One or more stones fell after sonic booms were heard around 8:00 A.M. near the village of Chassigny, France. Many small fragments were immediately recovered around a small impact hole. The total recovered weight of this first martian meteorite is estimated to be about 4 kg, although only about 800 g is accounted for today. Chassigny, NWA 2737, and NWA 8694 (photo courtesy of L. Labenne) constitute the only known samples of martian dunite.
Chassigny consists of 90 vol% Mg-rich olivine (Fo68), resembling a terrestrial cumulate dunite (although more FeO-rich), with the remaining constituents comprising the pyroxenes pigeonite and augite (5%), plagioclase feldspar (2%), and chromite (1.4%), along with minor pyrite and both fluorapatite (in melt inclusions) and chlor-fluorapatite (in interstitial maskelynite). Chassigny contains noble gases different from the martian atmosphere, and is presumed to originate from the martian mantle.
The Cl-enrichment of the interstitial material is thought to have developed as a result of the upward migration of a high-temperature, Cl-rich fluid, which then lost Cl and gained water as it percolated through the cumulus pile along interstitial pathways (McCubbin et al., 2007). McCubbin and Lindsley (2006) inferred from the structural formula of apatite that the water content was likely very low (<0.4 wt%) during crystallization of Chassigny; apatites from basaltic shergottites are more H2O-rich. In a contrasting study, McCubbin et al. (2009) found that kaersutite and Ti-biotite in Chassigny melt inclusions contain higher abundances of water than previously measured, a value which correlates to a parental source magma water content between ~460 and 840 ppm (0.50.8 wt%), while lower abundances of Cl and F were observed.
In support of the conclusion that martian parental magmas were water-poor, results of analyses by Filiberto and Treiman (2009) of both apatite and kaersutite in martian meteorites are consistent with a parental melt volatile content that was Cl- and S-rich and water-poor, properties which would affect mineral compositions and crystallization temperatures similar to the way in which a hydrous parental magma would. They cite the Cl-rich surface rocks revealed in analyses by the Mars Exploration Rovers in support of their theory. Employing volatile partitioning models, Giesting and Filiberto (2013) calculated the halogen content of the magma that was precursor to the two chassignites based on the Cl/Fl ratios in existing hydrous minerals (apatite) and in magmatic melt inclusions hosted by olivine (apatite, kaersutite, and biotite). A crystallization sequence of kaesutite ⇒ apatite was found to be consitent with the data, while biotite is thought likely to have crystallized from a late-stage, halogen-rich metasomatic fluid, the latter also considered a factor in the composition of ferromagnesian minerals and an influence upon the acidic nature of the martian environment.
Three different types of glass-bearing inclusions (up to 0.2 mm) are present in Chassigny: 1) multi-crystalline or multiphase (glass + several mineral inclusions); 2) crystal + glass or monocrystal (glass + a single mineral inclusion); and 3) pure glass (Monkawa et al., 2003; Varela and Zinner, 2015). Multiphase inclusions typically contain low- and high-Ca pyroxene, feldspar- and Si-rich glasses, sometimes low-H, Ti-rich kaersutite, and more rarely, chlorapatite, FeS, chromite, anorthite and albite. Mineralogical evidence suggests that these inclusions could have formed in the martian mantle under more reducing conditions than those of the Chassigny parent magma (Monkawa et al., 2006). Similar magmatic inclusions are present in nakhlites, shergottites, and lherzolites, which all typically contain AlTi augite. Monkawa et al. (2003) argue that both the reverse zoning present in some augites in Chassigny inclusions, and the absence of Ti-rich phases other than kaersutite, provide evidence for a late impact-shock event on Chassigny. This impact created more reducing conditions (causing reverse zoning) and higher temperatures and pressures (causing Ti-rich phases other than kaersutite to melt). Rapid cooling of the magmatic inclusions ensued.
