Student-Faculty Programs 2011 Abstract Book - Summer ...

Student Student-Faculty Student Faculty Programs 

2011 Abstract Book

STUDENT-FACULTY PROGRAMS 

2011 Abstract Book 

This document contains the abstracts of the research projects 

conducted by students in all programs coordinated by Caltech’s 

Student-Faculty Programs Office for the summer of 2011. 

Table of Contents 

Summer Undergraduate Research Fellowships (SURF) 1 

MURF Undergraduate Research Fellowships 113 

Amgen Scholars Program 121 

Laser Interferometer Gravitational-Wave Observatory (LIGO) 129 

NASA/JPL Programs 139 

� Planetary Geology and Geophysics 

Undergraduate Research Program (PGGURP) 

� The National Space Grant College 

and Fellowship Program (Space Grant) 

� Undergraduate Student Research Program (USRP) 

� JPL Student Internship Program (JPLSIP)

SURF 

SUMMER UNDERGRADUATE 

RESEARCH FELLOWSHIPS 

S 

U 

R 

F

In-Plane Thin Film Lattices: Thermally Stable Mirrors 

Deeksha Agrawal 

Mentors: Chiara Daraio and Eleftherios Gdoutos 

In devices that experience large temperature changes such as high speed aerospace vehicles, substantial thermal 

stresses develop when high temperature components are connected to lower temperature structures. This leads to 

failure by yielding, fracture or fatigue, as well as the formation of gaps that require sealing and extreme forces on 

attachments to other structures. To suppress such failures, we fabricate and study expansion thin film bimaterial 

lattices with tunable thermal expansion. The lattice configuration is composed of Aluminium and Titanium which 

have widely different individual coefficients of thermal expansion (CTEs) and intervening spaces. According to 

theoretical predictions the higher CTE internal Aluminium expands more rapidly and pushes outward against the 

lower CTE exterior lattice of Titanium. This causes the latter to rotate, thereby counteracting the lengthwise 

elongation of the material. We fabricate our lattice structure on a silicon wafer, coat the wafer with photoresist, 

expose with UV and create the pattern, coat with Aluminium by E-beam evaporation, lift off the photoresist and 

then iterate the procedure with Titanium. We then etch the sacrificial gold layer to separate the Al-Ti lattice. We 

use techniques like Finite Element Analysis using ABAQUS 6.9-2and X-ray diffraction to test the various properties 

of this lattice. 

Thermal Analysis of the Solar Decathlon House 

Sara Ahmed 

Mentor: Melany Hunt 

A group of Caltech and SCI-Arc students have teamed together to build an energy efficient, innovative and most 

importantly a Net-Zero home. To achieve this goal, we are using solar energy for generating electricity and day 

lighting as well as reducing energy consumption throughout the home. The thermal load is a major factor of energy 

use, including the heat emitted by lights, sunlight through windows, heat transfer from the environment to the 

house, domestic hot water needs and heat absorbed by PV (photovoltaic) panels. 

During my SURF I have been working on the thermal analysis and building of our solar house, CHIP. We have 

integrated our HVAC (heating, ventilation, air conditioning) unit and hot water system which are responsible for all 

the heating and cooling needs of the house. We have accomplished this by utilizing the waste heat of the outdoor 

unit of the HVAC unit to preheat the domestic hot water. For optimizing the operation of the integrated heat 

exchange system, we have done comprehensive temperature control and dehumidifying tests in the house. 

Through iterative tests and analysis, we have significantly improved its performance. I have also researched on 

improving the energy efficiency of the house using heat sinks to the PV Solar panels and light fixtures. 

C-H Bond Activation Using Organometallic Iridium Metal Complexes 

Tonia Ahmed 

Mentors: John Bercaw and Ian Tonks 

A large part of petroleum used in today’s society consists mainly of methane. However, few methods have been 

found to efficiently and selectively convert methane to other valuable products. Moreover, because of the cost and 

difficulty of transporting gases, oil companies generally resort to burning methane gas instead of utilizing it. As an 

alternative, methane could be converted to liquid methanol through selective C-H bond activation. C-H activation 

was studied in this project using a variety of iridium hydroxyl complexes and observed using methods of nuclear 

magnetic resonance (NMR) spectroscopy and X-ray crystallography. From the data provided by these methods, the 

C-H activation of a variety of substrates was confirmed. To determine the mechanism of the reactions, several 

kinetics experiments were performed also using NMR spectroscopy. The rate determining step was determined to 

be substrate coordination based on kinetic and Eyring analysis. 

The Impact of High Frequency Trading on Correlations 

Emanuelle Alm 

Mentor: Benjamin Gillen 

From the onset of the financial crisis through late 2010, soaring correlations were observed in the financial 

markets. Since 2010, correlation levels have gone down to some extent. However, these high correlations have 

persisted longer than expected given the current macroeconomic environment and incipient recovery. The 

expansion of high-frequency trading (HFT) activity is often mentioned as one possible cause of this persistence, 

likely due to significantly changes in the underlying market microstructure. HFT has experienced strong growth 

over the past several years and makes up a large amount of the trading volume in the Unites States. This paper 

empirically identifies and evaluates causal implications unique to the hypothesis that HFT is driving increased 

correlations. We present a mathematical model that illustrates the mechanism by which HFT would drive increase 

correlations and empirically test the model. Our analysis uses data on 23 Exchange Traded Funds to compare the 

ratio between short-term and long-term correlation for a period with low incidence of High Frequency Trading and a 

1

period from latter years when High Frequency Trading is considered to make up a notable amount of the trading 

volume. While most of the preliminary data preparation and analysis is complete, much formal statistical 

estimation and testing remains to be performed over the following month. 

Mechanical Properties of Nanocrystalline Ni Nanopillars 

Eli E. Alster 

Mentors: Julia Greer and Jie Lian 

Materials sized on the order of nanometers possess mechanical properties far divergent from their bulk analogs. In 

order to quantify the effects of microstructural dimension and external scale on materials’ mechanical properties, 

both grain size and sample size were varied across the nanometer range and plotted against yield strength to help 

form a three dimensional plot. Fabrication of the nanopillars was accomplished through electrodeposition of Ni in 

template microarrays and mechanical properties tested through uniaxial compressions by a Hysitron nanoindenter 

with a flat punch tip. Original and post-deformation images were taken with either the Nova-200 or Nova-600 

scanning electron microscope (SEM). Grain size was determined through the lift-out procedure and transmission 

electron microscopy (TEM). Molecular dynamics simulation work was also undertaken using the LAMMPS software 

package to investigate the nucleation and movement of dislocations, which are difficult to capture from 

experimental observations. These efforts were designed to gain insight into the not yet understood mechanisms 

which drive deformation at nanoscale. 

Optimization of Dreiding Hydrogen Bond Parameters 

Anthony Alvarez 

Mentor: William A. Goddard III 

In chemistry, accurately modeling small systems on the order of a few dozen atoms can be done using quantum 

mechanics. For large systems, such as proteins, these simulations take an incredibly long time. To be able to model 

these systems quickly and accurately a method called force fields is used rather than quantum calculations to 

model these structures. However, these force fields require optimized parameters to be able to accurately 

reproduce the atomic level interactions being described by quantum mechanics. One force field, Dreiding, has 

parameters optimized for high dielectrics, like water, but does poorly in low dielectric environments, such as a 

protein. Hydrogen bond parameters are particularly affected by the dielectric change. Since hydrogen bonds play a 

large role in the folding of proteins those parameters are particularly important. This project is aimed at obtaining 

high level quantum mechanics data on a set of molecules that represent the interactions found in a protein and use 

that data to fit the Dreiding hydrogen bond parameters. Once the parameters are optimized, Dreiding will be able 

to more accurately describe proteins and their interactions with ligands. 

Hyperbolic Versus Euclidean Embedding of Real-World Networks 

Sadaf Amouzegar 

Mentors: Babak Hassibi and Elizabeth Bodine-Baron 

Recent research has suggested that social networks can be described in terms of a hidden metric space, which 

often can be used to improve applications requiring navigation, such as routing or search. Different algorithms 

exist to embed networks in either Euclidean or hyperbolic space – in this work we evaluate two representative 

algorithms to determine the optimal embedding for various types of networks. We use Landmark Multidimensional 

Scaling as representative of Euclidean embedding. This method measures the length of the shortest path between 

any two nodes in the network and assigns a coordinate to every node in Euclidean space based on this distance. 

However, since Euclidean space has a geometric growth rate and most real-world networks exhibit low diameters, 

implying exponential growth, we also employ Crovella’s “online greedy graph algorithm” as representative of 

hyperbolic embedding. This method uses a spanning tree algorithm to calculate the appropriate coordinates for 

each node in hyperbolic space. To determine which type of hidden space is optimal for different types of networks, 

we evaluate these algorithms on both real world and online social network data. 

Supersonic Relative Velocities of Baryonic Fluids and Dark Matter 

Nikhil Anand 

Mentor: Christopher Hirata 

At the epoch of recombination, the sound speed of the Universe dropped precipitously from a relativistic limit 

($\sim c/\sqrt{3}$) to that of the thermal RMS velocities of hydrogen atoms ($\sim 10^{-5} c$). Because the 

dark matter does not suffer Thomson scattering, the relative velocities between the baryonic fluids and the dark 

matter becomes supersonic (coherent over several Mpc, comoving), effectively raising the Jeans mass necessary to 

form the earliest structures. This delays the growth and population of these structures by an amount that depends 

on the magnitude of the relative velocity. We examine the effect of supersonic relative velocities, particularly in 

regards to the galaxy power spectrum and its consequences for future galaxy surveys and BAO measurements. 

2

Contribution of Organosulfur Compounds to Atmospheric Aerosol in the United States 

Avin Andrade 

Mentors: John Seinfeld and ManNin Chan 

Atmospheric aerosol—tiny particles suspended in the air—forms from a variety of sources such as vegetation, 

transportation, industry, agriculture, and urban land use. It not only poses serious problems to human health but 

also has an effect on weather and climate. By understanding its properties and composition, we can determine the 

most effective means of minimizing its production. Organosulfates (R’OSO4) and nitrooxyorganosulfates 

(O2NOR’OSO4) compose a new and important class of organosulfur and organic compounds that has recently been 

detected in laboratory generated aerosol and ambient aerosol. Its importance can be estimated by looking at the 

difference in concentration of total sulfur and inorganic sulfur. Using data from the the Environmental Protection 

Agency's Interagency Monitoring of Protected Visual Environments (IMPROVE) database it has been shown that 

organosulfur makes up a significant contribution to aerosol composition across the United States. The 

concentrations are on the order of 0.1 ug/m 3 or around 10 percent of the total organic carbon concentration. These 

numbers vary significantly by season and year with the highest concentrations occurring during summer and the 

years 2005 and 2006. 

The Firefighter Problem on Planar Graphs 

Zhi Yuan Ang 

Mentor: Niranjan Balachandran 

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 

vertices that are not on fire and these vertices are being protected permanently. The fire then spreads from a burnt 

vertex to all its unprotected neighbours and the process repeats itself. Denote snk(v) as the maximum number of 

vertices that can be saved if the fire starts out at v. The k-surviving rate ρk(G) of G is defined to be 

k (v) / n 2 , which is the average proportion of saved vertices. Let C be the class of all planar graphs. In this 

report, we investigate the infimum of ρ3(G) taken over all G C. In other words, we want to know if 3 firemen 

suffice to save a good proportion of the graph. The ultimate goal is to solve or refute the conjecture that 2 firemen 

would suffice. 

Condensation of Water on Superhydrophobic Carbon Nanotube Arrays 

Patrick Arran 

Mentors: Mory Gharib and Indrat Adrianos 

It was thought that superhydrophobic vertically aligned carbon nanotube arrays would, uniquely among known 

superhydrophobic materials, not be subject to condensation of water. This hypothesis was made on the basis of the 

surface feature scale being orders of magnitude smaller than that for previously investigated materials. The study 

did not attempt to characterise the mechanisms of condensation, merely to observe the macro-level results. CNT 

array samples that were treated to be either superhydrophobic or hydrophilic along with traditional control samples 

such as polymers, glass and metals, were cooled with liquid nitrogen and exposed to humid air. The subsequent 

phenomena were recorded as video and results compared. The details of the methodology have evolved over the 

course of the study. Preliminary results indicate much smaller amounts of condensed water and much faster 

evaporation times for the nanotubes than for the control materials. It is hoped that more precisely controlled 

experiments will provide a more quantifiable, consistent result and an understanding of the effect of varying 

relative humidity. It is also hoped to more accurately determine the effect of the CNT treatments, which have so 

far made negligible differences. 

Data Reduction and Analysis Software for Robo-AO LGS Adaptive Optics System 

Ankit Arya 

Mentors: Christoph Baranec and Reed Riddle 

I developed and coded a library of reduction and analysis software for the telemetry system of a robotic adaptive 

optics telescope system (Robo-AO). In addition, I created a status page for the system, which helps the engineers 

and astronomers to monitor the overall functioning of the system instruments. The status page represents live 

information about the telescope instruments in the form of live charts and 2D heatmap images, making it a very 

useful tool for debugging and for representing errors quickly. First, the analysis software finds the minimum, 

maximum and median values for the deformable mirror actuator positions as well as the focus shape of the mirror. 

It also calculates the root mean square and leaky average of the input data. The results of calculations are written 

to data files, which are read by the status page asynchronously to display it live. The system code is in C++, so I 

used the same to integrate my library within the software. The status page makes use of scripting languages like 

JavaScript, jQuery, JSON, AJAX and other plotting API to represent information. All API used to design the page are 

open source to help the distribution of the software publically at a later stage. 

3

Analysis of Interpolation Schemes in the Quasicontinuum Method 

Hyunji Bae 

Mentors: Michael Ortiz and Malena I. Español 

The quasicontinuum method is used to effectively and accurately approximate the mechanics of solid materials at 

the atomistic level. The interpolation scheme is an important part of the method. The purpose of this research is to 

investigate the effect of the interpolation schemes in the quasicontinuum method. Specifically, we compare 

different interpolation schemes and the effect it has on approximating energies and displacement fields. We 

present numerical examples in the one-dimensional case. 

High Performance DSP on FPGAs Using Python 

Amit Bansod 

Mentors: John Carpenter and Glenn Jones 

The CASPER tools provide a design flow to rapidly develop complex digital signal processing (DSP) systems for the 

radio astronomy instrumentation using Field Programmable Gate Arrays (FPGAs). CASPER libraries provide high 

performance DSP functionality for current radio astronomy applications but require further optimizations to 

efficiently implement future wide bandwidth instrumentation like the CARMA (A combined array of 23 antennas 

with a maximum bandwidth of 8 GHz) correlator. In this project, it is attempted to develop a better toolflow using 

a python library called MyHDL which can use the existing library of optimized functions and an architectural 

paradigm that works well for radio astronomy applications. To test the functionality of MyHDL, a Biplex pipelined 

Fast Fourier Transform (FFT) function block was designed using MyHDL. Performance of the designed block was 

then compared with the CASPER FFT block. The results showed a better performance by the MyHDL block in terms 

of the maximum operating clock frequency and a similar resource usage as that of CASPER design. By making a 

user friendly design toolflow, MyHDL can be a promising tool for future radio astronomy instrumentation. 

Clamping Losses in Nanomechanical Resonators 

Aniruddha Bapat 

Mentors: Michael Roukes and Luis G. Villanueva 

The field of nanotechnology is an active area of research today. One important application of nanotechnology is the 

field of Nanoelectromechanical systems, or NEMS. The focus of this project is to study a particular loss mechanism, 

the clamping loss mechanism, in NEMS resonators by simulation using the software COMSOL and to try to minimize 

this loss by varying the clamping geometry. The simulation was used to calculate the ratio of strain energies in the 

resonator and in the clamps, for the desired eigenmodes. This parameter is proportional to the clamping-based 

Quality factor Q of the system. Different resonator geometries, such as square, disk and ring bulk and rectangular 

flexural resonators are analyzed versus changes in dimensions and clamp geometry. 

Analysis of a Crater Detection Algorithm Applied to the Moon 

Allison Haley Barnes 

Mentors: Yang Cheng, Matthew Golombek, and Fred Calef 

This project aims to analyze and optimize the result of an algorithm that fully automates crater detection on the 

Moon. The algorithm uses images taken by two Narrow Angle Cameras (NACs) on the Lunar Reconnaissance 

Orbiter (LRO) to detect a large number of craters (~200,000 craters per image). The Moon has the potential to aid 

further space exploration, reveal information about Earth, and provide resources. Impact cratering produces the 

majority of observed surface morphologies, reveals the age and composition of surfaces, serves as a mechanism 

for deposition and removal of material, and provides information about the solar system. The objective of this 

project is to develop an automatic evaluation procedure for the algorithm’s data product and apply this procedure 

to all of the NAC images of the Apollo 15 landing site. Accomplishing this objective is significant, as validating an 

automated procedure that detects craters will save researchers the arduous effort of manually measuring craters 

while also providing insight into the relationship between craters and their surrounding terrain. 

Synthesis of an Asymmetric Bimetallic Complex for Carbon Dioxide Reduction 

Marissa Barrientos 

Mentor: Jonas Peters 

The goal is to synthesize a symmetric bimetallic complex that can electrocatalytically reduce carbon dioxide. The 

Peters group has envisioned an asymmetric heptadentate, monoanionic dinucleating ligand based on a μ-phenolate 

scaffold. A soft site is created using a tris-chelating diphosphino amine “arm” which may be able to stabilize a lowvalent, 

nucleophilic metal center that can potentially bind to the carbon atom of carbon dioxide. A dipyridal amine 

“arm” serves as a hard site supporting a Lewis acidic metal center that can facilitate the binding of carbon dioxide 

through coordination of an oxygen atom. The ligand precusor HL py, Cl was synthesized from a known synthetic 

procedure within the group. After attaching the BH3 protected diphosphino amine “arm,” several deprotection 

routes were explored. After achieving a reaction mixture of HL py,PNP and organic byproducts, we were still able to 

metallate the compound. We believe we have synthesized the desired [Ni2(HL py,PNP )(OAc)2][BPh4]. 

4

Cryogenic Noise Measurements of Amplifiers for the Q/U Imaging ExperimenT (QUIET) Phase II 

Nathaniel Baskin 

Mentors: Anthony Readhead and Kieran Cleary 

The principle objective of the QUIET Phase II project is to detect the B-mode polarization of the Cosmic Microwave 

Background, a characteristic signature of universal inflation. Detecting these B-modes requires the development of 

lower noise amplifiers. Caltech and JPL are currently collaborating on a program to improve the fundamental 

performance of coherent amplifiers. The goal of this SURF project is to perform a full characterization on packaged 

35 nm amplifiers developed for this program. In particular, this involves measuring the DC bias characteristics and 

noise temperatures of these amplifiers. By doing so, we can provide feedback to the designers and the fabrication 

process; this will facilitate the optimization of the amplifiers’ cryogenic performance. 

Iron Glyoximes as Catalysts in Artificial Photosynthesis Systems 

Geoffrey Beck 

Mentor: Harry Gray 

Developing a viable artificial photosynthesis system that could generate fuel from sunlight and water remains one 

of the main challenges of 21 st century chemistry. While many catalysts based on different metals have been 

proposed, the use of iron for the catalytic center could allow for more economically viable devices to be developed. 

This project used density functional calculations to investigate the viability of iron glyoximes for the catalytic 

production of molecular hydrogen and the activation of carbon dioxide. Properties of intermediate structures for 

several potential reaction mechanisms were determined such that the feasibility of these pathways could be 

established and the utility of this and other iron based catalytic systems could be ascertained. 

Spectra Simulation for ULXs NGC 1313 X-1 and NGC 1313 X-2 

Juliette Becker 

Mentors: Fiona Harrison and Kristin Kruse Madsen 

ULXs (ultra-luminous X-ray sources) are accreting point-sources with unexpectedly high rates of luminosity. Not 

much is known about ULXs and their internal mechanisms. NGC 1313 X-1 and NGC 1313 X-2 are two such ULXs 

and thus both sources of interest as targets for observation by the NuSTAR telescope (a high energy x-ray 

telescope to be launched in February of 2012). Whether or not they are worth observing depends on if the spectral 

interference from nearby point sources is too high and masks the spectra of the ULXs. A NuSIM simulation will 

determine whether NGC 1313 X-1 and NGC 1313 X-2 are too obscured for an observation by NuSTAR to be useful. 

Feynman Graphs and the Field With One Element 

Dori Bejleri 

Mentor: Matilde Marcolli 

In quantum field theory, Feynman graphs parameterize the perturbative expansion of the Green’s function. 

To each Feynman graph G, there is an associated integral U(G) that gives a term in this expansion. Let 

be the Kerchoff polynomial, where the sum runs over all spanning trees T of G and each te is an indeterminate 

corresponding to the edge e. This polynomial is homogenous and it defines a projective variety XG. It is well known 

that U(G) can be can be written as an integral of an algebraic differential form on the projective complement of XG. 

Often U(G) is a mixed zeta value of a mixed motive suggesting XG is a mixed motive. This is not true in general, 

but considering when this holds motivates an interesting problem to determine when XG corresponds to a scheme 

over the field with one element. There are many different but related approaches to geometry over the field with 

one element. In this paper we investigate necessary and sufficient conditions on the graph G for XG to be defined 

over the field with one element under several of these approaches. Additionally, we explore what XG being defined 

over the field with one element and G satisfying these conditions implies for the physical interpretation of the 

Feynman graph. Further implications of this research lead to deeper studies and understanding of the link between 

the physical and algebro-geometric interpretations of Feynman diagrams. 

5

Bound States of Fourth Generation Quarks and Antiquarks 

Tanvir Ahamed Bhuyain 

Mentor: Mark B. Wise 

Recently it has been pointed out that fourth generation quarks and antiquarks can form long lived bound states 

that do not decay to a lower generation, if they have very small mixing with the lower generations. In this paper 

we further investigate some properties of such possible bound states. We focus on getting better estimates for the 

spectrum of these states using variational techniques. We study the effects of relativistic and perturbative 

corrections to the Higgs couplings in all spin, color and flavor channels and several trial wave functions. 

Identification of Endogenous Inflammatory Cells in the Mammalian Hypothalamus 

Evan R. Biggs 

Mentors: Daniel Marks and Ellen Rothenberg 

Cachexia, characterized by involuntary weight loss and loss of muscle mass, is a comorbidity commonly observed 

with numerous chronic inflammatory diseases. Cachexia has been repeatedly linked to chronic inflammation, and 

central inflammation of the hypothalamus has been repeatedly implicated as a critical component in its pathology. 

Accordingly, characterization of the molecular mechanisms underlying inflammation in the hypothalamus 

represents an essential step towards comprehending and treating this condition. It is well-established that 

peripheral immunologic challenge with lipopolysaccharide produces a cachexia-like phenotype in rodents, and 

hypothalamic expression of IL-1β has been implicated as an integral part of this pathophysiology. However, the 

molecular and neuroanatomic mechanisms underlying this induction of IL-1β signaling remain largely undescribed. 

In this study, we determine the cell-type responsible for the observed IL-1β signaling in the arcuate nucleus of the 

hypothalamus in response to peripheral immunologic challenge. Double-label in situ hybridization was utilized to 

observe colocalization of IL-1β mRNA with either an astrocyte or microglial marker in order to determine the 

generative cell-type. Results will be discussed as data become available. 

Deep UV Microbial Detection On Weathered Basalts 

Alexa Bilek 

Mentors: Luther Beegle, Rohit Bhartia, and William Abbey 

The Raman laser system is a sophisticated machine used to detect mineral biology on the surface of rock samples. 

My partner and I, through this summer internship program, have been working with this laser for the past 4 weeks 

in order to familiarize ourselves with the common graph spectra associated with the minerals found within Basalts 

and Gabbros. Developing a good understanding of this technique will ultimately provide a stable background of 

knowledge, which will be helpful in the field when we collect live samples from the Mojave Desert in two weeks. 

Since organics are not fond of hot surfaces, it will be important for us to scan the entire rock, including the 

shadowed areas, throughout the day and night in order to come up with a large range of data. We will then be able 

to hopefully make a correlation between anthropogenic sources (pollution containing CO2 and other organic 

compounds released from semi-trucks and cars) and the sustainability of microbiotic life (microbes without 

chlorophylls vs. microbes using photosynthesis). 

Insight Into the Growth of Garnet Porphyroblasts Through Isotopic Analysis of Iron 

Suzanne Birner 

Mentors: Nicolas Dauphas and Paul Asimow 

Large chemically zoned garnet porphyroblasts are commonly found in high-temperature/pressure metamorphic 

systems. While this concentric zonation-- specifically of Mg 2+ and Fe 2+ -- is well-documented, the processes 

governing crystal growth that lead to such gradients are less well understood. Recently, it has been shown that an 

isotopic analysis can help constrain the timing and conditions of crystal growth in igneous and metamorphic rocks. 

The growth of garnet porphyroblasts may be limited by grain-boundary diffusion in the country rock. The extent to 

which Fe and Mg isotopes will be fractionated during such growth will depend on the growth rate, thus providing a 

geospeedometer of garnet growth in metamorphic rocks. In this study, a large garnet from the Zillertal region of 

the Austrian Alps is analyzed compositionally using a scanning electron microscope. The resulting chemical 

zonation data are plotted as a contour map overlying a back-scattered electron image to show the concentric 

nature of the gradient. An isotopic analysis of iron, carried out using a Neptune inductively coupled plasma mass 

spectrometer, then produces a profile of the isotopic fractionation with respect to radius. These two results are 

combined to gain unique insight into the chemical and diffusive processes that govern crystal growth in 

metamorphic rocks. 

6

Improving Automated Segmentation for Statistical Analysis of Bacterial Filaments 

John Blackwood 

Mentors: Grant Jensen and Songye Chen 

The Jensen Laboratory at Caltech has proven the value of using electron cryotomography (ECT) for gaining key 

insights into cellular structures and cytoskeletal support systems of bacteria. While significant discoveries have 

already been made, there is still a need for substantial advancements in processing and analysis of ECT data sets. 

Very low signal-to-noise ratios resulting from limitations of the ECT experimental setup create challenges in noise 

reduction and automated segmentation that are unique to the field. The Jensen lab has recently developed a 

program to automatically segment bacterial filaments in three-dimensional reconstructed tomograms that come 

from ECT data collection and processing. Segmentation programs such as this are generally computationally 

expensive and have long execution times. This project explores roughly segmenting out the cell membrane in 

bacterial tomograms with a fast algorithm, so that more computationally intensive segmentation programs search 

in a reduced volume. Volume reduction can be as great as 70 percent, significantly reducing execution time and 

the necessary computational resources. In addition, preliminary results of the Jensen lab’s filament segmentation 

program were analyzed. The lab’s long term goal is to efficiently search for filaments in a large database of 

tomograms for quantitative, statistical results. 

Computational Study of Solution Accuracy in Finite Volume Methods Using OpenFOAM 

Daniel E. Blado 

Mentor: Dan Meiron 

OpenFOAM is an open source C++ library primarily used solving specific problems in continuum mechanics via the 

finite volume method. Prior to the full implementation and integration of OpenFOAM into computational fluid 

dynamics research, however, the accuracy of the library’s algorithms must be acceptable. The accuracy of 

OpenFOAM’s solver methods was investigated via a convergence study on a driven cavity problem, in which a 

square cavity of incompressible fluid has its top lid moving at a constant velocity. Using OpenFOAM’s calculated 

velocity data for this problem, velocities were extrapolated at the boundary. The spatial derivatives for each 

velocity component and pressure at the boundary were computed as well, which are needed to solve for important 

quantities such as fluid stresses. So far the solutions for a Reynolds number of 10 have been worked out with mesh 

resolutions 20x20x1 and 40x40x1, and such numerical data is being compared with expected results. Future work 

entails analyzing boundary solutions for both higher mesh resolutions and higher Reynolds numbers. Should the 

convergence rate of error in OpenFOAM be adequate, we can later utilize OpenFOAM for the study of separated 

flows over airfoils. 

Gene Expression Levels as Markers of Irradiation Damage to the Brain 

Michelle E. Bobrow 

Mentor: Robert J. Brown and Marianne Bronner 

There are currently no available means of treating or preventing the long-term effects of therapeutic brain 

irradiation, and the molecular mechanisms of such damage are not well characterized. This study determined the 

effects of radiation on gene expression levels in the brain, in order to identify markers of radiation injury. Slices of 

brain tissue containing the hippocampus were harvested from irradiated and non-irradiated mice for extraction of 

RNA. Reverse transcription followed by real-time PCR at several post-irradiation time points was used to analyze 

the changes in expression of various inflammation-related genes. A database of genes known to be expressed in 

the brain was created and annotated to allow for future real-time PCR analysis of genes involved in neurological 

disease, apoptosis, vascular formation and damage, hypoxia, oxidative stress, radiation response, and other 

processes relevant to radiation injury. 

Using Riboswitches to Understand Stem Cell Maintenance 

Stephanie Bohaczuk 

Mentors: Elliot Meyerowitz and Paul Tarr 

Synthetic biology aims to engineer biological systems that elicit preprogrammed outputs. An application is 

riboswitches, which are RNA molecules that control translation of an mRNA. A riboswitch consists of a ligand 

binding aptamer, which is coupled to a hammerhead ribozyme that mediates mRNA degradation. Binding of the 

ligand induces a conformational change in the riboswitch, inactivating the ribozyme and allowing translation. We 

will use riboswitches to study how dynamic changes in the CLAVATA/WUSCHEL (CLV/WUS) genetic regulatory 

network and auxin-signaling pathway regulate stem cell differentiation and specification in the shoot apical 

meristem (SAM). Utilizing live imaging, we will follow stem cell proliferation and differentiation over several days 

after induction of a tetracycline or theophylline riboswitch to recover CLV or WUS expression in clv3 and wus 

mutants, respectively. The role of auxin in plant organ initiation in the SAM has been difficult to study due to the 

lack of a direct reporter. We will implement a riboswitch to act as a direct reporter of auxin localization. We use the 

SELEX protocol to develop de novo apatamers responsive to auxin, which will be coupled to a fluorescent protein. 

Preliminary results reveal similar motifs in the aptamer candidates, which require improvements in their affinities 

for auxin. 

7

Synthesis of New Fluorophores for Super Resolution Microscopy 

Umed Boltaev 

Mentors: Long Cai and Timur Zhiyentayev 

Super resolution microscopy techniques overcome the diffraction limit of traditional microscopy. However, there is 

a demand for photostable fluorophores with high contrast ratio in order to achieve the desired resolution and high 

labeling density. In order to use different fluorophores simultaneously one need to activate only one kind during 

imaging. The goal of this project was to synthesize new photoswitchable fluorophores with desired spectral 

properties. We have obtained 2 coumarins that are able to cage rosamine derivative with excitation maximum at 

550 nm. My final goal is to prepare rosamine with excitation maximum at 595 nm, cage it with spectrally distinct 

coumarins and test it as super resolution probes. 

The Characterization of Carbon Dots, Gold Nanorods, and Graphene as Components of Photothermal 

Tumor Treatment 

Carly Bond 

Mentor: Giyoong Tae 

Previous studies have shown that it is possible to use the photothermal effect of various materials as a method of 

killing tumor cells. Current objectives of researching this method include maximizing the accumulation of the 

photothermal material in the tumor, minimizing the non-specific cytotoxicity of the materials, and finding 

photothermal materials that can be heated using light (preferably NIR light). Several nanomaterials including 

carbon dots (CDs), gold nanorods (GNRs), and graphene will be compared as possible photothermal materials for 

tumor treatment. Loading materials into functional nanocarriers that can concentrate photothermal materials and 

deliver selectively to the target tissue before injecting them into the bloodstreamis a way to increase the efficacy of 

the photothermal therapy Chitosan-conjugated, pluronic-based nanocarriers have demonstrated the selective 

tumor targeting of loaded GNRs and thus highly effective tumor ablation upon laser irradiation. Therefore, the 

encapsulation efficiency and release of other materials from the nanocarriers and the photothermal heating of 

nanocarriers loaded with. CDs, GNRs, or graphene will be characterized for potential in vivo applications. 

Network for Collection of Seismic Data 

Prashant Borde and Prathamesh Juvatkar 

Mentors: Julian J. Bunn and K. Mani Chandy 

Earthquakes are natural disasters responsible for destruction, injuries and loss of life and property. However, 

earthquakes are not directly responsible for loss of life. The secondary events triggered by earthquakes such as 

structure collapse, fires, tsunamis and volcanoes cause actual human disasters. The losses due to earthquakes can 

be greatly reduced if we can give early warnings of earthquake or at least detect an earthquake’s early seismic 

activity. For building a system capable of early detection of seismic activity, we need high density arrays of 

spatially distributed motion sensors. Our project is to develop a private network of portable low cost seismometers 

that consume minimal resources. Each seismometer consists of a Single Board Computer (SBC), “Phidget” 

accelerometer and a Home-plug device for power line communication. Each seismometer sends the data to next 

device in linear topology using power-line communication. A base-station in the network sends the collected data to 

a central server via the internet. The operation of the seismometer network is completely automated after 

installation. Also the network is robust against intermediate node failure. Large numbers of such private networks 

will ensure sufficient amounts of data available for the detection of earthquakes with higher accuracy. 

A Criterion for Arithmeticity of Subgroups of PGL(2,C) 

Zarathustra E. Brady 

Mentor: Dinakar Ramakrishnan 

We attempt to generalize a result of Piatetski-Shapiro and Jiang (Geom. Funct. Anal., 8(3):586–605, 1998) from 

PGL(2,R) to the group PGL(2,C), giving a criterion for the arithmeticity of a discrete subgroup with finite covolume. 

We show that under the right assumptions on a subgroup Γ of PGL(2,C), Γ is arithmetic if it has an 

automorphic cusp form f such that f is a Hecke eigenform with respect to a prime p, and that this is equivalent to 

L(s,f) admitting an Euler factor at a prime p. 

Multiple Mirror Space Telescope Arrays: A Study of Imaging Applications and the Determination of 

Deformable Mirror Shape From Optical Wave Measurements 

Liana Braun 

Mentors: Sergio Pellegrino and Keith Patterson 

It has been proposed that a multi-mirror, self-attuning space telescope may be more cost-effective than a 

telescope with a large monolithic, high-mass mirror if the various smaller mirrors can produce results that rival the 

accuracy of currently used space telescope technologies. This project aims to optimize the focusing abilities of each 

individual deformable mirror in the telescope’s array in order to successfully image stellar objects. Using various 

references, an assortment of ideal stellar imaging targets was collected and plotted to determine accessibility by 

8

the spacecraft. Once this collection was completed, work was geared toward focusing light reflected from a 

deformable mirror in the lab. This was done through manipulation of voltages passed across the PVDF membrane 

of the mirror, coupled with an imaging process to detect changes in the image size. After a successful image 

focusing process is determined, it will be possible to redirect the image to a Shack-Hartmann sensor, which will 

provide more detailed image information, allowing for better shape sensing. 

Water on the Moon? A Quantitative Study of Water Solubility 

Alex Brett 

Mentors: E. Stolper and M. Newcombe 

Lunar water is a controversial subject: ongoing debate concerns its origin and quantity. Lunar glasses containing 

up to 40ppm water and melt inclusions containing up to 1500ppm water give broad constraints, but correct 

interpretation of these samples requires a thorough understanding of the diffusive behaviour of water at low 

concentrations in appropriate melt compositions. Here I present an experimental study on the solubility of H2O in 

anorthite-diopside and lunar-composition glasses at 1350°C and 1 atm (CO2-H2-H2O-CO). The results show that OH 

and not H2 is likely to be the dominant diffusive species, and establish empirical relationships between equilibrium 

concentrations and partial pressure of atmospheric water. 

Characterizing the Properties of Copper in Silicon Microwire Arrays 

Reuben Britto 

Mentor: Nathan Lewis 

Silicon microwires, grown from SiCl4 through the vapor-liquid-solid (VLS) process, have been shown to possess 

high material quality enabling solar cells to reach efficiencies that rival many wafer-based crystalline Si 

technologies. The VLS growth process requires annealing copper-patterned silicon wafers at 1000� C before 

introducing the Si precursor. Copper, due to its high diffusion coefficient, readily mixes with silicon at this high 

temperature, forming many different compounds in different phases. Selective etching, EDS-SEM, and XRD were 

used to determine what compounds and phases are forming, where they’re forming, and in what amounts. Copper 

oxide was detected and located on planar samples through SEM-EDS, and its cause of formation – sample 

exposure to atmosphere at >100� C – was established. Copper silicide (Cu3Si) was also detected through X-Ray 

diffraction on annealed planar samples. 

After identifying two phases of Cu impurity on planar samples, wire arrays were analyzed in the same manner, and 

several Cu-containing regions were identified. These results motivate further investigation of how changing growth 

and cooling conditions can minimize precipitation of Cu-rich phases in Si microwire arrays. 

Bioremediation of Endocrine Disruptors: Isolation of an Endocrine-Disruptor-Degrading Bacterium 

Julia Brown 

Mentor: Richard Murray 

Endocrine disruptors are compounds, such as estrogen or compounds that mimic estrogen, that interfere with the 

normal action of an organism’s endocrine system. The effects of this are most significant when exposure occurs 

during the organism’s development, as endocrine-disrupting compounds can disrupt the signaling pathways that 

direct the various stages of development. The Caltech iGEM team is developing a system, housed in E. coli, that is 

capable of detecting and degrading these compounds (specifically, bisphenol A, DDT, nonylphenol, or 17-αethinylestradiol). 

Using soil samples from the L.A. river as a source, we attempt to isolate a bacterium capable of 

degrading our compounds and determine the gene(s) responsible for degradation. We hope to express this gene in 

E. coli in conjunction with a modified human estrogen receptor to form a biological circuit capable of removing and 

degrading an endocrine-disrupting compound from water. 

Chemical Effects of Ice Crystal Growth 

Nina Budaeva 

Mentor: Kenneth Libbrecht 

The prediction of ice crystal growth rates and morphologies is a difficult problem which depends on nucleation 

dynamics, temperature history, chemical makeup of the sample, and many other parameters. This project 

investigates the effects of solute chemistry on the growth of ice crystals. By varying the physical parameter space 

of nucleation temperature and solution concentration, I am able to observe changes in the ice crystal growth of 

various water solutions. 

The ice crystals are grown in a wire loop and are recorded on camera so that their growth rates and morphologies 

can be observed from the recordings. Solutions of particular interest are salt and sucrose solutions as well as 

nanoparticle suspensions and antifreeze protein solutions. The results of these experiments, besides providing 

scientific insight into an insufficiently-explored field, may also prove useful for direct applications such as in freezedrying 

of pharmaceuticals. 

9

Meter-Scale Evolution of the Martian Residual South Polar Cap 

Peter B. Buhler 

Mentors: James W. Head III and Andrew P. Ingersoll 

We analyze high resolution imagery of the Residual South Polar Cap (RSPC) on Mars to document seasonal and 

interannual ablation and accumulation processes of CO2 ice. We observe vertical and horizontal changes in the 

RSPC and find that ablation and accumulation at meter-scales vary both spatially and temporally. 

This investigation details the mesoscopic (decimeter to decameter scale) processes by which the RSPC evolves on 

multi-day to annual timescales. We find that seasonal CO2 frost persists longer into the summer during Martian 

Year (MY) 30 than it persists in MY 29 in some locations. We postulate that this gradual accumulation offsets the 

sublimation of the RSPC, acting to stabilize the RSPC on timescales shorter than solar or orbital climate cycles. 

Solar Cell Enhancement by Whispering Gallery Modes From Dielectric Nanosphere Array Coatings 

Colton Bukowsky 

Mentors: Harry Atwater and Dennis Callahan 

Thin-film photovoltaic performance can be improved by increasing the absorption capability of the active layer. 

An innovate method utilizing dielectric-resonant nanospheres placed on top of the active layer has been shown by 

simulations to excite whispering gallery modes (WGMs) in the nanospheres, which then couple incident light into 

the active layer. This coupling increases the absorption at these wavelengths, and it is desired to experimentally 

show this phenomenon occurs in real solar thin-film solar cells. Efforts to create hexagonally close packed 2D 

colloidal crystals of silica nanospheres have been successful in creating long range order on various substrates with 

greater than 90% area coverage using the Langmuir-Blodgett technique. These arrays were coated onto semiconducting 

surfaces and functioning PV cells. Absorption measurements will be used to demonstrate absorption 

peaks at certain wavelengths and matched to those in the models in order to show that the WGMs are excited. 

The Receptor mGluR1 Is Reduced in Postsynaptic Densities of Mice With a Deletion of Densin 

Sarah R. Cantor 

Mentor: Mary B. Kennedy 

Densin is a scaffold protein that is abundant in the postsynaptic density. Mice with a homozygous knockout of 

densin exhibit abnormal behaviors that are characteristic endophenotypes of schizophrenia and autism spectrum 

disorders. Past research has shown that densin knockout mice have a 30% reduction of mGluR5, a metabotropic 

glutamate receptor, in the postsynaptic density compared to wildtype mice. Type I metabotropic glutamate 

receptors comprise two individual receptors, mGluR1 and mGluR5. Currently, little is known about the binding 

partners of mGluR1 and whether this protein associates with densin. Through the use of quantitative immunoblots, 

I showed that densin knockout mice have a 30% reduction in levels of mGluR1 in the postsynaptic density fraction, 

whereas levels in the total homogenate are unchanged. This indicates that densin plays an important role in the 

localization of both mGluR5 and mGluR1 to the postsynaptic density. It may be part of a protein complex that 

associates with both mGluR1 and mGluR5 at the postsynaptic membrane. 

Cell Phone Earthquake Detection: Small Memory Variations on Expectation Maximization for Gaussian 

Mixture Models 

Alejandro Carbonara 

Mentor: Andreas Krause 

Most smartphones come equipped with accelerometers, providing a wealth of untapped data. The Caltech Seismic 

Network project aims to use this data in order to detect earthquakes as they are happening and warn cell phone 

users as early as possible. Our part of this project is the client side abnormality detection. Since we have a limit on 

how much data can be transmitted, each cell phone needs to detect which of its recordings are statistically 

significant in order to only transmit only the most important information. We decided early on in our project that 

we wanted to use Gaussian Mixture models to model our data and that we wanted to use Expectation Maximization 

to fit these models. Due to the memory limitations of a cell phone, we examined various ways to compress our 

data, including randomization, coresets, and a streaming model. Using cell phone data collected to represent an 

average day, we coded up and ran several experiments to see how well these various small memory 

implementations worked. 

Forming a Porous Scaffold With Clover-Leaf Shaped Microgels for Tissue Engineering Formulations 

Jacqueline Chan 

Mentors: David A. Tirrell, Julie A. Kornfield, Amy Fu, Larry Dooling, and Loddie Foose 

Three-dimensional protein polymer scaffolds have been engineered to promote specific cell behaviors such as 

migration, differentiation, and proliferation. This work describes a method of fabricating clover-leaf shaped 

polyethylene glycol diacrylate microgels that can be modularly assembled into a macroscopic scaffold material with 

high porosity, permeability, and pore interconnectivity. A secondary crosslinking reaction is used to stabilize the 

10

assembled microgels. This bottom-up method allows for the rapid and scalable fabrication of microgels with defined 

geometrical and biological features, which can potentially be used for microscale tissue engineering applications. 

Future work entails fabricating these microgels with photocrosslinkable artificial extracellular matrix protein. 

Fluorescent Double Labeling of the Signal Recognition Particle to Study Interdomain Interaction 

Sandhya Chandrasekaran 

Mentors: Shu-ou Shan and Ishu Saraogi 

Proper cellular localization of proteins is essential for survival and errors in this process could lead to cell 

malfunction or cell death. About one-third of the proteins in the cell are delivered to their correct cellular 

destination by the Signal Receptor Particle (SRP). During targeting, SRP forms a heterodimeric complex with its 

receptor (SR), and the SRP-SR complex goes through a series of conformational changes that provides the spatial 

and temporal cues for the targeting reaction. The protein component of SRP, called Ffh, is a multi-domain protein 

comprising a substrate-binding domain connected to a GTPase domain via a flexible linker. The mechanism by 

which the movement of these two domains with respect to each other coordinates downstream signaling events is 

poorly understood. Here, we developed a double fluorescence-labeling scheme for Ffh to study these interdomain 

movements with Forster Resonance Energy Transfer (FRET). By optimizing labeling conditions, Ffh double-cysteine 

mutants were differentially labeled using cysteine-maleimide chemistry with complementary donor and acceptor 

fluorophores. Subsequent FRET experiments with these labeled proteins can provide important kinetic and 

thermodynamic information about events in the protein targeting process that involve inter-domain movements of 

Ffh. 

Simulac: A Stochastic Simulator for Modeling Biomolecular Circuits 

Arjun Chandrasekhar 


The primary purpose of this research project is to develop and modify Simulac, a stochastic simulator for creating 

and simulating models of various biomolecular circuits. We want to add and modify different features of the 

simulator, and then use it to simulate environments expressing bimodal and graded gene expression to determine 

the factors that affect noise in gene expression. The first goal is to make the program easier to use by 

incorporating the graphical user interface from TinkerCell. Next we want to implement tau leaping algorithms to 

make the program run faster and more efficiently. Finally we want to expand the set of reactions that the simulator 

can account for, including anti-sense RNA binding and various polymerase interactions. In the future, once all of 

these features are in place, the next step will be to set up various environments and modify features that we think 

are correlated to noisiness in gene expression (combinatorial promotes, transcriptional efficiency, chromatin 

remodeling, etc.) in order to determine what causes bimodal and graded responses. 

Thioester-Functionalized Silica Microspheres as Substrates for Covalent Immobilization 

Christine H. Chang 

Mentors: Harry B. Gray and Paul J. Bracher 

This project focuses on the development of a binding system based on the thiol–thioester exchange reaction that 

uses thioester-functionalized silica microspheres as a substrate to bind thiols. The covalent immobilization of 

molecules to a solid phase is a key step in a wide range of applications, such as solid-phase organic synthesis, 

affinity chromatography, and assays for biological activity. While several binding systems are currently in use— 

each with its advantages and disadvantages—the system under investigation in this project is particularly 

advantageous due to its use of small and noninvasive molecular tags as binding partners, its ability to function in 

biologically compatible conditions, and the availability of several methods to cleave the tightly bound partners 

without the use of harsh reagents. The silica microspheres were fabricated by the Stöber process, the binding 

interactions were measured by fluorimetry, and the susceptibility of the microspheres to aminolysis and hydrolysis 

was also studied. Specific areas of interest included the effects of phosphorus(V) versus EDC coupling reagents in 

the preparation of the functionalized microspheres, metal species (e.g., Cu 2+ , Pb 2+ , and NaAuCl4) on the rates of 

aminolysis, and increased thiol functionalization on binding. 

11

Identification of Cis-Regulatory Regions in L. variegates DNA 

James Chang 

Mentor: Andrew Cameron 

Gene regulatory networks consist of elements that control gene expression and the connections between them. 

Computational genomics is a useful tool for finding non-coding but functional cis-regulatory regions which affect 

gene regulation in the Strongylocentrotus purpuratus (Sp) sea urchin genome. To find similarities, Sp sequence is 

juxtaposed with that of Lytechinus variegates (Lv), a species 50 million years diverged. Divergence of 

nonfunctional areas increases with evolutionary distance, but functional areas will be conserved. Thus, possible 

regulatory regions are defined as conserved areas upstream and downstream from genes. The search for 

regulatory regions begins by pairing similar genes as initial references. Then, the surrounding sequence from both 

genomes can be compared. Although similarities between known Sp genes and Lv sequences have been found, 

work on corresponding upstream and downstream sequences has not been completed. Once scripts have been 

adapted to large-scale sequence files, the process will be streamlined and potential cis-regulatory regions will be 

identified. 

An Exact Solution for the Time-Dependent Behavior of the Kitaev Toric Code Under a Rapid Quench 

Adrian Chapman 

Mentors: John Preskill and Steve Flammia 

Topologically ordered quantum systems have recently been of interest as potential mechanisms for the 

implementation of fault-tolerant quantum computation. We examine the time-dependent behavior of one such 

system, the Kitaev toric code, following the sudden application of a local transverse magnetic field -- a quantum 

quench. By mapping the transverse toric code to a series of decoupled one-dimensional Ising models, we exactly 

solve for the energy spectrum and eigenstates of the original model and use the result to calculate the time 

evolution of the expectation values of contractible and non-contractible loop operators, along with the topological 

entanglement entropy. We expect these quantities to remain constant over a timescale short compared to the 

system size divided by the Lieb-Robinson velocity. This study is useful for the understanding of topological phases 

and the construction of reliable quantum memories. 

Modeling Nonlinear Response in a Dispersion-Engineered Traveling-Wave Kinetic Inductance 

Parametric Amplifier 

Saptarshi Chaudhuri 

Mentor: Jonas Zmuidzinas 

Recently, Zmuidzinas and Day proposed a new concept for a low-noise cryogenic amplifier known as a Dispersion- 

Engineered Traveling-Wave Kinetic Inductance (DTWKI) amplifier. This superconducting transmission line device 

uses nonlinear kinetic inductance to convert pump wave photons into signal wave photons, producing significant 

signal power gain. In theory, the DTWKI amplifier should operate close to the noise limit imposed by quantum 

mechanics. However, the nonlinearity leads to the production of pump harmonics, and unless countermeasures are 

taken, shock waves form, preventing signal amplification. In the DTWKI device, shock front formation is prevented 

by periodically perturbing the width, and hence the impedance, of the transmission line to create a stop band at 

three times the pump frequency. Using a harmonic-balance computation scheme, a model of the amplifier was 

developed. Wave deformation and the growth of pump harmonics in uniform transmission lines were studied 

extensively. The effects of nonlinear kinetic inductance on traveling waveforms in periodically-loaded transmission 

lines were compared to the effects on waveforms in uniform lines. 

Sampling Tool Concepts for Comet Surface Sample Return 

Allen Chen 

Mentor: Paul Backes 

Three sampling tools for sampling of comet surfaces in a touch-and-go (TAG) mission architecture were developed 

and tested and compared with test results using a Brush Wheel Sampler (BWS) sampling tool. The BWS was 

developed previously and tested across a range of stimulants so it was used as a reference to compare the new 

tool concepts with. The three tools were a drive tube (DT), comet corer (CC), and comet auger (CA). The primary 

anticipated benefits of the DT, CC, and CA compared with the BWS were that they could acquire subsurface 

samples to 10cm depth and potentially enable a simpler sample acquisition, measurement, and transfer (SAMeT) 

capability compared to the BWS. The BWS has been shown to reliably acquire a large surface sample but it is of 

interest to determine whether an alternative tool could acquire a subsurface sample to at least 10cm depth with 

similar reliability and the same or reduced system complexity. 

12

Patterns of Nanog Sub-State Switching in Single Mouse Embryonic Stem Cells 

Chih-Ping Chen 

Mentors: Michael Elowitz and John Yong 

Nanog is a transcription factor thought to safeguard embryonic stem cells from differentiating. However, it is found 

to be expressed heterogeneously even in undifferentiated populations between a high-Nanog state and a low- 

Nanog state. Of particular interest for this project, it has been observed that individual cells can switch between 

high-Nanog and low-Nanog states while remaining undifferentiated. Using time-lapse fluorescent imaging, it is 

possible to track Nanog transcription in single cells through multiple cell cycles. From analyzing these tracks, we 

seek to characterize the frequency, time-scale and pattern of inheritance of the switch. These patterns are being 

compared to establish the uniqueness of the way Nanog transcription is being regulated. This analysis will serve as 

the basis for further study of the phenomenon under various culture conditions. 

Doped Polycrystalline Silicon Templates for Epitaxial Crystalline Thin-Film Solar Cells 

Clare Chen 

Mentors: Harry A. Atwater and Michael G. Deceglie 

We demonstrate the process of metal induced lateral crystallization (MILC) of phosphorus and boron doped 

amorphous silicon films to produce templates for epitaxial growth of thin-film crystalline silicon solar cells. In 

developing these thin-film crystalline templates, a low-temperature, low cost process enables the use of low-cost 

substrates such as glass. Upon annealing with nickel nanoparticles, amorphous silicon was crystallized by MILC. 

These thin films were characterized by optical microscopy, atomic force microscopy, and scanning electron 

microscopy to analyze MILC growth. Then surface nickel deposits were etched off the crystallized template before 

electrical characterization by Hall measurements and compositional analysis by secondary ion mass spectrometry 

(SIMS). These measurements were used to investigate the carrier concentration in this doped template layer to see 

if this polycrystalline template can provide back surface field for the solar cell, preventing flow of minority carriers 

to the rear of the cell thus minimizing surface recombination losses. 

A Modified Cellular Automata Model for Dendritic Computation 

Daniel Chen 

Mentors: Christof Koch and Adam Shai 

There exist many proposed models for post-synaptic integration in a typical pyramidal neuron. Many, for the sake 

of simplicity, do little, if anything, to account for the nonlinearities that exist throughout the neuron. Other more 

sophisticated models abstract the tree into a large set of complicated circuits, which require thousands of 

parameters. Consequently, it may be difficult and computationally intensive to use these models to perform 

analysis beyond small networks. In this paper, we propose a model that accounts for all of the nonlinearites while 

remaining relatively simple and easily adjustable to account for additional parameters. We have made several 

simplifications; particularly, abstracting the voltage transients induced by single synapses to discrete signals which 

interact with each other according to simple rules. The time and space evolution of these are governed according to 

known parameters. With this model, we hope to gain an understanding of the fundamental components of the 

single cell mechanism and thus be able to develop more refined experiments that focus on the more important 

factors. 

Establishing Ground Truths for Automated Monitoring and Analysis of Aggressive Behavior in Female 

Drosophila 

John Chen 

Mentors: David Anderson and Rod Lim 

Aggression is an innate social behavior observed in many species. Recently, Drosophila has emerged as a model 

system for studying aggression. Previous studies demonstrated that fruit flies, as in many other species, exhibit 

gender-specific patterns of aggressive behaviors. Although machine vision algorithms have been adapted for 

automatic detection of male-male social behaviors, no such tool exists for female-female behaviors. In order to 

characterize female behaviors in a systematic way, we seek to develop a machine vision system that can quantify 

and analyze various social behaviors in Drosophila. Currently these female behaviors are measured manually, 

which is slow and laborious. A machine algorithm that can detect and analyze these behaviors will be highly useful 

in studying the genetic and neural bases behind female aggression. By carefully annotating female-female 

aggression using high resolution cameras, we will establish definitive behavioral ground truths that, through 

collaboration with the Perona Lab, can help establish a machine algorithm for evaluating aggression in female 

Drosophila. 

13

Identifying Genes Involved in FGF-Dependent Cell Migrations in Drosophila 

Rebecca Chen 

Mentors: Angelike Stathopoulos and Young Bae 

Embryonic development of multicellular organisms is a complicated process of cell movement and cell 

differentiation. Drosophila serves as a model organism in which intricate gene regulatory networks that orchestrate 

the complex process of gastrulation can be studied. This project focuses on the study of two dynamic events that 

are regulated by fibroblast growth factor (FGF) signaling: migration of early mesoderm and caudal visceral 

mesoderm (CVM). To identify genes that play a role in mesoderm migration and differentiation, the gene 

expression profile of sorted mesodermal cells was obtained by RNA-seq. We are verifying gene expression in the 

mesoderm by in situ hybridization and testing if expression of any target genes is dependent on FGF signaling. 

Phenotypic analyses of those with specific expression in the early mesoderm will be performed. Similarly, for CVM 

cell migration, we confirmed expression of candidate genes through in situ hybridization. We are examining what 

role these candidate genes play in CVM cell migration using RNAi hairpins. In addition, a technique for cell 

transplantation has been developed so that cells can be transplanted and observed in an embryo of different 

genetic background. 

Parylene-C Characterization 

Rujian Chen 

Mentors: Yu-Chong Tai and Chun-Hui Ling 

Parylene-C is widely used as a biocompatible material for many kinds of micro analysis systems. As-deposit 

parylene coatings feature low thickness, high uniformity, good mechanical strength, barrier properties and 

adherence to substrates. However, parylene used in practical devices will go through many treatment processes 

which may alter its microstructure and hence its mechanical and chemical properties. During the fabrication of 

MEMS devices, it is observed that parylene films become more brittle after heat treatment, which shortens the 

film's lifetime and degrades its performance. This project investigates the cause and mechanism of the 

deterioration of mechanical properties. Parylene samples annealed under different conditions(air/vacuum, different 

temperatures and annealing times) are prepared, and structural changes are studied, by means of fluorescence, 

IR, XPS, XRD methods. Experimental results have shown evidence of the onset of densification, oxydization, and 

crystallization under different annealing conditions. 

Microfluidic Polymerase Chain Reaction (PCR) Platform for Pathogen Detection 

Samson Chen 

Mentor: Axel Scherer 

Microfluidic medical diagnostic systems, due to their ability to integrate incredibly complicated procedures into a 

single device, have the potential to drastically reduce the cost and increase the availability of modern medicine 

worldwide. However, there are still many significant barriers to achieving this goal with current microfluidic 

fabrication techniques. We present the first working application of a novel microfluidics platform which overcomes 

many of these barriers. Specifically, we fabricated a fully integrated Polymerase Chain Reaction (PCR) thermal 

cycler, a device which permits the amplification of very small concentrations of specific sequences of DNA, often 

found as part of many pathogen detection schemes. Additionally, we designed and optimized a low cost circuit on 

the same device to thermally regulate the PCR chamber, demonstrating the ability of this platform to tightly 

integrate electronics and fluidics in the same package at very low cost. 

Stereological Sampling of von Economo Neurons (VENs) in Aging Human Brain Tissue 

Victor Chen 

Mentor: John Allman 

The principal objective and scope of the project was to investigate the role of VENs in the human brain. It has been 

hypothesized that these neurons serve as homeostatic mechanisms to ensure proper functioning of core bodily 

functions, and therefore the project focused on determining whether there was a correlation between the number 

of VENs and longevity. The methodology of the project was based on the use of a stereoinvestigator to make 

accurate estimates of the number of VENs within the fronto-insular cortex. Only very preliminary results have been 

obtained so far, as very few aging brains have been completely counted with stereology for VENs. The project will 

be definitely continued with subsequent work focusing on enlarging the data pool so that reliable conclusions can 

be made about the importance of VENs in the brain. 

Identifying a Protein Lattice on the Outer Spore Membrane in Acetonema longum 

Wesley Chen 

Mentors: Grant Jensen and Elitza Tocheva 

During sporulation, an elaborate protein coat essential for spore viability assembles around the outer spore 

membrane. One protein of interest, SpoIVA forms the foundation of the protein coat to which all other proteins 

bind to. However, the exact structure of SpoIVA is unknown. While the protein coat has been studied extensively in 

14

Bacillus subtilis, there are no high-resolution images of SpoIVA in vivo during early stages of sporulation. One 

possible technique, electron cryo-tomography (ECT), is a powerful new tool that is capable of producing threedimensional 

reconstructions of microorganisms and proteins in their near-native state in close to atomic resolution. 

Acetonema longum is a sporulator suitable for ECT studies. Previously, a series of concentric protein rings has been 

imaged on the outer spore membrane of A. longum that we hypothesize to be SpoIVA. Here, we use a variety of 

biochemical assays, imaging techniques, and genetic manipulations to identify the protein lattice found on the 

protein surface of the endospore. We anticipate that these approaches will provide opportunities for identifying the 

protein lattice formed during early stages of sporulation in A. longum. 

Light Curve Classification of Time-Variable Astronomical Phenomena 

Yutong Chen 

Mentors: George Djorgovski, Baback Moghaddam, and Ashish Mahabal 

Recent development in digital synoptic sky surveys presents researchers with a more in-depth examination of 

known and newly discovered time-variable astronomical phenomena, or transients. This opportunity comes in the 

form of an overwhelming amount of data, and at the same time requiring identification and prioritization for followup 

observations. We employ a Decision Tree (DT) classifier for automatic classification of transients based on the 

transient’s light curve. Given a transient, we convert the transient’s light curve to its corresponding Δm vs. Δt 

histogram. We can generalize a particular class of transients’ Δm vs. Δt histogram; and thus building a probability 

density function (prototype histogram) for that class. Classification of an unknown transient is accomplished by 

sequentially cascading the transient’s Δm vs. Δt histogram against the DT’s prototype histograms. We assess the 

reliability of our DT classifier using a 5-fold cross-validation on a comprehensive confusion matrix. Our DT classifier 

performs well in correctly identifying CV, RR, and SN1a; but the reliability of our DT classifier significantly 

decreases for Blazar, Mira, SN1bc, and SN2np. A correlation with the reliability of our DT classifier is that wellperforming 

classes contains at least 350 examples in their training sets while poor-performing classes contains less 

than 100 examples in their training sets. 

Reorganization of Nanoparticles on Origami Surfaces by DNA Random Walkers 

Zibo Chen 

Mentors: Erik Winfree, Lulu Qian, Niranjan Srinivas, and Damien Woods 

The specificity of Watson-Crick base pairing has been shown to be useful when constructing nanostructures using 

DNA. Since the development of DNA nanotechnology, DNA structures with various functions have been 

constructed, such as DNA tiles that can process information, DNA origami that can serve as molecular assembly 

lines, etc. This study, conducted together with Shayan Doroudi, Gregory Izatt, Yae Lim Lee and Sarah Wittman, is 

part of the upcoming BIOMOD competition. In this study, we developed a technique to reorganize randomly placed 

nanoparticles using random walk. Using the algorithm NUPACK, we designed single-stranded DNA walkers that 

were capable of walking on DNA origami and picking up cargoes with DNA strands attached; they would continue 

the random walk until encountering goal strands and transferring the cargoes to the goals by strand displacement. 

The mechanisms were verified using polyacrylamide gel electrophoresis, fluorescent spectroscopy and atomic force 

microscopy. This system is useful when being integrated into a DNA assembly line to deliver DNA labeled 

nanoparticles, and it opens up the possibility of sorting different types of cargoes at the same time. 

Synthesis of Nickel Complexes of Redox-Active Protic Ligands for CO2 Reduction 

Christine Cheng 

Mentors: Theodor Agapie and Emily Tsui 

Carbon dioxide is a major cause of global warming and results from many common processes. Since CO2 is known 

to bind to dinuclear nickel complexes, a dinickel complex supported by a diphosphine p-terphenyl ligand was 

designed as a potential electrocatalyst for the reduction of CO2. In this project, variants of the ligand containing 

electron-donating methoxy and hydroxyl groups on the central arene were synthesized. Monometallic nickel(0) 

complexes and bimetallic nickel(I) complexes were synthesized using these diphosphine p-terphenyl ligand 

frameworks. These complexes were characterized by NMR and other methods and studied for CO2 reactivity. 

Using E. coli to Bioremediate Endocrine Disrupting Chemicals 

Puikei Cheng 

Mentors: Richard Murray, Nathaniel Glasser, Toni Lee, Joseph Meyerowitz, and Emmanuel Lorenzo de los Santos 

Endocrine disrupting chemicals, or EDC’s, disrupt the reproductive abilities of aquatic wildlife by imitating hormones 

present in their bodies. Traditional efforts to remove these chemicals have faced problems such as cost, disposal of 

products, and unforeseen consequences. This year, the Caltech iGEM team has decided to try a better method of 

eliminating these pollutants using bioengineering. Evolution has already selected for bacteria which metabolize 

EDC’s in polluted rivers, so a simple search for genes which can degrade toxic chemicals in the environment has 

proven fruitful. Two main approaches used were to either first select for these unusual bacteria by plating collect 

15

mud samples on media which only contain our chosen EDC’s as a carbon source, then sequence the gene, or the 

extraction DNA from the soil first and insertion into E. coli using a special type of plasmid afterwards. In addition, 

we explored the ability of a class of enzymes called cytochrome p450’s to degrade these chemicals. 

Towards the Use of Formal Methods for Dexterous Robotic Manipulation 

Sandeep Chinchali 

Mentors: Joel Burdick, Scott Livingston, and Ufuk Topcu 

We apply advances in formal methods and symbolic motion planning to synthesize robust, generalized control 

strategies for the canonical problem of finger gaiting in dexterous manipulation. In this scheme, fingers on a hand 

move continuously to re-orient an object until they encounter joint limitations, at which point the constrained 

fingers must be repositioned while always maintaining a force closure (FC) grasp. 

The problem of gaiting an object in the midst of external perturbations is modeled as a two person Generalized 

Reactivity (1) game between a control system (the robotic hand) and adversarial environment. Desired behavior of 

the system is specified using linear temporal logic (LTL) and a correct-by-construction control strategy in the form 

of an automaton is synthesized. Finally, results from a kinematics simulation of a planar robotic hand manipulating 

an object are presented to illustrate the efficacy of this approach. 

Purification of a Photon State Using Linear Optics 

Soonwon Choi 

Mentors: John Preskill and Netanel Lindner 

Many optical implementations of quantum information processing protocols require high quality on-demand singlephoton 

sources. Such photon sources must exhibit high collection efficiency and high purity for output photons. A 

promising route to such sources is based on solid state structures. A solid state based photon source exhibits good 

single-photon collection efficiency, but yields photonic states which are mixed in their spectral degrees of freedom. 

In this research, we demonstrated a simple method to improve the purity of a photon state using only linear optical 

devices, measurements, and post-selection conditioned on measurement outcomes. This restricted set of 

operations considered was chosen for simplicity and experimental relevance. The method we study utilizes the 

bosonic statistics of photons. The simplest example in which an improvement is achieved relies on interfering two 

photons in a 3 port interferometer. We shall further study possible optimizations and generalizations of this 

method. 

Numerical Simulation and Analysis of Darcy's Law in Fuel Cells 

David Choy 

Mentors: Guillaume Blanquart and Gerry Della Rocca 

Fuel cells are one of the most promising renewable energy technologies. Through liquid and gas transport, 

electricity is generated by a reduction reaction. The dynamics behind the gaseous molecular transport in microscale 

porous media are not well-understood in that an often-used law, Darcy’s law, is only an empiricism derived 

from simple channels and pipes. The validity of Darcy’s law in all structures, including complex geometries are the 

focus of this project. By adding a mask to a grid of cells, thereby creating solid and porous sections in a structure, 

various porous media were generated in which Darcy’s law was tested. With the aid of Fick’s law, Darcy’s law was 

compared to results from Monte Carlo simulations in the same structures. The flux through the structure provided a 

diffusive constant and a random walk Monte Carlo algorithm found the tortuosity used in the comparison. Data 

shows that Darcy’s law is valid for all structures at all logical porosities. 

Investigation of Different Isotope and Chemical Patterns of Pyrite Oxidation From the Río Tinto SW, 

Spain 

Justin Christensen 

Mentor: Max Coleman 

The Río Tinto, a river in southwest Spain, has become a premier site for Mars-analog research due to its highly 

acidic waters and sulfate rich nature. Influences from extremophile organisms on the isotope composition of sulfur 

and oxygen provide a unique opportunity for biomarker characterization. Our objective was to use oxygen isotope 

data collected from various sources on the Río Tinto system to determine the source of sulfate oxygen (either 

water or atmospheric), and the biological influence on the oxidation pathway and resulting isotopic signature. 

Preliminary data show ∆ 17 O ranging in values from +0.1‰ to +0.13‰. This is unexpected, since it indicates 

neither oxygen incorporation in sulfate from atmospheric (-0.35‰) nor water oxygen (0‰), the only two possible 

sources. Future δ 18 OSO4 measurements are planned and the results used as an additional indicator of source 

oxygen. As further data become available the relationship between δ 18 O and δ 17 O can be compared against the 

existing literature, δ 17 O = 0.52 δ 18 O + ∆ 17 O, and a refined slope specific to sulfate processes defined. In addition 

these data suggest the presence of sulfite intermediaries and isotope exchange processes during oxidation. Further 

data will help identify exact oxidation pathways and microbial influences. 

16

Constructing a Flexible Phage Display System for Characterizing the Peptide Substrate Specificity of the 

Bacterial Oligosaccharyltransferase 

Tinyi Chu 

Mentor: William Clemons 

The N-linked glycosylation is a post-translational modification of proteins containing the conserved sequon Asn-X- 

Ser/Thr. It is implicated in protein folding, trafficking and quality control, and ultimately exerts an impact on 

diverse physiological processes. The oligosaccharyltransferase (OST) is a membrane protein involved in 

transferring the oligosaccharide from the lipid-anchored oligosaccharide to the sequon on target protein. In 

general, All OSTs have their glycan donor specificity and acceptor peptide specificity. Recently, several bacteria 

species have been found to contain the N-linked glycosylation system, and their OST homologs, PglB, have been 

shown to have higher selectivity at the acceptor peptide sequon in addition to the requirement of Ser/Thr at the +2 

site. By constructing a flexible glycophage display system with high efficiency and controllability, we aim at 

characterizing the acceptor peptide sequon specificity of these various PglBs, including the pglBs with their saltbridge 

artificially mutated. The components of the phage display system are under construction and the efficacy of 

the system will be verified by in vivo glycosylation. The data provided by this phage display system will underlie 

the downstream characterization of the N-glycan structures from the corresponding species and their roles in 

virulence and cellular physiology. Furthermore, it may hopefully provide insight for developing the bacterial 

glycoprotein synthesizing system. 

AMMOS Cost Model Development 

Akash Chudasama 

Mentor: Eleanor Basilio 

AMMOS Cost Model Environment (ACME) is a model development project being undertaken by the GST (Ground 

Segment Team). The main purpose of ACME is to provide models that allow for the costing of the AMMOS 

(Advanced-Multi-Mission Operations System) tools and services suite to systematically implement mission 

operations to strategically enhance mission efficiency, reliability, and cost effectiveness. The ACME models will be 

used primarily during the Formulation phase to provide general cost estimates to allow for the appropriate design 

of a mission. 

This project focuses on the development of the Data Management element of Mission Operations, which is 

responsible for the end-to-end data flow during the life-cycle of the mission. Through detailed research and 

customer interviews, an algorithm is developed focusing on key mission elements and cost drivers. Having 

formulated an algorithm, a cost model is constructed and implemented into the larger ACME suite to yield cost 

estimates in the form of Full Time Employees (FTEs). This tool is relevant in providing insight to determine the 

required resources and feasibility of a proposed mission concept before the mission is implemented. 

Determination of Simple 3-Dimensional Abelian Surfaces With Complex Multiplication With Field of 

Moduli as Q 

Daniel Chun 

Mentor: Matthias Flach 

In this paper, I will be generalizing the results of Murabayashi's paper "The field of moduli of abelian surfaces with 

complex multiplication" to the case of 3-dimensional abelian varieties with complex multiplication. Furthermore, I 

will be showing some computational results involving computations of Hecke chracters for the 3-dimensional 

abelian varieties with complex multiplication which satisfy the condition of having its field of moduli coincide with 

the field of rational numbers. Basic results from number theory and paraphrasing of Murabayashi's arguments have 

been enough in achieving these aims. How- ever, for n-dimensional abelian varieties with complex multiplication 

where n is greater than or equal to 4, our approach here might not apply so well as there are far more caseworks 

and complications to be accounted for. Thus, for arbitrary n, a new method of attack seems well-advised. 

Predicting Lift-Off Height in Turbulent Lifted Flames Using Stochastic Flamelets 

Emmet M. Cleary 

Mentor: Guillaume Blanquart 

Autoignition is often considered the most crucial mechanism in predicting lift-off height. In this work, lift-off height 

of a turbulent ethylene jet flame in hot coflow is modeled numerically using a stochastic flamelet approach. 

Chemistry is modeled using a reduced mechanism for ethylene combustion, and is used to track ignition with a 

progress variable measuring CO2, CO, H2O, and H2 species. This progress variable is shown to be sufficient to track 

ignition even in the diffusion combustion regime, contributing only small errors to the ignition delay time. Modeling 

with stochastic flamelets takes temperature and the progress variable as deterministic parameters with a stochastic 

scalar dissipation rate. The results from the Direct Numerical Simulations (DNS) of a turbulent lifted ethylene jet 

17

flame reported by Yoo et al. (Proc. Comb. Inst., 2011) are used to find the decay rate of the mean dissipation rate 

and the normalized variance. The stochastic flamelet approach is found to predict the mean lift-off height and its 

fluctuations reasonably well. 

An Examination of the Spatial Distribution of Carbon Dioxide and Systematic Errors Within 

Algorithmic Retrievals 

Brennan Coffey 

Mentor: Mike Gunson 

With the advent of technologies such as vehicles compounded with general population increases, the 

concentrations of CO2 have steadily risen. This increase, which is being closely studied, has ramifications 

throughout the biological world. Through applications of spectroscopy, and radiative transfer amongst others, it is 

possible to calculate how many molecules of CO2 lie in a defined column of air. However, other molecules are 

present in the atmosphere, such as aerosols and water, which diffract the light. Understanding the diffraction 

geometry and path length is vital to computing the concentration of CO2. Comparing these satellite readings with 

ground-truth data (TCCON) the systematic error contingent on the presence of other particles can be assessed. 

Once the error is understood, it can be scaled for in the data retrieval algorithms to acquire a set of data, which is 

closer to the TCCON readings. Using this process, the algorithms are being developed to be perfect, within .1% 

worldwide of the true value. At this stage, the accuracy is within 1%, but through amending small issues contained 

in the algorithms, such as accounting for the nonlinear path of sunlight, the desired accuracy can be achieved. 

Using Rubredoxin From Pyrococcus furiosus as a Model for Determining Oxidation States of Individual 

Fe Atoms in Nitrogenase as a Means for Probing the Mechanism of Biological Nitrogen Reduction 

Connor W. Coley 

Mentors: Douglas C. Rees and Limei Zhang 

Nitrogenase is the enzyme responsible for biological nitrogen reduction. The oxidation states of individual Fe in 

nitrogenase can be assigned using the multi-wavelength anomalous diffraction technique, which is critical to 

understand the functional mechanism of nitrogenase. The single Fe-bound Rubredoxin from Pyrococcus furiosus is 

used as a model for optimizing data collection and data analysis for such a goal. In this project, the wild type 

Rubredoxin from Pyrococcus furiosus has been successfully overexpressed, and purified with high homogeneity 

using anion exchange and size-exclusion chromatography. Crystallization methods have also been refined, yielding 

single crystals up to 1 mm in diameter through a combination of microseeding and macroseeding. Any high quality 

crystals produced in this study will be applied to refining data collection strategy in the future. Analysis of multiwavelength 

anomalous diffraction data of PfRd collected previously will also help to improve the algorithms for data 

analysis in the subsequent studies. 

Structure Determination of Two Commensal Colonization Factors in Bacteroides fragilis 

Skylar Cook 

Mentors: Sarkis Mazmanian and Sung-Eun Lee 

The human gut is home to an immense group of symbiotic bacterial species. Bacteroides fragilis is among the most 

numerically prevalent colonizers of the human gastrointestinal tract, however the mechanism of bacterial 

colonization remains unknown. Recent studies by the Mazmanian group have identified a highly conserved genetic 

locus required for colonization in B. fragilis, comprised of 5 “commensal colonization factor” (ccf) genes. The 

Mazmanian group has shown that B. fragilis ccfC and ccfD gene knockouts are easily outcompeted when challenged 

by wild-type B. fragilis; they are unable to maintain saturation of their anatomic niche. In this study we attempt to 

determine the structure of these two proteins. 

Optimization of Vertical Axis Wind Turbine Arrays Using Numerical Simulations 

Anna Craig 

Mentor: John O. Dabiri 

Vertical axis wind turbines (VAWTs) are currently under investigation as a potential means by which to make wind 

energy a more economically, socially, and environmentally viable option. Although individual VAWTs are less 

efficient than other types of turbines, VAWTs can be placed closely together due to their small aerodynamic 

footprint. Additionally, the disturbances generated by the VAWTs may be able to be beneficially employed via 

proper turbine arrangement in order to increase the individual power coefficients of the VAWTs in the array. The 

resulting high power output per unit land area of a VAWT wind farm eliminates the need to build very tall turbines 

in order to reach the greater wind resources at higher altitudes. A VAWT wind farm therefore has less aesthetic, 

acoustic, radar, and environmental impacts than wind farms with other types of turbines. 

Our work has focused on creating a computationally simple potential flow model of the time-averaged flow around 

a single turbine. Using this model, we are able to rapidly optimize different arrangements of turbines in order to 

find one that maximizes the power output per unit land area of the total array. 

18

Enantioselective Friedel-Crafts Alkylation Reaction Substrate Scope 

Nadine P. Currie 

Mentors: Sarah Reisman and Lindsay Repka 

The Reisman laboratory is interested in the development of enantioselective Friedel-Crafts alkylation reactions to 

prepare unnatural tryptophan derivatives. Recently, Reisman and coworkers found that C(3)-substituted indoles 

and 2-amidoacrylates react in the presence of Tin(IV)Chloride and +(R)-BINOL to give pyrroloindolines with high 

enantioselectivity in a single step. The proposed mechanism for the formal (3+2) cycloaddition includes an 

important enantioselective protonation which increases the enantioselectivity of the pyrroloindolines formed. Based 

upon this hypothesis of the pyrroloindoline mechanism, it was anticipated that Friedel-Crafts alkylation reactions 

could be studied using other kinds of substrates. Since the Friedel-Crafts products only have one stereocenter, the 

protonation must be enantioselective. Thus, using a variety of C(2)-substituted indoles, the enantioselectivity, and 

therefore the reactivity, of the Friedel-Crafts alkylation reactions can be better understood. In order to carry out 

these enantioselective Friedel-Crafts reactions, certain C(2)-substituted indoles must be prepared first. Many of 

these substrates are difficult to prepare, for instance, some substrates form dimers, or decompose on silica gel. 

Many of the C(2)-substituted indoles have been shown to react with the acrylate to give Friedel-Crafts alkylation 

products with high enantioselectivity. 

List-Coloring Graphs Fixing a Particular Union 

Megan Cutrofello 

Mentor: Richard Wilson 

There are many known results regarding list-colorings of graphs. These results, however, only consider the size of 

each list used in the assignment; list-chromatic number does not take into account the total number of colors used 

in the assignment. This project fixes the union and asks what the consequences are when each list must have m 

elements and also there must be exactly n colors used in total. I call this the list-chromatic number of a graph G 

fixing n. Many questions arise from this definition. For example, is there always a graph G that has list-chromatic 

number m fixing n for any m

Small Crater Analysis of the Mars Science Laboratory Landing Site 

Mackenzie Day 

Mentor: John Grotzinger and Fred Calef 

Gale Crater was chosen as the landing site for the Mars Science Laboratory (MSL) Curiosity rover, due to arrive at 

Mars in early August, 2012. The landing ellipse contains a number of interesting features for exploration, 

documented by Anderson and Bell (2010). These include an alluvial fan marked by low thermal inertia, a more 

distal alluvial fan marked by high thermal inertia, and less well defined materials of the surrounding plains. We 

explored the stratigraphic relationships between the high and low thermal inertia alluvial fans and surrounding 

geologic units within the 20x25 km landing ellipse at Gale Crater by analyzing over 50,000 small craters (1-500 m 

diameter). With this data, we created a crater retention age map of the landing site based solely on crater 

observations to help calibrate the geologic map units delineated by Anderson and Bell (2010). By calculating crater 

retention ages, we estimated formation time of the various units to constrain the geologic history of the region. 

Comparing calculated ages with the morphology and stratigraphy, we inferred which surfaces are more indurated 

and the time of their formation. Information gathered in this project about the terrain of the landing site could help 

in planning the MSL Curiosity rover traverses and targeting strategies. 

Examining the Neurobiology of Facial Processing: A Pattern-Classification Based Approach 

Archy de Berker 

Mentors: Doris Tsao and Sebastian Moeller 

Primates have a dedicated neural system for facial recognition. This system contains a series of 'face patches' in 

the temporal and frontal lobes, identified with functional MRI (fMRI) and elucidated by microelectrode recordings 

from single cells. Single-cell analyses have shown neurons responsive only to faces and tuned to a variety of 

dimensions, allowing a population response to encode information such as the identity and orientation of the 

viewed face. These recordings also reveal qualitative differences in the representation of faces in the different 

patches; in some, the representation of identity appears to be view-specific, and others, view-invariant. This 

project applies classification algorithms to extant single-unit recordings and demonstrates that the identity and 

orientation of faces viewed by a Macaque monkey can be decoded from the population response of neurons in each 

face patch. We also confirmed that the patch AM (anterior-medial) encodes identity independent of view- a 

classifier trained on one view can decode identity from a response to a different view, a property not seen in the 

other patches. We then conducted an fMRI experiment, imaging the neural response to pictures of four individuals 

over five orientations, and applied similar classification techniques (multi-voxel pattern analysis, MVPA) to the 

scans. We again achieved decoding of orientation and identity, and demonstrated differences in decoding potential 

across the face patches, albeit with significantly lower classifier performance. Our results highlight the potential 

benefits and drawbacks of lower-resolution, non-invasive imaging techniques when attempting to functionally 

characterize interesting regions of the brain. 

Scanning Tunneling Spectroscopic Studies of Impurity Resonances in Epitaxial Films of Topological 

Insulator Bi2Se3 on Si(111) 

Charles de Bourcy 

Mentor: Nai-Chang Yeh 

Bi2Se3 is a three-dimensional topological insulator whose electronic properties are characterized by a large bulk 

bandgap and a single gapless Dirac cone (linear energy versus momentum dispersion relation for all directions of 

momentum) on the surface. Our group’s scanning tunneling spectroscopic studies of Bi2Se3 samples grown by 

molecular beam epitaxy on Si(111) substrate have revealed resonances in the local density of electronic states 

versus energy at isolated impurities. Prompt vanishing of the resonance with recovery of the Dirac cone as we 

move away from an impurity site illustrates the topological protection of the electronic surface state. Consistent 

with theoretical expectations for Dirac fermion tunneling spectra, the intensity of the impurity resonances 

decreases rapidly with decreasing sample thickness because of the increasing difference between Fermi energy and 

Dirac-point energy. Consistent with the expected breakdown of the Dirac-type dispersion relation in the thinsample 

limit due to coupling of boundary modes from the opposite surfaces, no impurity resonances are observed 

for samples thinner than 6 quintuple layers. To complete our study, we measure the precise Dirac energy in our 

samples by placing them in magnetic fields of different strengths and identifying the field-independent zeroth 

Landau level, and we fit theoretical simulations to the resonance peaks. 

Multi-Phase Coexistence in Fatty Acid Membrane Systems 

Anik Debnath 

Mentors: Jack Szostak, Itay Budin, and Rob Phillips 

Membranes composed of monoacyl lipids such as fatty acids, are putative models for primitive cell membranes. 

Compared to phospholipids, fatty acids feature significant aqueous solubility, which allow them to rapidly exchange 

between discrete membranes. These exchange processes have been proposed as a mechanism by which early cells 

could have competed with each other and evolved in a Darwinian manner. Here we ask if fatty acid exchange 

occurs purely via a solution phase of monomers, or also involves a coexisting micellar phase. We introduce a novel 

20

fluorescent assay based on the spectral shift of the fatty acid analogue Laurdan, which can distinguish between 

bilayer and micellar aggregates based on these structures’ head group solvation. We use this assay, along with 

surface tension measurements, to characterize the aggregate phase behavior of fatty acids as functions of 

concentration and pH. We find that fatty acid membranes do feature a coexisting micellar phase and that the 

equilibrium between these two aggregates is dependent both on enthalpic (head group protonation state) and 

entropic (the total fatty acid concentration) factors. We also use this assay to characterize the kinetics of the rapid 

pH-driven transition between micelles and vesicles. Our results indicate that fatty acid membranes exist as 

complex solutions of heterogenous aggregates in rapid equilibrium. 

Determining the Link Between Tortuosity and Recoverability in Vertically Aligned Carbon Nanotubes via 

Image Analysis 

Elizabeth R. Decolvenaere 

Mentors: Julia Greer and Shelby Hutchens 

Vertically aligned carbon nanotubes (VACNTs) are nominally vertically aligned arrays of carbon nanotubes (CNTs) 

grown perpendicular to a substrate on the order of tens of microns tall. Multiple different techniques exist for 

growing such structures, and many of these produce materials with markedly different microstructures. One aspect 

of this microstructure is the tortuosity or ‘curviness’ of the individual CNTs. This tortuosity is expected to play a 

role in both the intertube interactions and load-bearing capabilities, thus affecting the resulting mechanical 

properties of the structure. The mechanical properties of VACNTs are of great interest in applications related to 

energy absorption and dissipation (such as drop-shock cushioning for electronics). Therefore determining a link 

between observable physical geometry and mechanical properties, especially recoverability, is of great importance 

in the ultimate goal of tuning mechanical properties toward match specific design constraints. In this study, 

VACNTs grown by three different methods in three different labs were obtained and tested mechanically via 

nanoindentation using a flat-punch, and the results show highly variant properties across these three growth 

methods. Visual inspection reveals qualitative differences in the tortuosity of each type of sample. Using highmagnification 

scanning electron microscope (SEM) images, a semi-quantitative figure of merit for the tortuosity of 

the nanotubes in these structures is determined both by human analysis of the images and by an automated 

program. The results of each method are compared. 

Electrochemical Characterization of Model Thin-Film Metal Electrodes for Solid Acid Fuel Cells 

Rachel Deghuee and Jeffrey Kowalski 

Mentors: Sossina M. Haile and Aron Varga 

To date, platinum is the catalyst of choice in intermediate temperature fuel cells including solid acid fuel cells 

(SAFCs), due to its compatibility with the solid acid electrolyte (CsH2PO4) and high catalytic activity. We can infer 

mechanistic pathways and extract fundamental rate parameters for hydrogen electro-oxidation and oxygen electroreduction 

via AC impedance spectroscopy measurements of electrode performance of model thin film electrodes. 

Thin films were deposited via sputtering and metal organic chemical vapor deposition (MOCVD) with varying 

thicknesses. In addition, asymmetric cells were fabricated with high performance counter electrodes and small 

working electrodes by masking the electrolyte with Teflon tape. A good isolation between the electrodes allows for 

the assumption that all measured electrode resistance is due to the reaction at the working electrode. Both the 

forward and backward electrochemical reactions were sampled via biased measurements. The thin film morphology 

and microstructure were characterized via SEM, EDS, and XRD in order to correlate with the specific catalytic 

activity. 

The Hopf Algebra of Rooted Trees With Natural Growth by Fan Graphs 

Colleen R. Delaney 

Mentor: Susama Agarwala 

In Connes and Moscovici (1998), the authors define a Hopf algebra of formal vector fields on the frame bundle of 

an n-dimensional manifold. In the dimension one case, this Hopf algebra on formal vector fields is closely related to 

the Hopf algebra of rooted trees used in the renormalization calculations of quantum field theory. We present a 

new Hopf algebra that generalizes the Hopf algebra of rooted trees under growth by one vertex to allow growth by 

families of fan graphs of any number of vertices. 

We identify the Hopf algebra from both the perspective of the rooted trees and the formal vector fields. This new 

Hopf algebra can be studied using noncommutative geometry. 

