1. Introduction
In recent years, the frequency of warnings and advisories issued for fine particulate matter (PM10, PM2.5) in South Korea has steadily increased, raising significant public health concerns. Numerous studies have investigated the formation mechanisms of particulate matter and strategies for its reduction. Particulate matter (PM) is known to comprise the majority of atmospheric particles by number and mass. In recent years, attention has increasingly focused on anthropogenic sources rather than natural origins. The chemical composition of PM reveals that secondary pollutants-such as sulfate and nitrate—formed through atmospheric reactions involving nitrogen oxides (NOx), sulfur oxides (SOx), and ammonia, constitute a substantial portion of its mass. PM contains various cations (Na+, NH4+, K+, Ca2+, Mg2+) and anions (Cl−, NO3−, SO42−), many of which are water-soluble. These ionic species can serve as essential nutrients for ecosystems; however, excessive concentrations may exert harmful effects. Nitrate (NO3−) in the atmosphere is believed to form primarily through secondary processes, such as the homogeneous reaction between gaseous HNO3 and NH4+-both products of photochemical reactions-resulting in compounds like ammonium nitrate (NH4NO3) [1]. Moreover, the emission of combustion-related air pollutants, including NOx, SOx, organic compounds, and fine particulate matter, has led to a significant increase in the number of low-visibility days annually [2]. As a major air pollutant, particulate matter (PM) has been classified as a Group 1 carcinogen by the International Agency for Research on Cancer (IARC) under the World Health Organization (WHO) since 2013 [3]. Additionally, research has demonstrated that PM has far-reaching impacts on climate, visibility, ecosystems, and human health [4].
Fine particulate matter (PM10) and ultrafine particulate matter (PM2.5) refer to airborne particles with diameters less than 10 μm and 2.5 μm, respectively. These particles can penetrate deep into the alveoli of the lungs, increasing the risk of respiratory and cardiovascular diseases. The health risks associated with PM are not solely determined by its physical properties but also by its chemical composition [5]. PM consists of a complex mixture of water-soluble and water-insoluble compounds, originating from various natural and anthropogenic sources [6]. Among the water-soluble components, organic nitrogen compounds such as amino acids and amines pose significant environmental and health risks due to their high solubility, enabling interactions with water, acids, and bases during rainfall. These reactions can have detrimental effects on terrestrial and aquatic ecosystems [4].
Amino acids in atmospheric aerosols exist in two forms: free amino acids (FAA) and combined amino acids (CAA), derived from biological materials, biomass combustion, and marine environments [7]. Certain amino acid compounds exhibit hygroscopic properties, contributing to ice nucleation and cloud formation, thereby influencing the Earth’s radiative energy balance and climate. Amino acids and protein-like compounds also constitute a significant fraction of atmospheric organic nitrogen and organic carbon, affecting the buffering capacity, acidity, and alkalinity of aerosols, which in turn may impact global geochemical cycles [8]. In the atmosphere, amino acids undergo chemical transformations through reactions with nitrogen oxides (NOx) and ozone (O3), which can contribute to allergic reactions and other adverse health effects. CAA, considered secondary pollutants, are formed through the photochemical breakdown and recombination of primary compounds such as FAA [9]. The ratio of FAA to CAA serves as an indicator of atmospheric processing: regions with a high FAA/CAA ratio are dominated by primary emissions, while areas with a low FAA/CAA ratio indicate extensive photochemical transformation leading to CAA formation [10]. Due to these concerns, recent studies have focused on amino acids and proteins among the water-soluble nitrogen compounds present in particulate matter. According to Kang et al. (2012) [11], the Micro BCA Protein Assay Kit was used to analyze proteins in particulate matter samples collected in Hefei, China. Similarly, Song et al. (2017) [12] collected particulate matter in Guangzhou, China, and analyzed proteins and free amino acids (FAA) using the Micro BCA Protein Assay Kit and HPLC. In these studies, commercially available protein quantification kits—such as the Bradford reagent kit, NanoOrange protein quantitation kit, and Micro BCA protein assay kit—should be used with caution when interpreting the results. These methods may fail to accurately determine the protein content of samples due to significant interference from humic substances [8]. Additionally, Mandalakis et al. (2010) [13] analyzed FAA and CAA in particulate matter collected in Finokalia, Greece, using GC-MS. GC–MS is not commonly used for amino acid analysis, except in certain applications such as aerosols and biological samples. Moreover, GC–MS requires a derivatization step prior to analysis, which is a notable limitation [7]. To overcome these drawbacks, liquid chromatography–tandem mass spectrometry (LC–MS/MS) has recently gained widespread use as an alternative approach. Baek et al. (2022) [14] conducted a study in Seoul, South Korea, where particulate matter samples were collected and analyzed for FAA, combined amino acids (CAA), and amines using LC-MS/MS. Although research on proteins and FAA has been actively conducted in recent years, studies on CAA remain limited. This is primarily because amino acids in the atmospheric environment exist at extremely low concentrations, making it challenging to identify and quantify individual amino acid components [4]. Furthermore, the determination of CAA requires the hydrolysis of peptides and proteins followed by total amino acid analysis, which involves a complex analytical procedure [15].
