Fractionation of sulfur isotopes during heterogeneous oxidation of SO<sub>2</sub> on sea salt aerosol: a new tool to investigate non-sea salt sulfate production in the marine boundary layer

Abstract

Abstract. The oxidation of SO 2 to sulfate on sea salt aerosols in the marine environment is highly important because of its effect on the size distribution of sulfate and the potential for new particle nucleation from H 2 SO 4 (g). However, models of the sulfur cycle are not currently able to account for the complex relationship between particle size, alkalinity, oxidation pathway and rate -which is critical as SO 2 oxidation by O 3 and Cl catalysis are limited by aerosol alkalinity, whereas oxidation by hypohalous acids and transition metal ions can continue at low pH once alkalinity is titrated. We have measured 34 S/ 32 S fractionation factors for SO 2 oxidation in sea salt, pure water and NaOCl aerosol, as well as the pH dependency of fractionation.Oxidation of SO 2 by NaOCl aerosol was extremely efficient, with a reactive uptake coefficient of ≈0.5, and produced sulfate that was enriched in 32 S with α OCl = 0.9882±0.0036 at 19 • C. Oxidation on sea salt aerosol was much less efficient than on NaOCl aerosol, suggesting alkalinity was already exhausted on the short timescale of the experiments. Measurements at pH = 2.1 and 7.2 were used to calculate fractionation factors for each step from SO 2 (g) → multiple steps → SO 2− 3 . Oxidation on sea salt aerosol resulted in a lower fractionation factor than expected for oxidation of SO 2− 3 by O 3 (α seasalt = 1.0124±0.0017 at 19 • C). Comparison of the lower fractionation during oxidation on sea salt aerosol to the fractionation factor for high pH oxidation shows HOCl contributed 29 % of S(IV) oxidation on sea salt in the short experimental timescale, highlighting the potential importance of hypohalous acids in the marine environment.The sulfur isotope fractionation factors measured in this study allow differentiation between the alkalinity-limited pathways -oxidation by O 3 and by Cl catalysis (α 34 = 1.0163 ± 0.0018 at 19 • C in pure water or 1.0199 ± 0.0024 at pH = 7.2) -which favour the heavy isotope, and the alkalinity non-limited pathways -oxidation by transition metal catalysis (α 34 = 0.9905±0.0031 at 19 • C, Harris et al., 2012a) and by hypohalites (α 34 = 0.9882±0.0036 at 19 • C) -which favour the light isotope. In combination with field measurements of the oxygen and sulfur isotopic composition of SO 2 and sulfate, the fractionation factors presented in this paper may be capable of constraining the relative importance of different oxidation pathways in the marine boundary layer.

Investigating Source Contributions of Size‐Aggregated Aerosols Collected in Southern Ocean and Baring Head, New Zealand Using Sulfur Isotopes

“…Overall, because of the low contribution of SO 4 2À anth , the seasonal variation of sulfate concentration at Baring Head is mainly controlled by the variation of SO 4 2À DMS . However, uncertainties exist in our source appointment because (1) the sampling intervals were~1 month for the marine collector and 1 week for the ambient collector, thus the marine collector could not record the accurate δ 34 S DMS end-member; (2) the atmospheric chemistry of sulfate formation in the urban areas and the open ocean was different because of the differences in NO x , volatile organic carbon, and O 3 concentrations, hence the isotopic fractionation between SO 2 and sulfate may be different (Harris et al, 2012); and (3) the δ 34 S of anthropogenic sulfate could display a wider range (Calhoun et al, 1991;Norman et al, 1999;Wadleigh, 2004). Nevertheless, these uncertainties would not significantly impact our calculation.…”

Seasonality in the Δ<sup>33</sup>S measured in urban aerosols highlights an additional oxidation pathway for atmospheric SO<sub>2</sub>

“…This shows a depletion of Cl in most of our samples, which could be explained by a reaction between NaCl and sulfate/nitrate that forms NaSO4 or NaNO3, a process also known as the "aging of sea-salt particles" (Sinha et al, 2008;Song and Carmichael, 1999 of secondary aerosols precursors, which would form sulfates in a marine environment. In this context sulfates could either result from the oxidation of SO2 formed by the oxidation of DMS or from the oxidation of anthropogenic SO2 (Harris et al, 2012c). DMS is characterized by a mean δ 34 S-value of 18‰ and a mean Δ 33 S-value of 0.03‰ (Oduro et al, 2012) and anthropogenic SO2 by variable δ 34 S-values and a mean Δ 33 S-value of -0.018‰ (Lin et al, 2018b).…”

“…Chemical analyses of our samples show a correlation between sulfate and Na concentrations, suggesting sodium sulfate is probably the main form of sulfate in our samples. The formation of sodium sulfate result from the SO2 oxidation coming either from direct emission or DMS oxidation on sea-salt aerosols surface (Harris et al, 2012d;Sinha et al, 2008). Thus, the aging process, which induces a Cl depletion, might explain such Δ 33 S-values.…”

“…The alternative interpretation relies on a constant SO2 isotope composition. The variations observed in the sulfur multiple isotope compositions (δ 34 S and Δ 33 S) of rural sulfate aerosols reflect changes in the atmospheric concentrations of SO2 oxidants, each having distinct fractionation factors (Harris et al, 2012b;Harris et al, 2012d;Harris et al, 2013a;Harris et al, 2013b). In this case, high (up to ~7‰) and low (down to ~1‰) δ 34 S-values are predicted during winter and summer, respectively (Harris et al, 2013a).…”

Background atmospheric sulfate deposition at a remote alpine site in the Southern Canadian Rocky Mountains

“…Oxidation of atmospheric SO 2 to SO 4 2À can occur by homogenous (gas phase) or heterogeneous (gas and aqueous phase) pathways, and based on the studies by Harris et al [2012aHarris et al [ , 2012cHarris et al [ , 2013b, this can lead to distinct isotope differences for δ 34 S-SO 4 2À in the atmosphere and in precipitation. The most significant homogenous SO 2 oxidation pathway is initiated by the hydroxyl radical (OH) [Atkinson et al, 2004] and follows several steps with isotopic fractionation occurring in the slowest step (R1) [Tanaka et al, 1994].…”