Fine Particle pH and its Impact on PM 2.5 Control in a Megacity of Central China

Fine particle (PM 2.5 ) acidity greatly affects the formation of secondary aerosol, and the drivers of PM 2.5 pH variation are vital in understanding its effects. Moderate PM 2.5 acidity was found in Wuhan, a megacity of Central China, wherein 80% of PM 2.5 hold pH values ranging from 2 – 4. Total ammonia (NH x ) and sulfate contributed 79.1 – 93.7% to pH changes in spring and winter, while relative humidity was the largest contributor (33.7 – 36.3%) in summer and fall. By sensitivity simulations, PM 2.5 remained acidic with pH changes less than 0.5 units in spring, summer and fall in foreseeable future even when the concentration varied by two orders of magnitude. While pH changes in winter were three times those in the other seasons, and NH x changes was suggested as the indicator of PM 2.5 pH variation in winter. Furthermore, the impact of pH on PM 2.5 responses to emissions control was evaluated. The pH has opposite influences on the effect of SO 2 and NO x control in reducing PM 2.5 , the former being more effective at low pH and the latter being more effective at high pH due to ammonium and nitrate gas-particle partitioning. The effect of NH x control on PM 2.5 reduction is nonlinearly affected by pH. It is directly effective at low pH, but more ammonia control is required before achieving effectiveness at high pH. For the current particulate pH of 3 – 4 in Wuhan, both SO 2 and NO x control are beneficial for PM 2.5 reduction. However, NH x control is less effective before it is reduced by approximately 20%.

The reduction of atmospheric particulate matter has always been a concern, especially after frequent haze pollution events occurring in China (Zhou et al., 2022;An et al., 2019;Bai et al., 2019;Tian et al., 2018).Many studies have explored the responses of PM 2.5 to precursor emissions control, such as SO 2 , NO x , and NH 3 (Pinder et al., 2008;Blanchard et al., 2000;Ansari and Pandis, 1998).Recently, the sensitivity of particulate nitrate to ammonia or nitric acid has become a research hotspot (Xu et al., 2019;Guo et al., 2018;Wu et al., 2016).Nitrate in PM 2.5 could be effectively reduced by ammonia emission control in North China (An et al., 2019;Liu et al., 2019), but unsensitive in other regions such as South China (Liu et al., 2019), Sichuan Basin (Liu et al., 2019), and East China (Xu et al., 2020).Moreover, studies have also focused on the evaluation methods and indicators of the effectiveness of PM 2.5 reduction in emission control (Nenes et al., 2020;Zheng et al., 2019;Xu et al., 2019;Wu et al., 2016).However, the effect of particulate pH on the sensitivity of PM 2.5 to precursor emissions have not been considered widely.This may lead to the ineffectiveness of emission control measures for particles reduction.
Based on the hourly observations of water-soluble inorganic ions in PM 2.5 and precursor gases in 2014 and 2018 in Wuhan, the driving factors of particulate pH changes during this period were identified and their contributions were quantified using a thermodynamic model.The variations of future PM 2.5 pH for each season were predicted.Besides, the effect of particulate pH on the response of PM 2.5 to precursor reductions was explored.Results here can deepen the understanding of aerosol acidity properties and provide a reference for fine particle reduction policy making.

Observations
The observation site (114.36°N,30.53°E) is approximately 20 m above ground and is located in a commercial/residential mixed area without obvious industrial emission sources (Fig. 1).Watersoluble ions (WSI), including NH 4 + , K + , Ca 2+ , Na + , Mg 2+ , SO 4 2-, NO 3 -, and Cl -in PM 2.5 , and atmospheric HNO 3 , NH 3 , and HCl were synchronously observed hourly using an online ions analyzer (Marga ADI 2080).The details about the instrument can be found in Text S1 and previous research (Zheng et al., 2019).Continuous monitoring was conducted from January to December in 2014 and 2018, except for data missing due to equipment maintenance.Hourly temperature (Temp) and relative humidity (RH) were obtained from the local observatory (Text S1) (Zheng et al., 2019).At this observation site, SO 2 and NO 2 were monitored hourly by ultraviolet fluorescence and chemiluminescence online monitoring equipment from 2013, respectively.
In Section 3.2, a series of sensitivity simulations were conducted to explore the key factors and their contribution in each season.The chemical and meteorological factors, including SO 4 2-, NH x (NH 3 + NH 4 + ), TNO 3 (NO 3 -+ HNO 3 ), TCl (Cl -+ HCl), RH, and temperature were evaluated.Each factor was subjected to interannual replacement.For instance, firstly based on 2014 observation,

