Model study of multiphase DMS oxidation with a focus on halogens

International audienceWe studied the oxidation of dimethylsulfide (DMS) in the marine boundary layer (MBL) with a one-dimensional numerical model and focused on the influence of halogens. Our model runs show that there is still significant uncertainty about the end products of the DMS addition pathway, which is especially caused by uncertainty in the product yield of the reaction MSIA + OH. Under cloud-free conditions a MSA yield of only 5% in this reaction could make the addition pathway in the gas phase dominant for MSA formation whereas a yield of 0% would make the gas phase unimportant. Under cloudy conditions the uptake of DMSO and MSIA to droplets results in a contribution of the gas phase of only about 2% to the total formation rate of MSA. The aqueous phase reaction MSIA + OH is the main source for total MSA when gas phase production of MSA is unimportant. BrO strongly increases the importance of the addition branch in the oxidation of DMS even when present at mixing ratios smaller than 0.5 pmol mol?1. The inclusion of halogen chemistry leads to higher DMS oxidation rates and smaller DMS to SO2 conversion efficiencies. The DMS to SO2 conversion efficiency is also drastically reduced under cloudy conditions. In clouds especially during winter the aqueous phase reaction DMS + O3 contributes 4?18% to total DMS oxidation. In cloud-free model runs between 5 and 15% of the oxidized DMS reacts further to particulate sulfur, in cloudy runs this fraction is almost 100%. In general, more particulate sulfur is formed when halogen chemistry is included. A possible enrichment of HCO3 in fresh sea salt aerosol would increase pH values enough to make the reaction of S(IV)* with O3 dominant for sulfate production. It leads to a shift from MSA to nss-SO4?2 production but increases the total nss-SO4?2 only somewhat because almost all available sulfur is already oxidized to particulate sulfur in the base scenario. We discuss how realistic this is for the MBL. We found the reaction MSAaq + OH to contribute about 10\% to the production of nss-SO4?2 in clouds. It is unimportant for cloud-free model runs. Sulfate production by HOClaq and HOBraq is important in cloud droplets even for small Br? deficits and related small gas phase halogen concentrations. We found differences in the diurnal variation of the Br? in sea salt aerosol with a peak in the morning when the loss of Br? from the sea salt is small and a peak during day when the loss is almost complete. Overall we find that the presence of halogens lead to processes that decrease the albedo of stratiform clouds in the MBL