Amino acid analysis is commonly performed using liquid chromatography (LC) methods that take advantage of their high solubility and low volatility. These methods involve derivatization reagents such as 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate (AQC), 9-fluorenylmethyl chloroformate (FMOC), o-phthaldialdehyde (OPA), and phenylisothiocyanate (PITC), followed by detection using spectrophotometric, chemiluminescent, or conductivity detectors [16]. However, challenges with derivatization include reagent instability, interference, and the need for time-consuming procedures [17].
Water-soluble substances and carbon compounds in particulate matter play a crucial role in the biogeochemical cycles of the Earth and are considered significant contributors to visibility impairment [2]. Additionally, the hygroscopic nature of water-soluble compounds, such as nitrates and sulfates, influences particle size under varying humidity conditions, while many essential chemical reactions occurring in atmospheric circulation are regulated by the solubility of water-soluble ionic components [18]. In South Korea, studies on water-soluble compounds and carbon compounds in atmospheric particulate matter have been conducted in response to visibility issues in various regions, including Incheon [2], Gwangju [18], and Daejeon [1].
According to the “Air Korea” annual report by the Korea Environment Corporation (https://www.airkorea.or.kr/), the Jinju air quality monitoring station in Gyeongsangnam-do recorded the highest annual average particulate matter concentrations in the region over the past three years. During this period, the average annual concentrations were as follows: PM10 - 30 ~ 39 μg/m3 and PM2.5 - 15 ~ 17 μg/m3. Moreover, research by Matos et al. (2016) [8] suggests that the concentration and composition of amino compounds in atmospheric aerosols vary significantly due to their residence time in the atmosphere and the complexity of associated processes. This variability largely depends on the sources of these compounds and the influence of meteorological conditions.
Primary biological aerosol particles, including viruses and microorganisms (e.g., algae, fungi, bacteria), are likely major contributors to atmospheric protein content. Additionally, bioaerosols can originate from anthropogenic sources such as industrial facilities (e.g., textile factories), agricultural activities (e.g., fertilizers), and wastewater treatment plants, which are key components of urban-rural complex areas.
This study aimed to measure the concentrations of particulate matter (PM10 and PM2.5) and characterize their composition. Specifically, organic carbon and nitrogen, amino acids (classified as FAA and CAA), amines, and inorganic ions (e.g., Na+, Ca2+, Cl−, NO2−) were analyzed for each PM size fraction. Amino acids were quantified using high-performance liquid chromatography–tandem mass spectrometry (HPLC-MS/MS), a method that eliminates the need for derivatization and minimizes interference in trace analysis. The results were statistically evaluated to provide insights into the composition and formation mechanisms of PM.
