Title: Oxygen isotope fractionation in the portlandite–water and brucite–water systems from 125 to 450 °C, 50 MPa
Abstract: Equilibrium oxygen isotope fractionation factors were determined for the portlandite–water and brucite–water systems from 125 to 425 °C, 50 MPa using the partial exchange technique. Reagent grade cryptocrystalline Ca(OH)2 and amorphous Mg(OH)2 were reacted with three waters having different initial δ18O compositions. Isotope exchange occurred via recrystallization with exchange varying from 40% to 95% at 200 to 425 °C, respectively. Equilibrium 18O brucite–water fractionation factors (103lnα) increase from −4.7 ± 3.5‰ at 200 °C to −3.5 ± 2.5‰ at 425 °C. These data connect smoothly with previous experimental calibrations at lower and higher temperatures to define a single function valid from 15 to 450 °C, as follows:103lnαbrucite-water=4.39×106T2-16.95×103T+11.19where T is temperature in Kelvin. These results confirm the existence of a broad minimum in the fractionation factor for brucite at ∼250 °C. The equilibrium 18O fractionation factor for portlandite–water varies from −11.1 ± 2.7‰ at 125 °C to −6.6 ± 0.1‰ at 425 °C, and can be described by the following function:103lnαportlandite-water=5.61×106T2-26.29×103T+19.72where T is temperature in Kelvin. These experimental results indicate that brucite favors 18O relative to portlandite with brucite–portlandite fractionation decreasing from 8‰ to 3‰ from 125 to 425 °C. A significant temperature dependent cation mass effect is therefore indicated for cation–OH bonds in hydroxide minerals. The observed fractionation is consistent with quantum theory which predicts that bonds with less massive cations have higher vibrational frequencies and will display a relative affinity for 18O to stabilize the structure. Brucite–portlandite 18O fractionation predicted using the increment method is extremely small, opposite in sign (−0.1‰ to −0.2‰), and shows very little dependence on temperature, in poor agreement with the experimental calibration. This indicates that the method does not adequately account for the effect of cation mass on 18O fractionation within hydroxide minerals. It is suggested that cation-specific parameters within the increment method could be fit to the experimental calibrations reported here to improve prediction of fractionation factors for hydroxides and hydroxyl-bearing aluminosilicates, particularly at low temperate where the cation-mass effect is more significant.