Experimental investigation of Ca isotopic fractionation during abiotic gypsum precipitation

Experiments investigating Ca isotopic fractionation during gypsum precipitation were undertaken in order to elucidate the mechanisms and conditions that govern isotopic fractionation during mineral precipitation. Both stirred and unstirred free drift gypsum precipitation experiments were conducted at constant initial ionic strength (0.6M) and variable initial saturation states (4.8-1.5) and Ca²⁺:SO₄^2- ratios (3 and 0.33). Experimental durations varied between 0.5 and 190h, while temperature (25.9-24.0°C), pH (5.8-5.4) and ionic strength (0.6-0.5) were relatively constant. In all experiments, 20-80% of the initial dissolved Ca reservoir was precipitated. Isotopically light Ca preferentially partitioned into the precipitated gypsum; the effective isotopic fractionation factor (δ^44/40Ca_s-f=δ^44/40Ca_solid-δ^44/40Ca_fluid) of the experimental gypsum ranged from -2.25‰ to -0.82‰. The log weight-averaged, surface area normalized precipitation rates correlated with saturation state and varied between 4.6 and 2.0μmol/m²/h. The crystal size and aspect ratios, determined by SEM images, BET surface area, and particle size measurements, co-varied with precipitation rate, such that fast growth produced small (10-20μm), tabular crystals and slow growth produced larger (>1000μm), needle shaped crystals.Mass balance derived δ⁴⁴Ca_s and δ⁴⁴Ca_s-f, calculated using the initial fluid δ⁴⁴Ca and the mass fraction of Ca removed during precipitation (f_Ca) as constraints, suggest that the precipitate was not always sampled homogeneously due to the need to preserve the sample for SEM, surface area, and particle size analyses. The fractionation factor (α_s-f), derived from Rayleigh model fits to the fluid and calculated bulk solid, ranged from 0.9985 to 0.9988 in stirred experiments and 0.9987 to 0.9992 in unstirred experiments. The α_s-f demonstrated no clear dependence on either precipitation rate or initial saturation state in stirred reactors, but exhibited a positive dependence on rate in unstirred experiments. The differences in α_s-f between stirred and unstirred reactors, as well as a general correlation between α_s-f and crystal morphology, led us to hypothesize that growth on different crystal faces controls the isotopic composition of gypsum. We also explore the idea that speciation in solution explains the difference between experiments in which the only major difference was the Ca²⁺ to SO₄^2- ratio in solution.The importance of understanding the environmental controls on the fractionation factor during mineral precipitation is highlighted in this study. The fractionation factor of gypsum precipitation near chemical equilibrium was found to be ~0.9995, rather than 1, indicating that even at near equilibrium conditions, the δ⁴⁴Ca of minerals are not likely to record the δ⁴⁴Ca of the solution directly. However, the measurable isotopic fractionation associated with gypsum formation does suggest that a gypsum-based proxy may be useful in constraining Ca cycling in marginal environments over geologic time scales. Model examples are provided that demonstrate how such a proxy would operate.