Comparison of Carbon Analysis Protocols in PM2.5 Source Apportionment

SUNGHWAN SHIM, Seung-Muk Yi, Kwon Ho Jeon, Seoul National University, Seoul, Korea

     Abstract Number: 329
     Working Group: Source Apportionment

Abstract
Thermal-optical methods are widely used to quantify carbonaceous species in PM2.5, which serve as marker species for source apportionment, particularly for mobile sources such as gasoline and diesel vehicles. Two commonly applied protocols of carbon analysis are Thermal-Optical Transmittance (TOT) and Thermal-Optical Reflectance (TOR) methods, both of which provide quantitative information on organic carbon (OC) and elemental carbon (EC) fractions. Both TOT and TOR methods analyze carbon fractions by gradually heating filter sample, first in an inert atmosphere (100% He) to volatilize organic carbon, then in an oxidizing atmosphere (2% O2 + 98% He) to combust elemental carbon. The key difference is that TOT monitors the decrease in laser light transmitted through the filter, whereas TOR tracks changes in light reflected from its surface. These optical differences result in discrepancies in how carbon fractions are classified: TOT (OC1-OC5, POC, EC1-EC6), and TOR (OC1-OC4, EC1-EC3). Such differences can affect how carbon marker species are interpreted in receptor models like Dispersion-Normalized Positive Matrix Factorization (DN-PMF). In this study, we conducted a direct comparison of DN-PMF results using input data from both TOT and TOR to assess how analytical protocols influence PM2.5 source apportionment. A total of 450 PM2.5 samples were collected on the rooftop of Seoul National University’s building over 23-hour periods from September 2020 to August 2025. Mass concentrations were measured by microbalance. Trace elements (20 species) were analyzed using Energy-Dispersive X-ray Fluorescence (ED-XRF), and ionic species (6 species) were quantified through ion chromatography. The comparison highlights how protocol-specific differences in carbon fractionation affect the estimation of mobile source contributions. These findings may help guide analytical method selection in future source apportionment studies.

Acknowledgement

This study was supported by the National Institute of Environment Research, funded by the Ministry of Environment (NIER).

This research was supported by Particulate Matter Management Specialized Graduate Program through the Korea Environmental Industry & Technology Institute (KEITI) funded by the Ministry of Environment (MOE).