On Phobos & Deimos composition:
Believe it or not, we do know chemical and mineral composition of Phobos quite well, because we have
sample in hand. Yes, with Phobos-Grunt failed, nevertheless, we have 2.5 lbs of Phobos material, as meteorite called KAIDUN.
This meteorite fell some 35 years ago in Yemen right on the territory of Soviet military base, it was lucky from the very beginning. Because nobody believed it was natural, they thought it was a failed NATO missile or, worse, some spy device. Therefore, it was collected, preserved, and studied in best possible way with huge federal funding. Seriously, it was transferred to Moscow within hours of impact via urgently dispatched supersonic jet (the only case in the entire history of Russian meteorite studies).
It was preliminary classified as “very strange chondrite”. Which was lucky again – the officials concluded it was a spy thing developed by CIA and smartly disguised by NASA as an innocent meteorite. See, it fell too close to some object X, which never was in Yemen, moreover, Soviet Union never produced it. So, the funding for studying did not cease but tripled, and the rock was investigated millimeter by millimeter literally, with best instruments available. (This is a true story from the first hand – the PI for Kaidun study was A.V.Ivanov, and I was his grad student only 5 years later). No CIA microchips nor cameras were found, but Kaidun become the best studied meteorite in Russian collections. And it still keeps this record, lucky guy.
Later, it was classified as “very strange carbonaceous chondrite”, and finally, as “very strange CR chondrite”.
Now, the main part:
In 2003, at Lunar Conference in Houston, Andrei Ivanov and Michael Zolensky suggested their explanation for strange nature of Kaidun:
http://www.lpi.usra.edu/meetings/lpsc2003/pdf/1236.pdf (two last paragraphs in the abstract)
They concluded that it originates from the closest Mars moon, Phobos. I was there, in the conference room, and I can tell you that the general reaction was severe skepticism plus some fierce (and loud) sarcasm. I proudly note that I was not among critics; I liked the idea from the beginning. Not because Andrei was my dissertation advisor, I just liked the idea, although at the time it did sound crazy.
In following 10 years this "crazy" idea got several confirmations from different sides independently. The very last confirming data came just months ago – from Curiosity, and they are strong enough for me to call it "proof".
It is an interesting chain of reasoning (I mean, how they concluded that Kaidun = Phobos), but it is far from this thread, so I just go to
COMPOSITION OF PHOBOS.
FIRST, it has low density. It is so low, that some people suggested it is not a rock, but some artificial object (alien ship). Well, unfortunately it is not. Then, it was suggested, that low density of Phobos is due to large water ice core. Data from MRO and Mars Express do NOT confirm this, spectral results of all types in good agreement show no water ice, no crystallization water, not any traces of OH-groups on Phobos surface. In full agreement, Kaidun data strongly suggest that Phobos is very dry – on the surface AND below. (In Kaidun papers you can see phyllosilicates frequently mentioned – trust me, these are minor amounts). Low density of Phobos is most likely due to high porosity (micro-porosity), which is in good agreement with recent data on outer (captured) Jupiter moons. So, most likely Phobos has no water in any form, although you can read the opposite, especially in old articles - disregard these.
SECOND thing, which is also sad: Among all carbonaceous chondrites, CR-chondrites are on the very low end in carbon content (graphite is accessory mineral in CRs). So, Phobos is an unlikely source for carbon too.
THIRD, Kaidun consists mainly of chondrules (65-70% by volume) – little molten droplets of primordial matter in Solar Nebula, which were formed by high voltage discharges (static or electromagnetic) in protoplanetary disc, and these discharges were high-T too, at least 1500 K. So that these guys, chondrules, are depleted in everything, which boils below.
FINALLY, mining expectations for volatiles on Phobos are close to zero. But, maybe it’s a good place to look for iridium, ruthenium, or tungsten, who knows.
Now,
COMPOSITION OF DEIMOS:
We have no Deimos samples (yet), but we do have pretty good spectral data. They are actually the best in reproducibility, as I heard. They do show small water lines, too low for practical extraction, although hope remains for higher water content below surface. The good thing, there are some distinct carbon spectral lines, so there is a good chance for significant graphite in Deimos regolith.
As whole, spectral data place Deimos really close in composition to major carbonaceous chondrite types, CM and CO, with better confidence than typically for asteroids.