In order to study the component characteristics and sources of PM2.5 in the northern suburbs of Nanjing, 191 PM2.5 samples were collected and quantitatively analyzed for watersoluble ions (WSI), organic carbon (OC), elemental carbon (EC) and water-soluble organic carbon (WSOC). The annual average mass concentrations of OC, EC, and WSOC were 5.30 μg m-3, 0.96 μg m-3, and 3.09 μg m-3, respectively. The variation characteristics of their seasonal abundance were basically the same, which reached the maximum in winter and the lowest in summer. Based on the ratio of organic carbon to elemental carbon, we find that the air in the northern suburbs of Nanjing is affected by coal burning, exhaust emissions and biomass combustion. The annual mean mass concentration of WSI in PM2.5 was 23.00 μg m-3, and the order of seasonal concentration was winter > autumn > spring > summer. The WSI in PM2.5 are mainly SO4 2-, NO3 - , and NH4 +. Through the concentration ratio of NO3 - /SO4 2-, we found that mobile sources were the main sources of pollution in autumn and winter, while stationary sources were the main sources of pollution in summer. Through the iterative calculation of PMF, five sources of PM2.5 are determined, which are second generation (37.5%), traffic-related source (36.7%), combustion source (9.6%), marine source (8.5%), and industrial emissions (7.7%). In different seasons, PM2.5 is significantly correlated with SO4 2-, NO3 - , and NH4 +, indicating that PM2.5 mainly comes from secondary generation.

• PM2.5

• Carbon components

• Water-soluble ions

• Source analysis

Chen M, Yu W, Cao M (2022) Pollution Characteristics and Sources of Carbon Components and Water-Soluble Ions in Atmospheric Particulate Matter in Nanjing, China. JSM Environ Sci Ecol 10(2): 1082.

OC: Organic Carbon; EC: Elemental Carbon; CCN: Cloud Condensation Nuclei; WSI: Water-Soluble Inorganic ions; TCA: Total Carbon Aerosol; WSOC: Water-Soluble Organic Carbon; POC: Primary Organic Carbon; SOC: Secondary Organic Carbon

In recent years, China often appears haze weathers, severe haze will reduce atmospheric visibility, affect climate change and even affect human health [1-4]. PM2.5 was the main monitoring index, and secondary inorganic aerosols (SO4 2-, NH4 + , and NO3 - ), organic carbon (OC), elemental carbon (EC) and other trace organic compounds were the main components [5-7].

Carbon aerosol is an important part of atmospheric aerosol, accounting for 10-70% of PM2.5 [8,9]. It is mainly divided into organic carbon (OC), elemental carbon (EC) and inorganic carbon (IC) [1,10]. OC is mainly generated from biomass and coal combustion, direct emissions from tail gas and industrial emissions, or through photochemical oxidation of organic gas [11]. The former is usually defined as primary organic carbon and the latter as secondary organic carbon according to the mode of source [9,12]. It can also affect cloud condensation nuclei (CCN) which then impacts earth radiation balance and global climate change [13,14]. When biomass and fossil fuels are burned incompletely, large amounts of EC are produced. Because of its good adsorption activity, it can adsorb secondary pollutants and have a negative impact on the environment [15,16].

Water-soluble inorganic ions (WSI), as important components, account for about 50% of PM2.5 [17]. They are often used to identify the source of PM2.5. For example, K+ can be used as a tracer to evaluate biomass combustion; SO4 2-, NH4 + , and NO3 - are related to secondary generation; high abundances of Na+ and Cl- are generally found in marine aerosols, which can be transported to coastal and inland areas due to air mass transport [18,19]. Previous studies have shown that water-soluble ions are hygroscopic and can affect aerosol composition, size and PH [17,20].

The Yangtze River Delta region, China’s largest economic zone, has a large population and dense cities. This area is not only an important agricultural production base in China, but also a large number of chemical enterprises. A large number of gaseous pollutants are discharged in the process of industrial production and vehicle driving, which makes the air pollution situation increasingly serious and seriously affects people’s health and daily life. The Yangtze River Delta is faced with serious air pollution due to a large number of human activities. As a megalopolis in the Yangtze River Delta region and the capital city of Jiangsu Province,Nanjing also faces serious air pollution problems. Therefore, it is necessary to monitor carbon components and water-soluble ions in the atmosphere of Nanjing.

