Diurnal Variations in High Time-Resolved Molecular Distributions and Formation Mechanisms of Biogenic Secondary Organic Aerosols at Mt. Huang, East China

The molecular characteristics and formation mechanism of biogenic secondary organic aerosols (BSOAs) in the forested atmosphere are poorly known. Here, we report the temporal variations in and formation processes of BSOA tracers derived from isoprene, monoterpenes, and β caryophyllene in PM2.5 samples collected at the foot of Mt. Huang (483 m a. s. l) in East China during the summer of 2019 with a 3 h time resolution. The concentrations of nearly all of the detected species, including organic carbon (OC), elemental carbon (EC), levoglucosan, and SIA (sum of SO42−, NO3−, and NH4+), were higher at night (19:00–7:00 of the next day) than in the daytime (7:00–19:00). In addition, air pollutants that accumulated by the dynamic transport of the mountain breeze at night were also a crucial reason for the higher BSOA tracers. Most of the BSOA tracers exhibited higher concentrations at night than in the daytime and peaked at 1:00 to 4:00 or 4:00 to 7:00. Those BSOA tracers presented strong correlations with O3 in the daytime rather than at night, indicating that BSOAs in the daytime were primarily derived from the photo-oxidation of BVOCs with O3. The close correlations of BSOA tracers with SO42− and particle acidity (pHis) suggest that BSOAs were primarily derived from the acid-catalyzed aqueous-phase oxidation. Considering the higher relative humidity and LWC concentration at night, the promoted aqueous oxidation was the essential reason for the higher concentrations of BSOA tracers at night. Moreover, levoglucosan exhibited a robust correlation with BSOA tracers, especially β-caryophyllinic acid, suggesting that biomass burning from long-distance transport exerted a significant impact on BSOA formation. Based on a tracer-based method, the estimated concentrations of secondary organic carbon (SOC) derived from isoprene, monoterpenes, and β caryophyllene at night (0.90 ± 0.57 µgC m−3) were higher than those (0.53 ± 0.34 µgC m−3) in the daytime, accounting for 14.5 ± 8.5% and 12.2 ± 5.0% of OC, respectively. Our results reveal that the BSOA formation at the foot of Mt. Huang was promoted by the mountain-valley breezes and anthropogenic pollutants from long-range transport.


Introduction
Secondary organic aerosols (SOAs) are important components of atmospheric fine particulate matter (PM 2.5 ), which have adverse effects on global climate change, atmo-Molecules 2023, 28 spheric visibility, cloud condensation, and human health [1][2][3]. The homogeneous and heterogeneous reactions of volatile organic compounds (VOCs) are essential in the formation of SOAs [4,5]. It is known that vegetation can release a large quantity of biogenic VOCs (BVOCs) into the atmosphere, including monoterpenes (15%), isoprene (50%), and sesquiterpenes (3%) [6,7]. BVOC emissions are estimated to be an order of magnitude higher than anthropogenic VOCs (AVOCs) on the global scale, contributing significantly to the increase in SOA concentration levels [8]. Therefore, biogenic secondary organic aerosols (BSOAs) have attracted much more attention in recent years and have become the frontier and hot topic of air pollution research. However, the formation and transformation mechanisms of BSOAs are still less well known to date. Investigations on the characteristics of BSOA tracers derived from isoprene, monoterpene, and β-caryophyllene have been conducted in various environments, such as rural [9,10], urban [11], mountainous [12][13][14], and marine [15] regions. There have been relatively few studies on BSOAs in forested highland regions [16], especially with a high time resolution. Field observations about BSOAs at a high time resolution are helpful to understand the evolution processes, formation mechanisms, and controlling factors of BSOAs. Mochizuki et al. [17] pointed out that the photochemical oxidation of BVOCs such as α-pinene and isoprene with O 3 was the essential source for BSOAs in the forest ecosystem with a 3 h time resolution. Li et al. [18] carried out an observation campaign with a 3 h interval at a rural site in the North China Plain and reported that regional biomass burning contributed significantly to the terpene-derived SOA production. A few studies demonstrated that mountain-valley breezes and anthropogenic pollutants (e.g., biomass burning) from long-distance transport played an important role in the BSOA formation at the summit of mountains [12,14,19,20]. However, there is little information about the temporal variations, evolution mechanisms, and determining factors of BSOAs in mountainous forested regions at a high time resolution, especially in East China.
Compared with the plain areas, mountains provide a unique environment for investigating BSOAs due to the stronger solar radiation and higher relative humidity in high-altitude areas [13]. Air pollution is still a primary environmental issue in East China, despite a continuous improvement in air quality due to multiple regulatory policies [21,22]. Mt. Huang, the highest mountain in Eastern China (1840 m, a. s. l), provides an ideal sampling site for investing the molecular characteristics and formation mechanisms of BSOAs in PM 2.5 samples in the background region of Eastern China. To the best of our knowledge, research on the characteristics and evolution mechanisms of BSOAs at Mt. Huang has not been carried out. In this study, we conduct high time-resolution sampling with a 3 h interval and investigate the temporal variations in molecular characteristics, sources, and formation pathways of BSOA tracers derived from isoprene, monoterpene, and β-caryophyllene at Mt. Huang and then explore the effect of meteorological factors, aerosol aqueous properties (e.g., liquid water content (LWC) and particle acidity (pH is )), anthropogenic pollutants, and mountain-valley breezes on their formation processes in the forested atmosphere.

