Asteroid Spectroscopy

Lecture #2
Asteroid Spectroscopy
Principles and Practice

Characterizations of Asteroid Surface
Materials
Only one in situ characterization has been
carried out for an asteroid surface
– Gamma Ray Spectrometer (XGRS) on the NEAR
Mission after the spacecraft landed
– Successful “target of opportunity”
All other characterizations of asteroid surface
materials have been by remote sensing

HAYABUSA Sample Return Mission
A large number of microscopic particles
have been recovered from the sample return
capsule.
At this time, it was not known whether these
were collected from asteroid Itokawa.

Most “Compositional” Determinations
rely on the Interaction of Electromagnetic
Radiation with a Target
Nuclear
Interactions
Inner Shell
Electron
Interactions
Outer Shell
Electron
Interactions
Vibrational
Interactions
Electromagnetic Spectrum

Remote Sensing Techniques for
Characterizing the “Composition” of
Asteroid Surfaces
Visible-Near Infrared (VNIR) Spectroscopy
– Sunlight reflected from surface
Thermal Infrared (TIR) Spectroscopy
– Infrared radiation emitted by surface
Radar
– Reflected “radio wavelength” radiation
X-ray and Gamma-ray spectroscopy
– High energy photons emitted by excited atoms
and excited / unstable atomic nuclei.

Why are surface characterizations
important
The surface material of a body is generally a
mixture of all near-surface geologic units
– In “disrupted-and-reaccreted” bodies, it samples
all internal geologic units.
Surface composition (Chemistry)
– Element abundances
– Element ratios
– Isotopes
Surface Mineralogy
– Mineral types
– Mineral Abundances and Compositions

Why Mineralogy rather than
Composition
Compositional determinations such as
γ-ray spectroscopy can only be made
from a spacecraft near an asteroid.
Facies Concept
– Different conditions (e.g., pressure,
temperature, etc.) will produce different
minerals from the same bulk composition.
– Identification of the minerals constrains:
Formation conditions
Chemical composition

Al 2 SiO 5 Minerals

Mineralogical Characterizations
Mineralogy can:
– Generally distinguish between different
meteorite types and can often identify
specific meteorite types
Chondritic subtypes
Achondrite types
“Metal-rich” assemblages
– Provide insight into the chemical and thermal
history of the parent body (e.g., oxidation,
hydration, melting, differentiation, etc.)

Spectroscopy
When many materials
interact with light
(electromagnetic radiation)
they impose a characteristic
signature on that light.
The most obvious
manifestation is color.
Spectroscopy provides a
means of quantitatively
assessing the nature of the
material which interacted
with the light.

VNIR Spectra
Detection and compositions of transition
metal-bearing minerals
Detection and identification of minerals
which incorporate molecular species such as
H 2 O, OH, CO 2 , C-H, etc.
Relative abundance of mineral species
Some petrographic information (particle size,
mixture state, etc.)

What are Transition Metals
Periodic Table of
the Elements

Transition metal elements commonly have
multiple valence states:
Example – Iron ⇒ Fe 2+ or Fe 3+

Electronic Structure of Atoms
Electron Orbitals
– Electrons in an atom are distributed into
electronic shells (or levels)
– Each shell contains one or more sets of orbitals

Shell #3
Shell #2
Shell #1
Nucleus
Electronic Structure of Atoms
1
s
3
s 3 p 3 d
2
s 2 p
Electronic Shells
– Shell #1
s-orbital: 1 s
– Shell #2
s-orbital: 2 s
p-orbital: 2 p
– Shell #3
s-orbital: 3 s
p-orbital: 3 p
d-orbital: 3 d
– Shell #4
s-orbital: 4 s
p-orbital: 4 p
d-orbital: 4 d
f-orbital: 4 f

Spectra of Transition Metal-
Bearing Minerals
“Crystal Field Theory”
Transition metals (Fe, Ni, Co, etc.) are
characterized by an unfilled outer electron
shell.
– 3 d-shell for 1 st series transition metals
A d-shell contains 5 electron orbitals

First Series Transition metals
Periodic Table of
the Elements

d-shell Electron Orbitals
Each orbital can accept two
(2) electrons
Two of the orbitals lie along
the Cartesian axes
Three of the orbitals lie
between the axes:
In a uniform electrical field,
all of these orbitals have
equal energy, and cannot be
distinguished (degenerate).

In minerals, the metal cations (e.g., Fe 2+ ) are
often surrounded by oxygen anions (O 2- )
A common structure
is 6 oxygen ions
surrounding 1 metal
ion (octahedral
coordination)
⇒ olivine, pyroxene
In such a site, the “on-axis” d X
2 -Y
2 and d Z
2 are closer to
the oxygen anions than the three “between the axes”
orbitals (d xy , d xz , d yz ).

