Characterizing the Amount and Chemistry of Biogenic SOA Formation from Pine Forest Air Using a Flow Reactor

The amount and chemistry of biogenic secondary organic aerosol (SOA) formation was characterized as a function of oxidant exposure using a Potential Aerosol Mass (PAM) oxidative flow reactor, sampling air in a terpene- and MBO-dominated pine forest during the 2011 BEACHON-RoMBAS field campaign at the U.S. Forest Service Manitou Forest Experimental Observatory in the Colorado Rocky Mountains. In the reactor, a chosen oxidant (OH, O3, or NO3) was generated and stepped over a range of values up to 10,000 times ambient levels, accelerating the gas-phase and heterogeneous oxidative aging of volatile organic compounds (VOCs), inorganic gases, and preexisting aerosol. The resulting SOA formation was measured using an Aerodyne HR-ToF-AMS, a TSI SMPS and a PTR-TOF-MS. Oxidative processing in the flow reactor was equivalent to a few hours up to ~20 days of atmospheric aging during the ~4-min reactor residence time. During BEACHON-RoMBAS, OH oxidation led to a net production of up to several μg/m3 of SOA at intermediate exposures (1-10 equivalent days) but resulted in net loss of OA mass (up to ~30%) at higher OH exposures (10-20 equivalent days), demonstrating the competing effects of functionalization/condensation vs. fragmentation/evaporation reactions as OH exposure increased. O3 and NO3 oxidation led to smaller (up to 0.5 μg/m3) SOA production, and loss of SOA mass due to fragmentation reactions was not observed. OH oxidation resulted in f44 vs. f43 and Van Krevelen diagram (H:C vs. O:C) slopes similar to ambient oxidation, suggesting the flow reactor oxidation pathways are similar to those in ambient air. Organic nitrate SOA production was observed from NO3 radical oxidation only. New particle formation was observed from OH oxidation, but not O3 or NO3 oxidation under our experimental conditions. An enhancement of SOA production under the influence of anthropogenic pollution (Denver) was also observed. High-resolution AMS measurements showed that the O:C and H:C elemental ratios of the SOA mass added from 1-5 days exposure to each oxidant were approximately 0.54 and 1.56, respectively, consistent with known monoterpene-derived SOA constituents (e.g., pinic acid = C9H14O4). At higher OH exposures, the O:C of new SOA mass increased while the H:C decreased, and evidence of heterogeneous oxidation was observed. PTR-TOF-MS measurements of oxidized air showed that some compounds were depleted (e.g., monoterpenes) while some compounds were produced (e.g., acetaldehyde) due to oxidation. When applying laboratory yields measured with this reactor, assuming complete consumption of the most abundant VOCs measured at the forest site (monoterpenes, sesquiterpenes, and toluene), a simple model underpredicted the amount of SOA formed in the reactor in the field observations by a factor of ~6. This suggests one or more issues, including a large SOA source from oxygenated VOCs (e.g., monoterpene oxidation products) that were not included in our simple model, or from other VOCs not considered in the model, or limitations of the extrapolation of single precursor lab yields to multiple-precursor ambient air.