VANADIUM OXIDATION STATES IN ALLENDE CAI FASSAITE AND SYNTHETIC PYROXENE. S.R.
Sutton1,2, S. Simon1, L. Grossman1, J.S. Delaney3, J. Beckett4, M. Newville2, P. Eng2, and M. Rivers1,2. 1Department
of Geophysical Sciences and 2Consortium for Advanced Radiation Sources, The University of Chicago, Chicago, IL
60637 (sutton@cars.uchicago.edu); 3Geology Department, Rutgers University, New Brunswick, NJ; 4Geological &
Planetary Sciences, Caltech, Pasadena, CA 91125
[More sample description here?]
The microXANES analyses were performed with
the GeoSoilEnviroCARS x-ray microprobe at the Advanced Photon Source (APS, Argonne National Laboratory, IL). The apparatus consisted of an APS undulator x-ray source, silicon (111) cryogenic monochromator, Kirkpatrick-Baez microfocusing mirrors (ref) and a
germanium solid state x-ray fluorescence detector. The
x-ray beam size was 5 m. XANES spectra were typically measured over the V K absorption edge (5465
eV) from 5450 to 5500 in 0.2 eV steps (~10 sec dwell)
with additional measurements made well below (5400
eV) and well above (5700 eV) the absorption edge for
normalization purposes (absorption step defined to be
1000). The pre-edge peak (near 5470 eV) was characterized by the energy and normalized intensity at the
maximum. Previous work has shown these parameters
to be useful for oxidation state determinations.
Energy calibration was accomplished using a V
metal foil spectrum and assigning the first inflection
point to be 5465 eV. Standardization in terms of preedge peak energy and intensity versus oxidation state
was accomplished using glasses characterized by optical spectrometry (H. Schreiber, private communication) and measurements on oxides (Wong et al. 1984).
In addition, spectra were obtained on 2 pyroxene crystallization experiment charges produced by J. Beckett
(Caltech) at the C-CO buffer (CMAS @ 1280C;
log(fO2) ~= -17).
Results: A scan of V concentration versus position
was done (2 m step) across the “V spike” observed
previously near analysis spot 5. This scan showed the
presence of the narrow (<~10 m), high-vanadium
zone. The concentration “step” of about 10:1 is comparable to that observed in electron microprobe data.
Allende Fassaite Vanadium XANES: XANES spectra were obtained as spot analyses at pre-selected positions near those chosen for electron microprobe analysis. Qualitatively, the fassaite spectra were similar exhibiting very small pre-edge peaks indicative of a highly reduced oxidation state. The pre-edge peak intensi-
Pre-edge Peak Energy (eV rel. 5465)
Introduction: Refractory inclusions in carbonaceous chondrite meteorites contain cosmochemical
clues about the processes at work in the early evolution
of the Solar System. Some refractory inclusions may
be direct gas-to-solid condensates, but many underwent
melting as individual bodies in the nebula. A major
mineral in one of the dominant types of refractory inclusions is fassaite, a Ti-, Al-rich clinopyroxene. Many
fassaite grains are stongly zoned in Ti and V abundance with abundances decreasing from core to rim
and concentration spikes are common. These two elements are also the only major elements in this phase
with multivalent chemical characteristics. One possibility is that the zonation resulted from variable redox
conditions at the time of crystallization. This interpretation is particularly compelling for titanium because
Ti3+ is known to be more compatible in fassaite than
Ti4+(ref). In fact, the Ti3+/Ti4+ ratio in these fassaites
has been shown by stoichiometric treatment of electron
microprobe data to increase at the abundance spikes
(ref).
We report here on our initial results using x-ray absorption near edge structure (XANES) spectroscopy
(refs) for determining the oxidation state of vanadium
with ~ micrometer spatial resolution. Measurements
were conducted both on Allende fassaite and synthetic
pyroxene. The ultimate goal is to place constraints on
the oxygen fugacity of the solar nebula during the crystallization of individual fassaite grains by also determining the pyroxene/liquid partitioning behavior of V
as a function of valence state.
Methods: The Allende fassaite studied here was
grain ZF2 on thin section TS34 described previously (ref).
6
Schreiber
Glass
Wong 1984
Oxides
Allende
Fassaite
Beckett
pyroxene 1c
Beckett
melilite 1c
5
4
3
2
2
3
4
5
V Oxidation State
Figure 2: Peak energies vs oxidation state. Oxidation
states for the Allende and Beckett samples are those
determined using the intensities (figure 1).
SHORT TITLE HERE: A. B. Author and C. D. Author
ties fall between those of V3+ dominated oxides (e.g.,
Schreiber 112 and 113) and V2+ oxides. The latter are
expected to show no pre-edge intensity (as in VO) due
to the expected high symmetry of the octahedral site.
Thus, the presence of a pre-edge peak indicates vanadium in the fassaite is not totally divalent.
XANES spectra can also be sensitive to crystallo-
140
Schreiber
Glass
Pre-edge Peak Intensity
120
Oxides
100
80
Allende
Fassaite
60
40
Beckett
pyroxene 1c
20
0
2
2.5
3
3.5
Beckett
melilite 1c
V Oxidation State
Figure 1: Pre-edge peak intensity versus V oxidation state
showing V2+-V3+ region only.
graphic orientation relative to the polarization direction
of the synchrotron radiation, but this effect is expected
to be less important as site symmetry increases. To
check for this effect, the two spot analyses were repeated with the thin section rotated 90 degrees. These
spectra were identical to those obtained initially indicating negligible orientation effect.
