Quantifying noble gas contamination during terrestrial alteration in Martian meteorites from Antarctica


Corresponding author. E-mail: s.p.schwenzer@open.ac.uk


We investigated exterior and interior subsamples from the Martian shergottite meteorites Allan Hills (ALH) A77005 and Roberts Massif (RBT) 04261 for secondary minerals, oxygen isotopes, Ar-Ar, and noble gas signatures. Electron microprobe investigations revealed that RBT 04261 does not contain any visible alteration even in its most exterior fractures, whereas fracture fillings in ALHA77005 penetrate into the meteorite up to 300 μm, beyond which the fractures are devoid of secondary minerals. Light noble gases seem to be almost unaffected by terrestrially induced alteration in both meteorites. Thus, a shock metamorphic overprint of 30–35 GPa can be deduced from the helium measurements in RBT 04261. Oxygen isotopes also seem unaffected by terrestrially weathering and variations can easily be reconciled with the differences in modal mineralogy of the exterior and interior subsamples. The measurements on irradiated samples (Ar-Ar) showed a clear Martian atmospheric contribution in ALHA77005, but this is less apparent in our sample of RBT 04261. Exterior and interior subsamples show slight differences in apparent ages, but the overall results are very similar between the two. In contrast, krypton and xenon are severely affected by terrestrial contamination, demonstrating the ubiquitous presence of elementally fractionated air in RBT 04261. Although seemingly contradictory, our results indicate that RBT 04261 was more affected by contamination than ALHA77005. We conclude that irrespective of on which planet the alteration occurred, exposure of Martian rocks to atmosphere (or brine) introduces noble gases with signatures elementally fractionated relative to the respective atmospheric composition into the rock, and relationships of that process with oxygen isotopes or mineralogical observations are not straightforward.


Terrestrial alteration and the associated contamination of Martian signatures by those of the Earth's surface are a significant issue because few meteorites are observed falls; and even for falls, adsorption of air and changes caused by moisture cannot be excluded entirely. For Martian meteorites, about 10% (with the recent fall Tissint [Irving et al. 2012] included in the count) can be categorized as falls, 50% have been found in hot deserts, >30% have been found in Antarctica (US and Japanese collections only); and a few have unclear provenance (Meyers 2012). Antarctica is a major source of meteorites for scientific investigations because they are collected by expeditions exclusively dedicated to meteorite recovery, are well documented, and are appropriately stored and curated, for example by NASA (2012) and the Japanese National Institute of Polar Research (NIPR 2012). It is therefore critically important to understand the changes that Martian meteorites undergo while on/in the Antarctic ice sheet.

Weathering in a polar desert environment is caused by mechanical and temperature stresses and occasional wetting. Some examples: at Haughton impact crater (Devon Island, Nunavut, in northern Canada), mechanical weathering is found to dominate in the surface layers with a shift to chemical weathering at greater depth (Lacelle et al. 2008). In Wright Valley, Antarctica, soil profiles contain a salt deposition zone, where salts with seawater composition and some elements leached from local rocks are deposited. Leaching can also be confirmed in Wright Valley by strontium isotopes in local water, but chemical weathering is generally considered slow (Gibson et al. 1983).

Weathering in Antarctic meteorites is thought to be caused by thin films of water/brine that can exist even below the freezing point of pure water (Gooding 1986a). This happens while the meteorite is buried in ice through capillary water and during freeze-thaw cycles, when the meteorite is exposed on the surface of the ice sheet (see review by Koeberl and Cassidy 1991 and references therein). For meteorites, weathering scales have been developed (e.g., Wlotzka 1993). For the A-B-C weathering index used for Antarctic meteorites, Solberg and Burns (1989) found that these indices relate to 4–5%, 7–10%, and up to 30% ferric phases, respectively. Reflectance spectra of chondrites reveal that Antarctic finds have a wide diversity of 3 μm band (i.e., OH) intensities, but are characterized by high intensities on the 7.4 μm band (i.e., carbonate) compared with falls (Miyamoto 1991). Evaporite formation on meteorites collected by the ANSMET expeditions shows differences related to geographic location; percentages of meteorites with evaporites range from 0.7% to 13.5% (Losiak and Velbel 2011). Other controlling factors are the composition of the meteorite itself and the year of collection, but no correlation with the weathering index based on oxidation could be found (Losiak and Velbel 2011).

The effects of weathering on the mineralogy of several meteorite groups have been studied intensively. Mineralogical changes in stony meteorites start with the occurrence of some “rust,” because of the disintegration of sulfides and oxides (Gooding 1986a), with clay minerals likely to form from silicates (Gooding 1986b). In chondrites, several oxides (e.g., magnetite, hematite), hydroxides (e.g., goethite, ferrihydrite), sulfates (e.g., melanterite, jarosite, kieserite), and hydrated carbonates (e.g., nesquehonite, zaratite) have been identified (Gooding 1981). Other minerals observed are zeolites and carbonates (Gooding 1986b). The occurrence of characteristic evaporitic minerals has been found to be associated with specific host meteorite types (Velbel et al. 1991 and references therein). In the case of ordinary chondrites, Mg-carbonates form as a product of terrestrial water and atmospheric CO2 with Mg derived from the meteorite itself rather than from a terrestrial source (Velbel et al. 1991).

In terms of chemical changes caused by Antarctic weathering, Nobuyoshi et al. (1997), in their study on Antarctic chondrites, find that limonite formation occurred in a “closed system.” But some transport is required to account for Ce and Eu anomalies (Mittlefehldt and Lindstrom 1991; Strait et al. 1991). Iodine and minor Cl enrichment has been measured in Antarctic shergottites (Dreibus et al. 1986), whereby iodine is surface bound (Heumann et al. 1990). For noble gas studies, this results in an interesting aspect of polar desert weathering: iodine is an airborne contaminant.

Although mobility of most elements requires transport as particles or through a solvent, noble gases are constituents of air. Fractionated adsorption of air in the Antarctic environment has indeed been found in a variety of meteorites and analog materials (Schelhaas et al. 1990; Scherer et al. 1994; Patzer and Schultz 2001; Schwenzer et al. 2007b, 2008, 2009; Nagao and Park 2008; Nagao et al. 2008; Mohapatra et al. 2009; Okazaki et al. 2010). Another Antarctic weathering effect is the loss of noble gases (Graf and Marti 1995; Alexeev 1998; Patzer and Schultz 2001; Schwenzer et al. 2008). Studying noble gases adsorbed on analog materials has led to the insight that those adsorbed noble gases neither show air elemental ratios nor are removed by preheating or during a first, low-temperature measurement step (Schwenzer et al. 2012a). A more detailed analysis of weathering effects on the noble gas budget in Antarctic meteorites, including literature examples, is provided in the discussion section of this article.

A study of three Martian meteorites from the Japanese Antarctic collection revealed some of them to be pervasively contaminated by elementally fractionated terrestrial air (Schwenzer et al. 2009). This finding motivated a more in-depth study to answer the questions “what determines the extent of contamination” and “which measurements could predict weathering and air contamination of those meteorites.” Following from this previous study, here we are using the approach that Gooding and Muenow (1986) took to distinguish between Martian and terrestrial S-species and apply it to noble gases: we obtained exterior and interior subsamples to assess terrestrial weathering in Antarctic meteorites. In the present study, preliminary results of which were previously presented in abstract form (Schwenzer et al. 2012b), we performed mineralogical, oxygen, Ar-Ar, and noble gas measurements on outside and inside portions of sampled Antarctic finds to understand potential weathering effects and how effects in the various systems are related.

