Measurements of SOA photolysis rates and their atmospheric implications

 

Authors

John E Shilling — Pacific Northwest National Laboratory
Maria Anna Anna Zawadowicz — Brookhaven National Laboratory
Connor J. Flynn — University of Oklahoma School of Meteorology
Manishkumar Shrivastava — Pacific Northwest National Laboratory
Ben Hwan Lee — University of Washington
Joel Thornton — University of Washington

Category

Secondary organic aerosol

Description

Secondary organic aerosol (SOA) accounts for a large fraction of tropospheric particulate matter and is largely responsible for the growth of particles into climate-relevant sizes. Accurate representation of SOA life cycle in models is needed to quantify its effects on radiative forcing and air quality. While SOA formation and growth mechanisms have been and continue to be extensively investigated, loss pathways typically receive less attention. Most current large-scale models account for dry and wet deposition, but not for reactive losses, such as photolysis. Laboratory measurements of SOA photolysis rates are necessary to better inform models. This study reports a series of environmental chamber experiments quantifying mass loss rates and chemical composition changes due to UV photolysis of SOA formed from representative precursors, including isoprene, alpha-pinene, and caryophyllene. We find the highest photolysis mass loss rates for isoprene SOA, with alpha-pinene SOA displaying reduced, but still measurable photolysis mass loss rates. Significant photolytic loss of caryophyllene SOA mass was not observed. Changes in aerosol chemical composition were tracked in real-time using an AMS and a FIGAERO-CIMS. In general, multifunctional peroxide species were lost from the SOA condensed phase upon photolysis, which resulted in a complex evolution of AMS markers such as m/z 43, m/z 44, and the mean carbon oxidation state. Finally, species dependent photolysis rates derived from the experiments were incorporated into the WRF-Chem model and simulations run over the Amazon region show significant reductions in the predicted SOA loadings.