40th Lunar and Planetary Science Conference (2009)
1419.pdf
THERMAL INFRARED AND VISIBLE TO NEAR-INFRARED SPECTRAL ANALYSIS OF CHERT AND
AMORPHOUS SILICA. M. L. McDowell1, V. E. Hamilton2, S. L. Cady3, and P. Knauth4. 1Hawai’i Institute of
Geophysics and Planetology, University of Hawai’i, 1680 East-West Rd, Honolulu, HI 96822 (mcdowell@hawaii.edu), 2Southwest Research Institute, Boulder, CO, 3Portland State University, Portland, OR, 4Arizona
State University, Tempe, AZ.
Introduction: Recent spectral observations of the
Martian surface have discovered materials more rich in
crystalline and amorphous silica (SiO2) than previously
known to be present [e.g., 1-6]. In the past, limited
effort has been made to study and document the characteristics and variability of these phases in the thermal
infrared (TIR) and visible to near-infrared (VNIR)
wavelength ranges. This information is necessary to
accurately identify silica phases in our data sets from
Mars and to subsequently draw inferences about the
geologic conditions or environments they represent.
In our study we consider these silica phases, focusing primarily on the spectral characteristics of various
forms of chert and amorphous silica. Chert and
opalline silica are targets of astrobiological interest
because of their exceptional potential for microfossil
preservation. They signify past activity of water and
could represent deposits from surficial, diagenetic, or
hydrothermal fluids.
Previous spectral studies of chert and amorphous
silica include analysis of TIR emission spectra of four
cherts by Michalski [7], TIR transmission spectra by
Long et al. [8], and VNIR reflectance of lithic artifacts
by Hubbard et al. [9].
Because the structures of chert and amorphous silica are different from that of plutonic quartz, their TIR
spectra differ in shape and contrast, as described in
Michalski [7, 10]. Figure 1 shows examples of the
spectra of the samples analyzed. Plutonic quartz and
chert exhibit a double emissivity minimum at ~10001300 cm-1 (region in grey) due to Si-O stretching vibrations. However, the shape of these two minima are
round or angle inward toward each other for plutonic
quartz, while the two minima become more pointed
and angle outward away from each other in chert.
Other Si-O features at ~800 cm-1 and ~500 cm-1 are
similar with the exception of the emissivity minimum
at ~550 cm-1, which may not be present in chert.
The more disordered structure of amorphous silica
leads to a broader emissivity minimum in the ~10001300 cm-1 region and a lack of the minima at ~550 and
400 cm-1. The feature at ~800 cm-1 may be absent.
Though the Si-O bond has no features in the VNIR
region, it is likely that features attributable to hydration
and other minerals in chert may be observable.
Approach: We improve upon the previous studies
by including a more extensive sampling of geologic
chert in hand sample representing various sources,
Figure 1: A comparison of quartz, chert, and opal
spectra.
methods of formation, surface textures, and crystallinities. We analyzed the samples using TIR emission
spectroscopy and VNIR reflectance spectroscopy. TIR
emission spectra were collected over the range of 2000
- 200 cm-1 with a ThermoElectron Nexus 470 FTIR
spectrometer [11]. VNIR reflectance spectra were
measured over the range of 350 - 2500 nm using an
ASD field spectrometer. The following sections show
results from our spectral studies. Further characterization (e.g. XRD, SEM) of these samples is currently
underway.
TIR Spectra:
Spectral Variation in Chert and Amorphous Silica.
Variation in spectral character between different cherts
also exists, as shown by the samples in our study. Features that may vary include the following: 1) relative
depth of the two minima in the ~1000-1300 cm-1 region, 2) angle of slope of the two minima in the
~1000-1300 cm-1 region, 3) addition of local maxima/
minima superposed on the larger emissivity minima in
the ~1000-1300 cm-1 region, 4) presence and depth of
the feature at ~550 cm-1, 5) depth and shape of the feature at ~350 cm-1, 6) inclusion of a minimum at ~900
cm-1 due to carbonate minerals within the sample.
Orientation Effects. The spectra of minerals may
differ depending on the orientation of crystal axes relative to the viewing angle (e.g. [12,13]). In rocks where
the crystals have a preferred orientation, spectral features vary when measured in differing orientations.
