Inter-comparison of online and offline methods for measuring ambient heavy and trace elements and water-soluble inorganic ions (NO<sub>3</sub><sup>−</sup>, SO<sub>4</sub><sup>2−</sup>, NH<sub>4</sub><sup>+</sup>, and Cl<sup>−</sup>) in PM<sub>2.5</sub> over a heavily polluted megacity, Delhi

Abstract

Abstract. Characterizing the chemical composition of ambient particulate matter (PM) provides valuable information on the concentration of secondary species and toxic metals and assists in the validation of abatement techniques. The chemical components of PM can be measured by sampling on filters and analyzing them in the laboratory or using real-time measurements of the species. It is important for the accuracy of the PM monitoring networks that measurements from the offline and online methods are comparable and biases are known. The concentrations of water-soluble inorganic ions (NO3−, SO42−, NH4+, and Cl−) in PM2.5 measured from 24 h filter samples using ion chromatography (IC) were compared with the online measurements of inorganics from an aerosol mass spectrometer (AMS) with a frequency of 2 min. Also, the concentrations of heavy and trace elements determined from 24 h filter samples using inductively coupled plasma mass spectroscopy (ICP-MS) were compared with the online measurements of half-hourly heavy and trace metal concentrations from an Xact 625i ambient metal mass monitor. The comparison was performed over two seasons (summer and winter) and at two sites (Indian Institute of Technology Delhi (IITD) and Indian Institute of Tropical Meteorology, Delhi (IITMD)) which are located in the Delhi National Capital Region (NCR), India, one of the most heavily polluted urban areas in the world. Collocated deployments of the instruments helped to quantify the differences between online and offline measurements and evaluate the possible reasons for positive and negative biases. The slopes for SO42− and NH4+ were closer to the 1:1 line during winter and decreased during summer at both sites. The higher concentrations on the filters were due to the formation of particulate (NH4)2SO4. Filter-based NO3− measurements were lower than online NO3− during summer at IITD and winter at IITMD due to the volatile nature of NO3− from the filter substrate. Offline-measured Cl− was consistently higher than AMS-derived Cl− during summer and winter at both sites. Based on their comparability characteristics, elements were grouped into three categories. The online element data were highly correlated (R2>0.8) with the offline measurements for Al, K, Ca, Ti, Zn, Mn, Fe, Ba, and Pb during summer at IITD and winter at both the sites. The higher correlation coefficient demonstrated the precision of the measurements of these elements by both the Xact 625i and ICP-MS. Some of these elements showed higher Xact 625i elemental concentrations than ICP-MS measurements by an average of 10 %–40 % depending on the season and site. The reasons for the differences in the concentration of the elements could be the distance between two inlets for the two methods, line interference between two elements in Xact measurements, the sampling strategy, variable concentrations of elements in blank filters, and the digestion protocol for ICP-MS measurements.