NAKHLA

In the village of El Nakhla el Baharia in the Nile Delta of Egypt, at 9:00 in the morning, numerous stones having a combined weight of ~10 kg fell over an area of 4.5 km. The friable stones varied in weight from 20 g to 1813 g, and most had complete or partial glossy-black fusion crust covering the greenish-gray crystalline interior. The minimum pre-atmospheric diameter of the meteoroid was calculated to have been 44 cm, and its low mass ablation is consistent with a very high entry angle close to verticle.

The green cumulate crystals found in Nakhla are composed largely of the high-Ca clinopyroxene augite (76–85 vol%) with lesser amounts of Fe-rich olivine (6–15 vol%), together with plagioclase, K-feldspar (~1%), Fe-Ti oxides, pyrrhotite, pyrite, chalcopyrite, gypsum, a reddish-brown aqueous alteration phase similar to iddingsite (consisting of smectite clays, ferrihydrides, and iron oxides), and a halite–siderite–anhydrite–chlorapatite assemblage derived from evaporite deposits and incorporated into silicate melt. In addition, a halite–clay assemblage was identified that was probably formed by percolating fluids (Rost et al., 2005). The water-soluble ions of Cl, K, Na, S, C, Ca, and some others that are found within Nakhla match those that would be expected to precipitate through the low-temperature evaporation of an acidic brine.

At the center of olivine fractures in all nakhlites are secondary mineral assemblages in the forms of siderite, clay-like phases, and an amorphous hydrated Fe–Mg–Al silicate gel (Changela and Bridges, 2010). These alteration assemblages were produced in the highest abundances at the bottom of the cumulate pile as a result of impact-induced heating, melting of permafrost, rapid progressive cooling, and oxidation. Thereafter, upward percolation of the hydrothermal fluids through fractures led to fractionation of the secondary mineral assemblages, and the fluids ultimately evaporated from the surface.

Interestingly, Mn carbonates were also precipitated during brine evaporation, a process which on Earth, and perhaps on Mars, is microbially mediated (Bailey et al., 2003). A new weathering product was recently discovered within olivine melt inclusions in Nakhla (Rost et al., 2003). It is a brown microcrystalline phase, thought to be associated with precursor amphibole. Water contained within hydrous minerals accounts for over 0.11 wt% of Nakhla. Based on these findings, it is a reasonable assumption that Nakhla resided in a locale, possibly a crater, which enclosed a hydrothermal liquid.

In addition to Nakhla, seven other nakhlites have been discovered—Lafayette, Governador Valadares, NWA 817, Y-000593/749/802, NWA 998, MIL 03346, and NWA 5790. Most of these nakhlites have undergone detailed analyses which indicate that they all share similar mineralogies and petrologies, as well as crystallization and cosmic-ray exposure ages (~1.3 b.y. and ~12 m.y., respectively)(Misawa et al., 2003; Park et al., 2008). Correction of the Ar–Ar crystallization age of Nakhla and NWA 998 was accomplished by Cartwright et al. (2010) by accounting for a trapped component of 40Ar. The resulting ages are consistent with the ages determined from other chronometers. In addition, aqueous alteration products from Y-000593 and Lafayette have similar Rb–Sr ages of ~650 m.y. These age correlations infer that the many nakhlites are probably source crater paired. However, some important compositional differences do exist among them.

For example, a pre-terrestrial, reddish alteration product present in NWA 817 has been identified as a water-bearing ferromagnesian phyllosilicate of the smectite group, different from the iddingsite-like alteration phase present in Nakhla (Gillet et al., 2002). Multiple lines of evidence suggest that this smectite-like alteration product formed by the percolation of a sequestered, mantle-derived, aqueous fluid through crustal rock. This hypothesis is supported by the unfractionated H isotope value (very low D/H ratio) of the alteration phase, which suggests that the alteration fluid had not been equilibrated with the martian atmosphere, in contrast to the water from other nakhlites. The enrichment of Fe and LREE in the alteration phase of NWA 817 is attributed to the infiltration of the alteration fluid through country rock, incorporating a dissolved plagioclase component. Interestingly, NWA 5790 contains the highest Th, U and REE concentrations reported thus far in a nakhlite, exhibiting a strong LREE enrichment.

