Abstract— The newly found meteorite Northwest Africa 6234 (NWA 6234) is an olivine (ol)-phyric shergottite that is thought, based on texture and mineralogy, to be paired with Martian shergottite meteorites NWA 2990, 5960, and 6710. We report bulk-rock major- and trace-element abundances (including Li), abundances of highly siderophile elements, Re-Os isotope systematics, oxygen isotope ratios, and the lithium isotope ratio for NWA 6234. NWA 6234 is classified as a Martian shergottite, based on its oxygen isotope ratios, bulk composition, and bulk element abundance ratios, Fe/Mn, Al/Ti, and Na/Al. The Li concentration and δ7Li value of NWA 6234 are similar to that of basaltic shergottites Zagami and Shergotty. The rare earth element (REE) pattern for NWA 6234 shows a depletion in the light REE (La-Nd) compared with the heavy REE (Sm-Lu), but not as extreme as the known “depleted” shergottites. Thus, NWA 6234 is suggested to belong to a new category of shergottite that is geochemically “intermediate” in incompatible elements. The only other basaltic or ol-phyric shergottite with a similar “intermediate” character is the basaltic shergottite NWA 480. Rhenium-osmium isotope systematics are consistent with this intermediate character, assuming a crystallization age of 180 Ma. We conclude that NWA 6234 represents an intermediate compositional group between enriched and depleted shergottites and offers new insights into the nature of mantle differentiation and mixing among mantle reservoirs in Mars.
Most ol-phyric shergottites, however, do not represent unmodified primitive mantle melts. Instead, they represent magmas that accumulated excess minerals (mainly olivine), and/or magmas that experienced either magmatic or hydrothermal contamination (Goodrich 2002; Shearer et al. 2008; Filiberto and Dasgupta 2011). Therefore, it is vital to fully investigate those shergottites that may represent mantle-derived melts, and also gather all available information about changes that happened to the original melt, and the processes behind such changes.
Meteorite Northwest Africa 6234 (NWA 6234) was found in 2009 at an undisclosed location in Mali, western Africa and purchased by an anonymous collector in February 2010. Based on the description in the Meteoritical Bulletin, NWA 6234 was a 55.7 g partly fusion-crusted stone that experienced moderate shock and limited weathering (Bulletin 2011). For this study, we purchased a 3.31 g slice of the meteorite, which includes a possible shock vein, from Marmet Meteorites, and confirmed that it matched the formal description of NWA 6234 (Bulletin 2011; Irving et al. 2011). The sample has been split and we have assembled an international consortium to investigate this meteorite; this work represents the first paper from this consortium (Filiberto et al. 2011).
Mineralogy of NWA 6234
NWA 6234 is a new ol-phyric shergottite that has some unusual features compared with other ol-phyric shergottites. It is relatively unaltered, fine-grained, and is composed of olivine (up to 1 mm in length) set in an even finer grained ground mass of pigeonite (with rare subcalcic augite), maskelynite, and accessory minerals including spinel, titanomagnetite, ilmenite, merrillite, apatite, and iron sulfide (Bulletin 2011; Gross et al. 2012). The olivine compositions are generally more ferroan than in other ol-phyric shergottites (Core: Fo79, Rim: Fo60, Groundmass: Fo40–47, Gross et al. 2012). The core olivine composition has an Mg# consistent with chemical equilibrium with a melt of the meteorite’s bulk composition, suggesting that NWA 6234 represents a magma composition (Gross et al. 2012). For a full description of the mineralogy, textures, oxygen fugacity, and crystallization history of NWA 6234, see Gross et al. (2012) and the Meteoritical Bulletin (Bulletin 2011). NWA 6234 is thought to be paired, based on mineralogy and textures, with NWA 2990, NWA 5960, and NWA 6710 (Bulletin 2011; Irving et al. 2011).
Previous Work on NWA 6234 and Paired Stones NWA 2990, 5960, and 6710
To date, work on this group of paired stones has been conducted mainly on NWA 2990 (Bunch et al. 2009; Lapen et al. 2009; Filiberto and Dasgupta 2011; Irving et al. 2011). Based on the original bulk chemical analyses, NWA 2990 was suggested to represent a magma composition (Filiberto and Dasgupta 2011), consistent with its fine-grained textures. This original bulk composition is much more ferroan than those of the ol-phyric shergottites NWA 5789 and Yamato 980459 (Y98), which are thought to represent mantle-derived melts (Musselwhite et al. 2006; Gross et al. 2011). This suggests that NWA 2990 may represent either (1) a fractionated melt and not a primitive mantle-derived melt or (2) a melt from a Martian mantle reservoir with high Fe/Mg. However, recent work on NWA 6234 suggests that the preliminary data on NWA 2990 may not be representative (Filiberto et al. 2011; Irving et al. 2011). The original analyses were performed on a powder that may have undersampled the olivine phenocrysts and so was not representative of the bulk meteorite (Irving et al. 2011). This discrepancy motivated the present study, where we present new bulk major-element abundances for NWA 6234 to reappraise the original bulk major-element data. In addition, we present new trace element abundances, Li abundance and isotope composition, oxygen isotope composition, highly siderophile element (HSE; Os, Ir, Ru, Pt, Pd, Re) abundances, and Re-Os isotope data to constrain the formation of NWA 6234.