Samples and Methods

This study focuses on two Martian shergottite meteorites found in Antarctica: RBT 04261 and ALHA77005. We chose shergottites to evaluate the effects, especially for Martian basalts. The two meteorites were chosen specifically because of their large size and the possibility to sample material from both the rim and a presumably more pristine interior.

RBT 04261 and RBT 04262 were found close to each other and have a similar Martian ejection age of 2.9 Myr (Nishiizumi and Caffee 2010). Although this suggests pairing of the two stones, the terrestrial residence ages of the two meteorites are very different (Nishiizumi and Caffee 2010): <60 kyr for RBT 04261 and 0.7 Myr for RBT 04262. These results are in agreement with the cosmic ray exposure age of 2.1 Ma calculated from noble gas isotopes for RBT 04262 (Nagao and Park 2008). The pairing conundrum cannot be resolved from geochemical evidence, where overall major and trace elements point toward pairing (Usui et al. 2010), but the poikilitic olivines suggest otherwise (Alpert et al. 2012). Both RBT meteorites are described as lherzolithic shergottites, and have two textures—a poikilitic olivine lithology and a nonpoikilitic lithology consisting of pyroxene, olivine, and maskelynite (Usui et al. 2010). Shock effects include extensive fracturing and the presence of maskelynite (Anand et al. 2008), but no shock melt veins are present. Niihara et al. (2010) determined an U-Pb age on baddeleyite of 203 ± 30 Ma for RBT 04261. Niihara (2011) using the same technique reports approximately 200 Ma, and a 174 ± 14 Ma Sm-Nd isochron age has been obtained (Shih et al. 2009). In contrast, Albarède et al. (2009) find the Pb-Pb age to be consistent with their earlier 4.1 Ga age determinations for other shergottites. Terrestrial alteration has been suspected from δD measurements on gypsum and jarosite found in RBT 04262 (Greenwood et al. 2009).

ALHA77005 is also a lherzolitic shergottite. It consists of olivine, pyroxene, maskelynite, and a “pervasive presence of shock melt” (Treiman et al. 1994 and references therein). Its crystallization age is 154 ± 6 Ma and the shock event can be dated at less than 20 Ma (Jagoutz 1989). Sm and Gd isotopes point to a simple preterrestrial exposure history (Hidaka et al. 2009), and 36Cl to evidence for solar cosmic irradiation (SCR; Nishiizumi et al. 1994). The terrestrial age is reported as 190 ± 70 and 210 ± 70 ka, respectively (Schultz and Freundel 1984; Nishiizumi et al. 1994). Studying cosmogenic nuclides indicates that ALHA77005 was ejected from Mars as a small object (5–6 cm radius) 2.7 Ma ago (Nishiizumi et al. 1986), but exposure ages as high as 3.6 Ma have been reported (Eugster et al. 1996).

We obtained three bulk sample splits and a thin and thick section from each Martian meteorite, ALHA77005 and RBT 04261. In detail, the following samples were allocated to us: ALHA77005,225 (exterior), 280 mg; ALHA77005,226 (rind), 230 mg; ALHA77005,227 (interior), 310 mg; RBT 04261,34 (exterior), 270 mg, RBT 04261,35 (rind), 220 mg; RBT 04261,37 (interior), 220 mg; in addition, we received an ALHA77005,54 thin section, an ALHA77005,230 thick section, a RBT 04261,43 thin section, and RBT 04261,42 thick section. “Exterior” is from the meteorite's outside including fusion crust (if available), “rind” is close to the exterior, but does not include portions of the outside or fusion crust, and “interior” is taken from an interior area, and thus represents a sample that is normally taken from any meteorite for studies that do not require material from the outside area.

Electron Microprobe Measurements

Two thin sections, both comprising the outer surface of the meteorite, were investigated using the Open University's CAMECA SX100 electron microprobe. Quantitative analyses on spots and lines were accompanied by X-ray element distribution maps. Analytical conditions applied were standard protocols, i.e., 20 kV, 20 nA, beam diameter of 10 μm for quantitative analysis.

Oxygen Isotope Measurements

Oxygen isotope analysis was carried out at the Open University using an infrared laser-assisted fluorination system following the methods described in Miller et al. (1999). Samples were analyzed both untreated and after leaching in a solution of ethanolamine thioglycollate (EATG), which has proved to be efficient at removing terrestrial weathering products, without significantly disturbing the primary oxygen isotope composition of the sample (Martins et al. 2007). All analyses were obtained on whole-rock samples (0.5–2 mg) that were heated with a 10.6 μm laser 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 long-term averages of replicate analyses 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). All sample powders analyzed in this study were drawn from larger homogenized aliquots, which had a mass of >200 mg (see above). The precision (1σ) quoted for individual meteorite samples analyzed in this study is based on replicate analyses.

Oxygen isotopic analyses are reported in standard δ notation, where δ18O has been calculated as: δ18O = [(18O/16Osample)/(18O/16Oref) − 1] × 1000(‰) and similarly for δ17O using the 17O/16O ratio. Δ17O, which represents the deviation from the terrestrial fractionation line, has been calculated using the conventional format: Δ17O = δ17O − 0.52 δ18O.

Ar-Ar Measurements

For Ar analysis, whole-rock pieces were handpicked for single grain fusion, and overall approximately 40 grains per sample were used. The samples were irradiated at McMaster Reactor (Canada) for 100 MWh. J-values were obtained from simultaneous irradiation of the biotite standard GA1550, which has an age of 98.8 Myr (Renne et al. 1998), and are given as a footnote of Table 4. We used the decay constants by Steiger and Jäger (1977). The samples were melted using an infrared laser. Melting was determined by optical inspection, i.e., by the observation that the grain contracted to a spherical glass bead. The released gases were cleaned by a cold-finger and getter sequence and measured with a MAP 215–50 noble gas mass spectrometer. Blanks were monitored throughout the day by alternating blank and sample measurements. Day averages of blanks on 40Ar were generally <7.2 × 10−12 ccSTP and on 36Ar <4.3 × 10−14 ccSTP. Daily averages for the blank were used. 37Ar decay and neutron-induced interference reactions from Ca and K were corrected using the correction factors (39Ar/37Ar)Ca of 0.00065, (36Ar/37Ar)Ca of 0.000264, and (40Ar/39Ar)K of 0.0085.

Five Noble Gases

All five stable noble gases were measured on unirradiated material in the noble gas laboratory of the Max-Planck Institute in Mainz as described by Schelhaas et al. (1990) and Schwenzer (2004). Approximately 50 mg of ground bulk sample was wrapped in a platinum foil, loaded into the sample tree, and degassed under vacuum at 130 °C for 24 h to remove adsorbed terrestrial gases. The samples were measured in four temperature steps (500, 1000, 1400, 1800 °C). The gases were cleaned by Ti- and Al-Zr getters; separated into four fractions (He+Ne, Ar, Kr, Xe); and measured in the order Ne, He, Ar, Kr, and Xe. Typical 1800 °C blanks gave 36Ar = 1.1 × 10−12, 84Kr = 4 × 10−14, and 132Xe = 7 × 10−15 (in cm³ STP units). Data were corrected for mass spectrometric interferences using standard correction protocols (Schelhaas et al. 1990; Mohapatra et al. 2009).