21

Characterizing and Investigating Spatio-Temporal Expression of Genes Critical to Strongylocentrotus 

purpuratus Coelomic Pouch Development in Eucidaris tribuloides, a Distantly Related Cidaroid Sea 

Urchin 

Neal Desai 

Mentors: Eric H. Davidson and Eric Erkenbrack 

The Hedgehog signaling pathway plays an important role in coelomic pouch development during embryogenesis of 

the purple sea urchin Strongylocentrotus purpuratus. To determine its role within the echinoderm class Echinoidea, 

we aim to study this pathway in a distantly related sea urchin, the pencil urchin Eucidaris tribuloides. We are 

investigating where and when genes in this pathway are expressed and how those expression patterns change 

when cultured in the presence of cyclopamine, a chemical inhibitor that binds to and stabilizes the ligand of the 

Hedgehog signaling pathway. Spatio-temporal expression profiles were obtained for the genes of interest using the 

quantitative polymerase chain reaction (qPCR) and whole-mount in situ hybridization (WMISH). 

Eclipsing Binary White Dwarfs in Archival Palomar Transient Factory Data 

Suhail Dhawan 

Mentors: Shrinivas Kulkarni and Shriharsh Tendulkar 

This project aims to find binary systems of white dwarfs (WDs) in which the primary and the companion eclipse 

each other. These are of great interest to astronomers since they offer accurate mass-radius relations and can be 

used as chronometers for galaxies. Such systems are exciting because their population distribution can help us 

determine whether or not they are progenitors of type Ia supernovae. Binaries with short orbital periods can be 

evidence for gravitational radiation and plausible candidates for space based gravity wave detectors. In this project 

we used Palomar Transient factory (PTF) photometry cross-matched with SDSS white dwarfs and looked for 

candidates of eclipsing binary systems. We found three white dwarf M-Dwarf binary systems that are possibly 

eclipsing and three variable WD-WD binaries. We also found candidates for pulsating WDs with DQ, DZ and DC 

spectral classification. 

Performance Optimization of Large-Area Bulk Heterojunction Solar Cells 

Marino Di Franco 

Mentors: Kwanghee Lee and Sooncheol Kwon 

Bulk heterojunction solar cells are easily fabricated layer-by-layer from solution, with the commonly available 

polymer poly(3-hexylthiophene) (P3HT) and fullerene [6,6]-phenyl-C61 butyric acid methyl ester (PC60BM) used as 

electron donor and acceptor in the device’s active layer. However, scaling the dimensions of these devices becomes 

problematic as the large sheet resistance of the indium-tin-oxide (ITO) electrode is rather high, leading to drops in 

the power conversion efficiency of the solar cell. Thus, a module based approach is taken to minimize resistive 

loses in each substrate, and fabrication conditions were optimized such that the solvent for deposition of the active 

layer, the thickness of the active layer and the time and temperature of thermal annealing led to the highest 

attainable device performance. 

Magnetostratigraphy of the Late Cretaceous Rocks in the James Ross Basin, Antarctica, to Investigate 

the K/Pg Mass Extinction 

Matthew Ross Diamond 

Mentors: Joseph Kirschvink, Tim Raub, and Sarah Slotznick 

During the end of the Cretaceous, Antarctica was a place of rapid sediment collection, giving rise to high resolution 

stratigraphy. Taking advantage of this opportunity to study the Cretaceous-Paleogene Mass Extinction that 

occurred globally at the end of the Mesozoic, a scientific team with researchers from Caltech, University of 

Washington, University of Buenos Aires, and Laboratorio de Geología Andina has worked since 2008 to recover 

fossils and rock samples of this carbonate sequence. Both block and core samples were collected to be analyzed in 

the Paleomagnetics lab at Caltech. Using the documented Maastrichtian magnetic reversals known from other parts 

of the world, we are using a 2G SQUID magnetometer to study the paleomagnetic directions of the cores to 

correlate them with specific magnetic chrons, helping pin down the stratigraphic carbonate sequence to a global 

timeline. The steps for uncovering the magnetic direction during lithification including liquid nitrogen cooling, 

running alternating field (AF) demagnetization, and baking samples by 20 degree intervals up to 260 degrees 

Celsius. This analysis is important for understanding the biostratigraphic and geochemical data that the researchers 

at the University of Washington are studying. Geologists are currently interested in knowing when individual 

species went extinct at this time and understanding the global environmental conditions leading up to the Cenozoic 

in order to investigate the factor leading up to the K/Pg mass extinction. While many scientists link the volcanic 

eruptions of the Deccan traps or the Chicxulub meteor impact to this mass extinction, we still do not understand 

the kill mechanism that caused 75% of the species on Earth to die. Knowing the paleomagnetic directions of these 

carbonates will allow us to piece together information in a relative time frame to better interpret ongoing and 

future findings. 

22

Localization of Lipid Biosynthesis Enzymes in the Asymmetrically Dividing Bacterium 

Rhodopseudomonas palustris Utilizing Fluorescent-Tagged Proteins 

Michael Dieterle 

Mentors: Dianne K. Newman, David M. Doughty, and Gargi Kulkarni 

Hopanoids are triterpenoid lipids that are structurally very similar to sterols, their proposed eukaryotic 

counterparts. Fossils of triterpenoid lipids called hopanes are preserved in ancient rocks, some of which are as old 

as 2.7 billion years. Hopanoids are produced by both aerobic and anaerobic bacteria, such as Rhodopseudomonas 

palustris. While the role of these lipids is not understood, there are indications that they may play a role in 

asymmetrical cell division, membrane permeability and stability, and protection of the cells under stress. Based on 

earlier work, the Newman Lab has proposed a hopanoid biosynthetic pathway containing several genes associated 

with the production and transportation of hopanoids in Rhodopseudomonas palustris. Fluorescent proteins such as 

mCherry will be used to tag these biosynthetic proteins to determine their intracellular localization, shedding light 

on where hopanoids are produced and where they are integrated into the cell membranes (outer, cytoplasmic or 

inner-cytoplasmic) during the cell cycle. The localization patterns of the different biosynthetic proteins range from 

uniform throughout the cell to dense points near the poles of the wild type cell. If these patterns are affected for 

each protein by the absence of other proteins in the pathway, this suggests that the proteins may be forming 

complexes. 

Expression of Neuromedin B in the Anterior Cingulate Cortex and Fronto-Insular Cortex 

Monisha Dilip 

Mentors: John Allman and Nicole Tetreault 

Neuromedin B (NMB) is a peptide found in the gastrointestinal tract as well as in the central nervous system in 

humans. It works to regulate feeding, the release of thyroid stimulating hormone, as well as internal body 

temperatures. Both the Anterior Cingulate Cortex (ACC) and Fronto-insular Cortex (FI) are regions of the brain that 

have previously been found to have an effect on appetite regulation pathways. More specifically, Von Economo 

Neurons (VENs) are found in both the FI and the ACC and appear to carry out these pathways. Thus, NMB 

expression was monitored in the VENs in the FI of multiple human brains from Mount Sinai. No concrete results 

were observed in the SURF time period, but the VENs themselves were counted in brains from Chicago, so that 

VENs could be both mapped out within the brains as well as quantified. 

Identification and Isolation of Novel Long Noncoding RNAs Involved in Myogenesis 

Race DiLoreto 

Mentors: Barbara Wold and Georgi Marinov 

Advances in functional genomics have allowed the discovery of a number of interesting classes of noncoding RNAs. 

Long noncoding RNAs (lncRNAs) have become a focus of study because several lncRNAs have been identified as 

playing a role in biological processes, such as X chromosome inactivation and cell cycle regulation. I am using 

computational methods in to identify potential novel lncRNAs involved in muscle differentiation by evaluating 

conservation and protein coding capacity in novel transcripts sequenced from polyA selected myogenic mouse cell 

lines. I am also working to isolate several lncRNAs adjacent to MyoD, the master regulator of myogenesis, which 

have a putative role in regulation of myogenesis. Once isolated, I can clone these transcripts and use the 

constructs to assay biological function of the lncRNAs. 

Test Setup to Measure Properties of the Crystalline Lens 

Hyung Wan Do 

Mentors: Yu-Chong Tai and Charles DeBoer 

Presbyopia is a condition in which the eye’s ability to focus on near objects is weakened due to the natural course 

of aging. Accommodation is the process by which the eye changes its optical power in order to focus on objects of 

different distances. Change in the optical power is accomplished by adjusting the curvature of the lens with a ring 

shaped ciliary muscle which surrounds the lens. It is known that presbyopia is most affected by the accommodation 

ability of the lens. In order to measure accommodation in an enucleated eye, a test setup with high repeatability 

and accuracy was built. An intraocular lens was tested to calibrate the system and response of the natural lens to 

stretching inside the capsule was observed. In particular, emphasis was placed on in vitro measurement of 

accommodation of porcine lenses as the ciliary muscle is stretched. Eventually, this could lead to better 

understanding of accommodation of the eye in relation to presbyopia. 

Saturn Ring Data Analysis and Thermal Modeling 

Coleman Dobson 

Mentors: Linda Spilker and Estelle Deau 

CIRS, VIMS, and ISS, (Cassini’s Composite Infrared Spectrometer, Visual and Infrared Mapping Spectrometer, and 

Imaging Science Subsystem, respectively), have each operated in a multidimensional observation space and have 

acquired scans of the lit and unlit rings at multiple phase angles. To better understand physical and dynamical ring 

23

particle parametric dependence, we co-register profiles from these three instruments, taken at a wide range of 

wavelengths, from ultraviolet through the thermal infrared, to associate changes in ring particle temperature with 

changes in observed brightness, specifically with albedos inferred by ISS and VIMS. From this compilation of 

multiple wavelength data, we construct and fit phase curves using independent dynamical thermal models for ring 

structure. From these fits we test our dynamical thermal models; and from the phase curves we derive ring 

albedos, and derive properties of the ring particle surfaces. 

Analysis of Plant Wax δD to Evaluate the Conservation of Rainfall δD in Higher Plant Fatty Acids in the 

Indonesian Archipelago 

Sabrina Dodgin 

Mentors: Alex Sessions and Eva Niedermeyer 

A key part of any paleoclimate study is understanding the regional hydrologic. This cycle is reflected in the 

hydrogen isotopic composition (δD) of environmental water and this δD value is recorded in higher plant fatty 

acids. However, how the δD values are recorded in the plant waxes vary. My research focuses on comparing the 

δD values of marine-surface sediments’ fatty acids extracts taken from the coasts of Sumatra and Java with the 

δD values of the rainfall in the same area. Leaf waxes are extracted and separated; their fatty acids are stabilized, 

identified using mass spectrometry and analyzed using isotope-ratio mass spectrometry. The δD values are then 

compared to the δD values of rainfall in their respective areas. The results may be used in a paleoclimate study 

and will further enable a study of terrestrial plant waxes in the area. A study of terrestrial plant waxes will give 

information about the preceding conditions of the hydrologic cycle and past climate changes of the Indonesian 

Archipelago. 

Performance of Linear Error Correcting Codes With Anytime Reliability 

Yishun Dong 

Mentors: Babak Hassibi and Ravi Teja Sukhavasi 

New upper bounds on the error probability of casual linear codes under maximum-likelihood (ML) decoding are 

proposed. Three types of channels considered here are the additive white Gaussian noise (AWGN) channel, the 

binary symmetric channel (BSC), and the binary erasure channel (BEC). The new bounds have shown to widen the 

ranges of rates for which the union bound analysis applies for all three channels. For the AWGN channel, the 

Divsalar bound for block codes is adapted to the casual coding scheme using the distance spectrum of code 

obtained from the Toeplitz ensemble. It is shown that using the Divsalar bound, the error exponent can be 

improved over the union bound exponent at all rates and all channel signal-to-noise ratios (SNR). For the BSC and 

BEC, similar techniques are used to give tighter upper bounds on the error probability, and it is shown that the 

error exponents can be improved at rates that are higher than some critical rate. 

Due to the complexity of ML decoder over BSC, practical decoder needs to be considered. In particular, linear 

programming (LP) decoder by Feldman is explored. It is known that LP decoder gives error exponent for block 

codes. We show that Toeplitz parity check matrix with banded or log-log sparsity cannot be anytime reliable. 

Furthermore, extensive simulations have been done with log sparse Toeplitz parity check matrix. Results have 

shown that this particular matrix structure gives relatively poor performance. Hence, we search for alternative 

parity check matrix structures that are non-Toeplitz. 

Deterministically Reorganizing DNA on a Nanoscale DNA Surface Using Random Walking 

Shayan Doroudi 

Mentors: Erik Winfree, Lulu Qian, Niranjan Srinivas, and Damien Woods 

Aside from its essential role as the carrier of genetic information in all forms of life, DNA has proven itself as a 

powerful tool for creating nanoscale structures and systems in the past few decades. We (Zibo Chen, Gregory Izatt, 

Yae Lim Lee, Sarah Wittman, and I) propose one such system built entirely out of DNA where a single strand of 

DNA takes an unbiased random walk on a DNA origami surface, picks up cargo molecules it encounters, and 

deposits those molecules at predetermined locations on the surface. In this fashion, we can add order to a 

seemingly random system, fueled only by the addition and exchange of base pairs between strands. To verify that 

the nanoscale mechanisms work as planned, we will primarily rely on (a) atomic force microscopy, which can 

indicate the position of certain strands at different points in time, and (b) fluorescence measurements using a 

spectrofluorometer, which shows the progression of reactions over time. While our system only uses one type of 

cargo, it can theoretically be scaled to reorganizing a surface with many different types of cargoes, moving each 

cargo to its own destination. 

24

Ensuring Socially Beneficial Equilibration in Competitive Loan and Insurance Markets: Theory and 

Experiment 

Kathrin Eichinger 

Mentor: Peter Bossaerts 

Asymmetric information, where one side of the market knows more than the other, presents a significant issue in 

economics, as for a competitive market to generate optimal resource allocations for sure, it must not exist. 

Economic theory has sharp predictions for equilibria in these markets where imperfect information is present, such 

as insurance and bank lending, but they may not necessarily hold true. This project puts the theory to the test by 

recreating a loan market environment in the laboratory, where lenders offer loan contracts (price and interest rate) 

and borrowers may accept them in order to fund their projects. As there are borrowers with higher and lower 

chances of payoff and without any influence over which type of borrower will accept their contracts, lenders need to 

rely solely on the terms of the contracts to distinguish between them in order to make profit. Observing public 

behaviour in these markets provides information about equilibration dynamics, where nothing is known about the 

dynamics in a setting where convergence involves both prices and contract specifications. 

Graph Theoretic Approaches to the Static Membership Problem 

Sam Elder 


The static membership problem is concerned with the following question: How many bits of memory are required 

to store a set S from a finite universe U such that membership queries of the form, “Is x in S?” can be answered 

with a small number of probes to individual bits in the memory? For example, one may wish to encode the 

correctly spelled words of a given language from the universe of letter strings of a fixed length, and one wishes to 

easily ask whether a given string is a word. We hope to understand the asymptotic dependence of the size of 

required memory on the size of U. In this work, we investigate the static membership problem for particular cases 

when the size of S and the number of probes allowed are both small, hoping to extend the results of Alon and Feige 

(2008) and Radhakrishnan, Shah, and Shannigrahi (2010). These cases lend themselves to a graph theoretic 

interpretation, corresponding in particular to questions about colored bipartite graphs and hypergraphs. 

Synthesizing a Functional Pseudomonas aeruginosa Methionyl-tRNA Synthetase Mutant Recognizing 

Azidonorleucine 

Clara Eng 

Mentors: David Tirrell and Brett Babin 

BONCAT (bio-orthogonal non-canonical amino acid tagging) is a cell-specific protein labeling technique that utilizes 

functionalized non-canonical amino acids to enrich proteins produced by a subpopulation of interest in a mixed cell 

population. This project studies the feasibility of extending the functionality of the mutations L13N, Y260L, and 

H301L, used to engineer an Escherichia coli methionyl-tRNA mutant (NLL-MetRS) that selectively incorporates the 

methionine analog azidonorleucine (Anl) instead of methionine, to other biological systems. Anl contains an azide 

group that reacts with alkynyl probes, facilitating isolation of tagged proteins via affinity purification, analysis of 

cell lysates using a protein gel, and visualization of whole cells with fluorescence microscopy. The corresponding P. 

aeruginosa NLL-MetRS mutant was engineered and was demonstrated to selectively incorporate Anl, though with a 

lower efficiency than that of the E. coli NLL-MetRS. This is an important first step toward broadening the range of 

NLL-MetRS mutants as a tool for cell-specific labeling. 

The Neutron Electric Dipole Moment Experiment 

Joshua Escribano-Fontanet 

Mentors: Brad Filippone and Adrián Pérez Galván 

Studies of magnetic fluctuations of the synchrotron laboratory at the California Institute of Technology with and 

without magnetic shielding from high permeability materials show that to increase the magnetic shielding it is 

necessary to use a superconducting shield. For this purpose we started work on the optimization design of a large 

superconducting shield made out of lead. The superconducting shield belongs to a half scale magnet package that 

is a prototype for a new search of the electric dipole moment of the neutron to be carried out at Oak Ridge National 

Laboratory. The magnet package will provide the magnetic fields for the experiment. We present studies of the 

cooling rates of the half scale apparatus in temperatures between 300K to 5K. We show that by reducing the 

masses and improving the thermal connections of the superconducting shield we increase the cooling rates and 

decrease the amount of cryogenic liquid needed to cool down the lead shield as compared to a previous version of 

the apparatus. 

25

Low-Energy Limits of Noncommutative Calabi-Yau Manifolds 

Chris Estrada 


In this project, we consider algebraic formulations of Calabi-Yau manifolds that allow us to move into 

noncommutative territory, with the goal of constructing a spectral triple for a Calabi-Yau algebra, and interpreting 

the geometric model of Chamseddine, Connes, and Marcolli as a low-energy limit of a noncommutative Calabi-Yau 

manifold. In particular, we focus on a similar construction for an associative algebra obtained as a quotient of a 

path algebra for a quiver, and use it together with an embedding of the path algebra into the graph C*-algebra in 

order to obtain an algebra representation on a Hilbert space, and thus a spectral triple. 

Development and Testing of a Multifunctional Computer Fan Wind Tunnel for Wind Energy Applications 

Yuyang Fan 

Mentors: Morteza Gharib and Daegyoum Kim 

It is important to study non-uniform wind velocity profiles such as gust and wind shear to better understand flow 

physics and to develop new wind energy technologies. However, lacking of appropriate facilities that can generate 

realistic wind profiles in laboratory environment makes it hard to study the energy extraction from the wind. In 

order to overcome this problem, an inexpensive multifunctional wind tunnel with a test section area of 16 ft 2 

(4ft x 4ft) has been developed. This tunnel uses a 10 x 10 array of powerful computer fans to achieve different 

wind velocity profiles by controlling the speed of different rows of fans using pulse width modulation (PWM) 

through a computer. 

Telemedicine and the 10-Cent Checkup 

Alex Fandrianto 

Mentors: K. Mani Chandy and Julian J. Bunn 

We construct the framework for a remote medical checkup by developing an otoscope, stethoscope, and 

electrocardiogram (ECG) device. The otoscope and stethoscope devices are adapted to Nexus One Android phones. 

The otoscope casing covers the phone’s camera and contains a small lens and fiberoptic cable to shine the phone’s 

camera light out of its aperture. Acoustic signals sensed by the stethoscope head are converted to electrical signals 

by a microphone in a casing attached to the stethoscope; the electrical signals are fed to the microphone input of a 

phone or computer. ECG recordings from the 3-electrode ECG are taken by a laptop. Recordings are saved locally 

with associated medical metadata and can be uploaded to the Amazon Web Server (AWS) where they can be 

viewed securely from a browser. The ECG data is contaminated with 60 Hz power line noise; various filters are 

applied to improve the signal-to-noise ratio (SNR). The filtered data can then be used to detect pulse via 

triggering. The system performance of each device is determined by comparing its recordings against the medical 

ideal. 

Enhancing Performance of GaAs Solar Cells at High Incident Angles With Plasmonic Surface Structures 

Alta Fang 

Mentors: Harry Atwater and Ragip Palas 

Improving the performance of solar cells at large incident angles is valuable because it would allow solar cells to 

produce the same amount of electricity throughout the day as the sun travels across the sky, rendering mechanical 

solar-tracking mechanisms unnecessary. By adding metal or dielectric nanostructures to the surface of thin film 

GaAs solar cells, the solar cells' performance can be improved due to near-field absorption enhancements, 

waveguiding effects, and anti-reflection-coating effects. Finite-difference time-domain (FDTD) simulations are run 

to find the surface structure parameters that maximize light absorption in GaAs films at large angles. The optimal 

structures are then fabricated and tested in order to experimentally verify the simulation results and demonstrate 

the absorption enhancements. 

Thermal Bubble Generation for Flexible Microfluidics Control 

Kelvin C. Fang 

Mentors: Changhuei Yang, S. Pang, and C. Han 

Although the incredible accuracy of modern silicon fabrication processes allows for much flexibility in microfluidic 

channel design, the difficulty in post-fabrication manipulation of the channel currently limits microfluidic chips to 

relatively static operations. To provide runtime control over microfluidic chips, the Yang Biophotonics group is 

exploring the utilization of laser-generated channel bubbles as flexible fluid actuators. By selectively heating 

resistive metal pads underlying the channel, we produce adjustable microvalves and micropumps during chip 

operation. With the development of an integrated laser sensor and focusing system, we provide complete discretion 

in bubble placement and growth. This technology enhances the capabilities of present and future chip designs and 

creates numerous possibilities in microfluidics research and applications. 

26

Study of Hypervelocity Impacts Using Large-Scale, Long-Term Molecular Dynamics 

Chi Feng 

Mentors: William A Goddard III and Andres Jaramillo-Botero 

In the hypervelocity impact regime, where the impact velocity is tens of kilometers per second, the inertial stresses 

of the projectile and target are very large compared to the yield stress of both target and projectile materials which 

causes parts of the configuration to undergo solid-liquid-vapor phase transitions, ionization, plasma formation, and 

hydrodynamic instabilities. We use the electron force field (eFF) molecular dynamics simulation method developed 

at the Goddard Group at Caltech to model hypervelocity impact. This approach uses a mixed quantum-classical 

approach that enables us to track the long-term dynamics of large-scale electron excitations which overcomes the 

scale limitations of QM methods and the adiabatic nature of conventional force field methods. These simulations 

should reveal the effect of impact energy on the physics of hypervelocity impact with the ultimate goal of providing 

experimental observables such as plasma conductivity measurements for hypervelocity impact. 

Laser Desorption Infrared Spectroscopy for Icy Moon Surfaces 

Nate Figlewski 

Mentor: Luther W. Beegle 

The mechanics of Laser Desorption are constantly being researched and updated. Mass Spectroscopy is closely tied 

with Laser Desorption. Higher energy desorption techniques, such as Laser Induced Breakdown Spectroscopy 

(LIBS), and Matrix Assisted Laser Desorption/Ionization (MALDI) have been used to analyze the elemental and 

chemical makeup of extraterrestrial surfaces. However, processes like LIBS restrict the spectroscopic process to 

reviewing elements, as more complex compounds are broken apart during the desorption process. Laser 

Desorption Infrared Spectroscopy (LD-IR) takes advantage of water’s high absorption levels in the mid IR range, 

requiring less energy, thus simultaneously reducing the energy cost on a mission as well as allowing larger 

compounds to remain intact. First, this experiment will find the peak absorption wavelength of water using an 

Opotek Opolette 3034 tunable IR OPO. Then, using the wavelength found, we will find the relationship between 

pulse energy and quantity of material desorbed. 

Search for Blazars by Detection of Optical Transients Near Gamma Ray Sources 

Sveinbjörn Finnsson 

Mentors: S. George Djorgovski and Ashish Mahabal 

Blazars are active galactic nuclei (AGN), powered by material that falls onto a supermassive black hole. This 

material forms an accretion disc close to the black hole which generates energy. An AGN is defined as a blazar 

when the accretion disc produces a pair of relativistic jets and one of them is pointed towards the Earth. The jets 

are fast outflows of energetic plasma, perpendicular to the accretion disc. The energy ranges in form, from radio to 

gamma rays. We searched for blazars by looking for objects that show optical variability near known gamma ray 

sources with no known optical counterparts. The gamma ray sources were detected by the Fermi satellite and fields 

near them explored. The light curves (plots of magnitude of brightness vs. time) and images of objects in these 

fields were analyzed to identify variable objects. This data came from the Palomar Quest (PQ) and the Catalina 

Real-Time Transient Survey (CRTS). Interesting objects were then further inspected by referring to the NASA/IPAC 

Extragalactic Database (NED) to see if there was a known radio or x-ray source near any of them. The result is a 

list of blazar candidates that hopefully will later be identified as such. 

Microstructures and Liquidus Projection of Ternary Ag-Pb-Te System 

Wei Jian Foo 

Mentors: G. Jeffrey Snyder and Hsin-Jay Wu 

Thirty Ag-Pb-Te alloys were prepared, and their primary solidification phases were analyzed. The projection of 

liquidus troughs of the ternary Ag-Pb-Te system were determined using phase diagrams of the three constituent 

binary systems and experimental results. There are six primary solidification phase regions. In total, there exist 

three terminal solid solution phases and two binary compounds, with no ternary compound. Unique microstructures 

were observed during precipitation of PbTe in Ag2Te primary solidification phase. High Seebeck coefficients were 

also observed along the primary solidification phase boundaries between PbTe and Te. 

Neural Representations of Value and Loss: Prediction Error, Contrast Effects, and Loss Aversion 

Suzannah Fraker 

Mentor: John O’Doherty 

How people learn causal relationships is a central question of behavioral psychology. When an outcome is 

unexpected, the contrast between what was expected and reality is called a prediction error. It is believed that 

associations between predictive cues and their results are strengthened with each prediction error. We have 

studied the effects of contrasts between expectations and outcomes through multiple avenues. First, we have 

looked at the hypothesis that dopamine release encodes prediction error in the brain and is responsible for forming 

associations. This would imply that inducing extrinsic dopamine release using drugs of abuse, for example, would 

27

cause learning to continue even after prediction error has been reduced to zero and thus blocking would be 

inhibited. We test this in nicotine addicted subjects learning to predict receipt of nicotine and natural rewards 

(juice). Second, we looked at the effect of reference frames on people’s willingness to take risks or exert effort for 

rewards of various magnitudes. We hypothesize that decreases in reward magnitude are perceived as losses, and 

thus that subjects who are more averse to losing money will also show greater drops in effort level when faced with 

a negative reward contrast. 

Multiphase Flow in the Presence of Hydrophilic and Hydrophobic Surfaces 

Joshua W. Fromm 


Effective simulation of water droplets in multiphase environments is important to understand the countless 

applications of multiphase flow; this study, inspired by the hope of improving reactant transfer in fuel cells, 

examines the methods used in multiphase simulations and the implications of those simulations. Direct numerical 

simulations using the NGA code are used in a wide range of environments with varying parameters. Due to errors 

in NGA, multiphase flow initially was unable to be properly simulated. This led to the implementation and 

modification of different fluid front tracking methods to optimize multiphase flow simulation. Comparisons of 

simulated dynamic contact angles to the theoretical mathematical value that should be obtained using a given set 

of parameters are also carried out. Once multiphase flow is properly simulated, the movement of water through 

channels with hydrophilic and hydrophobic surfaces is compared to the movement of water through non polar 

channels and the results are considered with respect to fuel cell applications. 

Effect of Surface Chemistry on Trapped Air Phenomenon in Flat Plate Impacts 

Matthew K. Fu 

Mentor: Morteza Gharib 

A study was conducted to better understand the dynamics of hull slamming on the water free surface. A novel 

Slingshot Impact Testing System (SITS) was developed to simulate hull impacts by slamming flat plates at various 

impact angles and velocities. The loads on the plates was observed and used to determine the maximum force of 

the impact. Previous studies have shown that air can become trapped between the plates and water at small 

deadrise angles (deadrise angles less than 5 O ) resulting in a cushioning effect. This study sought to characterize 

the overall effect of surface chemistry on the presence of the trapped air, specifically, if hydrophobic plate surfaces 

can help stabilize the trapped air beneath the plate and reduce the overall impact force. Various hydrophobic and 

hydrophilic surfaces were attached to composite plates and slammed into the free surface at impact velocities 

between 4m/s and 5 m/s and deadrise angles between 0 o and 5 o . While preliminary results show some differences 

in the size of cushioning effect experienced by the hydrophilic and hydrophobic plates, further analysis and testing 

is needed. 

Building Vectors to Test the Roles of miR-146a Targets Traf6, Irak1, and Stat1 in HSC Biology 

Prakriti Gaba 

Mentors: David Baltimore, Ryan M. O’Connell, and Jimmy Zhao 

The immune system is an intricate system that aims to destroy any harmful invaders that enter an organism. An 

effective immune response typically involves a process called inflammation, which coordinates the recruitment of 

the proper immune cell types to eradicate infection. MicroRNAs (miRNAs) are small, non-coding RNAs that have 

recently been shown to be important players in the immune system. miRNAs function by targeting mRNAs leading 

to repression of their expression. Regulation of miRNA expression and function is controlled by transcription factors 

that regulate the production of miRNA containing primary transcripts in specific cell types during development or in 

response to various environmental signals. Specifically, we will be studying the transcription of a particular miRNA, 

miR-146a, which is present in immune cells and upregulated further in response to inflammatory stimuli such as 

Toll-like receptor ligands or pro-inflammatory cytokines. Hematopoietic stem cells (HSCs) give rise to all blood cell 

types, including those found in the immune system. They are located primarily in the bone marrow. In prior 

studies, mice that are lacking in ARS2, which contributes to pri-miRNA processing, have bone marrow failure that is 

thought to be due to defective HSC function. This experiment suggested that the miRNA pathway is probably 

important in HSC function. However, the identity of the responsive miRNAs remains elusive. Now, members of the 

Baltimore lab have found that miR-146a functions to promote proper HSC function and may be involved in how 

HSCs respond to inflammatory stress conditions. However, the mechanistic basis for miR-146a’s function in HSCs is 

presently unclear. Here, we focus on studying whether specific targets of miR-146a, including IRAK1, TRAF6, and 

STAT1, contribute to its biology in HSCs before and after inflammation, and assessing whether this occurs in a 

combinatorial manner. Our aim is to create retroviral vectors that produce siRNAs against one, two, or all three 

targets of miR-146a and determine which combination can rescue the miR-146a HSC phenotype. Our current 

results show that we are able to successfully knock down STAT1, TRAF6, and IRAK1 using the newly designed 

siRNAs. 

28

Implicit and Explicit Time-Stepping Fourier Continuation Solvers for the Burgers’ Equation 

Emmanuel Garza Gonzalez 

Mentors: Oscar Bruno and Edwin Jimenez 

Burgers’ equation is sometimes used as a simplified model of the incompressible Navier-Stokes equations to test 

numerical schemes. Since there is no general numerical stability theory for nonlinear problems, the quadratic 

convective term that appears in Burgers’ equation presents a unique challenge to formulate a robust numerical 

method. Solvers for the Burgers’ equation were implemented by using several time step schemes and a novel 

methodology referred as Fourier Continuation to approximate the spatial derivatives with high order accuracy. The 

diffusive term was treated implicitly since it would otherwise impose severe stability constraints. On the other 

hand, the convective term was treated with a variety of explicit and implicit approaches. For the 1D case a solver 

that is second order accurate in time was implemented which does not have an implicit quadratic term as in the 

case of a direct application of the Crank-Nicolson scheme, leading to a variable coefficients ODE instead of a 

nonlinear ODE. A 2D system of Burgers’ equations was solved by using Fourier Continuation along with the 

Alternating Direction Implicit methodology with different formulations for the convective term. The stability 

properties of these schemes are analyzed and the results suggest numerical formulations for the nonlinear 

convective term that yield robust solvers for similar nonlinear PDEs. 

Micro Measurements of 3-D Deformations Using Digital Image Correlation 

Alexandra Gdoutou 

Mentors: G. Ravichandran and Shuman Xia 

Digital Image Correlation (DIC) provides a full-field non-contact optical method for the accurate measurement of 

displacements and, therefore, strains during deformation of materials, devices and structures. The method is based 

on the comparison of speckle images taken before and after deformation. In the present work, we have developed 

an innovative 3D-DIC technique, which utilizes a transmission grating and a single camera, and is capable of 3D 

displacement measurement with a 2D-DIC algorithm. The accuracy and validity of the method has been 

demonstrated through measurement of deformations of a thin inflated membrane. The pressure chamber for 

performing the inflation of the membrane was designed and fabricated and the experimental setup was assembled. 

Calibration and translation experiments were performed to establish the resolution of the technique and effect of 

various speckle patterns on the measurements were established. Membrane inflation experiments were performed 

and the digital images were acquired. The acquired digital images were analyzed using a 2D-DIC software (VIC-2D, 

Correlated Solutions, Inc.) to obtain the 3-D displacement data that was compared favorably with the results from 

finite element simulations. 

Symmetries of Unrooted Polygons and Multiple Logarithms 

Meng Ge 

Mentor: Susama Agarwala 

Multiple polylogarithms are important algebraic objects. They have applications in number theory (mixed –Tate 

motives) and they also apply in the calculations of Feynman diagrams in physics. Previous papers have calculated 

dihedral group actions on rooted polygons, namely, σ which rotates the index of the sides of the polygon by 1, and 

which is the action of reflection. In this paper, we study the effect of transposition, denoted as on rooted 

polygons. We know that for a three-gon Q, Q= Q. Using induction, we can relate the action of 

polygon, which is defined in previous published papers, to that of its subpolygons. 

of a 

Conditional RNA Interference in a Mammalian Cell-Free System 

Tianjia J. Ge 

Mentors: Niles Pierce, Lisa Hochrein, and Ma’ayan Schwarzkopf 

RNA interference (RNAi) allows for specific downregulation of almost any endogenous gene. However, gene 

silencing by traditional, constitutive RNAi can lead to undesired effects when studying essential genes or cellspecific 

silencing. Conditional RNAi would regulate gene silencing, having applications ranging from studying gene 

function to therapies for human disease. Here we report a system of small engineered RNA molecules that execute 

conditional RNAi, in which gene A is silenced only in the presence of mRNA from gene B. The mRNA from gene B 

triggers a conformational change in the RNA molecules, leading to RNAi-mediated silencing of gene A. The system 

is designed for complete sequence independence between genes A and B, allowing its application to any 

combination of genes. We demonstrate the conditional RNAi system in HEK293 cell lysate, which should reflect in 

vivo conditions, with the intent of observing conditional siRNA production in the cell-free extract. Insights into the 

system from this study will hopefully help with an in vivo implementation of conditional RNAi. 

29

Exploring Optimal Electromagnetic Follow-Up Campaigns for Inspiralling Neutron Star Binaries 

Detected Using Ground-Based Gravitational-Wave Detectors 

Alexandra Georgieva 

Mentors: Tom Prince and Samaya Nissanke 

Merging neutron star binary systems are amongst the most promising sources for the network of gravitationalwave 

(GW) ground-based interferometric observatories. Observing an electromagnetic (EM) counterpart will 

provide complementary information which will help constrain further the physics of these sources. Possible EM 

counterparts are short hard bursts as well as optically faint short-lived transients (powered by radioactive decay 

ejecta) or longer-lived radio transients. Therefore, sky location and distance measurements (including their 

associated errors) from the GW signal are crucial when planning any EM follow-up campaign. We assume that the 

underlying distribution of neutron star binaries is such that: i) their density is uniform in constant comoving volume 

or ii) they are associated with a galaxy and their number is proportional to the galaxy luminosity. Using the 3.5PN 

approximation to model the binaries’ inspiral GW signature, we produce sets of detected events for different 

networks comprising combinations of possible detectors (including LIGO Australia and a proposed Indian detector 

INDIGO). We explore how the orientation of INDIGO influences the prior distribution of detected binaries. Then, 

using MCMC techniques, we compute error cubes for the sky positions of the detected systems. We then construct 

subsamples of neutron star binaries whose EM counterparts should be detectable as GRBs and/or transients in the 

optical and radio. This will allow us to consider various strategies for coordinated EM follow-up campaigns. 

Effect of Globally Imposed Reactant Gradients on Osmotic Propulsion 

Anirban Ghosh 

Mentors: John F. Brady and Nicholas Hoh 

One area of interest in the growing field of nanodevices and nanotechnology is the autonomous motion of 

molecular motors. The osmotic motor is one such model which uses a surface chemical reaction as its method of 

propulsion. By creating a concentration gradient in a bath of reactive particles, the motor is able to propel itself 

using the induced osmotic pressure difference. Computer simulations are designed to study the motion of osmotic 

motors in a half reactive, half nonreactive bath at different volume fractions of bath particles and initial orientations 

of the motor. The simulations will implement the Brownian dynamics method to model particle interactions in the 

colloidal dispersion. The average displacement and change in orientation of the motor are studied over time. The 

results will provide an understanding for how motor particles will behave in heterogeneous, non-uniform 

dispersions, and may allow for more control over osmotic motor movement through the use of reactant gradients 

in a mixture. 

Barium Tagging for the Enriched Xenon Observatory 

Lauren Gilbert 

Mentors: Martin Briedenbach, Liang Yang, and Maria Spiropulu 

One key component of the full Enriched Xenon Observatory (EXO) will be the in-situ tagging of xenon double beta 

decay daughter nucleus, barium. Such a tagging technique can significantly reduce radioactivity-induced 

background and increase the experiment sensitivity. Current R&D focuses on firstly, removing the barium from a 

platinum surface with a 750 mW Nd:YAG laser, and secondly, correctly identifying the barium ions with a time of 

flight spectrometer. While there is not enough data to draw any substantive conclusions, the early results look 

promising. 

Coherence of Majorana Qubit in a Topological Superconducting Wire Undergoing a Phase Slip Event 

Samuel Goldberg 

Mentors: Gil Refael, Doron Bergman, and David Pekker 

Recently, it has been suggested that topological states of quantum matter, those determined by the global 

properties of the system, may provide a decoherence-free realization of quantum computing. We are investigating 

the properties of one possible candidate: the spin-polarized p-wave superconductor wire. The quantum information 

is stored as a pair of Majorana modes located at the ends of the wire that together form a delocalized fermion state 

of zero energy. The goal of this project is to calculate the effect of superconductor order parameter fluctuations on 

the state of the Majorana qubit. In particular, we investigated the effect of a phase slip event, a fluctuation where 

the order parameter vanishes at a point, allowing the superconducting phase to unwind. We have found that a 

phase slip changes the delocalized fermion number and creates a bulk quasiparticle excitation with the energy of 

the superconducting gap. A second phase slip returns the delocalized fermion number to its initial value and 

destroys the quasiparticle. However, the unoccupied component of the Majorana qubit gains a Berry phase of 

relative to the occupied component. Phase slip events, therefore, will compromise the coherence of the qubit. 

30

Thermometry of He 3 Fridge 

Chen (Chris) Gong 

Mentors: Keith Schwab and Emma Wollman 

The Schwab group is in the process of building and testing a He 3 fridge to measure the properties a microwave 

amplifier and graphene bolometer. The He 3 fridge is designed to reach 300mK in its coldest stage. My work consists 

of the electronics of the fridge, which are the thermometry and the cryogen level detecting circuit. The precise 

measurement of the temperatures at various places in the fridge is a very important diagnostic tool for 

understanding the heat conduction circuit in the fridge. Since the measurements are done at very cool 

temperature, the design principle of the thermometry circuit is to limit the heat dissipation from the sensor as well 

as the heat conducted through the wires. Using four-wire sensing and lock-in amplification in this circuit we are 

able to maintain a high sensitivity while limit the current which might generate heat. The two circuits are currently 

still in testing stage. 

Temperature Variations of Jupiter’s Atmosphere From Mid-Infrared Imaging 

Jennifer J. Greco 

Mentor: Glenn Orton 

Over the last several years, Jupiter has been undergoing massive atmospheric changes, in what meteorologists 

classify as a period of ‘global upheaval.’ One aspect of these ongoing meteorological changes involves Jupiter’s 

North Temperate Belt (NTB), which turned a much lighter color in December 2002. In February and March 2007, 

the NTB ejected 2 plumes and, in the wake of the second, returned to a darker color, distinct from its typical 

shade. The normally dark brown South Equatorial Belt lightened in early 2007 and returned to its normal color later 

that year. It lightened again in 2009 – 2010, and began reviving in late 2010. Very little is understood about these 

atmospheric changes. In order to better understand them, images of Jupiter taken over the past decade at midinfrared 

wavelengths between 8.5 and 24.8 microns are being analyzed to examine the temperature in both the 

upper and deeper layers of Jupiter’s troposphere, together with changes in cloud properties and the distribution of 

ammonia gas and the fraction of molecular hydrogen in the para- vs. ortho states. The cloud and trace-gas 

distributions are useful as indirect indicators of vertical motions. Results have revealed some interesting 

temperature and gas profile changes including a decrease in temperature of the Equatorial Zone in 2000-2002, 

followed by a slow temperature rise, which correlates to an increase in the temperature of the North Equatorial 

Belt, followed by a slow decrease. Both of these are correlated to an increase in the para-H2 fraction in the 

Equatorial Zone. 

Metal-Decorated, Nanostructured Sm0.2Ce0.8O2-� Thin Films by PLD as High-Performance Anodes 

Kevin L. Gu 

Mentors: Sossina M. Haile and WooChul Jung 

The primary long-term goal for solid-oxide fuel cells (SOFCs) in general is to reduce operating temperatures to 

intermediate ranges (600–700°C) for reasons of material stability, cost, and wider applicability. Preliminary results 

involving platinum catalysis on vertically-aligned columns of Sm0.2Ce0.8O2-��(SDC) have shown significantly reduced 

electrode impedance when compared with other morphologies such as Pt on dense SDC for intermediate 

temperatures. The two primary goals of this project are to explore the suitability of Ni, Ti, or Cu as an inexpensive 

replacement for Pt, and investigate the dependence of performance on electrode morphology. Columnar thin film 

electrodes of SDC are grown on Y0.16Zr0.84O1.92 (YSZ) single crystals by pulsed-laser deposition. Metal catalyst 

particles are then deposited by means of sputtering. Results of electrochemical and morphological analyses carried 

out via AC impedance spectroscopy and scanning electron microscopy are presented. 

Proton Exchange Membranes With Through-Plane Aligned Ion Conducting Channels: A Study of Ionic 

and Electric Conductivity 

Mengyu (Kelly) Guan 

Mentors: Seung-Hyeon Moon and Sung-Hyun Yun 

Proton exchange, or polymer electrolyte, membranes (PEM) are semi-permeable membranes used in proton 

exchange membrane fuel cells (PEMFC). PEMs function to separate the reactants and transport protons across the 

membrane while blocking electrons. Current research focuses on improving membrane properties by (i) using 

different ionomers and (ii) modifying the structural properties of the membranes. Sulfonated poly(phenylene oxide) 

(SPPO) has been shown to be a promising PEM candidate, while applying an electric field perpendicular to the 

membrane surface may improve proton conductivity and PEMFC performance. In order to study the effects of 

applying an electric field along the membrane thickness direction, a series of SPPO membranes, with a target 

thickness of ~50 μm, was fabricated under an applied electric field and characterized. Since electrical insulation 

becomes an issue with thin membranes, this study will also attempt to measure the effects of electron transport 

across PEMs by electrical impedance spectroscopy. Finally, the optimal membrane will be tested for PEMFC 

performance against a commercially available Nafion membrane to compare and further optimize membrane 

properties. 

31

Studying the ABCB7 Transport Protein From Bacteria 

Trisha Guchait 

Mentors: Doug Rees and Jens Kaiser 

The ABCB7 trans-membrane protein is an ATP- binding cassette transporter thought to be responsible for 

transporting Fe-S compounds from the mitochondria to the cytosol. To learn more about the structure and function 

of this transporter in order to better understand iron metabolism in humans, homologues to the ABCB7 protein 

were studied in a wide variety of bacteria. First large colonies of the bacteria were grown and DNA was extracted 

using a phenol/chloroform extraction. Once the genomes were properly identified using the 16s RNA sequence, a 

PCR was performed on the ABCB7 sequence. The gene is then cloned and expressed using E-coli. 

Amplification of DNA-Mediated Redox Signals Using Enzyme-Mediated Electrocatalysis 

Luis F. Guerra 

Mentors: Jacqueline K. Barton and Catrina G. Pheeney 

The ability to rapidly and accurately detect the abundance of DNA sequences and lesions is necessary for medically 

relevant procedures such as genomic sequencing and the detection of genetic mutations that could lead to disease. 

Electrochemical biosensors achieve many of these criteria at low cost, and in particular, electrochemical DNA 

detection via DNA-mediated charge transfer (CT) has been shown by the Barton group here at the California 

Institute of Technology to have unique, robust advantages over other detection technologies. However, the signals 

obtained by DNA-mediated CT based biosensors—in which a redox reporter is intercalated into the distal end of 

gold-bound DNA and is then reduced via DNA-mediated CT—can be too weak to accurately detect low abundance 

oligonucleotides, such as miRNA. The feasibility of using an enzyme to electrocatalyze the redox reaction and thus 

enhance the sensitivity of these assays was investigated; in particular, hemoglobin—an important redox-active 

oxygen transporter in biological systems—was utilized due to its efficiency and lack of reactive side products. It 

was found that hemoglobin avoided many of the pitfalls of other electrocatalytic systems (such as direct reaction 

with the electrode) and catalyzed DNA-mediated CT redox reactions—important prerequisites for a successful 

electrocatalytic system. 

Design of a Program for Shear Induced Polymer Crystallization Control 

Ashley Guo 

Mentors: Richard Flagan, Ruoyu Zhang, and Xerxes Lopez-Yglesias 

By applying an external force to melted polymer, unique properties of the resulting crystallized polymer can be 

observed. We present a program developed in the software platform LabVIEW® for the control of a shear induced 

polymer crystallization experiment. In this system, the shear force is applied by the control of pressurized gas, and 

temperature is controlled by a proportional–integral–derivative controller (PID controller). All features of the 

program are accessible from a single interface which contains controls for the entire shear experiment. The 

program adjusts the temperature and pressure applied to the system based on user-defined values and both 

collects and saves temperature, pressure, and optical data. Additional features in the program include calibration 

functionality, flexibility in choice of hardware, live graphing of data, and the ability to perform a shear multiple 

times. A report file can be saved at any point while the program is running, containing the basic details of the 

experiment and custom comments by the user. 

Cost Model Comparison: A Study of Internally and Commercially Developed Cost Models in Use by NASA 

Garima Gupta 

Mentors: Eleanor Basilio and Greg Welz 

NASA makes use of numerous cost models to accurately estimate the cost of various components of a 

mission—hardware, software, mission/ground operations—during the different stages of a mission’s lifecycle. The 

purpose of this project was to survey these models and determine in which respects they are similar and in which 

they are different. The initial survey included a study of the cost drivers for each model, the form of each model 

(linear/exponential/other CER, range/point output, capable of risk/sensitivity analysis), and for what types of 

missions and for what phases of a mission lifecycle each model is capable of estimating cost. The models taken into 

consideration consisted of both those that were developed by NASA and those that were commercially developed: 

GSECT, NAFCOM, SCAT, QuickCost, PRICE, and SEER. Once the initial survey was completed, the next step in the 

project was to compare the cost models’ capabilities in terms of Work Breakdown Structure (WBS) elements. This 

final comparison was then portrayed in a visual manner with Venn diagrams. All of the materials produced in the 

process of this study were then posted on the Ground Segment Team (GST) Wiki. 

32

Iron Based Bulk Amorphous Alloy Sample Preparation for Annealing in Magnetic Field 

Karan Gupta 

Mentor: Marios D. Demetriou 

This project involved the preparation of samples of inductor and transformers cores of Iron based bulk amorphous 

alloys for annealing in a magnetic field. Two methods of preparation were employed. The first involved casting 

these alloys in an RF induction melter to make disks and the other used the RDF method (Rapid Discharge forming) 

to prepare toroidal samples. Amorphous alloy samples were obtained from both these methods. By annealing these 

samples in a magnetic field we hope to further reduce energy losses from already effecient bulk iron based 

amorphous alloy cores. 

Bootstrapping Robotic Systems: Implementation and Analysis 

Magnus Hakansson 

Mentors: Richard M. Murray and Andrea Censi 

This project covers a very extreme form of machine learning. An agent should start without having any information 

at all about neither its actuators nor its sensors. The agent should then learn about its dynamics by generating 

uninterpreted commands and receiving uninterpreted observations. Importantly, the learning algorithm does not 

depend on the actual robot configuration. Following the learning phase is the active phase where the learned 

dynamics are used for the agent to perform a specific task. The first part of the project has been spent on creating 

a set of programs that was used for the experiments. The convergence of the learning has been shown from 

loggings. One feasible task for the agent is the detection of sensors that fails. A strategy for accomplishing this is 

implemented and tested. Future work also includes examining whether it is possible to create an adaptive 

command generator for the agent. 

Electrical-Mechanical Properties of Graphene Cantilever Structures 

Jeff N. Han 

Mentors: Julia Greer and Mingyuan Huang 

Graphene, a 2-dimensional material composed of a single layer of graphite, has shown great promise in future 

electronic devices and Nano electromechanical systems. Due to its ultra-high electron mobility, low mass density, 

and exemplary mechanical properties, graphene has the potential to become the ideal material for future 

developments in transistor technology, sensors, and resonators. In mass, pressure and position sensing, graphene 

has proven to be accurate and robust in operation. Due to high strength, stiffness, and conductivity along the basal 

plane, a graphene cantilever can support itself with no built-in tension and exhibits a high resonance frequency. 

The fabrication of the cantilever structure involves mechanical exfoliation of graphite onto a SiO2 wafer The 

graphene is cut into a nanoribbon and is plated with electrodes before the SiO2 beneath the graphene is etched 

away. This suspended graphene is cut in half with a focus ion beam subsequently creating the cantilever 

structures. A voltage bias between the cantilever and the Si wafer below will be modulated and corresponding 

frequencies will be measured in a scanning electron microscope. 

Ab initio Quantum Computation of THz Vibrational Modes 

Matthew Harrigan 

Mentors: Geoffrey Blake and Matthew Kelley 

Terahertz time-domain spectroscopy (THz-TDS) uses light in the terahertz frequency region (3-300 cm -1 ) to excite 

low energy modes of molecules and materials. It promises to offer a new tool to investigate molecules with 

applications in pharmaceuticals, security and defense, and cosmochemistry. Whereas infrared (IR) vibrational 

modes can be assigned empirically or with simple free-molecule quantum mechanical computations, the observed 

THz absorptions cannot be linked to purely vibrational modes associated with individual species. Instead, the 

difficulty in assignment arises from the intermolecular nature of THz vibrations. Quantum mechanical codes which 

employ periodic boundary conditions were used to calculate intermolecular phonon modes. Both Gaussian basis 

sets and plane-wave basis sets with pseudopotentials were investigated. 

Optimizing Forging of Bulk Metallic Glasses Utilizing Rapid Discharge Capacitive Heating 

Thomas N. Harris 

Mentors: Marios Demetriou and Georg Kaltenboeck 

Bulk metallic glasses (BMG’s) possess equal or superior characteristics compared to modern engineering materials. 

However, BMG’s can be thermoplastically formed, enabling manufacturing processes to occur at much lower 

temperatures and stresses. In this study, we aim to forge amorphous GHDT into different shapes using a novel 

device, the Rapid Discharge Forming machine. After forming, we will analyze the quality of forging and use such 

analysis to refine future BMG manufacturing techniques. 

33

Cold Friends of Hot Jupiters 

Monica He 

Mentor: Sasha Hinkley 

With further Doppler monitoring and a better-characterized radial velocity curve of transiting exoplanets, we may 

establish the existence of additional planets on ~28% of our sample. By either confirming or placing strong limits 

on the presence of outer companions in these systems, we can directly test migration mechanisms involving fewbody 

interactions, such as planet-planet scattering, Kozai migration, as well as generic secular perturbations. 

Indeed, even the absence of these longer-period, eccentric objects would render migration processes, driven by 

secular instabilities, unlikely. A dynamical analysis of such a multiplanetary system, combined with models of the 

inner planet’s interior density structure, can also constrain the core mass and tidal quality factor, Q, of the 

transiting planet. Our objective of optimizing the observing strategy to be carried out by the California Planet 

Search (CPS) was approached by a Monte Carlo analysis, drawing for the period of the orbit, the planet mass, 

eccentricity, and a time baseline. A scaling relationship for the sensitivity for the planet’s mass was found, 

dependent on the dispersion of the planet’s linear radial velocities. With new observations taken over the next few 

years collated with the radial velocity measurement database of CPS, we hope to constrain the existence of 

companion planets. 

THz Emission From Photoinduced Charge Transfer in Re(I) Carbonyl Diimines 

Thomas Heavey 

Mentors: Geoff Blake and Marco Allodi 

Rhenium(I) carbonyl diimines absorb light in the near ultraviolet to generate a photoexcited state with a different 

dipole than the ground state of the molecule. The dipole change associated with this excitation causes the 

surrounding polar solvent molecules to rearrange. This rearrangement perturbs the electric field on the timescale of 

a few picoseconds, which corresponds to electromagnetic radiation in the terahertz region of the spectrum. Using 

ultrafast pulsed lasers, a solution containing these molecules can be photoexcited, and the corresponding terahertz 

radiation can be measured. The emitted electric field can be detected with a high signal-to-noise ratio using 

electro-optic sampling. Additionally, the emitted radiation can be simulated and consequently analyzed using 

computer models. The simulation model will be improved by being made more efficient, general, and user friendly. 

Using Phospolipid Bilayer Nanodiscs as Platforms for Small-Angle Scattering Investigations of a 

Membrane Protein 

Adriana Hertel-Wulff 

Mentors: Lise Arleth and William M. Clemons, Jr. 

Nanodiscs provide unique opportunities for structural and functional characterization of membrane proteins in 

solution. These systems consist of a form-maintaining membrane scaffold protein (MSP) surrounding a 

phospholipid bilayer, into which a membrane protein, such as bacteriorhodopsin (bR), can be inserted. Using this 

approach and verifying the results with spectrophotometry, circular dichroism, as well as small-angle x-ray 

scattering, we investigate the integration of trimeric bacteriorhodopsin into nanodiscs, which we then use to study 

the structural and functional characteristics of the proton pump. We thus find that the ideal conditions for 

successfully creating monodisperse, well-defined nanodiscs with bacteriorhodopsin include additional, non-native 

lipids, such as POPC; an appropriate concentration of a mild detergent, such as octyl glucoside; and the correct 

stoichiometry between the MSP, lipids, and bR. 

Spectral Gap Scaling of One Dimensional Quantum Spin Chains 

Matthew T. E. Heydeman 

Mentor: John Preskill 

Frustration-free Hamiltonians which are the sum of two-qubit projection operators have been considered as models 

for adiabatic quantum computation. The run time of such computation is determined by the scaling of the spectral 

gap with system size; in general, this property is not well understood. We show that the number of free 

parameters in any such Hamiltonian may be reduced to three real parameters and consider the decay of the gap 

with respect to the number of qubits. We present numerical results from a DMRG algorithm. For certain values of 

the real parameters, the scaling is shown to be an inverse polynomial of the system size. 

Investigation of Chemotrophic Sulfur and Iron Metabolisms Under Simulated Europa Ocean Conditions 

Samuel H. Ho 

Mentor: Robert T. Pappalardo 

Galileo data points to the probable existence of a subsurface ocean beneath the icy crust of the Jovian moon 

Europa. The water ocean provides a potential model for low-temperature redox cycling involving carbon, sulfur, 

and iron as energy sources for chemotrophic microorganisms. The presence of a solvent (water), a molecular 

scaffold (organic carbon) and an energy source (geochemical disequilibria) may provide the necessary conditions 

for life at Europa. In this study, the terrestrial organisms Shewanella putrefaciens and Shewanella oneidensis strain 

34

MR-1 (facultative anaerobic bacteria isolated from high-pressure, low-temperature, and nutrient-poor 

environments) are tested for growth and survival through sulfur and iron reduction under simulated Europa ocean 

conditions. Capillary electrophoresis can be used to monitor the increasing concentrations of reduced sulfur and 

iron during the course of cultivation experiments and compared with negative controls. Biomolecules generated 

from such reactions can be characterized using gas chromatography-mass spectrometry. Scanning electron 

microscopy can evaluate morphologies and distributions of biominerals and extracellular components that may be 

generated through the activities of these microorganisms. In the interests of studying reaction rates within feasible 

time-scales, free energies of Europa ocean geochemical reactions were calculated using the computer simulation 

program REACT within Geochemist’s Workbench (GWB) 6.0. Results from this project may provide valuable 

information about potential habitability and target biosignatures relevant to further in situ exploration of Europa. 

Using RGD Peptide Polymers to Promote the Attachment of Stem Cells to Parylene Surfaces 

Wilson Ho 

Mentors: Robert H. Grubbs and Paresma Patel 

Stem cells have the ability to differentiate into a variety of fetal or adult cell types and thus hold great potential as 

a medical treatment for tissues that do not regenerate when damaged. Stem cells need to attach and differentiate 

on a surface, and then be held in close contact with the damaged tissue for this treatment to work. A thin sheet of 

parylene, a bioinert material that is FDA-approved for use in medical implants, could serve as the solid support to 

hold the cells in place, but stem cells will not attach to and proliferate on this surface without modification or 

treatment. The RGD peptide sequence has been shown to bind integrin proteins expressed on the outer 

membranes of cells, and coating the parylene sheet with these peptides could promote stem cells to attach to its 

surface. Our approach is to embed the RGD peptide in a hydrogel polymer matrix, and then coat this matrix onto 

parylene prior to cell seeding. A number of methods for this coating process will be tested, and the RGDfunctionalized 

parylene surfaces will be sent to our collaborators for biological evaluation. 

Instrument Design and Construction for Photoacoustic Spectroscopy and Frequency Stabilized Cavity 

Ringdown Spectroscopy 

Daniel Hogan 

Mentor: Mitchio Okumura 

Based on earlier designs of photoacousic spectroscopy (PAS) and frequency stabilized cavity ringdown spectroscopy 

(FS-CRDS), highly accurate PAS and FS-CRDS instruments are constructed. The PAS apparatus was designed to 

operate near 770 nm, which is near the A-band of molecular oxygen. PAS will enable measurements at high 

pressures, where line mixing can be observed. The FS-CRDS apparatus is design to produce spectra around 2 um, 

near important carbon dioxide transitions. The spectra produced by the FS-CRDS instrument will be sufficiently 

detailed to precisely observe line shapes, and may be used to find deviations from revised theories. The spectra 

collected will be used as reference data for remote sensing projects, such as the Orbiting Carbon Observatory, 

which require extremely precise line parameters to achieve their desire precision. 

The Heroic Leaders of Renaissance Epics 

Nerissa Hoglen 

Mentor: Kristine Haugen 

Renaissance writers, particularly Ariosto, Tasso, Spenser, and Milton borrowed from Virgil and each other to write 

heroic epics, which practice allows us to analyze patterns emerging among their works. One trend involves 

portraying the hero of the epic in a leadership role. Though heroes are usually considered extraordinary figures, 

emphasizing their leadership qualities draws attention to the essential role other people, particularly the hero’s 

community and the audience of readers, play in defining the character as a hero. The Aeneid makes clear the 

importance of the leader’s community of followers, particularly his family. Experimentation with the role of the 

hero, particularly Tasso’s three-way division of personalities and responsibilities, pave the way for Milton to create 

two very different heroic figures with Adam and Satan in Paradise Lost. Milton explores their emotions and their 

relationships to their communities. The contrast between the two raises philosophical questions about what it 

means to be a hero. 

Rhesus Macaque Training for Neuroprosthetic Project 

Erin Hoops 

Mentor: Richard Andersen 

This neuroprosthetic project has two main objectives. First, to control a robotic arm using movement-planning 

signals. Second, to investigate simulation of tactile feedback using somatosensory stimulation. Our experiments 

involve three Rhesus Macaques (macaca mulatta) performing reaching tasks in 3D virtual reality. We compare 

control of a robotic arm using magnetic sensor information and actual arm movements with control directly from 

posterior parietal signals. Obtaining similar results from both tasks will verify the functionality of our signal 

recording hardware and software processing. To test the plausibility of simulating tactile feedback, we will repeat 

the aforementioned brain-control task while stimulating the somatosensory cortex upon completion of a reach. If 

35

the trials with stimulation are more accurate or lead to faster learning, more research will need to be done in order 

to determine the most effective way to incorporate this feedback into neuroprosthetic systems. Completion of this 

project will lead to a more effective neuroprosthetic device, which will help many people, including amputees, 

paraplegics and people with amyotrophic lateral sclerosis (ALS). 

Developing a Microfluidic Delivery System for Silicon Nanopillar Ion Sensors 

Maxwell Horton 

Mentors: Axel Scherer and Andrew Homyk 

Silicon nanopillars are nanoscale cylindrical features that can be fabricated on a silicon chip. Typically, they are 

about 10nm in diameter and over 100nm tall. Their high surface area to volume ratio makes them extremely 

sensitive to their environment. For example, if nanopillars are placed in a solution, the presence of ions in the 

solution can affect the current flow through the pillars when a voltage is applied across the pillars. By fabricating a 

sensor on a chip and delivering a test fluid to it, and applying a voltage across the sensor, the ion concentration of 

the test fluid can be deduced. Delivering test fluid is not easy. A simple PDMS mold cannot be used, because it is 

difficult to align the mold with the small features on the ion sensor. A microfluidic delivery system was designed to 

deliver test fluid to the nanopillars. SU8 was spun onto a chip and used to define channels, which were accurately 

aligned with the sensor using a mask aligner. A PDMS lid was placed on top to cover the channels. 

Faces in Motion: The Effect of Motion on the Saliency of Faces 

Gladia Hotan 

Mentors: Christof Koch and Blythe Towal 

The study of human visual saliency is key to the understanding of human visual attention. It is also important for 

the development of biologically feasible computer vision algorithms. Although much work has been done on the 

saliency of both low-level features such as color and intensity, and high-level features such as faces and text, most 

current saliency algorithms only apply to still images. Since a biologically feasible saliency algorithm must be 

applicable to real-life stimuli, we investigated how the motion of faces, which most people see on a daily basis, 

affects their saliency. Subjects performed a free-viewing experiment in which they viewed videos of people walking 

in natural and choreographed settings. Analysis of eye-tracking data from the experiment indicated which factors 

attracted visual attention the most. The data will be used to extend the current Itti-Koch saliency algorithm for use 

in natural video stimuli. 

Investigating the Sufficiency of “Leaky Gut” to Induce Autism-Like Symptoms 

Sophia Hsien 

Mentors: Paul H. Patterson and Elaine Y. Hsiao 

In the past few decades, there has been a dramatic increase in the number of autism cases diagnosed. The recent 

California Autism Twin Study estimates that 58% of autism development risk can be attributed to non-genetic 

factors; within these, there is accumulating evidence for a gastrointestinal (GI) disorder. Common GI complaints of 

autistic patients include chronic constipation and abdominal pain, and studies of some autistic individuals show 

eroded GI epithelium and increased intestinal permeability. Currently, one of the most widely used mouse models 

for autism is the maternal immune activation (MIA) model, in which the injection of viral mimic, poly(I:C), in a 

pregnant female yields offspring with the core symptoms of autism: deficits in communication, increased repetitive 

behavior, and impaired social interactions. Our laboratory has shown that these poly(I:C) offspring have increased 

intestinal permeability, commonly known as “leaky gut”. I am investigating whether leaky gut is sufficient for the 

development of autism-like behaviors. Postnatal day 21 mice were orally administered 0%, 1%, or 2% dextran 

sulfate sodium (DSS), which is commonly used to induce experimental colitis in mice. The resulting leaky gut was 

verified by monitoring diminished weight gain, scoring stool sample consistency, assessing colon pathology in H&Estained 

sections, and performing a leaky gut assay (to assay leakage of fluorescently-labeled dextran, from the GI 

tract into the serum). Using these assays, 2% DSS was determined to be the optimal dosage, and a second cohort 

of mice was started on this regimen. To characterize the behavior of DSS-treated mice, five assays with relevance 

to autism are being used: open field test (OFT), pre-pulse inhibition (PPI), marble burying, social preference, and 

sociability. Preliminary results show no difference in PPI. 

Characterization and Structural Study of a Zinc Exporting P1B-Type ATPase 

Jessica Hsu 

Mentors: Douglas Rees and Gabriele Meloni 

P1B-type ATPases actively transport transition metals across the cell membrane to regulate the concentrations of 

essential and toxic metal ions. However, the 3D atomic structure of this class of proteins and the mechanism by 

which the energy generated by ATP hydrolysis is coupled to metal transport and induces conformational changes in 

protein structure remain to be elucidated. Deletion mutants of ZntA from P. aeruginosa have been expressed, 

purified, and prepared for crystallization trials in a lipidic environment to identify the conditions under which the 

protein can be crystallized for X-ray diffraction structural determination. Nucleotide analogs and inhibitors have 

also been introduced to lock the ZntA into its intermediate states to map out the catalytic cycle. Its hydrolytic 

36

activity has been studied in various chemical environments, including its insertion back into proteoliposomes to 

recover the coupling of ATP hydrolysis to the transport of its substrate. By discerning the structural properties of 

transition metal transporters, we will add to our knowledge of the regulation of trace element homeostasis and the 

origins of diseases that arise from perturbations of such activity. Furthermore, the understanding of how metal 

selectivity is achieved will allow for the possibility to design modulators of P1B-type ATPase pumps. 

Dinosaur Body Temperatures From Clumped Isotope Paleothermometry in Carbonate From Eggshells 

David Hu 

Mentors: John Eiler and Rob Eagle 

The nature of the physiology and thermal regulation of non-avian dinosaurs is the subject of debate and arguments 

have been made for both endothermic and ectothermic metabolisms. Here clumped isotope thermometry of 

carbonate is used to show modern bird and reptile eggshells record the body temperatures the female experienced 

prior to egg laying. The method is applied to well preserved cretaceous oviraptor eggshells. The body temperatures 

of oviraptors are determined to be 29-33C, similar to modern reptiles. Soil nodules found near the cretaceous 

shells estimate paleoclimate temperatures at 21-25C. Evolutionarily, oviraptors are close relatives of modern day 

birds suggesting avian endothermy developed late on the evolutionary timeline of birds. It also suggests most 

dinosaurs were ectothermic and previous endothermic-like results in larger dinosaurs are the product of 

gigantothermy. 

Engineered Underdominance: A Gene Drive System for Drosophila melanogaster 

Jennifer Hu 

Mentors: Bruce A. Hay and Kelly J. Dusinberre 

A proposed strategy to control insect borne diseases such as malaria is replacement of insect vectors with 

transgenic counterparts that are incapable of disease transmission. In general, however, transgenic insects have no 

fitness advantage over wild type, and hence it is naïve to suggest that transgenes conferring disease refractoriness 

will be efficiently spread through natural selection. This project proposes the utilization of engineered 

underdominance to drive transgenes into populations. In engineered underdominance, we construct a pair of 

unlinked lethal genes, each of which is associated with a repressor of the other lethal gene. This effectively means 

that only if both engineered chromosomes are present will the individual survive. The head involution defective 

gene (Hid) of D. melanogaster functions in programmed cell death by triggering apoptosis. We have designed 

underdominance constructs in which the toxin is expression of engineered Hid, and the rescue consists of 

specifically targeted microRNA complementary to engineered Hid. Currently, efforts are being made to build each 

chromosome and generate transgenics with blocking sequences such that the killing component is off, so that both 

chromosomes can be brought together in a single individual to test drive. 

Sensing Module for CMOS Based Magneto Bio-Sensor 

Yunqing (Alexander) Hu 

Mentors: Ali Hajimiri and Alex Pai 

Current molecular level diagnostics are primarily carried out using stationary lab-based facilities, which require 

large equipment, high cost and extreme complexity. The CHIC Lab has designed a frequency-shift based CMOS 

based magnetic biosensor chip that can detect 1-μm diameter magnetic labels. By detecting the magnetic label, 

biological molecules can be detected and quantified in the biosensor’s proximity. However, a supporting system has 

yet to be designed to make the device handheld, easy to use, and cost effective. Thus, we have developed a 

solution consisting of an FPGA-PCB based design to give the test setup a handheld form factor and a Graphic User 

Interface (GUI) to ease the user’s experience. In our design, the PCB board is directly socketed on top of the FPGA, 

giving the final product the size of 2.1 inches by 3 inches, with a single USB wire providing both power and data 

channel. The software interface is made so that non-technical individuals can conduct assays for point of care 

applications. 

Modeling and Characterizing a Novel Nitric Oxide Donor Used as a Treatment for Dry Eye 

Tiffany A. Huang 

Mentors: Julia A. Kornfield and Dennis Ko 

Kertoconjunctivitis sicca, or dry eye, is a common eye condition affecting millions of people worldwide that arises 

from a malfunction of the tear film, resulting in low levels of tear production or tears that evaporate too quickly. 

Research has shown that nitric oxide (NO) can stimulate blood vessels in the eye to produce their own autologous 

serum, which increases tear production. To improve upon current treatments for dry eye, which are often 

inconvenient and uncomfortable, a protein that can periodically release NO and can be administered via contact 

lenses would be desirable. NO was conjugated to bovine serum albumin (BSA) to produce reduced nitrosylated 

bovine serum albumin (r-NO-BSA). Fluorimeter analysis and gel simulations showed that NO was released for a 

duration of up to 3 hours in vitro. The effects of pH, thiolene ratios, and amounts of BSA on r-NO-BSA’s 

biodegradability were shown by SDS-PAGE analysis. S-nitrosoglutathione (GSNO) was also synthesized and its 

release kinetics were studied as a standard for the NO release assays. 

37

Comparison of Lattice Boltzmann and Direct Numerical Simulations for Single Phase Flow 

Yifei Huang 


Due to their benign emissions and applicability to a broad range of applications, fuel cells serve as an alternative to 

combustion engines for energy conversion. However, slow removal of water from the cathode has been a limiting 

factor to their performance. Currently, the lattice Boltzmann method (LBM), based on statistical mechanics, and 

direct numerical simulation (DNS), based on Navier-Stokes continuum, are two methods used to simulate liquid 

flow in porous fuel cells. However, we desire a thorough comparison of the two methods. We studied both 

simulation methodologies in two cases with existing analytical solutions: flow through a channel (Pouseuille flow, 

analogous to flow through a pore), and flow around a cylinder (analogous to flow around an obstacle). Various 

Reynolds and Knudsen numbers were used to determine the range of applicability of each simulation. Specifically, 

we compared the entry length and velocity profiles of the two simulations with each other and with analytical 

solutions. This comparison provides a better understanding of the limitations of each simulation to improve flow 

simulations in fuel cells. 

Effect of Aging on the Foam Fractionation of Lactoferrin 

Yuehan Huang 

Mentors: Julia Kornfield and Robert Tanner 

Foam fractionation is an inexpensive and simple technique for concentrating proteins. The foamability of a protein 

can drastically change with the age of the protein. The foamability of solutions created from ten year old bovine 

lactoferrin (bLF) protein was investigated with varying concentration protein, air flow velocity, and the pH of the 

solution. The results suggest the foamability of the aged protein decreased to an insignificant level except at high 

pH with a protein concentration of 0.1 mg/mL. 

Linear Programming and Clustering 

Aditya Huddedar 

Mentor: Leonard Schulman 

We study an affine invariant approach to Strongly polynomial algorithm for linear programming by Vempala- 

Barasz. We prove that the method of combination of the improving rays (with non-negative coefficients) in a 

particular step does not matter as far as the analysis of the algorithm for Klee-Minty and Goldfarb-Sit cubes is 

considered. We present a counter-example in which a weaker version of their algorithm is forced to visit some 

facets of the polytope more than once, unlike in Klee-Minty or Goldfarb-Sit cases. We implement and analyse the 

efficiency of the seeding algorithm described in “The effectiveness of Lloyd-type Methods for the k-Means Problem” 

by Ostrovsky, Rabani, Schulman and Swamy. 

Optimizations of the Quasicontinuum Method on Lattice Structure Computation 

Ka Kin Kenneth Hung 

Mentors: Michael Ortiz and Malena Español 

In simulation of impacting materials with nanoindentors, it is crucial to compute the resultant displacements of 

atoms. These displacement of atoms in a lattice structure in response to an externally applied force field can be 

either computed atomistically by solving a system of non-linear equations or continuously by assuming even 

distribution of mass. However, fully atomistic calculation requires solving millions of non-linear equations which 

incur a huge computational cost while continuous calculations are not accurate. In this project, I have analyzed 

several possible methods based on the quasicontiuum method to approximate the total energy of the system more 

efficiently. 

Determining the Neutron-Star Equation of State From Mass and Radius Observations 

Nathaniel Indik 

Mentor: Lee Lindblom 

Neutrons stars are former main sequence stars that have undergone gravitational collapse to a state where the 

density is so high that the star is primarily composed of neutrons. Since this density is impossible to recreate in a 

lab, the high density equation of state is poorly understood. Many models have been proposed to describe Neutron 

star matter, but they disagree with each other by an order of magnitude. The goal of this SURF is to develop and 

improve theoretical tools that can be used to infer information about the high density equation of state governing 

the cores of neutron stars from the current and future Neutron star observations. 

The interior structures of these stars are governed by the Oppenheimer -Volkoff equation. By integrating this 

system of differential equations we can find the total mass and radius of the star. By assuming that the underlying 

equation of state is determined by a few parameters, we can use it together with the Oppenheimer –Volkoff 

38

equations to produce a mass-radii curve to compare with the observational mass-radii curve. When there are more 

measurements of neutron star mass and radii, the equation of state parameters that best fit the observational 

mass-radii curve will tell us which model for the neutron star equation of state best fits the data. 

Tracking and Analyzing Avalanche-Like Flows in Granular Media 

Robert (Rivers) Ingersoll and Wesley Swank 

Mentors: José Andrade and Carlos Avila 

Although phenomenological models for steady flows in granular media exist, the underlying mechanics of less 

predictable situations, such as avalanches, are difficult to simulate and predict with existing methods. By creating a 

link between the precision of the discrete element method (DEM) at the grain level and the flexibility of continuum 

models, the mentor is developing algorithms that can simulate sudden deformations with a high degree of 

accuracy. The goal of this project is to provide essential physical evidence to quantitatively assess the accuracy of 

the mentor’s theoretical framework and of simulations developed by the co-mentor. To generate this data, a highspeed 

camera captures thousands of images per second of the progression of short, recurring avalanches of steel 

beads in a steadily rotating plexiglass drum. These images are then reduced to an array of coordinates 

corresponding to the centroids of each bead in the frame of view, providing velocity and displacement values in a 

two-dimensional plane. With analysis, this data will be used to refine and eventually confirm the mentor’s model for 

the early progression of an avalanche. 

Neural Correlate to Attractiveness Leakage From Face to Hair 

Janis Intoy 

Mentors: sShinsuke Shimojo and Chihiro Saegusa 

The attractiveness of the face plays a major role in social interactions and can influence economic mobility and 

mate choice. However, few studies have delved into the interactions between face and its natural surrounding, 

hair. In this study, we used functional magnetic resonance imaging (fMRI) to determine brain activity in the event 

that attractiveness leaks from the face to the hair. The brain regions that may be active when leakage is occurring 

are related to the areas of the brain that represent facial attractiveness, including the orbitofrontal cortex, fusiform 

face area, amygdala, ventral striatum, and superior temporal sulcus. The prefrontal cortex is also expected to be 

involved in leakage because it plays a role in inhibition and suppression. 

Modification of SAMMS for Increased Hydrophilicity and Host-Guest Interactions 

Anna Ivanova 

Mentors: Darren Johnson, Sean Fontenot, and Robert Grubbs 

A previously-synthesized self-assembled monolayer on mesoporous supports (SAMMS) material was modified, 

replacing phenyl scaffolding with beta-cyclodextrin groups in hopes of increasing the material’s hydrophilic 

properties. The cyclodextrin groups were also intended to improve the interaction between the scaffolding and 

active layer by utilizing host-guest chemistry. After successful assembly of the cyclodextrin scaffolding, thiol guests 

were tested with varying degrees of success. The stability of the final material in water was evaluated by 

monitoring thiol leaching using Ellman’s reagent. 

A Mechanism for Random-Walk-Based Deterministic Object Sorting on DNA Origami 

Gregory Izatt 

Mentors: Erik Winfree, Lulu Qian, Niranjan Srinivas, and Damien Wood 

The specific base-pairing behavior of DNA, when combined with relatively cheap and easy-to-access base-specific 

commercial oligonucleotide synthesis, is a powerful tool for precisely designing and constructing nanoscale 

structures. Through careful design, these structures can be programmed to behave in productive ways. The Caltech 

BIOMOD team, comprised of Zibo Chen, Shayan Doroudi, Yae Lim Lee, Sarah Wittman, and I, presents work done 

to prove the functionality of such a mechanism for the spatial sorting of DNA-tagged cargo on the surface of a 

100nm by 70nm folded DNA rectangle -- hereby referred to as "DNA origami" -- through the parallel work of many 

random-walking single-stranded oligonucleotides cooperating with cargo-specific goal strands. We used a 

combination of gel analysis, fluorescence experiments, and atomic force microscopy to consecutively verify that the 

many steps and components of our mechanism function correctly, independently and together. 

Standing Before Stepping: A Home Stand Frame for Spinal Cord Injury Patients 

Anum Jang Sher 

Mentor: Joel Burdick 

About 250,000 people in US today suffer from severe spinal cord injury and approximately 11,000 new injuries 

occur each year. Hence, a tremendous amount of effort is been put in the research of reestablishing the connection 

between the brain and the body or to mimic the signals that the brain sends to the body to trigger motion after a 

spinal cord injury. A team of researchers from Caltech, UCLA and University of Louisville is investigating the use of 

high density electrical epidural stimulation to facilitate locomotion recovery after severe spinal cord injury. 

39

However, this technique has to be coupled with rigorous physical training to be effective. In this paper, we will 

focus on the key design features of a stand frame that people implanted with epidural stimulating arrays can use at 

home to extend their physical training outside of a clinic. As this stand frame is used at home, it is designed to be 

safely used without a trained therapist and with minimum assistance. The main structure is stable but light for 

convenience. There is a foot plate, various straps for support and adjustability features for people with different 

physique. Sensors are installed on this stand frame to collect data to analyze the force exerted by the user and the 

support provided by the fame. 

Development of Genetic Tools in Microtubule Possessing Bacteria 

Pierre M. Jean 

Mentors: Grant J. Jensen and Martin Pilhofer 

The genus Prosthecobacter is a bacterial group with increasing interest since the discovery of tubulin genes 

encoded in its genome. These tubulins were shown to form bacterial microtubules; however, their function remains 

unknown. Furthermore, their study is difficult due to the lack of genetic tools available in Prosthecobacter. A 

random transposon, noted for its high efficiency on a broad spectrum of bacteria, was selected for transformation 

of Prosthecobacter vanneervenii through electroporation. An electroporation protocol was designed and optimized 

for the transformation, and a transposon mutant library consisting of 63 P. vanneervenii clones was generated. The 

transposon was amplified using specific primers to confirm the insertion in the genome of the mutants. Using the 

two-step gene walking method, the sequences of the regions flanking the transposon were determined. These 

flanking sequences were BLASTed against the P. vanneervenii genome to determine the location of the insert and 

whether a gene was knocked-out during the process. It has been demonstrated that it is possible to introduce 

exogenous DNA in Prosthecobacter, providing opportunities to further manipulate these organisms and elucidate 

the function of genes of interests in these peculiar and poorly understood bacteria. 

Growth and Characterization of Silicon Germanium Microwires for Application in Multijunction Wire- 

Array Solar Cells 

Julie D. Jester 

Mentors: Harry Atwater and Dan Turner-Evans 

Microwire arrays have the potential to revolutionize the solar energy industry by reducing solar cell production cost. 

Additionally, multijunction arrays can increase the efficiency of non-concentrator solar cells. SiGe alloys are lattice 

matched to a variety of semiconductor materials of different band-gaps, allowing them to be used in conjunction 

with those materials in multijuntion devices. Thus, we have set out to grow SiGe wire arrays to serve as the base 

of multijunction wire solar cells. As a starting point, Si microwire arrays have been grown through Vapor-Liquid- 

Solid (VLS) growth with some success using copper catalyst substrates. The wire array growth fidelity degrades 

with water or oxygen contamination and thus great pains have been taken to remove these elements from the 

growth reactor. The reactor has been entirely rebuilt to remove leaks, and a Residual Gas Analysis (RGA) has been 

used to determine the amount of contaminants present. Furthermore, the copper catalyst may form a stable 

silicide rather than initiating growth, resulting in degradation of array fidelity. By looking at the phase diagrams for 

Cu-Si and Au-Si, it is apparent that Au catalysts should not experience the same silicide producing effect. Thus, we 

have begun to grow with a Au catalyst. In the future, the growth parameters, including temperature and gas flow, 

will be modified in order to produce high fidelity (111) oriented microwire arrays. These Si microwire arrays should 

then be able to be used to produce SiGe alloy microwires by introducing GeCl4 gas during the growth process. 

These microwires can then be doped with boron to give them appropriate electrical characteristics. Producing and 

characterizing these boron doped SiGe microwire arrays should increase understanding so that they can be used in 

multijunction solar cells. 

Engineering cTPR-Linked Antibody-Like Reagents to Improve Neutralization of HIV 

Siduo (Stone) Jiang 

Mentors: Pamela Bjorkman, Rachel Galimidi, and Anthony West 

The Human Immunodeficiency Virus (HIV) infects CD4+ T-helper cells via the binding of the viral surface 

glycoprotein gp120 with CD4. Clinical screens have identified a class of broadly-neutralizing antibodies that are 

capable of binding to and neutralizing many HIV strains. The natural design of these antibodies is in the IgG form, 

composed of two binding sites with a separation distance of 15nm. Due to the low surface density of gp120 on HIV 

however, natural IgGs may not be large enough to achieve bivalent binding, or binding of both antigen recognition 

sites at the same time to the virus. To improve the structure of these antibodies, we have engineered with a vast 

new library of all-protein reagents using cTPR linkers. cTPR proteins are all-helical structural motifs consisting of 

34 amino acid residues per repeating unit that demonstrate high stability. Our antibody-like reagents employ noncanonical 

combinations of Fabs and single-chain variable fragments of some of the most potent antibodies to date, 

and at the same time, are designed to have increased separation distance between their dual binding sites. As a 

result, we expect them to have increased breadth and improved potency of HIV neutralization. When put into an 

adeno-associated virus, an eternal supply of these reagents can be generated and may provide an alternative to 

costly conventional drug therapy. 

40

Design and Fabrication of Optically-Linked CMOS Neural Probes 

Raymond Jimenez 

Mentors: Axel Scherer and Aditya Rajagopal 

We present a fabricated, robust CMOS neural probe with a radical new level of integration. Our new design contains 

an optical power source (PiN diode), an optical transmitter (VCSEL), and a glass pipette electrode fully integrated 

into one die, with a total die area of 600 microns by 600 microns square. This is a breakthrough in standalone 

neural probes, as prior designs have never incorporated their own power source or sensing electrodes in their 

design area. We also present a new, stable oscillator design for sensing the action potential while still minimizing 

transistor count (and area). The resulting neural probe has wide applications in neurobiology, enabling scientists to 

sense the state of tens, if not hundreds of neurons simultaneously, several orders of magnitude more than 

currently possible. Further work could easily extend our design to triggering action potentials, which could enable a 

new realm of control in neuroscience. 

Post-WWI Scientific Communication Networks as Seen Through Albert Einstein’s Correspondence 

Xiao (Dawn) Jin 

Mentor: Diana Kormos-Buchwald 

Einstein’s correspondence during the post WWI period grew exponentially. His correspondents also diversified, 

ranging from Zionist leader Chaim Weizmann to celebrated author Stefan Zweig. In my project I utilize Einstein’s 

immense amount of letters as my primary sources, and examine both the reach and depth of his connections with 

colleagues, family, politicians et al. The most important method I use is statistics. I record the identities of 

Einstein’s correspondents, including nationality, occupation, family backgrounds, and relations to Einstein, in 

addition to frequency of correspondence. Then I carry out some typical statistical analyses such as sorting and 

distribution analysis. My project will expectedly reveal something new about the process of the science center 

shifting from Germany to the outside (possibly America), and the post-WWI scientific community in general. 

Regulation of Attraction by Small Signaling Molecules in Caenorhabditis elegans 

Yea-ra Jo 

Mentors: Paul Sternberg and Jagan Srinivasan 

Small molecule metabolites are known to play important roles in biology of many organisms including 

Caenorhabditis elegans. In Caenorhabditis elegans, ascarosides in varying amounts represent a structurally diverse 

set of signaling molecules that regulate its major life history traits by urging it to undergo specific responses. At 

very low concentrations, this chemical acts as a potent male attractant which allows the male nematode to search 

for potential hermaphrodites as mates. However, the specific genes that control response to ascarosides are still 

unidentified. Therefore, by running behavior assays on the male worms with various neuropeptide mutations and 

comparing mutant nematodes’ responses to previously identified ascarosides with those of wild-type male worms, 

we would like to identify the genes and the proteins encoded by the genes involved in the responses to sex 

pheromone in Caenorhabditis elegans and gain further insight into genetic pathways governing male recognition, 

and furthermore, general animal communication, and self-identification among other species. 

Interactive Image Segmentation Toolset: An Application for Crowdsourcing a New Benchmark 

Segmentation Dataset 

Alex Jose 

Mentor: Pietro Perona 

Image segmentation is an area of study in computer vision, usually referring to computer-generated partitioning of 

an image into regions. The goal of segmentation is typically to simplify the information in an image from a large set 

of unorganized pixels to a set of more manageable and meaningful areas. In creating algorithms for image 

segmentation, it is useful to have a base set of reference images such that the performance of multiple algorithms 

can be compared. This project aims to create a segmentation toolset that will enable the Vision Lab to generate 

such a dataset. This toolset will consist of programs enabling humans to quickly segment images, and will be 

distributed through Amazon’s Mechanical Turk crowdsourcing program. The tasks will be divided among 

participants who will complete small portions of the overall segmentation - such as outlining part of a single image 

- for a small amount of money. Ideally, from this process we will not only generate a new dataset but will elucidate 

some of the intricacies of human approaches to segmentation. 

Understanding Tumor Vascular Heterogeneity Using Whole Mount Optical Microscopy and MRI 

Devashish Joshi 

Mentors: Scott Fraser, David Koos, and Thomas Ng 

Tumors are heterogeneous in structure and a simple biopsy is often not sufficient to capture the complex 

relationship between the vasculature and tumor cells. Since many drugs are administered via the blood stream, it 

is essential to understand the tumor's vasculature to predict drug delivery. We are developing methods to correlate 

mobility maps identified via MRI with the 3D spatial organization of cells revealed by high resolution optical 

41

imaging. MRI imaging of intravenously delivered contrast agents demonstrates the fluid dynamics within the tumor, 

potentially identifying regions of vasculature leakage. Optical imaging of fluorescently labeled vasculature allows us 

to visualize the cellular architecture of the tumor. Correlation of the images obtained from the two modalities will 

allow us to identify the cellular underpinnings of the fluid mobility as seen through MRI. This study will lay forth the 

foundation for future studies aimed at quantifying and enhancing the uptake of drugs within tumors. 

