Understanding the Applicability of Gas/Particle Partitioning Theory to SOA using Novel Chemically-Speciated Partitioning Measurements

Principal Investigator(s):
Jose-Luis Jimenez, University of Colorado, Boulder

PJ Ziemann, University of Colorado
DA Day, University of Colorado

Aerosol particles have important but very uncertain climate forcing. Secondary organic aerosols (SOA) are formed from gas-to-particle conversion of volatile organic compounds. SOA has been recently shown to be a major component of the tropospheric submicron aerosol, but its sources, properties, aging, and removal processes remain poorly characterized. Almost all computer models that predict  SOA concentrations employ gas/particle (G/P) absorptive partitioning theory to predict the distribution of chemical components between the gas and particle phases. However, the applicability of this theory has been mostly studied for very reduced species such as alkanes and polycyclic aromatic hydrocarbons (PAHs), rather than with the highly oxidized species that comprise most SOA. Whether and when G/P partitioning equilibrium is a good approximation for SOA has recently become very controversial, with several groups reporting that laboratory and ambient SOA are kinetically hindered in reaching equilibrium on atmospheric timescales due to particle-phase effects, while other groups report the opposite, and still others report that timescales depend strongly on conditions (e.g., higher vs lower RH). However, all recent studies questioning the validity of partitioning theory have relied on indirect measurements, such as particle evaporation rates or viscosity, with none directly quantifying the G/P partitioning of representative chemical species through measurements of their gas and particle phase concentrations.

Recently our group has collaborated in demonstrating several techniques for simultaneous chemically-resolved gas and particle measurements that can be used to directly quantify G/P partitioning of SOA species. The technique uses a collector-thermal desorption High-Resolution Time-of-Flight Chemical-Ionization Mass Spectrometer (CIMS), and has been demonstrated with field data. Here we propose to apply this technique in the laboratory to investigate the applicability, timescales, and controlling processes of G/P partitioning to SOA. The focus of this proposal will be a series of chamber studies with systems of increasing complexity to probe the dynamics and equilibria of chemically-speciated G/P partitioning of SOA. Oxidized semi/low-volatility probe gases will be generated and monitored in situ, allowing for investigation of partitioning dynamics occurring at fast timescales.

The following tasks will be carried out: (1) Partitioning with probe particles composed of single compounds and simple SOA systems where well-known products are produced (without and with oligomer formation), will be measured and compared with model results. Dependence on particle viscosity/phase state, composition, and OA concentrations will be investigated. Perturbations (T, OA, dilution, RH) of systems in equilibrium will be used to probe re-equilibration time scales and for evidence of irreversible effects. The limitations of the experimental techniques (e.g., partitioning to/from chamber walls, limitations of CIMS sampling inlets) and their influence on the results will be characterized and modeled. The consistency of results from three thermal desorption approaches (FIGAERO-CIMS, thermal desorption particle beam mass spectrometer (TDPBMS), and thermal denuder-aerosol mass spectrometer (TD-AMS)) will be evaluated. (2) Partitioning using SOA systems of atmospheric importance (terpenes, isoprene, and aromatics, with OH, O3), highly aged SOA (produced by Cl  oxidation in a chamber or in an OH oxidation flow reactor) and ambient aerosol as probe particles will be investigated with the same methods as Task 1 and compared to field observations from GoAmazon2014/15. (3) Particle size distribution growth dynamics of selected systems will be modeled with MOSAIC to investigate the properties of the condensing gases and their diffusivity in the particles. (4) The observed behavior across the systems will be compared to equilibrium, kinetically-limited, and other recently-proposed models, and the best combination for regional and global models will be identified.

These combined laboratory and modeling investigations will provide unique data on chemically-resolved partitioning for a wide range of systems and conditions. The combination of mechanistic-level understanding and parameterizations that can be applied to regional and global modeling will help reduce the uncertainties in climate forcing by aerosols and on air quality prediction.