Experimental Study of Water Transport across an Adsorbing Regolith

Introduction: The presence of subsurface water ice on Mars has now been recognized by several instruments, including in situ observation by the Phoenix Lander in high-latitude region [1]. This ice could represent a significant water reservoir and might participate actively in the global Mars water cycle [2]. Subsurface water ice is supposed to be buried below an ice-free regolith layer, produced by the propagation of diurnal, seasonal and secular surface temperature fluctuations. The exchange of water molecules between the subsurface and the atmosphere then occurs by diffusion through a dry regolith, which will control the coupling between the atmosphere and the subsurface. Here, we report on water adsorption isotherm measurements of Martian soil analogs which we use to infer the transfer properties of the Martian subsurface. In addition, we have measured kinetics of water adsorption for the same materials, which sheds light on the possible timescale for subsurface/atmosphere exchange of water. Methods: A simulation chamber in relation with an adsorption setup was designed and built as a complement to the LPG reflectance spectrometer. Extensive description of the chamber can be found in [3]. The source of water vapor is a volume of ultra pure, demineralized and carefully outgassed liquid water maintained at a temperature of +20°C. Prior to any injection of water, continuous reflectivity measurement at one wavelength inside a hydration band is started (at 1.93 or 3.10 μm). Both thermodynamics and spectroscopy are used to determine the amount of water in the sample at each step of the hydration / dehydration process. For each adsorption step, the quantity of water adsorbed by the sample can simply be calculated from the difference of pressure in the chamber before and after sorption of water vapor on the sample. This calculation requires a good knowledge of the different volumes (which were previously measured). In order to quantify the water content of the sample surface, reflectance was monitored inside one of the strong hydration band (at 1.93 or 3.10 μm). Because these wavelengths correspond to very strong absorption levels, only a few grains are involved in the reflection process. We then obtained the “intrinsic” adsorption kinetics on the grains, i.e. a kinetic that is not biased by chemical diffusion though the sample porosity. The conversion of reflectance at a single wavelength into water content is made by using an empirical relationship between the strength of absorption bands and the water content from our previous study [3]. Samples: Six different types of materials were studied. They are: a basaltic volcanic tuff from the Massif Central, a synthetic ferrihydrite sample, the SWy-2 Na-Montmorillonite, the JSC-Mars 1 palagonitic soil, a Mg-sulfate/basaltic tuff mixture, and a dunite powder. All of these materials are either suspected to be major mineral components of the Martian regolith or are seen as good analogs of Martian surface materials. For each sample, the thickness in the holder is 1 mm and sample mass ranges between 0.4 and 1.0 mg. Results: The isotherms obtained have been described elsewhere [3] and will not be discussed in details. For each material, these isotherms were fitted successfully according to the Langmuir theory [4]. ρa= C[αP/(1+αP)] (eq. 3) Surface kinetic: Timescales for the adsorption to reach equilibrium range between a few tens of seconds and a few minutes for the samples studied. In order to characterize and compare the kinetics of the process, we modelled each of the adsorption step by a second-order kinetic law. This choice was made because it was found to show the best fit among typical kinetical laws. ( ) i f surf i f t t ρ ρ τ ρ ρ − + + = (eq. 4)