Correlations Between Surficial Sulfur and a REE - USRA

46th Lunar and Planetary Science Conference (2015)
77058 ([email protected]), 2CRESST, NASA/GSFC, Greenbelt, MD 20771
Introduction: Compared to terrestrial basalts, the
martian shergottite meteorites have an extraordinary
range of Sr and Nd isotopic signatures. In addition, the
S isotopic compositions of many shergottites show
evidence of interaction with the martian surface/atmosphere through mass-independent isotopic
fractionations (MIF, positive, non-zero ∆33S) that must
have originated in the martian atmosphere, yet ultimately were incorporated into igneous sulfides (AVS
— acid-volatile sulfur; [1]). These positive ∆33S signatures are thought to be governed by solar UV photochemical processes. And to the extent that S is bound
to Mars and not lost to space from the upper atmosphere, a positive ∆33S reservoir must be mass balanced
by a complementary negative reservoir.
There are two current explanations for the large
spread in shergottite Sr and Nd isotopic compositions.
The first model assumes that this isotopic diversity
represents a martian mantle heterogeneity with respect
to a KREEP-like component produced from the crystallization of a martian magma ocean [2]. The second
model assumes that all shergottites originated from a
very depleted mantle and that the observed Nd and Sr
isotopic dispersion results from assimilation of an ancient, enriched basaltic crust [3].
Newly acquired S isotopic data [1], coupled with
REE whole-rock analyses [4], serve to constrain these
two disparate models. We introduce here a parameter
La/Sm* [(La/Sm)sample/(La/Sm)QUE], a proxy for crustal
assimilation, where QUE refers to QUE 94201, the
shergottite with the smallest La/Sm ratio. In this notation, therefore, the La/Sm* for QUE is unity (Fig. 1).
And the AVS ∆33S of QUE is indistinquishable from
∆33S vs. La/Sm*: Figure 1 shows that there are at
least two distinct ∆33S-La/Sm* trends reflected in martian basalts. First, there is a trend of extremely variable La/Sm* with no obvious change in ∆33S. A second trend shows a positive correlation between ∆33S
and La/Sm*. These trends constrain, for an individual
shergottite, the amount of surficial MIF S that the rock
has incorporated.
Some shergottites, with large amounts of lithophile
crustal assimilants (high La/Sm*), have no ∆33S anomaly (e.g., Shergotty and Zagami), but no shergottites
yet analyzed have yielded low La/Sm* and high ∆33S.
The latter observation is most consistent with a crustal
assimilation model.
Discussion: The two models for the physical nature of the enriched component in shergottites make
very different predictions about the systematics of Fig.
1. If the enriched component resides in the mantle,
there will be no correlation between ∆33S and La/Sm*.
Conversely, if the enriched component resides in the
crust, those shergottites with the least crustal interaction would also be predicted to have the smallest ∆33S
anomalies. Taken at face value, the systematics of Fig.
1 are much more consistent with a crustal reservoir
being the enriched component.
Our physical interpretation of these observations is
that positive ∆33S anomalies were produced in the atmosphere by UV irradiation and that these isotopically
anomalous compounds eventually made their way to
the martian surface. The speciation of this high-∆33S
sulfur is unknown, but given the oxidized nature of the
martian surface, we postulate that that the high-∆33S
sulfur was either oxidized or became oxidized. Fluids
may then have incorporated this sulfur and transported
it deeper into the crust. Some shergottites then assimilated this high-∆33S component on their ascent to the
surface. The oxidized high-∆33S component was then
reduced to sulfide, probably by oxidizing Fe2+ to Fe3+,
but other redox reactions are also possible.
Figure 1. ∆33S vs. (La/Sm)sample/(La/Sm)QUE. Error
bars are one-sigma [1]. All analyses are for AVS (sulfide) only. No sulfate analyses are represented here.
Arrows are illustrative and are not regressions.
Another possibility for transferring a high ∆33S signature from an oxidized species to a reduced species is
46th Lunar and Planetary Science Conference (2015)
isotopic exchange, without a bulk change of redox
state. Sulfur in mantle-derived martian basalts is
thought to be dominated by the sulfide species, not
sulfate [5]. If isotopic exchange occurred between
mantle-derived sulfide and assimilated sulfate, bulk
conversion of sulfate to sulfide may not be necessary.
Higher Order Terms. Although we believe our
physical model is reasonable, there are several features
that suggest more complexity.
First, there are several shergottites with very modest La/Sm* (~2-3) have have a small ∆33S anomaly and
larger than the ∆33S values of several highly enriched
Second, Los Angeles, which is nearly indistinguishable in its Sr and Nd isotopic characteristics from
Shergotty and Zagami, has a large ∆33S anomaly,
whereas these latter shergottites have none. Clearly,
there can be a decoupling between various crustal and
surficial components during the assimilation process.
Third, the shergottites NWA 2990 and NWA 5960
are believed to be paired. And while both of these
samples show large, positive ∆33S anomalies, the
analyses do not agree within their 1σ analytical errors.
And while there would be a slight overlap of the 2σ
errors, there is a suggestion that the anomalous ∆33S
component may be heterogeneously distributed at the
hand-specimen scale.
Fourth, there is presently very little sulfur in the
martian atmosphere. Therefore, current UV production of ∆33S anomalies is likely to be very inefficient.
The easiest means of overcoming this problem is for
there to have been more sulfur in the atmosphere at
times in the past, perhaps during volcanic eruptions.
However, accomplishing this probably requires a
multi-stage process.
As noted above, mantle-derived sulfur is dominantly sulfide, not sulfate, and is generally much less
volatile than oxidized sulfur species. Therefore, for
large volumes of oxidized sulfur to be erupted by volcanism, mantle sulfide must first be oxidized. This is
certainly possible if mantle-derived basalts assimilated
oxidized crustal materials [e.g., 6], but volcanic outgassing of sulfur is not necessarily a given. Oxidation
state must be specified.
We do note, however, that there is a substantial
surficial reservoir of sulfur in the martian regolith.
And solar UV should penetrate to the martian surface.
But at present, we do not understand whether MIF reactions are likely for UV-solid interactions.
Complementary reservoirs. We noted above that
mass balance probably requires a negative ∆33S reservoir to complement the positive reservoir observed in
the shergottites.
References: [1] Franz H.B. et al. (2014) Nature
508, 364-367. [2] Borg L.E. and Draper D.S. (2003)
Met. Planet. Sci. 38, 1713-1731. [3] Jones J.H. (2003)
Met. Planet. Sci. 38, 1807-1814. [4] Meyer C. (2012)
Mars-Meteorite-Compendium. http://curator.jsc.nasa.
gov/antmet/mmc/. [5] Righter K. et al. (2008) Met.
Planet. Sci. 43, 1709-1723. [6] Wadhwa M. (2001)
Science 291, 1527-1530.