Concentrations and Major Chemical Compounds of PM 2.5
The diurnal variations in PM 2.5 concentrations, major chemical compounds (e.g., OC, EC, and SIA (sum of SO 4 2− , NO 3 − , and NH 4 + )), meteorological parameters, and gaseous pollutants are illustrated in Table 1 and Figure 1. The mean mass concentration of PM 2.5 (3 h) was 13.3 ± 7.3 µg m −3 in the whole observation period, which was significantly lower than that at Mt. Tai (37 ± 16 µg m −3 ) in the summer [23] and in surrounding megacities (e.g., Hefei, Shanghai, Nanjing, and Hangzhou) (19.3 ± 2.3 µg m −3 ) of the YRD during the corresponding period of observation (https://www.aqistudy.cn/, accessed on 21 November 2022). According to the Chinese National Ambient Air Quality Standard (35 µg m −3 ), the air quality at Mt. Huang was good. Therefore, Mt. Huang can be considered as an ideal rural background site for studying pollution characteristics in East China. Interestingly, the concentration level of PM 2.5 over Mt. Huang was almost equal to that at Mt. Wuyi (14 ± 7.8 µg m −3 ) in summer [12], representing a representative background level in the atmosphere of China. The daytime (7:00-19:00) concentrations of both OC and EC were lower than those at night (19:00-7:00 of the next day), but the ratios of OC/EC were similar during the daytime and night-time (Table 1). Secondary organic carbon (SOC) was associated with various physical and chemical transformation processes of aerosols, which can be used to assess SOA pollution levels [24]. As shown in Table 1, the night-time concentration of SOC was 2.6 times higher than that in the day, indicating more SOA production from the aqueous oxidation at night. Levoglucosan is principally derived from the pyrolysis of cellulose and hemicellulose at temperatures higher than 300 • C, and thus it can be recognized as a significant tracer of biomass burning [25]. The average concentration of levoglucosan was as low as 2.4 ± 2.6 ng m −3 , and the percentage of C from levoglucosan was 0.018 ± 0.016% in OC. Similar to the diurnal variations in PM 2.5 , OC, EC, and SOC concentrations, concentrations of SIA at night were also higher than those in the day, largely because of the dynamic transport of air pollutants via mountain breezes [26]. The difference in solar radiation leads to a large temperature difference between the peak and the foot of the mountain during the day and night, which in turn generates a difference in atmospheric pressure and forms the mountain-valley winds. The mountain breeze prevailed at night, and atmospheric pollutants were more likely to accumulate in the valley. In addition, the higher humidity (75.6 ± 12.2%) at night and larger daily range of temperature (10 • C) (Table 1, Figure S1) were conducive to the gas-to-particle transformation of organic compounds [27]. Similarly, SO 2 and NO 2 at night-time also presented higher concentrations than those in daytime ( Table 1). The LWC of aerosol is controlled by both SIA concentration and RH [9]. Considering the higher SIA concentration and RH at night, LWC presented a higher concentration at night than that in the day (Table 1). Conversely, pH is showed a higher value in the day than that at night (

Diurnal Variations in BSOA Tracers
A total of 13 BSOA tracers were identified in the PM 2.5 samples, which can be classified into three categories: SOA I , SOA M , and SOA C ( Table 2). The total concentration of BSOA tracers at night was 1.7 times higher than that in the day. SOA I tracers were the most abundant species, accounting for 56.1% and 61.5% of the total determined BSOA tracers during the daytime and night-time, respectively, followed by SOA M tracers (41.8% and 36.6%) and SOA C tracers (2.1% and 1.9%).
Diurnal variations (3 h interval) in the BSOA tracers, O 3 concentrations, and the PBL are shown in Figure 2. Nearly all of the BSOA tracers, except for 2-methylglyceric acid, presented higher concentrations at night than during the day, and culminated at 1:00 to 4:00 or 4:00 to 7:00. RH, LWC, SO 4 2− , and NO 3 − also peaked during the above time periods ( Figure S1). Higher RH and LWC can reduce the viscosity of the aerosol particles and enhance the hygroscopicity, which in turn promotes the generation of BSOAs [28]. In addition, according to the backward trajectory ( Figure 3), pollutants from long-distance transport were transported and accumulated at the foot of the mountain with the effect of the prevailing mountain breeze. According to Table 2, the ratio of BSOA tracers to OC was greater at night than in the day, indicating that the PBL height was not a major reason for the increase in BSOA tracers at night. The daytime concentrations of SOA I and SOA M tracers reached the highest values at 10:00 to 13:00 or 13:00 to 16:00, which was consistent with the variation inn O 3 , largely because of the enhanced photochemical oxidation driven by O 3 [11].

