Composition of Meridiani Hematite-rich Spherules: A Mass-Balance Mixing-Model Approach

2005·0.25 MB·English

by  jOLLIFF, b. l

Lunar and Planetary Science XXXVI (2005) 2269.pdf
COMPOSITION OF MERIDIANI HEMATITE-RICH SPHERULES: A MASS-BALANCE MIXING-
MODEL APPROACH. B. L. Jolliff1 and the Athena Science Team. 1Department of Earth and Planetary Sciences
and the McDonnell Center for the Space Sciences, Washington University, St. Louis, MO 63130 ([email protected])

Introduction: One of the great surprises of the Mars portions of components. This abstract presents a model
Exploration Rovers (MER) mission is the discovery at for determining the composition of the spherules by
Meridiani Planum that the surface hematite signature combining the proportions of components determined
observed from orbit is attributable largely to a surface from the microscopic image with a mass-balance calcu-
enrichment of hematite-rich spherules, thought to be lation, showing as an example the results for Jack Rus-
concretions, that have weathered out of rocks similar to sell. The same analysis for the Berry Bowl target yields
the underlying sulfate-rich rock formation [1]. A strong similar results for Fe O . According to the mass-balance
2 3
hematite signature has been observed by the Mini-TES model and taking the APXS compositions at face value,
[2] and by in-situ measurements of spherule-rich targets the spherules are not pure hematite, but they instead
by the Mössbauer spectrometer (MB) [3] and the alpha- contain a significant fraction (40-50% by weight)
particle X-ray spectrometer (APXS) [4]. The Mini-TES of other components.
derived spectrum of spherule-rich targets on the plains
is consistent with nearly pure coarse-grained hematite,
40
with perhaps as little as 5-10 areal % of other compo- Spherule- Soil
rich Bright soil
nents [2]. The occurrence and abundance of the spher- 35 Spherule-rich
Fract fill/vein
ules as the bearer of the widespread hematite signature
observed by MGS TES over much of Meridiani Planum Fe) 30 RRkk nwa/st osiulrf
is significant for global remote sensing, and their occur- tal 25 BarbertonSoil Bright RBka sRaAltT'd
o Soil
rence as concretions in the outcrop lithology is signifi- %, t 20 Bounce
cfoarnmt afotiro nth oef dthiaeg seendeitmice nhtiasrtoyr ryo cakn df orromlea toiofn w [5at]e. r in the O (wt3 15 BSaasnadltic SRooiclk on Nsuartf aRcke RRAocTk'ds
Setting: In the outcrop, the spherules make up only Fe2 10
1-2% of the volume of the rocks in which they occur,
5
judging by their abundance in RAT’d surfaces. How-
ever, in some locations loose spherules are concentrated 0
in depressions, and on the plains, they are concentrated 0 5 10 15 20 25 30
SO3 (wt.%, total S)
as a lag deposit because they are hard and relatively
resistant to wind abrasion, and too dense to be moved Figure 1. FeO(T) vs. SO(T), APXS. Lines projecting
2 3 3
easily by the wind. In RAT’d exposures, cross sections to high Fe O illustrate a 2-component mixing model.
2 3
of the spherules appear to be quite homogeneous.
Method: The model presented here involves (1) spa-
Combined APXS and Mössbauer measurements in-
tial analysis of the microscopic image of the Jack Rus-
clude the Berry Bowl (Sols 46-48) in Eagle Crater and
sell target, (2) conversion of components from area pro-
on the Plains at Jack Russell (Anatolia trench site, Sol
portions to mass percent using likely component
80), Fred Ripple (Sol 91), and Leah’s Choice (Sol 100).
densities, and (3) a compositional mass-balance calcula-
Each of these were targets significantly richer in spher-
tion using the APXS data. Results are presented for a
ules and rock fragments than typical soil, and the corre-
two-component solution (soil-spherules) and a three
sponding results reflect very hematite-rich targets (MB)
component solution (soil-spherules-rocks). In the three
and high concentrations of total Fe (APXS; Fig. 1).
component model, the rock fragment component can be
Although each of these occurrences represent mixed
represented by any mixture of basalt and evaporite rock
targets including soil, spherules, and rock or rock frag-
compositions. The data for Fe(T) as Fe O are shown in
ments, the targets can be evaluated in terms of the pro- 2 3
Figure 1, where the lines represent the concept of two-
portions of components using Pancam and microscopic
component mixing. Figure 2 shows just the data for Fe3+
images. The Berry Bowl experiment was designed to
as Fe O , determined from Mössbauer results [3], and
determine the composition of the spherules by compar- 2 3
illustrates the concept of three-component mixing. The
ing the analysis of a spherule-rich target with an adja-
possible solutions indicated are the most plausible ones
cent spherule-free rock surface; however, the experi-
based on the model described here.
ment was complicated by the presence of a thin layer of
Figure 3 shows the Jack Russell target. Spherules
soil in the “bowl.” Still, the Berry Bowl experiment can
range from about 1 to 4 mm diameter and, along with
