network for any given number<br />
of erasures, and provide an intrasession coding scheme that achieves it. We establish a relation between erasure<br />
correction capacity and error correction capacity regions under arbitrary number of erasures or errors. We then<br />
consider a sliding window erasure model as an approximation to probabilistic erasure model. We give an upper<br />
bound for the erasure correction capacity, and provide an intrasession coding scheme that achieves a constant<br />
multiple of the erasure correction capacity region.<br />
Fracture Toughness Enhancement of Elastically Heterogeneous Solids<br />
Ishan Tembhekar<br />
Mentors: G. Ravichandran and Shuman Xia<br />
Various brittle and quasi-brittle materials, like ceramics, cast iron, concrete and cast iron, find immense<br />
applications in industrial and engineering fields. These materials are often observed to fail at stress levels much<br />
lower than their theoretical values because of the presence of mechanical flaws or cracks in their microstructures.<br />
It is thus very important to increase the resistance to crack growth of such materials to prevent unknown failure<br />
mechanical structures. Existing work in this direction involves embedding the material system with tougher<br />
particles having higher intrinsic fracture toughness values and increasing crack resistance by crack trapping and<br />
crack pinning mechanisms. This project proposes that the same crack trapping can be achieved by introducing<br />
elastic heterogeneity only, and consequently leads to fracture toughness enhancement in an alternative way. The<br />
proposed enhancement mechanism is demonstrated on an experimental model system, in which elastic<br />
heterogeneity is achieved by embedding sapphire half-ball lenses in a polydimethylsiloxane (PDMS) matrix molded<br />
on a polyester substrate. A crack is initiated and propagated along the PDMS-polyester interface by application of a<br />
mechanical load. The entire event of crack propagation, pinning and depinning is captured by optical imaging. By<br />
using image processing techniques and load data, plots of fracture toughness versus crack length are obtained for<br />
two samples – one having elastic heterogeneity and other being homogeneous. The toughness values for the<br />
former sample are observed to be markedly higher thus conclusively proving the enhancement of fracture<br />
toughness of the system due to elastic heterogeneity.<br />
Bioremediation of Endocrine Disruptors From Water Using Genetically Modified Escherichia coli<br />
Nicole Thadani<br />
Mentors: Richard Murray, Nathaniel Glasser, Toni Lee, Joseph Meyerowitz, and Emmanuel Lorenzo de los Santos<br />
Endocrine disruptors, or substances that mimic estrogen in the body, have disastrous biological effects on the<br />
reproduction of several species of wild fish and birds. For the iGEM synthetic biology competition, the Caltech team<br />
is focusing on bioremediation of these toxins. The goal of this project is to create a system housed in E. coli that<br />
can be deployed in bodies of water to detect and remove endocrine disruptors. We identified enzymes that can<br />
degrade endocrine disruptors from previous research and from our own exploration of the genetic material from<br />
microbes in the Los Angeles river and characterized the behavior of these enzymes within E. Coli. We also isolated<br />
organisms that can grow on endocrine disruptors in minimal media as new sources of degradation enzymes. In<br />
addition, we explored the potential of certain cytochrome p450s to initiate degradation of these chemicals. We also<br />
explored the functionality of the human estrogen receptor as a signal for the presence of endocrine disruptors, by<br />
characterizing and modifying its toxicity and developing signal pathways that this enzyme can trigger through<br />
binding to endocrine disruptors and to a target genetic sequence. Finally, we characterized the functionality of E.<br />
coli protein processing when E. coli is deployed as an easily containable biofilm on plastic in aqueous environments.<br />
A Study of the Initiation Mechanism of Grubbs-Hoveyda Metathesis Catalysts<br />
Jordan C. Theriot<br />
Mentors: Robert Grubbs and Keith Keitz<br />
Despite the extensive kinetic studies performed on the first- and second-generation Grubbs catalysts, a conclusive<br />
initiation mechanism has not yet been determined for the Grubbs-Hoveryda type olefin metathesis catalyst. This is<br />
largely due to the limitations of standard NMR techniques: the substrate concentration cannot be increased enough<br />
to definitively observe saturation kinetics. We seek to overcome this obstacle by inducing saturation kinetics<br />
through an alternate method. A series of Hoveyda-type ruthenium catalysts have been prepared with varying<br />
substituents replacing the isopropoxy group of the benzylidene ether ligand. A study of their initiation kinetics gives<br />
further insight into the mechanism of this class of olefin metathesis catalysts.<br />
93
Entropy and the Witt Construction (or, Universal Structures in Elementary Algebra)<br />
Ryan Thorngren<br />
Mentor: Matilde Marcolli<br />
Variables, considered formally—without any reference to an underlying ring or algebra—have certain properties like<br />
associativity, commutativity, etc. which can be defined regardless of what we actually consider the variables to be.<br />
These formal properties are the object of study of universal algebra, which does not consider particular groups,<br />
modules, or lie algebras as interesting as their theories. One interesting structure is that of the Witt ring, an<br />
endofunctor in the category of formal rings which always produces a characteristic zero ring. The Witt functor is<br />
completely determined by formal properties of variables, and so we say it is a universal structure of elementary<br />
algebra. In application, the Witt ring can be used in characteristic p to produce unramified extensions of the p-adic<br />
numbers. In a sense it allows us to study characteristic p with characteristic 0. There is an obscure characteristic:<br />
one, whose objects if one would “write down” would be degenerate. Nevertheless, we can define structures over<br />
these objects by taking suitable limits of structures over characteristic p fields. A “limit geometry” is currently<br />
studied in hope of producing a solution to the Riemann hypothesis. Recently, Alain Connes constructed a Witt<br />
functor for characte.ristic one, whose inner workings depend critically on the Shannon entropy. In the present<br />
work, we argue that the same formal structures that force the form of the Witt functor also force the forms of<br />
Shannon and Tsallis entropies in information theory, and a coherent dictionary between characteristic zero Witt<br />
operads and information measures is presented.<br />
Split Stream Flow Past a Blunt Trailing Edge With Application to Combustion Instabilities<br />
Vicky Tian<br />
Mentors: Beverley McKeon and Ivett Leyva<br />
In a coaxial injector of a liquid rocket engine propellants traveling at different velocities are separated by the inner<br />
jet post before they come into contact with each other, mix, combust and power the rocket. Knowing how the<br />
fluids mix and how susceptible they are to instabilities is paramount for a successful liquid rocket engine. In this<br />
study, the wake behind a blunt trailing edge of a long plate, similar to an unwrapped co-axial injector, was studied<br />
using two fluid streams of different velocities introduced on opposite sides of the plate, and PIV was used to<br />
visualize the instabilities in the wake in order to determine the influence of the velocity ratio of the split stream.<br />
Measurements of the vortex shedding frequency were measured at various velocity ratios and compared with<br />
uniform free stream conditions to better understand of the mixing in coaxial jets typically used in cryogenic liquid<br />
rocket engines.<br />
Optical Counterparts to Unknown ROSAT (Röntgensatellit) X-Ray Sources<br />
Suk Sien Tie<br />
Mentors: Shrinivas Kulkarni and Varun Bhalerao<br />
The ROSAT Bright Source Catalog contains 18806 bright soft (0.1-2.4 keV) X-ray sources, of which around 50%<br />
are still unidentified. Using the database from the Palomar Transient Factory (PTF), we searched for the optical<br />
counterparts to these unknown X-ray sources to discover rare X-ray sources and/or new astronomical class(es).<br />
We look for sources with variable optical lightcurves within the positional error circle of each unidentified ROSAT<br />
source using a variability cut. Due to the huge amount of data, this process is being developed into an automated<br />
pipeline, of which PTF candidates of at least 300 unknown ROSAT sources pass the variability cut. As the filtering<br />
pipeline improves, we expect to cut down on the number of multiple PTF candidates to an unknown ROSAT source.<br />
In the presence of multiple candidates after the variability cut, they will be ranked accordingly. We report details of<br />
the matched counterparts.<br />
Pauli-Based Local Commuting Projector Codes<br />
Mohit Tiwari<br />
Mentors: John Preskill and Jeongwan Haah<br />
If Ai and Bi are commuting Pauli operators, then P = ¼(1 + Ai + Bi + AiBi) is a projector. We can define a quantum<br />
code whose code space is the ground space of a governing Hamiltonian H = ΣPi. We first provide a worked example<br />
with A1 = ZXZI, B1 = YXYI, A2 = IZXZ, B2 = IYXY. We then examine different choices of Pi and the Hamiltonian,<br />
which does not necessarily need to be taken to be the sum of the Pi, with the operators Ai and Bi taken to be the<br />
star and plaquette operators of the Toric Code proposed by Kitaev. We are interested in the logical operator<br />
algebra of such codes and implications for code distance and the existence of topological order in these systems.<br />
W+/Z+ Jets Cross-Section Ratio Produced With Muon Final States in pp Collisions at the LHC<br />
Baojia Tong<br />
Mentors: Maria Spiropulu, Yi Chen, and Christopher Rogan<br />
The study of jets produced in association with W and Z bosons provides important tests of perturbative QCD<br />
calculations, and the production of vector bosons with jets constitutes an important background in searches for<br />
new physics. In this project, we consider cases where W and Z bosons decaying into muons and the events are<br />
94
econstructed with particle-flow techniques. This study is done with the full 2010 data sample and current <strong>2011</strong><br />
data recorded with the CMS experiment at the LHC. The distribution of the number of jets produced in association<br />
to W and Z boson is processed after correction for muon efficiency, jet-energy scale and pile-up contributions. The<br />
analysis focuses on ratio measurements to avoid luminosity uncertainty and the impact of experimental<br />
systematics. We acquired the normalized inclusive rates of jets produced as well as the different jet multiplicity<br />
cross section ratios with different jet transverse momentum thresholds. From this we are able to compare the<br />
scaling behavior of the jets with different Monte Carlo simulations and current theoretical predictions.<br />
Implementation of a Distributed Camera Control Protocol<br />
Stephanie Tsuei<br />
Mentors: Richard Murray and Necmiye Ozay<br />
Current security camera networks are composed of many closed caption televisions (CCTVs). While CCTVs have<br />
aided in criminal investigations and increased security in public locations, they are still not very cost effective. We<br />
have created a small camera network made up of small tracking cameras with the ability to track people in an area<br />
and pan-tilt-zoom (PTZ) cameras with the ability to pan, tilt, and zoom to acquire high-resolution photographs of<br />
those walking through the arena. In order to do so, we modeled the interaction between the cameras and people<br />
as a “game” between the “environment”, or the people, and the “system”, or the PTZ cameras. If the cameras are<br />
able to acquire the necessary high-resolution photographs, then the system “wins” the game; otherwise the<br />
environment wins the game. By making a few assumptions about how people move and knowing how the cameras<br />
can move, we synthesize a controller that can control the cameras’ motions. We have implemented the protocols<br />
and collected tracking data and pictures on scripted targets and present those results here.<br />
Calculating Star Formation Histories Using Near Infrared Images of the M33 Galaxy<br />
Gautam Upadhya<br />
Mentors: Tom Soifer and Jason Melbourne<br />
We use Near Infrared (NIR) images of four fields in the nearby M33 galaxy to derive their star formation history<br />
(SFH). NIR images of galaxies are dominated by Asymptotic Giant Branch (AGB) and Red Helium Burning (RHeB)<br />
stars. AGB stars are red stars burning helium in shells outside their core, while RHeB stars are bluer and burn<br />
helium in their cores. Their properties are difficult to model and thus it is difficult to derive an SFH that produces<br />
correct populations of these stars. Using the CALCSFH program (which finds bestfit SFH’s for given sets of data),<br />
we have tried to reproduce the optically derived SFH's using the M33 NIR data. Previously, in other galaxies, this<br />
method yielded SFH's where the AGB population was overproduced, and the RHeB stars were under-produced. The<br />
main sequence and red giant branch stars in the SFH matched the real data however. It is likely that the stellar<br />
evolution codes used in the CALCSFH program need to be reevaluated to properly account for the AGB and RHeB<br />
behavior. Being able to derive SFH’s just from NIR flux will be of great help in finding SFHs of very distant galaxies<br />
where individual stars are not resolvable.<br />
Applications of the Dawid and Skene Model to Discrete Image Annotation<br />
Sara Venkatesh<br />
Mentors: Pietro Perona and Peter Welinder<br />
Crowdsourcing through online services such as Amazon Mechanical Turk can be an effective means of annotating<br />
images. We address the problem of collecting reliable annotations of discrete-labeled image sets through the<br />
implementation of the probabilistic model for the annotation process proposed by Dawid and Skene (1979). We<br />
explore the annotator parameters under which the model can provide accurate estimates of annotator competence<br />
and attain a higher accuracy in image labeling than majority ruling. We furthermore present possible advantages of<br />
supervising the algorithm by keeping image labels fixed at their ground truth values.<br />
Structured Illumination Microscopy for Improved Lateral and Axial Resolution<br />
Malvika Verma<br />
Mentors: Chin-Lin Guo, Jiun-Yann Yu, and Mingxing Ouyang<br />
This project involves constructing a structured illumination microscope based on previous models and then<br />
designing novel combinations of structured illumination with confocal, epifluorescent, or multi-photon microscopes<br />
such that the lateral and axial resolution of biological samples can be improved. To observe three-dimensional<br />
biological samples using microscopes, both surface and depth profiles - lateral information (x, y) and depth (z),<br />
respectively, are required. Scanning confocal microscopes are the most commonly used optical instrument for such<br />
purposes. They convey optical cross-sections, at different depths, by restricting light to the focal plane and<br />
rejecting out-of-focus light such that the final image contains only the in-focus information of the specimen at<br />
different focal plane depths. The confocal microscope does point-scanning across a cross-section, which is very<br />
expensive and time consuming. To speed up this process, higher excitation intensity is needed, but that is<br />
potentially damaging to cells. Structured illumination—projection of a one-dimensional, single spatial frequency<br />
fringe pattern onto the sample—is another optical sectioning technique. Image processing of three images at three<br />
different spatial positions can remove the out-of-focus image and fringe pattern.<br />
95
Development of PCR (Polymerase Chain Reaction) in Microfluidics<br />
Eugene Vinitsky<br />
Mentors: Michael Roukes and Jessica Lynn Arlett<br />
The goal of this project is to incorporate PCR (Polymerase Chain Reaction) into microfluidic chips. The ability to<br />
perform PCR in our dielectrophoretic chips will allow us to perform DNA and protein analysis on single cells in<br />
picoliter volumes of fluid. The main hurdle in utilizing PCR in microfluidics is the evaporation of the fluid in the<br />
channels due to the high heat needed to disassociate the strands of DNA. We hope to solve this by developing an<br />
appropriate hydration layer. Failing that we will manipulate the dimensions of the chambers to alter the surface<br />
area to volume ratio. If we succeed, we intend to make calibration curves to test the efficacy of PCR in our system<br />
versus the normal protocol. Finally, we intend to attempt to measure protein expression via reverse transcription<br />
mRNA measurements and integrate this with our dielectrophoretic chips.<br />
Near-IR Laser Cavity Ringdown Spectroscopy of Chloroalkylperoxy Radicals From 2-butene, 3-methyl-<br />
1-butene, and MBO-232<br />
Matthew Voss<br />
Mentors: Mitchio Okumura and Leah Dodson<br />
Peroxy radicals play an important role in the chemistry of the troposphere. Substituted alkylperoxy radicals have<br />
not been as well-studied as unsubstituted alkylperoxy radicals. We directly detected chloroalkylperoxy radicals<br />
using pulsed laser cavity ringdown spectroscopy in the near infrared region (7100 cm -1 to 8850 cm -1 ), observing<br />
the A-X electronic transition and the O-O vibronic stretch. The chloroalkylperoxy radicals were generated from<br />
chlorine oxidation of the alkenes 3-methyl-1-butene, MBO-232, and 2-butene. The effect of the chlorine substituent<br />
was examined in comparison to the corresponding alkylperoxy spectra. In addition, the chloroalkylperoxy spectrum<br />
of 2-butene was compared against the corresponding nitrobutylperoxy spectrum to inspect the effect of different<br />
substituents.<br />
Characterization of Small-Molecules That Elevate Reactive Oxygen Species in Cancer Cell Lines<br />
Alex Wang<br />
Mentors: David Tirrell, Stuart Schreiber, and Drew Adams<br />
Cancer is a collection of genetic diseases united by a set of common traits, including limitless replicative potential,<br />
ability to metastasize, and oxidative stress. Recently, the small-molecule piperlongumine has been found to<br />
selectively kill cancer cells, apparently by increasing reactive oxygen species (ROS). As a result of isolating<br />
piperlongumine, a high-throughput screen of over 40,000 compounds identified 902 small molecules that increased<br />
ROS with piperlongumine as the positive control. The objective of this project was to further characterize these hit<br />
compounds as well as piperlongumine. The hit compounds were characterized using various methodologies such as<br />
toxicity in the EJ cell line, antioxidant rescue of cell death, superoxide staining, ability to synergize with<br />
piperlongumine, and auto-fluorescence. Furthermore, four known bioactive compounds were found to antagonize<br />
piperlongumine induced cell death, generating new hypotheses regarding the small-molecule’s biological<br />
mechanism of action. The results indicate that ROS increase does not guarantee cell toxicity and that likely only a<br />
minority of mechanisms causing enhanced ROS lead to cell death. These findings give insight into the relationship<br />
between ROS elevation and cancer cell mortality, providing a basis for future prioritization of the screening hits.<br />
Test Bench for a Retinal Prosthesis Neurostimulator IC<br />
Angie Wang<br />
Mentors: Azita Emami and Manuel Monge<br />
Patients affected by diseases like retinitis pigmentosa (RP) and age-related macular degeneration (AMD) incur<br />
severe vision loss as retinal photoreceptor cells degrade, causing less visual information to be transmitted to the<br />
brain. However, it is possible to bypass damaged photoreceptors and stimulate the remaining retinal neurons with<br />
visual information directly through the use of charge-balanced biphasic currents generated by a retinal prosthesis<br />
implant, restoring vision to the blind. A system to test the functionality of the current-based neurostimulator IC<br />
that drives the retinal implant’s 1024-channel electrode array is being developed. Digital logic blocks designed in<br />
VHDL and implemented on an Altera Cyclone II FPGA along with radio-frequency analog components are utilized to<br />
wirelessly transmit and receive data between the test system and the retinal implant. Image data from a 5megapixel<br />
camera placed in close proximity to the implant is encoded for use as stimulus control signals that are<br />
sent to the implant using a quadrature phase-shift keying (QPSK) modulation scheme. The resultant stimulus<br />
driver output is verified by the test bench.<br />
96
Expression of a Thermostable TPP-Dependent Enzyme and Characterization of Its Decarboxylase<br />
Activity<br />
Eric E. Wang<br />
Mentors: Frances H. Arnold and Kersten S. Rabe<br />
Biofuels such as isobutanol have shown promise as an efficient renewable energy source. One potential resource to<br />
produce these biofuels might be plant waste material, where the abundant resource cellulose can be converted<br />
through consecutive reactions to feed host cells to produce isobutanol. One important step in the reaction cascade<br />
is the decarboxylation of 2-Keto-isovalerate to Isobutraldehyde which can be catalyzed by the enzyme Ketoisovalerate<br />
Decarboxylase (KIVD) from Lactococcus lactis. However this enzyme is not active at elevated<br />
temperatures (e.g. 60° C) which presents an obstacle for application to several biotechnological areas. Especially,<br />
the chemical decomposition of plant waste material involves heat and acid treatment; therefore, a thermophilic<br />
host organism would be very beneficial. To adapt the non-thermostable KIVD protein to work at higher<br />
temperatures, libraries of mutant versions of the KIVD are screened for endurance at higher temperatures as well<br />
as higher activity.<br />
In order to establish a second, parallel approach, genes that code for proteins with a predicted similar function to<br />
KIVD have been identified in thermophilic organisms and the corresponding genes have been transformed into<br />
E. coli utilizing appropriate vector systems. The proteins will be overexpressed and tested for any KIVD related<br />
enzymatic activity.<br />
These experiments are part of the effort to establish a metabolic pathway in a thermophilic host to ultimately<br />
convert cellulose into isobutanol at elevated temperatures e.g. 60° C.<br />
Scattering of Giant Radio Pulses From the Crab Pulsar by the Interstellar Medium<br />
Jing Wang<br />
Mentors: Anthony Readhead and Glenn Jones<br />
The interstellar medium scatters radio pulses emitted by pulsars, resulting in elongated pulse profiles.<br />
Unfortunately, the existing models of interstellar scattering fail to explain features seen in pulses from the Crab<br />
pulsar. At high frequencies, the observed pulse decay times scale as f −2 rather than the f −4 predicted by the thin<br />
screen model. We confirm the f −2 trend in our wide bandwidth 2 GHz to 10 GHz observations from the Goldstone<br />
Apple Valley Radio Telescope. Using simulations, we explore possible causes of this anomalous decay time behavior<br />
and place geometrical constraints on the ISM and the Crab nebula. We believe this anomaly can be explained by<br />
scattering in localized structures specific to the Crab nebula. Further analysis could be done on possible scatteringbased<br />
explanations of non-exponential pulse microstructure.<br />
Optical Design Optimization and Testing of an Intrasurgical Microscope-Mounted Optical Coherence<br />
Tomography System<br />
Kening (Connie) Wang<br />
Mentors: Joe Izatt, Justin Migacz, and Scott Fraser<br />
Spectral domain optical coherence tomography (SD-OCT) is an imaging technique which obtains high-resolution<br />
cross-sectional images of the retina noninvasively, and is currently the clinical standard of care for diagnosis of<br />
many retinal pathologies. Several research groups are working to extend SDOCT technology for use during<br />
intraocular surgery, where it could aid in visualization of delicate retinal structures. A handheld scanner exists for<br />
intrasurgical uses, but this requires the surgical instrument and optical microscope to be moved away from the<br />
patient. The Duke group has previously prototyped a microscope-mounted SD-OCT system (MMOCT) that would<br />
allow for real-time imaging without disrupting the surgeon’s view of the surgical microscope. This project involved<br />
using a professional optical design software, ZEMAX, to model the existing microscope-mounted SD-OCT system<br />
and to recommend improvements upon that system. In addition, optical testing techniques were used to measure<br />
quantitative performance metrics of the existing system and recommended improvements in order to verify the<br />
accuracy of the model and validate recommended design changes.<br />
Verification of Protein Receptors Using Double Florescent Immunohistory on VENs<br />
Yuchen (Carrie) Wang<br />
Mentors: John Allman, Nicole Tetreault, Brian William, and Nina Ng-Quinn<br />
Located in fronto-insular cortex (FI) and anterior cingulate cortex (ACC) of the brain, von Economo neurons (VENs)<br />
play significant roles involving the integration of interoception, emotion, and cognition. VENs’ regulation of several<br />
biological processes is dependent on several protein receptors: Interleukin 4 receptor (IL4R), Interleukin 10<br />
receptor (IL10R), dopamine receptor 3 (DRD3), and serotonin 2b receptor (5HT2B). In this study, the expressions<br />
of these four receptors on VENs are established by chromatic immunohistory and double fluorescent<br />
immunohistory. IL4R is linked to IgE mediated immunity and mutations are related to asthma and other allergic<br />
reactions. Interestingly, there is a linkage between the activity of the VENs containing area and the immune<br />
response, including asthma. IL10R has paracrine and autocrine actions as an anti-inflammatory responder and<br />
97
egulates IL1B. The evidence that these two receptors are strongly expressed on VENs suggests that VENs may be<br />
involved in the regulation of the immune response. DRD3 is the highest affinity dopamine receptor. When there’s a<br />
negative feedback or loss, there is a reduction in the dopamine signal that the receptor detects, activating VENs.<br />
5HT2B is strongly expressed in the human stomach and intestines, and may signal danger or punishment.<br />
Verification that 5HT2B, and DRD3, are expressed on the VENs suggests that VENs are involved in the intuitive<br />
decision-making in a given situation.<br />
Alkoxy Radical Cavity Ringdown Spectroscopy and Kinetics: Direct Detection of HOC4H8OO•, HOC4H8•,<br />
HOC5H10OO• and HOC5H10•<br />
Marissa Weichman<br />
Mentors: Mitchio Okumura and Matthew Sprague<br />
Alkoxy radicals are closely tied to the HOx and NOx reaction cycles, which impact air quality in the lower<br />
atmosphere. As alkoxy radicals are common derivatives of atmospheric hydrocarbons, determining the kinetics of<br />
their reactions is of great interest in modeling the impact of carbon emissions on Earth’s climate. An alkoxy radical<br />
can react in one of three ways: isomerization (if it can form a transition state with at least six atoms),<br />
decomposition, or reaction with O2. The isomerization product (HOR•) quickly reacts with O2 to form a<br />
hydroxyalkylperoxy radical (HOROO•). In the current work, we study the n-butoxy and 2-pentoxy systems with IR<br />
cavity ringdown spectroscopy. n-butoxy is the smallest alkoxy radical that can isomerize, and thus makes a good<br />
model for basic isomerization chemistry; 2-pentoxy provides a model for the behavior of more complex alkoxy<br />
radicals, but is simple enough to have only one pathway to isomerization. For each system, we have measured the<br />
OH vibrational spectra of primary isomerization products (HOR•, HOROO•), and use the HOROO• spectra to<br />
determine the relative kinetics of the isomerization and O2 reaction pathways. We use kinetic and quantum<br />
mechanical modeling for prediction and analysis of data.<br />
Sub-Nyquist Sampling of Action Potentials<br />
Alex Wein<br />
Mentors: Lakshminarayan Srinivasan and David Rutledge<br />
In signal processing, the Nyquist rate is the minimum sampling rate necessary to reconstruct a continuous-time<br />
bandlimited signal from discrete time-series samples. For signals containing quick pulses such as neuron action<br />
potentials, the Nyquist rate can be very high - more than 10,000 samples per second. However, such signals often<br />
have what is called finite rate of innovation (FRI), meaning they can be described by a finite number of parameters<br />
per second. For instance, a stream of action potentials can be approximately described by the spike times and<br />
amplitudes, and the rest of the signal can be treated as noise. If we seek only to reconstruct the spike times and<br />
amplitudes, the theoretical lower bound on the necessary sampling rate drops to the rate of innovation, which is<br />
orders of magnitude lower than the Nyquist rate. We discuss a variety of existing sub-Nyquist sampling techniques<br />
for action potentials and present a new algorithm that combines maximum likelihood estimation with a fast<br />
iterative procedure in order to reconstruct neural signals.<br />
Acoustic Levitation and Field Induced Droplet Ionization Mass Spectrometry for Studies of Chemical<br />
Reactions in “Wall-less” Reactors<br />
Paul Adrian Weitekamp<br />
Mentor: Jesse Beauchamp<br />
An acoustic levitation device has been constructed to create an airborne droplet reactor designed specifically for<br />
mass spectrometric analysis of molecular species present in the liquid phase. An ultrasonic transducer is employed<br />
with a conical reflector to produce an acoustic field in which liquid droplets can be trapped, creating a “wall-less”<br />
reactor for studying chemical and biological processes in volumes that range from 1 nanoliter to several microliters.<br />
To sample molecular species present in the droplet, an electric field is introduced along the axis of levitation to<br />
effect field induced droplet ionization (FIDI) ionization. Analyte ions are extracted through an aperture in the<br />
reflector of the acoustic levitation instrument and analyzed by ion trap mass spectrometry. This methodology<br />
allows for rapid sampling of complex molecular species to monitor ongoing reaction in the suspended droplet. This<br />
avoids any interference which surface contact might introduce. This becomes especially important with the<br />
unfavorable surface area to volume ratios in nanodroplets. This setup also provides a unique opportunity to study<br />
the air-liquid interfacial chemistry of the droplet.<br />
98
Derandomizing AM Using a Multivariate Generalization of Parvaresh-Vardy Codes<br />
Benjamin Weitz<br />
Mentor: Chris Umans<br />
We give a new proof of local list-decoding properties of Reed-Muller codes, independent of the fact that their local<br />
codes are Reed-Solomon codes. The proof uses techniques from low-degree testing to compile many, possibly<br />
inconsistent low-degree codes into a single global codeword. It is our hope that the proof in this paper may be<br />
extended to Parvaresh-Vardy codes, and thus obtain a locally-decodable multivariate generalization of PV codes.<br />
Using existing theorems, this would give an optimal uniform hardness vs. randomness tradeoff for derandomizing<br />
AM.<br />
An Analysis of the Presence of Hydrogen Peroxide on Mars From Mariner 9 UVS Data<br />
Danika Wellington<br />
Mentors: Leslie Tamppari and Amanda Hendrix<br />
The ultraviolet signature of H2O2 may be present in reflectance spectra obtained by the Mariner 9 ultraviolet<br />
spectrometer (UVS). In this study we isolate the contribution from surface reflectance in the UVS data by modeling<br />
and accounting for the atmospheric contribution to the total reflectance spectrum. We use the DISORT radiative<br />
transfer code to take into account the absorption and scattering properties of atmospheric gases and aerosols<br />
relevant in the wavelength region of interest (200-350nm), in particular CO2 Rayleigh scattering, dust and cloud<br />
aerosol scattering and absorption, and the Hartley absorption band of ozone. An analysis of the surface spectrum<br />
allows the abundance of hydrogen peroxide to be characterized, as well as its seasonal and latitudinal variations.<br />
Additionally, we investigate the expected correlation of hydrogen peroxide with water vapor, from which peroxide is<br />
produced via photochemistry, and the expected anti-correlation of hydrogen peroxide with ozone, whose<br />
production is inhibited by H2O2 and other photochemical products of water vapor.<br />
Analysis of Performance for Adaptive Mesh Refinement<br />
Jody Wen<br />
Mentor: Dan I. Meiron<br />
Adaptive mesh refinement is a computational technique used to solve problems that change in local complexity. For<br />
this project, we utilize a set of C++ template class to develop a series of tests essential to performing operations<br />
on discrete functions defined on a quadtree (in 2D) or an octree (in 3D). These tests are then implemented to<br />
measure performance and scalability by timing the various operations and then analyzing how the time varies with<br />
the processor count of a distributed memory system. Specifically, we examine the efficiency of look-up costs, cost<br />
of storage and retrieval of information in the tree, and the ability to modify the tree dynamically as requirements<br />
for adaptivity vary or aspects of our function evolve.<br />
Characterizing a Resonator Bolometer Detector Array<br />
Rebecca Wernis<br />
Mentor: Jonas Zmuidzinas<br />
Kinetic Inductance Detectors (KIDs) are a new type of radiation detector currently being developed at Caltech<br />
based on superconducting resonators. Their simple fabrication process and capacity for high-density frequency<br />
multiplexing enable drastically larger array sizes than would be possible with traditional submillimeter radiation<br />
detectors. Currently, both direct detection and absorption via antennas provide a means for coupling radiation to<br />
the array. We are investigating placement of the resonators on bolometric islands as a new coupling mechanism.<br />
This should enable warmer operating temperatures and an ability to be fabricated with a wider variety of materials,<br />
making them promising candidates for a compact, on-chip spectrometer as well as for satellite-based observations<br />
of Earth. In order to be established as a viable candidate for such projects, this type of detector must be tested in<br />
a lab. An array was fabricated at the Jet Propulsion Laboratory and brought to Caltech to be tested. Key properties<br />
of this array such as the resonant frequencies and quality factors of the individual pixels and the intrinsic noise<br />
were measured under various loading conditions and temperatures. The results were compared with theory to<br />
provide a crucial understanding of the detector and a clear pathway for device optimization.<br />
Enhanced Absorption in Thin Film Photonic Crystal Silicon Solar Cells<br />
Kelsey A. Whitesell<br />
Mentors: Harry Atwater, Dennis Callahan, and Jonathan Grandidier<br />
Cost of solar cells can be significantly reduced if a silicon solar cell that requires significantly less material than<br />
current designs but still has competitive efficiency could be designed. Increasing the density of available optical<br />
states (DOS) across the bandwidth of the solar spectrum and populating those states is key to increasing<br />
absorption per volume. This research explores the potential of using thin (100nm-800m), two-dimensional photonic<br />
crystals (PC) as solar cells and couplers to achieve these goals. The best-enhancements thus far were achieved by<br />
99
a 2D PC with cone-shaped unit cells and an array of different coupled defects in a PC cell, which absorbed 2-3<br />
times more power and current than a planar silicon slab of the same thickness, while requiring only 50-60 percent<br />
of the volume of material per thickness.<br />
Spectroscopic Analysis of a Plasma for an Astrophysical Jet Experiment<br />
Hannah V. Willett<br />
Mentors: Paul Bellan and Vernon Chaplin<br />
We are investigating the properties of a small plasma source which will be used to improve the research potential<br />
of the Caltech astrophysical jet experiment. Radio frequency power sources are used to break down argon gas into<br />
plasma, but varying conditions such as pressure and applied magnetic field affects the ionisation fraction of the<br />
plasma formed. Spectroscopic analysis of the light emitted by the plasma allows us to calculate its ionisation<br />
fraction and study the effects of changing the breakdown conditions. Preliminary data has indicated that the<br />
apparatus functions correctly and that one theoretical model (local thermodynamic equilibrium) does not apply to<br />
the plasma. We have found the most effective way of supplying power to the plasma and looked briefly at the<br />
effects of varying both pressure and magnetic field. Improvements to the circuit which creates the magnetic field<br />
are being made and we are investigating the suitability of a second model (coronal equilibrium) for calculating the<br />
ionisation fraction. Once an appropriate model is found, additional data will be taken to study the effects of varying<br />
the plasma conditions in greater depth.<br />
Low-Temperature Seebeck Measurement System<br />
Alexander Wilson<br />
Mentors: Jeff Snyder and Nicholas Heinz<br />
There has been recent interest in low temperature thermoelectric devices, especially that of possible cryogenic<br />
thermoelectric cooling, due to their scalability and robust solid state nature. In order to try to develop such<br />
materials, their ability to turn heat to electricity, or vice versa, needs to be quantified. Thermoelectric materials are<br />
rated by their Figure of Merit (zT), which is a measure of a material’s overall thermoelectric efficiency. The Figure<br />
of Merit is comprised of the Seebeck coefficient (S), the electrical conductivity (σ), and the electronic and lattice<br />
portions of the thermal conductivity (κe+κl), in the form zT = S 2 σ/(κe+κl). The Seebeck coefficient is a property of a<br />
material that relates the voltage (ΔV) that develops due to a temperature difference (ΔT) in a material<br />
(S = ΔV/ΔT). The low-temperature Seebeck system described here is a simple low-cost setup designed to allow the<br />
measurement of the Seebeck coefficient of thermoelectric materials at cryogenic temperatures without rare or<br />
expensive components. With relatively cheap parts, basic instruments, and minor programming, this lowtemperature<br />
Seebeck measurement setup is intended to be able to be easily constructed and modified for<br />
necessary requirements.<br />
Hydrogen Physisorption at High Pressure<br />
Andrew Wilson<br />
Mentors: Brent Fultz and Nicholas Stadie<br />
One possible method of storing hydrogen for an automotive application is physical adsorption (physisorption) on<br />
the surface of a porous material. Materials have been developed which show attractive hydrogen uptake and<br />
performance characteristics at cryogenic temperatures, but uptake for physisorbent materials at ambient<br />
temperature is typically less than 1 weight%, and this is not an improvement over compressed gas at the same<br />
pressure. It was reported in 2009 that zeolite-templated carbons (ZTCs) show remarkable hydrogen uptake at high<br />
pressures (10 to 34 MPa). No attempt has so far been made to reproduce this due to the difficulty of measuring<br />
physisorption at such high pressures. The Fultz group at Caltech has constructed an instrument to measure<br />
hydrogen uptake by the Sieverts method at pressures up to 70 MPa, in order to corroborate these results and to<br />
investigate the little-understood mechanisms governing high-pressure physisorption. The instrument has been<br />
commissioned up to 40 MPa, and initial adsorption measurements for the standard commercial adsorbent Maxsorb<br />
MSC-30 are consistent with literature results. I shall report hydrogen uptake measurements for MSC-30 and a<br />
zeolite-templated carbon.<br />
Sorting on the Surface of DNA Origami<br />
Sarah Wittman<br />
Mentors: Erik Winfree, Lulu Qian, Damien Woods, and Niranjan Srinivas<br />
Recently, DNA has shown itself to be a viable medium for use in developing computationally complex systems.<br />
Much work is being done to formulate ways of exploiting DNA’s natural properties and behaviors to form a new<br />
model of computation. One of the more computationally fundamental methods is the ability to sort and arrange<br />
data structures. In the case of DNA computation, this data is represented as a physical object with some feature<br />
indicative of the data it represents. Now, consider a 2D plane. On this plane is a random assortment of objects that<br />
have some meaning to us, something that we take some interest in keeping track of. Our goal for the summer is to<br />
100
develop and test a DNA system that is capable of retrieving and sorting these objects in a meaningful way, to<br />
provide some order to what is currently a relatively chaotic system, and give us some ability to track these objects<br />
of interest.<br />
Electrochemical Studies of Mono-, Bi-, and Trimetallic Glyoxime Complexes<br />
Justin M. Wolfe<br />
Mentors: Jonas Peters and Chris Uyeda<br />
Several monometallic macrocyclic glyoxime complexes have been shown to electrocatalyze the reduction of various<br />
substrates, and this project has analyzed the influence of additional metals. Electrochemical studies were<br />
conducted on 1) a monometallic copper complex, 2) related bimetallic species constructed from the copper<br />
complex with iron or zinc, 3) trimetallic complexes involving nickel dimethyl glyoxime and either copper or zinc.<br />
Their reactivity towards protons and carbon dioxide in the presence of an electric current was analyzed.<br />
Dynamically Optimized Sequential Experimentation (DOSE) for Estimating Time Preference Parameters<br />
Christina A. Wong<br />
Mentors: Colin F. Camerer and Stephanie W. Wang<br />
Time preference refers to the phenomenon of placing higher value on immediate gains than on future gains. Much<br />
literature exists evidencing the ubiquity of time preference; however, the current models of time’s effect on utility<br />
are not well defined and not generalizable to larger economic situations. To create a more robust model,<br />
parameters such as discount rate must be more accurately estimated.<br />
In order to efficiently gather the necessary data, we designed and coded a DOSE system to dynamically analyze<br />
the data during its collection, allowing us to ask subjects questions that will provide us more information. The<br />
program begins with an estimate of parameters based on existing expectations and updates that estimate with<br />
each question that the subject answers. The system chooses each question by determining how useful the answer<br />
will be in generating a finer estimation. The three most common models are the exponential, hyperbolic, and quasihyperbolic<br />
models. We plan to determine which of these three models fits best with our results.<br />
Prediction of Bubbles and Crashes Through Classification of Traders in an Experimental Market<br />
Katherine M. Wong<br />
Mentors: Colin F. Camerer and Alec C. Smith<br />
Two phenomena that may occur in a market environment are bubbles and crashes. Bubbles arise when traders buy<br />
assets at prices which are greater than fundamental values. Crashes occur when traders cease to buy assets at a<br />
higher value than the fundamental value, and prices fall to fundamentals. The idea that bubbles and crashes may<br />
occur even as a result of rational trading has been well-established in the literature. We have experimentally<br />
simulated markets that allow traders full information of fundamental value and where bubbles and crashes can be<br />
observed.<br />
By classifying traders based on their trading behaviors during the experiment, we fit a model of how the evolution<br />
of bubbles and crashes depends on the makeup of the trader population. Traders are classified as one of three<br />
main types: feedback traders, who trade based on past changes in market prices; passive traders, who trade<br />
based on fundamental values; and rational speculators, who trade based on estimates of future prices. We run<br />
simulations of markets composed of corresponding proportions of each type of trader to check that the simulations<br />
yield bubble and crash patterns similar to those observed in the experiment.<br />
On the Choice Number of the Hoffman-Singleton Graph and the Union of a Graph With a Matching<br />
Nicholas D. Woodward<br />
Mentor: Niranjan Balachandran<br />
The choice number of a graph is the smallest number k such that any assignment of color lists of size k to the<br />
vertices of the graph allows for a proper vertex coloring. We provide bounds on the choice number of the wellknown<br />
Hoffman-Singleton graph. Alon and Tarsi’s seminal paper on oriented graphs and graph colorings provides<br />
foundation for this study. While studying the Hoffman-Singleton graph another question arose: could the addition<br />
of a matching to a graph increase the choice number by at most one? We explain the interest in this question as<br />
well as explain the difficulties in understanding it.<br />
101
Visualizing Intracellular Protein Trafficking and Chaperoning of α4β2α5 nAChR Protein Subunits in N2A<br />
Cells and Primary Cortical Neurons With Fluorescent Confocal Microscopy<br />
Adela Wu<br />
Mentors: Henry A. Lester and Crystal H. Dilworth<br />
Smoking is a major public health concern and contributes to many premature deaths. Recent human genomic<br />
studies attempting to classify possible factors for nicotine addiction have identified several single nucleotide<br />
polymorphisms (SNPs) that contribute to smoking-related behaviors. One particular SNP involves an asparagine-toaspartic<br />
acid substitution in the nicotinic α5 subunit at amino acid position 398 (rs16969968; α5(N398)). This<br />
α5(N398) mutant variant may be associated with an increased risk for greater nicotine use.<br />
In humans, the α5 subunit is expressed as pentameric α4β2α5 nicotinic acetylcholinergic receptors (nAChRs) in the<br />
central nervous system and as α3β4α5 nAChRs in the periphery. Of particular interest to us is the mechanism by<br />
which α4β2α5 nAChR protein subunits influence pathological states. We aim to answer two questions—first,<br />
whether inclusion of a wildtype α5 subunit into α4β2 receptors show differences in intracellular protein trafficking;<br />
and, second, whether inclusion of a variant α5(N398) subunit into α4β2 receptors show differences in intracellular<br />
protein trafficking. With expression of fluorescently-labeled receptor subunits in both N2A cells and primary mouse<br />
cortical neurons and imaging using confocal microscopy, we are able to visualize the localizations and life cycles of<br />
these nAChRs’ assembly within the cell.<br />
Growth and Characterization of Crystalline Titanium Dioxide Thin Films for Photocatalytic Applications<br />
Anjian Wu<br />
Mentors: Harry A. Atwater and Christopher T. Chen<br />
Anatase titanium dioxide (TiO2) is a large band gap semiconductor with many appealing properties such as its<br />
durability and chemical stability. Conformal TiO2 coatings on glass, silicon (111), and Si microwire arrays<br />
substrates were investigated because of this metal oxide’s powerful oxidiation capabilities when illuminated with UV<br />
