Meteoritics & Planetary Science 48, Nr 10, 1958–1980 (2013) doi: 10.1111/maps.12211

“Sweating meteorites”—Water-soluble salts and temperature variation in ordinary chondrites and soil from the hot desert of Oman

Florian J. ZURFLUH1*, Beda A. HOFMANN2, Edwin GNOS3, and Urs EGGENBERGER1

1Institut fur€ Geologie, Universitat€ Bern, Baltzerstrasse 1 + 3, Bern CH-3012, Switzerland 2Naturhistorisches Museum der Burgergemeinde Bern, Bernastrasse 15, Bern CH-3005, Switzerland 3Museum d’histoire naturelle de la Ville de Geneve, 1 Route de Malagnou, CP 6434 CH-1211 Geneve 6, Switzerland *Corresponding author. E-mail: zur0fl[email protected] (Received 20 June 2012; revision accepted 05 September 2013)

Abstract–The common appearance of hygroscopic brine (“sweating”) on ordinary chondrites (OCs) from Oman during storage under room conditions initiated a study on the role of water-soluble salts on the weathering of OCs. Analyses of leachates from OCs and soils, combined with petrography of alteration features and a 11-month record of in situ meteorite and soil temperatures, are used to evaluate the role of salts in OC weathering. 2+ 2 + Main soluble ions in soils are Ca ,SO4 ,HCO3 ,Na , and Cl , while OC leachates are 2+ 2+ 2 dominated by Mg (from meteoritic olivine), Ca (from soil), Cl (from soil), SO4 (from meteoritic troilite and soil), and iron (meteoritic). “Sweating meteorites” mainly contain Mg2+ and Cl. The median Na/Cl mass ratio of leachates changes from 0.65 in soils to 0.07 in meteorites, indicating the precipitation of a Na-rich phase or loss of an efflorescent Na-salt. The total concentrations of water-soluble ions in bulk OCs ranges from 600 to 9000 lgg1 (median 2500 lgg1) as compared to 187–14140 lgg1 in soils (median 1148 lgg1). Soil salts dissolved by rain water are soaked up by meteorites by capillary forces. Daily heating (up to 66.3 °C) and cooling of the meteorites cause a pumping effect, resulting in a strong concentration of soluble ions in meteorites over time. The concentrations of water-soluble ions in meteorites, which are complex mixtures of ions from the soil and from oxidation and hydrolysis of meteoritic material, depend on the degree of weathering and are highest at W3. Input of soil contaminants generally dominates over the ions mobilized from meteorites. Silicate hydrolysis preferentially affects olivine and is enhanced by sulfide oxidation, producing local acidic conditions as evidenced by jarosite. Plagioclase weathering is negligible. After completion of troilite oxidation, the rate of chemical weathering slows down with continuing Ca-sulfate contamination.

INTRODUCTION Langenauer and Krahenb€ uhl€ 1993) and the efflorescence of salts on meteorite surfaces (e.g., Jull et al. 1988; Water-soluble salts were found in the ordinary Velbel 1988; Velbel et al. 1991; Losiak and Velbel chondrites (OCs) Monahans and Zag that both fell in 2011). In hot deserts, daily temperature variations are 1998 (Grossman 1998, 1999). The halite and sylvite much larger and liquid water and water-soluble salts are grains found in these meteorites were identified to be of much more abundant and play an important role during extraterrestrial origin (e.g., Rubin et al. 2002). So far, weathering of meteorites (e.g., Al-Kathiri et al. 2005; these are the only OCs from which extraterrestrial salts Bland et al. 2006). However, the only investigation were described. However, meteorites are subject to where salt contamination in hot desert meteorites was terrestrial alteration and contamination from the first studied focused on the halogen contamination of moment they reach the Earth. eucrites and OCs from Western Australia (Krahenb€ uhl€ A few studies focused on the halogen and Langenauer 1994, 1995). So far, a systematic survey contamination of meteorites from Antarctica (e.g., of salts within hot desert meteorites is missing.

© The Meteoritical Society, 2013. 1958 Water-soluble salts in chondrites from Oman 1959

In this study, we have analyzed the water-soluble salt concentration of 15 samples from the JaH 091 meteorite shower (Gnos et al. 2006) and of 16 individual OC samples collected in the hot desert of Oman during the Omani-Swiss meteorite search project. This study was initiated after the common observation of a hygroscopic behavior (“sweating”) of freshly sawn surfaces of Oman meteorites (Fig. 1). We selected samples from various geographic locations and different macroscopic appearances to be able to interpret the variability in salt concentrations. We recognized several types of salt contamination in meteorites (Fig. 2) (1) hygroscopic (“sweating”) meteorites as shown in Figs. 1, 2a, and 2b; (2) “dark green and dry” meteorites, which appear macroscopically relatively unaltered, but generally belonging to W ≥ 3 weathering grade (Wlotzka 1993) showing only minor efflorescence Fig. 1. Cut surface of meteorite JaH 091 (individual 0703-703) of salts (Fig. 2c); (3) “dry rusty meteorites,” typically of with brine droplets “sweating” caused by the presence of weathering degree W4 (Wlotzka 1993) often show hygroscopic salts. Even clear droplets (indicated by arrows) macroscopically visible white pore space fillings are Mg2+ and Cl rich. Some ferrous iron is mobilized from (Fig. 2d). the meteorite interior and oxidized to iron hydroxides on the Special attention was paid to solve the enigma of surface. Note only minor hygroscopic behavior of the dark, slightly leached outermost rim of the meteorite. the “sweating meteorites,” i.e., the hygroscopic behavior of freshly cut meteorite surfaces (Fig. 1). In a relatively humid environment (room condition and 50–60% which is the lowest out of 10 recording stations in relative humidity), salts in meteorites attract water and Oman (Fisher 1994). Additionally, fog may serve as a droplets of brine appear on the surface. This effect is source of moisture in this area. During our fieldwork, described by Nininger (1929) for iron meteorites, but we often observed fog in mornings, causing wetting of otherwise apparently unobserved. The “sweating” of the desert soil surface, especially with wind from iron meteorites is linked to the presence of the mythic southerly directions, i.e., the Arabian Sea. mineral lawrencite (FeCl2) that was claimed to be Daily temperature fluctuations play an important nonexistent (Buchwald and Clarke 1988). Extreme role in the development of the observed salt hygroscopic behavior as seen in Fig. 1 is commonly contamination. In this study, we present data for water- observed on OCs recovered from Oman, but has not soluble salts and temperature variation in an OC from been reported from other deserts like the Sahara, Oman and discuss their role in the weathering of Omani Roosevelt County, or Australia (personal meteorites. communication with persons investigating meteorites from these provenances, e.g., Rainer Bartoschewitz, SAMPLES AND ANALYTICAL TECHNIQUES Alex Bevan, Addi Bischoff, Anne Black, Luici Folco, Marc Jost, Martin Lee, Peter Marmet, Alex Ruzicka, Sample Selection and Leaching Experiments Jochen Schluter,€ and anonymous collectors). Liquid water is a prerequisite to mobilize salts in As base for the study, we used samples collected meteorites and initiate weathering. Precipitation in the during fieldwork in 2007, 2009, and 2010 of the joint Oman inland desert is sparse, but occurs nearly Oman-Swiss meteorite research team. The annually. Maps displaying the distribution of concentrations of water-soluble ions were determined by precipitation show a gradual decrease in mean annual leaching experiments using soil and meteorite samples. precipitation from the coastal regions of Oman toward In addition, six water samples from Oman were inland where typical meteorite recovery surfaces receive analyzed for comparison. In Table 1, an overview of the less than 70 mm per year of precipitation (Edgell 2006). performed experiments is given. Mostly, the rains occur as episodic events during February and March. Meteorological data from Meteorite Samples Yalooni station (19°56′ N, 57°06′ E), the only station All Oman meteorites used here were collected with located within the meteorite recovery area, yielded a tweezers, packed into polypropylene bags, and mean annual precipitation of 38.9 mm for 1979–1990, unpacked in the lab wearing gloves to avoid further 1960 F. J. Zurfluh et al.

Fig. 2. Cut slabs of JaH 091 samples showing the distribution of salts. Images (a) and (b) are from sample 0703-331. a) The major part shows presence of MgCl-rich brine with only minor iron hydroxide. Some spots are free of brine, as indicated by “dry.” b) The same cut surface as (a), but treated with AgNO3 solution. White precipitation of AgCl shows distribution of Cl . c) Sample 0603-239 (JaH 091) has a zoned occurrence of different types of salts. In the middle of the sample, a hygroscopic MgFeCl-rich brine is present. A greenish area surrounds it with efflorescence of whitish, hairy MgSO4, which is not hygroscopic. The outermost part shows no salt efflorescence or hygroscopic behavior and displays a rusty color. d) Meteorite 0702-241 (JaH 091) is a typical W4 sample with white minerals, most likely Ca-sulfates, in pores. After treatment with AgNO3,noCl rich spots are observed. contamination. Samples were cleaned from attached soil produced from meteorite interiors at least one material with pressured air and stored under controlled centimeter below natural surfaces. Parts of the conditions (20 °C, approximately 40% rel. humidity) at meteorites free of calcite veins or cracks were the Natural History Museum Bern. The samples were selected. Typical sample size was about cut using isopropanol as cooling agent to reduce 2 9 1 9 0.8 cm (2.02–7.17 g, mean 4.60 g). After leaching of the soluble pore minerals. weighing, samples were transferred to small Teflon Three experiments consisting of leaching series were flasks and filled with deionized Millipore-Q water performed to determine the water-soluble salts in (meteorite/water ratio approximate 1–4 by mass). meteorites (Table 1) The meteorites were leached over a period of 1. ML091: The aim of the first part of this study was 15 days at room temperature and agitated to investigate the range of water-soluble salts in occasionally to reach equilibrium between the rock meteorites from the large L5 strewn field JaH 091, a and water. single fall (Gnos et al. 2006). Fifteen meteorites 2. MLT: The second section of this study was aimed representing locations from entire approximately to optimize the leaching procedure. Fourteen 50 km long strewn field were selected for the subsamples from one individual stone of the JaH leaching experiment. The size of the individuals 091 strewn field were leached for periods of 10 min, varied from 47.7 to 8.8 kg with an average mass of 1 h, 8 h, 24 h, 4 days, 15 days, and 30 days. For about 120 g. Representative test blocks were each time span, two samples were leached and Water-soluble salts in chondrites from Oman 1961

Table 1. Compilation of experimental work and methods. Leach time Year of Year of Label Samples Goal (days) Analyses sampling experiment ML091 JaH 091, ordinary Water-soluble salts 15 IC, alkalinity, microscopy, 2007 2007/2008 chondrite (L5 S2) contamination in samples SEM, Raman, EMP with identical terrestrial age (strewn field) MLT JaH 091, ordinary Test of leaching experiment Variable IC, alkalinity, ICP-OES 2010 2010 chondrite (L5 S2) set up (time) MLO Ordinary chondrites, Water-soluble salts in OCs 48 IC, alkalinity, ICP-OES, 2009/2010 2011 Oman, Tamdakht, from Oman and Tamdakht microscopy, SEM, H chondrite fall Raman, EMP, XRD SL091 Soil samples, Variation of salts in soil over 7 IC, alkalinity, sieve 2007 2007/2008 JaH 091 region a small area SLO1 Soil samples Salts in soils from Oman 7 IC, alkalinity, sieve 2009/2010 2010 SLO2 Soil samples Salts in soils from Oman 7 IC, alkalinity, sieve 2009/2010 2011 IC = ion chromatography; ICP-OES = inductively coupled plasma - optical emission spectroscopy; Raman = Raman spectroscopy; EMP = electron microprobe; XRD = powder X-ray diffraction; SEM = scanning electron microscope.

