Reformulating the atmospheric lifecycle of SOA based on new field and laboratory data

Abstract. Atmospheric loadings of secondary organic aerosols (SOA) are significantly under-predicted by climate models. In these models, SOA particles are assumed liquid-like droplets at equilibrium with the gas-phase. In sharp contrast, our recent laboratory and field measurements show that SOA particles are non-rigid, highly viscous, spherical, quasi-solids, and do not behave like liquid droplets. They evaporate at rates much lower than predicted by models, and are consequently not at equilibrium with the gas phase. In addition, our data show that SOA particles trap hydrophobic organics, whose presence further reduces evaporation rates, and that aging these particles nearly stops evaporation. Measurements of the evaporation kinetics of ambient SOA particles under vapor-free conditions at room temperature showed that less than 20 % of particle mass evaporates in 4 h.

In this study, we examine, for the first time, these groundbreaking observations to present a new, experimentally based picture of the phase and evaporation behavior of SOA particles. We conclude that to first order SOA can be reasonably approximated to be non-evaporating. We use a simplified approach to investigate the implications of this near-irreversible gas-particle partitioning behavior in a box model and a 3-D chemical transport model, both of which, for the first time, include multi-generational gas-phase chemistry with functionalization and fragmentation reactions, and compare them to traditional reversible partitioning models. Results indicate that the revised irreversible partitioning approach yields slightly higher SOA loadings than traditional reversible partitioning approach when functionalization reactions, pushing SOA species to lower volatility bins are dominant. However, when fragmentation reactions play a major role, the revised irreversible partitioning approach predicts significantly higher SOA than the traditional approach. In addition to irreversibility, functionalization, and fragmentation, we explore the utility of lower activity coefficient to account for complex molecular interactions within particles and show that this approach predicts considerably higher SOA loadings. Using the 3-D Weather Research and Forecasting (WRF) model, coupled with Chemistry (WRF-Chem) modeling example of the Mexico City region, we demonstrate that when fragmentation is taken into account, irreversible partitioning increases predicted SOA loadings and lifetimes significantly compared to traditional models. When lower activity coefficient is also included, predicted SOA loadings in the Mexico City plateau increase by more than a factor of 3.