Isotopic Studies of Water from Martian Meteorites

Wide ranging, light element isotopic studies of Martian meteorites have been carried out for many years with the majority of early analyses concentrating on carbon, nitrogen and oxygen extracted from silicates and other anhydrous phases. Analysis of these elements from apparent Martian weathering products gave the first clues to the presence of products derived from hydrous alteration [e.g. 38]. However, the first analyses of water [39, 40], in studies that focussed on D/H ratios, found only terrestrial signatures. These studies were probably hampered by the relatively small indigenous water contents of the meteorites and restricted sample availability as subsequent studies found water with non-terrestrial signatures. The relatively anhydrous character of Martian meteorites, coupled with their precious nature, has limited the number of viable techniques available and therefore also the number of completed studies. Despite this, several other groups have made D/H measurements of water in these meteorites [e.g. 41, 42], each identifying hydrogen indigenous to the meteorites that would initially have been in the form of water. Studies reporting the measurement of oxygen isotopes from water include those by [33] and [31]. Both of these studies identified indigenous water in addition to terrestrial contamination.

More recently, measurements of D/H in Martian meteorites have been made using ion microprobes [27, 42, 29], which allow an appraisal of spatial variations in solid samples. This technique is not applicable to oxygen analysis due to the high oxygen content of host silicates. However, while the ion microprobe offers an unrivalled ability to analyse well characterised phases in polished thin sections, it is of little use when minerals are part of complex mixtures, as is often the case with hydrous, aqueous alteration phases. For such phases, ratios of D/H together with 17O/16O and 18O/16O are more accurately characterised using a technique of stepped heating, where crushed samples are subjected to progressively higher temperatures (typically increments of 100 - 200°C are used) either in a vacuum or in a flow of helium. It is important to note at this point that while a relatively small number of hydrated minerals has been identified in many of the meteorites (Section 1.3), the water extracted upon stepped heating may originate from several different sites within these phases. In all such relatively anhydrous samples, a large proportion of the water lost will be that adsorbed to sample surfaces. This is usually lost at low temperatures during stepped heating and will inevitably be of terrestrial origin. The next water to be lost from samples is that existing as water molecules held between layers in clay minerals. Only smectite-type or expanding clays hold such water, but this can represent a considerable proportion of the total yield. This inter-layer water can move relatively freely and thus is always liable to reflect the latest environment in which the samples were kept and thus will also produce a terrestrial signature. Finally all hydrated minerals contain structural OH groups that are bound more tightly within the minerals and thus generally liberate water at temperatures in excess of 250°C. It is these that can potentially retain an indigenous isotopic signature.

The variable results of the earliest investigations of the D/H ratios present in Martian meteorites [39, 40] probably reflected degrees of contamination by terrestrial reservoirs during analysis. The first study that recognised a significant deuterium enrichment [41] was completed using large samples of 2.0 and 2.8 g from Shergotty and Lafayette. Values for 5D of up to +800 % were measured in water released at temperatures between 450 and 1050°C (water extracted up to 450°C was assumed to be largely terrestrial in origin and so was discarded). A more extensive study of D/H ratios [27] was completed using an ion microprobe to allow targeting of individual magmatic phases in three Martian meteorites. Water contents of the phases were also measured. The minerals targeted were kaersutite (an amphibole), biotite and apatite and in each case analyses showed large enrichments in deuterium ranging from +500 to +4400 %, much greater than had been measured previously. The water contents of these phases did not conform with terrestrial counterparts either, with each of the phases measured yielding only around 10 % of the expected water.

A later study [34] used a more conventional method of analysis. Large whole-rock samples (0.42 - 2.56 g) of eight Martian meteorites were subjected to a regime of step heating to extract volatiles with D/H measurements on the resulting water. A range of 5D values from + 250 to + 2100 % were found in seven of the meteorites, with the highest values measured from the higher temperature steps. All measured compositions were assumed to be minimum values with a variable contribution from terrestrial contamination. One meteorite, Chassigny, was found to have water possessing a terrestrial value from all but the highest temperature step, possibly reflecting nearly complete replacement of indigenous water with that derived from terrestrial reservoirs.

In a return to microprobe analysis, meteorite QUE 94201 was analysed in an attempt to better characterise the D/H ratio of primary water extracted from magmatic minerals [42]. Apatite grains were again targeted and produced a range of §D values of between +1700 and +3600 %o. The observed range in §D seemed to show an anti-correlation with the amount of water present in the apatite samples, with a minimum in §D coinciding with a maximum initial water content.

The analyses of water in SNC meteorites [27, 41] revealed a wide range of hydrogen isotopic compositions that were nearly always enriched in deuterium relative to Earth-derived reservoirs, but did not result in any specific predictions of the water inventory on Mars. However, they did go a long way to confirming the large positive §D values of atmospheric measurements made in remote sensing studies and suggested that a link existed between the atmosphere and the hydrosphere. Interpretation of the deuterium excess found in these studies can be explained using a water inventory derived from one or a number of the above mentioned sources (Section 1.2).

