The Characteristics of the Chemical Composition of PM_2.5 during a Severe Haze Episode in Suzhou, China

Abstract

During the past decade, the air quality has been greatly improved in China since the implementation of the “Clean Air Act”. However, haze events are still being reported in some regions of China, and the pollution mechanism remains unclear. In this study, we investigate the chemical characteristics of the pollution mechanism of the PM_2.5 composition in Suzhou from October 18 to December 15, 2020. A notable declining trend in temperature was observed from 18 to 27 November, which indicates the seasonal transition from fall to the winter season. Four representative periods were identified based on meteorological parameters and the PM_2.5 mass concentrations. The heavy pollution period had the typical characteristics of a relatively low temperature, a high relative humidity, and mass loadings of atmospheric pollutants; nitrate was the dominant contributor to the haze pollution during this period. The nitrate formation mechanism was driven by the planetary boundary layer dynamics. The potential source contribution function model (PSCF) showed that the major PM_2.5 composition originated from the northwest direction of the sampling site. The aerosol liquid water content presented increasing trends with an increasing relative humidity. The pH was the highest during the heavy pollution period, which was influenced by the aerosol liquid water content and the mass loadings of NO₃⁻, SO₄²⁻, NH₄⁺, and Cl⁻. The comprehensive analysis in this paper could improve our understanding of the nitrate pollution mechanism and environmental effects in this region.

1. Introduction

With rapid industrialization and urbanization in China, severe haze events have occurred frequently in recent decades. Atmospheric aerosols have critical effects on, for example, visibility, climate change, and human health [1,2,3,4,5,6]. For example, long-term exposure to particulate matter (PM_2.5) chemical components could increase the risk of cardiovascular disease prevalence in Chinese older adults, and more targeted control of PM_2.5 chemical components may help alleviate this disease burden [7]. Since the implementation of the “Clean Air Act” in 2013, the air quality has been greatly improved in China [8]. However, haze events have still been reported in Jing-Jin-Ji (JJJ), the Pearl River Delta (PRD), the Sichuan Basin (SCB), and other regions of China [9,10,11,12]. Thus, it is great of importance to know the potential sources and pollution mechanisms of PM.

Seasonal variations, meteorological conditions, regional transport, primary emission, and secondary formation are the key factors affecting the PM quality [13,14,15,16]. For example, the average PM₁ was the highest during winter (81.7 ± 72.4 µg m⁻³), resulting from intense emissions due to adverse meteorological conditions and secondary formation, and the lowest during the summertime (37.5 ± 31.0 µg m⁻³) due to frequent washout in Beijing [17]. At the remote Pha Din–Global Atmosphere Watch monitoring station in Vietnam, a backward trajectory analysis showed that cleaner air masses were arriving from the northeast, while more polluted periods were characterized by trajectories from the southwest direction during the sampling period [18]. Another study also demonstrated that haze events in Fujian occurred either due to calm weather conditions or the southward transport of the pollutants along the coastline from inland China [10]. Restriction on emission control is an effective way to alleviate air pollution. Although strict emission controls were implemented during the Hangzhou G20 summit, nighttime regional transport can aggravate secondary PM pollution (OA was more aged, 48.7% and 13.7% higher than that before and after the G20), and there are unfavorable synoptic situations (i.e., subsidence airflows, wind, and vertical temperature gradient) [19,20]. Furthermore, the chemical mechanism and pH value are also critical factors influencing particle formation. The formation of less and more oxidized oxygenated organic aerosol in Baoji is mainly driven by photochemical reactions but significantly influenced by aqueous-phase chemistry under the condition of a low atmospheric oxidative capacity [21]. Under heavily polluted conditions, more secondary ions accumulate in the coarse mode, leading to the acidity of the coarse-mode aerosols shifting from neutral to weakly acidic [22]. On the basis of its influencing factors, the pollution mechanism of PM can be further clarified.