analyzed in parallel for control. To limit the of water-soluble salts that might be a source for the ions oxidation of iron, the meteorite cubes were vacuum- introduced into the meteorites (Al-Kathiri et al. 2005). sealed immediately after cutting and transported We have investigated 16 samples from the SL091 series into a N2-filled glove box where they were collected in 2007 and 19 from the SLO1 and SLO2 series transferred to the Teflon flasks. In addition, oxygen- collected in 2009/2010 (Table 1). Soil samples were free deionized Millipore-Q water was used as taken from each surface type distinguished on satellite solvent. images or on ground. In the ideal case, they were 3. MLO: The aim of the third series of analyses was to sampled beneath or in the vicinity of meteorite finds. study the geographic influence on the salt Leaching was performed under laboratory atmosphere contamination of the meteorites. Fifteen samples for 7 days in automatic shakers. Most samples were from meteorites with total masses ranging from leached with solid to liquid ratios of 1:1, i.e., 30 g of soil 144.4 to 4700 g, covering a large spread of find sample and 30 g by weight of deionized Milli-Q water; localities, soil types, weathering grades, and exceptions are noted in Table S1. meteorite types were selected. The sampled meteorites were collected along a profile from the Water Samples coast to the interior. A sample of the H5 S3 W0 Complementary, one rainwater sample and five chondrite fall Tamdakht in Morocco (Chennaoui- surface water samples from the central Oman desert Aoudjehane et al. 2009; Weisberg et al. 2009) was were analyzed. They are possible sources of water for also included. The samples of the MLO series were transporting ions into the meteorites. Two samples are leached for 48 days under the same conditions as from artificial pools between dunes fed by shallow for MLT samples. groundwater, one of reddish color that showed active To visualize the distribution of chloride in 16 precipitation of gypsum and the other from a greenish individuals of the JaH 091 strewn field, cut surfaces part of the pool (Table S2). Three samples were + were treated with a solution of AgNO3.Ag reacts collected at the natural oasis of Wadi Muqshin (Jupp with Cl to form a white precipitate of AgCl (Figs. 2b et al. 2008; Abed et al. 2011), one sample from the and 2c). Dark terrestrial rocks found in the meteorite surface of a brine pool, another at a depth of 20 cm in recovery area were also treated with AgNO3. the same pool, and the third at the bottom of a shallow To obtain data on the temperature fluctuations in an irrigation well (Table S2). Rainwater was collected at Al OC, we performed a field experiment using a meteorite Awad (20°52.696′ N58°11.631′ E) during night of equipped with a temperature logger and placed in a February 7/8, 2010, on a clean plastic foil. typical meteorite recovery area for nearly 1 year. Measurements of Water-Soluble Ions Soil Samples Thirty-five surface soil samples from Oman (fraction The leacheates from soil and meteorite samples <0.15 mm) were leached to determine the concentration were measured by ion chromatography (IC) on a 1962 F. J. Zurfluh et al.

Metrohm 861 compact ion-chromatograph in the Determination of Pore-Filling Minerals laboratory for rock-water interaction at the Institute of Geological Sciences, University of Bern, using Secondary minerals in pores of the meteorites were appropriate standards and blanks for each ion. Highly identified by transmitted and reflected light microscopy, concentrated samples had to be diluted 1:2, 1:10, or Raman spectroscopy, a Jeol JXA-8200 electron 1:100. Before dilution, the samples were filtered microprobe (EMP), and a Zeiss EVO50 scanning (0.2 lm). The leachates from the soils, meteorites, and electron microscope (SEM) with attached energy the water samples were analyzed for Na+,Mg2+,K+, dispersive X-ray spectroscope (EDS) at the Institute of 2+ 2+ 2 Ca ,Sr ,F,Cl,Br,SO4 , and NO3 .Intwo Geological Sciences at the University of Bern. Meteorite + meteorite samples, a peak corresponding to NH4 was samples used for aqueous leaching were studied in detail observed and the concentration was determined for pore-filling minerals. In addition, contamination semiquantitatively. The analytical error is about 5% minerals recorded during classification studies in other based on repeat measurements of standard solutions. meteorites from Oman are also included in this study. The detection limit for most of the anions and cations is Some of the pore space fillings, salt efflorescences 0.5 mg L1. The pH and alkalinity were measured using on the meteorite surfaces, and evaporated brine drops a Metrohm 785 DMP Titrino by titration with HCl. were studied using the SEM. Only JaH 478 contained Alkalinity is later in the text referred to HCO3 . sufficient secondary material for X-ray diffractometric Selected leachates from meteorites were additionally determination. The finely powdered material was analyzed by inductively coupled plasma-optical emission mounted on a silicon wafer and measured with a Philips spectroscopy (ICP-OES) with a Varian 720 ES PW1800 diffractometer. instrument. Meteorite leachates from the time series were analyzed for Na, Mg, Ca, Fe, Co, Ni, and Sr; Temperature Logging meteorites from ML series for Al, Si, Mn, Fe, Co, Ni, Sr, and Ba. Even though the experiments were To study the temperature variations in dark performed in a glove box flushed with N2, iron meteorites on bright soil in the desert of Oman, a hydroxides precipitated during several experiments. All 173.7 g fragment (dimensions 7 9 5.5 9 4 cm) from the ML samples were treated with 5% HNO3 before ICP- JaH 091 L5 S2 chondrite strewn field (stone 0703-416; OES measurement. Blank samples treated the same way (Gnos et al. 2006) was used. A thermocouple was as the experimental samples were measured in the same placed in a borehole of 6 mm diameter and 3 cm depth, run for control. 1 cm beneath the outer surface of the meteorite. The The mass of the solid material, the water for the soil thermocouple was connected with a lHOBO U23-003 leaches, and the wet masses of the meteorites were (version 1.0.9) data logger. This experiment was placed determined with balances to 0.01 g. After the in the Jiddat al Harasis near Hayma (19°49′ N, experiments, the meteorites were dried at 105 °Cto 55°50′ E) with the lower fifth of the meteorite buried in constant weight and the open (connective) porosity was soil. The temperature-dependent resistor was fixed with calculated from the difference of the wet and dry sample thermo conductive paste. A second temperature sensor masses. For the calculation of the concentration of the was placed in an aluminum disk and buried at a depth ions in the pore space, the solid/liquid (S:L) corrected of 30 cm in the soil for recording the temperature values were adjusted using the determined porosity and variation in the soil. Temperature was recorded at a an assumed density of the occupied pore space of rate of three readings/h for the first measuring period 1.2 g cm3. For simplification, we used a bulk density of (208 days) and two readings/h for the second period 3.3 g cm3 for all meteorites, a mean of unaltered H and (374 days). The thermistor of the meteorite was L chondrites (Consolmagno et al. 2008). disconnected in May 2010 during the second period To extend the data set of salt contents in soil probably by an animal bite. samples, a further 63 soil samples were analyzed by using Aquamerck and Merck Mikroquantâ fast tests for RESULTS chlorine contents in addition to samples analyzed by ion chromatography. For evaluating the reliability of the Water-Soluble Salts in Ordinary Chondrites fast test, all IC-analyzed samples were also tested with the fast test. The time series experiment (MLT; Tables 1, S3, Approximately 10 g of sieved soil (fraction Fig. S1) showed that even 30 days is at the low end for <0.15 mm) was mixed with 10 g deionized water and reaching equilibrium between rock and water, especially stirred for 1 min. After 5 min of sedimentation, the for Na+,K+,andCl (Table S3). The other species 2+ 2+ 2 supernatant was used for the test. such as Mg ,Ca , and SO4 apparently reached Water-soluble salts in chondrites from Oman 1963

Fig. 3. Ternary diagrams of the main cation and anion concentrations in various materials. Data displayed in meq L1. Ocean 2+ + water values are from Pilson (1998). The trend for terrestrial samples is from Ca to Na and HCO3 to Cl , respectively, and is indicated by the soil leachates. Meteorites have distinct salt compositions, dominated by the cations Mg2+ and Ca2+. Among 2 the anions, either Cl or/and SO4 is dominant. equilibrium. Therefore, the meteorite samples were Mg2+ from artificial weathering during the experiment is leached for 48 days during the main experiments minor as Tamdakht yielded 4 lgg1, by far the lowest 2 1 (MLO series; Tables 1, S4). By then, most of the value. The same is true for the anions SO4 (58 lgg ) samples showed precipitates and the leaching and Cl (8 lgg1), which are low in Tamdakht. These procedure was stopped. For all meteorites F, Br, Cr, anions can be enriched during weathering and are Cu, Co, and Zn were below detection (Table S4). Sr especially high in SaU 523, a W3 chondrite, yielding and Ba were also mostly below detection. Close to the 3367 lgg1 Cl. The sample with the highest total 1 2 detection limit were Si, Al, Mn, and Ni (Table S4). dissolved ions (UaS 011, W3) has 3939 lgg SO4 The values of Si and Al have to be treated with and 1461 lgg1 Mg2+. The elemental pattern shows 2 caution as these elements are known to occur in scatter, but a slight trend to either Cl or SO4 colloids depending on pH condition. It is difficult to dominated accumulation is observed (Fig. 3). Depending determine the amount of iron involved in weathering. on the degree of weathering, the leachates are dominated It can be either overestimated due to mobilization either by Mg2+ or Ca2+ with generally low contents of during the aqueous extraction or underestimated as it Na+ (Fig. 3). Leachates from samples with hygroscopic precipitated as insoluble iron hydroxide. Leachates of behavior (Figs. 1, 2a, and 2b) are rich in Mg2+ and Cl, RaS 295 and RaS 316 showed a greenish precipitation, while the “dry,” rusty samples (Fig. 2d) are dominated 2+ 2 2+ probably “green rust,” i.e., ferrous (mixed with some by Ca ,SO4 and Mg . Sixteen cut surfaces of L5 ferric) iron hydroxides produced during leaching of chondrites from the JaH 091 strewn field treated with metallic iron. AgNO3 (Figs. 2b and 2d) showed that Cl generally is The patterns of ion concentrations in the analyzed concentrated in those parts of the meteorites that were meteorites are relatively heterogeneous. The totals of not buried in soil during sampling. According to the leached salt concentrations (in solid) range from 596 to visible appearance of weathering and the contamination 8817 lgg1 (Table S4). The Tamdakht meteorite, a patterns, there is no significant difference in the fresh fall, yielded a total of 851 lgg1 of soluble salts, hygroscopic behavior of L and H chondrites. The 15 obviously due to leaching of primary minerals. As most hygroscopic samples (Fig. 1) comprise eight H and expected, Tamdakht yielded the lowest concentrations seven L chondrites, ranging from petrographic grade 4 2+ 2 2+ for Mg ,SO4 , and Ca , otherwise very abundant in to 6 (Van Schmus and Wood 1967), shock grade S1–S5 weathered samples. The measured values of Na+ in (Stoffler€ et al. 1991), and weathering degree W2 to W4 Oman meteorites are in the range of 10 to 485 lgg1 (W3 dominant) (Wlotzka 1993). The masses of the with a mean of 48 lgg1 (Table S4). From Tamdakht, sampled meteorite individuals show a wide range from 151 lgg1 of Na+ was leached. Calcium varies between 47.7 to 8.2 kg and salt efflorescence and hygroscopic 11 and 656 lgg1 with a mean of 217 lgg1; where, behavior was observed in small and big samples. The Tamdakht is among the lowest with 17 lgg1. size of the meteorites does not seem to control the Magnesium, an abundant element in meteorites, shows degree of salt contamination. In the 15 analyzed samples strong variations. But it seems that the contribution of of the JaH 091 meteorite shower, we observed a high 1964 F. J. Zurfluh et al.