Isoprene SOA Tracers
The night-time concentration (87.0 ± 61.2 ng m −3 ) of SOA I tracers was 1.9 times higher than that (46.6 ± 32.1 ng m −3 ) in the day ( Table 2). The molecular distribution of SOA I tracers was characterized by the dominance of 2-methyltetrols during the day and C 5 -alkene triols during the night, and 3-MeTHF-3,4-diol and 2-methylglyceric acid were the two lowest species during the daytime and night-time ( Figure 2).
3-MeTHF-3,4-diols, 2-methyltetrols, and C 5 -alkene triols can be primarily produced from the reactive uptake of isoprene epoxydiols (IEPOX) formed by the oxidation of isoprene with RO 2 and HO 2 ·radicals in the absence of NO x [11]. IEPOX can be partitioned from the gaseous phase into the aerosol phase and ultimately generate 3-MeTHF-3,4-diols and/or C 5 -alkene triols via acid-catalyzed intermolecular rearrangement reactions [27]. C 5 -alkene triols can be derived from the oxidation of 3-MeTHF-3,4-diol with acidic particles [29]. Therefore, C 5 -alkene triols presented a strong linear correlation with 3-MeTHF-3,4-diol during the day (r = 0.81, p < 0.01) and night (r = 0.87, p < 0.01). It is worth noting that the slope (ratio of C 5 -alkene triols to 3-MeTHF-3,4-diol) of the regression line at night (16.6) was almost twice higher than that (8.9) during the daytime ( Figure S2), indicating the stronger ring-opening reactions of the yield of C 5 -alkene triols from 3-MeTHF-3,4-diol at night [30]. The enhanced aqueous-phase oxidation of IEPOX derived from the oxidation of isoprene in the acidic/neutral particles at night was responsible for the higher night-time concentration of 2-methyltetrols [31,32]. Different from other detected BSOA tracers, the concentration of 2-methylglyceric acid in the day was 1.2 times higher than that at night (Table 2), which might be ascribed to the lower relative humidity conditions in the day. Field measurements and chamber studies pointed out that 2-methylglyceric acid is mainly derived from the oxidation of isoprene under high NO x conditions and is facilitated into the aerosol phase under the less humid environment [13,32,33]. As shown in Figure 4, there were robust correlations between SOA I tracers and SO 4 2− during the daytime (r = 0.80, p < 0.01) and night-time (r = 0.83, p < 0.01), suggesting that the aqueous-phase chemistry reaction was an important pathway for the formation of SOA I tracers in the atmosphere.

Isoprene SOA Tracers
The night-time concentration (87.0 ± 61.2 ng m -3 ) of SOAI tracers was 1.9 times higher than that (46.6 ± 32.1 ng m -3 ) in the day ( Table 2). The molecular distribution of SOAI tracers was characterized by the dominance of 2-methyltetrols during the day and C5alkene triols during the night, and 3-MeTHF-3,4-diol and 2-methylglyceric acid were the two lowest species during the daytime and night-time ( Figure 2).

Isoprene SOA Tracers
The night-time concentration (87.0 ± 61.2 ng m -3 ) of SOAI tracers was 1.9 times higher than that (46.6 ± 32.1 ng m -3 ) in the day ( Table 2). The molecular distribution of SOAI tracers was characterized by the dominance of 2-methyltetrols during the day and C5alkene triols during the night, and 3-MeTHF-3,4-diol and 2-methylglyceric acid were the two lowest species during the daytime and night-time ( Figure 2). mainly derived from the oxidation of isoprene under high NOx conditions and is facilitated into the aerosol phase under the less humid environment [13,32,33]. As shown in Figure 4, there were robust correlations between SOAI tracers and SO4 2-during the daytime (r = 0.80, p < 0.01) and night-time (r = 0.83, p < 0.01), suggesting that the aqueousphase chemistry reaction was an important pathway for the formation of SOAI tracers in the atmosphere. . Correlation analysis (r) of SOA tracers with levoglucosan (Lev), liquid water content (LWC), NO3 -, SO4 2-, in situ particle pH (pHis), relative humidity (RH), temperature (T), and O3 during the daytime and nighttime, respectively. Correlation coefficients with an asterisk (*) indicate statistically significant relationships at 0.01 < p < 0.05, two asterisks (**) means highly significant relationships at 0.001 < p < 0.01, and no asterisk indicates not significant relationships at p > 0.05.

