Kinetic Partitioning with Effective Mass Accommodation for Secondary Organic Aerosol Simulations in Chemical Transport Modeling

Abstract. Secondary organic aerosols (SOA) can adopt liquid, amorphous semi-solid, or glassy solid states in the atmosphere. Previous model simulations have revealed that SOA adopt semisolid or glassy solid phase states over dry regions and in the upper troposphere. A highly viscous phase state in SOA particles can cause kinetic limitations in SOA partitioning, challenging traditional SOA treatments with equilibrium partitioning in chemical transport modeling. In this study, we develop a method to simulate kinetic partitioning of SOA accounting for the impacts of SOA phase state by utilizing effective mass accommodation coefficient (α_eff). We initially test this method in box modeling to calculate equilibration timescale of SOA partitioning for a wide parameter space for bulk diffusivity, volatility, and total aerosol mass loading. Comparison with the chemical timestep of 20 min in GEOS-Chem reveals that equilibrium partitioning is still applicable for particles with bulk diffusivity > 10^-15 cm² s^-1, whereas kinetic partitioning is required for partitioning of semi-and low-volatile compounds into highly viscous particles. To reduce computational burden, we developed analytical equations with gas-particle mass transfer coefficient by effectively accounting for gas and bulk diffusion limitations. The application of the new α_eff kinetic partitioning scheme into GEOS-Chem predicts lowers SOA mass over drylands (e.g., Western US) in the boundary layer due to kinetically limited growth of viscous SOA particles, whereas SOA mass can be described well with equilibrium partitioning in most other areas at the surface and at 850 hPa. In the upper troposphere at 500 hPa, the α_eff -kinetic scheme predicts up to 50% higher SOA mass as evaporation is suppressed for glassy particles. The developed method is easily applicable, improving description and treatments of SOA in chemical transport models.