Fig. 4. Chloride concentrations in soil, water, and meteorite samples as a function of geographic location. Stars represent meteorites with hygroscopic behavior as displayed in Fig. 1. Diamonds show find localities of meteorites analyzed in the MLO series. Pyramids represent Cl concentration measured in soil samples determined by fast tests. Locations of saline water samples are marked with circles of variable sizes depending on Cl content measured by ion chromatography. Toward the coast of the Arabian Sea (bottom right), the effect of salt derived by sea spray is well visible near Ras Madrakah. Inland depressions are also generally richer in salt (e.g., Umm as Samim). Areas with dense meteorite accumulations such as the Sayh al Uhaymir (SaU), Ramlat as Samah (RaS), and JaH have relatively low concentrations of Cl. variation in absolute and relative concentrations of 6509 lgg1 (median = 324 lgg1) and varies with water-soluble salts, consistent with earlier observations geographic provenance (Fig. 4). of different degrees of weathering in different individuals Similar to soil bulk elemental composition (Al- of a single meteorite shower (Gnos et al. 2009). The Kathiri et al. 2005), the ionic composition of water- concentration range of the ML091 series is similar to the soluble salts from soils is fairly uniform (Fig. 5), but the samples from the MLO series. concentrations are variable. A difference observed is that highly concentrated soil leaches have low Ca2+/ Water-Soluble Salts in Soils and Saline Waters HCO3 ratios of about 0.3, while soils with low salt 2+ concentrations have high Ca /HCO3 ratios up to 40. Chloride concentrations were determined with Usually, highly concentrated samples have higher Aquamerck and Merck Mikroquantâ fast tests for 63 sulfate content. Na+/Cl ratio of higher concentrated soil samples collected in all major meteorite recovery samples is closer to the Na/Cl ratio of halite. Dominant 2+ 2 + areas over much of Oman, with a special focus on a ions are Ca ,SO4 , HCO3 ,Na , and Cl that can cost-inland transect from Ras Madrakah via Hayma be interpreted as being derived from evaporite and toward the dunes of the Rub al K’hali (Fig. 4). The bedrock minerals such as halite (NaCl), gypsum results of the chloride fast test are generally in good (CaSO4*2H2O), and calcite (CaCO3). Potassium, agreement with ion chromatography data. The analyzed magnesium, strontium, and nitrate are present in minor values were corrected for solid to liquid ratio. The Cl amounts, while bromine and fluorine are normally concentration of the soils covers a range from 12 to below detection. The soil samples of the first analyzed Water-soluble salts in chondrites from Oman 1965

Saline waters have seawater (Pilson 1998) like composition and Cl concentrations vary from 12 to 87 g L1 (Table S2). Our measurements of the pool water samples from Muqshin are comparable to data in literature (Jupp et al. 2008; Abed et al. 2011). To test for possible chloride accumulations within terrestrial rocks in the field, drops of AgNO3 solution were applied to broken surfaces of dark terrestrial rocks, found during meteorite searches (sandstones, limestones and cherts with desert varnish, gabbros, rhyolites, and basalts). In contrast to meteorites, none of the inspected terrestrial samples showed obvious precipitation of AgCl, indicating that the strong accumulation of chloride is restricted to meteorites.

Mineralogy of Pore Fillings in Meteorites

A wide variety of secondary minerals formed in Fig. 5. Water-soluble salts in soil samples normalized to pore space of weathered meteorites are known (Jull seawater (Pilson 1998). The elements are sorted by increasing et al. 1988; Velbel 1988; Velbel et al. 1991; Stelzner and median values to the right. Two types with relative uniform Heide 1996; Rubin 1997; Lee and Bland 2004; Al- compositions occur. One type has generally higher + 2+ Kathiri et al. 2005; Bland et al. 2006; Losiak and Velbel concentrations, low Na /Cl , and high Ca /HCO3 ratios, whereas the other type shows opposite values. 2011). Minerals found during this study in meteorites from Oman are listed in Table 2. The secondary minerals can be classified as contamination minerals series (SL091, Table S5) were quite uniform compared (main elements are from outside the meteorite, e.g., salts with the later analyzed (SLO1, SLO2; Table S1, S6). from evaporation) and alteration minerals (the source of Two explanations are plausible: (1) the covered the main elements are primary meteorite minerals, e.g., geographic area was quite small for the SL091 samples alteration products of iron metal or troilite). Some ions and (2) the way of sample collection was slightly present in evaporation products are from decomposition different for the SLO1 and SLO2 samples. SL091 of the meteorites itself (e.g., Velbel et al. 1991). samples represent a surface area of about 40 cm2, while The most abundant secondary minerals, besides the SLO samples are random samples collected over an iron hydroxides and iron oxides, are Ca-sulfates (most approximate area of 10 m2. common is gypsum) and calcite (Table 2). Carbonates Concentrations of water-soluble salts (based on may contain a few wt% of Sr, Mg, or Fe. Ca-sulfates assumed filling of pore space during rain events, liquid typically occupy pore spaces, whereas calcite mostly to solid approximately 1:4) in soils are low compared forms veins penetrating the meteorites. Iron oxides and with ocean water, saline waters, and meteorites from -hydroxides occupy the pore space as intimate mixtures Oman and range between 187 lgg 1 in solid and of different phases, typically associated with gypsum. 14140 lgg 1 in solid. Excluding the SL091 samples, They often show some Si content, in some cases pore which are from a relatively small geographic region, the fillings of opal or microcrystalline quartz were observed. median value for the total dissolved ions is 1148 lgg 1 Magnesium salts are also abundant and often visible by in solid. The content of water-soluble ions depends on naked eye. The most dominant phases are Mg-sulfates geographic and geologic factors. Close to the Arabic sea such as epsomite, MgSO4*7H2O, and kieserite, at Ras Madrakah, concentrations are higher due to the MgSO4*H2O, that tend to effloresce on the meteorite input of salts from sea spray (Fig. 4). Inland salt surfaces or in the pore space (Fig. 6a). Additional contents in the soil are higher close to depressions, identified minerals in pore space are magnesite, where water accumulates during occasional rainstorms celestine, and barite. Jarosite, (K,Na)Fe3(SO4)2(OH)6, and evaporates. A prominent depression is the sabkhah was observed in a few samples, mostly in cleavages of Umm as Samim (“mother of poison”; e.g., Fookes and weathered troilites (Fig. 6b). Some jarosite occurs also Lee 2009). High-density meteorite recovery areas such in vein fillings (Fig. 6c) or in pore space (Fig. 6d). as the Sayh al Uhaymir (SaU), Jiddat al Harasis (JaH), Most contamination minerals are easily dissolved in Dhofar or the Ramlat as Sahmah (RaS) are generally water and do not survive water-assisted thin section low in salt. preparation. Due to its deliquescent behavior, the 1966 F. J. Zurfluh et al.

Table 2. Secondary minerals in OCs from Oman. Mineral Chemical composition RLM TLM SEM (EDS) Raman XRD EMP (III) (II) Akaganeite (Fe ,Ni )8(O, OH)16Cl1.25*nH2Ox x Anhydrite CaSO4 xx x Ankerite Ca(Fe, Mg)(CO3)2 x Aragonite CaCO3 xx Barite BaSO4 x Bassanite CaSO4*0.5H2Oxx a “Bischofite” MgCl2*6H2Ox Calcite CaCO3 xx Celestine SrSO4 x Dolomite CaMg(CO3)2 xx (III) Ferrihydrite 5Fe 2O3*9H2Ox Epsomite MgSO4*7H2Ox Goethite a-Fe(III)O(OH) x x x x x Gypsum CaSO4*2H2Oxx (III) Hematite a-Fe 2O3 xx x Hexahydrite MgSO4*6H2O x (III) Jarosite KFe 3(SO4)2(OH)6 xx Kieserite MgSO4*H2Oxx a (II) “Lawrencite” (Fe ,Ni)Cl2 x Lepidocrocite c-Fe(III)O(OH) x x x (III) Maghemite c-Fe 2O3 xx Magnesite MgCO3 x (II) (III) Magnetite Fe Fe 2O4 xx x (III) Natrojarosite NaFe 3(SO4)2(OH)6 x Nepouite Ni3Si2O5(OH)4 xx Nickel-bischofite NiCl2*6H2Ox Opal SiO2*nH2Oxx Pecoraite Ni3Si2O5(OH)4 xx Quartz/Chalcedony SiO2 xx Sulfur S x (II-III) Unspecified Fe sulfate Fe -SO4-nH2Ox RFL = reflected light microscopy; TLM = transmitted light microscopy; SEM = scanning electron microscopy by use of EDS analysis for qualitative and semiquantitative element abundances; Raman = Raman spectroscopy; XRD = powder X-ray diffraction; EMP = electron microprobe analysis. aBischofite and lawrencite were not found as crystalline material under room conditions and are not considered true minerals.

presence of bischofite as crystalline phase could not be soil. For the meteorite, the maximum recorded verified. However, semiquantitative analysis of brine temperature was 66.3 °C on July 11, 2009 at 1:00 pm; droplets removed from “sweating meteorites” (Fig. 1) for the soil 54.8 °C on July 13, 2009 at 2:40 pm. yielded Mg/Cl molar ratios close to 2 as in bischofite. Minimum temperatures were measured on January 7, Iron chloride with traces of Ni is also part of the suite 2010 with 4.8 °C for the meteorite at 6:40 am and of salts identified in dried brine droplets. A dried iron 9.1 °C for the soil at 7:20 am. hydroxide rich residue of brine from meteorite Dhofar Since the meteorite was placed on the surface, it 1010, analyzed by XRD, yielded akaganeite as main heated faster and reached daily peak temperature earlier phase, similar to observations by Buchwald and Clarke than the soil resistor (Fig. 7a). The maximum (1989). Against expectation, halite was not observed temperatures in the meteorite were recorded around within the studied meteorites. 1:30 pm, in the soil about 1 h later, at 2:30 pm. Minimum temperatures show a similar behavior: in the Temperature Fluctuations in Meteorite and Soil meteorite, they are recorded at 6:20 am, just before sunrise, about 20 min earlier than in the soil. The mean The first period of temperature measurements was daily temperature variations are much larger in the from June 20, 2009 to May 13, 2010 with 24,360 meteorite (34.3 °C) than in the soil (21.6 °C) and temperature readings both in the meteorite and in the remain similar over the whole measuring period Water-soluble salts in chondrites from Oman 1967