Rugged Encapsulation Structures and Acoustic Instrument Development for Inexpensive Medical 

Sensors 

Theresa Juarez 

Mentors: K. Mani Chandy and Julian Bunn 

Public health in developing countries is limited by insufficient and inaccessible medical care. An innovative way to 

address this issue is by means of a health system capable of operating without direct interaction with a physician. 

The method proposed here would consist of portable cellphone based medical sensing devices that can be used by 

members of the local society in need. The instruments collect medical data and upload them into the Cloud where 

they are evaluated and can be assessed by physicians. Two such devices are heart auscultation and 

electrocardiograph (ECG) tools, both of which exist as prototypes. Problems with these devices include structural 

issues. The ECG and stethoscope lack robust casings and can break easily. The key elements of this study include 

creating an encapsulation for the current auscultation unit and developing a combined stethoscope and ECG device. 

The project will also outline an investigation to find acoustic shapes better suited to microphone recording in the 

stethoscope. 

Network for Collection of Seismic Data 

Prathamesh Juvatkar and Prashant Borde 

Mentors: Julian J. Bunn and K. Mani Chandy 

Earthquakes are natural disasters responsible for destruction, injuries and loss of life and property. However, 

earthquakes are not directly responsible for loss of life. The secondary events triggered by earthquakes such as 

structure collapse, fires, tsunamis and volcanoes cause actual human disasters. The losses due to earthquakes can 

be greatly reduced if we can give early warnings of earthquake or at least detect an earthquake’s early seismic 

activity. For building a system capable of early detection of seismic activity, we need high density arrays of 

spatially distributed motion sensors. Our project is to develop a private network of portable low cost seismometers 

that consume minimal resources. Each seismometer consists of a Single Board Computer (SBC), “Phidget” 

accelerometer and a Home-plug device for power line communication. Each seismometer sends the data to next 

device in linear topology using power-line communication. A base-station in the network sends the collected data to 

a central server via the internet. The operation of the seismometer network is completely automated after 

installation. Also the network is robust against intermediate node failure. Large numbers of such private networks 

will ensure sufficient amounts of data available for the detection of earthquakes with higher accuracy. 

Directed Evolution of Cytochrome P450 for Enantioselective Meso-Desymmetrization 

Karolina Kalbarczyk 

Mentors: Frances H. Arnold and Ryan Lauchli 

Cytochrome P450 monoxygenases (P450s) are a superfamily of enzymes responsible for a range of vital functions, 

from the biosynthesis of secondary metabolites to the degradation of toxic substances in the liver. Engineering 

cytochrome P450 enzymes is a budding area of research, with the goal of optimizing the critical functions 

performed by P450s, in this case with a focus on the potential for enantioselective meso-desymmetrization. Testing 

a variety of enzymes demonstrated that several had high activity and enantioselectivity, with substrates ranging 

from cyclopentanes to cyclooctanes. Particularly high enantioselectivity was present with the P450 9-10A F87I, 

which engineered for thermostability provides an enzyme with a backbone robust to mutation. Through the process 

of directed evolution with high-throughput screening assays, random mutations were generated in the 9-10ATS 

F87I to make a library which can subsequently be screened for higher enantioselectivity with cis-1,2dimethoxycyclooctane. 

Analysis by gas chromatography confirms the enantioselectivity of various mutants. 

Identifying, Analyzing, and Creating a Minimalistic in vitro System for the piRNA Pathway 

Amol Kamat 

Mentor: Alexei Aravin 

The Piwi-piRNA pathway is a unique, lesser understood process for silencing of selfish, repetitive transposable 

elements (TEs) of the genome via de novo DNA methylation in the germline. While it is generally thought that Piwi 

protein is involved in the biogenesis of piRNAs, the other components are not well known, especially in mammals. 

We sought to determine the basic components needed for piRNA biogenesis to occur, create a heterologous in vitro 

system for expressing the piRNA pathway, and study the characteristics of RNAs in the pathway. We expressed 

three proteins (Tdrd12, MOV10L1, and mitoPLD) paired with Mili (Piwi from mice) in stable cells. Their interactions 

with Mili (Piwi from mice) and associated RNAs were to be analyzed via immunoprecipitation, western blotting, RNA 

extraction, and creating and sequencing libraries of that RNA. Western blots of the Tdrd12/Mili IP indicated some 

42

interaction, and RNA gel analysis suggested the presence of small RNAs in the correct size range (23-30 nt) in this 

interacting complex. If sequencing confirms this result, we will be on an encouraging route towards identifying a 

minimalistic approach to creating an in vitro system for piRNA biogenesis, which will be an incredible tool for 

further study. We will continue the project with the remaining two proteins in order to form a more complete 

result. 

Computational Studies of Active Site Hydration in Cytochrome P450BM3 for the Development and 

Understanding of P450 Catalyzed C-H Amination 

Arvind Kannan 

Mentor: Frances Arnold 

The design of new enzymes capable of catalyzing novel reactions not yet observed in nature is a challenging 

multidisciplinary task. This project seeks to demonstrate that a combination of directed evolution and 

computational methods can be used to convert a fatty acid hydroxylase, cytochrome P450BM3, into an efficient 

catalyst for C-H nitrene insertion – an important synthetic reaction that lacks an enzymatic counterpart. We have 

thus far identified P450BM3 mutants that bind and activate aryl sulfonyl azides (one example of nitrene precursors), 

but analysis of the reaction products suggests that the desired amination reaction is outcompeted by hydrolysis of 

the key reactive intermediate, termed aza compound I. Here we have used computational molecular dynamics 

(MD) simulations in order to study active site hydration in P450BM3. Our simulations have identified a long hydrogen 

bonding network of waters in a channel that links the active site to the exterior solvent. A set of polar amino acids 

that stabilize this network of waters have been identified as candidates for future targeted mutagenesis 

experiments, with the ultimate goal of minimizing active site hydration in the presence of the bound nitrene 

intermediate. 

Entropy-Based Systematic Clustering for Constrained Molecular Dynamics 

Jonas T. Karlsen 

Mentors: William A. Goddard and Andres Jaramillo-Botero 

We propose using local entropy to investigate the dynamics of proteins, and we explore a simple way of utilizing 

this approach to develop systematic molecular clustering for constrained molecular dynamics. We interpret the 

notion of local entropy in terms of the physics of the quantum harmonic oscillator approximation, which is used to 

calculate entropies from the density of states that is the Fourier transform of the velocity autocorrelation function, 

which is obtained from a 20 ps molecular dynamics run. Based on our studies we propose a working hypothesis, 

stating that the local entropy of the backbone of a protein is a relative measure for how dynamic that part of the 

protein is, and that high entropy areas are the most likely to undergo large conformational changes in long term 

dynamics. This hypothesis implies that a short simulation from which entropies are calculated, contain information 

about the system on timescales several orders of magnitude larger. 

Object Recognition and Localization Using Audio and Tactile Sensors 

Robert Karol 

Mentor: Joel Burdick 

Due to the rapid increase in the number of robotic devices used in society, there is a strong desire to design 

hardware and software capable of being used in a variety of different situations. Reusable hardware and software 

would allow for robots to be more independent, and allow them to be integrated in a variety of situations. The 

Defense Advanced Research Projects Agency (DARPA) is sponsoring a two part competition to advance the state of 

research into the autonomous sensing, grasping and manipulation of objects. JPL/Caltech is one of six teams 

competing in this event. Prior work by the JPL/Caltech team has been done on using vision systems to perceive 

objects, estimate the position of each object, and plan paths to grasp the object, as well as controls for moving the 

arm along the planned path and manipulating the object. My current work focuses on the feasibility of identifying 

different sounds through the use of a Fast Fourier Transform, Principal Component Analysis, and Support Vector 

Machines. This approach should allow the robot to recognize the successful completion of a task. Additionally these 

techniques have been applied to tactile sensor data to determine the type of contact being made with an object as 

well as localize the object’s position within the hand. 

Gravitational Lensing of the Cosmic Microwave Background 

Ryan Keeley 

Mentors: Marc Kamionkowski and Fabian Schmidt 

Massive astronomical bodies deflect light causing a discrepancy between the actual position of the source and the 

observed position of the source. This gravitational lensing can be used to study the mass distribution of these 

astronomical bodies. The Cosmic Microwave Background (CMB) is useful as a source in such an analysis as its 

position (the last scattering surface) is known. This project will produce profiles of a lensed CMB distribution from a 

given unlensed distribution. This information will then be used to reconstruct the mass distribution of the lensing 

43

astronomical body. In the past, such analysis has been done with the assumption that the CMB can be modeled as 

a simple gradient on small scales. This project will do the full calculation for some general distribution in the 

unlensed CMB. 

Experimental Design for Two-Plasmon Quantum Interference Measurements 

Yousif A. Kelaita 

Mentors: Harry A. Atwater and Jim Fakonas 

In the last several decades, experiments in the field of quantum optics have opened new windows into the 

quantum nature of light. With a relatively simple setup, one can perform a suite of experiments that collectively 

demonstrate various quantum effects including quantum interference and entanglement. It is not yet known, 

however, whether surface plasmons—quanta of light that are sustained by the collective oscillation of a metal’s free 

electrons—exhibit all the same quantum effects as photons in free space do. This project proposes a set of 

experiments aimed at filling the gaps in our current understanding of plasmons at the quantum limit. To this end, 

the project requires work in precise laser alignment and calibration of collection optics, the design and 

implementation of a full fabrication process for surface plasmon waveguides on silicon nitride, and the design and 

construction of various custom parts. As a first goal, we aim to demonstrate two-plasmon quantum interference, an 

effect that has been predicted by analogy to two-photon interference but not yet observed. In the long term, we 

hope that this experiment and others will answer some of the many questions regarding the quantum nature of the 

plasmon. 

15 

N NMR Study of Intramolecular Hydrogen Bonding in Imidazole-4-Propanoic Acid 

Michael J. Kenney 

Mentors: John D. Roberts, William Carroll, and Mrinmoy Nag 

The presence of an intramolecular hydrogen bond in imidazole-4-propanoic acid in acetonitrile has been verified 

using 15 N NMR and 1 H NMR. Hydrogen bonding (deprotonated propanoic acid group) and non-hydrogen bonding 

(protonated propanoic acid group) cases were studied in acetonitrile-d3 using a 15 N-labeled 2-mercaptoimidazole- 

4-propanoic acid compound. Studies showed a downfield shift difference for N3 of 9 ppm, likely the result of an 

intramolecular hydrogen bond between H3 and the carboxylate anion. A CD3CN/H2O mixed solvent study, where 

the fraction gauche was measured as water was added to a solution of imidazole-4-propanoic acid in CD3CN in 5 μL 

increments, revealed the amount of water necessary to break the intramolecular hydrogen bond in imidazole-4propanoic 

acid. At 12 equivalents of water to one equivalent of imidazole-4-propanoic acid, the fraction of gauche 

conformers becomes statistical (0.66) indicating that the hydrogen bond is broken. 

Integration of Vertical-Cavity Surface-Emitting Lasers on CMOS Chips 

Aroutin Khachaturian 

Mentor: Axel Scherer 

Integration of vertical-cavity surface-emitting lasers (VCSEL) with the CMOS chip is an important step toward 

communicating with implantable biological sensors with high data transfer rates. VCSEL and CMOS chips have been 

previously bonded together through various fabrication techniques and an alternative approach is examined during 

this project. The initial goal of the project was to bond a 200µm thick VCSEL to a CMOS chip. This was achieved 

through a combination of photolithography and electroplating techniques so that the CMOS chip’s contact can be 

raised to the top contact of the VCSEL in order to make a connection between the two. Afterwards, they were 

tested in contact with skin, blood, and neural tissue using different types of data modulation techniques (e.g. 

amplitude and frequency modulation) and their efficiency is examined in terms of power and chip area consumed. 

Understanding the Role of Cilia in Hedgehog Signaling Pathway 

Sohini Khan 

Mentors: Rajat Rohatgi and David Tirrell 

The Hedgehog signaling pathway is known for its importance in embryonic development and cancer from 

medulloblastoma to basal cell carcinoma. Additionally, it requires primary cilia for full activation; however, in this 

pathway, it is still unknown which specific parts of the pathway require cilia. This project focuses on whether the 

interaction between Protein Kinase A (PKA) and Gli transcription factors depends on cilia. For this study, stable cell 

lines were generated containing a dominant negative form of Kif3A, a kinesin motor protein. These cells, in which 

primary cilia are absent, will then be used to explore the interaction between PKA and Gli proteins. Specifically, 

dominant negative PKA was tested for its ability to induce the Hedgehog pathway in the absence of cilia which 

proved to be ineffective. These cells were, however, able to activate the Hedgehog pathway after overexpression of 

Gli2. Also, PKA-insensitive mutants of Gli2 and Gli3 will be tested in the future for their ability to translocate into 

the nucleus and induce the expression of target genes. 

44

Design and Analysis of an in vitro Exact Adapter 

Ishan Khetarpal 

Mentors: Richard Murray and Jongmin Kim 

In the strict biological sense, by manipulating increasingly complex circuits, we gain much insight into how cells 

function. In terms of Computer Science, it has been shown that these biological processes are Turing complete, 

meaning that in theory, these systems can be used for computation. Several circuits have already been created, 

including oscillators, switches, and logical gates. The purpose of this research was to create and analyze an exact 

adapter. 

An exact adapter is a system whose outputs vary only transiently on its inputs. To implement the exact adapter in 

vitro, a gene was used that coded both for the creation of a strand fluorescent in the presence of malachite green 

(rMG), and its inhibitor (iMG). Effectively, the inhibitor degrades the fluorescent molecule after a transient period of 

visibility. One of the enzymes required for this process is ribonuclease R (RNase R). This enzyme is not fully 

understood, as it fails to operate on select fragments that are seemingly unrelated. Tests of degradation of rMG 

and iMG by RNase R were performed and analyzed via spectroflurometer and gel electrophoresis, to ensure the 

components would work correctly in the circuit. The circuit was then engineered with these components. 

Synthesizing a Pyrroloindoline Derivative Using a Formal (3+2) Cycloaddition Reaction 

Brian Kim 

Mentors: Sarah Reisman and Jane Ni 

A formal (3+2) cycloaddition reaction catalyzed by an (R)-BINOL · SnCl4 complex can be used to make different 

pyrroloindoline derivatives, the core structure of several biologically active products. 1,3-dimethyl-5-methoxyindole 

and methyl 2-trifluoroacetamidoacrylate were good substrates under the given reaction conditions, while 

1,3-dimethyl-5-bromoindole did not react due to the electron withdrawing effects of the bromine substituent. 

Thermodynamic Properties of Modified LiFePO4 for Li-ion Batteries 

Laura Kim 

Mentors: Brent Fultz, Hongjin Tan, and Hillary Smith 

Rechargeable lithium-ion batteries are one of the most popular batteries in consumer electronics owing to their 

high specific energy, long life cycles and reduced weight and volume. LiFePO4 has attracted significant interest 

because it is inexpensive, nontoxic, and made from naturally occurring minerals. Improvements to the intrinsic low 

electronic conductivity of LiFePO4 can be made by ball-milling to reduce the particle size and coating the material 

with carbon. We seek to understand the effect of these modifications on the thermodynamic properties of these 

materials. First, nano-sized LiFePO4 was prepared by ball milling. Then, the bulk and the nano-sized materials were 

characterized by x-ray diffraction. Next, coin cells were assembled in a glove box and their rechargeability and 

functionality were tested by a cell cycler. Finally, the coin cells were placed on an Electrochemical Thermodynamic 

Measurement System (ETMS) to compare the thermodynamics properties of the bulk, nano-sized, and carbon 

coated materials. Increased electrical conductivity in nano-sized particles and carbon-coated particles versus bulk 

and uncoated material has been observed, and ETMS measurements indicate that the configurational entropy of 

these materials has changed as a result of these modifications. 

Solid-State NMR Probe and Rotor Design at High Pressure and Corrosive Environment for Extended 

in situ Experiments 

Sangkyu Kim 

Mentors: Sonjong Hwang and S. I. Zones 

Zeolites make up an enormous class of three-dimensional aluminosilicate materials. The acid form of most zeolites 

is catalytic due to its ability to confine molecules in small spaces. The proposed mechanism under the condition of 

hydrofluoric acid is the following: 

SiO2 + SDAF + HF + H2O -> SDAF-MS + SDAF + HF + H2O 

where SDAF stands for the structure directing agent with fluorine, and MS for the molecular sieve. 

The in-situ method, synthesizing zeolite inside the NMR spectrometer, can offer important clues to understand 

zeolite frameworks and crystallization conditions. In this paper, the construction and performance (in zeolite 

syntheses) of high pressure magic angle spinning (MAS) inserts and rotor bodies made from the polymers PEEK 

and Teflon are described. The inserts are designed to fit inside standard commercial Bruker MAS rotors, and rotor 

bodies are designed as substitutes for the Bruker rotors. 

45

Progress Towards the Total Synthesis of Acutumine 

Seohyun (Chris) Kim 

Mentors: Sarah Reisman and Raul Navarro 

Acutumine is a propellane alkaloid natural product that possesses interesting biological properties, including 

selective T-cell cytotoxicity and antiamnesic activity. Structurally, acutumine comprises a highly fuctionalized azapropellane 

core, making it a challenging target for total synthesis. The goal of our studies is to achieve an efficient 

enantioselective synthesis of acutumine. Our lab recently reported a highly diastereoselective 1,2- addition of 

organometallic reagents to benzoquinone-derived sulfinyl imines, a strategy that allows for the preparation of 

enantioenriched dihydroindolones. These dihydroindolones serve as versatile intermediates in the preparation of 

propellane alkaloids; therefore, their gram-scale synthesis will facilitate synthetic studies en route to acutumine. 

We have achieved a multi-gram synthesis of the benzoquinone imine and grignard reagent required for our 

synthesis of acutumine. Current focus lies in the preparation of multi-gram quantities of sulfinamide addition 

product which can be readily converted to the dihydroindolone. By preparing larger amounts of addition product, 

we can thoroughly study the key steps in our proposed synthesis. 

Efficient Monte Carlo Methods for Radiative Transfer Modeling of Protoplanetary Disks 

Stacy Kim 

Mentor: Neal Turner 

When a protostar is born from a swirling, contracting cloud of dense gas, it leaves behind a rotating dusty and 

gaseous disk, produced by accretion of the surrounding interstellar material and the demands of angular 

momentum conservation. Translating observations of the disks into information about the conditions for planet 

formation and migration requires model temperature profiles and spectral energy distributions, which can be 

obtained relatively straightforwardly by applying the Monte Carlo method due to the stochastic nature of radiative 

transfer processes and the method’s adaptability to complex disk shapes. In practice, certain regions of the disk 

are under-sampled, such as the optically thick interior, or are of particular interest, such as the snow line (where 

water vapor condenses into ice) and the area surrounding a protoplanet. To improve the sampling, photon packets 

can be preferentially scattered and reemitted toward the preferred locations. This comes at the cost of deweighting 

packet energies by the same factor that the probability a packet travels in a particular direction was 

augmented. I present an analysis on the effectiveness of a variety of weighting schemes in terms of improved 

sampling, accuracy, and simulation speed. 

Electroporation of Engineered Bacterial Artificial Chromosomes (BACs) in Early Chick Embryos 

WoongBin Kim 

Mentors: Marianne Bronner and Marcos Simoes Costa 

Chicken embryos are important in embryology due to its similarity with human embryos. Making transient 

transgenic chicken embryos is the object of this project and it is an important tool for finding regulatory sequences 

of chicken embryos. Electroporation which increases the permeability and the electrical conductivity of a cell 

membrane by external electric field is the key technique to make transgenic chicken embryos. Electroporation of 

chicken embryos with different plasmid constructs has been established in high efficiency while electroporating 

BACs is much more difficult due to its big size. A [pCAGGS-eGFP (enhanced green fluorescent protein)-Kanamycin 

resistance gene] was made by ligation of C8 vector and PCR (polymerase chain reaction) product which were 

modified by XhoI and NotI restriction enzymes. The cassette was amplified with PCR. Purified PCR product of the 

cassette and BAC were electroporated into a strain of bacteria that expresses a recombinase. Bacterial cells which 

contain the BAC were selected with Kanamycin. Engineered BACs were extracted and purified with a BAC Maxiprep 

Kit. Electroporation of chicken embryos with BACs were processed under various conditions (different voltage and 

number of pulses). Survival of chicken embryos and expression of eGFP were observed after culturing chicken 

embryos. 

Limits on the Baryonic Contribution of Dark Matter via Astrometric Microlensing of Quasars 

Christina M. King 

Mentors: Charles J. Naudet and Christopher S. Jacobs 

Dark matter makes up 83% of the universe, though its composition remains a mystery. It is likely that dark matter 

partly consists of Massive Compact Halo Objects (MACHOs). The preferred methods for detecting MACHOs are 

photometric and astrometric microlensing. Existing MACHO surveys use nearby stars to detect photometric lensing, 

but it is more likely to observe an event when using extragalactic sources to search for astrometric lensing. 

Decades of Very Long Baseline Interferometry (VLBI) measurements have provided an extensive dataset of 

extremely fine measurements of positions of extragalactic radio sources. An upper limit for the precision required 

to detect microlensing events is established by comparing a trendline of a source’s motion with various Gaussian 

curves, which simulate expected microlensing signatures. Using VLBI data, is possible to set upper limits for 

detection of microlensing events on time scales of 1–30 years. The dominant systematic error is expected to arise 

46

from source structure, and can be isolated by rotating the coordinate system to analyze data parallel and 

perpendicular to the jet. Simulated astrometric events can distinguish smooth, predictable positional changes 

associated with gravitational lensing from erratic signatures produced by structural changes. 

Self Anchoring Drilling/Coring/Sampling System for in situ Planetary Applications 

Jonathan King 

Mentor: Aaron Parness 

The objective of future NASA exploration missions is in situ sampling and analysis of various astronomical bodies in 

the solar system (including Mars, Europa, Titan, comets and asteroids), and the return of samples to Earth for 

further study. Conventional drills require a relatively high axial preload or weight-on-bit, which limits their use in 

microgravity and on inverted surfaces. To resist these large reaction forces during drilling, an anchoring 

mechanism that effectively clings to rough and rocky surfaces has been integrated with a coring drill. The system 

consists of a compact rotary-percussive drill with a custom coring bit, a linear translation mechanism and spring 

system that will maintain preload and weight-on-bit, and a compliant anchoring mechanism that employs an omnidirectional 

array of micro-spines that cling to a surface by opportunistically catching on surface asperities. A 

prototype based on this design work is currently in fabrication, to be followed by testing and data collection. 

Additional efforts to develop a comprehensive model of the system are recommended so that further developments 

and predictions can be corroborated both theoretically and experimentally. 

The Phase and Proton Conduction Behavior of Solid Acids of the Form CsIx(H2PO4)1-x and 

Ba(3-x)MxHx(PO4)2 for M = Na, K, Rb, Cs 

Daniil Kitchaev 

Mentors: Sossina M. Haile and Ayako Ikeda 

The solid acid electrolyte that is currently believed to be optimal for fuel cell applications, CsH2PO4 (CDP), suffers 

from a low dehydration temperature, which limits its operating range, and water solubility, which causes problems 

during fuel cell startup and shutdown. We investigated the feasibility of two alternative materials: a doped variant 

of CDP that was expected to have a higher dehydration temperature, and water insoluble electrolytes based on 

Ba3(PO4)2. To investigate these samples, two atmosphere controlled furnaces were constructed. Then, electrolytes 

of various compositions were prepared by precipitation and high-temperature annealing, and examined via x-ray 

diffractometry, SEM/EDS, AC impedance spectroscopy and thermogravity. The solubility limits of CsIx(H2PO4)1-x, 

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 

these compounds was confirmed. Furthermore, the preferred dopant site occupancy in the latter four compounds 

and its relation to the materials’ conductivity was determined. Finally, it was found that the effect of the type of 

dopant ion on the conductivity of the barium phosphate compounds was much less than could be expected. 

Characterizing and Investigating Spatio-Temporal Expression Patterns of Gene Expression in the Veg1 

and Veg2 Regions of E. tribuloides 

Rebekah Z. Kitto 

Mentors: Eric H Davidson and Eric Erkenbrack 

The goal of this project is to temporally and spatially profile early endomesodermal patterning events in the slate 

pencil urchin Eucidaris tribuloides. These data will be compared to those collected for other echinoderms in order to 

cast light on the conservation of early endomesodermal patterning events within the Echinodermata. 

The aim of this project is to produce qPCR primers and WMISH probes for otx, gataE, foxA, hox 11/13b, eve, 

blimp1, and wnt8, and then employ qPCR and whole-mount in situ hybridization (WMISH) to characterize the 

expression of these genes during E. tribuloides development. The data collected can be compared to S. purpuratus 

and other echinoderms to determine the similarities and differences in development between these two and other 

species despite their evolutionary divergence. Additionally, we aim to determine the role of Notch signaling in early 

endomesodermal patterning. To do this we will knockdown Notch signaling by treating developing embryos with the 

GSK-3 inhibitor DAPT. Ideally, the results obtained will provide information linking S. purpuratus and E. tribuloides 

that can shed light on the evolutionary divergence of these two species. 

Asbestos-Induced Alveolar Epithelial Cell Death: An Investigation Into the Role of the ASK1/JNK 

Pathway in Apoptosis 

Kathryn B. Knister 

Mentors: David Kamp, Paul Cheresh, and David Chan 

Asbestos fibers cause pulmonary fibrosis and cancer by mechanisms not fully established. Asbestos induces 

alveolar epithelial cell (AEC) death (apoptosis) via the p53- and mitochondria-regulated (intrinsic) death pathway. 

Mitochondrial and endoplasmic reticulum (ER) interactions regulate intrinsic apoptosis by activating ASK1/JNK 

signaling, but it is unknown whether asbestos induces this response. We examined whether a JNK inhibitor 

attenuates asbestos- and ER stress (using thapsigargin)-induced AEC DNA fragmentation (signifying apoptosis) and 

ASK1/JNK signaling (gauged by measuring protein levels with Western blotting). Our preliminary results were 

47

inconclusive because the JNK inhibitor alone was slightly toxic. This should be overcome by adjusting culture 

conditions and JNK inhibitor dose/diluent. Using transient transfection of an ASK1 dominant/negative plasmid (to 

inhibit the ASK1 gene), we will address similar endpoints. Future experiments will examine whether a p53 inhibitor, 

pifithrin, blocks asbestos-induced ASK1/JNK signaling and ER stress response. If the inhibitors block asbestos- and 

ER stress-induced AEC apoptosis and ASK1/JNK signaling, this will implicate a critical role for the ASK1/JNK 

pathway in mediating mitochondrial-ER-p53 crosstalk that can lead to intrinsic apoptosis. Our studies may also 

show that ER stress is an adaptive response to mitochondrial free radicals but independent of downstream 

ASK1/JNK and p53 activation. 

The Role of Elastic Interactions in Bacterial Chemoreceptor Organization and Function 

Peter D. Koch 

Mentors: Rob Phillips and Christoph A. Haselwandter 

We investigate the role of membrane-mediated elastic interactions in the spatial organization and function of 

bacterial chemoreceptors. It is believed that the smallest functional units of chemoreceptors are trimers of dimers 

which, upon binding to periplasmic ligands, initiate the chemotaxis signaling cascade. Many types of membrane 

proteins, including chemoreceptors, exhibit a hydrophobic mismatch with the surrounding lipid bilayer, thereby 

inducing thickness and curvature deformations of the lipid membrane. The energies of the membrane deformations 

induced by chemoreceptor trimers can be calculated using elasticity theory. We numerically compute these 

energies using a finite difference scheme. We first study the energetics of two interacting chemoreceptor trimers 

and find that our numerical results are in good agreement with previous analytical calculations, as well as previous 

numerical studies of related problems. Next, we study the energetics of chemoreceptors in lattice structures, as 

recent electron cryo-tomography studies have shown that chemoreceptor trimers form regular lattices which 

appear to be conserved across the bacterial kingdom. In particular, we examine how the total membrane 

deformation energy depends on the type of lattice symmetry and on the orientations of the chemoreceptor trimers 

within the lattice. Lastly, recent in vivo FRET studies have shown that bacterial chemoreceptors engage in 

cooperative signaling, and various theoretical models have suggested that this cooperative behavior is crucial for 

the high sensitivity of the chemotaxis signaling pathway. We study whether elastic interactions may promote such 

cooperative behavior among chemoreceptors, and whether lattice symmetry and trimer orientation may provide a 

mechanism for the observed size of cooperative signaling teams of chemoreceptors. 

Investigation of the Reported Separation of Nuclear Spin Isomers of Water Through Selective 

Adsorption 

Theodore Koenig 

Mentors: Geoff Blake and Daniel Holland 

Water exists as two nuclear spin isomers, ortho- and para-water, having different magnetic properties. At ambient 

temperatures, the isomers remain in thermodynamic equilibrium with a 3:1 ortho/para ratio. Various experiments 

claim to have shown significant enrichment of one isotope over the other through selective adsorption of water 

vapor over various sorbent materials. However, few if any of these experiments have been independently verified, 

and the existence of the claimed effect is still contested. The purpose of this project is to rigorously test the 

claimed effect, using various spectroscopic techniques upon which the effect was originally claimed to have been 

observed. To this end, a vacuum apparatus with sorbent and sounding cells was constructed, heeding claimed 

material incompatibilities with the claimed effect (i.e. no magnetic parts). A tunable diode laser absorption 

spectrometer was built. Further, an FT-IR was setup to complement the near-IR diode measurements and a 

procedure for data analysis of these results was chosen. Laser absorption specific to each of ortho- and para-water 

has been observed. The integrated areas of these absorption peaks will be used to measure the relative ratios of 

ortho- and para-water over the course of an adsorption experiment, as was performed in the original literature 

claims. The results of these experiments should inform the debate about these claims, with potential impact in the 

astronomical and earth sciences. 

Damage Mitigation and Performance Improvement of Vertical Axle Wind Turbines Using Compliant 

Wind Blades 

Chan-Hee Koh 

Mentors: Morteza Gharib and Julia Cossé 

We investigate the effect of using compliant blades utilizing unsteady aerodynamics on drag based vertical axle 

wind turbines. A model wind turbine was built using various compliant materials for the blades, potentially 

increasing the efficiency, performance, and durability beyond what conventional drag based wind turbines could 

achieve. The design is a simple vertical axle wind turbine with two flat plates as blades. The blades are designed so 

that their angle can be changed relative to the frame. The blade angles were changed to see the various effects the 

wind had on the drag. These angles would show how the flexible materials deform in the wind, resulting in different 

areas that face the wind. After measuring the angular velocity and the torque of the turbine in a wind tunnel, we 

were able to calculate the power output and the power coefficient, also known as the efficiency of the turbine, to 

48

get an idea of how our model compared with conventional turbines. We used two materials of different flexibility, 

aquaplast plastic and rubber, as well as a rigid control, aluminum sheet. We compared the obtained values with 

one another as well as the control case, which would not rotate on its own for any configuration. 

The turbine showed a periodic inconsistency in the angular velocity as the power output oscillated through each 

rotation due to flapping of the flexible blades. The unsteady effects caused by the flexible flapping significantly 

improved the performance of the wind turbine. This is evident in the fact that the flexible materials were able to 

rotate in configurations that the rigid material was not. However, the design had a number of problems that 

prevented the rigid material from turning. One aspect was that the design made the blades symmetric, which 

meant that there would be an equal amount of drag force on both sides of the blades in the opposing direction, 

resulting in stall. Conventional savonius blades would be curved to ensure that there is a difference in drag forces 

so that the turbine would rotate. Another problem was that the alignment of the axle was not optimum, and we 

undoubtedly lost efficiency to friction. Nevertheless, the results delineate the potential improvements in 

performance of current turbines. 

Neuroeconomic Foundations of Social Learning in Economic Choice 

Swadhruth Komanduri 

Mentors: Christof Koch and Michael Hill 

Modern neuroeconomic experiments primarily consist of studying reward-related systems by having subjects make 

decisions that will directly alter their future reward. These studies have shown that the orbitofrontal cortex, the 

amygdala, and the anterior cingulate have neurons which encode for relevant variables, such as the prediction 

error, the value of a choice, or the winnings of a certain trial. Such studies uncover underlying neural pathways for 

learning from one’s own mistakes and triumphs. These studies, however, do not take into account the impact of 

others’ choices and outcomes which help in the decision making process. To understand this impact, I have been 

working on an experiment to discover the neural basis of social learning. In this experiment, each subject plays a 

multiplayer game with two computer controlled characters. Each player must choose from one of two decks, 

knowing that one deck has a 70% chance of winning money and the other a 30% chance. The outcomes are 

revealed publicly, allowing the subject to gain information and learn while not actively participating. The goal of the 

project is to understand the neural network underlying this type of social learning and to compare this network 

with those pertaining to personal reward-related learning. Preliminary analysis shows that there are responsive 

neurons in the orbitofrontal cortex and amygdala, as expected. Moreover, there are also responsive neurons 

present in the entorhinal cortex, parahippocampal gyrus, and anterior hippocampus, areas potentially involved in a 

distinct neural pathway for social learning. 

Elucidating NF- κB Oscillation at a Single Cell Level 

Philip Kong 

Mentors: David Baltimore and Devdoot Majumdar 

Nuclear factor kappa-B (NF-κB) is a transcription factor that orchestrates the activation of several immune 

response genes in response to inflammation. It is found in virtually all animal cells and plays a key role in 

responding to infection. One hallmark of NF-κB is its oscillation from the cytoplasm to the nucleus on an hourly 

timescale. The dynamics of this periodic nuclear entry of NF-κB and its resulting gene expression pattern have 

been studied in populations of cells. However, there exists considerable asynchrony in the timing of NF-κB 

oscillations among members of a population of cells; therefore, the study of the NF-κB transcriptional response is 

complicated by cell-to-cell heterogeneity. Here, we ask: what is the NF-κB induced gene expression profile as a 

function of time at the single cell level? By using real-time quantitative PCR (RT-qPCR) and time-lapse microscopy, 

we connect NF-κB oscillatory movement and NF-κB mediated temporal gene expression profiles in response to 

inflammatory stimuli. Our results indicate that we are able to develop a qPCR system that can detect genes at a 

single cell level, which has only few picograms of RNA. With the methods developed herein, we explore the 

functional relevance of NF-κB’s oscillatory behavior. 

Search for Highly Dispersed Transient Radio Pulses 

Michael Kossin 

Mentor: Walid Majid 

This project is part of an effort to develop and test an accurate and efficient dispersed transient radio source 

detection system. Certain phenomena eject radiation that may only be observable for a fraction of a second. As 

this radiation travels through space, it may undergo dispersion before it arrives at our detectors. Dispersion is an 

effect experienced by waves as they travel through an ionized medium: their group velocities become altered by an 

amount related to their frequencies. If a burst of radiation across a range of frequencies undergoes dispersion, the 

higher frequency radiation will arrive at an observer sooner than lower frequency radiation does. 

49

To detect radio signals received after experiencing an unknown amount of dispersion, we use a program utilizing a 

trial and error process to search for dispersed radio signals in a data set by varying a guessed dispersion measure. 

The program translates signals in the dataset in time by a function of their frequencies and searches for instances 

of pulses existing at the same time across these adjusted frequency bands. We test the efficiency of this program’s 

ability to detect pulses with different dispersion and noise levels with artificial datasets. 

Sputter Deposition and Electrochemical Characterization of Model Thin-Film Metal Electrodes for Solid 

Acid Fuel Cells 

Jeffrey Kowalski 

Mentors: Sossina Haile and Aron Varga 

Solid Acid Fuel Cells (SAFCs) are promising alternative energy conversion devices since their high efficiency helps 

to address the dwindling supplies of fossil fuels and their effects on the environment. The specific problem 

examined in this project is the performance of several metal electrocatalysts (platinum, palladium and nickel) in 

SAFCs constructed with CsH2PO4 as the electrolyte. Differences in electrochemical activity under conditions of 

varying electrocatalyst film thickness were investigated. The thickness of the film was controlled by varying 

deposition times. The physical attributes of the films were then characterized using scanning electron microscopy 

and x-ray diffraction. A.C. impedance spectroscopy was used to elucidate the performance for hydrogen oxidation. 

Cyclic voltammetry was also used to determine the voltage required to overcome the formation of Pt oxide when 

oxygen is present in the atmosphere of measurement. A comparison of XRD patterns before and after keeping an 

intimate mixture of NiO and CsH2PO4 powder at fuel cell operating conditions for 18 hours suggests a reaction 

occurring between the components. These findings suggest that Ni cannot be used as a catalyst for SAFCs that use 

CsH2PO4 as the electrolyte. 

Characterization of High Speed Analog-to-Digital Converters for Deep Space Network Downlink 

Array IF Sampling 

Kara Kundert 

Mentors: Robert Navarro and Melissa Soriano 

The NASA Deep Space Network provides services for navigation and telemetry of spacecraft beyond Earth orbit. 

Currently, NASA is starting work on a new downlink array—DSN Downlink Array (DDA)—which will be focused on 

bringing more of the processing from the analog to the digital domain. This focus will take advantage of modern 

signal processing techniques and improve the accuracy of the DSN data. This project is primarily focused on the 

evaluation of several candidate analog-to-digital converters to determine which provides the best overall 

performance for the needs of the DDA. Of particular concern is the flatness of the gain and group delay of the 

converter over the IF bandwidth, as excessive variation will interfere with the beam forming that occurs when 

combining the signals from many antennas. In addition, converter nonlinearity and noise will be evaluated as these 

could limit the DDA’s ability to resolve weak signals from our satellites in deep space, particularly in the presence 

of large interferers. 

d-Wave Superfluidity in Two Dimensional Optical Lattices of Ultracold Polar Molecules 

Kevin Kuns 

Mentors: John Preskill and Alexey Gorshkov 

It has long been proposed that simulations of the t-J model by ultracold atoms in two dimensional optical lattices 

could be used to shed light on the physics of high-temperature superconductors and the underlying d-wave 

superfluidity. Recent work on ultracold polar molecules, governed by the t-J-V-W model, suggests that molecules 

may be better suited for studying d-wave superfluidity due to stronger interactions and larger tunability of the 

system. We compute the phase diagram for polar molecules in a two dimensional square lattice in a certain 

perturbative limit. In the simplest experimentally realizable case where there is only tunneling and an XX type 

spin-spin interaction, we identify the parameter regime where d-wave superfluidity occurs. We also study the 

effects of density-density, spin-density, and Ising type spin-spin interaction terms. Our findings can be used as a 

guideline for tuning the experimental system of polar molecules into the d-wave superfluid phase. 

Effect of Nanostructured Topologies on Air Flow Properties Over Vertical Axis Wind Turbine Blade 

Surface 

Adela Kuzmiakova 

Mentors: Beverley McKeon and Julia Greer 

During its lifetime, a vertical axis wind turbine (VAWT) encounters a variety of meteorological conditions – be they 

icing, thunderstorms, or heavy rains. Yet still, little remains known about the aerodynamic influence on VAWT’s 

performance due to heavy rain. One of the key mechanisms responsible for degradation of VAWT’s efficiency is the 

“effective” time-dependent roughness on the airfoil's surface due to uneven distribution of water film. Here, we 

propose nano-structuring of thin layers on the top of the turbine’s surface in order to improve its energy efficiency. 

The thin films are exposed to a simulated precipitation design and the change in their surface distribution is 

50

measured. Then, the layers are applied on the airfoil’s surface and tested under comparable conditions. The 

preliminary study of air flow properties over the nano-structured topologies is related to their wetting capabilities in 

order to gage the material selection and processing for optimal VAWT’s energy efficiency. 

Asthma and Thunderstorms: Observing Pollen Grain Rupture in Laboratory Simulated Conditions 

Sung Min Kwon 

Mentors: Richard Flagan and James House 

Thunderstorms have been known to be associated with widespread asthma attacks in recent years. Whole pollen 

grains are too large (20 to 100μm in diameter) to reach the lower airways, but at high relative humidity (~96%), 

they can rupture through osmotic shock and release cytoplasmic material small enough (0.2 to 3.5μm in diameter) 

to invade smaller bronchial tubes of the lung. With intermolecular van der Waals forces sufficiently lowered after 

evaporation, winds deposit these fragments in the human respiratory tract, likely causing asthma. Studies have 

found that a synergic combination of electricity and humidity leads to more pollen rupture than humidity alone. 

The purpose of this study is to determine the optimal conditions for pollen grains to rupture. An apparatus was 

designed from acrylic resin to observe the behavior of pollen grains subjected to humid air and varying electric 

fields under a light microscope. The humidity will be measured by a chip hygrometer/thermistor inserted into a 

controlled airflow. The electric field will be calculated from input voltage and distance between two copper wire 

leads. The pollen samples will be taken from the Chinese elm (Ulmus parvifolia), a major source of airborne 

allergens in Pasadena that blooms in August. 

Building Mechanical Actuators From C2C12 Skeletal Muscle Cells 

Thomas Kwong 

Mentors: Mory Gharib and Hesham Azizgolshani 

Muscle cells are linear contractile actuators that cause mechanical motion in biological systems. The directional 

action of muscles derives from coordination on all levels of muscle structure, from the sarcomeres in each cell to 

the myofibers in the tissue. That directionality is crucial for allowing muscles to efficiently deliver forces. However, 

C2C12 cells grown in culture lack the organization of sarcomeres. In order to design mechanical actuators powered 

by muscle cells, a method must be developed in order to coax cells grown in culture to adopt the directionality 

found in the muscle tissue of organisms. To this end microcontact printing (µCP) enables linear patterning of 

fibronectin onto a culture surface, thereby promoting selectively favorable cell attachment in specific linear 

patterns. Muscle cells cultured on fibronectin µCP surfaces display linear directionality when differentiated into 

myofibers. µCP is the primary method for patterning cell growth, but its limitations motivate study of other 

approaches to expand patterning capabilities. In specific, µCP is limited to planar surfaces. One promising approach 

utilizes the fact that cells placed in contact with a favorable surface can transfer to it from their culture surface. 

This method invites the possibility of transferring cell patterns achieved through µCP to untreated surfaces by 

contact. Further studies are being performed with the goal of patterning cells on cylindrical surfaces. 

Determining the Stellar Properties of Herbig-Haro Energy Sources 

Stephanie Laga 

Mentor: Lynne Hillenbrand 

Optical spectroscopy of T Tauri stars, or very young stars, continues to reveal evidence used in developing theories 

about the early stages of low mass star formation. In particular, this project focuses on those Class I objects that 

power Herbig-Haro flows, spatially resolved jets or jet shocked materials that signify strong outflow from the young 

star. In a study of a small set of such objects in the Taurus Molecular Cloud, White and Hillenbrand (2004) found 

that the stars that power these Herbig-Haro flows have stellar and circumstellar properties generally 

indistinguishable from similar mass stars that do not power these flows. For this project, the protostars of interest 

were observed from 6300 to 8800 Angstroms in wavelength using the Keck telescope. The emission and absorption 

lines in the spectra produced by these objects will reveal specific properties of the enveloped protostars. Stellar 

properties related to the central star like spectral type the rotational velocity and circumstellar properties like 

spectral veiling tell us about accretion rates and various emission line strengths, important to determining theories 

about star formation. 

Elastostatic Solutions for Realistic Slip and Stress Around Mode II and III Cracks 

Valère R. Lambert 

Mentors: Jean-Philippe Avouac and Sylvain Barbot 

A common practice for analyzing the displacement and stress caused by slip on buried faults is to discretize the 

fault area into finite size patches of assumed uniform slip and use the corresponding Green’s functions to predict 

strain at the surface or within the bulk of the medium. This formalism is commonly used to invert observation of 

co-seismic deformation for fault slip, in afterslip analyses, stress transfer studies and in numerical models of the 

seismic cycle, such as the Uniform California Earthquake Rupture Forecast (UCERF). The discretization of the fault 

into areas of uniform slip introduces stress singularities at the edges of each patch, which might make the results 

51

quite sensitive to the choice of the discretization scheme. Yet, little attention has been paid to this issue so far. 

Here, we derive analytic expressions of the Green’s functions, for which the Okada’s solution is a special case. We 

derive a full-space formulation for linearly interpolated fault slip distributions, which allows discretization of any 

fault slip distributions without introducing stress singularities. We investigate the case of a finite crack with a 

uniform distribution of stress on the fault. We show that the Okada solution provides an inadequate representation 

of the stress distribution for mode II and III cracks, even in the limit of infinitely small discretization. We introduce 

a critical size of fault discretization below which stress singularities dominate and bias the distribution. Our 

tapered-slip solution does not suffer from these shortcomings, suppresses all singularities and converges to a 

uniform stress with smaller discretization. Our results demonstrate the need to resolve the areas of stress 

concentration in our models of fault slip evolution. 

Quantitative Analysis of the Phototaxic Response of Artemia 

Megan C. Larisch 

Mentor: John O. Dabiri 

Biomixing- ocean mixing brought about by the movement of animals- has been the source of much debate over the 

past several years. Recent work has shown that biomixing via vertical migration may be a significant contributor to 

overall ocean mixing. The implications of such mixing are far-reaching. If biomixing were found to be significant, it 

would affect climate models and our understanding of global circulation. Additionally, vertical mixing by small 

crustaceans, such as krill, could potentially bring deeper, more nutrient-rich water to the surface, feeding the 

plankton the krill eventually eat. 

It has been shown that certain models of vertical migration result in significant ocean mixing. However, the work 

has been mainly theoretical, and has yet to be replicated in a quantitative study. As a result, the viability of 

biomixing as a significant contributor to overall ocean mixing is still undetermined. Since vertical migration holds 

the most promise for biomixing, we have set up an experiment to simulate the migration of bottom-dwelling 

animals to the surface. A 1-meter tall graduate cylinder will be filled with different salinities of salt water in a 

stratified manner, so that the densest, saltiest water makes up the bottom layer and the lightest, least salty water 

makes up the top layer. Artemia (brine shrimp) will be made to swim from the bottom to the top of the column by 

taking advantage of their phototaxic behavior. 

While some studies have addressed the phototaxic behavior of brine shrimp little has been done to quantitatively 

study this attraction. To best understand any mixing by the shrimp in the above experiment, we must first 

understand the phototaxic behavior of the shrimp. 

Patterned Quantum Dots for Nanotechnology 

Pawel Latawiec 

Mentors: Axel Scherer and Andrew Homyk 

Quantum dots in solid state systems have been an active area of research for years now, showing tremendous 

potential in applications to next-generation communication and computing systems. Traditional methods of 

fabrication of quantum dots on semiconductor substrates have largely been unable to place them selectively, 

resulting in poor device performance. This project explores novel methods for the design and construction of 

patterned quantum dot systems. Of particular interest are applications to transistors and photonic crystals. Novel 

vertical transistor devices were fabricated which operated via quantum mechanical resonant tunneling. This was 

achieved by a self-terminating oxidation of a scalloped silicon nanopillar followed by vertical gate construction. 

Measurements of preliminary structures indicate results consistent with theoretical predictions. Patterned quantum 

dots constructed for photonic crystals seek to maximize crystal-dot coupling and bring interaction into the cavity 

quantum electrodynamic regime. This is achieved by wet etching of quantum wells. Devices which take advantage 

of the positioning control are proposed. 

Investigating How the Brain’s Sensitivity to the Eyes in a Face Associates With Social Gaze During Real 

Conversations in Autism 

Valerie Latimore 

Mentors: Ralph Adolphs and Michael Spezio 

The ability to gather and process social information such as emotional expression is an important skill for all social 

animals to possess. Facial signaling is a particularly important medium of transmittance of social information for 

humans during conversation. However, people with autism exhibit deficits in many aspects of social behavior, 

including face gaze and the use of information in that face, such as from the eyes and mouth, in order to make 

judgments about a person’s facial expression. Because it is known that high functioning adults and children with 

autism show difficulties in social judgment and interaction, but retain normal perceptual and cognitive systems, 

research on autism has been focused on understanding how people with autism process salient social cues, notably 

from faces. The novelty of this project is that it seeks to establish a link between how the brains of people with 

52

autism process facial features, on the one hand, and the social gaze of those same people as they engage in real 

conversations, on the other. The project combines fMRI data with high-speed eyetracking in naturalistic contexts to 

yield new insight into the biology of autism. 

Determining Cloud Base and Thickness From Spaceborne Spectroscopic Imaging and Lidar Profiling 

Techniques 

Wingyee (Marie) Lau 

Mentors: Yuk L. Yung and Dong L. Wu 

Spaceborne stereoscopic imaging from the Multi-angle Imaging SpectroRadiometer enables global wide cloud base 

height and thickness determination, which cannot be achieved by in situ ground based Lidar height profiling 

instruments. This new technique enables more accurate determination of cloud base and thickness of boundary 

layer clouds, which in turn provides an important observational constraint to climate and weather models. 

To effectively compare height profiles determined from spaceborne images and Lidar profiling, we focus on an 

island far from the equator, i.e. Graciosa of the Azores, and we do not study clouds that are too thick for 

spaceborne cameras to see to the cloud base. We have found sufficient cases where height profiling from 

spaceborne imaging agrees with that from Lidar instruments. 

Analysis of Lunar Roughness and Thermal Emission 

Michael Lauria 

Mentors: Oded Aharonson and Antoine Lucas 

Lunar roughness, or variance in slope as a function of scale, reveals information concerning impacts and other 

processes which shaped the surface of the Moon. Roughness of a surface also affects temperature because it 

modifies the amount of incident solar illumination as well as reradiation from nearby surfaces. Lunar temperature is 

an important criterion for volatile ices stablity, and hence, relevant for future exploration. In this study, the 

relationship between lunar roughness on the one meter scale and the bolometric temperature of the lunar surface 

is evaluated through detailed elevation data in the form of Digital Elevation Models (DEMs). These models are 

constructed from high resolution stereo images acquired by Lunar Reconnaissance Orbiter Camera, and analyzed 

for their statistical roughness properties. Initial results indicate correlations between the roughness and 

temperature fields, dependent on the azimuth and elevation of the sun during the temperature measurement, as 

well as on large-scale features such as craters. This study provides a targeted analysis of specific areas of the lunar 

surface and establishes a method that can be applied more generally. 

Further Consequences of the Gross-Zagier Formulae 

Brian Lawrence 

Mentor: Dinakar Ramakrishnan 

A quadratic twist of the L-function associated with a modular form is known to satisfy a functional equation, which 

may be even or odd. A result due to Gross and Zagier explicitly computes the central value of the L-function or its 

derivative. For modular forms in prime level, in cases when the functional equation is even, Michel and 

Ramakrishnan have used an averaging method, averaging over new forms while keeping the quadratic character 

fixed, to prove several consequences of the Gross-Zagier formulae, including a non-vanishing result. The present 

research generalizes this result to forms of prime-squared level, for which the functional equation is always odd. 

Such an L-function necessarily has a central value of zero; the Gross-Zagier formulae compute the central value of 

its derivative. We compute the average value of these derivatives over different L-functions and investigate the 

consequences of this formula. In particular, we show that for sufficiently large level there exists an L-function with 

only a simple zero at the center of symmetry. Some of our results are conditional upon a conjectured nonnegativity 

for the central value of the derivative of a twisted L-function. 

Social Judgments at Varying Distances 

Dennis Lazar 

Mentors: Ralph Adolphs and Laura Loesch 

This experiment looks at how our social judgments about other people are influenced by our distance from them. 

People can make rapid judgments from faces, but it is still unclear what information in faces allows for these 

Spontaneous Trait Inferences. This experiment is performed by having viewers judge pictures of actors taken at 

different distances: far (20 meters), mid (10 meters), near (1 meter). For each stimulus, a viewer is asked to rate 

that stimulus in terms of trust, threat, and dominance. Additionally, a subset of subjects’ behavioral data is 

supplemented by eye tracking data. The data obtained shows good test-retest reliability within subjects, but 

further analysis still needs to be done. To conclude the project, I will analyze the effect of distance on the 

judgments made, as well as on where the subjects were looking while making their judgments. 

53

Change-Point Detection for the Community Seismic Network 

Tuan Anh Le 

Mentors: Andreas Krause and Matthew Faulkner 

Change-point detection is a technique used to finding anomalies in time-dependent data series, which can be 

applied to detecting natural phenomena. One of the examples in which change-point detection could be used in is 

earthquake data for the Community Seismic Network. Currently, the Community Seismic Network applies a simple 

binary picking algorithm for earthquake detection. Here, we consider the problem of applying multivariate changepoint 

detection on a dense, noisy sensor network for earthquake detection in the Community Seismic Network. We 

give an estimate of the performance of the method in general, and especially its feasibility for the Community 

Seismic Network. Performance is demonstrated on several sets, including synthetic and real world data series. 

Applied Statistics and Digit Analysis to Identify Fraud in Financial Data Sets 

Jetson Leder-Luis 

Mentor: Jean Ensminger 

In investigating fraudulent expenditure, it is useful to be able to determine the probability of the presence of fraud 

within a set of expenditure data simply by analyzing the numbers. Digit analysis is a tool in which the digits of a 

data set are compared to expected digit frequencies to indicate the probability that human tampering was present. 

Data produced by real-world expenditures follow different digit frequency patterns than those made up by humans; 

namely, human-produced data shows preference for certain numbers and fails to be significantly random or to 

follow Benford’s law, which gives an expectation for certain digits in certain circumstances. For this SURF project, I 

produced tools to analyze a financial data set of development expenditures for evidence of human tampering, then 

applied these tools to show the significance of the fraudulent patterns. The results indicate that the data set with 

which I have been working has a statistically significant probability of human tampering within certain categories, 

which signals fraud. 

Application of High-Throughput Screening to the Generation of a Thermodynamic Stability Database of 

Protein Mutant Libraries 

Ernest Lee 

Mentors: Stephen L. Mayo and Alex Nisthal 

The application of computational protein design to protein engineering is a burgeoning field. The potential of 

sophisticated protein-based drugs with specialized structures and unique chemical properties is at a premium. 

Therefore, it is essential to understand the basis of protein thermodynamics in the context of force-field 

development for computational models. In order to do this, the generation of a high quality, self-consistent protein 

stability database is required. Protein stability can be measured quantitatively by calculating the difference in free 

energy between a mutant and the wild type. Studying single-mutant proteins allows us to understand how amino 

acid mutations directly affect protein stability. Stability data for a variety of proteins was generated by utilizing a 

high-throughput screening method to assay the complete mutagenesis of a protein domain. From this data, the 

quality of this database was compared to those of ones derived from existing computationally generated models. In 

addition, a quantitative analysis of the physical effects of amino acid hydrophobicity, polarity, and volume change 

on protein stability was carried out. Ultimately, this analysis allows for a clearer understanding of the correlation 

between the structure and stability of proteins. 

Three Lead EKG Glove With Stethoscopes 

EunJee Lee 

Mentor: Mani Chandy 

According to India Today, in both urban and rural areas of India, heart diseases have emerged as the number one 

cause of death. For this reason, it is urgent to introduce heart monitoring device especially to the areas without 

strong health facility. EKG (electrocardiography) is useful source to identify abnormal heart function or damaged 

area of heart. EKG detects electrical activity due to the contraction and expansion of heart muscle. 

This project is for inexpensive yet complete device to operate as a medical device in the third world countries. To 

meet affordable price, the device is simply built with two quadruple operational amplifiers, resistors, diodes, and 

capacitors. To test and reduce noise, hardware band pass filter and software filter within the recording program are 

used. Also, the device is combined with stethoscopes for detecting heart sound simultaneously with the EKG signal. 

The device can be made with less than 10 dollars and can function as complete EKG and stethoscopes for heart 

sound. 

54

Increasing Epithelial Permeability to Enhance Topical Drug Delivery to the Cornea 

Sinwook Lee 

Mentors: Julie Kornfield and Dennis Ko 

The drug administration to the cornea is required for the treatment of certain eye diseases such as keratoconus 

and post-LASIK ectasia. Keratoconus is the corneal thinning disorder that affects approximately 1 in 2000. But the 

corneal epithelium blocks the drug penetration because of the tight junctions in the epithelial layer. The most 

common method to improve it is to remove the epithelium, which has side effects such as tissue scarring, infection, 

and blurred vision. We believe some chemicals can make the tight junctions open wide enough temporarily for the 

drugs to penetrate the epithelium. And in order to determine how much the epithelium is permeabilized by those 

chemicals, we used eosin Y as a model drug, which is used for the treatment of keratoconus. If our project is 

successful, we will be able to administer drugs using eye drops instead of removing the epithelium. 

Demonstration of a DNA-Based Random Walking Molecular Robot That Reorganizes 

DNA-Tagged Objects 

Yae Lim Lee 

Mentors: Erik Winfree, Lulu Qian, Niranjan Srinivas, and Damien Wood 

Biomolecular robotics is a relatively recent research field with visible and intuitive mechanisms using DNA 

molecules. Many walkers have been demonstrated to walk on 1-dimensional track, but just a few walk on 

2-dimensional track and even fewer perform specific functions such as transferring gold nanoparticle species as 

cargos while traversing the pathway. This SURF project aims to incorporate both 2-dimensional walking and a 

specialized function into a DNA-based robot. More specifically, a molecular-scale DNA-based robot will reorganize 

cargos on 2dimensional fields. The scheme will include DNA- tagged cargos scattered on 2-dimensional origami, 

and a DNA-based robot randomly walking to sort them into different piles. The 2-dimensional track, cargo, and 

cargo goals will be planted on rectangular DNA origami [P. W. K. Rothemnd, Nature, 2006]. The motion of the 

walker and cargos will be examined by atomic force microscopy imaging. Bulk behavior of the system, kinetics of 

walking, and mechanisms of cargo picking up, and cargo dropping off will be analyzed by SPEX experiment. 

Quantifying Uncertainty in a Spacecraft Deployable Structure Latch 

James Leet 

Mentor: Lee Peterson 

A spacecraft must often deploy an additional structure, such as a solar array or a scientific instrument, in order to 

fulfill its mission objectives. While such tasks are mechanically simple, one must ensure that the devices 

responsible for deploying such structures can operate within mission parameters while subjected to the harsh 

demands of space. In this case, we employed computational modeling techniques to evaluate a latch for the 

planned SWOT (Surface Water Ocean Topography) mission. Tasked with deploying an array to focus returning 

microwave signals that have been reflected off of surface water on Earth, this latch has the potential to undermine 

the accuracy of measurements obtained if it is not deployed within certain boundary conditions. Utilizing a highfidelity, 

finite element model of the latch prepared with the Sierra software suite, we quantified the uncertainty 

present in the design using a DAKOTA (Design Analysis Kit for Optimization and Terascale Applications) interface. 

By quantifying the uncertainty present in the latch, mission planners can validate the current latch as a viable 

design or attempt to redesign it in order to provide less uncertainty. 

Improve Syntheses for Tracking the Decomposition of Silica-Supported Olefin Metathesis Catalysts 

Jomya Lei 

Mentors: Robert H. Grubbs and Matthew Van Wingerden 

Synthesizing the second generation Grubbs catalyst with a triethoxysilyl-functionalized NHC ligand grafted onto 

silica gel creates an efficient heterogeneous solid catalyst that can be recycled many times and does not leach 

ruthenium during olefin metathesis. Unfortunately, this catalyst can only be studied by ICP-MS which destroys the 

catalyst. A photolabile linker was synthesized to attach the catalyst to the silica gel which is easily cleaved off with 

55

light – taking the silica gel with it. Thus, the decomposition of the catalyst could now be studied with 

nondestructive techniques like NMR. New techniques were also implemented to synthesize the catalyst with higher 

purity and yield. 

Cellular Delivery of an Inhibitory Branched Capture Agent Against Akt1 via Cell-Penetrating Peptides 

Erica Leung 

Mentors: James Heath and Ryan Henning 

The non-specific serine/threonine-protein kinase Akt1 inhibits cell apoptosis; therefore, it is an important factor in 

the study of cancerous cells’ survival abilities. We further developed a branched capture agent against Akt1 for 

delivery within a cell. To study its effect within a living cell, known cell-penetrating peptide sequences were added 

to the capture agent. Two sequences were used with spacers of GSG: polyarginine consisting of nine arginines and 

the HIV-1-TAT sequence consisting of RKKRRQRRRPPQQ. The cell-penetrating peptides were synthesized on the 

carboxy-terminus of the anchor ligand using solid phase peptide synthesis, and the branched secondary and 

tertiary ligands were appended using the copper catalyzed Huisgen 1,3-dipolar cycloaddition reaction. The cell 

permeable capture agent will be delivered into cells to assess AKT inhibition and cellular toxicity. 

Mechanical Properties of Nanopillars Composed of Two Metals: A Simple Model for Metal Matrix 

Composites 

Jarvis Li 

Mentors: Julia Greer and Qiang Guo 

With growing interest in the properties of nano-scaled structures, we attempt to formulate a more detailed model 

of metal matrix composites to better understand the properties that govern the microscopic interactions among 

different materials. In this project, using electrodeposition on templates made from electron beam lithography- . 

we fabricated bi-material Cu-Fe and Cu-Ni nano-pillars containing a single heterogeneous interface, with diameters 

ranging from 50nm to 500nm. We then performed both compression and tension tests on the pillars to determine 

their mechanical properties, such as yield strengths, flow stresses, failure strain, work-hardening behaviors and 

etc. Also, we characterized the pillars using transmission electron microscope (TEM) to study the their 

microstructures, and scanning electron microscopy (SEM) to study the morphology before and after deformation. 

These bi-material pillars represent a prototype metal-matrix composite with highly controllable microstructure, high 

strengths, as well as strong-metal-metal interfaces. The isolation and subsequent engineering of a single interface 

in a pillar allows for fundamental insights gained into their deformation mechanisms, leading to more intelligent 

design of interfaces. 

Analyzing the Effects of Ice on Stratospheric Water 

Jingyuan Li 

Mentors: Geoffrey A. Blake and Ke Zhang 

How water enters the stratosphere is a process that is largely unknown. Despite the low temperatures at the 

tropopause which lead to small mixing ratios, stratospheric water levels show no correlation with temperature. 

Known methods of transport across the tropopause point to ice as an important factor. This project concentrates on 

any possible correlation between the amount of ice and the water vapor by using isotopic ratios. Due to the lower 

vapor pressure of HDO, the H/D ratio decreases as the water vapor condenses. The data is taken from the Keck 

telescopes, which can be used to extract both emission spectra and atmospheric transmission. Data reduction was 

done using a pipeline and then fitted to a lineshape model. Different line profiles (namely, the Voigt and the 

Galatry) were compared using CO fits to see if a simpler model is enough for the HDO/H2O fits. An atmospheric 

model gives us the values of HDO and H2O with altitude, and the results are tested with water vapor levels above 

Mauna Kea. 

Back-Action Evasion Measurement and Squeezed State of a Mechanical Oscillator 

John Li 

Mentors: Keith Schwab and Junho Suh 

Back-action evasion measurement avoids introducing noise when measuring a system. Kalman filter is a statistical 

algorithm that produces a better estimation from the measurement. I investigated Kalman filter in order to perform 

the optimal control of cantilever in back-action evasion measurement. I derived and verified the technique to 

implement Kalman filter for this specific experiment into discrete algorithm that DSP or FPGA can run, as well as 

the possible hardware means, including both Digital signal processor and Field programmable gate array, to 

perform digital signal processing. 

56

Improving the Toughness of Responsive Double Physical Network Injectable Hydrogels 

Shuaili Li 

Mentors: Bradley D. Olsen, Matthew Glassman, and David A. Tirrell 

Hydrogels show significant promise in minimally invasive tissue engineering and drug delivery because of their high 

water content and tunable mechanical properties. Triblock copolymers with a gel-forming protein midblock and a 

thermoresponsive poly(N-isopropylacrylamide) (PNIPAM) endblock form a double physical network hydrogel that is 

both injectable and responsively stiffens post-injection, but control of endblock relaxation for enhanced toughness 

remains a challenge.To engineer the thermodynamic and dynamic behavior of the endblocks, variants of PNIPAM 

polymers have been synthesized via Reversible Addition Fragmentation Chain Transfer (RAFT) with variable alkyl 

chain ends and with hydrophobic comonomers. Copolymers with N-tert-butyl acrylamide (NtBAM) are expected to 

have particularly promising characteristics due to depressed lower critical solution temperature (LCST) and higher 

glass transition temperature than the homopolymer. PNIPAM copolymerized with photocrosslinkable benzophenone 

monomer has also been explored for biomedical testing purposes. Fundamental investigations have been 

performed in purely synthetic triblock copolymer models as well as in bioconjugates with gel-forming protein. 

Linear oscillatory shear rheology was used to probe the relaxation behavior of the double-network gels via 

frequency sweeps as a function of temperature. Nonlinear rheology was used to look at the yielding behavior. 

Progress towards demonstration of the responsively tough, injectable biomaterial will involve compression testing 

for fracture toughness, assessments of injectability, and cell viability and adhesion studies. 

Identification of Piwi-Binding Loci in the Drosophila Germline Genome 

Susan E. Liao 

Mentor: Katalin Fejes Toth 

Suppressing mobile DNA elements in the germline is important for protecting genome integrity. One protein, Piwi, 

has been shown to partner with small RNA segments to silence effects of these mobile DNA elements in the 

cytoplasm. Since Piwi is localized to the nucleus in all metazoans, we suspect that it may also regulate expression 

of mobile DNA elements in the nucleus. Using DNA adenine methyltransferase identification (DamID), we 

investigate whether Piwi localizes to to specific chromatin loci. Then, through reverse transcription polymerase 

chain reaction (RT-PCR) and chromatin immunoprecipitation (ChIP), we test Piwi's effects on gene expression 

levels and epigenetic histone marks. 

Directed, in situ Growth of Conductive Nanowires Between the Photoelectrodes in a Tandem Microwire 

Array Fuel-Forming Device 

Ben Lieber 

Mentors: Nathan S. Lewis and Shane Ardo 

The primary focus of my research was to fabricate a highly conductive electrical conduit between two silicon 

microwire arrays through electric-field directed growth of 7,7,8,8-tetracyanoquinodimethane (TCNQ) anion chargetransfer 

crystals. This is important, because tandem microwire array solar photoelectrosynthetic cells require highly 

conductive interconnects to maximize efficiency. [Cu(TCNQ)] nanowires were grown in situ via physical vapor 

deposition from solid TCNQ precursors. Successful nanowire formation was demonstrated on photolithographically 

patterned 3 µm diameter copper islands, and on the copper vapor-liquid-solid (VLS) growth catalyst found at the 

top of silicon microwires. Scanning electron microscopy (SEM) and powder X-ray diffraction (XRD) confirmed the 

structure and conductive phase of the [Cu(TCNQ)] nanowires. Sandwich-cell devices, employing two planar 

patterned wafers, were constructed to model the back-to-back arrangement of polymer-embedded silicon 

microwire arrays envisioned for the ultimate device; [Cu(TCNQ)] nanowires were grown to bridge the halves of 

these sandwich-cell devices. Ongoing research is directed toward integrating these results to a tandem silicon 

microwire-array device and directing the [Cu(TCNQ)] growth with an applied electric field. 

Faster Compression Leads to Higher Recovery and Better Energy Absorption in Carbon Nanotube 

Micro-Pillars: An in situ Study 

Ee Jane Lim 

Mentors: Julia R. Greer and Siddhartha Pathak 

While individual carbon nanotubes (CNTs) have been announced as the strongest material known, layers of 

vertically aligned CNT (VACNT) bundles are more likely to find use in applications requiring large compliance and 

deformability, such as in microelectromechanical systems (MEMS) and impact mitigation. In this study we report on 

the mechanical behavior of such a VACNT bundle, grown in the form of 30/30 (height/diameter) micron-sized 

pillars using the chemical vapor deposition method at 750 mbar pressure and ~750 o C. Compression tests on these 

CNT pillars revealed 3 distinct regimes in their mechanical deformation—a short initial elastic section followed by a 

sloped plateau regime corresponding to buckle formation in the pillars, and a final densification regime. Each 

regime is characterized by its own distinct mechanical properties, for example a 6 fold increase in modulus and 

drastic decrease in recoverability were observed in the densification stage as compared to the elastic regime. Loadunload 

cycles at different strain rates reveal the gradual accumulation of damage in the CNT pillars at lower rates, 

57

which leads to lower recovery (

molds and extracted as free-standing particles in solution. Ultimately, clover-leaf microgels will be fabricated using 

artificial extracellular matrix proteins and assembled into porous constructs to determine the impact of porosity on 

corneal epithelial cell migration. 

Self-Assembling Artificial Extracellular Matrix Proteins for Cell Encapsulation 

Erik Liu 

Mentors: David A. Tirrell and Maren E. Buck 

The development of new biomaterials is important for tissue engineering and cell-based therapies. In this project, 

we use recombinant DNA technology to generate artificial extracellular matrix (aECMs) proteins. We have targeted 

two multidomain protein polymers composed of domains derived from elastin and fibronectin sequences that 

alternate with heterodimeric leucine zipper domains. Mixing of multiblock protein polymers containing 

complementary leucine zippers should permit the spontaneous formation of hydrogels under mild conditions. The 

DNA sequences encoding these proteins have been cloned and sequence verified. Initial expressions of the proteins 

resulted in low yields. Currently, we are pursuing a dual plasmid approach for co-expressing multidomain proteins 

with the complementary leucine zipper as a potential route to improving protein yields. 

Determination of the Dihedral Angles in Predominantly Trans-Ethanes Using NMR Spectrometry 

Tianxiang Albert Liu 

Mentor: John D. Roberts 

Two predominantly trans-ethane systems (2-(N,N-dimethylamino)-N-1-ethylammonium chloride and 3,3dimethylbutyl-N,N,N-trimethylammonium 

iodide) were synthesized in relatively high yield. Variable temperature 

and solvent studies suggested their strong trans conformational preference. Two ordered media (aqueous lyotropic 

and 7% PMA gel) were used to obtain 1 H, 13 C and HSQC spectra of both trans-ethane systems, which were then 

used to extract scalar, dipolar and residual dipolar coupling constants. This shed light on trans dihedral angles of 

both systems, which were previously undetermined. 

Cell ECM Network Study for Application in Tissue Engineering 

Ashley Lo 

Mentors: Chin-Lin Guo and Yuhwa Lo 

Cells have an extracellular matrix (ECM) on their exterior surface, which regulates processes such as growth, 

reparation, intercellular communication, support, and anchorage. When the ECM’s of each cell “detects” the 

presence of other cells within their vicinity (600 μm), these ECM’s will extensively branch out and connect to those 

of neighboring cells. 

MDCK and MCF-10A cells were grown in a cell culture, and injected into microfluidic channels and polygonal wells 

of different shapes, containing a medium solution and dissolved collagen, in attempt to stimulate the in vitro 

formation of ECM and epithelial (cellular) structures. 

In living organisms, the ECM plays a critical role in the ability of cells to self-organize and form functional 

structures. This process is evidenced in the formation of connective tissues, such as organs, joints, ligaments, 

fibrils, and tendons. Because cellular organization is such a fundamental aspect of structure and survival, research 

on the factors that influence and enable it would provide valuable insight into medical field. Ideally, the ability to 

induce cellular self-organization and differentiation in vitro would provide potential solutions to the current 

shortages in regenerative medicines and organ transplants. 

Atmospheric Features at the Jupiter North Pole From Cassini Images 

Daniel Y. Lo 

Mentors: Andrew P. Ingersoll and Shawn P. Ewald 

Newly reprojected mosaics of the Jupiter North Pole from Cassini images taken with the CB2 (756 nm), 

MT2 (727 nm) and MT3 (889 nm) filters has allowed the first detailed observations of the dynamical features and 

processes that occur at the high latitudes (>60 deg N). This study found that spots and filamentary regions (FRs) 

also occur at the high latitudes as in the lower latitudes. However, interactions between spots leading to merger 

take place in a very different manner, and this, together with the splitting behavior of spots, suggests that 

dynamics at the high latitudes are very different. In addition, this study also discovered that FRs originate from 

cyclonic spots that migrate into shear zones, and found evidence of the asymmetry in behavior between cyclonic 

and anticylonic features in these shear zones. Furthermore, this study also verified that the spot budget at the high 

latitudes was balanced over the observational period, and characterized the size of the spots at various latitudes. 

These findings will pave the way and serve to inform the first direct observations of the Jupiter North Pole by Juno 

when it enters Jovian orbit in 2016. 

59

N-Terminal Functionalization of Engineered Calmodulin Constructs for Surface Immobilization 

Megan Lo 

Mentors: David A. Tirrell and Chethana Kulkarni 

Calmodulin is a regulatory protein that plays an important role in learning and memory. The goal of this work is to 

immobilize engineered calmodulin constructs onto surfaces to create microarrays for high-throughput studies of 

calmodulin and the proteins which it regulates. We are working towards this goal by taking advantage of 

myristoylation, a well-studied, natural labeling process in eukaryotes. Myristoylation is the transfer of myristic acid 

onto the N-terminus of protein substrates via the eukaryotic enzyme N-myristoyl transferase (NMT). This process 

can be utilized as part of a versatile site-specific labeling technique due to NMT’s tolerance towards reactive 

myristic acid analogs. By engineering calmodulin to display a short sequence that NMT recognizes, we can achieve 

N-terminal functionalization of the protein which enables it to be conjugated to other materials such as beads and 

glass slides; specifically, calmodulin can be coupled to surfaces to create microarrays. After the construction of 

these microarrays, high-throughput studies can be performed to study the interaction of calmodulin with its binding 

partners to allow for a deeper understanding of the protein. 

An Analysis of Properties of Galaxies With Close Pairs 

Cassandra Lochhaas 

Mentor: Nick Scoville 

Merging galaxy systems are very important in the creation of new galaxies. Eventually, a system of merging 

galaxies can become a single, new galaxy. During the early stages of merging, galaxies can be at a variety of 

separation distances, and these distances give clues to how the merging process evolves. Due to the difficulties of 

identifying merging galaxies, this study considers close pairs of galaxies (instead of attempting to classify actual 

merging systems). The data was taken from the COSMOS study, which contains hundreds of thousands of galaxies 

in a small portion of the sky out to distances that correspond to half the age of the universe. The number of 

galaxies with close pairs as a function of separation distance and cosmic age of the universe (equivalently redshift) 

was also studied, with implications to how galaxies merged together in the far past to form galaxies in the more 

recent past. Differences in properties such as star formation rate, mass, and absolute magnitude of galaxies were 

analyzed. A. Petric also helped in analyzing spectra of a sample of nearby luminous infrared galaxies to better 

understand the effects of merging. 

Core Formation in SuperEarths 

Simon J. Lock 

Mentor: David J. Stevenson 

It is usually assumed in the study of SuperEarths (rocky planets with a mass of a few Earth masses) that they are 

essentially Earth-like, i.e. they have a differentiated (mostly liquid) core with a solid mantle and crust. However 

SuperEarths are significantly more massive than the Earth and so this may not always be the case. When 

SuperEarths are first formed it is believed that they are hot enough that their constituents (mainly silicates and 

iron) are fully mixed, down to the atomic level. As the planet cools the rock and iron demix, with the iron rich 

components sinking to the centre of the planet. This releases a lot of gravitational potential energy, providing a 

significant buffer against cooling. This process is being studied by constructing a model of the energy balance for a 

SuperEarth as it cools, in order to determine the effect on planet formation. A basic model indicates that significant 

core formation does not occur for planets of a few Earth masses. A more advanced model is being developed to 

further explore the problem. This work has significant implications for modelling the internal structures of 

SuperEarths and its results will need to be considered in future attempts. 

Developing an Algorithm to Pinpoint Deep Sea Corals 

Paige Logan 

Mentor: Jess Adkins 

Studies of deep-ocean sediments have revealed a clear connection between deep ocean circulation patterns and 

global climate change. A replacement for these sediments has been found in the deep-sea coral species 

Desmophyllum dianthus. These corals can provide more accurate oceanic circulation rate data for past climates 

using radiocarbon and U-series dating. This data may help us further understand current global climate change. 

Since these corals grow thousands of meters underwater and are only a few centimeters tall, they are difficult to 

find. My goal is to find an easy way to locate the corals. Using Amazon’s Mechanical Turk service, we quantified 

thousands of seafloor pictures to resolve the presence or absence of corals at various depths. By combining this 

information with bathymetry data from the autonomous rover that took the pictures, we can study the relationship 

between coral presence and the depth, slope, and curvature of the seafloor. This allows us to create an algorithm 

to estimate the likelihood of finding corals in a given area based on the shape of the seafloor there. 

60

The Role of Pavlovian Processes and Payment Methods on Decision Making and Reinforcement Learning 

Tong Lu 

Mentors: Antonio Rangel and Vanessa Janowski 

In contrast to the economic assumption that consumers are rational in valuation of goods, consumer choice is 

affected by product presentation, payment method, and other situational variables, as explored in two related 

experiments. The first simulated a seller choosing prices for various items, with the hypothesis being that the 

physical, proximal presentation of money in cash form significantly increases its value, as opposed to usage of nonmonetary 

equivalents or the presence of a physical barrier, resulting primarily from Pavlovian consummatory 

mechanisms. The second involved subjects given either cash or tokens to bet on the win or lose outcome of a 

series of colored cards in a gambling game. Their objective was to maximize their winnings by tracking the 

stochastically varying probabilities associated with the different card colors, generated by a Gaussian walk 

simulation. Preliminary results are consistent with the hypothesis that subjects are more willing to take risks with 

tokens versus cash. Identification of the variables that affect consumer decision making have immediate 

applications to marketing strategy. 

Design, Fabrication, and Characterization of Low Threshold Photonic Quasicrystal Lasers 

Tsung-Ju J. Lu 

Mentors: Axel Scherer and Se-Heon Kim 

Photonic crystal lasers have a spectral gap in the electromagnetic wave spectrum in which electromagnetic wave 

propagation is forbidden in all directions, which makes them attractive optical materials for controlling and 

manipulating the flow of light. However, there are some fabrication challenges in creating two-dimension periodic 

photonic crystals, specifically two-dimensional slab structures. In the present time, only thin slab periodic photonic 

crystals are useful since they can support a high Q factor. However, thin slabs have many disadvantages when 

used for real-life applications. Thus, the goal of this project is to fabricate a thick slab Quasicrystal cavity laser that 

has a quasi-periodic arrangement of air cylinders, which would potentially have a higher Q factor compared to 

photonic lasers with periodic arrangements of air cylinders. The design and simulation of the Quasicrystal structure 

were done by using Finite Difference Time Domain method and a program called “harminv”. The fabrication of the 

photonic Quasicrystal lasers was done by using a scanning electron microscope and etching machine. The lasers 

were then tested to make sure that they do indeed have spectral band-gaps and operate with a low threshold. 

Conformational Analysis of Succinate Dianion in Liquid Crystal Media 

Yichen Lu 

Mentors: John D. Roberts, Mrinmoy Nag, and William Carroll 

Anisotropic phases of liquid crystals restrict the freedom of motion of molecules in normal isotropic phase of 

solution. Nuclear Magnetic Resonance Spectroscopy (NMR) can be used to obtain geometrical information and other 

solution order parameters. Most importantly, dipolar couplings that contain information about the relative angles of 

the nuclei in a molecule can be obtained by fitting the oriented NMR spectrum. The average dihedral angle rotating 

around the central C-C bond of a specific molecule can then be calculated using dipolar coupling constants. In this 

project, both succinic acid and succinate dianion have been studied in an environment of oriented aprotic media 

such as liquid crystals. Both H-NMR and 13 C-NMR spectrum have been taken at nematic phase temperatures to find 

dipolar coupling constants. Mixtures of gauche and trans conformers have been considered. A MATLAB program has 

been constructed to perform the calculation. 

The cpSRP54-cpFtsY Interaction in the Chloroplast SRP Pathway 

Zeyu (Mike) Lu 

Mentors: Shu-ou Shan and Sowmya Chandrasekar 

The universally conserved signal recognition particle (SRP) and SRP receptor (SR) mediate the cotranslational 

targeting of proteins to cellular membranes. In contrast, a recently discovered unique chloroplast SRP in green 

plants is primarily dedicated to the post-translational targeting of light harvesting chlorophyll a/b binding (LHC) 

proteins to the thylakoid membranes. The pathway involves cpSRP54 and its receptor cpFtsY, two GTPases that are 

closely homologous to the cytosolic SRP54 and SR GTPases. However, intriguingly enough, the otherwise 

universally conserved SRP RNA which accelerates complex formation between the two GTPases has not been found 

in the chloroplast pathway. Instead, a new 43-kDa protein, cpSRP43, forms the chloroplast SRP together with 

cpSRP54. Despite the early hypothesis that cpSRP43 might serve the same purpose as the SRP RNA, kinetic 

analyses have shown that it does not significantly affect the rate of the cpSRP54·cpFtsY complex formation or the 

GTP hydrolysis rate of the complex. Instead, cpSRP43 facilitates the substrate recognition through interactions with 

the cargo proteins, LHCPs (1-3). My overall goal for the 2011 SURF is to understand how cpSRP54 and cpFtsY are 

able to associate so much more efficiently than their bacterial homologues in the absence of the SRP RNA by 

studying their complex structure using crystallography and lysine footprinting mass spec. 

61

The Impact of Driving Force on Rates of Electron Transfer in Ruthenium-Labeled Cytochrome P450 

Katja E. Luxem 

Mentors: Harry Gray, Jeff Warren, and Maraia Ener 

This project aimed to examine the high oxidation state intermediates in the catalytic cycle of P450 by studying the 

impact of driving force on the rate of electron transfer in our protein model system. By covalently labeling P450BM3 

with four different ruthenium-based photosensitizers and photo-exciting them, we were able to generate the 

species of interest. This research sheds light on redox events involved in the reactions of the high oxidation state 

intermediates in the catalytic cycle of cytochrome P450, knowledge which can be applied to modifying the protein 

for human use and studying the health impact of the protein. We have finished synthesis and characterization 

(NMR, mass spectrometry, cyclic voltammetry, transient and steady state fluorometry) of the photosensitizers 

attached to P450 and free in solution. 

Development of a Physical Model of E. coli Carbon Metabolism for Predicting Cell Growth Parameters 

Brian Ma 

Mentors: Kristala J. Prather, Kevin Solomon, and Richard Murray 

Understanding how cells use nutrients is the first step in formulating a systematic practice of metabolic 

engineering. In order to optimize the activity of a recombinant pathway and optimize the yield of a desired product, 

it must be decided how much carbon a cell can direct towards its various tasks, and therefore how much should be 

optimally allocated towards an artificial pathway. Using glucose as the sole carbon source, a simple physical model 

has been developed that dynamically links the expression of glucokinase (a glucose-intake, and therefore growthlimiting 

protein) and glucokinase mRNA, and the overall growth rate of E. coli KTS strains developed by the Prather 

Group. The model incorporates parameters such as promoter strength, anti-sense RNA and induced repressor 

expression to provide insight into how glucokinase activity can be controlled through different means and predict 

the results. A further expansion of this model calculates from first principles the use of carbon and ATP by the cells 

to pursue growth, providing both greater accuracy and a possible physical explanation for the observed growth 

trends of E. coli strains. 

Heterogeneous Nucleation of Organic Compounds on Crystalline Substrates 

Xiaoya Ma 

Mentors: Bernhardt L. Trout and David A. Tirrell 

Crystallization enjoys diverse applications, including the purification of pharmaceutical ingredients, food 

engineering, construction design, and the manufacture of semiconductors. One of the key challenges that arise in 

any industry, however, is controlling the properties of the crystalline products. To do so requires a better 

understanding of nucleation, the critical step in the crystallization process. In reality, all crystallizations occur by 

heterogeneous means. In this project, the heterogeneous nucleation of various organic crystalline materials onto 

crystalline surfaces was studied. Epitaxial studies of these systems helped shed light on the fundamental 

mechanisms behind heterogeneous nucleation, which are not well understood at present. A long-term consequence 

of greater control over nucleation will be greater control over properties of the crystalline product, such as 

polymorphism. 

Study of Z boson + Higgs � Z boson W boson W* boson � leptons + neutrinos at the International 

Linear Collider 

James Macdonald 

Mentors: Maria Spiropulu and Barry Barish 

This project studies the Higgs decay channel with either two or four lepton plus neutrino final states at a 

proposed International Linear Collider (ILC), an electron positron collider with center of mass energy 250 GeV and 

an integrated luminosity of as assumed in the ILC Letters of Intent. To accomplish this we produced Monte 

Carlo data for Higgs production events as well as standard model background. The objective is to extract the Higgs 

signal from the dominant standard model background of ZZ and WW dibosons and possibly from “Higgs lookalikes.” 

Ultimately we would like to develop handles to distinguish a standard model Higgs from its look-alike states 

and improve on the selection criteria by using angular and other new kinematic variables. 

Argonaute Protein From Rhodobacter sphaeroides and Its Nucleic Acid Interactions 

Vishnu Manoranjan 

Mentors: Alexei Aravin and Ivan Olovnikov 

RNA interference plays an important role in controlling gene activity. RNA interference is also known to be 

mediated by small-interfering RNAs in eukaryotes. These small-interfering RNAs are derived from double-stranded 

precursors, and Argonaute proteins aid in the actual target cleavage. Here, we examine Argonaute proteins from 

bacteria and the nucleic acids that they interact with. So far, bacterial and archaea Argonaute x-ray structures 

have been studied. In vivo functions of these proteins, however, remain a mystery. Here, we observe the 

Argonaute protein in Rhodobacter sphaeroides and find that this protein binds not only to small RNAs but also to a 

62

number of small DNAs. Preliminary deep sequencing results show that the small RNAs observed have a strong 1- 

Uracil bias and a size distribution of 17-18 nucleotides. This research sets the tone for further study of the role of 

Argonaute proteins in processing of small nucleic acids 

Investigating Mechanisms of Isotopic Fractionation in Microbial Sulfate Reduction 

Taylor Martin 

Mentors: Alex Sessions and Jacob Waldbauer 

Changes in S isotope fractionation can indicate changes in bacterial evolution, the productivity of sulfate reducing 

bacteria, and even the oxidation state of the ocean throughout the geologic record. Though the biochemical 

mechanism of reduction of sulfate to sulfite is well understood, the physiological basis of S isotope fractionation is 

unknown. Desulfobacterium autotrophicum, a marine sulfate-reducing bacterium, was grown under several 

different conditions expected to impact fractionation. One culture was grown autotrophically using H2 and CO2, 

whereas the other cultures were grown heterotrophically using 10 different carbon sources. The cultures were 

sampled over the course of a month, and each sample was analyzed for sulfate isotope composition, sulfide isotope 

composition, sulfate concentration, sulfide concentration, and cell density. The isotopic composition of sulfate and 

sulfide was compared for each sample to provide a value for the isotopic fractionation that occurs via sulfate 

reduction. The different fractionation values from the cultures grown on different substrates, in conjunction with 

how each of the substrates is processed by the bacteria, provide insight into the physiological basis of S isotope 

fractionation. 