light and fast charge transfer kinetics. These properties make TiO2 a worthy candidate for applications such as<br />
water splitting, dye sensitized solar cells, and artificial photosynthesis. TiO2 thin film samples were prepared using<br />
sol-gel deposition with different recipes, substrates, and annealing parameters. The films were characterized with<br />
scanning electron microscopy (SEM), atomic force microscopy (AFM), x-ray diffractometry (XRD), and Hall Effect<br />
measurements. We found that the crystallinity and phase of the films are mainly influenced by annealing duration<br />
and temperature, while the quality and morphology of the films can be controlled by the spin-coating speed,<br />
number of coated layers, sol-gel composition, and substrate type. Anatase films were observed around annealing<br />
temperatures of 500◦ C for at least one hour. The highest quality films were found at spin-coat speeds of 3000 RPM<br />
of fewer than three layers with sol-gels composed of lower vapor pressure alcohols.<br />
How to Tile a Rectangle With Rectangles of Algebraic Side Ratios<br />
Jianqiu Wu<br />
Mentor: Richard Wilson<br />
The tiling problem has been around for more than 100 years. A reasonable question we can ask is how to tile<br />
rectangles with rectangles. A recent paper studies this problem through its close connection to planar electrical<br />
networks and alternating circuits. The paper presented three necessary conditions for a rectangle to be tilable by<br />
rectangles of ratios For , these conditions are proved sufficient, but for , whether they are<br />
sufficient remains a conjecture. My focus was not to solve this problem because one of the conditions requires a<br />
rational complex function with integer coefficients to be positive-real, which means the function takes nonnegative<br />
real parts when all inputs have nonnegative real parts. This is not easy to characterize. Instead I studied a lot of<br />
examples of specific algebraic ratios. The most important results are the rectangles tilable by<br />
where are distinct square-free integers and those tilable by ,<br />
where is a cube-free integer.<br />
Spin Polarized Scanning Tunneling Microscopy Studies of Graphene in Finite Magnetic Fields<br />
Teng-Pao (Renee) Wu<br />
Mentor: Nai-Chang Yeh<br />
When graphene is in the presence of a magnetic field, it displays what is known as the anomalous quantum Hall<br />
effect (QHE). The QHE is observed in two-dimensional electronic systems, when the orbits of electrons under a<br />
magnetic field are quantized at discrete energy values known as the Landau levels. Scanning tunneling microscopy<br />
(STM) techniques were used to resolve the issue of whether applying a magnetic field to graphene lifts spin<br />
degeneracy or pseudospin degeneracy. STM studies of the tunneling conductance spectra of graphene under finite<br />
magnetic fields applied along the c-axis were done using two different types of tips: regular Pt/Ir and Cr-coated<br />
Pt/Ir tips. The two different types of tips allow for both regular STM and spin-polarized STM (SP-STM) to be<br />
performed. Comparisons of the two types of spectroscopic studies were used to reveal which degree of degeneracy<br />
is lifted under the presence of an external magnetic field.<br />
102
Engineering Thermostable Synthetic Cellulase Mixtures<br />
Timothy Wu<br />
Mentor: Frances Arnold<br />
In a world where fossil fuels are rapidly dwindling, the production of biofuels offers a promising partial solution to<br />
the alternative energy problem. Previously, novel thermostable enzymes were created which operated individually<br />
at a higher temperature than their natural counterparts. However, in nature, organisms create mixtures of<br />
enzymes that act synergistically; this results in a higher net activity than the cumulative activities of each<br />
component enzyme in an equivalent concentration. In this experiment, the synergistic effects of substituting<br />
natural enzymes with thermostable chimeras in a minimal two- and three-enzyme mixture are examined. It is<br />
hypothesized that, at higher temperatures, the effect of synergy will increase due to greater diffusion of enzymes<br />
on the substrate surface. The relationship between activity and temperature of the enzymes, both individually and<br />
in mixtures, is examined, and a relationship between synergy and temperature is investigated.<br />
NuSTAR Pixel Characterization<br />
Yue Wu<br />
Mentors: Fiona Harrison and Varun Bhalerao<br />
NuSTAR, a hard x-ray focusing telescope launching into orbit early February 2012, consists of two parallel<br />
telescopes and two focal planes with four pixelated detectors each. Pixel-to-pixel variations in the detectors arise<br />
from CZT (Cadmium Zinc Telluride) impurities, grain boundaries and imperfect readout contacts. We developed IDL<br />
(a programming language) code to characterize these pixels using data from x-ray generator scans in the lab. We<br />
look at the regions in each pixel that detect events in neighboring pixels and those which detect events only inside<br />
that pixel. We also examine pixel edge effects on x-ray spectra and account for noise in order to map the effective<br />
pixel boundaries for all detectors.<br />
Extracellular Sulfatase-2 Levels Are Differentially Expressed in Idiopathic Pulmonary Arterial<br />
Hypertension<br />
Zhaoying Xian<br />
Mentors: Marlene Rabinovitch, Christopher Rhodes, and Dennis Dougherty<br />
Objective: An early event in the pathology of idiopathic pulmonary arterial hypertension (IPAH) is the apoptosis<br />
and dysfunction of endothelial cells (ECs). The goal of my project was to determine whether changes in EC gene<br />
expression might provide novel insights into the pathobiology of IPAH, and thus lead to better treatments. One<br />
candidate, extracellular sulfatase 2 (SULF2) regulates heparan sulfate proteoglycan (HSPG) function. HSPGs affect<br />
the activity of growth factors such as wingless (Wnt), which is important for cell maintenance and survival.<br />
Methodology: Twenty-four genes differentially expressed in ECs from control vs. IPAH patients were identified by<br />
RNA-Seq analysis, and a subset confirmed by quantitative PCR. EC protein expression was investigated by<br />
immunoblotting and immunofluorescence.<br />
Results: RNAseq identified six genes of potential relevance to IPAH in IPAH-ECs. Of those six, qPCR analysis<br />
confirmed only that SULF2 mRNA was downregulated (p=0.0015). Protein levels of SULF2 were, however,<br />
upregulated (p=0.013), suggesting increased stability of the protein despite low mRNA levels.<br />
Conclusions: An increase in SULF2 in IPAH-ECs may cause more HSPGs to release or activate growth factors, thus<br />
leading to increased proliferation and survival of pulmonary vascular cells, as seen in the pathology of IPAH.<br />
Testing the Functional Capacity of Long Distance Candidate Regulatory Elements<br />
Guoning Xiao<br />
Mentors: Barbara Wold and Katherine Fisher-Aylor<br />
All cells in an organism have the same DNA; gene regulation is responsible for the difference between various cell<br />
types. The general mechanism of this is thought to be protein transcription factors binding to cis-regulatory<br />
modules (CRMs) of DNA and then interacting with the promoter of the gene, causing a change in gene expression.<br />
This project attempts to explore the interactions of candidate cis-regulatory modules (CRMs) more deeply by<br />
determining the relationship between CRMs associated expression pattern of a gene in an artificial transfection<br />
assay. First, candidate CRMs are selected using ChIP-Seq data. C2 muscle cells are cultured and transfected with<br />
bacterial plasmids that contain the candidate CRMs driving the reporter gene luciferase. Transfection results will be<br />
collected by reading the amount of luciferase produced by the cells. After collecting the transfection results, the<br />
goals are 1) to use the data from the experiments to determine if connected candidate CRMs are more likely to<br />
drive expression than the gene with no CRM, and 2) to compare the expression patterns of the in vitro plasmid<br />
with the patterns of the nearby gene in vivo.<br />
If the CRMs are more likely to drive expressions, then the candidate CRM is a transcriptional enhancer. If<br />
expression decreased, the candidate CRM can function as a silencer. If the expression level of the gene stays the<br />
same, there is no conclusive result about the “secret” of expression.<br />
103
Profiling the Proteomic Response to the Incorporation of Synthetic Amino Acids in E. coli With Pulsed<br />
SILAC<br />
Yushu Joy Xie<br />
Mentors: David Tirrell and John Bagert<br />
Proteomics is a largely increasing field of study, as it accounts for translation-regulating elements within a cell,<br />
such as small-RNAs and micro-RNAs whereas genomics does not. Recently, a proteomics method called bioorthogonal<br />
non-canonical amino acid tagging (BONCAT) was developed that allows the separation and identification<br />
of newly synthesized proteins in response to internal and external cues, permitting a more sensitive analysis of<br />
cellular response. The process utilizes the synthetic methionine surrogate azidohomoalanine (Aha) which, following<br />
a pulse, is incorporated into protein synthesis by the endogenous cellular translational machinery. Aha serves as a<br />
chemical handle that can be selectively conjugated to affinity tags with the copper-catalyzed azide alkyne<br />
cycloaddition, allowing separation of newly made proteins from old. This method has the potential to be a useful<br />
tool in proteomics and other metabolic labeling applications such as in situ fluorescence labeling, but it is still<br />
unclear whether the use of non-canonical amino acids alters natural protein synthesis. In this study we use pulsed<br />
stable isotope labeling by amino acids in cell culture (pSILAC) to analyze the proteomic effects of incorporating the<br />
synthetic amino acid Aha into the cellular metabolism of Escherichia coli.<br />
Towards the Production of a BMP Morphogen Gradient<br />
Melissa Xu<br />
Mentors: Michael Elowitz and Nagarajan Nandagopal<br />
Bone morphogenetic proteins (BMPs) are essential for regulating proper patterning during embryonic development<br />
through creation of a morphogen gradient. Though the mechanisms of this process are unknown, it is possible that<br />
the formation of a BMP4 gradient is comparable to that of its homologue Dpp in Drosophila. The Dpp gradient is<br />
created by spreading of Dpp from a central stripe of cells in the imaginal wing disc, aided by interactions with the<br />
cell-surface protein Dally, a Drosophila glypican. Using BMP4-mCherry secreting cells and BMP4 responsive cells,<br />
we aim to create a BMP4 gradient in Chinese Hamster Ovary cells (CHO). Under the assumption that glypican-3<br />
interacts with BMP4 in the same fashion as its Drosophila homologue Dally and Dpp, we use glypican-3 to aid in the<br />
sequestering and shuttling of BMP4 along the cell surface. The successful creation of a BMP gradient using<br />
glypicans would allow us to further understand BMP signaling and its role in embryonic development.<br />
Terahertz Imaging With a 4x4 Pixel Focal Plane Imager in 0.13µm SiGe BiCMOS Technology<br />
Lita F. Yang<br />
Mentors: Ali Hajimiri and Kaushik Sengupta<br />
Terahertz (THz; 10 12 Hz) radiation exhibits high molecular specificity and high penetration capabilities, allowing for<br />
applications in nondestructive package inspection and various biomedical imaging applications such as displaying<br />
contrast between diseased and healthy tissue. In this project, we present a low-cost, compact THz imaging system<br />
which can be readily implemented for clinical or outdoor applications. A 4x4 pixel focal plane THz imager for 0.25-<br />
0.3 THz radiation has been fully integrated in low-cost 0.13 µm SiGe BiCMOS process technology. Without the use<br />
of external silicon lens or post-processing techniques, a pixel of the imager achieves high responsivity and low<br />
noise equivalent power (NEP). Imaging of objects through optically opaque materials with transmission-mode<br />
imaging demonstrates the potential for future cost-effective security screening applications. Biomedical THz<br />
imaging in transmission and reflection modes were also demonstrated with dehydrated skin and tissue samples.<br />
Tissue and water distributions of different regions of the tissue samples were clearly resolved, which exhibits the<br />
system’s potential in reflecting tissue conditions and for analysis of histological tissue features.<br />
Durability of Superhydrophobic Carbon Nanotube Arrays Under Various Aqueous Conditions<br />
Perren Yang<br />
Mentors: Morteza Gharib and Adrianus Aria<br />
Recent study by Adrianus Aria and Morteza Gharib showed that carbon nanotube (CNT) arrays exposed to heat<br />
under vacuum develop extreme hydrophobicity. My project investigates the durability of this quality; a durable<br />
superhydrophobic surface opens many avenues of technological development, ranging from specialized nonstick<br />
and drag-reducing coatings to sterile surfaces and novel fluid mechanics testbeds. My first experiment examines<br />
the decay of superhydrophobicity while CNTs are submerged in water, and whether this time might be extended by<br />
periodic removal into the air. Having established a baseline, I examine the changes in the rate of decay under<br />
increased pressures. Experiments with differing temperature, dissolved solutes, and dissolved air will also be<br />
conducted.<br />
104
A Flip-Out Method to Obtain Drosophila Transformants With a Mitochondrial Killer Construct<br />
Ran Yang<br />
Mentors: Bruce Hay and Nikolai Kandul<br />
Mitochondria are known to undergo fission and fusion to maintain healthy genomes, but genetic pathways<br />
regulating these mechanisms are not well characterized. Furthermore, the mitochondrial bottleneck suspected to<br />
serve as a natural screen to discard unhealthy mitochondria is not well understood. A cassette-regulated construct<br />
that uses a constitutive promoter to drive a restriction enzyme tagged for mitochondria was designed to control<br />
when deletions are induced in the mitochondrial genome. Markers were used to test for and identify major<br />
components of the construct; RFP and GFP showed the corrected function of the promoter and cassette in cell<br />
culture and after the constructs were injected into flies, mitochondrial deletions were screened using PCR. Previous<br />
work in this lab has suggested that not only can we control when deletions occur, but that these mitochondrial<br />
deletions can be passed through the germline to future generations. The cassette-based constructs have provided a<br />
more controlled environment for inducing deletions, which can then be used in conjunction with gene knockdown to<br />
test the function and relation of genes thought to regulate mitochondrial health, human aging, and Parkinson’s<br />
Disease.<br />
Combining Radial Velocity Measurements and High Contrast Imaging to Characterize Long Period<br />
Planetary Systems<br />
Scott Yantek<br />
Mentor: Justin Crepp<br />
Radial velocity (RV) measurements are the most successful ground-based method of detecting extrasolar planets.<br />
This method relies on a star’s motion about the center of mass of a star-planet system. Lines in a star's spectrum<br />
are shifted in wavelength periodically, and from the amplitudes of these oscillations we calculate a lower limit for<br />
the mass of a companion. Currently, there exist vast collections of radial velocity measurements, but in analyzing<br />
this data, the focus has generally been on shorter period RV trends (up to a few months), so the possibility of<br />
detecting longer period companions has not been fully explored. Many RV trends with periods of a few years or<br />
more, such as those caused by planets farther from their stars, may not have been recognized. We have used<br />
existing radial velocity data, among other resources, to identify and rank stars with promising longer period RV<br />
trends. The best of these candidates will be observed using high-contrast imaging with NIRC-2 at Keck. The quality<br />
of these images taken in the L band ( =3.8 m) will be improved by the inclusion of a recently developed method<br />
of background subtraction (background noise is high in this wavelength range). With these observations we will<br />
fully characterize the orbits of these sub-stellar companions, calculating their actual masses, which will significantly<br />
contribute to improving planetary atmosphere models by fixing mass independently of spectra (using dynamics)<br />
and removing it as an unknown.<br />
Deciphering How Mitochondria Activity Supports and Modulates Nervous System Function in the<br />
Nematode Caenorhabditis elegans<br />
Alex Yeh<br />
Mentors: Paul Sternberg and Amir Sapir<br />
Mitochondria are the powerhouses of cells and play a crucial role in the functioning of organisms. To study the<br />
molecular circuits regulating mitochondria activity in nerves and muscles as well as the consequences of<br />
mitochondria dysfunction, we knock down genes of interest through RNA interference (RNAi) and then analyze the<br />
effect on the nervous and muscular systems’ functions. With our collaborators at Tel Aviv University, we identified<br />
and focused on one gene that is localized to the mitochondria of muscles and specific neurons in a stage-specific<br />
manner. Knocking down this gene resulted in age-dependent attenuation of worms’ movement, pumping, and egglaying:<br />
activities that presumably stem from muscular or neuronal dysfunction. Surprisingly, inhibiting<br />
mitochondria respiration and inducing mitochondria stress by knocking down atp-3, a gene which codes for<br />
adenosine triphosphate synthase subunit, resulted in ectopic translocation of the protein to the mitochondria. The<br />
age-dependent localization and stress-dependent recruitment of the protein to mitochondria suggest that this<br />
protein is required for the normal function of mitochondria. Future plans include performing behavioral assays that<br />
can pinpoint the genetic factors encoding proteins that interact with this protein in mitochondria. A further goal is<br />
to find additional molecules and circuits required to prevent age-dependent mitochondrial dysfunction in the<br />
nervous system and muscles.<br />
The Development of Luminescent Probes for DNA Base Mismatches<br />
Jessica S. Yeung<br />
Mentors: Jacqueline K. Barton and Anna McConnell<br />
The accuracy of DNA replication is vital to cellular development and survival. Failure of the cell to repair DNA bases<br />
mismatches arising from polymerase errors or DNA damage can lead to diseases such as cancer. Therefore, the<br />
development of a luminescent probe targeting DNA base mismatches may lead to a diagnostic tool for cancer. A<br />
successful fluorescent probe ideally will act as a light switch for DNA mismatches and only fluoresce when bound to<br />
a mismatch. Therefore, the probe must be highly fluorescent and selective for mismatched DNA in comparison to<br />
105
matched DNA. Two ruthenium complexes based on the 5,6-benz[c]acridinquinone (acri) ligand, [Ru(bpy)2(acri)] 2+<br />
and [Ru(phen)2(acri)] 2 , have been synthesized, characterized, and investigated as potential luminescent probes.<br />
Fluorescence studies in different solvents reveal that both complexes are fluorescent in acetonitrile, but not in<br />
water suggesting light switch behavior. The fluorescent and binding properties of both complexes were examined<br />
by further florescence studies and binding studies with matched and mismatched DNA.<br />
Analysis of the Electrical Characteristics of Micro-Superconducting QUantum Interference<br />
Devices (μ-SQUIDs) Fabricated Under a Trilayer Nb-AlOx-Nb Process<br />
Joshua Yoon<br />
Mentors: Kathryn Moler and Keith Schwab<br />
Superconducting quantum interference devices (SQUIDs) are known for its high field sensitivity, which has allowed<br />
magnetic and electronic materials research to take a leap forward in pursuit of new discoveries in many different<br />
areas. Although there is a growing number of books and articles that well-describe intrinsic properties of SQUIDs, it<br />
is of our interest to look at both its characteristic current-voltage (I-V) and flux-current (Φ-I) curves in our efforts<br />
to better understand its relationship with its unique fabrication process. Since SQUIDs reach a superconducting<br />
state only at low temperatures, we immerse it in liquid 4 He, and using a four-terminal configuration, we apply a<br />
sinusoidal current through the device loop and measure its voltage. By extracting its primary electrical parameters<br />
from the data gathered, the modulation amplitude tells us the size of the signal available from the SQUIDs and the<br />
I-V curves provides us with critical current values and its dependence on external flux. Data analysis show critical<br />
current vs. applied flux curves which are shifted relative to each other for certain devices, indicating asymmetry in<br />
the junction arms. In addition, critical current values seem higher for some devices than others. These results<br />
indicate that the SQUIDs that were tested underwent a non-uniform fabrication process, consequently producing<br />
undesirable devices that have either too high or low of a critical current or too small of a modulation amplitude. By<br />
adjusting the fabrication procedure in some way, which is yet to be determined, a higher yield of SQUIDs with a<br />
critical current range of ~10-15uA is possible in the future.<br />
Statistical Genomics of MRSA Infection and Disease Progression<br />
Yanwen You<br />
Mentors: Richard Wilson, Michael Wendl, and Katalin Fejes Toth<br />
Community associated Methicillin-resistant Staphylococcus aureus (CA-MRSA) infections are difficult to treat and<br />
pose a serious health concern. Some special cases escalate into a life-threatening infection. We hypothesize that<br />
the outcome of community MRSA acquisition is determined by interactions between the genetic composition of the<br />
infecting strain and polymorphisms in relevant host genes involved in immune and inflammatory responses.<br />
Samples of bacteria DNA and host DNA were collected from 6 sites in the U.S. Each sample was sequenced and<br />
compared to the reference genome to find small mutations in both the bacteria and its host. No significant<br />
differences were found among the samples of bacteria, indicating that the host genome is wholly responsible. The<br />
data is fitted with a logistic regression, controlled for co-variates, to determine the genes that contribute to the<br />
severity of the infection. Based on a preliminary data set of 49 human samples, the analysis returned several<br />
suspected genes. But with only 49 samples, any statistical test is underpowered. Much more data is required<br />
before we see a statistically significant effect and draw conclusions.<br />
Intracellular Trafficking of Nicotinic Receptors (α4β2)<br />
Caroline Yu<br />
Mentors: Henry Lester and Chris Richards<br />
Understanding the mechanisms of nicotine addiction is key to preventing tobacco-related diseases. Nicotine is<br />
known to upregulate specific subtypes of nicotinic acetylcholine receptors (nAChRs), leading to increased<br />
expression of receptors on the plasma membrane. The mechanism of upregulation is not well understood, but<br />
nicotine may affect rates of intracellular trafficking rates. Photoactivatable fluorophores were employed to observe<br />
trafficking of α4β2 subunits in neuroblastoma (N2A) cells from ER exit sites, and the Trans Golgi Network, as well<br />
as to quantify membrane turnover rates. Interestingly, nicotine was found to downregulate dynamin in the absence<br />
of nAChRs, but upregulate dynamin in the presence of α4β2. These observations support the hypothesis that<br />
nicotine affects protein expression at the plasma membrane likely through modification of intracellular trafficking.<br />
Low Frequency Impedance Measurement: Primary Coil of the Retinal Prosthesis System<br />
Chia-Chen (Debbie) Yu<br />
Mentors: Yu-Chong Tai and Yu Zhao<br />
Research shows that the retinal prosthesis has great potential for treating blindness diseases due to outer retinal<br />
degeneration and helping patients regain partial vision. Our retinal prostheses system incorporates an extraocular<br />
and an intraocular component that are inductively coupled to transfer data and power efficiently. This work<br />
presents the construction and characterization of the primary coil, which serves as the wireless power transferring<br />
link in the extraocular unit. Primary coils of different material parameters and winding methods have been<br />
constructed and characterized. In addition to standard 2-point and 4-point impedance measurements, a self-built<br />
106
4-point measurement setup is built to correctly characterize the coil parameters. A quality factor of 116 is reached<br />
at the operating frequency of 10 MHz with the braided litz coil design. Different winding methods of the coil and<br />
compensation of the data are the next steps to optimize the system’s performance. Ultimately, we wish to build a<br />
complete working primary circuit that minimizes power loss for the retinal prosthesis system.<br />
Capture Agent for the Akt2 Protein<br />
Mary Beth Yu<br />
Mentors: James Heath, Arundhati Nag, and Samir Das<br />
The protein Akt2 is associated with a signaling pathway for cell survival and overexpressed in some cancers. An<br />
inexpensive compound that could detect Akt2 in a sensitive and selective fashion could assist in early detection of<br />
cancer. Previously, a capture agent had already been found for a chemically synthesized segment of this protein.<br />
Here, we attempted to improve on this capture agent, two hexapeptides connected by a triazole, by adding a third<br />
hexapeptide. A biligand anchor, which is the extant capture agent with an azide added to the end, was synthesized.<br />
Biologically produced Akt2 and the biligand anchor were incubated with a C terminal alkyne amino acid containing<br />
library of hexapeptides. Then, library peptides that were both bound to Akt2 and formed triazoles with the anchor<br />
were isolated. In the future, the affinity and specificity of the resulting triligand for the Akt2 should be tested.<br />
Almost Strongly Full-Ranked Matrices Over Finite Fields<br />
Chi Ho Yuen<br />
Mentor: Richard Wilson<br />
An almost strongly full-ranked matrix is a t � u matrix with the property that one and only one of its t � t<br />
submatrices is singular. In this project, we study almost strongly full-ranked matrices over finite fields, as they<br />
appear in some linear algebraic constructions for certain combinatorial designs.<br />
Given a fixed finite field F and an integer t , denote by<br />
q<br />
u �q, t�<br />
the exact upper bound of u for which a t � u<br />
almost strongly full-ranked matrix over F exists. Using tools from finite geometry, number theory, linear algebra<br />
q<br />
and combinatorics, the structure of almost strongly full-ranked matrices over finite fields is studied and upper<br />
bounds and lower bounds for �qt� of u �q, t�<br />
for small q and t .<br />
u , are provided. Also, a computer search has been done to find the exact values<br />
The existence of large almost strongly full-ranked matrices over finite fields has been used to give some<br />
combinatorial designs that answer several problems in design theory and finite geometry.<br />
The Effect of Strategic Exits on the Incumbency Effect<br />
Jenny Z. Yung<br />
Mentor: Jonathan Katz<br />
Although elections in America are known to be democratic, statistics have shown that incumbents have an<br />
advantage in elections. However, previous analysis failed to take into account strategic exits. It’s possible that<br />
these incumbents win more because they only join races in which they are more likely to succeed, inflating the<br />
incumbency effect. Using a dataset of open US House of Representatives seats from the last 7 decades, we used<br />
statistical analysis to evaluate the difference between voluntary and involuntary exits and to determine the effect<br />
strategic entry and exits. We find that strategic exits have a significant effect on the incumbency effect.<br />
Predictive Statistical Modeling of piRNA Biogenesis From Progenitor Transcripts<br />
Alexander Zahn<br />
Mentor: Alexei Aravin<br />
Piwi-interacting RNAs (piRNAs) form a relatively recently discovered class of small non-coding RNAs that are critical<br />
to the suppression of transposon activity. piRNAs are known to comprise of short, twenty to thirty nucleotide<br />
subsequences of much longer non-coding single stranded RNA precursor transcripts through a poorly understood<br />
mechanism. We observe that the distribution of piRNA production events ( a “reads spectrum”) over precursor<br />
strands exhibits highly deterministic behavior, and in particular that the abundance of piRNA production at a given<br />
transcript position is a combined function of the immediately surrounding sequence of approximately twenty<br />
nucleotides as well as much longer sequences many hundreds of nucleotides distant along the transcript. We<br />
algorithmically analyze reads spectra in mice and Drosophila in search of patterns correlated with varying levels of<br />
piRNA production at particular transcript positions in hopes of forming a predictive model for piRNA production and<br />
also develop a quantitative metric for the level of similarity between arbitrary spectra over the same precursor for<br />
the purposes of model evaluation.<br />
107
Experimental Verification of the Theory of Dynamic Market Completeness<br />
Eric Zhang<br />
Mentor: Peter Bossaerts<br />
We sought to experimentally compare prices and portfolio choices for complete and incomplete asset markets, as<br />
explained in “Dynamically Complete Experimental Asset Markets” (Bossaerts, Meloso, and Zame). In this<br />
experimental setup, we created an incomplete market by shutting down one market and prohibiting trade of that<br />
asset. However, there was one additional provision: halfway through the trading process, we would announce that<br />
one of three possible future states would not occur. This additional flow of information allows, at least theoretically,<br />
for the incomplete market to fulfill the necessary conditions to be dynamically complete. This means that individual<br />
portfolios and asset prices should have exactly replicated those of the analogous complete market setup. Using<br />
electronic trading software, we tested this theory by implementing experiments among willing participants.<br />
Clustering Astronomical Event Notification via Natural Language<br />
Melissa L. Zhang<br />
Mentors: George Djorgovski and Matthew Graham<br />
Astronomical Event Notifications (ATels) are brief natural-language HTML reports on observations of dynamic<br />
sources in the sky. As sky survey technologies improve dramatically, these astronomer-reported text-based data<br />
become increasingly popular as a mode of communication. However, the data remain non-machine-processable,<br />
and so we cannot organize and interpret the information in large quantities. In this project, we use naturallanguage<br />
processing techniques to achieve automated classification of ATels by topic. We first removed all HTML<br />
tags, common words, punctuation, numerical tokens, and named entities from the ATel, stemmed the remaining<br />
words using the Porter Algorithm, and counted the stem frequencies to construct a vector space containing<br />
frequency-based characteristic vectors for each ATel. We then clustered these ATels via hierarchical clustering and<br />
used the top stems in each cluster to form a natural taxonomy of astronomical concepts, showing that these event<br />
notifications have an inherent classification that allows us to automatically organize ATels according a natural<br />
concept scheme. We conclude that we can use simple algorithms to process and extract meaning from<br />
astronomical textual data.<br />
The Use of Carbon Nanotube Arrays to Direct Stem Cell Differentiation<br />
Vivian Zhang<br />
Mentors: Morteza Gharib and Derek Rinderknecht<br />
Carbon nanotubes have unusually strong resistance to high mechanical strains. Therefore, they can be used to<br />
simulate a large range of mechanical properties of the extracellular environment around differentiating stem cells.<br />
Previous research shows a definite correlation between the differentiation of stem cells and the mechanical forces<br />
in the cells’ environment. However, for CNTs, there have only been qualitative data regarding the properties of the<br />
matrix on cell differentiation. To determine a quantitative relationship, CNTs will be anchored vertically in a matrix.<br />
The depth at which the CNTs are anchored will vary the effective length of each nanotube. This will alter the<br />
stiffness of the matrix as a whole, which is calculated with two types of force transducers by measuring the<br />
displacement of the tubes resulting from the amount of force applied. In the first mechanism, force-sensor probes<br />
will apply force along the side of a microarray, perpendicular to the direction of the nanotubes. Depending on the<br />
distance the nanotube was displaced, at what height the probe was relative to the base, and the amount of force<br />
applied, the stiffness of the matrix can be determined by pressing the probe into the side of the array.<br />
Microgrippers will also be inserted in the array and apply force to both sides of a bundle of CNTs. Undifferentiated<br />
stem cells will then be incubated on the array until differentiation occurs. The results should support existing<br />
qualitative research – stiffer matrices lead to bone cells while softer arrays yield adipose cells.<br />
Analysis of Noise Processes in Nanoelectromechanical Systems<br />
Yingqi (Helena) Zhang<br />
Mentors: Michael L. Roukes and Luis G. Villanueva<br />
Nanoelectromechanical systems (NEMS) integrate semiconductor electronics and mechanical devices on the<br />
nanometer scale to create sensitive, high-fidelity resonators with important applications in diverse fields ranging<br />
from precision quantum measurement to biotechnology. Such applications require resonators that are highly<br />
sensitive, down to the single molecule level. This can only be achieved with low noise levels and high frequency<br />
stability. Frequency fluctuation and noise measurements on several types of resonators with varying input<br />
parameters were conducted in order to better understand the relevant noise processes for such systems and<br />
possible problems with current models.<br />
108
Route Decisions in a Simple Network: A Pilot Study<br />
Yuanjun Zhang<br />
Mentor: Soo Hong Chew<br />
Travelers navigating through a road network must make route decisions prior to and during travel. We model the<br />
decisions of travelers choosing between routes with certain and uncertain time costs using traditional decision<br />
theory. Results from a small-scale experiment suggest that people behave differently depending on the information<br />
available to them at the time of their choice. Our results suggest that a more comprehensive theory is required to<br />
understand the decision-making behaviors of travelers in a road network.<br />
Crowdsourcing for Event Classification in NOvA Particle Physics Experiment<br />
Shiyu Zhao<br />
Mentor: Ryan B. Patterson<br />
The NuMI Off-axis � e Appearance Experiment (NOvA) aims to probe the nature of neutrinos through the<br />
phenomenon of neutrino oscillation. Its main goal is to search for the oscillation of muon neutrinos ( � � ) to electron<br />
neutrinos ( � e ), which requires accurate discrimination between � e Charged Current (CC) events and other<br />
background events. Some of the background events, such as some Neutral Current (NC) events, may produce very<br />
similar patterns to � e CC events in the liquid scintillator detector, and are thus difficult to recognize by computers.<br />
We present a method that uses crowdsourcing for event identification. This method first filters out computerrecognizable<br />
events, such as most � � CC events and some NC events, and then gathers the most ambiguous<br />
events into an event library for users to classify. We implement a prototype of such a crowdsourcing platform as a<br />
web application, capable of training and incentivizing users, and classify events according to user judgment. Upon<br />
further testing, this method may be used to improve the accuracy of event identification in NOvA.<br />
Autonomous Assembly, Capture, and Latching Strategies for Reconfigurable Spacecrafts in Low Earth<br />
Orbit<br />
Zipeng Zhao<br />
Mentors: Sergio Pelligrino and Marin Kobilarov<br />
There are many advantages to an active deployment scheme in an autonomous, reconfigurable, near-Earth<br />
satellite. In order to ensure that relevant docking maneuvers are performed successfully, a reliable docking<br />
mechanism must be implemented. Experiments evaluating the performance of several design iterations of docking<br />
mechanisms were conducted. All six degrees of freedom of the satellites’ surrogates were monitored and recorded<br />
at about 14 frames per second through an overhead camera stationed above an air table, which simulates zero-g<br />
vacuum, albeit in only two dimensions. Angle, velocity, and angular velocity at impact for each trial were extracted<br />
and amalgamated, and a plot of docking outcome for the three aforementioned parameters was generated. A pin<br />
and cup design that kinematically constrains the two spacecrafts with large cup diameters proves most promising<br />
for a typical CubeSat. Various theoretical latching strategies were also studied. Designs for potential mechanical<br />
latching devices were proposed and discussed. Further studies involving the use of variably controlled<br />
electromagnets on-board the spacecrafts during a docking maneuver and alternate docking performance evaluation<br />
methods are suggested.<br />
Sensory Substitution for Blind Rehabilitation and Cross-Modal Plasticity: A Study of Visual Substitution<br />
With Audition<br />
Yuqian C. Zheng<br />
Mentors: Shinsuke Shimojo and Noelle R. B. Stiles<br />
The brain is metamodal. While each portion of the brain analyzes best the signals from a particular sense, this does<br />
not preclude its analysis of other senses. We investigated the visual cortex's ability to analyze auditory input by<br />
using two sensory substitution devices (SSDs): the vOICe and the Raindrop. The vOICe generates a gray-scale<br />
image from the visual scene and encodes for brightness and location for each pixel with sound. After training pilot<br />
subjects for only five hours, they can localize a two-inch radius white dot and identify line conformations, simple<br />
shapes, and everyday objects without sight. We are also developing the Raindrop, an SSD designed to digitally<br />
mimic the sound of rain on a surface to determine location and surface qualities. To determine coding parameters<br />
and to gather baseline psychophysical data, six coplanar speakers called the Hoop is used. Pilot tests have shown<br />
that subjects can accurately pinpoint a single sound that lasts for twelve milliseconds and can also localize multiple<br />
sequential speaker activations. This data indicates that each raindrop can be just twelve ms, increasing the<br />
information bandwidth of the Raindrop. Studying cross-modal plasticity advances our understanding of sensory<br />
processing, while helping the blind to “see” again.<br />
109
Searching for New Generation Probes for Super Resolution Microscopy<br />
Ivanka R. Zhivkova<br />
Mentors: Long Cai and Timur Zhiyentayev<br />
The rapid development of fluorescence microscopy over the last few years made super-resolution imaging possible.<br />
However, the fluorophores that are currently used have low contrast ratio and low photo stability, which ultimately<br />
limits both labeling density and resolution. In this project we’re trying to synthesize carbospirans, photoactivatable<br />
fluorophores with enhanced contrast ratio and photo stability. Together with another SURF student I’ve been able<br />
to synthesize two coumarins with distinct absorbance spectra, both can be used as a caging moiety of<br />
carbospirans. This may allow multicolor super resolution imaging with carbospirans. Also if time allows I plan to<br />
obtain a water soluble carbospiran with excitation maximum at 650 nm.<br />
Using X-ray Crystallography to Unlock the Structure of the 2G12 Wild Type Dimer<br />
Yu (Daniel) Zhou<br />
Mentors: Pamela Bjorkman and Yunji Wu<br />
The 2G12 wild type dimer (WTD) is highly efficient at neutralizing the Human Immunodeficiency Virus (HIV) due to<br />
its high binding affinity to the low density gp120 spike. Because HIV constantly mutates into new strains, current<br />
pharmaceutical approaches aim to stop proliferation of the virus through a lifelong cocktail of medication. Solving<br />
the structure of 2G12 WTD reveals exactly how the dimer complexes with the gp120 envelop protein, which is<br />
crucial in developing highly efficient, broadly neutralizing antibodies to counteract the HIV infection with only a<br />
single shot. Attempts at discovering the protein structure have been made through various crystallization<br />
techniques involving: (1) hanging drop trays of 2G12 WTD complexed with 2G12.1 peptide in 5% tacsimate, 10%<br />
polyethylene glycol monomethyl ether (PEGMME), and 100mM HEPES, (2) isomorphous phasing methods using<br />
tantalum bromide (TaBr), and (3) crystal screenings of 2G12 WTD complexed with HxBC2 (construct with 7<br />
glycosylation sites removed from gp120 spike). Crystals have been observed in (1) but do not diffract to a high<br />
enough resolution for an accurate analysis while (2) and (3) both have shown encouraging results. Further studies<br />
varying the crystallization conditions will hopefully yield crystals of high enough diffraction resolution to produce a<br />
clear structure.<br />
A Route for Enhanced Strength Materials Through Three-Dimensional Nanostructured Design<br />
Jennifer Zhu<br />
Mentors: Julia R Greer and Andrew Jennings<br />
The goal of this research project is to develop a massive meta-material that utilizes the superior strength of<br />
nanoscale metals. In order to create a macroscopic sample that harness properties found only at the nanoscale, a<br />
fabrication route that preserves nanoscale features needs to be developed. As such, the initial goal of this research<br />
is to understand the mechanical properties of the component pieces: the individual iron nanopillars. This knowledge<br />
will allow for the design of macroscopic test specimens. In order to study these iron pillars, the first step is to<br />
electroplate iron nanopillars that range from 50 nm to 500 nm in diameter. Samples are then chosen with a<br />
scanning electron microscope (SEM). Mechanical properties are determined through small-scale compression tests<br />
in an Agilent G200, followed by post-compression analysis. Three pieces of data—pre-compression images,<br />
compression data, and post-deformation images—are then used to understand how iron plastically deforms as a<br />
function of size.<br />
Characterizing Maternal Drosophila Promoters for Use in Aedes<br />
Mario V. Zubia<br />
Mentors: Bruce Hay and Omar Akbari<br />
Several approaches are being made to help tackle the problem of mosquito-borne illnesses (e.g. malaria and<br />
dengue fever). One such approach implements a drive system that can help genetically engineered, diseaseresistant<br />
mosquitos replace the wild type population. One drive system, Medea, utilizes a maternal toxin and<br />
zygotic antidote. When progeny of a Medea-bearing mother receive this Medea gene, they survive; those that do<br />
not receive the gene, die. This dynamic requires the Medea gene is eventually needed for survival, and thus<br />
Medea-bearing individuals are given a fitness advantage. Thus after release into a wildtype population, these<br />
transformants are able to overtake and replace the population. Several Medea constructs were created and tested<br />
in Drosophila with varying degrees of success and failure. To understand why, components of this genetic construct<br />
were studied more closely. One such component, the maternal promoter, has different expression patterns in<br />
oogenesis. This may explain the effectiveness of the toxin and of the overall genetic construct. Further analysis<br />
needs to be completed, but if this association holds, then more effective systems can be developed and transferred<br />
to Aedes.<br />
110
MURF<br />
MURF UNDERGRADUATE<br />
RESEARCH FELLOWSHIPS<br />
M<br />
U<br />
R<br />
F
Collective and Rigid Body Behavior of Osmotic Motors<br />
Christian A. Aponte Rivera<br />
Mentors: John F. Brady and Nicholas Hoh<br />
Diffusiophoretic motion of a colloidal particle is possible if the particle creates a chemical potential gradient about<br />
its surface. We examine the behavior of multiple particles capable of inducing local concentration gradients, called<br />
osmotic motors. The particles are studied using Brownian dynamics simulations in untethered and rigid-assembly<br />
configurations. Simulations are verified in certain configurations by theoretical predictions. The effect of the rate of<br />
reaction, in the form of the Damköhler number, on the translational diffusivity is also studied. The average Péclet<br />
number is calculated as a function of the Damköhler number.<br />
Liquid Crystal Emulsion Generation With Micofluidic Techniques<br />
Gabriela Bernal<br />
Mentors: Julie Kornfield and Paul Pirogovsky<br />
Liquid Crystals (LCs) have been used as biosensors for the optical detection of bacteria and in photonics. Such<br />
sensors typically are thin film based and require special processing. Encapsulating the active LC element in a<br />
droplet with a responsive outer layer would allow for more robust, portable and easily manufactured sensing<br />
systems as well as self contained optical elements. We attempted to use a technique introduced by D.A. Weitz et<br />
al. using glass micro-capillaries to make monodisperse LC droplets using 4-cyano-4’-pentylbiphenyl (5CB).<br />
Preliminary calculations for relevant flow rates were performed taking account the geometry of the system and the<br />
rheological properties of the LC. The behavior of flowing liquid crystals was studied under temperature gradients<br />
and in the presence of a magnetic field, with results pending. We hope to be able to use the results of this study to<br />
create LC droplets of the desired size and shape. The next step is to incorporate a block copolymer surfactant<br />
consisting of a Side-Group Liquid Crystalline Polymer conjugated with a hydrophilic polymer to stabilize the<br />
droplets and control the anchoring condition at the interface, thus tailoring the structure and optical properties of<br />
the droplets.<br />
Data Quality Monitoring for Jets and Missing Transverse Energy, and Algorithms for Tagging<br />
b-Jets at CMS<br />
Dinah Simone (Simca) Bouma<br />
Mentors: Maria Spiropulu, Artur Apresyan, and Alex Mott<br />
The offline Data Quality Monitoring (DQM) framework is run on all the data taken at the Compact Muon Solenoid<br />
(CMS) experiment, in order to ensure that all the detector components are fully functional and performing<br />
according to expectations, and that the reconstruction of physics objects is performing as expected. To prepare for<br />
the increased instantaneous luminosity delivered by the LHC, this framework needs to optimize both its CPU and its<br />
memory usage. The JetMET DQM code was written during the commissioning period of CMS, when the first<br />
collisions occurred in 2009, so it contained some obsolete or unnecessary monitoring elements. These elements<br />
were removed, while the remaining elements was optimized according to the detector resolution. These changes<br />
