Abstract View
Quantitative Source Apportionment of Organic Aerosol by Combined Factor Analysis: Extractive Electrospray Ionization and Aerosol Time-of-Flight Mass Spectrometry (EESI-TOF/AMS)
Yandong Tong, Lu Qi, Giulia Stefenelli, Dongyu S. Wang, Francesco Canonaco, Urs Baltensperger, André S. H. Prévôt, JAY G. SLOWIK, Paul Scherrer Institute
Abstract Number: 376
Working Group: Source Apportionment
Abstract
Primary and secondary organic aerosol (POA and SOA) are significant components of aerosol pollution and are linked to adverse health effects. Efficient mitigation strategies require quantitative characterization of POA and SOA sources. Although the aerosol mass spectrometer (AMS) has proven to be a reliable method for quantitative source apportionment of POA, it usually lacks the chemical resolution to distinguish different SOA sources due to its extensive thermal decomposition and ionization-induced fragmentation. In contrast, softer ionization techniques such as extractive electrospray ionization time-of-flight mass spectrometry (EESI-TOF) or chemical ionization mass spectrometry (e.g., FIGAERO-CIMS) have the chemical resolution to resolve SOA sources, but are unable to provide quantitative results due to uncertainties in molecule-dependent sensitivities.
Here we introduce a method for quantitative source apportionment of the total organic aerosol, including both POA and SOA sources, by conducting positive matrix factorization (PMF) on a single dataset including both AMS and EESI-TOF mass spectra. Because each factor profile includes both AMS and EESI-TOF components, quantitative mass concentrations and EESI-TOF factor sensitivities are direct outputs of the mode. In this way, the model optimally combines the strengths of the AMS (quantification) and EESI-TOF (chemical resolution). We demonstrate the model’s ability to incorporate factors detectable by only a single instrument vs. both instruments, incorporate profile constraints for selected factors, and provide uncertainty analysis.
The combined PMF method is applied to summer and winter datasets from Zurich, Switzerland. Resolved sources include traffic, cooking, biomass burning, and cigarette smoke POA, as well as SOA from daytime and nighttime oxidation of biogenic emissions and biomass burning. The retrieved EESI-TOF sensitivities agree well with measurements of chemical standards, laboratory-generated SOA from known precursors, and parameterizations based on the chemical formulae of molecular ions. Comparison of these results to standalone EESI-TOF PMF highlights the importance of accounting for factor-dependent sensitivities.