Abstract View
Elemental Analysis of Oxygenated Organic Coatings on Black Carbon Particles using a Soot-Particle Aerosol Mass Spectrometer
MUTIAN MA, Laura-Helena Rivellini, Yuxi Cui, Megan Willis, Rio Wilkie, Jonathan Abbatt, Manjula Canagaratna, Junfeng Wang, Xinlei Ge, Alex Lee, National University of Singapore
Abstract Number: 278
Working Group: Instrumentation and Methods
Abstract
Chemical characterization of organic coatings is important to advance our understanding of the physio-chemical properties and environmental fate of black carbon (BC). Soot-particle aerosol mass spectrometer (SP-AMS) has been utilized for this purpose in field studies. The laser vaporization (LV) scheme of SP-AMS can heat up BC cores gradually until they are completely vaporized, during which organic coatings can be vaporized at temperatures lower than that of thermal vaporizer (TV) used in a standard high-resolution aerosol mass spectrometer (HR-AMS).
This work investigates the effects of vaporization schemes on fragmentation and elemental analysis of oxygenated organic species (30 in total) using three SP-AMS. We show that LV can reduce the fragmentation of organic molecules. Substantial enhancement of C2H3O+/CO2+ and C2H4O2+ signals were observed for most species vaporized by the LV scheme, suggesting that the observational frameworks developed based on HR-AMS field data may not be directly applicable for evaluating the chemical evolution of oxygenated organic aerosol (OOA) coated on ambient BC. The uncertainties of H:C and O:C determined by the improved-ambient (I-A) method for both LV and TV approaches were similar, and scaling factors of 1.10 for H:C and 0.89 for O:C were determined to facilitate more direct comparisons between observations from the two vaporization schemes. Applying the scaling factors to ambient data, we found that even the time series of OOA components determined by LV and TV scheme are strongly correlated at the same location, OOA coatings were likely less oxygenated compared to those externally mixed with BC. Lastly, the I-A method was updated based on the multilinear regression model for the LV scheme measurements, which can reduce the relative errors of O:C from -26% to 6%, and the relative errors of H:C remain similar.