resulted in a 70% decrease in memory usage, and were included in the next version of the CMS Software<br />
Framework.<br />
The second part of the project is to implement a new tracking algorithm for b-tagging at HLT. Jets originating from<br />
the hadronization of b quarks (b-jets) are interesting for studies of standard model physics, such as top quark<br />
properties, and physics beyond the standard model, such as supersymmetry (SUSY). Due to a very high crosssection<br />
of QCD multi-jet production, the trigger rates of purely hadronic triggers are extremely high. In order to<br />
efficiently collect events containing b-jets, a dedicated b-tagging algorithm is needed, prior to event reconstruction.<br />
The events tagged by this algorithm are saved and reconstructed for physics analysis. In order to improve the btagging<br />
triggers for SUSY, Exotica, and other analysis groups, a new algorithm for tracking the jets in the detector<br />
had been developed. Tentatively, the performance of this algorithm shows an improvement in efficiency of 6%<br />
compared to the old algorithm.<br />
Visual Representation of Chemical Species Over Southern California<br />
Denise Castellanos<br />
Mentors: John Seinfeld and Joseph Ensberg<br />
Data visualizations of the concentrations of atmospheric chemical species are meant to serve as a tool for<br />
understanding how the many processes in the atmosphere coincide. This understanding in turn serves as the basis<br />
for the actions that are undertaken to help prevent major shifts in the atmosphere’s composition which may lead to<br />
unintended consequences. Both the qualitative and quantitative nature of the data are used in the creation of<br />
public policy that is meant to protect the health of people and help reduce the impact of human activity on the<br />
earth’s climate. There are various types of graphs that are used to evaluate theoretical data and experimental data<br />
in different ways. For example, a time series graph is appropriate for comparing the experimental data and<br />
113
theoretical data of one pollutant type. In this way the validity of the model can be checked against actual data that<br />
has been collected. Also, contour maps can be utilized to compare potential theoretical models. The data can be<br />
manipulated in various ways which would offer a visualization that could explain the differences in the data sets. In<br />
this way it can help in determining the validity of one set or help prove that they are both false. Investigations into<br />
how data can be best represented is a necessary part of atmospheric research. It provides a means of finding the<br />
connection between numbers and the important concepts that will govern the future.<br />
Notch Toughness and Glass-Forming Ability of Vitreloy 101 and Its Variants<br />
Wei J. Chen<br />
Mentors: Marios Demetriou and Glenn Garrett<br />
The addition of Sn and Si to the Be-free Cu-based bulk metallic glass Vitreloy 101 is studied for their effects on<br />
glass-forming ability (GFA), notch toughness, yield strength, and elastic moduli. The glass-forming ability is<br />
determined by injection casting in to copper molds of different diameters and checking for crystallinity using X-ray<br />
diffractometry. The notch toughness is performed using 3-point bending of 3mm diameter rods. The yield strength<br />
and elastic moduli are measured from compression tests and ultrasonic measurements, respectively. Substituting<br />
Sn into the Vitreloy101 system improves GFA, but degrades the toughness.<br />
Tetranuclear Complexes for the Reduction of Dioxygen<br />
Michael Desanker<br />
Mentors: Theodor Agapie and Davide Lionetti<br />
There has been an emphasis on the development of alternate sources of energy that not only have the potential to<br />
replace the use of gasoline, but are also environmentally friendly. Fuel cells have this potential. The Pt/C catalyst<br />
currently used in fuel cells is expensive. Therefore development of a catalyst that utilizes cheaper metals will make<br />
fuel cells economically viable. Multinuclear complexes that reduce dioxygen are present in nature and serve as<br />
inspiration for the synthesis of a multinuclear complex capable of replicating this reactivity. Ascorbate oxidase in<br />
particular, yielded the idea of a dioxygen molecule binding at a trimetallic site. Ligand frameworks that would bind<br />
three iron or cobalt centers and a fourth yttrium center were synthesized. The goal of these complexes is to hold<br />
the cobalt or iron atoms in close enough proximity to each other to react cooperatively and limit excess reactivity.<br />
Metallation of these frameworks with iron, cobalt and yttrium has been studied. Further characterization of these<br />
complexes involves spectroscopy, electrochemistry and reactivity with dioxygen.<br />
Ground Motion Data Analysis in Southern California for Ruptures on the Southern San Andreas Fault<br />
Fernando (Estefan) Garcia<br />
Mentors: Swaminathan Krishnan and Ramses Mourhatch<br />
A large-magnitude earthquake in the range of 6.0-8.0 is expected to occur in southern California in the near future.<br />
An earthquake in this magnitude range is likely to have significant and hazardous impact on large city-centers such<br />
as downtown Los Angeles. High-performance computer simulations are used to obtain ground motion data at<br />
different stations in southern California from several earthquake scenarios. Using parameters of distance to rupture<br />
plane and magnitude, potential peak ground velocities can be analyzed in southern California. The mean and<br />
standard deviation of peak ground velocity can be found at different distance ranges, and confidence intervals for<br />
peak ground velocity can be created for each distance range. Graphs of peak ground velocity versus distance and<br />
magnitude can be used as tools to determine the level of shaking as determined by both parameters. To assess the<br />
accuracy of this analysis, peak ground velocity confidence intervals from simulations will be compared to those<br />
from an attenuation relationship developed as part of the PEER Next Generation Attenuation project.<br />
Investigating the Role of Conserved Signaling Pathways in Ascaroside Sensing in the Nematode<br />
Trapping Fungus Arthrobotrys oligospora<br />
Sherif Gerges<br />
Mentor: Paul Sternberg<br />
Arthrobotyrs oligospora is a nematophagous fungus which traps nematodes via three dimensional hyphal networks.<br />
Ongoing research has been focused on determining the molecular mechanisms responsible for nematode detection.<br />
Previous data have shown that the ascarosides, which are C. elegans pheromones, trigger trap formation in A.<br />
oligospora. Two conserved signaling pathways, the cyclic adenosine monophosphate – protein kinase pathway<br />
(cAMP-PKA) and mitogen-activating protein kinase (MAPK) pathways are known to govern fungal morphogenesis in<br />
fungal species across different phyla; thus, key components in the pathways including the protein kinase A PKA,<br />
MAPKK STE7, the G-protein coupled receptor STE3, and the G-beta subunit GPB1 were selected as candidate genes<br />
for targeted gene deletion to test whether these pathways play a role in ascaroside sensing. We use biolistics and<br />
protoplasts as two methods for transformation. Putative drug-resistant colonies will be screened by PCR and<br />
Southern blot hybridization to confirm gene deletion. Once the mutants have been obtained, we will test their<br />
ability to respond to ascarosides on a plate assay.<br />
114
Expression Patterns of Msi2a in Neural Crest Cells of Wild-Type Zebrafish Embryos<br />
Anna Ruth Gonçalves<br />
Mentors: Marianne Bronner and Tatiana Hochgreb<br />
Studies demonstrate that the Musashi family of genes is co-expressed in proliferating embryonic pluripotent neural<br />
precursors and in cells that are likely to lead to postnatal and adult CNS stem cells. Imbalanced expression of<br />
genes within the Musashi family (e.g. Msi1 and Msi2) has been linked to the following diseases: Pick and<br />
Alzheimer’s disease, Dementia and Chronic Myeloid Leukemia. Preliminary work in the Bronner lab has shown that<br />
Msi2 expression has been segregated between Msi2a and Msi2b in Zebrafish embryos, with Msi2a expressed in<br />
migrating neural crest cells, while Msi2b is localized to neurons in the ventral neural tube, olfactory region and<br />
pineal gland. This study involves performing Whole Mount in-situ Hybridization to characterize the expression<br />
pattern of Musashi 2a on wild-type Zebrafish embryos in order to visualize where the mRNA of Musashi2a within<br />
the embryo is expressed. Considering that not much is known about the Musashi gene family -- which is a family of<br />
post-transcriptional regulation genes -- beyond understanding the Musashi gene family, further studies can also<br />
explore how other genes may be regulated by the Musashi gene family and lead to various diseases and<br />
consequently, therapeutic interventions for these diseases.<br />
Power Flow of Electrical Distribution Feeders Modeled by Open Distribution System Simulator<br />
Jocelyn Graf<br />
Mentors: Steven Low and Dennice Gayme<br />
After electric power has been generated and transmitted on high power lines to substations, distribution feeders<br />
transport the power at lower voltages to individual customers. Many of these systems have not changed in 50-100<br />
years, but new technology is rapidly creating the need for new systems. For example, electric vehicles will demand<br />
new levels of power. Furthermore, a much higher proportion of electric power must be generated from renewables<br />
like wind and solar generators that produce intermittently, in turn, requiring power storage to smooth power flow.<br />
Some of this renewable generation will be distributed as smaller units throughout the grid instead of being<br />
concentrated at large power plants. These changes require a retrofit of power systems to include faster and more<br />
comprehensive monitoring, control, communication, and security systems, which will combine to create a “smart<br />
grid”. Energy can be conserved and generation thereby reduced if power flow optimization algorithms are<br />
developed to take advantage of the smart grid’s new capabilities. A distribution feeder was selected and modeled<br />
using the Open Distribution System Simulator. Power flow analyses were performed with and without distributed<br />
generation. The test feeder was characterized for power flow optimization research incorporating distributed<br />
photovoltaics and battery storage.<br />
Malley Probe: An Optical Instrument That Will Facilitate the Study of Thermally Perturbed Boundary<br />
Layers and Their Effects on Optical Systems<br />
Edward B. Guzman<br />
Mentors: Beverley J. McKeon and Rebecca Rought<br />
Collimated laser beams have been used in many aerospace applications, as in the case of communication and<br />
targeting. Unfortunately, when a collimated laser beam traverses a turbulent boundary layer containing density<br />
fluctuations, its wavefront becomes aberrated, potentially affecting the ability of the laser beam to be used in the<br />
far field. Yet, the causes of these aberrations are not entirely known, but only that they are produce by density<br />
fluctuations originating from either compressibility effects or the addition of heat to the boundary layer. The Malley<br />
Probe is an optical instrument that provides accurate information about the distortion of a laser beam in a turbulent<br />
boundary layer whose information will be used to calculate the convective velocity of the turbulent structures in the<br />
flow and the Optical Path Difference of the laser beam. The assembly of a new Malley probe and preliminary results<br />
from the turbulent boundary layer will be discussed.<br />
Transforming Drosophila Through in vivo Injection of DNA and Transfection Reagent<br />
Dustin Harris<br />
Mentors: Bruce Hay and Kelly Dusinberre<br />
Creating transgenic insects has been possible for decades. These experiments have helped us learn more about<br />
genetics and can lead to solving problems in agriculture, medicine and basic biology, but not all insects are easily<br />
transformed. The goal of our experiment is to inject both male and female adult Drosophila with DNA and<br />
transfection reagent to achieve transformation in the germ line. The plasmid DNA containing the gene of choice is<br />
co-precipitated and mixed with transfection reagent which encapsulates it in a cationic liposomes. These are then<br />
injected into the abdomen of living adult Drosophila and transported from the host hemolymph to the germ line.<br />
The liposomes fuse with the cell lipid bilayer and release the DNA, enabling transposition or integration of plasmid<br />
DNA into the genome. The first generation progeny of the injected males are screened along with the first and<br />
second generation progeny of the injected females. Transformation has not been seen in the first or second<br />
generation progeny of the first few DNA/transfection reagent combinations, but screening is not complete. More<br />
experimentation with the injection of larvae, composition of the injected material and volume injected can be<br />
carried out.<br />
115
The Planet-Metallicity Correlation in the Kepler Field<br />
Keith A. Hawkins<br />
Mentors: John Johnson and Tim Morton<br />
One of the most interesting correlations that has come out of large statistical studies of exoplanets is a relationship<br />
between the occurrence rate of giant planets planets and the metallicity of their host stars. The primary goal of this<br />
project is to develop an automated pipeline to measure stellar parameters (e.g. effective temperature, surface<br />
gravity and metallicity) of a large number of stars. We use a spectral index-based method that employs a Bayesian<br />
approach to determine the stellar parameters of stars. Preliminary results of this method will be presented. This<br />
method will be applied to a large number of stars in the Kepler field in order to determine the planet-metallicity<br />
correlation for small planets for the first time.<br />
Do Ascarosides Affect Egg-Laying Behavior in Caenorhabditis elegans?<br />
Circe Lassegue<br />
Mentor: Paul Sternberg<br />
Caenorhabditis elegans, a small transparent nematode that lives in temperate soil environments, is one of the<br />
simplest eukaryotic organisms with a nervous system to be studied in great detail. Over recent years, a large<br />
number of ascarosides have been identified as signaling molecules in C. elegans. Ascaroside levels are affected by<br />
worm concentration and available food when developed in “worm water”. Ascarosides have been shown to regulate<br />
a large number of behaviors in C. elegans including dauer formation, mating behavior, aggregation, and olfaction.<br />
Additionally, environmental and homeostatic cues are now being explored to see how these affect nematode egglaying<br />
habits. We studied the modulatory effect of several ascarosides on egg-laying behavior and brood size in<br />
adult female C. elegans.<br />
This study aims to determine the effect of ascarosides on egg-laying behavior in adult C. elegans. A range of<br />
concentrations of several synthetic ascarosides as well as natural worm water produced by C. elegans were<br />
studied. Standard egg-laying assays and known positive and negative controls were utilized.<br />
Perfluoroalkyl-Poly(ethylene glycol)-Poly(acrylic acid) Triblock Copolymers for Mucoadhesive Drug<br />
Delivery<br />
Yulan Lin<br />
Mentors: Julie Kornfield and Jeremy Wei<br />
Hydrophobic drugs can be loaded into the core of polymeric micelles, which form from diblock or triblock<br />
amphipathic polymers. Poly(ethylene glycol) based copolymers are of interest because PEG is nontoxic and<br />
nonimmunogenic, and fluoroalkyl ended PEG (Rf-PEG) has been shown to form a gel in solution. However,<br />
previously synthesized Rf-PEGs have not shown mucoadhesive properties. The desired mucoadhesive properties<br />
could potentially be conferred by adding poly(acrylic acid) (PAA), a compound characterized by its mucoadhesive<br />
properties. A tri-block Rf-PEG-PAA will be synthesized by polymerizing tert-butyl acrylate on the end of single<br />
ended Rf-PEG modified with a reversible addition-fragmentation chain transfer (RAFT) agent. The poly(tert-butyl<br />
acrylate) will then be converted to poly(acrylic acid). The resulting material will be characterized by gas permeation<br />
chromatography, NMR spectroscopy, a mucoadhesiveness test, and rheology. These tests will allow for the<br />
hydration kinetics, mucoadhesiveness, and rheological properties to be studied. Further tests will be run to<br />
determine its usability in targeted, controlled, drug delivery systems.<br />
Exploring the Spectrum of Multi-Dimensional Quasi-Periodic Shrödinger Operators<br />
Marissa Loving<br />
Mentors: Barry Simon and Helge Krüger<br />
Operators can be viewed as generalizations of matrices and their spectrum as the generalization of their<br />
corresponding eigenvalues. Shrödinger operators are a special class used in the description of quantum mechanics.<br />
In particular, their spectrum describes the allowed energy levels in such a system. It is the spectrum of a smaller<br />
subset of the one-dimensional Schrödinger operators known as the Almost Mathieu or Harper's operator that leads<br />
to the Hofstadter Butterfly a recursive geometric object known as a fractal.<br />
The Hofstadter Butterfly was computed two ways, by computing the eigenvalues of finite tridiagonal matrices or,<br />
alternatively, by finding the roots of the polynomial produced by computing the discriminant, D(E), where V(n) = 2<br />
cos(2 π ((p/q) n + x)). Due to the complexity of the computations and the sheer number required, Mathematica<br />
was used both to perform the computations and produce the final graph. A report was generated detailing the<br />
processes used to produce the Hofstadter Butterfly in order to simplify the process for future mathematicians<br />
performing research in this area.<br />
116
Ice Crystal Growth in Solutions Containing Salts, Polymers, Nanoparticles, and Antifreeze Proteins<br />
Icon Mazzaccari<br />
Mentor: Kenneth Libbrecht<br />
While there exists an established theory for ice crystal growth in pure water that can predict the growth rate and<br />
radius of the dendritic tip, there is no such theory for ice crystal growth in solutions. As the solution is supercooled<br />
and the ice interface begins to grow, it must release latent heat to transition into the solid ice phase.<br />
Simultaneously, the growing ice lattice must reject the solutes in the liquid, causing an increase in concentration<br />
near the ice interface. Solutes also lower the freezing temperature of solutions by disrupting the interactions<br />
between the solvent’s molecules. An investigation of how solutions containing chemical compounds including<br />
sucrose, salts, like NaCl and LiCl, polymers, nanoparticles and antifreeze proteins affect the growth of ice crystals is<br />
discussed, as well as the preliminary work towards a comprehensive theory of dendritic growth in solution.<br />
Experiments have been performed with the simpler compounds, and results and future plans will be described.<br />
Microstructural and Mechanical Characterization of Cementitious Materials<br />
Luis A. Miramontes<br />
Mentors: Julia Greer and Jane Lian<br />
The objective of the project was to measure the elastic modulus and hardness of Carbon Silicate Hydrates, C-S-H,<br />
at its different phases. C-S-H gels have two main phases, an inner sold bulk area and an outer more porous area.<br />
When ettringite forms within the nanopores of the outer phases, nanocracks are generated, and the mechanical<br />
properties could be quite different from the inner solid bulk phase. Four different cement samples were sent over<br />
by our collaborators at Tennessee Technological University to test in 1”x1”x ~2” blocks. The samples were then<br />
machined into 1” Diameter cylinders with a height of around 1” and set into an Epoxy mold to fit into our G200<br />
nanoindenter. As nanoindentation measurements are highly sensitive to sample roughness, the sample surfaces<br />
need to be flat and smooth so that reliable mechanical properties can be obtained. To achieve this, all samples<br />
were successfully grinded and polished down to a 0.05�m finish. Nanoindentation experiments will be conducted on<br />
different phases of the sample, and a comparison of the elastic modulus and hardness between the inner phases<br />
and outer phases will be obtained.<br />
Characterization of Adh5 Expression and Function in the Chick Neural Crest<br />
Ryan D. Morrie<br />
Mentors: Marianne Bronner and Crystal D. Rogers<br />
Adh5 is a class III alcohol dehydrogenase that may be required for normal embryonic development. In situ<br />
hybridization for Adh5 in Gallus gallus embryos showed a spatiotemporal expression pattern with Adh5 enriched in<br />
the developing neural tube and the migratory neural crest, suggesting an involvement in neural crest formation<br />
and migration initiation. Cryosectioning also revealed that Adh5 is expressed around the entire apical area of the<br />
neural tube. Here we analyzed the role of Adh5 in neural crest development through loss-of-function experiments<br />
using a translation blocking morpholino oligonucleotide against Adh5 (Adh5MO). Gallus gallus embryos were<br />
injected and electroporated with various concentrations of Adh5MO. At high concentrations Adh5MO appeared to<br />
induce cell death. Therefore, we analyzed the phenotype caused by lower doses of Adh5MO with<br />
immunohistochemistry and in situ hybridization against proteins and genes important in neural crest formation and<br />
migration, as well as general cell adhesion molecules. My results show that Adh5 is expressed in developing neural<br />
crest cells and may be necessary for neural crest specification and development.<br />
Determination of Binding Mechanism of a Beetle Antifreeze Protein and Crystalline Carbohydrates<br />
Edward (Ted) Ro<br />
Mentors: William A. Goddard III, Ravinder Abrol, and Xin Wen<br />
Antifreeze proteins (AFPs) refer to a class of polypeptides found in fish, plants, and insects that lower the freezing<br />
point of water without affecting the melting point. An AFP from the beetle Dendroides canadensis (DAFP-1) exhibits<br />
high antifreeze activity and demonstrates the ability to hinder crystal growth and prevent precipitation of naturally<br />
occurring carbohydrates at lower temperatures in vitro. The focal of the research is to identify the ligand binding<br />
sites of DAFP-1 by simulating protein and ligand docking. The hypothesis is that DAFP-1 will bind to hydroxyl<br />
groups of carbohydrate molecules in solution and interact with growing crystalline carbohydrates through either<br />
threonine or arginine residues. Through the use of GenDock software, a docking program that generates a large<br />
number of ligand binding poses and ranks them according to energy, the most favorable binding poses of DAFP-1<br />
and various trehalose ligands were simulated. The side chains of Arg54, Arg25, Gln19, and Thr56 are important<br />
residues in the binding of a single trehalose dihydrate molecule to DAFP-1. Docking a single trehalose anhydrous<br />
molecule to DAFP-1 with GenDock revealed the backbone atoms of Asp42 as important in the protein and ligand<br />
binding along with the side chains of Arg54, Thr56, Ser32, and Gln19. A trehalose unit cell and DAFP-1 docking<br />
simulation suggested that the protein interacts with the ligand on the corner of the unit cell at the AB and BC faces<br />
through the side chains of Glu50 and Arg25. Arginine, threonine, and glutamine, along with a few backbone atoms,<br />
appear to be important residues in DAFP-1 for trehalose binding. Future research will include the addition of a<br />
solvent to the docking environment and exploring different carbohydrate and DAFP-1 binding interactions.<br />
117
Cryogenic Impact Testing of Tin (Sn) Solders and Bulk Metallic Glass Composites<br />
Adam R. Ullah<br />
Mentors: Marios Demetriou, Doug Hofmann, and Scott Roberts<br />
Providing a protective shielding for space-deployed technologies is crucial when trying to ensure multi-component<br />
function and communication at different positions in space, in orbit and in other planets. At the California Institute<br />
of Technology, Vitreloy 1 (Zr41.2, Ti13.8, Cu12.5, Ni10.0, Be22.5), GHDT (Zr59.6, Ti26.8, Cu8.90, Be4.63), DH3 (Zr39.6, Ti33.9,<br />
Nb7.60, Cu6.40, Be12.5) and Tin (Sn) were cryogenically tested on a Charpy Impact Tester to determine their<br />
mechanical performance on a temperature scale from (-186.5�C to +22�C). At these temperature ranges, DH3 was<br />
compared to the monolithic glasses (Vitreloy 1 and GHDT). The dislocation motion in the BCC-phase of DH3<br />
decreases at lower temperatures but does not perform worse than monolithic glasses Vitreloy 1 and GHDT. There is<br />
a ductile to brittle transition in the mechanical performance of DH3; however, the ductile to brittle transition is<br />
small in comparison to other contending BCC metals. DH3 is capable of protective cryogenic shielding.<br />
Strong and Tough Graphene-Polymer Composites<br />
Amy Wat<br />
Mentors: Julia R. Greer and Wendy Gu<br />
The strength and toughness of a graphene-polymer composite are optimized by emulating the nacre structure,<br />
which calls for graphene platelets dispersed in a polymer film. The graphene was grown on a copper rod using<br />
chemical vapor desposition and extracted off the copper rod by sonication. Raman Spectrometry reveals that a<br />
small flake of graphene was obtained. The surface of the copper rod was also etched off, but there is no indication<br />
that graphene was obtained. Graphene was also obtained through liquid exfoliation, which is sonication of graphite<br />
in a solvent followed by vacuum filtration. This created a mixture consisting of a low yield by weight of graphene<br />
platelets. The composites were created by spincoating a thin polymer film layer and then applying a graphite layer,<br />
which interfered with creating consistent uniform thin films. Composites were also made by applying graphite onto<br />
a polyisoprene disc, which was cut into four pieces and stacked on top of each other then flattened, which allows<br />
the creation of many thin bilayers of graphene and polymer each time the process is repeated. Through this<br />
project, we found the difficulty of mass producing graphene and how graphite affects the polymer strength and<br />
toughness.<br />
118
AMGEN<br />
AMGEN SCHOLARS PROGRAM<br />
A<br />
M<br />
G<br />
E<br />
N
Endospore Spore Viability Fractions in Ice Using Microscopic Endospore Viability Assay<br />
Judson G. Alderman<br />
Mentors: Adrian Ponce and Aaron Noell<br />
Spore forming bacteria are considered one of the hardiest forms of life. Such resilient bacteria could survive in<br />
Martian ice or travel aboard an icy comet. With reported lifetimes ranging from thousands to millions of years, no<br />
consensus exists concerning the longevity of bacterial endospores. A Microscopic Endospore Viability assay<br />
(MicroEVA), which functions by detecting a UV-excitable complex that forms when spore-specific dipicolinic acid<br />
(DPA) binds luminescent terbium ions, has been developed to investigate the true longevity of endospores.<br />
Measuring the total number of spores by killing them to release their DPA for imaging is crucial to discovering<br />
endospore longevity. In an evaluation of the MicroEVA’s new 280nm LED light source, we found that increased<br />
exposure times and image integration improved spore visualization. The new light source detects 40% more spores<br />
and exhibits a more even distribution of light intensity than the previous light source. Autoclaved total spore<br />
images suffer from physical disruption and uneven DPA distribution due to the action of steam. Conversely total<br />
spore images from spore samples ruptured in an oven have an average 2-fold increase in light intensity versus<br />
autoclaved samples and experience no physical disruption. Discerning true endospore longevity relies upon<br />
establishing a well-understood method of obtaining total endospore counts, allowing for the eventual determination<br />
of viability fractions in environmental samples.<br />
The Determination of the Dihedral Angles of 1,2-Disubstituted Ethanes in Aqueous Lyotropic Solutions<br />
Andrew J. Bendelsmith<br />
Mentor: John D. Roberts<br />
Two 1,2-disubstituted ethanes of biological interest, the succinate dianion and β-alanine, have been studied as<br />
oriented molecules in aqueous lyotropic solutions. To prepare the lyotropic solutions, myristyltrimethylammonium<br />
bromide (MTAB), decanol, and the molecule of interest were added to D2O. 1 H and 13 C NMR spectra were taken,<br />
yielding dipolar and scalar coupling constants. A MATLAB program was written, which was used to analyze the<br />
coupling constants to calculate the dihedral angle about the central carbon bond for each X-CH2-CH2-Y system.<br />
Characterization of Cation-π Binding Interactions Between Varenicline and Homopentameric α7<br />
Nicotinic Acetylcholine Receptors<br />
Emily Blythe<br />
Mentors: Dennis Dougherty and Ethan Van Arnam<br />
Nicotinic acetylcholine receptors (nAChRs) are ligand-gated ion channels that have been linked to numerous<br />
neurological syndromes, including Alzheimer’s disease, schizophrenia and addiction. Numerous subtypes of nAChRs<br />
exist, and subtype-selective pharmacology has been exploited in the design of pharmaceuticals to treat specific<br />
conditions. Varenicline, currently on the market as the smoking cessation drug Chantix, was developed to be a<br />
selective agonist at α4β2 receptors. However, the psychological side effects associated with varenicline suggest<br />
that it may be affecting receptors other than its target. In this study we focus on characterizing the binding of<br />
varenicline to homopentameric α7 nAChRs since varenicline is a full agonist at α7. Using unnatural amino acid<br />
mutagenesis, we probed the subunit interface “aromatic box” binding site to identify the residues that participate in<br />
cation-π binding interactions with varenicline. Surprisingly, our results indicate that the binding of varenicline more<br />
closely resembles that of acetylcholine than that of epibatidine, which, like varenicline, is a nicotine analogue.<br />
Traffic Jam piRNA Localization in Drosophila Ovarian Somatic Cells<br />
Yi Cai<br />
Mentors: Alexei Aravin and Edward Perkins<br />
Genomic information in the germline must be tightly regulated for accurate transmission to the next generation.<br />
Piwi-interacting RNA (piRNA), a class of 23-30 nucleotide small RNA, function in germline cells through RNA<br />
interference to silence retroelement activity and maintain genome integrity. Primary piRNAs are produced from<br />
specific genomic loci, piRNA clusters, and are further amplified in a secondary amplification loop. The primary<br />
piRNA production pathway is not yet clearly understood. Recently, traffic jam (tj) was identified as a new Piwiassociated<br />
piRNA cluster in Drosophila. One full-length tj transcript not only serves as a source of piRNAs but also<br />
encodes the TJ protein. TJ is strongly expressed in ovarian somatic cells (OSCs). To determine where primary<br />
piRNA processing occurs in the cell, an RNA tethering system has been used to detect tj transcript localization in<br />
OSCs. We have created a selectable construct containing MS2 hairpin sequences in the 3’ untranslated region of tj.<br />
A second imaging construct encodes an MS2 capsid protein (MCP) and fluorophore. Upon co-transfection in OSCs,<br />
the MCP will bind to the MS2 hairpins of the tj transcript. This system allows us to determine the location of tj<br />
transcript and primary piRNA processing.<br />
121
Identification and Characterization of Genes Involved in the Control of Stem Cell Populations in Plant<br />
Shoot Apical Meristems<br />
Margaret J. Chiu<br />
Mentors: Elliot Meyerowitz and Yun Zhou<br />
Normal growth of shoot meristem involves two opposinhg processes, differentiation and stem cell maintenance. In<br />
Arabidposis, stem cell populations in the shoot apical meristem (SAM) require the homodomain protein, WUSCHEL<br />
(WUS), to maintain their identity. The genes Clavata 1, 2, and 3 (CLV) are responsible for restricting the stem cell<br />
population by controlling differentiation. Previous research has shown that WUS and CLV interact to establish a<br />
feedback loop, although the exact mechanism is not understood. By creating a number of mutant lines for target<br />
genes expressed in SAM stem cells, we look to uncover the roles that these genes play in feedback between WUS<br />
and CLV. T-DNA insertional knock-out mutants and 35s overexpression mutants are confirmed using RT-PCR to<br />
ensure the decrease or increase in gene transcript. Plants are examined for their phenotypes using microscopy.<br />
Genetic crosses are also conducted to determine how these genes interact.<br />
Plasmid Design for an Assay to Study the Role of DNA-Mediated CT In-Vivo<br />
Keshia Dykes<br />
Mentor: Jacqueline K. Barton<br />
Base excision repair (BER) proteins repair damaged or mismatched bases in the DNA. BER proteins that contain<br />
[4Fe-4S] clusters are able to send and receive charge, which aids in locating damage. Endo III and MutY are DNA<br />
glycosylases, which are enzymes that specifically aid in the recognition of damage within the genome. There is an<br />
interest in finding proteins that help MutY search for damage. In this study, we have constructed a reporter (LacZ<br />
CC104) and control (LacZ MG1655) plasmids to examine MutY activity in vivo. The plasmids were designed as<br />
reporters utilizing the LacZ gene. If the LacZ gene (CC104) is mutated, it causes β-gal to be inactive. If the BER<br />
protein MutY is not functioning properly and unable to recognize the LacZ mutation, it will lead to transversion to of<br />
A:8-oxo-G to an A:T base pair. By using molecular biology and cloning techniques, we were able to successfully<br />
construct the reporter and control plasmids that can be used in in vivo studies with E.coil to examine how BER<br />
proteins search for damage.<br />
Improving Anti-Tumor Immunity by Engineering T Cell Receptors to Prevent Mispairing<br />
Marvin H. Gee<br />
Mentors: David Baltimore and Michael Bethune<br />
Anti-tumor immunity can be engineered in cancer patients by delivering the genes encoding the α and β chains of a<br />
tumor-specific T cell receptor (TCR). Exogenous α/β TCR chains introduced into patient T lymphocytes have<br />
potential to mispair with endogenous α/β TCR chains, resulting in reduced expression of tumor-specific TCRs and<br />
potential autoimmunity by mispaired TCRs. We are modifying tumor-specific α/β TCR chains to prevent their<br />
mispairing with wild-type chains as follows: 1) swapped domains between the α/β chains; 2) domains connected by<br />
glycine and serine linkers within α/β chains; or 3) domains substituted with γ and/or δ domains from γ/δ TCRs.<br />
When correctly paired, these modified α/β TCR chains retain all domains necessary to recruit CD3 components and<br />
express on the cell surface. When mispaired with wild-type chains, however, all domains needed for surface<br />
expression are not present. Using flow cytometry, 293T cells transfected with constructs are analyzed for cell<br />
surface expression by tetramerized peptide-MHC. Constructs containing both modified α/β chains simulate correctly<br />
paired TCRs and constructs encoding only one modified chain and one wild-type chain simulate mispaired TCRs.<br />
Several constructs appear to prevent mispairing while retaining function and warrant further investigation as viable<br />
candidates for gene therapy.<br />
Progress Toward the Enantioselective Synthesis of Variecolin: Synthesis of the B-Ring Precursor<br />
Angela M. Guerrero<br />
Mentors: Brian M. Stoltz and Kun-Liang Wu<br />
Variecolin is a tetracyclic potent anti-HIV-I natural product. An expeditious synthesis of the B-ring precursor has<br />
been achieved through a five-step process. These efforts highlight the formation of a pyrone from ethyl propenyl<br />
ether and the production of an iron complex utilizing photochemical technology. This B-ring precursor will be<br />
employed in the synthesis of the AB-ring fragment and will be ultimately used toward the total synthesis of<br />
variecolin.<br />
Lipid D/H Variations as a Possible Paleometabolic Proxy<br />
Yang Hu<br />
Mentors: Alex Sessions and Xinning Zhang<br />
Stable hydrogen-isotopic (D/H) variations have been shown to reflect the central metabolic pathways of bacteria.<br />
In addition, D/H content has been shown to persist in lipids of rock records on the order of millions of years<br />
allowing for the potential of D/H content in lipids to become a biomarker. An understanding of the mechanism by<br />
which D/H variations occur is necessary before these lipids can actually be reliably used as biomarkers. It is<br />
122
hypothesized that lipid D/H correlates to central metabolism and the process is determined by NADPH, the major<br />
reductant for cell anabolic processes. We seek to demonstrate that changes in NADPH metabolism directly affect<br />
lipid D/H using a laboratory culture approach. We used several strains of Escherichia coli with single gene<br />
mutations that perturb major NADPH-related reactions within glycolysis, TCA, or pentose-phosphate metabolic<br />
pathways. These strains were grown separately on the diastereomers glucose and galactose. Even though glucose<br />
and galactose have similar structural formulas, they are metabolized in dramatically different ways in E. coli.<br />
Metabolism by glucose increases activity in glycolysis than metabolism by galactose, which increases activity in the<br />
TCA cycle. We hypothesize that metabolism by galactose will yield lipids more enriched in deuterium than<br />
metabolism by glucose due to an isotope effect associated with TCA cycle activity. Our results support our<br />
hypothesis, and we observe the fatty acids of E. coli grown on galactose to contain around 10 per mille more<br />
deuterium than for E. coli grown on glucose. In conclusion we demonstrate that growth on substrates within the<br />
same macromolecular class yield fatty acids with different D/H ratios and that this is correlated with TCA cycle<br />
activity in a way that is consistent with previous studies, which show that an increase in TCA cycle activity yields<br />
heavier lipids.<br />
Regioselective Electrochemical C–H Activation<br />
Megan N. Jackson<br />
Mentors: Harry Gray and Alec Durrell<br />
The regioselective chlorination of benzoquinoline has been achieved electrocatalytically through binuclear palladium<br />
complexes. Through bulk electrolysis, the Pd II –Pd II species is oxidized at modest potentials and reacts with chloride<br />
to form a Cl–Pd III –Pd III –Cl intermediate. In the presence of excess benzoquinoline, the Cl–Pd III –Pd III –Cl species<br />
undergoes reductive elimination of 10-chlorobenzoquinoline, regenerating the initial Pd II –Pd II complex. The<br />
chlorinated product has been characterized by NMR and GC–MS. Additionally, the scope of these palladium<br />
electrocatalysts in C–H functionalization reactions has been explored.<br />
Investigating AND Gate Topologies to Aid in Proteomic Analysis<br />
Granton Jindal<br />
Mentors: David Tirrell and Alborz Mahdavi<br />
Proteins are vital cellular components that are used for diverse functions including mobility, signaling, and<br />
metabolism. Understanding protein expression levels and dynamics is critical to determining how cells function.<br />
Examining the proteome of cells which are in environments defined by multiple conditions (temperature, pH) can<br />
be done using a split MetRS-NLL system. The MetRS-NLL protein has been shown to be able to incorporate<br />
Azidonorleucine (ANL), a homolog of Methionine (Met) with an azide group. Although the split system of MetRS has<br />
been examined, the split system of MetRS-NLL has not been evaluated to date. When split at the same residue<br />
367 that the MetRS has been split at, it is expected for each part of the protein to have no activity, while gaining<br />
activity when both parts are expressed in the cell. This can be tested by seeing if ANL is incorporated using the<br />
Click-iT reaction with the TAMRA alkyne. From another library of split sites, the MetRS enzyme split at residue<br />
269 was determined to have significant activity. To quicken the self-assembly of the two parts of the MetRS-NLL,<br />
key amino acid mutations could be made in the protein such that the two parts are attracted together with greater<br />
thermodynamic stability.<br />
4D Gene Expression of S. purpuratus Throughout Embryogenesis<br />
Jennifer Kang<br />
Mentors: Eric Davidson and Emmanuel Faure<br />
The validated gene regulatory network (GRN) that encodes the specification functions of the sea urchin embryo (S.<br />
purpuratus) is approaching global dimensions in that it now includes all but one embryonic territory. The canonical<br />
cell lineage allows for the construction of a 4D (3D + time) digitally imaged embryo on which spatial patterns of<br />
regulatory gene expression are superimposed at different developmental stages. Mapping the gene regulatory<br />
networks involved in sea urchin embryogenesis allows researchers to determine the sequential processes of cell<br />
specification, and provides information on combinatorial interactions between different genes that may result in a<br />
particular cell fate. However, gene expression in the spherical embryo often presents a misleading depiction of the<br />
cellular processes that are going on when analyzed in 2D. We present a computational method permitting a 4D<br />
visualization of the GRN linkages responsible for any current regulatory state in the anatomical context of the<br />
embryo morphology. This is done by whole mount double fluorescence in situ hybridization on fixed embryos from<br />
6-40 hours post fertilization, as this time frame contains the gene activity of interest. The particular<br />
endomesodermal genes we are working on include FoxA, Hox11/13B, Blimp, GCM and Hnf6. The embryos will be<br />
analyzed using a confocal microscope and the images can then be stacked to created a 4D representation of gene<br />
expression during embryogenesis.<br />
123
Role of Extraneous Layer in Enhanced UV Resistance<br />
Jordan E. Krebs<br />
Mentors: Kasthuri Venkateswaran and Parag Vaishampayan<br />
Bacillus horneckiae, an extraneous layer-bearing spore-forming bacteria, was isolated from a clean room where<br />
Phoenix mission components were assembled. Previous studies demonstrated that B. horneckiae spores are<br />
resistant to UV radiation up to 1000 J m -2 . This extraneous layer may act as a passive shield enhancing higher UV<br />
resistance in these spores. The electron microscopy results showed that this layer enables the spores of this<br />
bacterium to attach to each other and formed biofilm-like monolayers when dried on spacecraft-qualified metals<br />
which would further shield the spores from UV radiation. To further characterize biochemically, procedures to purify<br />
the extraneous layer with a French Press were optimized. The survival of the spores with and without extraneous<br />
layer under UV irradiation condition was performed. Preliminary results showed that French Press treated spores<br />
exhibited greater survival against UV irradiation than untreated spores but more study is warranted. Furthermore,<br />
to elucidate whether the layer plays any part in the enhanced UV resistance in B. horneckiae spores, purified<br />
extraneous layer will be mixed with B. subtilis, that lack this layer. Determining what structures, biological byproducts,<br />
or molecular pathway enable great UV resistance is a key to developing strategies to eradicate<br />
extremophiles associated with spacecraft, thus potentially prevent forward contamination.<br />
Mapping the Neural Circuitry of the Drosophila Fly Brain<br />
Christine M. Mak<br />
Mentors: David J. Anderson and Allan Wong<br />
As a model organism, the fruit fly Drosophila melanogaster provides a powerful research tool to study the neural<br />
circuitry underlying behaviors exhibited by higher organisms, from simple reflexive behaviors to more complex<br />
social behaviors. Previous studies have established primary wind sensory neurons in the antenna that map to the<br />
fly brain. Here, we focus on identifying the second-order wind neurons and their functional role in the neural<br />
circuit. We express a genetically encoded calcium indicator, GCaMP3, which labels the antennal mechanosensory<br />
and motor center (AMMC) of the fly brain for neural activity during the delivery of a wind stimulus. Using calcium<br />
imaging under a two-photon microscope, we measure change in fluorescence corresponding to neuronal activation<br />
or suppression. Further examination of these neurons’ projections is done using a photo-activatible green<br />
fluorescent protein (PA-GFP), which allows for visualization of the neural circuit via protein diffusion. From a<br />
behavioral perspective, we also carry out a behavioral assay to test for the result of silencing these wind neurons.<br />
Normally, flies experience wind-induced suppression of locomotion (WISL), but inactivating the second-order wind<br />
neurons of interest should eliminate this phenomenon. Our results will contribute to an understanding of how<br />
neural circuits in the brain work to induce a specific behavioral response.<br />
Characterizing the T Cell Response to Cancer Immunotherapies<br />
Leanna Morinishi<br />
Mentors: James Heath, Gwen Owens, and Chao Ma<br />
Adoptive cell transfer (ACT) immunotherapy is the transfer of immune-derived cells to a host, with the intention of<br />
transferring immunologic characteristics. It has been used to treat patients with advanced tumors, including<br />
melanoma. Due to the high level of heterogeneity in functional immune cells, identifying the efficacy of the<br />
treatment is difficult, but made possible through the Single Cell Barcode Chip (SCBC) based on the DNA-encoded<br />
antibody library (DEAL) approach to spatial immuno-sandwich assays. A noted problem (through studies involving<br />
SCBC) with the ACT immunotherapy is the short-lived maximum tumor-killing function of the infused engineered T<br />
cells. Ongoing in vitro studies and work in mouse models in Dr. David Baltimore’s lab are focused on improving the<br />
acquired immunologic response. A new SCBC with an array of mouse DEAL conjugates is necessary to assess the<br />
progress of these experiments. This summer, I prepared a functional SCBC for mice secretion proteins for initial<br />
testing with hematopoietic mouse stem cells, and for eventual use in probing the efficacy of in vivo ACT<br />
experiments in mouse models.<br />
Searching for the Genes: Could New Species of Wild Microbes Provide Improved Alcohol Tolerance?<br />
Nathan D. Morison<br />
Mentor: Jared R. Leadbetter<br />
Using basic enrichment techniques, novel microbes tolerant to the important biofuel isobutanol were enriched from<br />
natural sources. The details of solvent tolerance to isobutanol, n-butanol, and other potential biofuels are<br />
discussed. Comparisons between aerobes and anaerobes isolated from similar environments through the creation<br />
of different environments are considered. Genetic information is provided where it is available.<br />
124
DNA Ejection Mechanisms of Bacteriophage Lambda<br />
Caitlin Regan<br />
Mentors: Rob Phillips and Yi-Ju Chen<br />
Bacteriophages have long provided a means for studying viral infection into bacteria. In this project, we aim to<br />
focus on the basic mechanism of infection: viral DNA ejection. By staining the viral DNA, we can watch the transfer<br />
of fluorescence from the viral capsid into an E. coli cell. In this single cell in vivo experiment we can quantify the<br />