Implementation of Motion Simulation Software and Visual-Auditory Electronics for Use in a Low Gravity 

Robotic Testbed 

William C. Martin 

Mentor: John Leichty 

The Jet Propulsion Laboratory (JPL) is developing the All-Terrain Hex-Limbed Extra-Terrestrial Explorer (ATHLETE) 

to assist in manned space missions. One of the proposed targets for this robotic vehicle is a Near-Earth Asteroid 

(NEA), which typically exhibit a surface gravity of only a few micro-g. In order to properly test ATHLETE in such an 

environment, the development team has constructed an inverted Stewart platform testbed that acts as a robotic 

motion simulator. This project focused on creating physical simulation software that is able to predict how ATHLETE 

will function on and around a NEA. The corresponding platform configurations are calculated and then passed to 

the testbed to control ATHLETE’s motion. In addition, imitation attitude control thrusters were designed and 

fabricated for use on ATHLETE. These utilize a combination of high power LEDs and audio amplifiers to provide 

visual and auditory cues that correspond to the physics simulation. 

Microenvironments for Studying Extracellular Cues to Cell Behavior 

Jordan Maslov 

Mentor: Chin-Lin Guo 

Cell behavior, such as branching morphogenesis and structure formation, depends on mechanical and geometrical 

cues from the extracellular environment. Here, we develop multiple techniques and microenvironments to examine 

how these processes take place based on the cells’ external factors. One type of environment uses PDMS stamps to 

imprint traps into an agarose and gelatin mixed gel. Cells can be seeded and washed into the traps, and collagen 

can be added to induce either branching for MCF-10A cells or structure formation for MDCK cells. We also develop 

the use of channels of PDMS microfluidic devices as entrapments to shape and control the formation of cellular 

structures. 

Semi-Annual Oscillations and Seasonal Variations in Saturn’s Atmosphere 

Charlotte A. Mason 

Mentor: Glenn Orton 

Ground-based telescopes can provide us with images of thermal emission of Saturn in the mid-infrared spectral 

range that are useful in tracking various atmospheric properties. Using such data from observations carried out by 

my mentor and his colleagues with different mid-infrared wavelength filters, it is possible to study the variability of 

emission from different gases in Saturn’s atmosphere. I present the most recent study of low-latitude stratospheric 

variability of methane and ethane emission and discuss how the recent data fit with predictions for a semi-annual 

oscillation structure. Variability of strong emission from methane, being well mixed in Saturn’s atmosphere, tracks 

the variability of stratospheric temperatures. Variability of strong emission from ethane can, with the temperature 

information from the methane emission, be used to detect variability of the abundance of ethane itself. Besides the 

low-latitude oscillation, I also examine the high-latitude stratospheric variability of methane and ethane, and the 

variability of thermal emission in the troposphere as a function of time and compare these data with models for 

seasonal forcing of the atmosphere. 

63

Electromyographic Control of Tactile-Sensing Enabled Prosthetic Hand 

Blaine Matulevich 

Mentors: Gerald Loeb, Jeremy Fishel, and Joel Burdick 

Myoelectric prosthetic hands generally provide poor, if any, tactile feedback to their users; as a result, gripping 

objects is often both strenuous and difficult. Attaching multimodal tactile sensors to a currently available hand can 

provide tactile feedback for transmission to the user as well as enabling automated slip-detecting grip-control 

algorithms. While reflex-like grip control would improve the ability to grasp objects, fully autonomous grip control 

may be undesirable; for this project, we sought to integrate the user’s intent to scale the magnitude of the 

response. A circuit was designed and constructed that filtered raw electromyographic (EMG) signals from electrodes 

attached to the user’s forearm, outputting a smooth envelope indicating the approximate level of effort exerted by 

the user. A servo-amplifier was connected that enabled the user to drive the hand directly with EMG signals. A 

force gauge was designed, built, and utilized to create curves that correlate the voltage of the user’s signal with the 

force output by the hand. In future work, the user’s EMG signal will provide a measure of the intentions of the user 

and will be combined with tactile information about object loads and slip to determine the actual force generated 

by the actuator. This should result in a user-based control system that feels more natural and provides significantly 

better tactile feedback than current designs. 

Understanding the Effects of the Pulse Charging Method on Dendrite Growth in Lithium Metal Batteries 

Matthew Mayers 

Mentors: Thomas Miller and Jakub Kaminski 

The availability of powerful, lightweight, and long-lasting battery technologies is an essential prerequisite for the 

design and development of new technologies and devices. Lithium metal is an ideal anodal component for these 

battery technologies because of its high redox potential and low molecular weight. However, the realization of 

secondary, or rechargeable, lithium metal batteries has been frustrated by the formation of metal dendrites at the 

anode during recharging, leading to reduced lifetimes and catastrophic failures of the cells. Using coarse-grained 

molecular simulation methods, we develop a three-dimensional model to directly simulate the non-equilibrium 

processes of lithium reduction, plating, and dendrite formation on a nanoparticle. By observing the conditions 

under which dendrites nucleate and subsequently elongate, we hypothesize that the pulse charging method could 

aid in preventing the onset of dendrite formation. We show that the pulse charging method can be used with high 

overpotentials to cause plating densities identical to those seen in the case of low overpotentials. Finally, we see 

that although battery materials give rise to an inherent, relatively low upper limit on the pulsing frequencies that 

may be used, that upper limit still allows denser plating in the presence of a healthy SEI, a result that reinforces 

the importance of the SEI in helping to prevent dendrite formation. 

Online Real-Time Prediction of Motor Decisions Using Intracranial Brain Recordings in Humans 

Naomi McArthur 

Mentors: Christof Koch and Uri Maoz 

Neuroscientific research of free will research over the last 30 years has revealed that binary motor decisions might 

be predicted well in advance of a person’s conscious intention to move. However, such research was never 

conducted in a live setting – online and in realtime – with a patient before. The goal of this project was to gather 

data from patients implanted with intracortical electrodes as they press a button to advance through a series of 

images, and then incorporate the data gathered into a learning model that would predict the timing of the button 

press. This learning model will be used in real time to predict the movement of the patient before movement 

occurs. During the course of the project, new data was collected, the parameters of the patient’s task were 

changed, and the learning model was tested against available data. Future directions would involve refining the 

learning model further, testing in real time with new patients, and observing the effects on the patient’s agency 

and decision-making when they realize that their movements are being predicted. 

Seismocloud: An Investigation Into Efficient SCEDC Earthquake Data Manipulation 

Stephen Meisenhelter 

Mentors: Robert Clayton and Ellen Yu 

The Southern California Seismic Network currently stores all collected data to a server bank on Caltech campus and 

a redundant system in another on campus building. While this redundancy does provide some protection from 

failure, the entire system is still vulnerable to data loss from earthquakes in Pasadena. The system also suffers 

from bottlenecking leading to server slow-downs and denial of service during large events. 

Seismocloud is an experimental cloud based data server that utilizes the Google App Engine as a replacement for 

the current SCEDC server system. This system will solve the data redundancy and bottlenecking problems 

associated with maintaining physical servers while also reducing costs. Due to design changes in the Google App 

Engine, it became possible this year to drastically increase the efficiency of Seismocloud by precompiling a single 

unified record for each seismometer’s submitted data. This simplifies storage and allows for a more consistent data 

set. These changes make Seismocloud feasible for large-scale deployment. 

64

Improving Redox Couple Mass Transport in Silicon Wire-Array Liquid Junction Solar Cells 

Andrew C. Meng 

Mentor: Nathan S. Lewis 

Silicon wire-array liquid junctions show promise in solar to fuel conversion applications. However, high aspect ratio 

wire array geometries present additional mass transport complications relative to their planar analogs. Redox 

couple depletion at the bottom of the wire arrays was observed and accounted for carrier collection loss. To 

facilitate redox couple mass transport, a novel architecture is proposed to achieve higher efficiencies in liquid 

junction solar cells. Metal films were deposited onto the bottom of Si wire arrays by electron-beam evaporation. To 

prevent direct contacts to the p-Si wire array from the metal films, electrically insulating layers of SiO2 and Al2O3 

were prepared by thermal oxidation and electron-beam evaporation, respectively. During cell operation, the metal 

films served as additional counter electrodes, which regenerate redox couple species where it is most likely to 

deplete. Computational modeling suggested that the supply of redox species from the metal films minimizes 

concentration overpotential and solution resistance. In this study, the energy-conversion properties of Si wire-array 

photocathodes with metal film counter electrodes were characterized in contact with cobaltocene (CoCp2 +/0 ) redox 

couple. Significant improvements of current-voltage characteristics were observed compared to the conventional 

cell architecture. Under high illumination intensities or low stirring conditions, the new electrode design 

demonstrated superior performance in short-circuit current and fill factor. Spectral response also confirmed near 

unity collection of photogenerated carriers at the bottom of the wire array. The proposed structure, which 

essentially minimizes the effective ion transport path length to the photoelectrode, also provides design guidance 

for future tandem cells for water splitting reactions. 

Syntheses of Fluorinated Cobalt Diglyoxime Complexes for Electrocatalytic Hydrogen Evolution 

Rocio Mercado 

Mentors: Harry B. Gray and Michael J. Rose 

Work on various cobaloximes has shown them to be promising catalysts for electrocatalytic hydrogen evolution at 

low overpotentials. However, while cobaloximes of the forms Co(dmgBF2)2L2 and Co(dpgBF2)2L2 exhibit Co II/I 

reduction potentials in organic solvents close to NHE (-2.4 V vs SCE), placing them in water increases the 

magnitude of this potential enough to render the catalyst ineffective. Because of this, we are studying various 

cobalt complexes containing the fluorinated diphenylglyoxime ligand (“dAr F gH2”). We are studying these complexes 

because the electron withdrawing pentafluorophenyl groups will shift the Co II/I reduction potential in the positive 

direction. As a consequence, under aqueous conditions the potential of the catalyst will fall within an adequate 

range. So far, we have synthesized Co(dAr F gH)2(py)2, as was confirmed by X-ray crystallography and infrared 

spectroscopy. We have studied this compound by cyclic voltammetry, and found the Co II/I reduction potential to be 

-0.88 V vs. Ag/AgCl in methylene chloride. After -1 V, the CV is not well defined and suggests rearrangement of 

the Co I complex. These possible rearrangements could be due to the weak proton bridge linking the two ligands. 

Because of this, the proton bridges in the complex have been substituted by more stable difluoroboryl and 

diphenylboryl bridges, allowing for simple comparison of the BF2/BPh2-bridged cobaloxime to the proton-bridged 

cobaloxime. The syntheses, X-ray structures, and electrochemical characterizations of Co(dAr F gBF2)2(py)2, and 

Co(dAr F gBPh2)2(py)2 are underway at this time. 

Committee Karate: Divide and Conquer 

Brian Merlob 

Mentor: Charles Plott 

The outcome of a group decision making process is inherently dependent on the voting system employed, the 

availability of information, and the presence of veto players. This dependence potentially provides key members of 

the committee—particularly agenda-setters—with near-dictatorial power over the outcome. We explore through 

experimentation and simulation what additional influence over the outcome is gained by an agenda-setter that can 

divide committees into subcommittees. 

Novel Computational Methods for High Speed Radio Transient Searches 

Omar Mezenner 

Mentor: Walid Majid 

Radio transients are celestial objects that emit radiation in short time windows. One example is a pulsar, a rotating 

neutron star that emits a beam of electromagnetic radiation. Detection and classification of radio transients 

requires us to boost signal-to-noise on large datasets acquired from radio telescopes observing the radio sky. 

Errors in the datasets are due to dispersion, an effect of electromagnetic propagation through plasma in the 

interstellar medium that causes the signal to be dispersed across multiple frequency channels unevenly, with lower 

frequency components lagging behind higher frequency components. This causes a widened pulse width in the 

time-domain, and therefore a lower signal-to-noise ratio. To correct dispersive effects, dedispersion techniques 

may be applied incoherently (after detection) or coherently (before detection). This process involves applying a 

correctional time shift to lower frequency channels to align the signal with that of higher frequency channels, and 

65

then summing across all channels. The dedispersed time-series then possesses a narrower pulse width and higher 

signal-to-noise. We are exploring novel techniques to implement dedispersion strategies on GPU platforms because 

of the highly parallelizable nature of vector computations, as high-end GPUs hold two to three orders of maginitude 

more cores than their contemporary CPU counterparts. 

Effects of DNA Mismatches on DNA-Protein Crosslinking via Guanine Oxidation 

Kelsey Miller 

Mentor: Eric Stemp 

8-oxoguanine is a common oxidation product in DNA and can lead to mutation. The metallointercalator 

Ru(phen)2dppz 2+ is a useful luminescent probe for DNA that has also found use as a guanine-selective oxidizing 

agent via the flash-quench technique. Here, we use its osmium analogue as a selective way to oxidize 8oxoguanine 

in double-stranded DNA, as visualized by oxidative DNA-protein crosslinking. With the 3+/2+ couple at 

1.15 V for Os (phen)2dppz 2+ , this metallointercalator should be able to oxidize 8-oxo-G (~0.7 V) without oxidizing 

guanine (~1.3V). In plasmid DNA where 8-oxo-G has been incorporated, flash-quench treatment with Os 

(phen)2dppz 2+ and Co(NH3)5Cl 2+ leads to crosslinking with histone. Furthermore, in gel shift experiments with a 

duplex of the oligonucleotide 5’-ATATGATAT8GATATGATAT-3’ (8 = 8-oxo-G) and its complement, flash-quench 

treatment with Ru(phen)2dppz 2+ in the presence of histone leads to a band of intermediate mobility (presumably 

1:1 crosslink) and to a band of well-shifted material. In contrast, analogous treatment with Os(phen)2dppz 2+ 

produces only the band of intermediate mobility, consistent with the presence of only site that is oxidizable by the 

osmium complex. 

The Solar Decathlon: Improving House Performance Through Appliance Optimization 

Zeke Millikan 

Mentor: Melany Hunt 

The Solar Decathlon is a biennial competition sponsored by the Department of Energy in which college teams from 

around the world design and build solar powered homes. The houses are brought to Washington D.C. where they 

are judged and opened to the public, showcasing new strategies of energy efficiency in residential living. The 

California Institute of Technology and the Southern California Institute of Architecture have joined together to 

make an innovative and energy conscious house for the competition. The home, CH:IP (Compact Hyper-Insulated 

Prototype), makes energy efficiency a priority through interconnected systems and careful energy budgeting. More 

than forty percent of the energy budget is devoted to appliances; thus their performance has a large impact on the 

overall home performance. The appliances were tested extensively for energy and water use in different settings 

and configurations. The resultant numbers were used to choose optimal operational strategies and to verify overall 

energy performance of the home. 

Exploiting the Weak Temperature Gradient Approximation for Climate Theory on Slowly Rotating, 

Tidally Locked Planets 

Sean M. Mills 

Mentors: Dorian Abbot, Raymond Pierrehumbert, and Yuk Ling Yung 

Tidally locked planets are subject to extreme temperature variations due to the stellar flux directly warming only 

one side of the planet. This is important because planets in the habitable zone around M dwarf stars are likely to be 

tidally locked. Such planets are unlikely to be habitable if their antistellar temperatures are low enough that CO2 

will condense. This problem has previously been investigated using GCMs, which explicitly solve for atmospheric 

dynamics. In order to gain a greater understanding of the effect of different mechanisms on the temperature 

profile, we use a lower-order energy balance model here. We consider tidally locked planets that rotate slowly 

enough that we can neglect the Coriolis force, which allows us to assume that atmospheric temperature at any 

given height is independent of horizontal position (the weak temperature gradient approximation). This allows us to 

easily isolate and contrast the effects of different physical mechanisms, such as greenhouse gas level and surface 

turbulent exchange, on the resulting temperature profile. We find that the effect of turbulent exchange on climate 

saturates at fairly low values (very smooth planets), whereas CO2 has a consistently strong effect on climate. 

Investigating the Role of miR125b in Hematopoiesis and Development of a Myeloid Leukemia 

Aarathi Minisandram 

Mentors: David Baltimore and Alex So 

miR125b has been shown to be an important player in hematopoiesis. When miR125b is overexpressed in mice, 

the mice develop severe leukemia. We identified this leukemia to be primarily myeloid in nature, although in 

certain cases, it can be lymphoid. The primary cause of the myeloid cancer was a rearrangement of the 

hematopoietic system favoring myeloid progenitors and decreasing the number of lymphoid progenitors. We 

developed a loss-of function system for miR125b expression in mice and confirmed that miR125b has a regulatory 

effect on hematopoiesis. Finally, we investigated the primary target of miR125b, lin28, and show that miR125b 

causes a rearrangement of the hematopoietic system by decreasing the expression of lin28. 

66

Proteomic Analysis of Histone Post-Translational Modifications and Their Cellular Targets 

Neeli Mishra 

Mentors: Benjamin A. Garcia and Jack L. Beauchamp 

Histones are small proteins that function to package genomic DNA into the basic unit of chromatin, the 

nucleosome. Post-translational modifications (PTMs) of histones are called epigenetic (heritable changes in gene 

expression that occur without alterations in gene sequence). Histone PTMs occur on multiple but specific sites, 

suggesting that histones can act as a signaling platform for nuclear events. A “histone code” hypothesis has been 

put forward to explain how these PTMs are responsible for the regulation of transcriptional activity through the 

alteration of chromatin states. In the first study, we seek to investigate the impact of epigenetic gene expression 

on both histone modifications and the nuclear proteome. The project focuses on the knockdown cell line of EZH2, 

a clinically significant histone methyltransferase associated with developing tumors. Using proteomics to directly 

investigate the cellular function, we combine SILAC labeling and mass spectrometry (MS) to characterize the 

alterations in abundances for nuclear proteins in these epigenetically modified cells. Histones are typically studied 

using electron capture/transfer dissociation or collision-induced dissociation MS. The second study focuses on 

testing an alternate, recently developed free-radical initiated peptide sequencing (FRIPS) technique to selectively 

cleave peptides at specific amino acid residues and potentially improve histone PTM analysis. 

Progress Toward the Synthesis of Gagunin E, a Polyoxygenated Terpene Isolated From the Sponge 

Phorbas sp. 

Austin Moehle 

Mentor: Brian Stoltz 

Gagunin E, one of a large set of polyoxygenated terpenes isolated from the sponge Phorbas sp., exhibits high 

cytotoxicity toward the human leukemia cell-line K562 with an LC50 of 0.03 μg/mL (50 nM). Despite its potent 

biological activity, a synthesis of gagunin E has yet to be reported. The initial goal of the project was to construct 

the tricyclic core of gagunin E, which contains fused 5-, 6-, and 7-membered rings. Although cyanthiwigin F, a 

diterpene recently synthesized by the Stoltz group, possesses a similar tricyclic core, gagunin E is more complex 

due to oxygenated stereocenters bearing various ester groups. An array of ruthenium catalysts were screened for 

the 7-membered ring-closing metathesis and several approaches to the construction of the three rings were 

attempted, ultimately resulting in the desired synthesis of the tricyclic core of gagunin E from succinic acid, a 

commercially available precursor. Future work on this project will involve performing C-H oxidations on this core to 

attach the appropriate hydroxyl groups and ester moieties, followed by an investigation of the effects of various 

hydroxyl functionalization patterns on this compound’s biological activity. 

Characterization of the Partial Duplication of the �7 Nicotinic Acetylcholine Gene (CHRFAM7A) and Its 

2-Base Pair Deletion Polymorphism 

Mahati Mokkarala 

Mentors: Henry A. Lester and Ying Wang 

Previous research indicates that the �7 nicotinic acetylcholine receptor mediates P50 auditory gating; a defect in 

the P50 auditory response is a common symptom for schizophrenia. Of particular interest is the partial duplication 

of exons 5-10 of the �7 nicotine acetylcholine receptor gene (CHRFAM7A) prevalent only in humans. A recent study 

found that CHRFAM7A has a 2bp deletion (CHRFAM7A2bpdel) polymorphism linked to schizophrenia. Unfortunately, 

the structure and function of the gene products of CHRFAM7A (dup�7) and CHRFAM7A2bpdel (dup��7) are 

unknown. We are studying the subcellular localization of dup�7 and dup��7 in neuronal culture by overexpressing 

fluorescently tagged �7, dup�7 and dup��7 in rat hippocampus neuron cultures and comparing the distribution of 

receptor subunit with stained cellular structures such as the endoplasmic reticulum. Receptor subunit localization 

patterns may help us understand the function of dup�7 and dup��7 in mammalian neurons. We are also 

investigating the N-terminal sequence of dup�7 and dup��7 overexpressed in IMR-32 human neuroblastoma cells 

by applying the Edman degradation technique. Optimization of transfection efficiency for rat hippocampus neuron 

and human IMR-32 neuroblastoma cell cultures are ongoing. 

Conformational Equilibria of 1,4-Butanedioic Acid and Its Anions as a Function of the Deuterium Oxide 

Mole Fraction in Mixed Solvent Environments of Deuterium Oxide and Dimethyl Sulfoxide as 

Determined Through NMR Spectroscopy 

Greg Moore 


Scalar J13/J14 vicinal proton couplings of 2,3- 13 C labeled 1,4-butanedioic acid and its anions were used to evaluate 

conformational equilibria as the mole fraction of deuterium oxide in mixed solvents of deuterium oxide and 

dimethyl sulfoxide was progressed from 0 to1. Results for all compounds considered showed a surprisingly gradual 

transition from the conformational preference exhibited in pure dimethyl sulfoxide to that seen in pure deuterium 

oxide. This gradual change suggests that deuterium oxide is not as effective at disrupting the intramolecular 

hydrogen bonding of monohydrogen 1,4-butanedioate or protonating the carboxylate groups of 1,4-butanedioate in 

67

dimethyl sulfoxide solutions as was initially thought. Additionally, the calculated fraction of gauche conformers for 

1,4-butanedioate in pure anhydrous dimethyl sulfoxide was ~.20, a lower figure than the ~.40 previously reported 

and more consistent with computational predictions. 

Activation Responses to Hallucinogen Administration in a Mouse Model of Maternal Immune Activation 

Marlyn J. Moore 

Mentors: Paul H. Patterson and Natalia Malkova 

Hallucinations are a hallmark symptom of schizophrenia. Moreover, schizophrenic patients show a greater response 

to hallucinogenic drugs than control subjects. Epidemiologic studies have found that maternal immune activation 

(MIA) increases the risk of schizophrenia in the offspring. In a mouse model of this risk factor we analyzed the 

behavioral and molecular response of MIA offspring to an injection of the hallucinogenic drug DOI. Compared to 

controls, these mice display greater behavioral responses and higher levels of the immediate early gene (IEG) 

egr-1, a marker of neuronal activity, in prefrontal cortex. Thus, as in human schizophrenia, MIA offspring appear to 

be more sensitive to a hallucinogen. 

Mist Control Due to the Complementary Interactions of Polymer End Groups 

Rachel S. Moore 

Mentors: Julia A. Kornfield and Jeremy Wei 

The reason airplane crashes are so deadly is that the, during the crash, the fuel forms a highly combustible, fine 

mist. Adding a certain polymer to the fuel can control this fine mist by making the droplet size bigger, rendering 

the mist non-explosive. Past polymers have been able control misting, but they were not usable due to either shear 

degradation or insolubility in fuel. My polymer is designed to associate via reversible, complementary associating 

groups on the polymer ends. This reversibility is what is unique about this polymer, and the complementary nature 

of the association allows for efficacy at low concentrations without drastically altering the rheological properties of 

the fuel. In my studies, I utilized different methods for characterizing the polymers, including rheology, a spray 

test, a combustion test, and a shear degradation test. All tests so far indicate that the polymer associates in the 

predicted way and that the interaction between polymers is very strong. This strong interaction means that only 

four associating groups on each end need to be used. In turn, this means that the synthesis and mass production 

of the polymers will be feasible. 

Low Cost, High-Density Protein Arrays for Sensors and Fuel Cells 

Melody A. Morris 

Mentors: Bradley Olsen, Carla Thomas, Matthew Glassman, and David Tirrell 

Compared to traditional chemical synthesis, the use of enzymes has dramatically expanded the diversity of 

chemical conversions possible. Enzymes have many advantages: substrate specificity, high catalytic rate, mild 

operating conditions, biodegradability and renewability, and the ability to operate on non-traditional substrates. 

However, enzymatic catalysts and sensors face three key limitations that this project seeks to remedy: low density 

of active sites, short enzyme lifetimes or limited operating conditions, and high costs of enzyme purification and 

device construction. Recent work has shown that conjugates of globular protein and thermoresponsive polymer will 

self assemble into lamellar nanostructures with high protein density and preservation of folding. Here carbonic 

anhydrase (hCAII), an enzyme that converts carbon dioxide and water into bicarbonate and protons, was chosen 

as a model catalytic globular protein due to its potential application in emission reduction and environmental 

remediation. A thermoresponsive elastin-like protein (ELP) was fused to hCAII as a tool for low-cost purification 

and to direct self-assembly of the enzyme into nanostructured plastics and hydrogels. This project aims to develop 

methods for self-assembly of hydrated gels and solid-state nanostructures of ELP-enzyme fusion proteins, to 

explore the effects of ELP secondary structure on the formation of the nanostructures, and to enhance stability and 

activity of the enzyme. 

Multiple Laser Frequency Modulated Ranging System 

Alexander T. Mouschovias 

Mentors: Amnon Yariv and Arseney Vasilyev 

We investigate the combination of swept frequency optical waveforms for resolution enhancement in a diode laser 

frequency-modulated ranging experiment. To this end we build an optoelectronic feedback system capable of 

stitching four swept-frequency laser sources. We use this system to concatenate chirped vertical cavity surfaceemitting 

lasers (VCSELs) at wavelengths of 1550 and 1310 nm, and demonstrate a four-fold improvement in range 

resolution over a single source ranging system. 

68

The Incidence of Active Galactic Nuclei in Galaxy Groups 

Eric Mukherjee 

Mentors: Christine Jones, Andrew Goulding, and Fiona Harrison 

Properties of Active Galactic Nuclei (AGNs) are linked to properties of their host galaxies. Reseach suggests AGNs 

are the result of accreting black holes at centers of galaxies, and affect properties like star formation rate, stellar 

mass, and galaxy color. Knowing how the large scale environment influences AGN evolution may further our 

understanding of this relationship. We seek to constrain AGN environments in the 9 square degree Chandra Bootes 

survey. We used ~15,000 optically observed galaxies in the AGN and Galaxy Evolution Survey (AGES) catalog to 

determine whether AGN form in groups. We developed a friends-of-friends algorithm to locate galaxy groups 

associated by distance and velocity, and a matching program to associate groups with possible X-ray detections. 

We found 4,758 (~30%) galaxies in 1,554 groups in the AGES catalog with 0.01

Density of States of Nematic Emulsion 

John C. Napp 

Mentors: Arjun Yodh, Matthew Lohr, and Fiona Harrison 

Knowledge of the vibrational density of states of a material is a fundamental key to understanding its physical 

properties. In this project, the vibrational density of states of an emulsion of glycerol in the nematic liquid crystal 

(NLC) 4'-pentyl-4-biphenylcarbonitrile (5CB) is investigated. A thin layer of NLC is placed on top of a thick layer of 

glycerol by dissolving the 5CB in hexane, which spreads evenly over the glycerol and quickly evaporates. The 

system is then heated, allowing the 5CB to enter the isotropic phase and the glycerol to diffuse through the liquid 

crystal. When cooled, the 5CB reenters the nematic phase, and the glycerol condenses into droplets in a hexagonal 

lattice pattern. Some of these droplets are trapped at the NLC-air interface, and their vibrational motion can be 

tracked via video microscopy and standard particle tracking algorithms. Measurement of the displacement 

correlations between particles yields information on the vibrational properties of the system which allows 

calculation of the vibrational density of states. 

Phosphino–Borane Transition Metal Complexes for Energy Storage Reactions 

Luis Navarro 

Mentors: Jonas C. Peters and William H. Harman 

This project’s goals have been the synthesis and characterization of first row transition metal complexes using 

diphosphino-borane ligands and studying their reactivity as they relate to activation of hydrogen and carbon 

dioxide. Altering the sterics of these systems has been shown to dramatically alter their reactivity and particularly 

the nature of the boron-metal bonds. 

Investigating the Protein and RNA Components of the Chromatoid Body 

Pushpa Neppala 

Mentors: Alexei Aravin and Evelyn Stuwe 

Small non-coding RNAs are known to associate with the Argonaute (Ago) family of proteins, and their presence 

usually results in the silencing of the target gene expression. A just recently discovered class of small noncoding 

RNAs, the piRNAs (Piwi-interacting RNA), interacts with Piwi proteins (a clade of the Ago family) and is involved in 

repression of transposable elements in germ line cells. In mammalian male germ cells, a cloudlike cytoplasmic 

structure called the chromatoid body has been thought be composed of proteins and small RNAs involved in the 

piRNA pathway. Elucidating the nature of the chromatoid body could help to improve our understanding of 

translation and the highly complex process of spermatogenesis. The objective of this research is to further study 

and identify the protein and RNA components of the chromatoid body, through developing a novel isolation protocol 

that couples cross-linking techniques with immunoprecipitation in mouse testes. After formaldehyde treatment and 

immunoprecipitation of a known protein component of the chromatoid bodies, MILI, proteins were identified 

through Western blot analysis. RNA molecules were also isolated and identified through quantitative PCR. In the 

future, the RNA will be sequenced and we hope to perform mass spectrophotometry on the chromatoid bodies. This 

involves also developing an isolation protocol in which the chromatoid bodies are not cross-linked but can still be 

immunoprecipitated while kept intact. 

Social Judgments of Other People: Comparing 2D, 3D, and Live Formats 

HongAn Nguyen 

Mentors: Ralph Adolphs and Laura Loesch 

We make inferences about people every day based on how they present themselves. People are free to post 2D 

videos of themselves as well as make video calls via the internet. With the recent introduction of consumer-level 

3D technology, they have yet another format for interacting with the greater community. We were interested in 

whether the inferences people make about each other in terms of attributes such as trustworthiness, social 

dominance, and physical threat, as well as the facial information they used to make those judgments, were 

influenced by the format of the interaction. Participants’ eyes were tracked as they looked at facial stimuli and gave 

explicit social judgment ratings. The facial stimuli varied across these different formats of interaction: 2D and 3D 

video, live one-way video, live two-way video, actors through a one-way mirror, and actors interacting face-toface. 

We plan on looking at the ratings using ANOVA and at eye tracking information through the SHINE toolbox in 

MATLAB to see if there are statistically significant differences among these formats of presentation. 

Programmable Genetic Regulation by sRNA-mRNA Hybridization 

Paul D. Nguyen 

Mentors: Richard M. Murray and Joseph Meyerowitz 

Small non-coding ribonucleic acids (sRNAs) play key roles in regulatory pathways in an extensive range of 

prokaryotes and eukaryotes. Drawing inspiration from this natural device, we have begun an investigation to 

imitate the mechanism of regulation in vivo with Escherichia coli, and develop engineered and programmable 

biological circuitry that utilizes sRNA-mRNA interaction for regulation. We constructed a simple genetic circuit that 

70

is designed to give a delayed response of a reporter gene for green fluorescent protein (GFP). This circuit can be 

used in a programmable way to give a desired regulation of the GFP gene by some intended temporal delay in 

expression, as shown by our computer simulations of the system. To confirm the intended delayed regulatory 

capabilities of our genetic circuit, we characterized this system by taking various fluorescence measurements of the 

cells with this system against other cells. 

Non-Covalent, Self-Assembled Polyads for Multistep Electron and Energy Transfer 

Eva Nichols 

Mentor: Daniel T. Gryko and Harry Gray 

Systems for multistep electron and energy transfer are important to the field of solar energy because they provide 

access to long-lived charge-separated states, allowing electrochemical potential to be harnessed for catalysis. 

Covalently bonded donor-acceptor systems have been documented, but a polyad system capable of self-assembly 

via intermolecular interactions still remains an elusive target. The design of three- and four-component noncovalent 

supramolecular systems for electron and energy transfer is described. Individual donor and acceptor 

moieties have been designed and synthesized with groups that allow them to self-assemble in a complementary 

and site-specific fashion (via hydrogen bonding and ionic interactions). Preliminary data concerning the selfassembly 

and electron or energy transfer of each supramolecular complex will be presented. 

State Space Partitioning Algorithm in TuLiP 

Lars Petter Nilsson 

Mentors: Richard Murray, Necmiye Özay, and Ufuk Topcu 

The Temporal Logic Planning (TuLiP) Toolbox is a Python software package implementing the correct-byconstruction 

controller synthesizer method Receding Horizon Temporal Logic Planning. One step in this method is 

to partition the continuous state space into discrete states and establish reachability relations between the discrete 

states. Cells in the partition are described by convex sets with flat sides, called polytopes. State space 

discretization requires intersections, set differences and projections to be performed on polytopes, the optimal 

choice of method for these operations depends on the characteristics of the polytopes. The algorithm used for 

discretization can be tuned with respect to cell requirements and termination criteria in order to produce a good 

partition. 

We also investigate if using overlapping instead of disjoint polytopes can decrease computational complexity. The 

discretization algorithm is evaluated by letting TuLiP synthesize a controller for a fuel tank problem. 

Experimental Investigation Into Shock Wave Boundary Layer Interactions of Reflected Detonation 

Waves 

John J. Odell 

Mentors: Joseph E. Shepherd and Jason Damazo 

Accurate prediction and modeling of the damage caused by gaseous detonation necessitates understanding the 

physics of shock wave reflection as it pertains to supersonic combustion waves. Previous experimental evidence 

suggests that detonation reflection deviates from one-dimensional inviscid wave reflection theory thus 

inhibiting accurate prediction of material deformation. We propose that this deviation is caused by interaction 

between the boundary layer induced by the incident detonation with the shock wave created when a detonation 

undergoes normal reflection analogous to the classic phenomenon of shock wave bifurcation studied in shock 

tubes. This shock wave boundary layer (SWBL) interaction is experimentally investigated using optical techniques 

alongside pressure and thermocouple measurements for the case of nitrous oxide--hydrogen detonations of varying 

stoichiometry. These measurements allow us to examine the boundary layer development behind gaseous 

detonations and the boundary layer effects on both the fluid dynamics of the reflected shock wave and the 

destructive potential of these waves: It is shown that viscosity plays a key role in the behavior of reflected 

detonation waves. Analytical theory is developed corresponding to non-reacting shock wave bifurcation as 

pioneered by Mark to explain the observed phenomena and highlight the differences between shock and detonation 

waves. 

Microfluidic Sensors With Integrated Germanium Photodetectors 

Nicholas Ogden 

Mentors: Axel Scherer and Jinqing Wang 

Microfluidic devices have great potential to supplement or replace current biochemical analysis machines such as 

those used for blood tests. For this to happen, microfluidic and electronic technology must be compatible. This 

requires accurate and reliable sensors to be developed for use in microfluidic devices. In this case, an optical 

sensor is considered due to the small size, good sensitivity, and lack of required ancillary equipment. Our sensor 

uses a silicon waveguide to channel 1.55um light (commonly used in telecommunications) into a germanium 

photodetector. Because, the size and shape of the germanium is critical in the performance of the sensor, Finite- 

Difference Time-Domain software is employed to test potential designs. Construction starts with a silicon-on- 

71

insulator substrate, on which a silicon waveguide is deposited. A germanium strip is then deposited and the device 

undergoes thermal annealing to improve its crystallinity. Lastly, a layer of gold is deposited on both sides of the 

germanium to serve as electrical contacts for analysis circuitry. 

In situ Amplification for Imaging miRNAs and mRNA Splice Variants 

Kanenori (Kane) Okamoto 

Mentors: Niles Pierce, Aneesh Acharya, and Harry Choi 

RNA plays a key role in gene regulation; it is involved in almost every DNA-related cellular process. In addition to 

providing transcripts for protein synthesis, RNA can promote gene expression, silence genes, and catalyze various 

cellular reactions 1 , among many other functions. Imaging mRNA expression patterns in intact cells or organisms 

with in situ hybridization technology has helped bring about numerous discoveries in the field of embryonic 

development and gene regulation. Recent developments such as locked nucleic acid (LNA) probes have enabled 

in situ imaging of mRNA splice variants 2 and microRNAs (miRNAs) in single mammalian cells 3 and whole-mount 

embryos 4-6 . Despite these advances, multiplexed in situ imaging of mRNA splice variants and miRNA within a single 

sample has not yet been demonstrated. Multiplexing minimizes sample variations and allows us to study the 

relationships among genes simultaneously in the same sample. The Pierce group has developed a multiplexed 

fluorescent in situ amplification using Hybridization Chain Reactions (HCR) in intact zebrafish embryos under the 

microscope 7 . In our studies, we have improved and adapted this technology for use with multiple initiator probes 

and with flow cytometry for quantitative analysis and propose a multiplexed in situ procedure to image miRNAs and 

splice variants in mammalian cells. 

Redesigning IgG Antibody 3BN60 to Improve Neutralization of HIV 

Alejandra I. Olvera 

Mentors: Pamela Bjorkman and Jennifer Keeffe 

Human Immunodeficiency Virus (HIV) has several properties that make it successful at evading the immune 

system, thus preventing antibody-mediated neutralization. One property is HIV’s ability to rapidly mutate and 

conceal epitopes on its trimeric gp120-gp41 envelope spikes. Recently, Pamela Bjorkman and colleagues suggested 

a novel evasion strategy by HIV: that the low density of HIV spikes impedes antibodies from binding with avidity, 

the ability to bind two epitopes simultaneously. Avidity enhances the affinity of an antibody for its antigen and has 

been shown to be important in the neutralization of other viruses. To address this problem, the Bjorkman Lab 

seeks to reengineer the structure of antibodies to enable multivalent binding. As part of this goal, we have 

designed novel antibody-like reagents with two Fabs joined by a variable length linker. We expressed the Fabs 

separately and ligated them with Sortase A, an enzyme that catalyzes the creation of a novel peptide bond. By 

changing the length of the linker we can optimize the distance between the antigen binding sites and potentially 

achieve inter- or intra-spike cross-linking, creating an avidity effect that may enhance the antibody’s potency 

against HIV. These type of antibody-like reagents could be administered via passive immunization or gene therapy 

to HIV infected individuals and inform efforts to develop a vaccine. 

Lewis Acid Centers in Zeolites: On the Route to Mechanistic Understandings of Carbohydrate 

Conversions in Aqueous Media 

Arna Pálsdóttir 

Mentor: Mark E. Davis 

There is a growing emphasis on conversion of biomass to fuels and chemicals. Thus, reactions of carbohydrates in 

aqueous media are of importance. The Davis lab has discovered that sugars can be isomerized by zeolites with 

Lewis acid centers. The Lewis acid centers in the zeolite involve Sn or Ti atoms that are four coordinated with 

oxygen. The oxygens are either all coordinated to silicon (denoted the closed site) or three silicon atoms and a 

proton (denoted the open site). The open site has an adjacent silanol group. I investigated whether this silianol 

group is required for the reaction mechanism. The reason is that previously, the Davis group showed that this 

silianol group participates in oxidation reactions with Ti containing zeolites. Here, I analyzed Ti and Sn zeolites by 

IR methods to observe the presence of the silanol group and when the proton was substituted by a sodium ion. 

Upon sodium ion substitution, the Ti zeolites are not active for oxidation reactions. However, when erythrose is 

used as a model sugar, the erythrose is converted to erythrulose (Lewis acid mediated isomerization). Thus, the 

silanol group is not required for the isomerization reaction. Based upon these results, we are not able to distinguish 

whether the active center is the open or closed site or both. 

Monitoring Searching Behavior and Arousal in Drosophila melanogaster 

Ketaki Panse 

Mentor: David Anderson 

Arousal, a state characterized by increase in activity and sensitivity, is fundamental to many animal behaviors, but 

its neuronal mechanism is poorly understood. The purpose of this project is to study the neuronal mechanism of 

arousal in Drosophila melanogaster. To do so, we are using sugar to induce arousal by quantifying the effects of 

sucrose on the searching behavior of starved Drosophila melanogaster. Here, we show that when starved flies 

72

come in contact with water, they do not show local searching behavior. However, when the flies come in contact 

with sucrose solution, searching behavior increases, and their path consists of several turns and loops near the 

original food source. Similar behavior has been documented with C. elegans and blow flies, suggesting that there is 

a fundamental behavior common for several species. To ensure that this behavior is caused by activity of sugar 

neurons and not feeding, we will repeat the experiments by expressing Channelrhodopsin 2, a light sensitive cation 

channel, in sugar neurons to induce the behavior with light. To understand the neuronal mechanism and detect 

which neuromodulator is responsible for this behavior, we will perform screening for mutants and use calcium 

imaging. 

Intelligent Queue Scheduling for Robo-AO 

Athanasios Papadopoulos 

Mentors: Christoph Baranec and Reed Riddle 

This presentation discusses the design and implementation of the intelligent queue scheduling software for the 

Robo-AO adaptive optics system. Robo-AO will be the first robotic laser guide star adaptive optics system and will 

be able to deliver diffraction limited resolution in both the visible and the near-infrared spectrum. Due to the fact 

that Robo-AO is completely autonomous, a queue scheduler will be used to prioritize a set of scientific targets, 

making the choice of which target to observe after each observation is completed. The queue scheduling software 

will make the choice of its next observation by taking into account all the important parameters and input data for 

each target. These include the scientific importance, the current observing conditions and feedback about the 

current performance of the adaptive optics system. The logic of the optimization is, in essence, balancing the 

quality of the observations, the slewing time overhead and the scientific priority of a target. 

Atomic Force Microscopy Studies of the Effect of ATP on XPD Helicase Interactions With DNA 

Alison Parisian 

Mentors: Jacqueline Barton and Pamela Sontz 

DNA-mediated charge transport (DNA CT) is a proposed mechanism used by repair proteins to cooperatively search 

for lesions on DNA. Since DNA CT is significantly attenuated over a strand containing mismatches, proteins 

containing a redox-active [4Fe4S] cluster may use the strength of signal sent across a strand by another protein to 

determine the presence of a mismatch. XPD is a 5'-3' DNA helicase involved in the nucleotide excision repair 

pathway which contains a [4Fe4S] cluster and binds to ATP. It has been shown to redistribute onto mismatched 

strands, but it was unclear as to the effect of ATP on the protein's activity and redistribution. In our experiments, 

atomic force microscopy was used to visualize and quantify the binding of XPD to strands of varying lengths, some 

containing a mismatch, in the presence of ATP. It was found that over a 24 hour incubation with ATP the helicase 

activity of the protein was activated, overriding the CT mechanism, but without the incubation time, ATP has little 

effect on the redistribution of XPD. Future studies will focus on visualizing the helicase activity of XPD and its 

activation by ATP. 

Introduction of the Azide Chemical Reporter to Newly Synthesized Fatty Acids for Lipid Detection and 

Antibiotic Screening 

Sungjin Park 

Mentors: David A. Tirrell and Janek Szychowski 

Once thought to be simple in nature, lipids are now known to be incredibly diverse, and to play key roles in many 

biological processes. However, the lack of non-radioactive tools and methods make the study of lipids difficult. The 

modern “chemical reporter strategy,” which involves the labeling of biomolecules by chemical reporter groups, was 

implemented to effectively label newly synthesized lipids within cells. Vibrio harveyi was fed with N3(CH2)5CO2H, a 

short chain azido fatty acid, and analysis of click-reacted cellular lipid extract by HPLC-UV-MS confirmed the 

successful elongation and incorporation of N3(CH2)5CO2H by the cellular lipid biosynthetic machinery. In order to 

improve incorporation of the azide chemical reporter in lipids, the effects of concentration, pulse length, and 

introduction time point of N3(CH2)5CO2H on azido fatty acid incorporation were studied. The level of incorporation of 

the azide group was determined by measuring the fluorescence of cells treated with a coumarin-cyclooctyne 

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 

able to incorporate the azide chemical reporter into their newly synthesized lipids most efficiently. Using these 

optimized conditions for incorporation of the azide group in lipids, fatty acylated proteins were detected by SDS- 

PAGE in-gel fluorescence using a TAMRA-cyclooctyne conjugate. In addition, the observation of the significant 

disruptive effect of cerulenin, a fatty acid biosynthesis inhibitor, on N3(CH2)5CO2H incorporation in cellular lipids was 

used to take steps towards developing a high-throughput cell-based antibiotic screen for identifying antibiotics 

which specifically interfere with fatty acid synthesis. 

73

Transcription Factor Sox10 May Be Required for Microvillous Neuron Formation in the Zebrafish 

Olfactory System 

Brian Peng 

Mentors: Marianne Bronner and Ankur Saxena 

The zebrafish olfactory system has three distinct types of sensory neurons: ciliated, crypt, and microvillous. 

Microvillous neurons are an important component of olfaction across species, but their origin has remained unclear. 

In the Bronner lab, previous observations have shown transgenic Sox10:eGFP-labeled cells, thought to be neural 

crest, migrating into the olfactory region and differentiating into microvillous olfactory neurons. To determine the 

relevance of Sox10 in microvillous neuron formation, the expression profile of Sox10 was analyzed. Embryos were 

examined at different stages of development for Sox10 protein expression and assayed for colocalization with a 

variety of proteins including TRPC2, which is a marker for microvillous neurons. Results indicate that Sox10 protein 

is indeed highly expressed in neural crest cells that differentiate into microvillous neurons. In addition, loss-offunction 

experiments were performed using morpholino injections to determine Sox10’s role in microvillous neuron 

formation. Analysis was done primarily with confocal microscopy; preliminary results indicate a requirement for 

Sox10 in the differentiation of neural crest into microvillous neurons. Future experiments will investigate Sox10 

mRNA expression in order to fully determine the spatiotemporal span of Sox10 induction during neural crest 

migration into the olfactory region. 

Siegel Modular Varieties: Arithmetic Invariants, Cusps, and Real Points 

Christopher Perez 

Mentor: Andrei Jorza 

Siegel modular forms are a generalization of elliptic modular forms to several complex variables. As in the case of 

elliptic modular forms, there is an associated modular variety known as a Siegel modular variety. We attempt to 

generalize the notion of modular symbols to Siegel modular varieties and use these to generate a basis for Siegel 

modular forms. 

Hot Dust in Bright Galaxies: Using Mid-Infrared Photometry to Determine the Sources of Extragalactic 

Radiation and Spectroscopy 

John Pharo 

Mentor: Lee Armus 

Ultraluminous Infrared Galaxies (ULIRGs) have some of the highest energy output in the known universe. This is 

usually the result of some combination of stellar light, widespread star formation, and the accretion of matter by a 

central, supermassive black hole. These galaxies, however, contain vast amounts of dust, which absorb much of 

the energy produced and re-emit it as infrared radiation. By measuring the flux through the galaxy at different 

wavelengths of the infrared spectrum, and then by fitting and decomposing the spectrum, one can determine the 

respective energy contributions of star formation and accreting black holes. Many known ULIRGs are missing data 

between 2 and 5 microns, which can have significant effects on the shape and composition of the fit, as emissions 

from hot dust occur in this region. The Spitzer Space Telescope recently took IRAC images of 86 such objects, 

allowing us to measure fluxes at 3.6 and 4.5 microns. We are currently using using these data together with 

archived spectra and photometry from other telescopes to fit the spectra and estimate the energy contribution of 

an active, central black hole. 

Elk3 Is Required for Neural Crest Specification 

Jacquelyn Phillips 

Mentors: Marianne Bronner and Crystal Rogers 

Elk3/Net/Sap2 (here referred to as Elk3) is involved in tumor angiogenesis in mice, but is expressed in various 

tissues throughout embryonic development. Elk3 expression was characterized using whole mount in situ 

hybridization, and we found that Elk3 expression is very dynamic. During development, Elk3 is expressed in the 

developing head folds, the head mesenchyme, the developing somites, and in the migratory cranial neural crest. 

Because Elk3 is expressed in premigratory and migratory neural crest cells, we believe that it is involved in cranial 

crest specification and migration and may be required for the formation of many neural crest derivatives. Neural 

crest cells migrate throughout embryos and become various types of cells, including neurons, enteric ganglia, and 

melanocytes. Based on the Elk3 expression pattern, we wanted to explore the effect of Elk3 knockdown on neural 

crest cell development. Here we show that loss of Elk3 protein decreased the number of neural crest cells marked 

by Sox10 expression, which suggests a role for Elk3 in neural crest specification. 

74

Elucidation of the Architecture of a Protein Aggregate: The Role of Binding Interactions and 

Intermolecular Packing Interactions Within the Aggregate 

Samantha Piszkiewicz 

Mentors: Shu-ou Shan and Thang Nguyen 

Protein homeostasis is essential for all cells and requires the proper control of the folding, localization, and 

interactions of all proteins. The misfolding and aggregation of proteins are detrimental to cells and have been found 

to be the root cause of numerous age-related diseases. To study the general mechanism of disaggregase of a 

typical hydrophobic membrane protein, we will use LHCP as a model system to define the nature of such 

aggregates. The Shan Lab has recently shown the robust disaggregase activity of plant protein, the 43 kDa 

chloroplast signal recognition particle (cpSRP43), for the disassembly of aggregates of the family of the multitransmembrane 

light harvesting chlrophylla/b –binding (LHC) protein. We have shown through alkylation 

experiments of single-cysteine constructs of the LHC protein that L18 is a dominant sequence within the LHC 

aggregate, meaning that it is always presented on the surface of the aggregate, poised for disaggregase 

recognition. It appears that the third transmembrane domain of LHC is most involved in intermolecular packing 

interactions in the aggregate. 

Developing and Optimizing Polymer-Actuated Microfluidic Valves for Portable Applications 

Ahalya Prabhakar 

Mentors: Michael Roukes and Blake Axelrod 

The field of microfluidics has become a focus of nanobiotechnology due to the potential for lab-bench based 

biological experiments on a microscale. From large-scale arrays to single-cell assays, microfluidics allows 

researchers to conduct the same experiments in the lab for significantly less reagent cost, space, and time. Many 

groups have been working on portable microfluidic systems, but have focused on passive, valveless control. We are 

working on developing a portable, active-control microfluidic system through the use of electroactive polymers. I 

have focused on optimizing the polymer growth, chip design, and efficient and consistent voltage delivery for the 

control of the valve actuation. We are working towards successful long-term actuation of independent valves, which 

will work reliably for practical use. 

Corneal Fibroblast Migration on Cross-Linked Gelatin Hydrogels With Nanofibrous Scaffolds 

Amy C. Proctor 

Mentors: Julie Kornfield and Amy Fu 

The cornea is the clear protective surface that covers the front of the eye and aids in focusing and refracting light. 

Corneal endothelial cells cannot be regenerated if they are destroyed by disease or trauma, and blindness can 

result from the loss of too many endothelial cells. Synthetic corneal prostheses have successfully been used to 

treat corneal injuries; however, epithelialization of the implant surface remains a challenge. Although 

epithelialization is not necessary to restore vision, epithelial cell coverage can increase retention rates and reduce 

the risk of infection after implantation. Here we investigate the in vitro cellular migration of rabbit corneal 

fibroblasts on cross-linked gelatin hydrogels covered with cross-linked gelatin nanofibers of various directionalities 

as a first step toward developing a protocol to promote epithelization in a clinical setting. 

Identification of Regulatory Genes in the Endodermal Gene Regulation Network in Strongylocentrotus 

purpuratus During Gastrulation 

Ralph Edward Pursifull 

Mentors: Eric H. Davidson and Isabelle Peter 

Hnf1aLike and Hnf4 are transcription factor genes that are expressed in the later stages of development, from 30 

hours post fertilization (hpf) to 72 hpf and are potential components of the endoderm Gene Regulatory network. 

This project aims to detect the temporal and spatial localization of the expression of the targeted genes, Hnf1aLike 

and Hnf4 through in-situ hybridization. Expression patterns indicated by the in-situ hybridization will indicate 

whether the changes in expression of the genes within the endoderm cause changes to the regulatory state of the 

endoderm region. If this is the case, the expression patterns can be analyzed alongside other transcription factors, 

indicating potential factors that alter the expression of the target genes, as well as genes that experience altered 

expression based on the presence of Hnf1aLike and Hnf4. This will serve as a basis for further work to establish 

relations between the target genes and other genes in the late endoderm Gene Regulation Network of sea urchins. 

Surface Imaging of Mixed Methyl/Thienyl Monolayers on Si(111) 

Suyeon Pyo 

Mentors: Nathan S. Lewis and Leslie O’Leary 

Methyl monolayers and mixed methyl/thienyl monolayers on Si(111) were synthesized through a two step 

chlorination/alkylation procedure. The terrace size and surface roughness were investigated using transmission 

fourier transform infrared spectroscopy (FTIR) and atomic force microscopy (AFM). The surface composition was 

characterized via X-ray photoelectron spectroscopy (XPS) and the distribution of functional groups on mixed 

75

monolayers was determined using scanning tunneling microscopy (STM). The mixed methyl/thienyl monolayers 

were composed of 5-10% of thienyl groups, 85-90% of methyl groups and few hydrogen groups. AFM images 

showed straight terraces of 100-200nm width which were parallel to each other. The statistical distribution of 

functional groups will give deeper insight into monolayer formation via halogenation/alkylation, reactivity of 

monolayers, efficiency of secondary functionalization and catalyst attachment. 

Regulation of Huntingtin and REST by the IκB Kinase Complex in Huntington’s Disease 

Mike Qian 

Mentors: Paul H. Patterson and Ali Khoshnan 

Expansion of a polyglutamine repeat at the N-terminus of the huntingtin (Htt) protein causes Huntington’s disease 

(HD), a neurodegenerative disorder characterized by progressive cognitive decline, chorea, and psychiatric 

disturbances. Although the molecular pathways that regulate the disease remain poorly understood, many of its 

deleterious effects in the brain have been attributed to aberrant nuclear accumulation of neurotoxic N-terminal 

fragments of mutant Htt. The zinc-finger protein RE1 silencing transcription factor (REST), a master regulator of 

neuronal differentiation, is normally sequestered in the cytoplasm by Htt but becomes activated in HD, causing 

neuronal genes to be incorrectly silenced. REST levels are elevated in HD animal and cellular models. Many studies 

in these models have also implicated the IκB kinase (IKK) complex in HD pathogenesis. Interestingly, the REST 

promoter contains several binding sites for the transcription factor NF-κB. Since IKKβ plays a key role in activating 

NF-κB, we began testing a panel of REST promoter constructs in a cellular model to determine if blocking NF-κB 

activity could abolish abnormal REST expression. Preliminary results indicate that mutating NF-κB binding sites in 

the REST promoter decreases promoter activity, thus suggesting a possible link between IKKβ and REST 

expression. Future experiments will focus on direct inhibitors of IKKβ, which should help to establish whether this 

represents a novel pathway in the neuropathology of HD that could pave the road toward a promising new 

therapeutic strategy. 

Testing Properties of Parylene-C With X-Ray Photoelectron Spectroscopy 

Chenxi Qiu 

Mentors: Yu-Chong Tai and Jeffrey Lin 

Parylene-C is a class of poly(p-xylylene) polymers widely used for many biomedical implantable devices due to its 

high biocompatibility. The main use of a parylene is through a vapor deposition polymerization (VDP) process 

developed by Gorham, where a thin film of parylene is developed around usually sensitive materials to prevent 

them from corroding in harsh environments like the human body. This project focuses on understanding what 

causes parylene to become brittle after being processed at 100C to 200C and discovering methods to prevent it. 

X-Ray Photoelectron Spectroscopy (XPS) is a machine that irradiates the surface of a material with X-rays while 

simultaneously measuring the kinetic energy and number of electrons that escape from the material. With the use 

of XPS, it has been determined that oxidation might not be one of the causes of the brittleness for parylene 

processed at 100C in air because no significant amount of oxygen was discovered. Further work will involve finding 

a curve over temperature and time for the densification of parylene and verifying the possibility of crystallization as 

a cause for the brittleness. 

Stochastic Frequency Modulation of Bacillus subtilis Stress Response 

Sofia Quinodoz 

Mentors: Michael Elowitz and James Locke 

Genetic networks employ dynamic strategies to respond to different conditions. For example, it was recently shown 

that cells can activate gene expression in “pulses” under stress conditions. In particular, the sigB alternative sigma 

factor in B. subtilis exhibits discrete pulses of activation whose frequency is modulated by external stresses. To 

follow up on this discovery, I examined the dynamics of twelve other B. subtilis alternative sigma factors using 

quantitative single cell time-lapse microscopy. Alternative sigma factors control key regulatory pathways in 

bacteria, including chemotaxis response and antibiotic resistance. My investigations are providing a 

characterization of the dynamics of induction of these diverse alternative sigma factors and the dependence of 

these dynamics on the magnitude of activating inputs for each factor. 

A Study of the Binding Affinities of the Family I Carbohydrate-Binding Modules (CBM) and the 

Characterization of the CBM Chimeras Constructed Using SCHEMA Recombination 

Misha Raffiee 

Mentors: Frances H. Arnold and Indira Wu 

The skyrocketing cost of fossil fuels has given rise to the need to reduce America’s dependence on petroleum. By 

converting cellulosic biomass to glucose, and in turn, converting the glucose to biofuels, it is possible to reduce our 

dependence on petroleum-based fuels. Cellulose is known to be degraded by certain enzymes produced by specific 

bacterial and fungal organisms. Trichoderma reesi, a filamentous fungus with this ability, contains an array of 

cellulases which act as catalysts to the hydrolysis of the insoluble crystalline cellulose, which yields glucose as a 

product. Cellobiohydrolase, one such cellulase, has a unique structure consisting of a cellulose binding module 

76

(CBM) and a catalytic domain coupled by a linker region. Using overlap extension PCR, seven different CBMs were 

successfully fused with a previously engineered catalytic domain. The binding curves of the CBMs were examined in 

an effort to engineer and improve the binding affinities of the cellobiohydrolases. This research also utilized 

SCHEMA recombination to identify specific beneficial amino acid blocks in the CBMs. The goal of the study is to 

improve the binding affinity of the Family I CBMs, and allow for the successful and efficient conversion of biomass 

to biofuel. 

Automated Information Retrieval System (AIRS): An Economic Solution to Data Recovery From High 

Altitude Balloons 

Anthony Ransford 

Mentors: Jason Rhodes and Jeffery Booth 

Large amounts of data from high altitude balloons can be accrued during astronomic research such as the High 

Altitude Lensing Observatory (HALO). By confining all data to a balloon until mission termination, total data loss is 

risked in the event of undesirable landings. Conventional data transmissions are not feasible for large data rate 

missions due to low-data rate transmission capabilities. AIRS resolves these issues by deploying low-cost gliding 

aircraft, containing solid state drives with mission data, from the balloon. The AIRS vehicles autonomously guide 

themselves to within 300 ft. of designated GPS coordinates, automatically compensating for winds during descent. 

Initially velocity and lift predictions were made to determine the flight characteristics of different designs. The AIRS 

design was chosen and constructed from glass-fiber materials. A previously developed hobby autopilot was 

modified and used for the project needs. Low and high altitude, impact and velocity testing were performed with 

the vehicle. This project has showed it is feasible to build a safe, reliable and low cost data retrieval system that 

can be easily located. 

Accelerating Neuronal Model Simulations Using GPGPU Techniques 

Ben Razon 

Mentors: Alon Korngreen, Thomas Miller, and Roy Ben-Shalom 

Neuronal models are a critical tool that empower scientists to translate experimental results into knowledge of the 

underlying biophysical mechanisms. A common method involves optimizing kinetic parameters of the ion channels 

to match currents found in voltage-clamp protocol experiments. Both the genetic algorithm and model used in this 

technique are highly compute intensive, with execution times typically lasting several days even on large CPU 

clusters. By parallelizing these processes on graphics cards using the CUDA programming language, large 

performance gains can be realized. To fully accelerate these solutions a variety of optimization techniques such as 

utilizing shared memory, coalescing global memory accesses, and reducing memory transfers were implemented. 

Furthermore, the affordability of modern GPUs makes these solutions widely accessible, bringing super-computing 

capabilities to a new audience. 

Learning Representations of Astronomical Data With Missing Observations and Uneven Sampling 

Colorado Reed 

Mentors: Umaa Rebbapragada, David Thompson, and Kiri Wagstaff 

Radio telescopes are progressively observing candidate astronomical events at rates that are too high for human 

experts to visually inspect and analyze. As a result, automated statistical techniques for event detection and 

analysis are becoming commonplace for modern observatories. These methods often require the temporal data 

(i.e. light curves) to be robust and evenly sampled, while actual observational data usually contains missing or 

unevenly sampled data. Typical solutions include interpolating missing data (usually via linear interpolation) and 

marginalization (removing sections of data that are not robust). Missing values can be physically meaningful in 

astronomical data; however, and simply ignoring or interpolating this data may discard valuable information. In 

this work we present a novel technique for using contextual information to reconstruct probability distributions of 

missing values. The benefits of this technique are twofold: (1) using contextual observational data to impute 

missing data can outperform standard interpolation techniques, (2) the inferred probability distributions of the 

missing data provide an estimate of the uncertainty in the data representation that can be used in further analysis 

of the data. We demonstrate the effectiveness and generality of our technique with time-series data from the 

Green Bank Interferometer and the Very Long Baseline Array. 

Calibration of Mechanical/Electronic Platforms for Deep UV Imaging of Microbes on Natural Samples 

David Reid 

Mentors: Rohit Bhartia, William Abbey, and Luther Beegle 

The objective of my project is to help prepare the deep UV team for two major field excursions where the goal is to 

detect microbes and their distribution in the natural environment. One field sample will be basalt cores from the 

deep biosphere. For this I am reassembling and calibrating the current ultraviolet fluorescence scanner to reduce 

the need to refocus and to prevent sample contamination. The other field trip is to map microbial distributions on 

rocks in the Mojave. For this I will be calibrating a stand-off ultraviolet fluorescent scanners mounted on motorized 

tip/tilt. The tip/tilt will enable scans over 0.5x0.5m. This requires calibrations that enable assessment of the 

77

distance to the rock and determining the resolution capability of the instrument as well as the reproducibility to 

retarget a spot. With calibration of these instruments in the, the technology that has been demonstrated in the lab 

to non-invasively detect microbes in-situ can be successfully tested in the field. 

Algebraic Hecke Characters Which Are Rational Over Q 

Stephanie Reyes 

Mentor: Matthias Flach 

The motivation for this project comes from papers by Schütt (CM newforms with rational coefficients), Murabayashi 

(Determination of simple CM abelian surfaces defined over Q) and Anderson (Cyclotomy and a covering of the 

Taniyama group). The purpose of this project is to determine for which in Murabayashi’s list of degree four abelian 

fields there exists weight one or two Jacobi sum Hecke characters, to possibly find new degree four abelian fields 

for which there exists weight two Jacobi sum Hecke characters and to check if certain easily constructed weight two 

Hecke characters of imaginary quadratic fields are Jacobi sum Hecke characters. 

Precision Measurement of Growth Rates and Attachment Kinetics in Ice Crystals 

Mark E. Rickerby 


The basal growth of ice crystals between -2 o C and -40 o C was examined by using a purposely built convection 

chamber equipped with optical instruments. The obtained data was found to be in agreement with the model of 

nucleation limited growth. From this, the critical supersaturation σ0 was determined for a large number of crystals. 

This showed a very small nucleation barrier for the basal facet in the region -6 o C to -10 o C. An explanation of 

these findings by the effects of ice surface melting at temperatures below freezing point is presented. These results 

are consistent with the snow morphology diagram observed in nature that has remained unaccounted for since it 

was first proposed. Ideas on how to further develop the apparatus in order to also measure the radial growth are 

suggested. 

Paleoclimate Insights From the U-Th Isotope Geochemistry of the Buda Cave System 

Alexander D. Rider 

Mentors: Jess Adkins and Adam Subas 

The chemical and physical information preserved within speleothems (stalagmites and stalactites) can be an 

important tool for understanding past climates. However, this information is only useful if there is an understanding 

of the relationship between climate, speleothems, and the cave systems in which speleothems grow. Our goal is to 

incorporate the isotopes of the uranium decay series, often used for dating speleothems, into this understanding 

for the Buda Cave System. Using Caltech’s “Neptune” MC-ICP-MS we have made highly precise uranium series 

measurements of karst collected from the surface of the cave system, and drip waters collected from within. To 

complement these uranium series measurements we have also performed a broad-spectrum elemental analysis of 

the same samples using an Agilent 7500 series ICP-MS. This approach allows us to understand how the chemistry 

of drip water-rock interaction relates to the isotopes of the uranium series. We can then use our understanding of 

the uranium series geochemistry of the cave system to shed light on the record preserved within previously 

collected and analyzed stalagmite samples. 

Hot Flares From Cool Stars 

Joanna Robaszewski 

Mentors: George Djorgovski and Andrew Drake 

The Catalina Real-Time Transient Survey is a synoptic sky survey that observes variable objects across the entire 

sky and publishes information on them directly after observation. Many of these objects are flare stars, young stars 

that produce larges stellar flares. These flares can increase a star’s brightness by an order of magnitude and often 

only last for a few minutes. By examining the light cures of known flare stars, criteria to classify a light curve as 

belonging to a flare star or not can be determined. Once these parameters have been decided, they can then be 

applied to other transient objects that the Catalina Real-Time Transient Survey has discovered. After the objects 

have been classified, the frequency with which flare stars appear in our galaxy can be measured. Examining 

common characteristics in flare star light curves and looking correlations with spectral type can help us learn about 

the nature of stellar flares, as well as help us classify flare stars for future synoptic sky surveys. 

Error Quantification in Simulations of Variable Density Low Mach Number Turbulent Flows 

Nicholas Robertson 


In numerical simulations of low Mach number flows, the density is often expressed as a function of one or several 

transported scalar quantities through the use of an equation of state (EOS). As a result, these scalars play a critical 

role in controlling the accuracy of the overall simulation. This study aims at understanding and quantifying the 

sources of errors introduced in performing simulations of variable density low Mach number laminar and turbulent 

78

flows. As a first step, the relative importance of the order of accuracy of the scalar transport scheme and the 

velocity scheme is analyzed in canonical variable density test cases. Then, various implementations of the EOS are 

investigated in laminar flows. Particular importance is placed on the robustness, the convergence, and the accuracy 

of these implementations. Finally, the different scalar transport schemes and implementations of the EOS are 

combined and evaluated in simulations of turbulent planar jets. Following this analysis, guidelines for performing 

accurate and robust simulations of variable density low Mach number laminar/turbulent flows are proposed. 

Using Uncertainty Quantification to Maximize the Efficiency of Parallel Resources 

Andrew C. Rodriguez 

Mentor: Mike McKerns 

When dealing with any batch computation across parallel resources, a common problem is encountered: the 

computation time for the batch is extended by calculations that take longer than the rest. More so, we find that 

with parallel computation for the process to continue, we must allow for the final computations to finish. Though 

each calculation within the batch may not take much time individually, compounded, we find that, overall, a lot of 

time is taken waiting for the final few computations. We can reduce overall computation time through use of 

completion criteria. By deeming a batch “complete” when certain criteria are met, we find that we both make a 

better use of resources and save time. And so, we are creating a “Completion” data structure that shall manage 

such features for use within mapping functions. Criteria shall include, but not be limited to, a simple “percent 

completion” condition and use of statistical methods to terminate based on the confidence of the collected results. 

This would increase efficiency and save overall computation time. This is important since even a simple 10% time 

reduction allows for us to save 2.4 hours on a computation that would normally take 24 hours. 

Creating a Peptide Capture Agent Against IgG Fc 

Errika C. Romero 

Mentors: Jim Heath and Jessica Pfeilsticker 

Inexpensive and robust HIV tests are needed for effective point-of-care diagnostics in many developing countries. 

Current HIV diagnostic assays employ monoclonal antibodies, which are expensive and inconsistent. Instead, 

peptides, which are low-cost and whose chemical properties are easily tuned, can be designed to bind with 

comparable affinity and selectivity to anti-HIV antibodies similar to monoclonal antibodies that were raised against 

the same target. Previous research has identified a 13 residue cyclic peptide, monikered Wells’ peptide, which 

binds to the IgG Fc region of HIV antibodies. There are two binding pockets on every Fc molecule, so we created a 

dimer of Wells’ peptide using standard amide coupling to interact with both pockets simultaneously. Affinity and 

selectivity assays are performed to determine the ideal dimer linker length and to measure the dimer’s efficacy in 

binding to IgG Fc. 

Investigating the Expression Patterns of Potential Regulators and Homologues of FoxD3 

Shourya Sonkar Roy Burman 

Mentors: Marianne Bronner and Marcos Simões-Costa 

Neural crest cells are a population of transient migratory cells unique to the vertebrate embryos, which give rise to 

a wide variety of derivatives like the peripheral nervous system and the facial skeleton. They serve as a model to 

study multipotency, epithelial to mesenchymal transition, cell migration, embryonic induction and cell-fate 

determination. Of specific interest to this project is a transcriptional repressor in the gene regulatory network 

governing these processes, FoxD3. FoxD3 is expressed early in the presumptive neural crest region, and changes 

the adhesion properties of these cells leading to their epithelial to mesenchymal transition and migration. The first 

part of this project involves identification of potential regulators of FoxD3 in the chicken embryo using a candidate 

gene approach. A set of transcription factors was identified based on binding site studies on an enhancer element 

of FoxD3. The expression patterns of these genes were compared to the region where the enhancer is activated, 

and a refined list of candidates was prepared. In the second part of the project, differences in the formation of 

neural crest between an early vertebrate, the lamprey and other vertebrates was explored by studying the 

expression pattern of FoxD3 and a related homologue, FoxD1. 

Keck Spectroscopy of X-Ray Sources in the Ssa22 Field 

Iva Rreza 

Mentors: Fiona Harrison and Dan Stern 

We present the results of a deep optical spectroscopic survey of X-ray sources in the SSA22 field. After the Great 

Observatories Origins Deep Survey (GOODS), SSA22 is the next deepest extragalactic field observed by the 

Chandra X-ray Observatory, with a total integration time of XXX~ks. We observed six slitmasks with the Keck 

telescopes, obtaining 90 new high quality redshifts. We report on the properties of these sources, including their 

redshift distribution, absorbing column distribution, and the range of optical properties for the X-ray host galaxies. 

We compare our results to the published literature, and discuss several interesting sources in more detail. 

79

Sunk Costs and Reference-Dependent Preferences 

Alexander Manek Karl Runkel 

Mentors: Shinsuke Shimojo and Vikram Chib 

Human decision-making is influenced by expectations and aversions of loss, ambiguity and risk. A reference point 

represents a person’s rational expectations about past events, determined in an endogenous equilibrium. In a 

deterministic setting humans tend to maximize the utility derived from their actions – but in case of uncertainty, 

decisions are influenced by aversions as proposed by the Prospect Theory. In order to understand how these 

expectations are induced and how they influence human behavior, we performed several fMRI experiments to 

determine, firstly, how real effort is encoded as value; secondly, how this value is transcribed into a point of 

reference and thirdly how an established reference point shifts if uncertainties or risks are introduced. We trained 

subjects with visual cues associated with different levels of effort and a separate visual stimulus upon successful 

completion of the specific effort task. We found that subjects prefer the low-effort cues but paradoxically are more 

partial to the success-stimuli associated with the higher effort and a lower average payoff yield – thus indicating 

that humans are emotionally invested in incurred costs that cannot be recovered. In economic terms we can state 

that humans are attached to sunk costs and that the classical consumption utility function has to be refined by 

introducing a model of gain-loss-utility. 

C. elegans Lethargus and Molting Behavior 

Elizabeth Ryan 

Mentors: Paul Sternberg and Julie Cho 

Lethargus is a sleep-like condition exhibited by the nematode C. elegans between successive developmental 

stages. During lethargus, the nematode enters a state of quiescence in which locomotive activity and feeding 

behavior decrease. Subsequently, the nematode molts its exterior cuticle, signaling the end of lethargus. To 

understand the relationship between availability of food and its effect on behavioral states such as lethargus and 

molting, we placed worms in their fourth larval stage, known as L4, inside a microfluidic tracking chip with and 

without OP50 E. coli available for sustenance. Although the time taken for C. elegans nematodes to shed their 

cuticle in the liquid environment of a microfluidic chip is comparable to the time taken on an agar plate, results 

indicate that molting lasts longer when the worms are starving. Additionally, we want to use this sleep-like 

behavior to test the effects of sleep-altering drugs. Therefore, we are exploring the effect on lethargus when L4 

worms were exposed to the stimulant dipropylcyclopentylxanthine (DPCPX). Using tracking software, worms placed 

in a rectangular field of view are monitored during their transition into lethargus. DPCPX works as an antagonist for 

the adenosine A1 receptor and has been shown to cause wakefulness in Danio rerio (zebrafish). However, 

because C. elegans pores are more resistant to chemical penetration, dimethyl sulfoxide and Triton X-100 are used 

to carry compounds across the nematode membrane, rendering it more permeable. Findings may provide a 

connection between C. elegans sleep-like state and that of other species in the Animalia kingdom due to similarities 

in biological systems. 

Carbon Nanotube Arrays for Drug Delivery and Transport Studies 

Stephanie Rae P. Samson 

Mentors: Morteza Gharib and Adrianus Aria 

Carbon nanotubes, rolled seamless graphene sheets, can be functionalized non-covalently or covalently with 

different chemical groups to modify or add to their already unique intrinsic properties. Pristine carbon nanotubes 

are generally unreactive, but by taking advantage of existing defects or by introducing defects without destroying 

the nanotubes’ structures via oxidation with UV/Ozone or Oxygen/Plasma treatments, it is possible to add carboxyl, 

carbonyl, and hydroxyl groups on the carbon nanotube surface. The resulting oxygen-containing groups can be 

utilized to covalently add functionality to the carbon nanotubes. The ultimate goal is to create an ester linkage 

between vertically aligned carbon nanotube arrays (nanocarpets) and large molecules, such as proteins, drugs, and 

nanoparticles. To test the reactivity of oxidized nanocarpets, they are subjected to esterifications with nitrogencontaining 

compounds using Fischer esterification, Steglich esterification, and esterification after conversion of 

carboxyl groups to acid chlorides. The reacting nanocarpets will then be analyzed using Energy-Dispersive X-ray 

Spectroscopy (EDS), Raman Spectroscopy, and Fourier Transform Infrared Spectroscopy (FTIR) to determine the 

resulting nitrogen content and the integrity of the carbon nanotubes. If those results show promise, then those 

esterifications will be repeated using larger molecules. 

Producing a Lab-on-a-Chip Microfluidics Device for Speedy Cell Diagnostics 

Troy Sandberg 

Mentor: James Heath 

The main advantage a microfluidics device has over its larger counterparts is the small volume of materials it 

requires and the speed at which it is able to analyze these small samples. A hospital patient could supply a single 

drop of blood from which the lab-on-a-chip could determine the various cells therein, checking for metabolic signs 

indicative of metastasizing cancer cells. This project focuses on fine-tuning the production of such a chip. Once a 

particular design is decided on and laser-etched into a silicon wafer, a silicone polymer is poured over the wafer 

80

and baked so that it hardens in the desired shape and surface geometry of the chip. This chip has multiple 

separated chambers which can be coated with protein substrates to which cells will adhere. Collaborating with a lab 

at UCLA, many different coatings and cell lines have been paired together on the chip to find the combination that 

provides optimal cell adherence. With the cells adhered to the chip, medium containing radiolabeled probes can be 

flowed through the chip’s chambers and their uptake by the cells quantified. 

An in vitro Study of the Hemodynamics of Valvulogenesis 

Laura F. Santoso 

Mentors: Morteza Gharib and Derek Rinderknecht 

Heart disease is a significant problem worldwide, and people often undergo valve replacement surgery to treat 

their condition. Valvulogenesis is being studied to improve upon existing replacement valves, which are currently 

produced from synthetic materials or animal tissue, have limited lifetimes, and require additional medication. 

Ample research has shown that shear forces from blood flow affect how cells grow. The shear stress forces caused 

by various flow rates have been measured in this research by particle image velocimetry. Regulation of shear 

forces allowed emulation of the flow conditions present during valvulogenesis, under which endothelial cell 

proliferation was observed. Additional studies from previous research have provided strong evidence that extra 

expression of Krueppel-like factor 2 (KLF2) is indicative of valvulogenesis. Studying expression of KLF2 ensured 

that the prolonged fluid shear stress applied was recognized by the observed cells. Subsequent studies analyzing 

proliferation characteristics under pulsatile flows would further this research. 

Design and Modeling of an Autoinhibiting Biomolecular Circuit Incorporating Scaffolding 

Keshav Sapatnekar 


In nature, protein scaffolds allow processes to proceed more efficiently by localizing signaling proteins and their 

downstream elements, thus decreasing the time for signals to propagate. This could be used to create protein-level 

feedback systems utilizing scaffolds. In this project, a circuit was designed incorporating a histidine kinase, a 

response regulator and two scaffold elements bound to target proteins. One scaffold binds the kinase to the 

regulator, thereby activating production of the second scaffold. This second protein would bind the first, preventing 

its further production, ideally giving an equal concentration of the target proteins at steady state. An ordinary 

differential equation model was used to describe the behavior of the system, incorporating transcription, 

translation, and binding of proteins and scaffolds. Future investigation would include implementation of the system 

in E. coli using fluorescent target proteins and a plate reader to determine concentrations. 

3D Scientific Data Visualization in an Immersive Virtual Reality 

Franz Sauer 

Mentor: George Djorgovski 

As society moves into an age where scientific data becomes more complex, humans need more powerful 

visualization tools that would not only need to revolutionize the way we perceive data, but also provide effective 

ways for scientists to collaborate with one another. The Meta-Institute for Computational Astrophysics (MICA) was 

therefore created to explore, develop, and promote virtual worlds as a means of professional research. This specific 

project focused on developing and testing new data visualization tools in immersive virtual reality environments in 

order to gain insight into what methods work better than others in terms of understanding and collaboration. 

Currently set up in ScienceSim, a virtual world run by Intel, this project is developing a number of tools such as a 

multi-dimensional plotting device and viewing area, an in world method to format and scale data files, a 

collaborative tool that links data points to more information located on web databases, and an in world method to 

display online images to a group of researchers. As this is still and ongoing project many more tools will be 

developed with the eventual goal of developing an efficient virtual world as a means of scientific research 

collaboration and visualization. 

New Design for a Vertical Wind Turbine 

Brad Saund 

Mentor: John Dabiri 

Recent analysis and experiments demonstrate that carefully placed vertical axis wind turbines (VAWTs) have the 

potential for an order of magnitude increase in the power produced per land area compared to horizontal axis wind 

turbines. By placing a VAWT adjacent to a counter-rotating VAWT the efficiency of each turbine can be improved to 

above the efficiency of an isolated turbine. Currently commercially available vertical wind turbines are used at the 

field site for these experiments. Interest in designing and constructing a new wind turbine for this site stems from 

the desire for greater control over the experimental setup, increasing the beneficial effects of the turbine 

placement, and lowering the cost the turbines. A proposed design was analyzed, models of the design and key 

components were tested, and the turbine was constructed. 

81

Spectral Gap Scaling of One Dimensional Quantum Spin Chains 

Travis Scholten 

Mentors: John Preskill and Spiros Michalakis 

Frustration-free Hamiltonians which are the sum of two-qubit projection operators have been considered as models 

for adiabatic quantum computation. The run time of such computation is determined by the scaling of the spectral 

gap with system size; in general, this property is not well understood. We show that the number of free 

parameters in any such Hamiltonian may be reduced to three real parameters and consider the decay of the gap 

with respect to the number of qubits. We present numerical results from a DMRG algorithm. For certain values of 

the real parameters, the scaling is shown to be an inverse polynomial of the system size. 

Boy Meets Boy: A Study of Male-Male Aggression in Drosophila melanogaster 

Jonathan Schor 

Mentors: David Anderson and Liming Wang 

In Drosophila melanogaster, the fruit fly, the nonvolatile cuticular hydrocarbon (CH) (Z)-7-tricosene (7-T) regulates 

aggressive behavior in an independent and inverse fashion to courtship. Given that 7-T is a gustatory pheromone, 

it seems likely that its primary mode of transmission is through physical contact with another fly’s gustatory 

sensilla, though the kinetics of 7-T are yet unknown. Through use of a gal4/gal80-UAS system, pairs of oenocyteminus 

(oe-) flies lacking in all CHs, including 7-T, were observed in parallel with wild-type Canton S (CS) flies for 

their aggressive behavior. Films of the interactions were annotated by hand, though in the future annotation will be 

performed through use of a machine-learning algorithm currently under development. Preliminary results indicate 

that physical contact precedes aggressive behavior, before which male-male courtship is observed. Future 

experimentation will examine why CS flies raised in groups (group-housed flies) are less aggressive than those that 

are single-housed, with a focus on the frequency at which they touch and their sensitivity to 7-T. 

Improving Temperature Stability of Himar1 Transposase 

Stanford Schor 

Mentor: Frances H. Arnold 

Thermophilic microbes and the heat stable enzymes they produce have many potential applications in 

biotechnology. However, the lack of available genetic tools for thermophiles has hindered their widespread use in 

industry. High-density mutagenesis is an efficient method for probing the biology of such poorly characterized 

organisms. Himar1 is a mariner-family transposase isolated from the horn fly Haematobia irritans. It inserts a short 

inverted terminal repeat (ITR) between thymine and adenine nucleotides randomly throughout the genome of its 

host and requires no external factors, making it a versatile tool for disrupting gene function. Wild-type Himar1 does 

not function at temperatures above 40�C, and we propose to use directed evolution to shift the optimal 

temperature range of Himar1 and generate a set of mutants for use in thermophilic organisms. We will randomly 

introduce point mutations into Himar1 using error-prone PCR, and then test mutants for activity at higher 

temperatures using a novel in vitro transposition assay. Stabilized, functional mutants will be subjected to iterative 

rounds of error-prone PCR and activity testing at higher temperatures in order to produce a family of mutants with 

functionality well above that of the wild type enzyme. 

Separating Planck Bolometers and Beams via Simulated Planet Observations 

Kasey W. Schultz 

Mentors: Sunil Golwala and Brendan Crill 

For any Cosmic Microwave Background (CMB) observation, in order to extract the full cosmological information the 

telescope beam must be accurately modeled, and the data must be corrected for the beam’s filtering. The CMB 

satellite Planck scans the full sky by rotating on its spin axis at a constant rate. This causes a degeneracy between 

the detector time response and the beam, and necessitates detailed modeling to identify relationships between 

their respective parameters. This modeling allows their effects to be separated, and produces data that has been 

properly corrected. Utilizing and updating the Planck team’s in-place software pipeline, I have run planet 

observation simulations and detector and beam model reconstruction. Through identifying possible bias in the 

fitting routines and any degeneracy between parameters, my work provides results that will help separate these 

effects as much as possible. By analyzing best fit parameters of over 3000 simulations and fits, I have shown that 

the pipeline reconstructs the input parameters without bias, that proper planet motion introduces no bias, and 

hope to show that prior knowledge of the beam does not bias the fitting routines or parameters; an important 

result for analyzing real Planck data. 

82

Analysis of Energy Estimation From NOvA Detector Data 

James Scott 

Mentor: Ryan Patterson 

Determining the energy of a neutrino accurately is important for studying neutrino oscillations. The detector for 

NOvA, which is made of many cells of scintillator, is able to detect the outgoing particles from a neutrino 

interaction, which can then be used to reconstruct the event and thus the neutrino energy. Any reconstruction will 

have uncertainties introduced by limits in the resolution of the detector, as well as random variations and noise. 

Knowing the magnitude of the uncertainties is important not only for assigning uncertainty in measurements, but 

also in guiding detector calibration approaches and in determining the theoretical and practical limitations of the 

oscillation measurements. By analyzing simulated detector data the uncertainties can be determined for the 

various interactions. 

Fourier Continuation Method for Eigenvalue Problems on Arbitrary Two-Dimensional Domains 

Ayon Sen 

Mentors: Oscar Bruno and Timothy Elling 

The eigenvalues of the 2D Laplacian correspond to the fundamental modes of vibration of a membrane with given 

geometry. These eigenvalues can be determined by using an iterative eigenvalue solver on the numerical 

differential operator. To evaluate the operator, we propose a method for calculating numerical derivatives on 

arbitrary domains by decomposing the geometry into overlapping patches that can be mapped to rectangles and 

then applying Fourier series techniques to compute derivatives under the change of variables. The Fourier 

continuation method extends non-periodic data so that the periodic extension is smooth, meaning Fourier series 

can be used to accurately represent the extension. Thus, derivatives can be computed accurately and quickly (i.e. 

using fast Fourier transforms). Extending this methodology to differential operators with variable coefficients 

(corresponding to inhomogeneous membranes) is simple, and test cases suggest that eigenvalues can be 

computed accurately for various geometries. 

Fabrication of Silicon-Based Nanopore Membranes for DNA Nucleotide Characterization 

Bhargav Setlur 

Mentors: Axel Scherer and Aditya Rajagopal 

In this ongoing project, we focus on the design and fabrication of a silicon-based nanopore for use in single 

stranded DNA sequencing. We start by defining the nanopore geometry using a multiphysics modeling program, 

which automatically calculates various electrical and capacitive parameters relating to the structure. Then, based 

on the idealized design, we demonstrate fabrication of the silicon nanopore using nanofabrication equipment 

located in the Kavli Nanoscience Institute in the sub-basement of Steele Building. Fabrication processes involve 

boring nanoscale holes in silicon using a transmission electron microscope to bore holes in silicon, as well as an 

SEM to beam-write patterns. Finally we fit the nanopore structures with electrical contacts and integrate them into 

the structure. 

Functionalization and Biological Application of Carbon Nanotubes 

Juying Shang 

Mentor: Mory Gharib 

Carbon nanotubes (CNTs) are allotropes of carbon with a cylindrical nanostructure. Besides exhibiting extraordinary 

strength and unique electrical and thermal properties, functionalized CNTs are useful for a multitude of biological 

and chemical applications. For example, functionalized CNTs can be fashioned into new scaffolds for cells, 

supporting three-dimensional tissue formation. In this project, CNTs are grown using the chemical vapor deposition 

process and functionalized using the UV/O3 treatment. SEM/EDS and FTIR analysis were conducted before and 

after treatment to determine the effects of UV/O3 exposure. It was found that the UV/O3 treatment temporarily 

adds carboxyl groups to the surface of CNTs; the effects disappear after 4 days. In order to understand cell 

behavior on CNTs, proliferation and differentiation assays will be conducted with human mesenchymal stem cells 

(hMSCs) grown on CNTs. In addition, CNTs’ ability to absorb proteins will be determined. The same experiments 

will be performed using functionalized CNTs. From the combined results of cell study and functionalization 

characterization, cell adhesion to and affinity for CNTs can be evaluated, illustrating the possibilities of biological 

applications of carbon nanotubes. 