Therefore, this study conducted one-year sampling in the northern suburbs of Nanjing from March 2017 to February 2018, and analyzed PM2.5 samples by ion chromatography. The mass concentrations of five kinds of cations (Na+ , NH4 + , K+ , Mg2+, Ca2+) and four kinds of anions (F- , Cl- , SO4 2-, NO3 - ) were obtained, and the concentration level and correlation of water-soluble ions as well as the ratio of NO3 - /SO4 2- were analyzed. In order to comprehensively understand the pollution characteristics of PM2.5 in the northern suburbs of Nanjing, the OC, EC and WSOC in the samples were measured by the elemental carbon organic carbon analyzer and total organic carbon analyzer, and the values of POC and SOC were estimated, so as to comprehensively understand the pollution characteristics and sources of carbon components in PM2.5 in the northern suburbs of Nanjing.

Experimental MethodsSampling location and time

The sampling site is set on the roof of the library of Nanjing University of Information Science and Technology (N32°12′9′′, E118°42′49′′, Figure 1),

Figure 1: The location of sampling site.

about 20 m from the ground, located in the northern suburbs of Nanjing, 23 km from the urban area. There are no obstructing objects around him, so the air flows well. It is Jiangbei Avenue Expressway 1 km to the east, and the traffic flow is large. The sampling time is from March 2017 to February 2018, 24 hours continuous sampling from 12:00 to 12:00 the next day. The PM2.5 large flow sampler (TISCH, USA) is used, and the flow rate is 1.13 m3 /min. The sampling period is mainly sunny and cloudy. In case of rain and snow, the sampler will be turned off. Periodically clean the sampler and calibrate the flow.

Sample collection and storage

The sampling membrane was quartz filter membrane (QM/A; Whatman; Middlesex, UK), 20.30 cm × 25.40 cm. Before sampling, the quartz filter membrane was placed in a Muffle furnace and roasted at 450 ? for 4 h to remove residual organic matter, and then balanced in a constant temperature tank for 24 h. The quartz filter film before and after sampling was weighed by 1/100,000th electronic balance, which was weighed three times, and the average value of the three times was recorded as the weight of the filter film. The quality difference between quartz filter film before and after sampling is the quality of collected particles. After sample collection, the filter film was wrapped with tin foil and stored in a refrigerator at -20? for subsequent sample analysis (Figure 1).

OC and EC Analysis

The OC and EC in the sample were determined by the elemental carbon organic carbon analyzer (Sunset Model, United States), which consists of organic carbon/elemental carbon analyzer host, vacuum pump and computer. Before injection and analysis, the filter membrane and blank filter membrane with a diameter of 17 mm collected PM2.5 samples were cut with a punch. Also, the instrument should be turned on for an hour to stabilize and calibrated with a 99.9% sucrose solution. The main calibration uses sucrose solution, which can be used for online monitoring or offline monitoring, and can analyze and measure elemental carbon and organic carbon at the same time. Its detection limit was 0.2 μg/cm2 and the detection range was 0.2-1000 μg/cm2 . In addition, during the determination of aerosol samples, a blank sample analysis was performed for every ten samples.

Water-soluble inorganic ions analysis

Anions (F- , Cl- , NO3 - , SO4 2-) were analyzed by using the Ion chromatograph (ICS-2000) produced by Diane Company in the United States, and cations (NH4 + , Ca2+, Na+ , Mg2+, K+ ) were analyzed by using the ion chromatograph (ICS-3000). First analyze the standard sample solution prepared with chromatographically pure pharmaceutical grade, then create a standard curve, and finally inject the sample for analysis. The sample pretreatment steps are: take a filter membrane with a diameter of 3.6 cm, cut it into pieces with ceramic scissors, put it into a 50 mL centrifuge tube, add 45 mL of ultrapure water, tighten the cap, seal it with tin foil, and add an ice pack for sonication. Extraction was carried out for 60 min, during which the water temperature could not exceed 20°C. After sonication, a 0.22 μm PTFE syringe-type microporous membrane was used to filter the samples for analysis. The correlation coefficient of the configured standard curve is above 0.999. In order to reduce the error of the blank sample and make the result more reliable, a blank control experiment is set up. Two blank samples and five samples to be tested were ndomly selected, and quantitative standard solution was added. The recoveries were all between 90% and 110% after instrument determination. The detection limits of each ion are shown in Table S1.