be modeled as a three-component system. The Jack
spherule fragments, take up some 15 to 30 % of the area
Russell target, which had the highest Fe(total) concen-
of the scene. Fine soil makes up 60% of the scene and
tration measured by the APXS, had a co-located micro-
the total of spherules plus rock fragments is 40%. By
scopic image well-suited for determining the areal pro-
comparison to similar targets that have Pancam multi-
Abstract for the XXXVI Lunar and Planetary Science Conference, Houston, March, 2005
Lunar and Planetary Science XXXVI (2005) 2269.pdf
There are several important caveats. One is that the
50
possible spherule Soil Av target is heterogeneous and involves several mm of to-
45 compositions Bright soil Av pography, thus violating the ideal conditions for the
Hema Trench
3+e only) 334050 BAPLeenlaarairnhtyo'ss lBi aCohwolice AvµamPrXi eaSst atmhse ea a lsfouuwnre-cemtnioeennrgt .oy fA rXlasn-org,a eyt h aeen ndeef rf5ge0yc tµi(vomen sataht meh ipoglrhidn-egern doeefr pg1tyh0,
%, F 25 Sphriecrhule- Basaltic Sand Av e.g., Fe [4]). Thus if the spherules in the Jack Russell
wt
eO (23 1250 Beerrmy pBtoywl ttahrigcke,t thhaev es pdhuesrut lec ocaotimngpso soitnio nth ew oourdlde r nooft h1a0v-4e0 b eµemn
F
Soil measured directly.
10
Bright Lab measurements on
5 Basaltic Soil
Sand heterogeneous targets
0
0 2 4 6 8 10 12 14 16 18 20 with simulated mass
SO3 (wt.%, total S as SO3) and Z distribution
such as those meas-
Figure 2. FeO [Fe3+/Fe(T), MB × Fe O(T), APXS]
2 3 2 3 ured on Mars will
showing selected compositions and graphical represen-
help to constrain po-
tation of three-component mixing model.
tential sources of
spectral coverage, the rock fragments consist of both the
measurement error.
evaporite outcrop lithology and basalt [6].
The potential effects
The distribution of spherules within the field of view
of coatings on spher-
of the in-situ instruments is a potential source of error.
ule and fragment sur-
The APXS working diameter is 2.5 cm, but most of the
faces are allayed
signal comes from a somewhat smaller area than this.
somewhat by the ob-
Taking five ~1.5 cm diameter circles and subsampling
servation of Pancam
the 20% distribution shown in Fig. 3b gives an average
multispectral data that
of 21± 2% coverage.
spherules perched on
The mass-balance equation for three components is:
the soil and exposed
Csph×mfsph + Crk×mfrk + Csoil×mfsoil = Cmix to wind abrasion are
where C is the concentration (wt%) of an element, mf is less likely to be
the mass fraction of the component, and sph=spherule, coated [6].
rk=rock, and mix=Jack Russell. Rearranging the equa- The mass-balance
tion allows solving for C : model indicates, in
sph
addition to Fe O , Figure 3. Jack Russell, sol 80
C = (C - (C ×mf + C ×mf )) / mf 2 3
sph mix rk rk soil soil sph significant FeO (Fe2+) MI ~3 cm across. In (b), blue
The conversion from area (assume area=vol) to mass covers 20 % of the area.
in the spherules (9-11
fraction assumes densities of 1.5 g/cc for soil, 2.5 g/cc
wt%), and this is not consistent with Mössbauer data.
for rock, and 4.4 g/cc for spherules.
Here, the Berry Bowl experiment provides a constraint.
Results and Discussion: Taking the APXS results
In the Berry Bowl, spherules make up some 25-30% of
at face value, the concentration of Fe O in the spher-
2 3 the area, and a three-component mass-balance model
ules may be calculated using proportions of components
indicates spherule Fe O concentrations of 42-47 wt%,
constrained by the microscopic image (Fig. 3). Results 2 3
but only 3-4 wt% FeO. In the case of Jack Russell, the
are shown in the table below.
non-Fe2O3 component has a basaltic composition, but
Table. Mass-balance mixing model results for Jack Russell depleted in CaO, whereas in the model for the Berry
Case model area, sph mf, sph Fe2O3 sph Evap:Bas Bowl, the non-Fe2O3 component of the spherules is en-
1 2-Comp 15 vol% 0.341 54 wt% riched in silica and other oxides relative to FeO.
2 2-Comp 20 vol% 0.423 45 wt% Acknowledgments: The APXS team is thanked, especially
3 2-Comp 30 vol% 0.527 36 wt% Ralf Gellert for his efforts with the APXS data. The Engineer-
4 3-Comp 15 vol% 0.302 60 wt% 1:1.3 ing and Science teams of the Opportunity Rover are thanked
5 3-Comp 20 vol% 0.386 48 wt% 1:1 for their tireless and highly skilled work and their dedication
6 3-Comp 30 vol% 0.534 37 wt% 2:1 to the mission. Funding for this work was through NASA
The ratio of evaporite to basalt in the rock component can be varied, but support of the MER Athena science team.
is constrained by the need to prevent negative S concentrations in the References: [1] Squyres et al., Science 305, 2004;
spherule composition, e.g., Case 4. [2] Christensen et al., Science 305, 2004; [3] Klingelhöfer et
al., Science 305, 2004; [4] Gellert et al., Science 305, 2004;
Note that more area coverage by spherules corre- [5] McLennan et al., LPS 36, this vol., 2005 [6] Weitz et al.,
sponds to lower spherule Fe O content. LPS 36, this vol., 2005.
2 3
tract for the XXXVI Lunar and Planetary Science Conference, Houston, March, 2005