American Association for Aerosol Research - Abstract Submission

AAAR 35th Annual Conference
October 17 - October 21, 2016
Oregon Convention Center
Portland, Oregon, USA

Abstract View


Quantitative Off-line Particulate Matter Composition Analysis with Thermal Desorption Mass Spectrometry

XIAOLIANG WANG, Gustavo Riggio, Xufei Yang, Laxmi Narasimha Yatavelli, L.W. Antony Chen, Judith Chow, John Watson, Desert Research Institute

     Abstract Number: 329
     Working Group: Instrumentation and Methods

Abstract
On-line aerosol mass spectrometry has provided insights into atmospheric aerosol formation, sources, and transformation processes. However, wide deployment of on-line mass spectrometers is impractical due to cost, maintenance, and expertise requirements. Integrated filter samples are being collected in aerosol speciation networks with wide spatial and temporal coverage and a decades-long sample archive. A thermal desorption-mass spectrometry (TD-MS) method has been developed to quantify major constituents of particulate matter on quartz-fiber filters from these networks. A particle-laden filter punch is heated at predefined temperature steps in a pure helium atmosphere. Species evolve at different temperatures depending on their volatility and thermal stability, and the products of this evolution are separated and quantified by an electron-impact ionization mass selective detector (MSD). Chemical standards are used to establish calibration relations that include fragmentation patterns along with transfer, ionization, and detection efficiencies. TD-MS also provides an estimate of the ratios of oxygen (O), hydrogen (H), and nitrogen (N) to carbon (C), i.e., O/C, H/C, and N/C, and therefore the ratio of organic matter to organic carbon. With reasonable assumptions about oxidation states, this TD-MS approach has demonstrated quantification of particulate sulfate, nitrate, ammonium, and organic carbon mass concentrations in a single analysis. Comparison with other methods for ambient samples shows that the differences are within +/-5%.