fluorescence transfer, and thus have a metric for measuring DNA ejection. By adding a transcription inhibiting<br />
antibiotic, Rifampin, we aim to determine if transcription is involved in the viral DNA ejection mechanism.<br />
Improving the KM-Value for Isobutyraldehyde of Lactococcus lactis Alcohol Dehydrogenase Variant<br />
Ll_AdhA RE1-his6 in Yeast by Directed Evolution<br />
Tatyana E. Saleski<br />
Mentors: Frances H. Arnold and Sabine Bastian<br />
Isobutanol is well suited for replacement or supplementation of fossil fuels. A recently engineered high-yield, highspecificity<br />
pathway enables glucose to isobutanol conversion in a recombinant organism. For this pathway to be<br />
economically competitive with bioethanol production, it must be able to operate under anaerobic conditions. In<br />
order to resolve a cofactor imbalance that hindered anaerobic performance, the pathway’s alcohol dehydrogenase<br />
(ADH) was replaced with Lactococcus lactis ADH, AdhA, which was improved for its role in the pathway using<br />
directed evolution methods. The ADH’s substrate affinity is currently believed to be one of the limiting factors in the<br />
pathway’s efficiency. Here we have used directed evolution methods to engineer variants with lower substrate<br />
affinity. We have characterized variants in the lineage leading to our current best variant Ll_AdhA 29C8-his6 to<br />
investigate their substrate specificity change over the course of the evolution project.<br />
The Efficacy of Fe(II) Chelation Versus the Growth and Dispersal of Pseudomonas aeruginosa Biofilms<br />
Tahoura Samad<br />
Mentors: Dianne Newman and Ryan Hunter<br />
In cystic fibrosis patients, Pseudomonas aeruginosa is a significant cause of mortality. This is due, in part, to the<br />
ability of this bacterium to form biofilms –bacterial communities that are notoriously resistant to antibiotics. An<br />
attractive target to supplement current therapeutics is iron acquisition, which is essential to many aspects of<br />
bacterial metabolism. Recent studies have focused on use of Fe(III) chelators to reduce P. aeruginosa biofilm<br />
formation. However, P. aeruginosa is able to secrete small molecules known as phenazines which may change iron<br />
speciation, from predominately Fe(III) to Fe(II). We therefore wanted to test whether Fe(II) chelators can prevent<br />
the formation of biofilms and/or promote the dispersal of those already established. As a control, we first<br />
ascertained whether Fe(II) chelators prevented growth altogether. Consistent with prior experiments, the Fe(III)<br />
chelators lactoferrin, conalbumin, and Desferoxamine and Fe(II) chelators 2,2 dipyridyl and Ferrozine did not<br />
impact planktonic growth. Now, using a high-throughput biofilm growth, we are investigating whether the Fe(II)<br />
chelators described above can disperse and/or prevent the formation of P. aeruginosa biofilms. If so, they should<br />
be considered as a possible novel therapy that might be more effective than Fe(III) chelators at later stages of CF<br />
disease progression.<br />
Elucidating the Physiological Role of Mir-125b in Hematopoiesis and Immunity<br />
Nikita Sinha<br />
Mentor: David Baltimore<br />
MicroRNAs (miRNAs) have been identified as important regulators of gene expression, and they function by<br />
repressing specific target genes at the post-transcriptional level. MicroRNA-125b (miR-125b) is up-regulated in<br />
patients with leukemia. Preliminary data suggests that over-expression of miR-125b disrupts the homeostasis<br />
between myeloid and lymphoid immune cells. Here we used a miR-125b sponge loss-of-function approach to<br />
uncouple the role of miR-125b in tumorigenesis from its potential physiological function in hematopoiesis. For our<br />
loss of function system, we designed a high copy vector expressing several anti-sense microRNA sites that should<br />
act as a decoy, attracting the microRNA and keeping it away from its natural targets. We found that the<br />
reconstituted sponge-expressing mice had significantly less white blood cells, myeloid cells, granulocytes, B cells, T<br />
cells, and pre-erythrocytes. Overexpression of miR-125b target Lin28 in the mouse hematopoietic system<br />
mimicked the phenotype observed upon inhibiting miR-125b functions, leading to a decrease in hematopoietic<br />
development. Since endogenous expression of miR-125b targets Lin28, Trp53inp1, and BMF was de-repressed by<br />
the miR-125b sponge decoy in progenitor-enriched bone marrow, these pro-apoptotic genes may also collaborate<br />
with Lin28 inhibition during miR-125b tumorigenesis. In conclusion, we show here that miR-125b physiologically<br />
regulates hematopoiesis.<br />
125
Measuring Dynamic Phosphorylation of Spo0A in Bacillus subtilis<br />
Arvind Thiagarajan<br />
Mentors: Michael Elowitz and Joseph Levine<br />
Sporulation is an intricately controlled process in Bacillus subtilis, promising a potential wealth of novel network<br />
motifs. It has previously been shown that the key regulator involved in the induction of sporulation is the<br />
phosphorylated form of the transcription factor Spo0A. We have developed an experimental paradigm to measure<br />
Spo0A P levels dynamically in vivo. A strain of B. subtilis was engineered in which the Spo0A gene had been<br />
replaced by a Spo0A-GFP fusion and into which an array containing 256 copies of a Spo0A P binding site had been<br />
inserted. Imaging of this strain revealed clusters of localized fluorescence amidst diffuse background fluorescence,<br />
corresponding respectively to bound Spo0A P and unbound Spo0A (P) . In order to determine relative levels of Spo0A P<br />
and Spo0A from such images, a machine learning algorithm for identifying localized fluorescence and quantifying<br />
the corresponding intensities was developed. A probabilistic model was developed to estimate intracellular<br />
deviations in cluster intensities and consequently determine the ratio between observed fluorescence and number<br />
of molecules present. This paradigm will be used to study Spo0A P dynamics during sporulation, and so a media was<br />
optimized to induce sporulation. Finally, a microfluidic chemostat was optimized for the study of sporulation in B.<br />
subtilis.<br />
Establishing an Efficient Cell Sorting and Release Protocol<br />
Christine E. Wu<br />
Mentors: James R. Heath and Young Shik Shin<br />
Between each cell in human tumors, many genetic and epigenetic differences exist. Consequently, every cancer cell<br />
may respond differently to treatment, with some resistant to certain drugs. In order to fully study the drug<br />
resistant pathways in primary cancers, different cells must be sorted to be examined separately. The Heath Lab<br />
has developed the technique of DNA Encoded Antibody Libraries, which uses a chip-based microarray platform for<br />
cell sorting. Through this procedure, an antibody array is generated by hybridizing single-stranded DNA antibody<br />
conjugates onto slides patterned with complementary single-stranded DNA. The cell mixture is then applied to the<br />
antibody array and cells with surface markers of interest will adhere to the chip. While this technique has<br />
successfully isolated cells which overexpressed the epidermal growth factor receptor from primary glioblastoma<br />
biopsies, the releasing of the desired cells from the plate was met with some problems. With a combination of<br />
trypsin and DNase, coupled with a wash by pipetting, cells from a mouse model were successfully released from<br />
the slides and then applied to a single-cell barcode chip for protein detection. With the ability to release captured<br />
cells, isolating and studying certain cell types within a cancer can be easier to accomplish.<br />
Modeling of [FeFe]-Hydrogenase Using a Peptide Ligand<br />
Yuanchi (Victor) Zhao<br />
Mentors: Theodor Agapie, Kyle Horak, and Sandy Suseno<br />
Hydrogenases are enzymes that catalyze the interconversion of protons and dihydrogen using inorganic centers<br />
containing iron or nickel. The study of hydrogenases could inspire the development of novel catalysts based on<br />
earth-abundant, first-row transition metals capable of supplanting unsustainable platinum electrocatalysts currently<br />
used as a means of hydrogen production. One method for learning about the hydrogenases is through synthetic<br />
active site mimics. Work done towards assembling an [FeFe]-hydrogenase H-cluster mimic using a peptide scaffold<br />
is ongoing. Short peptides designed from conserved amino acid sequences of the iron-sulfur cluster-containing<br />
ferredoxin proteins were synthesized. Additionally, a diiron subsite mimic from the literature has been synthesized.<br />
Future work includes the incorporation of a [4Fe4S] cluster into the peptide scaffold. Linking the peptide-bound<br />
cluster to the diiron subsite mimic via a cysteinyl bridge will complete the [FeFe]-hydrogenase H-cluster mimic.<br />
Further characterization efforts include EPR and IR spectroscopy, electrospray mass spectrometry, and cyclic<br />
voltammetry.<br />
126
LIGO<br />
THE LASER INTERFEROMETER<br />
GRAVITATIONAL-WAVE OBSERVATORY<br />
L<br />
I<br />
G<br />
O
Crackling Noise in Blade Springs<br />
Vanessa Acon<br />
Mentors: Rana Adhikari, Tara Chalermsongsak, Alastair Heptonstall, Mingyuan Huang, and Seiji Kawamura<br />
Crackling noise arises in various physical systems, implying a nonlinear conversion of energy from slow changing<br />
external conditions into discrete 'crackling' events at higher frequencies. The mirrors in Advanced LIGO are<br />
suspended with a series of blade springs and mechanical flexures. At the sensitivity levels of Advanced LIGO, we<br />
are concerned with any noise in that suspension system, in this case with crackling noise in the blade springs. Our<br />
goal is to measure and characterize this crackling noise and determine its relevance to LIGO measurements. To do<br />
so, we construct a low-noise table-top Michelson interferometer with blade springs such that when they are driven<br />
in unison, any change in the output signal will be due to crackling. In order to extract a crackling signal on the<br />
order of the sensitivity of LIGO itself, we demodulate the signal using a "chopping" technique. Currently we have<br />
measured characteristic values for the blade springs, designed the low-noise Michelson apparatus, designed and<br />
simulated the chopping technique, and calculated an initial noise budget for the experiment. Subsequent work will<br />
be to continue to improve the low-noise design, complete the low-noise apparatus assembly, and collect crackling<br />
noise data. We can then derive the characteristic coefficients telling us the relationship between crackling and<br />
driving force, and compare our data with the current LIGO noise budget to see if crackling will pose a significant<br />
source of noise for LIGO measurements.<br />
Gravitational Radiation From Compact Binaries in the Massive Brans-Dicke Theory of Gravity<br />
Justin A. Alsing<br />
Mentors: Emanuele Berti and Yanbei Chen<br />
Most high-energy physics extensions to General Relativity predict the existence of scalar fields. For this reason, the<br />
investigation of gravitational radiation in scalar-tensor theories may provide important insights into high-energy<br />
modifications of Einstein's theory.<br />
We have computed the scalar and tensor contributions to the gravitational radiation from compact binary systems<br />
in the massive Brans-Dicke theory, and used recent observations of radiation damping in mixed binary systems to<br />
put stringent bounds on the parameters of the theory. The calculation divides neatly into four sections; the field<br />
equations of the massive Brans-Dicke theory are obtained, and linearized in the weak-field limit. Post-Newtonian<br />
expressions for the scalar and tensor fields are then derived from the resulting field equations. Using these results,<br />
the post-Newtonian equations of motion and the periastron shift of compact binaries are derived. The scalar and<br />
tensor gravitational waveforms are then obtained by solving the linearized field equations, and using these results<br />
together with those obtained in previous stages of the calculation we have derived an expression for the<br />
gravitational radiation damping of compact binary systems. Finally, with the result for the radiation damping in<br />
hand, we have used recent observations of radiation damping in mixed binary systems to put tight bounds on the<br />
parameters of the theory.<br />
Study of Fundamental Quantum Noise in Advanced Gravitational-Wave Detectors<br />
Anthony Bartolotta<br />
Mentors: Yanbei Chen and Haixing Miao<br />
Gravitational waves are weak phenomena; therefore, the effectiveness of laser interferometer gravitational wave<br />
detectors is very susceptible to noises present in the system. The ultimate limit on the effectiveness of the detector<br />
is noise produced by quantum effects. There are two primary sources of this quantum noise, the shot noise and the<br />
radiation pressure noise. Shot noise is decreased as the laser power is increased; however, this increases the<br />
radiation pressure noise. As such, these two sources of quantum noise, if not correlated, result in the Standard<br />
Quantum Limit. Advanced gravitational-wave detector schemes are designed to surpass the Standard Quantum<br />
Limit at their most sensitive detection band; of these proposed configurations, we have been considering the local<br />
readout scheme. The local readout scheme operates by measuring the local motion of the test mass with an<br />
auxiliary carrier light. The investigation of this scheme began with an analysis of Fabry-Perot cavities. This was<br />
expanded upon with an analysis of coupled Fabry-Perot cavities. This analysis was done by determining the<br />
transfer functions of each component and using MatLab to numerically determine the total transfer function. This<br />
provided the framework for analyzing the local readout detection scheme.<br />
Real-Time Decision-Making for Gravitational Wave Detection<br />
Christopher M. Biwer with Nick Fotopolous, Larry Price, and Leo Singer<br />
Mentors: Alan Weinstein and Roy Williams<br />
We have designed a real-time coincidence detector, a program that searches notifications of recent astrophysical<br />
observations for coincident events, that can be connected to the LIGO low-latency pipeline. The principle objectives<br />
were to determine time, and spatial models for gamma-ray bursts, and then implement methods to identify these<br />
coincidences in real-time using the existing advanced LIGO prototype infrastructure. The primary functions of the<br />
129
coincidence detector are to query remote databases, decide if any coincidences exist, and publish notifications; a<br />
modular design was implemented to allow easy manipulation, and addition of new models. Time coincidences are<br />
found by searching time windows that encompass the range of expected time delays between different<br />
messengers. Coincidences in space are determined by computing the likelihood ratio statistic for shared areas of<br />
the sky. Subsequent analysis of trial runs, and injected coincidences were conducted to evaluate performance.<br />
Properties of Neutron Stars, Their Pulsational Modes, and Gravitational Wave Emission<br />
Iryna Butsky<br />
Mentors: Christian Ott and Jeff Kaplan<br />
In this project, we investigate the different proposed equations of state (EOS) for neutron stars. We create mass -<br />
radius plots of the different EOS, along with constraints to find if any are inconsistent with theory or observation.<br />
Using the Spectral Einstein Code (SpEC), we evolve a spherical neutron star model to observe the oscillations in<br />
central density. Performing a Fourier transform on these oscillations allows us to extract the pulsational frequencies<br />
which are perturbed with a quadrupole perturbation. The theoretical f-mode frequencies are then obtained using<br />
perturbation theory.<br />
Ultra-Stable Fabry-Perot Cavities for Coating Thermal Noise Characterization<br />
Raphael Cervantes<br />
Mentors: Frank Seifert, Rana Adhikari, and Tara Chalemsongsak<br />
The performance of high-precision optical interferometric experiments is compromised by many noise sources. One<br />
of these noise sources is the coating thermal noise. The characterization of the coating thermal noise limit in<br />
various coatings could be used to verify the current theoretical model and to help minimize it for future<br />
experiments. Currently, the reference cavity noise experiment aims to use a small table-top setup with a fast turn<br />
around for measuring the coating thermal noise. The experiment consists locking a laser to a reference Fabry-Perot<br />
cavity using Pound-Drever-Hall locking. The same laser is sent to another cavity and the difference in frequencies<br />
between the output beams is measured, and thus the length fluctuations of the cavity are characterized. Although<br />
there are other noise sources, they will be minimized so that the coating thermal noise is dominant in the<br />
frequency region of interest. Cavity mirrors with different coatings can then be tested to characterize the<br />
fluctuations caused by various coating materials and/or designs.<br />
Third-Generation Gravitational Wave Detector Placement<br />
Michael Coughlin<br />
Mentor: Jan Harms<br />
One of the most immediate challenges associated with third-generation gravitational wave detectors is to select<br />
site candidates. There are numerous factors that must be taken into consideration, including surface topography,<br />
seismicity, population density and many more. The project analyzes a number of US-wide data sets, including<br />
seismic, wind, topography, and geology, as well as seismic data from around the world. The combination of these<br />
data sets indicates the suitability of possible site locations. Other analyses with the seismic data include tracking<br />
seismic noise levels at various frequencies over large periods of time as well as microseismic peak correlation in<br />
various locations.<br />
Probing Strong-Field Gravity: Constraining the Number of Gravitational Wave Polarizations<br />
Bryant Garcia<br />
Mentors: Larry Price and Stephen Privitera<br />
General Relativity predicts only two tensorial gravitational wave polarizations. More general theories, however,<br />
contain up to six independent polarizations. Observations of these four non-tensorial polarizations could severely<br />
constrain alternative theories of gravity and provide a unique strong-field test for GR. As the 2 nd generation<br />
gravitational wave detectors, advanced LIGO and VIRGO, come online, we expect to begin compiling a list of<br />
known GW signals. We present a model selection analysis for constraining the number polarizations present in the<br />
gravitational field. We investigate the possibility of detecting these alternate polarizations using the next<br />
generation of gravitational wave detectors.<br />
Investigation of Noise in Photodiodes Meant for a Gravitational Wave Interferometer<br />
Matthew S. Gilmer<br />
Mentors: Valery Frolov and Chris Guido<br />
This research focuses on the isolation of excess noise in photodiodes from other noise sources through noise<br />
cancellation techniques. The noise spectra from two photodiodes were simultaneously monitored; a beam splitter<br />
facilitated the investigation by allowing the same laser source to be incident on the two detectors at once. The<br />
research includes the use of a feedback system and several other noise reduction techniques. These were<br />
130
performed for many different photodiodes. The experiment was done in a clean room, the Optics Lab at the LIGO<br />
Livingston Observatory. Data was taken and analyzed in the context of the 4km Gravitational Wave Interferometer<br />
to be used in the upcoming project, Advanced LIGO.<br />
Spin Effects in the Gravitational-Wave Memory for Quasi-Circular Inspiralling Compact Objects<br />
Xinyi Guo<br />
Mentors: Marc Favata and Yanbei Chen<br />
The gravitational-wave memory effect is a time-varying but non-oscillatory contribution to the gravitational-wave<br />
amplitude. When a gravitational-wave with memory passes through an ideal detector, it will cause a permanent<br />
displacement of the test masses. The nonlinear form of the memory arises from the gravitational waves produced<br />
by previously emitted gravitational waves, and is present in virtually all gravitational-wave sources. It is also<br />
known to affect the gravitational-waveform at leading order despite the fact that it originates from higher-order<br />
interactions. Thus understanding the memory is important for building our accurate knowledge of the gravitationalwaveforms<br />
in order to probe the nonlinearity of general relativity. Previous calculations of the memory have only<br />
considered non-spinning binaries. However, most compact binaries have spinning components; these spins will<br />
significantly modulate the gravitational-wave amplitude and phase. We studied the effect on the memory from the<br />
spin-orbit interaction and calculated the corrections to the memory waveform through 1.5 post-Newtonian (PN)<br />
order. We found that the spin correction starts to enter at 1PN order and can contribute a ~20% correction to the<br />
memory.<br />
Measuring the Vibration Isolation of the Suspended Tip-Tilt Mirrors in the LIGO 40m Prototype<br />
Nicole L. Ing<br />
Mentors: Koji Arai, Rana Adhikari, and Jameson Graef Rollins<br />
The Tip-Tilt (TT) mirrors are used in the LIGO 40m prototype to geometrically fold the power recycling cavities. We<br />
will evaluate the effectiveness of the Tip-Tilt suspension in reducing vibration isolation by injecting horizontal and<br />
vertical motion and measuring the relative displacement between the mirror and the suspension frame. Since<br />
current plans for Advanced LIGO include the use of a similar type of Tip-Tilt suspensions, our evaluation of will be<br />
useful for the Advanced LIGO design.<br />
Advanced LIGO Output Mode Cleaner Piezo Actuator Noise<br />
Brian J. Koopman<br />
Mentors: Valera Frolov and Ryan DeRosa<br />
The Laser Interferometer Gravitational Wave Observatory (LIGO) in Livingston, Louisiana was built with the<br />
intention of one day detecting gravitational waves caused by cosmological events such as binary inspiral and<br />
supernovae. The interferometer is a 4km long Fabry-Perot Michelson interferometer. The readout of the<br />
interferometer is located in the Output Mode Cleaner (OMC), within which a piezo actuator is attached to one of the<br />
mirrors. The piezo actuator can introduce a new source of noise to the interferometer readout, a noise that we<br />
currently know little about. By utilizing a small Michelson interferometer with two piezo actuators, one in each arm,<br />
this source of noise can be studied.<br />
By driving the two piezo actuators in common mode, keeping the differential arm length the same, a peak in the<br />
amplitude spectral density can be created at 1Hz. This creates harmonic peaks at higher frequencies. When the 1<br />
Hz peak is suppressed by locking signal the harmonics remain, demonstrating a phenomenon known as up<br />
conversion. Understanding this upconversion will lead to understanding how much noise the piezo actuators<br />
present in the differential arm (DARM) readout.<br />
Testing General Relativity Using Gravitational Waves From Spinning Neutron Stars<br />
Christina Lee<br />
Mentor: Alan Weinstein<br />
The direct detection of a gravitational wave with the Advanced LIGO detectors provides the opportunity to measure<br />
departures from General Relativity. These departures can arise in the speed of the gravitational wave, existence of<br />
alternate polarizations, and parity violation. To measure these, I simulated a single detector measurement of a<br />
continuous gravitational wave from a well defined pulsar source, for example the Crab pulsar, due to an<br />
asymmetry in the moment of inertia. The Doppler frequency modulation allows determination of the speed to a<br />
part in 10 -6 for a signal with a sufficiently high signal to noise ratio. If the speed differs by much more than this<br />
value, the signal extraction technique no longer works. I also measured the ability to distinguish two additional<br />
polarizations, the breathing and longitudinal, from the standard plus and cross. If in propagation one polarization<br />
handedness is enhanced and the other suppressed due to parity violation it would result in an anomalous<br />
measurement of the inclination angle of the neutron star spin with respect to the line of sight, which is known from<br />
X-ray observations.<br />
131
Modeling the Advanced-LIGO Response to Gravitational Waves and Calibration<br />
Benjamin C. Li<br />
Mentor: Alan Weinstein<br />
The response of the Advanced LIGO detectors (currently under construction) to gravitational waves depends in a<br />
complex way on the optical configuration of the laser interferometer at the heart of the detector. This can change<br />
due to intentional and unintentional changes to the configuration parameters. Precise calibration of that response,<br />
required for optimal extraction of information about the gravitational wave source, will require us to carefully<br />
monitor and understand those configuration parameters. This project aims to identify the most important<br />
parameters and the best methods for measuring them. This will be done through careful modeling and simulation,<br />
using existing Matlab-based software.<br />
Investigating Gravitational Waves From the Ringdown of Accreting Black Holes<br />
Ryan C. Lynch<br />
Mentor: Gregory Mendell<br />
The characteristic strain produced by black hole ringdowns is well-known, taking the form of a damped sinusoid.<br />
The frequency of oscillation of this strain function is dependent upon the inverse of the source’s mass, and<br />
therefore it should decrease over time as the source black hole accretes mass. Currently, the data pipelines used to<br />
conduct match-filtered searches for ringdowns implement static-mass templates: templates with constant<br />
frequencies. Here, signal templates are designed with dynamically changing mass to account for mass accretion,<br />
allowing the expected frequency of oscillation in strain to vary over time. Simulations are also run to search for<br />
ringdowns over single and multiple detectors, using both static-mass and dynamic-mass templates. By comparing<br />
the results of these searches, it is clear that while mass accretion does make static-mass templates less efficient<br />
than dynamic-mass ones, this difference is only significant in cases of extremely high accretion rates that greatly<br />
exceed the Eddington limit. Thus, static-mass templates should be capable of detecting black hole ringdowns as<br />
they are currently understood.<br />
Identifying Correlated Detector Noise Contaminating Searches for Stochastic Gravitational Wave<br />
Backgrounds<br />
Gabriella Martini<br />
Mentors: Nick Fotopoulos and Alan Weinstein<br />
The Stochastic Gravitational Wave Background (SGWB) is the undetected gravitational radiation counterpart to the<br />
cosmic microwave background radiations (CMBR). The theoretically most sensitive SGWB searches involve colocated<br />
pairs of detectors, like the LIGO Hanford interferometers. Unfortunately these detectors are immersed in a<br />
common noisy environment which induces correlations in the data that can mimic that of an SGWB. Much of the<br />
correlation can be identified and removed using data by cross-correlating physical environment monitors (PEMs)<br />
such as seismometers, voltmeters, etc. with the gravitational wave channels. In the past, stochastic searches have<br />
employed crude approximations of the environmental contamination to avoid the computational heavy-lifting<br />
involved in the complete calculation. This project analyzes the full NxN cross-coherence matrix for the improved<br />
calculation of the environmental contribution to detector correlation. Preliminary results show that clustering<br />
algorithms identify a block structure amongst the PEM coherences and indicate the possibility of a significant<br />
reduction in computation.<br />
Optical Gyroscopes for the LIGO Gravitational Wave Detectors<br />
Zoe L. Masters<br />
Mentors: Rana Adhikari, Alastair Heptonstall, and W. Zach Korth<br />
Seismic isolation of the mirrors in the LIGO detectors can improve the detectors’ performance at low frequencies.<br />
Because seismometers confuse rotation with horizontal motion, however, optical gyroscopes will be added to the<br />
advanced LIGO detectors to measure rotation alone. The goal of this project is to improve electronics and reduce<br />
mechanical noise in a laser-based optical gyroscope operating on the Sagnac principle and locked using the Pound-<br />
Drever-Hall technique. Specific improvements include a new circuit board containing generic filters for use in the<br />
locking system, prototype tests and simulations of a possible thermal stabilization method, and tests of the<br />
gyroscope’s current functionality.<br />
An Optical Follower Servo for the LIGO Photon Calibrators<br />
Rolf Minton<br />
Mentors: Richard Savage and Paul Schwinnberg<br />
The LIGO Photon Calibrator currently uses a power modulated auxiliary laser to induce periodic displacements of<br />
the 40 kg end test masses on the order of 10 -18 m via radiation pressure. The optical follower servo reduces<br />
unwanted modulation harmonics, lowers relative power noise, and significantly increases the usable fraction of the<br />
auxiliary laser power. The response of the servo actuator, an acousto-optic modulator, saturates by a factor of 30<br />
132
over the range of injected laser power modulations desired. Modeling and lab measurements indicate that variable<br />
gain in the servo electronics can compensate for this actuator’s saturation. A prototype servo amplifier has been<br />
constructed and tested. The experimental results are presented and discussed.<br />
Diagnosis of the 40m Prototype Interferometer With Auxiliary Laser-Beam Injection<br />
Sonali Mohapatra<br />
Mentors: Rana Adhikari and Suresh Doravari<br />
The aim of this project is the deterministic locking and characterization of the arms of the LIGO 40m prototype<br />
interferometer. In order to obtain low-loss high-reflectance mirrors, LIGO uses dielectric coatings on the test<br />
masses. The Pound-Drever-Hall technique is employed to lock the laser to the cavity or vice-versa. The high<br />
sensitivity of the PDH lock limits its linear range which makes it difficult to bring the cavity into resonance with the<br />
laser. As a solution to this problem, this report describes an auxiliary-beat lock (green or IR) to give us a much<br />
broader linear range. This enables us to smoothly change the cavity length and thus lock the laser to the cavity by<br />
obtaining a resonance. The 532 nm green laser light is obtained by frequency doubling of the 1064nm NdYAG<br />
auxiliary laser. This frequency doubled green beam from the AUX laser has been previously used to obtain a greenbeat<br />
lock between the Pre-Stabilized-Laser and the AUX laser. We use the IR beam from the AUX laser, before it is<br />
frequency doubled and obtain an IR beat lock. The frequency noise of the IR-beat lock and the Green-Beat lock are<br />
compared.<br />
Exploration of the Latest Numeric-Relativity-Inspired Waveforms for Compact Binary Coalescence<br />
Jessie Otradovec<br />
Mentors: L. Santamaría, P. Ajith, and A. Weinstein<br />
Coalescing compact binaries are among the most promising sources of gravitational waves (GWs), which makes<br />
them a particularly interesting problem for GW astrophysics. The signature of these systems encodes a great deal<br />
of information about their parameters: masses, spins, orbital angular momentum and sky position. Although there<br />
is no analytically exact solution of such systems in full General Relativity, recent breakthroughs in Numerical<br />
Relativity have allowed the computation of handfuls of waveforms at the late stages of the binary coalescence<br />
(near the merger and ringdown) for multiple points in the parameter space. Together with post-Newtonian<br />
methods and black hole perturbation theory, it is now possible to construct waveforms for the full coalescence<br />
process. In turn, these waveforms are used in search algorithms and parameter estimation in GW data analysis.<br />
This study seeks to validate different waveform families by comparing the overlap, normalized difference in signalto-noise<br />
ratio (SNR) and chi-square. In addition, visual inspection for expected behavior is performed. Preliminary<br />
findings document differences in conventions, and ~6%-15% maximum disagreements in SNR. Additionally, worse<br />
agreement is found at smaller symmetric mass ratios. Some unexpected behavior of the generic-spin approximant<br />
family has also been discovered.<br />
A New Magnetar Gravitational Wave Search: GW Stacking Using Timing Information From<br />
Electromagnetic Light Curves<br />
Bryance Oyang<br />
Mentor: Peter Kalmus<br />
Soft gamma repeaters (SGR) are thought to be magnetars, young neutron stars with extremely strong magnetic<br />
fields (10 14 -10 15 Gauss). These objects sporadically release large bursts of soft gamma rays when energy in the<br />
magnetar's magnetic field is released by crustal deformation and fracture. This could excite non-radial modes (fmodes)<br />
on the magnetar, which would emit gravitational waves (GW). This project expands a database of bursts'<br />
light curves collected by satellites and uses the information collected to run a search for GWs in LIGO data. Since<br />
we expect the GWs and the gamma rays to arrive in our solar system at roughly the same time, we find the start<br />
times of each burst from the light curves. Then we run a “stacking” search for GWs in LIGO data, where we<br />
combine time-frequency tilings of LIGO data at the times of each burst for hundreds of bursts. Stacking<br />
dramatically improves the sensitivity for GWs under the assumption that the frequencies of the GWs are the same<br />
from burst to burst.<br />
Prototyping Adaptive Feed-Forward Seismic Noise Cancellation at the 40m Interferometer<br />
Ishwita Saikia<br />
Mentors: Rana Adhikari and Jenne Driggers<br />
Advanced LIGO is an upcoming upgrade to the LIGO Project, which we expect to be able to detect gravitational<br />
waves with frequencies as low as 10 Hz. At such low frequencies the detectors will encounter seismic noise,<br />
displacement noise which moves the mirrors (present in the detectors) and thus changes the path length. This<br />
noise is due to the ground motion that occurs by volcanic/seismic activity, ocean tides, human activities, etc. This<br />
project is intended to reduce seismic noise at the 40m Prototype Interferometer with the help of online adaptive<br />
filtering. The technology developed in the 40m laboratory will be applied in Advanced LIGO.<br />
133
Extracting Information About the Central Engine of Core Collapse Supernovae Using Gravitational Wave<br />
Signals<br />
James H. C. Scargill<br />
Mentor: Peter Kalmus<br />
The complete and detailed mechanism which governs the core collapse in certain supernova explosions is unknown.<br />
Different possible mechanisms will cause different waveforms of gravitational radiation to be emitted, and thus<br />
information on the mechanism can be deduced from the received waveform. This project will develop existing ways<br />
in which this information can be extracted. Principal component analysis can be used to identify key waveform<br />
features from a library of known waveforms; Bayesian statistical techniques can then be used to analyse a new<br />
waveform (modelled as received by the detector, i.e. with noise) in terms of these principal components. Current<br />
algorithms can be improved by considering the coherent response of multiple detectors, which would give higher<br />
accuracy and durability in model selection and parameter estimation, as well as extra polarisation and sky-location<br />
information. This in turn will allow quantitative comparison to previous algorithms. Having developed such an<br />
algorithm, we will then test it on new waveforms to determine its efficacy in distinguishing the various mechanisms<br />
which govern core collapse, as the signal-to-noise ratio is reduced, in order to know what science could be done<br />
with gravitational waves in the event of a (galactic) supernova.<br />
Thermal Coupling in Cryogenic Fabry-Pérot Cavities<br />
Jennifer Schloss<br />
Mentor: David Yeaton-Massey<br />
The measurement of extremely weak signals such as gravitational waves requires highly sensitive detectors with<br />
exceptionally low noise levels. In an effort to reduce the contribution of thermal noise due to Brownian motion and<br />
fundamental temperature fluctuations, cryogenic Fabry-Pérot reference cavities are being built. To improve the<br />
temperature-stability of these cavities, it is useful to understand the couplings of outside temperature fluctuations<br />
to the cavity interior. Conductive and radiative heat transfer through cryogenic and ambient-temperature cavities<br />
was modeled using FEA simulation software. Room-temperature simulation output was compared with an<br />
equivalent experimental measurement. The level of agreement provides an indication of the extent to which the<br />
model cryogenic cavity simulations predict the thermal couplings in the physical cryogenic cavities.<br />
Directional Gravitational Wave Search<br />
Clio Sleator<br />
Mentor: Peter Kalmus<br />
Ground-based observatories such as LIGO are expected to directly detect gravitational waves (GWs), which will<br />
begin the age of gravitational wave astronomy. The search methods depend on the astrophysical sources of GWs.<br />
Some sources (specifically violent events in the universe), such as core-collapse supernovae, neutron star collapse,<br />
merging compact binaries, star-quakes, pulsar glitches, and cosmic string cusps, are thought to make bursts of<br />
GWs that last less than 1s. As detector glitches look like burst signals, it is difficult to determine the difference<br />
between them. In this project, a search method will be developed that looks for sky directions with a statistical<br />
excess of triggers. Upon finding such a sky direction, it should be easier to determine whether the glitches seen in<br />
the data are only glitches or if they are actually GW signals. In order to find a sky direction with a statistical excess<br />
of triggers, we will have to calculate the sky direction of each event using time-of-flight triangulation and plot these<br />
sky directions onto skymaps. Using simulated signals will help to characterize the search.<br />
Optimal Telescope Tilings for Electromagnetic Followup of Gravitational Wave Events<br />
Antony J. Speranza<br />
Mentors: Larry R. Price and Leo Singer<br />
With the enhanced detection capabilities of Advanced LIGO, the prospect of multi-messenger astronomy has<br />
become an exciting means for investigating some of the most violent events in the universe. Gravitational wave<br />
detectors are inherently omnidirectional, while EM telescopes can image only small portions of the sky at one time.<br />
In order to make a joint observation of a GW event and an EM counterpart, we need an efficient method for<br />
determining EM telescope tilings that most effectively image the probable sky location of the source, as determined<br />
by the GW detector. For the single telescope case, this can be viewed as a convolution of the telescope's field of<br />
view with the source location skymap. We discuss an algorithm for efficiently computing tilings for a single<br />
telescope, and explore optimization techniques for quickly determining tilings that maximize the detection<br />
probability when multiple telescopes are involved. For both cases, we characterize the detection efficiency as a<br />
function of tiling resolution using a network of EM telescopes that will likely be available during upcoming scientific<br />
runs with aLIGO. The efficiency gains for the multiple telescope algorithms makes a case for coordinated tilings<br />
across the entire EM telescope network.<br />
134
Characterization of Noise in Driven Cantilever Blade Springs<br />
Larisa Thorne<br />
Mentors: Rana Adhikari and Alastair Heptonstall<br />
In the search for gravitational waves, detection is difficult to all but the most accurate and sensitive devices.<br />
Among the myriad of noises that could affect device readings is the “crackling” noise in driven blade springs,<br />
characterization of which is crucial to determining its status as a limiting factor in the shape of the LIGO noise<br />
curve, and ultimately detection and measurement of gravitational waves.<br />
Crackling noises are characteristic of systems which are subject to an outside driving force, and which respond in a<br />
series of discrete and abrupt jumps. Part of the crackling noise effect is due to some sort of hysteresis in the<br />
system. This is logical, considering that the excess energy lost due to hysteresis causes (nonlinear) energy<br />
upconversion in the blade springs which is transitioning from slow driving force applied into the crackling noise at<br />
higher frequencies.<br />
Simulations can only offer so much information. In order to comprehend the underpinnings of the crackling, an<br />
actual experiment must be run. A downscaled Michelson interferometer must be constructed to quantify the<br />
crackling noise. This noise should be revealed in the output signal (after it has been demodulated via a ‘chopping’<br />
technique) when the cantilever blade springs are driven in unison by magnetic actuators, and the set-up has either<br />
been adjusted to its lowest noise level (we will use vacuum tanks and seismic isolation stacks to accomplish this)<br />
or all other sources of noise are known and accounted for. Once the crackling noise has been identified, it will be<br />
possible to determine if it will limit future gravitational wave detection in advanced LIGO (aLIGO).<br />
Study of PowerFlux Response to Continuous Wave Signals of Non-Standard Form and Locations of<br />
Potential Sources in the Milky Way Galaxy<br />
Gregory E. Vansuch<br />
Mentor: Vladimir Dergachev<br />
We present results on the robustness of the PowerFlux algorithm to injected signals with various parameters, in<br />
particular the speed and period of a gravitational wave, as well as the ability of the algorithm to reconstruct<br />
parameters from simulations. We will also describe potential locations of continuous gravitational radiation from<br />
various star clusters and the arms of the Milky Way Galaxy.<br />
Implementing Digital Control for Quad-Maglev Suspension<br />
Yi Xie<br />
Mentors: Haixing Miao and Rana Adhikari<br />
Low-frequency seismic noise prevents us from detecting gravitational waves from many interesting astrophysical<br />
sources. The maglev suspension system is decided to create a soft suspension by using a specially designed<br />
magnetic field. And the digital control system produces feedback control force to make the balance come true.<br />
135
NASA/JPL <strong>Programs</strong><br />
PGGURP<br />
Planetary Geology and Geophysics<br />
Undergraduate Research Program<br />
SPACE GRANT<br />
The National Space Grant College<br />
and Fellowship Program<br />
USRP<br />
Undergraduate <strong>Student</strong> Research Program<br />
JPLSIP<br />
JPL <strong>Summer</strong> Internship Program<br />
N<br />
A<br />
S<br />
A
Transition From Conventional Software Applications to a Web Application Based Platform<br />
Stirling S. Algermissen<br />
Mentors: Maher Hanna, Douglass Alexander, and Helen Mortensen<br />
The Multimission Instrument Processing Lab (MIPL) uses a variety of tools across multiple platforms, languages,<br />
and systems. This diversification is a burden for MIPL and creates overhead when switching between systems for<br />
different tools. A more accessible web interface was explored for decreasing this overhead and making MIPL’s<br />
software tools more versatile and more easily accessible. A website was created to allow the MIPL team to meet<br />
these goals. The marsviewer software, used for viewing image data products produced by the Mars Science<br />
Laboratory and other Mars rovers, is converted to a client-server AJAX interface. This allows access to the tool from<br />
any modern web browser. Other aspects of MIPL’s operation were also moved to the web, such as coordination<br />
with other teams and access to internal use software. The results of this transition are explored, along with a<br />
demonstration of the move of conventional software to the web browser.<br />
Programming Tools for MSL Flight Software Testing<br />
Daniel Alkalai<br />
Mentor: Jim Chase<br />
I am programming tools used to help test the MSL Flight Software as part of the extensive Verification and<br />
Validation process under the cross-cutting subdivision, which deals with system-wide testing concerns rather than<br />
with specific components. Includes an interface allowing users to query the MPCS ground database for operations<br />
and testbed session information, using a configuration file to keep track of important information for the user, as<br />
well as a script that parses through session information and returns statistics about the Flight Software commands<br />
run. The interface to MPCS data would be especially helpful during actual mission operations, so it would still be<br />
used after Flight Software testing has finished.<br />
Maintaining Hardware and Software Cohesiveness for the Cassini Program<br />
Robert Alland<br />
Mentor: Diane Conner<br />
To successfully perform its Solstice Mission until 2017, the Cassini program requires its hardware and software<br />
configurations to match an expected state. Of interest to this project are two areas of system cohesiveness: the<br />
configuration of computers that are part of the Cassini team’s Ground Data System (GDS) and the parameter and<br />
constraint checks performed by the Sequence Generator tool (SEQGEN), which validates and models a proposed<br />
sequence of commands for the spacecraft. The first objective of this project is an adaptation of the Online Asset<br />
software from the JPL Multi Mission Office to automatically produce audits of system configurations – both<br />
hardware and software – and a web interface for viewing and analyzing of the audits. This assists the Cassini team<br />
in satisfying requirements for the JPL IT Security Database and allows it to more easily manage and diagnose<br />
problems with its computers. The second objective of this project is a correction of command checking for a subset<br />
of Attitude and Articulation Control Subsystem (AACS) commands in SEQGEN. This update permits uplinking the<br />
full range of allowable commands to Cassini and prevents commanding errors. This project illustrates the difficulty<br />
and importance in maintaining system cohesiveness for the Cassini program.<br />
Simulations and GUIs: A New Look to Radio Science<br />
Michael Andonian and Carlos Gonzalez<br />
Mentors: Sami Asmar and Kamal Oudrhiri<br />
Radio waves constitute a large portion of the EM spectrum and as such are significant to the process of acquiring<br />
knowledge of the vast universe we live in. The Radio Science Systems Group (RSSG) devotes itself to the analysis<br />
of radio wave phenomena and developing analysis tools to assist. At the present, the RSSG is developing a<br />
command line demodulation tool for processing and analyzing spacecraft signal data from an open-loop radio<br />
science receiver (RSR) while continuing its interest in signal interference amongst other tasks. However, watching<br />
the command line analyze data or drawing figures on a whiteboard to convey notions tends to be monotonous at<br />
best. Surely any user would prefer a graphical user interface (GUI) to work with for processing data or a visual<br />
simulation effectively showing communication between satellites and ground stations. Within this presentation will<br />