WSOC Analysis

The total organic carbon/total nitrogen analyzer (TOC/TN instrument, TOC-L type, Shimadzu Corporation, Japan) was used to analyze the content of WSOC in the sample, and the oxidative combustion infrared analysis method was used to measure the total carbon (TC) and Inorganic carbon (IC), the value of WSOC is obtained by subtracting the two. The operation was as follows: 3.6 cm diameter sample filter membrane and blank filter membrane were cut with a hole punch, 45 mL ultrapure water was added, ultrasonic extraction was carried out for 3 times, 20 minutes each time, then filtration was carried out, and finally 30 mL sample was taken for analysis. Before WSOC analysis, ensure that the correlation coefficient required by the standard curve reaches 99.9%, and analyze 30 mL of ultrapure water before analyzing the atmospheric sample, and do blank sample analysis for every ten samples.

Data Analysis

SPSS software was used to analyze the correlation between carbon components, water-soluble ions and PM2.5, so as to further explore the source of PM2.5.

In this study, OC, EC and water-soluble ions were used to determine the source of PM2.5. A detailed description of the PMF software and a user’s guide can be found on this website: www.epa.gov/air-research/positive-matrix-factorizationmodelenvironmental-data-analyes. Two data files are required to run the PMF model: concentration files and uncertainty files. The error rate was set at 15%. For the uncertainty value, if the concentration is less than or equal to the method detection limit: Unci =5/6MDL

If the concentration is greater than the method detection limit:

Ambient Concentrations of Carbonaceous Aerosols in PM2.5 During 2017

Overview of Carbon Components Concentrations: The concentrations of OC, EC and WSOC in atmospheric PM2.5 samples from the northern suburbs of Nanjing were obtained, as well as the mean, maximum and minimum values of each season from March 2017 to February 2018. Total carbon aerosol (TCA) can be estimated by the mass concentration of OC and EC, the specific formula is TCA=1.6×OC+EC.

As can be seen from Table 1,

Table 1: Annual seasonal concentrations of carbon components and water-soluble ions in PM2.5 in Nanjing in 2017.

the average concentrations of OC, EC and WSOC in PM2.5 in the northern suburbs of Nanjing in 2017 were 5.30 μg m-3, 0.96 μg m-3, and 3.09 μg m-3, respectively. The maximum value of OC (20.71 μg m-3) appeared on November 23, while the minimum value (1.08 μg m-3) appeared on October 17. The maximum peak value of EC (3.46 μg m-3) also occurred on October 17, but the minimum peak value (0.22 μg m-3) occurred on October 15. It could be seen from Table 1 that the annual average mass concentration of TCA was 9.31 μg m-3. The maximum concentration was 36.6 μg m-3 on November 23 in autumn, which was consistent with the maximum value of OC and EC. The minimum value also appeared in autumn, and the mass concentration was 1.96 μg m-3 (Table 1).

Table 2 shows the comparison of the mass concentrations and ratios of OC and EC in the atmospheric PM2.5 in the northern suburbs of Nanjing and others cities.

Table 2: Comparison of OC and EC concentrations (µg m−3) measured in worldwide.

It can be seen that compared with the northern cities such as Handan and Beijing, the atmospheric concentration level of OE and EC in Nanjing is lower. Among them, the mass concentrations of OC and EC in Handan were higher than those in major cities in China, which were 3 times and 10 times higher than those in the northern suburbs of Nanjing in this study, respectively. This may be due to the industrial structure of resource-based heavy industry in the region and the use of coal for heating in winter. The mass concentrations of OC and EC in Shanghai was significantly lower than those in the inland cities in the table, which may be due to the influence of oceanic airflow. The mass concentration of OC in Shanghai is slightly higher than that of this sampling point, and the mass concentration of EC is about 3.5 times that of this sampling point, and the source of EC is related to vehicle exhaust emissions [21]. This may be due to the developed traffic in Shanghai and the large number of motor vehicles, resulting in increased exhaust emissions.

Seasonal Characteristics of Carbon Components: In the northern suburbs of Nanjing, the concentration of OC was the highest in winter (8.29 μg m-3), followed by spring (6.45 μg m-3) and autumn (4.42 μg m-3), and the lowest in summer (3.47 μg m-3). The average mass concentration of OC in winter was 2.4 times that of summer. OC showed obvious seasonal variation, which may be closely related to seasonal climate change. Because the sampling site is close to the industrial zone, there is a large number of industrial combustion phenomenon, and residents also burn wood for heating. In winter, the mixed layer height is low, and PM2.5 tends to gather to a higher concentration, resulting in the highest OC concentration in winter. The average mass concentration of OC was the lowest in summer, which may be due to the influence of East Asian monsoon in Nanjing, which resulted in more atmospheric precipitation, which was conducive to the diffusion and removal of pollutants.

