Prediction of the pure component and mixture properties of organic compounds
The potential presence of thousands of organic compounds in the atmospheric aerosol presents difficulties in both the prediction of those components present and of the properties of the resulting mixtures. Air quality and global climate models need to realistically capture the formation and behaviour of multicomponent aerosol if they are to realistically represent the atmospheric particulate burden and its effects. Particles may be classified as “primary”; those directly emitted as solid or liquid particles, or “secondary”; those formed by condensation from gaseous components. Conventional treatments of secondary organic aerosol (“SOA”) have largely relied upon yields fitted to smog chamber experiments. This approach is difficult to validate in the real atmosphere and carries limited mechanistic information. We are currently developing an alternative approach based on the physico-chemical properties of potential organic aerosol (“OA”) components.Using the Master Chemical Mechanism (MCM), an explicit model of the oxidation of emitted pollutant Volatile Organic Compounds (“VOC”s) or a similar realistic model traceable to laboratory-measured degradation rates, it is possible to predict the distribution of organic compounds in the atmosphere. Such models predict that there may be several thousand compounds which could potentially condense into particulate material. A previous approach using the MCM coupled to absorptive partitioning theory to simulate the abundance and dominant contributors to the OA mass, whilst substantially underpredicting the loading, identified several tens of compounds which contributed the majority of the OA (Johnson et al., 2006).
We are using a number of approaches to refine the previous representation. We are starting from the MCM predicted molecular abundances of some 3000+ organic compounds formed by oxidation of primary emitted VOCs to identify which are the most likely contributors to the organic fraction of atmospheric aerosol. The aim is to provide a set of surrogate compounds that can be used to represent the collective properties of the condensed species in reduced model schemes suitable for large-scale simulations.
A key driver of the partitioning of a compound into a condensed phase is its vapour pressure. Our current effort is to focus on identification of the best estimation methods for predicting vapour pressures for atmospherically relevant compounds to atmospherically-realistic aerosol which will include both hydrophobic and hydrophilic particles. We are further investigating appropriate techniques for weighting by predicted abundance to select of the most appropriate surrogate compounds for use in parameterised schemes for large-scale models. These schemes will be developed by incorporating the compounds into ADDEM to predict properties of, and vapour pressure of all components over, multicomponent aerosol particles. This will be used to develop reduced partitioning schemes for large-scale models within the NERC-funded QUAAC and EU FP6-funded EUCAARI projects and is being carried out by Mark Barley and Dave Topping.