To date, some field studies about the pollution mechanism have been reported around the world [23,24,25,26,27]. Nitrate was reported to be the major chemical composition in PM₁ (16.3% and 24.8% in Beijing in winter and summer), and photochemistry was the dominant pathway in winter, while thermodynamics was the dominant pathway in summer [28]. Additionally, N₂O₅ uptake in aerosols and clouds is the dominant nitrate production pathway in the wintertime in Beijing, but its formation rate is limited by the ozone under high NO_x, high PM_2.5 conditions [29]. A high relative humidity significantly contributes to the severe haze pollution. An elevated relative humidity leads to the increase in the aerosol liquid water content (ALWC), which, in turn, promotes a series of liquid-phase or surface chemical reactions. For example, the self-amplifying process of the ALWC and secondary inorganic aerosols and the aqueous-phase formation of secondary organic carbon enhance aerosol formation, contributing to air pollution and a reduction in visibility [30]. Furthermore, gaseous pollutants, such as NH₃ and HNO₃, can influence the aerosol water content and pH of aerosols, thereby affecting the formation of particulate matter [31]. A previous study showed that controlling the emissions of NH₃ would effectively reduce the ALWC and PM_2.5, while the mass loadings of SO₄²⁻ and NO₃⁻ were high in Beijing [32]. However, studies about nitrate formation and its potential influencing factors (i.e., RH, ALWC, HNO₃, and NH₃) during the haze episode in the Yangtze River Delta region are limited, making it an issue which needs to be further discussed.

Suzhou is a typical industrial city in eastern China; various air pollutants are emitted from industrial factories and motor vehicles [33]. Unfavorable weather conditions were conducive to haze formation in Suzhou, and the haze formation mechanism was different during the different haze periods in January 2013 [34]. During the past decade, studies in Suzhou have been mainly concentrated on the changes in seasonal chemical composition and the light absorption of PM_2.5. However, studies about the real-time chemical characteristics and evolution mechanism of PM_2.5 haze after the “Clean Air Act” in September 2013 are relatively rare, making it an issue which needs to be further discussed. In this study, we report the real-time measurements of the chemical composition, gaseous species, and meteorological conditions at an industrial site in eastern China from 28 October to 15 December of 2020. The characteristics of the atmospheric chemical composition and its evolution processes during the sampling period are comprehensively discussed. Meanwhile, the industrial structure, meteorological conditions, as well as the aerosol liquid water content and pH will be investigated. The results from this study could provide a better understanding of this topic to help with regional atmospheric pollution control and policy formulation.

2. Materials and Methods

2.1. Sampling Site and Instrumentation

The field campaign was carried out at the Qingjian Lake Site (31.37° N, 120.72° E), which belonged to a suburban industrial park in Suzhou, China. Suzhou is located at the center of the Yangtze River Delta region between Nanjing, Shanghai, and Hangzhou. The sampling site is surrounded by factories, residential communities, schools, and major roads. It is notable that many large high-tech enterprises, including electronics, pharmaceuticals, purification equipment, and precision instruments, are situated to the eastern and southern directions of the sampling site. Direct emissions as well as regional transport potentially affected the atmospheric quality, formation, and evolution mechanism at the suburban site [35]. The measurement was conducted from October 28 to December 15 in 2020, which covered the fall and winter seasons.