The average mass concentration of EC in PM2.5 in spring, summer, autumn and winter were 0.96 μg m-3, 0.68 μg m-3, 0.70 μg m-3, and 1.18 μg m-3, respectively. It can be seen that the average mass concentration of EC is the highest in winter and the lowest in summer, of which the average mass concentration of EC in winter is 1.7 times that in summer. This may be due to the fact that the sampling site is close to Jiangbei Expressway, and the large traffic flow in winter leads to larger automobile exhaust emissions. In addition, under cold meteorological conditions, it is not conducive to the diffusion of pollutants, so the concentration of EC is higher in winter. In contrast, the average mass concentration of EC in summer is the lowest, which may be because of more rainfall in summer, which can remove pollutants in the atmosphere to some extent. The seasonal variation trend of TCA and WSOC was the same as that of OC and EC, with the highest average mass concentration in winter and the lowest in summer. In addition, the average mass concentration of TCA in winter was 4.2 times that in summer.

OC/EC: OC/EC ratio is not only related to the source of carbon aerosols, but also can be used to judge the existence of secondary organic aerosols [22]. If the ratio of OC/EC is greater than 2, it indicates the presence of SOC in the atmosphere [23]. As can be seen from Table S2, the average annual ratio of OC/EC in PM2.5 in the northern suburbs of Nanjing in this study is 6.79, and the ratios of OC/EC in spring, summer, autumn and winter are 7.65, 5.71, 6.67 and 7.74, respectively. Compared with major cities in China, the OC/EC ratio in the northern suburbs of Nanjing is higher than that in Beijing and Shanghai (Table 2).

Studies have shown that the OC/EC ratio of gasoline and diesel vehicle exhaust is 1.0-4.2, and the OC/EC ratio of coal is 2.5-10.5. In addition, the OC/EC ratio is between 8.1-12.7, indicating the existence of biomass combustion emissions. As can be seen from Table S2, during the sampling period, the annual mean value of OC/EC in the northern suburbs of Nanjing was 6.79. The ratio of OC/EC in spring, summer, autumn and winter was 7.65, 5.71, 6.67 and 7.74, respectively, with the highest value in winter, followed by spring and autumn, and the lowest value in summer. In addition, the minimum OC/EC ratio was 1.79 and the maximum was 12.92 in the whole year. Therefore, Nanjing may be affected by vehicle exhaust, coal burning and biomass combustion.

SOC Estimation: OC includes primary organic carbon (POC) and secondary organic carbon (SOC). SOC is mainly generated by the photochemical reaction of volatile organic compounds (VOCs) in the atmosphere. In this study, the following formula is used to calculate SOC and POC [24]:

SOC=OC-POC

POC=EC× (OC/EC)min

where, (OC/EC)min is the minimum value of OC/EC during the sampling period. Table S2 shows that the annual mean concentration of POC is 2.16 μg m-3. The average concentration of POC was 3.62 μg m-3, 2.73 μg m-3, 1.80 μg m-3, 1.21 μg m-3 in the four seasons from high to low, which was consistent with the seasonal trend of EC. In addition, the annual average proportion of POC in OC was 41.24%, less than 50%, indicating that OC was less affected by direct discharge of pollution sources. As can be seen from Table S2, the annual average mass concentration of SOC is 3.13 μg m-3, and the average mass concentration in spring, summer, autumn and winter is 3.72 μg m-3, 2.26 μg m-3, 2.62 μg m-3, 4.67 μg m-3, respectively. Meanwhile, the annual average SOC/OC ratio was 58.76%, and the SOC/OC ratio was 57.20%, 64.99%, 55.54%, and 55.88% in spring, summer, autumn and winter, respectively. The concentration of SOC was the highest in winter, about twice that in summer. This may be because the atmosphere is relatively stable in winter, which is not conducive to the diffusion of pollutants, and because of the increase of coal and biomass burning for winter heating, which leads to the highest concentration of SOC in winter. In summer, Nanjing is affected by monsoon, and precipitation weather increases. Precipitation can play a certain role in cleaning suspended particulate matter and gaseous pollution in the atmosphere, so the concentration of SOC in summer is the lowest (Table 2).

