Direct observations of the atmospheric processing of Asian mineral dust

Abstract. The accumulation of secondary acid products and ammonium on individual mineral dust particles during ACE-Asia has been measured in real-time using ATOFMS. Changes in the amounts of sulphate, nitrate, and chloride mixed with dust particles corresponded to different air mass source regions. During volcanically influenced periods, dust mixed with sulphate dominated. This rapidly switched to dust predominantly mixed with chloride when the first Asian dust front reached the R/V Ronald Brown. We hypothesise that the high degree of mixing of dust with chloride was caused by the prior reaction of NOy(g) and volcanic SO2(g) with sea salt particles, reducing the availability of nitrate and sulphate precursors while releasing HCl(g), which then reacted with the incoming dust front. The segregation of sulphate from nitrate and chloride in individual dust particles is demonstrated for the first time. This is likely caused by the dust plume encountering elevated SO2(g) in the Chinese interior before reaching coastal urban areas polluted by both SO2(g) and NOx(g). This caused the fractions of dust mixed with nitrate and/or chloride to be strongly dependent on the total dust loadings, whereas dust mixed with sulphate did not show this same dust concentration dependence. Ammonium was also significantly mixed with dust and the amount correlated strongly with the total amount of secondary acid reaction products in the dust. Submicron dust and ammonium sulphate were internally mixed, contrary to frequent statements that they exist as an external mixture. The size distribution of the mixing state of dust with these secondary species validates previous models and mechanisms of the atmospheric processing of dust. The uptake of secondary acids was also dependent on the individual dust particle mineralogy; nitrate accumulated on calcium-rich dust while sulphate accumulated on aluminosilicate-rich dust. Oxidation of S(IV) to S(VI) by iron in the aluminosilicate-rich dust is a probable explanation for this result, with important consequences for dust as a vector for the fertilization of remote oceans by soluble iron. This series of novel results has important implications for improving the treatment of dust in global chemistry models and highlights several key processes requiring further investigation through laboratory and field studies.