Crystal Field Splitting
This configuration produces greater repulsion
for the electrons in the two “on axis” orbitals
than the electrons in the three “off axis”
orbitals.
The result is that the
“on axis” orbitals are
“split” to higher
energy
and the
“off axis” orbitals are
“split” to lower
energy.
dx 2 -y 2
dz 2
d xy d yz d xz

Consider Fe 2+ in an octahedral site
⇒Three lower energy orbitals (groundstate)
⇒Two higher energy orbitals
Fe 2+ ⇒ six d-electrons
Electrons have a spin orientation (spin up, spin
down)
Pauli Exclusion Principle
Only two electrons (with opposite spins) can
occupy any individual orbital
First five electrons each occupy one of the five
d-orbitals

But Fe 2+ has six d-electrons!
The 6 th electron can occupy any of
the orbitals
Electrons normally reside in the
lowest possible energy level
Because of their lower energy, the 6 th
electron normally occupies one of
the ground-state orbitals.

∆ O is the energy difference between the
upper (excited) state and the groundstate
A photon of energy (hν) equal to ∆ O can
be absorbed and the 6 th (spin down)
electron kicked up to the excited state.
∆ O
hν
⇒

Absorption of these photons produces a
feature in the spectrum of the interacting
light at the wavelength corresponding
to the energy ∆ O
0.3
1 micron region of pyroxene spectrum
Reflectance
0.2
0.1
∆ O
0
0.3 0.5 0.7 0.9 1.1 1.3
Wavelength (um)
Absorption feature is characterized by:
central wavelength, width and depth

Most cation sites in transition metalbearing
minerals are not perfectly
symmetric
Distortions of an Octahedral Site

Such Distortions Remove the
Degeneracy from the Energy Levels
Each electron transition
from the groundstate to an
excited state produces a
separate feature in the
spectrum.
The wavelength of the
resulting absorption band is
at the photon energy
corresponding to the
difference between the
energy levels.

The result is multiple absorption features
in many mineral spectra
0.3
Reflectance
0.2
0.1
0
Pyroxene
0.3 0.8 1.3 1.8 2.3
Wavelength (um)

Band positions in pyroxene spectra are
correlated and restricted
Pyroxene Band-Band Plot
(after Adams 1974)
Band I Center Wavelength
1.08
1.06
1.04
1.02
1.00
0.98
0.96
0.94
0.92
0.90
0.88
Opx
Cpx
1.7 1.8 1.9 2.0 2.1 2.2 2.3 2.4
Band II Center Wavelength

Spectra of different HED meteorite
types
0.4
Eucrite
High Fe & Ca Pyroxene
0.3
Reflectance
0.2
0.1
0
Diogenite
Low Fe & Ca Pyroxene
1.885 µm
0.3 0.8 1.3 1.8 2.3
Wavelength (um)
Note the shift in the pyroxene band centers!

Pyroxenes:
Band Position vs. Composition
2.2
2.1
Wo < 11
Wavelength (Microns)
2.0
1.9
1.8
0 20 40 60 80 100
Molar Iron Content (Fs)

The broad 1 µm olivine feature is composed of
three overlapping absorptions at ~0.75, 1.04 &
1.35 µm
1
0.8
Reflectance
0.6
0.4
0.2
0
0.3 0.8 1.3 1.8 2.3
Wavelength (um)

Olivine vs Pyroxene
0.6
0.5
Reflectance
0.4
0.3
0.2
0.1
0
0.3 0.8 1.3 1.8 2.3
Wavelength (um)

Olivine-Pyroxene Mixture Spectra
With increasing olivine content, the area of Band I
increases while the area of Band II decreases.

Band Area Ratio (BAR)
Spectral Reflectance
0,7
0,6
0,5
0,4
BI
BII
0,3
0,2
0,5 1 1,5 2 2,5
Wavelength (um)
We use linear continua to calculate the
areas of the bands.

In Olivine-Pyroxene mixtures, the ratio of the
areas of the 1 and 2 µm bands is proportional to
the relative abundance of the mineral species.
2.50
2.00
BAR @ MSI(0.76) & NIS Wavelengths
BAR(Nom) 63-90
BAR(Act) 63-90
Band Area Ratio
1.50
1.00
0.50
0.00
0 25 50 75 100
% Pyroxene
Cloutis et al. (1986)

Diagnostic Parameters
The position of the absorption features is a
function of crystal structure.
The intensity of the features is related to the
abundance of the absorbing species (e.g., Fe 2+ )
in the mineral
and in mixtures to the relative abundance of the
different minerals.
A mineral is defined by its structure and its
composition.
The positions and intensities of the absorption
features in a spectrum are uniquely related to
specific minerals.

Molecular Vibrations Also Produce
Spectral Features
S.J. Gaffey (1988)
10.5 × 10 13 Hz =
= 9.8 × 10 13 Hz
Overtones occur at sums or multiples of
the fundamental vibration frequencies.

Clay Mineral Spectra
S. J. Gaffey et al. (1993)
The narrow absorption features
near 1.4, 1.9, 2.2-2.3, 2.5 µm (etc.)
are produced by vibrational
overtones of the H 2 O and OH
molecules.
Their shape and exact position
depends on the crystal structure
and the attached cations.
These spectra parameters are
diagnostic of specific minerals.