During the campaign, the inorganic components of PM_2.5 (Na⁺, NH₄⁺, K⁺, Mg²⁺, Ca²⁺, Cl⁻, NO₃⁻, and SO₄²⁻) and water-soluble gaseous species (HCl, HONO, SO₂, HNO₃, and NH₃) were measured by the Monitor for AeRosols and Gases in Ambient air (MARGA, Model ADI 2080, Metrohm Inc., Herisau, Switzerland), which is an online analyzer for semicontinuous measurements. In brief, MARGA consists of sampling and analytical boxes. Ambient air is drawn into the sampling box with a flow rate of 1 m³ h⁻¹. The sampling box comprises one wet rotating denuder (WRD) for gas sampling and one steam jet aerosol collector (SJAC) for aerosol collecting. Gaseous components were scavenged and dissolved in the liquid film (0.0035% H₂O₂) formed by the WRD, and then aerosol species were continuously collected by the SJAC, where the condensation of steam makes aerosols rapidly become water droplets. Then, the two collected liquid samples were analyzed by ion chromatographic analysis, respectively [36,37,38]. The time resolution for the MARGA is in a one-hour format. Carbonaceous aerosols, including organic carbon (OC) and elemental carbon (EC), were monitored using a Sunset OC/EC online analyzer (Model RT-41, Sunset Lab. Inc., Tigard, OR, USA) following the standard protocol [39]. Mixing ratios of gas species, such as SO₂, NO₂, CO, O₃, and PM_2.5, were collected parallelly by a series of monitors (43i, 42i, 49i, 48i, and 5030i, Thermo Fisher, Waltham, MA, USA). In addition, meteorological data, including temperature (T), relative humidity (RH), wind speed (WS), and wind direction (WD), were also monitored at the same site (WS600, Lufft, Fellbach, Germany). All of the data were averaged to a one-hour format.

2.2. Data Analysis

The aerosol liquid water content and pH were simulated using the ISORROPIA-II model by using the inorganic species and meteorological data [40]. ISORROPIA-II is an inorganic thermodynamic model, which included forward mode and reverse mode. In the forward mode, the input data included Na⁺, K⁺, Ca²⁺, Mg²⁺, SO₄²⁻, TNO₃ (TNO₃ = NO₃⁻ + HNO₃), TCl (TCl = Cl⁻ + HCl), TNH₃ (TNH₃ = NH₄⁺ + NH₃), RH, and T. In the reverse mode, the gaseous HNO₃, HCl, and NH₃ were not considered. The forward mode was used in this study because this mode considers gaseous HNO₃, HCl, and NH₃. Here, the aerosol pH is calculated as

$$\mathrm{p}\mathrm{H}=-{\mathrm{l}\mathrm{o}\mathrm{g}}_{10}{\displaystyle \frac{1000{\mathrm{H}}_{\mathrm{a}\mathrm{i}\mathrm{r}}^{+}}{\mathrm{A}\mathrm{L}\mathrm{W}\mathrm{C}}}$$

where the ALWC is the aerosol liquid water content uptake by inorganic species (μg m⁻³), and ${\mathrm{H}}_{\mathrm{a}\mathrm{i}\mathrm{r}}^{+}$ is the equilibrium particle hydronium ion concentration per volume of air (μg m⁻³).

The molar equivalent ratio of cations (CE) to anions (AE) is an important indicator to assess the balance between cations and anions, which can be calculated using the following formulas [41]:

$$\mathrm{C}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{s}={\displaystyle \frac{\left[{\mathrm{N}\mathrm{a}}^{+}\right]}{23}}+{\displaystyle \frac{\left[{\mathrm{N}\mathrm{H}}_4^{+}\right]}{18}}+{\displaystyle \frac{\left[{\mathrm{K}}^{+}\right]}{39}}+2\times {\displaystyle \frac{\left[{\mathrm{C}\mathrm{a}}^{2+}\right]}{40}}+2\times {\displaystyle \frac{\left[{\mathrm{M}\mathrm{g}}^{2+}\right]}{24}}$$

$$\mathrm{A}\mathrm{n}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{s}=2\times {\displaystyle \frac{\left[{\mathrm{S}\mathrm{O}}_4^{2-}\right]}{96}}+{\displaystyle \frac{\left[{\mathrm{N}\mathrm{O}}_3^{-}\right]}{62}}+{\displaystyle \frac{\left[{\mathrm{C}\mathrm{l}}^{-}\right]}{35.5}}$$

$$\mathrm{E}\mathrm{q}\mathrm{u}\mathrm{i}\mathrm{v}\mathrm{a}\mathrm{l}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}=\left[\mathrm{C}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{s}\right]/\left[\mathrm{A}\mathrm{n}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{s}\right]$$

where the concentrations of water-soluble ions are expressed in μg m⁻³. When the equivalent ratio of cations to anions is greater than 1, this indicates a deficiency of anions. Conversely, when the equivalent ratio is less than 1, this signifies a shortage of cations. This ratio reflects the balance of ions in the sample, which is essential for understanding the chemical composition of aerosols.