Quantitative oxidation states were computed from
the pre-edge peak intensities using a linear interpolation between the intensity measured for Schreiber glass
112 (oxidation state determined by optical spectrometry of 3.12) and zero (expected for V2+). The results
are shown graphically in figure 2.
The data in figure 2 is expanded in figure 3 to show
only V2+-V3+ space. The mean V oxidation state for
the fassaite measurements is 2.4 with a standard deviation of 0.1. The small standard deviation indicates
that the large V concentration variation in this grain
(order of magnitude) is unlikely to be due to variations
in V redox state. The fairly tight range of intensity
values is comparable to the typical measurement precision. The variation in oxidation state along the traverse is uncorrelated with vanadium content (figure 4).
Synthetic Materials: Beckett’s samples VT01-1a
(rapid quench) and VT01-1c (cooled at 50 deg/hr to
1000) were measured. Both showed well developed
crystals and XANES spectra were obtained on grains
identified beforehand by Simon. These results are also
shown on figures 2 and 3.
The pyroxene grain noted on Steve’s VT01-1c BSE
photo yielded an oxidation state indistguishable from
the mean of the Allende fassaite measurements (2.4).
Two other phases were analyzed, melilite (2.4) and a
dark-gray (BSE) phase tentatively denoted “glass”
(2.7). Together these results suggest the pyroxene
crystallation experiments reproduced the crystallization
conditions of the Allende fassaite grain fairly closely.
Beckett’s sample VT01-1a was very interesting …
and puzzling. A line scan of the V content was done
across a region that included pyroxene, melilite and
glass, and the locations of the grain boundaries were
easily recognized by steps in the V K fluorescence intensity profile. However, the XANES spectrum for
each of these regions was the same and consistent with
glass spectra we’ve obtained previously on the reduced
glasses from the Hanson, Schreiber and Canil synthetic
suites. The only interpretation that comes to mind at
the moment is that the grains contain melt inclusions.
Such inclusions are not obvious viewing in the petrographic scope…if they’re there, they’re likely to be
tiny (sub micron?). The pre-edge intensity for these
spectra yielded an oxidation state of 3.2, i.e., significantly more oxidized than the other measurements.
This sample needs to be investigated further.
No XANES measurements were done on sample
VT01-1b that appeared to have no crystals (Simon’s
observations).
Figure 5 shows the pre-edge peak energies plotted versus oxidation state (determined from intensities). This
plot shows that the measured energies for the pyroxenes are consistent with the oxidation state assignments
based on the intensities. A reasonably good correlation
exists between the pyroxenes, the Schreiber glasses and
the oxides measured by Wong 1984.
Discussion: The oxidation state standards used in
this work are a FAD glass containing predominately
V3+ and a V2+ oxide (VO). An underlying assumption
is that the site geometry is the same in the standard and
the unknown. This is likely to be a good assumption
here because octahedral coordination of V dominates.
VO consists of regular octahedral VO6 units. Optical
spectroscopy on V3+ in albite-diopside glass shows
vanadium in octahedral sites (Keppler 1992). In diopside, based on ionic size arguments, V3+ is likely to go
into the M1 octahedral site but V2+ is large enough that
it may be forced into the M2 site substituting for Ca2+.
The coordination of V2+ in M2 is uncertain. Calcium is
8-fold but divalent cations that substitute for it (e.g.,
Mn2+ and Fe2+) are octahedral. The coordination geometry of V in these samples needs to be investigated
further perhaps using extended x-ray absorption fine
structure spectroscopy.
References:
Delaney, J.S., Jones J.H., Sutton
S.R., Simon S. and Grossman L. (1999) In situ microanalysis of vanadium, chromium, and iron oxidation
states in extraterrestrial samples by synchrotron microXANES (SMX) spectroscopy. Meteoritics & Planet. Sci. 34, A32.
Delaney J.S., Sutton S.R., Newville M., Jones
J.H., Hanson B., Dyar M.D. and Schreiber H. (2000)
Synchrotron micro-XANES measurements of vanadium oxidation state in glsses as a function of oxygen
fugacity: experimental calibration of data relevant to
partition coefficient determination. Lunar Planet. Sci.
XXXI, abstr. # 1806 (CD-ROM).
Quartieri, S., G. Antonioli, G. Artioli, and P.
Lottici (1993) XANES study of titanium coordinatioin
in natural diopsidic pyroxenes. Eur. J. Mineral. 5,
1101-1109.
Simon S.B. and Grossman L. (1991) Profiles
of Ti3+/Titot ratios in zoned fassaite in Allende refractory inclusions. Meteoritics 26, 395.
Simon S.B., Davis A.M. and Grossman L.
(1992) Evidence for changes in redox state during crystallization of Allende type B1 inclusions. Meteoritics
27, 289-290.
Waychunas, G. (1987) Synchrotron radiation
XANNES spectroscopy of Ti in minerals: Effects of Ti
bonding distances, Ti valence and site geometry on
absorption edge structure. Am. Mineral. 72, 89-101.