2. Materials and Methods
2.1. Site Selection and Study Period
As shown in Fig. 1, Jinju City, Gyeongsangnam-do, is located between 127°53′ and 128°22′ longitude and 35°03′ and 35°26′ latitude. Situated in the western part of Gyeongsangnam-do, the city is traversed by the Nam River, which flows through Jinju after forming Jinyang Lake. Surrounded by mountains, Jinju exhibits basin-like topography, contributing to frequent occurrences of dense fog. According to Kim et al. (2017) [19], visibility in Jinju decreased from 16.3 km in 2009 to 13.4 km in 2015. The annual average PM10 concentrations, based on data from the air monitoring network, were 53 μg/m3 in 2014 and 51 μg/m3 in 2015, exceeding the national air quality standard for annual average concentration (50 μg/m3) and ranking the highest in the Gyeongsangnam-do region. Organic aerosol formation is typically elevated during the summer due to active secondary production via photochemical reactions. However, Baek et al. (2022) [14] conducted seasonal analyses of free amino acids (FAAs) and aliphatic amines using samples collected from the Korea Institute of Science and Technology in Seoul. Their results indicated that FAA concentrations were highest from late winter to spring, while aliphatic amines tended to increase during the fall and winter. Based on these findings, this study conducted a total of 24 measurements over a three-month period from November 2022 to January 2023, corresponding to the winter season. Sampling locations included points situated 3 km (A) and 9 km (B) north of Jinju City Hall, as well as 4 km (C), 11 km (D), and 15 km (E) east of the city hall. PM2.5 samples were collected from locations A, B, and C, and each sample was collected simultaneously from 5 locations (PM10) and 3 locations (PM2.5) for 24-hours.
2.2. Sample Collection Analysis Methods
2.2.1. Particular matter (PM10, PM2.5)
Sampling was conducted at five sites located north and east of Jinju City Hall: 3 km (Site A: PM10, PM2.5) and 9 km (Site B: PM10, PM2.5) to the north, and 4 km (Site C: PM10, PM2.5), 11 km (Site D: PM10), and 15 km (Site E: PM10) to the east. The mass concentrations of fine dust were measured at five sites for PM10 and three sites for PM2.5 in Jinju. Sampling was conducted on building rooftops in accordance with the Atmospheric Pollution Process Test Standard (ES 01605.2, ES 01606.1). A Low Volume Air Sampler (PMS-204, Korea) was used to collect particulate matter at a flow rate of 16.7 L/min over 24 hours, yielding an air sample volume of approximately 24 m3. Quartz filters (47 mm, WHATMAN, 2.0 μm, United Kingdom) were pretreated in an electric furnace at 550°C for 24 hours to minimize organic contamination. Before use, they were stored in a temperature- and humidity-controlled chamber for 24 hours. The mass concentration of PM was determined by weighing the filters before and after sampling using an electronic balance (ATX224, Japan). After concentration measurements, the filters were sealed in aluminum foil, stored at −20°C, and divided into four sections for subsequent analysis of water-soluble organic matter, ions, amino acids, and amines (Table 1).
2.2.2. Water-soluble organic carbon (WTOC) and Water-soluble organic nitrogen (WSON)
A section of the divided filters was placed in 40 mL vials, and 30 mL of distilled water was added. A 30 mL portion of the filter extract was subjected to ultrasonic extraction at 25°C for 1 hour. To eliminate particles that could interfere with the analysis of dissolved compounds, the samples were filtered through a 0.22 μm polypropylene (PP) filter. Total Organic Carbon (TOC) and Total Nitrogen (TN) were analyzed using a TOC/TN analyzer (TOC-L + TNM-L, Shimadzu, Japan). High-purity air was used as the carrier gas at a flow rate of 80 mL/min for sample injection. TOC was quantified using a non-dispersive infrared detector (NDIR) following thermal catalytic oxidation, while TN was measured using an ozone chemiluminescence detector. The measured TOC and TN values were considered the mass concentrations of water-soluble total organic carbon (WTOC) and water-soluble total nitrogen (WTN), respectively. The difference between WTN and water-soluble inorganic nitrogen (WSIN) concentrations was regarded as the water-soluble organic nitrogen (WSON) mass concentration. The WSIN concentration was considered equivalent to the sum of NH3+, NO3−, and NO2− concentrations, measured as described below [4], [10].