Fig. 6. Pore space fillings. a) Macroscopic view of a large partly filled pore in meteorite JaH 478. The whitish-yellowish area is a mixture of anhydrite (Anh), epsomite (Eps), kieserite (Ksr), and jarosite (Jrs). Scale bar is 1 mm. b) BSE image of a weathered troilite grain in meteorite RaS 408 (H6 S2 W4). Most of the troilite is replaced by iron hydroxide. The iron oxides to the left are altered metal grains. Troilite is weathering along (001) twinning plains and, at latest stages of weathering, produces new pore space filled here with Ca-sulfate (CaSO4) and jarosite (Jrs). Some celestine (SrSO4) precipitated in the neighborhood. The olivine grain to the right is slightly attacked (w-Ol) and the resulting pore space is filled with minor Ca-sulfate and iron hydroxide. Scale bar is 20 lm. c) BSE image of completely filled pore in OC RaS 408. The majority of the pore filling is jarosite, which also produced veins that crack iron oxide veins. Sparse Ca-sulfate occurs close to a weathered troilite grain. Scale bar is 100 lm. d) BSE image of a weathered metal grain with jarosite efflorescence in sample RaS 395 (L6 S5 W4). Scale bar is 50 lm. e) BSE image of salt crystals from efflorescence on a cut OC surface (Dhofar 1684, H5 S2 W3). Minerals visible are Mg-sulfates with traces of Na, Cl, and Ni occurring as hairy minerals and aggregates of small grains. The bright mineral at the left is troilite with attached silicate material that was accidentally picked. Scale bar is 200 lm. f) BSE image of a weathered phosphate grain in a JaH 091 sample (L5 S2 W3, sample # 0703-335). The bright mineral in the neighborhood is partially weathered troilite. Ca-phosphate is partly replaced by Fe-phosphate. Scale bar is 50 lm. 1968 F. J. Zurfluh et al.

the meteorite (long-term experiment), several tens of km away, at the same time.

DISCUSSION

In the following sections, we discuss the source of the contamination in the meteorites, the process of salt uptake, and the relevance to weathering. Although the extraction series for meteorites were performed under various conditions, they are in agreement with each other and provide a first data basis on a previously unrecognized effect.

Source of the Ions

Potential sources of water-soluble ions are water- soluble primary meteoritic minerals, ions mobilized during weathering of meteoritic minerals (hydrolysis), and contamination from the terrestrial environment.

Chlorine The concentration of primary water-soluble salts is generally very low in freshly fallen OCs as observed in leaching experiments of Bensour (Lee et al. 2006) and Tamdakht (this study). Primary halite and sylvite were detected in just two meteorite falls both of which curiously occurred in 1998 (Rubin et al. 2002). One is Monahans, a H5 chondrite fall in Texas (Grossman Fig. 7. Temperatures recorded by the data logger. a) Typical 1998; Zolensky et al. 1999) and the other is Zag, a H3-6 pattern of daily temperature evolution as recorded in July regolith breccia fall in Western Sahara (Grossman 2009. The meteorite is heated faster and cools down more 1999). rapidly than the thermistor in the soil. The daily temperature Garrison et al. (2000) calculated elemental chlorine variation reaches 34.3 °C in the meteorite and is about 21.6 °C in the soil. b) Monthly average temperatures of soil concentrations of 132 samples from 94 meteorites (most and meteorite temperatures show only small differences and of them finds, including some from Antarctica) using are larger in summer. The meteorite reaches higher maximum the nuclear reaction of 37Cl (n,c) 38Ar achieved during and lower minimum temperatures. neutron irradiation of meteorites. For ordinary chondrites, they found 15–177 lgg1 of chlorine. This (Fig. 7b). However, daily and monthly average is in partial agreement with 13 analyses of OCs by temperatures in soil and meteorite are very similar combustion ion chromatography of Tarter et al. (1980) (Fig. 7b). The largest monthly deviations are observed yielding 9–345 lgg1 of chlorine. Most of their studied in summer (June: meteorite-soil 1.7 °C); in winter; they samples were falls, but no systematic higher chlorine are almost identical (January: 0.1 °C). concentrations in finds were observed. During leaching Surface temperatures were also measured with a experiments of Zag samples (Bridges et al. 2004), a Cl thermoprobe during fieldwork. On 2 days at the end of excess relative to Na+ was observed, indicating the January 2010, when air temperatures reached 28 °C presence of another phase delivering Cl beside halite. (1:15 pm) and 29 °C (14:45 pm), meteorite surface The only other Cl bearing phase in Zag is chlorapatite, temperatures were 34 °C and 41 °C. The top 1 cm of which is another source of dissolved Cl. The consensus the soil was at temperatures of 37–46 °C. The of all these studies is the concentration of Cl in OCs is temperatures measured on the surface of the soil are variable and ranges between 9 and about 350 lgg1. significantly higher than those measured at a depth of In our samples from Oman, we observed a range of 5 cm. This indicates that the magnitude of diurnal extractable chloride concentrations from 10.8 to temperature fluctuations rapidly decreases with 3367 lgg1 with one outlier at 6036 lgg1. Most of increasing depth in the soil. These meteorite surface the Oman samples are significantly higher than temperatures are very similar to those measured inside 177 lgg1, the upper limit of the Cl concentrations Water-soluble salts in chondrites from Oman 1969 postulated by Garrison et al. (2000). The average Cl samples and meteoritic Ca-minerals are relatively concentration of the MLO series is 721 lgg1 and the resistant to weathering, the majority of leachable Ca2+ median 356 lgg1. Such elevated Cl concentrations has probably been transported from the soil into the are most likely the result of terrestrial contamination. meteorite. Sources for the halite, the major Cl carrier phase in A similar behavior is observed in the case of the Oman soil, are bedrock weathering, evaporation of strontium. It is introduced in the meteorites and groundwater, wind-transported salts from sabkhas and precipitates as celestine, a highly insoluble mineral, sea spray (Chapman 1980). Meteorites from Oman which accumulates over terrestrial residence time (Al- show a large scatter in Cl, the contamination occurring Kathiri et al. 2005). mainly in the interior of the samples (Fig. 2c). This is in agreement with Cl depth profiles measured in three H5 Sulfur chondrite finds from Western Australia (Krahenb€ uhl€ Sulfur in salt phases probably is partly derived from and Langenauer 1995) where contamination patterns the weathering of meteoritic troilite, but it also occurs in are random with lower values toward the meteorite the soil. The sulfate found in pore space most likely has a surface. The highest detected Cl concentrations in combined origin from soil and meteorite. In the vicinity these three chondrites were 262, 503, and 1528 lgg1. of a weathered Ca-phosphate and a partly altered troilite, The lowest concentrations are systematically higher than an alteration assemblage of Ca-sulfate, Fe-phosphate, in meteorite falls. Therefore, the contamination of hot and iron hydroxides was observed (Fig. 6f), which might desert meteorites with Cl can be considered a common be the result of spatial limited ion redistribution within phenomenon. the meteorite. However, contamination of hot desert meteorites by Ca-, Sr-, and Ba-sulfates is common and Sodium and Potassium occurs also in troilite poor meteorites such as lunar and The source of the measured Na and K contents in Martian meteorites (e.g., Crozaz and Wadhwa 2001; meteorites could be primary halite or sylvite, which Nazarov et al. 2004; Korotev 2012). Therefore, a both have a high solubility (equilibrium constant of significant part of the sulfate within meteorites must be NaCl is log K 0.21; Palandri and Kharka 2004) or derived from the desert environment. contaminants from desert soil. An alternative source could be meteoritic feldspar susceptible to weathering Mass Balance (Lee et al. 2006). However, feldspar dissolution rates are much lower as the solubility is lower (log k 11.8, To interpret the relative abundance patterns of the (Palandri and Kharka 2004). water-soluble ions in the meteorites, a mass balance approach was applied. The concentrations of the Magnesium leachates from 15 Oman meteorites from the MLO The interpretation of the origin of magnesium series were converted to mmol g1of the meteorite. seems simpler. Extracts from the soil are low in Mg2+, Then ions in proportions corresponding to the common whereas meteorites show high concentrations. As salt minerals and hydrolysis reactions were subtracted ordinary chondrites are composed of 50% to 75% of as normative minerals in seven steps (Table 3). After the Mg-rich minerals, olivine and orthopyroxene the mass balance, the median values of the contribution (Mason 1965; Dunn et al. 2010), which are relatively of each constituents were calculated for the weathering susceptible to weathering (solubility of olivine log k 7, grades W ≤ 2, W = 3, and W ≥ 4 (Table 3). pyroxene log k 11; Palandri and Kharka 2004), it is Step 1: Subtraction of NaCl. It is assumed that all assumed that most of the detected Mg2+ is meteoritic in Cl analyzed derived from the soil since OC falls origin. Moreover, thin section investigations show that generally have negligible concentrations as shown in the only olivine shows clear alteration. This idea is Tamdakht meteorite. In addition, primary halite could be supported by the observation of Mg-carbonates in leached out during an early stage of weathering. On the Antarctic chondrites with Na/Mg, K/Mg, and Ca/Mg basis of analyzed soil extracts, we assume that chloride values closer to chondrites than Sea Water indicating a introduced into meteorites was initially accompanied by meteoritic origin (Velbel et al. 1991). Na and K with a molar Cl:Na:K ratio of 100:93:7. After subtraction of all Cl and corresponding amounts of Na Calcium and Strontium and K, a negative balance for Na+ results in 8 of 15 Less clear is the situation in the case of calcium, weathered meteorites. Extreme negative balance of Na+ abundant in the soil (e.g., calcite, gypsum) and in is found in samples of W = 3 (Table 4). meteorites (feldspar, Ca-rich pyroxenes, and Ca- Step 2: Subtraction of the contribution from phosphates). As Ca2+ is easily leached from soil weathering of meteoritic feldspar. For this simplified 1970 F. J. Zurfluh et al.

Table 3. Normative mineral contributions to salt extracts from meteorites based on mass balance calculations. Step “Mineral” Used formula Used ion for calculation W ≤ 2W= 3W≥ 4 1 Halite Na0.93K0.07Cl Cl 3.60 36.08 7.38 2 Feldspar Na0.10K0.06Ca0.84Si2.16Al1.84O8 Al 0.03 0.03 0.04 3 Calcite CaCO3 HCO3 0.14 0.14 0.71 2+ 4 Ca-sulfate CaSO4 Ca 0.47 4.58 10.15 2+ 5 Olivine Fe0.2Mg0.8SiO4 Mg 0.56 5.77 3.23 6 Metal Fe15Ni Ni 4.44 7.39 0.47 2 7 Troilite FeS SO4 1.70 9.75 1.83 W = Weathering grade (Wlotzka 1993). Number of samples: for W ≤ 2 n = 4, for W = 3 n = 5, and for W ≥ 4 n = 7. Median values based on ion concentration in (mmol g1).