Component Partitioning for Noble Gases

To evaluate the contribution of cosmogenic, radiogenic, and Martian versus terrestrial atmospheric components, partitioning of the noble gases between trapped (plus radiogenic in the case of 4He and 40Ar) and cosmogenic components was undertaken. In the following, “c” refers to cosmogenic, “tr” to trapped (+ radiogenic), “rad” to the pure radiogenic component in 4He and 40Ar. To disentangle the radiogenic from the cosmogenic 4He, 3He was assumed to be 100% cosmogenic and the cosmogenic ratio (4He/3He)c was taken as 4.1 (Schwenzer et al. 2007a, 2008). Abundances of trapped 20Ne and cosmogenic 21Ne and estimates for the cosmogenic (22Ne/21Ne)c ratio were derived using (20Ne/22Ne)tr = 10 ± 1, (21Ne/22Ne)tr = 0.029 ± 0.002 (these values cover both air and most published values for Martian atmosphere), and (20Ne/22Ne)c = 0.83 ± 0.03 (GCR, chondritic range; Wieler 2002). If the low (20Ne/22Ne) value of 7 for Martian atmosphere suggested by Garrison et al. (1995) and Park and Nagao (2006) is used, the resulting cosmogenic 21Ne and (22Ne/21Ne)c become slightly lower. For argon, air and Martian atmosphere are possible choices as the trapped component; different choices for (38Ar/36Ar)tr within the range given in the literature for Martian atmosphere (Wiens 1988; Bogard 1997) change calculated 38Arc and 36Artr abundances to various degrees. As our heavy noble gas data show contamination by terrestrial air (see below), we took (38Ar/36Ar)tr = 0.1885 ± 0.02 (terrestrial atmosphere) and (38Ar/36Ar)c = 1.54 ± 0.05.



Thin sections for both samples containing rim portions were examined for any mineralogic changes caused by Antarctic weathering. In RBT 04261, no secondary minerals were found (Fig. 1a). The sample shows pristine fusion crust and clean fractures. This is in agreement with previous studies (Usui et al. 2010). Results of mineral analysis along fractures did not differ from those away from fractures.

Figure 1.

a) Backscattered electron image (left column, middle panel) and element maps for an approximately 1 mm wide part of the rim of RBT 04261. Note the fusion crust in the lower right corner of the image. Apatite is preserved and fractures are devoid of secondary minerals. Scale bar is 500 μm. Numbers on the right of each panel are count intensities. b) Backscattered electron image (left column, middle panel) and element maps for an approximately 700 μm wide part of the rim of ALHA77005. Note the Mg and Fe depletion (upper two panels), which is accompanied by elevated Si, S, and Ca amounts. Scale bar is 200 μm. Numbers on the right of each panel are count intensities.

In contrast, some terrestrial alteration was observed along the exterior of ALHA77005 (Fig. 1b). The backscattered electron image shows a fracture, which is filled with finely crystalline material. X-ray element maps of the area reveal Mg and Fe depletion along the fracture, which is accompanied by elevated Si, S, and Ca contents. Deeper into the sample, Fe-S-filled fractures also occur. The alteration continues up to about 300 μm into the sample after which fractures appear clean and are devoid of any secondary materials. All quantitative measurements on individual minerals were carried out close to but farther than 10 μm away from any visible alteration and then compared with analyses from the interior part of the sample. Measurements of minerals in the exterior away from fractures and the interior were indistinguishable; thus, the alteration is confined to the cracks of the very outer portion of the meteorite.


Full results of oxygen isotope analysis of ALHA77005 and RBT 04261 are given in Table 1 and shown in (Figs. 2a–c). In Fig. 2a, oxygen isotope analyses for ALHA77005 and RBT 04261 (average of EATG-washed analyses, Table 1) are plotted in relation to the Martian Fractionation Line (MFL) of Franchi et al. (1999) and laser fluorination data for other Martian samples (compilation of data published in Meteoritical Bulletins: 80–98). ALHA77005 and RBT 04261 plot close to the MFL and have oxygen isotope compositions that are well within the broad cloud of data defined by other shergottite samples.

Table 1. Oxygen isotope results
Average 2.4804.1830.305
Average 2.4964.2080.308
RBT 04261-untreated
Average 2.5834.3830.304
RBT 04261-EATG
Average 2.6264.4350.320
Figure 2.

a) Oxygen isotope composition of ALHA77005 and RBT 04261 (average EATG-washed analyses, Table 1) compared with other Martian samples (data: Meteoritical Bulletin and Franchi et al. 1999; see text for further details). b) EATG-washed and untreated fractions of ALHA77005 and RBT 04261 compared (E = exterior; R = rind; I = interior). c) ALHA77005 and RBT 04261 (average of untreated analyses, Table 1) compared with untreated and EATG-washed Antarctic winonaite samples (data: Greenwood et al. 2012). MFL = Mars fractionation line (Franchi et al. 1999); TFL = terrestrial fractionation line. Indicative 2σ error bars shown based on long-term reproducibility reported by Miller et al. (1999). Analytical uncertainties on the measurements made during the course of this study are likely to be less than those given by Miller et al. (1999).

RBT 04261 shows only minor oxygen isotope variation between its various fractions, which are generally no more than about 0.1‰ and therefore below the level of precision of the measurements, which for δ18O is ±0.16‰ (2σ) (Miller et al. 1999). The EATG-washed fractions of RBT 04261 show a slight difference between the interior and exterior of the sample of just over 0.2‰ with respect to δ18O. As EATG treatment should largely remove terrestrial weathering products, this variation may reflect a slight compositional difference between the interior and exterior of the sample. In conclusion, and in keeping with the mineralogical observations, oxygen isotope analysis suggests that the level of terrestrial alteration experienced by RBT 04261 was small. Similarly, there are only minor differences between the untreated and EATG-washed fractions of ALHA77005, again indicating that its oxygen isotope composition has been largely unaffected by terrestrial weathering processes.

Compared with meteorites with a high metal and sulfide content, such as chondrites and primitive achondrites, Martian meteorites generally do not display significant shifts in their oxygen isotope composition as a result of Antarctic terrestrial weathering processes. This is illustrated in Fig. 2c where the untreated analyses for ALHA77005 and RBT 04261 are shown with respect to untreated and EATG-washed Antarctic winonaite analyses (Greenwood et al. 2012). Untreated winonaites can show shifts of almost 7‰ compared with their respective EATG-washed fractions (Fig. 2c) (Greenwood et al. 2012). This behavior reflects the fact that Fe,Ni metal, and sulfides are significantly more susceptible to Antarctic weathering than silicates (Gooding 1986a; Bland et al. 2006) (see further discussion below).

Helium and Neon

Helium measurements in both meteorites yield quite different results (Table 2). ALHA77005 has a 3He/4He ratio of approximately 0.22 (Fig. 3). This is close to the ratio expected from cosmic ray production in Martian meteorites (Schwenzer et al. 2008). Rim-interior differences are subtle to nonexistent. The average cosmic ray exposure age calculated from 3He (T3) of all three subsamples is 4.52 ± 0.07 Ma. This is significantly higher than the previous highest report of 3.6 Ma by Eugster et al. (1996). In fact, our sample contained more 3He than is to be expected from theoretical calculations (Leya et al. 2000, 2004): for a sample with ALHA77005 composition irradiated as small (5 cm radius) object in space, which should contain no more than 4.6–5.1 ccSTP g−1 3He after 3.6 Ma cosmic irradiation. This compares with the average of 7.1 ± 0.2 × 10−8 ccSTP g−1 for our three subsamples. Our 3He concentration is similar in all samples, and it is well inside the range covered by literature values (e.g., 6.2 × 10−8 ccSTP, Bogard et al. 1984; 6.0 × 10−8 ccSTP, Miura et al. 1995, 7.49 × 10−8 ccSTP, Eugster et al. 2002). We also note that ALHA77005 has lost almost all radiogenic 4He, which is in accord with the high shock pressure this meteorite experienced (Fritz et al. 2005), and the complete helium loss expected from that event (Schwenzer et al. 2008).

