Theoretical estimates of sulfoxyanion triple-oxygen equilibrium isotope effects and their implications

Abstract

Triple-oxygen isotope (δ18O and Δ′17O) analysis of sulfate is becoming a common tool to assess several biotic and abiotic sulfur-cycle processes, both today and in the geologic past. Multi-step sulfur redox reactions often involve intermediate sulfoxyanions such as sulfite, sulfoxylate, and thiosulfate, which may rapidly exchange oxygen atoms with surrounding water. Process-based reconstructions therefore require knowledge of equilibrium oxygen-isotope fractionation factors (18α and 17α) between water and each individual sulfoxyanion. Despite this importance, there currently exist only limited experimental 18α data and no 17α estimates due to the difficulty of isolating and analyzing short-lived intermediate species. To address this, we theoretically estimate 18α and 17α for a suite of sulfoxyanions—including several sulfate, sulfite, sulfoxylate, and thiosulfate species—using quantum computational chemistry. We determine fractionation factors for sulfoxyanion “water droplets” using the B3LYP/6-31G+(d,p) method; we additionally calculate higher-order method (CCSD/aug-cc-pVTZ and MP2/aug-cc-pVTZ) scaling factors, and we qualitatively estimate the importance of anharmonic zero-point energy (ZPE) corrections using a suite of gaseous sulfoxy compounds. Methodological scaling factors greatly impact 18α predictions, whereas ZPE corrections are likely small (i.e., ⩽1‰) at Earth-surface temperatures; existing experimental data best agree with 18α predictions when including redox state-specific CCSD/aug-cc-pVTZ scaling factors. Theoretical pH- and temperature-specific bulk-solution (i.e., abundance-weighted average of all species) 18α values yield root-mean-square errors for sulfate/water, sulfite/water, and thiosulfate/water equilibrium of 4.5‰ (n=18 experimental conditions), 3.7‰ (n=27), and 2.2‰ (n=3), respectively. However, sulfate- and sulfite-system agreement improves considerably when comparing experimental results only to SO3(OH)−/H2O (RMSE = 1.6‰) and SO2(OH)−/H2O (RMSE = 2.2‰) predictions, rather than bulk solutions. This is particularly true for the sulfite system at high and low pH, when SO2(OH)− is not the dominant species. We discuss potential experimental and theoretical biases that may lead to this apparent improvement. By combining 18α and 17α predictions, we additionally estimate that sulfate, sulfite, sulfoxylate, and thiosulfate species can exhibit Δ′17O values as much as 0.199‰, 0.205‰, 0.101‰, and 0.186‰ more negative than equilibrated water at Earth-surface temperatures (reference line slope = 0.5305). This theoretical framework provides a foundation to interpret experimental and observational triple-oxygen isotope results of several sulfur-cycle processes including pyrite oxidation, microbial metabolisms (e.g., sulfate reduction, thiosulfate disproportionation), and hydrothermal anhydrite precipitation. We highlight this with several examples.