The potential source contribution function (PSCF) model was deployed to identify the potential regional source based on the HYSPLIT model [42,43]. In this study, 36 h back trajectories per hour arriving at the sampling site at a height of 500 m above sea level were used to calculate the PSCF values. The value of PSCF is defined through the following equation [43,44]:

$${PSCF}_{ij}={\displaystyle \frac{m_{ij}}{n_{ij}}}$$

where $n_{ij}$ is the total number of trajectory endpoints in the cell ijth, and $m_{ij}$ is the number of trajectory endpoints in the cell ijth for which the corresponding trajectories arrive at a receptor site when the concentration is higher than a pre-specified criterion value. In order to minimize the uncertainties of PSCF analysis, a weighing function $W\left(n_{ij}\right)$ was employed in this study [45]:

$$\begin{array}{l}W(n_{ij})\\ =\left\{\begin{array}{c} 1 \mathrm{for} \log \left(n_{ij}+1\right)/{max}_{\log \left(n_{ij}+1\right)}\ge 0.85 \\ 0.725 \mathrm{for} 0.6\le \log \left(n_{ij}+1\right)/{max}_{\log \left(n_{ij}+1\right)}<0.85 \\ 0.475 \mathrm{for} 0.35\le \log \left(n_{ij}+1\right)/{max}_{\log \left(n_{ij}+1\right)}<0.6 \\ 0.175 \mathrm{for} \log \left(n_{ij}+1\right)/{max}_{\log \left(n_{ij}+1\right)}<0.35 \end{array}\right.\end{array}$$

The bivariate polar plot was employed to evaluate potential local sources of the measured pollutant at the sampling site, considering the effects of both wind direction and wind speed [46,47]. The R statistical software (version 3.6.0, R Core Team, 2019) with the openair package was employed. In this study, pollutants, such as K⁺, Na⁺, Ca²⁺, and Mg²⁺, were measured by MARGA.

3. Results and Discussions

3.1. An Overview of the Meteorology and Gaseous Pollutants

Figure 1 illustrates the temporal variations in meteorological parameters, trace gases, and PM_2.5 mass concentrations in 2020. During the field campaign, the average temperature (T) was 13.1 ± 5.1 °C. Between 28 October and 17 November, the temperature was relatively high, averaging 17.1 ± 2.7 °C. In contrast, the mean temperature dropped to 8.1 ± 2.8 °C from 28 November to 15 December. A notable decline in ambient temperature occurred from 18 to 27 November, indicating a transition from fall to winter. The average relative humidity (RH) during the sampling period was 78.2 ± 14.6%. Wind speeds ranged from 0.3 to 3.5 m s⁻¹ with an average of 1.8 ± 0.8 m s⁻¹. The prevailing wind direction was from the east and northwest (see Figure S1). Average concentrations were 2.1 ± 0.9 ppb for SO₂, 23.3 ± 12.9 ppb for NO₂, 0.5 ± 0.2 ppm for CO, 20.9 ± 15.2 ppb for O₃, and 35.6 ± 32.2 µg m⁻³ for PM_2.5. Visibility averaged 12.9 ± 7.8 km throughout the sampling period, which showed a negative correlation with PM_2.5 (r = −0.59).