Table 2. Results of the helium measurements of RBT 04261 and ALHA77005. Exterior, rind, and interior samples are subsamples 34, 35, and 36 for RBT 04261 and 225, 226, and 227 for ALHA77005. Concentrations are in 10−8 ccSTP/g. T3 in Ma
RBT 04261
Figure 3.

3He versus 4He for both meteorites. The dotted line highlights a 4He/3He ratio of 4.1, which is the cosmic 4He/3He ratio as found in Martian meteorites (see Schwenzer et al. 2008 and discussion therein).

The situation is different for RBT 04261, which shows clear differences between the subsamples and a much wider scatter in T3. The interior sample has approximately 50% less 4He compared with the exterior sample, but at the same time approximately 55% more 3He (Table 2). T3 scatters over a wider range increasing from 1.79 to 2.74 Ma from the exterior to the interior. Using chemical data of RBT 04262 (Anand et al. 2008) and theoretical cosmic ray production calculations (Leya et al. 2000, 2004), we find that the (22Ne/21Ne)cos of RBT 04261 agrees with theoretical values for the exterior and interior of a small object (approximately 5 cm radius) in space. With this, it becomes possible to estimate the theoretical amount of 3He produced in RBT 04261 for exterior and interior to be 1.19 and 1.31 × 10−8 ccSTP (g−1 Ma−1). Factoring in the cosmic ray exposure age of 2.1 Ma for RBT 04262 (Nagao and Park 2008) leads to expecting 2.5 and 2.8 × 10−8 ccSTP g−1 of 3He in exterior and interior, respectively; and using an exposure age of 2.9 Ma (Nishiizumi and Caffee 2010) leads us to expect 3.5 and 3.8 × 10−8 ccSTP g−1 of 3He in exterior and interior, respectively. Our subsamples contain between 2.8 and 4.5 × 10−8 ccSTP g−1 of 3He (Table 2). Given the uncertainties in chemistry, sampling depth, and the likelihood of helium loss in the outer portions of the stone, we note that our measurements are in fair agreement with the expected values.

Using the only analysis available to date (Th = 0.257 ppm, U = 0.058 ppm; Anand et al. 2008 on RBT 04262) and the baddeleyite age of 203 Ma (Niihara et al. 2010), 2.95x10-6 ccSTP g−1 of radiogenic 4He would be expected in RBT 04261 (assuming that it is paired with RBT 04262 and thus contains the same U and Th concentrations). In contrast to our findings in ALHA77005, the measured concentrations in RBT 04261 show that 36, 29, and 13% of the expected radiogenic 4He are still contained in the exterior, rind, and the interior portions of the meteorite, respectively (Table 2). Taking the potentially large uncertainties from the different samples and the small measured sample size into account, this reflects a loss of radiogenic 4He of about 70–80%. As weathering loss (Graf and Marti 1995; Alexeev 1998; Patzer and Schultz 2001; Schwenzer et al. 2008) is expected to act more severely on the exterior than the interior, we exclude this possibility. The difference between the two rim samples (exterior and rind), which were sampled in close proximity to the interior sample, could be a result of inhomogeneous distribution of the U and Th carrier phase, because the loss calculation is based on an average value. Petrologically, RBT 04261 contains two lithologies; therefore, differences in trace phase distribution are to be expected. Another possibility could be shock heating, which is notoriously inhomogeneous on a small scale, because it is more intense where high density contrast occurs in the sample. The average loss of 4He can be related to a shock metamorphic overprint of the sample at 30 to 35 GPa using the correlation of Schwenzer et al. (2008). This is similar to shock pressures observed in Shergotty and Zagami Martian meteorites. Both meteorites contain maskelynite, but after many years of exhaustive research, in Shergotty, shock melt has only been found on grain boundaries between pyroxene and opaque minerals, whereas, in Zagami, shock melt is found in the dark mottled areas, but not in the main lithology (Meyers 2012). Shock effects have been described as maskelynitization and extensive fracturing for RBT 04262 (Anand et al. 2008). The petrographic similarity supports our estimate of comparable shock pressure exposure of RBT 04261 to Shergotty and Zagami.

Table 3. Results of the neon measurements. Exterior, rind, and interior samples are subsamples 34, 35, and 36 for RBT 04261 and 225, 226, and 227 for ALHA77005. Concentrations are in 10−10 ccSTP g−1. T21 in Ma
RBT 04261
RBT 04261

20Ne/22Ne versus 21Ne/22Ne ratios for all samples fall close to, but toward slightly lower 21Ne/22Ne ratios, than the spallation field for chondrites (Fig. 4; Table 3), which has been interpreted as SCR contribution (Garrison et al. 1995; Schwenzer et al. 2007). Note also, however, that the expected range of GCR- 21Ne/22Ne for ALHA 77005 (Treiman et al. 1994) and RBT 04261 (Anand et al. 2008) chemistry falls to lower 21Ne/22Ne ratios than for chondritic abundances, reflecting the higher Na-content of Martian meteorites compared with chondrites. There are subtle differences (see inset of Fig. 4) between exterior and interior of each sample. The exterior sample of RBT04261 has the highest 20Ne/22Ne, which may indicate terrestrial contamination, but cannot be distinguished from a Martian atmospheric contribution. Additionally, all ALHA77005 subsamples show higher than chondritic (or calculated from their respective chemistries) 20Ne/22Ne. Also, two samples are lower in their 21Ne/22Ne ratio than expected from chondritic ratios and closer to the calculated range (Fig. 4). To further investigate the cosmic irradiation component in the neon isotopes, we plot (3He/21Ne)c versus (22Ne/21Ne)c (similar to the Bern-plot after Eberhardt et al. 1966; Fig. 5) and compare the measured values with theoretical values calculated with the data from Leya et al. (2004) and with the isotopic ratios calculated with the formula for diogenites after Eugster and Michel (1995). Chemistry is taken from the paired sample RBT 04262 (Anand et al. 2008) and from Treiman et al. (1994) for ALHA77005. For RBT, all subsamples match the (22Ne/21Ne)c range for a small object, but fall off the theoretical values toward lower (3He/21Ne)c. This might indicate that helium loss has occurred after cosmic ray exposure. Such loss has been reported for chondrites (Scherer et al. 1994). In contrast, one subsample of ALHA77005 falls onto the theoretical values for a small (up to 10 cm diameter) object. The two other subsamples are within range in their (3He/21Ne)c, but seem to have “too much” 22Ne, which is similar to the observation in the neon three isotopes plot. Cosmic ray exposure ages derived from 21Ne are in agreement with the 3He age derived from the most interior part for RBT 04261, which would confirm preferential helium loss from the outer parts of the meteorite. For ALHA77005, the rind sample—the one in best agreement with the calculated cosmic ray exposure isotopic ratios—gives 4.39 Ma for T21. This is higher than exposure ages reported in the literature as 2.45 (Miura et al. 1995) and 1.9 Ma (Bogard et al. 1984), and the preferred value by Eugster et al. (2002) of 3.6 ± 0.7 Ma, but it is in accordance with our T3, which itself is within the literature range for this meteorite. We therefore prefer T21 = 4.4 Ma. Combined with the T3 of the same sample, this results in an exposure age of 4.5 Ma from helium and neon.

Figure 4.