Ambient Concentrations of Water-Soluble Ions in PM2.5 During 2017

Overall Pollution Levels of Water-Soluble Ions: In this study, PM2.5 samples were collected in the northern suburbs of Nanjing from March 2017 to February 2018, and WSI was measured and analyzed. The annual average mass concentration of total water-soluble ions is 23.00 μg m-3, the minimum value is 5.55 μg m-3 in autumn, and the maximum value is 71.52 μg.m-3 in winter. It can be found that during the sampling period from February 2017 to March 2018, the maximum value of total watersoluble ions was 12.8 times the minimum value, indicating that the mass concentration of water-soluble ions in the atmospheric PM2.5 in the northern suburbs of Nanjing changed greatly. The annual average mass concentration of WSI is in the order of NO3 - > SO4 2- > NH4 + > Ca2+ > Na+ > Cl- > K+ > Mg2+ > F- . The average mass concentrations were 7.47 μg m-3, 6.68 μg m-3, 4.39 μg m-3, 1.36 μg m-3, 0.99 μg m-3, 0.95 μg m-3, 0.75 μg m-3, 0.35 μg m-3, and 0.06 μg m-3, respectively (Figure 2).

Figure 2: The proportion of water-soluble ions in different seasons (a) spring, (b) summer, (c) autumn, (d) winter.

Seasonal Characteristics of Water-Soluble Ions: The average mass concentrations of WSI in PM2.5 in the northern suburbs of Nanjing in spring, summer, autumn and winter were 18.60 μg m-3, 14.06 μg m-3, 24.08 μg m-3, and 36.95 μg m-3, respectively. In terms of seasons, the mass concentration of WSI was the highest in winter, and decreased in order of autumn, spring and summer. It can be seen from Figure 2 that the proportion of WSI is arranged in different order in each season. The proportion of different ions to total ions in spring is SO4 2- > NO3 - > NH4 + > Ca2+ > K+ > Na+ > Cl- > Mg2+ > F- . The proportion of WSI in summer is SO4 2- > NO3 - > NH4 + > Na+ > Ca2+ > K+ > Mg2+ > Cl- > F- . The ratio of Na+ was higher in summer than in spring. The proportion of WSI in autumn is in the order of NO3 - > SO4 2- > NH4 + > Ca2+ > Na+ > K+ > Cl- > Mg2+ > F- . The proportion of different ions to total ions in winter is NO3- > SO4 2- > NH4 + > Cl- > Ca2+ > Na+ > K+ > F- . It can be seen that although the contributions of different ions are different in the four seasons, secondary inorganic ions (SO4 2-, NO3 - , and NH4 + ) occupy more than 70% of all ions in any season, which further indicates that secondary pollution occurs in the northern suburbs of Nanjing.

Ratio Analysis of NO3 - /SO4 2-: At present, the mass concentration ratio of NO3 - /SO4 2- is used to judge the contribution of mobile sources and fixed sources to relative atmospheric pollution. If the mass concentration ratio of NO3 - /SO4 2- is greater than 1, it indicates that the contribution of mobile sources is greater; if the mass concentration ratio of NO3 - /SO4 2- is less than 1, it indicates that the contribution of fixed sources is greater [25-29]. From Table S3 NO3 - /SO4 2--ratios in all seasons and throughout the year, it can be concluded that the annual average NO3 - /SO4 2- ratio in PM2.5 in the northern suburbs of Nanjing is 1.18. The average NO3 - /SO4 2- ratios in autumn and winter are all greater than 1, which are 1.19 and 1.97, respectively. This indicates that the contribution of mobile sources is greater than that of fixed sources in autumn and winter. Mobile sources are the main source of pollution, and mobile source pollution in winter is more serious than that in autumn. In summer, the average NO3 - / SO4 2- ratio is less than 1, indicating that the contribution of mobile sources is smaller than that of fixed sources in summer, and fixed sources are the main sources of pollution (Figure 3).

Figure 3 :(a) Profile of each factor by PMF analyses (b) Source contribution of the five factors to the total components in PM2.5 of Nanjing in 2019.

Possible Sources of PM2.5

PMF Analysis: Using EPA PMF5.0 model and obtained data (191 samples × 11 species), the source contribution of PM2.5 was explored. The exploration of PMF optimal scheme is determined by the number of factors of the change scheme. We simulated the schemes with 4-6 factors in turn, and each scheme was run and calculated 20 times. Finally, the scheme of 5 factors is determined as the best scheme, because it has more physical significance than others. Other schemes are detailed in Figure S1,2. Therefore, the optimal solution is when the number of analytical factors of PMF is 5. Their source profiles and percentage contributions are depicted in (Figure 3).