To obtain a representative homogenous sample of NWA 6234, we powdered 240 mg of the interior of the stone. We chose a sample of the interior to minimize any potential terrestrial weathering influence and to avoid the aforementioned shock vein. The sample was powdered under ethanol for 1 h with an agate mortar and pestle. The resulting powder was placed in a drying oven at 100 °C for 24 h. A 15 mg split from the original powder was then made into a glass, as is routine in high-pressure experimental petrology, using an end-loaded piston-cylinder apparatus in the high-pressure laboratory at Rice University, using established techniques (Tsuno and Dasgupta 2011). The experiment was brought to 1 GPa and 1600 °C, held for 1 h, and rapidly quenched by turning the power off. A thick section mount was made of this sample for electron microprobe (EMP) and laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS). The experimental charge contains large glass patches with small amounts of quench crystals, which are common in experimental results on synthetic powders with Mg-rich bulk compositions. The quench crystals were analyzed by both EMP and LA-ICP-MS, and show no discernible compositional differences from the glass, except for lower analytical precision.
The rest of the material (homogenized powder and remaining chips) was sent to team members in the consortium for analyses of mineral chemistry; Li abundance and isotope ratio (reported here); Re-Os isotopes and HSE abundance data (reported here); oxygen isotope ratios (reported here); Ar isotope ratio (including a split of the impact melt vein); abundances and isotope ratios for all noble gases (for resolution of Martian interior and atmosphere components using step heating); stepped combustion analyses of C and N to determine current residual and initial magmatic volatile components; melt inclusion mineralogy and chemistry to further constrain the original magma composition and crystallization sequence; and Sm-Nd isotope ratios to further elucidate the mantle source and age of the meteorite (Filiberto et al. 2011).
Electron Microprobe Analyses
The major-element composition of the quenched glass was analyzed using the CAMECA SX-100 electron probe micro-analyzer in the ARES directorate at NASA Johnson Space Center. Analysis conditions were routine for this laboratory: 15 kV accelerating potential, 20 nA beam current, and defocused electron beam (5 μm diameter) to reduce volatile loss during the analyses. Peak and background counting times were 20–40 s per element. Analytical standards included well-characterized synthetic and natural glasses and minerals including diopside (Si, Ca, Mg), oligoclase (Na, Al), hematite (Fe), rutile (Ti), chromite (Cr), rhodochrosite (Mn), orthoclase (K), and apatite (P). Data quality was ensured by analyzing the standard materials as unknowns as well as by repeated analyses on the NWA 6234 glass (n = 55).
Oxygen Isotope Analyses
Oxygen isotope analysis was carried out at the Open University using an infrared laser-assisted fluorination system (Miller et al. 1999). A whole-rock chip of approximately 50 mg of NWA 6234 was powdered and homogenized and from this approximately 2 mg of powder was loaded for each replicate analysis. Oxygen was liberated from the sample by laser-heating in the presence of BrF5. After fluorination, the O2 released was purified by passing it through two cryogenic nitrogen traps and over a bed of heated KBr. O2 was analyzed using a MAT 253 dual-inlet mass spectrometer. Analytical precision (1σ), based on replicate analysis of international (NBS-28 quartz, UWG-2 garnet) and internal standards, is approximately ±0.04‰ for δ17O; ±0.08‰ for δ18O; ±0.024‰ for Δ17O (Miller et al. 1999). The quoted precision (2σ) for NWA 6234 is based on the results obtained from the mean of two analyses. To examine whether the oxygen isotope composition of NWA 6234 had been disturbed by terrestrial weathering, a fraction of the meteorite was leached in a solution of ethanolamine thioglycollate (EATG) prior to analysis. EATG is known to be efficient at removing terrestrial alteration products, such as iron-hydroxides, carbonates, and sulfates from a rock (Cornish and Doyle 1984; Martins et al. 2007).
Results for the oxygen isotope analysis of NWA 6234 are reported in standard δ notation, where δ18O has been calculated as: δ18O = [(18O/16Osample)/ (18O/16Oref) − 1] × 1000(‰) (ref = VSMOW: Vienna Standard Mean Ocean Water) and similarly for δ17O using the 17O/16O ratio. Δ17O, which represents the deviation from the terrestrial fractionation line, has been calculated as: Δ17O = δ17O − 0.52 δ18O.