Fine Particle pH in 2014 and 2018
The annual average PM 2.5 concentration significantly decreased by 45.7% in Wuhan from 2014 (92.2 ± 57.6 µg m -3 ) to 2018 (50.4 ± 33.6 µg m -3 ), while both of them still exceeded the PM 2.5 annual value of the secondary air quality standard in China (35 µg m -3 ).The proportion of SNA (sulfate, nitrate and ammonium) in PM 2.5 increased from 34% in 2014 to 47% in 2018, indicating the gradual dominance of inorganic components in fine particles, which is consistent with other studies (Li et al., 2019).The strong relationship among SO 4 2-, NO 3 -, and NH 4 + in 2014 (r 2 (square of correlation coefficient) of SO 4 2--NO 3 -= 0.73, r 2 of NH 4 + -NO 3 -= 0.91, r 2 of SO 4 2--NH 4 + = 0.86) (Table S1) exhibited partial homology (e.g., combustion emission).While the correlation among SO 4 2-, NO 3 -, and NH 4 + decreased in 2018 (r 2 values of SO 4 2 -NO 3 -, NH 4 + -NO 3 -, and SO 4 2--NH 4 + were 0.40, 0.87 and 0.59, respectively) (Table S1).The variation of dominant emission sources was identified from the changes in the [NO 3 -]/[SO 4 2-] mass ratio (Xing et al., 2021), which increased from 0.85 in 2014 to 1.17 in 2018, implying the decrease in stationary source (mainly coal combustion) or increase in mobile source (mainly traffic emissions).Compared to SNA, the contributions of non-volatile cations (NVCs, including K + , Na + , Ca 2+ , and Mg 2+ ) to PM 2.5 were only 3.0% and 4.0% in 2014 and 2018, respectively.1), which has the greatest effect on pH among NVCs and was mainly originated from sand or road dust and construction activities (Hegde et al., 2016).Considering the weak effect of NVCs on the aerosol acidity in Wuhan and the randomness of sand-dust event, the NVCs effect was subtracted from fine particle pH and AWC in the subsequent discussion.
Fig. 3 showed the frequency proportions in different particle pH ranges during the four seasons.In both 2014 and 2018, the fine particle pH distribution was in accordance with the norm, and both of the peaks were at pH values of 3-3.5 with frequencies of 27.4% and 24.6%, respectively.In Wuhan, over 80% of the fine particulate was at a pH of 2-4.Higher pH mainly appeared in winter (3.5-4.5) and lower particulate pH was observed in summer (below 3.5) (Fig. 3).From the pH frequency distribution between 2014 and 2018, the particulate pH in each season dropped in Wuhan.Fine particle pH also changed hourly, with an early morning peak at approximately 07:00 (local time, pH up to 3.67 in 2014 and 3.27 in 2018), and a subsequent decrease during the daytime, reaching a minimum in the afternoon at around 16:00 (local time) (Fig. S1), which was similar to the results of Guo et al. (2015).There was a great consistency in the diurnal variation trend between particulate pH and AWC both in 2014 and 2018 (Fig. S1).The following section explored the driving factors of pH variation in Wuhan during 2014-2018.

