Volatility of Primary and Secondary Organic Aerosols in the Field Contradicts Current Model Representations

A novel combination of a temperature-stepping thermodenuder (TD) in front of an Aerosol Mass Spectrometer (AMS) has allowed rapid quantification of chemically-resolved volatility in the field for the first time. The TD consists of a tube heated to a uniform temperature, followed by a carbon charcoal denuder to adsorb volatilized gases, leaving a core of less volatile material for detection by the AMS. The TD temperature can be changed quickly (1 to 10 min) to allow rapid measurements of volatility spectra. The volatility of organic aerosols (OA) influences the concentration and lifetime of OA, secondary OA (SOA) formation, and OA measurements. The TD- AMS system was used to characterize urban OA, primary OA (POA) from biomass burning, and chamber- generated SOA. Almost all atmospheric models represent POA as non-volatile and SOA as semi-volatile. Our results indicate nearly the opposite: urban POA is semi-volatile, while urban SOA is much less volatile. Biomass burning POA exhibited a wide range of volatilities, but was often similar in volatility to urban OA. Chamber- generated SOA was significantly more volatile than urban OOA, challenging the extrapolation of laboratory volatility measurements to the atmosphere. Our results indicate that the representation of OA volatility in most current models is highly inaccurate and should be revised.