Table 4. Residuals of mass balance calculations. Element W ≤ 2W= 3W≥ 4 Possible explanation Na+ 1.79 12.68 5.42 Precipitation of jarosite or lost as Na-sulfate? K+ 0.25 1.38 0.53 Precipitation of jarosite? Fe 1.50 1.09 0.92 Precipitation of iron hydroxides Si 0.43 7.98 3.76 Precipitation of SiO2, silicates Balance 0.97 23.13 10.63 Negative balance due to Na+,K+, Fe, and Si underabundance W = Weathering grade (Wlotzka 1993). All values in (mmol g1). mass balance approach, it was assumed that all of the assumed to result from hydrolysis of olivine with an aluminum measured in the extracts is derived from average composition of (Fe0.2Mg0.8)Si1O4. Pyroxene feldspar as it is the only major Al-bearing mineral and hydrolysis was ignored, as olivine seems to be artificial weathering experiments showed that it is preferentially weathered as seen in thin section (Fig. 8). attacked (Lee et al. 2006). For the mass balance Step 6: Based on the measured Ni concentration, approach, a normative feldspar composition of the contribution derived from dissolution of metallic (Na0.10K0.06Ca0.84)(Si2.16Al1.84)O8 was subtracted until iron was subtracted. A metal composition of Fe15Ni1 Al reached a value of zero. Another Al-bearing mineral was used. in OCs would be chromite that is extremely resistant Step 7: At last, using the remaining excess of 2 against weathering as it is preserved in meteorites where SO4 , the contribution of troilite (Fe1S1) oxidation was all other primary minerals have been altered or replaced calculated. (Thorslund et al. 1984; Schmitz and Haggstrom 2006). The results of these calculations are listed in Table 3 Step 3: All the measured alkalinity is interpreted to and illustrated in Fig. 9. Table 4 shows the residuals. If be derived from calcite dissolution, a contaminant from we assume that all Cl found in the meteorites was the environment. In this step, pure calcite with the introduced as NaCl in proportions as it is present in the 2+ + building blocks 1Ca and 1HCO3 was removed in soil, we have a strong deficit in Na and to some extent the mass balance approach. Its contribution is relatively K+. The samples that are rich in Cl and show small. hygroscopic behavior are usually most depleted in Na+. Step 4: The residual Ca2+ is thought to be from This issue is discussed later. Also, Si should have been terrestrial Ca-sulfates, which have precipitated in pore detected in higher concentrations in the leachates, space as gypsum or anhydrite. Therefore, Ca1(SO4)1 was especially at W = 3. But it is possible that silica removed in the mass balance. Besides feldspar, calcite, mobilized by the weathering of the silicates precipitated and Ca-sulfate, there might be additional sources for as stable chalcedony and was not dissolved completely Ca2+ like Ca-rich pyroxenes or Ca-phosphates. But since during the leaching procedure. In addition, Si could have these are minor minerals, their contribution was ignored. built colloids together with Al, which would also have an Calcium-phosphates are usually highly weathered in finds influence on the estimated contribution of feldspar. from Oman (Fig. 6f), but their modal occurrence is The measured values for iron were not considered typically <1% (Mason 1965). The contamination with in the mass balance as they have to be treated carefully. Ca-sulfates increases over weathering history. On one hand, artificial weathering during leaching Step 5: All Mg2+ is derived from the weathering of experiments can mobilize iron, which was not yet the meteorite. For simplicity, all magnesium was involved in weathering, and on the other hand, iron Water-soluble salts in chondrites from Oman 1971

Fig. 8. Petrographic indications of leaching of meteoritic minerals. a) Reflected light image of RaS 397 (L6 S5 W4). In the vicinity of a weathered troilite, olivine grains are attacked in boxwork style and iron hydroxides are produced. A feldspar grain (Fsp) is not attacked at all; only some fractures are filled with iron hydroxides. Scale bar is 50 lm. b) BSE image of RaS 316. This W4 chondrite is highly friable and is cross cut by several cracks. Olivines surrounding the weathered troilite are leached in boxwork style, producing new pore space. Troilite is altered to iron hydroxides and partly replaced by natrojarosite. A former metal grain is replaced by iron oxide (FeO). Pyroxenes (dark gray) and feldspars (darkest gray) show no or only minor leaching. Scale bar at bottom left is 100 lm. c) Reflected light image (oil immersion) of a JaH 091 sample (L5 S2 W3, #0603-249). A mafic silicate grain next to a mixture of iron oxides (FeO) and iron hydroxides (FeOOH) shows etch pits. The brightly reflecting mineral is a partly altered troilite. Scale bar is 20 lm. d) BSE image of a weathered metal and troilite grain in RaS 316 (L5 S3 W4). Note the leached parts in the surrounding olivine (“boxwork”).

released during natural weathering usually precipitates the dominant processes, but also olivine is attacked and as insoluble iron hydroxides and would be not detected a significant amount of ions is introduced from outside. during leaching experiments. During the latest stages of weathering (W ≥ 4), the The input of terrestrial contaminants is significant remaining troilite and metal are oxidized and olivine is and generally dominates over the ions mobilized from leached. During weathering history, the contamination the meteorite by weathering (Table 3; Fig. 9). by Ca-sulfates increases continuously. Weathering of Especially, the input of Cl is highest during the W = 3 feldspar is negligible. stage. At the first stages of weathering (W ≤ 2), weathering of metal and to a lesser degree of troilite are Weathering Reactions the main processes (e.g., Al-Kathiri et al. 2005). Contamination is present, but minor. During the In this section, we link water-soluble salt data with weathering stage, W3 metal and troilite oxidation are observed and/or inferred weathering reactions. 1972 F. J. Zurfluh et al.

0 III 4Fe þ 2H2O þ 3O2 ! 4Fe OOH (iron-hydroxide) (4)

(2) Weathering of troilite typically leads to the precipitation of iron hydroxides. Sulfur is mobile or precipitates as sulfate minerals. The weathering reaction can be described as in Hofmann et al. (2000):

III 4FeS þ 6H2 O þ 9O2 ! 4Fe OOH (Fe-hydroxide) þ 2 þ þ 4SO4 8H (5)

It is important to note that hydrogen ions are produced by this reaction, lowering the pH, similar to pyrite oxidation in acid mine waste (e.g., Nordstrom 2011). This is consistent with our findings: We measured pH values of 4 to 7 in meteorite leachates; most samples of stage W3 were at pH 4. This indicates a local disequilibrium between an acid pore fluid and the mafic silicates. (3) The weathering of mafic silicates (olivine and Fig. 9. Median contribution of the most important water- pyroxene treated together) is dominated by hydrolysis, soluble ions represented as normative “minerals” as function here the example for forsterite: of weathering history (W = weathering grade; Wlotzka 1993). Contamination constituents are to the left, while the þ þ ! 2þ þ þ contribution of weathering of meteoritic material is to the Mg2SiO4 4H 2Mg SiO2 2H2O (6) right. Contamination is important over the whole history and generally increases with weathering degree. Most ions are This reaction is important as it consumes hydrogen liberated from the meteorite during W = 3. ions. Weathering of troilite produces enough protons at stage W3 for dissolution of Mg2+ from olivine to be (1) The alteration of metal (kamacite, taenite) is a consistent with the calculated mass balance approach multistage process producing mobile ferrous and finally (Table 3). We take the typically higher degree of ferric iron that precipitates as iron oxides or iron weathering of olivines around altered troilites as hydroxides (Buchwald and Clarke 1989). Ferrous iron supporting evidence for this scenario (Figs. 6b and 8a–d). can be transported within the meteorites and even Together with Cl, the liberated Mg2+ forms a reaches the meteorite surface where its oxidation hygroscopic brine, which causes the macroscopically produces rusty encrustations, commonly incorporating visible “sweating” effect (Figs. 1 and 2). sand grains. Within the meteorites, a network of Weathering of the fayalite component of olivine is a complex iron oxides and iron hydroxides is forming. multistage process, similar to the oxidation of troilite As learned from thin section studies of OCs from (Bland et al. 2006). The final, stable phases are iron Oman, metal grains are first altered to fine-grained hydroxides, often containing minor but significant iron hydroxides followed by iron oxides such as amounts of SiO , as demonstrated by our EDS analyses magnetite and maghemite at a later stage (Figs. 6 and 2 and earlier work (e.g., Lee and Bland 2004). Therefore, 8) (Al-Kathiri et al. 2005). Hematite occurs only we write the weathering equation (fayalite to iron rarely. Taking this into account, we can write the hydroxide) as follows: following weathering reactions for the oxidation of iron: 2Fe SiO þ O þ 2H O ! 4FeIIIOOH þ 2SiO (7) 0 þ ! II III 2 4 2 2 2 3Fe 2O2 Fe Fe2 O4 (magnetite) (1)

Weathered olivines are most commonly associated 0 þ ! II III þ 3Fe H2O Fe Fe2 O4 (magnetite) 4H2 (2) with oxidized troilites (Figs. 6b and 8a). “Boxwork” olivines are very abundant where Mg2+ is leached out and a fine framework of iron hydroxides is left behind 0 þ ! III 4Fe 3O2 2Fe2 O3 (maghemite) (3) (Figs. 8b and 8c). In vicinity, SiO2 phases can occur in Water-soluble salts in chondrites from Oman 1973 pore space or along cracks. In a few meteorites from Oman, we found SiO2 minerals. They are also described in meteorites from other localities, e.g., the Wolf Creek iron meteorite (White et al. 1967) or the polymict ureilite EET 83309 (Beard et al. 2009, 2011; Downes et al. 2011). In contrast to the authors of the latter studies, we interpret the occurrence of opal as low temperature terrestrial alteration. It typically forms during advanced weathering degrees, when the silicates start to be affected (W4; (Wlotzka 1993).

