Particle characterization and quantification of organic and inorganic compounds from Chinese and Iranian aerosol filter samples using scanning laser desorption/ionization mass spectrometry

Abstract

Besides their influence on climate and cloud formation, many organic and inorganic substances in aerosol particles pose a risk to human health. Namely, polycyclic aromatic hydrocarbons (PAH) and heavy metals are suspected to be carcinogenic or acutely toxic. The detection and quantification of such compounds is difficult if only small amounts of particulate matter (PM) are available. In addition, filter samples are often complex and time-consuming to prepare for chromatographic measurements and elemental analysis. Here, we present a method based on high-resolution atmospheric pressure laser desorption ionization mass spectrometry imaging (AP-LDI-MSI) and statistical analysis which allows the analysis and characterization of very small sample quantities (< 30 µg) without any sample preparation. The power and simplicity of the method is demonstrated by two filter samples from heavily polluted mega cities. The samples were collected in Tehran (Iran) and Hangzhou (China) in February 2018. In the course of the measurement, more than 3200 sum formulae were assigned, which allowed a statistical evaluation of colocalized substances within the particles on the filter samples. This resulted in a classification of the different particle types on the filters. Finally, both megacities could be distinguished based on characteristic compounds. In the samples from Tehran, the number of sulphur-containing organic compounds was up to 6 times as high as the samples from Hangzhou, possibly due to the increasing efforts of the Chinese government to reduce sulphur emissions in recent years. Additionally, quantification of 13 PAH species was carried out via standard addition. Especially, the samples from Tehran showed elevated concentrations of PAHs, which in the case of higher-molecular-weight species (> m/z 228) were mostly more than twice as high as in Hangzhou. Both cities showed high levels of heavy metals and potentially harmful organic compounds, although their share of total particulate matter was significantly higher in the samples from Tehran. The pre-treatment of the samples was reduced to a minimum with this method, and only small amounts of particles were required to obtain a comprehensive picture for a specific filter sample. The described method provides faster and better control of air pollution in heavily polluted megacities.

Keywords: Aerosol composition; Inorganic particles; Laser desorption ionization mass spectrometry; Megacities; Organic particles; Pollution.

Fig. 1

Total number of assigned sum formulae features represented in Venn diagrams for positive (a) and negative (b) ion mode

Fig. 2

Assigned sum formulae grouped into 11 classes in negative- (a–d) and positive-ion mode (e–h). Measurements of the Chinese filters on day 1 (FPI17B0807, a, e) and day 2 (FPI17B0811, b, f) as well as for the Iranian filters on day 1 (287, c, g) and day 2 (289, d, h) exhibit characteristic features for each sampling location

Fig. 3

Plots showing the number of double-bond equivalents for 9 classes of organic molecules in negative- (a–d) and positive-ion mode (e–h). The Chinese filters on day 1 (FPI17B0807, a, e) and day 2 (FPI17B0811, b, f) as well as the Iranian filters on day 1 (287, c, g) and day 2 (289, d, h) are shown. The area of each data point corresponds to the number of compounds

Fig. 4

Representative mass spectral patterns for larger data clusters (> 30 sum formulas) for the Hangzhou sample FPI17B0807. A total of 15 data clusters, 10 for positively charged ions (a–j) and 5 for the negatively charged ions (k–o) were found and assigned to different particle types

Fig. 5

Representative mass spectral patterns for larger data clusters (> 30 sum formulas) for the Tehran sample 287. A total of 20 data clusters, 9 for positively charged ions (a–i) and 11 for the negatively charged ions (j–t) were found and assigned to different particle types

Fig. 6

Calibration curves for dried droplet standard addition method on Hangzhou samples FPI17B0807 (a), FPI17B0811 (b) and Tehran samples 287 (c) and 289 (d). Calibration curves are shown for high-molecular-weight PAHs

Fig. 7

Calculated concentrations of 4 selected PAH groups BaA/Chr, BbF/BkF/BaP, BghiP/IND and DBA for each Tehran and Hangzhou sample compared to values measured in Beijing during winter 2016 [88]