LA-ICP-MS Trace-Element Analysis
Trace element and selected major-element data were obtained in situ from the fused glass prepared at Rice using a New Wave 213 nm laser ablation system coupled to a ThermoFinnigan Element 2 ICP-MS at Rice University. The ablated sample was carried via He gas out of the ablation cell and mixed with Ar gas before entering the plasma. Prior to ablation, approximately 30 s (approximately 10 measurements) of background are collected with the laser off, followed by 2–3 min of sample ablation (30–50 measurements). The following elements were analyzed in medium resolution (m/Δm = 3000): Na, Mg, Al, P, K, Ca, Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Nb, and Zr. The rare earth elements (La through Lu), Y, Li, Be, Rb, Sr, Cs, Ba, Pb, Th, Ta, and Tl were analyzed in low resolution (m/Δm = 300). BHVO-1, BIR-1, and BCR-2 glasses were used as external standards and analyzed at the beginning and end of the analytical session. 25Mg was used as an internal standard to correct for drift; the concentration of Mg in the fused glass was taken from EPMA measurements. Prior to analyzing the sample, the instrument was tuned manually by controlling sample gas (Ar) to obtain a sensitivity of approximately 200,000 cps per 15 ppm La in BHVO2g for 55 μm spot size and 10 Hz laser frequency. All analyses were performed using either 80 or 110 μm spot sizes and 10 Hz laser frequency at 100% laser power. Laser fluence was approximately 15 J cm−2.
A 50.3 mg split of well-homogenized powder of NWA 6234 was digested in a 4:1 mixture of concentrated Teflon-distilled (TD) HF and HNO3 inside a Teflon Parr bomb at >180 °C for >72 h in an oven along with separate digestions of standard rock powders and a total procedural blank. Digestion vessels were then allowed to cool, were uncapped, and solutions were taken to incipient dryness. Residues were repeatedly taken into solution with concentrated TD HNO3 and HCl and dried down to remove any residual fluorides. Gravimetrically weighed aliquots representing 25% of the final 4 mL 4 M HCl solution were separated for trace element abundance measurements, with the remaining solution being used for Li isotope and abundance measurements. Trace-element analysis solutions were doped with indium for internal normalization and drift correction, and diluted by a factor of 60 in 1 M HCl. Basaltic international standards BCR-2 and BHVO-2 were prepared using the same procedure as the sample for matrix matching. Trace element measurements of the same suite of elements as analyzed using LA-ICP-MS with the addition of Mo and exception of Co and Ga were accomplished using the Finnigan Element 2 ICP-MS at the University of Maryland, College Park (UMCP). Solutions were desolvated and introduced using an Elemental Scientific Inc. Apex with oxide production <2%. Raw data were corrected offline for blank contributions (low [<0.1% of measured concentration] and consistent), drift (negligible during analysis), and were calibrated and monitored relative to accepted values of international standards BHVO-2 and BCR-2 for accuracy. Reproducibility (as 1σ relative standard deviation) was <8% for most elements with the exception of Cr, Ni, Cu, and Zn for “unknown” runs of basalt standard BCR-2.
Li Isotope and Abundance Analysis
For Li isotope and concentration analysis, 25% aliquots from the solution trace-element digestion of NWA 6234, synchronously processed standard reference materials (see section above), and total procedural blank, were removed from the 4 mL 4M HCl solutions as described above. Lithium concentrations and isotope compositions were measured at the University of Maryland, College Park (UMCP). Procedures for separation and isotopic analytical procedures are the same as those reported in Penniston-Dorland et al. (2010). Each 1 mL aliquot was passed through three sets of columns for chromatographic separation, each containing 1 mL of cation exchange resin (BioRad AG50W-x12) following the first three column procedures described by Moriguti and Nakamura (1998) to separate Li for analysis. The yield was monitored by taking aftercuts and measuring their Li contents on the Finnigan Element 2 ICP-MS. Combined loss of Li for all columns was <2% Li for each of the three repeat analyses of NWA 6234. The samples were analyzed by standard bracketing (Tomascak et al. 1999) using the L-SVEC standard (Flesch et al. 1973) on the Nu Plasma Multi Collector Inductively-Coupled Plasma Mass Spectrometer (MC-ICP-MS) at UMCP. Solutions were desolvated using either a Cetac Aridus or Elemental Scientific Inc. Apex introduction system. The total procedural blank associated with the measurement of samples was <500 pg, which is ≤1% of the total Li concentration of the samples.
The long-term reproducibility of Li isotope analyses at UMCP, based on 2σ of repeat runs of pure Li standard solutions, is ≤±1.0‰ (Teng et al. 2006). For this study, measured values of δ7Li of standard solution Li-UMD-1 average +55.1 ± 0.7‰ (2σ, n = 6) and of IRMM-016 average +0.2 ± 0.4‰ (2σ, n = 4). Previously reported values for Li-UMD-1 and IRMM-016 range from +54.2 to +54.8‰ and −0.8 to +0.2‰, respectively. The measured δ7Li and Li concentration of rock standard BHVO-2 were +3.6‰ ± 0.6‰ and 5.1 ± 0.9 ppm (2σ, n = 3). Previously reported measurements of δ7Li for BHVO-2 range from +4.3 to +4.8‰ with reported concentrations of 4.2 and 4.8 ppm. Measured values of δ7Li and Li concentration of rock standard BCR-2 average +3.5 ± 1.2‰ and 9.8 ± 3 ppm (2σ, n = 4). Previously reported measurements of δ7Li for BCR-2 include values of +4.4, +2.6, +2.6, and +2.9‰ (see Penniston-Dorland et al.  for a full list of references for standard measurements).