Classifying Things That Go *BANG!* in the Night 

Nihar Sharma 

Mentors: S.G. Djorgovski and Ciro Donalek 

Digital synoptic sky surveys promise to shed a whole new light on existing as well as previously unknown timevariable 

astronomical phenomena. The challenge is to automate the classification and prioritization of transients for 

follow-up observations within the data that we receive in such a way that minimizes the need to use expensive 

spectroscopic resources. Due to an inevitably large amount of missing input data, a Bayesian Network (BN) was 

83

initially used to generate probabilistically inferred results that placed transients into different classes by computing 

the distribution of a given subset of variables where other variables were observed. Effort is being directed to 

incorporate continuous nodes into the existing BN, so that the need for artificial binning of continuous data can be 

eliminated, avoiding loss of information that is currently inherent in the classifier. An ensemble of learners was 

then constructed by prototyping Decision Trees as well as Multi-Layer Perceptrons and integrating it into the overall 

classification framework. This framework was rigorously tested on parameters extracted from time-series data of 

light curves so that it could potentially be employed in the real-time variable classification pipeline. 

Branching Messaging for Anonymous Communication 

Isaac C. Sheff 

Mentor: Tracey Ho 

We consider anonymous communication between pairs of nodes in the presence of an adversary who can observe 

all network traffic. Existing schemes involve partially trusted central servers, a large amount of cover traffic, or 

high latency. We propose a scheme which improves upon the trade-off, in which messages require only logarithmic 

time to deliver, as well as computational time to send. In this scheme, any node receiving a message applies a 

private decryption key to discover content, or instructions to forward the message to zero or more other nodes. 

Furthermore, when a node applies its decryption key to a message not encrypted with that node's encryption key, 

the result is indistinguishable from a message with forwarding instructions. A sender wraps a message in an onion 

route of logarithmic length, branching (forwarding to multiple nodes) at random points, resulting in a tree of 

forwarded messages. Nodes wait to initiate messages of their own until they have received a message, so it is 

indistinguishable whether they are sending or forwarding. Through analysis and simulation, we show this system 

preserves a high degree of sender and receiver anonymity, as well as unlinkability between communicating pairs. 

Bioremediation of Endocrine Disrupting Compounds Using the Synthetic Biology BioBrick Standard 

Amanda Shelton 

Mentors: Richard Murray, Nate Glasser, Toni Lee, Joe Meyerowitz, and Emzo de los Santos 

Endocrine disruptors are organic pollutants that cause reproductive harm to fish, birds and possibly humans. The 

2011 Caltech iGEM Team aims to use synthetic biology to create a biosensing and bioremediation system in 

Escherichia coli for any of four compounds: bisphenol A (BPA), dichlorodiphenyltrichloroethane (DDT), nonylphenol 

and 17α-ethynylestradiol. To detect the presence of these chemicals we characterize the human estrogen receptor 

in E. coli and then modify the protein for reduced toxicity to the chassis and to improve its ligand-binding specificity 

to some of the endocrine disruptors. We also have identified genes that degrade these compounds and attempt to 

characterize their function in E. coli. Using both Gibson and standard assembly to create plasmids, we put the 

modified estrogen receptor and the degradation genes into the BioBrick Standard, a way of formatting genes, 

promoters, plasmids and other parts for easy retrieval by iGEM teams and others to use in their own synthetic 

biology projects. 

A Catalytic Enantioselective Approach to the Synthesis of Physovenine 

Jeff Shen 

Mentors: Brian M. Stoltz and Boram D. Hong 

A convenient method for the construction of enantioenriched oxindoles bearing all carbon quarternary 

stereocenters is employed toward the synthesis of (�)-physovenine, a natural product containing a furoindoline 

moiety. The strategy involves a single step copper-catalyzed enantioselective stereoablative alkylation of the 

racemic halooxindole, synthesized from the commercially available p-anisidine, with up to 87% enantiomeric 

excess. Subsequent decarboxylation, N-methylation and reductive cyclization of the oxindole allow access to the 

desired furoindoline structure. The successful application of the enantioselective alkylation allows for future 

asymmetric syntheses of other biologically significant alkaloids such as physostigmine. 

3-D Visualization and Modeling of Building Shaking During Earthquakes 

Hong Sheng 

Mentor: Monica Kohler 

The goal of this project is to find a new way to visualize and model the shaking of buildings during earthquakes. 

Visualization provides a rapid portrayal of the extent of potentially damaging shaking during an earthquake and can 

be used for emergency response, loss estimation, and for public information through the media. For this project I 

used Millikan Library as our prototype building. I integrated the accelerations recorded by a 36-channel seismic 

network inside Millikan Library during four earthquakes (Mar 16, 2010, ML=4.4 Pico Rivera; Aug. 9, 2007, ML=4.6 

Chatsworth; Apr. 4, 2010 Mw=7.2 Baja California and July 29, 2008, ML=5.5 Chino Hills) and applied filters and 

numerical integration to get the displacements of each floor during the earthquakes. I then made movies from the 

images that resulted from applying the displacements to the initial building model. I also plotted the peak 

accelerations and velocities on each floor using a color scale that indicates the extent of shaking during the 

earthquake. The peak accelerations we compute from the data are an indirect measure of potential damage. 

84

Construction of a Polarization Sensitive Optical Coherence Tomography System 

Jeffrey D. Sherman 

Mentors: Scott Fraser and Reza Motaghian 

Neural degenerative diseases result in tissue damage and cause discomfort, disease, and death. Preemptively 

diagnosing these degenerative diseases and visualizing them will help in diagnostics as well as the search for cures 

and treatments. Recent advances in optical techniques have led to Optical Coherence Tomography (OCT), a 

noninvasive imaging technique without the use of ionizing radiation, particularly in the eye. The birefringence of 

tissue can act as a surrogate biomarker, allowing for an analysis of tissue structure and health. Polarization 

Sensitive OCT (PS-OCT) has emerged as an extension of OCT for capturing birefringence, particularly in the eye. 

Using PS-OCT, it is possible to detect the initiation of neural degeneration. Since the method is noninvasive, a 

longitudal analysis can be performed to monitor the progression of these degenerations and adjust therapies to 

better help the patient. A PS-OCT system was constructed in free space on optical breadboards for modularity. 

To verify the effectiveness of the system, eye scans taken by a Swept Source OCT system at 1060nm will be 

compared with scans that will be taken by the newly constructed PS-OCT system at 1060nm. If the PS-OCT system 

reveals tissue organization better than standard OCT, then the construction of the system will be considered a 

success. 

Analysis of Soot Evolution Using a Combined Fluid Mechanics and Chemical Model 

Kevin Shi 


The importance of hydrocarbon fuels necessitates the study of their emissions, and in particular their sooting 

tendencies. This work relies on two programs, FlameMaster and NGA, to simulate the overall combustion process. 

The first uses a detailed chemical mechanism with 185 species and 1903 reactions to generate counterflow 

diffusion flames using a flamelet approach. These data are tabulated and integrated into a full 2D fluid dynamics 

solver to simulate an axisymmetric coflow diffusion flame. Special emphasis is placed on evaluating the relative 

importance of radiation and incorporating a radiation model to account for heat losses. A soot model, built on top 

of the chemistry and fluid dynamics, calculates the soot production rate. Sooting tendency is studied for aliphatic 

and aromatic dopant species, including polycyclic aromatic hydrocarbons. These results are validated with existing 

experimental and computational studies. 

Electron Transfer Studies of the Double Tryptophan Azurin Mutant H107W108W110 (All Phe) 

Dong Woo Shin 

Mentors: Harry Gray and Heather Williamson 

Simplified protein structure can be studied to help understand the complex electron transfer system present in 

nature. In this study, Azurin mutant with hydrophobic and hydrophilic tryptophan at sites 108 and 110 was used to 

look for communication and electron transfer in the system. Circular dichroism and steady-state fluorescence was 

used to probe the environment and the communication line of the two tryptophans. Luminescence and transient 

absorption of the protein was studied with the nanosecond lasers to determine the nature of the pathway of 

electron transfer system. No clear communication between the tryptophans has been isolated. The laser studies did 

indicate weak interaction between the rhenium photo sensitizer and the tryptophans, but no communication 

between the tryptophans and the native copper. 

A 2.4 GHz Printed Circuit Board Proof-of-Concept of a High Efficiency mm-Wave CMOS Active Radiator 

Gunnar Atli Sigurdsson 

Mentors: Ali Hajimiri and Steven Bowers 

The immense popularity of wireless communications has left the common frequency bands crowded, prompting 

researchers to utilize available spectrum at ever higher frequencies. At mm-wave frequencies there is pronounced 

need for novel antenna designs that are tightly integrated with their driving circuitry in order to reduce power 

losses. A radiator concept for 94 GHz CMOS-technology was reviewed, scaled up, and redesigned to work at 2.4 

GHz on a FR-4 printed circuit board, in the interest of testing the concept. The radiator works in similar manner to 

an array of dipoles, and can connect directly to the last amplifier stage without impedance matching, due to loadpull 

matched input impedances, accomplishing all of its power combining in the air. 3D full-wave electromagnetic 

field simulations were performed on all transmission line structures and furthermore, various ways to achieve 

symmetric power splitting and shorted transmission line stubs with coupled lines were designed and experimented 

with, in order to achieve acceptable efficiency and radiation pattern of the radiating array. Simulated results predict 

the radiating section having 99.5% efficiency, 4.62 dBi maximum antenna gain and the power amplifier section 

delivering 2.28 W of power to the radiating section, making the whole structure 76% efficient with 22.6 dB power 

gain. 

85

Cosmic Sparklers: A Systematic Search for Cataclysmic Variables Using the Palomar Transient Factory 

Gregory Simonian 

Mentors: Thomas Prince and David Levitan 

I identified potential Catalysmic Variable (CV) sources in the Palomar Transient Factory (PTF) photometric 

database. Accomplishing this involved utilizing basic cuts based on magnitude and duration of outbursts to 

eliminate spurious sources, and cross-matching remaining candidates using SDSS and SIMBAD catalogs. This 

method yielded 44 new potential CVs, and out of these, four were followed up with classification spectra. In order 

to study one of these systems in detail, we undertook a case-study of a known special CV: CR Boo, an AM CVn 

system. We expect to confirm the connection between quiescent variability and the orbital period of such AM CVn 

systems which would be very useful for determining properties of these systems with photometric surveys of the 

sky. 

Reduction of Thermal Conductivity in Graphene 

Prastuti Singh 

Mentors: Jim Heath and Slobodan Mitrovic 

One of the greatest challenges in thermoelectric is increasing the effectiveness of heat-to-electricity conversions. 

One solution is to fabricate materials whose thermal conductivity, ��can be reduced while their electrical 

conductivity, �� is kept intact. Recent research suggests the answer may lie with nanostructures. High surface-tovolume 

ratio of nanostructures allows for greater scattering of phonons (heat carrying particles). Further research, 

led by Prof. Jim Heath’s lab, led to the development of a phononic silicon nanomesh. The nanomesh is patterned 

with spacings shorter than the mean free path of phonons, results in greater scattering and therefore, a lower 

��The purpose of this project was to examine this process using graphene, a material known to possess extremely 

high values of ��and ��Graphene was exfoliated onto SiO2 wafers, and then its location mapped using the SEM. 

Four electrodes were written on the graphene, and then a 200 nm mesh was written and etched upon the graphene 

using E-Beam Lithography. This process allows for the electrical conductivity of the graphene device to be 

measured. Further research is needed to develop a process for measuring the thermal conductivity of the graphene 

mesh. 

The Regulatory Role of wnt16 in Strongylocentrotus purpuratus Endoderm Specification 

Natnaree Siriwon 

Mentors: Eric H. Davidson and Isabelle Peter 

Cell fate is determined by the gene regulatory network. The endoderm precursor cells in sea urchin embryos derive 

from two cell lineages: veg1 and veg2. Endoderm specification in both cell lineages requires the expression of 

hox11/13b. The expression of Hox11/13b in veg2 derived cells was predicted to be required for the activation of 

itself in veg1 derived cells, according to a Hox11/13b perturbation experiment. Based on that observation, it was 

proposed that a wnt signaling ligand expressed downstream of hox11/13b in veg2 derived cells might be involved. 

QPCR experiment showed that the expression of wnt16 decreased when Hox11/13b was perturbed. We showed by 

whole mount in situ hybridization that wnt16 is expressed in veg2 derived cells at 24h. Moreover, when the 

expression of wnt16 was analyzed in 24h Hox11/13b perturbed embryos, a decrease of expression in veg2 derived 

cells was observed. These results show that hox11/13b activates wnt16 in veg2 derived cells. When the expression 

of wnt16 was perturbed, the expression of bra, blimp and foxA was lowered at 18h and 24h post-fertilization. The 

expression of eve was lowered only at 18h. However, the expression of hox11/13b increased at 18h, 21h and 24h. 

A QPCR analysis of 15h, 28h, 21h, 27h and 30h Wnt16MASO treated embryos showed that the rise of hox11/13b 

and foxA expression delayed by 6 hours due to the perturbation of wnt16. This result shows that Wnt16 is 

responsible for the activation of hox11/13b in veg2 endoderm precursor cells. However, it is still inconclusive 

whether Wnt16 is the activator of hox11/13b in veg1 endoderm precursor cells. 

Risk Mitigated Optimal Power Flow With High Wind Penetration 

Emma Sjödin 

Mentors: Steven Low and Dennice Gayme 

Increased penetration of renewable energy sources into the power grid poses new challenges, mostly due to their 

intermittency. To ensure a safe and steady supply of power, a substantial amount of ancillary services will need to 

be introduced to the grid. We consider a case where intermittent wind generation is taken as the primary energy 

source and where energy storage and spinning reserves from conventional sources are provided. The problem is 

formulated as a risk mitigating optimal power flow (OPF) problem, where the objective is to schedule the spinning 

reserves to minimize the cost of ancillary services while compensating for uncertainties in generation and load. 

Using a modified version of the IEEE 14 bus benchmark system, the OPF is solved both as a finite and receding 

horizon control problem, where the optimal strategy of pre-allocation of spinning reserves is studied with both 

time-invariant and demand based cost functions. We also investigate the optimal distribution of storage capacity 

across different network topologies to minimize transmission losses and unnecessary charge/discharge cycles. The 

results quantify the need for ancillary services and suggest a strategy for their scheduling and placement. 

86

Automated Information Retrieval System (AIRS): An Economical Safeguard Against Data Loss for High 

Altitude Balloon Missions 

Russell Smith 

Mentors: Jeff Booth and Jason Rhodes 

Long duration balloon missions such as the High Altitude Lensing Observatory (HALO) acquire such high quantities 

of data that they exceed the downlink bandwidth which can be guaranteed. AIRS provides a backup data transfer 

mechanism so that flights need not be limited by fear of total data loss. An autonomous glider with GPS and 

autopilot carries a solid state hard drive to an approved location near the balloon’s position. For maximum safety, 

mass and final kinetic energy have been minimized. Accurate waypoint targeting both simplifies recovery and 

further mitigates risk. A low drag, low aspect ratio wing provides rapid penetration through high crosswinds, yet 

supports slow landings at a high angle of attack. The 800g laminated thin wall fiberglass and foam capsule includes 

a solid state drive rated to withstand 1500G, more than double the capsule's impact force in an unlikely control 

system failure. A radio signal is broadcast after landing to allow AIRS to be easily located with a directional 

receiver. AIRS has been designed, fabricated, fitted with an autopilot, and tested for functionality, accuracy and 

safety both in simulation and from a weather balloon drop, demonstrating the feasibility of safe and easy data 

retrieval at low cost. 

Construction and Testing of Low-Energy α- and β-Particle-Tracking Detector Prototype 

Daniel Sotolongo 

Mentors: Sunil Golwala and Robert Nelson 

The Cryogenic Dark Matter Search is searching for evidence of Weakly-Interacting Massive Particles as a candidate 

for dark matter. The dominant background for this type of experiment is low-energy electrons emitted by 

radioactive contaminants on detectors’ surfaces. Therefore, a detector capable of screening these crystal detectors 

for contamination at each step in their manufacturing process was needed. The proposed detector, dubbed the 

Betacage, is a drift chamber with energy resolution of a few keV. The Betacage is composed of several wire grids 

charged to ~3kV, a voltage that creates an electric field strong enough to cause an electron avalanche around a 

given wire from only a few ionized particles. These avalanches are read as pulses on the grids, which are 

reconstructed into a 3D path. The construction of the prototype, composed of only 1 grid and smaller than the final 

design, has revealed and resolved problems that would have been considerably more problematic at full scale. The 

prototype has confirmed the feasibility of the design, and work will soon begin on a larger-scale fully-functional 

prototype. 

Formation of Membrane Protein Nanocrystals Through Laminar Flow 

Kelsey M. Spaur 

Mentors: Michael Stowell, Rob Phillips, and Ray (Hsin-Jui) Wu 

Membrane proteins play a critical role in transport across membranes. The ability to easily determine membrane 

protein structures can greatly benefit the pharmaceutical world as well as enhancing our knowledge of the human 

body in general. The project focuses on using a microfluidic device to mix protein and lipid and then induce the 

diffusion of detergent through laminar flow to allow the membrane protein to form. To make this methodology 

operational, tests were performed to prove that the protein and lipid mix at the designated ratio and to determine 

the mixing ratios and flow rates that best formed membrane proteins. Results indicate that the device reliably 

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, 

protein/lipid combination to buffer respectively, have produced membrane proteins. Determining the diffusion rate 

of the detergent would be valuable supporting information about the process. Tests should be run with ratios near 

these values, 1 to 90 and 1 to 120 for example, to narrow in on the best ratios. Tests should also be performed 

with different protein/lipid mixing ratios. 

Neural Predictors of Bargaining Decision Making Using Hyperscanning EEG 

Stasja Stanisic 

Mentors: Shinsuke Shimojo and Kyongsik Yun 

Striking is the act of not participating in beneficial circumstances with the intention of achieving a long-term goal. 

In this project we wanted to induce a possible strike situation using a bargaining game. In each experiment a 

subject pair would go through 120 trials of the game where they could reach a no deal (no payoff for either) or a 

deal (payoff for both) situation. Their brain activity was recorded using electroencephalography (EEG), EGI 128 

channel device. Their finger movement, with which they would coordinate their negotiations, was tracked using a 

Vicon infra-red camera. EEG data was processed with independent component analysis and standardized low 

resolution brain electromagnetic tomography for source localization. Brain activity and finger movement data could 

allow discovery of neural process that leads to a strike which can be used for prediction of strikes moments before 

the thought is fully formed in the brain. Comparing the brain activity of the two subjects and between the pairs, we 

aim to discover what leads to a strike. 

87

Do Tutte Polynomials Satisfy the Kontsevich Conjecture? 

Jessica Su 


The Kontsevich conjecture holds for a polynomial if its zeros are polynomially countable, i.e., if the number of zeros 

of that polynomial has polynomial dependence on the size of the field. Our aim is to determine which graphs yield 

Tutte polynomials that satisfy the Kontsevich conjecture. With computer software and logical reasoning we have 

indicated that very few polynomials satisfy the conjecture directly, but some satisfy a modified version of the 

conjecture. In those cases, the dependence is polynomial except at points where a certain parameter (and thus, 

the entire polynomial) is 0. Further work should focus on identifying qualities of graphs that yield deeper failures. 

The Mechanism Behind I-SceI Injection Method for Strongylocentrotus purpuratus 

Julia Y. Su 

Mentors: Eric Davidson and Miao Cui 

Understanding gene regulatory networks is crucial to comprehending the relationship between genes and their 

effects on embryonic development and evolution. Current research on gene regulatory networks has been impeded 

by the mosaic pattern that emerges from injecting reporter constructs. Although mosaic pattern does not affect the 

data of young embryos, it does impact the data gathered from older animals. Research has observed that this new 

I-SceI injection method has decreased the mosaic pattern in embryos. There is little if anything known about the 

mechanisms behind this injection method. If the mechanism of how the injected DNA is integrated is understood, 

then the knowledge can be used to further decrease the mosaic patterns and help scientists gather better data in 

their analysis of gene regulatory networks. In this study, TAIL-PCR is used to recover the adjacent sequence 

flanking the injected DNA which directly answer the question of where it is integrated into genome. In addition, It 

also indicates the form of exogenous insertion by obtaining injected DNA itself as the adjacent sequence. 

Gesture Based Interface for Solar Decathlon House 

Ka-Chuan Suen 


Energy use in homes and residence accounts for nearly 25% of the total US energy output per year. While 

appliances such as lights and media devices are not the largest energy consumers in a house, they are easier to 

control and turn off when not in use. Using the recent advances in 3D cameras and pairing it with a home 

controller system, Control4, we provide a gesture based interface to control lights and appliances. The Xbox kinect 

with its 3D video stream is used to determine where a person is pointing and compares it with a virtual 3D map of 

appliances to find which appliance is selected. Commands are then sent using Control4. The location of the user is 

also used as a separate system for controlling devices. With greater ease in controlling household appliances that 

use energy, it is hoped that overall energy use in a house can be decreased. 

C-H Activation of Methane Using Homogeneous Rh-Based Catalysts: A Computational Study 

Vincentius Jeremy Suhardi 

Mentors: William Goddard III and Robert Nielsen 

Despite the fact that methane is one of the main components of natural gas, it has not been used to its full 

potential due to difficulties associated with its selective functionalization. Current technologies for the conversion of 

natural gas to liquid products proceed by the energy intensive Fischer-Tropsch process (Temperature = 850 o C). In 

this project, we explored different Rhodium-based catalysts for direct catalytic partial oxidation and 

functionalization of methane’s C-H bond. The advantage of direct functionalization of methane is the increased 

efficiency due to the ability for the process to proceed at much lower temperature (T=150-200 o C). Using Density 

Functional Theory (DFT) and the polarizable continuum model of solvation, we screened various Rh-bound pincer 

ligands (L), and various combinations of nucleophiles (X)/solvents combinations (HX) (X=TFA - , Cl - , HSO4 - , SO4 2- , 

and OH2). Screening of free energy surfaces for CH activation and SN2 functionalization barriers predicted that 

when L is a CNC-derived ligand with X=Cl - (~ 1 mM) and water at pH~8 as solvent, a full catalytic cycle of 

conversion of CH4 to CH3Cl will be observed with turnover frequencies ~ 1 hr -1 at T�200 o C. 

88

Optical Feedback Cooling of a “Zipper” Optomechanical Cavity 

Aiden Sullivan 

Mentors: Oskar Painter and Alexander Krause 

In an optomechanical cavity, such as the “zipper” cavities used in this investigation, the circulating optical field 

gives rise to a radiation pressure force that can drive the mechanical modes of the cavity. In turn, the mechanical 

oscillations deform and detune the optical cavity and thus modulate the optical field. Such coupling of the 

electromagnetic and mechanical degrees of freedom allows energy and information transfer between the optical 

and mechanical modes. Cavity optomechanics allows for unprecedented tests of quantum mechanics at the 

mesoscale. However, the mechanical mode must be close to its quantum mechanical ground state for quantum 

effects to be apparent. 

Cryogenic cooling alone cannot sufficiently cool the mechanics. In order to cool further, we create a damping force 

with feedback. We measure the position of the mechanics with the light modulated by the cavity, and modulate the 

light going into the cavity to drive the mechanics with a force proportional to the velocity of the mechanics. We are 

working towards an implementation of such a feedback loop, and have developed a design that uses analog 

electronics to accomplish this. Further work includes implementation and tests of this design, as well investigating 

the possibility of using DSP technology to accomplish feedback. 

NO2 Mixing Ratio Determination by Topographic Target Light Scattering-Differential Optical Absorption 

Spectroscopy (ToTaL-DOAS) in Gwangju, South Korea 

Sylvia Sullivan 

Mentors: Young Joon Kim and Jihyo Chong 

Nitrogen dioxide is a primary pollutant from fossil fuel combustion, which acts as an anthropogenic fingerprint in 

the troposphere. The free-radical cycle of its chemical family (NOx) also has a direct influence on stratospheric 

formation and tropospheric removal of ozone. These roles make quantification of atmospheric NO2 presence 

especially important. In this study, differential optical absorption spectroscopy is coupled with the newly developed 

topographic target light scattering method (ToTaL-DOAS) to obtain NO2 mixing ratios in Gwangju. The sensitivity of 

ToTaL-DOAS to temporal and spatial variations in urban boundary layer NO2 levels makes it an especially promising 

technique. Three sun-illuminated targets of similar albedo at increasing distance (560 m, 920 m, and 1490 m) are 

scanned both horizontally and vertically using Multi-Axis DOAS instruments. NO2 fluctuations are considered in 

relation to different meteorological parameters, and the accuracy of the ToTaL-DOAS method is assessed by 

comparison with in-situ data from three nearby air quality monitoring stations. 

Fabrication and Studies of Overdoped High-Temperature Superconducting Cuprate Thin Films 

(Y1-xCax)Ba2Cu3O7-� 

Qunchao Sun 

Mentors: Nai-Chang Yeh and Zhenjie Feng 

Vortex liquid state is a state when the H field goes up and close to Hc2, the vortices in Type II superconductor will 

be mobilized, which gives rise to a tiny current. The curve that marks the transition from vortex solid state to 

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 

Pulsed Laser Deposition (PLD) to make four samples of (Y1-xCax)Ba2Cu3O7-� with different doping levels of Calcium, 

and get the phase boundary between the vortex solid state and the vortex liquid state in the magnetic field vs. 

temperature plot for each of them to find the correlation between the vortex phase boundary and the doping 

levels. 

Directed Evolution of Alcohol Dehydrogenases for Aldehyde Resistance 

Sabrina I. Sun 

Mentors: Frances Arnold and Devin Trudeau 

Detoxification of lignocellulosic degradation products is an important challenge in optimizing viability and sugar-toproduct 

yields of engineered biofuel pathways. Among these compounds, aldehydes interfere most with microbial 

functionality whereas alcohols are notably less toxic. Alcohol dehydrogenases reduce aldehydes to alcohols using 

NADPH as a cofactor, but existing enzymes lack catalytic activity on many industrially relevant toxins. We therefore 

aim to create a novel alcohol dehydrogenase with broad substrate specificity and high activity, to improve aldehyde 

resistance in a range of fermenting microorgansisms. We have cloned and overexpressed the Saccharomyces 

cerevisiae ADH6 (alcohol dehydrogenase 6) gene, and verified that this can improve resistance of the wildtype 

BY4741 to cinnamaldehyde and 5-hydroxymethylfurfural. Random mutagenesis was used to construct an ADH6 

mutant library of approximately 10 6 clones. To achieve both high enzymatic activity and promiscuity, variants are 

currently being screened on combinations of aldehydes at lethal concentrations. 

89

Detecting Dust Devil Tracks and Wind Streaks in the North Polar Region of Mars 

Vivian Sun 

Mentor: Leslie Tamppari 

This study utilizes images from the Mars Reconnaissance Orbiter’s (MRO’s) Context Camera (CTX) to detect dust 

devil tracks (DDT), windstreaks (WS), and active dust devils (DD) in the 65-75N latitude band of Mars’s North Polar 

Region. Such features are formed through aeolian processes and can be indicative of local wind directions. The CTX 

images were processed and map-projected using ISIS, a planetary image processing software. The dataset yielded 

roughly 1500 usable images, of which 75% contained at least one of DDT, WS, or DD, with over 60 occurrences of 

active dust devils. DDT and WS will be analyzed for their frequency of occurrence and orientation and distribution 

according to location, and overlapping images will be compared for temporal analysis to determine seasonal, 

annual, and/or interannual changes and to assess the lifetimes of DDTs and WS. 

CubeSat Radiometer Development 

Yichuan Sun 

Mentor: William R. Johnson 

NASA/JPL has developed two automated validation sites at Lake Tahoe CA/NV and Salton Sea CA to help calibrate 

and validate thermal infrared measurements from airborne and satellite platforms. Radiometers deployed at the 

sites are used to measure the surface temperature of the water and a variety of meteorological equipment is used 

to provide ancillary measurements including relative humidity, air temperature, wind speed and direction and 

atmospheric pressure. In order to provide National Institute of Standards and Technology (NIST) traceability 

between the radiometers and the remote measurements the field radiometers must first be calibrated in the 

laboratory in a controlled environment. This process involves changing multiple variables to simulate a range of 

conditions, and requires operating four instruments simultaneously: environmental chamber, water bath, 

thermistor, and a radiometer. In order to undertake the laboratory calibration I have developed a multithreaded 

application coded in Python to test all possible permutations of environmental conditions. The code developed 

allows this by controlling all the necessary hardware. Additionally the program will also be used to simulate real 

weather data obtained from existing setups. 

Investigating Bismuth Oxide as a Photocatalyst for Water-Splitting 

Dylan Sures 

Mentor: Jay Winkler 

The photocatalytic process of water-splitting has the potential to produce a “clean” energy source in the form of 

hydrogen gas. In order to accomplish this water-splitting, a semiconductor photocatalyst is necessary; bismuth 

oxide has been investigated as a possible semiconductor material. Through cyclic voltammetry and open circuit 

potential measurements, it was determined that bismuth oxide produces a moderate, albeit unstable photovoltage 

in neutral and acidic electrolytic solutions. In higher-pH electrolyte solutions, bismuth oxide exhibited larger 

photovoltages and better, but still suboptimal stability as a photocatalyst. By combining bismuth oxide with other 

transition metal oxides, there is great promise for the preparation of a strong, durable photocatalyst. 

Terahertz Time-Domain Spectroscopy and Far Infrared Spectroscopy of Minerals 

Kevin M. Sutherland 

Mentors: George R. Rossman and Geoffrey Blake 

Far infrared spectroscopy and terahertz time-domain spectroscopy are used to investigate the low-frequency 

vibrational modes and torsional modes of crystalline solids. This study presents far infrared spectra in the range of 

600-50 cm -1 and terahertz time-domain spectra (THz-TDS) in the range of 150-3 cm -1 of 100 minerals. The 

minerals used in the study represent a wide cross section of common and rare minerals including oxides, 

carbonates, phosphates, pyroxenes, sulfides, sulfates, native elements, clay minerals, tungstates, and various 

other minerals. Samples are prepared for terahertz and far infrared analysis by pressing a mixture of polyethylene 

powder and mineral sample into pellets. Far infrared absorption spectra were collected with a Fourier transform 

infrared spectrometer and terahertz absorption spectra were collected with a femtosecond laser-driven terahertz 

time-domain spectrometer; in each case, spectra were collected in a dry gas-purged environment. Terahertz 

spectra and far infrared spectra were collected for 100 different minerals phases, compiled into a continuous 

spectra from 600 cm -1 to 3 cm -1 for each mineral, and entered into our terahertz/far infrared spectral library. All 

minerals investigated in the study produced a unique spectral fingerprint in the 600-50 cm -1 range. 

Thermoelectric Optimization in Sr3AlSb3 and Ca5In2Sb6 Through Control of Carrier Concentration 

Jessica G. Swallow 

Mentors: G. Jeffrey Snyder and Alex Zevalkink 

Sr3AlSb3 and Ca5In2Sb6 are two attractive possibilities in the search for new, efficient, low-cost, nontoxic 

thermoelectric materials. With controlled doping, the conflicting and interdependent properties that determine the 

thermoelectric figure of merit of these materials, and therefore their efficiency, can be optimized. Both Zintl 

90

compounds were synthesized with several dopant concentrations using powder metallurgy and their transport 

properties, including Seebeck coefficient, thermal conductivity, and resistivity, were investigated to high 

temperature. Modeling based on these measurements allows prediction of the optimal carrier concentration at 

which the maximum figure of merit for both compounds can be achieved. 

Tracking and Analyzing Avalanche-Like Flows in Granular Media 

Wesley Swank and Rivers Ingersoll 

Mentors: José Andrade and Carlos Avila 

Although phenomenological models for steady flows in granular media exist, the underlying mechanics of less 

predictable situations, such as avalanches, are difficult to simulate and predict with existing methods. By creating a 

link between the precision of the discrete element method (DEM) at the grain level and the flexibility of continuum 

models, the mentor is developing algorithms that can simulate sudden deformations with a high degree of 

accuracy. The goal of this project is to provide essential physical evidence to quantitatively assess the accuracy of 

the mentor’s theoretical framework and of simulations developed by the co-mentor. To generate this data, a highspeed 

camera captures thousands of images per second of the progression of short, recurring avalanches of steel 

beads in a steadily rotating plexiglass drum. These images are then reduced to an array of coordinates 

corresponding to the centroids of each bead in the frame of view, providing velocity and displacement values in a 

two-dimensional plane. With analysis, this data will be used to refine and eventually confirm the mentor’s model for 

the early progression of an avalanche. 

HCAL Noise Rate Stability Using Fit-Based Filters 

Kelly Swanson 

Mentor: Maria Spiropulu 

Robust noise-filtering algorithms are crucial for analyzing new physics phenomena, particularly those characterized 

by missing transverse energy. This paper presents an overview of fit-based noise filters and their long-term 

behavior. Using 2011 data, we examined isolated, flat, spike, and triangle pulse-shape discriminants for HCAL 

barrel and endcap anomalous signals. From this, we determined the channels that were persistently noisy and 

examined each filter's performance in high pile-up runs. 

Cellphone Medicine: Medical Devices for a Mobile Platform 

Albert Tan 

Mentors: K. Mani Chandy and Julian Bunn 

Medical technologies and devices have advanced significantly in the past decades, but the cost of the equipment 

remains high. Many third world countries are unable to afford such quality devices and thus have a dire lack of 

medical resources. Cellphone technologies, widespread across the globe, offer a potentially inexpensive solution to 

this problem. This project developed third generation low cost electronic stethoscopes and electrocardiograms in 

addition to hardware and software required for the devices to interact with cellular phones. The systems developed 

allow heart auscultations and electrical activity to be recorded and analyzed concurrently on the cellphone 

platform. The data can subsequently be uploaded to the Cloud and downloaded by a medical professional for 

further review. By building upon existing cellphone technologies, the project aims to create an affordable 10 cent 

medical checkup and diagnostic system that can be used in countries with limited resources. 

Viral Targeting of Dopaminergic Neurons 

Alison Tan 

Mentors: Thanos Siapas and Maria Papadopoulou 

The objective of this project is to use viral tools and target gene expression selectively to dopaminergic neurons of 

the rat midbrain. These neurons play a critical role in reward processing, addiction and movement initiation. Their 

loss, as well as misregulation of pathways in which these neurons are involved, have been implicated in many 

neurological diseases and disorders; as such, targeting gene expression specifically to these neurons is of great 

clinical significance. To accomplish specific targeting, we are testing injections of three different adeno-associated 

viruses in the midbrain dopamine-rich areas of rats. In these viruses, GFP expression is placed under the influence 

of an aminergic promoter; we therefore expect infected cells in the midbrain expressing GFP to be dopaminergic. 

Success of viral targeting of dopaminergic neurons will be assessed by the co-localization of GFP expression and 

the labeling of dopaminergic cells through antibody staining. By comparing the three viruses, we aim to identify the 

virus that leads to the highest spread of infection while maintaining targeting specificity of gene expression to 

dopaminergic cells. 

91

Construction and Characterization of Artificial Vessels for in vitro Arterial Wave Dynamic Studies 

Michelle Tang 

Mentors: Morteza Gharib and Niema Pahlevan 

The study of the wave dynamics and hemodynamics of the arterial system will lead the medical community to a 

better understanding of the pathogenesis and prognosis of cardiovascular complications, some of which include 

arteriosclerosis, ventricular hypertrophy, and heart failure. These two cardiovascular biomechanical concepts are 

affected by such factors as heart rate, reflection sites, pumping characteristics, and aortic rigidity. Wave 

dynamics—a synthesis of the effects on both the fluid flowing through the vessels and the vessel walls—are 

specifically influenced by the different parameters associated with aortic rigidity. For the aorta, such dynamics 

affect the load on and perfusion of the heart; for other arterial vessels, such dynamics affect the fluid mechanics 

within the vessels. In this study, we are investigating the relationship between aortic rigidity and wave dynamics— 

more specifically, how aortic rigidity influences the wave dynamics within the cardiovascular system. We designed 

several aortic and other arterial vessels of various compliances with either latex or silicone and subsequently 

measured their volume compliance, local compliance, elastic modulus, arterial distensibility, pulse wave velocity, 

and fluid particle velocity. The calculated results were then compared to the corresponding clinical results 

generated from previous studies of males and females within the range of 20 to 80 years of age. From these 

comparisons, we were able to experimentally simulate the physiological wave conditions associated with the 

arterial systems of males and females of various ages. 

Acquiring the High Frequency Terahertz Spectra of Amino Acids and Sugars Using Terahertz Pulses 

Generated via Plasma Filamentation 

Justine Tawel 

Mentors: Geoffrey Blake and Marco Allodi 

This work examined various hydrated and dehydrated amino acids and sugars in the 0.1 to ~7 THz range of the 

electromagnetic spectrum, corresponding to the range used by astronomical observatories to collect data. The time 

domain THz spectrometer contained a plasma filament to generate the THz radiation and a GaP crystal was used 

for detection. We acquired spectra of D-arabinose, D-fructose, D-glucose, D-mannose, D-xylose, glycoaldehyde, 

sucrose, L-aspartic acid, L-glutamic acid, 2-aminoisobutyric acid, and γ-aminobutyric acid over a wide range of 

temperatures from 10 to 310 kelvin. This study will enable astronomers to identify the presence of amino acids and 

sugars in regions of the cosmos, as these molecules have unique spectral fingerprints in the THz range. Prebiotic 

compounds, like amino acids and sugars, are thought to have played a role in the origin of life on Earth. In order to 

better understand the process of how life originated on Earth, and to search for the precursors of life in other parts 

of the cosmos, it is necessary to be able to find and identify regions where these prebiotic molecules are present. 

FU Orionis Stars 

Jamie Tayar 

Mentor: Lynne Hillenbrand 

FU Orionis is the prototype of a class of young stellar objects most notable for their dramatic outbursts. The 

spectra of stars have several distinctive features including P Cygni profiles in lines such as Hydrogen, Helium, 

Sodium D, Calcium II H and K as well as optical spectra similar to F or G supergiants and near infrared spectra 

characteristic of K or M supergiants. The leading theory is that an instability in the circumstellar disk causes a 

dramatic jump in accretion. The purpose of this project was to examine high resolution optical spectra taken of 

several FU Orionis objects and compared them to the spectra of stars ranging from F supergiants to M dwarfs in 

order to better understand the physical state of these objects and compare them to the predictions of leading 

theoretical explanations. 

Correlation of Gamma-Ray Photons With Radio Giant Pulses From the Crab Pulsar 

Thanchanok Teeraratkul 

Mentors: Anthony Readhead and Glenn Jones 

To gain better understanding of giant pulse (GP) emission mechanism and test models of giant pulse emission, we 

have carried out observations of the Crab pulsar at radio wavelengths using the GAVRT telescope which can then 

be compared to simultaneous gamma-ray observations made by the Fermi space telescope. In 60.37 hours of 

simultaneous observations, we obtained 8950 giant pulses observed at radio frequency of 2.6 GHz and 847 Fermi 

photons with energies between 100 MeV and 153 GeV. One photon with energy 1872.87 MeV was found within 0.4 

ms window of the interpulse (IP) GP, and one with energy 134.56 MeV was found within 1.8 ms window of the 

main pulse GP. Higher energy photons are more likely to be found near GP window, indicating a weak correlation 

between GP and gamma-ray. 

92

Erasure Correction for Nested Networks 

Omer F. Tekin 

Mentor: Tracey Ho 

We consider the problem of finding erasure and error correction capacity regions of networks having a nested 

structure, where each sink receives a subset of the packets received by its successor. Such nested networks can be 

used to model streaming applications, where each sink represents a deadline by which a certain set of information 

is decoded. We first characterize the erasure correction capacity region of a nested network for any given number 

of erasures, and provide an intrasession coding scheme that achieves it. We establish a relation between erasure 

correction capacity and error correction capacity regions under arbitrary number of erasures or errors. We then 

consider a sliding window erasure model as an approximation to probabilistic erasure model. We give an upper 

bound for the erasure correction capacity, and provide an intrasession coding scheme that achieves a constant 

multiple of the erasure correction capacity region. 

Fracture Toughness Enhancement of Elastically Heterogeneous Solids 

Ishan Tembhekar 

Mentors: G. Ravichandran and Shuman Xia 

Various brittle and quasi-brittle materials, like ceramics, cast iron, concrete and cast iron, find immense 

applications in industrial and engineering fields. These materials are often observed to fail at stress levels much 

lower than their theoretical values because of the presence of mechanical flaws or cracks in their microstructures. 

It is thus very important to increase the resistance to crack growth of such materials to prevent unknown failure 

mechanical structures. Existing work in this direction involves embedding the material system with tougher 

particles having higher intrinsic fracture toughness values and increasing crack resistance by crack trapping and 

crack pinning mechanisms. This project proposes that the same crack trapping can be achieved by introducing 

elastic heterogeneity only, and consequently leads to fracture toughness enhancement in an alternative way. The 

proposed enhancement mechanism is demonstrated on an experimental model system, in which elastic 

heterogeneity is achieved by embedding sapphire half-ball lenses in a polydimethylsiloxane (PDMS) matrix molded 

on a polyester substrate. A crack is initiated and propagated along the PDMS-polyester interface by application of a 

mechanical load. The entire event of crack propagation, pinning and depinning is captured by optical imaging. By 

using image processing techniques and load data, plots of fracture toughness versus crack length are obtained for 

two samples – one having elastic heterogeneity and other being homogeneous. The toughness values for the 

former sample are observed to be markedly higher thus conclusively proving the enhancement of fracture 

toughness of the system due to elastic heterogeneity. 

Bioremediation of Endocrine Disruptors From Water Using Genetically Modified Escherichia coli 

Nicole Thadani 

Mentors: Richard Murray, Nathaniel Glasser, Toni Lee, Joseph Meyerowitz, and Emmanuel Lorenzo de los Santos 

Endocrine disruptors, or substances that mimic estrogen in the body, have disastrous biological effects on the 

reproduction of several species of wild fish and birds. For the iGEM synthetic biology competition, the Caltech team 

is focusing on bioremediation of these toxins. The goal of this project is to create a system housed in E. coli that 

can be deployed in bodies of water to detect and remove endocrine disruptors. We identified enzymes that can 

degrade endocrine disruptors from previous research and from our own exploration of the genetic material from 

microbes in the Los Angeles river and characterized the behavior of these enzymes within E. Coli. We also isolated 

organisms that can grow on endocrine disruptors in minimal media as new sources of degradation enzymes. In 

addition, we explored the potential of certain cytochrome p450s to initiate degradation of these chemicals. We also 

explored the functionality of the human estrogen receptor as a signal for the presence of endocrine disruptors, by 

characterizing and modifying its toxicity and developing signal pathways that this enzyme can trigger through 

binding to endocrine disruptors and to a target genetic sequence. Finally, we characterized the functionality of E. 

coli protein processing when E. coli is deployed as an easily containable biofilm on plastic in aqueous environments. 

A Study of the Initiation Mechanism of Grubbs-Hoveyda Metathesis Catalysts 

Jordan C. Theriot 

Mentors: Robert Grubbs and Keith Keitz 

Despite the extensive kinetic studies performed on the first- and second-generation Grubbs catalysts, a conclusive 

initiation mechanism has not yet been determined for the Grubbs-Hoveryda type olefin metathesis catalyst. This is 

largely due to the limitations of standard NMR techniques: the substrate concentration cannot be increased enough 

to definitively observe saturation kinetics. We seek to overcome this obstacle by inducing saturation kinetics 

through an alternate method. A series of Hoveyda-type ruthenium catalysts have been prepared with varying 

substituents replacing the isopropoxy group of the benzylidene ether ligand. A study of their initiation kinetics gives 

further insight into the mechanism of this class of olefin metathesis catalysts. 

93

Entropy and the Witt Construction (or, Universal Structures in Elementary Algebra) 

Ryan Thorngren 


Variables, considered formally—without any reference to an underlying ring or algebra—have certain properties like 

associativity, commutativity, etc. which can be defined regardless of what we actually consider the variables to be. 

These formal properties are the object of study of universal algebra, which does not consider particular groups, 

modules, or lie algebras as interesting as their theories. One interesting structure is that of the Witt ring, an 

endofunctor in the category of formal rings which always produces a characteristic zero ring. The Witt functor is 

completely determined by formal properties of variables, and so we say it is a universal structure of elementary 

algebra. In application, the Witt ring can be used in characteristic p to produce unramified extensions of the p-adic 

numbers. In a sense it allows us to study characteristic p with characteristic 0. There is an obscure characteristic: 

one, whose objects if one would “write down” would be degenerate. Nevertheless, we can define structures over 

these objects by taking suitable limits of structures over characteristic p fields. A “limit geometry” is currently 

studied in hope of producing a solution to the Riemann hypothesis. Recently, Alain Connes constructed a Witt 

functor for characte.ristic one, whose inner workings depend critically on the Shannon entropy. In the present 

work, we argue that the same formal structures that force the form of the Witt functor also force the forms of 

Shannon and Tsallis entropies in information theory, and a coherent dictionary between characteristic zero Witt 

operads and information measures is presented. 

Split Stream Flow Past a Blunt Trailing Edge With Application to Combustion Instabilities 

Vicky Tian 

Mentors: Beverley McKeon and Ivett Leyva 

In a coaxial injector of a liquid rocket engine propellants traveling at different velocities are separated by the inner 

jet post before they come into contact with each other, mix, combust and power the rocket. Knowing how the 

fluids mix and how susceptible they are to instabilities is paramount for a successful liquid rocket engine. In this 

study, the wake behind a blunt trailing edge of a long plate, similar to an unwrapped co-axial injector, was studied 

using two fluid streams of different velocities introduced on opposite sides of the plate, and PIV was used to 

visualize the instabilities in the wake in order to determine the influence of the velocity ratio of the split stream. 

Measurements of the vortex shedding frequency were measured at various velocity ratios and compared with 

uniform free stream conditions to better understand of the mixing in coaxial jets typically used in cryogenic liquid 

rocket engines. 

Optical Counterparts to Unknown ROSAT (Röntgensatellit) X-Ray Sources 

Suk Sien Tie 

Mentors: Shrinivas Kulkarni and Varun Bhalerao 

The ROSAT Bright Source Catalog contains 18806 bright soft (0.1-2.4 keV) X-ray sources, of which around 50% 

are still unidentified. Using the database from the Palomar Transient Factory (PTF), we searched for the optical 

counterparts to these unknown X-ray sources to discover rare X-ray sources and/or new astronomical class(es). 

We look for sources with variable optical lightcurves within the positional error circle of each unidentified ROSAT 

source using a variability cut. Due to the huge amount of data, this process is being developed into an automated 

pipeline, of which PTF candidates of at least 300 unknown ROSAT sources pass the variability cut. As the filtering 

pipeline improves, we expect to cut down on the number of multiple PTF candidates to an unknown ROSAT source. 

In the presence of multiple candidates after the variability cut, they will be ranked accordingly. We report details of 

the matched counterparts. 

Pauli-Based Local Commuting Projector Codes 

Mohit Tiwari 

Mentors: John Preskill and Jeongwan Haah 

If Ai and Bi are commuting Pauli operators, then P = ¼(1 + Ai + Bi + AiBi) is a projector. We can define a quantum 

code whose code space is the ground space of a governing Hamiltonian H = ΣPi. We first provide a worked example 

with A1 = ZXZI, B1 = YXYI, A2 = IZXZ, B2 = IYXY. We then examine different choices of Pi and the Hamiltonian, 

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 

star and plaquette operators of the Toric Code proposed by Kitaev. We are interested in the logical operator 

algebra of such codes and implications for code distance and the existence of topological order in these systems. 

W+/Z+ Jets Cross-Section Ratio Produced With Muon Final States in pp Collisions at the LHC 

Baojia Tong 

Mentors: Maria Spiropulu, Yi Chen, and Christopher Rogan 

The study of jets produced in association with W and Z bosons provides important tests of perturbative QCD 

calculations, and the production of vector bosons with jets constitutes an important background in searches for 

new physics. In this project, we consider cases where W and Z bosons decaying into muons and the events are 

94

econstructed with particle-flow techniques. This study is done with the full 2010 data sample and current 2011 

data recorded with the CMS experiment at the LHC. The distribution of the number of jets produced in association 

to W and Z boson is processed after correction for muon efficiency, jet-energy scale and pile-up contributions. The 

analysis focuses on ratio measurements to avoid luminosity uncertainty and the impact of experimental 

systematics. We acquired the normalized inclusive rates of jets produced as well as the different jet multiplicity 

cross section ratios with different jet transverse momentum thresholds. From this we are able to compare the 

scaling behavior of the jets with different Monte Carlo simulations and current theoretical predictions. 

Implementation of a Distributed Camera Control Protocol 

Stephanie Tsuei 

Mentors: Richard Murray and Necmiye Ozay 

Current security camera networks are composed of many closed caption televisions (CCTVs). While CCTVs have 

aided in criminal investigations and increased security in public locations, they are still not very cost effective. We 

have created a small camera network made up of small tracking cameras with the ability to track people in an area 

and pan-tilt-zoom (PTZ) cameras with the ability to pan, tilt, and zoom to acquire high-resolution photographs of 

those walking through the arena. In order to do so, we modeled the interaction between the cameras and people 

as a “game” between the “environment”, or the people, and the “system”, or the PTZ cameras. If the cameras are 

able to acquire the necessary high-resolution photographs, then the system “wins” the game; otherwise the 

environment wins the game. By making a few assumptions about how people move and knowing how the cameras 

can move, we synthesize a controller that can control the cameras’ motions. We have implemented the protocols 

and collected tracking data and pictures on scripted targets and present those results here. 

Calculating Star Formation Histories Using Near Infrared Images of the M33 Galaxy 

Gautam Upadhya 

Mentors: Tom Soifer and Jason Melbourne 

We use Near Infrared (NIR) images of four fields in the nearby M33 galaxy to derive their star formation history 

(SFH). NIR images of galaxies are dominated by Asymptotic Giant Branch (AGB) and Red Helium Burning (RHeB) 

stars. AGB stars are red stars burning helium in shells outside their core, while RHeB stars are bluer and burn 

helium in their cores. Their properties are difficult to model and thus it is difficult to derive an SFH that produces 

correct populations of these stars. Using the CALCSFH program (which finds bestfit SFH’s for given sets of data), 

we have tried to reproduce the optically derived SFH's using the M33 NIR data. Previously, in other galaxies, this 

method yielded SFH's where the AGB population was overproduced, and the RHeB stars were under-produced. The 

main sequence and red giant branch stars in the SFH matched the real data however. It is likely that the stellar 

evolution codes used in the CALCSFH program need to be reevaluated to properly account for the AGB and RHeB 

behavior. Being able to derive SFH’s just from NIR flux will be of great help in finding SFHs of very distant galaxies 

where individual stars are not resolvable. 

Applications of the Dawid and Skene Model to Discrete Image Annotation 

Sara Venkatesh 

Mentors: Pietro Perona and Peter Welinder 

Crowdsourcing through online services such as Amazon Mechanical Turk can be an effective means of annotating 

images. We address the problem of collecting reliable annotations of discrete-labeled image sets through the 

implementation of the probabilistic model for the annotation process proposed by Dawid and Skene (1979). We 

explore the annotator parameters under which the model can provide accurate estimates of annotator competence 

and attain a higher accuracy in image labeling than majority ruling. We furthermore present possible advantages of 

supervising the algorithm by keeping image labels fixed at their ground truth values. 

Structured Illumination Microscopy for Improved Lateral and Axial Resolution 

Malvika Verma 

Mentors: Chin-Lin Guo, Jiun-Yann Yu, and Mingxing Ouyang 

This project involves constructing a structured illumination microscope based on previous models and then 

designing novel combinations of structured illumination with confocal, epifluorescent, or multi-photon microscopes 

such that the lateral and axial resolution of biological samples can be improved. To observe three-dimensional 

biological samples using microscopes, both surface and depth profiles - lateral information (x, y) and depth (z), 

respectively, are required. Scanning confocal microscopes are the most commonly used optical instrument for such 

purposes. They convey optical cross-sections, at different depths, by restricting light to the focal plane and 

rejecting out-of-focus light such that the final image contains only the in-focus information of the specimen at 

different focal plane depths. The confocal microscope does point-scanning across a cross-section, which is very 

expensive and time consuming. To speed up this process, higher excitation intensity is needed, but that is 

potentially damaging to cells. Structured illumination—projection of a one-dimensional, single spatial frequency 

fringe pattern onto the sample—is another optical sectioning technique. Image processing of three images at three 

different spatial positions can remove the out-of-focus image and fringe pattern. 

95

Development of PCR (Polymerase Chain Reaction) in Microfluidics 

Eugene Vinitsky 

Mentors: Michael Roukes and Jessica Lynn Arlett 

The goal of this project is to incorporate PCR (Polymerase Chain Reaction) into microfluidic chips. The ability to 

perform PCR in our dielectrophoretic chips will allow us to perform DNA and protein analysis on single cells in 

picoliter volumes of fluid. The main hurdle in utilizing PCR in microfluidics is the evaporation of the fluid in the 

channels due to the high heat needed to disassociate the strands of DNA. We hope to solve this by developing an 

appropriate hydration layer. Failing that we will manipulate the dimensions of the chambers to alter the surface 

area to volume ratio. If we succeed, we intend to make calibration curves to test the efficacy of PCR in our system 

versus the normal protocol. Finally, we intend to attempt to measure protein expression via reverse transcription 

mRNA measurements and integrate this with our dielectrophoretic chips. 

Near-IR Laser Cavity Ringdown Spectroscopy of Chloroalkylperoxy Radicals From 2-butene, 3-methyl- 

1-butene, and MBO-232 

Matthew Voss 

Mentors: Mitchio Okumura and Leah Dodson 

Peroxy radicals play an important role in the chemistry of the troposphere. Substituted alkylperoxy radicals have 

not been as well-studied as unsubstituted alkylperoxy radicals. We directly detected chloroalkylperoxy radicals 

using pulsed laser cavity ringdown spectroscopy in the near infrared region (7100 cm -1 to 8850 cm -1 ), observing 

the A-X electronic transition and the O-O vibronic stretch. The chloroalkylperoxy radicals were generated from 

chlorine oxidation of the alkenes 3-methyl-1-butene, MBO-232, and 2-butene. The effect of the chlorine substituent 

was examined in comparison to the corresponding alkylperoxy spectra. In addition, the chloroalkylperoxy spectrum 

of 2-butene was compared against the corresponding nitrobutylperoxy spectrum to inspect the effect of different 

substituents. 

Characterization of Small-Molecules That Elevate Reactive Oxygen Species in Cancer Cell Lines 

Alex Wang 

Mentors: David Tirrell, Stuart Schreiber, and Drew Adams 

Cancer is a collection of genetic diseases united by a set of common traits, including limitless replicative potential, 

ability to metastasize, and oxidative stress. Recently, the small-molecule piperlongumine has been found to 

selectively kill cancer cells, apparently by increasing reactive oxygen species (ROS). As a result of isolating 

piperlongumine, a high-throughput screen of over 40,000 compounds identified 902 small molecules that increased 

ROS with piperlongumine as the positive control. The objective of this project was to further characterize these hit 

compounds as well as piperlongumine. The hit compounds were characterized using various methodologies such as 

toxicity in the EJ cell line, antioxidant rescue of cell death, superoxide staining, ability to synergize with 

piperlongumine, and auto-fluorescence. Furthermore, four known bioactive compounds were found to antagonize 

piperlongumine induced cell death, generating new hypotheses regarding the small-molecule’s biological 

mechanism of action. The results indicate that ROS increase does not guarantee cell toxicity and that likely only a 

minority of mechanisms causing enhanced ROS lead to cell death. These findings give insight into the relationship 

between ROS elevation and cancer cell mortality, providing a basis for future prioritization of the screening hits. 

Test Bench for a Retinal Prosthesis Neurostimulator IC 

Angie Wang 

Mentors: Azita Emami and Manuel Monge 

Patients affected by diseases like retinitis pigmentosa (RP) and age-related macular degeneration (AMD) incur 

severe vision loss as retinal photoreceptor cells degrade, causing less visual information to be transmitted to the 

brain. However, it is possible to bypass damaged photoreceptors and stimulate the remaining retinal neurons with 

visual information directly through the use of charge-balanced biphasic currents generated by a retinal prosthesis 

implant, restoring vision to the blind. A system to test the functionality of the current-based neurostimulator IC 

that drives the retinal implant’s 1024-channel electrode array is being developed. Digital logic blocks designed in 

VHDL and implemented on an Altera Cyclone II FPGA along with radio-frequency analog components are utilized to 

wirelessly transmit and receive data between the test system and the retinal implant. Image data from a 5megapixel 

camera placed in close proximity to the implant is encoded for use as stimulus control signals that are 

sent to the implant using a quadrature phase-shift keying (QPSK) modulation scheme. The resultant stimulus 

driver output is verified by the test bench. 

96

Expression of a Thermostable TPP-Dependent Enzyme and Characterization of Its Decarboxylase 

Activity 

Eric E. Wang 

Mentors: Frances H. Arnold and Kersten S. Rabe 

Biofuels such as isobutanol have shown promise as an efficient renewable energy source. One potential resource to 

produce these biofuels might be plant waste material, where the abundant resource cellulose can be converted 

through consecutive reactions to feed host cells to produce isobutanol. One important step in the reaction cascade 

is the decarboxylation of 2-Keto-isovalerate to Isobutraldehyde which can be catalyzed by the enzyme Ketoisovalerate 

Decarboxylase (KIVD) from Lactococcus lactis. However this enzyme is not active at elevated 

temperatures (e.g. 60° C) which presents an obstacle for application to several biotechnological areas. Especially, 

the chemical decomposition of plant waste material involves heat and acid treatment; therefore, a thermophilic 

host organism would be very beneficial. To adapt the non-thermostable KIVD protein to work at higher 

temperatures, libraries of mutant versions of the KIVD are screened for endurance at higher temperatures as well 

as higher activity. 

In order to establish a second, parallel approach, genes that code for proteins with a predicted similar function to 

KIVD have been identified in thermophilic organisms and the corresponding genes have been transformed into 

E. coli utilizing appropriate vector systems. The proteins will be overexpressed and tested for any KIVD related 

enzymatic activity. 

These experiments are part of the effort to establish a metabolic pathway in a thermophilic host to ultimately 

convert cellulose into isobutanol at elevated temperatures e.g. 60° C. 

Scattering of Giant Radio Pulses From the Crab Pulsar by the Interstellar Medium 

Jing Wang 

Mentors: Anthony Readhead and Glenn Jones 

The interstellar medium scatters radio pulses emitted by pulsars, resulting in elongated pulse profiles. 

Unfortunately, the existing models of interstellar scattering fail to explain features seen in pulses from the Crab 

pulsar. At high frequencies, the observed pulse decay times scale as f −2 rather than the f −4 predicted by the thin 

screen model. We confirm the f −2 trend in our wide bandwidth 2 GHz to 10 GHz observations from the Goldstone 

Apple Valley Radio Telescope. Using simulations, we explore possible causes of this anomalous decay time behavior 

and place geometrical constraints on the ISM and the Crab nebula. We believe this anomaly can be explained by 

scattering in localized structures specific to the Crab nebula. Further analysis could be done on possible scatteringbased 

explanations of non-exponential pulse microstructure. 

Optical Design Optimization and Testing of an Intrasurgical Microscope-Mounted Optical Coherence 

Tomography System 

Kening (Connie) Wang 

Mentors: Joe Izatt, Justin Migacz, and Scott Fraser 

Spectral domain optical coherence tomography (SD-OCT) is an imaging technique which obtains high-resolution 

cross-sectional images of the retina noninvasively, and is currently the clinical standard of care for diagnosis of 

many retinal pathologies. Several research groups are working to extend SDOCT technology for use during 

intraocular surgery, where it could aid in visualization of delicate retinal structures. A handheld scanner exists for 

intrasurgical uses, but this requires the surgical instrument and optical microscope to be moved away from the 

patient. The Duke group has previously prototyped a microscope-mounted SD-OCT system (MMOCT) that would 

allow for real-time imaging without disrupting the surgeon’s view of the surgical microscope. This project involved 

using a professional optical design software, ZEMAX, to model the existing microscope-mounted SD-OCT system 

and to recommend improvements upon that system. In addition, optical testing techniques were used to measure 

quantitative performance metrics of the existing system and recommended improvements in order to verify the 

accuracy of the model and validate recommended design changes. 

Verification of Protein Receptors Using Double Florescent Immunohistory on VENs 

Yuchen (Carrie) Wang 

Mentors: John Allman, Nicole Tetreault, Brian William, and Nina Ng-Quinn 

Located in fronto-insular cortex (FI) and anterior cingulate cortex (ACC) of the brain, von Economo neurons (VENs) 

play significant roles involving the integration of interoception, emotion, and cognition. VENs’ regulation of several 

biological processes is dependent on several protein receptors: Interleukin 4 receptor (IL4R), Interleukin 10 

receptor (IL10R), dopamine receptor 3 (DRD3), and serotonin 2b receptor (5HT2B). In this study, the expressions 

of these four receptors on VENs are established by chromatic immunohistory and double fluorescent 

immunohistory. IL4R is linked to IgE mediated immunity and mutations are related to asthma and other allergic 

reactions. Interestingly, there is a linkage between the activity of the VENs containing area and the immune 

response, including asthma. IL10R has paracrine and autocrine actions as an anti-inflammatory responder and 

97

egulates IL1B. The evidence that these two receptors are strongly expressed on VENs suggests that VENs may be 

involved in the regulation of the immune response. DRD3 is the highest affinity dopamine receptor. When there’s a 

negative feedback or loss, there is a reduction in the dopamine signal that the receptor detects, activating VENs. 

5HT2B is strongly expressed in the human stomach and intestines, and may signal danger or punishment. 

Verification that 5HT2B, and DRD3, are expressed on the VENs suggests that VENs are involved in the intuitive 

decision-making in a given situation. 

Alkoxy Radical Cavity Ringdown Spectroscopy and Kinetics: Direct Detection of HOC4H8OO•, HOC4H8•, 

HOC5H10OO• and HOC5H10• 

Marissa Weichman 

Mentors: Mitchio Okumura and Matthew Sprague 

Alkoxy radicals are closely tied to the HOx and NOx reaction cycles, which impact air quality in the lower 

atmosphere. As alkoxy radicals are common derivatives of atmospheric hydrocarbons, determining the kinetics of 

their reactions is of great interest in modeling the impact of carbon emissions on Earth’s climate. An alkoxy radical 

can react in one of three ways: isomerization (if it can form a transition state with at least six atoms), 

decomposition, or reaction with O2. The isomerization product (HOR•) quickly reacts with O2 to form a 

hydroxyalkylperoxy radical (HOROO•). In the current work, we study the n-butoxy and 2-pentoxy systems with IR 

cavity ringdown spectroscopy. n-butoxy is the smallest alkoxy radical that can isomerize, and thus makes a good 

model for basic isomerization chemistry; 2-pentoxy provides a model for the behavior of more complex alkoxy 

radicals, but is simple enough to have only one pathway to isomerization. For each system, we have measured the 

OH vibrational spectra of primary isomerization products (HOR•, HOROO•), and use the HOROO• spectra to 

determine the relative kinetics of the isomerization and O2 reaction pathways. We use kinetic and quantum 

mechanical modeling for prediction and analysis of data. 

Sub-Nyquist Sampling of Action Potentials 

Alex Wein 

Mentors: Lakshminarayan Srinivasan and David Rutledge 

In signal processing, the Nyquist rate is the minimum sampling rate necessary to reconstruct a continuous-time 

bandlimited signal from discrete time-series samples. For signals containing quick pulses such as neuron action 

potentials, the Nyquist rate can be very high - more than 10,000 samples per second. However, such signals often 

have what is called finite rate of innovation (FRI), meaning they can be described by a finite number of parameters 

per second. For instance, a stream of action potentials can be approximately described by the spike times and 

amplitudes, and the rest of the signal can be treated as noise. If we seek only to reconstruct the spike times and 

amplitudes, the theoretical lower bound on the necessary sampling rate drops to the rate of innovation, which is 

orders of magnitude lower than the Nyquist rate. We discuss a variety of existing sub-Nyquist sampling techniques 

for action potentials and present a new algorithm that combines maximum likelihood estimation with a fast 

iterative procedure in order to reconstruct neural signals. 

Acoustic Levitation and Field Induced Droplet Ionization Mass Spectrometry for Studies of Chemical 

Reactions in “Wall-less” Reactors 

Paul Adrian Weitekamp 

Mentor: Jesse Beauchamp 

An acoustic levitation device has been constructed to create an airborne droplet reactor designed specifically for 

mass spectrometric analysis of molecular species present in the liquid phase. An ultrasonic transducer is employed 

with a conical reflector to produce an acoustic field in which liquid droplets can be trapped, creating a “wall-less” 

reactor for studying chemical and biological processes in volumes that range from 1 nanoliter to several microliters. 

To sample molecular species present in the droplet, an electric field is introduced along the axis of levitation to 

effect field induced droplet ionization (FIDI) ionization. Analyte ions are extracted through an aperture in the 

reflector of the acoustic levitation instrument and analyzed by ion trap mass spectrometry. This methodology 

allows for rapid sampling of complex molecular species to monitor ongoing reaction in the suspended droplet. This 

avoids any interference which surface contact might introduce. This becomes especially important with the 

unfavorable surface area to volume ratios in nanodroplets. This setup also provides a unique opportunity to study 

the air-liquid interfacial chemistry of the droplet. 

98

Derandomizing AM Using a Multivariate Generalization of Parvaresh-Vardy Codes 

Benjamin Weitz 

Mentor: Chris Umans 

We give a new proof of local list-decoding properties of Reed-Muller codes, independent of the fact that their local 

codes are Reed-Solomon codes. The proof uses techniques from low-degree testing to compile many, possibly 

inconsistent low-degree codes into a single global codeword. It is our hope that the proof in this paper may be 

extended to Parvaresh-Vardy codes, and thus obtain a locally-decodable multivariate generalization of PV codes. 

Using existing theorems, this would give an optimal uniform hardness vs. randomness tradeoff for derandomizing 

AM. 

An Analysis of the Presence of Hydrogen Peroxide on Mars From Mariner 9 UVS Data 

Danika Wellington 

Mentors: Leslie Tamppari and Amanda Hendrix 

The ultraviolet signature of H2O2 may be present in reflectance spectra obtained by the Mariner 9 ultraviolet 

spectrometer (UVS). In this study we isolate the contribution from surface reflectance in the UVS data by modeling 

and accounting for the atmospheric contribution to the total reflectance spectrum. We use the DISORT radiative 

transfer code to take into account the absorption and scattering properties of atmospheric gases and aerosols 

relevant in the wavelength region of interest (200-350nm), in particular CO2 Rayleigh scattering, dust and cloud 

aerosol scattering and absorption, and the Hartley absorption band of ozone. An analysis of the surface spectrum 

allows the abundance of hydrogen peroxide to be characterized, as well as its seasonal and latitudinal variations. 

Additionally, we investigate the expected correlation of hydrogen peroxide with water vapor, from which peroxide is 

produced via photochemistry, and the expected anti-correlation of hydrogen peroxide with ozone, whose 

production is inhibited by H2O2 and other photochemical products of water vapor. 

Analysis of Performance for Adaptive Mesh Refinement 

Jody Wen 

Mentor: Dan I. Meiron 

Adaptive mesh refinement is a computational technique used to solve problems that change in local complexity. For 

this project, we utilize a set of C++ template class to develop a series of tests essential to performing operations 

on discrete functions defined on a quadtree (in 2D) or an octree (in 3D). These tests are then implemented to 

measure performance and scalability by timing the various operations and then analyzing how the time varies with 

the processor count of a distributed memory system. Specifically, we examine the efficiency of look-up costs, cost 

of storage and retrieval of information in the tree, and the ability to modify the tree dynamically as requirements 

for adaptivity vary or aspects of our function evolve. 

Characterizing a Resonator Bolometer Detector Array 

Rebecca Wernis 

Mentor: Jonas Zmuidzinas 

Kinetic Inductance Detectors (KIDs) are a new type of radiation detector currently being developed at Caltech 

based on superconducting resonators. Their simple fabrication process and capacity for high-density frequency 

multiplexing enable drastically larger array sizes than would be possible with traditional submillimeter radiation 

detectors. Currently, both direct detection and absorption via antennas provide a means for coupling radiation to 

the array. We are investigating placement of the resonators on bolometric islands as a new coupling mechanism. 

This should enable warmer operating temperatures and an ability to be fabricated with a wider variety of materials, 

making them promising candidates for a compact, on-chip spectrometer as well as for satellite-based observations 

of Earth. In order to be established as a viable candidate for such projects, this type of detector must be tested in 

a lab. An array was fabricated at the Jet Propulsion Laboratory and brought to Caltech to be tested. Key properties 

of this array such as the resonant frequencies and quality factors of the individual pixels and the intrinsic noise 

were measured under various loading conditions and temperatures. The results were compared with theory to 

provide a crucial understanding of the detector and a clear pathway for device optimization. 

Enhanced Absorption in Thin Film Photonic Crystal Silicon Solar Cells 

Kelsey A. Whitesell 

Mentors: Harry Atwater, Dennis Callahan, and Jonathan Grandidier 

Cost of solar cells can be significantly reduced if a silicon solar cell that requires significantly less material than 

current designs but still has competitive efficiency could be designed. Increasing the density of available optical 

states (DOS) across the bandwidth of the solar spectrum and populating those states is key to increasing 

absorption per volume. This research explores the potential of using thin (100nm-800m), two-dimensional photonic 

crystals (PC) as solar cells and couplers to achieve these goals. The best-enhancements thus far were achieved by 

99

a 2D PC with cone-shaped unit cells and an array of different coupled defects in a PC cell, which absorbed 2-3 

times more power and current than a planar silicon slab of the same thickness, while requiring only 50-60 percent 

of the volume of material per thickness. 

Spectroscopic Analysis of a Plasma for an Astrophysical Jet Experiment 

Hannah V. Willett 

Mentors: Paul Bellan and Vernon Chaplin 

We are investigating the properties of a small plasma source which will be used to improve the research potential 

of the Caltech astrophysical jet experiment. Radio frequency power sources are used to break down argon gas into 

plasma, but varying conditions such as pressure and applied magnetic field affects the ionisation fraction of the 

plasma formed. Spectroscopic analysis of the light emitted by the plasma allows us to calculate its ionisation 

fraction and study the effects of changing the breakdown conditions. Preliminary data has indicated that the 

apparatus functions correctly and that one theoretical model (local thermodynamic equilibrium) does not apply to 

the plasma. We have found the most effective way of supplying power to the plasma and looked briefly at the 

effects of varying both pressure and magnetic field. Improvements to the circuit which creates the magnetic field 

are being made and we are investigating the suitability of a second model (coronal equilibrium) for calculating the 

ionisation fraction. Once an appropriate model is found, additional data will be taken to study the effects of varying 

the plasma conditions in greater depth. 

Low-Temperature Seebeck Measurement System 

Alexander Wilson 

Mentors: Jeff Snyder and Nicholas Heinz 

There has been recent interest in low temperature thermoelectric devices, especially that of possible cryogenic 

thermoelectric cooling, due to their scalability and robust solid state nature. In order to try to develop such 

materials, their ability to turn heat to electricity, or vice versa, needs to be quantified. Thermoelectric materials are 

rated by their Figure of Merit (zT), which is a measure of a material’s overall thermoelectric efficiency. The Figure 

of Merit is comprised of the Seebeck coefficient (S), the electrical conductivity (σ), and the electronic and lattice 

portions of the thermal conductivity (κe+κl), in the form zT = S 2 σ/(κe+κl). The Seebeck coefficient is a property of a 

material that relates the voltage (ΔV) that develops due to a temperature difference (ΔT) in a material 

(S = ΔV/ΔT). The low-temperature Seebeck system described here is a simple low-cost setup designed to allow the 

measurement of the Seebeck coefficient of thermoelectric materials at cryogenic temperatures without rare or 

expensive components. With relatively cheap parts, basic instruments, and minor programming, this lowtemperature 

Seebeck measurement setup is intended to be able to be easily constructed and modified for 

necessary requirements. 

Hydrogen Physisorption at High Pressure 

Andrew Wilson 

Mentors: Brent Fultz and Nicholas Stadie 

One possible method of storing hydrogen for an automotive application is physical adsorption (physisorption) on 

the surface of a porous material. Materials have been developed which show attractive hydrogen uptake and 

performance characteristics at cryogenic temperatures, but uptake for physisorbent materials at ambient 

temperature is typically less than 1 weight%, and this is not an improvement over compressed gas at the same 

pressure. It was reported in 2009 that zeolite-templated carbons (ZTCs) show remarkable hydrogen uptake at high 

pressures (10 to 34 MPa). No attempt has so far been made to reproduce this due to the difficulty of measuring 

physisorption at such high pressures. The Fultz group at Caltech has constructed an instrument to measure 

hydrogen uptake by the Sieverts method at pressures up to 70 MPa, in order to corroborate these results and to 

investigate the little-understood mechanisms governing high-pressure physisorption. The instrument has been 

commissioned up to 40 MPa, and initial adsorption measurements for the standard commercial adsorbent Maxsorb 

MSC-30 are consistent with literature results. I shall report hydrogen uptake measurements for MSC-30 and a 

zeolite-templated carbon. 

Sorting on the Surface of DNA Origami 

Sarah Wittman 

Mentors: Erik Winfree, Lulu Qian, Damien Woods, and Niranjan Srinivas 

Recently, DNA has shown itself to be a viable medium for use in developing computationally complex systems. 

Much work is being done to formulate ways of exploiting DNA’s natural properties and behaviors to form a new 

model of computation. One of the more computationally fundamental methods is the ability to sort and arrange 

data structures. In the case of DNA computation, this data is represented as a physical object with some feature 

indicative of the data it represents. Now, consider a 2D plane. On this plane is a random assortment of objects that 

have some meaning to us, something that we take some interest in keeping track of. Our goal for the summer is to 

100

develop and test a DNA system that is capable of retrieving and sorting these objects in a meaningful way, to 

provide some order to what is currently a relatively chaotic system, and give us some ability to track these objects 

of interest. 

Electrochemical Studies of Mono-, Bi-, and Trimetallic Glyoxime Complexes 

Justin M. Wolfe 

Mentors: Jonas Peters and Chris Uyeda 

Several monometallic macrocyclic glyoxime complexes have been shown to electrocatalyze the reduction of various 

substrates, and this project has analyzed the influence of additional metals. Electrochemical studies were 

conducted on 1) a monometallic copper complex, 2) related bimetallic species constructed from the copper 

complex with iron or zinc, 3) trimetallic complexes involving nickel dimethyl glyoxime and either copper or zinc. 

Their reactivity towards protons and carbon dioxide in the presence of an electric current was analyzed. 

Dynamically Optimized Sequential Experimentation (DOSE) for Estimating Time Preference Parameters 

Christina A. Wong 

Mentors: Colin F. Camerer and Stephanie W. Wang 

Time preference refers to the phenomenon of placing higher value on immediate gains than on future gains. Much 

literature exists evidencing the ubiquity of time preference; however, the current models of time’s effect on utility 

are not well defined and not generalizable to larger economic situations. To create a more robust model, 

parameters such as discount rate must be more accurately estimated. 

In order to efficiently gather the necessary data, we designed and coded a DOSE system to dynamically analyze 

the data during its collection, allowing us to ask subjects questions that will provide us more information. The 

program begins with an estimate of parameters based on existing expectations and updates that estimate with 

each question that the subject answers. The system chooses each question by determining how useful the answer 

will be in generating a finer estimation. The three most common models are the exponential, hyperbolic, and quasihyperbolic 

models. We plan to determine which of these three models fits best with our results. 

Prediction of Bubbles and Crashes Through Classification of Traders in an Experimental Market 

Katherine M. Wong 

Mentors: Colin F. Camerer and Alec C. Smith 

Two phenomena that may occur in a market environment are bubbles and crashes. Bubbles arise when traders buy 

assets at prices which are greater than fundamental values. Crashes occur when traders cease to buy assets at a 

higher value than the fundamental value, and prices fall to fundamentals. The idea that bubbles and crashes may 

occur even as a result of rational trading has been well-established in the literature. We have experimentally 

simulated markets that allow traders full information of fundamental value and where bubbles and crashes can be 

observed. 

By classifying traders based on their trading behaviors during the experiment, we fit a model of how the evolution 

of bubbles and crashes depends on the makeup of the trader population. Traders are classified as one of three 

main types: feedback traders, who trade based on past changes in market prices; passive traders, who trade 

based on fundamental values; and rational speculators, who trade based on estimates of future prices. We run 

simulations of markets composed of corresponding proportions of each type of trader to check that the simulations 

yield bubble and crash patterns similar to those observed in the experiment. 

On the Choice Number of the Hoffman-Singleton Graph and the Union of a Graph With a Matching 

Nicholas D. Woodward 


The choice number of a graph is the smallest number k such that any assignment of color lists of size k to the 

vertices of the graph allows for a proper vertex coloring. We provide bounds on the choice number of the wellknown 

Hoffman-Singleton graph. Alon and Tarsi’s seminal paper on oriented graphs and graph colorings provides 

foundation for this study. While studying the Hoffman-Singleton graph another question arose: could the addition 

of a matching to a graph increase the choice number by at most one? We explain the interest in this question as 

well as explain the difficulties in understanding it. 

101

Visualizing Intracellular Protein Trafficking and Chaperoning of α4β2α5 nAChR Protein Subunits in N2A 

Cells and Primary Cortical Neurons With Fluorescent Confocal Microscopy 

Adela Wu 

Mentors: Henry A. Lester and Crystal H. Dilworth 

Smoking is a major public health concern and contributes to many premature deaths. Recent human genomic 

studies attempting to classify possible factors for nicotine addiction have identified several single nucleotide 

polymorphisms (SNPs) that contribute to smoking-related behaviors. One particular SNP involves an asparagine-toaspartic 

acid substitution in the nicotinic α5 subunit at amino acid position 398 (rs16969968; α5(N398)). This 

α5(N398) mutant variant may be associated with an increased risk for greater nicotine use. 

In humans, the α5 subunit is expressed as pentameric α4β2α5 nicotinic acetylcholinergic receptors (nAChRs) in the 

central nervous system and as α3β4α5 nAChRs in the periphery. Of particular interest to us is the mechanism by 

which α4β2α5 nAChR protein subunits influence pathological states. We aim to answer two questions—first, 

whether inclusion of a wildtype α5 subunit into α4β2 receptors show differences in intracellular protein trafficking; 

and, second, whether inclusion of a variant α5(N398) subunit into α4β2 receptors show differences in intracellular 

protein trafficking. With expression of fluorescently-labeled receptor subunits in both N2A cells and primary mouse 

cortical neurons and imaging using confocal microscopy, we are able to visualize the localizations and life cycles of 

these nAChRs’ assembly within the cell. 

Growth and Characterization of Crystalline Titanium Dioxide Thin Films for Photocatalytic Applications 

Anjian Wu 

Mentors: Harry A. Atwater and Christopher T. Chen 

Anatase titanium dioxide (TiO2) is a large band gap semiconductor with many appealing properties such as its 

durability and chemical stability. Conformal TiO2 coatings on glass, silicon (111), and Si microwire arrays 

substrates were investigated because of this metal oxide’s powerful oxidiation capabilities when illuminated with UV 

light and fast charge transfer kinetics. These properties make TiO2 a worthy candidate for applications such as 

water splitting, dye sensitized solar cells, and artificial photosynthesis. TiO2 thin film samples were prepared using 

sol-gel deposition with different recipes, substrates, and annealing parameters. The films were characterized with 

scanning electron microscopy (SEM), atomic force microscopy (AFM), x-ray diffractometry (XRD), and Hall Effect 

measurements. We found that the crystallinity and phase of the films are mainly influenced by annealing duration 

and temperature, while the quality and morphology of the films can be controlled by the spin-coating speed, 

number of coated layers, sol-gel composition, and substrate type. Anatase films were observed around annealing 

temperatures of 500◦ C for at least one hour. The highest quality films were found at spin-coat speeds of 3000 RPM 

of fewer than three layers with sol-gels composed of lower vapor pressure alcohols. 

How to Tile a Rectangle With Rectangles of Algebraic Side Ratios 

Jianqiu Wu 


The tiling problem has been around for more than 100 years. A reasonable question we can ask is how to tile 

rectangles with rectangles. A recent paper studies this problem through its close connection to planar electrical 

networks and alternating circuits. The paper presented three necessary conditions for a rectangle to be tilable by 

rectangles of ratios For , these conditions are proved sufficient, but for , whether they are 

sufficient remains a conjecture. My focus was not to solve this problem because one of the conditions requires a 

rational complex function with integer coefficients to be positive-real, which means the function takes nonnegative 

real parts when all inputs have nonnegative real parts. This is not easy to characterize. Instead I studied a lot of 

examples of specific algebraic ratios. The most important results are the rectangles tilable by 

where are distinct square-free integers and those tilable by , 

where is a cube-free integer. 

Spin Polarized Scanning Tunneling Microscopy Studies of Graphene in Finite Magnetic Fields 

Teng-Pao (Renee) Wu 

Mentor: Nai-Chang Yeh 

When graphene is in the presence of a magnetic field, it displays what is known as the anomalous quantum Hall 

effect (QHE). The QHE is observed in two-dimensional electronic systems, when the orbits of electrons under a 

magnetic field are quantized at discrete energy values known as the Landau levels. Scanning tunneling microscopy 

(STM) techniques were used to resolve the issue of whether applying a magnetic field to graphene lifts spin 

degeneracy or pseudospin degeneracy. STM studies of the tunneling conductance spectra of graphene under finite 

magnetic fields applied along the c-axis were done using two different types of tips: regular Pt/Ir and Cr-coated 

Pt/Ir tips. The two different types of tips allow for both regular STM and spin-polarized STM (SP-STM) to be 

performed. Comparisons of the two types of spectroscopic studies were used to reveal which degree of degeneracy 

is lifted under the presence of an external magnetic field. 

102

Engineering Thermostable Synthetic Cellulase Mixtures 

Timothy Wu 

Mentor: Frances Arnold 

In a world where fossil fuels are rapidly dwindling, the production of biofuels offers a promising partial solution to 

the alternative energy problem. Previously, novel thermostable enzymes were created which operated individually 

at a higher temperature than their natural counterparts. However, in nature, organisms create mixtures of 

enzymes that act synergistically; this results in a higher net activity than the cumulative activities of each 

component enzyme in an equivalent concentration. In this experiment, the synergistic effects of substituting 

natural enzymes with thermostable chimeras in a minimal two- and three-enzyme mixture are examined. It is 

hypothesized that, at higher temperatures, the effect of synergy will increase due to greater diffusion of enzymes 

on the substrate surface. The relationship between activity and temperature of the enzymes, both individually and 

in mixtures, is examined, and a relationship between synergy and temperature is investigated. 

NuSTAR Pixel Characterization 

Yue Wu 

Mentors: Fiona Harrison and Varun Bhalerao 

NuSTAR, a hard x-ray focusing telescope launching into orbit early February 2012, consists of two parallel 

telescopes and two focal planes with four pixelated detectors each. Pixel-to-pixel variations in the detectors arise 

from CZT (Cadmium Zinc Telluride) impurities, grain boundaries and imperfect readout contacts. We developed IDL 

(a programming language) code to characterize these pixels using data from x-ray generator scans in the lab. We 

look at the regions in each pixel that detect events in neighboring pixels and those which detect events only inside 

that pixel. We also examine pixel edge effects on x-ray spectra and account for noise in order to map the effective 

pixel boundaries for all detectors. 

Extracellular Sulfatase-2 Levels Are Differentially Expressed in Idiopathic Pulmonary Arterial 

Hypertension 

Zhaoying Xian 

Mentors: Marlene Rabinovitch, Christopher Rhodes, and Dennis Dougherty 

Objective: An early event in the pathology of idiopathic pulmonary arterial hypertension (IPAH) is the apoptosis 

and dysfunction of endothelial cells (ECs). The goal of my project was to determine whether changes in EC gene 

expression might provide novel insights into the pathobiology of IPAH, and thus lead to better treatments. One 

candidate, extracellular sulfatase 2 (SULF2) regulates heparan sulfate proteoglycan (HSPG) function. HSPGs affect 

the activity of growth factors such as wingless (Wnt), which is important for cell maintenance and survival. 

Methodology: Twenty-four genes differentially expressed in ECs from control vs. IPAH patients were identified by 

RNA-Seq analysis, and a subset confirmed by quantitative PCR. EC protein expression was investigated by 

immunoblotting and immunofluorescence. 

Results: RNAseq identified six genes of potential relevance to IPAH in IPAH-ECs. Of those six, qPCR analysis 

confirmed only that SULF2 mRNA was downregulated (p=0.0015). Protein levels of SULF2 were, however, 

upregulated (p=0.013), suggesting increased stability of the protein despite low mRNA levels. 

Conclusions: An increase in SULF2 in IPAH-ECs may cause more HSPGs to release or activate growth factors, thus 

leading to increased proliferation and survival of pulmonary vascular cells, as seen in the pathology of IPAH. 

Testing the Functional Capacity of Long Distance Candidate Regulatory Elements 

Guoning Xiao 

Mentors: Barbara Wold and Katherine Fisher-Aylor 

All cells in an organism have the same DNA; gene regulation is responsible for the difference between various cell 

types. The general mechanism of this is thought to be protein transcription factors binding to cis-regulatory 

modules (CRMs) of DNA and then interacting with the promoter of the gene, causing a change in gene expression. 

This project attempts to explore the interactions of candidate cis-regulatory modules (CRMs) more deeply by 

determining the relationship between CRMs associated expression pattern of a gene in an artificial transfection 

assay. First, candidate CRMs are selected using ChIP-Seq data. C2 muscle cells are cultured and transfected with 

bacterial plasmids that contain the candidate CRMs driving the reporter gene luciferase. Transfection results will be 

collected by reading the amount of luciferase produced by the cells. After collecting the transfection results, the 

goals are 1) to use the data from the experiments to determine if connected candidate CRMs are more likely to 

drive expression than the gene with no CRM, and 2) to compare the expression patterns of the in vitro plasmid 

with the patterns of the nearby gene in vivo. 

If the CRMs are more likely to drive expressions, then the candidate CRM is a transcriptional enhancer. If 

expression decreased, the candidate CRM can function as a silencer. If the expression level of the gene stays the 

same, there is no conclusive result about the “secret” of expression. 

103

Profiling the Proteomic Response to the Incorporation of Synthetic Amino Acids in E. coli With Pulsed 

SILAC 

Yushu Joy Xie 

Mentors: David Tirrell and John Bagert 

Proteomics is a largely increasing field of study, as it accounts for translation-regulating elements within a cell, 

such as small-RNAs and micro-RNAs whereas genomics does not. Recently, a proteomics method called bioorthogonal 

non-canonical amino acid tagging (BONCAT) was developed that allows the separation and identification 

of newly synthesized proteins in response to internal and external cues, permitting a more sensitive analysis of 

cellular response. The process utilizes the synthetic methionine surrogate azidohomoalanine (Aha) which, following 

a pulse, is incorporated into protein synthesis by the endogenous cellular translational machinery. Aha serves as a 

chemical handle that can be selectively conjugated to affinity tags with the copper-catalyzed azide alkyne 

cycloaddition, allowing separation of newly made proteins from old. This method has the potential to be a useful 

tool in proteomics and other metabolic labeling applications such as in situ fluorescence labeling, but it is still 

unclear whether the use of non-canonical amino acids alters natural protein synthesis. In this study we use pulsed 

stable isotope labeling by amino acids in cell culture (pSILAC) to analyze the proteomic effects of incorporating the 

synthetic amino acid Aha into the cellular metabolism of Escherichia coli. 

