Abstract. Particulate air pollution in China is influencing human health, ecosystem and climate. However, the chemical composition of particulate aerosol, especially of the organic fraction, is still not well understood. In this study, particulate aerosol samples with a diameter ≤ 2.5 μm (PM2.5) were collected in January 2014 in three cities located in Northeast, East and Southeast China, i.e., Changchun, Shanghai and Guangzhou, respectively. Organic aerosol (OA) in the PM2.5 samples was analyzed by ultrahigh performance liquid chromatography (UHPLC) coupled to high-resolution Orbitrap mass spectrometry in both negative mode (ESI−) and positive mode electrospray ionization (ESI+). After a non-target screening including molecular formula assignments, compounds were classified into five groups based on their elemental composition, i.e., CHO, CHON, CHN, CHOS and CHONS. The CHO, CHON and CHN compounds present the dominant signal abundances of 81–99.7 % in the mass spectra and the majority of these compounds were assigned to mono- and polyaromatics, suggesting that anthropogenic emissions are a large source of urban OA in all three cities. However, the chemical characteristics of these compounds varied among different cities. The degree of aromaticity and the number of polyaromatic compounds were significantly higher in samples from Changchun, which could be attributed to the large emissions from residential heating (i.e., coal combustion) during winter time in Northeast China. Moreover, the ESI− analysis showed higher H / C and O / C ratios for organic compounds in Shanghai and Guangzhou compared to samples from Changchun, indicating that OA in lower latitude regions of China experiences more intense photochemical oxidation processes. The majority of sulfur-containing compounds (CHOS and CHONS) in all cities were assigned to aliphatic compounds with low degrees of unsaturation and aromaticity. Again, samples from Shanghai and Guangzhou exhibit a larger chemical similarity but largely differ from those from Changchun.
Embedding Feynman integral (Calabi-Yau) geometries in weighted projective space