 ACIDIC ALTERATION OF THE JSC-1 SIMULANT DOWN TO -18°C: CONSEQUENCES FOR THE MARTIAN ALTERATION HISTORY.   G. Berger and L. Ayang-Nzamé, IRAP, CNRS, 14 Av. E. Belin, 31400 Toulouse, France (gilles.berger@irap.omp.eu)        Introduction: Phyllosilicates and sulfate minerals discovered in several places on the Mars surface are used, besides the classical geomorphologic observations, to constrain the water history. Phyllosilicates are generally observed from orbit in old Noachian terrain, reviewed in [1], and more recently trioctahedral smectites were even characterized in drilled samples by the Curiosity rover at Gale [2]. For sulfate, many different minerals have been detected on the surface of Mars in the Burns Formation, at the Phoenix landing site, in Endeavour Crater, in North Polar dunes and are reviewed in [3]. However, most of these sulfate occurences are not commonly associated with clays. Even at the observation scale of Curiosity in Gale crater, both clay and sulfate minerals are present but are attributed to different stages of the sediment's history [4].     A global view of climate versus alteration history was proposed in [5], suggesting that clays, sulfates and oxides correspond of successive geological periods in response to a global climate change. By contrast, and from a chemical point of view, sulfate and clay minerals may be co-eval by-products of a single reaction with acid fogs [6]. Given the potential mobility of soluble sulfates, compared with phyllosilicates, the observations made around the planet do not reflect necessarily the first conditions of formation. In this study we focus on the effect of an acid fog on a basaltic material having undergone a previous, not necessarily acidic, alteration history. We carried out experimental interactions of an H2SO4 brine with a material used as a simulant of the Martian regolith, the JSC-1, between -17 and +40°C. The originality of these experiments is to use a natural material previously altered and containing allophones, and under negative temperatures (in Celsius) where the brine has a consistence of ice. The motivation of this study was, besides the possible sulfate formation, to investigate if the crystalline phases are converted into clay minerals or amorphous phases, or if the allophanes recristallize in more stable hydrous phases under acidic conditions.    Methodology: The JSC-1 simulant is the <1 mm fraction of weathered volcanic ash from Pu'u Nene, Hawaii. It is considered as a Martian regolith analogue  [7].  The grains are composed of fine crystals of plagioclase, pyroxene and olivine partly cemented by a vitreous phase. Under the scanning electron microscopy (SEM) an altered rim can be observed at the surface of the grains and sometime larger areas of non-igneous material fill the cavity of porous grains (Fig.1). The simulant was reacted with 10-2, 10-1 and 1m H2SO4 solutions under a water-rock ratio of 5. The runs were conducted betwwen 40° and -18°C for 1 week or 8 months. At -18°C the solution was frizen: hard ice for 0.1m and 0.01m H2SO4, soft ice for 1 m H2SO4.   The relative proportions of allophane, plagioclase, pyroxene and olivine were empirically determined by their X-ray diffraction patterns ratio and monitored during the runs. The final fluid composition was determined by ion chromatography and optic ICP spectrometry. The degree of reaction was indicated by pH monitoring.    Fig. 1: Polished section of the JSC-1 simulant.     Mineralogical changes: XRD analyses of the altered grains did not provide evidence of a significant change in the relatice proportions of the starting phases, pristine minerals or allophane, whatever the duration and temperature of the runs. By contrast, secondary gypsum was identified in the first days of reactions in all of the 1m H2SO4 runs, even at -18°C (Fig.2).   Fig. 2: Gypsum crystals produced at -18°C at the surface of the JSC-1 simulant, reacted with 1m H2SO4. This fast reaction is accompanied by a concomitant increase of pH that increases very slowly afterwards. This pH increase likely accounts for the low degree of other mineral reactions, even after 8 months. Microanalyses by EDS under the SEM suggested that the composition of the altered rim at the grain surface or within the vacuoles changed and was the source of calcium for this reaction (Fig.3).   Fig. 3: Polished section of the simulant altered at 40°C, 12 days, 1m H2SO4. The altered rim are depleted in Ca and Mg.     Fluid composition: In addition to the removal of aquous sulfate, the brines and soft ice were enriched in Mg, as well as Al and Fe at low pH. An example is given in Fig.4.  The brines having been filtered at 0.2 microns just after sampling, the high concentration of Fe and Al in the more acidic (collored) samples suggests the presence of colloids. Ultrafiltration at 3000 dalton did not purify these brines that remained reddish. However, XRD analyses of the residue (Fig.5) revealed a very well cristallized product, still under investigation.     Fig. 4: Example of final brine compositions, as a function of starting acidity.        Discussion and Conclusions  These simple experiments conducted at a small laboratory scale may infom on what may have occurred at a larger scale on the Mars surface. Several conclusions may be drawn: 1- If the martian regolith has undergone a previous alteration stage under an humid climate (Noachian?) leading to the precipitation of allophanes, this amorphous constituent may act as an ionic exchanger and may quickly neutralize further acidic brines. The main mineral reaction is then the precipitation of gypsum, that may be itself easily transported by further dissolution-precipitation reactions. The remaining Mg-brine may be the source of Mg-sulfate deposits as observed in the Burns Formation. 2- The ultrafine particles produced by the acidic attack of the regolith and the precipitation of colloids may contribute to the martian dust under a dry climate. 3- Such a proton/cations exchange is possible in cryosphere provided that a small fraction of the brine is still liquid. But the precipitation of salt (sulfate here) decreases the sallinity of the brine and increases the viscosity of the ice/brine mixture. These reactions should be taken into account when modeling the fluid or ice dynamics.  References:[1] Ehlmann et al. 2013, Space Sci. Reviews 174, 329-364. [2] Vaniman et al., 2013, Science, doi:10.1126/ science.1243480. [3] Gaillard et al., 2013, Space Sci. Reviews 174, 251-300. [4] McLennan et al., 2013, Science, doi: 10.1126/science.1244734. [5] Biebring et al., 2006, Science 312, 400-404. [6] Berger et al., 2009, Amer. Mineral. 94, 1279-1282. [7] Morris et al. (2001), JGR 106, 5057-5083. 