be the story of transitioning from analysis in the command line to a graphical user interface, creating simulations<br />
for a radio science experiment involving the International Space Station (ISS) and the Tracking and Data Relay<br />
Satellite System (TDRSS) to form a three-way communication loop, and operating radio science tools, in particular<br />
performing a weekly "check-up" on the RSR of New Horizons.<br />
139
Mars Science Laboratory Wiki<br />
Laura Austin<br />
Mentors: Alicia Allbaugh and David Mohr<br />
The Mars Science Laboratory is currently planned to launch at the end of <strong>2011</strong>. A wiki has been selected as a<br />
repository of sorts to store information about procedures, Operational Interface Agreements (OIAs), etc. and links<br />
to the more secure documents stored elsewhere. This permits the editing of information in one place and every<br />
person being able to see a change, without the resending of documents. My goal is to help implement the wiki to<br />
meet the user’s requirements. I have met with various members of some of the different elements in order to<br />
understand their user requirements. This has led to the creation of a template and format for surface procedures<br />
and OIAs. Most users want to be able to edit information in one place and have it update all across the wiki. While<br />
this is possible, it is only achievable at the individual procedure or OIA level. With the design of the wiki, there has<br />
to be a tradeoff with the users deciding the optimum benefits versus cost analysis, which will influence the future<br />
direction of the wiki. The surface section of the wiki is being developed based upon the lessons learned from the<br />
cruise implementation.<br />
Thermal Conductivity Measurements on Icy Compositions<br />
Jaclyn A. Avidon<br />
Mentors: Martin Barmatz, Mathieu Choukroun, and Fang Zhong<br />
It is widely believed that many of the satellites in the outer solar system, such as Europa, have a liquid ocean<br />
surrounding their core, covered by an icy crust. Infrared spectral data returned by the Galileo and Cassini missions<br />
indicate that these crusts are made of icy compositions of salt hydrates, clathrate hydrates, and ammonia water.<br />
Models of the evolution of these satellites need accurate data on the thermal conductivity of the crustal materials,<br />
since these icy impurities depress the melting point of ice and alter the temperature dependence of the ice’s<br />
thermal conductivity. My project is to support the measurement of the thermal conductivity of these icy<br />
compositions. I will present my involvement in (1) preparing a polycrystalline water ice sample for measurement in<br />
a new cryogenic thermal conductivity probe and (2) calibrating the new probe using a quartz sample of known<br />
thermal conductivity.<br />
Analyzing the Necessary Power to Discover Extra-Terrestrial Habitability<br />
Kimberly Baird<br />
Mentor: Ray Crum<br />
Are we alone? Are there any galaxies, planets (besides Earth), or moons that hold characteristics of life? Many<br />
scientists suspect Jupiter’s moon Europa has high probability to hold the necessities to sustain extra-terrestrial life.<br />
NASA is determined to discover life outside of earth. With the help of Applied Physics Laboratory (ALP) and NASA’s<br />
own Jet Propulsion Laboratory (JPL), the mission to explore the surface and subsurface of Europa is being defined.<br />
With all the instruments to discover the attributes of Europa needing power, the power system is vital to the<br />
exploration of Europa, but what power source is the pivotal one? Analyzing and modeling of the architecture helps<br />
determine the best preference for the system. With the different advanced power systems, each have their pros<br />
and cons and with the mission still in the definition stage, there is no set load. To make a quick and reliable<br />
decision on the perfect power generation, as soon as the mission has been determined, a trade study of the<br />
different power sources is made.<br />
Mars Science Laboratory Flight Software Testing: Wake Cycle Robustness<br />
Payam Banazadeh<br />
Mentor: Danny Lam<br />
Mars Science Laboratory(MSL) is one of the most complex spacecrafts in the history of mankind. Due to the nature<br />
of its complexity, a large number of flight software requirements have been written for implementation . In practice,<br />
these requirements necessitate very complex and very precise flight software with no room for error. One of flight<br />
software’s responsibility is to be able to boot up and check the state of all devices on the spacecraft after the wake<br />
up process. This boot up and initialization is crucial to the mission since any misbehavior of different devices needs<br />
to be handled through the flight software. This summer, I am working with two other interns to exhaustively test<br />
the flight software under variety of different unexpected scenarios and validate that flight software can handle any<br />
situation after booting up. The test includes initializing different devices on spacecraft to different configurations<br />
and confirm at the end of the flight software boot up that the flight software has initialized those devices to what<br />
they are suppose to be in that particular scenario.<br />
140
Improving Single Event Effects Testing Through Software<br />
Mackenzie Banker<br />
Mentors: Steve McClure, Larry Edmonds, Philippe Adell, and Greg Allen<br />
Radiation encountered in space environments can be damaging to microelectronics and potentially cause spacecraft<br />
failure. Single event effects (SEE) are a type of radiation effect that occur when an ion strikes a device. Single<br />
event gate rupture (SEGR) is a type of SEE that can cause failure in power transistors. Unlike other SEE rates in<br />
which a constant linear energy transfer (LET) can be used, SEGR rates sometimes require a non-uniform LET to be<br />
used to be accurate. A recent analysis shows that SEGR rates are most easily calculated when the environment is<br />
described as a stopping rate per unit volume for each ion species. Stopping rates in silicon for pertinent ions were<br />
calculated using the Stopping and Range of Ions in Matter (SRIM) software and CRÈME-MC software. A reference<br />
table was generated and can be used by others to calculate SEGR rates for a candidate device. Additionally, lasers<br />
can be used to simulate SEEs, providing more control and information at lower cost than heavy ion testing. The<br />
electron/hole pair generation rate from a laser pulse in a semiconductor can be related to the LET of an ion.<br />
MATLAB was used to generate a plot to easily make this comparison.<br />
NASA’s Advanced Multimission Operations System: A Case Study in Formalizing Software Architecture<br />
Evolution<br />
Jeffrey M. Barnes<br />
Mentors: Brian Giovannoni and Oleg Sindiy<br />
All software systems of significant size and longevity eventually undergo changes to their basic architectural<br />
structure. Such changes may be prompted by evolving requirements, changing technology, or other reasons.<br />
Whatever the cause, software architecture evolution is commonplace in real-world software projects. Recently,<br />
software architecture researchers have begun to study this phenomenon in depth. However, this work has suffered<br />
from problems of validation; research in this area has tended to make heavy use of toy examples and hypothetical<br />
scenarios and has not been well supported by real-world examples. To help address this problem, I describe an<br />
ongoing effort at the Jet Propulsion Laboratory to re-architect the Advanced Multimission Operations System<br />
(AMMOS), which is used to operate NASA’s deep-space and astrophysics missions. Based on examination of project<br />
documents and interviews with project personnel, I describe the goals and approach of this evolution effort and<br />
then present models that capture some of the key architectural changes. Finally, I demonstrate how approaches<br />
and formal methods from my previous research in architecture evolution may be applied to this evolution, while<br />
using languages and tools already in place at the Jet Propulsion Laboratory.<br />
Modeling the Structure of Venusian Rift Systems Through Gravity Data Analysis<br />
Annabelle Batista<br />
Mentor: Suzanne Smrekar<br />
The Magellan Spacecraft (1990-1994) has provided scientists with high-resolution maps of Venus' topography and<br />
gravitational field. Magellan radar imagery of the planet reveals a surface marked by chains of rift systems<br />
(chasmata) stretching thousands of kilometers. The formation of these chasmata is puzzling, as there is no obvious<br />
geologic process that would cause extensional stresses. Venus lacks plate tectonics, the mechanism driving<br />
complimentary extensional and compressional processes on Earth. To better understand chasmata formation, we<br />
examine the lithospheric structure at four rift systems: Dali/Diana Chasmata System, Ganis Chasmata System,<br />
Perunitsa and Khosedem Fossae, and Devana Chasma. The four rifts vary in location, size, and geologic setting. We<br />
create maps of the chasmatas' crustal and elastic lithosphere thickness by modeling the admittance function for<br />
gravity and topography. This method matches observed admittance variations across each chasma system to<br />
elastic compensation top- and bottom-loading models of predicted admittance. This gravity analysis may result in<br />
better constraints on the internal structure of chasmata, and may provide insight into the tectonic processes<br />
responsible for chasmata formation.<br />
Photochemical Studies of N2/CH4 Ice and Formation of Radicals<br />
Sky Beard<br />
Mentors: Paul Johnson and Robert Hodyss<br />
Photochemical studies of cryogenic ice films analogous to the icy surfaces of Triton and Pluto were performed to<br />
simulate solar photochemistry. N2/ CH4 ices were exposed to a hydrogen discharge lamp to provide an intense<br />
source of ultraviolet light in an ultra high vacuum chamber, and the chemical evolution of the reactants over time<br />
was monitored with infrared spectroscopy. One product of interest, ethane, was formed from photochemically<br />
generated methyl radicals: · CH3 + · CH3 → C2H6. Experiments were repeated at different temperatures from<br />
14- 30 K, and the reaction rate as a function of temperature and activation energy of diffusion for methyl radicals<br />
in N2 ice was determined. Spectroscopic studies of the formyl radical, HCO, in N2 ice in the visible and IR were<br />
performed and the rate of destruction of the formyl radical under visible light was determined.<br />
141
Correlations Between Technical Performance and Cost Growth: A Case Study Using Technical<br />
Complexity<br />
Keith Becker<br />
Mentors: Kevin Rice, Kevin Endo, Eric Kwan, and John Jack<br />
Flight Projects historically exceed the available budgetary funds allocated by NASA HQ. This group research project<br />
investigates historical data from the 13 most recent JPL Flight Projects to determine trends and correlations<br />
between technical and programmatic parameters to predict cost growth. Each of the 13 projects is analyzed using<br />
5 major data sets: technical complexity, subcontractor performance, risk trends, margin analysis and schedule slip.<br />
My focus is the technical complexity data set. Activities in this analysis include: pointing accuracy, instrument count,<br />
redundancy level and mission life, along with 52 additional parameters. Once all technical and cost data is collected<br />
a numerical complexity matrix is created based on technical performance, followed by regression analysis and<br />
modeling to establish trends. It is expected that certain aspects of technical performance parameters will lend<br />
themselves to be more accurate predictors of cost growth than others. Conclusions derived from this research can<br />
be utilized to form a historical data base for future projects and compared against the current in-work flight<br />
projects to improve cost growth predictions.<br />
Software Analysis of an Automatic Sequence Processor<br />
Brandon Benjamin<br />
Mentor: Barbara Streiffert<br />
Jet Propulsion Laboratory’s planning and sequencing element provides space missions with multi-purpose software<br />
to plan spacecraft activities, sequence spacecraft commands, and then integrate these products and execute them<br />
on spacecraft. Currently, Jet Propulsion Laboratory is flying many missions and the processes for building,<br />
integrating, and testing the multi-mission planning and sequencing software is being revitalized to meet the needs<br />
of current and future missions, as well as the operations teams that command the spacecraft. A special team is<br />
responsible for collecting and processing the observations, experiments and engineering activities that are to be<br />
performed on a selected spacecraft. The collection of these activities, ultimately, becomes a sequence of spacecraft<br />
commands. There are commands that are guaranteed not to harm the spacecraft. As such, these commands may<br />
be processed automatically and without intervention by an operations team member. This task is identifying,<br />
understanding, mapping, and documenting the nature of the scripts responsible for these automated commands in<br />
order to facilitate the future missions operating on the revitalized software.<br />
Venusian Applications of 3D Convection Modeling<br />
Timary (Annie) Bonaccorso<br />
Mentor: Christophe Sotin<br />
This study models mantle convection on Venus using the 'cubed sphere' code OEDIPUS, which models one-sixth of<br />
the planet in spherical geometry. We are attempting to balance internal heating, bottom mantle viscosity, and<br />
temperature difference across Venus' mantle, in order to create a realistic model that matches with current<br />
planetary observations. This summer's central question is how to define a mantle plume. Traditionally, we have<br />
defined a hot plume as any contour with at least 40% of the maximum temperature, and a cold plume as any<br />
contour with 40% of the minimum. For less viscous cases (10 20 Pa-s), the plumes generated by that definition<br />
lacked vigor, displaying buoyancies 1/100 th of those found in previous simulations. As the mantle plumes with large<br />
flux are most likely to produce topographic uplift and volcanism, most of the low viscosity cases' plumes exhibit no<br />
surface expression. In an effort to eliminate the smallest plumes, we experimented with different heat flux<br />
parameters and temperature percentages. We plan to run both lower and upper mantle simulations to determine<br />
whether layered (as opposed to whole-mantle) convection might be a more appropriate model. Upper mantle<br />
simulations will be completed using OEDIPUS' Cartesian counterpart, JOCASTA.<br />
Model-Based Systems Engineering Approach for Capturing Mission Architecture Systems Processes<br />
With Application Case Study Problem: Orion Flight Test 1<br />
Kevin H. Bonanne<br />
Mentors: Oleg V. Sindiy and Brian J. Giovannoni<br />
Model-based Systems Engineering (MBSE) is an emerging methodology that can be leveraged to enhance many<br />
system development processes. MBSE allows for the centralization of an architecture description that would<br />
otherwise be stored in various locations and formats, thus simplifying communication among the project<br />
stakeholders, inducing commonality in representation, and expediting report generation. This paper outlines the<br />
MBSE approach taken to capture the processes of two different, but related, architectures by employing the<br />
Systems Modeling Language (SysML) as a standard for architecture description and the modeling tool MagicDraw.<br />
The overarching goal of this study was to demonstrate the effectiveness of MBSE as a means of capturing and<br />
designing a mission systems architecture. The first portion of the project focused on capturing the necessary<br />
system engineering activities that occur when designing, developing, and deploying a mission systems architecture<br />
142
for a space mission. The second part applies activities from the first to an application problem—the system<br />
engineering of the Orion Flight Test 1 (OFT-1) End-to-End Information System (EEIS). By modeling the activities<br />
required to create a space mission architecture and then implementing those activities in an application problem,<br />
the utility of MBSE as an approach to systems engineering can be realized.<br />
Magnetic Coolant Characterization<br />
Shannon Colleen Bourke<br />
Mentors: Warren Holmes and Talso Chui<br />
Infrared spectroscopy is a powerful tool that enables astrophysicists to determine the distance to other galaxies in<br />
the universe, as well as their components. Space borne infrared spectrometers for astrophysics require detectors<br />
that are cooled to temperatures lower than 1K to decrease the noise and increase sensitivity. A common way to<br />
achieve sub-Kelvin temperatures is an adiabatic demagnetization refrigerator (ADR). Future missions that will use<br />
an ADR include the Background Limited Infrared and Submillimeter Spectrometer (BLISS) that will fly on the<br />
Japanese space telescope SPICA. ADRs use a magnetic coolant, or salt pill, suspended in the bore of a<br />
superconducting magnet. The salt pill that we are studying is chromic cesium alum (CCA). We characterize the CCA<br />
by measuring its heat capacity as a function of magnetic field and temperature using the heat pulse technique. We<br />
also measure the thermal resistance between the point where the instrument is attached and the CCA. Our two<br />
CCA salt pills routinely achieve base temperatures of 20-30mK, and can therefore maintain BLISS at 50mK as<br />
required.<br />
California Precipitation Trends: Natural Variability Overwhelms Climate Change Trends<br />
Ann Bui<br />
Mentors: William Patzert and Joshua Willis<br />
Precipitation trends in California from approximately 1900 to present were identified in this study. In our analyses<br />
of 14 long-term precipitation records representing multiple climate types throughout the state, we have found<br />
distinct regional and seasonal trends. In the North, annual rainfall has increased over the time series, while<br />
Southern California stations show little or no trend at all. Seasonally, winter precipitation shows increases and<br />
higher frequencies while spring, fall and summer display little gain in the long term. However, these trends did not<br />
conform to accepted measures of significance because of high degree in variability. By quantifying the influence of<br />
the Pacific Decadal Oscillation (PDO), and El Nino- Southern Oscillation (ENSO) on the annual variability in<br />
precipitation, we also found that the coefficient of variability is increasing with time, making it more difficult to<br />
distinguish a trend from the data. In general we have found that precipitation trends in California are not as<br />
extreme as previously hypothesized by models at a global level, and because of high inter-annual variability, more<br />
data are needed for the trends to be significant.<br />
Innovations in Robotic Code<br />
James Burdick<br />
Mentors: Michael McHenry and Norm Aisen<br />
Creating comprehensive controlling computer code for robotic hardware is a challenge. Most robotics software is<br />
handwritten in lower level languages, like C or Fortran, which functions very well in platforms due to the simpler<br />
nature of these languages and the need for computational efficiency. However, the creation of this code is<br />
expensive in terms of time and labor; the situation suggests a more innovative approach.<br />
Higher level programming languages provide the possibility of creating the equivalent of reams of code, while<br />
simplifying the software creation process. Matlab-based Simulink and associated Stateflow products are well-suited<br />
for robotic controls. Simulink provides a continuous computing framework, while Stateflow is designed to aid the<br />
development of state-machine based software, both of which are valuable in any robotics control system. Stateflow<br />
also can create embeddable c-code. While handwritten c-code functions in a linear manner, in robotics disparate<br />
events can be introduced into the system at unpredictable times. The research conducted in this internship was<br />
directed towards understanding and demonstrating the use of MATLAB-based tools to generate such embeddable<br />
c-code, and to interface this code with external handwritten code. It is clear that code generated from these<br />
MATLAB-based programs can help to simplify the code-making process in robotics.<br />
Data Analysis and Archiving for Mars Exploration Rovers (MER)<br />
William J. Canan<br />
Mentors: David Mohr and Matthew Keuneke<br />
The Mars Exploration Rovers (MER) tactical team relies heavily on modeling software to determine the most<br />
effective way to explore the Martian environment. MER attempts to make the most of every minute of operation<br />
time and of every data byte they can downlink to Earth. The models used to make predictions of imaging<br />
acquisition duration and the amount of data volume available to downlink on a UHF pass are known to have<br />
inaccuracies. By going through the daily reports of the rover operations en masse using Python scripts, the<br />
143
inaccuracy of these models has been quantified. This information allows the team to make more informed decisions<br />
regarding the use of modeling results or to introduce refinements to the modeling software based on the collected<br />
data. The presentation focuses on the methodology used to determine the accuracy of the models.<br />
Computing Temperatures in Optically Thick Protoplanetary Disks<br />
Lawrence Capuder<br />
Mentor: Neal Turner<br />
I am working with a Monte Carlo radiative transfer code to simulate the transfer of energy through protoplanetary<br />
disks, where planet formation occurs. The code tracks photons from the star into the disk, through scattering,<br />
absorption and re-emission, until they escape to infinity. High optical depths in the disk interior dominate the<br />
computation time because it takes the photon packet many interactions to get out of the region. High optical<br />
depths also receive few photons and therefore do not have well-estimated temperatures. I will apply two methods<br />
for treating high optical depths, a modified random walk (MRW) and a partial diffusion approximation (PDA), to<br />
speed up the Monte Carlo calculations. The MRW is implemented by calculating the average number of interactions<br />
the photon packet will undergo in diffusing within a single cell of the spatial grid and then updating the packet<br />
position, packet frequencies, and local radiation absorption rate appropriately. The PDA is implemented by applying<br />
the time independent diffusion equation to a large region after each iteration of the code. The approximations will<br />
be tested for accuracy and speed compared to the original code.<br />
Testing of Fusing Sonar and Video Images in Automated Target Detection and Tracking<br />
Jeffrey Chiang<br />
Mentor: Thomas Lu<br />
Assisted Target Recognition (ATR) systems aim to automate certain aspects of surveillance tasks, with the end goal<br />
of fully automating target detection, recognition, and tracking. The current project applies this system to low<br />
resolution SONAR and camera videos taken from Unmanned Underwater Vehicles (UUVs). These SONAR images<br />
are inherently noisy and difficult to interpret, and pictures taken underwater are unreliable due to murkiness and<br />
inconsistent lighting. The ATR system breaks target recognition into two stages: 1) Videos of both SONAR and<br />
camera footage are broken into frames and preprocessed in the Fourier domain to enhance images. 2) Regions of<br />
Interest (ROIs) are determined by evaluating the processed images, and characteristic features are extracted.<br />
3) ROIs are then classified as true or false positives using a standard Neural Network based on the extracted<br />
features. Several preprocessing methods are explored, as well as integrating data from both images in order to<br />
maximize the efficiency of the program.<br />
Characteristics of the Cathode and Plume of a High Power Hall Thruster<br />
Emily Chu<br />
Mentor: Dan Goebel<br />
The plasma dynamics in the plume region of high current hollow cathodes and Hall thrusters for high power use are<br />
complex and the success of modeling efforts designed to identify the fundamental mechanisms of erosion and life<br />
of these devices are dependent on the support of experiments that provide plasma measurements in the nearthruster<br />
plume region for model validation. In this study, we use Langmuir and emissive probes to investigate the<br />
plasma parameters inside and in the near-plume of a 200 A hollow cathode and in the near-field plume of a<br />
6-to-10 kW Hall thruster that has been modified to reduce erosion rates and increase the power level to 10 kW.<br />
The cathode and thruster are run at operating conditions associated with their respective throttle tables and we<br />
measure the resulting plasma properties to determine plasma density, potential, and electron temperature profiles<br />
in front of the thruster. This data will be used to provide benchmarking data to thruster modelers for code<br />
predictions and life model validation.<br />
Analysis of the Instrument Management Layer of the MSL Flight Software<br />
Louis M. Cialdella<br />
Mentor: Marcel Schoppers<br />
The control of science instruments on the upcoming Mars Science Laboratory mission is implemented via the<br />
"Instrument Management Layer" (IML), an abstraction layer that handles transmission and reception of commands<br />
to and from instruments. After several years of ongoing development and numerous structural changes, the IML is<br />
functional but suboptimal in many ways – it is convoluted, poorly documented, and contains redundant states and<br />
code. This combination of things makes the code both error-prone and difficult to debug. By analyzing the current<br />
version of the abstraction layer, we will be able to get a concrete idea of how precisely it works. In particular, we<br />
will be examining the finite state machine that underlies the current implementation, allowing us an intuitive and<br />
powerful view of the way in which the software functions. After this has been done, methods for restructuring the<br />
state machine (ie, reductions by eliminating unnecessary states, merging similar states) will be considered in order<br />
to see how a restructuring could improve the clarity and integrity of the software.<br />
144
An Analysis of Database Replication Technologies With Regard to Deep Space Network Application<br />
Requirements<br />
Andrea M. Connell<br />
Mentors: Paul Wolgast and Silvino Zendejas<br />
The Deep Space Network (DSN) has three communication facilities which handle telemetry, commands, and other<br />
data relating to spacecraft missions. The network requires these three sites to share data with each other and with<br />
the Jet Propulsion Laboratory for processing and distribution. Many database management systems have<br />
replication capabilities built in, which means that data updates made at one location will be automatically<br />
propagated to other locations. This project examines multiple replication solutions, looking for stability, automation,<br />
flexibility, performance, and cost. After comparing these features, Oracle Streams is chosen for closer analysis.<br />
Two Streams environments are configured – one with a Master/Slave architecture, in which a single server is the<br />
source for all data updates, and the second with a Multi-Master architecture, in which updates originating from any<br />
of the servers will be propagated to all of the others. These environments are tested for data type support, conflict<br />
resolution, performance, changes to the data structure, and behavior during and after network or server outages.<br />
Through this experimentation, it is determined which requirements of the DSN can be met by Oracle Streams and<br />
which cannot.<br />
Solid Modeling and Structural Analysis of LNA (Low Noise Amplifier) Feed Platform for Type-2 BWG<br />
(Beam Wave Guide) Antennas<br />
Andrew Crawford<br />
Mentor: Jason Carlton<br />
The primary objective of this internship research project involves working in a collaborative environment on the<br />
detailed solid-modeling design and structural analysis of the LNA (low noise amplifier) feed platform to be built and<br />
installed on the new 34-Meter BWG (Beam Wave Guide) Type-2 DSN (Deep Space Network) tracking antenna at<br />
the CDSCC tracking facility in Canberra, Australia, as well as all future similar BWG-Type 2 antenna builds.<br />
The scope and methodology of the project involves obtaining the necessary mandated requirements from each<br />
interfacing stakeholder involved with the platform, using SolidWorks 3D modeling software of said requirements to<br />
design, model, and analyze the construction and assembly of the platform in the new 34 meter BWG-Type 2<br />
Antenna.<br />
By using pass/fail criteria software, analysis will be applied using techniques involving live-load, seismic, frequency<br />
vibration, and deflection tests, ensuring proper and acceptable design plans and models can be created for<br />
eventual peer review and released drawings. Material and structural analysis of the platform designs and models<br />
using industry standards and codes will be incorporated to meet national safety requirements. The deliverables<br />
include a complete solid model of the platform, detailed analysis results, and presentation materials ready for the<br />
peer review.<br />
Modeling Intermittent Contact With Complementarity<br />
Cory Crean<br />
Mentors: Abhinandan Jain and Marco Quadrelli<br />
One of the most important tasks in simulating a multibody system is to properly handle the constraints. There are<br />
well-developed methods for handling bilateral constraints in tree-topology systems, but handling such constraints<br />
in systems with closed loops still presents some difficulty. Ndarts provides two methods for handling loop<br />
constraints: the augmented approach, and constraint embedding. This paper validates these implementations by<br />
simulating classic mechanisms, and comparing the results produced by Ndarts to the results available from the<br />
literature. Also, it compares the two methods on the basis of efficiency and accuracy.<br />
Unilateral constraints are also present in many multibody system. For example, a legged robot's foot interacts with<br />
the ground via a unilateral contact. Ndarts makes use of the complementarity formulation to simulate such systems.<br />
This paper describes the algorithm used for simulating objects with one or more unilateral contacts, and describes<br />
some test problems to which the algorithm has been applied.<br />
An IR Analysis of Cryovolcanism at Sotra Facula on Titan<br />
Paul A. Dalba<br />
Mentor: Bonnie J. Buratti<br />
Investigations of cryovolcanic activity on Titan are numerous and could uncover important clues regarding<br />
processes taking place on and within Titan. Using the Synthetic Aperture RADAR (SAR) imager and the Visual and<br />
Infrared Mapping Spectrometer (VIMS) aboard the Cassini Spacecraft, many potential cryovolcanoes have been<br />
mapped. One of the leading candidates is an area known as Sotra Facula located at 39.8ºW 12.5ºS. In this region,<br />
SAR data identified a high peak located near a low basin which may have produced mass flows in the past. An<br />
investigation of Sotra Facula using VIMS data proved that the region is brighter than its immediate surroundings at<br />
145
2.01µm, the band which provides the clearest view of Titan's surface. This varying brightness was not detected at<br />
2.13µm and is, therefore, not due to methane clouds in Titan's atmosphere. Sotra Facula was detected in an image<br />
produced from coadded spectra between 4.9µm and 5.1µm, but the heat signature was not anomalous compared<br />
to the surrounding terrain. Furthermore, in a ratio of the 4.9µm to 5.1µm coadded spectra to the 2.01µm band,<br />
Sotra Facula was nearly undetectable. This spectroscopic investigation of Sotra Facula failed to yield any signature<br />
of a cryovolcano.<br />
Investigating the Wake of Capsule-Like Bodies Using Computational Fluid Dynamics and Fluid Structure<br />
Interaction<br />
Derek J. Dinzl<br />
Mentor: Anita Sengupta<br />
The performance of aerodynamic decelerators in the wake of bluff bodies is of particular interest for optimizing<br />
parachute design during and after planetary entry. This project focuses on computational simulations of the Multi-<br />
Purpose Crew Vehicle at 10% subscale test conditions. This will include the capsule alone, the capsule mounted on<br />
a sting, and the capsule with a disk or parachute in its wake. The code used to implement these cases is US3D with<br />
a Spalart-Allmaras Detached Eddy Simulation turbulence model. Simulation data of the capsule by itself will be<br />
compared to previously recorded data sets. Wind tunnel experiments of various capsule-parachute configurations<br />
are being carried out with a maximum dynamic pressure of 100 psf and a Reynolds number near 3×10 6 . Data from<br />
the simulations will be compared to particle image velocimetry calculations and results of these wind tunnel tests.<br />
Focus will be on lift and drag forces, velocity/mach and pressure contours, vortical wake structures, and stagnation<br />
point locations. Validation of simulation data with experiment is of prime importance; once the fidelity of<br />
computational methods is established one can emulate realistic flight conditions and observe otherwise complex<br />
phenomena to capture. To this end, a fluid-structure interaction model for the capsule’s drogue parachutes is<br />
introduced and developed. This will allow us to analyze the motion of the parachute along with its drag<br />
performance in the wake of the capsule. Benefits and drawbacks of both computational fluid dynamics and wind<br />
tunnel testing as they pertain to the subject will be explored and discussed.<br />
Deep Space Network-Wide Portal Development<br />
Silviya Doneva<br />
Mentor: Paul Wolgast<br />
The Deep Space Network (DSN) is an international network of antennas that supports interplanetary spacecraft<br />
missions and radio and radar astronomy observations for the exploration of the solar system and the universe.<br />
DSN provides the vital two-way communications link that guides and controls planetary explorers, and brings back<br />
the images and new scientific information they collect. In an attempt to streamline operations and improve overall<br />
services provided by the Deep Space Network a DSN-wide portal is under development. The project is one step in a<br />
larger effort to centralize the data collected from current missions including user input parameters for spacecraft to<br />
be tracked. This information will be placed into a principal repository where all operations related to the DSN are<br />
stored. Furthermore, providing statistical characterization of data volumes will help identify technically feasible<br />
tracking opportunities and more precise mission planning by providing upfront scheduling proposals. Business<br />
intelligence tools are to be incorporated in the output to deliver data visualization.<br />
Percussive Augmenter of Rotary Drills<br />
Chris Donnelly<br />
Mentor: Yoseph Bar-Cohen<br />
Hammering drills are effective in fracturing the drilled medium while rotary drills remove cuttings. The combination<br />
provides a highly effective penetration mechanism. Piezoelectric actuators were integrated into an adapter to<br />
produce ultrasonic percussion; augmenting rotary drilling. The drill is capable of operating at low power, low<br />
applied force and, with proper tuning, low noise. These characteristics are of great interest for future NASA<br />
missions and the construction/remodeling industry. The developed augmenter connects a commercially available<br />
drill and bit. Input power to the drill is read using a multi-meter and the augmenter receives a separate input<br />
voltage. The drive frequency of the piezoelectric actuator is controlled by a hill climb algorithm that optimizes and<br />
records average power usage and operates the drill at resonating frequency. Testing the rotary drill and augmenter<br />
across a range of combinations with total power constant at 160 Watts has shown improved results in concrete and<br />
limestone samples under low axial loading (4.2 kg). The drill rate was increased 1.5 to over 10 times compared to<br />
rotation alone. For a given power there is an optimal balance between the rate of producing cuttings and their<br />
removal.<br />
146
Processing of Mars Atmospheric Radio Occultation Data<br />
Ursula Doswell-Fotheringham<br />
Mentor: Sami Asmar<br />
JPL’s Radio Science Systems Group has been collecting and processing radio occultation data from the Mars<br />
Reconnaissance Orbiter for several years, as previously done with the Mars Global Surveyor mission. The scientific<br />
output of temperature-pressure profiles is of high interest to the planetary atmospheric science community. The<br />
raw data gets publically archived promptly while detailed processing is then carried out at JPL. With years of data<br />
in storage, a system to process and organize it had to be developed. Using UNIX and Matlab, the data from 1999 is<br />
downloaded from the Planetary Data System Atmospheres Node and a series of codes is run to produce a<br />
temperature and profile graph for approximately every four hour time period. Similarly, the data from 2008 thru<br />
<strong>2011</strong> has been processed and is being converted from a Matlab-specific format to one that can be viewed and<br />
analyzed in other programs. There are 30 graphs that have to go thru this per day for the four-year span that the<br />
Mars Reconnaissance Orbiter has been producing data. Once processed, this data can be made available for<br />
analysis of the ionosphere and atmosphere of Mars.<br />
MISR INteractive eXplorer (MINX): Production Digitizing to Retrieve Smoke Plume Heights and<br />
Validating Heights Against Lidar Data<br />
Ben Dunst<br />
Mentor: David Nelson<br />
Aerosols such as smoke from wildfires, ash from volcanoes, and desert dust can contribute to climate change. They<br />
typically rise into the atmosphere as plumes from identifiable sources and can be analyzed to determine their<br />
height and speed. Determining accurate heights for aerosols is important, because if they are injected above the<br />
planetary boundary layer, they can be transported great distances. MINX, an interactive software tool that uses<br />
stereoscopic imagery from the Multi-angle Imaging SpectroRadiometer (MISR) instrument, can retrieve heights and<br />
winds from aerosol plumes. More than 10,000 smoke plumes from around the world have been digitized to date<br />
with MINX and are compiled into a publicly available online database. Several hundred plumes will be added this<br />
summer from fires in North and South America. In addition, the height retrieval algorithms in MINX have not been<br />
formally validated and algorithms were recently added to allow for retrievals in different color bands and using<br />
different size pattern matchers. The accuracy of MINX height retrievals for specific aerosol events will be evaluated<br />
by finding time- and space-coincident Lidar retrievals and using those as ground truth to compare with MINX.<br />
Analysis of Subsurface Clathrates in the Upper Crust of Titan<br />
John Elliott<br />
Mentors: William Smythe and Mathieu Choukroun<br />
Titan has an atmosphere rich in methane, which should have long since been depleted unless a mechanism exists<br />
for storing these compounds below the surface. One hypothesis is that methane could be stored in the form of<br />
clathrate hydrates, which are compounds with an ice lattice forming molecular cages in which gases are trapped.<br />
They are stable at low temperatures and over a wide range of pressures, suggesting that clathrate hydrates may<br />
have stored methane on Titan from the beginning of its history. To find the conditions at which the clathrates<br />
dissociate, we will conduct various experiments using a cryogenic calorimeter. The calorimeter needed to be<br />
modified to allow us to create a high pressure system (100 bars) similar to the pressure on the subsurface of Titan.<br />
We have finalized the modifications for the calorimeter and will calibrate the device using the vapor-liquid<br />
equilibrium of CO2. Preliminary tests on CO2 clathrates will be completed before the end of the internship.<br />
Planetary Ices Spectroscopy<br />
Jeffrey T. Foster<br />
Mentors: Robert Carlson and Kevin Hand<br />
The goal of this research project is to develop a robust and highly sensitive spectrometer that can be used to scan<br />
water ice and liquid samples on Europa for organic compounds. The aim is to find signs of life on the ice covered<br />
moon orbiting Jupiter because it has been shown to have liquid water underneath its icy crust. To work towards<br />
this goal I am helping set up and perform spectral analyses on several different sample types relevant to Europa.<br />
The tasks I perform are varied but the focus is infrared spectroscopy of ice grains containing organic molecules.<br />
Progress is being made on obtaining spectra similar to those that would be obtained from Europa.<br />
Local Time Asymmetry in the Jovian Equatorial Current Sheet<br />
Alexander Freed<br />
Mentors: Henry Garrett and Michael Kokorowski<br />
Charged particle observations from multiple spacecraft have shown an asymmetry in the magnetic equatorial<br />
current sheet between the dawn and dusk local time regions. There is a thin current sheet at dawn and a thick<br />
sheet at dusk. Galileo spacecraft measurements show that thin current sheet crossings are regular and coincident<br />
with observed local maxima in high-energy (~ MeV) electron count rates. However, they also show that where the<br />
147
current sheet is thick, like in the dusk region, the crossings and electron count rate maxima are neither periodic or<br />
synchronized. This study investigates the local time asymmetry on a more global scale by using the energetic<br />
electron data from Galileo, the only spacecraft to orbit Jupiter. Count rate maxima in the middle and outer<br />
magnetosphere have been averaged over sections of local time to investigate the area in between the dawn and<br />
dusk regions. These results are then discussed in context of prevailing theories about the nature of this asymmetry.<br />
High-Speed Generation of Illumination Spectra for a Stereoscopic Endoscope<br />
Eric Fritz<br />
Mentors: Harish Manohara and Michael Shearn<br />
Traditional stereoscopic vision (3D) is achieved through use of two separate cameras, arranged to emulate human<br />
eyes. This method works on large scale projects, but runs into problems on small scale designs, such as surgical<br />
endoscopes. This project is focused on developing a stereoscopic endoscope, using a single camera and Conjugated<br />
Multiple-Bandpass Filters (CMBF) to create stereoscopic vision. Each half of filter is built to allow a distinct<br />
spectrum through, while blocking the complementary spectrum. A system with different left and right filters can<br />
produce stereoscopic images.<br />
To accomplish this, the light must be filtered at the light source to match the filters at the camera. Additionally, the<br />
light source and camera must be synchronized in a way that each image will show only one filter spectrum. In this<br />
talk, I will describe the design and characterization of the prototype electro-optical system, including optical<br />
throughput measurements and video produced using this method.<br />
Data Collection for Physical Modeling of a Small Dynamic Legged Robot Platform<br />
Xiao-Yu Fu<br />
Mentor: Rudranarayan Mukherjee<br />
OctoRoACH is a 25 gram, 20-cm long, eight-legged mobile robot platform. It is intended to serve as a base<br />
platform for testing the development of new technologies for control, communication, and sensing in small robots.<br />
Because OctoRoACH is a new platform undergoing active mechanical development, its physical behavior is not well<br />
understood. We collected physical performance data describing structural rigidity, dynamic behavior, power<br />
utilization, and control response using a combination of onboard sensors and external instruments in ideal and<br />
realistic conditions. This data is being compiled into dynamic model simulation that approximates robot behavior,<br />
with the goal of providing greater understanding about robot performance and providing directional insights for<br />
design improvements.<br />
Nontronite Dissolution Kinetics and Secondary Mineral Formation<br />
Seth R. Gainey<br />
Mentor: Joel A. Hurowitz<br />
Spectral observations from orbiter space craft have detected clay minerals in ancient rock deposits exposed at the<br />
Martian surface, formed form from the hydrolysis of silicate minerals, indicating the past presence of liquid water.<br />
It is the purpose of this research to alter clay minerals analogous to those observed on Mars, in order to determine<br />
their rate of dissolution under pedogenic conditions analogous to those common on the Earth’s surface, and<br />
possibly common on the ancient surface of Mars. These experiments determined the dissolution rate of nontronite<br />
(standard NAu-1) and the secondary minerals produced as a byproduct of this alteration. Mineralogy was confirmed<br />
using a Horiba Jobin Yvon LabRam HR confocal Raman microscope. Flow through reactors were modeled after<br />
Taylor et al. (2000), each flow through reactor consist consists of Teflon tubing where solution is pumped through<br />
the sample by syringe pump. Effluent was subsequently collected for pH measurement and ion concentration<br />
analysis via ICP-OES. Rate laws were determined from the dissolution rates as a function of pH and distance from<br />
equilibrium. Altered and unaltered mineral powders were analyzed using Raman spectroscopy, electron microprobe<br />
analysis, SEM combined with EDS to determine mineral phase identification, chemical composition, and morphology,<br />
respectively.<br />
Finding the Sun’s Closest Companion: Using WISE Observations and Follow-Up Data to Find the Closest<br />
and Coldest Brown Dwarfs<br />
Thomas Gautier<br />
Mentors: Amy Mainzer and Chris Gelino<br />
The Wide-Field Infrared Survey Explorer (WISE) Brown Dwarf Team has compiled a list of approximately 1600<br />
objects that it classifies as brown dwarf candidates. Many of these objects had partially or unprocessed follow-up<br />
data. This project compiled follow-up data from various telescopes and primarily used IDL routines to complete the<br />
processing and extract the precise astrometry and photometry from each observation. The result of the project is<br />
an easily accessible database with the astrometric and photometric data for each of the candidates, which greatly<br />
facilitates the analysis of by the Brown Dwarf Team to find objects with brown dwarf colors and that are quickly<br />
moving, indicating they are close to the Sun.<br />
148
Modeling Hydrothermal Vents on Europa<br />
Patricia Gavin<br />
Mentor: Steve Vance<br />
Hydrothermal vents occurring at sea-floor-spreading centers may have hosted the first transition from abiotic to<br />
biotic chemical processes, eventually resulting in the emergence of life. The possibility of these vents also being<br />
present on Europa’s subsurface ocean floor puts forth intriguing astrobiological implications relevant to both life on<br />
another planetary body as well as the origins of life on Earth. The studies presented here simulate conditions at<br />
potential hydrothermal vents on Europa to determine which minerals favorable to life form under these conditions.<br />