We can see from Figure 3(a) that Na+ (77.3%) has a high proportion in factor 1, and high abundances of Na+ is generally found in marine aerosols. So factor 1 is defined as marine source. Factor 2 is defined as secondary generation, in which the contribution ratio of SO4 2- (85.3%), NH4 + (37.5%), and NO3 - (22%) is higher, and these ions are related to secondary generation. The proportion of OC and EC in factor 3 was 79.8% and 68.9%, respectively. In addition, Ca2+ (46.9%) also made certain contribution. Because OC and EC mostly come from biomass combustion and coal combustion, factor 4 is defined as the combustion source. Factor 4 is characterized by a high proportion of contributions from F- (83.2%), Cl- (36.2%), and Ca2+ (27.8%), which may come from industrial emissions. Factor 5 is defined as a traffic-related source, because NO3 - (60.8%) and NH4 + (54.5%) have high contributions in this factor, because the exhaust emissions of vehicles such as cars will release pollutants such as nitrogen oxides. Figure 3(b) presents the contributions of five sources to mass PM2.5, where second generation, trafficrelated source, combustion sources, marine source, and industrial emissions are seen to account for 37.5%, 36.7%, 9.6%, 8.5%, and 7.7% of the total PM2.5 mass in the suburban region. In Nanjing, secondary generation and traffic emissions are the main sources of PM2.5 pollution, which together contribute 74.2%.

The Dominant Sources of PM2.5 in Different Seasons: In order to better understand the dominant sources of PM2.5 in different seasons, we conducted correlation analysis between PM2.5 and characteristic components. It can be seen from Table S4 that PM2.5 has a good correlation with Ca2+ (r=0.670, p=0.01), Cl- (r=0.666, p=0.01), and F- (r=0.657, p=0.01), which means that most of PM2.5 in Nanjing in spring may come from industrial emissions. At the same time, secondary ions (SO4 2-, NH4 + , and NO3 - ) also show a good correlation with PM2.5, and the contribution of secondary generation to PM2.5 cannot be ignored. The correlation between NH4 + and NO3 - was the highest with a correlation coefficient of 0.895, followed by NO3 - and Cl- with a correlation coefficient of 0.668. NH4 + is likely to exist in the form of NH4 NO3 , NH4 Cl, and (NH4 )2 SO4 . It can be seen from Table S5 that PM2.5 is significantly correlated with NO3 - , SO4 2-, and NH4 + at 0.01 level, with correlation coefficients of 0.460, 0.499 and 0.590, respectively, indicating that PM2.5 and SNA (NO3 - , SO4 2-, and NH4 + ) have the same source in summer. NH4 + was significantly correlated with SO4 2- and NO3 - , with correlation coefficients of 0.830 and 0.452, respectively. NH4 + may exist in the form of NH4 NO3 and (NH4 )2 SO4 . In addition, PM2.5 has a certain correlation with K+ (r=0.448, p=0.01) and Ca2+ (r=0.342, p=0.05). This means that there may be biomass burning during the season because K+ is a tracer of biomass burning. As can be seen from Table S6, PM2.5 is significantly correlated with K+ , SO4 2-, NO3 - , NH4 + , and Ca2+ at the level of 0.01, with correlation coefficients of 0.566, 0.602, 0.658, 0.620, and 0.566, respectively. Different from summer, the correlation between PM2.5 and K+ , Ca2+, and Mg2+ increased in autumn, but the correlation between PM2.5 and SNA did not change significantly. K+ generally comes from biomass combustion, while Ca2+ and Mg2+ come from natural sources such as road dust and soil, indicating that the influence of biomass combustion and natural sources is greater in autumn than in summer. In winter, PM2.5 is significantly correlated with SO4 2-, NO3 - , and NH4 + at 0.01 level, with correlation coefficients of 0.556, 0.566 and 0.637, respectively. Unlike in other seasons, there is no good correlation between other ions and PM2.5, which means that PM2.5 mainly comes from secondary generation in this season.

In this study, carbon components (OC, EC, and WSOC) and water-soluble ions in 191 PM2.5 samples were determined. In the PM2.5 samples collected in 2017, the average concentration of total carbon aerosol (TCA) was 9.31 μg m-3, and the average concentration of OC and EC were 5.30 and 0.96, respectively. The concentrations of OC and EC in Nanjing were lower than those in North China, but similar to those in Shanghai. This may be due to the industrial structure of resource-based heavy industries in North China and the use of coal for heating in winter. All carbon components have obvious seasonal variation characteristics, that is, the highest abundance in winter and the lowest in summer. The dominant carbon components in the other three seasons were different, which might be due to the different source contributions of different components. OC/EC ratios in different seasons were used to explore the source of PM2.5 in Nanjing, and it was found that may be affected by automobile exhaust, coal burning and biomass combustion.