Evolution of Contamination

Uptake of the Ions and Salt Crystallization The analyzed Oman rainwater is very low in 2+ 2 dissolved ions with HCO3 > Ca > Cl > SO4 > Na+, comparable to other rainwaters from Arabia (Alabdula’aly and Khan 2000). Over long periods of time and with extensive evaporation, rainwater might be sufficient to explain the observed contamination of meteorites by water-soluble salts. However, it is more likely that soil pore water, derived from rainwater by dissolution of salts, is transported into the meteorites by a combination of capillary forces and evaporative Fig. 10. Calculated concentrations of ions in water-saturated pumping. This saline water is further fed with ions from soil (assumed porosity 20%) and in pore space of meteorites compared with ion concentrations in saline waters, ocean, the meteorites and is distributed over most of the open and rain as a function of Cl . NaCl is the halite dissolution pore space of the meteorites due to the activity of line. All terrestrial samples plot close to this line, except soil capillary forces (Chapman 1980) and the introduced ions samples at low concentrations. Weathered meteorites have a mix with pre-existing salt concentrations by diffusion (Pel lower Na+/Cl -ratio. The meteorites with elevated Na+/Cl et al. 2003). Preferential heating of the dark meteorites as ratios are the Tamdakht sample and W1 chondrite JaH 578 compared with the bright soil leads to preferred (H6 S2). Here, the released sodium might be of primary meteoritic origin with only minor contamination by NaCl. evaporation from the meteorites, causing more soil water to be taken up into the meteorites. Figure 10 shows the concentration of Na+ in rainwater, soil leachates, saline waters, and the meteorite samples plotted against Cl. Fig. 11, where concentrations of the leached ions are Rainwater, soil leachates and saline waters have Na/Cl plotted against weathering grade. Cations that are ratios lying on the sodium chloride dissolution line (halite mainly introduced from the environment, such as Ca2+ Na/Cl = 0.64, median soil Na/Cl = 0.65, median saline and Sr2+, increase with terrestrial residence time water Na/Cl = 0.63) while meteorite leachates in most (Figs. 11a and 11b). The behavior of Na+ is more cases are depleted in Na (median meteorite Na/Cl = complex (Fig. 11c), as it is introduced together with 0.07). Most meteorites have similar concentrations of Cl, but leachates are strongly depleted in Na+ relative ions comparable to regional saline waters. In meteorites, to Cl as compared with soil leachates, possibly due to precipitation of evaporate minerals is expected due to the later loss as discussed below. The concentrations of oversaturation of the brine and the sequence of Mg2+ and Cl show the same variation in precipitating minerals is similar to terrestrial sea lakes concentration compared with weathering grade, starting with carbonates, which are followed by sulfates indicating that they are probably linked during the and chlorides (e.g., Kilic and Kilic 2006). Eventually, the history of weathering (Figs. 11d and 11e). Externally 2+ 2 brine becomes oversaturated in Mg ,Cl ,andSO4 , derived Cl is accumulated from stages W1 to W3. leading to the formation of hydrous Mg-sulfates and Chlorine is continuously lost from stage W3 to W4. The finally bischofite (Steiger et al. 2011). exact reason for this effect remains unclear, but it seems to be linked to the alteration of troilite. At the end of Salt Contents as a Function of Weathering Stages W3 and at W4, all of the troilite is oxidized and no The evolution of the OC contamination is more protons are produced leading to general neutral correlated with the degree of weathering as displayed in pH conditions. Similarly, Mg2+ is leached from olivine 1974 F. J. Zurfluh et al.

Fig. 11. Correlation of leachable ions with weathering degree (after Wlotzka [1993] with slight modifications). a) The good correlation of Ca2+ with the weathering degree is the result of a continuous contamination with calcite and Ca-sulfate. b) The contamination of hot desert meteorites with Sr2+ is continuous with terrestrial residence time and therefore also with increasing degree of weathering. c) No strong correlation is observed between Na+ and increasing degree of weathering. d) Most of the Mg2+ is leached form olivine at weathering stage W3. e) Samples at stage W3 have the highest Cl concentrations. f) The 2 amount of SO4 increases at higher degrees of weathering resulting in precipitation of Ca-, Sr-, and Ba-sulfates. At stage W3, there might be an increased contribution due to the weathering of the iron sulfide troilite. and pyroxene continuously from W1 to W3, but the transported outwards as weathering proceeds from W3 majority remains in the meteorite, mainly in to W4, where the mafic silicates, the main carriers of combination with Cl.Mg2+ is subsequently magnesium, are attacked. Water-soluble salts in chondrites from Oman 1975

2 The weathering history of SO4 is also complex The Sodium Deficit (Fig. 11f). A part of the sulfate is introduced into the meteorites and precipitates as relative stable phases like The composition of soil leachates demonstrates 2 CaSO4, BaSO4, or SrSO4. Simultaneously, SO4 is that sodium and minor potassium enter meteorites as released during weathering of troilite (Eq. 5) that starts contaminants in combination with Cl. The meteorite at W1, but has it maximum during W3 to W4. Indeed, leachates (median Na/Cl = 0.07) are strongly depleted 2 we observe an increase of SO4 up to late stage W3, in Na relative to soil leachates (median Na/Cl = but samples at W4 are significantly lower compared 0.65), however. This deficit in sodium could be 2 with high stage W3 samples indicating a loss of SO4 interpreted as an artifact of the leaching experiments during weathering stage W3 to W4. This is in or as analytical problems. But this is not the case. agreement with earlier studies (Lee and Bland 2004; We can exclude the precipitation of an Na phase Al-Kathiri et al. 2005; Saunier et al. 2010). Only a during leaching as untreated IC measurements are in 2 minor amount of the released SO4 remains in the perfect agreement with ICP-OES measurements where meteorites, as component of weathering products such the precipitated minerals were dissolved by adding as jarosite, celestine, barite, and Ca-sulfates. Buchwald HNO3. We have two hypotheses to explain the and Clarke (1989) detected traces of sulfur in terrestrial sodium deficit: (I) precipitation of a low-solubility akaganeite within Antarctic meteorites. But the main phase during weathering (e.g., magadiite, kenyaite, 2 part of the SO4 anions might be lost together with natrojarosite) and (II) crystallization of Na-salts on Mg2+ and Cl as proposed above and also with Na+ meteorite surfaces and successive erosion by wind or and K+ as discussed below. At weathering stage W4, rain. only low concentrations of soluble ions were detected. (I) Precipitation of an Na phase: Possible Most of the contamination minerals are precipitated as candidates are the hydrous Na-silicates magadiite, relative stable minerals and weathering of metal and NaSi7O13(OH)3*3H2O, and kenyaite, Na2Si22O41(OH)8 * troilite is complete and the chemical environment 6H2O, that were found in recent sediments of the within the meteorite reaches a near equilibrium, alkaline Lake Magadi, Kenya (Eugster 1967; Eugster which slows down chemical weathering. Accordingly, and Jones 1968). A possible reaction (Eq. 8) would be the leachates of highly weathered samples tend the hydrolysis of forsterite reacting with Na+and to become neutral in pH. Most likely, OCs have hydrogen ions, most likely released from the weathering reached a relatively passive stage at W4 with limited of troilite (Eq. 5). chemical weathering. During more than 10 years of þ þ þ þ ! meteorite searches in Oman, we found only one 7Mg2SiO4 Na 27H NaSi7O13(OH)3 3H2O meteorite at W5. 2þ þ 14Mg þ 9H2O (8) Degree of Burial and Salt Contamination Based on the common observation of soil particles During this reaction, also a lot of Mg2+ would be attached to meteorites’ surfaces by iron hydroxide liberated. However, no Na-silicate mineral was found cements, many meteorites in Oman must have been through extensive survey of ordinary chondrite samples completely buried in the soil for a significant part of and hence this reaction remains unsubstantiated. their terrestrial history. Nevertheless, contamination by Another poorly soluble sodium-bearing mineral is the soil probably was more efficient during partial natrojarosite (Eq. 9). exposure. We assume that stronger heating of meteorites leads to rapid evaporation of their contained þ 12FeSþ4Na þ 18H2O þ 27O2 pore water. Water containing salts will be replenished (9) ! IIIð Þ ð Þ þ 2 þ þ from the surrounding soil. This process will finally lead 4NaFe3 SO4 2 OH 6 4SO4 12H to a strong enrichment of salts in the meteorites. This is in contrast to proposed processes in more humid, Jarosite generally forms under acidic conditions, temperate climates, where weathering occurs dominantly which can locally occur in the meteorites due to troilite in the subsurface (Abreu and Brearley 2005). We found oxidation. Indeed, in several altered remnants of evidence in several samples for more extensive troilites, jarosite was found. However, 10 out of 12 weathering of the exposed meteorite parts as compared samples were potassium jarosite with minor traces of with buried parts. The meteorite samples with the sodium, and only in two cases, natrojarosite was highest concentrations of water-soluble salts, UaS 011 present. In samples having an Na+ deficit as calculated and 0703-703 (JaH 091), were not buried in the soil during the mass balance approach, no natrojarosite was (Tables S4 and S7). found. Therefore, reaction 10 might explain the 1976 F. J. Zurfluh et al. observed deficit of K+ in OCs, but cannot explain fully the deficit in Na+ (Table 4). (II) A second possible explanation is the formation of sodium sulfate efflorescences (mirabilite or related phases) on or near the meteorite surfaces, while Mg2+ and Cl remain concentrated in brine or bischofite in more central parts of the meteorites (Eq. 10). Iron hydroxides and silica are produced as additional phases.

þð þ Þ þ þ þ Mg2SiO4 14 X H2O 4NaCl 4FeS 20O2 ! þ þ 2MgCl2 6H2O 4FeOOH SiO2

þ 4NaSO4 XH2O (10)

Several Omani OCs stored in humidity- and temperature-controlled rooms showed efflorescence of white hairy minerals similar to descriptions of mirabilite salts from desert rocks or salt efflorescences during rock disintegration experiments with sodium sulfate (e.g., Walther 1900; Rodriguez-Navarro and Doehne 1999). Fig. 12. Porosity of meteorites used for the leaching Nevertheless, all analyzed hairy salts and aggregates experiments correlated with degree of weathering. Fresh OC of small grains found on meteorites from Oman were samples have relatively high porosities, 7.0% (H) and 5.6% Mg-sulfates and Mg-carbonates, and not Na-sulfates (L), respectively (Consolmagno et al. 2008). (Fig. 6e). The reason might be that Na+ was already lost from these meteorites in the desert. Mass balance cementation of pores by iron hydroxides (Consolmagno requires that Na+ is introduced into the meteorites et al. 2008) that can result in virtual zero porosity (Britt together with Cl , but then the two ions are somehow and Consolmagno 2003). The porosities of our samples separated. Chlorine remains in the meteorite, while Na+ show a distinct picture: W0 meteorite Tamdakht has the is transported with sulfate to the surface of the highest porosity, whereas W1 and W2 samples have the meteorites. When the meteorites cool down overnight, lowest. Interestingly, with increasing weathering degree, efflorescences of mirabilite might be formed, which then the porosity increases again (Fig. 12). A detailed study are removed by wind erosion or washed off by (rare) of thin sections reveals the reason for the porosity precipitation. However, we have not observed sodium variation during weathering history. First, an initial sulfates on meteorites collected in Oman. Thus, the decrease in pore space occurs due to sealing and exact reason for the deficit of extractable Na+ in clogging of microcracks and other primary pore space weathered meteorites from Oman remains somewhat by iron oxides and iron hydroxides (W1, W2, and enigmatic, but probably is due to a combination of the partially W3; Fig. 13a). Volume increase can partially two hypotheses. Moreover, the loss of sodium sulfate be buffered by the initial pore space (Al-Kathiri et al. would at least partly explain the bulk loss of sulfur 2005). The volume increase due to metal to iron from weathered meteorites (Lee and Bland 2004; Al- hydroxide transformation (approximately 2.0x) is much Kathiri et al. 2005; Saunier et al. 2010). higher compared with troilite alteration to iron hydroxide (approximately 1.1x; Al-Kathiri et al. 2005; Weathering Effects Lee and Bland 2004). In a second stage, formation of iron hydroxide- and calcite veins (Fig. 13b) can fracture Evolution of Porosity During Weathering the meteorites and produce new porosity (see also Porosity in ordinary chondrites occurs mostly as a Macke et al. 2010). Weathering of metal and troilite is network of microcracks along grain boundaries and associated with leaching of mafic silicates (Figs. 8b and cutting through grains, the latter produced by the 13b) and produces a new generation of secondary pore passage of shock waves from collisions on the parent space (W3 to W4). body (DeCarli et al. 2001). The mean porosity of H chondrite falls is 7.0 4.9% and 5.6 4.7% for L Salt Weathering chondrites falls (Consolmagno et al. 2008). Generally, it Salt weathering can cause disintegration of rocks is assumed that weathering lowers pore space due to (e.g., Goudie et al. 1997). Meteorite fragmentation due Water-soluble salts in chondrites from Oman 1977