Driving Factors of pH Decline during 2014-2018
The annual average concentrations of atmospheric SO 2 and NO x (Fig. S2) obviously decreased from 2014 to 2018 due to the implementation of control measures (Jin et al., 2016), while the particle pH still continued to decline.There were other driving factors affecting the pH variation during 2014-2018.
Fig. 4 showed the sensitivity results of the driving factors affecting the pH variation in Wuhan between 2014 and 2018 (please refer to Section 2.2 for the simulation details).NH x mainly contributed to the pH decline during 2014-2018 both in spring and winter.RH was the most important factor leading to the pH decreases in summer and fall.The contribution of NH x to pH variation were 44.2%-48.9% in spring and winter, and the RH contribution to pH changes was 33.7-36.3% in summer and fall (Fig. 4, pie chart).This differed from the North China Plain, where SO 4 2-+ Ca 2+ , NH x + RH, NH x + Temp, and SO 4 2-+ NH x were the driving factors of pH corresponding to spring, summer, fall and winter, respectively (Ding et al., 2019).Compared with North China, the higher RH (NBS, 2019) and relatively lower ammonia emission (Huang et al., 2012) in Central China resulted in the higher sensitivity of RH and NH x to particulate pH (Liu et al., 2019;Ding et al., 2019).Sulfate decline in Wuhan raised the particulate pH as expected (Fu et al., 2015), wherein the pH increased by 0.07-1.0units corresponding to the 12-63% decrease in sulfate.Similar to previous research (Ding et al., 2019), the sensitivity of fine particulate pH to TNO 3 variation was less than that of SO 4 2-, owing to the low volatility of the latter.In spring, summer, and winter, the particulate pH almost unchanged when TNO 3 decreased (7-12%).The pH in fall slightly decreased by 0.05 units with 61% TNO 3 increase.The insensitivity of pH to TNO 3 reduction partly resulted from the decrease in AWC (Fig. S3) and the gas-particle partitioning.Moreover, the removal of nitrate released ammonium to the gas phase, leading to hydroxyl decline and buffering changes in PM 2.5 pH (Blanchard et al., 2000;Dennis et al., 2008).NH x decreased by 22.3% and 44.1% in spring and winter, respectively, resulting in pH decrease by 0.2 and 1.0 units, respectively.The difference of pH decline between the two seasons was partly due to the discrepancy in AWC changes (Fig. S3).Lower particle phase distribution in spring (below 0.4 in both 2014 and 2018) compared to winter (about 0.7 in both 2014 and 2018) contributed to the low AWC changes in former.NH x reduction in the season with higher particle phase distribution caused a greater decline in NH 4 + water uptake (Guo et al., 2018).The negligible effect of NH x on pH changes in summer and fall was mainly due to the small NH x concentration variation and relatively low NH x particle phase distribution (Table 1).
Unlike SO 4 2-and TNO 3 , decreasing TCl in summer and fall reduced the particle pH, partly due to the decline in AWC (Fig. S3).TCl showed a negligible impact on the pH changes in winter, even though the AWC in winter also obviously decreased.Compared to the other seasons, TCl in winter was mostly distributed in the particle phase, with ε(Cl -) (particulate Cl -fraction, ε(Cl -) = Cl -/TCl) of 0.95 in 2014 (Table 1).Cl -in particle phase was in the form of NH 4 Cl, owing to excessive ammonia in the observation city (Zheng et al., 2019).Reduction of TCl released associated ammonium to the gas phase, buffering the particulate pH changes.
RH hold different impacts on the fine particulate pH in different seasons.Decreased RH reduced the PM 2.5 pH in summer and fall, but nearly exhibited no effect in spring.Moreover, increased RH in winter also presented a negligible impact on the pH changes.The effect of RH on fine particulate pH was determined by the competition of RH's impact on protons and AWC (Ding et al., 2019 ACP), including the process of water uptake, gas-to-particle conversion and liquid phase reaction (Seinfeld and Pandis, 2016;Guo et al., 2018).The Temp in summer and autumn in 2018 increased by 1.5-1.7°Ccompared with 2014, contributing to the decline in particulate pH.High Temp promote the conversion of semi-volatile components, such as ammonium nitrate and ammonium chloride, to the gas phase (Seinfeld and Pandis, 2016).Additionally, high Temp could also reduce the AWC (Fig. S3), further leading to the decrease in PM 2.5 pH (Guo et al., 2015).

