Abstract
Abstract. Hundreds of gas- and particle-phase organic acids were measured in a rural ponderosa pine forest in Colorado, USA, during BEACHON-RoMBAS (Bio-hydro-atmosphere interactions of Energy, Aerosols, Carbon, H2O, Organics &amp; Nitrogen – Rocky Mountain Biogenic Aerosol Study). A recently developed micro-orifice volatilization impactor high-resolution time-of-flight chemical ionization mass spectrometer (MOVI-HRToF-CIMS) using acetate (CH3C(O)O−) as the reagent ion was used to selectively ionize and detect acids semicontinuously from 20 to 30 August 2011, with a measurement time resolution of ~1.5 h. At this site 98% of the organic acid mass is estimated to be in the gas phase, with only ~2% in the particle phase. We investigated gas–particle partitioning, quantified as the fraction in the particle phase (Fp), of C1–C18 alkanoic acids, six known terpenoic acids, and bulk organic acids vs. carbon number. Data were compared to the absorptive partitioning model and suggest that bulk organic acids at this site follow absorptive partitioning to the organic aerosol mass. The rapid response (<1–2 h) of partitioning to temperature changes for bulk acids suggests that kinetic limitations to equilibrium are minor, which is in contrast to conclusions of some recent laboratory and field studies, possibly due to lack of very low ambient relative humidities at this site. Time trends for partitioning of individual and groups of acids were mostly captured by the model, with varying degrees of absolute agreement. Species with predicted substantial fractions in both the gas and particle phases show better absolute agreement, while species with very low predicted fractions in one phase often show poor agreement, potentially due to thermal decomposition, inlet adsorption, or other issues. Partitioning to the aqueous phase is predicted to be smaller than to the organic phase for alkanoic and bulk acids, and has different trends with time and carbon number than observed experimentally. This is due to the limited additional functionalization observed for the bulk acids. Partitioning to water appears to only play a role for the most oxidized acids during periods of high aerosol liquid water. Based on measurement–model comparison we conclude that species carbon number and oxygen content, together with ambient temperature, control the volatility of organic acids and are good predictors for partitioning at this site. Partitioning of bulk acids is more consistent with model predictions for hydroxy acids, hydroperoxyacids, or polyacids, and less so for keto acids.
Highlights
Organic aerosol (OA) is a substantial fraction of the submicron aerosol mass globally (Murphy et al, 2006; Zhang et al, 2007)
After multi-peak fitting, each ion peak signal is divided by two factors that account for the fact that aerosol was allowed to accumulate for 45 min and for the lower flow rate (2 L per minute (Lpm)) during the aerosol heating step compared to the sampling step (10 Lpm)
Hundreds of organic acids were measured in the gas and particle phase during this campaign
Summary
Organic aerosol (OA) is a substantial fraction of the submicron aerosol mass globally (Murphy et al, 2006; Zhang et al, 2007). It is not surprising that there are often major discrepancies between modeled and observed SOA concentrations in the atmosphere (de Gouw et al, 2005; Heald et al, 2005; Volkamer et al, 2006) Possible explanations for these large discrepancies include the lack of consideration by models of (1) semivolatile and intermediate volatility organic compounds (S/IVOCs), which can contribute substantially to SOA formation (Robinson et al, 2007); (2) secondary chemistry in the condensed phase (Kalberer et al, 2004); (3) aqueous-phase reactions in cloud or fog droplets and aerosol particle water (Ervens et al, 2011); (4) nonequilibrium partitioning due to changes in the phase state of SOA (Perraud et al, 2012; Vaden et al, 2011; Virtanen et al, 2010); and (5) enhancement of biogenic SOA by anthropogenic pollution (Spracklen et al, 2011). Errors reported here are variable due to changes in ambient temperature and OA, as well as estimated precision calculated using random errors from ion counting, random error due to variation of the single-ion signal intensities, and electronic noise
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