Re-Os Isotope and Highly Siderophile Element Abundance Analysis
An approximately 0.2 g chip of NWA 6234 was used for Re-Os isotope and highly siderophile element abundance analysis (HSE: Os, Ir, Ru, Pt, Pd, Re). Prior to disaggregation, exterior surfaces exposed to sawing or splitting implements were thoroughly buffed with carborundum, and the sample was then washed with high purity water. The sample powder was prepared by coarse-crushing using an alumina mortar and pestle, in a dust-free environment to avoid contamination. The mortar and pestle were leached for >48 h using approximately 0.5 M HCl, abraded with pure silica powder and washed with ultra-pure water and dried prior to use. Qualitative analysis (count rates versus pure 0.5 M HCl diluent and a 10 ppt multi-element standard) of leachate from the mortar and pestle shows negligible HSE concentrations, implying no cross-contamination to the sample during disaggregation.
Osmium isotope and HSE abundance analyses were performed at the University of Maryland, College Park, using the technique described in Day et al. (2010b). A 0.134 g split of homogenized powder was digested in a sealed borosilicate Carius tube, with isotopically enriched multi-element spikes (99Ru, 106Pd, 185Re, 190Os, 191Ir, 194Pt), and 5.5 mL of a 1:2 mixture of multiply distilled HCl and purged and multiply distilled HNO3. The sample was digested to a maximum temperature of 270 °C in an oven for 96 h. Osmium was triply extracted from the acid using CCl4 and then back-extracted into HBr, prior to purification by micro-distillation. Rhenium and the other HSE were recovered and purified from the residual solutions using standard anion exchange separation techniques (Day et al. 2010b). Isotopic compositions of Os were measured in negative ion mode on a ThermoFisher Triton thermal ionization mass spectrometer. Rhenium, Pd, Pt, Ru, and Ir were measured using an Cetac Aridus II desolvating nebulizer coupled to a ThermoElectron Element 2 ICP-MS. Offline corrections for Os involved an oxide correction (Nier 1950), an iterative fractionation correction using a 192Os/188Os ratio = 3.08271, a190Os spike subtraction, and finally, an Os blank subtraction. External precision for 187Os/188Os ratio, determined by repeated measurement of the UMCP Johnson-Matthey standard over the course of a month of measurement, was better than ±2.1‰ (2σ; 0.11372 ± 23; n = 25). Duplicate measurements of unknowns were also in this range. Measured Re, Ir, Pt, Pd, and Ru isotopic ratios for sample solutions were corrected for mass fractionation using the deviation of the standard average run on the day over the natural ratio for the element. External reproducibility (2σ) on HSE analyses using the Element 2 was better than 0.5% for 0.1 ppb solutions and 0.3% for 1 ppb solutions, and all reported values are blank corrected. The total procedural blank run with NWA 6234 had a 187Os/188Os ratio of 0.168 ± 0.003, with quantities (in picograms) of 1.1 [Re], 3.3 [Pd], 1.9 [Pt], 2.7 [Ru], 3.3 [Ir], and 0.14 [Os]; these blanks represented 0.6%, 0.11%, 0.05%, 1.1%, 0.23%, and 0.02% of these elements in the measured sample, respectively.
The bulk chemical composition of NWA 6234 (Table 1) is consistent with the composition of previously studied Martian meteorite basalts. NWA 6234 is an Al-poor, Fe- and Mg-rich basaltic rock, with 17.1 wt% MgO, and the highest measured MnO content among ol-phyric and basaltic shergottites (0.56 wt% MnO) (Fig. 1).
Table 1. Major- and trace-element abundance data for NWA 6234.
NWA 6234 (RU)
NWA 6234 (UMCP)
RU = Rice University by LA-ICP-MS analysis.
UMD = University of Maryland by solution ICP-MS.
Trace element abundances by LA-ICP-MS (Rice University) and by solution ICP-MS (University of Maryland) are quite consistent, except for Ni. The disparity in analyzed Ni abundances may reflect “nugget” heterogeneities of this element in the glass bead charge, resulting in lower Ni concentrations during LA-ICP-MS analysis; however, Ni nuggets typically seen in experimental charges (e.g., Filiberto et al. 2009; Herd et al. 2009; Longhi et al. 2010) are not evident. The higher Ni values obtained by solution ICP-MS for NWA 6234 are consistent with known compositions of shergottites (see Table S1) (e.g., Barrat et al. 2001; Treiman 2003; Schmidt and McCoy 2010).