In another study involving the melt inclusions in Chassigny, Filiberto et al. (2004) concluded that formation of Chassigny is more consistent with fractionation from an alkalic, silica-saturated parental liquid under elevated pressure, rather than from a low-pressure fractionation of an olivine tholeiite. In support of this scenario, it was experimentally demonstrated by Nekvasil et al. (2007) that a magma compositionally analogous to terrestrial tholeiite could be the parental liquid of the chassignites. They found that a tholeiitic magma can undergo crystallization at >4.3 kbar (equivalent to martian depths of ~35 km) and >0.4 wt% water to produce an alkalic, silica-saturated hawaiitic melt. This melt can evolve to form the polyphase melt inclusions in Chassigny olivines, including the highly evolved alkali-rich rhyolitic glass component. By a similar process, but constituting a less evolved, mildly alkalic basalt liquid, the trapped melt component found in the chassignite NWA 2737 could be formed. Crystallization at this significant depth would be followed by ascent to a near-surface location where impact excavation can occur.
Martian tholeiites have been identified by NASA's Mars Exploration Rover Spirit, examples of which are the rocks Adirondack, Mazatzal, and Humphrey; Spirit has also identified several hawaiitic rocks, including the silica-saturated Backstay, Wishstone, and Irvine. In a study by Elardo et al. (2008) the martian hawaiitic Backstay rock, which was studied by the MER Spirit at Columbia Hills of Gusev crater, was utilized as a model for the Chassigny parental melt. It was demonstrated that the accepted crystallization sequence for Chassigny is consistent with a parental magma water content of between 1.5 wt% and 2.6 wt% (Nekvasil et al., 2009). Models utilizing a Backstay-like rock as a parental melt were utilized to reveal similar phases known to exist in Chassigny, both cumulus phases and melt inclusions, would be produced during crystallization and loss of residual liquids at the base of a 5070 km thick crust (6.89.3 kbar) given a bulk water content of 2.6 wt% (1.5 wt% Cl, 0.3 wt% S, 0.4 wt% water; Ustunisik and Nekvasil, 2010).
Other experiments were conducted to determine the actual composition of the parental magma of the chassignites (Filiberto, 2008). They revealed that previous estimates of the parental magma composition based on the melt inclusions fail to produce some of the phases that are observed in the chassignites, and therefore this parental magma must instead have been more magnesium- and aluminum-rich than in previous estimates. In addition, a formation depth equivalent to ~9.3 kbar was found to be most reasonable, with the chassignite assemblages crystallizing only after 830% crystallization of mafic phases was achieved. This newly calculated composition for the chassignite parental liquid is similar in composition to the Gusev tholeiitic basalt rock named Humphrey, analyzed in situ on the martian surface.
Although it has been generally accepted that Chassigny and its glassy melt inclusions have a magmatic origin, studies by Varela et al. (2000) suggest that the composition of these melt inclusions is more consistent with having been trapped at sub-igneous temperatures concurrent with the formation of the olivine host. As the olivine precipitated from a chondritic fluid phase, the solid glass precursors became trapped within the crystallizing matrix, some serving as nucleation sites for other mineral phases. The non-equilibrium condition of the various components within polyphase inclusions supports such a non-igneous scenario (Varela et al., 2007). In still another study of trapped melt inclusions in Chassigny, Varela and Zinner (2015) argue that the non-homogeneous composition of the inclusions' glass, the chemical (both major and trace element) variability among the three types of inclusions, and the lack of equilibrium that exists both among crystalline phases within inclusions and between these crystalline phases and the host glass, are most consistent with open-system behavior. This significant variability is thought to be the result of post-entrapment modification. One such modification process likely involves metasomatic fluid conduction through radial fractures associated with the melt inclusions. Therefore, these melt inclusions might not represent the primary parental magma.
In a study of pyroxene lamellae widths, Monkawa et al. (2004) determined a cooling rate for Chassigny. The widths of augite exsolution lamellae of pigeonite in Chassigny was compared to those widths present in Zagami, which is considered to be from a thick lava flow or shallow dike. Chassigny is consistent with a cooling rate of 3543°C/yr, which is 45 times slower than that of Zagami, and corresponds to a burial depth of ~15 m. Furthermore, they studied the Ca zoning profiles in Chassigny olivine that resulted from atomic diffusion after cooling, and determined a cooling rate of 28°C/yr. This suggests a relatively shallow burial depth of ~15 m, consistent with that indicated by the pyroxene lamellae.