Using the modeling software Geochemist’s Workbench (GWB), several simple models were composed, each<br />
simulating a different geochemical scenario. Among the simulations derived were (1) water-rock interactions,<br />
(2) terrestrial hydrothermal fluids and (3) reduction-oxidation (redox) reactions. The first simulation was used to<br />
study how fluids inside a hydrothermal vent react with the rock that composes the vent. The second model<br />
simulated the resulting hydrothermal fluid being introduced and reacting with seawater. The third model<br />
investigated the chemical energy potentially available to microbes for use in organic and metabolic processes.<br />
Results from these efforts help in understanding what signatures of hydrothermal activity in Europa’s ocean might<br />
be found by exploration spacecraft observing its geologically active surface.<br />
Mars Topographical Data Analysis<br />
Hallie Gengl<br />
Mentors: Matt Golombek and Fred Calef<br />
The overarching objective of this project is to analyze Mars elevation models at various length scales. On a 1 km<br />
length scale, it involves working with digital elevation models (DEM) at the current MSL sites, as well as at<br />
candidate proposed future robotic missions to Mars like Mars Sample Return (MSR). Due to engineering constraints,<br />
relief cannot exceed 100 meters on the 1 km or smaller length scale in landing site ellipses. High Resolution Stereo<br />
Camera imager (HRSC at 50-300 m/pixel) and Mars Orbiter Laser Altimeter data (MOLA at ~300 m spacing), both<br />
DEM and shot-points, are used. The results of each are compared to evaluate the relief within the parameters of<br />
the landing site constraints. On a small length scale, DEMs are created using stereographic image pairs from the<br />
High Resolution Imaging Science Experiment (HiRISE) camera onboard the Mars Reconnaissance Orbiter (MRO).<br />
These products (25 cm/pixel) can be used for many projects, including high-resolution landing site analysis as well<br />
as small rayed crater morphology. One of the goals is to create a larger dataset of DEMs with small rayed craters<br />
and extract morphometric statistics to understand the change in crater and ejecta shape with changes in diameter<br />
and impact type.<br />
High Precision Propagation and Tracking of the International Space Station<br />
Corrina L. Gibson<br />
Mentor: Rob Witoff<br />
Optical Payload for Lasercomm Science (OPALS) is an experimental payload that will be mounted on the<br />
International Space Station (ISS) in 2012. The goal is to deliver video from the ISS to an optical ground terminal<br />
via an optical communications link. In order to optically downlink data during an ISS passover, high precision<br />
propagation followed by closed-loop tracking of the ISS is required. The ground station’s telescope has a limited<br />
field of view, requiring the ISS position error to be less than 100 meters for closed loop tracking to initiate. Precise<br />
ISS position predictions are used to align the telescope prior to ISS passover, ensuring sufficient view of the ISS<br />
for closed-loop tracking. Once the ISS is within the telescope’s field of view, a beacon is transmitted from the<br />
ground station to a camera on OPALS. A fast steering mirror is used to track OPALS from the ground. The ground<br />
station’s centroiding loop controls the fast steering mirror to remain pointed at OPALS. Techniques that will enable<br />
prediction of the ISS’s position to within 100 meters after 5 minutes of propagation are being developed. Thus far<br />
the error of various propagators has been characterized relative to a truth data set. Additional propagators will be<br />
investigated to identify an acceptable propagation technique.<br />
ISSARS Aerosol Database: An Incorporation of Atmospheric Particles Into a Universal Tool to Simulate<br />
Remote Sensing Instruments<br />
Michael B. Goetz<br />
Mentors: Simone Tanelli and Noppasin Niamsuwan<br />
The Instrument Simulation Suite for Atmospheric Remote Sensing (ISSARS) is entering its third and final year of<br />
development with an overall goal of providing a universal tool to simulate active and passive space borne<br />
atmospheric remote sensing instruments. These simulations focus on the atmosphere ranging from the UV to<br />
microwaves. ISSARS handles all assumption and uses various models on scattering and microphysics to fill the<br />
gaps left unspecified by the atmospheric models to create each instrument’s measurements. This will help benefit<br />
mission design and reduce mission cost, create efficient implementation of multi-instrument/platform Observing<br />
System Simulation Experiments (OSSE), and improve existing models as well as new advance models in<br />
development. In this effort we incorporate various Aerosol particles into the system, and a simulation of input<br />
wavelength and spectral refractive indexes related to each spherical test particle(s) can generate its single<br />
149
scattering albedo (SSA) and phase function (P( �)). Future development with a radiative transfer code will<br />
generate a series of results that can be validated with results obtained by the Multi-angle Imaging<br />
SpectoRadiameter (MISR) instrument.<br />
Verifying Flight Software Commands Generated by CHILL Against MSEQ for the MSL Rover<br />
Ryan Goetz<br />
Mentor: Jim A. Kurien<br />
The flight commands for the MSL rover have been adapted and changed during the years. The old way to generate<br />
these commands is with a program known as CHILL. As the mission draws closer more detail commands need to be<br />
generated for both cruise and land operations. MSEQ was developed to generate detail mission plans for the rover<br />
during operation. In order to switch to the MSEQ program one must verify that the commands generated with both<br />
programs are similar. A specific command is run through the CHILL and MSEQ program to generate a scmf file that<br />
is compared with one another. The results should be that both scmf files are similar thus verifying that the rover<br />
will receive and operate the commands from the newer MSEQ program.<br />
Correlations Between Technical Performance and Cost Growth: A Case Study Using Schedule Slips<br />
Austin Gomez<br />
Mentors: Kevin Rice, Kevin Endo, Eric Kwan, and John Jack<br />
Flight Projects historically exceed the available budgetary funds allocated by NASA HQ. This group research project<br />
investigates historical data from the 13 most recent JPL flight projects to determine trends and correlations<br />
between technical and programmatic parameters to predict cost growth. Each of the 13 projects is analyzed using<br />
5 major data sets: technical complexity, subcontractor performance, risk trends, margin analysis and schedule slips.<br />
Each data set is analyzed to determine which correlations exist. My area of focus for this group project is to analyze<br />
cost and schedule growth between each milestone using schedule slip information. The normalization process will<br />
investigate the changes between the planned and the actual schedule data. Additionally, all data will be divided<br />
into lower-level activities such as instruments and spacecraft for more detailed conclusions. Once all technical and<br />
cost data is collected, regression analysis and modeling is performed to establish trends. It is expected that certain<br />
aspects of technical performance parameters will lend themselves to be more accurate cost predictors than others.<br />
Conclusions derived from this group research can be utilized to form a historical data base for future projects and<br />
compared against the current in-work flight projects to predict cost growth.<br />
Using Wide Infrared Survey Explorer (WISE) and Ground Based Follow-Up Data to Find the Closest and<br />
Coldest Brown Dwarfs<br />
Stephanie Gomillion<br />
Mentors: Amy Mainzer and Michael Cushing<br />
The WISE Brown Dwarf Team has compiled a list of approximately 1600 objects that it classifies as brown dwarf<br />
candidates. Many of these had partially or unprocessed follow-up data. My project was to help gather and compile<br />
the dispersed ground based telescope data into a precise format, primarily using IDL and Emacs. The result of the<br />
project is an easily accessible database with the astrometric and photometric data for each of the candidates,<br />
which significantly facilitates the analysis by the Brown Dwarf Team to find objects with brown dwarf colors and<br />
that are quickly moving, indicating they are close to the Sun.<br />
Radio Science Software Tools for Improved Operations and Analysis<br />
Carlos E. Gonzalez<br />
Mentors: Sami Asmar and Kamal Oudrhiri<br />
The DSN open-loop telemetry demodulation method down-converts and records spacecraft signals in pre-selected<br />
bandwidth using predicted frequency profiles. Radio Science techniques provide more flexibility and robustness in<br />
the acquisition, processing, and interpretation than the closed loop system. We are designing a graphical user<br />
interface to run the demodulation quickly and efficiently as well as selecting particular demodulation cases. Task<br />
#2 involves performing science experiments using Low Earth Orbit Satellites such as the International Space<br />
Station. DSN antennas are strained due to ISS’s fast orbit so higher orbiting satellites form a three-way<br />
communication loop. There are multiple scenarios that can be evaluated using the Satellite Orbit Analysis Program.<br />
We will be able to provide the most desirable results. The primary configuration will have a DSN antenna<br />
communicating with the TDRS satellite, which then relays the signal to the ISS, and ISS will then relay the<br />
information back to the DSN. Task #3 involves the real time operation of the New Horizons Mission and controlling<br />
the DSN Radio Science Receivers remotely from JPL. We will test the real time monitoring and recording of the<br />
downlink signal from New Horizons in order to verify whether or not the spacecraft is working properly.<br />
150
ASTRA Command and Data Handling Flight Software Development<br />
Maximillian Gonzalez<br />
Mentor: Sam Gordon<br />
The Analog Site Testbed for Readiness Advancement (ASTRA) project is aiming to increase the technology<br />
readiness level (TRL) of two instruments, a hot-wire anemometer (HWA) and a rapid acquisition mass spectrometer<br />
(RAMS), so they can be considered for future missions to Mars. ASTRA is a high-altitude balloon project whose<br />
payload will be hoisted up to 120,000ft into a Mars analog test environment of 6 mbar and -35°C. Command and<br />
data handling (C&DH) software has been developed to guide the HWA through a 12-hour nominal mission duration<br />
from pre-launch operations through power-down before landing. The software uses a data sampling interrupt<br />
routine running at 20 kHz to command HWA electronics through a 1 Hz collection cycle and sample digital data<br />
coming in 16-bit serial packets from the HWA at 10 kHz. Once a full, time-stamped sample set, consisting of<br />
60 16-bit data packets, has been gathered, the software will store the sample set on its onboard flash memory as<br />
well as downlink the data to the ground station via the on-board Consolidated Instrument Package (CIP) radio.<br />
Improving Functionality of a Mechanical Control System for a Scanning Microwave Limb Sounder<br />
(SMLS) Through Program Development and Structural Modification<br />
Zachary Greene<br />
Mentor: Paul Stek<br />
The Microwave Limb Sounder (MLS) team is currently developing its newest instrument to probe the Earth's<br />
atmosphere. Known as a scanning microwave limb sounder (SMLS), this device will use microwaves to measure<br />
atmospheric variables including temperature, pressure, and chemical composition. Deployed via high altitude<br />
balloon or aircraft to the upper stratosphere, the SMLS will be the first MLS instrument to scan the horizon laterally<br />
in addition to ordinary vertical profiling performed by previous instruments. Incomplete code existed govern the<br />
scanning mechanism and problems with the scanner arose frequently. Furthermore, documentation left by previous<br />
interns was brief and unclear, forcing much of my initial time to be spent re-learning what they had already done.<br />
My tasks were to modify the existing code to fix nuances and programmatic errors, construct the circuitry and<br />
physical setup as well as troubleshoot and resolve any problems that occurred within, and leave adequate<br />
documentation for future interns and MLS team members who wish to modify or use the SMLS.<br />
Hyperspectral Imaging and the Building of a Near-Infrared Imaging Spectrometer<br />
Kate Grode<br />
Mentors: Bill Mateer and Marc Lane<br />
Hyperspectral imaging, or imaging spectroscopy, is the analysis of the electromagnetic spectra reflected from what<br />
is being studied, such as the ground, a rainforest, or an object, and then recorded as images. Objects give off<br />
distinctive wavelengths, so it possible to distinguish one object from another within the image. Imaging<br />
spectroscopy has many applications, such as in mineralogy, agriculture, environmental research, and surveillance.<br />
The spectrometer currently being built uses hyperspectral imaging and is similar to previous spectrometers. They<br />
are attached to airplanes and flown over a region to gather data. A past spectrometer built for the Carnegie<br />
Airborne Observatory records the wavelengths from the plants in the Amazon Rainforest in order to understand<br />
how the rainforest and the environment as a whole are affected by humans. The current instrument is in the build<br />
phase and is scheduled to be finished in January 2012.<br />
Acceleration of the KINETICS Integrated Dynamical/Chemical Computational Model Using MPI<br />
Jonathan (Max) Grossman<br />
Mentors: Karen Willacy and Mark Allen<br />
Understanding the evolution of a planet's atmosphere not only provides a better theoretical understanding of<br />
planetary physics and the formation of planets, but also grants useful insight into Earth's own atmosphere. One of<br />
the tools used at JPL for the modeling of planetary atmospheres and protostellar disks is KINETICS. KINETICS can<br />
simulate years of complex dynamics and chemistry. At the moment, KINETICS is in use by or planned to be used<br />
for:<br />
1. The ExoMars mission to study the Martian atmosphere in 2016.<br />
2. Modeling of the composition of the protostellar disks which lead to star formation.<br />
3. Examination of pollutants in Earth's atmosphere.<br />
4. Modeling of Titan's atmosphere<br />
Because of the complexity of KINETICS, it requires large amounts of computation to run simulations at a<br />
meaningful level of detail. Using MPI (Message Passing Interface), a popular multi-core programming model, will<br />
permit more useful experiments to be run in less time, accelerating the discovery process. The advantages of MPI<br />
lie in its exposure of communication, allowing programmers to manually tune the communication pattern of an<br />
application. MPI will be used with the existing FORTRAN code to parallelize and accelerate performance-critical<br />
sections of code in KINETICS.<br />
151
Improving the Thermoelectric Efficiency of Mg2Si<br />
George M. Hakimeh<br />
Mentors: Sabah K. Bux and Jean-Pierre Fleurial<br />
Thermoelectric devices have been a proven, established, and dependable power source technology for many<br />
unmanned terrestrial and deep space missions for the past 50 years. Although current thermoelectric generators<br />
have been both reliable and successful, a significant increase in their relatively low thermal-to-electric energy<br />
conversion efficiency is needed for a variety of future NASA science missions. Recent research suggests that<br />
nanostructuring materials can increase the dimensionless thermoelectric figure of merit ZT, a gauge of a material’s<br />
thermal-to-electric conversion efficiency, by reducing the lattice thermal conductivity via increased phonon<br />
scattering due to the nanoscale of the grains or the presence of nanostructured inclusions. As an example, of<br />
nanostructured bulk Si was enhanced by 250% at 1275 K compared to single crystal Si. Building upon the previous<br />
work on high temperature materials, nanostructured silicide-based materials are ideal for maximizing device<br />
performance at the lower temperature range (500-775 K) range. Magnesium silicide (Mg2Si) is particularly<br />
attractive due to its low density, low toxicity, thermal stability, relative abundance and low cost of production.<br />
However, the ZT of the n-type material remains low, about 0.7 at 775 K. Theory suggests that significant increases<br />
in ZT could be achieved by forming composites containing various nanoscale (< 10 nm) silicide inclusions. Here we<br />
present the impact of nanostructuring Mg2Si on the thermoelectric transport properties of magnesium silicide.<br />
System-Level Trade Analysis of the Entry, Descent, and Landing System and the Mobility System of the<br />
Mars Rover<br />
Bryan He<br />
Mentors: Yoshiaki Kuwata and Marco Pavone<br />
The potential Mars rover mission depends on two main systems: the entry, descent, and landing system; and the<br />
mobility system. The entry, descent, and landing system has an inherent uncertainty due to errors in sensor<br />
measurements, variations in atmospheric winds, and dispersions caused during travel through the atmosphere.<br />
Consequently, the target landing location is best represented as an ellipse rather than as a single point. Failure of<br />
the entry, descent, and landing system can occur due to impact with a hazard or landing in a location with a slope<br />
steeper than the maximum tolerance of the lander. The mobility system determines the movement of the rover<br />
after landing. Failure of the mobility system can occur when the rover lands in an area surrounded by hazards or<br />
due to mechanical failure of the rover. The probability that mechanical failure occurs is a function of the distance<br />
travelled and the proximity to hazards. The optimal landing ellipse is the location with the minimum combined<br />
probability of landing failure and mobility failure. The algorithm currently being developed accounts for the<br />
probabilistic nature of the initial entry phase by using a reverse dynamic programming method and determines the<br />
optimal landing target.<br />
Panosphere: Panoramic Martian Landscapes in an Immersive Environment<br />
Daniel L. Healy<br />
Mentor: Victor Luo<br />
NASA Outreach programs are important to get students and the public interested in space and Mars exploration.<br />
The OPS Lab is currently working on several outreach programs that deal with natural user interfaces including<br />
Microsoft’s Kinect. I'm working with the OPS Lab on an interactive 3D Mars outreach experience created in Unity.<br />
The application allows users to follow the path of the Spirit rover, learn more about Spirit's journey, and experience<br />
the Martian surface in 3D. The imagery and height maps for the terrain were captured by several orbiters at<br />
varying levels of detail, with emphasis on areas close to Spirit's path. This enables us to load images of the terrain<br />
at higher-resolution as the user rolls closer. To achieve even higher detail, I am working with a partner on adding a<br />
first-person panoramic experience which can be navigated in a way similar to Google's Street View. Future work for<br />
me includes completing the feature, optimizing the feature so it can be merged with the main experience, and<br />
contributing to a separate project for Kinect.<br />
DC Magnetics System Design<br />
Jared Daniel Helms<br />
Mentor: Pablo Narvaez<br />
DC Magnetics testing is vital for replicating the magnetic environment that spaceflight hardware might experience<br />
during its mission, from very low magnetic fields in deep space (about 2 nanoTeslas) to stronger fields of up to<br />
1 x 10^6 nanoTeslas (10 Gauss), such as that around Jupiter. The objective of this project is to develop a semiautomated<br />
system to do DC Magnetics testing in the environments experienced by flight hardware during its<br />
mission into outer space (from anywhere near Earth to orbits around Jupiter). The system required a Helmholtz coil<br />
to negate Earth’s magnetic field, magnetometers to measure the ambient magnetic field, a device to communicate<br />
with and control the coil, and a shaft encoder connected to a turn table to track the movement of hardware within<br />
the null-field as the hardware is rotated 360 degrees for mapping its magnetic field signature. All of the above were<br />
connected to a computer with LabVIEW software applications developed to process the data gathered by the<br />
152
instruments in order to automate the measurements. The fully functional system measures the magnetic field<br />
produced by hardware at multiple points corresponding to the angle at which it is turned. The resulting data can be<br />
used to characterize hardware’s magnetic field, find its magnetic dipole moment, and determine the impact it has<br />
on other hardware.<br />
ATHLETE: Tool and Test Bed Development<br />
Gabe Herz<br />
Mentor: Christopher McQuin<br />
The All-Terrain, Hex-Limbed, Extra-Terrestrial Explorer (ATHLETE) is a 6 limbed concept rover. Each limb has<br />
6 degrees of freedom and a wheel located at the end. ATHLETE is capable using the wheels to drive across gentle<br />
terrain. When rough terrain is encountered, the wheels lock and the limbs can be used for climbing. Each limb can<br />
also serve as a robotic arm using a tool attachment mechanism. These unique features allow ATHLETE to be used<br />
for transporting heavy cargo or to serve as a platform for astronaut mobility. The current focus of ATHLETE’s<br />
development is directed towards low gravity environments like asteroids and comets. For ATHLETE to operate in<br />
these conditions, it must be able to anchor itself. Once secured to the surface, ATHLETE will be able to use various<br />
end effectors for digging or drilling. The focus of this project is to complete the development of a counter rotating<br />
auger anchor, and develop test beds to evaluate it, and other end effectors’ efficiency.<br />
The Development of a Sub-Kelvin Grating Spectrometer Test Bed<br />
Jeremy Hodis<br />
Mentor: Charles M. Bradford<br />
We have researched the possibility of making a ground based far-IR silicon spectrometer using a grism as the main<br />
grating element. The spectral range of the spectrometer was based on the atmospheric window of CCAT’s future<br />
location on Cerro Chajnantor in Northern Chile. The wavelengths of interest range from 405-521 µm and 304-391<br />
µm. Some potential benefits of this far-IR spectrometer design include smaller size and fewer optical components.<br />
This spectrometer would be a good candidate as an instrument on the CCAT telescope so CCAT’s properties were<br />
parameters in our designs. Since this spectrometer isn’t of conventional design some research needs to be made<br />
into its fabrication process.<br />
SSCAE Systems Audit of Software Tools Used by Division 31<br />
Aaron Hoffman<br />
Mentors: David Mohr and Victoria Scarffe<br />
Systems and Software Computer Aided Engineering team provides support on tools used by Division 31. The<br />
SSCAE team has been working to understand what the division’s employees’ wants and problems are and their<br />
opinions on the software tools that SSCAE supports. SSCAE focuses on tools whose licenses are purchased instead<br />
of tools that are developed at JPL or NASA. SSCAE has also searching for ways to help overall production and<br />
communication. In order to obtain the data needed, a survey of the mission operations employees in Division 31<br />
and Sections 342 and 345 was conducted and concluded. Interviews of employees in Division 31 have also been<br />
conducted to supplement the data collected from the survey. Based on the data so far, there have been four main<br />
issues and requests: communication, basic productivity tools, Adobe products, and collaborative editing. The data<br />
collected shows the usage of the programs supported by SSCAE. Based on this data, reductions in the number of<br />
licenses of various products will take place, as well as the implementation of the requests by employees. After<br />
these changes occur, another survey and additional interviews of employees should be conducted.<br />
Construction and Testing of Active Precision Reflectors for the Purpose of Surface Shape Control and<br />
Deformation Compensation<br />
Samuel Hoffman<br />
Mentor: Samuel Case Bradford<br />
Three piezoelectrically active reflectors have been developed and tested. The first is a meter-scale hexagonal<br />
composite reflector consisting of a spherically curved graphite face sheet, aluminum honeycomb core composite<br />
panel with a network of 90 distributed piezoelectric composite actuators. The actuator system may be used for<br />
controlling the structural dynamic response of the reflector, or for correcting low-order thermally-induced quasistatic<br />
distortions in the panel. Thermally-induced surface deformations of 1 to 5 microns were deliberately<br />
introduced onto the reflector, measured using a speckle holography interferometer, and corrected using a<br />
piezoelectric composite actuator network distributed across the back face sheet of the reflector. The second<br />
reflector is a metal hexagon with an outer diameter of 23cm, a polished metal front surface, and a network of<br />
12 actuators on the back face. It is designed to test reflector scalability and performance. The third reflector is a<br />
circular borosilicate window 150mm in diameter with 15 actuators on the back face. Absolute deformation can be<br />
measured by a standard optical interferometer. Using this interferometer we have shown that the piezoelectric<br />
actuator network can correct the surface flatness of the window to better than the original 4λ/inch.<br />
153
Volcanic Ash Plumes: An Analysis With Multi-Angle Imaging SpectroRadiometer (MISR) Images<br />
Collin Holmquist<br />
Mentors: Veljko Jovanovic and Olga Kalashnikova<br />
The analyses of volcanic plumes are pertinent to the study of plume dynamics. Ash plume heights and thickness<br />
are important for analyzing ash air pollution and transport. Of interest for this project are the extents to which ash<br />
plumes rise, thus determining the probability of mixing within the Planetary Boundary Layer or moving into the free<br />
atmosphere. To help study the plumes, NASA’s Multi-angle Imaging SpectroRadiometer (MISR) instrument has the<br />
ability to create stereoscopic views of features below. Using the MISR images, the MISR Interactive eXplorer<br />
(MINX) program computes plume wind speed and heights. A database of MINX digitized volcanic eruptions and ash<br />
plumes will be created. It is possible that statistics of the eruptions can be then compiled in future, if not during the<br />
time at JPL, that will focus on plume height compared to eruptive strength. As a second part to the project, an indepth<br />
study of the plume heights of the Puyehue-Cordon eruption in Chile will be used in trajectory models. The<br />
results will then be compared to other findings by volcanologists.<br />
Granular Media Simulations Application in Mechanical Design<br />
Ryan Houlihan<br />
Mentors: Rudranarayan Mukherjee and Marc Pomerantz<br />
We explore the capabilities of massively parallel granular media simulations and its feasibility for its use in analysis<br />
and design of various mechanical systems. Properties of granular media and the interaction of different granular<br />
particles with mechanical systems and themselves is largely important in the design and optimization of such<br />
mechanical systems. To visualize and analyze the interactions and locomotion of mechanical systems on granular<br />
mediums massively parallel simulations are quite advantageous. The simulations are highly parallelized and highly<br />
scalable and can help us both visually as well as numerically analyze the effects certain mechanical systems have<br />
on these mediums through computing the various phenomenological first-order interaction forces in the system.<br />
Our work explores what is currently feasible with granular media simulations as well as the effects different<br />
granular materials have in terms of mechanical locomotion and interaction with these mediums<br />
Investigation of Visualization Tools for Integrated Spacecraft Analysis<br />
Cody A. Hyman<br />
Mentors: Yu-Wen Tung and Pierre Maldague<br />
Spacecraft modeling, a critically important portion in validating planned spacecraft activities, is currently carried<br />
out using a time consuming method of mission to mission model implementations and integration. A current<br />
project in early development, Integrated Spacecraft Analysis (ISCA), aims to remedy this hindrance by providing<br />
reusable architectures and reducing time spent integrating models with planning and sequencing tools. The<br />
principle objective of this internship was to develop a user interface for an experimental ontology-based structure<br />
visualization of navigation and attitude control system modeling software. To satisfy this, a number of tree and<br />
graph visualization tools were researched and a Java based hyperbolic graph viewer was selected for experimental<br />
adaptation. Early results show promise in the ability to organize and display large amounts of spacecraft model<br />
documentation efficiently and effectively through a web browser. This viewer serves as a conceptual<br />
implementation for future development but trials with both ISCA developers and end users should be performed to<br />
truly evaluate the effectiveness of continued development of such visualizations.<br />
Capabilities of the Bullet Physics Engine: Collision Detection for Physics-Based Simulation of Multi-Body<br />
System<br />
Rasheed Ibrahim<br />
Mentor: Rudra Mukherjee<br />
Collision Detection is critical to applications in physics-based simulation and robot modeling that includes the<br />
intersection of two or more (soft or rigid body) objects. In this paper, we present a collision detection algorithm<br />
implemented in the Bullet Physics Engine for the intersection of two or more multi-body system. Furthermore, we<br />
provide physics-based simulations for soft body and rigid body dynamic systems that demonstrate the capabilities<br />
of the Collision Detection library in the Bullet Physics Engine. First, we created a basic simulation of a rigid ball<br />
colliding into a static block using the built-in rigid body collision shapes provided by the physics engine. Then, we<br />
apply a convex hull generator algorithm written by Stan Melax to approximate a custom arbitrarily complex<br />
concave triangle mesh as a collection of rigid convex objects for use in the Bullet Physics Engine. Finally, we<br />
explored the soft body dynamics library in the physics engine. We created a basic simulation that shows two soft<br />
body objects colliding into one another and a separate simulation that shows a soft body object colliding into a rigid<br />
body object. We find the time, location and existence of contact between pairs of objects in simulation and it is<br />
expected that the Bullet Physics Engine will demonstrate good collision detection capabilities for our use.<br />
154
Flight Rule Identification, Implementation, and Verification for Mars Science Laboratory<br />
Lee Jasper<br />
Mentor: Andrew Mishkin<br />
Mars Science Laboratory is poised to launch late this year. Because launch is approaching, significant effort is being<br />
put into creation and completion of operations processes and software for the Launch, Cruise, and Approach (LCA)<br />
phases of the mission. Flight Rules (FRs) are one component of the operations processes that govern flight<br />
operations. FRs provide guidelines and constraints for safe and proper operation of the flight hardware. They are<br />
implemented in software and operations procedures to ensure that harmful events do not occur. LCA FRs are<br />
currently being implemented, verified, and validated. Implementation of each FR generally falls under procedural<br />
checks or the software program SEQGEN, which is used to model and verify commands prior to execution on the<br />
spacecraft. In an effort to aid FR verification and to understand how SEQGEN modeling and adaptation occurs,<br />
several FR tests have been conducted during this project to ensure that all commands that could violate a FR have<br />
been correctly modeled, and are caught by, SEQGEN. Also, the baseline allocation of LCA FRs to the responsible<br />
operations procedures has been completed. Allocations were reviewed by the operations teams and subsystems to<br />
ensure FR checks have been incorporated into the proper procedures.<br />
Characterization of the Phase-Shifting Zernike Wavefront Sensor for Telescope Applications<br />
Rebecca M. Jensen-Clem<br />
Mentor: J. Kent Wallace<br />
The dynamic Zernike wavefront sensor (ZWFS) is a common-path, phase-shifting interferometric wavefront sensor<br />
with superior broadband and noise rejection capabilities. The ZWFS will be integrated into the Mount Palomar 200”<br />
Hale Telescope adaptive optics systems in 2013. It will also be used to phase the segmented mirrors of Cerro<br />
Chajnantor Atacama Telescope (CCAT), a 25-m submillimeter telescope recommended by the 2010 decadal survey.<br />
This summer, the dynamic ZWFS testbed was completed. This work involved: 1) optical measurements of key<br />
components, 2) revised component specifications, and 3) procurements with vendors to fabricate optical<br />
components and internal JPL fabrication resources. A numerical simulation of the ZWFS was built from first<br />
principles to compare the sensor’s theoretical and testbed performances. Measurements of the phase perturbations<br />
of a glass microscope slide were made using both the ZWFS testbed and a Zygo interferometer. These<br />
experimental validation measurements and the numerical simulations allowed us to determine the accuracy of the<br />
ZWFS phase measurements.<br />
Enhancing the GUI-Based Visualization and Management iPanel Software for ISAAC<br />
Carrine M. Johnson<br />
Mentor: Yutao He<br />
The Instrument ShAred Artifact for Computing (ISAAC) is a modular and integrated FPGA-based common<br />
instrument control and computing platform for NRA Decadal instrument missions. ISAAC includes six components,<br />
iBoard, iBus, iCore, iPackage, iBench, and iTool. This project is focus on iPanel that is part of iPackage. iPanel is a<br />
modular, flexible and extendable GUI-based software written in Python that controls and communicates with an<br />
ISAAC iBoard. Currently there is an emerging need for iPanel to collect and visualize telemetries from the ADC/DAC<br />
daughter board of an iBoard as well as display real-time data and send out commands. The aim of this project is to<br />
enhance the current version of iPanel with a SPI interface. It will implemented using wx.Python and its array of<br />
widgets. The new interface allows the user to interact with the ADC/DAC peripherals via SPI and to update register<br />
values and display the telemetry data.<br />
Performing a Telemetry Audit and Developing a Power Model for the Mars Science Laboratory Rover<br />
Kristin P. Kagetsu<br />
Mentors: David Oh and James Chase<br />
During the verification and validation stage of the project, the Mars Science Laboratory (MSL) Flight Software team<br />
runs tests on the spacecraft and rover to ensure that they meet all of the requirements. One of the tests is to<br />
perform a telemetry audit which mines the testbed telemetry database for validation of all FSW-created telemetry<br />
channels. This requires running tests in the simulated FSW environment, which allows the user to run tests from a<br />
remote computer and allows for faster iterations of the testing procedures. Once the data is acquired from each<br />
session, statistics can be determined about which channels have been called and what range of values have been<br />
logged. The ATLO team is verifying the power consumption of various components of the spacecraft and rover to<br />
create a more accurate power model of them. The ATLO baseline tests have data on instruments used during the<br />
descent and cruise stages. The current of the various components can be acquired from these tests and then the<br />
voltage is determined from WMD based on the timestamp of the current data. This allows us to calculate the power<br />
of each component and compare it to the predictions in the Power Equipment List.<br />
155
A Survey of Efforts to Characterize Parachute Performance in the Wake of a Blunt Forebody<br />
Cole D. Kazemba<br />
Mentor: Ian G. Clark<br />
As the paradigm for robotic Mars missions shifts from the Viking class ballistic entry vehicles to larger and more<br />
complex vehicles on the order of Mars Science Laboratory (~3250 kg entry mass) or greater, there is increasing<br />
emphasis on improving the capabilities and understanding of the supersonic parachutes used during atmospheric<br />
descent. A new technology being developed at JPL, the Supersonic Inflatable Aerodynamic Decelerator (SIAD), will<br />
allow for even greater entry masses than MSL. However, the large diameter and mass of an entry vehicle utilizing a<br />
deployed SIAD will push the requirements of its parachute beyond the disk-gap-band (DGB) systems qualified by<br />
the Viking program. In order to address some of challenges of predicting the expected performance of a new<br />
parachute system, an extensive literature review and data collection effort was conducted. Multiple empirical and<br />
analytical methods to predict parachute performance in a trailing wake were implemented and compared to the<br />
historical data with very limited success. The poor performance of these prediction methods and large gaps in the<br />
available experimental data indicate that significant effort is required for developing more robust and capable<br />
methods and building a thorough experimental database to anchor new and existing computational tools.<br />
Correlations Between Technical Performance and Cost Growth: A Case Study Using Risk Trends<br />
Brad Kellogg<br />
Mentors: Kevin Rice, Kevin Endo, Eric Kwan, and John Jack<br />
Flight Projects historically exceed the available budgetary funds allocated by NASA HQ. Project lessons-learned<br />
indicated that there are common reasons for cost growth where strong correlations existed between technical<br />
performance and cost/schedule performance. This group research project investigates historical data from the<br />
13 most recent JPL flight projects to determine trends and correlations between technical and programmatic<br />
parameters to predict cost growth. Each of the 13 projects is analyzed using 5 major data sets: technical<br />
complexity, subcontractor performance, risk trends, margin analysis and schedule slip. Each data set is analyzed to<br />
determine which correlations exists. My specific assignment for this group project is to analyze monthly risk<br />
assessments to more accurately predict cost growth. A numerical risk factor is calculated using both new and<br />
proven methods. After compiling numerical risk values, a regression analysis will be performed with cost growth<br />
data to establish trends. It is expected that certain aspects of technical performance parameters will lend<br />
themselves to be more accurate cost predictors than others. Conclusions derived from this group research can be<br />
utilized to form a historical data base for future projects and compared against the current in-work flight projects<br />
to predict cost growth.<br />
Analysis and Predictions of Mission Downlink Data Volume for Mars Reconnaissance Orbiter Extended<br />
Missions<br />
Andrew K. Kennedy<br />
Mentors: Dan Johnston, Dave Herman, and Reid Thomas<br />
The Mars Reconnaissance Orbiter (MRO) has returned more scientific data to date than the combined totals of all<br />
other deep space missions. Dependable estimates of the amount of continuing downlink data from MRO are<br />
important in supporting science and ground system planning, as well as honoring the project’s commitments to the<br />
Mars Exploration Program office.<br />
For these reasons, a study was performed on the downlink process for MRO in order to update the project’s data<br />
volume analysis tools and provide data estimates for future extended missions. As part of this study, it was<br />
necessary to research in depth the MRO’s communications link through the Deep Space network, in addition to the<br />
spacecraft’s orbital geometry. This knowledge was used to develop streamlined algorithms to calculate data<br />
volumes, and update the previous analysis tools.<br />
The final result of this study was a documented knowledge of MRO’s “downlink algorithm”, as well as specific<br />
predictions of downlink volume for MRO’s current extended mission, and next planned extended mission. The<br />
documentation produced will also serve as a user’s guide for future mission planners to more effectively make data<br />
volume predictions.<br />
MSL Surface Operations: Remodeling of the Mission Operations System Tactical Processes<br />
Sheehan Khuu<br />
Mentors: Andrew Mishkin and Alicia Allbaugh<br />
Surface operations for the Mars Science Laboratory (MSL) rover, Curiosity, are split into two categories: strategic<br />
and tactical. Strategic processes, performed over multiple days, include, among others, determining<br />
communication windows and scheduling with the DSN. Tactical processes define the specific activity plans for the<br />
rover for the next Martian day (sol), and develop the command sequences that implement the activity plans<br />
onboard the rover. The MSL Mission Operations System (MOS) includes the tools, processes, procedures, and<br />
teams that perform the planning and execution for Curiosity. The MSL MOS is modeled to validate that the tactical<br />
156
processes produce the appropriate products for consistent uplink of the sequence load. Artifacts extracted from the<br />
model designed in Unified Modeling Language (UML) were converted to a PowerPoint based design. The UML model<br />
ineffectively communicated the higher level tactical processes to the applicable audience. The purpose of the<br />
remodeling is to remedy this problem. The model is expected to provide a robust high to mid-level view of the<br />
tactical processes and serve as a review product for the Integration, Planning and Execution (IPE) and mission<br />
operations team. This paper describes the approach and preliminary results of the conversion process from a UML<br />
to PowerPoint model.<br />
Fault Tolerant RCS Thruster and Thruster Drive Electronics Architectures for EHM Orbiter<br />
Brent A. Kisling<br />
Mentor: Ray Crum<br />
Results from the Vision and Voyages for Planetary Science in the Decade 2013-2022 report indicated that JEO<br />
(Jupiter Europa Orbiter) “should be flown only if changes to both the mission design and the NASA planetary<br />
budget make it affordable without eliminating any other recommended missions.” Due to this motivation, a need<br />
arose to examine alternative fault tolerant RCS (Reaction Control System) thruster and thruster drive electronics<br />
architectures that may lower the overall system cost of the revamped JEO mission, now entitled EHM (Europa<br />
Habitability Mission). Through my research, I will identify alternative RCS thruster and thruster drive electronics<br />
architectures and then evaluate those architectures using a set of weighted criteria; including architecture size,<br />
cost, mass, fault tolerance, and reliability. Based upon this criteria and corresponding weights, a decision matrix<br />
can be formed to compute the weighted score of each alternative architecture. The architecture with the highest<br />
weighted score will be suggested as the best solution for EHM.<br />
Multi-Mission Space Exploration Vehicle: Simulation of Operations in Proximity to a Near Earth Object<br />
Heather Kline<br />
Mentors: J. Balaram and Marco Quadrelli<br />
This project focuses on simulating the dynamics of an MMSEV (Multi-Missions Space Exploration Vehicle) on<br />
rendezvous to collect samples from a NEO (Near Earth Object), modeling the control characteristics of the<br />
spacecraft using DSENDS (Dynamics Simulator for Entry, Descent, and Surface landing). A mission of the MMSEV is<br />
to collect samples from an asteroid, a mission which requires the spacecraft to be able to navigate to an orbit<br />
which keeps it stationary over an area of a spinning asteroid while a robotic arm interacts with the surface. The<br />
capability to model reaction wheel and thruster control has been added. Additional capabilities added include<br />
station keeping, where the spacecraft holds itself steady over a spot on the rotating asteroid's surface. In addition<br />
to the added functionality the results of a parametric study of control gain sensitivity and Monte Carlo simulation of<br />
the behavior near the surface are discussed. The functions used to control the spacecraft with simulated reaction<br />
wheels and thrusters have been written in a modular way such that they could easily be used in other systems, and<br />
it is compatible with multiple thruster allocation methods.<br />
Jupiter’s Magnetosphere: An Analysis of Electron Pitch Angle and Heavy Ion Loss to Europa<br />
Nicholas Knezek<br />
Mentor: Michael Kokorowski<br />
Jupiter’s magnetosphere is the largest object in the solar system and it dominates the local space environment.<br />
Through past analysis of Pioneer 10 and 11, Voyager 1 and 2, Ulysses, and Galileo data, several regions and<br />
important transport processes have been defined within it. The project has two objectives: examine the transition<br />
region between the inner and outer magnetosphere and identify losses of heavy ions to Jupiter’s moons, Europa in<br />
particular. For the first objective, data from the magnetometer and Energetic Particle Detector on Galileo were<br />
utilized to reproduce previous results showing an electron pitch angle distribution change from a trapped<br />
distribution in the inner magnetosphere to a field-aligned distribution in the outer magnetosphere. Due to<br />
limitations of the Galileo orbiter transmission rate and reduced dataset, further results were impossible to obtain.<br />
Focus was then shifted to analysis of data from the Heavy Ion Counter. Using oxygen measurements at various<br />
energy levels, phase space density were calculated for ions with various magnetic moments, then used to examine<br />
whether significant losses of Heavy Ions occur near the orbit of Europa.<br />
Improving Measurement Systems and Instrumentation<br />
Stephen Krach<br />
Mentor: Lauren Halatek and Ian Ruiz<br />
The Thermal Energy Conversion Technologies group is in need of updated test systems. My objective is to convert<br />
the existing Visual Basic code that commands the PID controller into a maintainable LabVIEW program. By using<br />
flow charts, original code, and help from peers, I am on my way to completing this important sub-function upgrade.<br />
Electronic Ground Support Equipment consists of a rack complete with test equipment and custom electronics (aka<br />
chassis). My task is to create two templates; one will output the rack’s center of gravity, and the other will validate<br />
adequate cooling for the rack.<br />
157
A low cost breakout box (BOB) has been designed that is half the cost of the conventional BOB. My objective is to<br />
finalize this design by creating a bill of materials, receiving quotes, and gathering the parts for a technician to<br />
assemble. I will also be working with quality assurance to develop an appropriate testing procedure to ensure the<br />
BOB is qualified to connect to flight hardware.<br />
Orbit Determination Techniques for the Mars Reconnaissance Orbiter<br />
Sandeep Krishnan<br />
Mentor: Allen Halsell<br />
Orbit determination and maneuver design are critical components of navigation for a spacecraft. Orbit<br />