Towards the Production of a BMP Morphogen Gradient 

Melissa Xu 

Mentors: Michael Elowitz and Nagarajan Nandagopal 

Bone morphogenetic proteins (BMPs) are essential for regulating proper patterning during embryonic development 

through creation of a morphogen gradient. Though the mechanisms of this process are unknown, it is possible that 

the formation of a BMP4 gradient is comparable to that of its homologue Dpp in Drosophila. The Dpp gradient is 

created by spreading of Dpp from a central stripe of cells in the imaginal wing disc, aided by interactions with the 

cell-surface protein Dally, a Drosophila glypican. Using BMP4-mCherry secreting cells and BMP4 responsive cells, 

we aim to create a BMP4 gradient in Chinese Hamster Ovary cells (CHO). Under the assumption that glypican-3 

interacts with BMP4 in the same fashion as its Drosophila homologue Dally and Dpp, we use glypican-3 to aid in the 

sequestering and shuttling of BMP4 along the cell surface. The successful creation of a BMP gradient using 

glypicans would allow us to further understand BMP signaling and its role in embryonic development. 

Terahertz Imaging With a 4x4 Pixel Focal Plane Imager in 0.13µm SiGe BiCMOS Technology 

Lita F. Yang 

Mentors: Ali Hajimiri and Kaushik Sengupta 

Terahertz (THz; 10 12 Hz) radiation exhibits high molecular specificity and high penetration capabilities, allowing for 

applications in nondestructive package inspection and various biomedical imaging applications such as displaying 

contrast between diseased and healthy tissue. In this project, we present a low-cost, compact THz imaging system 

which can be readily implemented for clinical or outdoor applications. A 4x4 pixel focal plane THz imager for 0.25- 

0.3 THz radiation has been fully integrated in low-cost 0.13 µm SiGe BiCMOS process technology. Without the use 

of external silicon lens or post-processing techniques, a pixel of the imager achieves high responsivity and low 

noise equivalent power (NEP). Imaging of objects through optically opaque materials with transmission-mode 

imaging demonstrates the potential for future cost-effective security screening applications. Biomedical THz 

imaging in transmission and reflection modes were also demonstrated with dehydrated skin and tissue samples. 

Tissue and water distributions of different regions of the tissue samples were clearly resolved, which exhibits the 

system’s potential in reflecting tissue conditions and for analysis of histological tissue features. 

Durability of Superhydrophobic Carbon Nanotube Arrays Under Various Aqueous Conditions 

Perren Yang 

Mentors: Morteza Gharib and Adrianus Aria 

Recent study by Adrianus Aria and Morteza Gharib showed that carbon nanotube (CNT) arrays exposed to heat 

under vacuum develop extreme hydrophobicity. My project investigates the durability of this quality; a durable 

superhydrophobic surface opens many avenues of technological development, ranging from specialized nonstick 

and drag-reducing coatings to sterile surfaces and novel fluid mechanics testbeds. My first experiment examines 

the decay of superhydrophobicity while CNTs are submerged in water, and whether this time might be extended by 

periodic removal into the air. Having established a baseline, I examine the changes in the rate of decay under 

increased pressures. Experiments with differing temperature, dissolved solutes, and dissolved air will also be 

conducted. 

104

A Flip-Out Method to Obtain Drosophila Transformants With a Mitochondrial Killer Construct 

Ran Yang 

Mentors: Bruce Hay and Nikolai Kandul 

Mitochondria are known to undergo fission and fusion to maintain healthy genomes, but genetic pathways 

regulating these mechanisms are not well characterized. Furthermore, the mitochondrial bottleneck suspected to 

serve as a natural screen to discard unhealthy mitochondria is not well understood. A cassette-regulated construct 

that uses a constitutive promoter to drive a restriction enzyme tagged for mitochondria was designed to control 

when deletions are induced in the mitochondrial genome. Markers were used to test for and identify major 

components of the construct; RFP and GFP showed the corrected function of the promoter and cassette in cell 

culture and after the constructs were injected into flies, mitochondrial deletions were screened using PCR. Previous 

work in this lab has suggested that not only can we control when deletions occur, but that these mitochondrial 

deletions can be passed through the germline to future generations. The cassette-based constructs have provided a 

more controlled environment for inducing deletions, which can then be used in conjunction with gene knockdown to 

test the function and relation of genes thought to regulate mitochondrial health, human aging, and Parkinson’s 

Disease. 

Combining Radial Velocity Measurements and High Contrast Imaging to Characterize Long Period 

Planetary Systems 

Scott Yantek 

Mentor: Justin Crepp 

Radial velocity (RV) measurements are the most successful ground-based method of detecting extrasolar planets. 

This method relies on a star’s motion about the center of mass of a star-planet system. Lines in a star's spectrum 

are shifted in wavelength periodically, and from the amplitudes of these oscillations we calculate a lower limit for 

the mass of a companion. Currently, there exist vast collections of radial velocity measurements, but in analyzing 

this data, the focus has generally been on shorter period RV trends (up to a few months), so the possibility of 

detecting longer period companions has not been fully explored. Many RV trends with periods of a few years or 

more, such as those caused by planets farther from their stars, may not have been recognized. We have used 

existing radial velocity data, among other resources, to identify and rank stars with promising longer period RV 

trends. The best of these candidates will be observed using high-contrast imaging with NIRC-2 at Keck. The quality 

of these images taken in the L band ( =3.8 m) will be improved by the inclusion of a recently developed method 

of background subtraction (background noise is high in this wavelength range). With these observations we will 

fully characterize the orbits of these sub-stellar companions, calculating their actual masses, which will significantly 

contribute to improving planetary atmosphere models by fixing mass independently of spectra (using dynamics) 

and removing it as an unknown. 

Deciphering How Mitochondria Activity Supports and Modulates Nervous System Function in the 

Nematode Caenorhabditis elegans 

Alex Yeh 

Mentors: Paul Sternberg and Amir Sapir 

Mitochondria are the powerhouses of cells and play a crucial role in the functioning of organisms. To study the 

molecular circuits regulating mitochondria activity in nerves and muscles as well as the consequences of 

mitochondria dysfunction, we knock down genes of interest through RNA interference (RNAi) and then analyze the 

effect on the nervous and muscular systems’ functions. With our collaborators at Tel Aviv University, we identified 

and focused on one gene that is localized to the mitochondria of muscles and specific neurons in a stage-specific 

manner. Knocking down this gene resulted in age-dependent attenuation of worms’ movement, pumping, and egglaying: 

activities that presumably stem from muscular or neuronal dysfunction. Surprisingly, inhibiting 

mitochondria respiration and inducing mitochondria stress by knocking down atp-3, a gene which codes for 

adenosine triphosphate synthase subunit, resulted in ectopic translocation of the protein to the mitochondria. The 

age-dependent localization and stress-dependent recruitment of the protein to mitochondria suggest that this 

protein is required for the normal function of mitochondria. Future plans include performing behavioral assays that 

can pinpoint the genetic factors encoding proteins that interact with this protein in mitochondria. A further goal is 

to find additional molecules and circuits required to prevent age-dependent mitochondrial dysfunction in the 

nervous system and muscles. 

The Development of Luminescent Probes for DNA Base Mismatches 

Jessica S. Yeung 

Mentors: Jacqueline K. Barton and Anna McConnell 

The accuracy of DNA replication is vital to cellular development and survival. Failure of the cell to repair DNA bases 

mismatches arising from polymerase errors or DNA damage can lead to diseases such as cancer. Therefore, the 

development of a luminescent probe targeting DNA base mismatches may lead to a diagnostic tool for cancer. A 

successful fluorescent probe ideally will act as a light switch for DNA mismatches and only fluoresce when bound to 

a mismatch. Therefore, the probe must be highly fluorescent and selective for mismatched DNA in comparison to 

105

matched DNA. Two ruthenium complexes based on the 5,6-benz[c]acridinquinone (acri) ligand, [Ru(bpy)2(acri)] 2+ 

and [Ru(phen)2(acri)] 2 , have been synthesized, characterized, and investigated as potential luminescent probes. 

Fluorescence studies in different solvents reveal that both complexes are fluorescent in acetonitrile, but not in 

water suggesting light switch behavior. The fluorescent and binding properties of both complexes were examined 

by further florescence studies and binding studies with matched and mismatched DNA. 

Analysis of the Electrical Characteristics of Micro-Superconducting QUantum Interference 

Devices (μ-SQUIDs) Fabricated Under a Trilayer Nb-AlOx-Nb Process 

Joshua Yoon 

Mentors: Kathryn Moler and Keith Schwab 

Superconducting quantum interference devices (SQUIDs) are known for its high field sensitivity, which has allowed 

magnetic and electronic materials research to take a leap forward in pursuit of new discoveries in many different 

areas. Although there is a growing number of books and articles that well-describe intrinsic properties of SQUIDs, it 

is of our interest to look at both its characteristic current-voltage (I-V) and flux-current (Φ-I) curves in our efforts 

to better understand its relationship with its unique fabrication process. Since SQUIDs reach a superconducting 

state only at low temperatures, we immerse it in liquid 4 He, and using a four-terminal configuration, we apply a 

sinusoidal current through the device loop and measure its voltage. By extracting its primary electrical parameters 

from the data gathered, the modulation amplitude tells us the size of the signal available from the SQUIDs and the 

I-V curves provides us with critical current values and its dependence on external flux. Data analysis show critical 

current vs. applied flux curves which are shifted relative to each other for certain devices, indicating asymmetry in 

the junction arms. In addition, critical current values seem higher for some devices than others. These results 

indicate that the SQUIDs that were tested underwent a non-uniform fabrication process, consequently producing 

undesirable devices that have either too high or low of a critical current or too small of a modulation amplitude. By 

adjusting the fabrication procedure in some way, which is yet to be determined, a higher yield of SQUIDs with a 

critical current range of ~10-15uA is possible in the future. 

Statistical Genomics of MRSA Infection and Disease Progression 

Yanwen You 

Mentors: Richard Wilson, Michael Wendl, and Katalin Fejes Toth 

Community associated Methicillin-resistant Staphylococcus aureus (CA-MRSA) infections are difficult to treat and 

pose a serious health concern. Some special cases escalate into a life-threatening infection. We hypothesize that 

the outcome of community MRSA acquisition is determined by interactions between the genetic composition of the 

infecting strain and polymorphisms in relevant host genes involved in immune and inflammatory responses. 

Samples of bacteria DNA and host DNA were collected from 6 sites in the U.S. Each sample was sequenced and 

compared to the reference genome to find small mutations in both the bacteria and its host. No significant 

differences were found among the samples of bacteria, indicating that the host genome is wholly responsible. The 

data is fitted with a logistic regression, controlled for co-variates, to determine the genes that contribute to the 

severity of the infection. Based on a preliminary data set of 49 human samples, the analysis returned several 

suspected genes. But with only 49 samples, any statistical test is underpowered. Much more data is required 

before we see a statistically significant effect and draw conclusions. 

Intracellular Trafficking of Nicotinic Receptors (α4β2) 

Caroline Yu 

Mentors: Henry Lester and Chris Richards 

Understanding the mechanisms of nicotine addiction is key to preventing tobacco-related diseases. Nicotine is 

known to upregulate specific subtypes of nicotinic acetylcholine receptors (nAChRs), leading to increased 

expression of receptors on the plasma membrane. The mechanism of upregulation is not well understood, but 

nicotine may affect rates of intracellular trafficking rates. Photoactivatable fluorophores were employed to observe 

trafficking of α4β2 subunits in neuroblastoma (N2A) cells from ER exit sites, and the Trans Golgi Network, as well 

as to quantify membrane turnover rates. Interestingly, nicotine was found to downregulate dynamin in the absence 

of nAChRs, but upregulate dynamin in the presence of α4β2. These observations support the hypothesis that 

nicotine affects protein expression at the plasma membrane likely through modification of intracellular trafficking. 

Low Frequency Impedance Measurement: Primary Coil of the Retinal Prosthesis System 

Chia-Chen (Debbie) Yu 

Mentors: Yu-Chong Tai and Yu Zhao 

Research shows that the retinal prosthesis has great potential for treating blindness diseases due to outer retinal 

degeneration and helping patients regain partial vision. Our retinal prostheses system incorporates an extraocular 

and an intraocular component that are inductively coupled to transfer data and power efficiently. This work 

presents the construction and characterization of the primary coil, which serves as the wireless power transferring 

link in the extraocular unit. Primary coils of different material parameters and winding methods have been 

constructed and characterized. In addition to standard 2-point and 4-point impedance measurements, a self-built 

106

4-point measurement setup is built to correctly characterize the coil parameters. A quality factor of 116 is reached 

at the operating frequency of 10 MHz with the braided litz coil design. Different winding methods of the coil and 

compensation of the data are the next steps to optimize the system’s performance. Ultimately, we wish to build a 

complete working primary circuit that minimizes power loss for the retinal prosthesis system. 

Capture Agent for the Akt2 Protein 

Mary Beth Yu 

Mentors: James Heath, Arundhati Nag, and Samir Das 

The protein Akt2 is associated with a signaling pathway for cell survival and overexpressed in some cancers. An 

inexpensive compound that could detect Akt2 in a sensitive and selective fashion could assist in early detection of 

cancer. Previously, a capture agent had already been found for a chemically synthesized segment of this protein. 

Here, we attempted to improve on this capture agent, two hexapeptides connected by a triazole, by adding a third 

hexapeptide. A biligand anchor, which is the extant capture agent with an azide added to the end, was synthesized. 

Biologically produced Akt2 and the biligand anchor were incubated with a C terminal alkyne amino acid containing 

library of hexapeptides. Then, library peptides that were both bound to Akt2 and formed triazoles with the anchor 

were isolated. In the future, the affinity and specificity of the resulting triligand for the Akt2 should be tested. 

Almost Strongly Full-Ranked Matrices Over Finite Fields 

Chi Ho Yuen 


An almost strongly full-ranked matrix is a t � u matrix with the property that one and only one of its t � t 

submatrices is singular. In this project, we study almost strongly full-ranked matrices over finite fields, as they 

appear in some linear algebraic constructions for certain combinatorial designs. 

Given a fixed finite field F and an integer t , denote by 

q 

u �q, t� 

the exact upper bound of u for which a t � u 

almost strongly full-ranked matrix over F exists. Using tools from finite geometry, number theory, linear algebra 

q 

and combinatorics, the structure of almost strongly full-ranked matrices over finite fields is studied and upper 

bounds and lower bounds for �qt� of u �q, t� 

for small q and t . 

u , are provided. Also, a computer search has been done to find the exact values 

The existence of large almost strongly full-ranked matrices over finite fields has been used to give some 

combinatorial designs that answer several problems in design theory and finite geometry. 

The Effect of Strategic Exits on the Incumbency Effect 

Jenny Z. Yung 

Mentor: Jonathan Katz 

Although elections in America are known to be democratic, statistics have shown that incumbents have an 

advantage in elections. However, previous analysis failed to take into account strategic exits. It’s possible that 

these incumbents win more because they only join races in which they are more likely to succeed, inflating the 

incumbency effect. Using a dataset of open US House of Representatives seats from the last 7 decades, we used 

statistical analysis to evaluate the difference between voluntary and involuntary exits and to determine the effect 

strategic entry and exits. We find that strategic exits have a significant effect on the incumbency effect. 

Predictive Statistical Modeling of piRNA Biogenesis From Progenitor Transcripts 

Alexander Zahn 

Mentor: Alexei Aravin 

Piwi-interacting RNAs (piRNAs) form a relatively recently discovered class of small non-coding RNAs that are critical 

to the suppression of transposon activity. piRNAs are known to comprise of short, twenty to thirty nucleotide 

subsequences of much longer non-coding single stranded RNA precursor transcripts through a poorly understood 

mechanism. We observe that the distribution of piRNA production events ( a “reads spectrum”) over precursor 

strands exhibits highly deterministic behavior, and in particular that the abundance of piRNA production at a given 

transcript position is a combined function of the immediately surrounding sequence of approximately twenty 

nucleotides as well as much longer sequences many hundreds of nucleotides distant along the transcript. We 

algorithmically analyze reads spectra in mice and Drosophila in search of patterns correlated with varying levels of 

piRNA production at particular transcript positions in hopes of forming a predictive model for piRNA production and 

also develop a quantitative metric for the level of similarity between arbitrary spectra over the same precursor for 

the purposes of model evaluation. 

107

Experimental Verification of the Theory of Dynamic Market Completeness 

Eric Zhang 

Mentor: Peter Bossaerts 

We sought to experimentally compare prices and portfolio choices for complete and incomplete asset markets, as 

explained in “Dynamically Complete Experimental Asset Markets” (Bossaerts, Meloso, and Zame). In this 

experimental setup, we created an incomplete market by shutting down one market and prohibiting trade of that 

asset. However, there was one additional provision: halfway through the trading process, we would announce that 

one of three possible future states would not occur. This additional flow of information allows, at least theoretically, 

for the incomplete market to fulfill the necessary conditions to be dynamically complete. This means that individual 

portfolios and asset prices should have exactly replicated those of the analogous complete market setup. Using 

electronic trading software, we tested this theory by implementing experiments among willing participants. 

Clustering Astronomical Event Notification via Natural Language 

Melissa L. Zhang 

Mentors: George Djorgovski and Matthew Graham 

Astronomical Event Notifications (ATels) are brief natural-language HTML reports on observations of dynamic 

sources in the sky. As sky survey technologies improve dramatically, these astronomer-reported text-based data 

become increasingly popular as a mode of communication. However, the data remain non-machine-processable, 

and so we cannot organize and interpret the information in large quantities. In this project, we use naturallanguage 

processing techniques to achieve automated classification of ATels by topic. We first removed all HTML 

tags, common words, punctuation, numerical tokens, and named entities from the ATel, stemmed the remaining 

words using the Porter Algorithm, and counted the stem frequencies to construct a vector space containing 

frequency-based characteristic vectors for each ATel. We then clustered these ATels via hierarchical clustering and 

used the top stems in each cluster to form a natural taxonomy of astronomical concepts, showing that these event 

notifications have an inherent classification that allows us to automatically organize ATels according a natural 

concept scheme. We conclude that we can use simple algorithms to process and extract meaning from 

astronomical textual data. 

The Use of Carbon Nanotube Arrays to Direct Stem Cell Differentiation 

Vivian Zhang 

Mentors: Morteza Gharib and Derek Rinderknecht 

Carbon nanotubes have unusually strong resistance to high mechanical strains. Therefore, they can be used to 

simulate a large range of mechanical properties of the extracellular environment around differentiating stem cells. 

Previous research shows a definite correlation between the differentiation of stem cells and the mechanical forces 

in the cells’ environment. However, for CNTs, there have only been qualitative data regarding the properties of the 

matrix on cell differentiation. To determine a quantitative relationship, CNTs will be anchored vertically in a matrix. 

The depth at which the CNTs are anchored will vary the effective length of each nanotube. This will alter the 

stiffness of the matrix as a whole, which is calculated with two types of force transducers by measuring the 

displacement of the tubes resulting from the amount of force applied. In the first mechanism, force-sensor probes 

will apply force along the side of a microarray, perpendicular to the direction of the nanotubes. Depending on the 

distance the nanotube was displaced, at what height the probe was relative to the base, and the amount of force 

applied, the stiffness of the matrix can be determined by pressing the probe into the side of the array. 

Microgrippers will also be inserted in the array and apply force to both sides of a bundle of CNTs. Undifferentiated 

stem cells will then be incubated on the array until differentiation occurs. The results should support existing 

qualitative research – stiffer matrices lead to bone cells while softer arrays yield adipose cells. 

Analysis of Noise Processes in Nanoelectromechanical Systems 

Yingqi (Helena) Zhang 

Mentors: Michael L. Roukes and Luis G. Villanueva 

Nanoelectromechanical systems (NEMS) integrate semiconductor electronics and mechanical devices on the 

nanometer scale to create sensitive, high-fidelity resonators with important applications in diverse fields ranging 

from precision quantum measurement to biotechnology. Such applications require resonators that are highly 

sensitive, down to the single molecule level. This can only be achieved with low noise levels and high frequency 

stability. Frequency fluctuation and noise measurements on several types of resonators with varying input 

parameters were conducted in order to better understand the relevant noise processes for such systems and 

possible problems with current models. 

108

Route Decisions in a Simple Network: A Pilot Study 

Yuanjun Zhang 

Mentor: Soo Hong Chew 

Travelers navigating through a road network must make route decisions prior to and during travel. We model the 

decisions of travelers choosing between routes with certain and uncertain time costs using traditional decision 

theory. Results from a small-scale experiment suggest that people behave differently depending on the information 

available to them at the time of their choice. Our results suggest that a more comprehensive theory is required to 

understand the decision-making behaviors of travelers in a road network. 

Crowdsourcing for Event Classification in NOvA Particle Physics Experiment 

Shiyu Zhao 

Mentor: Ryan B. Patterson 

The NuMI Off-axis � e Appearance Experiment (NOvA) aims to probe the nature of neutrinos through the 

phenomenon of neutrino oscillation. Its main goal is to search for the oscillation of muon neutrinos ( � � ) to electron 

neutrinos ( � e ), which requires accurate discrimination between � e Charged Current (CC) events and other 

background events. Some of the background events, such as some Neutral Current (NC) events, may produce very 

similar patterns to � e CC events in the liquid scintillator detector, and are thus difficult to recognize by computers. 

We present a method that uses crowdsourcing for event identification. This method first filters out computerrecognizable 

events, such as most � � CC events and some NC events, and then gathers the most ambiguous 

events into an event library for users to classify. We implement a prototype of such a crowdsourcing platform as a 

web application, capable of training and incentivizing users, and classify events according to user judgment. Upon 

further testing, this method may be used to improve the accuracy of event identification in NOvA. 

Autonomous Assembly, Capture, and Latching Strategies for Reconfigurable Spacecrafts in Low Earth 

Orbit 

Zipeng Zhao 

Mentors: Sergio Pelligrino and Marin Kobilarov 

There are many advantages to an active deployment scheme in an autonomous, reconfigurable, near-Earth 

satellite. In order to ensure that relevant docking maneuvers are performed successfully, a reliable docking 

mechanism must be implemented. Experiments evaluating the performance of several design iterations of docking 

mechanisms were conducted. All six degrees of freedom of the satellites’ surrogates were monitored and recorded 

at about 14 frames per second through an overhead camera stationed above an air table, which simulates zero-g 

vacuum, albeit in only two dimensions. Angle, velocity, and angular velocity at impact for each trial were extracted 

and amalgamated, and a plot of docking outcome for the three aforementioned parameters was generated. A pin 

and cup design that kinematically constrains the two spacecrafts with large cup diameters proves most promising 

for a typical CubeSat. Various theoretical latching strategies were also studied. Designs for potential mechanical 

latching devices were proposed and discussed. Further studies involving the use of variably controlled 

electromagnets on-board the spacecrafts during a docking maneuver and alternate docking performance evaluation 

methods are suggested. 

Sensory Substitution for Blind Rehabilitation and Cross-Modal Plasticity: A Study of Visual Substitution 

With Audition 

Yuqian C. Zheng 

Mentors: Shinsuke Shimojo and Noelle R. B. Stiles 

The brain is metamodal. While each portion of the brain analyzes best the signals from a particular sense, this does 

not preclude its analysis of other senses. We investigated the visual cortex's ability to analyze auditory input by 

using two sensory substitution devices (SSDs): the vOICe and the Raindrop. The vOICe generates a gray-scale 

image from the visual scene and encodes for brightness and location for each pixel with sound. After training pilot 

subjects for only five hours, they can localize a two-inch radius white dot and identify line conformations, simple 

shapes, and everyday objects without sight. We are also developing the Raindrop, an SSD designed to digitally 

mimic the sound of rain on a surface to determine location and surface qualities. To determine coding parameters 

and to gather baseline psychophysical data, six coplanar speakers called the Hoop is used. Pilot tests have shown 

that subjects can accurately pinpoint a single sound that lasts for twelve milliseconds and can also localize multiple 

sequential speaker activations. This data indicates that each raindrop can be just twelve ms, increasing the 

information bandwidth of the Raindrop. Studying cross-modal plasticity advances our understanding of sensory 

processing, while helping the blind to “see” again. 

109

Searching for New Generation Probes for Super Resolution Microscopy 

Ivanka R. Zhivkova 

Mentors: Long Cai and Timur Zhiyentayev 

The rapid development of fluorescence microscopy over the last few years made super-resolution imaging possible. 

However, the fluorophores that are currently used have low contrast ratio and low photo stability, which ultimately 

limits both labeling density and resolution. In this project we’re trying to synthesize carbospirans, photoactivatable 

fluorophores with enhanced contrast ratio and photo stability. Together with another SURF student I’ve been able 

to synthesize two coumarins with distinct absorbance spectra, both can be used as a caging moiety of 

carbospirans. This may allow multicolor super resolution imaging with carbospirans. Also if time allows I plan to 

obtain a water soluble carbospiran with excitation maximum at 650 nm. 

Using X-ray Crystallography to Unlock the Structure of the 2G12 Wild Type Dimer 

Yu (Daniel) Zhou 

Mentors: Pamela Bjorkman and Yunji Wu 

The 2G12 wild type dimer (WTD) is highly efficient at neutralizing the Human Immunodeficiency Virus (HIV) due to 

its high binding affinity to the low density gp120 spike. Because HIV constantly mutates into new strains, current 

pharmaceutical approaches aim to stop proliferation of the virus through a lifelong cocktail of medication. Solving 

the structure of 2G12 WTD reveals exactly how the dimer complexes with the gp120 envelop protein, which is 

crucial in developing highly efficient, broadly neutralizing antibodies to counteract the HIV infection with only a 

single shot. Attempts at discovering the protein structure have been made through various crystallization 

techniques involving: (1) hanging drop trays of 2G12 WTD complexed with 2G12.1 peptide in 5% tacsimate, 10% 

polyethylene glycol monomethyl ether (PEGMME), and 100mM HEPES, (2) isomorphous phasing methods using 

tantalum bromide (TaBr), and (3) crystal screenings of 2G12 WTD complexed with HxBC2 (construct with 7 

glycosylation sites removed from gp120 spike). Crystals have been observed in (1) but do not diffract to a high 

enough resolution for an accurate analysis while (2) and (3) both have shown encouraging results. Further studies 

varying the crystallization conditions will hopefully yield crystals of high enough diffraction resolution to produce a 

clear structure. 

A Route for Enhanced Strength Materials Through Three-Dimensional Nanostructured Design 

Jennifer Zhu 

Mentors: Julia R Greer and Andrew Jennings 

The goal of this research project is to develop a massive meta-material that utilizes the superior strength of 

nanoscale metals. In order to create a macroscopic sample that harness properties found only at the nanoscale, a 

fabrication route that preserves nanoscale features needs to be developed. As such, the initial goal of this research 

is to understand the mechanical properties of the component pieces: the individual iron nanopillars. This knowledge 

will allow for the design of macroscopic test specimens. In order to study these iron pillars, the first step is to 

electroplate iron nanopillars that range from 50 nm to 500 nm in diameter. Samples are then chosen with a 

scanning electron microscope (SEM). Mechanical properties are determined through small-scale compression tests 

in an Agilent G200, followed by post-compression analysis. Three pieces of data—pre-compression images, 

compression data, and post-deformation images—are then used to understand how iron plastically deforms as a 

function of size. 

Characterizing Maternal Drosophila Promoters for Use in Aedes 

Mario V. Zubia 

Mentors: Bruce Hay and Omar Akbari 

Several approaches are being made to help tackle the problem of mosquito-borne illnesses (e.g. malaria and 

dengue fever). One such approach implements a drive system that can help genetically engineered, diseaseresistant 

mosquitos replace the wild type population. One drive system, Medea, utilizes a maternal toxin and 

zygotic antidote. When progeny of a Medea-bearing mother receive this Medea gene, they survive; those that do 

not receive the gene, die. This dynamic requires the Medea gene is eventually needed for survival, and thus 

Medea-bearing individuals are given a fitness advantage. Thus after release into a wildtype population, these 

transformants are able to overtake and replace the population. Several Medea constructs were created and tested 

in Drosophila with varying degrees of success and failure. To understand why, components of this genetic construct 

were studied more closely. One such component, the maternal promoter, has different expression patterns in 

oogenesis. This may explain the effectiveness of the toxin and of the overall genetic construct. Further analysis 

needs to be completed, but if this association holds, then more effective systems can be developed and transferred 

to Aedes. 

110

MURF 

MURF UNDERGRADUATE 

RESEARCH FELLOWSHIPS 

M 

U 

R 

F

Collective and Rigid Body Behavior of Osmotic Motors 

Christian A. Aponte Rivera 

Mentors: John F. Brady and Nicholas Hoh 

Diffusiophoretic motion of a colloidal particle is possible if the particle creates a chemical potential gradient about 

its surface. We examine the behavior of multiple particles capable of inducing local concentration gradients, called 

osmotic motors. The particles are studied using Brownian dynamics simulations in untethered and rigid-assembly 

configurations. Simulations are verified in certain configurations by theoretical predictions. The effect of the rate of 

reaction, in the form of the Damköhler number, on the translational diffusivity is also studied. The average Péclet 

number is calculated as a function of the Damköhler number. 

Liquid Crystal Emulsion Generation With Micofluidic Techniques 

Gabriela Bernal 

Mentors: Julie Kornfield and Paul Pirogovsky 

Liquid Crystals (LCs) have been used as biosensors for the optical detection of bacteria and in photonics. Such 

sensors typically are thin film based and require special processing. Encapsulating the active LC element in a 

droplet with a responsive outer layer would allow for more robust, portable and easily manufactured sensing 

systems as well as self contained optical elements. We attempted to use a technique introduced by D.A. Weitz et 

al. using glass micro-capillaries to make monodisperse LC droplets using 4-cyano-4’-pentylbiphenyl (5CB). 

Preliminary calculations for relevant flow rates were performed taking account the geometry of the system and the 

rheological properties of the LC. The behavior of flowing liquid crystals was studied under temperature gradients 

and in the presence of a magnetic field, with results pending. We hope to be able to use the results of this study to 

create LC droplets of the desired size and shape. The next step is to incorporate a block copolymer surfactant 

consisting of a Side-Group Liquid Crystalline Polymer conjugated with a hydrophilic polymer to stabilize the 

droplets and control the anchoring condition at the interface, thus tailoring the structure and optical properties of 

the droplets. 

Data Quality Monitoring for Jets and Missing Transverse Energy, and Algorithms for Tagging 

b-Jets at CMS 

Dinah Simone (Simca) Bouma 

Mentors: Maria Spiropulu, Artur Apresyan, and Alex Mott 

The offline Data Quality Monitoring (DQM) framework is run on all the data taken at the Compact Muon Solenoid 

(CMS) experiment, in order to ensure that all the detector components are fully functional and performing 

according to expectations, and that the reconstruction of physics objects is performing as expected. To prepare for 

the increased instantaneous luminosity delivered by the LHC, this framework needs to optimize both its CPU and its 

memory usage. The JetMET DQM code was written during the commissioning period of CMS, when the first 

collisions occurred in 2009, so it contained some obsolete or unnecessary monitoring elements. These elements 

were removed, while the remaining elements was optimized according to the detector resolution. These changes 

resulted in a 70% decrease in memory usage, and were included in the next version of the CMS Software 

Framework. 

The second part of the project is to implement a new tracking algorithm for b-tagging at HLT. Jets originating from 

the hadronization of b quarks (b-jets) are interesting for studies of standard model physics, such as top quark 

properties, and physics beyond the standard model, such as supersymmetry (SUSY). Due to a very high crosssection 

of QCD multi-jet production, the trigger rates of purely hadronic triggers are extremely high. In order to 

efficiently collect events containing b-jets, a dedicated b-tagging algorithm is needed, prior to event reconstruction. 

The events tagged by this algorithm are saved and reconstructed for physics analysis. In order to improve the btagging 

triggers for SUSY, Exotica, and other analysis groups, a new algorithm for tracking the jets in the detector 

had been developed. Tentatively, the performance of this algorithm shows an improvement in efficiency of 6% 

compared to the old algorithm. 

Visual Representation of Chemical Species Over Southern California 

Denise Castellanos 

Mentors: John Seinfeld and Joseph Ensberg 

Data visualizations of the concentrations of atmospheric chemical species are meant to serve as a tool for 

understanding how the many processes in the atmosphere coincide. This understanding in turn serves as the basis 

for the actions that are undertaken to help prevent major shifts in the atmosphere’s composition which may lead to 

unintended consequences. Both the qualitative and quantitative nature of the data are used in the creation of 

public policy that is meant to protect the health of people and help reduce the impact of human activity on the 

earth’s climate. There are various types of graphs that are used to evaluate theoretical data and experimental data 

in different ways. For example, a time series graph is appropriate for comparing the experimental data and 

113

theoretical data of one pollutant type. In this way the validity of the model can be checked against actual data that 

has been collected. Also, contour maps can be utilized to compare potential theoretical models. The data can be 

manipulated in various ways which would offer a visualization that could explain the differences in the data sets. In 

this way it can help in determining the validity of one set or help prove that they are both false. Investigations into 

how data can be best represented is a necessary part of atmospheric research. It provides a means of finding the 

connection between numbers and the important concepts that will govern the future. 

Notch Toughness and Glass-Forming Ability of Vitreloy 101 and Its Variants 

Wei J. Chen 

Mentors: Marios Demetriou and Glenn Garrett 

The addition of Sn and Si to the Be-free Cu-based bulk metallic glass Vitreloy 101 is studied for their effects on 

glass-forming ability (GFA), notch toughness, yield strength, and elastic moduli. The glass-forming ability is 

determined by injection casting in to copper molds of different diameters and checking for crystallinity using X-ray 

diffractometry. The notch toughness is performed using 3-point bending of 3mm diameter rods. The yield strength 

and elastic moduli are measured from compression tests and ultrasonic measurements, respectively. Substituting 

Sn into the Vitreloy101 system improves GFA, but degrades the toughness. 

Tetranuclear Complexes for the Reduction of Dioxygen 

Michael Desanker 

Mentors: Theodor Agapie and Davide Lionetti 

There has been an emphasis on the development of alternate sources of energy that not only have the potential to 

replace the use of gasoline, but are also environmentally friendly. Fuel cells have this potential. The Pt/C catalyst 

currently used in fuel cells is expensive. Therefore development of a catalyst that utilizes cheaper metals will make 

fuel cells economically viable. Multinuclear complexes that reduce dioxygen are present in nature and serve as 

inspiration for the synthesis of a multinuclear complex capable of replicating this reactivity. Ascorbate oxidase in 

particular, yielded the idea of a dioxygen molecule binding at a trimetallic site. Ligand frameworks that would bind 

three iron or cobalt centers and a fourth yttrium center were synthesized. The goal of these complexes is to hold 

the cobalt or iron atoms in close enough proximity to each other to react cooperatively and limit excess reactivity. 

Metallation of these frameworks with iron, cobalt and yttrium has been studied. Further characterization of these 

complexes involves spectroscopy, electrochemistry and reactivity with dioxygen. 

Ground Motion Data Analysis in Southern California for Ruptures on the Southern San Andreas Fault 

Fernando (Estefan) Garcia 

Mentors: Swaminathan Krishnan and Ramses Mourhatch 

A large-magnitude earthquake in the range of 6.0-8.0 is expected to occur in southern California in the near future. 

An earthquake in this magnitude range is likely to have significant and hazardous impact on large city-centers such 

as downtown Los Angeles. High-performance computer simulations are used to obtain ground motion data at 

different stations in southern California from several earthquake scenarios. Using parameters of distance to rupture 

plane and magnitude, potential peak ground velocities can be analyzed in southern California. The mean and 

standard deviation of peak ground velocity can be found at different distance ranges, and confidence intervals for 

peak ground velocity can be created for each distance range. Graphs of peak ground velocity versus distance and 

magnitude can be used as tools to determine the level of shaking as determined by both parameters. To assess the 

accuracy of this analysis, peak ground velocity confidence intervals from simulations will be compared to those 

from an attenuation relationship developed as part of the PEER Next Generation Attenuation project. 

Investigating the Role of Conserved Signaling Pathways in Ascaroside Sensing in the Nematode 

Trapping Fungus Arthrobotrys oligospora 

Sherif Gerges 

Mentor: Paul Sternberg 

Arthrobotyrs oligospora is a nematophagous fungus which traps nematodes via three dimensional hyphal networks. 

Ongoing research has been focused on determining the molecular mechanisms responsible for nematode detection. 

Previous data have shown that the ascarosides, which are C. elegans pheromones, trigger trap formation in A. 

oligospora. Two conserved signaling pathways, the cyclic adenosine monophosphate – protein kinase pathway 

(cAMP-PKA) and mitogen-activating protein kinase (MAPK) pathways are known to govern fungal morphogenesis in 

fungal species across different phyla; thus, key components in the pathways including the protein kinase A PKA, 

MAPKK STE7, the G-protein coupled receptor STE3, and the G-beta subunit GPB1 were selected as candidate genes 

for targeted gene deletion to test whether these pathways play a role in ascaroside sensing. We use biolistics and 

protoplasts as two methods for transformation. Putative drug-resistant colonies will be screened by PCR and 

Southern blot hybridization to confirm gene deletion. Once the mutants have been obtained, we will test their 

ability to respond to ascarosides on a plate assay. 

114

Expression Patterns of Msi2a in Neural Crest Cells of Wild-Type Zebrafish Embryos 

Anna Ruth Gonçalves 

Mentors: Marianne Bronner and Tatiana Hochgreb 

Studies demonstrate that the Musashi family of genes is co-expressed in proliferating embryonic pluripotent neural 

precursors and in cells that are likely to lead to postnatal and adult CNS stem cells. Imbalanced expression of 

genes within the Musashi family (e.g. Msi1 and Msi2) has been linked to the following diseases: Pick and 

Alzheimer’s disease, Dementia and Chronic Myeloid Leukemia. Preliminary work in the Bronner lab has shown that 

Msi2 expression has been segregated between Msi2a and Msi2b in Zebrafish embryos, with Msi2a expressed in 

migrating neural crest cells, while Msi2b is localized to neurons in the ventral neural tube, olfactory region and 

pineal gland. This study involves performing Whole Mount in-situ Hybridization to characterize the expression 

pattern of Musashi 2a on wild-type Zebrafish embryos in order to visualize where the mRNA of Musashi2a within 

the embryo is expressed. Considering that not much is known about the Musashi gene family -- which is a family of 

post-transcriptional regulation genes -- beyond understanding the Musashi gene family, further studies can also 

explore how other genes may be regulated by the Musashi gene family and lead to various diseases and 

consequently, therapeutic interventions for these diseases. 

Power Flow of Electrical Distribution Feeders Modeled by Open Distribution System Simulator 

Jocelyn Graf 

Mentors: Steven Low and Dennice Gayme 

After electric power has been generated and transmitted on high power lines to substations, distribution feeders 

transport the power at lower voltages to individual customers. Many of these systems have not changed in 50-100 

years, but new technology is rapidly creating the need for new systems. For example, electric vehicles will demand 

new levels of power. Furthermore, a much higher proportion of electric power must be generated from renewables 

like wind and solar generators that produce intermittently, in turn, requiring power storage to smooth power flow. 

Some of this renewable generation will be distributed as smaller units throughout the grid instead of being 

concentrated at large power plants. These changes require a retrofit of power systems to include faster and more 

comprehensive monitoring, control, communication, and security systems, which will combine to create a “smart 

grid”. Energy can be conserved and generation thereby reduced if power flow optimization algorithms are 

developed to take advantage of the smart grid’s new capabilities. A distribution feeder was selected and modeled 

using the Open Distribution System Simulator. Power flow analyses were performed with and without distributed 

generation. The test feeder was characterized for power flow optimization research incorporating distributed 

photovoltaics and battery storage. 

Malley Probe: An Optical Instrument That Will Facilitate the Study of Thermally Perturbed Boundary 

Layers and Their Effects on Optical Systems 

Edward B. Guzman 

Mentors: Beverley J. McKeon and Rebecca Rought 

Collimated laser beams have been used in many aerospace applications, as in the case of communication and 

targeting. Unfortunately, when a collimated laser beam traverses a turbulent boundary layer containing density 

fluctuations, its wavefront becomes aberrated, potentially affecting the ability of the laser beam to be used in the 

far field. Yet, the causes of these aberrations are not entirely known, but only that they are produce by density 

fluctuations originating from either compressibility effects or the addition of heat to the boundary layer. The Malley 

Probe is an optical instrument that provides accurate information about the distortion of a laser beam in a turbulent 

boundary layer whose information will be used to calculate the convective velocity of the turbulent structures in the 

flow and the Optical Path Difference of the laser beam. The assembly of a new Malley probe and preliminary results 

from the turbulent boundary layer will be discussed. 

Transforming Drosophila Through in vivo Injection of DNA and Transfection Reagent 

Dustin Harris 

Mentors: Bruce Hay and Kelly Dusinberre 

Creating transgenic insects has been possible for decades. These experiments have helped us learn more about 

genetics and can lead to solving problems in agriculture, medicine and basic biology, but not all insects are easily 

transformed. The goal of our experiment is to inject both male and female adult Drosophila with DNA and 

transfection reagent to achieve transformation in the germ line. The plasmid DNA containing the gene of choice is 

co-precipitated and mixed with transfection reagent which encapsulates it in a cationic liposomes. These are then 

injected into the abdomen of living adult Drosophila and transported from the host hemolymph to the germ line. 

The liposomes fuse with the cell lipid bilayer and release the DNA, enabling transposition or integration of plasmid 

DNA into the genome. The first generation progeny of the injected males are screened along with the first and 

second generation progeny of the injected females. Transformation has not been seen in the first or second 

generation progeny of the first few DNA/transfection reagent combinations, but screening is not complete. More 

experimentation with the injection of larvae, composition of the injected material and volume injected can be 

carried out. 

115

The Planet-Metallicity Correlation in the Kepler Field 

Keith A. Hawkins 

Mentors: John Johnson and Tim Morton 

One of the most interesting correlations that has come out of large statistical studies of exoplanets is a relationship 

between the occurrence rate of giant planets planets and the metallicity of their host stars. The primary goal of this 

project is to develop an automated pipeline to measure stellar parameters (e.g. effective temperature, surface 

gravity and metallicity) of a large number of stars. We use a spectral index-based method that employs a Bayesian 

approach to determine the stellar parameters of stars. Preliminary results of this method will be presented. This 

method will be applied to a large number of stars in the Kepler field in order to determine the planet-metallicity 

correlation for small planets for the first time. 

Do Ascarosides Affect Egg-Laying Behavior in Caenorhabditis elegans? 

Circe Lassegue 

Mentor: Paul Sternberg 

Caenorhabditis elegans, a small transparent nematode that lives in temperate soil environments, is one of the 

simplest eukaryotic organisms with a nervous system to be studied in great detail. Over recent years, a large 

number of ascarosides have been identified as signaling molecules in C. elegans. Ascaroside levels are affected by 

worm concentration and available food when developed in “worm water”. Ascarosides have been shown to regulate 

a large number of behaviors in C. elegans including dauer formation, mating behavior, aggregation, and olfaction. 

Additionally, environmental and homeostatic cues are now being explored to see how these affect nematode egglaying 

habits. We studied the modulatory effect of several ascarosides on egg-laying behavior and brood size in 

adult female C. elegans. 

This study aims to determine the effect of ascarosides on egg-laying behavior in adult C. elegans. A range of 

concentrations of several synthetic ascarosides as well as natural worm water produced by C. elegans were 

studied. Standard egg-laying assays and known positive and negative controls were utilized. 

Perfluoroalkyl-Poly(ethylene glycol)-Poly(acrylic acid) Triblock Copolymers for Mucoadhesive Drug 

Delivery 

Yulan Lin 

Mentors: Julie Kornfield and Jeremy Wei 

Hydrophobic drugs can be loaded into the core of polymeric micelles, which form from diblock or triblock 

amphipathic polymers. Poly(ethylene glycol) based copolymers are of interest because PEG is nontoxic and 

nonimmunogenic, and fluoroalkyl ended PEG (Rf-PEG) has been shown to form a gel in solution. However, 

previously synthesized Rf-PEGs have not shown mucoadhesive properties. The desired mucoadhesive properties 

could potentially be conferred by adding poly(acrylic acid) (PAA), a compound characterized by its mucoadhesive 

properties. A tri-block Rf-PEG-PAA will be synthesized by polymerizing tert-butyl acrylate on the end of single 

ended Rf-PEG modified with a reversible addition-fragmentation chain transfer (RAFT) agent. The poly(tert-butyl 

acrylate) will then be converted to poly(acrylic acid). The resulting material will be characterized by gas permeation 

chromatography, NMR spectroscopy, a mucoadhesiveness test, and rheology. These tests will allow for the 

hydration kinetics, mucoadhesiveness, and rheological properties to be studied. Further tests will be run to 

determine its usability in targeted, controlled, drug delivery systems. 

Exploring the Spectrum of Multi-Dimensional Quasi-Periodic Shrödinger Operators 

Marissa Loving 

Mentors: Barry Simon and Helge Krüger 

Operators can be viewed as generalizations of matrices and their spectrum as the generalization of their 

corresponding eigenvalues. Shrödinger operators are a special class used in the description of quantum mechanics. 

In particular, their spectrum describes the allowed energy levels in such a system. It is the spectrum of a smaller 

subset of the one-dimensional Schrödinger operators known as the Almost Mathieu or Harper's operator that leads 

to the Hofstadter Butterfly a recursive geometric object known as a fractal. 

The Hofstadter Butterfly was computed two ways, by computing the eigenvalues of finite tridiagonal matrices or, 

alternatively, by finding the roots of the polynomial produced by computing the discriminant, D(E), where V(n) = 2 

cos(2 π ((p/q) n + x)). Due to the complexity of the computations and the sheer number required, Mathematica 

was used both to perform the computations and produce the final graph. A report was generated detailing the 

processes used to produce the Hofstadter Butterfly in order to simplify the process for future mathematicians 

performing research in this area. 

116

Ice Crystal Growth in Solutions Containing Salts, Polymers, Nanoparticles, and Antifreeze Proteins 

Icon Mazzaccari 


While there exists an established theory for ice crystal growth in pure water that can predict the growth rate and 

radius of the dendritic tip, there is no such theory for ice crystal growth in solutions. As the solution is supercooled 

and the ice interface begins to grow, it must release latent heat to transition into the solid ice phase. 

Simultaneously, the growing ice lattice must reject the solutes in the liquid, causing an increase in concentration 

near the ice interface. Solutes also lower the freezing temperature of solutions by disrupting the interactions 

between the solvent’s molecules. An investigation of how solutions containing chemical compounds including 

sucrose, salts, like NaCl and LiCl, polymers, nanoparticles and antifreeze proteins affect the growth of ice crystals is 

discussed, as well as the preliminary work towards a comprehensive theory of dendritic growth in solution. 

Experiments have been performed with the simpler compounds, and results and future plans will be described. 

Microstructural and Mechanical Characterization of Cementitious Materials 

Luis A. Miramontes 

Mentors: Julia Greer and Jane Lian 

The objective of the project was to measure the elastic modulus and hardness of Carbon Silicate Hydrates, C-S-H, 

at its different phases. C-S-H gels have two main phases, an inner sold bulk area and an outer more porous area. 

When ettringite forms within the nanopores of the outer phases, nanocracks are generated, and the mechanical 

properties could be quite different from the inner solid bulk phase. Four different cement samples were sent over 

by our collaborators at Tennessee Technological University to test in 1”x1”x ~2” blocks. The samples were then 

machined into 1” Diameter cylinders with a height of around 1” and set into an Epoxy mold to fit into our G200 

nanoindenter. As nanoindentation measurements are highly sensitive to sample roughness, the sample surfaces 

need to be flat and smooth so that reliable mechanical properties can be obtained. To achieve this, all samples 

were successfully grinded and polished down to a 0.05�m finish. Nanoindentation experiments will be conducted on 

different phases of the sample, and a comparison of the elastic modulus and hardness between the inner phases 

and outer phases will be obtained. 

Characterization of Adh5 Expression and Function in the Chick Neural Crest 

Ryan D. Morrie 

Mentors: Marianne Bronner and Crystal D. Rogers 

Adh5 is a class III alcohol dehydrogenase that may be required for normal embryonic development. In situ 

hybridization for Adh5 in Gallus gallus embryos showed a spatiotemporal expression pattern with Adh5 enriched in 

the developing neural tube and the migratory neural crest, suggesting an involvement in neural crest formation 

and migration initiation. Cryosectioning also revealed that Adh5 is expressed around the entire apical area of the 

neural tube. Here we analyzed the role of Adh5 in neural crest development through loss-of-function experiments 

using a translation blocking morpholino oligonucleotide against Adh5 (Adh5MO). Gallus gallus embryos were 

injected and electroporated with various concentrations of Adh5MO. At high concentrations Adh5MO appeared to 

induce cell death. Therefore, we analyzed the phenotype caused by lower doses of Adh5MO with 

immunohistochemistry and in situ hybridization against proteins and genes important in neural crest formation and 

migration, as well as general cell adhesion molecules. My results show that Adh5 is expressed in developing neural 

crest cells and may be necessary for neural crest specification and development. 

Determination of Binding Mechanism of a Beetle Antifreeze Protein and Crystalline Carbohydrates 

Edward (Ted) Ro 

Mentors: William A. Goddard III, Ravinder Abrol, and Xin Wen 

Antifreeze proteins (AFPs) refer to a class of polypeptides found in fish, plants, and insects that lower the freezing 

point of water without affecting the melting point. An AFP from the beetle Dendroides canadensis (DAFP-1) exhibits 

high antifreeze activity and demonstrates the ability to hinder crystal growth and prevent precipitation of naturally 

occurring carbohydrates at lower temperatures in vitro. The focal of the research is to identify the ligand binding 

sites of DAFP-1 by simulating protein and ligand docking. The hypothesis is that DAFP-1 will bind to hydroxyl 

groups of carbohydrate molecules in solution and interact with growing crystalline carbohydrates through either 

threonine or arginine residues. Through the use of GenDock software, a docking program that generates a large 

number of ligand binding poses and ranks them according to energy, the most favorable binding poses of DAFP-1 

and various trehalose ligands were simulated. The side chains of Arg54, Arg25, Gln19, and Thr56 are important 

residues in the binding of a single trehalose dihydrate molecule to DAFP-1. Docking a single trehalose anhydrous 

molecule to DAFP-1 with GenDock revealed the backbone atoms of Asp42 as important in the protein and ligand 

binding along with the side chains of Arg54, Thr56, Ser32, and Gln19. A trehalose unit cell and DAFP-1 docking 

simulation suggested that the protein interacts with the ligand on the corner of the unit cell at the AB and BC faces 

through the side chains of Glu50 and Arg25. Arginine, threonine, and glutamine, along with a few backbone atoms, 

appear to be important residues in DAFP-1 for trehalose binding. Future research will include the addition of a 

solvent to the docking environment and exploring different carbohydrate and DAFP-1 binding interactions. 

117

Cryogenic Impact Testing of Tin (Sn) Solders and Bulk Metallic Glass Composites 

Adam R. Ullah 

Mentors: Marios Demetriou, Doug Hofmann, and Scott Roberts 

Providing a protective shielding for space-deployed technologies is crucial when trying to ensure multi-component 

function and communication at different positions in space, in orbit and in other planets. At the California Institute 

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, 

Nb7.60, Cu6.40, Be12.5) and Tin (Sn) were cryogenically tested on a Charpy Impact Tester to determine their 

mechanical performance on a temperature scale from (-186.5�C to +22�C). At these temperature ranges, DH3 was 

compared to the monolithic glasses (Vitreloy 1 and GHDT). The dislocation motion in the BCC-phase of DH3 

decreases at lower temperatures but does not perform worse than monolithic glasses Vitreloy 1 and GHDT. There is 

a ductile to brittle transition in the mechanical performance of DH3; however, the ductile to brittle transition is 

small in comparison to other contending BCC metals. DH3 is capable of protective cryogenic shielding. 

Strong and Tough Graphene-Polymer Composites 

Amy Wat 

Mentors: Julia R. Greer and Wendy Gu 

The strength and toughness of a graphene-polymer composite are optimized by emulating the nacre structure, 

which calls for graphene platelets dispersed in a polymer film. The graphene was grown on a copper rod using 

chemical vapor desposition and extracted off the copper rod by sonication. Raman Spectrometry reveals that a 

small flake of graphene was obtained. The surface of the copper rod was also etched off, but there is no indication 

that graphene was obtained. Graphene was also obtained through liquid exfoliation, which is sonication of graphite 

in a solvent followed by vacuum filtration. This created a mixture consisting of a low yield by weight of graphene 

platelets. The composites were created by spincoating a thin polymer film layer and then applying a graphite layer, 

which interfered with creating consistent uniform thin films. Composites were also made by applying graphite onto 

a polyisoprene disc, which was cut into four pieces and stacked on top of each other then flattened, which allows 

the creation of many thin bilayers of graphene and polymer each time the process is repeated. Through this 

project, we found the difficulty of mass producing graphene and how graphite affects the polymer strength and 

toughness. 

118

AMGEN 

AMGEN SCHOLARS PROGRAM 

A 

M 

G 

E 

N

Endospore Spore Viability Fractions in Ice Using Microscopic Endospore Viability Assay 

Judson G. Alderman 

Mentors: Adrian Ponce and Aaron Noell 

Spore forming bacteria are considered one of the hardiest forms of life. Such resilient bacteria could survive in 

Martian ice or travel aboard an icy comet. With reported lifetimes ranging from thousands to millions of years, no 

consensus exists concerning the longevity of bacterial endospores. A Microscopic Endospore Viability assay 

(MicroEVA), which functions by detecting a UV-excitable complex that forms when spore-specific dipicolinic acid 

(DPA) binds luminescent terbium ions, has been developed to investigate the true longevity of endospores. 

Measuring the total number of spores by killing them to release their DPA for imaging is crucial to discovering 

endospore longevity. In an evaluation of the MicroEVA’s new 280nm LED light source, we found that increased 

exposure times and image integration improved spore visualization. The new light source detects 40% more spores 

and exhibits a more even distribution of light intensity than the previous light source. Autoclaved total spore 

images suffer from physical disruption and uneven DPA distribution due to the action of steam. Conversely total 

spore images from spore samples ruptured in an oven have an average 2-fold increase in light intensity versus 

autoclaved samples and experience no physical disruption. Discerning true endospore longevity relies upon 

establishing a well-understood method of obtaining total endospore counts, allowing for the eventual determination 

of viability fractions in environmental samples. 

The Determination of the Dihedral Angles of 1,2-Disubstituted Ethanes in Aqueous Lyotropic Solutions 

Andrew J. Bendelsmith 


Two 1,2-disubstituted ethanes of biological interest, the succinate dianion and β-alanine, have been studied as 

oriented molecules in aqueous lyotropic solutions. To prepare the lyotropic solutions, myristyltrimethylammonium 

bromide (MTAB), decanol, and the molecule of interest were added to D2O. 1 H and 13 C NMR spectra were taken, 

yielding dipolar and scalar coupling constants. A MATLAB program was written, which was used to analyze the 

coupling constants to calculate the dihedral angle about the central carbon bond for each X-CH2-CH2-Y system. 

Characterization of Cation-π Binding Interactions Between Varenicline and Homopentameric α7 

Nicotinic Acetylcholine Receptors 

Emily Blythe 

Mentors: Dennis Dougherty and Ethan Van Arnam 

Nicotinic acetylcholine receptors (nAChRs) are ligand-gated ion channels that have been linked to numerous 

neurological syndromes, including Alzheimer’s disease, schizophrenia and addiction. Numerous subtypes of nAChRs 

exist, and subtype-selective pharmacology has been exploited in the design of pharmaceuticals to treat specific 

conditions. Varenicline, currently on the market as the smoking cessation drug Chantix, was developed to be a 

selective agonist at α4β2 receptors. However, the psychological side effects associated with varenicline suggest 

that it may be affecting receptors other than its target. In this study we focus on characterizing the binding of 

varenicline to homopentameric α7 nAChRs since varenicline is a full agonist at α7. Using unnatural amino acid 

mutagenesis, we probed the subunit interface “aromatic box” binding site to identify the residues that participate in 

cation-π binding interactions with varenicline. Surprisingly, our results indicate that the binding of varenicline more 

closely resembles that of acetylcholine than that of epibatidine, which, like varenicline, is a nicotine analogue. 

Traffic Jam piRNA Localization in Drosophila Ovarian Somatic Cells 

Yi Cai 

Mentors: Alexei Aravin and Edward Perkins 

Genomic information in the germline must be tightly regulated for accurate transmission to the next generation. 

Piwi-interacting RNA (piRNA), a class of 23-30 nucleotide small RNA, function in germline cells through RNA 

interference to silence retroelement activity and maintain genome integrity. Primary piRNAs are produced from 

specific genomic loci, piRNA clusters, and are further amplified in a secondary amplification loop. The primary 

piRNA production pathway is not yet clearly understood. Recently, traffic jam (tj) was identified as a new Piwiassociated 

piRNA cluster in Drosophila. One full-length tj transcript not only serves as a source of piRNAs but also 

encodes the TJ protein. TJ is strongly expressed in ovarian somatic cells (OSCs). To determine where primary 

piRNA processing occurs in the cell, an RNA tethering system has been used to detect tj transcript localization in 

OSCs. We have created a selectable construct containing MS2 hairpin sequences in the 3’ untranslated region of tj. 

A second imaging construct encodes an MS2 capsid protein (MCP) and fluorophore. Upon co-transfection in OSCs, 

the MCP will bind to the MS2 hairpins of the tj transcript. This system allows us to determine the location of tj 

transcript and primary piRNA processing. 

121

Identification and Characterization of Genes Involved in the Control of Stem Cell Populations in Plant 

Shoot Apical Meristems 

Margaret J. Chiu 

Mentors: Elliot Meyerowitz and Yun Zhou 

Normal growth of shoot meristem involves two opposinhg processes, differentiation and stem cell maintenance. In 

Arabidposis, stem cell populations in the shoot apical meristem (SAM) require the homodomain protein, WUSCHEL 

(WUS), to maintain their identity. The genes Clavata 1, 2, and 3 (CLV) are responsible for restricting the stem cell 

population by controlling differentiation. Previous research has shown that WUS and CLV interact to establish a 

feedback loop, although the exact mechanism is not understood. By creating a number of mutant lines for target 

genes expressed in SAM stem cells, we look to uncover the roles that these genes play in feedback between WUS 

and CLV. T-DNA insertional knock-out mutants and 35s overexpression mutants are confirmed using RT-PCR to 

ensure the decrease or increase in gene transcript. Plants are examined for their phenotypes using microscopy. 

Genetic crosses are also conducted to determine how these genes interact. 

Plasmid Design for an Assay to Study the Role of DNA-Mediated CT In-Vivo 

Keshia Dykes 

Mentor: Jacqueline K. Barton 

Base excision repair (BER) proteins repair damaged or mismatched bases in the DNA. BER proteins that contain 

[4Fe-4S] clusters are able to send and receive charge, which aids in locating damage. Endo III and MutY are DNA 

glycosylases, which are enzymes that specifically aid in the recognition of damage within the genome. There is an 

interest in finding proteins that help MutY search for damage. In this study, we have constructed a reporter (LacZ 

CC104) and control (LacZ MG1655) plasmids to examine MutY activity in vivo. The plasmids were designed as 

reporters utilizing the LacZ gene. If the LacZ gene (CC104) is mutated, it causes β-gal to be inactive. If the BER 

protein MutY is not functioning properly and unable to recognize the LacZ mutation, it will lead to transversion to of 

A:8-oxo-G to an A:T base pair. By using molecular biology and cloning techniques, we were able to successfully 

construct the reporter and control plasmids that can be used in in vivo studies with E.coil to examine how BER 

proteins search for damage. 

Improving Anti-Tumor Immunity by Engineering T Cell Receptors to Prevent Mispairing 

Marvin H. Gee 

Mentors: David Baltimore and Michael Bethune 

Anti-tumor immunity can be engineered in cancer patients by delivering the genes encoding the α and β chains of a 

tumor-specific T cell receptor (TCR). Exogenous α/β TCR chains introduced into patient T lymphocytes have 

potential to mispair with endogenous α/β TCR chains, resulting in reduced expression of tumor-specific TCRs and 

potential autoimmunity by mispaired TCRs. We are modifying tumor-specific α/β TCR chains to prevent their 

mispairing with wild-type chains as follows: 1) swapped domains between the α/β chains; 2) domains connected by 

glycine and serine linkers within α/β chains; or 3) domains substituted with γ and/or δ domains from γ/δ TCRs. 

When correctly paired, these modified α/β TCR chains retain all domains necessary to recruit CD3 components and 

express on the cell surface. When mispaired with wild-type chains, however, all domains needed for surface 

expression are not present. Using flow cytometry, 293T cells transfected with constructs are analyzed for cell 

surface expression by tetramerized peptide-MHC. Constructs containing both modified α/β chains simulate correctly 

paired TCRs and constructs encoding only one modified chain and one wild-type chain simulate mispaired TCRs. 

Several constructs appear to prevent mispairing while retaining function and warrant further investigation as viable 

candidates for gene therapy. 

Progress Toward the Enantioselective Synthesis of Variecolin: Synthesis of the B-Ring Precursor 

Angela M. Guerrero 

Mentors: Brian M. Stoltz and Kun-Liang Wu 

Variecolin is a tetracyclic potent anti-HIV-I natural product. An expeditious synthesis of the B-ring precursor has 

been achieved through a five-step process. These efforts highlight the formation of a pyrone from ethyl propenyl 

ether and the production of an iron complex utilizing photochemical technology. This B-ring precursor will be 

employed in the synthesis of the AB-ring fragment and will be ultimately used toward the total synthesis of 

variecolin. 

Lipid D/H Variations as a Possible Paleometabolic Proxy 

Yang Hu 

Mentors: Alex Sessions and Xinning Zhang 

Stable hydrogen-isotopic (D/H) variations have been shown to reflect the central metabolic pathways of bacteria. 

In addition, D/H content has been shown to persist in lipids of rock records on the order of millions of years 

allowing for the potential of D/H content in lipids to become a biomarker. An understanding of the mechanism by 

which D/H variations occur is necessary before these lipids can actually be reliably used as biomarkers. It is 

122

hypothesized that lipid D/H correlates to central metabolism and the process is determined by NADPH, the major 

reductant for cell anabolic processes. We seek to demonstrate that changes in NADPH metabolism directly affect 

lipid D/H using a laboratory culture approach. We used several strains of Escherichia coli with single gene 

mutations that perturb major NADPH-related reactions within glycolysis, TCA, or pentose-phosphate metabolic 

pathways. These strains were grown separately on the diastereomers glucose and galactose. Even though glucose 

and galactose have similar structural formulas, they are metabolized in dramatically different ways in E. coli. 

Metabolism by glucose increases activity in glycolysis than metabolism by galactose, which increases activity in the 

TCA cycle. We hypothesize that metabolism by galactose will yield lipids more enriched in deuterium than 

metabolism by glucose due to an isotope effect associated with TCA cycle activity. Our results support our 

hypothesis, and we observe the fatty acids of E. coli grown on galactose to contain around 10 per mille more 

deuterium than for E. coli grown on glucose. In conclusion we demonstrate that growth on substrates within the 

same macromolecular class yield fatty acids with different D/H ratios and that this is correlated with TCA cycle 

activity in a way that is consistent with previous studies, which show that an increase in TCA cycle activity yields 

heavier lipids. 

Regioselective Electrochemical C–H Activation 

Megan N. Jackson 

Mentors: Harry Gray and Alec Durrell 

The regioselective chlorination of benzoquinoline has been achieved electrocatalytically through binuclear palladium 

complexes. Through bulk electrolysis, the Pd II –Pd II species is oxidized at modest potentials and reacts with chloride 

to form a Cl–Pd III –Pd III –Cl intermediate. In the presence of excess benzoquinoline, the Cl–Pd III –Pd III –Cl species 

undergoes reductive elimination of 10-chlorobenzoquinoline, regenerating the initial Pd II –Pd II complex. The 

chlorinated product has been characterized by NMR and GC–MS. Additionally, the scope of these palladium 

electrocatalysts in C–H functionalization reactions has been explored. 

Investigating AND Gate Topologies to Aid in Proteomic Analysis 

Granton Jindal 

Mentors: David Tirrell and Alborz Mahdavi 

Proteins are vital cellular components that are used for diverse functions including mobility, signaling, and 

metabolism. Understanding protein expression levels and dynamics is critical to determining how cells function. 

Examining the proteome of cells which are in environments defined by multiple conditions (temperature, pH) can 

be done using a split MetRS-NLL system. The MetRS-NLL protein has been shown to be able to incorporate 

Azidonorleucine (ANL), a homolog of Methionine (Met) with an azide group. Although the split system of MetRS has 

been examined, the split system of MetRS-NLL has not been evaluated to date. When split at the same residue 

367 that the MetRS has been split at, it is expected for each part of the protein to have no activity, while gaining 

activity when both parts are expressed in the cell. This can be tested by seeing if ANL is incorporated using the 

Click-iT reaction with the TAMRA alkyne. From another library of split sites, the MetRS enzyme split at residue 

269 was determined to have significant activity. To quicken the self-assembly of the two parts of the MetRS-NLL, 

key amino acid mutations could be made in the protein such that the two parts are attracted together with greater 

thermodynamic stability. 

4D Gene Expression of S. purpuratus Throughout Embryogenesis 

Jennifer Kang 

Mentors: Eric Davidson and Emmanuel Faure 

The validated gene regulatory network (GRN) that encodes the specification functions of the sea urchin embryo (S. 

purpuratus) is approaching global dimensions in that it now includes all but one embryonic territory. The canonical 

cell lineage allows for the construction of a 4D (3D + time) digitally imaged embryo on which spatial patterns of 

regulatory gene expression are superimposed at different developmental stages. Mapping the gene regulatory 

networks involved in sea urchin embryogenesis allows researchers to determine the sequential processes of cell 

specification, and provides information on combinatorial interactions between different genes that may result in a 

particular cell fate. However, gene expression in the spherical embryo often presents a misleading depiction of the 

cellular processes that are going on when analyzed in 2D. We present a computational method permitting a 4D 

visualization of the GRN linkages responsible for any current regulatory state in the anatomical context of the 

embryo morphology. This is done by whole mount double fluorescence in situ hybridization on fixed embryos from 

6-40 hours post fertilization, as this time frame contains the gene activity of interest. The particular 

endomesodermal genes we are working on include FoxA, Hox11/13B, Blimp, GCM and Hnf6. The embryos will be 

analyzed using a confocal microscope and the images can then be stacked to created a 4D representation of gene 

expression during embryogenesis. 

123

Role of Extraneous Layer in Enhanced UV Resistance 

Jordan E. Krebs 

Mentors: Kasthuri Venkateswaran and Parag Vaishampayan 

Bacillus horneckiae, an extraneous layer-bearing spore-forming bacteria, was isolated from a clean room where 

Phoenix mission components were assembled. Previous studies demonstrated that B. horneckiae spores are 

resistant to UV radiation up to 1000 J m -2 . This extraneous layer may act as a passive shield enhancing higher UV 

resistance in these spores. The electron microscopy results showed that this layer enables the spores of this 

bacterium to attach to each other and formed biofilm-like monolayers when dried on spacecraft-qualified metals 

which would further shield the spores from UV radiation. To further characterize biochemically, procedures to purify 

the extraneous layer with a French Press were optimized. The survival of the spores with and without extraneous 

layer under UV irradiation condition was performed. Preliminary results showed that French Press treated spores 

exhibited greater survival against UV irradiation than untreated spores but more study is warranted. Furthermore, 

to elucidate whether the layer plays any part in the enhanced UV resistance in B. horneckiae spores, purified 

extraneous layer will be mixed with B. subtilis, that lack this layer. Determining what structures, biological byproducts, 

or molecular pathway enable great UV resistance is a key to developing strategies to eradicate 

extremophiles associated with spacecraft, thus potentially prevent forward contamination. 

Mapping the Neural Circuitry of the Drosophila Fly Brain 

Christine M. Mak 

Mentors: David J. Anderson and Allan Wong 

As a model organism, the fruit fly Drosophila melanogaster provides a powerful research tool to study the neural 

circuitry underlying behaviors exhibited by higher organisms, from simple reflexive behaviors to more complex 

social behaviors. Previous studies have established primary wind sensory neurons in the antenna that map to the 

fly brain. Here, we focus on identifying the second-order wind neurons and their functional role in the neural 

circuit. We express a genetically encoded calcium indicator, GCaMP3, which labels the antennal mechanosensory 

and motor center (AMMC) of the fly brain for neural activity during the delivery of a wind stimulus. Using calcium 

imaging under a two-photon microscope, we measure change in fluorescence corresponding to neuronal activation 

or suppression. Further examination of these neurons’ projections is done using a photo-activatible green 

fluorescent protein (PA-GFP), which allows for visualization of the neural circuit via protein diffusion. From a 

behavioral perspective, we also carry out a behavioral assay to test for the result of silencing these wind neurons. 

Normally, flies experience wind-induced suppression of locomotion (WISL), but inactivating the second-order wind 

neurons of interest should eliminate this phenomenon. Our results will contribute to an understanding of how 

neural circuits in the brain work to induce a specific behavioral response. 

Characterizing the T Cell Response to Cancer Immunotherapies 

Leanna Morinishi 

Mentors: James Heath, Gwen Owens, and Chao Ma 

Adoptive cell transfer (ACT) immunotherapy is the transfer of immune-derived cells to a host, with the intention of 

transferring immunologic characteristics. It has been used to treat patients with advanced tumors, including 

melanoma. Due to the high level of heterogeneity in functional immune cells, identifying the efficacy of the 

treatment is difficult, but made possible through the Single Cell Barcode Chip (SCBC) based on the DNA-encoded 

antibody library (DEAL) approach to spatial immuno-sandwich assays. A noted problem (through studies involving 

SCBC) with the ACT immunotherapy is the short-lived maximum tumor-killing function of the infused engineered T 

cells. Ongoing in vitro studies and work in mouse models in Dr. David Baltimore’s lab are focused on improving the 

acquired immunologic response. A new SCBC with an array of mouse DEAL conjugates is necessary to assess the 

progress of these experiments. This summer, I prepared a functional SCBC for mice secretion proteins for initial 

testing with hematopoietic mouse stem cells, and for eventual use in probing the efficacy of in vivo ACT 

experiments in mouse models. 

Searching for the Genes: Could New Species of Wild Microbes Provide Improved Alcohol Tolerance? 

Nathan D. Morison 

Mentor: Jared R. Leadbetter 

Using basic enrichment techniques, novel microbes tolerant to the important biofuel isobutanol were enriched from 

natural sources. The details of solvent tolerance to isobutanol, n-butanol, and other potential biofuels are 

discussed. Comparisons between aerobes and anaerobes isolated from similar environments through the creation 

of different environments are considered. Genetic information is provided where it is available. 

124

DNA Ejection Mechanisms of Bacteriophage Lambda 

Caitlin Regan 

Mentors: Rob Phillips and Yi-Ju Chen 

Bacteriophages have long provided a means for studying viral infection into bacteria. In this project, we aim to 

focus on the basic mechanism of infection: viral DNA ejection. By staining the viral DNA, we can watch the transfer 

of fluorescence from the viral capsid into an E. coli cell. In this single cell in vivo experiment we can quantify the 

fluorescence transfer, and thus have a metric for measuring DNA ejection. By adding a transcription inhibiting 

antibiotic, Rifampin, we aim to determine if transcription is involved in the viral DNA ejection mechanism. 

Improving the KM-Value for Isobutyraldehyde of Lactococcus lactis Alcohol Dehydrogenase Variant 

Ll_AdhA RE1-his6 in Yeast by Directed Evolution 

Tatyana E. Saleski 

Mentors: Frances H. Arnold and Sabine Bastian 

Isobutanol is well suited for replacement or supplementation of fossil fuels. A recently engineered high-yield, highspecificity 

pathway enables glucose to isobutanol conversion in a recombinant organism. For this pathway to be 

economically competitive with bioethanol production, it must be able to operate under anaerobic conditions. In 

order to resolve a cofactor imbalance that hindered anaerobic performance, the pathway’s alcohol dehydrogenase 

(ADH) was replaced with Lactococcus lactis ADH, AdhA, which was improved for its role in the pathway using 

directed evolution methods. The ADH’s substrate affinity is currently believed to be one of the limiting factors in the 

pathway’s efficiency. Here we have used directed evolution methods to engineer variants with lower substrate 

affinity. We have characterized variants in the lineage leading to our current best variant Ll_AdhA 29C8-his6 to 

investigate their substrate specificity change over the course of the evolution project. 

The Efficacy of Fe(II) Chelation Versus the Growth and Dispersal of Pseudomonas aeruginosa Biofilms 

Tahoura Samad 

Mentors: Dianne Newman and Ryan Hunter 

In cystic fibrosis patients, Pseudomonas aeruginosa is a significant cause of mortality. This is due, in part, to the 

ability of this bacterium to form biofilms –bacterial communities that are notoriously resistant to antibiotics. An 

attractive target to supplement current therapeutics is iron acquisition, which is essential to many aspects of 

bacterial metabolism. Recent studies have focused on use of Fe(III) chelators to reduce P. aeruginosa biofilm 

formation. However, P. aeruginosa is able to secrete small molecules known as phenazines which may change iron 

speciation, from predominately Fe(III) to Fe(II). We therefore wanted to test whether Fe(II) chelators can prevent 

the formation of biofilms and/or promote the dispersal of those already established. As a control, we first 

ascertained whether Fe(II) chelators prevented growth altogether. Consistent with prior experiments, the Fe(III) 

chelators lactoferrin, conalbumin, and Desferoxamine and Fe(II) chelators 2,2 dipyridyl and Ferrozine did not 

impact planktonic growth. Now, using a high-throughput biofilm growth, we are investigating whether the Fe(II) 

chelators described above can disperse and/or prevent the formation of P. aeruginosa biofilms. If so, they should 

be considered as a possible novel therapy that might be more effective than Fe(III) chelators at later stages of CF 

disease progression. 

Elucidating the Physiological Role of Mir-125b in Hematopoiesis and Immunity 

Nikita Sinha 

Mentor: David Baltimore 

MicroRNAs (miRNAs) have been identified as important regulators of gene expression, and they function by 

repressing specific target genes at the post-transcriptional level. MicroRNA-125b (miR-125b) is up-regulated in 

patients with leukemia. Preliminary data suggests that over-expression of miR-125b disrupts the homeostasis 

between myeloid and lymphoid immune cells. Here we used a miR-125b sponge loss-of-function approach to 

uncouple the role of miR-125b in tumorigenesis from its potential physiological function in hematopoiesis. For our 

loss of function system, we designed a high copy vector expressing several anti-sense microRNA sites that should 

act as a decoy, attracting the microRNA and keeping it away from its natural targets. We found that the 

reconstituted sponge-expressing mice had significantly less white blood cells, myeloid cells, granulocytes, B cells, T 

cells, and pre-erythrocytes. Overexpression of miR-125b target Lin28 in the mouse hematopoietic system 

mimicked the phenotype observed upon inhibiting miR-125b functions, leading to a decrease in hematopoietic 

development. Since endogenous expression of miR-125b targets Lin28, Trp53inp1, and BMF was de-repressed by 

the miR-125b sponge decoy in progenitor-enriched bone marrow, these pro-apoptotic genes may also collaborate 

with Lin28 inhibition during miR-125b tumorigenesis. In conclusion, we show here that miR-125b physiologically 

regulates hematopoiesis. 

125

Measuring Dynamic Phosphorylation of Spo0A in Bacillus subtilis 

Arvind Thiagarajan 

Mentors: Michael Elowitz and Joseph Levine 

Sporulation is an intricately controlled process in Bacillus subtilis, promising a potential wealth of novel network 

motifs. It has previously been shown that the key regulator involved in the induction of sporulation is the 

phosphorylated form of the transcription factor Spo0A. We have developed an experimental paradigm to measure 

Spo0A P levels dynamically in vivo. A strain of B. subtilis was engineered in which the Spo0A gene had been 

replaced by a Spo0A-GFP fusion and into which an array containing 256 copies of a Spo0A P binding site had been 

inserted. Imaging of this strain revealed clusters of localized fluorescence amidst diffuse background fluorescence, 

corresponding respectively to bound Spo0A P and unbound Spo0A (P) . In order to determine relative levels of Spo0A P 

and Spo0A from such images, a machine learning algorithm for identifying localized fluorescence and quantifying 

the corresponding intensities was developed. A probabilistic model was developed to estimate intracellular 

deviations in cluster intensities and consequently determine the ratio between observed fluorescence and number 

of molecules present. This paradigm will be used to study Spo0A P dynamics during sporulation, and so a media was 

optimized to induce sporulation. Finally, a microfluidic chemostat was optimized for the study of sporulation in B. 

subtilis. 

Establishing an Efficient Cell Sorting and Release Protocol 

Christine E. Wu 

Mentors: James R. Heath and Young Shik Shin 

Between each cell in human tumors, many genetic and epigenetic differences exist. Consequently, every cancer cell 

may respond differently to treatment, with some resistant to certain drugs. In order to fully study the drug 

resistant pathways in primary cancers, different cells must be sorted to be examined separately. The Heath Lab 

has developed the technique of DNA Encoded Antibody Libraries, which uses a chip-based microarray platform for 

cell sorting. Through this procedure, an antibody array is generated by hybridizing single-stranded DNA antibody 

conjugates onto slides patterned with complementary single-stranded DNA. The cell mixture is then applied to the 

antibody array and cells with surface markers of interest will adhere to the chip. While this technique has 

successfully isolated cells which overexpressed the epidermal growth factor receptor from primary glioblastoma 

biopsies, the releasing of the desired cells from the plate was met with some problems. With a combination of 

trypsin and DNase, coupled with a wash by pipetting, cells from a mouse model were successfully released from 

the slides and then applied to a single-cell barcode chip for protein detection. With the ability to release captured 

cells, isolating and studying certain cell types within a cancer can be easier to accomplish. 

Modeling of [FeFe]-Hydrogenase Using a Peptide Ligand 

Yuanchi (Victor) Zhao 

Mentors: Theodor Agapie, Kyle Horak, and Sandy Suseno 

Hydrogenases are enzymes that catalyze the interconversion of protons and dihydrogen using inorganic centers 

containing iron or nickel. The study of hydrogenases could inspire the development of novel catalysts based on 

earth-abundant, first-row transition metals capable of supplanting unsustainable platinum electrocatalysts currently 

used as a means of hydrogen production. One method for learning about the hydrogenases is through synthetic 

active site mimics. Work done towards assembling an [FeFe]-hydrogenase H-cluster mimic using a peptide scaffold 

is ongoing. Short peptides designed from conserved amino acid sequences of the iron-sulfur cluster-containing 

ferredoxin proteins were synthesized. Additionally, a diiron subsite mimic from the literature has been synthesized. 

Future work includes the incorporation of a [4Fe4S] cluster into the peptide scaffold. Linking the peptide-bound 

cluster to the diiron subsite mimic via a cysteinyl bridge will complete the [FeFe]-hydrogenase H-cluster mimic. 

Further characterization efforts include EPR and IR spectroscopy, electrospray mass spectrometry, and cyclic 

voltammetry. 

126

LIGO 

THE LASER INTERFEROMETER 

GRAVITATIONAL-WAVE OBSERVATORY 

L 

I 

G 

O

Crackling Noise in Blade Springs 

Vanessa Acon 

Mentors: Rana Adhikari, Tara Chalermsongsak, Alastair Heptonstall, Mingyuan Huang, and Seiji Kawamura 

Crackling noise arises in various physical systems, implying a nonlinear conversion of energy from slow changing 

external conditions into discrete 'crackling' events at higher frequencies. The mirrors in Advanced LIGO are 

suspended with a series of blade springs and mechanical flexures. At the sensitivity levels of Advanced LIGO, we 

are concerned with any noise in that suspension system, in this case with crackling noise in the blade springs. Our 

goal is to measure and characterize this crackling noise and determine its relevance to LIGO measurements. To do 

so, we construct a low-noise table-top Michelson interferometer with blade springs such that when they are driven 

in unison, any change in the output signal will be due to crackling. In order to extract a crackling signal on the 

order of the sensitivity of LIGO itself, we demodulate the signal using a "chopping" technique. Currently we have 

measured characteristic values for the blade springs, designed the low-noise Michelson apparatus, designed and 

simulated the chopping technique, and calculated an initial noise budget for the experiment. Subsequent work will 

be to continue to improve the low-noise design, complete the low-noise apparatus assembly, and collect crackling 

noise data. We can then derive the characteristic coefficients telling us the relationship between crackling and 

driving force, and compare our data with the current LIGO noise budget to see if crackling will pose a significant 

source of noise for LIGO measurements. 

Gravitational Radiation From Compact Binaries in the Massive Brans-Dicke Theory of Gravity 

Justin A. Alsing 

Mentors: Emanuele Berti and Yanbei Chen 

Most high-energy physics extensions to General Relativity predict the existence of scalar fields. For this reason, the 

investigation of gravitational radiation in scalar-tensor theories may provide important insights into high-energy 

modifications of Einstein's theory. 

We have computed the scalar and tensor contributions to the gravitational radiation from compact binary systems 

in the massive Brans-Dicke theory, and used recent observations of radiation damping in mixed binary systems to 

put stringent bounds on the parameters of the theory. The calculation divides neatly into four sections; the field 

equations of the massive Brans-Dicke theory are obtained, and linearized in the weak-field limit. Post-Newtonian 

expressions for the scalar and tensor fields are then derived from the resulting field equations. Using these results, 

the post-Newtonian equations of motion and the periastron shift of compact binaries are derived. The scalar and 

tensor gravitational waveforms are then obtained by solving the linearized field equations, and using these results 

together with those obtained in previous stages of the calculation we have derived an expression for the 

gravitational radiation damping of compact binary systems. Finally, with the result for the radiation damping in 

hand, we have used recent observations of radiation damping in mixed binary systems to put tight bounds on the 

parameters of the theory. 

Study of Fundamental Quantum Noise in Advanced Gravitational-Wave Detectors 

Anthony Bartolotta 

Mentors: Yanbei Chen and Haixing Miao 

Gravitational waves are weak phenomena; therefore, the effectiveness of laser interferometer gravitational wave 

detectors is very susceptible to noises present in the system. The ultimate limit on the effectiveness of the detector 

is noise produced by quantum effects. There are two primary sources of this quantum noise, the shot noise and the 

radiation pressure noise. Shot noise is decreased as the laser power is increased; however, this increases the 

radiation pressure noise. As such, these two sources of quantum noise, if not correlated, result in the Standard 

Quantum Limit. Advanced gravitational-wave detector schemes are designed to surpass the Standard Quantum 

Limit at their most sensitive detection band; of these proposed configurations, we have been considering the local 

readout scheme. The local readout scheme operates by measuring the local motion of the test mass with an 

auxiliary carrier light. The investigation of this scheme began with an analysis of Fabry-Perot cavities. This was 

expanded upon with an analysis of coupled Fabry-Perot cavities. This analysis was done by determining the 

transfer functions of each component and using MatLab to numerically determine the total transfer function. This 

provided the framework for analyzing the local readout detection scheme. 

Real-Time Decision-Making for Gravitational Wave Detection 

Christopher M. Biwer with Nick Fotopolous, Larry Price, and Leo Singer 

Mentors: Alan Weinstein and Roy Williams 

We have designed a real-time coincidence detector, a program that searches notifications of recent astrophysical 

observations for coincident events, that can be connected to the LIGO low-latency pipeline. The principle objectives 

were to determine time, and spatial models for gamma-ray bursts, and then implement methods to identify these 

coincidences in real-time using the existing advanced LIGO prototype infrastructure. The primary functions of the 

129

coincidence detector are to query remote databases, decide if any coincidences exist, and publish notifications; a 

modular design was implemented to allow easy manipulation, and addition of new models. Time coincidences are 

found by searching time windows that encompass the range of expected time delays between different 

messengers. Coincidences in space are determined by computing the likelihood ratio statistic for shared areas of 

the sky. Subsequent analysis of trial runs, and injected coincidences were conducted to evaluate performance. 

Properties of Neutron Stars, Their Pulsational Modes, and Gravitational Wave Emission 

Iryna Butsky 

Mentors: Christian Ott and Jeff Kaplan 

In this project, we investigate the different proposed equations of state (EOS) for neutron stars. We create mass - 

radius plots of the different EOS, along with constraints to find if any are inconsistent with theory or observation. 

Using the Spectral Einstein Code (SpEC), we evolve a spherical neutron star model to observe the oscillations in 

central density. Performing a Fourier transform on these oscillations allows us to extract the pulsational frequencies 

which are perturbed with a quadrupole perturbation. The theoretical f-mode frequencies are then obtained using 

perturbation theory. 

Ultra-Stable Fabry-Perot Cavities for Coating Thermal Noise Characterization 

Raphael Cervantes 

Mentors: Frank Seifert, Rana Adhikari, and Tara Chalemsongsak 

The performance of high-precision optical interferometric experiments is compromised by many noise sources. One 

of these noise sources is the coating thermal noise. The characterization of the coating thermal noise limit in 

various coatings could be used to verify the current theoretical model and to help minimize it for future 

experiments. Currently, the reference cavity noise experiment aims to use a small table-top setup with a fast turn 

around for measuring the coating thermal noise. The experiment consists locking a laser to a reference Fabry-Perot 

cavity using Pound-Drever-Hall locking. The same laser is sent to another cavity and the difference in frequencies 

between the output beams is measured, and thus the length fluctuations of the cavity are characterized. Although 

there are other noise sources, they will be minimized so that the coating thermal noise is dominant in the 

frequency region of interest. Cavity mirrors with different coatings can then be tested to characterize the 

fluctuations caused by various coating materials and/or designs. 

Third-Generation Gravitational Wave Detector Placement 

Michael Coughlin 

Mentor: Jan Harms 

One of the most immediate challenges associated with third-generation gravitational wave detectors is to select 

site candidates. There are numerous factors that must be taken into consideration, including surface topography, 

seismicity, population density and many more. The project analyzes a number of US-wide data sets, including 

seismic, wind, topography, and geology, as well as seismic data from around the world. The combination of these 

data sets indicates the suitability of possible site locations. Other analyses with the seismic data include tracking 

seismic noise levels at various frequencies over large periods of time as well as microseismic peak correlation in 

various locations. 

Probing Strong-Field Gravity: Constraining the Number of Gravitational Wave Polarizations 

Bryant Garcia 

Mentors: Larry Price and Stephen Privitera 

General Relativity predicts only two tensorial gravitational wave polarizations. More general theories, however, 

contain up to six independent polarizations. Observations of these four non-tensorial polarizations could severely 

constrain alternative theories of gravity and provide a unique strong-field test for GR. As the 2 nd generation 

gravitational wave detectors, advanced LIGO and VIRGO, come online, we expect to begin compiling a list of 

known GW signals. We present a model selection analysis for constraining the number polarizations present in the 

gravitational field. We investigate the possibility of detecting these alternate polarizations using the next 

generation of gravitational wave detectors. 