Four representative periods were identified based on meteorological parameters and PM_2.5 mass concentrations, as shown in Figure 1 and Figure 2. Two obvious moderate pollution processes were observed from October 28 to November 8 with an average PM_2.5 concentration of 32.5 ± 28.0 μg m⁻³, which was defined as the first moderate pollution period (Stage 1). This period is characterized by low RH, as well as the high T and mass loadings of O₃. The clean period (Stage 2) was observed from 9 to 17 November with an average PM_2.5 concentration of 22.0 ± 8.0 μg m⁻³. The mass loadings of SO₂, CO, and gaseous HNO₃ were the lowest (1.8 ± 0.4 ppb, 0.5 ± 0.1 ppm, and 0.59 ± 0.56 μg m⁻³, respectively) (Table S1). The mean visibility could reach 15.7 ± 6.7 km during this period. It was interesting that the T and RH both showed an increasing trend. Similar to Stage 1, the second moderate pollution period (Stage 3) was observed from November 28 to December 9 with an average PM_2.5 concentration of 40.8 ± 21.3 μg m⁻³. The mass concentrations of SO₂, NO₂, CO, and PM_2.5 were close to those in Stage 1. The period from 10 to 13 December was selected as a heavy pollution period (Stage 4) with an average PM_2.5 concentration of 106.7 ± 46.5 μg m⁻³, along with a relatively low T (8.7 ± 2.0 °C) and high RH (83.5 ± 11.4%), SO₂, NO₂, and CO concentrations (2.9 ± 1.2 ppb, 34.1 ± 12.7 ppb, and 0.9 ± 0.3 ppm, respectively). The mass loadings of gaseous HONO and HNO₃ were also the highest (3.5 ± 1.6 μg m⁻³, and 1.2 ± 1.1 μg m⁻³, respectively). O₃ concentrations during this period were the lowest (11.3 ± 15.2 ppb), which were likely influenced by the titration effect (O₃ + NO = NO₂ + O₂) at the surface. The NO concentration during Stage 4 was 13.0 ± 10.8 ppb, higher than other periods (Figure S2). In addition, the visibility during this period (3.7 ± 3.0 km) was lower than in any other periods.

3.2. Characteristics of PM_2.5 Composition

3.2.1. Major Composition

Figure 1f,g show the time series of OC and EC and the chemical composition of PM_2.5. NO₃⁻ (31.3%) was the most abundant composition in major PM_2.5 species during the whole sampling period, followed by OM (28.5%) (1.6 × OC (17.8%), [48]), SO₄²⁻ (15.6%), NH₄⁺ (14.2%), EC (4.3%), Cl⁻ (3.8%), K⁺ (0.8%), Na⁺ (0.8%), Ca²⁺ (0.6%), and Mg²⁺ (0.3%). In comparison, the mass loadings and fractions of major PM_2.5 species during the four periods are illustrated in Figure 3. During Stage 2, the mass concentrations of PM_2.5 chemical components were the lowest, while they reached the highest levels during Stage 4 (Figure 3a). Among these parameters, the most significant changes were observed in the mass loadings of NO₃⁻, SO₄^²−, NH₄⁻, Cl⁻, OM, and EC. Specifically, the mass loading of NO₃⁻ increased dramatically from 7.2 ± 5.5 μg m⁻³ during Stage 2 to 45.5 ± 18.3 μg m⁻³ during Stage 4. This substantial rise underscores a pronounced shift in nitrate concentrations over the study period. Similar trends were observed for SO₄²⁻, NH₄⁺, Cl⁻, OM, and EC, highlighting a marked increase in these chemical components as the stages progressed. A 36 h backward trajectory during Stage 4 illustrated that the air mass shifted from the east to the west and northwest, accompanied by the temporal variations of PM_2.5 and its major chemical components (Figure S3a). Additional strong evidence was provided by the fire spots detected in the western and northwestern directions of Suzhou (11 December 2020, Figure S3b), which is consistent with the trend of the 36 h backward trajectories (eastern direction excluded). The mass fractions of the chemical composition varied across the different stages (Figure 3b). Interestingly, the highest contribution of OM (34.1%) and a relatively low mass fraction of NO₃⁻ (26.3%) were observed in Stage 2. In contrast, the lowest concentration of OM (19.1%) and the highest NO₃^- concentration (41.3%) were observed in Stage 4. The pollution mechanism of NO₃⁻ will be further discussed in Section 3.3. The fraction differences of SO₄²⁻, NH₄⁺, and Cl⁻ during the four periods were not significant. As shown in Table S2, NH₄⁺ correlated strongly with SO₄²⁻, implying that they were likely present in (NH₄)₂SO₄ and NH₄HSO₄. In addition, Cl⁻ correlated well with NH₄⁺, indicating that the mass loading of NH₄Cl in the atmosphere was high.