Also of interest is the fact that the augite component crystallized from a compositionally distinct melt source than that from which the olivine and mesostasis were derived (Sautter et al., 2002). The mesostasis in NWA 5790 is present in the highest proportion (39.7%), while NWA 817 contains the next highest proportion (~20%) compared to lesser abundances in the other nakhlites (e.g., 10% in Nakhla and Y-000593). Northwest Africa 817 contains an Fe-rich feldspar component indicative of rapid crystallization following undercooling conditions. MIL 03346 contains the lowest abundance of olivine, and has pyroxenes that are strongly zoned with hedenbergite. For many other features, e.g., highly zoned olivines (Fa59–86 in NWA 817; Fa56–91 in MIL), olivine alteration products, and skeletal Ti-magnetite in the mesostasis, NWA 817 and MIL 03346 show very close similarities; the less well studied NWA 5790 appears to have many close similarities as well.

Notably, MIL 03346 contains melt inclusions within clinopyroxene with a trapped chlorine-rich fluid and its reaction product of potassic-chlorohastingsite (McCubbin et al., 2009). Successive oxidation of this fluid and its reaction with pyrrhotite led to acidification and subsequent formation of jarosite and hematite upon cooling below ~200°C.

Preliminary mineralogical analysis of NWA 998 indicates that it is an unusual, orthopyroxene-bearing nakhlite, which was permeated by deuterium-rich fluids on Mars. The three paired nakhlites from Antarctica, Y-000593, have a combined weight of ~15 kg, which constitutes the largest nakhlite found to date. The composition of the mesostasis in Y-000593 suggests a faster cooling rate than that of Nakhla, GV, and Lafayette, but slightly slower than that of NWA 817.

The fact that NWA 5790 is the sole member of the nakhlite group that has retained complex primary zoning features on augite grains, indicates it cooled the fastest and formed at the top of the cumulate pile. Further evidence for rapid cooling at this topmost location can be seen in the Cl contamination previously observed in MIL 03346, the high proportion of evolved mesostasis, and the lack of equilibrium among mineral phases. Based on texture and mineralogy (mesostasis [REE] abundance, plagioclase size, olivine composition, intercumulus porosity, and pyroxene composition), as well as Fe–Mg and Ca zoning profiles in olivine and pyroxene (reflecting cooling rates), and on crystal size distribution analysis, a comparative burial depth within the cumulate pile (lava flow or magma chamber) for each of these nakhlites has been proposed (Mikouchi et al., 2003, 2005, 2006; Lentz and McSween, 2003; Macrì et al., 2004; Imae et al., 2005; Day et al., 2006; Treiman and Irving, 2008; Jambon et al., 2010):

1. NWA 5790—top of cumulate pile
2. MIL 03346—near top; 1-2 m
3. NWA 817—near top; ~1-2 m
4. Y-000593/749/802—7 m
5. Nakhla—~10 m
6. Governador Valadares—~10 m
7. Lafayette—>30 m
8. NWA 998—bottom of the cumulate pile; >30 m

A study which proposes a different stratigraphic ordering of the nakhlites has been published by Grady et al. (2007). Without regard to the later recovery of the nakhlite NWA 5790, they argued that although MIL 03346 contains the least equilibrated olivine cores and has experienced the most rapid cooling of all the known nakhlites, both factors which are consistent with a shallow emplacement, it also contains the lowest abundance of carbonates with intermediate 13C isotopic composition, factors more consistent with minimal aqueous alteration at greater depth. Therefore, they envisage MIL 03346 as forming at the lowest zone of the flow and being cooled by circulating groundwater derived from melted ices. The source of the isotopically-light carbonates is considered to be from this groundwater.

In a contrasting view, Lentz et al (2005) find a lack of correlation between the olivine texture/distribution and the stratigraphic order, but rather, propose that the olivine texture/distribution is related to formation in different lava flows of variable composition. They suggest that at least two flows generated the various nakhlites—one which produced NWA 817, Nakhla, and Governador Valadares, and another which produced Y-000593/749/802, Lafayette, and NWA 998. A third flow may have generated MIL 03346. To account for the postulated gabbroic layer, which should have overlain the nakhlite unit and slowed its rapid cooling rate, they proposed that these rapidly cooled nakhlites were extruded late in evolved lava flows, but before plagioclase formation began.

There are several competing scenarios to explain the formation of nakhlites. Some scenarios places their origin in a plutonic environment, while others place it close to the surface in a thick lava flow or shallow intrusion. Following are details of three of the major competing scenarios consistent with most nakhlites.

