Martian Geochemistry 

Recent radar imaging has suggested large amounts of water lie beneath the surface particularly towards the poles.
Note cyclonic cloud formation.

The surface of our neighbouring planet Mars is partly hidden by thin dry dust deposits sometimes whirled about by high velocity wind creating obscuring dust clouds. At other times all is calm, but fine surface detail is often hidden from telescopes and orbiting satellites by the dust layer. The number and size of meteorite impact craters is distinctly less than on the moon, perhaps the combined mass of Earth and Moon has attracted more meteorites. There are no sign of island arcs, trenches, or subduction zones so we would be very surprised to find andesites, rhyolites or granites. Nor, apart from a small magnetic zone is there any sign of mid-ocean-type spreading centres, so we would not expect MORB type basalts.

There are some major faults and rifts and a few volcanoes of basaltic appearance, some of which are surprisingly large. Olympus Mons is 550km in diameter and 25 km high with half a dozen overlapping calderas at the summit, and one or two flank parasitic cones. The base of Olympus Mons is surrounded by a kind of collar which might be expected to be the result of subsidence into the crust of such a great mass as is found around Hawaii. However, some commentators describe it (and 3D projections show it) as a basalt cliff where the surrounding regolith has dropped away. Looking at the view of Olympus Mons with the sun on the left, it can be seen the the circumferential "collar" is a depression with a "squeeze-up" ridge on both sides. Other low angle Martian Orbiter photos do show definite high cliffs on one side and give the impression that the whole massif has tilted. Olympus Mons may be situated over a "Mantle Plume" and with no mantle circulation has remained in one place, so that instead of dozens of basaltic volcaoes being created in a chain as seen in Hawaii-Emperor Seamounts, we see a single monumental mass, much as though all the Hawaiian shield were combined. Possibly the uprising stationary "plume" explains the uplift.

Olympus Mons As Mars may be formed of impacted chondrite material one might assume a bulk composition much like the earth’s mantle prior to the partial melting of the oceanic and continental crusts. The Martian mantle should thus be close to the surface with perhaps a thin layer of enriched basalt and local volcanic cones over "hot spots" or deep-seated fractures of even more enriched type.


The Martian Lander was able to carry out some analytical work including gases and isotopes and this has allowed a number of meteorites found on Earth to be identified as originating on Mars. A total of 30 such, some found in stagnant and ablating sectors of the Antarctic ice-sheet and some in the desert of India, Oman, Arabia and Morocco and some seen to fall in Florida and in France, all have some characteristics in common and which set them apart from both terrestrial and lunar samples and include trapped gases comparable with Martian atmosphere.

The chemical data for these meteorites has recently been assembled by Dr Charles Meyer of NASA in the Mars Meteorite Compendium. The literature referred to is enormous and serious readers should investigate this source. Other data occur in the "MetBase" database.

The practical geochemist who is interested in ties between Martian, Lunar and terrestrial compositions will be both delighted that so much is already available but also disappointed. The latter because the Martian meteorites are small and up to a dozen investigators have been given at most a few hundred mgm of often uncrushed sample. Because of the coarse grainsize, interlaboratory difference are disappointingly large, differences of as great as 100% may be found for some elements. Intersample agreement in general is not as good as one might like.

Typical Martian scene, with sand-polished fine-grained basalt venifacts.
A cold desert, T about -80ºC
Probably a large number of the rocks seen are meteorites.
Photo: NASA.

The majority of the Martian meteorites are ultrabasic ferro-dunites, ferro-clinopyroxenites, ferro-orthpyroxenites, and lherzolites with a few ferro-basalts, the FeO content of all samples lying between 17 and 20%. At least, they have been termed "dunites" but terrestrial dunites have close to 6-7% total iron with 45 – 48 or even 50% MgO, whereas these Martian rocks have 17 – 21% FeOT, only 30% MgO and have a very iron-rich olivine, often about Fo65-70 instead of Fo 90-91, and very low Ni, often < 100 ppm cf 2500-3000 ppm in our earthly dunites. Were they found on Earth we might term them "Ferro-picrites". Both CaO and Al2O3 approach zero levels at only 32.5% MgO instead of the 50% MgO seen on Earth. At the same time the Cr is remarkably high, as much as 7000 ppm. This has never been found in a terrestrial ferro-pyroxenite with 20% FeOT and 28% MgO except in chromitite bands in much more basic rocks.
While one or two of the "basaltic" Shergottites (named for a village in India where the original was found) are somewhat similar to terrestrial ferro-basalt, yet there are significant differences. The P is very high, well beyond the range of terrestrial rocks but Ti very low (<1%) whereas in terrestrial ferro-basalts it would be at least 4-5%. No massive titanomagnetite is seen, the iron is all in the olivine, orthopyroxene, and clinopyroxene.
The Zr/Nb varies from 30, (indicating tholeiitic type) down to an alkaline 6.5. Plagioclase is always present in low amount, and always as maskelynite, a high pressure variety formed under shock pressure.
The Zr/Hf at 34.5 is lower than seen in good data for fresh terrestrial basaltic rocks where it averages 38-40. Nb/Ta is high at 20 though many samples show scatter. La/Lu is highly variable.
The REE show three main groups, one with high LREE at about 10 times terrestrial mantle, but lower HREE, a group with markedly depleted LREE of pyroxene type, and a very low, even distribution somewhat lower than chondrites, for a high olivine sample. The REE also show that the Shergottites are remarkably LREE depleted, more so than NMORBs.

|__| Ree diagram for Martian achondrite meteorites.
|__| Fingerprint diagram for the above.
The above are shown in the chapter on "Meteorites"