Monoterpene SOA Tracers
The concentration of total SOAM tracers at night was 1.5 times higher than that in the day. 3-HGA was the most abundant SOAM tracer, which accounted for 40.0% and 42.5% in the total SOAM tracers during the day and night, respectively, followed by MBTCA (37.2% versus 41.8%), cis-pinonic acid (13.7% versus 9.3%), and cis-pinic acid (9.1% versus 6.4%). Cis-Pinonic acid and cis-pinic acid are the first-generation products of SOAM tracers, which can be further photodegraded by HO· radicals to form MBTCA [34]. Thus, the P/M ratio (sum of cis-pinonic acid and cis-pinic acid to MTBCA) can be proposed as an essential tracer for the aging degree of SOAM [35]. The higher P/M ratio reflects the fresher organic aerosols [35]. A chamber study showed that the P/M ratio varied from 1.51 to 3.21 in fresh α-pinene SOA samples [36]. In the aerosols of Mt. Huang, the average ratio (1.1 ± 0.6) of P/M in the daytime was 1.4 times higher than that (0.8 ± 0.6) at night, suggesting that SOAM was more aged at night due to the enhanced aqueous oxidation. The lowest ratio (0.41) of − , SO 4 2− , in situ particle pH (pH is ), relative humidity (RH), temperature (T), and O 3 during the daytime and nighttime, respectively. Correlation coefficients with an asterisk (*) indicate statistically significant relationships at 0.01 < p < 0.05, two asterisks (**) means highly significant relationships at 0.001 < p < 0.01, and no asterisk indicates not significant relationships at p > 0.05.

Monoterpene SOA Tracers
The concentration of total SOA M tracers at night was 1.5 times higher than that in the day. 3-HGA was the most abundant SOA M tracer, which accounted for 40.0% and 42.5% in the total SOA M tracers during the day and night, respectively, followed by MBTCA (37.2% versus 41.8%), cis-pinonic acid (13.7% versus 9.3%), and cis-pinic acid (9.1% versus 6.4%). Cis-Pinonic acid and cis-pinic acid are the first-generation products of SOA M tracers, which can be further photodegraded by HO· radicals to form MBTCA [34]. Thus, the P/M ratio (sum of cis-pinonic acid and cis-pinic acid to MTBCA) can be proposed as an essential tracer for the aging degree of SOA M [35]. The higher P/M ratio reflects the fresher organic aerosols [35]. A chamber study showed that the P/M ratio varied from 1.51 to 3.21 in fresh α-pinene SOA samples [36]. In the aerosols of Mt. Huang, the average ratio (1.1 ± 0.6) of P/M in the daytime was 1.4 times higher than that (0.8 ± 0.6) at night, suggesting that SOA M was more aged at night due to the enhanced aqueous oxidation. The lowest ratio (0.41) of P/M appeared at 4:00-7:00 when SOA M was the most aged, driven by the relatively higher LWC and SO 4 2− ( Figure S1). The average ratio of P/M (1.0 ± 0.8) at the foot of Mt. Huang was comparable to that (0.89) in Hefei, which is nearly 200 km farther away from the sampling site, and was much lower than the ratios of the urban aerosols in North China and South China in summer [5]. Additionally, the 3-HGA/MBTCA ratio can be adopted to investigate the formation pathway of SOA M from α-pinene or β-pinene, because the oxidation of α-pinene relative to β-pinene can form more MBTCA than HGA [28]. The HGA/MBTCA ratio is as low as 1.0 when α-pinene makes a more significant contribution to SOA M [35,37]. The daytime (1.6 ± 0.9) and night-time (1.7 ± 1.1) ratios of 3-HGA/MBTCA were close to 1.0, indicating that α-pinene was a more important precursor of SOA M compared to β-pinene at Mt. Huang.

β-Caryophyllene SOA Tracers
As one of the predominant sesquiterpenes, β-caryophyllene can be emitted from pine trees, sunflower, corn, and other crops [38]. β-caryophyllinic acid is generated from the ozonolysis or photochemical oxidation of β-caryophyllene, and thus it can be proposed as an indicator for SOA from sesquiterpenes [10]. Interestingly, the pattern of diurnal changes in β-caryophyllic acid was consistent with levoglucosan (the tracer of biomass burning) (Tables 1 and 2), suggesting an important effect of biomass burning on the formation of β-caryophyllic acid. Consistent with the pattern of diurnal variations in SOA I and SOA M , the β-caryophyllinic acid concentration presented a higher concentration in the night-time than that in the daytime (Table 2) and peaked at 4:00-7:00 ( Figure 2).