determination is defined as obtaining the knowledge of a vehicle’s motion in space in a specified coordinate system.<br />
The first part of this process involves estimating the state of the spacecraft using a least squares method. This is<br />
accomplished by correcting a nominal trajectory in order to “best” fit observed measurements. Once an estimate<br />
has been made, the spacecraft’s state can be predicted using numerical integration. This entire process is repeated<br />
continually as the spacecraft’s motion evolves. The Mars Reconnaissance Orbiter (MRO) is a spacecraft developed<br />
in order to explore Mars through remote sensing observations. Launched in 2005, it reached its final science orbit<br />
in March 2006 and began its primary science phase. For deep space missions such as MRO, the primary method of<br />
tracking utilizes radiometric techniques. The MRO navigation team is responsible for reconstructing and predicting<br />
the spacecraft’s orbits based on Doppler data provided by the Deep Space Network (DSN). For MRO, the orbit<br />
determination process uses the Monte software. New versions of Monte and other software tools are currently<br />
being validated. This report details the testing procedure as well as the preliminary results.<br />
DARTS Simulation for a Dexterous Manipulator (ARM)<br />
Calvin Kuo<br />
Mentor: Abhinandan Jain<br />
The dynamics and real time simulations (DARTS) lab at the Jet Propulsion Laboratory (JPL) provides physics<br />
simulations for a number of rovers and spacecraft to supplement concept design and hardware testing. JPL<br />
currently has a project developing perception and grasping algorithms for a dexterous manipulator for DARPA.<br />
As the algorithms and tasks for the manipulator become more complex, testing on the physical hardware becomes<br />
more time consuming and potentially dangerous to both the hardware and the people operating it. Thus, the need<br />
for an offline simulator that can accurately simulate the physics and dynamics of the system become more<br />
prevalent. While creating models for simulations of multi-body systems in DARTS has been implemented for<br />
projects such as the Mars Science Laboratory, this simulation also requires an interface for the manipulator control<br />
software. Thus the focus of this project will be to create a simulator of the manipulator, provide an interface for the<br />
control software, and to verify that the simulated version of the manipulator behaves identically to its hardware<br />
counterpart.<br />
Performance Analysis of Magnetic Field Shapes in Hall Thrusters<br />
Maria C. Lang<br />
Mentor: Richard R. Hofer<br />
The shape and strength of the magnetic field in Hall thrusters is the dominant factor affecting plasma stability,<br />
thermal balance, performance, and life. Of particular interest to this project is the effect of changes in the shape of<br />
the magnetic field on thruster performance by modifying the applied field in a 6 kW laboratory Hall thruster.<br />
A thrust stand, ion flux probe, ExB probe, RPA, and emissive probe are used to measure the thrust, ion current<br />
density, ion species fractions, ion energy, and plasma potential for the two distinct magnetic field configurations.<br />
The differences in the fields are enabled through the use of an auxiliary “trim” coil that modifies the magnetic field<br />
in the near-anode region of the thruster. Data from the diagnostics is analyzed and applied to an analytical model<br />
of thruster performance. The model describes the dominant processes affecting thrust efficiency that is then used<br />
to identify which changes in the plasma structure are responsible for changes in thruster performance between the<br />
two magnetic field configurations.<br />
Framework for Testing Detection, Diagnostic, and Remediation Systems Within the Smart Grid<br />
Shing-hon Lau<br />
Mentors: Bryan Johnson and Kymie Tan<br />
America’s electrical grid is currently undergoing a significant paradigm shift aimed at producing a highly reliable<br />
critical national infrastructure for power – a Smart Grid. A key objective of this shift is to create a resilient electrical<br />
grid that is able to autonomously detect, diagnose, and remediate impediments to the delivery of power.<br />
Impediments can range from downed power lines to cyber-attacks to natural disasters. Measuring the effectiveness<br />
of such resilience lies solely in the domain of a rigorous testing process. Since the Smart Grid is a massive,<br />
complex cyber-physical system – one where computational and physical elements are closely intertwined –<br />
designing such a rigorous testing process will be challenging. The purpose of our work is to lay out a guiding<br />
158
framework and principles for testing the resilience of a complex, cyber-physical system. By clearly identifying the<br />
appropriate metrics, data characteristics, and error sources that comprise the critical components of a rigorous test<br />
regime, this work aims to make a significant contribution in the area of measuring the true resilience of a cyberphysical<br />
system.<br />
Radar and Lidar Remote Sensing of Forests<br />
Diana Liskovich<br />
Mentor: Marc Simard<br />
Using radar and lidar data, the aim is to improve 3D rendering of terrain, including digital elevation models (DEM)<br />
and estimates of vegetation height and biomass in a variety of forest types and terrains. The 3D mapping of<br />
vegetation structure and the analysis are useful to determine the role of forest in climate change (carbon cycle), in<br />
providing habitat and as a provider of socio-economic services. This in turn will lead to potential for development of<br />
more effective land-use management. The first part of the project was to characterize the Shuttle Radar<br />
Topography Mission DEM error with respect to ICESat/GLAS point estimates of elevation. We investigated potential<br />
trends with latitude, canopy height, signal to noise ratio (SNR), number of lidar waveform peaks, and maximum<br />
peak width. Scatter plots were produced for each variable and were fitted with 1st and 2nd degree polynomials.<br />
Higher order trends were visually inspected through filtering with a mean and median filter. We also assessed<br />
trends in the DEM error variance. Finally, a map showing how DEM error was geographically distributed globally<br />
was created. The second part of this project consisted of mapping vegetation height in Canada. A set of decision<br />
trees (DT) trained with ICEsat estimates of canopy height was produced. A set of input layers describing elevation,<br />
climate and vegetation type were used. These decision tree models were then applied to model canopy height<br />
throughout Canada.<br />
Sputtering Chamber Hardware Modification<br />
Christopher Matthes<br />
Mentor: Charles Hays<br />
Sputter deposition is the process of ejecting atoms from a target material due to the impact of gas ions, causing<br />
the resulting material to be deposited onto a substrate and form a thin film. JPL’s sputtering facility allows the<br />
creation of alloyed thin films using multiple targets, which are used for research of fuel cell cathode catalysts with<br />
improved oxygen reduction reaction (ORR) capabilities. The objective for this project was to enable better control<br />
of sputtering conditions through the implementation of computer-operated settings, which were previously<br />
manually controlled. This required hardware modifications for the gas-handling and high-vacuum systems, while<br />
introducing a computer control and data acquisition system to regulate various parameters of the sputtering<br />
process. Specifically, the introduction of an electronically interfaced pressure sensor and mass flow controller setup<br />
allowed for the regulation of desired parameters using the electronically obtained data. Installation of the new<br />
system hardware necessitated an application of design fundamentals to ensure proper assembly of all components<br />
and to address additional mechanical obstacles that surfaced through this process.<br />
Verification and Validation of Granular Media Simulation Capability With Applications in NEO Anchoring<br />
Hammad Mazhar<br />
Mentor: Marco Quadrelli<br />
This contribution describes an approach used to simulate and validate a simulation capability that is able to<br />
simulate granular dynamics simulations with millions of objects. This massively parallel granular dynamics<br />
framework relies on ubiquitous Graphics Processing Unit (GPU) cards that are capable of performing trillions of<br />
floating point operations per second and can simultaneously process hundreds of computational threads, making<br />
them ideally suited for the task at hand. Utilizing a variety of experimental data, the simulation capability was<br />
validated with existing literature. Tests included a lift/drag and ball drop test in a bed of granular material, a steady<br />
state vibration test, a brazil nut test and a penetration test. This parallel simulation capability was used to<br />
investigate two types of anchoring systems, and inflatable anchor and a helical anchor, when deployed on a Near<br />
Earth Object (NEO) in a low gravity environment. In order to successfully simulate the anchor and granular<br />
material in this environment, the simulation capability was augmented with a cohesion model capable of capturing<br />
the necessary behavior and short range forces associated with NEOs. Described herein are the results associated<br />
with the experimental validation and anchor tests along with a recommendation on which anchor design is most<br />
promising.<br />
Life Test Testbed Development for SMAP’s Attitude Control System Thruster<br />
Colleen McAvoy<br />
Mentors: David Vaughan, Corinne Gatto, and Barry Nakazono<br />
The attitude control system thrusters for the SMAP mission exceed the qualification through the number of cold<br />
starts, total number of pulses, and total propellant. A life test will be conducted within a year to analyze the<br />
performance of an individual thruster over a wide range of operating parameters including steady state, pulsed,<br />
and temperature extremes. A schematic representing the layout of the testbed for the hydrazine fueled<br />
159
monopropellant system is being developed to the scope of the test and modified for potential hazards arising from<br />
hydrazine’s toxicity and combustibility. In parallel, a virtual interface is being developed with NI LabVIEW and will<br />
be used to operate the life test. Upon completion of the schematic, the necessary hardware will be ordered and<br />
manufactured to begin the testbed assembly. In addition, the test procedure will need to be written based on the<br />
final schematic design. A future task will be to create an interactive graphic user interface for simulating the test<br />
procedure in conjunction with the test to ensure a safe testing environment and to visually confirm the required<br />
test actions.<br />
Martian Atmospheric Modeling of Scale Factors for MarsGRAM 2005 and the MAVEN Project<br />
Christopher McCullough<br />
Mentor: Allen Halsell<br />
For spacecraft missions to Mars, especially the navigation of Martian orbiters and landers, an extensive knowledge<br />
of the Martian atmosphere is extremely important. The generally-accepted NASA standard for modeling of the<br />
Martian atmosphere is called the Mars Global Reference Atmospheric Model (MarsGRAM), which was developed at<br />
Marshall Space Flight Center. MarsGRAM is useful for tasks such as aerobraking, performance analysis and<br />
operations planning for aerobraking, entry descent and landing, and aerocapture. For the Mars Atmosphere Volatile<br />
Environment (MAVEN) Project, the navigation team uses this atmospheric model to target a specified density<br />
corridor. Using data acquired from the Mars Reconnaissance Orbiter (MRO), Mars Global Surveyor (MGS), and<br />
Odyssey, the MarsGRAM model can be compared to the actual densities felt by the spacecraft by assigning<br />
MarsGRAM a scale factor. By examining these scale factors, a stochastic model with altitude, latitude, and seasonal<br />
dependences can be created that effectively models the variations in the atmospheric scale factor.<br />
DROP: The Durable Reconnaissance and Observation Platform<br />
Clifford F. McKenzie<br />
Mentors: Aaron Parness<br />
Robots can provide a remote presence in areas that are either inaccessible or too dangerous for humans. However,<br />
robots are often limited by their ability to adapt to the terrain or resist environmental factors. The Durable<br />
Reconnaissance and Observation Platform (DROP) is a lightweight robot that addresses these challenges with the<br />
capability to survive falls from significant heights, carry a useable payload, and traverse a variety of surfaces,<br />
including climbing vertical surfaces like wood, stone, and concrete. DROP is manufactured using a combination of<br />
rapid prototyping and shape deposition manufacturing. It uses microspine technology to create a new wheel-like<br />
design for vertical climbing. To date, DROP has successfully engaged several vertical surfaces, hanging statically<br />
without assistance, and traversed horizontal surfaces at approximately 30 cm/s. Unassisted vertical climbing is<br />
expected in the very near future. DROP can currently survive falls from up to 2 meters; a new design being<br />
assembled now should improve this substantially. Work on DROP is ongoing and both its mobility and survivability<br />
are still improving. Future efforts will focus on improving the microspine wheels, selecting more resilient materials,<br />
customizing the controls, and performing more rigorous and quantifiable testing.<br />
Optimization of PMA-PCR Protocol for Viability Detection of Pathogen Surrogates<br />
Brian Mikkelson<br />
Mentors: Adrian Ponce, Christine Lee, and Aaron Noell<br />
In the event of a bioterrorism attack or epidemiologic outbreak, the ability to determine whether an area is safe, as<br />
well as what species is present, in the shortest amount of time, is essential. This project’s objective was to optimize<br />
protocols for use in creating a portable instrument capable of determining viability and identification<br />
simultaneously. This is made possible by the use of the DNA-intercalating agent propidium monoazide (PMA)<br />
combined with molecular biology analyses including PCR. PMA selectively binds to dead cells, inhibiting polymerase<br />
activity, making DNA unable to be amplified by PCR. By inhibiting the amplification of DNA of any dead cells, and<br />
using specific primers, PMA-PCR results consist purely of viable samples of the target species. PMA-PCR protocol<br />
was optimized and the correlation between viable to dead cell ratio with PMA-PCR output was determined using<br />
methods of DNA extraction, PCR, phase-contrast microscopy, cell culture, and Bacillus spore germinability assay.<br />
Interference with PCR amplification by primer dimers was solved by keeping the reagents on ice until being placed<br />
in the thermal-cycler. The loss of culturability, by heat-killing, did not prevent amplified PCR products, which<br />
supports the use of PMA to prevent amplification and differentiate between viable and dead cells.<br />
Thermal Modeling of Asteroids Observed With the Wide-Field Infrared Survey Explorer (WISE)<br />
Wenli Mo<br />
Mentor: Amy Mainzer<br />
The Wide-field Infrared Survey Explorer (WISE) has observed ~157,000 minor planets, many of them previously<br />
unknown. With these observations, astronomers can determine important physical parameters, including the<br />
albedo and diameter of the object, through thermal modeling. I will explain the methods of the Near Earth Asteroid<br />
160
Thermal Model (NEATM) using the object’s observed data from visible and infrared wavelengths. Also, I will present<br />
the general results of the thermal modeling of near earth objects (NEOs) observed by WISE and future analysis to<br />
be conducted with the resulting physical parameters.<br />
Multi-Angle Rear Viewing Endoscopic TooL (MARVEL): Mechanical Design and Analysis<br />
Jack M. Mondry<br />
Mentors: Harish Manohara and Michael Shearn<br />
Minimally Invasive Surgery (MIS) provides the patient great advantages including shortened recovery time, less<br />
cosmetic scarring, and less risk of infection. However, this type of procedure is generally more complex than a<br />
traditional open surgery, especially when performing neurosurgery. A major factor of this complexity has to do with<br />
the reduced field of view the surgeon must operate in. Traditional endoscopes provide the surgeon a fixed<br />
viewpoint and often only in two dimensions which hinders depth perception and therefore decreases accuracy and<br />
increases surgery time. This project involves implementation of a stereoscopic (3D) camera to an actuated revolute<br />
joint that will provide the surgeon alternate viewing angles of the surgical site. Due to the working environment of<br />
the MARVEL and the unacceptable consequence of device failure, strict design requirements were applied. A<br />
concept was developed that allows the tool to be sterilized for multiple uses while maintaining a small form factor<br />
and meeting all design requirements. Multiple actuation methods were prototyped and analyzed for performance.<br />
Wideband Spectroscopy: The Design and Implementation of a 3 GHz, 2048 Channel Digital<br />
Spectrometer<br />
Ryan Monroe<br />
Mentor: Robert Jarnot and Paul Stek<br />
A state-of-the-art digital Fourier Transform spectrometer has been developed, with a combination of high<br />
bandwidth and fine resolution unavailable elsewhere. Analog signals consisting of radiation emitted by constituents<br />
in planetary atmospheres or galactic sources are downconverted and subsequently digitized by a pair of interleaved<br />
Analog-to-Digital Converters (ADC). This 6 Gsps (giga sample per second) digital representation of the analog<br />
signal is then processed through an FPGA-based streaming Fast Fourier Transform (FFT), the key development<br />
described below. Digital spectrometers have many advantages over previously used analog spectrometers,<br />
especially in terms of both accuracy and resolution, both of which are particularly important for the type of<br />
scientific questions to be addressed with next-generation radiometers. The implementation, results and underlying<br />
math for this spectrometer, as well as potential for future extension to even higher bandwidth, resolution, channel<br />
orthogonality and radio interference detection needed to support proposed future advanced atmospheric science<br />
and radioastronomy, are discussed.<br />
Mission Planning and Sequencing Investigation of Third Party Work Flow Software<br />
Michael Mozingo<br />
Mentor: Barbara Streiffert<br />
The Mission Planning and Sequencing (MPS) element of the Multi-Mission ground System and Services (MGSS)<br />
provides space missions with software that meets the demands of multiple missions. The requirements on this<br />
software set are to be able to plan spacecraft activities, sequence commands to the spacecraft, package the<br />
commands and notify the operations team that the command is ready to be radiated to the vehicle. The Automated<br />
Sequence Processor (ASP) is used to perform all of these functions if the command is a non-interactive command<br />
(NIC). Non-interactive commands are guaranteed not to harm the spacecraft. To update the ASP the use of a<br />
commercially available workflow engine has been investigated. Two workflow engines have been prototyped and<br />
the work has been documented with the pros and cons of various engines. In addition, a recommendation has been<br />
made. The following paper discusses the prototypes and the process used for evaluation.<br />
Conductor: Kinect Motion-Tracking for NASA Outreach and Robotics Potential<br />
Joshua Nuernberger<br />
Mentor: Victor Luo<br />
The scope and objective of Project Conductor is to apply natural user interfaces, utilizing the Microsoft Kinect, for<br />
NASA Outreach application and robotics control. The methodology of the project emerges from a collaborative<br />
process beginning in research, brainstorming, conceptualizing, prototyping, and then culminating in development.<br />
Current projects include the topics of Space Camp and NASA Robotics. From these, gameplay mockups and<br />
concepts are presented to the team for integration. My main area of contribution is focused on the conceptual<br />
potential of the experiences, including game design and establishing a visual style. However, it also includes some<br />
aspect of the developmental side of the project, using the Unity game engine. The final goal, working with the rest<br />
of the team, is to create fun and educational gameplay experiences for NASA Outreach.<br />
161
Spectral Emissions of a Pulsed Nitrogen Cold-Gas Thruster and a Hall Thruster System<br />
Kelsey S. O’Connor<br />
Mentors: Lee Johnson and David Conroy<br />
Quantitative spectral emission measurement were performed to identify and investigate the nature of charge<br />
transfer, recombination, and back-scattering created when a pulsed nitrogen cold-gas thruster and a hall thruster<br />
are simultaneously operated in close proximity. The measurements were taken with the specific target of<br />
identifying and investigating the nature of spectral changes in the ionized plasma plume of the hall thruster in<br />
order to gain a comprehensive understanding of the effect these interactions will have on the lifetime and<br />
effectiveness of the Hall thruster and the spacecraft it is attached to.<br />
Developing a Plug-and-Play Framework for the ISAAC iBoard<br />
Kevin Ortega<br />
Mentor: Yutao He<br />
iBoard is a common integrated instrument-control-and-computing FPGA-based board used to simultaneously<br />
process telemetry from attached instruments, such as cameras, radars, and so forth. Since iBoard will handle<br />
multiple types of instruments, there needs to be an autonomous way for detecting and communicating with its<br />
connected devices. To implement it, a plug-and-play framework has been developed for iBoard and its attached<br />
peripherals. We have performed a survey on other plug-and-play platform technology such as Universal Serial Bus<br />
(USB), Android software stack, Universal Plug-and-Play (UPnP), etc. Also, iBoard and peripheral interconnection<br />
requirements have been developed. The framework has been developed and the prototype has also been<br />
implemented. This framework allows for a compatible peripheral to be easily connected to and communicated with<br />
an iBoard.<br />
Processing and Analyzing Snow Cover Data From the Moderate Resolution Imaging Spectroradiometer<br />
(MODIS) Instrument<br />
Andrew Park, Jr.<br />
Mentor: Kevin Hand<br />
Moderate Resolution Imaging Spectroradiometer (MODIS) is an instrument aboard the Terra and Aqua satellites<br />
that images the Earth surface to provide scientists with invaluable data about Earth’s lands, oceans, and lower<br />
atmosphere. Of the many data products that have been transferred to ground stations for analysis, snow cover<br />
data is especially relevant to current research of icy moons such as Europa, which many scientists and<br />
astrobiologists believe to be habitable for extraterrestrial life. As part of an ongoing effort to better understand the<br />
chemical phenomena that occur during freezing, MODIS data from the Aqua satellite will be processed using<br />
modified Matlab scripts to look at snow cover from a 500 meter spatial resolution at 8-day coverage intervals. With<br />
this data, it will then be possible to look at changes in the snow cover in key regions such as Barrow, Alaska, where<br />
pockets of methane gas are beginning to form. By investigating these regions, scientists hope to find trends in<br />
these methane levels and apply this knowledge to current understanding of the complex chemistry that occurs in<br />
icy environments.<br />
Real-Time Event Management for the Mars Relay Operations Service<br />
Corey Peterson<br />
Mentors: Michael Wallick and Daniel Allard<br />
The Mars Relay Operations Service (MaROS) comprises a number of tools to coordinate, plan, and visualize various<br />
aspects of the Mars Relay Network. As part of MaROS, I have developed and implemented a feature set that<br />
operates on several levels of the software architecture. These levels include a web-based user interface, a backend<br />
“ReSTlet” built in Java, and databases which store the data as it is received from the network.<br />
My primary goal this summer was to develop a real-time event notification and management system, so mission<br />
teams can track and act upon events on a moment-by-moment basis. It additionally allows users to post important<br />
event-related messages for team members within the same feed.<br />
My workflow has been as follows: collaborate with the end-users and developers to identify the features they would<br />
like to see, assemble the back-end architecture to manage communication with the databases, as well as<br />
recognizing, filtering, and sending event information to the user interface. Lastly, I build the interface components<br />
that comprise all the elements of the event notification system. Once this basic process is completed, I iterate and<br />
expand upon my design by adding or changing features based on developer and user feedback.<br />
162
Ocean Forecasting in the Gulf of Mexico: User Interface and Accuracy Analysis<br />
Heidi K. Peterson<br />
Mentor: Yi Chao<br />
Currently JPL is creating an ocean forecasting system for the Gulf of Mexico and I was charged with identifying<br />
users and the user’s needs as well as checking the accuracy of the water velocity hindcasts conducted by the<br />
program. Since the most interested user demographic was oil and oil support companies I designed a user<br />
interface for oil operators and researched the ocean conditions of interest to oil companies. There was some<br />
concern for large changes in temperature and salinity; however the greatest concern was eddies. Eddies<br />
occasionally form and break free from the Loop Current that enters the Gulf between Cuba and the Yucatan<br />
Peninsula and leaves the Gulf between Florida and Cuba. These eddies cause changes in water velocity that are<br />
dangerous to offshore oil platforms and offshore construction. I created a user interface for oil platform operators,<br />
which will give them the forecasting information they need in an easy to use format. I am also developing an<br />
animated graph showing actual and model water column velocity data. I hope to also complete the statistics for<br />
variances and error in the model.<br />
Orbit Determination and Navigation Software Testing for the Mars Reconnaissance Orbiter<br />
Alex J. Pini<br />
Mentors: Allen Halsell and Premkumar Menon<br />
During the extended science phase of the Mars Reconnaissance Orbiter’s lifecycle, the operational duties pertaining<br />
to navigation primarily involve orbit determination. The orbit determination process utilizes radiometric tracking<br />
data and is used for the prediction and reconstruction of MRO’s trajectories. Predictions are executed twice per<br />
week and are used to recommend corrections to the satellite, such as ephemeris updates and maneuver delta-v’s.<br />
Reconstructions, which incorporate a batch estimator, provide precise information about the spacecraft state to be<br />
synchronized with scientific measurements. These tasks were conducted regularly to validate the results obtained<br />
by the MRO Navigation Team. Additionally, the team is in the process of converting to newer versions of the<br />
navigation software and operating system. The capability to model multiple densities in the Martian atmosphere is<br />
also being implemented. However, testing outputs among these different configurations was necessary to ensure<br />
compliance to a satisfactory degree.<br />
Coronae Formation on Venus via Extension and Lithospheric Instability<br />
Danielle Piskorz<br />
Mentor: Suzanne Smrekar<br />
Coronae are semi-annular volcanotectonic formations that dot the surface of Venus. There exist many hypotheses<br />
that explain their origin, but perhaps the least understood is the origin of the coronae located 150 to 1500 km from<br />
their associated rift. The goal of this project is to show that a mantle plume associated with a rift can melt the<br />
lower lithosphere and cause lithospheric dripping into the upper mantle, leading to extension and surface stresses.<br />
Using a finite element code that solves equations for the conservation of heat, mass, and momentum, we vary<br />
lithospheric thickness, mantle density, and the mantle's viscosity structure. We compare the predicted surface<br />
topographies, gravity anomalies, and melt compositions of our successful models to SAR and VIRTIS observations.<br />
By further characterizing the relationship between rifts and coronae, we may be able to better understand the<br />
subsurface dynamics of a single-plate planet.<br />
High Frequency Radiometer Testbed<br />
Stephen Pomes<br />
Mentors: Pekka Kangaslahti and Douglas Dawson<br />
JPL has been a leader in the development of internally calibrated spaceborne microwave radiometers. These<br />
radiometers are used to measure wet tropospheric path delay produced by the changing index of refraction of the<br />
atmosphere resulting from the variable concentration of water. Though these radiometers have all operated at<br />
frequencies below 40 GHz, JPL is currently investigating the feasibility of the internal calibration approach for<br />
microwave radiometers operating in the 90-200GHz range, particularly at the frequencies of 90GHz, 130GHz, and<br />
166GHz for the Surface Water and Ocean Topology mission. The radiometers consist of several high-performance<br />
broadband low-noise amplifiers, isolators, filters, a detection circuit, and a data acquisition unit. Since parameters<br />
such as gain and noise temperature are variable with time, it is necessary to test various internal calibration<br />
sources and approaches in order to assess their stability inherently and over a range of temperatures as well as the<br />
overall system impacts of such components.<br />
163
ATHLETE: A Rotary Percussion Drill Tool Package for Rock or Ice Drilling<br />
Jay Qi<br />
Mentor: Matthew Frost<br />
The All-Terrain Hex-Limbed Extra-Terrestrial Explorer (ATHLETE) is a vehicle concept developed at the Jet<br />
Propulsion Laboratory based on the wheel-on-limb approach to mobility, allowing it to drive, walk, climb, and even<br />
manipulate its environment with its limbs. Through a standardized fixture and a clamp mechanism on each limb,<br />
ATHLETE is able to extract various tool attachments from a "tool belt" as necessary during exploration missions. A<br />
rotary percussion drill tool package concept was designed to integrate with ATHLETE's quick-disconnect tool clamp<br />
with the purpose of allowing ATHLETE to drill into solid rock or ice surfaces. This tool concept was developed to be<br />
simple, robust, and versatile, using a commercial rotary percussion drill mechanism powered by an easily<br />
changeable rechargeable battery and controlled wirelessly. The design also includes a mechanism for ATHLETE to<br />
change drill bits. With such a drill tool, ATHLETE will be able to drill into rock or ice during missions on the Moon,<br />
other planets, or asteroids, whether for anchoring, obtaining core samples, or other tasks.<br />
Reducing Color Rivalry in Imagery for Complementary Multiple Bandpass Filter Based Stereo Endoscopy<br />
Allen Ream<br />
Mentors: Harish Manohara and Sam Bae<br />
A pair of complementary multiple bandpass filters (CMBF) can be used to create spatially separated pupils in a<br />
traditional lens and imaging sensor system, allowing for the passive capture of stereo video. This method is<br />
especially useful for surgical endoscopy where smaller cameras are needed to provide ample room for manipulating<br />
tools while also granting improved visualizations of scene depth. The salient issue in this process is that, due to the<br />
complementary nature of the filters, the colors seen through each filter do not match each other, and also differ<br />
from colors as seen under a white illumination source. A color correction model was implemented that included<br />
optimized filter selection, such that the degree of necessary post-processing correction was minimized, and a<br />
chromatic adaptation transformation that attempted to fix the imaged colors tristimulus indices based on the<br />
principle of color constancy. Due to fabrication constraints, only dual bandpass filters were feasible. The theoretical<br />
average color error after correction between these filters was still above the fusion limit meaning that rivalry<br />
conditions are possible during viewing. This error can be minimized further by designing the filters for a subset of<br />
colors corresponding to specific working environments.<br />
A Cloud Imager in the Thermal Infrared<br />
Brian J. Redman<br />
Mentor: Sabino Piazzolla<br />
Information about cloud patterns is useful for climate science studies and Earth-space optical communications<br />
research. Thermal infrared sky imaging is a technique that records cloud patterns by measuring the heat radiation<br />
emitted by the clouds. This method is particularly well suited for continuous ground-based measurements of cloud<br />
cover statistics because it functions equally well during day and night. There is an interest in exploring the<br />
capabilities of low-cost systems capable of imaging the entire sky. A system using a metal dome to reflect the sky<br />
fits these criteria. In order for this system to be used algorithms have been developed to compensate for<br />
distortions from the dome, in particular uneven heating of the dome.<br />
Mars Sensor System Electrical Cable Management and Distributed Motor Control Computer Interface<br />
Robin Reil<br />
Mentors: Mohammad Mojarradi and Colin McKinney<br />
The success of JPL’s Next Generation Imaging Spectrometer (NGIS) in Earth remote sensing has inspired a followon<br />
project, the Mars Sensor System (MSS). One of JPL’s responsibilities in the MSS project involves updating the<br />
documentation from the previous JPL airborne imagers to provide all the information necessary for an outside<br />
customer to operate the instrument independently. As part of this documentation update, detailed electrical cabling<br />
diagrams were created to provide JPL technicians with clear and concise build instructions, and a database was<br />
produced to track the status of cables from order to build to delivery.<br />
Simultaneously, a distributed motor control system is being developed for potential use on the 2018 Mars rover<br />
mission. This system would significantly reduce the mass necessary for rover motor control, making more mass<br />
space available to other important spacecraft systems. The current stage of the project consists of a desktop<br />
computer talking to a single “cold box” unit containing the electronics to drive a motor. In order to test the<br />
electronics, a graphical user interface (GUI) was developed using MATLAB to allow a user to send simple<br />
commands to the cold box and display the responses received in a user-friendly format.<br />
164
Dew Point Monitoring System for Waveguide Arc Testing<br />
Ryan D. Rickard<br />
Mentors: Juan Ocampo and Michael Britcliffe<br />
The objective of this project is to create a system for monitoring moisture contamination levels within the<br />
waveguide of an X band transmitter that is part of the Deep Space Network (DSN). The introduction of moisture<br />
contaminants within a waveguide is believed to produce a rare phenomenon of RF (radio frequency) arcing that can<br />
potentially damage and even destroy transmitter components. By modifying a previously created monitoring<br />
system’s code and hardware components, a sensor was integrated with a single board computer capable of<br />
displaying dew point readings onto an LCD screen and Ethernet connection. Following the design and build of the<br />
monitoring system, the sensor was placed at the end of a custom waveguide and monitored injected moisture<br />
searching for a correlation between dew point (moisture) and RF arcing. With early detection of moisture in the<br />
waveguide using a dew point sensor, DSN ground station operators can prevent damage to transmitter components<br />
and schedule needed repairs and maintenance.<br />
Earth Science Data Processing in the Cloud<br />
Andres Riofrio<br />
Mentors: Khawaja S. Shams and Tom Soderstrom<br />
Cloud computing offers exciting new opportunities to many kinds of scientific missions, offering a cost-effective<br />
alternative to in-house cluster-based systems. Low budget airborne missions such as such as the Carbon in Arctic<br />
Reservoirs Vulnerability Experiment (CARVE) mission are typically deprived of the resources required for timely<br />
processing; furthermore, the algorithm is highly parallel and processing is bursty, making cloud computing a<br />
natural fit. Along with the Airborne Cloud Computing Environment (ACCE) task at the Jet Propulsion Laboratory<br />
(JPL), we built and developed software systems that enable the Level 2 Full Physics processor to run in Amazon's<br />
Elastic Compute Cloud (EC2) using Polyphony, a workflow orchestration framework developed at JPL that supports<br />
job execution in a cloud environment, as well as in a local environment, and any combination thereof. This software<br />
is applicable to other earth science missions, including OCO-2. We also developed cloud management techniques<br />
that will be useful in the adoption of cloud computing by missions at NASA.<br />
Mars Science Laboratory Flight Software Boot Robustness Testing<br />
Brian Roth<br />
Mentors: Danny Lam, Steve Scandore, and John Lai<br />
On the surface of Mars, the Mars Science Laboratory will boot up its flight computers every morning, having<br />
charged the batteries through the night. This boot process is complicated, critical, and affected by numerous<br />
hardware states that can be difficult to test. The hardware test beds do not facilitate testing a long duration of<br />
back-to-back unmanned automated tests, and although the software simulation has provided the necessary<br />
functionality and fidelity for this boot testing, there has not been support for the full flexibility necessary for this<br />
task. Therefore to perform this testing a framework has been built around the software simulation that supports<br />
running automated tests loading a variety of starting configurations for software and hardware states. This<br />
implementation has been tested against the nominal cases to validate the methodology, and support for<br />
configuring off-nominal cases is ongoing. The implication of this testing is that the introduction of input<br />
configurations that have yet proved difficult to test may reveal boot scenarios worth higher fidelity investigation,<br />
and in other cases increase confidence in the robustness of the flight software boot process.<br />
Correlations Between Technical Performance and Cost Growth: A Case Study Using Historical Cost Data<br />
Dean Rowley<br />
Mentors: Kevin Rice, Kevin Endo, Eric Kwan, and John Jack<br />
Flight Projects historically exceed the available budgetary funds allocated by NASA HQ. Project lessons-learned<br />
indicated that there are common reasons for cost growth where strong correlations existed between technical<br />
performance and cost/schedule performance. This group research project investigates historical data from the<br />
13 most recent JPL flight projects to determine trends and correlations between technical and programmatic<br />
parameters to predict cost growth. Each of the 13 projects is analyzed using 5 major data sets: technical<br />
complexity, subcontractor performance, risk trends, margin analysis and schedule. Each data set is analyzed to<br />
determine which correlations exist. My area of focus for this group project is to establish the cost per month and to<br />
set a baseline for the 5 data sets that acts as a common comparison between the analyses. Once all technical and<br />
cost data is collected regression analysis and modeling is performed to establish trends. It is expected that certain<br />
aspects of technical performance parameters will lend themselves to be more accurate cost predictors than others.<br />
Conclusions derived from this group research can be utilized to form a historical data base for future projects and<br />
compared against the current in-work flight projects to predict cost growth.<br />
165
Development of Interactive Tools for Quantitative Spectral Analysis<br />
Garreth Ruane<br />
Mentor: J. Brad Dalton<br />
Infrared reflectance spectroscopy is a powerful tool for understanding the chemistry of planetary surface materials.<br />
Detailed analyses of specific absorption features can provide information that is uniquely diagnostic of composition.<br />
Modern spectroscopy relies on software applications to determine spectral properties such as absorption center<br />
frequencies (wavelengths), absorption band depths, and spectral feature shape. Removal of the continuum, defined<br />
by anchor points to either side of an absorption feature, permits analysis of a given feature independent of the<br />
influence of other absorption features - some of which may arise from other materials within the scene.<br />
Abundances of materials within a scene may be estimated by the application of linear mixing algorithms, which<br />
treat the spectrum as arising from a linear combination of endmember spectra. Many approaches to linear<br />
unmixing use matrix inversion techniques that limit the number of endmembers that may be included in a spectral<br />
library. I have developed software tools to enable researches to apply continuum-removal and linear mixing<br />
methods to analyze spacecraft spectral observations using spectral libraries of arbitrary size, thus expanding<br />
current capabilities and allowing the description of complex planetary surfaces. I will illustrate these approaches<br />
with example data from the Galileo spacecraft observations of Europa.<br />
F6 Space System: A Model of the Implementation and Operation of a Cluster<br />
Tyler Ryan<br />
Mentor: Steven Cornford<br />
DARPA has funded a number of teams to develop its Fractionated Spacecraft vision. A team led by JPL has been<br />
tasked to develop a tool for the evaluation of the Business Case for this fractionated system architecture. The<br />
purpose of this evaluation is to understand under what conditions and constraints the fractionated architecture<br />
makes more sense from the cost/benefit perspective than the traditional monolithic paradigm. One key issue is to<br />
create a model complete enough to provide an accurate representation of the Business Case, yet executes fast<br />
enough to allow a thorough tradespace exploration. Such a model has been created in a discrete event simulator<br />
called Arena, representing both the implementation and operation phases of a cluster of N spacecraft. The<br />
implementation model includes the design, build, assembly, and test phases of major subsystems, their integration,<br />
and subsequent launch sequencing . The operations model includes gathering and managing power, and taking,<br />
cross-linking, and downlinking data. The model also accounts for various degrees of fractionation, as well as<br />
options for responding to stimuli, including launch vehicle failure, component failure, technology obsolescence and<br />
changes in mission objectives. The model, how it was developed, and the results of its execution will be discussed.<br />
Mars Atmospheric Modeling for the MAVEN mission<br />
Carly Sakumura<br />
Mentor: Allen Halsell<br />
Accurate predictions of the atmospheric density on Mars are crucial for the Mars Atmosphere and Volatile Evolution<br />
(MAVEN) mission. In order to accomplish its scientific objectives, this mission is targeting a density corridor at<br />
periapsis of 0.05 to 0.15 kg/km3 during normal science operations with five additional “deep dips” to densities of<br />
2.0 to 3.5 kg/km3. The current model of the Martian atmosphere is the Mars Global Reference Atmospheric Model<br />
(Mars-GRAM). While this program is accurate in predicting most trends and an estimate of the atmospheric density,<br />
a more accurate model is desired in order to better meet the strict density requirement imposed upon this mission.<br />
In order to accomplish this goal the scale factor between the predicted and measured atmospheric density is<br />
analyzed from the previous Mars Reconnaissance Orbiter, Odyssey, and Mars Global Surveyor missions. A<br />
statistical model of the variation as well as trends in the scale factor due to altitude, latitude, and seasonal effects<br />
is developed to both improve and better understand this estimate.<br />
Standardization of Spore Inactivation Method for PMA-PhyloChip Analysis<br />
Michael Schrader<br />
Mentor: Kasthuri Venkateswaran<br />
Development of a more sensitive method that will enumerate viable microorganisms from spacecraft and<br />
associated surfaces is warranted to facilitate future life-detection missions. Current culture-based and nucleic-acidbased<br />
polymerase chain reaction (PCR) methods fail to do so. The Biotechnology and Planetary Protection group at<br />
JPL is currently involved in the development of a new molecular method based on the use of propidium monoazide<br />
(PMA), in combination with DNA microarray (PhyloChip). PMA can only penetrate the membrane of dead cells and<br />
then intercalates into DNA, making it unavailable for PCR.<br />
Two bacterial spores will be used: Bacillus subtilis 168 and Bacillus pumilus SAFR-032. Spores were harvested by<br />
repeated centrifugation and washing in buffer solutions at 4 °C. The standardization of the spore inactivation<br />
method with no detrimental effect on DNA will then be implemented. Results indicate that heat treatment at 90 0 C<br />
for 90 min, completely inactivates (non-viable) spores without damaging DNA, confirmed by the Quantitative PCR<br />
166
(Q-PCR) method. After successful standardization of heat inactivation method, we will subject the spores to PMA<br />
treatment followed by DNA extraction. This method will help NASA set planetary protection standards and improve<br />
techniques for reducing spacecraft bioburden for future missions.<br />
Sampling Tool Concepts for Comet Surface Sample Return<br />
Rafael Send<br />
Mentors: Paul Backes and Christopher McQuin<br />
Three sampling tools for sampling of comet surfaces in a touch-and-go (TAG) mission architecture were developed<br />
and tested and compared with test results using a Brush Wheel Sampler (BWS) sampling tool. The BWS was<br />
developed previously and tested across a range of stimulants so it was used as a reference to compare the new<br />
tool concepts with. The three tools were a drive tube (DT), comet corer (CC), and comet auger (CA). The primary<br />
anticipated benefits of the DT, CC, and CA compared with the BWS were that they could acquire subsurface<br />
samples to 10cm depth and potentially enable a simpler sample acquisition, measurement, and transfer (SAMeT)<br />
capability compared to the BWS. The BWS has been shown to reliably acquire a large surface sample but it is of<br />
interest to determine whether an alternative tool could acquire a subsurface sample to at least 10cm depth with<br />
similar reliability and the same or reduced system complexity.<br />
The Effect of Glow Discharge on Thermal Vacuum Vapor Deposition<br />
Theodore J. Sharp<br />
Mentor: Andrew Hanna<br />
Glow discharge is a process used in thermal vacuum vapor deposition to increase adhesion by cleaning the surface<br />
being coated. The glow discharge is done before the coating while the vacuum chamber is pumping down to high<br />
vacuum. For the small chamber in building 248, the glow discharge is started as the chamber pressure goes below<br />