2.2.3. Inorganic ions
To analyze five cations (Na+, K+, NH4+, Ca2+, Mg2+) and five anions (Cl−, NO2−, NO3−, SO42−, PO43−), a section of the divided filters was placed in 40 mL vials with the addition of 10 mL of distilled water. A 10 mL aliquot of the filter extract was sonicated at 25°C for 1 hour, followed by filtration through a 0.22 μm polypropylene (PP) filter to remove particles. The resulting filtrate was used for analysis. Anions were analyzed using an ion chromatograph (CBM-20A, Shimadzu, Japan) equipped with a Dionex IonPac™ AS14 RFIC™ (4 mm × 250 mm) column and an electrical conductivity detector. The flow rate was set at 1.0 mL/min, and the eluent consisted of 1.7 mM sodium bicarbonate (NaHCO3) and 1.8 mM sodium carbonate (Na2CO3) dissolved in distilled water. Cations were analyzed using an ion chromatograph (DIONEX ICS-1100, Thermo, USA) equipped with a Dionex IonPac™ CS12A RFIC™ (4 mm × 250 mm) column and an electrical conductivity detector. The flow rate was maintained at 0.7 mL/min, with 15 mM methanesulfonic acid (CH4O3S) used as the eluent.
2.2.4. Amino acids and amines
The final section of the divided filters was placed in 40 mL vials, and 10 mL of 0.1 N hydrochloric acid (HCl) was added. Ultrasonic extraction was performed at 25°C for 30 minutes, followed by filtration through a 0.22 μm polypropylene (PP) filter to remove particulate matter. The filtered extract was concentrated from 3 mL to 1 mL using a nitrogen concentrator and subsequently analyzed for free amino acids (FAA) and amines. Total amino acid (TAA) analysis was performed following the acid hydrolysis procedure outlined by Filippo et al. (2014) [15]. The hydrolyzed sample was then re-dissolved in 1 mL of 0.1 N HCl for analysis. Of the total filtered extract, 3 mL was allocated for FAA analysis, while another 3 mL underwent acid hydrolysis for TAA analysis. The hydrolyzed sample was then concentrated using a nitrogen concentrator and re-dissolved in 1 mL of 0.1 N HCl for further analysis. FAA, combined amino acids (CAA), and amines were analyzed using liquid chromatography-mass spectrometry (LC-MS/MS; LCMS-8040, Shimadzu, Japan) equipped with an ACE5 C18-PFP column (2.1 mm × 150 mm, 5 μm, Shimadzu, Japan). Water (Mobile Phase A) and acetonitrile (Mobile Phase B) were used as eluents, with a flow rate of 0.7 mL/min (Table S1). The CAA mass concentration was determined as the difference between the measured FAA and FA mass concentrations from the TAA and TA mass concentrations. The 24 amino acids were categorized into three groups based on their affinity for water: hydrophobic (7 types), hydrophilic (11 types), and neutral (6 types) (Table 1).
2.2.5. Quality assurance/quality control (QA/QC)
To ensure the reliability of this study, quality assurance and quality control (QA/QC) measures were conducted, including the evaluation of detection limits, quantitation limits, accuracy, precision, and calibration curve linearity for ions (10 types), amino acids (24 types), and amines (6 types). For ions, detection limits ranged from 0.002 μg/mL to 0.014 μg/mL, while quantitation limits ranged from 0.005 μg/mL to 0.056 μg/mL. Accuracy and precision were within the ranges of 87.97%–106.83% and 0.25%–2.21%, respectively, with correlation coefficients (R2) exceeding 0.99. For amino acids and amines, detection limits ranged from 0.935 ng/mL to 7.602 ng/mL, and quantitation limits ranged from 2.979 ng/mL to 24.211 ng/mL. Accuracy was within the range of 99.9%–134.3%, precision ranged from 2.6% to 19.3%, and correlation coefficients (R2) were greater than 0.98. The QA/QC results confirmed that the detection limits, quantitation limits, accuracy, precision, and calibration linearity were all within acceptable thresholds, ensuring the robustness and reliability of the analytical methods employed in this study (Table S2).