Comparison of BSOA Tracers with Other Sites
A comparison of the concentration level of BSOA tracers over Mt. Huang with other previous studies is shown in Figure 5. The molecular distribution of BSOAs at Mt. Huang followed the order of SOA I > SOA M > SOA C , which is consistent with that in other regions, except for Mt. Wuyi and Qinghai Lake where SOA M was the dominant species ( Figure 5). PM 2.5 samples were collected at the foot of Mt. Huang (483 m a. s. l) during the summer in this study, when the sampling site was mainly covered by evergreen and deciduous broad-leaved mixed forests [39]. However, the summit of Mt. Hua (29.8 ng m −3 ) in the North China Plain is covered with little vegetation and most of the land is barren [13]. Thus, the total concentration (112.3 ± 37.8 ng m −3 ) of BSOA tracers at the foot of Mt. Huang was much higher than that at Mt. Hua (29.8 ng m −3 ) in summer. Interestingly, we found that the concentration (66.8 ng m −3 ) of SOA I tracers at Mt. Huang was comparable to that (69.0 ng m −3 ) at Mt. Fuji of Japan [14], primarily because both are dominated by broad-leaved forests. Isoprene is mainly emitted from broad-leaved trees and herbaceous plants [28,32]. The concentration of SOA I tracers in this study was lower than that at the top of Mt. Gongga , as well as in the continental background regions in the Tibetan Plateau [43], Nam Co Lake (26.6 ng m −3 ), and Qinghai Lake (3.8 ng m −3 ) [32]. As shown in Figure 4, the concentration of SOA I tracers was higher in the eastern regions such as Mt. Tai, Mt. Changbai, and Mt. Wuyi than that in the western regions such as Mt. Hua and Mt. Himalayas, which was related to the different vegetation types dominated by the natural geographical environment characteristics (e.g., climate, precipitation, and solar radiation).

Effects of Anthropogenic Sources and Meteorological Parameters on BSOA Formation
The correlations between BSOA tracers and temperature, relative humidity, O 3 , pH is , LWC, SO 4 2− , NO 3 − , and levoglucosan are shown in Figure 4. The BSOA tracers derived from isoprene, monoterpene, and β-caryophyllene were correlated strongly with O 3 (r = 0.63-0.73, p < 0.01) in the daytime, whereas there was no correlation (−0.35 ≤ r ≤ −0.33, p > 0.05) at night, which was also observed over Mt. Tai [6]. O 3 and OH· radicals have been considered key atmospheric oxidants during the daytime, while NO 3 radicals are important oxidants at night [48,49]. Additionally, the chemical sink of O 3 by the residual NO titration at night may be an alternate reason for the insignificant correlation of BSOA tracers with O 3 at night [50]. Therefore, it can be concluded that BSOAs at the foot of Mt. Huang during the daytime were primarily generated from the photo-oxidation of BVOCs with O 3 , whereas BSOAs at night may be mainly derived from the oxidation with NO 3 radicals and other oxidants. Each BSOA tracer showed no correlation with temperature throughout the whole observation period (r from −0.13 to +0.20, p > 0.05, Figure 4), indicating that the effect of temperature on the production of BSOAs was negligible at Mt. Huang. The increased temperature can accelerate biogenic VOC emissions and further favor the BSOA formation, as reported in previous studies [13,51,52]. Nevertheless, the increase in temperature can promote the volatilization of BSOAs, resulting in the decreased concentration of BSOAs in the aerosol phase [53,54]. The mountain-valley breezes could also be responsible for such insignificant correlations. SO 4 2− is a representative product of aqueous oxidation, and thus the correlation between SOA tracers and SO 4 2− can be used to investigate the aqueous formation process of BSOAs [32,55]. BSOA tracers exhibited stronger correlations (r = 0.66−0.83, p < 0.01) with SO 4 2− in the night-time than those (r = 0.48-0.49, 0.01 < p < 0.05, and r = 0.80, p < 0.01) in the daytime (Figure 4), indicating the enhanced aqueous formation processes of BSOAs at night. Chamber studies and field measurements have demonstrated that the acidic condition of particles can promote BSOA formation [31,51]. Lin et al. [30] pointed out that acidic aerosols containing sulfate could promote the conversion of IEPOX into 2-methyltetrols and other IEPOX-derived SOA tracers. As shown in Figure 4, each BSOA tracer presented a significant negative correlation (−0.75 ≤ r ≤ −0.68, p < 0.01) with pH is during the daytime and night-time, confirming that the acidic environment of aerosols is beneficial to the BSOA generation through acid-catalyzed heterogeneous oxidation. Field measurements reported that BSOAs exhibited significant negative correlations with RH over Mt. Hua and Mt. Wuyi [12,13], while BSOAs displayed no relationship with LWC over Mt. Hua and Mt. Tai. Moreover, the chamber study reported that the lower RH favored the formation of 2-methylglyceric acid and its corresponding oligomers, whereas the effect of RH on 2methyltetrols was negligible [33]. In this study, BSOA tracers showed negative correlations with RH (−0.47 ≤ r ≤ −0.35, p < 0.05) during the night-time and moderate correlations with LWC (r = 0.37-0.48, p < 0.05) (Figure 4). High RH and LWC concentration could enhance the gas-to-particle partitioning of volatile organic precursors of BSOAs into the aqueous phase, and ultimately facilitate the BSOA formation. However, high RH and LWC concentration could inhibit the BSOA formation via acid-catalyzed oxidation due to the dilution of particle acidity. Additionally, BSOA tracers showed stronger correlations with RH and LWC during the night-time than those during the daytime (Figure 4).
Previous studies pointed out that anthropogenic pollutants could promote BSOA formation [12,55]. SO 4 2− originates from the aqueous oxidation of SO 2 emitted from the combustion of petroleum and coal [56], and thus it can be proposed as an important indicator for anthropogenic pollutants. There were positive correlations between BSOA tracers and SO 4 2− in the day and night as discussed above, indicating the enhancing effect of SO 4 2− on BSOA generation. One reason for the significant influence of SO 4 2− on BSOAs was that an enhancement of SO 4 2− could promote the ring-opening reaction of IEPOX and further SOA formation [32,57]. In addition, the salting-in effect of SO 4 2− could enhance the solubility of polar organic compounds such as IEPOX [45,55]. These results demonstrated that the higher concentration of SO 4 2− can augment SOAs by promoting gasparticle conversion and aqueous-phase oxidation [27]. Nevertheless, there were moderate correlations (r = 0.40-0.59, p < 0.05) between BSOA tracers and NO 3 − , primarily because of the volatilization of NO 3 − during long-distance transport [58]. Levoglucosan is principally derived from the pyrolysis of cellulose and hemicellulose at temperatures higher than 300 • C, and thus it can be recognized as a significant tracer of biomass burning [25]. There were moderate or strong correlations of levoglucosan with BSOA tracers (r = 0.55-0.78, p < 0.05, Figure 4), suggesting that biomass burning from long-distance transport exerted a significant effect on BSOAs. Because of its lower volatile property, β-caryophyllene is abundant in leaves and wood. It is emitted to the atmosphere only through biomass burning [59]. Thus, levoglucosan exhibited stronger correlations (r = 0.78, p < 0.01) with SOA C tracers than SOA I and SOA M (with r = 0.55, p < 0.05, or r = 0.61-0.75, p < 0.01), indicating that biomass burning exerted a more significant effect on SOA C than SOA I and SOA M . Such a phenomenon was consistent with our previous observation in Liaocheng on the North China Plain [60].

