A curved multi-component aerosol hygroscopicity model framework: Part 2 – Including organic compounds

摘要

This paper describes the inclusion of organic particulate material withinthe Aerosol Diameter Dependent Equilibrium Model (ADDEM) framework describedin the companion paper applied to inorganic aerosol components. Theperformance of ADDEM is analysed in terms of its capability to reproduce thebehaviour of various organic and mixed inorganic/organic systems usingrecently published bulk data. Within the modelling architecture alreadydescribed two separate thermodynamic models are coupled in an additiveapproach and combined with a method for solving the Kohler equation in orderto develop a tool for predicting the water content associated with anaerosol of known inorganic/organic composition and dry size. For developmentof the organic module, the widely used group contribution method UNIFAC isemployed to explicitly deal with the non-ideality in solution. The UNIFACpredictions for components of atmospheric importance were improvedconsiderably by using revised interaction parameters derived fromelectro-dynamic balance studies. Using such parameters, the model was foundto adequately describe mixed systems including 5–6 dicarboxylic acids, downto low relative humidity conditions. By comparison with electrodynamicbalance data, it was also found that the model was capable of capturing thebehaviour of aqueous aerosols containing Suwannee River Fulvic acid, astructure previously used to represent the functionality of complex oxidisedmacromolecules often found in atmospheric aerosols. The additive approachfor modelling mixed inorganic/organic systems worked well for a variety ofmixtures. As expected, deviations between model predictions and measurementsincrease with increasing concentration. Available surface tension models,used in evaluating the Kelvin term, were found to reproduce measured datawith varying success. Deviations from experimental data increased withincreased organic compound complexity. For components only slightly solublein water, significant deviations from measured surface tension depressionbehaviour were predicted with both model formalisms tested. A Sensitivityanalysis showed that such variation is likely to lead to predicted growthfactors within the measurement uncertainty for growth factor taken in thesub-saturated regime. Greater sensitivity was found for the value of drydensity used in the assumed form of the dried out aerosol. Comparison with acoupled thermodynamic approach showed that assumed values for interactionsparameters may lead to erroneous results where a simple additive approachmay provide more accurate results. However, where available, the use ofcoupled thermodynamics can better reproduce measured behaviour. Further work(and laboratory data) is required to assess whether this difference lieswithin the experimental uncertainty of observed hygroscopic behaviour for avariety of systems.