Volatile components in the moon: Abundances and isotope ratios of Cl and H in lunar apatites

Introduction: Observations of terrestrial levels of highly volatile elements in lunar apatite, including H [1-3], suggested that some lunar melts were at least “damp” if not “wet” at some point in their histories. On the other hand, the very high δCl values in lunar rocks and apatites (up to ~+30‰ [4]) have been taken to mean that their parent magmas were essentially anhydrous, because such large Cl isotope fractionations could not occur in water-bearing systems. In the laboratory, however, fractionations up to δCl = +9‰ have been seen in the system of HClliquid + HClgas [5], so there might have been lunar conditions that permitted large fractionations of Cl isotopes in the presence of H. Here we examine the relationship between [H] and both the concentration and isotopic composition ([Cl] and δCl) of lunar apatites from 9 thin sections of 8 different lunar samples: 10044,12; 10044,644; 12013,148; 12039,42; 12040,211; 14305,94; 75055,55; 76535,56; & 79215,51. Methods: All [H], [Cl], and δCl values reported here were measured on the Cameca 7f-GEO secondary ion mass spectrometer (SIMS) at Caltech. δCl analyses were performed in separate sessions from the concentration measurements, but in all cases were made on the same crystals. δD values for the first seven sample above are from [3,6], from different crystals in the same thin sections. δD values for 76535 and 79215 were taken at Caltech [7, unpublished data]. Isotopic measurements. δCl was measured with a ~0.7 nA, 10 kV focused Cs+ primary beam with normal-incidence electron gun charge compensation. Samples were pre-sputterred with a 25x25 μm rastering beam of ~2 nA for 10 minutes. Cl and Cl were collected with an electron multiplier (EM) under a mass resolving power (MRP) of ~2000. δD (for 76535 and 79215) was measured on the same apatite grains at an MRP of ~500, after concentration measurements. Concentration measurements. After Cl isotope analysis, C, H, F, S, and Cl were obtained with a 10x10 μm rastering Cs+ beam of ~3 nA. To avoid cracks and contamination, ion images of C, OH, P,and Cl were examined for each grains; the smallest field aperture and electronic gating were used. C, OH, O, F, P, S, and Cl were collected with an EM or a Faraday cup at an MRP of ~5000. Measured ion ratios were calibrated against in-house apatite standards; the largest component of uncertainty arises from blank corrections. H2O values < 100 ppm are within uncertainty of zero. Results: Hydrogen concentrations [H] range from within uncertainty of zero to ~3000 ppm H2O, while [Cl] ranges from 200-22000 ppm (Fig. 1). High [Cl] values (> 2000 ppm) are observed only in apatites with < 500 ppm H2O; conversely, apatites with > 500 ppm H2O have < 1200 ppm Cl (Fig. 1b). δCl values range from +1.6 to +32.8‰ (Fig. 2), consistent with previous observations of very high Cl/Cl ratios in lunar samples [4]. The highest δCl values are exclusively in apatite with low [H] (< 200 ppm H2O), but apatite grains with δCl < +18‰ show a wide range of [H], including the apatites from 12039,42 with 2000-3000 ppm H2O and δCl of +15 to +17‰. The relationship between δCl and [Cl] is somewhat the opposite, with greater dispersion in [Cl] for those apatites with high δCl, including those from 12013,148 and 76535,56 with [Cl]>1.5 wt.% and δCl≥ +25‰. 0 200 400 600 80