Source Apportionment
To assess the relative contributions of primary and secondary sources of organic aerosols, the SOA tracer method can be used to estimate their relative abundances in OC. In this study, the tracer yield method proposed by Kleindienst et al. [52] was applied to estimate the contribution of secondary organic carbon (SOC) derived from the oxidation of isoprene, monoterpene, and β-caryophyllene. The f SOC (conversion coefficient of BSOA tracers) values were reported to be 0.155 ± 0.039, 0.231 ± 0.111, and 0.023 ± 0.005 for isoprene, monoterpenes, and β-caryophyllene, respectively. According to Zhang et al. [61] and Zhang et al. [62], the contribution of biomass burning was calculated via the concentration of levoglucosan, and the f SOC was 0.080 ± 0.033. It is noteworthy that the SOA tracer method possesses a certain degree of uncertainties, as discussed by El Haddad et al. [63] and Yttri et al. [64] in detail. However, such an assessment was considered the most scientific available method thus far to provide meaningful insights into the diurnal trends of SOA tracers. The equation for calculating [SOC] i is given as follows: In the above equations, C 2-MTs , C C5-alkane , C 2-MG , C caryophyllinic , and C levoglucosan refer to the concentrations of 2-methyltetrols, C 5 -alkene triols, 2-methylglyceric acid, β-caryophyllinic acid, and levoglucosan in aerosols, respectively; ∑ BSOAM refers to the sum of cis-pinonic acid, cis-pinic acid, 3-hydroxyglutaric acid, and 3-methyl-1,2,3-butanetricarboxylic acid.
The calculated OCs are shown in Table 2 and Figure 6. The total calculated OC mass concentrations were 0.55 ± 0.11 µgC·m −3 and 0.93 ± 0.23 µgC·m −3 during the day and night, accounting for 12.7 ± 6.0% and 15.1 ± 8.4% of the sum of OC, respectively ( Table 2). The diurnal variations in primary OC and biogenic SOC were consistent with the concentration levels of the detected organic species characterized by the higher night-time values than in daytime. The concentration of biogenic SOC (0.71 ± 0.40 µgC·m −3 ) was more than 20 times higher than that (0.03 ± 0.03 µgC·m −3 ) of primary OC (biomass-burningderived OC). Isoprene-derived SOC was the most abundant species during the daytime and night-time, followed by monoterpene-, β-caryophyllene-derived SOC, and biomassburning-derived OC. These four species accounted for 6.8 ± 2.8%, 3.6 ± 1.4%, 1.8 ± 0.9%, and 0.5 ± 0.5% of OC during the daytime and 8.8 ± 5.6%, 3.7 ± 2.1%, 2.0 ± 1.4%, and 0.6 ± 0.4% during the night-time, respectively. The average concentration of biomass-burning-derived OC was 0.03 ± 0.03 µgC·m -3 , which was equal to that (0.03 ± 0.01µgC·m -3 ) over Mt. Fuji. However, the contribution of biomass-burning-derived OC to total OC concentrations was only 0.6 ± 0.5% during the entire sampling period, which was much less than that (24%) in early June over Mt. Tai where organic aerosols were significantly influenced by biomass burning, and smaller than that (2.3%) over Mt. Fuji of Japan where the effect of biomass burning was minor, indicating the negligible influence of biomass burning on organic aerosols at the foot of Mt. Huang. The relative abundance (13.4 ± 6.9%) of estimated biogenic SOC in aerosol OC at Mt. Huang was almost the same as that in the urban aerosol of Jinan (13.3%) [65] and higher than that over Mt. Tai in June (5.5%) [26] and a rural site (5.5%) in Shanxi Province of China [9], but lower than at Mt. Fuji (37.2%) [14], Mt. Gongga (16.6%) [38], and a rural site in the central Pearl River Delta (38.4%) [51] in the summer.