Investigation of Noise in Photodiodes Meant for a Gravitational Wave Interferometer 

Matthew S. Gilmer 

Mentors: Valery Frolov and Chris Guido 

This research focuses on the isolation of excess noise in photodiodes from other noise sources through noise 

cancellation techniques. The noise spectra from two photodiodes were simultaneously monitored; a beam splitter 

facilitated the investigation by allowing the same laser source to be incident on the two detectors at once. The 

research includes the use of a feedback system and several other noise reduction techniques. These were 

130

performed for many different photodiodes. The experiment was done in a clean room, the Optics Lab at the LIGO 

Livingston Observatory. Data was taken and analyzed in the context of the 4km Gravitational Wave Interferometer 

to be used in the upcoming project, Advanced LIGO. 

Spin Effects in the Gravitational-Wave Memory for Quasi-Circular Inspiralling Compact Objects 

Xinyi Guo 

Mentors: Marc Favata and Yanbei Chen 

The gravitational-wave memory effect is a time-varying but non-oscillatory contribution to the gravitational-wave 

amplitude. When a gravitational-wave with memory passes through an ideal detector, it will cause a permanent 

displacement of the test masses. The nonlinear form of the memory arises from the gravitational waves produced 

by previously emitted gravitational waves, and is present in virtually all gravitational-wave sources. It is also 

known to affect the gravitational-waveform at leading order despite the fact that it originates from higher-order 

interactions. Thus understanding the memory is important for building our accurate knowledge of the gravitationalwaveforms 

in order to probe the nonlinearity of general relativity. Previous calculations of the memory have only 

considered non-spinning binaries. However, most compact binaries have spinning components; these spins will 

significantly modulate the gravitational-wave amplitude and phase. We studied the effect on the memory from the 

spin-orbit interaction and calculated the corrections to the memory waveform through 1.5 post-Newtonian (PN) 

order. We found that the spin correction starts to enter at 1PN order and can contribute a ~20% correction to the 

memory. 

Measuring the Vibration Isolation of the Suspended Tip-Tilt Mirrors in the LIGO 40m Prototype 

Nicole L. Ing 

Mentors: Koji Arai, Rana Adhikari, and Jameson Graef Rollins 

The Tip-Tilt (TT) mirrors are used in the LIGO 40m prototype to geometrically fold the power recycling cavities. We 

will evaluate the effectiveness of the Tip-Tilt suspension in reducing vibration isolation by injecting horizontal and 

vertical motion and measuring the relative displacement between the mirror and the suspension frame. Since 

current plans for Advanced LIGO include the use of a similar type of Tip-Tilt suspensions, our evaluation of will be 

useful for the Advanced LIGO design. 

Advanced LIGO Output Mode Cleaner Piezo Actuator Noise 

Brian J. Koopman 

Mentors: Valera Frolov and Ryan DeRosa 

The Laser Interferometer Gravitational Wave Observatory (LIGO) in Livingston, Louisiana was built with the 

intention of one day detecting gravitational waves caused by cosmological events such as binary inspiral and 

supernovae. The interferometer is a 4km long Fabry-Perot Michelson interferometer. The readout of the 

interferometer is located in the Output Mode Cleaner (OMC), within which a piezo actuator is attached to one of the 

mirrors. The piezo actuator can introduce a new source of noise to the interferometer readout, a noise that we 

currently know little about. By utilizing a small Michelson interferometer with two piezo actuators, one in each arm, 

this source of noise can be studied. 

By driving the two piezo actuators in common mode, keeping the differential arm length the same, a peak in the 

amplitude spectral density can be created at 1Hz. This creates harmonic peaks at higher frequencies. When the 1 

Hz peak is suppressed by locking signal the harmonics remain, demonstrating a phenomenon known as up 

conversion. Understanding this upconversion will lead to understanding how much noise the piezo actuators 

present in the differential arm (DARM) readout. 

Testing General Relativity Using Gravitational Waves From Spinning Neutron Stars 

Christina Lee 

Mentor: Alan Weinstein 

The direct detection of a gravitational wave with the Advanced LIGO detectors provides the opportunity to measure 

departures from General Relativity. These departures can arise in the speed of the gravitational wave, existence of 

alternate polarizations, and parity violation. To measure these, I simulated a single detector measurement of a 

continuous gravitational wave from a well defined pulsar source, for example the Crab pulsar, due to an 

asymmetry in the moment of inertia. The Doppler frequency modulation allows determination of the speed to a 

part in 10 -6 for a signal with a sufficiently high signal to noise ratio. If the speed differs by much more than this 

value, the signal extraction technique no longer works. I also measured the ability to distinguish two additional 

polarizations, the breathing and longitudinal, from the standard plus and cross. If in propagation one polarization 

handedness is enhanced and the other suppressed due to parity violation it would result in an anomalous 

measurement of the inclination angle of the neutron star spin with respect to the line of sight, which is known from 

X-ray observations. 

131

Modeling the Advanced-LIGO Response to Gravitational Waves and Calibration 

Benjamin C. Li 

Mentor: Alan Weinstein 

The response of the Advanced LIGO detectors (currently under construction) to gravitational waves depends in a 

complex way on the optical configuration of the laser interferometer at the heart of the detector. This can change 

due to intentional and unintentional changes to the configuration parameters. Precise calibration of that response, 

required for optimal extraction of information about the gravitational wave source, will require us to carefully 

monitor and understand those configuration parameters. This project aims to identify the most important 

parameters and the best methods for measuring them. This will be done through careful modeling and simulation, 

using existing Matlab-based software. 

Investigating Gravitational Waves From the Ringdown of Accreting Black Holes 

Ryan C. Lynch 

Mentor: Gregory Mendell 

The characteristic strain produced by black hole ringdowns is well-known, taking the form of a damped sinusoid. 

The frequency of oscillation of this strain function is dependent upon the inverse of the source’s mass, and 

therefore it should decrease over time as the source black hole accretes mass. Currently, the data pipelines used to 

conduct match-filtered searches for ringdowns implement static-mass templates: templates with constant 

frequencies. Here, signal templates are designed with dynamically changing mass to account for mass accretion, 

allowing the expected frequency of oscillation in strain to vary over time. Simulations are also run to search for 

ringdowns over single and multiple detectors, using both static-mass and dynamic-mass templates. By comparing 

the results of these searches, it is clear that while mass accretion does make static-mass templates less efficient 

than dynamic-mass ones, this difference is only significant in cases of extremely high accretion rates that greatly 

exceed the Eddington limit. Thus, static-mass templates should be capable of detecting black hole ringdowns as 

they are currently understood. 

Identifying Correlated Detector Noise Contaminating Searches for Stochastic Gravitational Wave 

Backgrounds 

Gabriella Martini 

Mentors: Nick Fotopoulos and Alan Weinstein 

The Stochastic Gravitational Wave Background (SGWB) is the undetected gravitational radiation counterpart to the 

cosmic microwave background radiations (CMBR). The theoretically most sensitive SGWB searches involve colocated 

pairs of detectors, like the LIGO Hanford interferometers. Unfortunately these detectors are immersed in a 

common noisy environment which induces correlations in the data that can mimic that of an SGWB. Much of the 

correlation can be identified and removed using data by cross-correlating physical environment monitors (PEMs) 

such as seismometers, voltmeters, etc. with the gravitational wave channels. In the past, stochastic searches have 

employed crude approximations of the environmental contamination to avoid the computational heavy-lifting 

involved in the complete calculation. This project analyzes the full NxN cross-coherence matrix for the improved 

calculation of the environmental contribution to detector correlation. Preliminary results show that clustering 

algorithms identify a block structure amongst the PEM coherences and indicate the possibility of a significant 

reduction in computation. 

Optical Gyroscopes for the LIGO Gravitational Wave Detectors 

Zoe L. Masters 

Mentors: Rana Adhikari, Alastair Heptonstall, and W. Zach Korth 

Seismic isolation of the mirrors in the LIGO detectors can improve the detectors’ performance at low frequencies. 

Because seismometers confuse rotation with horizontal motion, however, optical gyroscopes will be added to the 

advanced LIGO detectors to measure rotation alone. The goal of this project is to improve electronics and reduce 

mechanical noise in a laser-based optical gyroscope operating on the Sagnac principle and locked using the Pound- 

Drever-Hall technique. Specific improvements include a new circuit board containing generic filters for use in the 

locking system, prototype tests and simulations of a possible thermal stabilization method, and tests of the 

gyroscope’s current functionality. 

An Optical Follower Servo for the LIGO Photon Calibrators 

Rolf Minton 

Mentors: Richard Savage and Paul Schwinnberg 

The LIGO Photon Calibrator currently uses a power modulated auxiliary laser to induce periodic displacements of 

the 40 kg end test masses on the order of 10 -18 m via radiation pressure. The optical follower servo reduces 

unwanted modulation harmonics, lowers relative power noise, and significantly increases the usable fraction of the 

auxiliary laser power. The response of the servo actuator, an acousto-optic modulator, saturates by a factor of 30 

132

over the range of injected laser power modulations desired. Modeling and lab measurements indicate that variable 

gain in the servo electronics can compensate for this actuator’s saturation. A prototype servo amplifier has been 

constructed and tested. The experimental results are presented and discussed. 

Diagnosis of the 40m Prototype Interferometer With Auxiliary Laser-Beam Injection 

Sonali Mohapatra 

Mentors: Rana Adhikari and Suresh Doravari 

The aim of this project is the deterministic locking and characterization of the arms of the LIGO 40m prototype 

interferometer. In order to obtain low-loss high-reflectance mirrors, LIGO uses dielectric coatings on the test 

masses. The Pound-Drever-Hall technique is employed to lock the laser to the cavity or vice-versa. The high 

sensitivity of the PDH lock limits its linear range which makes it difficult to bring the cavity into resonance with the 

laser. As a solution to this problem, this report describes an auxiliary-beat lock (green or IR) to give us a much 

broader linear range. This enables us to smoothly change the cavity length and thus lock the laser to the cavity by 

obtaining a resonance. The 532 nm green laser light is obtained by frequency doubling of the 1064nm NdYAG 

auxiliary laser. This frequency doubled green beam from the AUX laser has been previously used to obtain a greenbeat 

lock between the Pre-Stabilized-Laser and the AUX laser. We use the IR beam from the AUX laser, before it is 

frequency doubled and obtain an IR beat lock. The frequency noise of the IR-beat lock and the Green-Beat lock are 

compared. 

Exploration of the Latest Numeric-Relativity-Inspired Waveforms for Compact Binary Coalescence 

Jessie Otradovec 

Mentors: L. Santamaría, P. Ajith, and A. Weinstein 

Coalescing compact binaries are among the most promising sources of gravitational waves (GWs), which makes 

them a particularly interesting problem for GW astrophysics. The signature of these systems encodes a great deal 

of information about their parameters: masses, spins, orbital angular momentum and sky position. Although there 

is no analytically exact solution of such systems in full General Relativity, recent breakthroughs in Numerical 

Relativity have allowed the computation of handfuls of waveforms at the late stages of the binary coalescence 

(near the merger and ringdown) for multiple points in the parameter space. Together with post-Newtonian 

methods and black hole perturbation theory, it is now possible to construct waveforms for the full coalescence 

process. In turn, these waveforms are used in search algorithms and parameter estimation in GW data analysis. 

This study seeks to validate different waveform families by comparing the overlap, normalized difference in signalto-noise 

ratio (SNR) and chi-square. In addition, visual inspection for expected behavior is performed. Preliminary 

findings document differences in conventions, and ~6%-15% maximum disagreements in SNR. Additionally, worse 

agreement is found at smaller symmetric mass ratios. Some unexpected behavior of the generic-spin approximant 

family has also been discovered. 

A New Magnetar Gravitational Wave Search: GW Stacking Using Timing Information From 

Electromagnetic Light Curves 

Bryance Oyang 

Mentor: Peter Kalmus 

Soft gamma repeaters (SGR) are thought to be magnetars, young neutron stars with extremely strong magnetic 

fields (10 14 -10 15 Gauss). These objects sporadically release large bursts of soft gamma rays when energy in the 

magnetar's magnetic field is released by crustal deformation and fracture. This could excite non-radial modes (fmodes) 

on the magnetar, which would emit gravitational waves (GW). This project expands a database of bursts' 

light curves collected by satellites and uses the information collected to run a search for GWs in LIGO data. Since 

we expect the GWs and the gamma rays to arrive in our solar system at roughly the same time, we find the start 

times of each burst from the light curves. Then we run a “stacking” search for GWs in LIGO data, where we 

combine time-frequency tilings of LIGO data at the times of each burst for hundreds of bursts. Stacking 

dramatically improves the sensitivity for GWs under the assumption that the frequencies of the GWs are the same 

from burst to burst. 

Prototyping Adaptive Feed-Forward Seismic Noise Cancellation at the 40m Interferometer 

Ishwita Saikia 

Mentors: Rana Adhikari and Jenne Driggers 

Advanced LIGO is an upcoming upgrade to the LIGO Project, which we expect to be able to detect gravitational 

waves with frequencies as low as 10 Hz. At such low frequencies the detectors will encounter seismic noise, 

displacement noise which moves the mirrors (present in the detectors) and thus changes the path length. This 

noise is due to the ground motion that occurs by volcanic/seismic activity, ocean tides, human activities, etc. This 

project is intended to reduce seismic noise at the 40m Prototype Interferometer with the help of online adaptive 

filtering. The technology developed in the 40m laboratory will be applied in Advanced LIGO. 

133

Extracting Information About the Central Engine of Core Collapse Supernovae Using Gravitational Wave 

Signals 

James H. C. Scargill 


The complete and detailed mechanism which governs the core collapse in certain supernova explosions is unknown. 

Different possible mechanisms will cause different waveforms of gravitational radiation to be emitted, and thus 

information on the mechanism can be deduced from the received waveform. This project will develop existing ways 

in which this information can be extracted. Principal component analysis can be used to identify key waveform 

features from a library of known waveforms; Bayesian statistical techniques can then be used to analyse a new 

waveform (modelled as received by the detector, i.e. with noise) in terms of these principal components. Current 

algorithms can be improved by considering the coherent response of multiple detectors, which would give higher 

accuracy and durability in model selection and parameter estimation, as well as extra polarisation and sky-location 

information. This in turn will allow quantitative comparison to previous algorithms. Having developed such an 

algorithm, we will then test it on new waveforms to determine its efficacy in distinguishing the various mechanisms 

which govern core collapse, as the signal-to-noise ratio is reduced, in order to know what science could be done 

with gravitational waves in the event of a (galactic) supernova. 

Thermal Coupling in Cryogenic Fabry-Pérot Cavities 

Jennifer Schloss 

Mentor: David Yeaton-Massey 

The measurement of extremely weak signals such as gravitational waves requires highly sensitive detectors with 

exceptionally low noise levels. In an effort to reduce the contribution of thermal noise due to Brownian motion and 

fundamental temperature fluctuations, cryogenic Fabry-Pérot reference cavities are being built. To improve the 

temperature-stability of these cavities, it is useful to understand the couplings of outside temperature fluctuations 

to the cavity interior. Conductive and radiative heat transfer through cryogenic and ambient-temperature cavities 

was modeled using FEA simulation software. Room-temperature simulation output was compared with an 

equivalent experimental measurement. The level of agreement provides an indication of the extent to which the 

model cryogenic cavity simulations predict the thermal couplings in the physical cryogenic cavities. 

Directional Gravitational Wave Search 

Clio Sleator 


Ground-based observatories such as LIGO are expected to directly detect gravitational waves (GWs), which will 

begin the age of gravitational wave astronomy. The search methods depend on the astrophysical sources of GWs. 

Some sources (specifically violent events in the universe), such as core-collapse supernovae, neutron star collapse, 

merging compact binaries, star-quakes, pulsar glitches, and cosmic string cusps, are thought to make bursts of 

GWs that last less than 1s. As detector glitches look like burst signals, it is difficult to determine the difference 

between them. In this project, a search method will be developed that looks for sky directions with a statistical 

excess of triggers. Upon finding such a sky direction, it should be easier to determine whether the glitches seen in 

the data are only glitches or if they are actually GW signals. In order to find a sky direction with a statistical excess 

of triggers, we will have to calculate the sky direction of each event using time-of-flight triangulation and plot these 

sky directions onto skymaps. Using simulated signals will help to characterize the search. 

Optimal Telescope Tilings for Electromagnetic Followup of Gravitational Wave Events 

Antony J. Speranza 

Mentors: Larry R. Price and Leo Singer 

With the enhanced detection capabilities of Advanced LIGO, the prospect of multi-messenger astronomy has 

become an exciting means for investigating some of the most violent events in the universe. Gravitational wave 

detectors are inherently omnidirectional, while EM telescopes can image only small portions of the sky at one time. 

In order to make a joint observation of a GW event and an EM counterpart, we need an efficient method for 

determining EM telescope tilings that most effectively image the probable sky location of the source, as determined 

by the GW detector. For the single telescope case, this can be viewed as a convolution of the telescope's field of 

view with the source location skymap. We discuss an algorithm for efficiently computing tilings for a single 

telescope, and explore optimization techniques for quickly determining tilings that maximize the detection 

probability when multiple telescopes are involved. For both cases, we characterize the detection efficiency as a 

function of tiling resolution using a network of EM telescopes that will likely be available during upcoming scientific 

runs with aLIGO. The efficiency gains for the multiple telescope algorithms makes a case for coordinated tilings 

across the entire EM telescope network. 

134

Characterization of Noise in Driven Cantilever Blade Springs 

Larisa Thorne 

Mentors: Rana Adhikari and Alastair Heptonstall 

In the search for gravitational waves, detection is difficult to all but the most accurate and sensitive devices. 

Among the myriad of noises that could affect device readings is the “crackling” noise in driven blade springs, 

characterization of which is crucial to determining its status as a limiting factor in the shape of the LIGO noise 

curve, and ultimately detection and measurement of gravitational waves. 

Crackling noises are characteristic of systems which are subject to an outside driving force, and which respond in a 

series of discrete and abrupt jumps. Part of the crackling noise effect is due to some sort of hysteresis in the 

system. This is logical, considering that the excess energy lost due to hysteresis causes (nonlinear) energy 

upconversion in the blade springs which is transitioning from slow driving force applied into the crackling noise at 

higher frequencies. 

Simulations can only offer so much information. In order to comprehend the underpinnings of the crackling, an 

actual experiment must be run. A downscaled Michelson interferometer must be constructed to quantify the 

crackling noise. This noise should be revealed in the output signal (after it has been demodulated via a ‘chopping’ 

technique) when the cantilever blade springs are driven in unison by magnetic actuators, and the set-up has either 

been adjusted to its lowest noise level (we will use vacuum tanks and seismic isolation stacks to accomplish this) 

or all other sources of noise are known and accounted for. Once the crackling noise has been identified, it will be 

possible to determine if it will limit future gravitational wave detection in advanced LIGO (aLIGO). 

Study of PowerFlux Response to Continuous Wave Signals of Non-Standard Form and Locations of 

Potential Sources in the Milky Way Galaxy 

Gregory E. Vansuch 

Mentor: Vladimir Dergachev 

We present results on the robustness of the PowerFlux algorithm to injected signals with various parameters, in 

particular the speed and period of a gravitational wave, as well as the ability of the algorithm to reconstruct 

parameters from simulations. We will also describe potential locations of continuous gravitational radiation from 

various star clusters and the arms of the Milky Way Galaxy. 

Implementing Digital Control for Quad-Maglev Suspension 

Yi Xie 

Mentors: Haixing Miao and Rana Adhikari 

Low-frequency seismic noise prevents us from detecting gravitational waves from many interesting astrophysical 

sources. The maglev suspension system is decided to create a soft suspension by using a specially designed 

magnetic field. And the digital control system produces feedback control force to make the balance come true. 

135

NASA/JPL Programs 

PGGURP 

Planetary Geology and Geophysics 

Undergraduate Research Program 

SPACE GRANT 

The National Space Grant College 

and Fellowship Program 

USRP 

Undergraduate Student Research Program 

JPLSIP 

JPL Summer Internship Program 

N 

A 

S 

A

Transition From Conventional Software Applications to a Web Application Based Platform 

Stirling S. Algermissen 

Mentors: Maher Hanna, Douglass Alexander, and Helen Mortensen 

The Multimission Instrument Processing Lab (MIPL) uses a variety of tools across multiple platforms, languages, 

and systems. This diversification is a burden for MIPL and creates overhead when switching between systems for 

different tools. A more accessible web interface was explored for decreasing this overhead and making MIPL’s 

software tools more versatile and more easily accessible. A website was created to allow the MIPL team to meet 

these goals. The marsviewer software, used for viewing image data products produced by the Mars Science 

Laboratory and other Mars rovers, is converted to a client-server AJAX interface. This allows access to the tool from 

any modern web browser. Other aspects of MIPL’s operation were also moved to the web, such as coordination 

with other teams and access to internal use software. The results of this transition are explored, along with a 

demonstration of the move of conventional software to the web browser. 

Programming Tools for MSL Flight Software Testing 

Daniel Alkalai 

Mentor: Jim Chase 

I am programming tools used to help test the MSL Flight Software as part of the extensive Verification and 

Validation process under the cross-cutting subdivision, which deals with system-wide testing concerns rather than 

with specific components. Includes an interface allowing users to query the MPCS ground database for operations 

and testbed session information, using a configuration file to keep track of important information for the user, as 

well as a script that parses through session information and returns statistics about the Flight Software commands 

run. The interface to MPCS data would be especially helpful during actual mission operations, so it would still be 

used after Flight Software testing has finished. 

Maintaining Hardware and Software Cohesiveness for the Cassini Program 

Robert Alland 

Mentor: Diane Conner 

To successfully perform its Solstice Mission until 2017, the Cassini program requires its hardware and software 

configurations to match an expected state. Of interest to this project are two areas of system cohesiveness: the 

configuration of computers that are part of the Cassini team’s Ground Data System (GDS) and the parameter and 

constraint checks performed by the Sequence Generator tool (SEQGEN), which validates and models a proposed 

sequence of commands for the spacecraft. The first objective of this project is an adaptation of the Online Asset 

software from the JPL Multi Mission Office to automatically produce audits of system configurations – both 

hardware and software – and a web interface for viewing and analyzing of the audits. This assists the Cassini team 

in satisfying requirements for the JPL IT Security Database and allows it to more easily manage and diagnose 

problems with its computers. The second objective of this project is a correction of command checking for a subset 

of Attitude and Articulation Control Subsystem (AACS) commands in SEQGEN. This update permits uplinking the 

full range of allowable commands to Cassini and prevents commanding errors. This project illustrates the difficulty 

and importance in maintaining system cohesiveness for the Cassini program. 

Simulations and GUIs: A New Look to Radio Science 

Michael Andonian and Carlos Gonzalez 

Mentors: Sami Asmar and Kamal Oudrhiri 

Radio waves constitute a large portion of the EM spectrum and as such are significant to the process of acquiring 

knowledge of the vast universe we live in. The Radio Science Systems Group (RSSG) devotes itself to the analysis 

of radio wave phenomena and developing analysis tools to assist. At the present, the RSSG is developing a 

command line demodulation tool for processing and analyzing spacecraft signal data from an open-loop radio 

science receiver (RSR) while continuing its interest in signal interference amongst other tasks. However, watching 

the command line analyze data or drawing figures on a whiteboard to convey notions tends to be monotonous at 

best. Surely any user would prefer a graphical user interface (GUI) to work with for processing data or a visual 

simulation effectively showing communication between satellites and ground stations. Within this presentation will 

be the story of transitioning from analysis in the command line to a graphical user interface, creating simulations 

for a radio science experiment involving the International Space Station (ISS) and the Tracking and Data Relay 

Satellite System (TDRSS) to form a three-way communication loop, and operating radio science tools, in particular 

performing a weekly "check-up" on the RSR of New Horizons. 

139

Mars Science Laboratory Wiki 

Laura Austin 

Mentors: Alicia Allbaugh and David Mohr 

The Mars Science Laboratory is currently planned to launch at the end of 2011. A wiki has been selected as a 

repository of sorts to store information about procedures, Operational Interface Agreements (OIAs), etc. and links 

to the more secure documents stored elsewhere. This permits the editing of information in one place and every 

person being able to see a change, without the resending of documents. My goal is to help implement the wiki to 

meet the user’s requirements. I have met with various members of some of the different elements in order to 

understand their user requirements. This has led to the creation of a template and format for surface procedures 

and OIAs. Most users want to be able to edit information in one place and have it update all across the wiki. While 

this is possible, it is only achievable at the individual procedure or OIA level. With the design of the wiki, there has 

to be a tradeoff with the users deciding the optimum benefits versus cost analysis, which will influence the future 

direction of the wiki. The surface section of the wiki is being developed based upon the lessons learned from the 

cruise implementation. 

Thermal Conductivity Measurements on Icy Compositions 

Jaclyn A. Avidon 

Mentors: Martin Barmatz, Mathieu Choukroun, and Fang Zhong 

It is widely believed that many of the satellites in the outer solar system, such as Europa, have a liquid ocean 

surrounding their core, covered by an icy crust. Infrared spectral data returned by the Galileo and Cassini missions 

indicate that these crusts are made of icy compositions of salt hydrates, clathrate hydrates, and ammonia water. 

Models of the evolution of these satellites need accurate data on the thermal conductivity of the crustal materials, 

since these icy impurities depress the melting point of ice and alter the temperature dependence of the ice’s 

thermal conductivity. My project is to support the measurement of the thermal conductivity of these icy 

compositions. I will present my involvement in (1) preparing a polycrystalline water ice sample for measurement in 

a new cryogenic thermal conductivity probe and (2) calibrating the new probe using a quartz sample of known 

thermal conductivity. 

Analyzing the Necessary Power to Discover Extra-Terrestrial Habitability 

Kimberly Baird 

Mentor: Ray Crum 

Are we alone? Are there any galaxies, planets (besides Earth), or moons that hold characteristics of life? Many 

scientists suspect Jupiter’s moon Europa has high probability to hold the necessities to sustain extra-terrestrial life. 

NASA is determined to discover life outside of earth. With the help of Applied Physics Laboratory (ALP) and NASA’s 

own Jet Propulsion Laboratory (JPL), the mission to explore the surface and subsurface of Europa is being defined. 

With all the instruments to discover the attributes of Europa needing power, the power system is vital to the 

exploration of Europa, but what power source is the pivotal one? Analyzing and modeling of the architecture helps 

determine the best preference for the system. With the different advanced power systems, each have their pros 

and cons and with the mission still in the definition stage, there is no set load. To make a quick and reliable 

decision on the perfect power generation, as soon as the mission has been determined, a trade study of the 

different power sources is made. 

Mars Science Laboratory Flight Software Testing: Wake Cycle Robustness 

Payam Banazadeh 

Mentor: Danny Lam 

Mars Science Laboratory(MSL) is one of the most complex spacecrafts in the history of mankind. Due to the nature 

of its complexity, a large number of flight software requirements have been written for implementation . In practice, 

these requirements necessitate very complex and very precise flight software with no room for error. One of flight 

software’s responsibility is to be able to boot up and check the state of all devices on the spacecraft after the wake 

up process. This boot up and initialization is crucial to the mission since any misbehavior of different devices needs 

to be handled through the flight software. This summer, I am working with two other interns to exhaustively test 

the flight software under variety of different unexpected scenarios and validate that flight software can handle any 

situation after booting up. The test includes initializing different devices on spacecraft to different configurations 

and confirm at the end of the flight software boot up that the flight software has initialized those devices to what 

they are suppose to be in that particular scenario. 

140

Improving Single Event Effects Testing Through Software 

Mackenzie Banker 

Mentors: Steve McClure, Larry Edmonds, Philippe Adell, and Greg Allen 

Radiation encountered in space environments can be damaging to microelectronics and potentially cause spacecraft 

failure. Single event effects (SEE) are a type of radiation effect that occur when an ion strikes a device. Single 

event gate rupture (SEGR) is a type of SEE that can cause failure in power transistors. Unlike other SEE rates in 

which a constant linear energy transfer (LET) can be used, SEGR rates sometimes require a non-uniform LET to be 

used to be accurate. A recent analysis shows that SEGR rates are most easily calculated when the environment is 

described as a stopping rate per unit volume for each ion species. Stopping rates in silicon for pertinent ions were 

calculated using the Stopping and Range of Ions in Matter (SRIM) software and CRÈME-MC software. A reference 

table was generated and can be used by others to calculate SEGR rates for a candidate device. Additionally, lasers 

can be used to simulate SEEs, providing more control and information at lower cost than heavy ion testing. The 

electron/hole pair generation rate from a laser pulse in a semiconductor can be related to the LET of an ion. 

MATLAB was used to generate a plot to easily make this comparison. 

NASA’s Advanced Multimission Operations System: A Case Study in Formalizing Software Architecture 

Evolution 

Jeffrey M. Barnes 

Mentors: Brian Giovannoni and Oleg Sindiy 

All software systems of significant size and longevity eventually undergo changes to their basic architectural 

structure. Such changes may be prompted by evolving requirements, changing technology, or other reasons. 

Whatever the cause, software architecture evolution is commonplace in real-world software projects. Recently, 

software architecture researchers have begun to study this phenomenon in depth. However, this work has suffered 

from problems of validation; research in this area has tended to make heavy use of toy examples and hypothetical 

scenarios and has not been well supported by real-world examples. To help address this problem, I describe an 

ongoing effort at the Jet Propulsion Laboratory to re-architect the Advanced Multimission Operations System 

(AMMOS), which is used to operate NASA’s deep-space and astrophysics missions. Based on examination of project 

documents and interviews with project personnel, I describe the goals and approach of this evolution effort and 

then present models that capture some of the key architectural changes. Finally, I demonstrate how approaches 

and formal methods from my previous research in architecture evolution may be applied to this evolution, while 

using languages and tools already in place at the Jet Propulsion Laboratory. 

Modeling the Structure of Venusian Rift Systems Through Gravity Data Analysis 

Annabelle Batista 

Mentor: Suzanne Smrekar 

The Magellan Spacecraft (1990-1994) has provided scientists with high-resolution maps of Venus' topography and 

gravitational field. Magellan radar imagery of the planet reveals a surface marked by chains of rift systems 

(chasmata) stretching thousands of kilometers. The formation of these chasmata is puzzling, as there is no obvious 

geologic process that would cause extensional stresses. Venus lacks plate tectonics, the mechanism driving 

complimentary extensional and compressional processes on Earth. To better understand chasmata formation, we 

examine the lithospheric structure at four rift systems: Dali/Diana Chasmata System, Ganis Chasmata System, 

Perunitsa and Khosedem Fossae, and Devana Chasma. The four rifts vary in location, size, and geologic setting. We 

create maps of the chasmatas' crustal and elastic lithosphere thickness by modeling the admittance function for 

gravity and topography. This method matches observed admittance variations across each chasma system to 

elastic compensation top- and bottom-loading models of predicted admittance. This gravity analysis may result in 

better constraints on the internal structure of chasmata, and may provide insight into the tectonic processes 

responsible for chasmata formation. 

Photochemical Studies of N2/CH4 Ice and Formation of Radicals 

Sky Beard 

Mentors: Paul Johnson and Robert Hodyss 

Photochemical studies of cryogenic ice films analogous to the icy surfaces of Triton and Pluto were performed to 

simulate solar photochemistry. N2/ CH4 ices were exposed to a hydrogen discharge lamp to provide an intense 

source of ultraviolet light in an ultra high vacuum chamber, and the chemical evolution of the reactants over time 

was monitored with infrared spectroscopy. One product of interest, ethane, was formed from photochemically 

generated methyl radicals: · CH3 + · CH3 → C2H6. Experiments were repeated at different temperatures from 

14- 30 K, and the reaction rate as a function of temperature and activation energy of diffusion for methyl radicals 

in N2 ice was determined. Spectroscopic studies of the formyl radical, HCO, in N2 ice in the visible and IR were 

performed and the rate of destruction of the formyl radical under visible light was determined. 

141

Correlations Between Technical Performance and Cost Growth: A Case Study Using Technical 

Complexity 

Keith Becker 

Mentors: Kevin Rice, Kevin Endo, Eric Kwan, and John Jack 

Flight Projects historically exceed the available budgetary funds allocated by NASA HQ. This group research project 

investigates historical data from the 13 most recent JPL Flight Projects to determine trends and correlations 

between technical and programmatic parameters to predict cost growth. Each of the 13 projects is analyzed using 

5 major data sets: technical complexity, subcontractor performance, risk trends, margin analysis and schedule slip. 

My focus is the technical complexity data set. Activities in this analysis include: pointing accuracy, instrument count, 

redundancy level and mission life, along with 52 additional parameters. Once all technical and cost data is collected 

a numerical complexity matrix is created based on technical performance, followed by regression analysis and 

modeling to establish trends. It is expected that certain aspects of technical performance parameters will lend 

themselves to be more accurate predictors of cost growth than others. Conclusions derived from this research can 

be utilized to form a historical data base for future projects and compared against the current in-work flight 

projects to improve cost growth predictions. 

Software Analysis of an Automatic Sequence Processor 

Brandon Benjamin 

Mentor: Barbara Streiffert 

Jet Propulsion Laboratory’s planning and sequencing element provides space missions with multi-purpose software 

to plan spacecraft activities, sequence spacecraft commands, and then integrate these products and execute them 

on spacecraft. Currently, Jet Propulsion Laboratory is flying many missions and the processes for building, 

integrating, and testing the multi-mission planning and sequencing software is being revitalized to meet the needs 

of current and future missions, as well as the operations teams that command the spacecraft. A special team is 

responsible for collecting and processing the observations, experiments and engineering activities that are to be 

performed on a selected spacecraft. The collection of these activities, ultimately, becomes a sequence of spacecraft 

commands. There are commands that are guaranteed not to harm the spacecraft. As such, these commands may 

be processed automatically and without intervention by an operations team member. This task is identifying, 

understanding, mapping, and documenting the nature of the scripts responsible for these automated commands in 

order to facilitate the future missions operating on the revitalized software. 

Venusian Applications of 3D Convection Modeling 

Timary (Annie) Bonaccorso 

Mentor: Christophe Sotin 

This study models mantle convection on Venus using the 'cubed sphere' code OEDIPUS, which models one-sixth of 

the planet in spherical geometry. We are attempting to balance internal heating, bottom mantle viscosity, and 

temperature difference across Venus' mantle, in order to create a realistic model that matches with current 

planetary observations. This summer's central question is how to define a mantle plume. Traditionally, we have 

defined a hot plume as any contour with at least 40% of the maximum temperature, and a cold plume as any 

contour with 40% of the minimum. For less viscous cases (10 20 Pa-s), the plumes generated by that definition 

lacked vigor, displaying buoyancies 1/100 th of those found in previous simulations. As the mantle plumes with large 

flux are most likely to produce topographic uplift and volcanism, most of the low viscosity cases' plumes exhibit no 

surface expression. In an effort to eliminate the smallest plumes, we experimented with different heat flux 

parameters and temperature percentages. We plan to run both lower and upper mantle simulations to determine 

whether layered (as opposed to whole-mantle) convection might be a more appropriate model. Upper mantle 

simulations will be completed using OEDIPUS' Cartesian counterpart, JOCASTA. 

Model-Based Systems Engineering Approach for Capturing Mission Architecture Systems Processes 

With Application Case Study Problem: Orion Flight Test 1 

Kevin H. Bonanne 

Mentors: Oleg V. Sindiy and Brian J. Giovannoni 

Model-based Systems Engineering (MBSE) is an emerging methodology that can be leveraged to enhance many 

system development processes. MBSE allows for the centralization of an architecture description that would 

otherwise be stored in various locations and formats, thus simplifying communication among the project 

stakeholders, inducing commonality in representation, and expediting report generation. This paper outlines the 

MBSE approach taken to capture the processes of two different, but related, architectures by employing the 

Systems Modeling Language (SysML) as a standard for architecture description and the modeling tool MagicDraw. 

The overarching goal of this study was to demonstrate the effectiveness of MBSE as a means of capturing and 

designing a mission systems architecture. The first portion of the project focused on capturing the necessary 

system engineering activities that occur when designing, developing, and deploying a mission systems architecture 

142

for a space mission. The second part applies activities from the first to an application problem—the system 

engineering of the Orion Flight Test 1 (OFT-1) End-to-End Information System (EEIS). By modeling the activities 

required to create a space mission architecture and then implementing those activities in an application problem, 

the utility of MBSE as an approach to systems engineering can be realized. 

Magnetic Coolant Characterization 

Shannon Colleen Bourke 

Mentors: Warren Holmes and Talso Chui 

Infrared spectroscopy is a powerful tool that enables astrophysicists to determine the distance to other galaxies in 

the universe, as well as their components. Space borne infrared spectrometers for astrophysics require detectors 

that are cooled to temperatures lower than 1K to decrease the noise and increase sensitivity. A common way to 

achieve sub-Kelvin temperatures is an adiabatic demagnetization refrigerator (ADR). Future missions that will use 

an ADR include the Background Limited Infrared and Submillimeter Spectrometer (BLISS) that will fly on the 

Japanese space telescope SPICA. ADRs use a magnetic coolant, or salt pill, suspended in the bore of a 

superconducting magnet. The salt pill that we are studying is chromic cesium alum (CCA). We characterize the CCA 

by measuring its heat capacity as a function of magnetic field and temperature using the heat pulse technique. We 

also measure the thermal resistance between the point where the instrument is attached and the CCA. Our two 

CCA salt pills routinely achieve base temperatures of 20-30mK, and can therefore maintain BLISS at 50mK as 

required. 

California Precipitation Trends: Natural Variability Overwhelms Climate Change Trends 

Ann Bui 

Mentors: William Patzert and Joshua Willis 

Precipitation trends in California from approximately 1900 to present were identified in this study. In our analyses 

of 14 long-term precipitation records representing multiple climate types throughout the state, we have found 

distinct regional and seasonal trends. In the North, annual rainfall has increased over the time series, while 

Southern California stations show little or no trend at all. Seasonally, winter precipitation shows increases and 

higher frequencies while spring, fall and summer display little gain in the long term. However, these trends did not 

conform to accepted measures of significance because of high degree in variability. By quantifying the influence of 

the Pacific Decadal Oscillation (PDO), and El Nino- Southern Oscillation (ENSO) on the annual variability in 

precipitation, we also found that the coefficient of variability is increasing with time, making it more difficult to 

distinguish a trend from the data. In general we have found that precipitation trends in California are not as 

extreme as previously hypothesized by models at a global level, and because of high inter-annual variability, more 

data are needed for the trends to be significant. 

Innovations in Robotic Code 

James Burdick 

Mentors: Michael McHenry and Norm Aisen 

Creating comprehensive controlling computer code for robotic hardware is a challenge. Most robotics software is 

handwritten in lower level languages, like C or Fortran, which functions very well in platforms due to the simpler 

nature of these languages and the need for computational efficiency. However, the creation of this code is 

expensive in terms of time and labor; the situation suggests a more innovative approach. 

Higher level programming languages provide the possibility of creating the equivalent of reams of code, while 

simplifying the software creation process. Matlab-based Simulink and associated Stateflow products are well-suited 

for robotic controls. Simulink provides a continuous computing framework, while Stateflow is designed to aid the 

development of state-machine based software, both of which are valuable in any robotics control system. Stateflow 

also can create embeddable c-code. While handwritten c-code functions in a linear manner, in robotics disparate 

events can be introduced into the system at unpredictable times. The research conducted in this internship was 

directed towards understanding and demonstrating the use of MATLAB-based tools to generate such embeddable 

c-code, and to interface this code with external handwritten code. It is clear that code generated from these 

MATLAB-based programs can help to simplify the code-making process in robotics. 

Data Analysis and Archiving for Mars Exploration Rovers (MER) 

William J. Canan 

Mentors: David Mohr and Matthew Keuneke 

The Mars Exploration Rovers (MER) tactical team relies heavily on modeling software to determine the most 

effective way to explore the Martian environment. MER attempts to make the most of every minute of operation 

time and of every data byte they can downlink to Earth. The models used to make predictions of imaging 

acquisition duration and the amount of data volume available to downlink on a UHF pass are known to have 

inaccuracies. By going through the daily reports of the rover operations en masse using Python scripts, the 

143

inaccuracy of these models has been quantified. This information allows the team to make more informed decisions 

regarding the use of modeling results or to introduce refinements to the modeling software based on the collected 

data. The presentation focuses on the methodology used to determine the accuracy of the models. 

Computing Temperatures in Optically Thick Protoplanetary Disks 

Lawrence Capuder 

Mentor: Neal Turner 

I am working with a Monte Carlo radiative transfer code to simulate the transfer of energy through protoplanetary 

disks, where planet formation occurs. The code tracks photons from the star into the disk, through scattering, 

absorption and re-emission, until they escape to infinity. High optical depths in the disk interior dominate the 

computation time because it takes the photon packet many interactions to get out of the region. High optical 

depths also receive few photons and therefore do not have well-estimated temperatures. I will apply two methods 

for treating high optical depths, a modified random walk (MRW) and a partial diffusion approximation (PDA), to 

speed up the Monte Carlo calculations. The MRW is implemented by calculating the average number of interactions 

the photon packet will undergo in diffusing within a single cell of the spatial grid and then updating the packet 

position, packet frequencies, and local radiation absorption rate appropriately. The PDA is implemented by applying 

the time independent diffusion equation to a large region after each iteration of the code. The approximations will 

be tested for accuracy and speed compared to the original code. 

Testing of Fusing Sonar and Video Images in Automated Target Detection and Tracking 

Jeffrey Chiang 

Mentor: Thomas Lu 

Assisted Target Recognition (ATR) systems aim to automate certain aspects of surveillance tasks, with the end goal 

of fully automating target detection, recognition, and tracking. The current project applies this system to low 

resolution SONAR and camera videos taken from Unmanned Underwater Vehicles (UUVs). These SONAR images 

are inherently noisy and difficult to interpret, and pictures taken underwater are unreliable due to murkiness and 

inconsistent lighting. The ATR system breaks target recognition into two stages: 1) Videos of both SONAR and 

camera footage are broken into frames and preprocessed in the Fourier domain to enhance images. 2) Regions of 

Interest (ROIs) are determined by evaluating the processed images, and characteristic features are extracted. 

3) ROIs are then classified as true or false positives using a standard Neural Network based on the extracted 

features. Several preprocessing methods are explored, as well as integrating data from both images in order to 

maximize the efficiency of the program. 

Characteristics of the Cathode and Plume of a High Power Hall Thruster 

Emily Chu 

Mentor: Dan Goebel 

The plasma dynamics in the plume region of high current hollow cathodes and Hall thrusters for high power use are 

complex and the success of modeling efforts designed to identify the fundamental mechanisms of erosion and life 

of these devices are dependent on the support of experiments that provide plasma measurements in the nearthruster 

plume region for model validation. In this study, we use Langmuir and emissive probes to investigate the 

plasma parameters inside and in the near-plume of a 200 A hollow cathode and in the near-field plume of a 

6-to-10 kW Hall thruster that has been modified to reduce erosion rates and increase the power level to 10 kW. 

The cathode and thruster are run at operating conditions associated with their respective throttle tables and we 

measure the resulting plasma properties to determine plasma density, potential, and electron temperature profiles 

in front of the thruster. This data will be used to provide benchmarking data to thruster modelers for code 

predictions and life model validation. 

Analysis of the Instrument Management Layer of the MSL Flight Software 

Louis M. Cialdella 

Mentor: Marcel Schoppers 

The control of science instruments on the upcoming Mars Science Laboratory mission is implemented via the 

"Instrument Management Layer" (IML), an abstraction layer that handles transmission and reception of commands 

to and from instruments. After several years of ongoing development and numerous structural changes, the IML is 

functional but suboptimal in many ways – it is convoluted, poorly documented, and contains redundant states and 

code. This combination of things makes the code both error-prone and difficult to debug. By analyzing the current 

version of the abstraction layer, we will be able to get a concrete idea of how precisely it works. In particular, we 

will be examining the finite state machine that underlies the current implementation, allowing us an intuitive and 

powerful view of the way in which the software functions. After this has been done, methods for restructuring the 

state machine (ie, reductions by eliminating unnecessary states, merging similar states) will be considered in order 

to see how a restructuring could improve the clarity and integrity of the software. 

144

An Analysis of Database Replication Technologies With Regard to Deep Space Network Application 

Requirements 

Andrea M. Connell 

Mentors: Paul Wolgast and Silvino Zendejas 

The Deep Space Network (DSN) has three communication facilities which handle telemetry, commands, and other 

data relating to spacecraft missions. The network requires these three sites to share data with each other and with 

the Jet Propulsion Laboratory for processing and distribution. Many database management systems have 

replication capabilities built in, which means that data updates made at one location will be automatically 

propagated to other locations. This project examines multiple replication solutions, looking for stability, automation, 

flexibility, performance, and cost. After comparing these features, Oracle Streams is chosen for closer analysis. 

Two Streams environments are configured – one with a Master/Slave architecture, in which a single server is the 

source for all data updates, and the second with a Multi-Master architecture, in which updates originating from any 

of the servers will be propagated to all of the others. These environments are tested for data type support, conflict 

resolution, performance, changes to the data structure, and behavior during and after network or server outages. 

Through this experimentation, it is determined which requirements of the DSN can be met by Oracle Streams and 

which cannot. 

Solid Modeling and Structural Analysis of LNA (Low Noise Amplifier) Feed Platform for Type-2 BWG 

(Beam Wave Guide) Antennas 

Andrew Crawford 

Mentor: Jason Carlton 

The primary objective of this internship research project involves working in a collaborative environment on the 

detailed solid-modeling design and structural analysis of the LNA (low noise amplifier) feed platform to be built and 

installed on the new 34-Meter BWG (Beam Wave Guide) Type-2 DSN (Deep Space Network) tracking antenna at 

the CDSCC tracking facility in Canberra, Australia, as well as all future similar BWG-Type 2 antenna builds. 

The scope and methodology of the project involves obtaining the necessary mandated requirements from each 

interfacing stakeholder involved with the platform, using SolidWorks 3D modeling software of said requirements to 

design, model, and analyze the construction and assembly of the platform in the new 34 meter BWG-Type 2 

Antenna. 

By using pass/fail criteria software, analysis will be applied using techniques involving live-load, seismic, frequency 

vibration, and deflection tests, ensuring proper and acceptable design plans and models can be created for 

eventual peer review and released drawings. Material and structural analysis of the platform designs and models 

using industry standards and codes will be incorporated to meet national safety requirements. The deliverables 

include a complete solid model of the platform, detailed analysis results, and presentation materials ready for the 

peer review. 

Modeling Intermittent Contact With Complementarity 

Cory Crean 

Mentors: Abhinandan Jain and Marco Quadrelli 

One of the most important tasks in simulating a multibody system is to properly handle the constraints. There are 

well-developed methods for handling bilateral constraints in tree-topology systems, but handling such constraints 

in systems with closed loops still presents some difficulty. Ndarts provides two methods for handling loop 

constraints: the augmented approach, and constraint embedding. This paper validates these implementations by 

simulating classic mechanisms, and comparing the results produced by Ndarts to the results available from the 

literature. Also, it compares the two methods on the basis of efficiency and accuracy. 

Unilateral constraints are also present in many multibody system. For example, a legged robot's foot interacts with 

the ground via a unilateral contact. Ndarts makes use of the complementarity formulation to simulate such systems. 

This paper describes the algorithm used for simulating objects with one or more unilateral contacts, and describes 

some test problems to which the algorithm has been applied. 

An IR Analysis of Cryovolcanism at Sotra Facula on Titan 

Paul A. Dalba 

Mentor: Bonnie J. Buratti 

Investigations of cryovolcanic activity on Titan are numerous and could uncover important clues regarding 

processes taking place on and within Titan. Using the Synthetic Aperture RADAR (SAR) imager and the Visual and 

Infrared Mapping Spectrometer (VIMS) aboard the Cassini Spacecraft, many potential cryovolcanoes have been 

mapped. One of the leading candidates is an area known as Sotra Facula located at 39.8ºW 12.5ºS. In this region, 

SAR data identified a high peak located near a low basin which may have produced mass flows in the past. An 

investigation of Sotra Facula using VIMS data proved that the region is brighter than its immediate surroundings at 

145

2.01µm, the band which provides the clearest view of Titan's surface. This varying brightness was not detected at 

2.13µm and is, therefore, not due to methane clouds in Titan's atmosphere. Sotra Facula was detected in an image 

produced from coadded spectra between 4.9µm and 5.1µm, but the heat signature was not anomalous compared 

to the surrounding terrain. Furthermore, in a ratio of the 4.9µm to 5.1µm coadded spectra to the 2.01µm band, 

Sotra Facula was nearly undetectable. This spectroscopic investigation of Sotra Facula failed to yield any signature 

of a cryovolcano. 

Investigating the Wake of Capsule-Like Bodies Using Computational Fluid Dynamics and Fluid Structure 

Interaction 

Derek J. Dinzl 

Mentor: Anita Sengupta 

The performance of aerodynamic decelerators in the wake of bluff bodies is of particular interest for optimizing 

parachute design during and after planetary entry. This project focuses on computational simulations of the Multi- 

Purpose Crew Vehicle at 10% subscale test conditions. This will include the capsule alone, the capsule mounted on 

a sting, and the capsule with a disk or parachute in its wake. The code used to implement these cases is US3D with 

a Spalart-Allmaras Detached Eddy Simulation turbulence model. Simulation data of the capsule by itself will be 

compared to previously recorded data sets. Wind tunnel experiments of various capsule-parachute configurations 

are being carried out with a maximum dynamic pressure of 100 psf and a Reynolds number near 3×10 6 . Data from 

the simulations will be compared to particle image velocimetry calculations and results of these wind tunnel tests. 

Focus will be on lift and drag forces, velocity/mach and pressure contours, vortical wake structures, and stagnation 

point locations. Validation of simulation data with experiment is of prime importance; once the fidelity of 

computational methods is established one can emulate realistic flight conditions and observe otherwise complex 

phenomena to capture. To this end, a fluid-structure interaction model for the capsule’s drogue parachutes is 

introduced and developed. This will allow us to analyze the motion of the parachute along with its drag 

performance in the wake of the capsule. Benefits and drawbacks of both computational fluid dynamics and wind 

tunnel testing as they pertain to the subject will be explored and discussed. 

Deep Space Network-Wide Portal Development 

Silviya Doneva 

Mentor: Paul Wolgast 

The Deep Space Network (DSN) is an international network of antennas that supports interplanetary spacecraft 

missions and radio and radar astronomy observations for the exploration of the solar system and the universe. 

DSN provides the vital two-way communications link that guides and controls planetary explorers, and brings back 

the images and new scientific information they collect. In an attempt to streamline operations and improve overall 

services provided by the Deep Space Network a DSN-wide portal is under development. The project is one step in a 

larger effort to centralize the data collected from current missions including user input parameters for spacecraft to 

be tracked. This information will be placed into a principal repository where all operations related to the DSN are 

stored. Furthermore, providing statistical characterization of data volumes will help identify technically feasible 

tracking opportunities and more precise mission planning by providing upfront scheduling proposals. Business 

intelligence tools are to be incorporated in the output to deliver data visualization. 

Percussive Augmenter of Rotary Drills 

Chris Donnelly 

Mentor: Yoseph Bar-Cohen 

Hammering drills are effective in fracturing the drilled medium while rotary drills remove cuttings. The combination 

provides a highly effective penetration mechanism. Piezoelectric actuators were integrated into an adapter to 

produce ultrasonic percussion; augmenting rotary drilling. The drill is capable of operating at low power, low 

applied force and, with proper tuning, low noise. These characteristics are of great interest for future NASA 

missions and the construction/remodeling industry. The developed augmenter connects a commercially available 

drill and bit. Input power to the drill is read using a multi-meter and the augmenter receives a separate input 

voltage. The drive frequency of the piezoelectric actuator is controlled by a hill climb algorithm that optimizes and 

records average power usage and operates the drill at resonating frequency. Testing the rotary drill and augmenter 

across a range of combinations with total power constant at 160 Watts has shown improved results in concrete and 

limestone samples under low axial loading (4.2 kg). The drill rate was increased 1.5 to over 10 times compared to 

rotation alone. For a given power there is an optimal balance between the rate of producing cuttings and their 

removal. 

146

Processing of Mars Atmospheric Radio Occultation Data 

Ursula Doswell-Fotheringham 

Mentor: Sami Asmar 

JPL’s Radio Science Systems Group has been collecting and processing radio occultation data from the Mars 

Reconnaissance Orbiter for several years, as previously done with the Mars Global Surveyor mission. The scientific 

output of temperature-pressure profiles is of high interest to the planetary atmospheric science community. The 

raw data gets publically archived promptly while detailed processing is then carried out at JPL. With years of data 

in storage, a system to process and organize it had to be developed. Using UNIX and Matlab, the data from 1999 is 

downloaded from the Planetary Data System Atmospheres Node and a series of codes is run to produce a 

temperature and profile graph for approximately every four hour time period. Similarly, the data from 2008 thru 

2011 has been processed and is being converted from a Matlab-specific format to one that can be viewed and 

analyzed in other programs. There are 30 graphs that have to go thru this per day for the four-year span that the 

Mars Reconnaissance Orbiter has been producing data. Once processed, this data can be made available for 

analysis of the ionosphere and atmosphere of Mars. 

MISR INteractive eXplorer (MINX): Production Digitizing to Retrieve Smoke Plume Heights and 

Validating Heights Against Lidar Data 

Ben Dunst 

Mentor: David Nelson 

Aerosols such as smoke from wildfires, ash from volcanoes, and desert dust can contribute to climate change. They 

typically rise into the atmosphere as plumes from identifiable sources and can be analyzed to determine their 

height and speed. Determining accurate heights for aerosols is important, because if they are injected above the 

planetary boundary layer, they can be transported great distances. MINX, an interactive software tool that uses 

stereoscopic imagery from the Multi-angle Imaging SpectroRadiometer (MISR) instrument, can retrieve heights and 

winds from aerosol plumes. More than 10,000 smoke plumes from around the world have been digitized to date 

with MINX and are compiled into a publicly available online database. Several hundred plumes will be added this 

summer from fires in North and South America. In addition, the height retrieval algorithms in MINX have not been 

formally validated and algorithms were recently added to allow for retrievals in different color bands and using 

different size pattern matchers. The accuracy of MINX height retrievals for specific aerosol events will be evaluated 

by finding time- and space-coincident Lidar retrievals and using those as ground truth to compare with MINX. 

Analysis of Subsurface Clathrates in the Upper Crust of Titan 

John Elliott 

Mentors: William Smythe and Mathieu Choukroun 

Titan has an atmosphere rich in methane, which should have long since been depleted unless a mechanism exists 

for storing these compounds below the surface. One hypothesis is that methane could be stored in the form of 

clathrate hydrates, which are compounds with an ice lattice forming molecular cages in which gases are trapped. 

They are stable at low temperatures and over a wide range of pressures, suggesting that clathrate hydrates may 

have stored methane on Titan from the beginning of its history. To find the conditions at which the clathrates 

dissociate, we will conduct various experiments using a cryogenic calorimeter. The calorimeter needed to be 

modified to allow us to create a high pressure system (100 bars) similar to the pressure on the subsurface of Titan. 

We have finalized the modifications for the calorimeter and will calibrate the device using the vapor-liquid 

equilibrium of CO2. Preliminary tests on CO2 clathrates will be completed before the end of the internship. 

Planetary Ices Spectroscopy 

Jeffrey T. Foster 

Mentors: Robert Carlson and Kevin Hand 

The goal of this research project is to develop a robust and highly sensitive spectrometer that can be used to scan 

water ice and liquid samples on Europa for organic compounds. The aim is to find signs of life on the ice covered 

moon orbiting Jupiter because it has been shown to have liquid water underneath its icy crust. To work towards 

this goal I am helping set up and perform spectral analyses on several different sample types relevant to Europa. 

The tasks I perform are varied but the focus is infrared spectroscopy of ice grains containing organic molecules. 

Progress is being made on obtaining spectra similar to those that would be obtained from Europa. 

Local Time Asymmetry in the Jovian Equatorial Current Sheet 

Alexander Freed 

Mentors: Henry Garrett and Michael Kokorowski 

Charged particle observations from multiple spacecraft have shown an asymmetry in the magnetic equatorial 

current sheet between the dawn and dusk local time regions. There is a thin current sheet at dawn and a thick 

sheet at dusk. Galileo spacecraft measurements show that thin current sheet crossings are regular and coincident 

with observed local maxima in high-energy (~ MeV) electron count rates. However, they also show that where the 

147

current sheet is thick, like in the dusk region, the crossings and electron count rate maxima are neither periodic or 

synchronized. This study investigates the local time asymmetry on a more global scale by using the energetic 

electron data from Galileo, the only spacecraft to orbit Jupiter. Count rate maxima in the middle and outer 

magnetosphere have been averaged over sections of local time to investigate the area in between the dawn and 

dusk regions. These results are then discussed in context of prevailing theories about the nature of this asymmetry. 

High-Speed Generation of Illumination Spectra for a Stereoscopic Endoscope 

Eric Fritz 

Mentors: Harish Manohara and Michael Shearn 

Traditional stereoscopic vision (3D) is achieved through use of two separate cameras, arranged to emulate human 

eyes. This method works on large scale projects, but runs into problems on small scale designs, such as surgical 

endoscopes. This project is focused on developing a stereoscopic endoscope, using a single camera and Conjugated 

Multiple-Bandpass Filters (CMBF) to create stereoscopic vision. Each half of filter is built to allow a distinct 

spectrum through, while blocking the complementary spectrum. A system with different left and right filters can 

produce stereoscopic images. 

To accomplish this, the light must be filtered at the light source to match the filters at the camera. Additionally, the 

light source and camera must be synchronized in a way that each image will show only one filter spectrum. In this 

talk, I will describe the design and characterization of the prototype electro-optical system, including optical 

throughput measurements and video produced using this method. 

Data Collection for Physical Modeling of a Small Dynamic Legged Robot Platform 

Xiao-Yu Fu 

Mentor: Rudranarayan Mukherjee 

OctoRoACH is a 25 gram, 20-cm long, eight-legged mobile robot platform. It is intended to serve as a base 

platform for testing the development of new technologies for control, communication, and sensing in small robots. 

Because OctoRoACH is a new platform undergoing active mechanical development, its physical behavior is not well 

understood. We collected physical performance data describing structural rigidity, dynamic behavior, power 

utilization, and control response using a combination of onboard sensors and external instruments in ideal and 

realistic conditions. This data is being compiled into dynamic model simulation that approximates robot behavior, 

with the goal of providing greater understanding about robot performance and providing directional insights for 

design improvements. 

Nontronite Dissolution Kinetics and Secondary Mineral Formation 

Seth R. Gainey 

Mentor: Joel A. Hurowitz 

Spectral observations from orbiter space craft have detected clay minerals in ancient rock deposits exposed at the 

Martian surface, formed form from the hydrolysis of silicate minerals, indicating the past presence of liquid water. 

It is the purpose of this research to alter clay minerals analogous to those observed on Mars, in order to determine 

their rate of dissolution under pedogenic conditions analogous to those common on the Earth’s surface, and 

possibly common on the ancient surface of Mars. These experiments determined the dissolution rate of nontronite 

(standard NAu-1) and the secondary minerals produced as a byproduct of this alteration. Mineralogy was confirmed 

using a Horiba Jobin Yvon LabRam HR confocal Raman microscope. Flow through reactors were modeled after 

Taylor et al. (2000), each flow through reactor consist consists of Teflon tubing where solution is pumped through 

the sample by syringe pump. Effluent was subsequently collected for pH measurement and ion concentration 

analysis via ICP-OES. Rate laws were determined from the dissolution rates as a function of pH and distance from 

equilibrium. Altered and unaltered mineral powders were analyzed using Raman spectroscopy, electron microprobe 

analysis, SEM combined with EDS to determine mineral phase identification, chemical composition, and morphology, 

respectively. 

Finding the Sun’s Closest Companion: Using WISE Observations and Follow-Up Data to Find the Closest 

and Coldest Brown Dwarfs 

Thomas Gautier 

Mentors: Amy Mainzer and Chris Gelino 

The Wide-Field Infrared Survey Explorer (WISE) Brown Dwarf Team has compiled a list of approximately 1600 

objects that it classifies as brown dwarf candidates. Many of these objects had partially or unprocessed follow-up 

data. This project compiled follow-up data from various telescopes and primarily used IDL routines to complete the 

processing and extract the precise astrometry and photometry from each observation. The result of the project is 

an easily accessible database with the astrometric and photometric data for each of the candidates, which greatly 

facilitates the analysis of by the Brown Dwarf Team to find objects with brown dwarf colors and that are quickly 

moving, indicating they are close to the Sun. 

148

Modeling Hydrothermal Vents on Europa 

Patricia Gavin 

Mentor: Steve Vance 

Hydrothermal vents occurring at sea-floor-spreading centers may have hosted the first transition from abiotic to 

biotic chemical processes, eventually resulting in the emergence of life. The possibility of these vents also being 

present on Europa’s subsurface ocean floor puts forth intriguing astrobiological implications relevant to both life on 

another planetary body as well as the origins of life on Earth. The studies presented here simulate conditions at 

potential hydrothermal vents on Europa to determine which minerals favorable to life form under these conditions. 

Using the modeling software Geochemist’s Workbench (GWB), several simple models were composed, each 

simulating a different geochemical scenario. Among the simulations derived were (1) water-rock interactions, 

(2) terrestrial hydrothermal fluids and (3) reduction-oxidation (redox) reactions. The first simulation was used to 

study how fluids inside a hydrothermal vent react with the rock that composes the vent. The second model 

simulated the resulting hydrothermal fluid being introduced and reacting with seawater. The third model 

investigated the chemical energy potentially available to microbes for use in organic and metabolic processes. 

Results from these efforts help in understanding what signatures of hydrothermal activity in Europa’s ocean might 

be found by exploration spacecraft observing its geologically active surface. 

Mars Topographical Data Analysis 

Hallie Gengl 

Mentors: Matt Golombek and Fred Calef 

The overarching objective of this project is to analyze Mars elevation models at various length scales. On a 1 km 

length scale, it involves working with digital elevation models (DEM) at the current MSL sites, as well as at 

candidate proposed future robotic missions to Mars like Mars Sample Return (MSR). Due to engineering constraints, 

relief cannot exceed 100 meters on the 1 km or smaller length scale in landing site ellipses. High Resolution Stereo 

Camera imager (HRSC at 50-300 m/pixel) and Mars Orbiter Laser Altimeter data (MOLA at ~300 m spacing), both 

DEM and shot-points, are used. The results of each are compared to evaluate the relief within the parameters of 

the landing site constraints. On a small length scale, DEMs are created using stereographic image pairs from the 

High Resolution Imaging Science Experiment (HiRISE) camera onboard the Mars Reconnaissance Orbiter (MRO). 

These products (25 cm/pixel) can be used for many projects, including high-resolution landing site analysis as well 

as small rayed crater morphology. One of the goals is to create a larger dataset of DEMs with small rayed craters 

and extract morphometric statistics to understand the change in crater and ejecta shape with changes in diameter 

and impact type. 

High Precision Propagation and Tracking of the International Space Station 

Corrina L. Gibson 

Mentor: Rob Witoff 

Optical Payload for Lasercomm Science (OPALS) is an experimental payload that will be mounted on the 

International Space Station (ISS) in 2012. The goal is to deliver video from the ISS to an optical ground terminal 

via an optical communications link. In order to optically downlink data during an ISS passover, high precision 

propagation followed by closed-loop tracking of the ISS is required. The ground station’s telescope has a limited 

field of view, requiring the ISS position error to be less than 100 meters for closed loop tracking to initiate. Precise 

ISS position predictions are used to align the telescope prior to ISS passover, ensuring sufficient view of the ISS 

for closed-loop tracking. Once the ISS is within the telescope’s field of view, a beacon is transmitted from the 

ground station to a camera on OPALS. A fast steering mirror is used to track OPALS from the ground. The ground 

station’s centroiding loop controls the fast steering mirror to remain pointed at OPALS. Techniques that will enable 

prediction of the ISS’s position to within 100 meters after 5 minutes of propagation are being developed. Thus far 

the error of various propagators has been characterized relative to a truth data set. Additional propagators will be 

investigated to identify an acceptable propagation technique. 

ISSARS Aerosol Database: An Incorporation of Atmospheric Particles Into a Universal Tool to Simulate 

Remote Sensing Instruments 

Michael B. Goetz 

Mentors: Simone Tanelli and Noppasin Niamsuwan 

The Instrument Simulation Suite for Atmospheric Remote Sensing (ISSARS) is entering its third and final year of 

development with an overall goal of providing a universal tool to simulate active and passive space borne 

atmospheric remote sensing instruments. These simulations focus on the atmosphere ranging from the UV to 

microwaves. ISSARS handles all assumption and uses various models on scattering and microphysics to fill the 

gaps left unspecified by the atmospheric models to create each instrument’s measurements. This will help benefit 

mission design and reduce mission cost, create efficient implementation of multi-instrument/platform Observing 

System Simulation Experiments (OSSE), and improve existing models as well as new advance models in 

development. In this effort we incorporate various Aerosol particles into the system, and a simulation of input 

wavelength and spectral refractive indexes related to each spherical test particle(s) can generate its single 

149

scattering albedo (SSA) and phase function (P( �)). Future development with a radiative transfer code will 

generate a series of results that can be validated with results obtained by the Multi-angle Imaging 

SpectoRadiameter (MISR) instrument. 

Verifying Flight Software Commands Generated by CHILL Against MSEQ for the MSL Rover 

Ryan Goetz 

Mentor: Jim A. Kurien 

The flight commands for the MSL rover have been adapted and changed during the years. The old way to generate 

these commands is with a program known as CHILL. As the mission draws closer more detail commands need to be 

generated for both cruise and land operations. MSEQ was developed to generate detail mission plans for the rover 

during operation. In order to switch to the MSEQ program one must verify that the commands generated with both 

programs are similar. A specific command is run through the CHILL and MSEQ program to generate a scmf file that 

is compared with one another. The results should be that both scmf files are similar thus verifying that the rover 

will receive and operate the commands from the newer MSEQ program. 

Correlations Between Technical Performance and Cost Growth: A Case Study Using Schedule Slips 

Austin Gomez 


Flight Projects historically exceed the available budgetary funds allocated by NASA HQ. This group research project 

investigates historical data from the 13 most recent JPL flight projects to determine trends and correlations 

between technical and programmatic parameters to predict cost growth. Each of the 13 projects is analyzed using 

5 major data sets: technical complexity, subcontractor performance, risk trends, margin analysis and schedule slips. 

Each data set is analyzed to determine which correlations exist. My area of focus for this group project is to analyze 

cost and schedule growth between each milestone using schedule slip information. The normalization process will 

investigate the changes between the planned and the actual schedule data. Additionally, all data will be divided 

into lower-level activities such as instruments and spacecraft for more detailed conclusions. Once all technical and 

cost data is collected, regression analysis and modeling is performed to establish trends. It is expected that certain 

aspects of technical performance parameters will lend themselves to be more accurate cost predictors than others. 

Conclusions derived from this group research can be utilized to form a historical data base for future projects and 

compared against the current in-work flight projects to predict cost growth. 

Using Wide Infrared Survey Explorer (WISE) and Ground Based Follow-Up Data to Find the Closest and 

Coldest Brown Dwarfs 

Stephanie Gomillion 

Mentors: Amy Mainzer and Michael Cushing 

The WISE Brown Dwarf Team has compiled a list of approximately 1600 objects that it classifies as brown dwarf 

candidates. Many of these had partially or unprocessed follow-up data. My project was to help gather and compile 

the dispersed ground based telescope data into a precise format, primarily using IDL and Emacs. The result of the 

project is an easily accessible database with the astrometric and photometric data for each of the candidates, 

which significantly facilitates the analysis by the Brown Dwarf Team to find objects with brown dwarf colors and 

that are quickly moving, indicating they are close to the Sun. 

Radio Science Software Tools for Improved Operations and Analysis 

Carlos E. Gonzalez 

Mentors: Sami Asmar and Kamal Oudrhiri 

The DSN open-loop telemetry demodulation method down-converts and records spacecraft signals in pre-selected 

bandwidth using predicted frequency profiles. Radio Science techniques provide more flexibility and robustness in 

the acquisition, processing, and interpretation than the closed loop system. We are designing a graphical user 

interface to run the demodulation quickly and efficiently as well as selecting particular demodulation cases. Task 

#2 involves performing science experiments using Low Earth Orbit Satellites such as the International Space 

Station. DSN antennas are strained due to ISS’s fast orbit so higher orbiting satellites form a three-way 

communication loop. There are multiple scenarios that can be evaluated using the Satellite Orbit Analysis Program. 

We will be able to provide the most desirable results. The primary configuration will have a DSN antenna 

communicating with the TDRS satellite, which then relays the signal to the ISS, and ISS will then relay the 

information back to the DSN. Task #3 involves the real time operation of the New Horizons Mission and controlling 

the DSN Radio Science Receivers remotely from JPL. We will test the real time monitoring and recording of the 

downlink signal from New Horizons in order to verify whether or not the spacecraft is working properly. 

150

ASTRA Command and Data Handling Flight Software Development 

Maximillian Gonzalez 

Mentor: Sam Gordon 

The Analog Site Testbed for Readiness Advancement (ASTRA) project is aiming to increase the technology 

readiness level (TRL) of two instruments, a hot-wire anemometer (HWA) and a rapid acquisition mass spectrometer 

(RAMS), so they can be considered for future missions to Mars. ASTRA is a high-altitude balloon project whose 

payload will be hoisted up to 120,000ft into a Mars analog test environment of 6 mbar and -35°C. Command and 

data handling (C&DH) software has been developed to guide the HWA through a 12-hour nominal mission duration 

from pre-launch operations through power-down before landing. The software uses a data sampling interrupt 

routine running at 20 kHz to command HWA electronics through a 1 Hz collection cycle and sample digital data 

coming in 16-bit serial packets from the HWA at 10 kHz. Once a full, time-stamped sample set, consisting of 

60 16-bit data packets, has been gathered, the software will store the sample set on its onboard flash memory as 

well as downlink the data to the ground station via the on-board Consolidated Instrument Package (CIP) radio. 

Improving Functionality of a Mechanical Control System for a Scanning Microwave Limb Sounder 

(SMLS) Through Program Development and Structural Modification 

Zachary Greene 

Mentor: Paul Stek 

The Microwave Limb Sounder (MLS) team is currently developing its newest instrument to probe the Earth's 

atmosphere. Known as a scanning microwave limb sounder (SMLS), this device will use microwaves to measure 

atmospheric variables including temperature, pressure, and chemical composition. Deployed via high altitude 

balloon or aircraft to the upper stratosphere, the SMLS will be the first MLS instrument to scan the horizon laterally 

in addition to ordinary vertical profiling performed by previous instruments. Incomplete code existed govern the 

scanning mechanism and problems with the scanner arose frequently. Furthermore, documentation left by previous 

interns was brief and unclear, forcing much of my initial time to be spent re-learning what they had already done. 

My tasks were to modify the existing code to fix nuances and programmatic errors, construct the circuitry and 

physical setup as well as troubleshoot and resolve any problems that occurred within, and leave adequate 

documentation for future interns and MLS team members who wish to modify or use the SMLS. 

Hyperspectral Imaging and the Building of a Near-Infrared Imaging Spectrometer 

Kate Grode 

Mentors: Bill Mateer and Marc Lane 

Hyperspectral imaging, or imaging spectroscopy, is the analysis of the electromagnetic spectra reflected from what 

is being studied, such as the ground, a rainforest, or an object, and then recorded as images. Objects give off 

distinctive wavelengths, so it possible to distinguish one object from another within the image. Imaging 

spectroscopy has many applications, such as in mineralogy, agriculture, environmental research, and surveillance. 

The spectrometer currently being built uses hyperspectral imaging and is similar to previous spectrometers. They 

are attached to airplanes and flown over a region to gather data. A past spectrometer built for the Carnegie 

Airborne Observatory records the wavelengths from the plants in the Amazon Rainforest in order to understand 

how the rainforest and the environment as a whole are affected by humans. The current instrument is in the build 

phase and is scheduled to be finished in January 2012. 

Acceleration of the KINETICS Integrated Dynamical/Chemical Computational Model Using MPI 

Jonathan (Max) Grossman 

Mentors: Karen Willacy and Mark Allen 

Understanding the evolution of a planet's atmosphere not only provides a better theoretical understanding of 

planetary physics and the formation of planets, but also grants useful insight into Earth's own atmosphere. One of 

the tools used at JPL for the modeling of planetary atmospheres and protostellar disks is KINETICS. KINETICS can 

simulate years of complex dynamics and chemistry. At the moment, KINETICS is in use by or planned to be used 

for: 

1. The ExoMars mission to study the Martian atmosphere in 2016. 

2. Modeling of the composition of the protostellar disks which lead to star formation. 

3. Examination of pollutants in Earth's atmosphere. 

4. Modeling of Titan's atmosphere 

Because of the complexity of KINETICS, it requires large amounts of computation to run simulations at a 

meaningful level of detail. Using MPI (Message Passing Interface), a popular multi-core programming model, will 

permit more useful experiments to be run in less time, accelerating the discovery process. The advantages of MPI 

lie in its exposure of communication, allowing programmers to manually tune the communication pattern of an 

application. MPI will be used with the existing FORTRAN code to parallelize and accelerate performance-critical 

sections of code in KINETICS. 

151

Improving the Thermoelectric Efficiency of Mg2Si 

George M. Hakimeh 

Mentors: Sabah K. Bux and Jean-Pierre Fleurial 

Thermoelectric devices have been a proven, established, and dependable power source technology for many 

unmanned terrestrial and deep space missions for the past 50 years. Although current thermoelectric generators 

have been both reliable and successful, a significant increase in their relatively low thermal-to-electric energy 

conversion efficiency is needed for a variety of future NASA science missions. Recent research suggests that 

nanostructuring materials can increase the dimensionless thermoelectric figure of merit ZT, a gauge of a material’s 

thermal-to-electric conversion efficiency, by reducing the lattice thermal conductivity via increased phonon 

scattering due to the nanoscale of the grains or the presence of nanostructured inclusions. As an example, of 

nanostructured bulk Si was enhanced by 250% at 1275 K compared to single crystal Si. Building upon the previous 

work on high temperature materials, nanostructured silicide-based materials are ideal for maximizing device 

performance at the lower temperature range (500-775 K) range. Magnesium silicide (Mg2Si) is particularly 

attractive due to its low density, low toxicity, thermal stability, relative abundance and low cost of production. 

However, the ZT of the n-type material remains low, about 0.7 at 775 K. Theory suggests that significant increases 

in ZT could be achieved by forming composites containing various nanoscale (< 10 nm) silicide inclusions. Here we 

present the impact of nanostructuring Mg2Si on the thermoelectric transport properties of magnesium silicide. 

System-Level Trade Analysis of the Entry, Descent, and Landing System and the Mobility System of the 

Mars Rover 

Bryan He 

Mentors: Yoshiaki Kuwata and Marco Pavone 

The potential Mars rover mission depends on two main systems: the entry, descent, and landing system; and the 

mobility system. The entry, descent, and landing system has an inherent uncertainty due to errors in sensor 

measurements, variations in atmospheric winds, and dispersions caused during travel through the atmosphere. 

Consequently, the target landing location is best represented as an ellipse rather than as a single point. Failure of 

the entry, descent, and landing system can occur due to impact with a hazard or landing in a location with a slope 

steeper than the maximum tolerance of the lander. The mobility system determines the movement of the rover 

after landing. Failure of the mobility system can occur when the rover lands in an area surrounded by hazards or 

due to mechanical failure of the rover. The probability that mechanical failure occurs is a function of the distance 

travelled and the proximity to hazards. The optimal landing ellipse is the location with the minimum combined 

probability of landing failure and mobility failure. The algorithm currently being developed accounts for the 

probabilistic nature of the initial entry phase by using a reverse dynamic programming method and determines the 

optimal landing target. 

Panosphere: Panoramic Martian Landscapes in an Immersive Environment 

Daniel L. Healy 

Mentor: Victor Luo 

NASA Outreach programs are important to get students and the public interested in space and Mars exploration. 

The OPS Lab is currently working on several outreach programs that deal with natural user interfaces including 

Microsoft’s Kinect. I'm working with the OPS Lab on an interactive 3D Mars outreach experience created in Unity. 

The application allows users to follow the path of the Spirit rover, learn more about Spirit's journey, and experience 

the Martian surface in 3D. The imagery and height maps for the terrain were captured by several orbiters at 

varying levels of detail, with emphasis on areas close to Spirit's path. This enables us to load images of the terrain 

at higher-resolution as the user rolls closer. To achieve even higher detail, I am working with a partner on adding a 

first-person panoramic experience which can be navigated in a way similar to Google's Street View. Future work for 

me includes completing the feature, optimizing the feature so it can be merged with the main experience, and 

contributing to a separate project for Kinect. 

DC Magnetics System Design 

Jared Daniel Helms 

Mentor: Pablo Narvaez 

DC Magnetics testing is vital for replicating the magnetic environment that spaceflight hardware might experience 

during its mission, from very low magnetic fields in deep space (about 2 nanoTeslas) to stronger fields of up to 

1 x 10^6 nanoTeslas (10 Gauss), such as that around Jupiter. The objective of this project is to develop a semiautomated 

system to do DC Magnetics testing in the environments experienced by flight hardware during its 

mission into outer space (from anywhere near Earth to orbits around Jupiter). The system required a Helmholtz coil 

to negate Earth’s magnetic field, magnetometers to measure the ambient magnetic field, a device to communicate 

with and control the coil, and a shaft encoder connected to a turn table to track the movement of hardware within 

the null-field as the hardware is rotated 360 degrees for mapping its magnetic field signature. All of the above were 

connected to a computer with LabVIEW software applications developed to process the data gathered by the 

152

instruments in order to automate the measurements. The fully functional system measures the magnetic field 

produced by hardware at multiple points corresponding to the angle at which it is turned. The resulting data can be 

used to characterize hardware’s magnetic field, find its magnetic dipole moment, and determine the impact it has 

on other hardware. 

ATHLETE: Tool and Test Bed Development 

Gabe Herz 

Mentor: Christopher McQuin 

The All-Terrain, Hex-Limbed, Extra-Terrestrial Explorer (ATHLETE) is a 6 limbed concept rover. Each limb has 

6 degrees of freedom and a wheel located at the end. ATHLETE is capable using the wheels to drive across gentle 

terrain. When rough terrain is encountered, the wheels lock and the limbs can be used for climbing. Each limb can 

also serve as a robotic arm using a tool attachment mechanism. These unique features allow ATHLETE to be used 

for transporting heavy cargo or to serve as a platform for astronaut mobility. The current focus of ATHLETE’s 

development is directed towards low gravity environments like asteroids and comets. For ATHLETE to operate in 

these conditions, it must be able to anchor itself. Once secured to the surface, ATHLETE will be able to use various 

end effectors for digging or drilling. The focus of this project is to complete the development of a counter rotating 

auger anchor, and develop test beds to evaluate it, and other end effectors’ efficiency. 

The Development of a Sub-Kelvin Grating Spectrometer Test Bed 

Jeremy Hodis 

Mentor: Charles M. Bradford 

We have researched the possibility of making a ground based far-IR silicon spectrometer using a grism as the main 

grating element. The spectral range of the spectrometer was based on the atmospheric window of CCAT’s future 

location on Cerro Chajnantor in Northern Chile. The wavelengths of interest range from 405-521 µm and 304-391 

µm. Some potential benefits of this far-IR spectrometer design include smaller size and fewer optical components. 

This spectrometer would be a good candidate as an instrument on the CCAT telescope so CCAT’s properties were 

parameters in our designs. Since this spectrometer isn’t of conventional design some research needs to be made 

into its fabrication process. 

SSCAE Systems Audit of Software Tools Used by Division 31 

Aaron Hoffman 

Mentors: David Mohr and Victoria Scarffe 

Systems and Software Computer Aided Engineering team provides support on tools used by Division 31. The 

SSCAE team has been working to understand what the division’s employees’ wants and problems are and their 

opinions on the software tools that SSCAE supports. SSCAE focuses on tools whose licenses are purchased instead 

of tools that are developed at JPL or NASA. SSCAE has also searching for ways to help overall production and 

communication. In order to obtain the data needed, a survey of the mission operations employees in Division 31 

and Sections 342 and 345 was conducted and concluded. Interviews of employees in Division 31 have also been 

conducted to supplement the data collected from the survey. Based on the data so far, there have been four main 

issues and requests: communication, basic productivity tools, Adobe products, and collaborative editing. The data 

collected shows the usage of the programs supported by SSCAE. Based on this data, reductions in the number of 

licenses of various products will take place, as well as the implementation of the requests by employees. After 

these changes occur, another survey and additional interviews of employees should be conducted. 

Construction and Testing of Active Precision Reflectors for the Purpose of Surface Shape Control and 

Deformation Compensation 

Samuel Hoffman 

Mentor: Samuel Case Bradford 

Three piezoelectrically active reflectors have been developed and tested. The first is a meter-scale hexagonal 

composite reflector consisting of a spherically curved graphite face sheet, aluminum honeycomb core composite 

panel with a network of 90 distributed piezoelectric composite actuators. The actuator system may be used for 

controlling the structural dynamic response of the reflector, or for correcting low-order thermally-induced quasistatic 

distortions in the panel. Thermally-induced surface deformations of 1 to 5 microns were deliberately 

introduced onto the reflector, measured using a speckle holography interferometer, and corrected using a 

piezoelectric composite actuator network distributed across the back face sheet of the reflector. The second 

reflector is a metal hexagon with an outer diameter of 23cm, a polished metal front surface, and a network of 

12 actuators on the back face. It is designed to test reflector scalability and performance. The third reflector is a 

circular borosilicate window 150mm in diameter with 15 actuators on the back face. Absolute deformation can be 

measured by a standard optical interferometer. Using this interferometer we have shown that the piezoelectric 

actuator network can correct the surface flatness of the window to better than the original 4λ/inch. 

153

Volcanic Ash Plumes: An Analysis With Multi-Angle Imaging SpectroRadiometer (MISR) Images 

Collin Holmquist 

Mentors: Veljko Jovanovic and Olga Kalashnikova 

The analyses of volcanic plumes are pertinent to the study of plume dynamics. Ash plume heights and thickness 

are important for analyzing ash air pollution and transport. Of interest for this project are the extents to which ash 

plumes rise, thus determining the probability of mixing within the Planetary Boundary Layer or moving into the free 

atmosphere. To help study the plumes, NASA’s Multi-angle Imaging SpectroRadiometer (MISR) instrument has the 

ability to create stereoscopic views of features below. Using the MISR images, the MISR Interactive eXplorer 

(MINX) program computes plume wind speed and heights. A database of MINX digitized volcanic eruptions and ash 

plumes will be created. It is possible that statistics of the eruptions can be then compiled in future, if not during the 

time at JPL, that will focus on plume height compared to eruptive strength. As a second part to the project, an indepth 

study of the plume heights of the Puyehue-Cordon eruption in Chile will be used in trajectory models. The 

results will then be compared to other findings by volcanologists. 

Granular Media Simulations Application in Mechanical Design 

Ryan Houlihan 

Mentors: Rudranarayan Mukherjee and Marc Pomerantz 

We explore the capabilities of massively parallel granular media simulations and its feasibility for its use in analysis 

and design of various mechanical systems. Properties of granular media and the interaction of different granular 

particles with mechanical systems and themselves is largely important in the design and optimization of such 

mechanical systems. To visualize and analyze the interactions and locomotion of mechanical systems on granular 

mediums massively parallel simulations are quite advantageous. The simulations are highly parallelized and highly 

scalable and can help us both visually as well as numerically analyze the effects certain mechanical systems have 

on these mediums through computing the various phenomenological first-order interaction forces in the system. 

Our work explores what is currently feasible with granular media simulations as well as the effects different 

granular materials have in terms of mechanical locomotion and interaction with these mediums 

Investigation of Visualization Tools for Integrated Spacecraft Analysis 

Cody A. Hyman 

Mentors: Yu-Wen Tung and Pierre Maldague 

Spacecraft modeling, a critically important portion in validating planned spacecraft activities, is currently carried 

out using a time consuming method of mission to mission model implementations and integration. A current 

project in early development, Integrated Spacecraft Analysis (ISCA), aims to remedy this hindrance by providing 

reusable architectures and reducing time spent integrating models with planning and sequencing tools. The 

principle objective of this internship was to develop a user interface for an experimental ontology-based structure 

visualization of navigation and attitude control system modeling software. To satisfy this, a number of tree and 

graph visualization tools were researched and a Java based hyperbolic graph viewer was selected for experimental 

adaptation. Early results show promise in the ability to organize and display large amounts of spacecraft model 

documentation efficiently and effectively through a web browser. This viewer serves as a conceptual 

implementation for future development but trials with both ISCA developers and end users should be performed to 

truly evaluate the effectiveness of continued development of such visualizations. 

Capabilities of the Bullet Physics Engine: Collision Detection for Physics-Based Simulation of Multi-Body 

System 

Rasheed Ibrahim 

Mentor: Rudra Mukherjee 

Collision Detection is critical to applications in physics-based simulation and robot modeling that includes the 

intersection of two or more (soft or rigid body) objects. In this paper, we present a collision detection algorithm 

implemented in the Bullet Physics Engine for the intersection of two or more multi-body system. Furthermore, we 

provide physics-based simulations for soft body and rigid body dynamic systems that demonstrate the capabilities 

of the Collision Detection library in the Bullet Physics Engine. First, we created a basic simulation of a rigid ball 

colliding into a static block using the built-in rigid body collision shapes provided by the physics engine. Then, we 

apply a convex hull generator algorithm written by Stan Melax to approximate a custom arbitrarily complex 

concave triangle mesh as a collection of rigid convex objects for use in the Bullet Physics Engine. Finally, we 

explored the soft body dynamics library in the physics engine. We created a basic simulation that shows two soft 

body objects colliding into one another and a separate simulation that shows a soft body object colliding into a rigid 

body object. We find the time, location and existence of contact between pairs of objects in simulation and it is 

expected that the Bullet Physics Engine will demonstrate good collision detection capabilities for our use. 