The composition of Mars' initial inventory was assumed [41, 27] to be similar to that of the Earth and the increased D/H ratio was the result of atmospheric loss processes of the types discussed previously. A further assumption [27] was that pristine magmatic minerals should reflect this. However, to explain the large deuterium excesses found during subsequent analysis of the minerals, it was suggested that isotopic enrichment was by alteration on the Martian surface with deuterium-enriched fluids. This would require conditions similar to terrestrial hydrothermal circulation and alteration, and would also require close linkage between the atmosphere and hydrosphere. In this case analysis of magmatic phases would provide more information about water involved in a hydrological cycle than about the composition of primordial Martian water. In a subsequent study [42] the composition of hydrated, magmatic minerals was found to vary according to their presumed original water content. Those minerals with the greatest initial contents of water suffered less isotopic alteration during hydrothermal activity. The conclusion of this study was that the initial magmatic water composition on Mars actually had a D/H ratio approximately twice that of terrestrial water (i.e. §D of ~ + 900 %) and that this initial enrichment was the result of an earlier period of hydrodynamic escape resulting from an enhanced flux of UV from the developing sun (see [43] for a full explanation). If this were the case, then all estimates of the Martian water inventory based upon the assumption of a terrestrial-like starting composition would be in error. However, a recent, more detailed ion microprobe study of D/H ratios in a suite of Martian meteorites [29] revealed §D compositions with values ranging to as low as 0 %. The magmatic minerals targeted were found in melt inclusions in ALHA 77005 and Chassigny, and with no evidence of shock alteration the authors' conclusion was that the initial composition of Martian water was indeed similar to that measured on Earth.

A study using measured hydrogen isotopic compositions of hydrated minerals in Martian meteorites [44] and an estimate of the loss rate of hydrogen from the atmosphere both now and in the past, showed that a conservative estimate of the amount of water present 4.5 Ga ago was equivalent to a layer of 42 - 280 m. The range suggested depends upon the age at which the minerals in the Martian meteorite Zagami were at equilibrium with water in the Martian hydrosphere, assuming a steady loss rate. However, if as many believe, loss rates were higher early in Mars history, either as a result of higher atmospheric temperatures or pressures, a possibility that is not prohibited by the isotopic data, then the original water inventory may have been equivalent to a 2200 m layer. This, in turn, would suggest that the amount remaining today in the Martian crust may be as much as 190 m equivalent depth [44].

The idea that it was largely Mars' initial inventory of water that was retained and modified by loss processes to produce what we see today is not universally accepted. Thus while conclusions of most of the major studies of D/H in Martian meteorites include the effect of atmospheric loss processes, the measured excesses can also be accounted for by a variety of other explanations. A study of Martian magmatism [33] concluded that juvenile water alone may have produced a global equivalent depth of around 200 m. These calculations were based upon the assumption that pre-eruptive magmas on Mars contained around 1.4 wt % water, together with an estimate of the volume of magma erupted in the last 3.9 billion years. As this water would have been in addition to any retained and did not include that erupted prior to 3.9 Ga ago, it was considered a lower limit. However, the actual water content of the hydrated phases [27] was only around 10 % of that expected and consequently the estimate based upon juvenile water had to be reduced from 200 to 20 m. This is a similar quantity to a 10 - 20 m estimate which was also based upon the water content of Martian meteorites [45]. This study suggested that the remainder of any initial water inventory was consumed by total oxidation of available iron during accretion with loss of hydrogen to space.

The D/H excesses measured in different Martian meteorite studies prompted other theories to be developed as to the possible source of water. Impacting bodies with a range of compositions are thought by many to have played an important role in contributing to the atmospheres of the terrestrial planets [e.g. 46, 47]. After several studies of D/H in carbonaceous chondrites [48] a D/H excess in these meteorites is well established and carbonaceous chondrites, which can contain up to around 20 wt % water, have been proposed as a possible source of the deuterium-rich water on Mars.

An alternative theory [49] recently revised [46], suggests that the majority of volatiles, including water, originated from impacts of cometary bodies towards the end of the late, heavy bombardment. This theory also suggests that the original atmosphere resulting from accretion had been lost. The authors do not attempt to estimate a depth for the Martian water layer, indeed they favour a model where repeated cycles of development and loss of atmosphere occurred through Mars' evolution. In each cycle, the water composition may have been different depending on the exact nature of the comets, which were envisioned as being variable in volatile composition depending upon the location and temperature at which they formed.

