|
Abstract. Aromatic hydrocarbons represent a large fraction of anthropogenic volatile organic compounds and significantly contribute to tropospheric ozone and secondary organic aerosol (SOA) formation. Toluene photooxidation experiments were carried out in an oxidation flow reactor (OFR). We identified and quantified the gaseous and particulate reaction products at 280, 285 and 295 K using a proton-transfer reaction time-of-flight mass spectrometer (PTR-ToF-MS) coupled to a CHemical Analysis of aeRosol ONline (CHARON) inlet. The reaction products accounted for both ring-retaining compounds such as cresols, benzaldehyde, nitrophenols, nitrotoluene, bicyclic intermediate compounds, as well as ring-scission products such as dicarbonyls, cyclic anhydrides, small aldehydes and acids. The chemical system exhibited a volatility distribution mostly in the semi-volatile (SVOCs – semi-volatile organic compounds) regime. The saturation concentration (Ci*) values of the identified compounds were mapped onto the two-dimensional volatility basis set (2D-VBS). Temperature decrease caused a shift of Ci* towards lower values while there was no clear relationship between Ci* and oxidation state. The CHARON PTR-ToF-MS instrument identified and quantified approximately 70–80 % of the total organic mass measured by an aerosol mass spectrometer (AMS). The experiments were reproduced by simulating SOA formation with the SSH-aerosol box model. A semi-detailed mechanism for toluene gaseous oxidation was developed. It is based on the MCM and GECKO-A deterministic mechanisms modified following the literature in particular to update cresols and ring-scission chemistry. The new mechanism improved secondary species representation with an increment of the major identified species (+35 % in number). Light compounds formation (i.e. m/z < 100) is enhanced and accumulation of heavy compounds (i.e. m/z ≥ 100) is reduced, especially in the gas phase. Additional tests on (i) partitioning processes such as condensation into aqueous phase, (ii) interactions of organic compounds between themselves and with inorganics and (iii) wall losses were also performed. When all these processes were taken into account the simulated SOA mass concentration showed a much better agreement with the experimental results. Finally, an irreversible partitioning pathway for methylglyoxal was introduced and considerably improved the model results, opening a way to further developments of partitioning in models. Our results underline that the volatility itself is not sufficient to explain the partitioning between gas and particle phase: the organic and the aqueous phases need to be taken into account as well as interactions between compounds in the particle phase.
|