Sample Collection
Mt. Huang (30.1° N, 118.2° E, 1840 m a. s. l) is situated in the southwest of the Yangtze River Delta (YRD), East China. Climate in this area is heavily regulated by the East Asian monsoon with a marked seasonal variability in temperature, cloudiness, and relative humidity. More than 90% of the mountainous land is covered by vegetation, which comprises broad-leaved forests at low altitudes and coniferous forests at high altitudes. The climate of Mt. Huang varies significantly on a daily basis, owing to the contrasting topography of high mountains and deep valleys. The average concentration of biomass-burning-derived OC was 0.03 ± 0.03 µgC·m −3 , which was equal to that (0.03 ± 0.01µgC·m −3 ) over Mt. Fuji. However, the contribution of biomass-burning-derived OC to total OC concentrations was only 0.6 ± 0.5% during the entire sampling period, which was much less than that (24%) in early June over Mt. Tai where organic aerosols were significantly influenced by biomass burning, and smaller than that (2.3%) over Mt. Fuji of Japan where the effect of biomass burning was minor, indicating the negligible influence of biomass burning on organic aerosols at the foot of Mt. Huang. The relative abundance (13.4 ± 6.9%) of estimated biogenic SOC in aerosol OC at Mt. Huang was almost the same as that in the urban aerosol of Jinan (13.3%) [65] and higher than that over Mt. Tai in June (5.5%) [26] and a rural site (5.5%) in Shanxi Province of China [9], but lower than at Mt. Fuji (37.2%) [14], Mt. Gongga (16.6%) [38], and a rural site in the central Pearl River Delta (38.4%) [51] in the summer.

Sample Collection
Mt. Huang (30.1 • N, 118.2 • E, 1840 m a. s. l) is situated in the southwest of the Yangtze River Delta (YRD), East China. Climate in this area is heavily regulated by the East Asian monsoon with a marked seasonal variability in temperature, cloudiness, and relative humidity. More than 90% of the mountainous land is covered by vegetation, which comprises broad-leaved forests at low altitudes and coniferous forests at high altitudes. The climate of Mt. Huang varies significantly on a daily basis, owing to the contrasting topography of high mountains and deep valleys.
The monitoring site is located at the Anhui Academy of Forestry of Mt. Huang (30.1 • N, 118.2 • E, 483 m a. s. l, Figure 3), which is situated at the foot of Mt. Huang. There are not any obvious anthropogenic emissions near the sampling site, and thus it can reflect the air quality of alpine background sites in East China. The PM 2.5 sampling at a 3 h time resolution was conducted from July 27 to August 3 in 2019 using a high-volume air sampler (TE-6070-BLX, TISCH, New York, NY, USA) coupled with prebaked (450 • C for 6 h) quartz fiber filters at an airflow rate of 1.13 m −3 min −1 . Two blank samples were collected at the beginning and the end of the sampling period by installing the filters onto the sampler for about 15 min without pumping any air. A total of 53 PM 2.5 samples and 2 field blank samples were collected in this study. After sampling, each filter was sealed in an aluminum bag and stored at −20 • C prior to analysis.
Gaseous pollutants (e.g., CO, SO 2 , NO 2 , and O 3 ) and meteorological parameters including temperature, relative humidity, and wind speed/direction at the foot of Mt. Huang were retrieved from the website of Chinese Air Quality Monitoring System (http://www. aqistudy.cn, accessed on 8 September 2022) at a one-hour time resolution. Planetary boundary layer (PBL) height was obtained from the website of Copernicus Meteorological Center (https://cds.climate.copernicus.eu/cdsapp#!/dataset/reanalysis-era5-single-levels?tab, accessed on 15 October 2022).