Future pH
Based on the most significant positive and negative drivers of particulate pH changes, we expanded the ranges of NH x and SO 4 2-in spring and winter, and RH and SO 4 2-in summer and fall, for more sensitivity analyses by ISORROPIA-II.Total ammonia and sulfate were independently varied in steps of 0.1 µg m -3 in spring and winter.In summer and fall, RH and sulfate independently changed in steps of 1% and 0.1 µg m -3 , respectively.Other inputs were under the seasonal average conditions.The corresponding prediction results were shown in Fig. 5.
It showed that PM 2.5 remained acidic even with the significant reduction of sulfate (from 35 to 0.1 µg m -3 ), in consistent with Weber et al. (2016)'s study in the U.S.Moreover, the pH variations in summer and fall were smaller than those in spring and winter, which agreed well with the above section.
According to the changes during 2014-2018, the sulfate concentration decreased in each season, and all values were lower than 10 µg m -3 in 2018.In the springs of foreseeable future, within a sulfate concentration of 10 µg m -3 , the particulate pH will still be above 2.5, even if NH x is reduced by half from 2018.Furthermore, assuming an increase of sulfate in future spring, although this probability is low due to the continuous SO 2 control policy in China, the particulate pH would still be close to 2 (1.83) with a 50% increase in sulfate and a 50% NH x reduction.Additionally, assuming that the NH x increases by half in the future and the sulfate changes by +50%, 0% and -50%, the changes in pH are +0.19,+0.47, and +0.62, respectively.As mentioned in last section, sulfate and NH x were the most important factors affecting the pH in spring.While the Fig. 5. Sensitivity of the pH to total ammonia (NH x ) and sulfate (SO 4 2-) concentrations in spring and winter, and that to RH and SO 4 2-in summer and fall.particulate pH in future spring is generally less affected by sulfate and NH x .This shows that in the foreseeable future, the pH of aerosols in Wuhan will not change considerably (within 0.5) in spring.
In winter, at the sulfate level observed in 2018, the pH decreased by 1.7 units when NH x decreased by half.Even if NH x drops by 20%, the pH changes will still be more than 0.5 units.However, if NH x remains at its 2018 level, the pH only increases by 0.29 units when the sulfate decreases by half.The pH decreased by 1.28 units when both sulfate and NH x decreased by half.Clearly, the particulate pH was sensitive to NH x changes in winter, and NH x variation might be an indicator of fine particle pH in winter.In the foreseeable future, one may directly deduce the decline in wintertime particulate pH from the reducing NH x in Wuhan.
In fall, the pH value decreased by 0.7units when the RH decreased to 30%, and it increased by 0.27 units when the RH increased to 95%.When RH was at the 2018 annual value, the pH increased and decreased by 0.3 and 0.46 units when sulfate was halved and increased by half, respectively.The results showed that although sulfate and RH were the most important components and meteorological parameters affecting the particulate pH in fall, fine particle pH in future fall was less affected by them.In the foreseeable future, the particulate pH in this megacity of Central China will not change significantly (within 0.5 units), similar to that in spring.
In summer, Fig. 5 showed that the changes in RH could lead to significant changes in particulate pH.The pH decreased by 1.05 units when RH dropped to 30%, and it increased 0.63 units when RH increased to 95%.A decrease and increase by half in sulfate resulted in a pH increase of 0.15 and a decrease of 0.22 units, respectively.Compared to RH, the effect of sulfate on summertime PM 2.5 pH was less than that in the former.The pH changed by 0.2-0.3 units when the summertime RH changed by 10%.This implies that RH can be used as an indicator of fine particle pH in summer.However, considering the modest interannual variation of atmospheric RH, the future summer pH will not deviate considerably from the current levels.For the diurnal and hourly PM 2.5 variations in summer, the RH indicator might be more useful.