The rare earth element (REE) pattern of NWA 6234 (Fig. 2) is characterized by having similar relative and absolute abundances of the middle and heavy REE (MREE and HREE, respectively) to enriched ol-phyric shergottites such as Larkman Nunatak (LAR) 06319, NWA 1068, and Dhofar 378. However, light REE (La-Nd, LREE) in NWA 6234 are depleted relative to HREE and MREE, which is similar to the REE patterns of depleted ol-phyric shergottites like Dhofar 019, Yamato 980459, DaG476, and SaU 005/094. To compare the new meteorite data with the existing data on other basaltic and ol-phyric shergottites, we have compiled and averaged all published bulk chemistry analyses similar to previous studies (Lodders 1998) (Table S1). References for each meteorite are in Appendix A. For this paper, we have chosen to exclude lherzolitic shergottites from discussion because they are cumulate rocks and do not represent compositions of magmas (e.g., Ikeda 1998; Lin et al. 2005; Usui et al. 2010). Cumulate minerals will greatly affect the bulk chemistry and obscure the original igneous characteristics (Treiman 1996, 2003; Filiberto et al. 2006). This is especially problematic for rocks that contain xenolithic material (e.g., Elephant Moraine 79001 lithology A, Treiman et al. 1994; Goodrich 2003). Therefore, we have excluded all lherzolitic shergottites, as well as EETA lithology A from this discussion.
The average of the three measured δ7Li values of NWA 6234 is +3.2 ± 1‰ (long-term reproducibility, 2σ standard deviation for the three analyses is ±0.3‰), and the average Li concentration is 3.5 ± 1 ppm (2σ) (Table 2; Fig. 3). Previous bulk rock measurements of δ7Li values of Martian meteorites have mostly fallen within the range of +3.6 to +5.2‰ with the exception of orthopyroxenite ALHA84001, which has bulk rock δ7Li values of −0.6 and 1.2‰ (Magna et al. 2006b; Seitz et al. 2006). The δ7Li value of NWA 6234 is consistent within uncertainty with these values and with the estimates of Magna et al. (2006) and Seitz et al. (2006) for a homogeneous inner solar system δ7Li value of approximately 3.8 ± 0.4‰ (Magna et al. 2006a; Seitz et al. 2006). The reported Li concentrations of Martian meteorites are somewhat variable, with whole-rock Li concentrations ranging from 1.8 to 12.2 ppm. The measured value of 3.5 ppm Li is consistent with other basaltic shergottites, which range from 1.8 to 4.8 ppm Li (Magna et al. 2006b; Seitz et al. 2006).
Table 2. Oxygen, Re-Os, and Li isotope, and highly siderophile element and Li abundances for NWA 6234.
With the exception of Re, the HSE abundances of NWA 6234 are close to those reported for NWA 2990 (Fig. 4), a sample to which NWA 6234 is considered to be paired (e.g., Irving et al. 2011) (see below for a discussion of possible causes of this discrepancy). The chondrite-normalized HSE pattern for NWA 6234 is characterized by high (Pt, Pd, Re)/(Os, Ir, Ru) ratios, similar to the patterns observed for some ol-phyric and basaltic shergottites (e.g., Brandon et al. 2012), and for terrestrial alkali intraplate basalts (e.g., Day et al. 2010a). The abundances of the HSE lie within the range previously measured for ol-phyric and basaltic shergottites (Fig. 4). The measured 187Os/188Os value of 0.1367 for NWA 6234 is similar to that measured of 0.1387 for NWA 2990 and lies within the range of 187Os/188Os for other ol-phyric shergottites (0.1277–0.1481; with the exception of low Os ol-phyric shergottite Dhofar 019; Brandon et al. 2012).
A Martian Origin for NWA 6234: Oxygen Isotope Ratios and Bulk Chemistry
The mean of two untreated replicate oxygen isotope analyses gave the following composition for NWA 6234: δ17O = 2.652‰ ± 0.016 (2σ); δ18O = 4.475‰ ± 0.038 (2σ); Δ17O 0.325‰ ± 0.004 (2σ). The results are shown in Fig. 5 in relation to the data of Franchi et al. (1999) and laser fluorination analyses of other SNC samples published in the Meteoritical Bulletin. The analysis for NWA 6234 plots close to the Mars Fractionation Line (MFL) of Franchi et al. (1999) and confirms that the meteorite is a member of the SNC group, generally considered to be of Martian origin (McSween 1994; Treiman et al. 2000). Analysis of an EATG-leached fraction of NWA 6234 gave the following results: δ17O = 2.630‰ ± 0.090 (2σ); δ18O = 4.465‰ ± 0.134 (2σ); Δ17O = 0.308‰ ± 0.020 (2σ). While the precision obtained on the leached fraction is somewhat poorer than that of the untreated meteorite, both sets of data are within error of each other, indicating that the primary oxygen isotope composition of NWA 6234 has not been disturbed by terrestrial weathering processes.