3. Results and Discussion
3.1. Fine Particles (PM10, PM2.5)
The mass concentrations of PM10 and PM2.5, along with meteorological factors (temperature, maximum temperature, minimum temperature, humidity, and wind speed) obtained from the Korea Meteorological Administration’s Data Portal, are summarized in Table 2. During the three-month study period, the average temperature in Jinju was approximately 3.6°C, with a maximum of 11.3°C and a minimum of −2.8°C. The average PM10 mass concentration across the five sampling sites was 39 μg/m3, accounting for approximately 39% of the 24-hour standard (100 μg/m3) set by the Ministry of Environment. In contrast, the average PM2.5 concentration from the three sampling sites was 27 μg/m3, representing 76% of the 24-hour standard (35μg/m3). Furthermore, these concentrations were higher than those recorded at Jinju’s air quality monitoring stations during the same period, where PM10 and PM2.5 levels were measured at 36 μg/m3 and 20 μg/m3, respectively. This discrepancy is likely due to differences in monitoring locations; while the air quality monitoring stations are situated in urban areas, the sampling sites in this study were located in rural areas, where biomass combustion sources may have contributed to the elevated particulate matter levels [14].
3.2. Water-Soluble Organic Carbon (WTOC), Water-Soluble Organic Nitrogen (WSON)
The analyzed results of water-soluble total organic carbon (WTOC) and water-soluble total nitrogen (WTN) from the pretreated filters are presented in Table 3. Water-soluble inorganic nitrogen (WSIN) was calculated based on the mass concentrations of NH4+, NO3−, and NO2− ions, while water-soluble organic nitrogen (WSON) was derived as the difference between WTN and WSIN [10]. The average WTOC concentration in PM10 was 7.1 μg/m3, with the highest concentration observed at site E (8.1 μg/m3) and the lowest at site B (6.3 μg/m3). The average WTN concentration in PM10 was 4.8 μg/m3, with site C recording the highest value (5.5 μg/m3) and site A the lowest (4.4 μg/m3). For PM2.5, the average WTOC concentration was 6.3 μg/m3, with the highest at site A (6.6 μg/m3) and the lowest at site B (6.1 μg/m3). The WTOC and WTN concentrations in PM10 were approximately 0.8 μg/m3 higher than those in PM2.5.
The ratios of WTOC and WTN to the total fine particle mass were higher in PM2.5 than in PM10, indicating that smaller particles have a greater influence on WTOC and WTN levels. Additionally, the higher ratio of WTOC to PM compared to WTN to PM suggests that carbon compounds constitute a larger proportion of fine particles than nitrogen compounds. A study by Kim et al. (2022) [5] on the composition of fine particles near livestock farms in Gimhae, Gyeongsangnam-do, found that the carbon content (8–10%) was significantly higher than the sulfur (0.3–2%) and nitrogen (0.1–1%) contents. Similarly, Matsumoto et al. (2021) [10] reported that water-soluble organic carbon (WSOC) concentrations in fine particles in the inland city of Kofu, Japan, were higher than water-soluble total nitrogen (WSTN) concentrations.
The average WSON concentration was 0.31 μg/m3 in PM10 and 0.32 μg/m3 in PM2.5, accounting for approximately 6.5% of WTN in PM10 and 7.9% in PM2.5. Matsumoto et al. (2021) [10] and Baek et al. (2022) [14] reported that amino acids and amines-among the water-soluble organic nitrogen compounds-play a significant role in atmospheric chemistry. Due to their high solubility, they are ubiquitous in rainwater, and their hygroscopic properties and ability to form cloud condensation nuclei (CCN) contribute notably to global climate processes.
Therefore, these findings indicate that water-soluble nitrogen is highly hygroscopic, and the water absorption capacity of fine particles can significantly enhance light scattering, potentially affecting atmospheric visibility. Moreover, water-soluble organic nitrogen can play a crucial role in cloud condensation nuclei formation, contributing to indirect climate effects. Additionally, WSON may influence the acid-neutralizing capacity and light-absorbing properties of atmospheric particles [20]. Importantly, WSON can pose significant health risks due to its mutagenic, carcinogenic, and other toxic properties [4]. In addition, a study investigating the concentration and composition of water-soluble organic nitrogen (WSON) in fog water collected in Davis, California’s Central Valley, compared fine particles and fog water from the same location and period. The results showed that the composition of water-soluble organic carbon (WSOC) in fine particles closely resembled that in fog water [20]. These findings suggest that the frequent fog events in Jinju City may be closely linked to the presence of WSOC in fine particulate matter.