Scenario 1

1. the petrogenesis of Nakhla began with the formation of pyroxene-rich cumulates during an early Mars differentiation episode ~4.56 b.y. ago.
2. a late magmatic event ~1.3 b.y. ago resulted in the partial melting of the cumulate mantle.
3. Fe-poor pyroxenes and Fe-rich olivines were formed; rapid cooling prevented the exsolution of Ca-poor pyroxene.
4. intercumulus melt inclusions were entrapped within olivine crystals; this basaltic melt was rich in Fe and Ca, poor in Al and Ti, enriched in K relative to Na, and similar to alkaline basalts on Earth and martian rocks analyzed by the Spirit rover in Gusev crater.
5. the rapid growth of augite and olivine was followed by a period of slow cooling, gravitational settling, and equilibration within a plutonic environment, forming a cumulus-textured solid.
6. continued percolation of highly oxidizing, intercumulus magma resulted in late-magmatic re-equilibration.
7. a final slow cooling phase resulted in the formation of exsolution features; this final solidification possibly occurred nearer to the surface, perhaps in a lava flow or sill.
8. infiltration of liquid water produced iddingsite.

Scenario 2 (Imae et al., 2005)

1. accumulation of phenocrysts of the Fe-poor pyroxene augite occurred in a plutonic environment at depths commensurate with pressures of at least 3.0 GPa under high temperature, oxidizing, slow cooling conditions.
2. lesser amounts of olivine phenocrysts (~74% of these being of cumulate origin, the remainder having crystallized from the intercumulus melt) and cumulate titanomagnetite micro-phenocrysts were formed.
3. magmatic inclusions were trapped within both augite and olivine phenocryst cores; symplectites possibly formed at this time.
4. entrainment and intrusion of the cumulate material to the surface produced a thick lava flow or sill.
5. under rapid cooling conditions, plagioclase and other accessory minerals crystallized from intercumulus melt to form the mesostasis and the inner rims on augite phenocrysts.
6. Ca-depleted pyroxene then nucleated within the mesostasis, and also formed the outer rim on augite phenocrysts.
7. a late-stage, slow cooling phase set up the following formation sequence: 1) augite—magnetite aggregates, 2) symplectic intergrowths at olivine—augite boundaries, and 3) the formation of exsolution features in olivine cores.
8. infiltration of liquid water produced iddingsite.

Scenario 3

This scenario utilizes the 2.7 b.y. old, differentiated surface lava flow in Ontario, Canada, known as Theo's Flow, as an analog for the nakhlites. In particular, the differentiated, 60 m thick pyroxenite layer is so similar to the nakhlites that it is proposed they share the same formation processes.

1. a very mafic magma flow tens of meters thick began to pool.
2. pyroxene nucleation began in a cooler zone underneath a quenched crustal layer and began to link into clusters.
3. these pyroxene clusters sank through the low-viscosity magma only to be carried back up in convection currents.
4. during the few days the pyroxene clusters were in this convective cycle, each cycle lasting for a few hours, they grew larger each time they reached the cooler nucleation zone.
5. when the clusters became too heavy for convective forces, they settled out onto the cumulus pile below where the grains developed a preferred orientation.
6. finally, plagioclase crystallized from trapped low-Al melt, along with a possible unsampled gabbroic layer.
7. infiltration of liquid water produced iddingsite.

Nakhla was subsequently ejected from Mars 10.8 (±0.8) m.y. ago, along with the other nakhlites, from a crater over 100 km wide. The single olivine-rich chassignite, with its similar crystallization age of ~1.3 b.y. and similar cosmic ray exposure age of 11.1 (±1.6) m.y., may have been ejected during this same impact event. Nakhla was only weakly shocked during its ejection and experienced an estimated peak pressure of 14–20 GPa, among the lowest of the martian suite. Fritz et al. (2005) revealed that a correlation exists between the Mars-to-Earth transfer time and the shock stage of the material; i.e., fragments having a higher degree of shock also have a faster transit to an Earth-crossing orbit and vice versa. Therefore, the absence in our collections of highly shocked nakhlites may be reconciled by consideration of their short lifetimes on Earth (less than ~1 m.y.).

There is a legend, perpetuated since the Nakhla fall, regarding the death of a dog that was hit by one of the falling stones. This story was first reported in the Arabic Newspaper "El Ahali", but the account given can be discounted based on several points. The fall took place on June 28, not on the 29th as the impact witness stated. The reported village of Denshal is actually about 33 km southwest of the 5 km-long strewnfield in El Nakhla el Baharia. Denshal probably experienced sonic booms, but no other witnesses observed a meteor or any meteorites from the area. The main evidence, the dog, was never produced. Lacking any compelling evidence to the contrary, it may be presumed that this story of the killed dog is apocryphal, and can finally be laid to rest.

The International Quarterly, Meteorite!, published a two-part article in 1998, Vol. 4, Nos. 2 and 3, in which Kevin Kichinka presents an exhaustive review of the record concerning many aspects of the Nakhla meteorite. By permission, his article is presented here in its entirety.