20Ne/22Ne versus 21Ne/22Ne for all three samples from RBT 04261 and ALH77005. Earth atmospheric ratios plot within the range of measured Martian values. The trapped components air (gray star) and Martian atmosphere (gray dot; Bogard and Garrison 1998; Swindle 2002). The shaded area indicates the range of results for Martian atmospheric Ne given in the literature, which ranges from 20Ne/22Ne = 7 to 11 (Swindle et al. 1986; Wiens et al. 1986; Pepin 1991; Ott and Löhr 1992; Bogard and Garrison 1998; Mohapatra et al. 2003; Park and Nagao 2006; Schwenzer et al. 2008). Also shown is the SCR-Ne signature (curved gray line; after Garrison et al. 1995, data from Reedy 1992) as well as the spallation field for chondrites. Because the isotopic ratios as produced from Na spallation can account for chemical differences between chondrites and Martian meteorites, dotted lines indicate the spallation as calculated with Martian meteorite chemistry and the spallation data of Leya and co-authors (see text).

Figure 5.

Plots adapted from the “Bern-plots” of Eberhardt et al. (1966) for ALHA77005 (left) and RBT 04261 (right). Solid and dashed lines show isotopic ratios as predicted from model production rates using the production rates from Leya et al. (2004) and chemistry from Treiman et al. (1994) for ALHA77005 and Anand et al. (2008) for RBT 04261. Dotted lines are calculated after the formula of Eugster and Michel (1995) for diogenites. High (22Ne/21Ne)c indicates a trapped component (Martian atmosphere or terrestrial air) that has not been entirely corrected. Low (3He/21Ne)c indicates helium loss.

Heavy Noble Gases

We used two subsample splits for the heavy noble gases. One was irradiated and used for argon measurements, and the other remained unirradiated and was used for argon, krypton, and xenon measurements.

The irradiated argon measurements were carried out as single grain fusion on exterior and interior subsample splits (Table 4). Figure 6 shows histograms of single grain apparent ages and thus provides a first overview of the measured data. The crystallization age for ALHA77005 is 154 ± 6 Ma (Jagoutz 1989), and that for RBT 04261 is 174 ± 14 Ma (Sm-Nd isochron age; Shih et al. 2009). The most common apparent ages are 355 Ma and 189 Ma for ALHA77005 exterior and interior, and 246 Ma and 253 Ma for RBT 04261 exterior and interior. Note that those apparent ages are higher than the crystallization ages. Ar-Ar ages higher than crystallization ages obtained by stepped heating, and the frequently observed even higher than average apparent ages are a result of excess argon incorporation from both the Martian and the terrestrial atmospheres. Plots of 36Ar/40Ar versus 39Ar/40Ar (Fig. 7) and 36Ar/38Ar versus 1/38Ar (Fig. 8) of single grain fusion measurements of irradiated samples of ALHA77005 and RBT 04261 show no difference between exterior and interior samples, but clearly a Martian atmospheric signature is present in ALHA77005. Our highest 40Ar/36Ar measurement on ALHA77005 is 1382 ± 20. If this value is treated by the same partitioning procedure for trapped and cosmogenic gases (ignoring production of 38Ar from chlorine in the reactor, which is low based on the 38Ar/36Ar versus 37Ar/36Ar plot), an 40Ar/36Artr ratio of 1556 ± 89 results, which is lower than found for the Martian atmospheric contribution in this meteorite (approximately 1700–1800; Bogard and Garrison 1999). This maximum ratio obtained from single grain fusion is higher than any sum of unirradiated sample measurements, but it is lower than the high temperature steps measured on rind and interior subsamples (Table 5). The latter are in excellent agreement with the Martian atmospheric component of approximately 1700–1800, which demonstrates that air contamination of argon in ALHA77005 unirradiated samples can be corrected for by stepwise heating of interior samples.

Table 4. Results of the argon measurements on irradiated portions of the sample. Exterior and interior samples are subsamples 34 and 36 for RBT 04261 and 225, and 227 for ALHA77005. Abundances are in 10−10 ccSTP
  1. J-values are 1.331008E-2 ± 6.655E-05 for ALHA77005 and 9311325E-3 ± 4.65566E-05 for RBT 04261.

RBT 04261,34
RBT 04261,36
Table 5. Results of the argon measurements on unirradiated portions of the sample. Exterior, rind, and interior samples are subsamples 34, 35, and 36 for RBT 04261 and 225, 226, and 227 for ALHA77005. Concentrations are in 10−10 ccSTP g−1
RBT 04261
RBT 04261,34
RBT 04261,35
RBT 04261,36
Figure 6.

Apparent ages measured in irradiated exterior and interior subsamples of ALHA77005 (left) and RBT 04261 (right). Crystallization age for ALHA77005 is 154 ± 6 Ma (Jagoutz 1989), and for RBT 04261, 174 ± 14 Ma (Sm-Nd isochron age; Shih et al. 2009). Highest bar is centered at 355 and 189 Ma for ALHA77005 exterior and interior, and at 246 and 253 Ma for RBT 04261 exterior and interior. Note that these are higher than the crystallization ages and that much higher apparent ages are frequent, because of excess argon from both the Martian and the terrestrial atmospheres.

Figure 7.

36Ar/40Ar versus 39Ar/40Ar of single grain fusion measurements of irradiated samples of ALHA77005 (left) and RBT 04261 (right). Note that there is no difference between exterior and interior samples on this plot. The 39Ar/40Ar ratio corresponding to an age of 200 Ma is indicated by a gray arrow.

Figure 8.

36Ar/38Ar versus 1/38Ar of single grain fusion measurements of irradiated samples of ALHA77005 (left) and RBT 04261 (right). Note that there is no difference between exterior and interior samples on this plot. Only ALH77005 samples show a clear indication of the presence of Martian atmosphere.

Heavy noble gas concentrations measured in nonirradiated samples show clear and uniform patterns for all noble gases, except xenon in RBT 04261 (Tables 5 and 6; Fig. 9). For both samples, the exterior sample has the most trapped 36Ar and the lowest 40Ar/36Artr ratio (which point toward terrestrial air contribution), whereas the rind and interior samples are similar. 84Kr amounts are also highest in the exterior splits of both meteorites. But for 132Xe, only ALHA77005 shows a clear trend toward lower abundances and higher isotopic ratio (i.e., Martian atmospheric contribution) from exterior to interior. RBT 04261 has generally higher amounts of Xe than ALHA77005, but exterior and interior splits are fairly similar. Note that Fig. 9 shows sums of measurements only, temperature steps are given in Tables 5 and 6; Fig. 10. Our RBT 04261 data compare well with the extensive data set of Cartwright et al. (2010) who measured mineral separates, i.e., six olivine-pyroxene mixtures (ol-pyx), two maskelynite separates (mask), and one opaques separate (Fig. 11). Although their mineral separates allow for a clearer insight into the siting of the Martian atmospheric and interior components, both data sets reveal the high and serious contamination of RBT 04261 by terrestrial air.

Table 6. Results of the krypton and xenon measurements. Exterior, rind and interior samples are subsamples 34, 35, and 36 for RBT 04261 and 225, 226, and 227 for ALHA77005. Results are as measured and not corrected for cosmogenic contribution. Concentrations are in 10−12 ccSTP g−1
RBT 04261
RBT 04261,34
RBT 04261,35
RBT 04261,36
Figure 9.

40Ar/36Artr versus 36Artr, 86Kr/84Kr versus 84Kr, and 129Xe/132Xe versus 132Xe [ccSTP g−1]. Exterior, rind, and interior samples are subsamples 34, 35, and 36 for RBT 04261 and 225, 226, and 227 for ALHA77005. Note that for both samples, the exterior sample has the most trapped 36Ar and the lowest isotopic ratio, whereas the rind and interior samples are similar. For 84Kr amounts are highest in the exterior splits of both meteorites, too, but for 132Xe, only ALHA77005 shows a clear trend to lower amounts and higher isotopic ratio from exterior to interior. RBT 04261 has generally higher amounts of Xe than ALHA77005, but exterior and interior are fairly similar.