The annual average mass concentration of WSI in PM2.5 was 23.00 μg m-3, and the average mass concentration of WSI in spring, summer, autumn and winter were 18.60 μg m-3, 14.06 μg m-3, 24.08 μg m-3 and 36.95 μg m-3, respectively. The WSI in PM2.5 are mainly SNA, which account for 80.62% of the total WSI. According to the analysis of NO3 - /SO4 2- mass concentration ratio, the results show that mobile source is the main source of pollution in autumn and winter, and fixed source is the main source of pollution in summer. In order to further understand the source of PM2.5, the data of OC, EC and WSI were poured into PMF model to identify the source of PM2.5. PMF results show that PM2.5 mainly comes from five sources, namely transport-related sources, secondary generation, combustion sources, marine source and industrial emissions, with contribution rates of 37.5%, 36.7%, 9.6%, 8.5%, and 7.7%, respectively. The correlation between PM2.5 and ions is used to judge the source of advantage in different seasons. PM2.5 was significantly correlated with SNA in different seasons, especially in autumn. Therefore, secondary generation was the main source of PM2.5 in the northern suburbs of Nanjing in 2017.

This research was funded by the National Natural Science Foundation of China (grant number 21976094).

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

1. Haoming Yang, Junfeng Wang, Mindong Chen, Dongyang Nie, Fuzhen Shen, et al. Chemical characteristics, sources and evolution processes of fine particles in Lin’an, Yangtze River Delta, China. Chemosphere. 2020; 254: 126851.
2. Ostro B, Malig B, Broadwin R, Basu R, Gold EB, Bromberger JT, et al. Chronic PM2.5 exposure and inflammation: Determining sensitive subgroups in mid-life women. Environ Res. 2014; 132: 168-175.
3. Tao J, Gao J, Zhang L, Zhang R, Che H, Zhang Z, et al. PM2.5 pollution in a megacity of southwest China: source apportionment and implication. Atmos. Chem. Phys. 2014; 14: 8679-8699.
4. Maoyu Cao, Wenjing Li, Pengxiang Ge, Mindong Chen, Junfeng Wang. Seasonal variations and potential sources of biomass burning tracers in particulate matter in Nanjing aerosols during 2017-2018. Chemosphere. 2022; 303: 135015.
5. Min Zhou, Liping Qiao, Shuhui Zhu, Li Li, Shengrong Lou, Hongli Wang, et al. Chemical characteristics of fine particles and their impact on visibility impairment in Shanghai based on a 1-year period observation. J Environ Sci. 2016, 48: 151-160.
6. Kin Fai Ho, Guenter Engling, Steven Sai Hang Ho, Ru-Jin Huang, Senchao Lai, Junji Cao, et al. Seasonal variations of anhydrosugars in PM2.5 in the Pearl River Delta Region, China. Tellus B. 2014; 66: 22577.
7. Junfeng Wang, Xinlei ge, Yafei Shen, Zhang Qi, Jianzhong Xu, Huan Yu et al. Highly time-resolved urban aerosol characteristics during springtime in Yangtze River Delta, China: insights from soot particle aerosol mass spectrometry. Atmos Chem Phys. 2016; 16: 9109-9127.
8. Barbara J.Turpin, PradeepSaxena, ElisabethAndrews. Measuring and simulating particulate organics in the atmosphere: problems and prospects. Atmos Environ. 2000; 34: 2983-3013.
9. Zheng N, Song S, Jin X, Jia H, Wang Y, Ji Y, et al. Assessment of Carbonaceous Aerosols at Mount Tai, North China: Secondary Formation and Regional Source Analysis. Aerosol Air Qual Res. 2019; 19: 1708-1720.
10. Wang G, Cheng S, Li J, Lang J, Wen W, Yang X, et al. Source apportionment and seasonal variation of PM2.5 carbonaceous aerosol in the Beijing- Tianjin-Hebei Region of China. Environ Monit Assess. 2015; 187: 143.
11. Nguyen Thi Thu Thuy, Nghiem Trung Dung, Kazuhiko Sekiguchi, Ly Bich Thuy, Nguyen Thi Thu Hien, Ryosuke Yamaguchi. Mass Concentrations and Carbonaceous Compositions of PM0.1, PM2.5, and PM10 at Urban Locations in Hanoi, Vietnam. Aerosol Air Qual Res. 2018, 18: 1591-1605.