Recently, SEM and light microscope images were taken from interior chips of a previously intact, 640.8 g, 100% fusion-crusted individual donated by the British Museum. The mass was divided under sterile conditions at NASA–JSC, Houston, with about half of this distributed to 37 groups of investigators. Images taken by the team of Dr. David McKay reveal a network of cracks and veins filled with a brownish, low-temperature alteration product similar to smectite, which is thought to have formed no earlier than ~650 m.y. ago. Embedded within this phase is an iron-rich band with µm-sized, rounded particles that are theorized to be the mineralized products of bacteria similar to those of terrestrial iron-reducing bacteria. The images below show these possible fossilized martian microbial cells embedded in 700 m.y. old clay minerals within Nakhla (courtesy NASA–JSC):

Evidence supporting this claim of fossilized Martian microbes includes the following:

• The diameter of these particles, 0.2–2 µm, is within the size range of most terrestrial bacterial cells.
• The cells have a complex surface structure at high magnification resembling that of some terrestrial mineralized bacterial cells.
• Some cells are joined in a manner similar to dividing bacteria, while one is observed to have a fibril-like appendage.
• Material coating the surface of some cells resembles mineralized biofilm common to bacteria.
• The uneven distribution of these particles is consistent with that of other bacterial colonies.

However, in another study (PNAS 1999, vol. 96), researchers suggest that the amino acids detected in Nakhla have a D/L ratio similar to that found in association with terrestrial bacterial activity, carrying a fingerprint matching that of Nile River Delta sediments. New samples of Nakhla constituting a depth profile were analyzed utilizing ToF-SIMS, SEM, EDX, and culturing techniques (Toporski et al., 2000; Toporski and Steele, 2004). They revealed a significant degree of microbial contamination in all samples studied, including the presence of spores and hyphae from several filamentous bacterial and fungal species. Identification of these various species using genetic techniques such as PCR amplification of DNA is ongoing. The implication of rapid terrestrial contamination may have repercussions for future Mars sampling missions.

In a recent analysis of a well-preserved Nakhla stone, a new optically dark, carbon-rich component, lacking a terrestrial isotopic signature, was discovered filling fractures within (Gibson et al., 2006; McKay et al., 2006). This carbonaceous phase is associated with the secondary replacement material iddingsite, and it occurs in several forms (sub-µm- to µm-sized) including vein fillings, dendritic tubes connected to veins or intermingled with the host silicate within cavities, and as tiny blebs. It is proposed that some of the mineral corrosion that is observed is a result of microbe-produced organic acids. An isotopic analysis has been conducted on this reduced, C-bearing phase, and the composition was identified as 13C from an extraterrestrial source. The meteorite is currently under study to distinguish the source of this phase from among the likely candidates; e.g, delivery through impact of a carbonaceous impactor, or as a product of biogenic activity.

In a related study of a well preserved Nakhla specimen, conducted at Oregon State University (Fisk et al., 2006), tunnels measuring ~10 µm long were discovered, comparable to tracks formed within terrestrial igneous rocks and basaltic glasses from extreme environments by ~1 µm-sized iron-trophic bacteria. They also found elemental associations similar to some terrestrial microfossils. Their findings have raised the possibility that martian bacteria were extant 600 m.y. ago during a period when Nakhla was exposed to an aqueous environment.

Tunnels are resolved within fractures in a Nakhla sample (in boxes).
Photo courtesy of Oregon State University

Nasa's astobiology group examined pristine Nakhla samples and other nakhlites with scanning electron microscopy techniques and have identified additional potential martian biogenic fossil evidence (Spaceflight Now; C. Covault, 2010). Dating of these features show that they were formed during a warmer and wetter period on Mars 1.4 b.y. ago. These features are virtually identical to features found in the nakhlite Y-000593 and the martian orthopyroxenite ALH 84001, as well as to micro-fossils found in some terrestrial basalts. More advanced techniques including an ion microprobe with spectrographic analysis capability will be utilized in future studies to better resolve these micro-fossils. The three NASA photos shown below demonstrate what is believed by some to be the remnants of ancient martian bacterial mats and other biological signatures:

The nakhlite group presently comprises Nakhla, Lafayette, Governador Valadares, NWA 817, Y-000593/749/802, NWA 998, MIL 03346, and NWA 5790. The orthopyroxenite ALH 84001 was characterized by the Planetary Chemistry Laboratory at Washington University as a subgroup of the nakhlites. The specimen of Nakhla pictured at the top of this page is a 1.9 g fragment containing a patch of reddish-brown material (seen in the upper center of the image), probably an aqueous alteration product similar to iddingsite or smectite (and martian microbes?).

---