3.2.2. Metal Ions

Metal ions are key components in the atmosphere. A previous study showed that the non-photosensitive metal ions represented by Fe³⁺, Cu²⁺, Al³⁺, and Zn²⁺ have different effects on the radiative forcing effect and photochemical reaction process of BrC aerosols [49]. As shown in Figure 3 and Table S2, the mass concentrations and fractions of metal ions (K⁺: 0.23 ± 0.19 μg m⁻³, 1.1%; Na⁺: 0.27 ± 0.12 μg m⁻³, 1.2%; Ca²⁺: 0.15 ± 0.11 μg m⁻³, 0.8%; Mg²⁺: 0.10 ± 0.09 μg m⁻³, 0.5%) during Stage 1 were significantly higher than in any other periods. The metal ions during Stage 4 were the lowest (Table S2), implying that regional transport no longer contributed to increased metal ion concentrations. This result was different from those for SNA, Cl⁻, OM, and EC. The metal ions originated from a local source in Suzhou. A source analysis of the polar plot in metal ions was conducted in light of wind directions and speeds. The bivariate polar plots showed that high mass loadings of the metal ion Na⁺ during Stage 1 mainly originated from the western direction, influenced by two moderate pollution episodes (the 30 October–2 November and 6–8 November periods, respectively). K⁺, Ca²⁺, and Mg²⁺ primarily originated from the southeast (Figure S4).

3.2.3. Source Analysis

Based on the results of the trajectory analysis, the PSCF model was deployed to explore the likely regional sources of major chemical components in PM_2.5 during the whole observation period, including SO₄²⁻, NO₃⁻, NH₄⁺, Cl⁻, OM, and EC, as illustrated in Figure 4. Generally, high concentrations of the major composition (SO₄²⁻, NO₃⁻, NH₄⁺, Cl⁻, OM, and EC) originated from the northwest direction of the sampling site. According to a previous study by Li et al. (2020), the pollution sources mainly included the dust-related, more aged, and less aged PM_2.5 [50]. Meanwhile, our results were consistent with the previous studies reported in the Pha Din–Global Atmosphere Watch monitoring station (Vietnam) and Fujian (China) [10,18].

3.3. Analysis of Ion Balance

The ion balance was used to verify the reliability of the measured water-soluble ions as it is an effective method to estimate aerosol pH. The predicted NH₄⁺ values were calculated according to the following equation: NH₄⁺ predicted = 18 × (2 × SO₄²⁻/96 + NO₃⁻/62 + Cl⁻/35.5) [51]. The molar ratio of inorganic anions (sulfate, nitrate, and chloride) to cations (ammonium) is 1.10 (r² =0.99, Figure S5). It can be concluded that PM_2.5 was acidic throughout the study.

SO₄²⁻, NO₃⁻, and Cl⁻ were the critical components in the atmosphere. As shown in Figure 5a, the molar equivalent concentrations of measured ammonium had good correlations with the sum of nitrate, sulfate, and chloride (r²: 0.99; slope: 1.05). This result verifies that inorganic anions (nitrate, sulfate, and chloride) mostly exist in the form of NH₄NO₃, (NH₄)₂SO₄, and NH₄Cl. The linear correlation between [NH₄⁺]/[SO₄²⁻] and [NO₃⁻]/[SO₄²⁻] was often used to explore formation with different loadings of SO₄²⁻ [52]. When [NH₄⁺]/[SO₄²⁻] ≤ 1.5, nitrate formation was associated with crustal elements rather than ammonium; when [NH₄⁺]/[SO₄²⁻] > 1.5, the homogeneous gas-phase reactions between NH₃ and HNO₃ became the major pathway for atmospheric ammonia to form NH₄NO₃ [52,53]. As shown in Figure 5b, the molar ratios of NO₃⁻/SO₄²⁻ exceeded 1.5 (ammonium-rich region) during the sampling period, which correlated well with NH₄⁺/SO₄²⁻ (r² = 0.97), indicating that NH₄NO₃ was formed after the formation of (NH₄)₃H(SO₄)₂. Furthermore, we further discussed the relationship between NH₄⁺ and NO₃⁻. The molar concentrations of excess ammonium (NH₄⁺/SO₄²⁻ − 1.5)/SO₄²⁻ also correlated very strongly with nitrate (r² = 0.99; slope = 0.92; Figure 5c). This result proves that NH₄NO₃ formed when sulfuric acid was fully neutralized by ammonia, as found in a previous study [28]. The molar equivalent ratios of cations (CE) to anions (AE) were also calculated. As shown in Figure 5d, the cations correlated well with anions (r² = 0.99). The slope of CE vs. AE is 0.93, implying the shortage of the cations in Suzhou.