Figure 10.

Plot of the isotopic ratios 129Xe/132Xe and 136Xe/132Xe versus the elemental ratio of 84Kr/132Xe. The plot allows distinguishing between the endmember components Martian atmosphere (high 129Xe/132Xe), Martian interior (Ott 1988), and terrestrial air (Ozima and Podosek 2002). While the latter two have a similar 129Xe/132Xe ratio, they are distinguishable in their 136Xe/132Xe (Swindle 2002). The distinct isotopic signatures allow us to judge the nature of the elemental ratio: Martian interior has a 84Kr/132Xe of approximately 1. However, elemental fractionation of air noble gases can lead to a similarly low value (Mohapatra et al. 2009). While the high-T-steps and the sums of interior samples of ALHA77005 show Martian atmospheric component, RBT 04261 is entirely terrestrial air in its composition. Exterior, rind, and interior samples are subsamples 34, 35, and 36 for RBT 04261 and 225, 226, and 227 for ALHA77005. Note that the errors of the RBT 04261 measurements (sum of T-steps) are such that the error bars are smaller than the symbol plotted.

Figure 11.

Plot of the isotopic ratios 129Xe/132Xe and 136Xe/132Xe versus the elemental ratio of 84Kr/132Xe of RBT 04261 (this study) and mineral separate stepwise heating experiments of Cartwright et al. (2010). Note the high 129Xe/132Xe (Martian atmosphere) in the maskelynite separates (mask), which points at the source for the Martian atmospheric component in the 1000 °C measurement of our bulk rocks. Most data points in both measurements, however, are comparable to the signature of terrestrial air. Martian interior from Ott (1988) and terrestrial air from Ozima and Podosek (2002).

Noble gases in ALHA77005 have been measured repeatedly (Jessberger et al. 1981; Schaeffer and Warasila 1981; Bogard et al. 1984; Swindle et al. 1986; Nagao 1987; Garrison et al. 1995; Miura et al. 1995; Bogard and Garrison 1998, 1999; Eugster et al. 2002). In a Xe-3-isotopes plot of 129Xe/132Xe versus 136Xe/132Xe, our data for the exterior sample fall onto the terrestrial air signature, and so do the data by Swindle et al. (1986) and Miura et al. (1995). In contrast, mineral separates from an earlier study (Schwenzer 2004; Schwenzer et al. 2006) show contributions of Martian atmosphere in an olivine separate, whereas the pyroxene separate is contaminated by air (Fig. 12).

Figure 12.

Plot of the isotopic ratios 129Xe/132Xe and 136Xe/132Xe versus the elemental ratio of 84Kr/132Xe of ALHA77005 (this study), mineral separates from our earlier study on ALHA77005 (Schwenzer 2004; Schwenzer et al. 2006) and stepwise heating experiments of Swindle et al. (1986) and Miura et al. (1995). Martian interior from Ott (1988), Martian atmosphere from Swindle (2002), and terrestrial air from Ozima and Podosek (2002).


Good News from Mineralogy, Oxygen, and Light Noble Gases

We observed alteration minerals in ALHA77005, but not in RBT 04261. Primarily, this may reflect the difference in terrestrial residence age: <60 ka for RBT 004261 (Nishiizumi and Caffee 2010) and 210 ± 70 ka for ALHA77005 (Nishiizumi et al. 1994). But find location is another factor to be considered: Losiak and Velbel (2011) found significant differences in evaporite formation for meteorites from different fields. The RBT field has the lowest percentage of meteorites with evaporites (1.8%); the ALH field is intermediate with 5.2%. For comparison, the highest percentage of meteorites with evaporites occurs in the GRO field with 13.5%. Both observations lead to the expectation that air contamination in the noble gas budget should be less severe in RBT 04261 than in ALHA77005. However, RBT 04261 does contain more open fractures from the impact shock than ALHA77005, which contains melt (Treiman et al. 1994; Usui et al. 2010). Fewer open fractures such as in ALH77005 compared with RBT 04261—equaling fewer pathways—might help in preventing the interior from exposure to air, thus limiting contamination in the interior.

The oxygen isotope data for both samples suggest that they have experienced only minor terrestrial weathering. However, we note that oxygen isotope values of meteorite samples with low metal and sulfide contents (such as SNCs) are insensitive to alteration, and may have to undergo major levels of weathering to produce significant shifts in isotope compositions. This is in contrast to metal and sulfide-rich meteorites such as chondrites and primitive achondrites. Greenwood et al. (2012) analyzed leached and unleached sample pairs of metal-rich primitive achondrites and found that oxygen isotope compositions were disturbed even where alteration levels were quite low. Note that their finding contrasts with the model of Bland et al. (2000) who suggested that chondrites must be subjected to significant levels of alteration before there is a significant shift in their oxygen isotope composition. The contrasting behavior of SNCs and metal and sulfide-rich meteorites reflects the low stability of metal and sulfides during terrestrial weathering. The initial stages of meteorite weathering involve the conversion of metal and sulfide into iron oxides and oxyhydroxides (Wlotzka 1993; Bland et al. 2006). With respect to oxygen isotopes, the important feature of this process is that terrestrial oxygen becomes fixed into such samples at an early stage. In contrast, weathering of metal and sulfide-poor meteorites such as SNCs and HEDs requires alteration of silicates to take place for there to be distinct shifts in oxygen isotope compositions. Thus, in contrast to chondrites and primitive achondrites, in the case of differentiated achondrites, oxygen isotopes are less useful indicators of terrestrial weathering.

In the case of helium, no difference between rim and interior of ALHA77005 has been found and our values are within the range of previous results, although higher than expected from theoretical calculations (see the 'Results' section). This is in contrast to observations on chondrites (e.g., Scherer et al. 1994), where loss of helium (but also other light noble gases) has been reported. But it is in accordance with our mineralogical and oxygen isotope results: silicates as the carrier phases of both helium isotopes are not noticeably affected by the weathering. In contrast, the 3He content in RBT 04261 increases by about 21% from exterior to rind, despite being only few mm apart. The increase of 3He from exterior to interior is 55%. 3He production does increase with depth within the first few cm, but the difference is much smaller and measured concentrations are too high for the assumption we deduced from our neon measurements, i.e., that RBT 04261 was irradiated as a small object in space. If we take the match of our interior sample with the calculated ratio, then RBT 04261 was part of a larger piece in space, but must have lost a considerable amount of 3He in its exterior and rind portions. Especially for the exterior of this fusion crusted stone, this might be a result of entry-heating and weathering. Consequently, we conclude our T3 cosmic ray exposure age to be that of the interior sample, i.e., 2.7 Ma (Table 2). This matches our T21 of 2.9 Ma, is in excellent agreement with the exposure age of 2.9 Myr by Nishiizumi and Caffee (2010), but is slightly higher than the 2.1 Ma calculated from noble gas isotopes for RBT 04262 (Nagao and Park 2008). 4He should have approximately similar diffusion behavior to 3He, so should be victim of the same loss processes. However, its concentrations decrease significantly inward, for which the only explanation is that of siting of the two isotopes in different minerals. 3He is produced on main elements (O, Si, Al, Mg, Ni, Fe, in order of decreasing production rate; Leya et al. 2004) and distributed evenly in the main minerals, but radiogenic 4He is the product of U and Th decay, thus located in accessory phases, e.g., apatite and zircon. We exclude surface contamination by 4He itself, because of its adsorption coefficients (Ozima and Podosek 2002). Another source could be contaminant U and Th. Approximately 1 × 10−5 g U per g sample or 2.5 × 10−5 g Th per g sample would be required to produce 1 × 10−8 ccSTP g−1 4He, if a residence time of 10,000 yr is assumed since the contamination happened. This U and Th concentration is several orders of magnitude higher than the bulk rock concentration. Therefore, the 4He results on RBT 04261 seem to be best explained by heterogeneous distribution of the U-Th carrier phases, to allow for a sufficient decay time of the U and Th present in the phosphates. As RBT 04261 has two lithologies (Usui et al. 2010), the apparent correlation might be an artifact of sampling one lithology in the exterior and rind samples, which stem from nearby locations, and sampling the other lithology for the interior sample, which was taken at a distance from the first two.

