Atmospheric aerosol particles affect air pollution, public health, and uncertainty about radiative forcing. Their phase state and morphology have far ranging effects on gas uptake, particle phase rate of reaction and aging. Recent reports have demonstrated that atmospheric particles may exist in semi–solid or glassy solid states. The phase transition between these two states occurs at the glass transition temperature (Tg). The kinetic multi–layer model of gas particle interactions (KM-GAP) explicitly treats the temporal evolution of all species from the gas phase to the particle bulk. Investigations into the role of amines in new particle formation and nano–particle growth are essential to our understanding of cloud condensation nuclei. The goal of this work was to predict the phase state and viscosity of secondary organic aerosol (SOA) mixtures and apply kinetic modeling to advance our understanding of amine uptake by atmospheric particles.

First, a parameterization for estimating Tg based on molecular composition and a modeling approach to predict viscosity as a function of ambient temperature and relative humidity (RH) is presented and applied to a–pinene and isoprene SOA using marker compounds. Predictions agree well with measured viscosity. Additionally, we apply our method to high–resolution mass spectrometry data for toluene and biomass burning SOA and discuss the effect of different ionization techniques. Next, measurements of the reactive uptake of dimethylamine (DMA) by ammonium sulfate (AS) and mixed AS–sucrose particles at different RH are simulated with KM-GAP.The temporal evolution of particle mass increases and their humidity dependence are successfully reproduced, revealing that uptake is limited by diffusion through a viscous sucrose–rich shell at lower RH. Uptake coefficients increase when RH increases, but decrease when the molar fraction of sucrose increases at fixed RH. The model is extrapolated to emulate atmospheric conditions. At RH > 70% for liquid particles, amine uptake can lead to a mass increase of  ~20 to 60%. Finally, our viscosity prediction method is applied to SOA particles generated by photooxidation of diesel fuel vapors. Predicted and measured viscosity compare well. Additional experimental datasets would increase viscosity prediction accuracy.