Organic Compounds
Detailed methods for extraction, derivatization, and gas chromatography/mass spectrometry (GC/MS) analysis have been reported in our previous studies [40,60]. Half of a circle (a diameter of 90 mm) of each filter was cut into pieces and put into a sample bottle. Methylene chloride and a methanol mixed solution (2:1, v/v) were added and ultrasonic extraction was carried out three times (each time for 15 min). The extracts were filtered through glass wool in a pasteurized dropper and transferred to a pear-shaped bottle; they were then concentrated by a rotary evaporator in vacuum state and dried with pure nitrogen gas. After reacting with a 60 µL mixture of N,O-bis-(trimethylsilyl)trifluoroacetamide (BSTFA) and pyridine (5:1, v/v) at 70 • C for 3 h, the derivatives were diluted with 40 µL nhexane containing the internal standard (C 13 n-hexane) before GC/MS analysis. The derivatized fraction was determined by GC/MS (Agilent, 7890A-5977C, Santa Clara, CA, USA) equipped with a fused silica capillary column (HP-5MS, Santa Clara, CA, USA).

Organic Carbon (OC), Elemental Carbon (EC), and Inorganic Ions
OC and EC were analyzed using a DRI Model 2015 Carbon Analyzer following the Interagency Monitoring of Protected Visual Environments (IMPROVE) thermal/optical reflectance (TOR) protocol. A 0.5026 cm 2 filter membrane was placed in the analyzer sample boat and the PM 2.5 samples were heated sequentially to 140 • C (OC1), 280 • C (OC2), 480 • C (OC3), and 580 • C (OC4) in an oxygen-free pure helium (HE) environment. The temperature was gradually increased to 580 • C (EC1), 740 • C (EC2), and 840 • C (EC3) under the condition of He containing 2% oxygen. The limits of detection for OC and EC are 0.08 µg·m −3 and 0.06 µg·m −3 , respectively.
Inorganic ions were analyzed by ion chromatography (Dionex-Aquion, Dionex-600, Thermo Fisher, Waltham, MA, USA). An aliquot of each sample filter was extracted with 30 mL of Milli-Q water using an ultrasonic bath three times and then filtered through PTFE filters to remove particles and filter debris, and finally identified using by ion chromatography.

Quality Assurance and Quality Control
Field blanks were extracted and measured in the same way as the PM 2.5 samples, and no obvious target compounds (less than 3%) were found in the field blanks. The recovery experiment was performed by spiking the standard solution onto blank filters and recoveries of the target compounds ranged from 85% to 120%. Method detection limits for 2-methylthreitols, 3-methyl-2,3,4-trihydroxy1-butene, 2-methylglyceric acid, 3-HGA, and β-caryophyllinic acid are 0.04, 0.05, 0.08, 0.05, and 0.12 ng m −3 , respectively.

Aerosol Liquid Water Content and Particle In Situ pH
In this study, liquid water content (LWC) and particle in situ pH (pH is ) in the PM 2.5 samples were calculated using the ISORROPIA-II model. This model treats Na + -NH 4 + -K + -Ca 2+ -Mg 2+ -SO 4 2− -NO 3 − -Cl − as a system [27]. The forward and metastable mode in the ISORROPIA model was employed.

Conclusions
The diurnal variations in BSOA tracers derived from isoprene, monoterpene, and β-caryophyllene were analyzed for PM 2.5 samples collected at the foot of Mt. Huang during the summer. We found that SOA I tracers were the most abundant species, accounting for 56.1% and 61.5% of the total BSOA tracers during the day and night, followed by SOA M tracers (41.9% and 36.7%) and SOA C tracers (2.1% and 1.9%). The night-time concentrations of levoglucosan, OC, EC, and SIA were also higher at night than during the day. Nearly all of the BSOA tracers exhibited higher concentrations at night than in the day. In addition, the cumulative effect of the dynamic transport of atmospheric pollutants through the mountain breeze at night was another crucial reason for the higher BSOA tracers. There was a positive correlation between BSOA tracers and O 3 during the day rather than at night, indicating that BSOA tracers were primarily generated from the photo-oxidation of BVOCs with O 3 only during the day. BSOA tracers were strongly correlated with pH is and SO 4 2− , indicating that BSOAs were produced from the acid-catalyzed aqueous-phase oxidation. The robust correlation of levoglucosan with BSOA tracers, especially β-caryophyllinic acid, suggested that biomass burning from long-distance transport exerted a significant effect on BSOA formation.
Informed Consent Statement: Not applicable.
Data Availability Statement: The data for this paper will be made available on request from the corresponding author.