Earth's Fingerprint in Heavy Noble Gases


As a first indicator for adsorbed heavy noble gases, we compare the amount of trapped 36Ar with the 40Ar/36Artr ratio. Trapped 36Ar (36Artr) can result from the incorporation of either Martian or terrestrial atmosphere, depending on the accompanying 40Ar/36Artr ratio. In both cases, the exterior sample (RBT 04261,34 and ALHA77005,225) shows higher concentrations of trapped 36Artr and a lower isotopic ratio (Fig. 9). Further information comes from the data obtained from Ar-isotopic measurements of irradiated samples. They return apparent ages (uncorrected for terrestrial argon), which are higher than the crystallization ages of the meteorites, if the measured grain contains Martian or terrestrial atmospheric argon. The apparent ages show a wide variation in all samples; however, there is no clear difference in age distribution between the exterior and interior samples of any of the two meteorites (Fig. 6). But there is a difference between samples: apparent ages of ALHA77005 reach much higher values than in the RBT 04261 samples, indicating that a major source for the trapped component in ALHA77005 might be Martian atmosphere. The apparent ages for RBT 04261 range from 210 to 2340 Ma in the exterior and from 192 to 2647 Ma in the interior sample, presenting a wider variety in trapped component(s) concentration than the sample of Cartwright and Burgess (2011): approximately 310–1355 Ma on mineral separates. The 40Ar/36Ar and 38Ar/36Ar ratios vary widely in both meteorites (Table 5). Again there is a difference between samples: ALHA77005 36Ar/38Ar ratios indicate Martian atmospheric contribution, whereas RBT 04261 shows no evidence of contamination (Fig. 8).

Combining the Ar datasets obtained on irradiated samples (at The Open University) and unirradiated samples (at Max-Planck Institute Mainz) allows further insights into contaminating species because irradiation produces additional 40Ar on potassium and 38Ar on chlorine in the samples. For ALHA77005, a plot of the isotopic ratios 40Ar/36Ar versus 38Ar/36Ar (Fig. 13) shows that the first temperature steps of the unirradiated samples fall onto the terrestrial air point, whereas higher T-steps and interior samples deviate from air toward Martian atmosphere and cosmic ray contribution. Our values from interior samples are in accordance with the measurements by Bogard et al. (1984), whereas the exterior samples resemble those of Miura et al. (1995), supporting their interpretation of air contamination of their sample split. The irradiated samples are much higher in 38Ar/36Ar than the unirradiated samples and show a trend for grains with high 38Ar/36Ar to also have high 40Ar/36Ar. In addition, they form two groups, where the group with the highest 38Ar/36Ar has a much higher ratio than would result from cosmic irradiation only. We therefore conclude that this group of points was measured on grains carrying terrestrial alteration. Note that all irradiated grains have higher 38Ar/36Ar than the unirradiated grains, but the 40Ar/36Ar ratios are similar. This suggests that the irradiation contribution of terrestrial chlorine from the observed alteration has a larger effect on the 38Ar/36Ar than K contamination on the 40Ar/36Ar, because of the much higher abundance of 40Ar (especially where Martian atmosphere is present) than 38Ar. Our finding of two distinct groups, wherein only a few individual grains seem to carry larger amounts of chlorine, is in accordance with previous observations of existing but minor chlorine contribution to 38Ar in stepwise heating experiments (Bogard and Garrison 1999; Bogard et al. 2010), and the observation of air contamination alongside chlorine (observed as 38Ar) in other subsamples (Swindle et al. 1986).

Figure 13.

Plot of the isotopic ratios 40Ar/36Ar versus 38Ar/36Ar for the irradiated data set obtained at The Open University (OU) and the unirradiated data set obtained at Max-Planck Institute for Chemistry, Mainz (MPI). Irradiation produces additional 38Ar from Cl, and 40Ar from K. Plot for ALHA77005 on the left, and for RBT 04261 on the right. Note the different scales of the two plots in 38Ar/36Ar, whereas 40Ar/36Ar cover a similar range. Note also that the highest 38Ar/36Ar in ALHA77005 form a separate group and all have high 40Ar/36Ar.

The Ar contamination for RBT 04261 contrasts with ALH77005. The unirradiated samples blend in with the irradiated samples suggesting, at a first glance, no major K or Cl contamination. Moreover, no single grain measurement exceeds (taking uncertainties into account) the 38Ar/36Ar for cosmic irradiation; and the trend observed in ALHA77005 is absent. Instead, a clear difference between exterior and interior is visible, the exterior having lower 38Ar/36Ar. This points toward a higher contribution of terrestrial atmosphere in the grains from the exterior compared with the grains from the interior. 40Ar/36Ar is generally lower than in ALHA77005, but shows a clear indication of Martian atmosphere. Moreover, two trends are visible in our data set. This was observed before for RBT 04261 by Cartwright et al. (2010), who studied mineral separates and therefore could assign the high K and Martian atmosphere contribution to maskelynite, whereas a high 38Ar/36Ar, interpreted as chlorine, was found in the olivine-pyroxene fraction. The data set presented here (Fig. 13) is in accordance with this observation and interpretation.

Krypton and Xenon

84Kr abundance is highest in both exterior samples and decreases inward in all samples (Fig. 9), which—like the Ar results—points toward air contamination of the exterior, decreasing inward, for both meteorites. The story for xenon is the same only for ALHA77005, which shows highest abundances in the exterior sample and much less 132Xe in the rind and interior. This decrease in abundance is accompanied by a sharp increase in the 129Xe/132Xe ratio, clearly demonstrating that air incorporation with a low isotopic ratio is masking the Martian atmosphere's high isotopic fingerprint in the exterior sample. In contrast, RBT 04261 shows an even lower 129Xe/132Xe ratio in all three subsamples, and the highest amount in the interior sample.

To analyze the krypton and xenon results further, they are plotted on a diagram 129Xe/132Xe versus 84Kr/132Xe (Fig. 10). The three endmembers, Martian atmosphere (Swindle 2002), terrestrial atmosphere (Ozima and Podosek 2002), and Martian interior (Ott 1988), are clearly distinguishable on this plot. The data for ALHA77005 follow the pattern expected from the data analysis so far: rind and exterior are dominated by Martian gases, falling close to the mixing line between Martian interior and Martian atmosphere. Note especially that the high temperature steps (compare Fig. 10 and Table 6) fall onto this mixing line. The exterior sample, however, falls close to but toward the right of the Martian interior endmember. Looking at individual temperature steps reveals that they fall onto or below the mixing line between air and Martian interior. We have shown earlier for hot desert meteorites (Mohapatra et al. 2009) that especially low-T steps can be contaminated by air that is elementally fractionated (EFA). As the 129Xe/132Xe ratio is that of air (0.98) within error in the first, and dominating temperature step, we conclude that the exterior sample is dominated by EFA. In summary, the exterior sample is dominated by fractionated air contamination to the extent that any Martian signature becomes invisible, whereas in the rind and interior sample, the Martian component is obvious. The interior sample itself has a well-established Martian fingerprint in the 1400 °C step (Table 6) and could thus be used to study Mars.

