J. Longhi
Lamont-Doherty Earth Observatory
Palisades, NY 10964
[[email protected]]

It is widely, though not universally, accepted that the SNC (shergottites-nakhlites-chassignites) meteorites come from Mars. These basaltic achondrites have traces of water-bearing minerals and magnetite, which indicates that their oxidation states are similar to terrestrial basalts, yet their oxygen isotope compositions show that they are definitely extra-terrestrial. Belief in a martian origin is based on relatively young ( 1.3 Ga) crystallization ages - such "recent" igneous activity being likely only on a planet large enough to retain its heat - and on a distinct isotopic signature of the martian atmosphere in Ne and Ar gas trapped in an impact-melted glass in one of the shergottites. As with all basaltic rocks, the SNCs provide constraints on the composition, structure, and evolution of their parent body.

Composition of the Surface and Interior. Most of the SNCs are medium- to coarse-grained rocks that may have gained or lost crystals during slow solidification; so, unlike fine-grained basalts, bulk chemical analyses do not recover their parent magma compositions. Fortunately, a large data base derived from melting experiments allows us to estimate the composition of the parent magmas from the mineral compositions measured in the meteorites. Estimates of the parent magma compositions of the SNCs show that martian lavas have a wide range of low Al2O3 and high FeO concentrations with an average that is comfortably close to those of the Viking Lander soil analyses. Melting experiments also tell us that FeO is fractionated only weakly during planetary melting and limited subsequent crystallization of magma near the surface, so the average FeO concentration (~18 wt%) in the SNC parent magmas is probably a good approximation of the FeO concentration in the martian mantle.This value is more than twice that of the Earth's upper mantle. Another constraint on martian composition comes from the SNC K/La ratio, which is less than chondritic, but nearly twice as high as the Earth's. Because K and La are both highly incompatible elements, their ratio changes very little during most igneous processes, yet bulk K abundance is likely to vary from planet to planet because of its moderate relative volatility, whereas all of the rocky planets are expected to similar (i.e., chondritic) abundances of La and the other refractory elements (Ca, Al, Ti, Mg, Si, REE, U, Th).Thus the K/La ratio is an effective planetary probe which indicates that Mars has a richer endowment of the volatile elements (H, C, Na, S, P, etc.) than the Earth. Some of this endowment may have been spent, however, in oxidizing Fe that wound up as FeO bound in mantle silicates instead of metal in the core. Because the most likely oxidant is H2O, it is possible that Mars is actually much drier now than the Earth. Relatively dry magmatism is also indicated by the relatively sparse alteration in the SNC meteorites and by the low OH content of SNC amphiboles (ion probe analyses reveal that only about 10% of the hydroxyl sites contain OH groups).

The total amount of S in Mars that is consistent with the SNC K/La ratio is far in excess of what could be in the martian mantle, so it is likely that much of the sulfur is in the core. This inference is consistent with estimates of relatively low mantle abundances for elements such as Ni, Co, and Cu, which have strong affinities for Fe metal and sulfide. Placing most of the inferred S in the core and accepting as a trial proposition that Mars has overall chondritic abundances of Fe and Ni allows the composition and size of the martian core to be computed by mass balance (78% Fe, 8% Ni, 14% S; 1700 km). Data on the inclinations of Mars' axis of rotation obtained from the Pathfinder mission will yield an accurate moment of inertia which will be the first test of the geochemical model. A more rigorous test will come from seismic measurements (Mars Internet) which should allow direct determination of core size. Core composition also plays an important role in determining thermal structure. The absence of a significant martian magnetic field suggests that a strong core dynamo, driven by the release of latent heat of crystallization, is not operating. Thus, either the martian core is completely molten or completely solid; and the liquidus of the core, which depends on the S concentration, provides a minimum temperature for a "hot" martian geotherm , while the core solidus provides a maximum temperature for a "cold" geotherm.

