Equilibrium isotope fractionation factors of H exchange between steam and soil clay fractions.

Abstract

Steam equilibration overcomes the problem of the traditional measurements of H isotope compositions, which leave an arbitrary amount of adsorbed water in the sample, by controlling for the entire exchangeable H pool, including adsorbed water and hydroxyl-H. However, the use of steam equilibration to determine nonexchangeable stable H isotope compositions in environmental media (expressed as δ2 Hn values) by mathematically eliminating the influence of exchangeable H after sample equilibration with waters of known H-isotopic composition requires the knowledge of the equilibrium isotope fractionation factor between steam-H and exchangeable H of the sample (αex-w ), which is frequently unknown. We developed a new method to determine the αex-w values for clay minerals, topsoil clay fractions, and mica by manipulating the contributions of exchangeable H to the total H pool via different degrees of post-equilibration sample drying. We measured the δ2 H values of steam-equilibrated mineral and soil samples using elemental analyzer-pyrolysis-isotope ratio mass spectrometry. The αex-w values of seven clay minerals ranged from 1.071 to 1.140, and those of 19 topsoil clay fractions ranged from 0.885 to 1.216. The αex-w value of USGS57 biotite, USGS58 muscovite, and of cellulose was 0.965, 0.871, and 1.175, respectively. The method did not work for kaolinite, because its small exchangeable H pool did not respond to the selected drying conditions. Structurally different mineral groups such as two- and three-layer clay minerals or mica showed systematically different αex-w values. The αex-w value of the topsoil clay fractions correlated with the soil clay content (r=0.63, P= 0.004), the local mean annual temperature (r = 0.68, P= 0.001), and the δ2 H values of local precipitation (r = 0.72, P< 0.001), likely to reflect the different clay mineralogy under different weathering regimes. Our new αex-w determination method yielded realistic results in line with the few previously published values for cellulose. The determined αex-w values were similar to the widely assumed values of 1.00-1.08 in the literature, suggesting that the adoption of one of these values in steam equilibration approaches is appropriate.