154

Flight Rule Identification, Implementation, and Verification for Mars Science Laboratory 

Lee Jasper 

Mentor: Andrew Mishkin 

Mars Science Laboratory is poised to launch late this year. Because launch is approaching, significant effort is being 

put into creation and completion of operations processes and software for the Launch, Cruise, and Approach (LCA) 

phases of the mission. Flight Rules (FRs) are one component of the operations processes that govern flight 

operations. FRs provide guidelines and constraints for safe and proper operation of the flight hardware. They are 

implemented in software and operations procedures to ensure that harmful events do not occur. LCA FRs are 

currently being implemented, verified, and validated. Implementation of each FR generally falls under procedural 

checks or the software program SEQGEN, which is used to model and verify commands prior to execution on the 

spacecraft. In an effort to aid FR verification and to understand how SEQGEN modeling and adaptation occurs, 

several FR tests have been conducted during this project to ensure that all commands that could violate a FR have 

been correctly modeled, and are caught by, SEQGEN. Also, the baseline allocation of LCA FRs to the responsible 

operations procedures has been completed. Allocations were reviewed by the operations teams and subsystems to 

ensure FR checks have been incorporated into the proper procedures. 

Characterization of the Phase-Shifting Zernike Wavefront Sensor for Telescope Applications 

Rebecca M. Jensen-Clem 

Mentor: J. Kent Wallace 

The dynamic Zernike wavefront sensor (ZWFS) is a common-path, phase-shifting interferometric wavefront sensor 

with superior broadband and noise rejection capabilities. The ZWFS will be integrated into the Mount Palomar 200” 

Hale Telescope adaptive optics systems in 2013. It will also be used to phase the segmented mirrors of Cerro 

Chajnantor Atacama Telescope (CCAT), a 25-m submillimeter telescope recommended by the 2010 decadal survey. 

This summer, the dynamic ZWFS testbed was completed. This work involved: 1) optical measurements of key 

components, 2) revised component specifications, and 3) procurements with vendors to fabricate optical 

components and internal JPL fabrication resources. A numerical simulation of the ZWFS was built from first 

principles to compare the sensor’s theoretical and testbed performances. Measurements of the phase perturbations 

of a glass microscope slide were made using both the ZWFS testbed and a Zygo interferometer. These 

experimental validation measurements and the numerical simulations allowed us to determine the accuracy of the 

ZWFS phase measurements. 

Enhancing the GUI-Based Visualization and Management iPanel Software for ISAAC 

Carrine M. Johnson 

Mentor: Yutao He 

The Instrument ShAred Artifact for Computing (ISAAC) is a modular and integrated FPGA-based common 

instrument control and computing platform for NRA Decadal instrument missions. ISAAC includes six components, 

iBoard, iBus, iCore, iPackage, iBench, and iTool. This project is focus on iPanel that is part of iPackage. iPanel is a 

modular, flexible and extendable GUI-based software written in Python that controls and communicates with an 

ISAAC iBoard. Currently there is an emerging need for iPanel to collect and visualize telemetries from the ADC/DAC 

daughter board of an iBoard as well as display real-time data and send out commands. The aim of this project is to 

enhance the current version of iPanel with a SPI interface. It will implemented using wx.Python and its array of 

widgets. The new interface allows the user to interact with the ADC/DAC peripherals via SPI and to update register 

values and display the telemetry data. 

Performing a Telemetry Audit and Developing a Power Model for the Mars Science Laboratory Rover 

Kristin P. Kagetsu 

Mentors: David Oh and James Chase 

During the verification and validation stage of the project, the Mars Science Laboratory (MSL) Flight Software team 

runs tests on the spacecraft and rover to ensure that they meet all of the requirements. One of the tests is to 

perform a telemetry audit which mines the testbed telemetry database for validation of all FSW-created telemetry 

channels. This requires running tests in the simulated FSW environment, which allows the user to run tests from a 

remote computer and allows for faster iterations of the testing procedures. Once the data is acquired from each 

session, statistics can be determined about which channels have been called and what range of values have been 

logged. The ATLO team is verifying the power consumption of various components of the spacecraft and rover to 

create a more accurate power model of them. The ATLO baseline tests have data on instruments used during the 

descent and cruise stages. The current of the various components can be acquired from these tests and then the 

voltage is determined from WMD based on the timestamp of the current data. This allows us to calculate the power 

of each component and compare it to the predictions in the Power Equipment List. 

155

A Survey of Efforts to Characterize Parachute Performance in the Wake of a Blunt Forebody 

Cole D. Kazemba 

Mentor: Ian G. Clark 

As the paradigm for robotic Mars missions shifts from the Viking class ballistic entry vehicles to larger and more 

complex vehicles on the order of Mars Science Laboratory (~3250 kg entry mass) or greater, there is increasing 

emphasis on improving the capabilities and understanding of the supersonic parachutes used during atmospheric 

descent. A new technology being developed at JPL, the Supersonic Inflatable Aerodynamic Decelerator (SIAD), will 

allow for even greater entry masses than MSL. However, the large diameter and mass of an entry vehicle utilizing a 

deployed SIAD will push the requirements of its parachute beyond the disk-gap-band (DGB) systems qualified by 

the Viking program. In order to address some of challenges of predicting the expected performance of a new 

parachute system, an extensive literature review and data collection effort was conducted. Multiple empirical and 

analytical methods to predict parachute performance in a trailing wake were implemented and compared to the 

historical data with very limited success. The poor performance of these prediction methods and large gaps in the 

available experimental data indicate that significant effort is required for developing more robust and capable 

methods and building a thorough experimental database to anchor new and existing computational tools. 

Correlations Between Technical Performance and Cost Growth: A Case Study Using Risk Trends 

Brad Kellogg 


Flight Projects historically exceed the available budgetary funds allocated by NASA HQ. Project lessons-learned 

indicated that there are common reasons for cost growth where strong correlations existed between technical 

performance and cost/schedule performance. This group research project investigates historical data from the 

13 most recent JPL flight projects to determine trends and correlations between technical and programmatic 

parameters to predict cost growth. Each of the 13 projects is analyzed using 5 major data sets: technical 

complexity, subcontractor performance, risk trends, margin analysis and schedule slip. Each data set is analyzed to 

determine which correlations exists. My specific assignment for this group project is to analyze monthly risk 

assessments to more accurately predict cost growth. A numerical risk factor is calculated using both new and 

proven methods. After compiling numerical risk values, a regression analysis will be performed with cost growth 

data to establish trends. It is expected that certain aspects of technical performance parameters will lend 

themselves to be more accurate cost predictors than others. Conclusions derived from this group research can be 

utilized to form a historical data base for future projects and compared against the current in-work flight projects 

to predict cost growth. 

Analysis and Predictions of Mission Downlink Data Volume for Mars Reconnaissance Orbiter Extended 

Missions 

Andrew K. Kennedy 

Mentors: Dan Johnston, Dave Herman, and Reid Thomas 

The Mars Reconnaissance Orbiter (MRO) has returned more scientific data to date than the combined totals of all 

other deep space missions. Dependable estimates of the amount of continuing downlink data from MRO are 

important in supporting science and ground system planning, as well as honoring the project’s commitments to the 

Mars Exploration Program office. 

For these reasons, a study was performed on the downlink process for MRO in order to update the project’s data 

volume analysis tools and provide data estimates for future extended missions. As part of this study, it was 

necessary to research in depth the MRO’s communications link through the Deep Space network, in addition to the 

spacecraft’s orbital geometry. This knowledge was used to develop streamlined algorithms to calculate data 

volumes, and update the previous analysis tools. 

The final result of this study was a documented knowledge of MRO’s “downlink algorithm”, as well as specific 

predictions of downlink volume for MRO’s current extended mission, and next planned extended mission. The 

documentation produced will also serve as a user’s guide for future mission planners to more effectively make data 

volume predictions. 

MSL Surface Operations: Remodeling of the Mission Operations System Tactical Processes 

Sheehan Khuu 

Mentors: Andrew Mishkin and Alicia Allbaugh 

Surface operations for the Mars Science Laboratory (MSL) rover, Curiosity, are split into two categories: strategic 

and tactical. Strategic processes, performed over multiple days, include, among others, determining 

communication windows and scheduling with the DSN. Tactical processes define the specific activity plans for the 

rover for the next Martian day (sol), and develop the command sequences that implement the activity plans 

onboard the rover. The MSL Mission Operations System (MOS) includes the tools, processes, procedures, and 

teams that perform the planning and execution for Curiosity. The MSL MOS is modeled to validate that the tactical 

156

processes produce the appropriate products for consistent uplink of the sequence load. Artifacts extracted from the 

model designed in Unified Modeling Language (UML) were converted to a PowerPoint based design. The UML model 

ineffectively communicated the higher level tactical processes to the applicable audience. The purpose of the 

remodeling is to remedy this problem. The model is expected to provide a robust high to mid-level view of the 

tactical processes and serve as a review product for the Integration, Planning and Execution (IPE) and mission 

operations team. This paper describes the approach and preliminary results of the conversion process from a UML 

to PowerPoint model. 

Fault Tolerant RCS Thruster and Thruster Drive Electronics Architectures for EHM Orbiter 

Brent A. Kisling 

Mentor: Ray Crum 

Results from the Vision and Voyages for Planetary Science in the Decade 2013-2022 report indicated that JEO 

(Jupiter Europa Orbiter) “should be flown only if changes to both the mission design and the NASA planetary 

budget make it affordable without eliminating any other recommended missions.” Due to this motivation, a need 

arose to examine alternative fault tolerant RCS (Reaction Control System) thruster and thruster drive electronics 

architectures that may lower the overall system cost of the revamped JEO mission, now entitled EHM (Europa 

Habitability Mission). Through my research, I will identify alternative RCS thruster and thruster drive electronics 

architectures and then evaluate those architectures using a set of weighted criteria; including architecture size, 

cost, mass, fault tolerance, and reliability. Based upon this criteria and corresponding weights, a decision matrix 

can be formed to compute the weighted score of each alternative architecture. The architecture with the highest 

weighted score will be suggested as the best solution for EHM. 

Multi-Mission Space Exploration Vehicle: Simulation of Operations in Proximity to a Near Earth Object 

Heather Kline 

Mentors: J. Balaram and Marco Quadrelli 

This project focuses on simulating the dynamics of an MMSEV (Multi-Missions Space Exploration Vehicle) on 

rendezvous to collect samples from a NEO (Near Earth Object), modeling the control characteristics of the 

spacecraft using DSENDS (Dynamics Simulator for Entry, Descent, and Surface landing). A mission of the MMSEV is 

to collect samples from an asteroid, a mission which requires the spacecraft to be able to navigate to an orbit 

which keeps it stationary over an area of a spinning asteroid while a robotic arm interacts with the surface. The 

capability to model reaction wheel and thruster control has been added. Additional capabilities added include 

station keeping, where the spacecraft holds itself steady over a spot on the rotating asteroid's surface. In addition 

to the added functionality the results of a parametric study of control gain sensitivity and Monte Carlo simulation of 

the behavior near the surface are discussed. The functions used to control the spacecraft with simulated reaction 

wheels and thrusters have been written in a modular way such that they could easily be used in other systems, and 

it is compatible with multiple thruster allocation methods. 

Jupiter’s Magnetosphere: An Analysis of Electron Pitch Angle and Heavy Ion Loss to Europa 

Nicholas Knezek 

Mentor: Michael Kokorowski 

Jupiter’s magnetosphere is the largest object in the solar system and it dominates the local space environment. 

Through past analysis of Pioneer 10 and 11, Voyager 1 and 2, Ulysses, and Galileo data, several regions and 

important transport processes have been defined within it. The project has two objectives: examine the transition 

region between the inner and outer magnetosphere and identify losses of heavy ions to Jupiter’s moons, Europa in 

particular. For the first objective, data from the magnetometer and Energetic Particle Detector on Galileo were 

utilized to reproduce previous results showing an electron pitch angle distribution change from a trapped 

distribution in the inner magnetosphere to a field-aligned distribution in the outer magnetosphere. Due to 

limitations of the Galileo orbiter transmission rate and reduced dataset, further results were impossible to obtain. 

Focus was then shifted to analysis of data from the Heavy Ion Counter. Using oxygen measurements at various 

energy levels, phase space density were calculated for ions with various magnetic moments, then used to examine 

whether significant losses of Heavy Ions occur near the orbit of Europa. 

Improving Measurement Systems and Instrumentation 

Stephen Krach 

Mentor: Lauren Halatek and Ian Ruiz 

The Thermal Energy Conversion Technologies group is in need of updated test systems. My objective is to convert 

the existing Visual Basic code that commands the PID controller into a maintainable LabVIEW program. By using 

flow charts, original code, and help from peers, I am on my way to completing this important sub-function upgrade. 

Electronic Ground Support Equipment consists of a rack complete with test equipment and custom electronics (aka 

chassis). My task is to create two templates; one will output the rack’s center of gravity, and the other will validate 

adequate cooling for the rack. 

157

A low cost breakout box (BOB) has been designed that is half the cost of the conventional BOB. My objective is to 

finalize this design by creating a bill of materials, receiving quotes, and gathering the parts for a technician to 

assemble. I will also be working with quality assurance to develop an appropriate testing procedure to ensure the 

BOB is qualified to connect to flight hardware. 

Orbit Determination Techniques for the Mars Reconnaissance Orbiter 

Sandeep Krishnan 

Mentor: Allen Halsell 

Orbit determination and maneuver design are critical components of navigation for a spacecraft. Orbit 

determination is defined as obtaining the knowledge of a vehicle’s motion in space in a specified coordinate system. 

The first part of this process involves estimating the state of the spacecraft using a least squares method. This is 

accomplished by correcting a nominal trajectory in order to “best” fit observed measurements. Once an estimate 

has been made, the spacecraft’s state can be predicted using numerical integration. This entire process is repeated 

continually as the spacecraft’s motion evolves. The Mars Reconnaissance Orbiter (MRO) is a spacecraft developed 

in order to explore Mars through remote sensing observations. Launched in 2005, it reached its final science orbit 

in March 2006 and began its primary science phase. For deep space missions such as MRO, the primary method of 

tracking utilizes radiometric techniques. The MRO navigation team is responsible for reconstructing and predicting 

the spacecraft’s orbits based on Doppler data provided by the Deep Space Network (DSN). For MRO, the orbit 

determination process uses the Monte software. New versions of Monte and other software tools are currently 

being validated. This report details the testing procedure as well as the preliminary results. 

DARTS Simulation for a Dexterous Manipulator (ARM) 

Calvin Kuo 

Mentor: Abhinandan Jain 

The dynamics and real time simulations (DARTS) lab at the Jet Propulsion Laboratory (JPL) provides physics 

simulations for a number of rovers and spacecraft to supplement concept design and hardware testing. JPL 

currently has a project developing perception and grasping algorithms for a dexterous manipulator for DARPA. 

As the algorithms and tasks for the manipulator become more complex, testing on the physical hardware becomes 

more time consuming and potentially dangerous to both the hardware and the people operating it. Thus, the need 

for an offline simulator that can accurately simulate the physics and dynamics of the system become more 

prevalent. While creating models for simulations of multi-body systems in DARTS has been implemented for 

projects such as the Mars Science Laboratory, this simulation also requires an interface for the manipulator control 

software. Thus the focus of this project will be to create a simulator of the manipulator, provide an interface for the 

control software, and to verify that the simulated version of the manipulator behaves identically to its hardware 

counterpart. 

Performance Analysis of Magnetic Field Shapes in Hall Thrusters 

Maria C. Lang 

Mentor: Richard R. Hofer 

The shape and strength of the magnetic field in Hall thrusters is the dominant factor affecting plasma stability, 

thermal balance, performance, and life. Of particular interest to this project is the effect of changes in the shape of 

the magnetic field on thruster performance by modifying the applied field in a 6 kW laboratory Hall thruster. 

A thrust stand, ion flux probe, ExB probe, RPA, and emissive probe are used to measure the thrust, ion current 

density, ion species fractions, ion energy, and plasma potential for the two distinct magnetic field configurations. 

The differences in the fields are enabled through the use of an auxiliary “trim” coil that modifies the magnetic field 

in the near-anode region of the thruster. Data from the diagnostics is analyzed and applied to an analytical model 

of thruster performance. The model describes the dominant processes affecting thrust efficiency that is then used 

to identify which changes in the plasma structure are responsible for changes in thruster performance between the 

two magnetic field configurations. 

Framework for Testing Detection, Diagnostic, and Remediation Systems Within the Smart Grid 

Shing-hon Lau 

Mentors: Bryan Johnson and Kymie Tan 

America’s electrical grid is currently undergoing a significant paradigm shift aimed at producing a highly reliable 

critical national infrastructure for power – a Smart Grid. A key objective of this shift is to create a resilient electrical 

grid that is able to autonomously detect, diagnose, and remediate impediments to the delivery of power. 

Impediments can range from downed power lines to cyber-attacks to natural disasters. Measuring the effectiveness 

of such resilience lies solely in the domain of a rigorous testing process. Since the Smart Grid is a massive, 

complex cyber-physical system – one where computational and physical elements are closely intertwined – 

designing such a rigorous testing process will be challenging. The purpose of our work is to lay out a guiding 

158

framework and principles for testing the resilience of a complex, cyber-physical system. By clearly identifying the 

appropriate metrics, data characteristics, and error sources that comprise the critical components of a rigorous test 

regime, this work aims to make a significant contribution in the area of measuring the true resilience of a cyberphysical 

system. 

Radar and Lidar Remote Sensing of Forests 

Diana Liskovich 

Mentor: Marc Simard 

Using radar and lidar data, the aim is to improve 3D rendering of terrain, including digital elevation models (DEM) 

and estimates of vegetation height and biomass in a variety of forest types and terrains. The 3D mapping of 

vegetation structure and the analysis are useful to determine the role of forest in climate change (carbon cycle), in 

providing habitat and as a provider of socio-economic services. This in turn will lead to potential for development of 

more effective land-use management. The first part of the project was to characterize the Shuttle Radar 

Topography Mission DEM error with respect to ICESat/GLAS point estimates of elevation. We investigated potential 

trends with latitude, canopy height, signal to noise ratio (SNR), number of lidar waveform peaks, and maximum 

peak width. Scatter plots were produced for each variable and were fitted with 1st and 2nd degree polynomials. 

Higher order trends were visually inspected through filtering with a mean and median filter. We also assessed 

trends in the DEM error variance. Finally, a map showing how DEM error was geographically distributed globally 

was created. The second part of this project consisted of mapping vegetation height in Canada. A set of decision 

trees (DT) trained with ICEsat estimates of canopy height was produced. A set of input layers describing elevation, 

climate and vegetation type were used. These decision tree models were then applied to model canopy height 

throughout Canada. 

Sputtering Chamber Hardware Modification 

Christopher Matthes 

Mentor: Charles Hays 

Sputter deposition is the process of ejecting atoms from a target material due to the impact of gas ions, causing 

the resulting material to be deposited onto a substrate and form a thin film. JPL’s sputtering facility allows the 

creation of alloyed thin films using multiple targets, which are used for research of fuel cell cathode catalysts with 

improved oxygen reduction reaction (ORR) capabilities. The objective for this project was to enable better control 

of sputtering conditions through the implementation of computer-operated settings, which were previously 

manually controlled. This required hardware modifications for the gas-handling and high-vacuum systems, while 

introducing a computer control and data acquisition system to regulate various parameters of the sputtering 

process. Specifically, the introduction of an electronically interfaced pressure sensor and mass flow controller setup 

allowed for the regulation of desired parameters using the electronically obtained data. Installation of the new 

system hardware necessitated an application of design fundamentals to ensure proper assembly of all components 

and to address additional mechanical obstacles that surfaced through this process. 

Verification and Validation of Granular Media Simulation Capability With Applications in NEO Anchoring 

Hammad Mazhar 

Mentor: Marco Quadrelli 

This contribution describes an approach used to simulate and validate a simulation capability that is able to 

simulate granular dynamics simulations with millions of objects. This massively parallel granular dynamics 

framework relies on ubiquitous Graphics Processing Unit (GPU) cards that are capable of performing trillions of 

floating point operations per second and can simultaneously process hundreds of computational threads, making 

them ideally suited for the task at hand. Utilizing a variety of experimental data, the simulation capability was 

validated with existing literature. Tests included a lift/drag and ball drop test in a bed of granular material, a steady 

state vibration test, a brazil nut test and a penetration test. This parallel simulation capability was used to 

investigate two types of anchoring systems, and inflatable anchor and a helical anchor, when deployed on a Near 

Earth Object (NEO) in a low gravity environment. In order to successfully simulate the anchor and granular 

material in this environment, the simulation capability was augmented with a cohesion model capable of capturing 

the necessary behavior and short range forces associated with NEOs. Described herein are the results associated 

with the experimental validation and anchor tests along with a recommendation on which anchor design is most 

promising. 

Life Test Testbed Development for SMAP’s Attitude Control System Thruster 

Colleen McAvoy 

Mentors: David Vaughan, Corinne Gatto, and Barry Nakazono 

The attitude control system thrusters for the SMAP mission exceed the qualification through the number of cold 

starts, total number of pulses, and total propellant. A life test will be conducted within a year to analyze the 

performance of an individual thruster over a wide range of operating parameters including steady state, pulsed, 

and temperature extremes. A schematic representing the layout of the testbed for the hydrazine fueled 

159

monopropellant system is being developed to the scope of the test and modified for potential hazards arising from 

hydrazine’s toxicity and combustibility. In parallel, a virtual interface is being developed with NI LabVIEW and will 

be used to operate the life test. Upon completion of the schematic, the necessary hardware will be ordered and 

manufactured to begin the testbed assembly. In addition, the test procedure will need to be written based on the 

final schematic design. A future task will be to create an interactive graphic user interface for simulating the test 

procedure in conjunction with the test to ensure a safe testing environment and to visually confirm the required 

test actions. 

Martian Atmospheric Modeling of Scale Factors for MarsGRAM 2005 and the MAVEN Project 

Christopher McCullough 


For spacecraft missions to Mars, especially the navigation of Martian orbiters and landers, an extensive knowledge 

of the Martian atmosphere is extremely important. The generally-accepted NASA standard for modeling of the 

Martian atmosphere is called the Mars Global Reference Atmospheric Model (MarsGRAM), which was developed at 

Marshall Space Flight Center. MarsGRAM is useful for tasks such as aerobraking, performance analysis and 

operations planning for aerobraking, entry descent and landing, and aerocapture. For the Mars Atmosphere Volatile 

Environment (MAVEN) Project, the navigation team uses this atmospheric model to target a specified density 

corridor. Using data acquired from the Mars Reconnaissance Orbiter (MRO), Mars Global Surveyor (MGS), and 

Odyssey, the MarsGRAM model can be compared to the actual densities felt by the spacecraft by assigning 

MarsGRAM a scale factor. By examining these scale factors, a stochastic model with altitude, latitude, and seasonal 

dependences can be created that effectively models the variations in the atmospheric scale factor. 

DROP: The Durable Reconnaissance and Observation Platform 

Clifford F. McKenzie 

Mentors: Aaron Parness 

Robots can provide a remote presence in areas that are either inaccessible or too dangerous for humans. However, 

robots are often limited by their ability to adapt to the terrain or resist environmental factors. The Durable 

Reconnaissance and Observation Platform (DROP) is a lightweight robot that addresses these challenges with the 

capability to survive falls from significant heights, carry a useable payload, and traverse a variety of surfaces, 

including climbing vertical surfaces like wood, stone, and concrete. DROP is manufactured using a combination of 

rapid prototyping and shape deposition manufacturing. It uses microspine technology to create a new wheel-like 

design for vertical climbing. To date, DROP has successfully engaged several vertical surfaces, hanging statically 

without assistance, and traversed horizontal surfaces at approximately 30 cm/s. Unassisted vertical climbing is 

expected in the very near future. DROP can currently survive falls from up to 2 meters; a new design being 

assembled now should improve this substantially. Work on DROP is ongoing and both its mobility and survivability 

are still improving. Future efforts will focus on improving the microspine wheels, selecting more resilient materials, 

customizing the controls, and performing more rigorous and quantifiable testing. 

Optimization of PMA-PCR Protocol for Viability Detection of Pathogen Surrogates 

Brian Mikkelson 

Mentors: Adrian Ponce, Christine Lee, and Aaron Noell 

In the event of a bioterrorism attack or epidemiologic outbreak, the ability to determine whether an area is safe, as 

well as what species is present, in the shortest amount of time, is essential. This project’s objective was to optimize 

protocols for use in creating a portable instrument capable of determining viability and identification 

simultaneously. This is made possible by the use of the DNA-intercalating agent propidium monoazide (PMA) 

combined with molecular biology analyses including PCR. PMA selectively binds to dead cells, inhibiting polymerase 

activity, making DNA unable to be amplified by PCR. By inhibiting the amplification of DNA of any dead cells, and 

using specific primers, PMA-PCR results consist purely of viable samples of the target species. PMA-PCR protocol 

was optimized and the correlation between viable to dead cell ratio with PMA-PCR output was determined using 

methods of DNA extraction, PCR, phase-contrast microscopy, cell culture, and Bacillus spore germinability assay. 

Interference with PCR amplification by primer dimers was solved by keeping the reagents on ice until being placed 

in the thermal-cycler. The loss of culturability, by heat-killing, did not prevent amplified PCR products, which 

supports the use of PMA to prevent amplification and differentiate between viable and dead cells. 

Thermal Modeling of Asteroids Observed With the Wide-Field Infrared Survey Explorer (WISE) 

Wenli Mo 

Mentor: Amy Mainzer 

The Wide-field Infrared Survey Explorer (WISE) has observed ~157,000 minor planets, many of them previously 

unknown. With these observations, astronomers can determine important physical parameters, including the 

albedo and diameter of the object, through thermal modeling. I will explain the methods of the Near Earth Asteroid 

160

Thermal Model (NEATM) using the object’s observed data from visible and infrared wavelengths. Also, I will present 

the general results of the thermal modeling of near earth objects (NEOs) observed by WISE and future analysis to 

be conducted with the resulting physical parameters. 

Multi-Angle Rear Viewing Endoscopic TooL (MARVEL): Mechanical Design and Analysis 

Jack M. Mondry 

Mentors: Harish Manohara and Michael Shearn 

Minimally Invasive Surgery (MIS) provides the patient great advantages including shortened recovery time, less 

cosmetic scarring, and less risk of infection. However, this type of procedure is generally more complex than a 

traditional open surgery, especially when performing neurosurgery. A major factor of this complexity has to do with 

the reduced field of view the surgeon must operate in. Traditional endoscopes provide the surgeon a fixed 

viewpoint and often only in two dimensions which hinders depth perception and therefore decreases accuracy and 

increases surgery time. This project involves implementation of a stereoscopic (3D) camera to an actuated revolute 

joint that will provide the surgeon alternate viewing angles of the surgical site. Due to the working environment of 

the MARVEL and the unacceptable consequence of device failure, strict design requirements were applied. A 

concept was developed that allows the tool to be sterilized for multiple uses while maintaining a small form factor 

and meeting all design requirements. Multiple actuation methods were prototyped and analyzed for performance. 

Wideband Spectroscopy: The Design and Implementation of a 3 GHz, 2048 Channel Digital 

Spectrometer 

Ryan Monroe 

Mentor: Robert Jarnot and Paul Stek 

A state-of-the-art digital Fourier Transform spectrometer has been developed, with a combination of high 

bandwidth and fine resolution unavailable elsewhere. Analog signals consisting of radiation emitted by constituents 

in planetary atmospheres or galactic sources are downconverted and subsequently digitized by a pair of interleaved 

Analog-to-Digital Converters (ADC). This 6 Gsps (giga sample per second) digital representation of the analog 

signal is then processed through an FPGA-based streaming Fast Fourier Transform (FFT), the key development 

described below. Digital spectrometers have many advantages over previously used analog spectrometers, 

especially in terms of both accuracy and resolution, both of which are particularly important for the type of 

scientific questions to be addressed with next-generation radiometers. The implementation, results and underlying 

math for this spectrometer, as well as potential for future extension to even higher bandwidth, resolution, channel 

orthogonality and radio interference detection needed to support proposed future advanced atmospheric science 

and radioastronomy, are discussed. 

Mission Planning and Sequencing Investigation of Third Party Work Flow Software 

Michael Mozingo 

Mentor: Barbara Streiffert 

The Mission Planning and Sequencing (MPS) element of the Multi-Mission ground System and Services (MGSS) 

provides space missions with software that meets the demands of multiple missions. The requirements on this 

software set are to be able to plan spacecraft activities, sequence commands to the spacecraft, package the 

commands and notify the operations team that the command is ready to be radiated to the vehicle. The Automated 

Sequence Processor (ASP) is used to perform all of these functions if the command is a non-interactive command 

(NIC). Non-interactive commands are guaranteed not to harm the spacecraft. To update the ASP the use of a 

commercially available workflow engine has been investigated. Two workflow engines have been prototyped and 

the work has been documented with the pros and cons of various engines. In addition, a recommendation has been 

made. The following paper discusses the prototypes and the process used for evaluation. 

Conductor: Kinect Motion-Tracking for NASA Outreach and Robotics Potential 

Joshua Nuernberger 

Mentor: Victor Luo 

The scope and objective of Project Conductor is to apply natural user interfaces, utilizing the Microsoft Kinect, for 

NASA Outreach application and robotics control. The methodology of the project emerges from a collaborative 

process beginning in research, brainstorming, conceptualizing, prototyping, and then culminating in development. 

Current projects include the topics of Space Camp and NASA Robotics. From these, gameplay mockups and 

concepts are presented to the team for integration. My main area of contribution is focused on the conceptual 

potential of the experiences, including game design and establishing a visual style. However, it also includes some 

aspect of the developmental side of the project, using the Unity game engine. The final goal, working with the rest 

of the team, is to create fun and educational gameplay experiences for NASA Outreach. 

161

Spectral Emissions of a Pulsed Nitrogen Cold-Gas Thruster and a Hall Thruster System 

Kelsey S. O’Connor 

Mentors: Lee Johnson and David Conroy 

Quantitative spectral emission measurement were performed to identify and investigate the nature of charge 

transfer, recombination, and back-scattering created when a pulsed nitrogen cold-gas thruster and a hall thruster 

are simultaneously operated in close proximity. The measurements were taken with the specific target of 

identifying and investigating the nature of spectral changes in the ionized plasma plume of the hall thruster in 

order to gain a comprehensive understanding of the effect these interactions will have on the lifetime and 

effectiveness of the Hall thruster and the spacecraft it is attached to. 

Developing a Plug-and-Play Framework for the ISAAC iBoard 

Kevin Ortega 


iBoard is a common integrated instrument-control-and-computing FPGA-based board used to simultaneously 

process telemetry from attached instruments, such as cameras, radars, and so forth. Since iBoard will handle 

multiple types of instruments, there needs to be an autonomous way for detecting and communicating with its 

connected devices. To implement it, a plug-and-play framework has been developed for iBoard and its attached 

peripherals. We have performed a survey on other plug-and-play platform technology such as Universal Serial Bus 

(USB), Android software stack, Universal Plug-and-Play (UPnP), etc. Also, iBoard and peripheral interconnection 

requirements have been developed. The framework has been developed and the prototype has also been 

implemented. This framework allows for a compatible peripheral to be easily connected to and communicated with 

an iBoard. 

Processing and Analyzing Snow Cover Data From the Moderate Resolution Imaging Spectroradiometer 

(MODIS) Instrument 

Andrew Park, Jr. 

Mentor: Kevin Hand 

Moderate Resolution Imaging Spectroradiometer (MODIS) is an instrument aboard the Terra and Aqua satellites 

that images the Earth surface to provide scientists with invaluable data about Earth’s lands, oceans, and lower 

atmosphere. Of the many data products that have been transferred to ground stations for analysis, snow cover 

data is especially relevant to current research of icy moons such as Europa, which many scientists and 

astrobiologists believe to be habitable for extraterrestrial life. As part of an ongoing effort to better understand the 

chemical phenomena that occur during freezing, MODIS data from the Aqua satellite will be processed using 

modified Matlab scripts to look at snow cover from a 500 meter spatial resolution at 8-day coverage intervals. With 

this data, it will then be possible to look at changes in the snow cover in key regions such as Barrow, Alaska, where 

pockets of methane gas are beginning to form. By investigating these regions, scientists hope to find trends in 

these methane levels and apply this knowledge to current understanding of the complex chemistry that occurs in 

icy environments. 

Real-Time Event Management for the Mars Relay Operations Service 

Corey Peterson 

Mentors: Michael Wallick and Daniel Allard 

The Mars Relay Operations Service (MaROS) comprises a number of tools to coordinate, plan, and visualize various 

aspects of the Mars Relay Network. As part of MaROS, I have developed and implemented a feature set that 

operates on several levels of the software architecture. These levels include a web-based user interface, a backend 

“ReSTlet” built in Java, and databases which store the data as it is received from the network. 

My primary goal this summer was to develop a real-time event notification and management system, so mission 

teams can track and act upon events on a moment-by-moment basis. It additionally allows users to post important 

event-related messages for team members within the same feed. 

My workflow has been as follows: collaborate with the end-users and developers to identify the features they would 

like to see, assemble the back-end architecture to manage communication with the databases, as well as 

recognizing, filtering, and sending event information to the user interface. Lastly, I build the interface components 

that comprise all the elements of the event notification system. Once this basic process is completed, I iterate and 

expand upon my design by adding or changing features based on developer and user feedback. 

162

Ocean Forecasting in the Gulf of Mexico: User Interface and Accuracy Analysis 

Heidi K. Peterson 

Mentor: Yi Chao 

Currently JPL is creating an ocean forecasting system for the Gulf of Mexico and I was charged with identifying 

users and the user’s needs as well as checking the accuracy of the water velocity hindcasts conducted by the 

program. Since the most interested user demographic was oil and oil support companies I designed a user 

interface for oil operators and researched the ocean conditions of interest to oil companies. There was some 

concern for large changes in temperature and salinity; however the greatest concern was eddies. Eddies 

occasionally form and break free from the Loop Current that enters the Gulf between Cuba and the Yucatan 

Peninsula and leaves the Gulf between Florida and Cuba. These eddies cause changes in water velocity that are 

dangerous to offshore oil platforms and offshore construction. I created a user interface for oil platform operators, 

which will give them the forecasting information they need in an easy to use format. I am also developing an 

animated graph showing actual and model water column velocity data. I hope to also complete the statistics for 

variances and error in the model. 

Orbit Determination and Navigation Software Testing for the Mars Reconnaissance Orbiter 

Alex J. Pini 

Mentors: Allen Halsell and Premkumar Menon 

During the extended science phase of the Mars Reconnaissance Orbiter’s lifecycle, the operational duties pertaining 

to navigation primarily involve orbit determination. The orbit determination process utilizes radiometric tracking 

data and is used for the prediction and reconstruction of MRO’s trajectories. Predictions are executed twice per 

week and are used to recommend corrections to the satellite, such as ephemeris updates and maneuver delta-v’s. 

Reconstructions, which incorporate a batch estimator, provide precise information about the spacecraft state to be 

synchronized with scientific measurements. These tasks were conducted regularly to validate the results obtained 

by the MRO Navigation Team. Additionally, the team is in the process of converting to newer versions of the 

navigation software and operating system. The capability to model multiple densities in the Martian atmosphere is 

also being implemented. However, testing outputs among these different configurations was necessary to ensure 

compliance to a satisfactory degree. 

Coronae Formation on Venus via Extension and Lithospheric Instability 

Danielle Piskorz 


Coronae are semi-annular volcanotectonic formations that dot the surface of Venus. There exist many hypotheses 

that explain their origin, but perhaps the least understood is the origin of the coronae located 150 to 1500 km from 

their associated rift. The goal of this project is to show that a mantle plume associated with a rift can melt the 

lower lithosphere and cause lithospheric dripping into the upper mantle, leading to extension and surface stresses. 

Using a finite element code that solves equations for the conservation of heat, mass, and momentum, we vary 

lithospheric thickness, mantle density, and the mantle's viscosity structure. We compare the predicted surface 

topographies, gravity anomalies, and melt compositions of our successful models to SAR and VIRTIS observations. 

By further characterizing the relationship between rifts and coronae, we may be able to better understand the 

subsurface dynamics of a single-plate planet. 

High Frequency Radiometer Testbed 

Stephen Pomes 

Mentors: Pekka Kangaslahti and Douglas Dawson 

JPL has been a leader in the development of internally calibrated spaceborne microwave radiometers. These 

radiometers are used to measure wet tropospheric path delay produced by the changing index of refraction of the 

atmosphere resulting from the variable concentration of water. Though these radiometers have all operated at 

frequencies below 40 GHz, JPL is currently investigating the feasibility of the internal calibration approach for 

microwave radiometers operating in the 90-200GHz range, particularly at the frequencies of 90GHz, 130GHz, and 

166GHz for the Surface Water and Ocean Topology mission. The radiometers consist of several high-performance 

broadband low-noise amplifiers, isolators, filters, a detection circuit, and a data acquisition unit. Since parameters 

such as gain and noise temperature are variable with time, it is necessary to test various internal calibration 

sources and approaches in order to assess their stability inherently and over a range of temperatures as well as the 

overall system impacts of such components. 

163

ATHLETE: A Rotary Percussion Drill Tool Package for Rock or Ice Drilling 

Jay Qi 

Mentor: Matthew Frost 

The All-Terrain Hex-Limbed Extra-Terrestrial Explorer (ATHLETE) is a vehicle concept developed at the Jet 

Propulsion Laboratory based on the wheel-on-limb approach to mobility, allowing it to drive, walk, climb, and even 

manipulate its environment with its limbs. Through a standardized fixture and a clamp mechanism on each limb, 

ATHLETE is able to extract various tool attachments from a "tool belt" as necessary during exploration missions. A 

rotary percussion drill tool package concept was designed to integrate with ATHLETE's quick-disconnect tool clamp 

with the purpose of allowing ATHLETE to drill into solid rock or ice surfaces. This tool concept was developed to be 

simple, robust, and versatile, using a commercial rotary percussion drill mechanism powered by an easily 

changeable rechargeable battery and controlled wirelessly. The design also includes a mechanism for ATHLETE to 

change drill bits. With such a drill tool, ATHLETE will be able to drill into rock or ice during missions on the Moon, 

other planets, or asteroids, whether for anchoring, obtaining core samples, or other tasks. 

Reducing Color Rivalry in Imagery for Complementary Multiple Bandpass Filter Based Stereo Endoscopy 

Allen Ream 

Mentors: Harish Manohara and Sam Bae 

A pair of complementary multiple bandpass filters (CMBF) can be used to create spatially separated pupils in a 

traditional lens and imaging sensor system, allowing for the passive capture of stereo video. This method is 

especially useful for surgical endoscopy where smaller cameras are needed to provide ample room for manipulating 

tools while also granting improved visualizations of scene depth. The salient issue in this process is that, due to the 

complementary nature of the filters, the colors seen through each filter do not match each other, and also differ 

from colors as seen under a white illumination source. A color correction model was implemented that included 

optimized filter selection, such that the degree of necessary post-processing correction was minimized, and a 

chromatic adaptation transformation that attempted to fix the imaged colors tristimulus indices based on the 

principle of color constancy. Due to fabrication constraints, only dual bandpass filters were feasible. The theoretical 

average color error after correction between these filters was still above the fusion limit meaning that rivalry 

conditions are possible during viewing. This error can be minimized further by designing the filters for a subset of 

colors corresponding to specific working environments. 

A Cloud Imager in the Thermal Infrared 

Brian J. Redman 

Mentor: Sabino Piazzolla 

Information about cloud patterns is useful for climate science studies and Earth-space optical communications 

research. Thermal infrared sky imaging is a technique that records cloud patterns by measuring the heat radiation 

emitted by the clouds. This method is particularly well suited for continuous ground-based measurements of cloud 

cover statistics because it functions equally well during day and night. There is an interest in exploring the 

capabilities of low-cost systems capable of imaging the entire sky. A system using a metal dome to reflect the sky 

fits these criteria. In order for this system to be used algorithms have been developed to compensate for 

distortions from the dome, in particular uneven heating of the dome. 

Mars Sensor System Electrical Cable Management and Distributed Motor Control Computer Interface 

Robin Reil 

Mentors: Mohammad Mojarradi and Colin McKinney 

The success of JPL’s Next Generation Imaging Spectrometer (NGIS) in Earth remote sensing has inspired a followon 

project, the Mars Sensor System (MSS). One of JPL’s responsibilities in the MSS project involves updating the 

documentation from the previous JPL airborne imagers to provide all the information necessary for an outside 

customer to operate the instrument independently. As part of this documentation update, detailed electrical cabling 

diagrams were created to provide JPL technicians with clear and concise build instructions, and a database was 

produced to track the status of cables from order to build to delivery. 

Simultaneously, a distributed motor control system is being developed for potential use on the 2018 Mars rover 

mission. This system would significantly reduce the mass necessary for rover motor control, making more mass 

space available to other important spacecraft systems. The current stage of the project consists of a desktop 

computer talking to a single “cold box” unit containing the electronics to drive a motor. In order to test the 

electronics, a graphical user interface (GUI) was developed using MATLAB to allow a user to send simple 

commands to the cold box and display the responses received in a user-friendly format. 

164

Dew Point Monitoring System for Waveguide Arc Testing 

Ryan D. Rickard 

Mentors: Juan Ocampo and Michael Britcliffe 

The objective of this project is to create a system for monitoring moisture contamination levels within the 

waveguide of an X band transmitter that is part of the Deep Space Network (DSN). The introduction of moisture 

contaminants within a waveguide is believed to produce a rare phenomenon of RF (radio frequency) arcing that can 

potentially damage and even destroy transmitter components. By modifying a previously created monitoring 

system’s code and hardware components, a sensor was integrated with a single board computer capable of 

displaying dew point readings onto an LCD screen and Ethernet connection. Following the design and build of the 

monitoring system, the sensor was placed at the end of a custom waveguide and monitored injected moisture 

searching for a correlation between dew point (moisture) and RF arcing. With early detection of moisture in the 

waveguide using a dew point sensor, DSN ground station operators can prevent damage to transmitter components 

and schedule needed repairs and maintenance. 

Earth Science Data Processing in the Cloud 

Andres Riofrio 

Mentors: Khawaja S. Shams and Tom Soderstrom 

Cloud computing offers exciting new opportunities to many kinds of scientific missions, offering a cost-effective 

alternative to in-house cluster-based systems. Low budget airborne missions such as such as the Carbon in Arctic 

Reservoirs Vulnerability Experiment (CARVE) mission are typically deprived of the resources required for timely 

processing; furthermore, the algorithm is highly parallel and processing is bursty, making cloud computing a 

natural fit. Along with the Airborne Cloud Computing Environment (ACCE) task at the Jet Propulsion Laboratory 

(JPL), we built and developed software systems that enable the Level 2 Full Physics processor to run in Amazon's 

Elastic Compute Cloud (EC2) using Polyphony, a workflow orchestration framework developed at JPL that supports 

job execution in a cloud environment, as well as in a local environment, and any combination thereof. This software 

is applicable to other earth science missions, including OCO-2. We also developed cloud management techniques 

that will be useful in the adoption of cloud computing by missions at NASA. 

Mars Science Laboratory Flight Software Boot Robustness Testing 

Brian Roth 

Mentors: Danny Lam, Steve Scandore, and John Lai 

On the surface of Mars, the Mars Science Laboratory will boot up its flight computers every morning, having 

charged the batteries through the night. This boot process is complicated, critical, and affected by numerous 

hardware states that can be difficult to test. The hardware test beds do not facilitate testing a long duration of 

back-to-back unmanned automated tests, and although the software simulation has provided the necessary 

functionality and fidelity for this boot testing, there has not been support for the full flexibility necessary for this 

task. Therefore to perform this testing a framework has been built around the software simulation that supports 

running automated tests loading a variety of starting configurations for software and hardware states. This 

implementation has been tested against the nominal cases to validate the methodology, and support for 

configuring off-nominal cases is ongoing. The implication of this testing is that the introduction of input 

configurations that have yet proved difficult to test may reveal boot scenarios worth higher fidelity investigation, 

and in other cases increase confidence in the robustness of the flight software boot process. 

Correlations Between Technical Performance and Cost Growth: A Case Study Using Historical Cost Data 

Dean Rowley 



indicated that there are common reasons for cost growth where strong correlations existed between technical 

performance and cost/schedule performance. This group research project investigates historical data from the 

13 most recent JPL flight projects to determine trends and correlations between technical and programmatic 

parameters to predict cost growth. Each of the 13 projects is analyzed using 5 major data sets: technical 

complexity, subcontractor performance, risk trends, margin analysis and schedule. Each data set is analyzed to 

determine which correlations exist. My area of focus for this group project is to establish the cost per month and to 

set a baseline for the 5 data sets that acts as a common comparison between the analyses. Once all technical and 

cost data is collected regression analysis and modeling is performed to establish trends. It is expected that certain 

aspects of technical performance parameters will lend themselves to be more accurate cost predictors than others. 

Conclusions derived from this group research can be utilized to form a historical data base for future projects and 

compared against the current in-work flight projects to predict cost growth. 

165

Development of Interactive Tools for Quantitative Spectral Analysis 

Garreth Ruane 

Mentor: J. Brad Dalton 

Infrared reflectance spectroscopy is a powerful tool for understanding the chemistry of planetary surface materials. 

Detailed analyses of specific absorption features can provide information that is uniquely diagnostic of composition. 

Modern spectroscopy relies on software applications to determine spectral properties such as absorption center 

frequencies (wavelengths), absorption band depths, and spectral feature shape. Removal of the continuum, defined 

by anchor points to either side of an absorption feature, permits analysis of a given feature independent of the 

influence of other absorption features - some of which may arise from other materials within the scene. 

Abundances of materials within a scene may be estimated by the application of linear mixing algorithms, which 

treat the spectrum as arising from a linear combination of endmember spectra. Many approaches to linear 

unmixing use matrix inversion techniques that limit the number of endmembers that may be included in a spectral 

library. I have developed software tools to enable researches to apply continuum-removal and linear mixing 

methods to analyze spacecraft spectral observations using spectral libraries of arbitrary size, thus expanding 

current capabilities and allowing the description of complex planetary surfaces. I will illustrate these approaches 

with example data from the Galileo spacecraft observations of Europa. 

F6 Space System: A Model of the Implementation and Operation of a Cluster 

Tyler Ryan 

Mentor: Steven Cornford 

DARPA has funded a number of teams to develop its Fractionated Spacecraft vision. A team led by JPL has been 

tasked to develop a tool for the evaluation of the Business Case for this fractionated system architecture. The 

purpose of this evaluation is to understand under what conditions and constraints the fractionated architecture 

makes more sense from the cost/benefit perspective than the traditional monolithic paradigm. One key issue is to 

create a model complete enough to provide an accurate representation of the Business Case, yet executes fast 

enough to allow a thorough tradespace exploration. Such a model has been created in a discrete event simulator 

called Arena, representing both the implementation and operation phases of a cluster of N spacecraft. The 

implementation model includes the design, build, assembly, and test phases of major subsystems, their integration, 

and subsequent launch sequencing . The operations model includes gathering and managing power, and taking, 

cross-linking, and downlinking data. The model also accounts for various degrees of fractionation, as well as 

options for responding to stimuli, including launch vehicle failure, component failure, technology obsolescence and 

changes in mission objectives. The model, how it was developed, and the results of its execution will be discussed. 

Mars Atmospheric Modeling for the MAVEN mission 

Carly Sakumura 


Accurate predictions of the atmospheric density on Mars are crucial for the Mars Atmosphere and Volatile Evolution 

(MAVEN) mission. In order to accomplish its scientific objectives, this mission is targeting a density corridor at 

periapsis of 0.05 to 0.15 kg/km3 during normal science operations with five additional “deep dips” to densities of 

2.0 to 3.5 kg/km3. The current model of the Martian atmosphere is the Mars Global Reference Atmospheric Model 

(Mars-GRAM). While this program is accurate in predicting most trends and an estimate of the atmospheric density, 

a more accurate model is desired in order to better meet the strict density requirement imposed upon this mission. 

In order to accomplish this goal the scale factor between the predicted and measured atmospheric density is 

analyzed from the previous Mars Reconnaissance Orbiter, Odyssey, and Mars Global Surveyor missions. A 

statistical model of the variation as well as trends in the scale factor due to altitude, latitude, and seasonal effects 

is developed to both improve and better understand this estimate. 

Standardization of Spore Inactivation Method for PMA-PhyloChip Analysis 

Michael Schrader 

Mentor: Kasthuri Venkateswaran 

Development of a more sensitive method that will enumerate viable microorganisms from spacecraft and 

associated surfaces is warranted to facilitate future life-detection missions. Current culture-based and nucleic-acidbased 

polymerase chain reaction (PCR) methods fail to do so. The Biotechnology and Planetary Protection group at 

JPL is currently involved in the development of a new molecular method based on the use of propidium monoazide 

(PMA), in combination with DNA microarray (PhyloChip). PMA can only penetrate the membrane of dead cells and 

then intercalates into DNA, making it unavailable for PCR. 

Two bacterial spores will be used: Bacillus subtilis 168 and Bacillus pumilus SAFR-032. Spores were harvested by 

repeated centrifugation and washing in buffer solutions at 4 °C. The standardization of the spore inactivation 

method with no detrimental effect on DNA will then be implemented. Results indicate that heat treatment at 90 0 C 

for 90 min, completely inactivates (non-viable) spores without damaging DNA, confirmed by the Quantitative PCR 

166

(Q-PCR) method. After successful standardization of heat inactivation method, we will subject the spores to PMA 

treatment followed by DNA extraction. This method will help NASA set planetary protection standards and improve 

techniques for reducing spacecraft bioburden for future missions. 


Rafael Send 

Mentors: Paul Backes and Christopher McQuin 










The Effect of Glow Discharge on Thermal Vacuum Vapor Deposition 

Theodore J. Sharp 

Mentor: Andrew Hanna 

Glow discharge is a process used in thermal vacuum vapor deposition to increase adhesion by cleaning the surface 

being coated. The glow discharge is done before the coating while the vacuum chamber is pumping down to high 

vacuum. For the small chamber in building 248, the glow discharge is started as the chamber pressure goes below 

200 milliTorr. While still pumping down, the direct current transformer is slowly increased to a voltage setting of 

about 1.5 kilovolts. This transformer is connected to a high voltage wire that runs through the vacuum chamber. 

As the voltage increases the remaining air in the vacuum chamber is ionized separating the air into ions and 

electrons. With the high energy electrons released, the chamber and the surface to be coated is “scrubbed” as the 

electrons bombard the surface. This bombardment cleans the surface on a molecular level allowing for greater 

adhesion between the surface and the aluminum vapor. This is continued for approximately five (5) minutes at 

which time the transformer is shut down and the glow discharge ends. 

After continuing to pump down and reaching a high vacuum state, the tungsten filaments inside of the chamber are 

heated up melting the pure aluminum wire hung on the filaments. The molten aluminum then wicks up into the 

filaments and is evaporated which begins coating all surfaces inside of the chamber with a layer of aluminum vapor 

between 3-7 microns thick. To determine whether the coat is thick enough to be a good mirror, a simple process is 

used utilizing a transparent light bulb mounted just behind a sample slide of glass. Once coating begins, the 

operator can watch the filament in the light bulb slowly grow dimmer as the glass is covered with the highly 

reflective aluminum vapor. Once the filament can no longer be seen the coating process is stopped and the 

chamber is backfilled and opened. 

Once the samples of glass have been coated in aluminum, a “tape test” will be performed to check the adhesion of 

the aluminum to the glass. The glass plate is pulled from the vacuum chamber and placed on a flat surface to 

prevent it from fracturing. A piece of scotch tape is placed on the surface of the aluminum coat and pressed firmly 

down and is then sharply removed removing any aluminum that was not entirely adhered to the surface of the 

glass. By inspecting the tape and seeing how much aluminum was removed, the operator can determine whether 

the coat is well adhered or whether they must strip the aluminum and start over. 

A major goal for this research is to determine an optimal voltage to set the glow discharge to obtain the best 

adhesion. By finding an optimal voltage, a future operator may use that information to better use the glow 

discharge, to clean the surface being prepared for coating. To determine the ideal voltage a range of voltages will 

be used to prepare sample slides of glass which will then be coated using a standardized procedure with roughly 

the same amount of aluminum. They will then be removed and operators will perform a tape test removing any 

aluminum that had not adhered to the glass. The samples will then be marked with the voltage used for the glow 

discharge and they will be sent for reflectivity testing. By considering a blank slide of glass as a control to compare 

to the reflectivity of the sample should increase as more aluminum adheres correctly to the glass slide. By doing 

this, it will be possible to qualitatively determine what the optimum voltage setting is for the glow discharge. 

Analysis of Data in Accordance With Space Flight Mission Environmental Requirements 

Monica Shei 

Mentor: Kin F. Man 

The Environmental Assurance Program ensures that all flight hardware is compatible with the environments that 

will be encountered during its mission. It outlines the design, test, risk control, and documentation standards for 

the spacecraft and ensures its reliability in external and self-induced environments that it may experience. The 

167

Environment Requirements Document (ERD) provides the environmental design and test requirements for the 

project’s flight system, subsystems, assemblies, and instruments. It defines the system level environments and 

derived environments for the flight system from the ground, through launch and cruise, and science operations. 

It also specifies the method of verification for the environments, whether it is to be verified by analysis or by test. 

This summer’s objective was to complete the Environmental Assurance Program Summary Report for both the Juno 

Project and Mars Science Laboratory (MSL) Project. Environments-related Waivers and Problem Failure Reports 

(PFRs) were compiled along with statistics on the number of Environmental Test Information Summaries (ETIS’s) 

for each assembly/subsystem. These and other environmental assurance information will be presented in the final 

presentation. 

Constraining Fatty Acid Vesicle Formation in Simulated Deep-Ocean Conditions 

Aomawa Shields 

Mentors: Steve Vance, William Abbey, and Richard Kidd 

Modern cells use lipid membranes to selectively determine which molecules enter and exit the cell. These 

membranes are composed primarily of phospholipids, each consisting of two fatty acids and a phosphate group 

attached to a glycerol molecule. One theory of life's emergence on Earth holds that initial metabolic and replication 

reactions took place in protobiotic cell membranes. These may have been composed of n- monocarboxylic acids 

(fatty acids) alone, rather than more complex phospholipid structures. Our experiments in the lab confirm that 

aqueous decanoic acid self-assembles into bilayer vesicles at standard temperature and pressure (25°C, 1 atm) in 

the pH range between 7 and 8.5, as indicated by stepwise increases in fluorescence induced by ultraviolet 

excitation of dye conjugated to decanoic acid solution. Life’s chemical evolution in primordial oceans on Earth or 

elsewhere could have occurred over a range of temperatures, however. Further investigation will constrain the pH 

range for vesicle formation as a function of temperature. We also plan to explore the role of pressure, which is 

similarly important to whether life may have formed in deep-ocean environments on Earth, or on icy worlds in the 

solar system. 

Correlations Between Technical Performance and Cost Growth: A Case Study Using Margin Analysis 

Tyler Shin 



suggest that there are common reasons for cost growth where strong correlations existed between technical 

performance and cost/schedule performance. This research project investigates historical data from the 13 most 

recent JPL flight projects to determine trends and correlations between technical and programmatic parameters to 

predict cost growth. Each of the 13 projects is analyzed using 5 major data sets: technical complexity, 

subcontractor performance, risk trends, margin analysis and schedule slips. Each data set is analyzed to determine 

which correlations exist. My specific assignment for this group project is to use margin analysis to more accurately 

predict cost growth. After gathering the data, such as mass and power growth of each spacecraft, a regression 

analysis will be performed with cost growth data to establish trends. It is expected that certain aspects of technical 

performance parameters will lend themselves to be more accurate cost predictors than others. Conclusions derived 

from this group research can be utilized to form a historical data base for future projects and compared against the 

current in-work flight projects to predict cost growth. 

Development of an Electronics Package for In Situ Thermal Infrared Imaging: A Comparison of MPU 

and FPGA Based Platforms 

Steven Shoen 

Mentor: William Johnson 

Recent advances in infrared (IR) detector array technologies have made possible the development of IR imaging 

systems with unprecedented quality. One challenge to any portable imager is always the instrument’s ability to 

process the high-speed, high-density data generated by the detector accurately and within the instrument’s power 

budget. With fast serial I/O speeds (~250Mhz) and internal logic (600Mhz), Field Programmable Gate Arrays 

(FPGAs) are shown to have strong potential for handling and processing data at bus speeds as much as two orders 

of magnitude greater than Microprocessor Unit (MPU) based platforms. The FPGA architecture also supports 

multiple clocking domains, allowing a single device to perform data acquisition, hardware control and 

synchronization tasks. While Microprocessor Units (MPU) remain the current standard for remote systems, FPGAs 

are showing promising potential in this application. To this end, a thorough investigation was made into the 

performance and limitations of MPU and FPGA based systems. A Virtex-5 FXT FPGA was specified and set-up on an 

evaluation platform in the lab. Code was generated and executed on the device to evaluate the device’s expected 

field performance. Two instruments, the JPL Near Nulling Radiometer (JPLNNR) and the Hyperspectral Infrared 

Imager (HyspIRI) are used as bases for comparison. 

168

Determining Oxygen-18 Isotopic Fractionation in the Ferrous Sulfate, Melanterite 

Danielle Z. Shulaker 

Mentors: Max Coleman and Issaku Kohl 

Studying regions on Earth that are analogous to Mars serve as case studies for studying astrobiology and planetary 

surface rock formation processes. Río Tinto, Spain is very rich in iron sulfates, and has an environment that is 

possibly very similar to the former environment on Mars. Certain bacteria play significant roles in accelerating 

pyrite dissolution rates, which contribute to the formation of ferrous sulfates, such as melanterite. During mineral 

crystallization from aqueous solution, there are systematic isotopic differences between the solution and solid 

phases (fractionation). Quantifying fractionation enables isotopic analysis to be used to trace the original isotopic 

signature. Thus, determining the oxygen-18 isotopic fractionation of melanterite has many potential benefits for 

understanding processes that formed Mars and its past environment. To determine the fractionation of melanterite, 

acidic aqueous solutions saturated with dissolved hydrated ferrous sulfate were evaporated at 25°C and 40°C. 

During evaporation, the aqueous solution and crystallized melanterite that form were sampled. Aqueous and solid 

samples are converted to BaSO4 for oxygen-18 isotopic composition analysis. Further work will include determining 

the isotopic fractionation of other minerals common to Earth and Mars, and testing conditions, such as pressure or 

humidity, and comparing the different isotopic fractionations determined. 

Development of CCSDS DCT to Support Spacecraft Dynamic Events 

Anahita F. Sidhwa 

Mentor: Kar-Ming Cheung 

The Consultative Committee for Space Data Systems (CCSDS) Design Control Table (DCT) is a versatile link 

calculation tool to analyze different kinds of radio frequency links. It started out as an Excel-based program, and is 

now being evolved into a Mathematica-based link analysis tool. The Mathematica platform offers a rich set of 

advanced analysis capabilities, and can be easily extended to a web-based architecture. Last year the CCSDS DCT’s 

for the uplink, downlink, two-way, and ranging models were developed as well as the corresponding input and 

output interfaces. Another significant accomplishment is the integration of the NAIF SPICE library into the 

Mathematica computation platform. 

This summer, the attitude models for the CCSDS DCT are being developed along with a Smart GUI that 

incorporates the different components needed to model the spacecraft attitude. The attitude heuristics model the 

spacecraft dynamic events of earth-pointing, sun-pointing, spinning, conning, and lander orientation. The attitude 

heuristics will also include models for “Safemode” scenarios that depend on the mission phase (cruising, mapping, 

etc.). Based on the knowledge of antenna pointing relative to the spacecraft frame, the cone and clock angles of 

the antenna pattern with respect to the target of interest can be computed, and this provides the antenna pointing 

loss of the link equation. Thus modeling of spacecraft attitude behavior enables one to provide accurate link 

prediction during spacecraft dynamic events. 

Mars Lava Inflation Identification Through Visual Surveys 

Laura Sigelmann 

Mentors: Suzanne Smrekar and Serina Diniega 

Different styles of Martian lava eruption create distinctive surface formations, which provide insight into the past 

and present volcanic behavior of Mars. These formations develop under specific conditions, and their distribution 

can provide valuable information about eruption dynamics on Mars. For example, inflationary features are created 

when a thickened and rigid crust develops on top of a still-fluid lava interior. Pressure exerted by the moving fluid 

core uplifts the crust, and a positive topographic feature with distinct cleft systems develops. It is has been 

proposed that the distribution patterns of inflation features correlate with areas of higher and lower pressure within 

the magma core. Through a survey of images taken by the High Resolution Imaging Science Experiment (HiRISE) 

camera on the Mars Reconnaissance Orbiter (MRO), the location, orientation, and size of various inflation features 

such as tumuli, lava rises, and lava-rise pits were identified. A set of criteria for identifying such features was also 

established to aid in continued research. The analysis of the resulting data set will be used to test and develop 

models for eruption and flow dynamics of inflated flows. 

Applied Field Magnet Development and Magnetic Probe Testing in a Magnetoplasmadynamic Thruster 

Vritika Singh 

Mentors: James Polk and Robert Moeller 

Space propulsion began with powerful chemical rockets propelling the first missions into orbit around the Earth and 

Moon. Since those first missions, views of space and space exploration have evolved to include new possibilities, 

and space propulsion has evolved to include electric propulsion (EP). EP is a type of space propulsion that enables 

more efficient space travel and a greater capacity for scientific instruments on-board spacecraft. One high power, 

high specific impulse EP device is a magnetoplasmadynamic (MPD) thruster, which uses electric and magnetic 

forces to create thrust. MPD thruster performance generally increases with increasing operating current, but at 

higher current levels, performance and lifetime actually decreases due to an instability condition called “onset.” To 

169

mitigate onset, an applied field magnet system was developed and validated through bench testing and computer 

modeling. Early work consisted of mapping the self generated magnetic field inside the MPD thruster at varying 

current levels using a Bdot probe. The applied field magnet system will be implemented with the MPD thruster, and 

the magnetic field structure will be mapped with the influence of the magnets. This magnetic field map will be 

compared to the map produced without the applied magnets to study their influence on MPD thruster discharge 

characteristics. 

Hardware Testing for the Optical Payload for Lasercomm Science 

Amanda Slagle 

Mentors: Anthony Mannucci and Parker Fagrelius 

Optical communication is the use of visible light rather than radio-frequencies to send a data signal. The technology 

is valuable because it allows much higher data rates than radio communications can achieve. The Optical Payload 

for Lasercomm Science (OPALS) project will establish an optical link between the International Space Station (ISS) 

and Earth, which will allow for the testing of vital algorithms for controlling optical communication equipment. 

When OPALS is on the ISS, it will lock onto a laser beacon transmitted by a telescope on the ground. The payload 

will then transmit a video modulated onto a laser beam back to earth. 

The payload consists of an optical head containing the laser transmitter and a camera for detecting the ground 

telescope’s beacon, a gimbal to control the pointing of the optical head, and a sealed electronics box. Equipment 

for the gimbal, laser, and avionics subsystems was tested. For example, an inspection of the flight power board 

was conducted to ensure that all power and ground signals were isolated, that polarized components were correctly 

oriented, and that all components were intact and securely soldered. 

In Situ Analysis of Organics With a Portable Mass Spectrometer 

Santosh Soparawalla 

Mentor: Luther Beegle 

In order to find life on other planets, we start by examining life processes we understand on the earth. Signs of life 

can exist in the form of small organic molecules such as amino acids. Mass spectrometry (MS) is an analytical 

method applied to gain chemical information on a wide variety of chemical compounds. The project demonstrates a 

portable MS system, the Mini 10.5, for astrobiological applications including in situ hydrocarbon analysis and 

sediments analysis using an automated sample processing system (ASPS). The ASPS is a sampling device 

developed at JPL that uses a cation exchange resin to extract amino acids from sediments. Since the amino acid 

solution eluted from the ASPS is at pH 12 positive ion formation required for MS analysis is inhibited. The pH was 

reduced without forming excessive salt by adding a small amount of acetic acid to form an ammonium acetate 

buffer suitable for electrospray ionization. The second part of the study tested a sorbent trap system developed to 

preconcentrate hydrocarbon compounds from the air. Air sampling was done at sites in the Mojave desert and near 

roads to evaluate the role of automobile exhaust as a carbon source for lichens in the arid environment. 

Analysis and Design of OTM-24 for the Mars Reconnaissance Orbiter 

Rebekah Frances Sosland 

Mentors: Martin D Johnston, Reid Thomas, and Jessica Williams 

The Mars Reconnaissance Orbiter (MRO) Project is a major component of NASA’s Mars Exploration Program (MEP). 

Launched in August 2005, the MRO spacecraft is now in the Extend Mission phase and is on-station in its low 

altitude, Primary Science Orbit (PSO) at Mars. The MRO orbit altitude is maintained using Orbit Trim Maneuvers 

(OTMs) which are implemented at orbit apses once every eight weeks and are designed to maintain the MRO 

groundtrack walk. This paper describes the MRO maneuver design and implementation process and details a new 

OTM strategy to perform OTMs every sixteen weeks. Orbital mechanics theory and Navigation software was used to 

analyze maneuvers with various Density Scale Factors (DSF) in order to maintain the orbit altitude. Using delivered 

Orbit Determination (OD) solutions, OTM-24 was modeled and performed on the spacecraft using the OTM 

sequence block. The Maneuver was designed to increase orbit semi-major axis in order to target a GTW error 

staying within the GTW constraints. After analyzing OTM-24, the data showed a very small execution error 

resulting in a very successful OTM and confirming that the OTM-25 may potentially be postponed and performed in 

sixteen weeks. 

Cryogenic Dielectric Measurements of Liquid Hydrocarbons 

Charles H. Starr 

Mentor: Martin B. Barmatz 

Accurate interpretation of Cassini RADAR measurements of Titan’s liquid hydrocarbon lakes requires knowledge of 

the dielectric properties of hydrocarbon mixtures at 13.8 GHz and at cryogenic surface temperatures (~ 90 K). In 

particular, a value for the complex dielectric constants of hydrocarbon mixtures would allow determination of lake 

depths. An improved cavity perturbation technique using exact field solutions to Maxwell’s equations has been 

developed to study liquid samples contained in a quartz holder within a cylindrical cavity resonator. This technique 

170

equires the measurement of the cavity resonant frequency and quality factor with and without a sample inserted 

in the cavity. A cylindrical microwave cavity has been constructed that will initially be used to determine the 

complex dielectric constant of liquid methane, liquid ethane, and their mixtures. I am currently developing a userfriendly 

Mathematica program to obtain the complex dielectric constant from the experimental measurements. In 

the future, we intend to extend this new cavity perturbation technique by developing a similar Mathematica 

program to obtain the permeability values of magnetic samples. 

The Effects of High Vacuum on the Longevity of Tungsten Filaments 

Aaron Steinkraus 

Mentor: Andrew Hanna 

Thin film vapor deposition is a vital process in increasing the reflectivity of a material. During deposition the coating 

material is vaporized and deposited on to the substrate. One way to vaporize the coating metal is to hang loops of 

the coating metal on tungsten filaments and then run current through the filament until the metal wicks into the 

coils and vaporizes. The thermal cycling and oxidation of the tungsten causes the filaments to fail after only three 

to four coatings. In theory, by conducting coating at lower vacuum, the filaments would experience less oxidation 

and thereby extend the life of the filaments. To test this theory, the coating process will be run at two levels of 

vacuum and then cycling the filaments on and off until failure. The filaments that are operated at lower pressures 

are predicted to have a longer life than those at a higher pressure. 

Assessment of Bacterial Spores in Solid Materials 

Kamil B. Stelmach 

Mentor: Wayne Schubert 

The United States is part of the 1967 United Nations treaty called “Treaty on Principles Governing the Activities of 

States in the Exploration and Use of Outer Space, Including the Moon and Other Bodies” and is required to 

participate in planetary exploration. Planetary protection measures are required avoid accidental transfer of Earth 

life to other solar system bodies. To that end, studies are being conducted to enumerate the amount of microbial 

bioburden found in common spacecraft material. A wide variety of spacecraft materials were studied: various 

epoxies, Kapton tape, electrical components, and Lucite. A literature search was conducted into the possible 

dissolution of these materials. Based on the search, the use of the Hansen Solubility Parameters is recommended 

to be used for picking out future solvents. Currently, select polymeric materials are being artificially embedded with 

a known spore count of Bacillus atrophaeus. The materials are then to be cryogenically ground to a fine powder. 

Spore counts will be conducted from the ground material to enumerate the amount of viable spores extracted. 

Furthermore, quantitative PCR (qPCR) will be used to enumerate the amount of total bacterial DNA found within 

spacecraft material. 


David C. Sternberg 

Mentors: Paul Backes and Christopher McQuin 










Improving Data Accessibility 

Ben-han Sung 

Mentors: Paul Wolgast and Bach Bui 

The majority of the data from the Deep Space Network (DSN) is kept in a firewall-protected database and accessed 

by users through a web portal. This project aimed to improve the limited data accessibility by building more tools, 

such as mobile applications. The DSN Dashboard app, created for iOS devices last year, underwent improvement 

as the first part of this project. The app helps visualize tracking information such as antenna azimuth and elevation, 

and bridges the traditional gap between mission management and real-time data. New features include a view that 

shows the DSN’s released tracking schedule. Secondly, overall data accessibility was improved by the development 

of an Openfire module to facilitate communication between NASA employees. JPL and NASA use the Openfire XMPP 

server for collaborative work and sharing information across facilities. Previously, users had to make project or 

location specific accounts in order to use the service. The newly developed module authenticates users against 

171

multiple LDAP servers so they can log in if they have an account on any of those servers. Removing the 

inconvenience of multiple accounts will help open lines of communication between mission management and 

operators at remote DSN complexes. 

Design of a Robotic Ankle Joint for a Microspine-Based Robot 

Nitish Thatte 

Mentor: Aaron Parness 

Successful robotic exploration of near-earth asteroids necessitates a method of securely anchoring to the surface of 

these bodies without gravitational assistance. Microspine gripper arrays that can grasp rock faces are a potential 

solution to this problem. A key component of a future microspine-based rover will be the ankle used to attach each 

microspine gripper to the robot. The ankle’s purpose is twofold: 1) to allow the gripper to conform to the rock so a 

higher percentage of microspines attach to the surface, and 2) to neutralize torques that may dislodge the grippers 

from the wall. So far, two mock-up ankles that demonstrate the key mechanical features have been constructed. 

The manufacture of a final, actuated prototype is underway. Parts were developed using computer aided design 

and manufactured using a variety of methods including selective laser sintering, CNC milling, and traditional 

manual machining techniques. Upon completion of the final prototype, the gripper and ankle system will be tested 

to demonstrate robotic engagement and disengagement of the gripper and to determine the improvement in load 

bearing ability. The immediate application of this project is to outfit the Lemur IIb robot so it can climb and hang 

from rock walls. 

Planning Surface Interaction Operations for a Near-Earth Asteroid Rover 

Eric Timmons 

Mentors: David Mittman and Jay Torres 

With NASA shifting its focus to landing a person on a near-Earth object by 2025, the All-Terrain Hex-Limbed Extra- 

Terrestrial Explorer (ATHLETE) team at JPL is examining the feasibility of using ATHLETE to study an asteroid. Due 

to the low magnitude, non-radial gravity field of an asteroid, innovative techniques must be employed to ensure 

ATHLETE’s success. The anchoring system to ensure the robot stays on the surface of the asteroid is one 

innovation. It is integral to the mission that the anchors can supply the necessary forces to hold ATHLETE to the 

surface while the robot is performing its tasks. The maximum reaction force the anchors can apply is limited by 

local regolith properties such as cohesion. Current estimates of asteroid regolith properties have high uncertainty. 

Additionally, these properties vary with location and time due to anchor interactions. Using current sensors 

available on ATHLETE, this research will develop a system that will provide current state information for the 

anchors. State information includes in situ estimates of local regolith properties. These estimates will be used to 

determine if an action is safe and the uncertainty in that decision. The ultimate goal of this project is to present 

this information to the operator in a clear, useful manner. This software will be validated in simulation and is 

planned to be validated using ATHLETE and regolith simulant in the future. 

Plume Dynamics and Its Implications for Gravity and Topography for Venus 

Henry Tom 


In this study we numerically model Venus with a 3D spherical code, OEDIPUS, to try to both simulate surface 

expressions and track its thermal evolution. From OEDIPUS's output, MATLAB codes are used for analysis and 

various plots. In particular, plots of plumes allow determination of the number, buoyancy, and heat flux of each 

plume. Further, plumes are correlated with topographical features such as hotspots and coronae. However, the 

features of plumes differ from one run to another due to different input parameters into OEDIPUS. A balance 

between the input parameters, Rayleigh number and internal heating, has yet to be determined but recent 

developments suggest constraints on these values. In particular, we look at the variability of mantle plumes, such 

as their buoyancy, temperature difference, and radius, as we change these parameters. Moreover, Rayleigh 

number is dependent on both viscosity and temperature difference across the convecting domain. Decreasing 

bottom viscosity from 10 21 to 10 20 produces more hot plumes; however, internal heating reduces their buoyancies, 

therefore creating insignificant topographical features. Further, comparing topographies from analysis and from 

gravity and topography data will narrow our constraint values even further, allowing us to produce the 

approximately nine hotspots we observe on Venus. 

Development of the NEXUS Hardware Testbed 

Carson A. Umsted 


The goal of NEXUS is to develop a unified, scalable, and highly capable avionics interconnect for future spacecraft. 

To meet this goal, Serial RapidIO has been chosen as the baseline protocol to be integrated with the use of ISAAC 

technology. My task is to develop a NEXUS hardware testbed composed of a switch, and a number of FPGA-based 

NEXUS nodes using ISAAC iBoards. Each node includes a Bus Interface Unit (BIU) IP core based on the Serial 

RapidIO protocol and various types of backend such as ADCs and high-speed camera. The Xilinx Serial RapidIO IP 

172

core is used as the basic implementation for different configurations including, loop-back, point-to-point, and 

switch-based. The complete hardware testbed will be integrated and verified with real-time high-speed data 

sources. 

Fade and Revival of the South Equatorial Belt of Jupiter in the Mid-Infrared Wavelength Range 

Steffi B. Valkov 

Mentor: Glenn S. Orton 

Sporadic fluctuations in the color and reflectivity of Jupiter's South Equatorial Belt (SEB) have been observed over 

the last century. In the latest of these, from 2009 to 2010, the SEB experienced an episode of turning from a 

typical brown color to white (known as a “fade”); the previous episode of this type was in the early 1990's. Toward 

the end of 2010, the SEB began a return to a brown color (known as a “revival'). We used data collected from 

ground-based telescopes in the mid-infrared wavelength range in order to find pressure-temperature profiles, 

fractions of para hydrogen, and aerosols in the atmosphere of Jupiter. Our ultimate goal is to find what causes a 

fade and revival of the SEB and what changes in the atmosphere may occur before and after a fade. The 

instruments that were used in this research were MIRSI and VISIR, both sensitive to wavelengths between 

4.90 and 24.80 μm. After initial reduction of these mid-infrared thermal images, they were calibrated in absolute 

radiance using Cassini infrared data as a standard. New, 2011 Jupiter data were acquired and calibrated in late July. 

We used a FORTRAN-based program developed at the University of Oxford to determine properties of the 

atmosphere from the images of Jupiter's radiance collected at these telescopes. Initial results show that prior to 

the fade, temperatures in the troposphere began to drop, as opacity of aerosols increased. After the full fade took 

place, the para-ortho hydrogen fraction in the SEB increased by December 2010 and aerosols decreased. 

MSL OPGS Product Generation Pipeline’s Performance Assessment 

Sergeh Vartanian 

Mentors: Maher Hanna, Adrian Tinio, and Costin Radulescu 

The MSL rover "Curiosity" will be launching in 2011 and landing on Mars in August 2012. The Operation Product 

Generation Subsystem, OPGS, is responsible for generating critical products such as images captured through the 

rover’s cameras which are used by the rover planners in determining tactical and strategic rover activities. For my 

research project, I was responsible for measuring, assessing, and documenting the performance of the MSL OPGS 

product generation pipeline with respect to its operational timeline requirement of thirty minutes. My tasks included 

the measurement of the overall performance throughout the pipeline from raw data product to EDR/RDR generated 

files delivered to the operational team, as well as of the individual pipelines within the workflow. In addition, I have 

conducted a profound study and evaluation of the OPGS pipeline as part of the performance assessment. In order 

to meet the thirty minute requirement, I spent many hours running the pipeline in three different environments 

that involved virtual machines, physical standalone machines, and the Amazon Cloud with varies software and 

hardware configurations that I contemplated while analyzing and studying the behavior of the system. 

Analysis of Ultra High Resolution Sea Surface Temperature Data 

Grant T. Wagner 

Mentors: Edward Armstrong and Jorge Vazquez 

The Group for High Resolution Sea Surface Temperature (GHRSST) is an international project that distributes 

satellite derived sea surface temperatures (SST) data from multiple platforms and sensors. The goal of the project 

is to distribute these SSTs for operational uses such as ocean model assimilation and decision support applications, 

as well as support fundamental SST research and climate studies. Examples of near real time applications include 

hurricane and fisheries studies and numerical weather forecasting. The JPL group has produced a new 1 km daily 

global SST product, the Multiscale Ultrahigh Resolution (MUR), that blends SST data from 3 distinct NASA 

radiometers in both the microwave and infrared spectral bands, but requires further validation and accuracy 

assessment, especially in coastal regions. 

The goal of this project is to analyze, validate and characterize the newly developed GHRSST 1km (high-resolution) 

SST dataset for use in coastal studies with the focus on major upwelling regions of the world’s oceans, including 

the California Coast. This work will include comparisons of the new dataset with in situ fixed and drifting buoy SST 

measurements and as well as comparisons to a Low Resolution version of the product (25 km resolution). It also 

includes exploring applications of the new dataset in fishery and marine biology through evaluation of submesoscale 

oceanic features such as temperature fronts, gradients, and mesoscale eddies. 

173

Taxonomic Classification of NEOWISE Objects: Correlating Spectroscopy, Albedo, and Physical 

Properties 

Jessica A. Watkins 

Mentor: Amy Mainzer 

The influx of new data to the NEOWISE system offers the opportunity to physically characterize 157,000 newly 

discovered small planetary bodies (specifically Near-Earth Objects (NEOs) and Main Belt Asteroids (MBAs)). The 

goal is to understand the relationship between visible/near-IR (VNIR) spectral features and physical properties such 

as mineral composition, size, density, regolith, and measured visible and near-IR albedos. NEOWISE utilizes multiwavelength 

infrared spectra to more accurately determine diameters and compute albedos, allowing us to compare 

various taxonomic classes more accurately than previous taxonomic schemes derived from VNIR spectroscopy and 

photometry. We classified the thermally fitted spectral data into known, corresponding classification schemes 

(Tholen, Bus-Binzel, Bus-DeMeo, and Carvano) based on color to test their accuracy with the larger, less biased, 

and more accurate NEOWISE data set. There is a two-fold bias in these classification schemes, involving visible 

light surveys that favor high albedo objects, and taxonomic classifications with the NEOWISE survey will mitigate 

this bias. The processes that may affect the correlation between spectroscopy and albedo and cause peculiar 

absorption features in spectral signatures are examined (space weathering, grain size, hydration features, etc.) to 

explore precisely what information a given spectral signature or albedo can actually provide about an asteroid’s 

surface. 

Electrochemical Analysis of Fuel Cell Catalysts 

Ian R. Wessen 

Mentor: Thomas I. Valdez 

Polymer electrolyte membrane (PEM) fuel cells exhibit many exciting properties, not the least of which is their 

startling efficiency. For instance, internal combustion engines can theoretically convert energy with 39% efficiency 

(often averaging around 20%). Fuel cells do not burn their fuels, instead turning chemical potential directly into 

work. In theory, a fuel cell may do so at 100% efficiency. It is not uncommon for some of our PEM hydrogenoxygen 

fuel cells to perform upwards of 80% efficiency at low current densities. If we could find a fuel cell catalyst 

that can maintain 80% efficiency at higher current densities, we would have an extremely attractive and viable 

option for power generation, particularly with regards to future space application. We have analyzed various 

electrochemical properties of Pt-Ni electroless plated catalysts using cyclic voltammetry, potentiodynamic scans, 

and electrochemical impedance spectroscopy to deduce their respective efficiencies and effectiveness as oxygen 

reduction catalysts. The discovery of substantial improvements in fuel cell efficiency will have direct impact in 

tomorrow's generation of fuel cells. 

Ontological Modeling for Integrated Spacecraft Analysis 

Erica L. Wicks 


Current spacecraft work as a cooperative group of a number of subsystems. Each of these requires modeling 

software for development, testing, and prediction. It is the goal of me team to create an overarching software 

architecture called the Integrated Spacecraft Analysis (ISCA) to aid in deploying the discrete subsystems’ models. 

Such a plan has been attempted in the past, and has failed due to the excessive scope of the project. Our goal in 

this version of ISCA is to use new resources to reduce the scope of the project, including using ontological models 

to help link the internal interfaces of subsystems’ models with the ISCA architecture. 

I have created an ontology of functions specific to the modeling system of the navigation system of a spacecraft 

and linked this with an ontology modeling the attitude control subsystem. The resulting ontologies not only link, at 

an architectural level, two spacecraft subsystems, but also are also web-viewable and can act as documentation 

standards. 

These ontologies are proof of the concept that ontological modeling can aid in the integration necessary for ISCA to 

work, and can act as the prototype for future ISCA ontologies. 

A Prototyping Effort for the Integrated Spacecraft Analysis System 

Raymond Wong 


Computer modeling and simulation has recently become an essential technique for predicting and validating 

spacecraft performance. However, most computer models only examine spacecraft subsystems, and the 

independent nature of the models creates integration problems, which lowers the possibilities of simulating a 

spacecraft as an integrated unit despite a desire for this type of analysis. A new project called Integrated 

Spacecraft Analysis was proposed to serve as a framework for an integrated simulation environment. The project is 

still in its infancy, but a software prototype would help future developers assess design issues. The prototype 

explores a service oriented design paradigm that theoretically allows programs written in different languages to 

174

communicate with one another. It includes creating a uniform interface to the SPICE libraries such that different inhouse 

tools like APGEN or SEQGEN can exchange information with it without much change. Service orientation may 

result in a slower system as compared to a single application, and more research needs to be done on the different 

available technologies, but a service oriented approach could increase long term maintainability and extensibility. 

Automated Purgatoid Identification 

Steven B. Wood 

Mentor: Paolo Bellutta 

Purgatoids are small (< 1 m high, < 10 wide) eolian bedforms commonly found scattered across the Meridiani 

Planum region of Mars. First encountered by the Mars Exploration Rover Opportunity during sol 446 of that mission, 

the rover became stuck in the poorly consolidated soil of a purgatoid while attempting to traverse the feature. 

Critically, further driving was nearly impossible as most forward or backward commanded movement was 

translated into digging the wheels deeper into the loose, unstable bedform surface. Although Opportunity was 

ultimately freed on sol 484, from that day forward identification and avoidance of purgatoids became an important 

feature of rover path planning. Purgatoid identification is currently done manually using high-resolution HiRISE 

images from the Mars Reconnaissance Orbiter. This project, in support of the Mars Terrain Classifier, created and 

implemented an algorithm that uses the same HiRISE images as data and automatically identifies the location of 

potential purgatoids in any input image. The algorithm relies on regional shape descriptors and moment invariants, 

along with normalized cross-correlation with a selected group of kernels representative of purgatoids, for accurate 

identification. 

Visual Tracking With a Small Quadrotor UAV 

Jonathan Wu 

Mentor: Roland Brockers 

Unmanned micro air vehicles (UMAVs) are capable of low cost reconnaissance, search and rescue. To facilitate 

human operator control, these systems need autonomous capabilities to execute basic navigation tasks. For tasks 

like flying to a goal or tracking a moving target from a vantage point, robust image based tracking methods are 

needed that can be executed on the vehicle in real time. 

This project explores various methods and approaches in template based target tracking using imagery from an on 

board camera with extensions applicable to UMAVs. The methods explored survey the real-time stable performance 

of different multi-parameter warping models (affine, perspective, and quadratic) and their corresponding 

parameter estimators to track targets in image sequences over time. 

Mapping Hazards for Future Mars Rover Landing Sites 

Kimberly E. Yauk 

Mentors: Matthew Golombek and Fred Calef 

To help decide where to land future rover missions on Mars, hazards on the Martian surface were mapped at 

prospective landing sites. Hazards include rock fields, steep scarps, and inescapable landscapes such as steep 

craters and fields of intersecting aeolian bedforms. We investigated seven prospective landing sites identified for a 

possible rover mission in 2018. We downloaded HiRISE images and stereo anaglyphs into ArcMap and geospatially 

referenced the images onto lower resolution CTX images for each landing site. At resolutions on the order of 1:600, 

the terrain was assigned categories ranging from flat and safe (one) to non-traversable and dangerous (four) 

based on the apparent slope of the 3D anaglyph. Slopes greater than thirty degrees are considered dangerous to 

the rover and are evidenced by shadows cast from the rocks. The hazard maps, once completed in detail, will help 

determine the most appropriate landing site, could help fine tune the landing ellipse and give engineers a map on 

which to estimate landing site safety and traversability for the rover. 

Deep Space Network Portal for Android Tablet 

Jonathan Yee 

Mentor: Paul Wolgast 

The Deep Space Network (DSN) consists of three deep space communications facilities around the world that are 

constantly tracking information throughout the day. The DSN Portal was created as a website to make the data 

from the DSN accessible on any computer. Computers and laptops are not the most mobile devices to use to view 

data on the DSN Portal. Last year, a version of the DSN Portal was created for the iPad and has been successful. 

My work this summer involved developing an optimized version of the DSN Portal to run on an Android tablet. An 

Android DSN Portal app allows DSN operators to use a mobile interface to the DSN portal from an Android tablet. 

This gives the user live tracking information from each of the three facilities using an intuitive interface and touch 

screen gestures. As mobile devices such as tablets become more ubiquitous, they will allow DSN operators the 

convenience to view live data from the DSN Portal remotely and easily. 

175

SIAD Supersonic Flight Dynamics Test Engineering Support 

Regina Zmuidzinas 

Mentor: Brant Cook 

The Low Density Aerodynamic Decelerator (LDSD) research and technology development is currently working to 

develop two Mars atmospheric entry technologies: a 6-meter Supersonic Inflatable Aerodynamic Decelerator 

(SIAD) and a 30-meter Supersonic Ringsail Parachute (SSRS). However, before being implemented in a Mars entry 

environment, these technologies must be qualified in a series of tests on earth to verify that the SIAD and SSRS 

will function as expected. The Supersonic Flight Dynamics Test (SFDT) is one of three LDSD tests to confirm these 

results, using a vehicle that can obtain Mach 4 at an earth altitude of 180,000 feet to simulate a Mars entry 

environment. The SFDT vehicle has many design constraints to perform at these conditions, including mass and 

power consumption. In order to assist the SFDT vehicle design team, the harnessing for the instruments on the 

vehicle will be determined and routed within the vehicle in order to better assess harness mass. The vehicle 

electronic box will be preliminarily designed. The entire mass and power of the vehicle will also be closely tracked 

during the design phase so that the team can closely track how design changes affect system mass margin. 

176

Abstracts 

Student–Faculty Programs 

2 

0 

1 

1

Student-Faculty Programs 2011 Abstract Book - Summer ...

Create successful ePaper yourself

Delete template?

Save as template?