On a SmNd isochron plot reflecting the igneous crystallization age, Chassigny, the nakhlites, and the shergottites DaG 476, and QUE 94201 all show a linear match of ~1.4 b.y. The ArAr age of 1.4 b.y. calculated for Chassigny is consistent as well. Other similarities which suggest a petrogenetic link between Chassigny and the nakhlites include identical Sr isotopic ratios, similar whole rock REE abundance patterns that are LREE-enriched, and similar REE compositions of their respective parent melts. In support of this data, it was recently discovered that cumulus orthopyroxene in the form of enstatite is present in Chassigny, further evidence this meteorite was derived from a shergottite-like parent composition.
In a study by Fritz et al. (2005), they utilized numerical simulations (Artemieva and Ivanov, 2004) to determine the extent of the source region of martian meteorites. They found that oblique impacts are required to eject material, and that such material is confined to an area between 1 and 3 radii of the projectile to a depth of 0.2 radii. Mosaicism of olivine and maskelynization of feldspar, along with planar fractures and shear stresses are evidence of one and possibly two moderate shock events generating pressures of 2632 GPa and a post-shock temperature increase of 4060°C. A shock pressure of this magnitude leading to escape velocity (>5.4 km/s) would require that the target material be situated relatively close to the surface (e.g., within 20 m of the surface for a 200 m impactor). Several high-pressure phases, including the olivine high-pressure polymorph wadsleyite and ringwoodite, have been identified in melt pockets and along grain boundaries (Greshake and Fritz, 2009). Fritz et al. (2005) discovered 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. Therefore, the absence in our collections of certain highly shocked martian samples (such as nakhlites) might be reconciled by consideration of their short lifetimes on Earth.
The International Quarterly, Meteorite, published a two-part article in 2001, Vol. 7, Nos. 3 and 4, in which Kevin Kichinka presents an exhaustive review of the record concerning many aspects of the Chassigny meteorite. By permission, his article is presented here in its entirety.
As with the shergottites and nakhlites, Chassigny has a young crystallization age of 1.3 b.y. The cosmic-ray exposure age (and ejection age) based on Ar systematics of Chassigny (10.51 ±0.18 m.y.) is indistinguishable from that of Nakhla (10.35 ±0.06 m.y.) and the other nakhlites, supporting the likelihood that they all formed within a common cumulate pile. The chassignites crystallized first at the bottom of the magma chamber, and the nakhlites crystallized thereafter, some possibly following the eruption and emplacement of a lava flow (see the Nakhla page for further details on the stratographic sequence). The chassignites and nakhlites were launched toward Earth in a common ejection event. Recently, thermal emission spectrometry performed by the Mars Global Surveyor along with data from the Mars Odyssey THEMIS have led to the identification of craters with favorable characteristics in a few specific regions, particularly Nili Fossae in the Syrtis Major volcanic complex. Here there exists a basement basaltic unit that shows evidence for episodic occurrences of aqueous alteration, which is overlaid by an olivine-rich basalt unit containing Mg-rich olivine similar to that found in the chassignites (Amador and Bandfield, 2015). The presence of olivine in the Nili Fossae region is indicative of persistent dry conditions ever since the olivine was exposed through post-impact faulting probably over 3 b.y. ago (T. Hoefen, USGS, 2003). Importantly, in an adjacent region in eastern Syrtis Major, Fe-rich olivine and high-Ca pyroxene similar to nakhlite compositions have been identified (Harvey and Hamilton, 2005). It is conceivable that lava flowing from Syrtis Major onto the Nili Fossae region would form the kind of terrain from which both chassignites and nakhlites could have been launched in a single impact event. Tornabene et al. (2006) have recognized the 3.3 km-diameter rayed crater Zumba, located in Daedalia Planum south of Tharsis volcano, which reflects late Hesperian age volcanic terrain (spanning a period ~3.5 to ~1.8 b.y. ago). They consider this crater as a possible source of the 1.3 b.y. old chassignites and nakhlites.
The specimen of Chassigny shown above is a 0.76 g cut fragment with fusion crust along one edge. A large 312.406 g specimen of Chassigny is curated at the Muséum National d'Histoire de Paris and displayed on their website.