Fig. 13. Evolution of pore space in ordinary chondrites during weathering. Scale bars are 100 lm. (a) Reflected light image of the relatively fresh W1 meteorite JaH 578. Primary pore space along grain boundaries and open pore space due to shock events appears black, some are marked as “open cracks.” Around a metal grain with a slightly oxidized rim a radiation of iron (oxy) hydroxide veins following primary pore space is visible. (b) Reflected light image of RaS 297 (L6 S4 W4), a highly weathered chondrite. All of the metal and troilite is oxidized and primary pore space is filled by an early set of bright gray iron oxide veins, mainly grown along grain boundaries. A second generation of gray iron hydroxide veins cross cuts minerals. A late-stage dark gray calcite vein (Cal-vein) partly follows iron hydroxide veins. Due to leaching of olivine, additional new pore space is generated. to the alteration of iron metal to iron hydroxides and low humidity (approximately 40%) conditions. On some oxides is certainly more important than the formation of strongly salt-contaminated OC samples stored under cracks due to salt crystallization. During this work, salts such conditions, we observed efflorescence of white Mg- such as Ca- and Mg-sulfates that would have a potential rich minerals, identified as (hydrous) Mg-carbonates to crack the meteorites (Winkler 1987; Charola et al. and Mg-sulfates with minor amounts of Cl. We also 2007) were found in pores. However, these phases were noted that the magnesium chloride brines are very mostly located in the center of pore space or within corrosive, causing the formation of holes in aluminum partially weathered metal grains where their foil. crystallization pressures could be compensated by the pore space. Iron hydroxide veins propagate mostly along Temperatures of Meteorites and Soils in the Desert grain boundaries and along small fissures formed by shock events. In weathered samples, cracks are abundant In hot deserts, extreme temperatures can occur. For and some of them are only partially filled, mostly with example, in the Sahara, air temperatures vary between Ca-carbonates or Ca-sulfates. It is therefore difficult to 7 °C and 57 °C (Cooke and Warren 1973) and sand estimate the effect of salt weathering in meteorites. surface temperatures can reach up to 83 °C (Petrov The most friable rocks used in this study were W4 1976). Air temperatures in Arabia are commonly above rocks with abundant cracks, partially filled with Ca- 50 °C in summer with monthly averages of 41–42 °C carbonates and Ca-sulfates. They were found in the field (Edgell 2006). The surface temperature of black basalts disintegrated into a high number of fragments (Table (Walther 1900) and asphalt in Al Ain (UAE), close to S4). But we also found a significant number of highly Oman, was found to reach 73 °C (McGreevy and Smith weathered (W4) meteorites that are complete, solid 1982; Potocki 1978). The maximum temperature individuals, even though they were penetrated by several measured inside the meteorite during our experiment was generations of iron hydroxide veins. Such samples are 66.3 °C, approximately 15 °C above the air temperature. found as single pieces and do not break apart during Maximal ΔT observed during a day is 46.5 °C while the cutting in contrast to friable samples with abundant median ΔT per day is 34.3 °C. These temperatures and Ca-sulfates and Ca-carbonate veins. the daily fluctuations have an influence on physical and especially chemical weathering. Due to different thermal Salt-Related Weathering During Storage conductivities (Yomogida and Matsui 1983) and different Due to the hygroscopic force of the magnesium volume expansions of the meteoritic minerals, diurnal chloride brines, water is attracted even under relatively temperature variations weaken the rocks physically (e.g., 1978 F. J. Zurfluh et al.

Gomez-Heras et al. 2006) and salts can be introduced diurnal variations and a maximum temperature of into the meteorite. Temperatures <0 °C, critical for rock 66.3 °C. disintegration (Ruedrich et al. 2011) were not observed The concentrations of water-soluble ions in soil during our temperature recording period, but can occur samples have only a minor influence on the salt during cold winter nights in Arabia (Edgell 2006). contamination in associated meteorites. Rather, the We think that the temperature gradient in the degree of weathering controls the amount and type of meteorite and between meteorite and soil cause a salts found in OCs from Oman. pumping effect, resulting in a strong concentration of The presented data set of water-soluble salts and soluble ions in meteorites over time. Soil salts dissolved temperature variations in OCs from Oman has by rain water are soaked into the meteorites by capillary permitted us to understand in more detail weathering forces and mix with elements dissolved from the processes in hot desert meteorites. The reason why only meteorite. This eventually leads to formation of strongly meteorites from Oman are heavily contaminated with hygroscopic magnesium chloride brines in the meteorite. salts but apparently not meteorites from other hot desert areas with similar present and past climate histories and CONCLUSIONS geologic setting remains unclear. The clue may be in the particular climate history and proximity to the ocean. Chondrites showing strongly hygroscopic behavior (“sweating”) so far are only known from Oman. The Acknowledgments—We thank Mohammed Al-Battashi, pore space of ordinary chondrites recovered from the Ali Al-Rajhi, Salim Al-Buseidi, and Hilal Al Azri, hot desert of Oman is contaminated with water-soluble Ministry of Commerce and Industry, Muscat, for their salts that can occur in high concentrations (up to support and permission. The analytical work by Ruth 3400 lgg1 of Cl and 9000 lgg1 total soluble salts Mader,€ Priska Bahler,€ Stefan Weissen, Christine Lemp, in bulk meteorite samples). The major ions are Mg2+ Florian Eichinger, and Grabriel Chevallier is gratefully (from meteoritic olivine), Ca2+ (from soil), Cl (from appreciated. Ali Al-Kathiri placed the temperature 2 soil), SO4 (from meteoritic troilite and soil), and iron logger in the middle of the Oman desert on a hot (meteoritic). The observed salt concentrations and summer day. Manuel Eggimann and Nicolas Greber relative ion abundances are controlled by weathering conducted a part of the XRD analyses. Nick Waber is reactions. During weathering stages 1 and 2, metallic thanked for his help in designing the experiments. We iron and troilite oxidize and form weakly soluble acknowledge the helpful comments and suggestions of Fe-oxides and -hydroxides and salts. Highest salt associate editor Alex Ruzicka, Anna Losiak, and two concentrations in leachates are reached with weathering anonymous reviewers. This work was supported by the degree W3, characterized by oxidation of iron metal and Swiss National Science Foundation (grant 200020- especially troilite whereby acid is produced that 119937). promotes the hydrolysis of mafic silicates (mainly olivine) and the formation of jarosite. Chloride is Editorial Handling—Dr. Alex Ruzicka reaching high concentrations in pore space while sodium is strongly depleted relative to Cl . The exact process of REFERENCES sodium loss remains enigmatic, but is probably related to either efflorescence of sodium sulfates on meteorite Abed R. M. M., Dobrestov S., Al-Kharusi S., Schramm A., surfaces and subsequent erosion by wind or, less likely, Jupp B., and Golubic S. 2011. Cyanobacterial diversity by formation of Na-bearing jarosite. The “sweating” and bioactivity of inland hypersaline microbial mats from a desert stream in the Sultanate of Oman. Fottea 11:215– meteorites are usually at W3 and show high 2+ 224. concentrations of Mg and Cl that form hygroscopic Abreu N. M. and Brearley A. J. 2005. Carbonates in brines. At weathering degree W4, meteorites have vigarano: Terrestrial, preterrestrial, or both? Meteoritics & reached a relatively stable chemical state with ongoing Planetary Science 40:609–625. slow silicate hydrolysis. However, sulfate contamination Alabdula’aly A. I. and Khan M. A. 2000. Chemistry of rain water in Riyadh, Saudi Arabia. Archives of Environmental continues after complete oxidation of troilite. Some W4 Contamination and Toxicology 39:66–73. meteorites are physically weakened, mostly due to Al-Kathiri A., Hofmann B. A., Jull A. J. T., and Gnos E. fracturing as a result of weathering-related volume 2005. Weathering of meteorites from Oman: Correlation of increase. Fracturing of the meteorites and leaching of chemical and mineralogical weathering proxies with 14C olivine produce new pore space resulting in an increase terrestrial ages and the influence of soil chemistry. Meteoritics & Planetary Science 40:1215–1239. in porosity at higher weathering degrees. Beard A. D., Downes H., and Howard K. 2009. Hydrated Temperature variations, recorded for the first time in silica (opal) in a polymict ureilite, EET 83309. Meteoritics situ in a hot desert meteorite demonstrated extreme & Planetary Science 44:A31. Water-soluble salts in chondrites from Oman 1979

