Size dependent hygroscopicity of levoglucosan and D-glucose aerosol nanoparticles

Abstract. The interaction between water vapor and aerosol nanoparticles is of great significance in atmospheric processes. However, current knowledge of hygroscopicity of sub-10 nm organic nanoparticles and their concentration-dependent thermodynamic properties (e.g., water activity) in the highly supersaturated concentration range is scarcely available. In this study, we investigate the size dependence of hygroscopicity of organics (i.e., levoglucosan, D-glucose) in size down to 6 nm using a nano-hygroscopicity tandem differential mobility analyzer (nano-HTDMA). There is a weak size dependence of the hygroscopic growth factor observed for levoglucosan and D-glucose nanoparticles with diameters down to 20 nm. However, a clear size-dependent hygroscopic growth factor is observed for D-glucose nanoparticles down to 6 nm in size. A reduction in diameters of sub-20 nm levoglucosan is observed at the dry RHs, which is explained by partial levoglucosan evaporation into gas phase, indicting high impact of volatility of sub-20 nm levoglucosan aerosol nanoparticles. However, this also means that the hygroscopic growth factors of levoglucosan nanoparticles with diameters below 20 nm are not possible to be determined. The use of water activity parameterization models proposed by Kreidenweis et al. (2005) (KD, Köhler), the Extend-Aerosol Inorganic Model (E-AIM (standard UNIFAC), and Differential Köhler Analysis (DKA) method is to determine thermodynamic properties (e.g., water activity) of levoglucosan and D-glucose nanodroplets as a function of solute concentration, respectively. Predicated water activity for these aqueous organic solutions (i.e., levoglucosan, D-glucose) from the different methods are similar to observations from references in the low solute concentration (< 20 mol kg^-1), while a quite difference is found in the high solute concentration (> 20 mol kg^-1). In addition, we compare hygroscopicity measurements for levoglucosan and D-glucose nanoparticles with the E-AIM (standard UNIFAC), the ideal solution theory, and DKA predictions, respectively. The ideal solution theory describes well the measured hygroscopic growth factors of levoglucosan with diameters down to 20 nm and D-glucose nanoparticles with diameters higher than 60 nm, respectively, while the E-AIM (standard UNIFAC) model can successfully predict the growth factors of levoglucosan with diameters from 100 down to 6 nm at RH above 88–40 % (e.g., at RH above 88 % for 100 nm D-glucose, at RH above 40 % for 6 nm D-glucose). The use of the DKA method leads to a good agreement with measured hygroscopic growth factors of D-glucose aerosol nanoparticles with diameters from 100 down to 6 nm.