Inorganic Salt Interference on CO 2 + in Aerodyne AMS and ACSM Organic Aerosol Composition Studies

• Published in: Environmental science & technology 2016-10, Vol.50 (19), p.10494-10503
• Main Authors: Pieber, Simone M, El Haddad, Imad, Slowik, Jay G, Canagaratna, Manjula R, Jayne, John T, Platt, Stephen M, Bozzetti, Carlo, Daellenbach, Kaspar R, Fröhlich, Roman, Vlachou, Athanasia, Klein, Felix, Dommen, Josef, Miljevic, Branka, Jiménez, José L, Worsnop, Douglas R, Baltensperger, Urs, Prévôt, André S H
• Format: Article
• Language: English
• Subjects: Aerosols; Carbon; Engineering; Environmental Sciences & Ecology; Mass Spectrometry; Sodium Chloride; Sodium Chloride, Dietary
• ISSN: 0013-936X; 1520-5851
• DOI: 10.1021/acs.est.6b01035

Aerodyne aerosol mass spectrometer (AMS) and Aerodyne aerosol chemical speciation monitor (ACSM) mass spectra are widely used to quantify organic aerosol (OA) elemental composition, oxidation state, and major environmental sources. The OA CO fragment is among the most important measurements for such analyses. Here, we show that a non-OA CO signal can arise from reactions on the particle vaporizer, ion chamber, or both, induced by thermal decomposition products of inorganic salts. In our tests (eight instruments, n = 29), ammonium nitrate (NH NO ) causes a median CO interference signal of +3.4% relative to nitrate. This interference is highly variable between instruments and with measurement history (percentiles P = +0.4 to +10.2%). Other semi-refractory nitrate salts showed 2-10 times enhanced interference compared to that of NH NO , while the ammonium sulfate ((NH ) SO ) induced interference was 3-10 times lower. Propagation of the CO interference to other ions during standard AMS and ACSM data analysis affects the calculated OA mass, mass spectra, molecular oxygen-to-carbon ratio (O/C), and f . The resulting bias may be trivial for most ambient data sets but can be significant for aerosol with higher inorganic fractions (>50%), e.g., for low ambient temperatures, or laboratory experiments. The large variation between instruments makes it imperative to regularly quantify this effect on individual AMS and ACSM systems.