3.4. Formation Mechanism of Nitrate

Figure 3a shows that the mass loading of NO₃⁻ during Stage 4 was the highest. In order to investigate the transformation of NO₂ to NO₃^-, the nitrogen oxidation ratio (NOR) was defined as n[NO₃⁻]/(n[NO₃⁻] + n[NO₂]) [54]. The NOR values during the four periods were 0.16 ± 0.13, 0.11 ± 0.08, 0.22 ± 0.12, and 0.32 ± 0.12, respectively (Figure S6). The NOR trend was consistent with the mass loadings of PM_2.5 (r = 0.63). Previous studies have shown that low temperatures and high RH during the wintertime are generally favorable for the formation of SNA (the sum of sulfate, nitrate, and ammonium) aerosols, and the transformation of HNO₃ into a particle phase is generally enhanced at high RH and low temperature conditions [55,56]. During Stage 4, the low T, high RH, and mass loading of HNO₃, as well as the regional transport, contributed to the formation of NO₃^- haze. We further computed the theoretical dissociation constants of NH₄NO₃ (K_p) by assuming that the system was at thermodynamic equilibrium [57]. The mechanism mainly included the thermodynamics driven, photochemical driven, and planetary boundary layer (PBL) driven [28]. Figure 6 shows the average diurnal changes in T, RH, NO₂, O₃, HONO, NH₃, NO₃⁻, and K_p. The nitrate diurnal pattern during Stage 4 correlated well with the T (r = 0.83) and O₃ (r = 0.89), but it had an opposite result with K_p (r = −0.67) and RH (r = −0.71). Higher temperatures favor gas-phase reactions involving O₃ and NO_x that lead to the production of HNO₃, which then react with NH₃ to form NH₄NO₃. Sunlight drove the photochemical reactions between volatile organic compounds (VOCs) and NO_x to produce O₃ and then participate in further reactions that produce nitric acid. A higher relative humidity generally increased the aerosol liquid water content, facilitating heterogeneous reactions in the formation of nitrate particles. K_p decreased with decreasing temperature, favoring the condensation of NH₃ and HNO₃ into particulate NH₄NO₃ under cold conditions. Thus, we propose that this nitrate formation mechanism is driven by planetary boundary layer dynamics. This is consistent with a previous study reported in Lanzhou in the summertime [28]. In addition, more research is needed to clarify the formation mechanism of nitrate. Lower temperatures slow down the dissipation of pollutants by reducing vertical mixing in the PBL and shift the equilibrium of the gas–particle phase partitioning of nitrogen oxides (NO_x); a high RH facilitates the formation of aerosol liquid water, and the presence of liquid water on aerosols promotes reactions between HNO₃ and NH₃; and sulfate aerosols enhance the water-absorbing capacity to influence nitrate formation and can compete for ammonia to influence the distribution of ammonium between sulfate and nitrate aerosols. For example, sulfuric acid is formed from the oxidation of SO₂ via gaseous and multiphase reactions, and then it is subsequently fully or partly neutralized by gaseous NH₃ taken up on particles to further influence the formation of nitrate [58]. As shown in Table S3, NO₃⁻ correlated well with SO₄²⁻, NH₄⁺, OM, and EC during the field campaign (r = 0.85, 0.98, 0.76, and 0.81, respectively).