12. Begoña Artíñano, Pedro Salvador, Francisco J. Gomez-Moreno, Manuel Pujadas, Casimiro Adrião Pio. Short-term secondary organic carbon estimations with a modified OC/EC primary ratio method at a suburban site in Madrid (Spain). Atmos Environ. 2011; 45: 2496- 2506.
13. Ramírez Omar, Sánchez de la Campa A.M, de la Rosa Jesús. Characteristics and temporal variations of organic and elemental carbon aerosols in a high-altitude, tropical Latin American megacity. Atmos Res. 2018; 210: 110-122.
14. Zhao Q, Qin S, Zhao C, Sun Y, Panchal B, Chang X. Origin and geological implications of super high sulfur-containing polycyclic aromatic compounds in high-sulfur coal. Gondwana Res. 2021; 96: 219-231.
15. Tshehla CE, Wright CY. Spatial and Temporal Variation of PM10 from Industrial Point Sources in a Rural Area in Limpopo, South Africa. Int. J. Environ. Res. Public Health. 2019; 16: 3455.
16. Xue F, Niu H, Hu S, Wu C, Zhang C, Gao N, et al. Seasonal variations and source apportionment of carbonaceous aerosol in PM2.5 from a coal mining city in the North China Plain. Energy Exploration & Exploitation. 2022; 40: 834-851.
17. Hu G, Zhang Y, Sun J, Zhang L, Shen X, Lin W, et al. Variability, formation and acidity of water-soluble ions in PM2.5 in Beijing based on the semi-continuous observations. Atmos Res. 2014; 145: 1-11.
18. Wang B, Tang Z, Cai N, Niu H. The characteristics and sources apportionment of water-soluble ions of PM2.5 in suburb Tangshan, China. Urban Clim. 2021; 35.
19. Zhang X, Zhao X, Ji G, Ying R, Shan Y, Lin Y. Seasonal variations and source apportionment of water-soluble inorganic ions in PM2.5 in Nanjing, a megacity in southeastern China. J Atmos Chem. 2019; 76: 73-88.
20. Yue D, Hu M, Wu Z, Wang Z, Guo S, Wehner B, et al. Characteristics of aerosol size distributions and new particle formation in the summer in Beijing. J Geophys Res Atmos. 2009; 114.
21. Charron A, Polo-Rehn L, Besombes JL, Golly B, Buisson C, Chanut H, et al. Identification and quantification of particulate tracers of exhaust and non-exhaust vehicle emissions. Atmos Chem Phys. 2019; 19: 5187-5207.
22. Duan J, Tan J, Cheng D, Bi X, Deng W, Sheng G, et al. Sources and characteristics of carbonaceous aerosol in two largest cities in Pearl River Delta Region, China. Atmos Environ. 2007; 41: 2895-2903.
23. Zhang L, Huang Y, Liu Y, Yang F, Lan G, Fu C, et al. Characteristics of Carbonaceous Species in PM2.5 in Wanzhou in the Hinterland of the Three Gorges Reservior of Northeast Chongqing, China. Atmos. 2015; 6: 534-546.
24. Castro LM, Pio CA, Harrison RM, Smith D. Carbonaceous aerosol in urban and rural European atmospheres: estimation of secondary organic carbon concentrations. Atmos Environ. 1999; 33: 2771-2781.
25. Wang F, Guo Z, Lin T, Rose NL. Seasonal variation of carbonaceous pollutants in PM2.5 at an urban ‘supersite’ in Shanghai, China. Chemosphere. 2016; 146: 238-244.
26. Ding J, Huang W, Zhao J, Li L, Xiong G, Jiang C, et al. Characteristics and source origins of carbonaceous aerosol in fine particulate matter in a megacity, Sichuan Basin, southwestern China. Atmos Pollut Res. 2022; 13.
27. Choi J, Heo J, Ban S, Yi S, Zoh K. Chemical characteristics of PM2.5 aerosol in Incheon, Korea. Atmos Environ. 2012; 60: 583-592.
28. Niu Z, Zhang F, Chen J, Yin L, Wang S, Xu L. Carbonaceous species in PM2.5 in the coastal urban agglomeration in the Western Taiwan Strait Region, China. Atmos Res. 2013; 122: 102-110.
29. Arimoto R, Duce RA, Savoie DL, Prospero JM, Talbot R, Cullen JD, et al. Relationships among aerosol constituents from Asia and the North Pacific during PEM-West A. J Geophys Res. 1996; 101.