Beard A. D., Downes H., and Howard K. T. 2011. Significance Gnos E., Eggimann M., Al-Kathiri A., and Hofmann B. A. of opal in ureilites—Delivery of H(2)O to the inner solar 2006. The JaH 091 strewnfield. Meteoritics & Planetary system? Meteoritics & Planetary Science 46:A15. Science 41:A64. Bland P. A., Zolensky M. E., Benedix G. K., and Sephton M. Gnos E., Lorenzetti S., Eugster O., Jull A. J. T., Hofmann B. A. 2006. Weathering of chondritic meteorites. In A., Al-Kathiri A., and Eggimann M. 2009. The Jiddat al Meteorites and the early solar system II, edited by Lauretta Harasis 073 strewn field, Sultanate of Oman. Meteoritics & D. S. and McSween H. Y. Tucson, Arizona: The Planetary Science 44:375–387. University of Arizona Press. Gomez-Heras M., Smith B. J., and Fort R. 2006. Surface Bridges J. C., Banks D. A., Smith M., and Grady M. M. temperature differences between minerals in crystalline 2004. Halite and stable chlorine isotopes in the Zag H3-6 rocks: Implications for granular disaggregation of granites breccia. Meteoritics & Planetary Science 39:657–666. through thermal fatigue. Geomorphology 78:236–249. Britt D. T. and Consolmagno G. J. 2003. Stony meteorite Goudie A. S., Viles H. A., and Parker A. G. 1997. porosities and densities: A review of the data through Monitoring of rapid salt weathering in the central Namib 2001. Meteoritics & Planetary Science 38:1161–1180. Desert using limestone blocks. Journal of Arid Environ- Buchwald V. F. and Clarke R. S. 1988. Akaganeite, not ments 37:581–598. lawrencite, corrodes Antarctic iron-meteorites. Meteoritics Grossman J. N. 1998. The Meteoritic Bulletin, No. 82. 23:261. Meteoritics & Planetary Science 33:A221–A239. Buchwald V. F. and Clarke R. S. 1989. Corrosion of Fe-Ni Grossman J. N. 1999. The Meteoritical Bulletin, No. 83. alloys by Cl-containing akaganeite (beta-FeOOH)—The Meteoritics & Planetary Science 34:A169–A186. Antarctic meteorite case. American Mineralogist 74:656–667. Hofmann B. A., Nystrom J. O., and Krahenb€ uhl€ U. 2000. The Chapman R. W. 1980. Salt weathering by sodium chloride in Ordovician chondrite from Brunflo, central Sweden III. the Saudi Arabian desert. American Journal of Science Geochemistry of terrestrial alteration. Lithos 50:305–324. 280:116–129. Jull A. J. T., Cheng S., Gooding J. L., and Velbel M. A. 1988. Charola A. E., Puhringer J., and Steiger M. 2007. Gypsum: A Rapid growth of magnesium-carbonate weathering review of its role in the deterioration of building materials. products in a stony meteorite from Antarctica. Science Environmental Geology 52:207–220. 242:417–419. Chennaoui-Aoudjehane H., Jambon A., Denise M. B., Jupp B. P., Eichenberger U., and Cookson P. 2008. The Boudouma O., Gattacceca J., Weber P., and Bonte P. microbial domes of Wadi Muqshin pools, Sultanate of 2009. Tamdakht meteorite: The last Moroccan fall. Oman. International Journal of Environmental Studies Meteoritics & Planetary Science 44:A50. 65:685–703. Consolmagno G. J., Britt D. T., and Macke R. J. 2008. The Kilic O.€ and Kilic A. M. 2006. Recovery of salt co-products significance of meteorite density and porosity. Chemie der during the salt production from brine. Desalination Erde-Geochemistry 68:1–29. 186:11–19. Cooke R. and Warren A. 1973. Geomorphology in deserts. Korotev R. L. 2012. Lunar meteorites from Oman. Meteoritics London: Batsford Ltd. & Planetary Science 47:1365–1402. Crozaz G. and Wadhwa M. 2001. The terrestrial alteration of Krahenb€ uhl€ U. and Langenauer M. 1994. Distribution of Saharan shergottites Dar al Gani 476 and 489: A case halogens in eucrites and their weathering compared to study of weathering in a hot desert environment. chondrites. Meteoritics 29:486–486. Geochimica et Cosmochimica Acta 65:971–977. Krahenb€ uhl€ U. and Langenauer M. 1995. Comparison of the DeCarli P. S., Bowden E., and Seaman L. 2001. Shock- distribution of halogens in chondrites from Antarctica and induced compaction and porosity in meteorites. from Western Australia. In Workshop on meteorites from Meteoritics & Planetary Science 36:A47. hot and cold deserts, edited by Schultz L., Annexstad J. O., Downes H., Smith C. L., Beard A. D., and Ross A. J. 2011. and Zolensky M. LPI Technical Report 95-02. Houston, Early solar system processes revealed by an integrated Texas: Lunar and Planetary Institute. study of ureilite meteorites. Meteoritics & Planetary Langenauer M. and Krahenb€ uhl€ U. 1993. Halogen Science 46:A60. contamination in antarctic H5 and H6 chondrites and Dunn T. L., Cressey G., McSween H. Y., Jr., and McCoy T. relation to sites of recovery. Earth and Planetary Science J. 2010. Analysis of ordinary chondrites using powder Letters 120:431–442. X-ray diffraction: 1. Modal mineral abundances. Lee M. R. and Bland P. A. 2004. Mechanisms of weathering Meteoritics & Planetary Science 45:123–134. of meteorites recovered from hot and cold deserts and the Edgell H. S. 2006. Arabian deserts. Dordrecht: Springer. formation of phyllosilicates. Geochimica et Cosmochimica Eugster H. P. 1967. Hydrous sodium silicates from Lake Magadi Acta 68:893–916. Kenya—Precursors of bedded chert. Science 157:1177–1180. Lee M. R., Smith C. L., Gordon S. H., and Hodson M. E. Eugster H. P. and Jones B. F. 1968. Gels composed of 2006. Laboratory simulation of terrestrial meteorite sodium-aluminium silicate, Lake Magadi, Kenya. Science weathering using the Bensour (LL6) ordinary chondrite. 161:160–163. Meteoritics & Planetary Science 41:1123–1138. Fisher M. 1994. Another look at the variability of desert Losiak A. and Velbel M. A. 2011. Evaporite formation during climates, using examples from Oman. Global Ecology and weathering of Antarctic meteorites—A weathering census Biogeography Letters 4:79–87. analysis based on the ANSMET database. Meteoritics & Fookes P. G. and Lee M. E. 2009. Desert environments of Planetary Science 46:443–458. inland Oman. Geology Today 25:226–231. Macke R. J., Consolmagno G. J., Britt D. T., and Hutson M. Garrison D., Hamlin S., and Bogard D. 2000. Chlorine L. 2010. Enstatite chondrite density, magnetic abundances in meteorites. Meteoritics & Planetary Science susceptibility, and porosity. Meteoritics & Planetary 35:419–429. Science 45:1513–1526. 1980 F. J. Zurfluh et al.

Mason B. 1965. The chemical composition of olivine-bronzite Schmitz B. and Haggstrom T. 2006. Extraterrestrial chromite and olivine-hypersthene chondrites. American Museum in Middle Ordovician marine limestone at Kinnekulle, Noviatates 2223:1–38. southern Sweden–Traces of a major asteroid breakup McGreevy J. P. and Smith B. J. 1982. Salt weathering in hot event. Meteoritics & Planetary Science 41:455–466. deserts—Observations on the design of simulation Steiger M., Linnow K., Ehrhardt D., and Rohde M. 2011. experiments. Geografiska Annaler Series A—Physical Decomposition reactions of magnesium sulfate hydrates + 2+ Geography 64:161–170. and phase equilibria in the MgSO4-H2O and Na -Mg - 2 Nazarov M. A., Demidova S. I., Patchen A., and Taylor L. A. Cl -SO4 -H2O systems with implications for Mars. 2004. Dhofar 311, 730 and 731: New lunar meteorites Geochimica et Cosmochimica Acta 75:3600–3626. from Oman (abstract #1233). 35th Lunar and Planetary Stelzner T. and Heide K. 1996. The study of weathering Science Conference. CD-ROM. products of meteorites by means of evolved gas analysis. Nininger H. H. 1929. Notes on oxidation of certain Meteoritics & Planetary Science 31:249–254. meteorites: The formation of meteorodes. Transactions of Stoffler€ D., Keil K., and Scott E. R. D. 1991. Shock the Kansas Academy of Science 32:63–67. metamorphism of ordinary chondrites. Geochimica et Nordstrom D. K. 2011. Mine waters: Acidic to circumneutral. Cosmochimica Acta 55:3845–3867. Elements 7:393–398. Tarter J. G., Evans K. L., and Moore C. B. 1980. Chlorine in Palandri J. L. and Kharka Y. K. 2004. A compilation of rate meteorites (abstract). Meteoritics 15:373–374. parameters of water-mineral interaction kinetics for Thorslund P., Wickman F. E., and Nystrom J. O. 1984. The application to geochemical modeling. Menlo Park, Ordovician chondrite from Brunflo, central Sweden, 1. California: U. S Geological Survey. General description and primary minerals. Lithos 17:87–100. Pel L., Huinink H., and Kopinga K. 2003. Salt transport and Van Schmus W. R. and Wood J. A. 1967. A chemical- crystallization in porous building materials. Magnetic petrologic classification for the chondritic meteorites. Resonance Imaging 21:317–320. Geochimica et Cosmochimica Acta 31:747–765. Petrov M. P. 1976. Deserts of the world. New York: Wiley and Velbel M. A. 1988. The distribution and significance of Sons. evaporitic weathering products on Antarctic meteorites. Pilson M. E. Q. 1998. An introduction to the chemistry of the sea. Meteoritics 23:151–159. Upper Saddle River, New Jersey: Prentice Hall. 431 p. Velbel M. A., Long D. T., and Gooding J. L. 1991. Terrestrial Potocki F. P. 1978. Road temperatures and climatological weathering of Antarctic stone meteorites: Formation of observations in the Emirate of Abu Dhabi. U.K. Transport Mg-carbonates on ordinary chondrites. Geochimica et and Road Research Laboratory Supp. Rept. 412. Cosmochimica Acta 55:67–76. Rodriguez-Navarro C. and Doehne E. 1999. Salt weathering: Walther J. 1900. Das Gesetz der Wustenbildung€ Dietrich Influence of evaporation rate, supersaturation and Reimer (Ernst Vohsen). Berlin. crystallization pattern. Earth Surface Processes and Weisberg M. K., Smith C., Benedix G., Folco L., Righter K., Landforms 24:191–209. Zipfel J., Yamaguchi A., and Aoudjehane H. C. 2009. The Rubin A. E. 1997. Mineralogy of meteorite groups. Meteoritical Bulletin, No. 95. Meteoritics & Planetary Meteoritics & Planetary Science 32:733–734. Science 44:429–462. Rubin A. E., Zolensky M. E., and Bodnar R. J. 2002. The White J. S., Henderson Ep, and Mason B. 1967. Secondary halite-bearing Zag and Monahans (1998) meteorite minerals produced by weathering of the Wolf Creek breccias: Shock metamorphism, thermal metamorphism meteorite. American Mineralogist 52:1190–1197. and aqueous alteration on the H-chondrite parent body. Winkler E. M. 1987. Weathering and weathering rates of natural Meteoritics & Planetary Science 37:125–141. stone. Environmental Geology and Water Sciences 9:85–92. Ruedrich J., Kirchner D., and Siegesmund S. 2011. Physical Wlotzka F. 1993. A weathering scale for the ordinary weathering of building stones induced by freeze-thaw chondrites. Meteoritics 28:460. action: A laboratory long-term study. Environmental Earth Yomogida K. and Matsui T. 1983. Physical-properties of ordinary Sciences 63:1573–1586. chondrites. Journal of Geophysical Research 88:9513–9533. Saunier G., Poitrasson F., Moine B., Gregoire M., and Zolensky M. E., Bodnar R. J., Gibson E. K., Nyquist L. E., Seddiki A. 2010. Effect of hot desert weathering on the Reese Y., Shih C. Y., and Wiesmann H. 1999. Asteroidal bulk-rock iron isotope composition of L6 and H5 ordinary water within fluid inclusion-bearing halite in an H5 chondrites. Meteoritics & Planetary Science 45:195–209. chondrite, Monahans (1998). Science 285:1377–1379.

SUPPORTING INFORMATION Ordinary chondrites from Oman and Tamdakht. Additional supporting information may be found in Table S5: Leaching experiments results: SL091 series, the online version of this article: soil samples from JaH 091 region. Table S1: Leaching experiments results: SLO1 series, Table S6: Leaching experiments results: SLO2 series, soil samples from Oman. soil samples from Oman. Table S2: Oman water composition. Table S7: Leaching experiments results: ML091 series, Table S3: Leaching experiments results: MLT series, meteorites from JaH 091 shower. time series with samples from the JaH 091 (L5 S3 W3/4) Fig. S1: Example of time series experiment for Cl .It shower. shows that even 30 days is at the low end for reaching Table S4: Leaching experiments results: MLO series, equilibrium between rock and water.