A Martian origin for a meteorite (or at least kinship with the SNC meteorite group) can be verified based on element abundance ratios like Fe/Mn, Al/Ti, Na/Al (e.g., Dreibus and Wänke 1982, 1985, 1987; McSween 1985, 2002; Treiman et al. 2000; Barrat et al. 2002). However, recent analyses of basalts on Mars, and of terrestrial basalts, show that many of the element abundance ratios considered as characteristic of a planetary body (e.g., Fe/Mn, Na/Al, Al/Ti, K/Th) actually can vary significantly from basalt to basalt from the same planet (e.g., Filiberto 2008, 2011; McSween et al. 2009; Schmidt and McCoy 2010). With this caveat in mind, Fig. 1 shows that the Fe/Mn, Al/Ti, and Na/Al ratios for NWA 6234 are essentially identical to those of other ol-phyric and basaltic shergottites. Thus, the bulk composition of NWA 6234 is consistent with the oxygen isotope classification as a Martian basalt––more specifically a shergottite.
Weathering in NWA 6234
NWA 6234 is reported to be essentially unweathered (Bulletin 2011; Gross et al. 2012), but it is important to document the extent of its geochemical interactions with the environment of the Sahara. In general, weathering in hot deserts has been shown to affect elements whose compounds dissolve readily in water, like Ca, Sr, and Ba (Barrat et al. 2001; Crozaz et al. 2003; McLennan 2003). To evaluate the effects of hot desert weathering on NWA 6234, we focus on these elements as has been done for other Martian meteorites (Barrat et al. 2002).
Weathering of basaltic meteorites in terrestrial hot desert environments will not add or remove aluminum, but will dramatically increase fluid mobile elements such as Sr, Ba, and Ca. Therefore, plotting these elements against Al and comparing NWA 6234 to ol-phyric and basaltic shergottites found in both hot and cold deserts can help evaluate the amount of terrestrial alteration in NWA 6234 (Fig. 6). For NWA 6234, the Ca/Al ratio is similar to that of other ol-phyric shergottites, the Sr/Al ratio is rather elevated, and the Ba/Al ratio is significantly elevated (Fig. 6). Thus, it seems likely that the desert weathering of NWA 6234 has added some fluid-mobile elements (possibly as veinlets of Ba and Sr sulfates). However, there is no evidence of alteration mineralogy, such as sulfates, within NWA 6234 (Gross et al. 2012); therefore, we cannot rule out Ba addition by contamination during storage, shipping, or processing of this meteorite.
Furthermore, while the rest of the HSE in NWA 6234 are similar to those measured in NWA 2990 (Fig. 4b), Re––typically the most mobile of the HSE, is lower in NWA 2990 than NWA 6234, possibly indicating recent significant Re loss from NWA 2990, which has a similar 187Os/188Os composition (0.1387) to NWA 6234 (0.1367). We tentatively suggest rhenium loss from NWA 2990 due to weathering, rather than rhenium gain into NWA 6234, because the low Re/Os ratio in NWA 2990 contrasts with the measured enrichments in Pt and Pd relative to Os, Ir, and Ru.
Finally, weathering and the formation of clays and soil from rock can lower the δ7Li value of rock samples (Rudnick et al. 2004). NWA 6234 has a δ7Li composition in the range of Martian meteorite falls and finds, which suggests that (like Re-Os isotopes) the Li isotope composition of NWA 6234 has not been significantly affected by terrestrial processes. We conclude that, while NWA 6234 has presumably seen terrestrial alteration, the effect it has had on fluid immobile trace elements and on O, Li, and Os isotopes is limited. A possible test of this is through analysis of the noble gas signature of NWA 6234, as contamination would be clearly detected even after minimal exposure to terrestrial alteration (Mohapatra et al. 2009; Schwenzer et al. 2011).
Major-Element Composition of NWA 6234 and Its Relation to NWA 2990
The bulk composition of NWA 6234, as reported here, is significantly different from that of the presumably paired meteorite NWA 2990 reported by Irving et al. (2011) (Table 1). To explain this difference, it was hypothesized that the reported NWA 2990 composition represents a sample with an unrepresentatively low proportion of olivine phenocrysts (Irving et al. 2011). To test this hypothesis, we use a simple mass balance calculation for all major elements and compare the composition of NWA 6234 with mixtures of the reported NWA 2990 composition and that of the most forsteritic olivine in NWA 6234 (Bulletin 2011; Irving et al. 2011; Gross et al. 2012). To produce the NWA 6234 bulk composition, approximately 30–40 wt% forsteritic olivine must be added to the published NWA 2990 bulk composition. This is significantly more olivine than is actually present in NWA 6234 (approximately 20–25%, Gross et al. 2012), or most olivine-phyric shergottites (e.g., Goodrich 2002; Filiberto and Dasgupta 2011), and implies that the compositional differences between these two meteorites are not caused by underrepresentation of the abundance of olivine phenocrysts in the analyzed sample. It remains possible that NWA 2990 and NWA 6234 are not paired, but their similarity in 187Os/188Os and HSE abundances possibly implies a close relationship. To resolve this discrepancy, additional bulk chemical (major and trace) analyses of NWA 2990 are needed.