3.5. Aerosol Liquid Water Content and pH

The aerosol liquid water content plays a critical role in the formation and evolution process of atmospheric aerosol. The inorganic compound conversion was enhanced via aqueous phase chemistry on moist particles, and heterogeneous reactions may play a more important role in aggravating haze events [59,60]. As shown in Figure 7a, the aerosol liquid water content presented increasing trends with an increasing relative humidity (RH). This study further found that when the RH remains constant, the aerosol liquid water content and the mass loadings of PM_2.5 are high. This is consistent with a previous study reported in Hohhot, where Xie et al. (2023) summarized that the inorganic compound conversion was enhanced via aqueous-phase chemistry on moist particles owing to sulfuric acid and nitric acid being formed in the gas phase and then partitioning to the particles, which then leads to more water absorption by the particles [61]. In addition, the potential role of organics in the water content cannot be ignored. For example, the ALWC contributed by organics accounted for 30% ± 22% of the total ALWC in Beijing in winter, which plays an important role in the formation of secondary aerosols through multiphase reactions at the initial stage of a heavy haze episode [62].

Studies on the acidity and alkalinity of aerosols are critical for understanding the aerosol formation mechanisms and the multiphase formation pathways of semi-volatile components, as well as for providing effective strategies to control PM_2.5 pollution. The pH value during the whole sampling period was 3.5 ± 0.3, which was lower than that in Beijing [22] and some areas in Henan Province, China [63]. Figure 7b presented the pH values across the four stages. The highest pH was observed during Stage 4, while the smallest pH was observed at Stages 1 and 2. This trend was consistent with the ALWC as well as the mass loadings of NO₃⁻, SO₄²⁻, NH₄⁺, and Cl⁻.

4. Conclusions

In this study, we conducted continuous measurements of the chemical composition, gaseous species, and meteorological conditions in suburban Suzhou during the cold season of 2020. The mean temperature during the sampling period was 13.1 ± 5.1 °C with a relative humidity of 78.2 ± 14.6%. The predominant wind direction ranged from the east to the northwest. It is noteworthy that visibility was inversely correlated with PM_2.5 mass concentrations, indicating that higher pollution levels led to reduced visibility. Among the major PM_2.5 species, nitrate (NO₃⁻) was the most abundant, followed by organic matter (OM), ammonium (NH₄⁺), sulfate (SO₄²⁻), elemental carbon (EC), chloride (Cl⁻), potassium (K⁺), sodium (Na⁺), calcium (Ca²⁺), and magnesium (Mg²⁺). Based on the meteorological parameters and PM_2.5 mass concentrations, four distinct periods were identified. The period of heavy pollution was characterized by lower temperatures, higher humidity, and elevated concentrations of air pollutants. During this stage, nitrate dominated haze pollution, influenced by factors such as HNO₃ mass loadings, regional transport, and the height of the planetary boundary layer. The potential source contribution function (PSCF) model revealed that the major PM_2.5 components (dust-related, more aged, and less aged sources) originated primarily from the northwest of the sampling site. Additionally, the aerosol liquid water content showed an increasing trend with a rising relative humidity, corresponding with increased PM_2.5 mass loadings. The pH trend was consistent with aerosol liquid content as well as the mass loadings of major components in PM_2.5.

However, it is important to acknowledge several limitations that may influence the interpretation of these results. First, the data were collected over a limited period during the cold season, and comparative observations and analyses during the warmer months are needed. Furthermore, the continuous measurements were conducted at a single suburban site in Suzhou, and we should increase the number of observation stations to enhance the robustness of the findings. Finally, an analysis of pollution mechanisms needs to be combined with the air quality model. Overall, this study enhances our understanding of haze formation mechanisms and pollutant sources, providing valuable insights to inform effective pollution control strategies for governments and policymakers.