For the samples in our study the spectral features
from different orientations are generally similar, but
with slight differences in the depth and/or shape of the
40th Lunar and Planetary Science Conference (2009)
~450 cm-1 band or the ~1000-1300 cm-1 bands. The
overall similarity of the spectra suggest that the grains
in these samples have little preferred orientation, or
alternatively, any crystal orientation present produces
only slight differences in spectral shape for chert.
Effect of Viewing Angle. TIR spectra of the same
surface may vary when measured at different angles, as
illustrated in Figure 2. The same features are present
in the spectra from all angles, but the overall and relative depth of the minima at ~1000-1300 cm-1 change,
as does the shape of these minima. A systematic
change based on the degree of slope is not apparent,
however. A similar characteristic has been observed
previously in the spectra of amorphous silica [14].
Ruff et al. [3] attribute the spectral shape of the high
silica material in Gusev Crater to a variation of an
amorphous silica spectrum caused by differences in
emission angle and sample porosity.
Figure 2: Spectra from the same surface of sample
MC, each measured with the surface at different
angles.
Surface Roughness Effects. Differences in surface
roughness cause differences in spectral features for the
same material, as discussed by Michalski [7]. Contamination of rough or natural surfaces with weathering products is likely, but not significant enough to be
apparent in these spectra. We observe: 1) shallowing
of features with increased surface roughness (or
deepening of features in one case, sample 613), 2)
change in relative depth of the minima in the ~10001300 cm-1 region, 3) absence of the ~550 and ~350
cm-1 feature in surfaces of rougher texture.
VNIR Spectra: VNIR spectra of our chert and
amorphous silica samples show that the material is not
featureless in this wavelength range (e.g., Fig 3). The
number and position of observed features varies between samples. Although their strengths vary considerably, almost all spectra appear to have features at
~1410 and ~1910 nm, which are attributable to H-O-H
and O-H bonds in the samples. The additional fea-
1419.pdf
Figure 3: Examples of chert and amorphous silica
VNIR spectra from our study.
ture(s) at ~2200-2300 nm may be a result of M-OH
bonds. Though a few samples have obvious narrow
absorptions in this area indicative of clay minerals,
most samples have a shallow, broad convex shape.
Some samples also have features at shorter wavelengths (<1000 nm) that may be due to ferric oxides
such as hematite.
These features suggest that hydration in chert and
amorphous silica can be observed in VNIR data, along
with other phases which my potentially be included in
the material (e.g., clay minerals and ferric oxides),
even though features from the Si-O bonds of the chert
cannot be seen.
Summary & Ongoing Work: The TIR spectra of
chert samples from a variety of geologic settings show
that they have a range in spectral characteristics indicating differences in structure, physical character,
and/or mineral inclusions. VNIR spectra show that
most chert and amorphous silica samples exhibit hydration features. The spectra also indicate that some
samples contain clay minerals in addition to silica.
Future work will focus on attributing spectral variation observed in chert to specific physical (e.g. surface
roughness, grain size) or chemical characteristics (e.g.
silica phase, mineral inclusions).
References: [1]Bandfield et al. (2004) JGR, 109,
doi:10.1029/2004JE002290. [2]Hamilton, V.E. (2005) Eos
Trans. AGU, Fall Meeting Suppl., Abst. P24A-08. [3]Ruff et
al. (2007), Eos Trans. AGU, 88(52), Fall Meeting Suppl.,
Abst. P23A-1097. [4] Morris et al. (2007), Eos Trans. AGU,
88(52), Fall Meeting Suppl., Abst. P21C-02. [5] Bandfield et
al. (2008), GRL, 35, doi:10.1029/2008GL033807. [6] Milliken et al. (2008), Geology, 36, doi:10.1130/G24967A.1.
[7]Michalski, J. R. (2005) PhD Diss., ASU, Tempe. [8]Long,
et al. (2001) Canadian Archaeological Assoc., 33rd Meeting.
[9]Hubbard et al. (2003) GSA, Abstracts with Programs,
35(6). [10] Michalski et al. (2003) GRL, 30(19),
doi:10.1029/2003GL018354. [11]Hamilton and Lucey.
(2005), LPSC, Abst. 1272. [12]Lyon and Burns (1963) Econ.
Geol., 58. [13]Ruff, S.W. (1998) PhD Diss., ASU, Tempe.
[14]Almeida, R. M. (1992) Phys. Rev. B, Vol. 45.