Li Isotope Constraints on Mantle Source Characteristics
The whole-rock abundance and isotopic composition of Li in NWA 6234 are comparable to those of other Martian meteorites, including basaltic shergottites (Fig. 3) (Magna et al. 2006b; Seitz et al. 2006). While the Li concentration of Martian nakhlites is higher than that of NWA 6234, the Li isotope composition of NWA 6234 also falls within uncertainty of these clinopyroxenites (Magna et al. 2006b; Seitz et al. 2006). Processes that might affect the isotope composition of meteorites after igneous crystallization include cosmic ray spallation reactions, solar wind effects, and weathering. Cosmic ray spallation has the potential to decrease the δ7Li value of a rock, as spallation reactions create 6Li and 7Li in equal abundances, while 7Li is generally approximately 12 times more abundant than 6Li in most Solar System materials. The effects of spallation are more evident in samples that contain little intrinsic Li. In previous studies, the possible effects of cosmic ray spallation on Li isotope compositions have been estimated based on the production of 3He in a sample (see Seitz et al.  for description). These calculated effects in Martian meteorites have been <0.09‰ (Seitz et al. 2006). Similar calculations for the effects of cosmic ray spallation on other types of meteorites have produced estimates in most cases <0.3‰ (Seitz et al. 2007). These estimates suggest that the effect of cosmic ray spallation on δ7Li of Martian materials is minor compared with Li isotopic signatures generated from igneous and surface processes on Mars. A study of lunar soil demonstrated the effects of interaction with isotopically heavy solar wind on the Li isotope composition (Seitz et al. 2006). These effects appear to be limited to samples with small particle size (<42 mm) that have surfaces bombarded by solar wind (e.g., soils). As Martian rocks are not soils and are relatively pristine, the effects of solar wind on δ7Li are likely minimal. The effects of weathering through clay formation and soil development were investigated by Rudnick et al. (2004). This study demonstrated depletions in δ7Li ranging up to approximately 15‰ associated with the degree of weathering (expressed as the chemical index of alteration). As there is no evidence for clay formation, and the range of δ7Li for NWA 6234 falls within the range for other Martian meteorites, the effects of weathering on Li are likely to be small. Thus, the measured δ7Li of NWA 6234 is interpreted to have been unaffected by secondary processes.
Rare Earth Element Constraints on Enriched or Depleted Source Region
Basaltic shergottites are typically categorized as either enriched or depleted in REE (e.g., Laul 1987; Longhi 1991; Norman 1999; Borg and Draper 2003; Jones 2003; Bridges and Warren 2006; Shearer et al. 2008). One commonly used method for characterizing the degree of enrichment or depletion of a magma in light REE is to ratio CI-normalized abundances of a light REE to a middle REE, e.g., the La/Sm ratio (e.g., Barrat et al. 2002b). With the exception of NWA 1068, LAR 06319, and Dhofar 378, which are enriched (La/Sm = 0.84–0.91 × CI), all other ol-phyric shergottites are depleted in the LREE (La/Sm = 0.14–0.3 × CI). In contrast, most basaltic shergottites are enriched (La/Sm = 0.76–0.88 × CI) with only QUE 94201 falling into the depleted category (La/Sm = 0.10 × CI). NWA 6234 is neither enriched nor depleted, but instead falls into an intermediate category with a La/Sm = 0.47–0.5 × CI. No other ol-phyric shergottite falls into this category; however, basaltic shergottite NWA 480 has a similarly intermediate composition and within error the same La/Sm ratio (La/Sm = 0.495 × CI). This suggests a unique source region for NWA 6234 that has similarities to the source region of NWA 480, but has not been previously sampled by other ol-phyric shergottites.
A Possible Link Between Ol-phyric and Basaltic Shergottites
To explore the possible link between ol-phyric shergottite NWA 6234 and basaltic shergottite NWA 480, we can look at the full range of REE and compare the REE pattern for NWA 6234 to ol-phyric (Fig. 2a) and basaltic (Fig. 2b) shergottites. NWA 6234 has a REE pattern that is unique among ol-phyric shergottites, because its light-middle REE (La through Nd) abundances are intermediate between the enriched olivine-phyric shergottites (NWA 1068, Dhofar 378, and LAR 06319) and the depleted olivine-phyric shergottites (Dhofar 019, SAU 005/094, Yamato 980459, and DaG476); however, NWA 6234 has a heavy REE pattern similar to NWA 1068 and Dhofar 378.
The trace element abundances of NWA 6234 have several similarities with basaltic shergottites. As noted above, NWA 6234 has a similar pattern of the REE to NWA 480, but it is shifted to lower abundances of all the REE. As olivine contains very low abundances of the REE, addition of olivine (such as in NWA 6234) would have a dilutional effect, shifting the entire bulk REE pattern to lower abundances than in an olivine-free magma (such as NWA 480). Addition of olivine would have an insignificant effect on relative REE abundances and ratios (e.g., La/Sm). Thus, the lower REE abundances in NWA 6234 compared with those in NWA 480 could merely reflect addition of olivine to the former (or removal from the latter). Moreover, the similarities of REE patterns of NWA 6234 and NWA 480 suggest a possible genetic relationship between them. Two simple models to explain this would be (1) NWA 480 and NWA 6234/NWA 2990 come from the same Martian mantle source, but NWA 6234 crystallized from a more primitive melt than NWA 480, or (2) NWA 6234 formed from a magma that was similar to NWA 480 and then accumulated, and possibly partially digested, excess olivine. Similar scenarios have been proposed for NWA 1068 and LAR 06319 (Barrat et al. 2002b; Shearer et al. 2008; Basu Sarbadhikari et al. 2009, 2011; Filiberto et al. 2010b; Filiberto and Dasgupta 2011). Because NWA 6234 appears to represent a magma composition (Gross et al. 2012), scenario 2 seems unlikely. Thus, we accept scenario 1––that NWA 6234 and NWA 480 share a mantle source region, but that NWA 6234 is a more primitive (less fractionated) magma from that source.