The picture is different for RBT 04261. All three subsamples (sums of T-steps) fall below the mixing line between Martian interior and air (Fig. 10). Note that the precision of our measurement is good enough to distinguish both components from each other. To check whether the RBT 04261 sample really is pervasively contaminated by elementally fractionated air and no Martian interior component is present, we show a plot 129Xe/132Xe versus 136Xe/132Xe (Fig. 12), which allows us to distinguish between those two components. The exterior of ALHA77005 and all RBT 04261 results plot around air, not Martian interior. Also visible in this plot is a fission contribution on 136Xe, shifting the rind and interior results toward a higher 136Xe/132Xe ratio than expected from pure Martian interior–atmosphere mixing. This has been observed previously for ALHA77005 (Fig. 12; Swindle et al. 1986; Mathew et al. 1998).

Thus, the exterior subsample of ALHA77005 and all three subsamples of RBT 04261 show terrestrial contamination to an extent that Martian components become invisible. Any attempt to correct for the incorporated contamination would result in very large errors, especially as the Martian components are atmosphere and interior, and the cosmogenic contribution complicates the picture further. Unfortunately, the data show that stepwise heating cannot solve the problem of such intense contamination as observed in RBT 04261.

At this point, there is a conundrum to solve: ALHA77005 has the longer terrestrial residence time and shows visible alteration in its outermost approximately 300 μm, whereas RBT 04261 appears unaltered. ALHA77005 does show signs of K or Cl contamination in its irradiated Ar isotopic signature, while RBT 04261 has no such contamination, but RBT 04261 is pervasively contaminated by terrestrial air and ALHA77005 is not. There are two potential explanations, both of which are concerned with permeability: ALHA77005 has experienced much higher shock pressures (50 GPa; Fritz et al. 2005) compared with RBT 04261, for which we estimate—based on He loss (see above)—a shock pressure of 30–35 GPa. Consequentially, ALHA77005 contains shock melt, which occurs in melt pockets and vein fillings, while RBT 04261 only has open fractures. The heavily fractured nature of RBT 04261 might have enhanced the permeability of the rock and thus air contamination is more severe and reaches the inside of the stone. Another factor reducing the permeability for air and water is the deposition of alteration minerals in the outside of the ALHA77005 meteorite. Thus, the seemingly better conditions from less observed weathering in the RBT than the ALH Antarctic ice sheets (Losiak and Velbel 2011) could result in a disadvantage for noble gas measurements, because alteration minerals in the outside portions seal the pathways for gas exchange. It also supports the assumption that the heavy noble gases are adsorbed onto surfaces, potentially aided by moisture during freeze-thaw cycles, independently of the occurrence of evaporite minerals. As fractionation is enhanced by repeated adsorption and desorption, the open-fracture network of RBT 04261 might have enhanced noble gas exchange compared with the more sealed nature of the fractures in ALHA77005. In any case, no clear correlation between the observed mineralogy and noble gas contamination can be deduced.

Considering Mars, the effects observed in cold desert weathering on Earth can inform us about effects that might be responsible for the fractionated Martian heavy noble gas signatures observed in the nakhlite meteorites (Ott 1988; Drake et al. 1994; Gilmour et al. 1999; Swindle 2002; Mathew and Marti 2005). As fractionation of heavy noble gases has been observed for Martian meteorites from hot (Mohapatra et al. 2009) and cold (Schwenzer et al. 2009; this study) deserts in a very similar way, it can be assumed that the difference in atmospheric pressure, temperature, and moisture between the different locations on Earth does not result in a major difference in adsorption and dissolution behavior. Therefore, incorporation through adsorption and mineral formation on Mars should also lead to fractionated noble gas signatures in Martian meteorites. Incorporation of unfractionated noble gas signatures, such as observed in impact melt glasses in shergottites, needs a different mechanism, which has been demonstrated to be the shock process itself (Bogard et al. 1986, 1987, 1989).

Application of the above arguments to Mars suggests that incorporation of fractionated Martian atmosphere is a potential mechanism to explain the observations in the nakhlite and ALHA84001 meteorites. The presence of Martian alteration phases has been observed in the nakhlites (Treiman 2005; Changela and Bridges 2010), and fractionated Martian atmosphere has been found in them (Ott 1988; Drake et al. 1994; Gilmour et al. 1999; Swindle 2002; Mathew and Marti 2005). Therefore, if the secondary mineral forming fluid ever had contact with the Martian atmosphere, this would directly lead to incorporation of fractionated Martian atmospheric signatures into the nakhlite Martian meteorites. Lastly, the low 84Kr/132Xe ratio in ALH 84001 (Murty and Mohapatra 1997; Gilmour et al. 1998; Miura and Sugiura 2000; Mathew and Marti 2001) could similarly have its source in the alteration processes that caused carbonate precipitation (Harvey and McSween 1996; Kring et al. 1998). Terrestrial contamination, a nuisance when studying Martian signatures in shergottites, therefore can provide an excellent analog for understanding the nakhlite (and potentially ALH 84001) fractionated Martian atmospheric signatures.


Our study shows that terrestrial loss or contamination in the noble gas signatures of any Martian meteorite is dependent on a variety of factors and is not always directly related to the presence of secondary minerals. Helium is least affected, but has been found to be lost during Antarctic weathering. Neon adsorption is also minor, but might affect the component partitioning into cosmogenic and trapped fractions, and thus the determination of the cosmic ray exposure age. Oxygen is known to be a sensitive indicator of terrestrial weathering in iron metal-bearing meteorites, with oxide formation rapidly incorporating a terrestrial oxygen component into the sample. Here, it is almost unaffected, because silicate weathering is a prerequisite for changing oxygen isotopic signatures in meteorites, which only contain accessory sulfides and no metal.

The heavy noble gases are most seriously affected by alteration, in some subsamples to the extent that no Martian signatures can be detected anymore. Our results demonstrate that a visibly clean interior portion of a meteorite is not necessarily free from terrestrial contamination in the heavy noble gases. RBT 04261, which is essentially free from any visible alteration at electron microprobe investigation scale, is contaminated by fractionated terrestrial noble gases (EFA) throughout, with the most serious effects on xenon. Even in stepwise heating experiments, the amount and degassing behavior of this contamination seriously compromise the interpretation of the Martian signature in the meteorite.

Nevertheless, the effects of cold desert weathering on Earth are key indicators of the potential for weathering effects upon Martian heavy noble gas signatures observed in the shergottite and nakhlite Martian meteorites. If heavy noble gases fractionate upon incorporation through adsorption or mineral formation on Earth, they are expected to do so on Mars. If we assume that the differences in temperature and moisture between Mars and Earth do not result in a major difference in adsorption and dissolution behavior, then incorporation through adsorption and mineral formation should lead to fractionated noble gas signatures in Martian rocks. Unfractionated incorporation, such as observed in impact melt glasses in shergottites, requires a different mechanism such as the shock process itself. Turning the arguments around, from the presence of alteration phases, as observed in the nakhlites and ALH 84001, one should expect fractionated Martian atmospheric signatures in them, if the secondary mineral-forming fluid ever had contact with the Martian atmosphere.


This study was only possible through careful and project-specific sampling of the two meteorites. We thank NASA and the Meteorite Working Group for allocating the samples to us, and especially Kevin Righter for his advice in meteorite selection. We also thank Sarah Sherlock and Alison Halton for their support in the Ar-laboratory. We acknowledge reviews by Jisun Park and Tim Swindle. I. A. F., R. C. G., and J. M. G. thank STFC for Rolling Grant funding (Grant no. ST/I001964/1).

Editorial Handling

Dr. A. J. Timothy Jull