Modelling the Gas-Particle Partitioning and Water Uptake of Isoprene-derived Secondary Organic Aerosol at High and Low Relative Humidity

We present a unique model-measurement comparison study aiming for a better characterization of the hygroscopic growth behaviour and effects of different inorganic seed particles on the formation of secondary organic aerosols (SOA) from the dark ozone-initiated oxidation of isoprene at low NO_x conditions. We performed simulations using a gas-phase chemical reaction mechanism based on the Master Chemical Mechanism (MCM) in combination with an equilibrium gas-particle partitioning model. The equilibrium model accounts for non-ideal mixing in liquid phases, including liquid-liquid phase separation (LLPS), and is based on the AIOMFAC model for mixture non-ideality and the EVAPORATION model for pure compound vapour pressures. Measurements from the Cosmics leaving outdoor droplets (CLOUD) chamber experiments conducted at the European Organization for Nuclear Research (CERN) for isoprene ozonolysis cases, were used to aid in parameterizing the SOA yields at different atmospherically-relevant temperatures and reacted isoprene levels. To represent the isoprene ozonolysis-derived SOA, we introduce a selection of organic surrogate species in the coupled modelling system in the form of higher-generation oxidation products (compounds not directly predicted by MCM). The model predicts a single, mixed particle phase at all relative humidity (RH) levels for the seed-free case. In the presence of sulfuric acid or ammonium bisulfate seed particles (liquid solution at high RH), the model predicts LLPS to occur below 82 % RH, where the particles consist of an inorganic-rich liquid phase and an organic-rich liquid phase; however, with significant amounts of bisulfate and water remaining in the organic-rich phase. The measurements and model calculations show an enhancement in the SOA amount at 85 % compared to 35 % RH. We will discuss the model-measurement comparison and how inorganic water uptake and SOA partitioning affect SOA mass enhancement aside from other potential factors.