Distribution of gaseous and particulate organic composition during dark α-pinene ozonolysis
1School of Geography, Earth & Environmental Science, University of Birmingham, Birmingham, B15 2TT, UK
2Department of Chemistry, University of York, York, YO10 5DD, UK
3Department of Chemistry, University of Leicester, Leicester, LE1 7RH, UK
4National Centre for Atmospheric Science, University of Leeds, Leeds, LS2 9JT, UK
Abstract. Secondary organic aerosol (SOA) affects atmospheric composition, air quality and radiative transfer. However major difficulties are encountered in the development of reliable models for SOA formation. Constraints on processes involved in SOA formation can be obtained by interpreting the speciation and evolution of organics in the gaseous and condensed phase simultaneously. In this study we investigate SOA formation from dark α-pinene ozonolysis with particular emphasis upon the mass distribution of gaseous and particulate organic species. A detailed model for SOA formation is compared with the results from experiments performed in the EUropean PHOtoREactor (EUPHORE) simulation chamber, including on-line gas-phase composition obtained from Chemical-Ionization-Reaction Time-Of-Flight Mass-Spectrometry measurements, and off-line analysis of SOA samples performed by Electrospray Ionisation Ion Trap Mass Spectrometry. The temporal profile of SOA mass concentration is relatively well reproduced by the model. Sensitivity analysis highlights the importance of the choice of vapour pressure estimation method. Comparisons of the simulated gaseous- and condensed-phase mass distributions with those observed show a generally good agreement. The simulated speciation has been used to (i) propose a chemical structure for the principal gaseous semi-volatile organic compounds and condensed monomer organic species and (ii) explore the possible contribution of a range of accretion reactions occurring in the condensed phase. We find that oligomer formation through esterification reactions gives the best agreement between the observed and simulated mass spectra.