Photolytic processing of secondary organic aerosols dissolved in cloud droplets

The effect of UV irradiation on the molecular composition of aqueous extracts of secondary organic aerosol (SOA) was investigated. SOA was prepared by the dark reaction of ozone and d-limonene at 0.05–1 ppm precursor concentrations and collected with a particle-into-liquid sampler (PILS). The PILS extracts were photolyzed by 300–400 nm radiation for up to 24 h. Water-soluble SOA constituents were analyzed using high-resolution electrospray ionization mass spectrometry (HR-ESI-MS) at different stages of photolysis for all SOA precursor concentrations. Exposure to UV radiation increased the average O/C ratio and decreased the average double bond equivalent (DBE) of the dissolved SOA compounds. Oligomeric compounds were significantly decreased by photolysis relative to the monomeric compounds. Direct pH measurements showed that acidic compounds increased in abundance upon photolysis. Methanol reactivity analysis revealed significant photodissociation of molecules containing carbonyl groups and the formation of carboxylic acids. Aldehydes, such as limononaldehyde, were almost completely removed. The removal of carbonyls was further confirmed by the UV/Vis absorption spectroscopy of the SOA extracts where the absorbance in the carbonyln → π* band decreased significantly upon photolysis. The effective quantum yield (the number of carbonyls destroyed per photon absorbed) was estimated as ∼0.03. The total concentration of peroxides did not change significantly during photolysis as quantified with an iodometric test. Although organic peroxides were photolyzed, the likely end products of photolysis were smaller peroxides, including hydrogen peroxide, resulting in a no net change in the peroxide content. Photolysis of dry limonene SOA deposited on substrates was investigated in a separate set of experiments. The observed effects on the average O/C and DBE were similar to the aqueous photolysis, but the extent of chemical change was smaller in dry SOA. Our results suggest that biogenic SOA dissolved in cloud and fog droplets will undergo significant photolytic processing on a time scale of hours to days. This type of photolytic processing may account for the discrepancy between the higher values of O/C measured in the field experiments relative to the laboratory measurements on SOA in smog chambers. In addition, the direct photolysis of oligomeric compounds may be responsible for the scarcity of their observation in the field.