Abstract: Objective　Over the past decade, the concentrations of particulate matter (PM2.5, PM10) and gaseous primary pollutants (SO2, NO2, etc.) in China have decreased significantly. However, secondary pollution represented by ozone has become increasingly prominent, and secondary aerosols generated through processes such as photo-oxidation have become a major source of urban atmospheric particulate matter. Taizhou, located in central Jiangsu, is a typical industrialized city. To achieve sustained reduction in PM2.5 concentrations, it is urgent to formulate locally tailored air-pollution response plans, based on scientific evidence regarding local PM2.5 composition and pollution sources. Nevertheless, systematic research on PM2.5 characteristics and sources in Taizhou remains limited. Therefore, investigating the pollution characteristics and sources of PM2.5 in Taizhou has important demonstration and reference significance for precise air-pollution control in the Yangtze River Delta region. For this purpose, this study conducted a full-year collection of PM2.5 filter samples in Taizhou during 2021, analyzed their chemical composition and seasonal variation patterns, and quantified the contribution of different pollution sources and their seasonal characteristics. Methods　The sampling site was located at the Taizhou Municipal Bureau of Justice, approximately 200 m from Taizhou Lianhua National Automatic Monitoring Station. Meteorological parameters (temperature, relative humidity, wind direction, and wind speed), PM2.5 and PM10 concentrations, and conventional gaseous pollutants (O3, NO2, SO2, and CO) concentrations from Taizhou Lianhua National Automatic Monitoring Station were used for data analysis. The sampling site is surrounded by residential areas and is representative of a typical urban setting. Moreover, there are no large industrial sources within a 2 km radius of the site. A Thermo Fisher Scientific 2000i atmospheric particulate matter sampler was used to collect PM2.5 samples. The sampling period was from January 1, 2021, to March 31, 2022. Daily sampling duration was 23 hours, and samples were collected every 3 days, so in total, 152 filter samples were manually collected. After extraction from the filters, the water-soluble ionic components in PM2.5 were analyzed using a Metrohm 940 (Switzerland), including five cations (NH4 +, Na+, K+, Ca2+, and Mg2+) and four anions (NO3−, SO4 2−, F−, and Cl−). Organic carbon (OC) and elemental carbon (EC) in PM2.5 were analyzed using a Desert Research Institute (DRI) 2015 instrument (USA) via the thermal oxidation method. Trace elements in PM2.5 were analyzed using a BRUKER S8 Tiger (Germany), yielding 24 elements including Na, K, Mg, Ca, Ba, Cd, Sn, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, As, Se, Pb, Al, S, Cl, Sb, and Si. Source apportionment of PM2.5 filter samples was conducted using Positive Matrix Factorization (PMF) 5.0 of U.S. Environmental Protection Agency (USEPA). In the PMF calculations, the number of factors k was tested sequentially from 3 to 10, and five factors were ultimately selected. Results and Discussion　(1) PM2.5 pollution in Taizhou exhibits distinct seasonal characteristics that the concentrations of PM2.5 are higher in spring and winter, with water-soluble inorganic salts and organic matter as the dominant chemical components. Seasonal comparisons show that the proportion of nitrate increases in winter and decreases in summer, whereas sulfate shows the opposite trend. On polluted days (daily PM2.5 concentration > 35 μg/m³), nitrate increases significantly and becomes the primary component. (2) PMF source apportionment identifies five major sources of PM2.5, including secondary sources, industrial sources, combustion sources, dust sources, and traffic sources. Secondary sources are dominated by secondary inorganic salts. Industrial sources are characterized by metals such as Ba, Cd, and V. Combustion sources are traced according to species including K+ and Cl− , with primary contributions from fireworks and biomass burning. Dust sources are mainly composed of crustal elements (e.g., Ca2+ and Al). Traffic sources make a prominent contribution to EC. Pollution sources such as regional transport, local vegetation emissions, and residential cooking emissions are categorized as "other sources". (3) Seasonal source apportionment results indicate that secondary sources dominate throughout the year, with the highest contribution in winter (44.0%) and a decrease in summer (36.7%). Industrial sources and dust sources show similar seasonal trends, with their proportions increasing in summer. Combustion sources are more significant in autumn and winter, associated with straw burning and heating activities. Due to the promotion of new energy vehicles and pollution prevention and control measures in recent years, contributions from traffic sources and other sources remain below 10%. (4) Winter pollution analysis shows that the contribution of secondary sources on polluted days is significantly higher than that on clean days (48.7% vs. 34.2%), indicating that secondary transformation is the key process leading to severe pollution. Besides secondary sources, combustion sources, dust sources, and industrial sources each contribute over 15% on clean days. In contrast, on polluted days, contributions from dust sources and industrial sources decrease markedly, both falling below 10%. (5) Compared with the previous winter, the contributions of combustion sources and dust sources increased in the winter of 2022, whereas those of industrial sources and secondary sources decreased slightly. Conclusions　(1) Increases in gaseous precursors such as NO2 and SO2 provide necessary conditions for generating more secondary particulate matter. (2) The main components of PM2.5 in Taizhou are organic matter, NO3− , SO4 2− , NH4 + , and mineral components. Organic matter and sulfate exhibit the highest proportions in summer and the lowest in winter, while nitrate shows the opposite pattern. In summer, intense sunlight promotes O3 formation via NO2 photolysis, increasing atmospheric oxidation capacity and favoring the formation of secondary aerosols such as secondary organic aerosols and sulfates. Therefore, coordinated control of PM2.5 and O3 pollution is necessary to address air pollution in Taizhou. (3) PMF analysis identified six pollution-source categories. During the sampling period, the average contribution rates of secondary sources, industrial sources, combustion sources, dust sources, traffic sources, and other sources were 40.4%, 13.5%, 15.3%, 14.9%, 7.9%, and 7.8%, respectively. Secondary sources always dominated, while traffic sources and other sources contributed the least. In addition, combustion and dust contributions in Taizhou were relatively significant, whereas the proportion of industrial sources in the urban area was relatively low.

Key words: fine particulate matters, source apportionment, secondary sources