Effect of pH on PM 2.5
Here, we assessed the effect of pH on PM 2.5 response to SO 2 , NO x , and NH 3 control in Wuhan using a thermodynamic model.The observation data were divided into four groups according to particulate pH: pH > 4, pH 3-4, pH 2-3 and pH < 2. Changes in the SNA were assessed when the SO 4 2-, TNO 3 , and NH x were changed in each pH group, representing the control of SO 2 , NO x , and NH 3 emissions, respectively (Guo et al., 2018).Each of them was individually changed in steps of 20%, while the other inputs remained constant.
The effect of SO 4 2-reduction on SNA changes at different pH groups in Fig. 6 indicated that SO 2 control was more beneficial to SNA reduction at lower particulate pH.This was mainly contributed from the increasing NH 4 + decrease owing to relatively higher ammonia particle phase partitioning at low pH.The loss of associated NH 4 + was attributed to both the decrease in sulfate and the volatilization caused by reduced AWC (Guo et al., 2018).This implied that the efficiency of reducing SNA by SO 2 control in winter was lower than that in other seasons.It also showed that the particulate pH changed slightly (within 0.5 units) when SO 4 2-changed by 80% (Fig. S4), partly due to buffering by ammonia gas-particle partitioning (Guo et al., 2017;Weber et al., 2016).
In the case of NH x reduction, at low pH (nearly pH < 3, Fig. 6), NH x reduction was directly effective for the decline in SNA.At higher particulate pH, NH x control was not immediately effective for Fig. 6.The effect of fine particle pH on the SNA response to sulfate, total nitrate and total ammonia changes.SNA reduction, and more NH x reduction was required.For instance, considering a pH greater than 4, it could be seen that a 20% reduction in NH x had little effect on the SNA concentration.While the reduction of SNA was almost linear with NH x control when the NH x reduction reduced by more than 40%.At low pH, x reduction resulted in more NO 3 -shifting to the gas phase, partly due to the decrease in ε(NO 3 -).At high pH, TNO 3 remained in the particle phase at 40% reduction of NH x .However, once the partitioning between NO 3 -and HNO 3 was noticeably toward the gas phase due to the decline in pH and AWC since more NH x reduction, NO 3 -sharply decreased (Fig. S5).Simultaneously, NH 4 + also rapidly declined as the ε(NH 4 + ) value close to 1.
Fine particle pH is an objective condition that cannot be ignored when implementing strategies to reduce inorganic fine particles.For current observation with a pH of 3-4, both SO 2 and NO x control are effective for fine particle reduction.However, NH x control is less effective before approximately 20% NH x reduction in Wuhan, which almost consistent with previous studies (Zheng et al., 2019) suggesting a 25% effectiveness critical point of ammonia control on PM 2.5 .

CONCLUSIONS
Using hourly chemical composition observation data collected in a megacity of Central China, the driving factors of fine particle pH changes between 2014 and 2018 were analyzed.Moderate aerosol acidity was observed in Wuhan, with mean pH values of 3.63 and 3.38 in 2014 and 2018, respectively, showing a slight decline.In both 2014 and 2018, over 80% of particulate pH was at 2-4, with the largest frequency proportion occurring at a pH of 3-3.5.NH x mainly contributed to the pH decline during 2014-2018 in both spring and winter.RH was the most important factor leading to a decrease of pH in summer and fall.Further evaluation of the contribution of each factor to particulate pH changes showed that the contribution of NH x + SO 4 2-was 79.1-93.7% in spring and winter, and RH was the largest contributor (33.7-36.3%) in summer and fall.
Sensitivities simulations of exploring the future particulate pH were conducted by expanding the concentration ranges of driving factors in each season.PM 2.5 will remain acidic even when concentrations varied by two orders of magnitude.The interannual pH fluctuation in future spring, summer and fall (within 0.5 units in the foreseeable future) is less than that in winter.NH x can be suggested as an indicator of the changes of PM 2.5 pH in winter.The decline of wintertime particulate pH can be inferred from the decrease of NH x in Wuhan.
Owing to higher NH x particle phase partitioning at low pH leading to a higher NH 4 + decrease, SO 2 control is more beneficial to SNA reduction at lower particulate pH.In Contrast, NO x control is more effective for SNA reduction at higher particulate pH owing to high ε(NO 3 -) (close to 1).NH x reduction is directly effective for SNA decline at low pH (nearly pH < 3).At higher particulate pH, more NH x reduction will be required before effectiveness is achieved.For the observation city with a pH of 3-4, both SO 2 and NO x control are effective for fine particle reduction, while NH x control is less effectiveness before approximately 20% NH x reduction.

Fig. 1 .
Fig. 1.Location of the observation site in Wuhan, a megacity of Central China.

Fig. 3 .
Fig. 3. Frequency proportions of different pH ranges in four seasons (the number of data n = 4173 in 2014 and n = 5123 in 2018).