Login / Signup

Synthesis of Carboxylic Acid and Dimer Ester Surrogates to Constrain the Abundance and Distribution of Molecular Products in α-Pinene and β-Pinene Secondary Organic Aerosol.

Christopher M KensethNicholas J HafemanYuanlong HuangNathan F DalleskaBrian M StoltzJohn H Seinfeld
Published in: Environmental science & technology (2020)
Liquid chromatography/negative electrospray ionization mass spectrometry [LC/(-)ESI-MS] is routinely employed to characterize the identity and abundance of molecular products in secondary organic aerosol (SOA) derived from monoterpene oxidation. Due to a lack of authentic standards, however, commercial terpenoic acids (e.g., cis-pinonic acid) are typically used as surrogates to quantify both monomeric and dimeric SOA constituents. Here, we synthesize a series of enantiopure, pinene-derived carboxylic acid and dimer ester homologues. We find that the (-)ESI efficiencies of the dimer esters are 19-36 times higher than that of cis-pinonic acid, demonstrating that the mass contribution of dimers to monoterpene SOA has been significantly overestimated in past studies. Using the measured (-)ESI efficiencies of the carboxylic acids and dimer esters as more representative surrogates, we determine that molecular products measureable by LC/(-)ESI-MS account for only 21.8 ± 2.6% and 18.9 ± 3.2% of the mass of SOA formed from ozonolysis of α-pinene and β-pinene, respectively. The 28-36 identified monomers (C7-10H10-18O3-6) constitute 15.6-20.5% of total SOA mass, whereas only 1.3-3.3% of the SOA mass is attributable to the 46-62 identified dimers (C15-19H24-32O4-11). The distribution of identified α-pinene and β-pinene SOA molecular products is examined as a function of carbon number (nC), average carbon oxidation state (OS¯C), and volatility (C*). The observed order-of-magnitude difference in (-)ESI efficiency between monomers and dimers is expected to be broadly applicable to other biogenic and anthropogenic SOA systems analyzed via (-) or (+) LC/ESI-MS under various LC conditions, and demonstrates that the use of unrepresentative surrogates can lead to substantial systematic errors in quantitative LC/ESI-MS analyses of SOA.
Keyphrases