To address this, large parts of the work carried out within MICROSCOPE have been related to investigating the phase behaviour of organic aerosol particles containing SOA and how it changes as a function of relative humidity (RH). In one activity, the phase behaviour of parti-cles denoting internal mixtures of primary organic aerosol (POA) and secondary organic aero-sol (SOA) was investigated. Single-component, commercially available organic com-pounds were used as proxies of atmospheric POA and SOA components, covering relevant properties of ambient organic aerosol such as functional groups, carbon numbers or elemental oxygen-to-carbon (O/C) ratios. The phase behaviour of individual POA+SOA particles was investigated using optical and fluorescence microscopy tools. Specifically, by embedding trace amounts of a solvatochromic dye within the POA+SOA particles, the number of phases was directly visualized and detectable from microscopy images of the particles, as the solva-tochromic nature causes the dye molecule to fluoresce at different wavelength (colours) de-pending on the polarity of the environment (aerosol phase). By measuring the phase behaviour of 273 unique POA+SOA mixtures, for humidities between 90% and 0% RH, the results of this activity considerably expanded the organic aerosol phase behaviour data available in the literature. A key result of these experiments was that POA+SOA mixtures can either form single-phase or phase separated particles, and that the number of phases is largely dependent on the polarity of the POA and SOA component being internally mixed within a given particle, as approximated by their elemental oxygen-to-carbon (O/C) ratios. In a related activity, and in an attempt to assess to what extent the previous findings transfer to atmospheric organic aerosols, similar phase behaviour experiments were carried out for POA+SOA particles, but replacing the single-component, commercial SOA proxies with complex SOA material that was generated in the laboratory from oxidation of precursor gases. The results of this activity largely confirmed the previous results that the phase behaviour is governed by the polarity (O/C ratio) of the components being internally mixed. In addition, these experiments with more realistic, complex SOA demonstrated that the difference in the average O/C ratio of the POA and SOA material, i.e. the ΔO/C value, is a powerful predictor of the number of phases in internally mixed POA+SOA particles. Specifically, a single ΔO/C threshold of 0.265 was found to successfully predict the phase behaviour of 92% of the 77 POA+SOA mixtures studied. These results have implications for policy strategies to reduce SOA mass concentrations in urban environments, where POA concentrations are often high, and internally mixed POA+SOA particles can form through a range of processes. In another activity, the number of phases in particles containing different types of SOA was investigated. This was done in order to verify if the ΔO/C framework can be unified and used as a general tool to describe the phase behaviour of internally mixed organic aerosols. An assumption frequently made in chemical transport models is that different SOA types form a single condensed-phase when internally mixed in individual aerosol particles. Predictions of total ambient or-ganic mass concentrations for air quality would be substantially altered, if this assumption were incorrect. To address this, six different SOA types were generated in the laboratory using environmental chambers. The SOA types had O/C ratios from ~0.34 to ~1.05 thus covering a broad range of O/C ratios observed in the atmosphere. Mixing each SOA type with each other and performing phase behaviour experiments on the SOA+SOA particles showed that not all SOA types mix. In fact, if the ΔO/C of a SOA+SOA mixture was low, the polarities of the SOA components were found to be similar enough to mix completely, while phase separated SOA+SOA particles were observed if the ΔO/C of a SOA+SOA mixture was high. The results have implications for interpreting other (measured) SOA properties and thus for predicting SOA properties and processes in the atmosphere. For instance, knowledge under what conditions a single condensed phase exists in internally mixed organic aerosol particles will help to constrain the validity of other aerosol properties, such as diffusivities estimated from the average chemical particle properties. Thus, the improved understanding of organic aerosol phase behaviour established within the MICROSCOPE project could greatly benefit the representation of aerosol processes in models and improve predictions of air quality and climate.
