Differential behavior of tungsten stable isotopes during sorption to Fe versus Mn oxyhydroxides at low ionic strength

Abstract

Tungsten is an emerging contaminant of interest with serious health implications whose environmental transport and fate have scarcely been studied. Sorption to mineral particles is likely the primary process controlling tungsten's mobility in oxidizing soils and aquifers, but the few papers published thus far have not yet resolved the relationships between aqueous speciation, substrate phases, and sorption complexes at conditions relevant to these settings. In many cases, these relationships are best investigated with techniques such as X-ray spectroscopy, but the low concentrations of tungsten in typical contaminated settings limit the applicability of that approach. Here, we lay initial groundwork for eventual use of the stable isotope geochemistry of tungsten to infer sorption mechanisms at field-relevant conditions, because metal stable isotope systematics are sometimes very sensitive to speciation and surface complexation geometry. We report the first tungsten stable isotope fractionations for soil-relevant conditions using simple batch experiments, in which tungsten interacted with either birnessite (Mn oxyhydroxide) or ferrihydrite (Fe oxyhydroxide) nanoparticles, at pH 5 or 8 and very low ionic strength.In 120-h experiments, in which proportions of sorbate and sorbent were varied, lighter isotopes of W preferentially sorbed on ferrihydrite. On a plot of δ183/182W against % W sorbed, parallel lines fit the pH 8 data from ferrihydrite experiments well, with a best fit Δ183/182Wdissolved-sorbed of +0.39‰ (R2 = 0.95), indicating an equilibrium fractionation (reversible sorption in a closed system). Parallel, linear trends also fit the data from ferrihydrite experiments at pH 5 well, with a best-fit fractionation of +0.32‰ (R2 = 0.93). When error bars are propagated rigorously, these two Δ183/182Wdissolved-sorbed values are indistinguishable, indicating that fractionation is insensitive to pH in the W-ferrihydrite system over the relevant range of pH and suggesting that sorption mechanism may also not vary with pH. In every birnessite experiment, lighter isotopes preferentially sorbed, but fractionation magnitude varied systematically with fraction of W sorbed, with larger fractionations at larger fraction of total W sorbed. At pH 8, the range in Δ183/182Wdissolved-sorbed was +0.12 to +0.40‰, with an average of 0.31‰, and at pH 5, the range was +0.29 to +0.67‰, and the average was +0.47 ± 0.10‰. Isotope behavior and sorption mechanisms appear to be different for birnessite and more complicated compared to ferrihydrite; neither parallel lines (equilibrium trends) nor Rayleigh trends obviously fit all the data well. This observation led us to consider the possibility of a kinetic isotope effect superimposed on equilibrium fractionation, where the expressed mixture of kinetic and equilibrium fractionation varies with both surface loading and time.To explore further, we conducted experiments at pH 5 with fixed W/birnessite proportions and durations up to 504h. Results indicate that the amount of W sorbed to birnessite continues to increase gradually for at least 3 weeks. The magnitude of isotope fractionation also appears to increase gradually over time, from ~0.3‰ at ≤48 h. to ~0.6‰ at >100 h. We use multiple lines of indirect evidence to hypothesize that W initially sorbs directly to birnessite surfaces, but gradually a polymeric W-O surface precipitate phase assembles and grows, as observed in previous publications for W sorption on other substrates. In future work, we will attempt to confirm this hypothesis with X-ray spectroscopic analysis. Once the relationship between isotopic fractionation and sorption mechanism(s) are constrained, use of tungsten isotopes to probe sorption mechanisms at field-relevant concentrations of tungsten may be possible. In addition, inclusion of sorption-driven isotope systematics in reactive transport models may be useful in tracing W mobility in contaminated soils and sediments.