HSE Abundance and Osmium Isotope Composition of NWA 6234: Implications for Depleted and Enriched Sources
Given the similarity of Os, Ir, Ru, Pt, Pd abundances in NWA 2990 and NWA 6234, and the overall similarity of HSE CI-chondrite normalized patterns with other intermediate MgO shergottites (e.g., EETA 79001; Fig. 4), these data are consistent with earlier suggestions that NWA 2990 and NWA 6234 are likely to be paired. Of great importance, the HSE abundances measured for NWA 6234 suggest that the Martian mantle has HSE abundances much like those in the Earth’s mantle (Riches et al. 2011; Brandon et al. 2012). This inference suggests either that the Earth and Mars had nearly identical conditions of core formation and mantle differentiation, or that the mantles of both planets incorporated the same sort of “broadly chondritic veneer” after core formation.
Brandon et al. (2012) measured a 187Re/188Os ratio of 0.639 ± 0.042 for NWA 2990, and calculated its γ187Osi of +8.7 ± 0.4 (where γ187Osi is the age-corrected difference, in percent, of the sample over CI-chondrite), assuming an age of 180 Ma (like most other shergottites). This γ187Osi value is similar to that of other shergottites, and so suggests that Re and Os abundances in NWA 2990 have not been significantly disturbed.
Abundances of Re are problematic; they are significantly different in NWA 6234 and its supposed pair NWA 2990. Despite the similarity in 187Os/188Os and the relative and absolute abundances of Os, Ir, Ru, Pt, and Pd in NWA 6234 and NWA 2990, the Re concentrations measured in these samples are substantially different. NWA 6234 has 187Re/188Os = 1.707 ± 0.052, and a calculated γ187Osi of +4.1 ± 0.1 assuming an age of 180 Ma (if its age is 450 Ma, then it would have had γ187Osi = −0.6 ± 0.2; if its age were 4.0 Ga, then γ187Osi = −81 ± 1). Very tentatively, Brandon et al. (2012) suggested that the γ187Osi value for NWA 2990 may represent the highest value for any sample in the intermediate shergottite group and would be consistent with NWA 2990, representing a unique type of shergottite with respect to its isotopic composition and potentially its mantle source. However, it is possible that the low Re measured in NWA 2990 may represent recent Re removal from the sample, leading to higher apparent γ187Osi. Depleted ol-phyric shergottites have γ187Osi of between −3.4 and +4.3 (DaG 476; Sau 005/094, Y 98; Brandon et al. 2012) and enriched ol-phyric shergottites have γ187Osi of between +14.7 and +15.2 (NWA 1068, LAR 06319; Brandon et al. 2012). Regardless of the potential issue of rhenium-disturbance in NWA 2990, the values of γ187Osi at 180 Ma for NWA 6234 (+4.1) and NWA 2990 (+8.7; Brandon et al. 2012) are broadly intermediate to these other olivine-phyric shergottite γ187Osi values, consistent with the notion that NWA 6234/2990 are intermediate ol-phyric shergottites based on REE abundance data.
From its oxygen isotopic composition and major- and trace-element budget, NWA 6234 is clearly a new sample of the Martian SNC suite of meteorites. NWA 6234 is an ol-phyric shergottite with trace element characteristics similar to basaltic shergottite NWA 480 and isotopic characteristics similar to ol-phyric shergottite NWA 2990. In terms of the REE, it (along with its putative paired stones) is neither enriched nor depleted, and thus is distinct from any other ol-phyric shergottite. NWA 6234 is a unique meteorite that represents the first geochemically intermediate olivine-phyric shergottite and suggests a unique source region for NWA 6234, which has a similar trace element signature to the source region of NWA 480, but has not been previously taped by other ol-phyric shergottites.
Acknowledgments–– We are grateful to R. Dasgupta and C.T. Lee for help with experimental and analytical equipment and useful discussions. We thank N. Shirai and an anonymous reviewer as well as associate editor A. Yamaguchi for helpful reviews of this manuscript. This work was supported by NASA MFR grant # NNX09AL25G to A.H. Treiman and J. Filiberto, an SIU startup package to J. Filiberto, NASA LASER grant # NNX11AG34G and NASA Cosmochemistry grant # NNX12AH75G to J. Day, and NSF grant EAR 0911100 to S.C. Penniston-Dorland. This is LPI Contribution 1669.