3.3. Ionic Species
Table 4 presents the mass concentrations of ions in PM10 and PM2.5. The ratio of cations to anions in PM10 was 0.98, with a charge balance error of approximately 2%, while in PM2.5, the ratio was 1.06, with a charge balance error of about 6%. Among the anions in PM10, sulfate (SO42−) exhibited the highest average concentration at 4.3 μg/m3, followed by nitrate (NO3−) at 1.0 μg/m3 and chloride (Cl−) at 0.8 μg/m3. A similar trend was observed in PM2.5, with SO42− displaying the highest concentration (3.7 μg/m3), followed by NO3− (0.7 μg/m3) and Cl− (0.5 μg/m3). A study by Jeoung et al. (2018) [21] analyzing the ion composition of PM2.5 in a rural area reported that NO3− (6.5 ± 4.6 μg/m3) and SO42− (4.4 ± 3.2 μg/m3) were the dominant anions. Compared to Jinju, SO42− levels were comparable, whereas NO3− concentrations were significantly lower in Jinju, showing a tenfold difference (approximately 5.8 μg/m3 lower). Kim et al. (2022) [5] investigated fine particles near livestock farms in Gimhae, Gyeongsangnam-do, and found that SO42− had the highest concentrations (2.8 μg/m3 in PM10 and 2.5 μg/m3 in PM2.5), followed by Cl− (1.6 μg/m3 in PM10 and 1.3 μg/m3 in PM2.5) and NO3− (1.3 μg/m3 in PM10 and 1.1 μg/m3 in PM2.5). Thus, in Nonsan, Chungcheongnam-do, which is located near the Seoul metropolitan area, NO3− concentrations were high due to the influence of vehicular traffic and industrial complexes. In contrast, Gimhae, Gyeongsangnam-do, a rural area farther from the metropolitan region, exhibited NO3− concentrations similar to those in Jinju, likely due to the reduced influence of traffic emissions and the predominance of other regional sources. The Jinju area is geographically situated near Jinyang Lake and the Nam River Dam, with the Nam River flowing through the city center. The region experiences an average of 88.5 foggy days per year—the highest in the country [19]. Geng et al. (2010) [2] reported that soluble inorganic ions (NO3−, SO42−, NH4+) and carbonaceous compounds in particulate matter (PM) are major contributors to visibility reduction during urban haze events. Therefore, the high number of foggy days in Jinju is likely attributable to the influence of these soluble inorganic species.
3.4. Amino Acids and Amine
Total amino acids (TAA) can be expressed as the sum of free amino acids (FAA), which are primary pollutants directly and indirectly emitted from biological particles and biomass combustion, and combined amino acids (CAA), which are secondary pollutants formed through photochemical reactions involving primary pollutants in the atmosphere. Areas with high FAA/CAA ratios are dominated by primary pollutants, whereas areas with low FAA/CAA ratios indicate a higher prevalence of photochemical processes leading to CAA formation [10]. In this study, the FAA/CAA ratios were 0.59 for PM10 and 0.62 for PM2.5, suggesting that CAA levels exceeded FAA levels in Jinju. This indicates that secondary pollutants formed via photochemical reactions played a more significant role than primary emissions in determining amino acid composition. Similar trends have been observed in other regions. In Kofu, Japan, the FAA/CAA ratios were 0.65 for PM10 and 0.45 for PM2.5 [10]. In Rome, FAA and CAA were measured by particle size in both summer and winter, yielding FAA/CAA ratios of 0.34 in summer and 0.28 in winter [15]. Zhang and Anastasio (2003) [9] reported an FAA/CAA ratio of 0.27 when measuring amino acids in PM2.5 and fog water in Davis, California, while Mandalakis et al. (2011) [22] found an FAA/CAA ratio of 0.24 in fine particulate matter collected at the Finokalia Oceanographic Observatory in Greece. The results from Japan, which is geographically closer to Jinju, were more comparable, while lower FAA/CAA ratios were observed in Europe and the United States. Photochemical degradation of combined amino acids (CAA) present in atmospheric