Oxygen isotope fractionation between water and the aqueous hydroxide ion

Abstract

The stable oxygen isotope fractionation factor between water and the aqueous hydroxide ion is a fundamental geochemical parameter widely used in the Earth Sciences and other fields. Studies from the 1960s reported α(H2O(l)-OH-(aq)), the fractionation factor between liquid H2O and aqueous OH−, theoretically as 1.046 and ∼1.048 (46‰ and 48‰ at 15 °C and 25 °C) and experimentally as 1.045 (45‰) at 15 °C. These, or similar values have been used in the literature for almost 60 years. Here I present quantum-chemical (QC) calculations, showing that the above theoretical values are fundamentally wrong as they pertain to free OH− (incorrectly assumed equivalent to aqueous OH−) and ignore intermolecular vibrational modes in solution. As a result, the theoretical values from the 1960s are off by a factor of ∼2 (when expressed in ‰), suggesting that the experimental value is also wrong. QC computations of OH−-water clusters with up to n=22 water molecules demonstrate that hydrogen bonding in solution significantly affects the fundamental vibrational modes associated with OH− and substantially reduces the oxygen isotope fractionation between water and OH-(aq), compared to the fractionation between water and free OH−. The most accurate QC methods tested here yield values for the fractionation factor between H2O and OH− in water clusters with n=7 to 22 water molecules of ∼1.019 to ∼1.024 (19 to 24‰) at 25 °C based on the harmonic approximation. Estimated effects due to anharmonicity (from numerically demanding computations) could add uncertainties of up to ∼3‰ to these values.