New Insights Into the Composition of Organics in the Atmosphere Enabled by Advanced Processing Techniques for Existing Chromatographic Datasets

SUNGWOO KIM, Lindsay Yee, Allen Goldstein, Nathan Kreisberg, Gabriel Isaacman-VanWertz, Virginia Tech

     Abstract Number: 566
     Working Group: Meet the Job Seekers

Abstract
Ambient aerosols may contain hundreds or thousands of unique organic compounds, and even minor components may provide valuable insight into particle sources and formation chemistry. However, due to the complexity of this data and the highly dynamic nature of aerosol composition, conventional analysis methods tend to be highly time consuming and also result in a substantial fraction of data never being analyzed or interpreted. We developed new methods to more comprehensively identify the full suite of analytes in a complex dataset by applying Positive Matrix Factorization (PMF) coupled with a peak-filtering decision tree. The analysis results find cataloged information of roughly 1000 potential analytes of interest. Single-ion chromatogram (SIC) based automatic peak fitting and integration method has been used to generate time series and complete the analysis cycle. This method requires a mass-to-charge ratio (m/z) that best represents the analyte peak signal, a quantifier ion, for integration process. An automated quantifier ion selection method based on peak resolution and signal intensity was developed for the completion of fully automated analysis process. In this presentation, we apply these methods to existing datasets collected by the Semi-Volatile Thermal Desorption Aerosol Gas chromatograph (SV-TAG) in Manacapuru, Brazil during GoAmazon 2014/15 campaign to identify new analytes and interpret their variability and transformations in the atmosphere. We apply two dimensionality reduction techniques, hierarchical cluster analysis (HCA) and spherical K-means, to the resulting analyte concentration time series to systematically investigate the underlying patterns of the observation and yield new insights into the sources of previously overlooked analytes. In particular, we aim to understand the characteristics of the categorized analytes in the context of atmospheric aerosol properties.