NGUYEN VAN KIEN (Mass Spectrometry Convergence Research Institute, Daegu, 41566, Republic of Korea)

Sunghwan Kim (Mass Spectrometry Convergence Research Institute, Daegu, 41566, Republic of Korea)

NGUYEN VAN KIEN (Mass Spectrometry Convergence Research Institute, Daegu, 41566, Republic of Korea)

Geondo Park (Department of Chemistry, Kyungpook National University, Daegu, 41566, Republic of Korea)

Seong Weon Lee (Department of Applied Chemistry, Kyungpook National University, Daegu, 41566, Republic of Korea)

Chan Sik Cho (Department of Applied Chemistry, Kyungpook National University, Daegu, 41566, Republic of Korea)

Ho-Jin Lim (Department of Environmental Engineering, Kyungpook National University, Daegu, 41566, Republic of Korea)

Sunghwan Kim (Mass Spectrometry Convergence Research Institute, Daegu, 41566, Republic of Korea)

Secondary organic aerosol (SOA) compounds derived from both anthropogenic and biogenic volatile organic compounds (VOCs) generated in smog chamber experiments have been characterized at the molecular level using high-resolution mass spectrometry (HRMS). However, elucidating SOA formation pathways based on HRMS data remains time-consuming and relies heavily on manual analysis. Here, we developed a cheminformatics-based framework to simulate reaction pathways leading to SOA formation from VOC precursors. The framework integrates predefined SMARTS reaction rules representing key atmospheric transformation mechanisms with complete reaction pathways extracted from the Master Chemical Mechanism (MCM) for the studied VOCs, generating an extensive chemical space of chemically plausible oxidation products. Validation of the simulated products against experimental HRMS data based on exact mass (within 5ppm) and molecular formula demonstrated that the framework successfully reproduced a large proportion of experimentally observed products, with a molecular formula considered a match when it was present in both the simulated and experimentally assigned datasets. Structural-level comparisons further confirmed the model’s ability not only to reproduce experimental observed products but also to reconstruct multistep pathways leading to these compounds, consistent with formation mechanisms proposed in the literature. Overall, this framework provides a complementary approach for elucidating plausible SOA formation pathways and supports mechanistic interpretation of atmospheric oxidation processes.​