Water derived from impact delivery either of carbonaceous or cometary material is inherently deuterium-rich. This is the result of incorporation of deuterium enriched molecules from the cold outer parts of the solar nebula. The deuterium enrichments themselves were imparted by ion-molecule reactions in cold parts of the inter-stellar medium prior to formation of the Solar System [5, 50]. However, as explained in Section 1.2, water in some comets may have undergone isotopic exchange with a deuterium depleted reservoir close to the sun, and consequently may differ markedly from other comets formed at greater heliocentric distances. Thus the measured deuterium enrichment may have resulted from comets derived from a combination of sources, each having a unique deuterium signature which may then have been subject to further evolution resulting from preferential loss of hydrogen to space.

A major problem for estimates of Mars' inventory based upon isotopic compositions as measured in Martian meteorites is the assumption that all available water is active within any Martian hydrological cycle. Recent Mars Odyssey data suggest that a considerable amount of water exists just below the Martian surface (Chap. 5 by Mitrofanov) and this may well be actively participating in the atmosphere/hydrosphere system within geological timescales (Chap. 8 by Tokano). However, it is entirely possible that a great deal more may lie deeper within the crust, isolated from the active contingent. Suggestions as to how such a situation may have developed [51] seem quite plausible and it may, therefore, be that isotopic ratios of O2 and H2 are only providing a measure of the surface, exchangeable reservoir. Assessing the extent of interaction of the total water inventory during Mars history is crucial. If, in the past, a larger proportion of the Martian crustal inventory was actively exchanging with the atmosphere, then the implication is that the total water inventory will be small. If, however, the majority of crustal water has been isolated, then D/H ratios do not preclude the possibility that much greater quantities of water exist in subsurface regions. The recent report [52] that Mars lacks any concentrated deposits of carbonates, at face value seems to preclude the presence, in the past, of any large-scale water-bearing bodies at the surface (oceans etc.). This is yet a further constraint on models of water evolution to take into account.

1.5.2 Oxygen Isotopic Studies

Studies of the oxygen composition of Martian meteorites have been dominated by those looking at isotopic abundance in silicates. As described in Section 1.3, these measurements were vital in establishing the provenance of the meteorites.

The importance of making oxygen isotopic measurements of water is that the data provide a way of distinguishing unequivocally between water of terrestrial origin and that which is indigenous to the meteorite. The first study to make such measurements from Martian meteorites, that included the important 17O measurement, were completed by Karlsson et al. [33]. This work looked at six individual meteorites with water being extracted from large samples of between 2.0 and 3.4 g, over a range of temperatures from 0 to around 1000°C. Despite the large size of the samples, their anhydrous nature meant that the heating profile was restricted to 4 individual steps. Total water contents ranged from only 0.04 wt % in Zagami to 0.4 wt % in Lafayette. The results of the study produced 518O values that, while far from identical, did display some inter-sample consistency. The A17O values (reproduced in Fig. 1.2) show a more consistent pattern, with most samples typically releasing water close to the terrestrial fractionation line in the first one or sometimes two steps (150 and 350°C) before rising to positive values at higher temperatures (650 and 1000°C). Three of the meteorites produced A17O values of particular note, Shergotty and EETA 79001 because they were exceptions to the general trend and Chassigny because it did follow the trend and as such contradicted the results of D/H measurements on bulk rock [34]. In fact the oxygen isotope data from Chassigny seem to indicate the presence of water indigenous to Mars whereas the results of D/H measurements on bulk rock were only able to detect water of terrestrial origin. The other two meteorites (both Shergottites) do not appear to show any convincing evidence of Martian water, a direct contradiction of the D/H bulk rock measurements that clearly indicated the presence of water derived from indigenous sources. A further observation resulting from the study was the lack of isotopic equilibrium between the water and the silicates, a point discussed below. A more sophisticated study of water in Martian meteorites used a continuous flow technique that required considerably less gas for analysis and consequently has greatly reduced sample requirements. This study [31], analysed samples from four Martian meteorites; EETA 79001, Nakhla, ALH 84001 and DaG 476. Samples analysed were all around 50 mg but were usually able to produce sufficient water to allow at least six separate temperature steps and so allow clearer distinction between terrestrial and indigenous water.

Two of these meteorites were also included in the previous study, however, ALH 84001 which is notable for its great age, around 4 billion years [53, 54] and DaG 476 a desert meteorite, had not previously been analysed and provided interesting new data. The results from this study, shown in Fig. 1.3, are consistent with those previously gained [33] with low temperature water displaying a terrestrial composition and higher temperature water indicating indigenous reservoirs. The greater number of temperature steps afforded by the improved sensitivity [31], produced an improvement in resolution and in all cases allowed clearer distinction between terrestrial water and indigenous water and also suggested a mixture at intermediate temperatures. In the case of DaG 476 the large water content of the meteorite allowed many individual steps but suggested that the results of weathering in harsh desert conditions had overprinted most of the indigenous signature with terrestrial water. The analysis of ALH 84001 provided the most interesting data with a large anomaly in the A17O value being recorded in water driven off at 300°C.

0 0

Post a comment