Magmatic Style. In addition to constraints on Mars' bulk composition and structure, the SNC meteorites provide insights into Mars' melting and differentiation processes. The low-Al2O3 contents of the SNC parent magmas are consistent with differentiated source regions, either lunar style (magma ocean cumulates) or terrestrial style (depletion by basalt extraction). The absence of negative Eu-anomalies in the rare earth patterns of the SNC parent magmas obviates formation of an ancient anorthositic crust and complementary Al-poor mantle. Patterns of extreme depletion of the light rare earths in chondrite-normalized plots of the Antarctic shergottite REE abundances are consistent with multiple episodes of previous melt extraction. Because shock effects have disrupted some of the isotopic systems, crystallization ages for most of the shergottites remain ambiguous. One interpretation, supported by a Sm-Nd whole rock isochron, is that most of the shergottites formed at the same time as the nahklites and chassignites (~ 1.3 Ga) and were then remelted by an impact at ~ 0.18 Ga. Sm-Nd isotope systematics require that the nahklite source region had long-term, strong light-REE depletion prior to melting. If 1.3 Ga was the time of shergottite formation, then Nd isotopes require that their source was long-term light REE enriched (crust) and that the observed patterns of light-REE depletion were produced in the melting process. Although improbable, variable amounts of fractional fusion could produce these relations. If 0.18 Ga was the crystallization age of most of the shergottites, then their source was similar to that of the nahklites, long-term REE-depleted, and the apparent 1.3 Ga whole rock isochron is an artifact of assimilating a light REE-enriched component (crust?) into melt from a depleted mantle. Besides being more conventional, the 0.18 Ga scenario is also consistent with complementary trace element patterns of rare earth and high field strength elements (HFSE = Nb, Hf, Zr) in nahklite and shergottite parent magma compositions: it appears that a Nahkla-like melt component was extracted from the shergottite source prior to melting (Nd isotopes specifically prohibit this relation if the shergottites formed at 1.3 Ga). Another clue to martian magmatism lies with Ba and Sr abundances. Ba and Sr usually behave similarly to Th and Ce, respectively), yet in the Antarctic shergottites Ba and Sr are dramatically enriched with respect to Th and Ce. These enrichments in concert with nonchondritic ratios of HFSE to REE are characteristic of convergent plate margin volcanics on Earth. This similarity presumably does not imply plate tectonics on Mars, but probably does suggest that fluids preferentially extracted Ba and Sr with respect to the REE and preferentially extracted REE with respect to the HFSE. And subsequently these fluids metasomatized depleted mantle, possibly fluxing melting in the process.

A final clue provided by SNC meteorites lies in the 4.50 Ga crystallization age of one shergottite (E84001). The very age of this plutonic rock, which is apparently well constrained, is grounds for amazement - finding similarly old and intact pristine rocks on the Moon takes considerable effort. That the rock is relatively unaltered is also amazing in light of the popular notion of an ancient warmer, wetter epoch replete with hydrothermal activity on Mars. Most significantly, perhaps, is the implication that Mars' magmatic style did not change over geologic time: the texture and orthopyroxene composition indicate close affinities to the other shergottites, which are much younger. Thus Mars' crust may consist primarily of a pile of SNC-like volcanic and hypabyssal rocks with no distinction in composition between old crust and young volcanics, unlike the Moon or Earth. Of course, acceptance of this unique style of planetary differentiation is contingent upon finding more rocks with similar petrological characteristics.

In summary, the SNC meteorites tell us a great deal about Mars, and there will even greater understanding once the moment of inertia and core size are measured accurately; but the SNCs' stories are inherently limited by the lack of geological context and by the the potential disruption of isotopic systems and magnetic domains by shock in the impact events that blasted these rocks off of Mars. Returned samples of igneous rock offer the potential of not only enhancing our understanding of the Martian interior significantly, but also hold promise of calibrating the absolute ages of Mars' surface features - knowledge that should prove vital to any systematic plan of exploration.


Figure 1: Rationale for Fresh Igneous Samples

Figure 2: SNC Meteorites

Figure 3: SNC Parent Magmas: Major Elements

Figure 4: SNC Parent Magmas: Trace Elements

Figure 5: Sampling Goals

Figure 6: Precursor Measurements

Figure 7: Composition of SNC parent magma compositions with solar system hypersthene-normative basalts

Figure 8: Comparison of the internal structures of the Earth and Mars to scale.

Figure 9: Schematic phase equilibria relations for the Dreibus-Wänke Mars model and calculated temperature- depth relations

Figure 10: Refractory incompatible element pattern of SNC parent magmas.

Figure 11: Refractory incompatible element pattern of SNC parent magmas including.

Figure 12: Refractory incompatible element pattern of terrestrial volcanic rocks from convergent plate margins.

View comments