资源环境科技发展态势分析平台

The volatility distribution of the organic compounds present in secondary organic aerosol (SOA) at different conditions is a key quantity that has to be captured in order to describe SOA dynamics accurately. The development of the Filter Inlet for Gases and AEROsols (FIGAERO) and its coupling to a chemical ionization mass spectrometer (CIMS; collectively FIGAERO–CIMS) has enabled near-simultaneous sampling of the gas and particle phases of SOA through thermal desorption of the particles. The thermal desorption data have been recently shown to be interpretable as a volatility distribution with the use of the positive matrix factorization (PMF) method. Similarly, volatility distributions can be inferred from isothermal particle evaporation experiments when the particle size change measurements are analyzed with process-modeling techniques. In this study, we compare the volatility distributions that are retrieved from FIGAERO–CIMS and particle size change measurements during isothermal particle evaporation with process-modeling techniques. We compare the volatility distributions at two different relative humidities (RHs) and two oxidation conditions. In high-RH conditions, where particles are in a liquid state, we show that the volatility distributions derived via the two ways are similar within a reasonable assumption of uncertainty in the effective saturation mass concentrations that are derived from FIGAERO–CIMS data. In dry conditions, we demonstrate that the volatility distributions are comparable in one oxidation condition, and in the other oxidation condition, the volatility distribution derived from the PMF analysis shows considerably more high-volatility matter than the volatility distribution inferred from particle size change measurements. We also show that the Vogel–Tammann–Fulcher equation together with a recent glass transition temperature parametrization for organic compounds and PMF-derived volatility distribution estimates are consistent with the observed isothermal evaporation under dry conditions within the reported uncertainties. We conclude that the FIGAERO–CIMS measurements analyzed with the PMF method are a promising method for inferring the volatility distribution of organic compounds, but care has to be taken when the PMF factors are analyzed. Future process-modeling studies about SOA dynamics and properties could benefit from simultaneous FIGAERO–CIMS measurements.