Martian Pathfinder Data

NASA have now released the data obtained by the Mars Pathfinder Alpha Proton X-Ray spectrometer aboard the tiny crawler tractor that "went ashore" and pushed its detectors into a series of littered rocks of vesicular basaltic boulder appearance. It was announced some years ago that the rocks, given names like "Yogi" and "Barnacle Bill" and "Scooby Doo" were andesites which seemed highly unlikely. Impressive as the feat of analysing a rock several million miles away and transmitting the data back to Earth is, we are not here to praise technology but to take a critical look at the results. Considerable changes have taken place since the first "preliminary" data was released and estimated uncertainties eg, for silica of 52% +- 2.5% are on a par with those for the "SNC" meteorites, where for one sample has been assigned silica values of 47.06, 43.5, 44.3, 48 and 47.02% by different laboratories. Pedantic we may be but this leaves one feeling a little unhappy. However no one has ever claimed this was a perfect world.

However the Pathfinder data does confirm that the so-called SNC (for the three main types, Shergottite, Nakhla, Chassigny etc) meteorites are extremely similar even though ultrabasic rather than basic while the ones analysed in situ are ferro-basalt to ferro-icelandite. There is the same high iron in both rocks and soils (13 – 17%), the alumina is still low, ( 7.4 to 10.8%, TiO2 is also low (0.8 – 1.3%) even though the silica is somewhat higher in the range of 48.2 – 58.6%

If found on Earth we might term these rocks ferrobasalts or ferro Icelandites but find them puzzling as the higher silica rocks do not have higher TiO2, while the CaO is actually LOWER than that found in the ultrabasic rocks and there is no change as silica increases in the rocks-soils. We have recently been working on averages of about 20,000 analysed orogenic andesites and though some may share a normative andesine feldspar, no terrestrial orogenic andesite has such high iron, MnO, or such low alumina with an MgO level of 3.0 – 8.3%. The six "soil samples should be little changed from the original rocks, but could be an average from the whole planet, in fact the soils are all more magnesian than the three rocks implying an ultrabasic contribution.

The data may not be precise enough to allow computer simulation of possible derivation of the ferro-basalts by fractionation from the ferro-ultrabasic meteorite-type rock.

We can only conclude that the Martian rocks, while in general are somewhat similar to those of earth, the processes to which they have been subject to over the last 3.8 billion years have been quite different, with a slow static cooling and expulsion of a little residual magma vertically. It is dollars to doughnuts that Olympus Mons has become more alkaline and enriched with time, but the greater mass or shield stage may be close to Hawaiian type.


REE Diagram for Martian SNC meteorites normalised to Terrestrial EMORB.
The data are very incomplete, the majority of the REE determinations being for 4-6 elements only. By eliminating most of those where two or more adjacent elements are missing, and by interpolating all Pm and other single missing elements, we are able to get at least an approximate picture.
The bulk of the samples are Shergottites, which probably represent the bulk of the Martian unmodified surface, a high iron peridotite with an LILE content ranging from roughly EMORB to NMORB though the shape of the REE distribution is not one commonly seen on Earth, being rather more domed.
Two other trends both with elevated relative LREE are seen. The three samples with the lowest HREE at about 0.04 EMORB with La = x .07-x .08 EMORB are Chassignites.
The other group with Lu at 0.08 Emorb and La at ~0.4 EMORB are the Nakhlites.
The only process definitely known to change the La/Lu ratio is differing degrees of partial melting. We might guess that the Chassignites and Nakhlites are derived from Martian basaltic volcanoes.

Chassignites & Nakhlites

Here the Martian Chassignites and Nakhlites are shown separately from the Shergottites. The data is very patchy with many unfortunate gaps, but with some interpolation we can see the most depleted Chassignites which show elevation of the LREE suggesting an origin due to partial melting. The Nakhlites including Lafayett are the most enriched with even more elevated LREE but still on average, well below terrestrial EMORB.
A sample also labelled ALH 84001 is included. It is a type of Shergottite, but is reported as being an orthopyroxenite. Only a single complete analysis is available. Other types may be included in the Shergottites but data are too incomplete at this point to differentiate.


As with the Lunar samples it would be a rash scientist who would pass judgment on a score of samples, they may well prove to be quite atypical. However, for a planet apparently lacking any continents, subduction or orogeny, it is perhaps not surprising that the samples are so basic, as the meteorite impact causing the ejection of the samples should only effect a few hundred feet of the surface at the most. The moon which also lacks continents, spreading centres and subduction zones appears to have fractionated extensively during cooling forming massive anorthosites, troctolites and high Ti ferrobasalt. Why are no anorthosites on Mars, as even with the dust layer they should be visible?

The smaller mass of Mars compared to the combined mass of Moon + Earth, might mean that incoming velocities of accreting meteors were lower and the initial planetary body much cooler. A mass of chondritic composition only partially molten, might cool with the lower temperature ferriferous differentiates working their way towards the surface with only a minimum of local volcanism. This does not explain the low ages obtained for some of the meteorites of about 175 myr, nor does it explain the high Cr content. The REE average about chondrite x 10 and the La/Lu and Lile/HFSE ratios differ in different meteorites so there are differences in degree of melt at least. And why should the greatest known volcano occur in such a setting? The heat source must be remarkable.

It will be many years before a deep drilling program will be carried out on Mars, but we can be sure (if our civilisation survives) that it will occur!

To see the Martian orbital path relative to the earth, see under "Meteorites".


Copyright © Dr B.M.Gunn 1998-2006

The Moon