200 milliTorr. While still pumping down, the direct current transformer is slowly increased to a voltage setting of<br />
about 1.5 kilovolts. This transformer is connected to a high voltage wire that runs through the vacuum chamber.<br />
As the voltage increases the remaining air in the vacuum chamber is ionized separating the air into ions and<br />
electrons. With the high energy electrons released, the chamber and the surface to be coated is “scrubbed” as the<br />
electrons bombard the surface. This bombardment cleans the surface on a molecular level allowing for greater<br />
adhesion between the surface and the aluminum vapor. This is continued for approximately five (5) minutes at<br />
which time the transformer is shut down and the glow discharge ends.<br />
After continuing to pump down and reaching a high vacuum state, the tungsten filaments inside of the chamber are<br />
heated up melting the pure aluminum wire hung on the filaments. The molten aluminum then wicks up into the<br />
filaments and is evaporated which begins coating all surfaces inside of the chamber with a layer of aluminum vapor<br />
between 3-7 microns thick. To determine whether the coat is thick enough to be a good mirror, a simple process is<br />
used utilizing a transparent light bulb mounted just behind a sample slide of glass. Once coating begins, the<br />
operator can watch the filament in the light bulb slowly grow dimmer as the glass is covered with the highly<br />
reflective aluminum vapor. Once the filament can no longer be seen the coating process is stopped and the<br />
chamber is backfilled and opened.<br />
Once the samples of glass have been coated in aluminum, a “tape test” will be performed to check the adhesion of<br />
the aluminum to the glass. The glass plate is pulled from the vacuum chamber and placed on a flat surface to<br />
prevent it from fracturing. A piece of scotch tape is placed on the surface of the aluminum coat and pressed firmly<br />
down and is then sharply removed removing any aluminum that was not entirely adhered to the surface of the<br />
glass. By inspecting the tape and seeing how much aluminum was removed, the operator can determine whether<br />
the coat is well adhered or whether they must strip the aluminum and start over.<br />
A major goal for this research is to determine an optimal voltage to set the glow discharge to obtain the best<br />
adhesion. By finding an optimal voltage, a future operator may use that information to better use the glow<br />
discharge, to clean the surface being prepared for coating. To determine the ideal voltage a range of voltages will<br />
be used to prepare sample slides of glass which will then be coated using a standardized procedure with roughly<br />
the same amount of aluminum. They will then be removed and operators will perform a tape test removing any<br />
aluminum that had not adhered to the glass. The samples will then be marked with the voltage used for the glow<br />
discharge and they will be sent for reflectivity testing. By considering a blank slide of glass as a control to compare<br />
to the reflectivity of the sample should increase as more aluminum adheres correctly to the glass slide. By doing<br />
this, it will be possible to qualitatively determine what the optimum voltage setting is for the glow discharge.<br />
Analysis of Data in Accordance With Space Flight Mission Environmental Requirements<br />
Monica Shei<br />
Mentor: Kin F. Man<br />
The Environmental Assurance Program ensures that all flight hardware is compatible with the environments that<br />
will be encountered during its mission. It outlines the design, test, risk control, and documentation standards for<br />
the spacecraft and ensures its reliability in external and self-induced environments that it may experience. The<br />
167
Environment Requirements Document (ERD) provides the environmental design and test requirements for the<br />
project’s flight system, subsystems, assemblies, and instruments. It defines the system level environments and<br />
derived environments for the flight system from the ground, through launch and cruise, and science operations.<br />
It also specifies the method of verification for the environments, whether it is to be verified by analysis or by test.<br />
This summer’s objective was to complete the Environmental Assurance Program Summary Report for both the Juno<br />
Project and Mars Science Laboratory (MSL) Project. Environments-related Waivers and Problem Failure Reports<br />
(PFRs) were compiled along with statistics on the number of Environmental Test Information Summaries (ETIS’s)<br />
for each assembly/subsystem. These and other environmental assurance information will be presented in the final<br />
presentation.<br />
Constraining Fatty Acid Vesicle Formation in Simulated Deep-Ocean Conditions<br />
Aomawa Shields<br />
Mentors: Steve Vance, William Abbey, and Richard Kidd<br />
Modern cells use lipid membranes to selectively determine which molecules enter and exit the cell. These<br />
membranes are composed primarily of phospholipids, each consisting of two fatty acids and a phosphate group<br />
attached to a glycerol molecule. One theory of life's emergence on Earth holds that initial metabolic and replication<br />
reactions took place in protobiotic cell membranes. These may have been composed of n- monocarboxylic acids<br />
(fatty acids) alone, rather than more complex phospholipid structures. Our experiments in the lab confirm that<br />
aqueous decanoic acid self-assembles into bilayer vesicles at standard temperature and pressure (25°C, 1 atm) in<br />
the pH range between 7 and 8.5, as indicated by stepwise increases in fluorescence induced by ultraviolet<br />
excitation of dye conjugated to decanoic acid solution. Life’s chemical evolution in primordial oceans on Earth or<br />
elsewhere could have occurred over a range of temperatures, however. Further investigation will constrain the pH<br />
range for vesicle formation as a function of temperature. We also plan to explore the role of pressure, which is<br />
similarly important to whether life may have formed in deep-ocean environments on Earth, or on icy worlds in the<br />
solar system.<br />
Correlations Between Technical Performance and Cost Growth: A Case Study Using Margin Analysis<br />
Tyler Shin<br />
Mentors: Kevin Rice, Kevin Endo, Eric Kwan, and John Jack<br />
Flight Projects historically exceed the available budgetary funds allocated by NASA HQ. Project lessons-learned<br />
suggest that there are common reasons for cost growth where strong correlations existed between technical<br />
performance and cost/schedule performance. This research project investigates historical data from the 13 most<br />
recent JPL flight projects to determine trends and correlations between technical and programmatic parameters to<br />
predict cost growth. Each of the 13 projects is analyzed using 5 major data sets: technical complexity,<br />
subcontractor performance, risk trends, margin analysis and schedule slips. Each data set is analyzed to determine<br />
which correlations exist. My specific assignment for this group project is to use margin analysis to more accurately<br />
predict cost growth. After gathering the data, such as mass and power growth of each spacecraft, a regression<br />
analysis will be performed with cost growth data to establish trends. It is expected that certain aspects of technical<br />
performance parameters will lend themselves to be more accurate cost predictors than others. Conclusions derived<br />
from this group research can be utilized to form a historical data base for future projects and compared against the<br />
current in-work flight projects to predict cost growth.<br />
Development of an Electronics Package for In Situ Thermal Infrared Imaging: A Comparison of MPU<br />
and FPGA Based Platforms<br />
Steven Shoen<br />
Mentor: William Johnson<br />
Recent advances in infrared (IR) detector array technologies have made possible the development of IR imaging<br />
systems with unprecedented quality. One challenge to any portable imager is always the instrument’s ability to<br />
process the high-speed, high-density data generated by the detector accurately and within the instrument’s power<br />
budget. With fast serial I/O speeds (~250Mhz) and internal logic (600Mhz), Field Programmable Gate Arrays<br />
(FPGAs) are shown to have strong potential for handling and processing data at bus speeds as much as two orders<br />
of magnitude greater than Microprocessor Unit (MPU) based platforms. The FPGA architecture also supports<br />
multiple clocking domains, allowing a single device to perform data acquisition, hardware control and<br />
synchronization tasks. While Microprocessor Units (MPU) remain the current standard for remote systems, FPGAs<br />
are showing promising potential in this application. To this end, a thorough investigation was made into the<br />
performance and limitations of MPU and FPGA based systems. A Virtex-5 FXT FPGA was specified and set-up on an<br />
evaluation platform in the lab. Code was generated and executed on the device to evaluate the device’s expected<br />
field performance. Two instruments, the JPL Near Nulling Radiometer (JPLNNR) and the Hyperspectral Infrared<br />
Imager (HyspIRI) are used as bases for comparison.<br />
168
Determining Oxygen-18 Isotopic Fractionation in the Ferrous Sulfate, Melanterite<br />
Danielle Z. Shulaker<br />
Mentors: Max Coleman and Issaku Kohl<br />
Studying regions on Earth that are analogous to Mars serve as case studies for studying astrobiology and planetary<br />
surface rock formation processes. Río Tinto, Spain is very rich in iron sulfates, and has an environment that is<br />
possibly very similar to the former environment on Mars. Certain bacteria play significant roles in accelerating<br />
pyrite dissolution rates, which contribute to the formation of ferrous sulfates, such as melanterite. During mineral<br />
crystallization from aqueous solution, there are systematic isotopic differences between the solution and solid<br />
phases (fractionation). Quantifying fractionation enables isotopic analysis to be used to trace the original isotopic<br />
signature. Thus, determining the oxygen-18 isotopic fractionation of melanterite has many potential benefits for<br />
understanding processes that formed Mars and its past environment. To determine the fractionation of melanterite,<br />
acidic aqueous solutions saturated with dissolved hydrated ferrous sulfate were evaporated at 25°C and 40°C.<br />
During evaporation, the aqueous solution and crystallized melanterite that form were sampled. Aqueous and solid<br />
samples are converted to BaSO4 for oxygen-18 isotopic composition analysis. Further work will include determining<br />
the isotopic fractionation of other minerals common to Earth and Mars, and testing conditions, such as pressure or<br />
humidity, and comparing the different isotopic fractionations determined.<br />
Development of CCSDS DCT to Support Spacecraft Dynamic Events<br />
Anahita F. Sidhwa<br />
Mentor: Kar-Ming Cheung<br />
The Consultative Committee for Space Data Systems (CCSDS) Design Control Table (DCT) is a versatile link<br />
calculation tool to analyze different kinds of radio frequency links. It started out as an Excel-based program, and is<br />
now being evolved into a Mathematica-based link analysis tool. The Mathematica platform offers a rich set of<br />
advanced analysis capabilities, and can be easily extended to a web-based architecture. Last year the CCSDS DCT’s<br />
for the uplink, downlink, two-way, and ranging models were developed as well as the corresponding input and<br />
output interfaces. Another significant accomplishment is the integration of the NAIF SPICE library into the<br />
Mathematica computation platform.<br />
This summer, the attitude models for the CCSDS DCT are being developed along with a Smart GUI that<br />
incorporates the different components needed to model the spacecraft attitude. The attitude heuristics model the<br />
spacecraft dynamic events of earth-pointing, sun-pointing, spinning, conning, and lander orientation. The attitude<br />
heuristics will also include models for “Safemode” scenarios that depend on the mission phase (cruising, mapping,<br />
etc.). Based on the knowledge of antenna pointing relative to the spacecraft frame, the cone and clock angles of<br />
the antenna pattern with respect to the target of interest can be computed, and this provides the antenna pointing<br />
loss of the link equation. Thus modeling of spacecraft attitude behavior enables one to provide accurate link<br />
prediction during spacecraft dynamic events.<br />
Mars Lava Inflation Identification Through Visual Surveys<br />
Laura Sigelmann<br />
Mentors: Suzanne Smrekar and Serina Diniega<br />
Different styles of Martian lava eruption create distinctive surface formations, which provide insight into the past<br />
and present volcanic behavior of Mars. These formations develop under specific conditions, and their distribution<br />
can provide valuable information about eruption dynamics on Mars. For example, inflationary features are created<br />
when a thickened and rigid crust develops on top of a still-fluid lava interior. Pressure exerted by the moving fluid<br />
core uplifts the crust, and a positive topographic feature with distinct cleft systems develops. It is has been<br />
proposed that the distribution patterns of inflation features correlate with areas of higher and lower pressure within<br />
the magma core. Through a survey of images taken by the High Resolution Imaging Science Experiment (HiRISE)<br />
camera on the Mars Reconnaissance Orbiter (MRO), the location, orientation, and size of various inflation features<br />
such as tumuli, lava rises, and lava-rise pits were identified. A set of criteria for identifying such features was also<br />
established to aid in continued research. The analysis of the resulting data set will be used to test and develop<br />
models for eruption and flow dynamics of inflated flows.<br />
Applied Field Magnet Development and Magnetic Probe Testing in a Magnetoplasmadynamic Thruster<br />
Vritika Singh<br />
Mentors: James Polk and Robert Moeller<br />
Space propulsion began with powerful chemical rockets propelling the first missions into orbit around the Earth and<br />
Moon. Since those first missions, views of space and space exploration have evolved to include new possibilities,<br />
and space propulsion has evolved to include electric propulsion (EP). EP is a type of space propulsion that enables<br />
more efficient space travel and a greater capacity for scientific instruments on-board spacecraft. One high power,<br />
high specific impulse EP device is a magnetoplasmadynamic (MPD) thruster, which uses electric and magnetic<br />
forces to create thrust. MPD thruster performance generally increases with increasing operating current, but at<br />
higher current levels, performance and lifetime actually decreases due to an instability condition called “onset.” To<br />
169
mitigate onset, an applied field magnet system was developed and validated through bench testing and computer<br />
modeling. Early work consisted of mapping the self generated magnetic field inside the MPD thruster at varying<br />
current levels using a Bdot probe. The applied field magnet system will be implemented with the MPD thruster, and<br />
the magnetic field structure will be mapped with the influence of the magnets. This magnetic field map will be<br />
compared to the map produced without the applied magnets to study their influence on MPD thruster discharge<br />
characteristics.<br />
Hardware Testing for the Optical Payload for Lasercomm Science<br />
Amanda Slagle<br />
Mentors: Anthony Mannucci and Parker Fagrelius<br />
Optical communication is the use of visible light rather than radio-frequencies to send a data signal. The technology<br />
is valuable because it allows much higher data rates than radio communications can achieve. The Optical Payload<br />
for Lasercomm Science (OPALS) project will establish an optical link between the International Space Station (ISS)<br />
and Earth, which will allow for the testing of vital algorithms for controlling optical communication equipment.<br />
When OPALS is on the ISS, it will lock onto a laser beacon transmitted by a telescope on the ground. The payload<br />
will then transmit a video modulated onto a laser beam back to earth.<br />
The payload consists of an optical head containing the laser transmitter and a camera for detecting the ground<br />
telescope’s beacon, a gimbal to control the pointing of the optical head, and a sealed electronics box. Equipment<br />
for the gimbal, laser, and avionics subsystems was tested. For example, an inspection of the flight power board<br />
was conducted to ensure that all power and ground signals were isolated, that polarized components were correctly<br />
oriented, and that all components were intact and securely soldered.<br />
In Situ Analysis of Organics With a Portable Mass Spectrometer<br />
Santosh Soparawalla<br />
Mentor: Luther Beegle<br />
In order to find life on other planets, we start by examining life processes we understand on the earth. Signs of life<br />
can exist in the form of small organic molecules such as amino acids. Mass spectrometry (MS) is an analytical<br />
method applied to gain chemical information on a wide variety of chemical compounds. The project demonstrates a<br />
portable MS system, the Mini 10.5, for astrobiological applications including in situ hydrocarbon analysis and<br />
sediments analysis using an automated sample processing system (ASPS). The ASPS is a sampling device<br />
developed at JPL that uses a cation exchange resin to extract amino acids from sediments. Since the amino acid<br />
solution eluted from the ASPS is at pH 12 positive ion formation required for MS analysis is inhibited. The pH was<br />
reduced without forming excessive salt by adding a small amount of acetic acid to form an ammonium acetate<br />
buffer suitable for electrospray ionization. The second part of the study tested a sorbent trap system developed to<br />
preconcentrate hydrocarbon compounds from the air. Air sampling was done at sites in the Mojave desert and near<br />
roads to evaluate the role of automobile exhaust as a carbon source for lichens in the arid environment.<br />
Analysis and Design of OTM-24 for the Mars Reconnaissance Orbiter<br />
Rebekah Frances Sosland<br />
Mentors: Martin D Johnston, Reid Thomas, and Jessica Williams<br />
The Mars Reconnaissance Orbiter (MRO) Project is a major component of NASA’s Mars Exploration Program (MEP).<br />
Launched in August 2005, the MRO spacecraft is now in the Extend Mission phase and is on-station in its low<br />
altitude, Primary Science Orbit (PSO) at Mars. The MRO orbit altitude is maintained using Orbit Trim Maneuvers<br />
(OTMs) which are implemented at orbit apses once every eight weeks and are designed to maintain the MRO<br />
groundtrack walk. This paper describes the MRO maneuver design and implementation process and details a new<br />
OTM strategy to perform OTMs every sixteen weeks. Orbital mechanics theory and Navigation software was used to<br />
analyze maneuvers with various Density Scale Factors (DSF) in order to maintain the orbit altitude. Using delivered<br />
Orbit Determination (OD) solutions, OTM-24 was modeled and performed on the spacecraft using the OTM<br />
sequence block. The Maneuver was designed to increase orbit semi-major axis in order to target a GTW error<br />
staying within the GTW constraints. After analyzing OTM-24, the data showed a very small execution error<br />
resulting in a very successful OTM and confirming that the OTM-25 may potentially be postponed and performed in<br />
sixteen weeks.<br />
Cryogenic Dielectric Measurements of Liquid Hydrocarbons<br />
Charles H. Starr<br />
Mentor: Martin B. Barmatz<br />
Accurate interpretation of Cassini RADAR measurements of Titan’s liquid hydrocarbon lakes requires knowledge of<br />
the dielectric properties of hydrocarbon mixtures at 13.8 GHz and at cryogenic surface temperatures (~ 90 K). In<br />
particular, a value for the complex dielectric constants of hydrocarbon mixtures would allow determination of lake<br />
depths. An improved cavity perturbation technique using exact field solutions to Maxwell’s equations has been<br />
developed to study liquid samples contained in a quartz holder within a cylindrical cavity resonator. This technique<br />
170
equires the measurement of the cavity resonant frequency and quality factor with and without a sample inserted<br />
in the cavity. A cylindrical microwave cavity has been constructed that will initially be used to determine the<br />
complex dielectric constant of liquid methane, liquid ethane, and their mixtures. I am currently developing a userfriendly<br />
Mathematica program to obtain the complex dielectric constant from the experimental measurements. In<br />
the future, we intend to extend this new cavity perturbation technique by developing a similar Mathematica<br />
program to obtain the permeability values of magnetic samples.<br />
The Effects of High Vacuum on the Longevity of Tungsten Filaments<br />
Aaron Steinkraus<br />
Mentor: Andrew Hanna<br />
Thin film vapor deposition is a vital process in increasing the reflectivity of a material. During deposition the coating<br />
material is vaporized and deposited on to the substrate. One way to vaporize the coating metal is to hang loops of<br />
the coating metal on tungsten filaments and then run current through the filament until the metal wicks into the<br />
coils and vaporizes. The thermal cycling and oxidation of the tungsten causes the filaments to fail after only three<br />
to four coatings. In theory, by conducting coating at lower vacuum, the filaments would experience less oxidation<br />
and thereby extend the life of the filaments. To test this theory, the coating process will be run at two levels of<br />
vacuum and then cycling the filaments on and off until failure. The filaments that are operated at lower pressures<br />
are predicted to have a longer life than those at a higher pressure.<br />
Assessment of Bacterial Spores in Solid Materials<br />
Kamil B. Stelmach<br />
Mentor: Wayne Schubert<br />
The United States is part of the 1967 United Nations treaty called “Treaty on Principles Governing the Activities of<br />
States in the Exploration and Use of Outer Space, Including the Moon and Other Bodies” and is required to<br />
participate in planetary exploration. Planetary protection measures are required avoid accidental transfer of Earth<br />
life to other solar system bodies. To that end, studies are being conducted to enumerate the amount of microbial<br />
bioburden found in common spacecraft material. A wide variety of spacecraft materials were studied: various<br />
epoxies, Kapton tape, electrical components, and Lucite. A literature search was conducted into the possible<br />
dissolution of these materials. Based on the search, the use of the Hansen Solubility Parameters is recommended<br />
to be used for picking out future solvents. Currently, select polymeric materials are being artificially embedded with<br />
a known spore count of Bacillus atrophaeus. The materials are then to be cryogenically ground to a fine powder.<br />
Spore counts will be conducted from the ground material to enumerate the amount of viable spores extracted.<br />
Furthermore, quantitative PCR (qPCR) will be used to enumerate the amount of total bacterial DNA found within<br />
spacecraft material.<br />
Sampling Tool Concepts for Comet Surface Sample Return<br />
David C. Sternberg<br />
Mentors: Paul Backes and Christopher McQuin<br />
Three sampling tools for sampling of comet surfaces in a touch-and-go (TAG) mission architecture were developed<br />
and tested and compared with test results using a Brush Wheel Sampler (BWS) sampling tool. The BWS was<br />
developed previously and tested across a range of stimulants so it was used as a reference to compare the new<br />
tool concepts with. The three tools were a drive tube (DT), comet corer (CC), and comet auger (CA). The primary<br />
anticipated benefits of the DT, CC, and CA compared with the BWS were that they could acquire subsurface<br />
samples to 10cm depth and potentially enable a simpler sample acquisition, measurement, and transfer (SAMeT)<br />
capability compared to the BWS. The BWS has been shown to reliably acquire a large surface sample but it is of<br />
interest to determine whether an alternative tool could acquire a subsurface sample to at least 10cm depth with<br />
similar reliability and the same or reduced system complexity.<br />
Improving Data Accessibility<br />
Ben-han Sung<br />
Mentors: Paul Wolgast and Bach Bui<br />
The majority of the data from the Deep Space Network (DSN) is kept in a firewall-protected database and accessed<br />
by users through a web portal. This project aimed to improve the limited data accessibility by building more tools,<br />
such as mobile applications. The DSN Dashboard app, created for iOS devices last year, underwent improvement<br />
as the first part of this project. The app helps visualize tracking information such as antenna azimuth and elevation,<br />
and bridges the traditional gap between mission management and real-time data. New features include a view that<br />
shows the DSN’s released tracking schedule. Secondly, overall data accessibility was improved by the development<br />
of an Openfire module to facilitate communication between NASA employees. JPL and NASA use the Openfire XMPP<br />
server for collaborative work and sharing information across facilities. Previously, users had to make project or<br />
location specific accounts in order to use the service. The newly developed module authenticates users against<br />
171
multiple LDAP servers so they can log in if they have an account on any of those servers. Removing the<br />
inconvenience of multiple accounts will help open lines of communication between mission management and<br />
operators at remote DSN complexes.<br />
Design of a Robotic Ankle Joint for a Microspine-Based Robot<br />
Nitish Thatte<br />
Mentor: Aaron Parness<br />
Successful robotic exploration of near-earth asteroids necessitates a method of securely anchoring to the surface of<br />
these bodies without gravitational assistance. Microspine gripper arrays that can grasp rock faces are a potential<br />
solution to this problem. A key component of a future microspine-based rover will be the ankle used to attach each<br />
microspine gripper to the robot. The ankle’s purpose is twofold: 1) to allow the gripper to conform to the rock so a<br />
higher percentage of microspines attach to the surface, and 2) to neutralize torques that may dislodge the grippers<br />
from the wall. So far, two mock-up ankles that demonstrate the key mechanical features have been constructed.<br />
The manufacture of a final, actuated prototype is underway. Parts were developed using computer aided design<br />
and manufactured using a variety of methods including selective laser sintering, CNC milling, and traditional<br />
manual machining techniques. Upon completion of the final prototype, the gripper and ankle system will be tested<br />
to demonstrate robotic engagement and disengagement of the gripper and to determine the improvement in load<br />
bearing ability. The immediate application of this project is to outfit the Lemur IIb robot so it can climb and hang<br />
from rock walls.<br />
Planning Surface Interaction Operations for a Near-Earth Asteroid Rover<br />
Eric Timmons<br />
Mentors: David Mittman and Jay Torres<br />
With NASA shifting its focus to landing a person on a near-Earth object by 2025, the All-Terrain Hex-Limbed Extra-<br />
Terrestrial Explorer (ATHLETE) team at JPL is examining the feasibility of using ATHLETE to study an asteroid. Due<br />
to the low magnitude, non-radial gravity field of an asteroid, innovative techniques must be employed to ensure<br />
ATHLETE’s success. The anchoring system to ensure the robot stays on the surface of the asteroid is one<br />
innovation. It is integral to the mission that the anchors can supply the necessary forces to hold ATHLETE to the<br />
surface while the robot is performing its tasks. The maximum reaction force the anchors can apply is limited by<br />
local regolith properties such as cohesion. Current estimates of asteroid regolith properties have high uncertainty.<br />
Additionally, these properties vary with location and time due to anchor interactions. Using current sensors<br />
available on ATHLETE, this research will develop a system that will provide current state information for the<br />
anchors. State information includes in situ estimates of local regolith properties. These estimates will be used to<br />
determine if an action is safe and the uncertainty in that decision. The ultimate goal of this project is to present<br />
this information to the operator in a clear, useful manner. This software will be validated in simulation and is<br />
planned to be validated using ATHLETE and regolith simulant in the future.<br />
Plume Dynamics and Its Implications for Gravity and Topography for Venus<br />
Henry Tom<br />
Mentor: Suzanne Smrekar<br />
In this study we numerically model Venus with a 3D spherical code, OEDIPUS, to try to both simulate surface<br />
expressions and track its thermal evolution. From OEDIPUS's output, MATLAB codes are used for analysis and<br />
various plots. In particular, plots of plumes allow determination of the number, buoyancy, and heat flux of each<br />
plume. Further, plumes are correlated with topographical features such as hotspots and coronae. However, the<br />
features of plumes differ from one run to another due to different input parameters into OEDIPUS. A balance<br />
between the input parameters, Rayleigh number and internal heating, has yet to be determined but recent<br />
developments suggest constraints on these values. In particular, we look at the variability of mantle plumes, such<br />
as their buoyancy, temperature difference, and radius, as we change these parameters. Moreover, Rayleigh<br />
number is dependent on both viscosity and temperature difference across the convecting domain. Decreasing<br />
bottom viscosity from 10 21 to 10 20 produces more hot plumes; however, internal heating reduces their buoyancies,<br />
therefore creating insignificant topographical features. Further, comparing topographies from analysis and from<br />
gravity and topography data will narrow our constraint values even further, allowing us to produce the<br />
approximately nine hotspots we observe on Venus.<br />
Development of the NEXUS Hardware Testbed<br />
Carson A. Umsted<br />
Mentor: Yutao He<br />
The goal of NEXUS is to develop a unified, scalable, and highly capable avionics interconnect for future spacecraft.<br />
To meet this goal, Serial RapidIO has been chosen as the baseline protocol to be integrated with the use of ISAAC<br />
technology. My task is to develop a NEXUS hardware testbed composed of a switch, and a number of FPGA-based<br />
NEXUS nodes using ISAAC iBoards. Each node includes a Bus Interface Unit (BIU) IP core based on the Serial<br />
RapidIO protocol and various types of backend such as ADCs and high-speed camera. The Xilinx Serial RapidIO IP<br />
172
core is used as the basic implementation for different configurations including, loop-back, point-to-point, and<br />
switch-based. The complete hardware testbed will be integrated and verified with real-time high-speed data<br />
sources.<br />
Fade and Revival of the South Equatorial Belt of Jupiter in the Mid-Infrared Wavelength Range<br />
Steffi B. Valkov<br />
Mentor: Glenn S. Orton<br />
Sporadic fluctuations in the color and reflectivity of Jupiter's South Equatorial Belt (SEB) have been observed over<br />
the last century. In the latest of these, from 2009 to 2010, the SEB experienced an episode of turning from a<br />
typical brown color to white (known as a “fade”); the previous episode of this type was in the early 1990's. Toward<br />
the end of 2010, the SEB began a return to a brown color (known as a “revival'). We used data collected from<br />
ground-based telescopes in the mid-infrared wavelength range in order to find pressure-temperature profiles,<br />
fractions of para hydrogen, and aerosols in the atmosphere of Jupiter. Our ultimate goal is to find what causes a<br />
fade and revival of the SEB and what changes in the atmosphere may occur before and after a fade. The<br />
instruments that were used in this research were MIRSI and VISIR, both sensitive to wavelengths between<br />
4.90 and 24.80 μm. After initial reduction of these mid-infrared thermal images, they were calibrated in absolute<br />
radiance using Cassini infrared data as a standard. New, <strong>2011</strong> Jupiter data were acquired and calibrated in late July.<br />
We used a FORTRAN-based program developed at the University of Oxford to determine properties of the<br />
atmosphere from the images of Jupiter's radiance collected at these telescopes. Initial results show that prior to<br />
the fade, temperatures in the troposphere began to drop, as opacity of aerosols increased. After the full fade took<br />
place, the para-ortho hydrogen fraction in the SEB increased by December 2010 and aerosols decreased.<br />
MSL OPGS Product Generation Pipeline’s Performance Assessment<br />
Sergeh Vartanian<br />
Mentors: Maher Hanna, Adrian Tinio, and Costin Radulescu<br />
The MSL rover "Curiosity" will be launching in <strong>2011</strong> and landing on Mars in August 2012. The Operation Product<br />
Generation Subsystem, OPGS, is responsible for generating critical products such as images captured through the<br />
rover’s cameras which are used by the rover planners in determining tactical and strategic rover activities. For my<br />
research project, I was responsible for measuring, assessing, and documenting the performance of the MSL OPGS<br />
product generation pipeline with respect to its operational timeline requirement of thirty minutes. My tasks included<br />
the measurement of the overall performance throughout the pipeline from raw data product to EDR/RDR generated<br />
files delivered to the operational team, as well as of the individual pipelines within the workflow. In addition, I have<br />
conducted a profound study and evaluation of the OPGS pipeline as part of the performance assessment. In order<br />
to meet the thirty minute requirement, I spent many hours running the pipeline in three different environments<br />
that involved virtual machines, physical standalone machines, and the Amazon Cloud with varies software and<br />
hardware configurations that I contemplated while analyzing and studying the behavior of the system.<br />
Analysis of Ultra High Resolution Sea Surface Temperature Data<br />
Grant T. Wagner<br />
Mentors: Edward Armstrong and Jorge Vazquez<br />
The Group for High Resolution Sea Surface Temperature (GHRSST) is an international project that distributes<br />
satellite derived sea surface temperatures (SST) data from multiple platforms and sensors. The goal of the project<br />
is to distribute these SSTs for operational uses such as ocean model assimilation and decision support applications,<br />
as well as support fundamental SST research and climate studies. Examples of near real time applications include<br />
hurricane and fisheries studies and numerical weather forecasting. The JPL group has produced a new 1 km daily<br />
global SST product, the Multiscale Ultrahigh Resolution (MUR), that blends SST data from 3 distinct NASA<br />
radiometers in both the microwave and infrared spectral bands, but requires further validation and accuracy<br />
assessment, especially in coastal regions.<br />
The goal of this project is to analyze, validate and characterize the newly developed GHRSST 1km (high-resolution)<br />
SST dataset for use in coastal studies with the focus on major upwelling regions of the world’s oceans, including<br />
the California Coast. This work will include comparisons of the new dataset with in situ fixed and drifting buoy SST<br />
measurements and as well as comparisons to a Low Resolution version of the product (25 km resolution). It also<br />
includes exploring applications of the new dataset in fishery and marine biology through evaluation of submesoscale<br />
oceanic features such as temperature fronts, gradients, and mesoscale eddies.<br />
173
Taxonomic Classification of NEOWISE Objects: Correlating Spectroscopy, Albedo, and Physical<br />
Properties<br />
Jessica A. Watkins<br />
Mentor: Amy Mainzer<br />
The influx of new data to the NEOWISE system offers the opportunity to physically characterize 157,000 newly<br />
discovered small planetary bodies (specifically Near-Earth Objects (NEOs) and Main Belt Asteroids (MBAs)). The<br />
goal is to understand the relationship between visible/near-IR (VNIR) spectral features and physical properties such<br />
as mineral composition, size, density, regolith, and measured visible and near-IR albedos. NEOWISE utilizes multiwavelength<br />
infrared spectra to more accurately determine diameters and compute albedos, allowing us to compare<br />
various taxonomic classes more accurately than previous taxonomic schemes derived from VNIR spectroscopy and<br />
photometry. We classified the thermally fitted spectral data into known, corresponding classification schemes<br />
(Tholen, Bus-Binzel, Bus-DeMeo, and Carvano) based on color to test their accuracy with the larger, less biased,<br />
and more accurate NEOWISE data set. There is a two-fold bias in these classification schemes, involving visible<br />
light surveys that favor high albedo objects, and taxonomic classifications with the NEOWISE survey will mitigate<br />
this bias. The processes that may affect the correlation between spectroscopy and albedo and cause peculiar<br />
absorption features in spectral signatures are examined (space weathering, grain size, hydration features, etc.) to<br />
explore precisely what information a given spectral signature or albedo can actually provide about an asteroid’s<br />
surface.<br />
Electrochemical Analysis of Fuel Cell Catalysts<br />
Ian R. Wessen<br />
Mentor: Thomas I. Valdez<br />
Polymer electrolyte membrane (PEM) fuel cells exhibit many exciting properties, not the least of which is their<br />
startling efficiency. For instance, internal combustion engines can theoretically convert energy with 39% efficiency<br />
(often averaging around 20%). Fuel cells do not burn their fuels, instead turning chemical potential directly into<br />
work. In theory, a fuel cell may do so at 100% efficiency. It is not uncommon for some of our PEM hydrogenoxygen<br />
fuel cells to perform upwards of 80% efficiency at low current densities. If we could find a fuel cell catalyst<br />
that can maintain 80% efficiency at higher current densities, we would have an extremely attractive and viable<br />
option for power generation, particularly with regards to future space application. We have analyzed various<br />
electrochemical properties of Pt-Ni electroless plated catalysts using cyclic voltammetry, potentiodynamic scans,<br />
and electrochemical impedance spectroscopy to deduce their respective efficiencies and effectiveness as oxygen<br />
reduction catalysts. The discovery of substantial improvements in fuel cell efficiency will have direct impact in<br />
tomorrow's generation of fuel cells.<br />
Ontological Modeling for Integrated Spacecraft Analysis<br />
Erica L. Wicks<br />
Mentors: Yu-Wen Tung and Pierre Maldague<br />
Current spacecraft work as a cooperative group of a number of subsystems. Each of these requires modeling<br />
software for development, testing, and prediction. It is the goal of me team to create an overarching software<br />
architecture called the Integrated Spacecraft Analysis (ISCA) to aid in deploying the discrete subsystems’ models.<br />
Such a plan has been attempted in the past, and has failed due to the excessive scope of the project. Our goal in<br />
this version of ISCA is to use new resources to reduce the scope of the project, including using ontological models<br />
to help link the internal interfaces of subsystems’ models with the ISCA architecture.<br />
I have created an ontology of functions specific to the modeling system of the navigation system of a spacecraft<br />
and linked this with an ontology modeling the attitude control subsystem. The resulting ontologies not only link, at<br />
an architectural level, two spacecraft subsystems, but also are also web-viewable and can act as documentation<br />
standards.<br />
These ontologies are proof of the concept that ontological modeling can aid in the integration necessary for ISCA to<br />
work, and can act as the prototype for future ISCA ontologies.<br />
A Prototyping Effort for the Integrated Spacecraft Analysis System<br />
Raymond Wong<br />
Mentors: Yu-Wen Tung and Pierre Maldague<br />
Computer modeling and simulation has recently become an essential technique for predicting and validating<br />
spacecraft performance. However, most computer models only examine spacecraft subsystems, and the<br />
independent nature of the models creates integration problems, which lowers the possibilities of simulating a<br />
spacecraft as an integrated unit despite a desire for this type of analysis. A new project called Integrated<br />
Spacecraft Analysis was proposed to serve as a framework for an integrated simulation environment. The project is<br />
still in its infancy, but a software prototype would help future developers assess design issues. The prototype<br />
explores a service oriented design paradigm that theoretically allows programs written in different languages to<br />
174
communicate with one another. It includes creating a uniform interface to the SPICE libraries such that different inhouse<br />
tools like APGEN or SEQGEN can exchange information with it without much change. Service orientation may<br />
result in a slower system as compared to a single application, and more research needs to be done on the different<br />
available technologies, but a service oriented approach could increase long term maintainability and extensibility.<br />
Automated Purgatoid Identification<br />
Steven B. Wood<br />
Mentor: Paolo Bellutta<br />
Purgatoids are small (< 1 m high, < 10 wide) eolian bedforms commonly found scattered across the Meridiani<br />
Planum region of Mars. First encountered by the Mars Exploration Rover Opportunity during sol 446 of that mission,<br />
the rover became stuck in the poorly consolidated soil of a purgatoid while attempting to traverse the feature.<br />
Critically, further driving was nearly impossible as most forward or backward commanded movement was<br />
translated into digging the wheels deeper into the loose, unstable bedform surface. Although Opportunity was<br />
ultimately freed on sol 484, from that day forward identification and avoidance of purgatoids became an important<br />
feature of rover path planning. Purgatoid identification is currently done manually using high-resolution HiRISE<br />
images from the Mars Reconnaissance Orbiter. This project, in support of the Mars Terrain Classifier, created and<br />
implemented an algorithm that uses the same HiRISE images as data and automatically identifies the location of<br />
potential purgatoids in any input image. The algorithm relies on regional shape descriptors and moment invariants,<br />
along with normalized cross-correlation with a selected group of kernels representative of purgatoids, for accurate<br />
identification.<br />
Visual Tracking With a Small Quadrotor UAV<br />
Jonathan Wu<br />
Mentor: Roland Brockers<br />
Unmanned micro air vehicles (UMAVs) are capable of low cost reconnaissance, search and rescue. To facilitate<br />
human operator control, these systems need autonomous capabilities to execute basic navigation tasks. For tasks<br />
like flying to a goal or tracking a moving target from a vantage point, robust image based tracking methods are<br />
needed that can be executed on the vehicle in real time.<br />
This project explores various methods and approaches in template based target tracking using imagery from an on<br />
board camera with extensions applicable to UMAVs. The methods explored survey the real-time stable performance<br />
of different multi-parameter warping models (affine, perspective, and quadratic) and their corresponding<br />
parameter estimators to track targets in image sequences over time.<br />
Mapping Hazards for Future Mars Rover Landing Sites<br />
Kimberly E. Yauk<br />
Mentors: Matthew Golombek and Fred Calef<br />
To help decide where to land future rover missions on Mars, hazards on the Martian surface were mapped at<br />
prospective landing sites. Hazards include rock fields, steep scarps, and inescapable landscapes such as steep<br />
craters and fields of intersecting aeolian bedforms. We investigated seven prospective landing sites identified for a<br />
possible rover mission in 2018. We downloaded HiRISE images and stereo anaglyphs into ArcMap and geospatially<br />
referenced the images onto lower resolution CTX images for each landing site. At resolutions on the order of 1:600,<br />
the terrain was assigned categories ranging from flat and safe (one) to non-traversable and dangerous (four)<br />
based on the apparent slope of the 3D anaglyph. Slopes greater than thirty degrees are considered dangerous to<br />
the rover and are evidenced by shadows cast from the rocks. The hazard maps, once completed in detail, will help<br />
determine the most appropriate landing site, could help fine tune the landing ellipse and give engineers a map on<br />
which to estimate landing site safety and traversability for the rover.<br />
Deep Space Network Portal for Android Tablet<br />
Jonathan Yee<br />
Mentor: Paul Wolgast<br />
The Deep Space Network (DSN) consists of three deep space communications facilities around the world that are<br />
constantly tracking information throughout the day. The DSN Portal was created as a website to make the data<br />
from the DSN accessible on any computer. Computers and laptops are not the most mobile devices to use to view<br />
data on the DSN Portal. Last year, a version of the DSN Portal was created for the iPad and has been successful.<br />
My work this summer involved developing an optimized version of the DSN Portal to run on an Android tablet. An<br />
Android DSN Portal app allows DSN operators to use a mobile interface to the DSN portal from an Android tablet.<br />
This gives the user live tracking information from each of the three facilities using an intuitive interface and touch<br />
screen gestures. As mobile devices such as tablets become more ubiquitous, they will allow DSN operators the<br />
convenience to view live data from the DSN Portal remotely and easily.<br />
175
SIAD Supersonic Flight Dynamics Test Engineering Support<br />
Regina Zmuidzinas<br />
Mentor: Brant Cook<br />
The Low Density Aerodynamic Decelerator (LDSD) research and technology development is currently working to<br />
develop two Mars atmospheric entry technologies: a 6-meter Supersonic Inflatable Aerodynamic Decelerator<br />
(SIAD) and a 30-meter Supersonic Ringsail Parachute (SSRS). However, before being implemented in a Mars entry<br />
environment, these technologies must be qualified in a series of tests on earth to verify that the SIAD and SSRS<br />
will function as expected. The Supersonic Flight Dynamics Test (SFDT) is one of three LDSD tests to confirm these<br />
results, using a vehicle that can obtain Mach 4 at an earth altitude of 180,000 feet to simulate a Mars entry<br />
environment. The SFDT vehicle has many design constraints to perform at these conditions, including mass and<br />
power consumption. In order to assist the SFDT vehicle design team, the harnessing for the instruments on the<br />
vehicle will be determined and routed within the vehicle in order to better assess harness mass. The vehicle<br />
electronic box will be preliminarily designed. The entire mass and power of the vehicle will also be closely tracked<br />
during the design phase so that the team can closely track how design changes affect system mass margin.<br />
176
<strong>Abstract</strong>s<br />
<strong>Student</strong>–<strong>Faculty</strong> <strong>Programs</strong><br />
2<br />
0<br />
1<br />
1