Secondary Organic Aerosol Mass Yields from NO 3 Oxidation of α-Pinene and Δ-Carene: Effect of RO 2 Radical Fate.

Dark chamber experiments were conducted to study the SOA formed from the oxidation of α-pinene and Δ-carene under different peroxy radical (RO 2 ) fate regimes: RO 2 + NO 3 , RO 2 + RO 2 , and RO 2 + HO 2 . SOA mass yields from α-pinene oxidation were <1 to ∼25% and strongly dependent on available OA mass up to ∼100 μg m -3 . The strong yield dependence of α-pinene oxidation is driven by absorptive partitioning to OA and not by available surface area for condensation. Yields from Δ-carene + NO 3 were consistently higher, ranging from ∼10-50% with some dependence on OA for <25 μg m -3 . Explicit kinetic modeling including vapor wall losses was conducted to enable comparisons across VOC precursors and RO 2 fate regimes and to determine atmospherically relevant yields. Furthermore, SOA yields were similar for each monoterpene across the nominal RO 2 + NO 3 , RO 2 + RO 2 , or RO 2 + HO 2 regimes; thus, the volatility basis sets (VBS) constructed were independent of the chemical regime. Elemental O/C ratios of ∼0.4-0.6 and nitrate/organic mass ratios of ∼0.15 were observed in the particle phase for both monoterpenes in all regimes, using aerosol mass spectrometer (AMS) measurements. An empirical relationship for estimating particle density using AMS-derived elemental ratios, previously reported in the literature for non-nitrate containing OA, was successfully adapted to organic nitrate-rich SOA. Observations from an NO 3 - chemical ionization mass spectrometer (NO 3 -CIMS) suggest that Δ-carene more readily forms low-volatility gas-phase highly oxygenated molecules (HOMs) than α-pinene, which primarily forms volatile and semivolatile species, when reacted with NO 3 , regardless of RO 2 regime. The similar Δ-carene SOA yields across regimes, high O/C ratios, and presence of HOMs, suggest that unimolecular and multistep processes such as alkoxy radical isomerization and decomposition may play a role in the formation of SOA from Δ-carene + NO 3 . The scarcity of peroxide functional groups (on average, 14% of C 10 groups carried a peroxide functional group in one test experiment in the RO 2 + RO 2 regime) appears to rule out a major role for autoxidation and organic peroxide (ROOH, ROOR) formation. The consistently substantially lower SOA yields observed for α-pinene + NO 3 suggest such pathways are less available for this precursor. The marked and robust regime-independent difference in SOA yield from two different precursor monoterpenes suggests that in order to accurately model SOA production in forested regions the chemical mechanism must feature some distinction among different monoterpenes.