Triple oxygen isotope insight into terrestrial pyrite oxidation

Abstract

The mass-independent minor oxygen isotope compositions (Δ′17O) of atmospheric O2andCO2are primarily regulated by their relative partial pressures,pO2/pCO2. Pyrite oxidation during chemical weathering on land consumesO2and generates sulfate that is carried to the ocean by rivers. The Δ′17O values of marine sulfate deposits have thus been proposed to quantitatively track ancient atmospheric conditions. This proxy assumes directO2incorporation into terrestrial pyrite oxidation-derived sulfate, but a mechanistic understanding of pyrite oxidation—including oxygen sources—in weathering environments remains elusive. To address this issue, we present sulfate source estimates and Δ′17O measurements from modern rivers transecting the Annapurna Himalaya, Nepal. Sulfate in high-elevation headwaters is quantitatively sourced by pyrite oxidation, but resulting Δ′17O values imply no direct troposphericO2incorporation. Rather, our results necessitate incorporation of oxygen atoms from alternative,17O-enriched sources such as reactive oxygen species. Sulfate Δ′17O decreases significantly when moving into warm, low-elevation tributaries draining the same bedrock lithology. We interpret this to reflect overprinting of the pyrite oxidation-derived Δ′17O anomaly by microbial sulfate reduction and reoxidation, consistent with previously described major sulfur and oxygen isotope relationships. The geologic application of sulfate Δ′17O as a proxy for pastpO2/pCO2should consider both 1) alternative oxygen sources during pyrite oxidation and 2) secondary overprinting by microbial recycling.

Sulfate sulfur isotopes and major ion chemistry reveal that pyrite oxidation counteracts CO2 drawdown from silicate weathering in the Langtang-Trisuli-Narayani River system, Nepal Himalaya

“…8d) formation by the fractionating process of microbial sulfate reduction (Thode et al, 1951;Sim et al, 2011), and could feasibly be 34 S-depleted relative to the sulfide minerals of the HHC. Previous work in the Marsyandi river has shown that SO 4 2derived from FeS 2 oxidation in the TSS reaches δ 34 S values as low as approximately -25‰, consistent with but lower than our minimum estimates of δ 34 S FeS2 in downstream stations (Turchyn et al, 2013;Hemingway et al, 2020).…”

“…The elevated SO 4 2-/Σ + and low δ 34 S SO4 of samples thus requires an additional source of 34 S-depleted sulfur. As previously suggested for the Langtang River (Bhatt et al, 2000; and shown in the Marsyandi catchment (Turchyn et al, 2013;Hemingway et al, 2020), a likely source of this sulfur is the oxidation of FeS 2 . The oxidation of FeS 2 to H 2 SO 4 and subsequent weathering of minerals could increase river SO 4 2-/Σ + and, depending on the value of δ 34 S FeS2 , decrease δ 34 S SO4 to observed values (Fig.…”

“…No direct observations of halite or gypsum are reported for TSS in the Narayani catchment, although some work suggests its presence (Tshering & Bhandari, 1973in Fort, 1996Wolff-Boenisch et al, 2009; Appendix 2). The TSS was previously found to contain FeS 2 (Bordet et al, 1971) and FeS 2 oxidation is reported in the Marsyandi River (Turchyn et al, 2013;Hemingway et al, 2020), which is included in the headwaters of the catchment area captured by the most downstream sampling site in this study, LNS-1. The fraction of each catchment occupied by TSS decreases downstream from LNS-6, although the particular fraction depends on how much unmapped area is assigned to TSS.…”

“…Unlike [HCO 3 -], dissolved major ions are largely conservative in most river systems and can be used to reconstruct the ΔALK/ΔDIC ratio of weathering. To do so for samples from the Langtang-Trisuli-Narayani River system, the contribution of distinct lithologies to the river dissolved load was calculated using a Monte-Carlo inversion model (Négrel et al, 1993;Gaillardet et al, 1999;Torres et al, 2016;Burke et al, 2018;Hemingway et al, 2020). This inversion is analogous to prior studies of river chemistry both globally (Gaillardet et al, 1999) and in the Himalaya specifically (Bickle et al, 2015), but differs from the forward-modeling common in some prior work in the region (Galy & France-Lanord, 1999;Tipper et al, 2006).…”

“…Sulfur isotope ratios of dissolved riverine SO 4 2-(δ 34 S SO4 = 34 R sample / 34 R VCDT -1, commonly reported in ‰ relative to Vienna Canyon Diablo Troilite (VCDT) and where 34 R is the 34 S/ 32 S ratio of SO 4 2-) are a useful tool for studying FeS 2 oxidation. Previously, δ 34 S SO4 was measured in the Marsyandi River of the Nepal Himalaya (Turchyn et al, 2013;Hemingway et al, 2020). Because the studied region of the Marsyandi represents only a relatively small fraction of the total Narayani catchment, the Nepal Himalaya remains largely uncharacterized for δ 34 S SO4 .…”