Some potentially important asteroid
minerals do not exhibit characteristic
absorption features.
0.8
0.6
Spectra of Iron metal and NiFe
Meteorites
Reflectance
0.4
0.2
Iron Metal
Butler
Casey County
0.0
0.5 1.0 1.5 2.0 2.5
Wavelength (um)
General “reddish” slope but no absorption features

Analysis of Asteroid Spectra
Absorption features identify the
presence of specific transition metal or
H 2 O (etc.) bearing minerals.
The compositions of the mineral species
are derived from the wavelengths of the
band centers.
The relative abundance of the mineral
species are derived from the relative
intensities of the absorption features.

Mineralogical Analysis of Vesta
Gaffey (1997)
Spectral parameters indicate abundant pyroxene. The
mineral chemistry indicates an HED surface assemblage

Limitations of Reflectance Spectroscopy
Spectroscopy is unable to detect minerals
which lack diagnostic VNIR features (e.g.
quartz).
Spectroscopy has limited capabilities for
weakly featured minerals or for minerals
whose features are obscured by interfering
species.
Spectroscopy only samples a thin surface
layer (optical surface).
– Surface alteration processes (space weathering –
lecture 13) can modify the spectrum, potentially
leading to inaccurate interpretations
– Is a thin (< few hundred microns) surface layer
representative of an asteroidal body

Is a Thin Surface Layer Representative
Concept of a “split”
– Grind up a rock, homogenize the bulk sample, and
take a small sample for analysis.
Asteroid surfaces layers as “splits”
– Surface area of 100 km diameter spherical
asteroid = 31,415 km 2 = 3.14 x 10 10 m 2
– Volume of 100 µm (10 -4 m) surface layer
= area x thickness = 3.14 x 10 10 m 2 x 10 -4 m
= 3.14 x 10 6 m 3 ~ 3 million metric tons
Thin asteroid surface layers sample very
large masses of material!

Current State-of-the-Art
1459 Magnya H 2 O
V mag = 15.7
H 2 O
Hardersen et al. (2004)
Unsmoothed, unedited average spectrum.
This should be the current expectation for NIR
asteroid spectra.

Observational Methodologies
Standard star (S) observations
interspersed with asteroid (A)
observations.
5S 10A 5S 10A 5S 10A 5S
Allows extinction coefficients
to be calculated (slopes).
Slope & intercept allows flux of
star to be calculated at each
channel at the same airmass
as each asteroid observation.
Dividing the raw asteroid flux
curve by the calculated star
flux curve removes the
atmospheric absorptions.
A slope-intercept pair
for each channel

Effects of orographic winds
Upwind / downwind asymmetry in
atmospheric absorption

Typical instrumental flux measurement
SpeX Instrument @ IRTF in “Asteroid” Mode

Ratio of two raw flux curves

Correction for extinction using a
SPECPR Starpack
Extinction coefficients were calculated from standard
star observations without correction for channel
offsets

Pattern for 1 channel Offset on SpeX
Smoothing does not remove this pattern
13 Pt Smooth
~0.05 µm Interval
25 Pt Smooth
~0.10 µm Interval
37 Pt Smooth
~0.15 µm Interval
51 Pt Smooth - Edited
~0.15 µm Interval

Extinction correction with channel
offsets
Most of the “noise” in the 1.4 & 1.9 µm regions was
not random, but due to uncorrected channel offsets

Effects of Uncorrected 0.5 Channel Offset
+0.5 Chan.
No Shift
-0.5 Chan.
A small offset produces major deviations in Band II

Implications for Analysis
Irrespective of the analysis technique
Curve matching
Gaussian fitting
Parameter extraction
Interpretations would differ for these spectra.

Correction Process - I
The pattern identifies presence of an offset.
Channel offset determined for each set of
observations relative to some reference set.
Offsets are derived by using the steep edge
of the 1.4 µm atmospheric water vapor
feature.

Correction Process - II
Offsets should be established to ~0.1-0.2
pixels.
Pixel offset corrections are applied to the raw
standard star spectra prior to calculation of
the extinction coefficients (or their use in
ratios).
Pixel offset corrections are applied to the raw
object spectra prior to extinction corrections.

Conclusions
All medium resolution NIR asteroid spectral
data should be corrected for channel offsets
as the initial reduction step.
– Preferably the offsets should be determined from
the data itself.
Extinction should use “objective” criteria.
– Standard extinction coefficients adjusted until the
spectrum “looks right” should be avoided.
Smoothing of spectra should be avoided
unless the previous steps have been
accomplished.

The Upside --
Asteroid spectra obtained with medium
resolution NIR spectrographs must be
routinely checked and corrected for channel
offsets.
The parameters needed to make the offset
correction are derivable from the raw data
itself.
Routines exist to make the corrections as a
regular step in data reduction.