A field measurement based scaling approach for quantification of major ions, organic carbon, and elemental carbon using a single particle aerosol mass spectrometer. (October 2016)

• Record Type: Journal Article
• Authors: Zhou, Yang; Huang, X.H. Hilda; Griffith, Stephen M.; Li, Mei; Li, Lei; Zhou, Zhen; Wu, Cheng; Meng, Junwang; Chan, Chak K.; Louie, Peter K.K.; Yu, Jian Zhen
• Is Part Of: Atmospheric environment. Volume 143(2016)
• Page Start: 300
• Page End: 312
• Publication Date: 2016-10
• Subjects: Single particle -- PM2.5 -- Laser ionization -- Scaling factor -- Ions -- OCEC; Air -- Pollution -- Periodicals; Air -- Pollution -- Meteorological aspects -- Periodicals; 551.51
• DOI: 10.1016/j.atmosenv.2016.08.054
• Languages: English
• ISSNs: 1352-2310

Abstract

Abstract: Single Particle Aerosol Mass Spectrometers (SPAMS) have been increasingly deployed for aerosol studies in Asia. To date, SPAMS is most often used to provide unscaled information for both the size and chemical composition of individual particles. The instrument's lack of accuracy is primarily due to only a fraction of particles being detected after collection, and the instrumental sensitivity is un-calibrated for various chemical species in mixed ambient aerosols. During a campaign from January to April 2013 at a coastal site in Hong Kong, the particle number information and ion intensity of major PM2.5 components collected by SPAMS were scaled by comparing with collocated bulk PM2.5 measurements of hourly or higher resolution. The bulk measurements include PM2.5 mass by a SHARP 5030 Monitor, major ions by a Monitor for Aerosols & Gases in ambient Air (MARGA), and organic carbon (OC) and elemental carbon (EC) by a Sunset OCEC analyzer. During the data processing, both transmission efficiency (scaled with the Scanning Mobility Particle Sizer) and hit efficiency conversion were considered, and component ion intensities quantified as peak area (PA) and relative peak area (RPA) were analyzed to track the performance. The comparison between the scaled particle mass assuming a particle density of 1.9 g cm −3 from SPAMS and PM2.5 concentration showed good correlation (R 2 = 0.81) with a slope of 0.814 ± 0.004. Regression analysis results suggest an improved scaling performance using RPA compared with PA for most of the major PM2.5 components, including sulfate, nitrate, potassium, ammonium, OC and EC. Thus, we recommend preferentially scaling these species using the RPA. For periods of high K + concentrations (>1.5 μg m −3 ), under-estimation of K + by SPAMS was observed due to exceeding the dynamic range of the acquisition board. When only applying the hit efficiency correction, data for sulfate, nitrate, ammonium, potassium and OC were in reasonably good correlation (R 2 = 0.56–0.79) with the bulk measurements, suggesting semi-quantified data by the hit efficiency correction can be utilized when transmission efficiency correction is not available. More inter-comparisons of this type are needed to assess the inter-instrument variability among different SPAMS instruments. Highlights: Developed several scaling procedures for SPAMS based on a field campaign datasets. Determined a set of correction parameters to link SPAMS data with the bulk measurements. Obtained improved correlation of ion intensity by SPAMS with bulk measurements of main PM2.5 components. SPAMS could underestimate K + under high influence of biomass burning emissions.