particulate matter has been identified as a potential source of free amino acids (FAA) [12]. Based on this, Matsumoto et al. (2021) [10] reported a significant correlation between the concentrations of FAA and Ox (oxidants) in particulate matter. Therefore, a higher FAA/CAA ratio may indicate a greater extent of CAA degradation and, consequently, a higher potential for FAA formation. In Jinju, the average FAA concentration in PM10 was 54.8 ng/m3, with a maximum of 78.8 ng/m3 and a minimum of 44.0 ng/m3. The average free amine (FA) concentration was 25.9 ng/m3, with a maximum of 30.7 ng/m3 and a minimum of 23.5 ng/m3. For PM2.5, the average FAA concentration was 52.7 ng/m3, with a maximum of 56.1 ng/m3 and a minimum of 48.5 ng/m3, while the average FA concentration was 27.8 ng/m3, with a maximum of 33.6 ng/m3 and a minimum of 23.9 ng/m3 (Table 5). According to Baek et al. (2022) [14], free amine concentrations (17.0 ng/m3) were higher than FAA concentrations (11.6 ng/m3) in winter in Seoul, with FAA levels increasing from late winter to spring, while amino acid concentrations peaked in fall and winter. The FAA concentration in Jinju was approximately twice that of FA, likely due to the timing of sample collection in late winter.
The composition of FAA in PM10 was primarily hydrophobic amino acids (23.5 ng/m3), followed by hydrophilic (16.0 ng/m3) and neutral amino acids (15.3 ng/m3). A similar trend was observed in PM2.5, where hydrophobic amino acids had the highest concentration (22.3 ng/m3), followed by neutral (15.4 ng/m3) and hydrophilic (15.0 ng/m3) amino acids. The ratio of hydrophobic to hydrophilic amino acids in both PM10 and PM2.5 was approximately 1.5, suggesting a significant contribution of hydrophobic amino acids such as alanine (Ala), valine (Val), leucine (Leu), and isoleucine (Ile), which are known to originate from biological sources associated with soil surfaces in rural and forested areas [23]. According to Barbaro et al. (2011) [12], temporal trends and correlation analyses of amino acid concentrations indicated that phenylalanine (Phe), threonine (Thr), Leu, Ile, methionine (Met), tryptophan (Trp), Val, and serine (Ser) likely originate from similar sources, including terrestrial emissions such as plant debris, bacteria, spores, and pollen. Zhu et al. (2022) [23] examined the seasonal variations of amino acids in urban, rural, and forested areas, identifying glycine (Gly) as a marker of biomass combustion, hydrophobic amino acids (Ala, Val, Leu, and Ile) as indicators of soil sources, and hydrophilic amino acids (glutamic acid (Glu), lysine (Lys), and aspartic acid (Asp)) as being abundant in plant-derived aerosols. These findings align with the results of this study. The dominance of hydrophobic over hydrophilic amino acids in Jinju suggests characteristics typical of rural environments, where soil surfaces and biological particles contribute significantly to aerosol composition. These findings provide insight into the sources and atmospheric transformation of amino acids in fine particulate matter, particularly in regions influenced by natural and agricultural emissions.
As shown in Table 6, the ratio of total combined amino acids (CAA) to total free amino acids (FAA) was approximately twice as high. This finding is consistent with the results of Matsumoto et al. (2021) [10], which reported that CAA concentrations were 1.7 to 3.3 times higher than FAA concentrations in the Kofu region of Japan. Regions with low CAA/FAA ratios are typically dominated by primary pollutants, such as FAA directly emitted from various sources. In contrast, regions with high CAA/FAA ratios indicate a greater influence of secondary pollutants, such as CAA, which are formed through photochemical reactions. The higher CAA concentrations observed in Jinju suggest that secondary pollutants formed via photochemical transformations played a more significant role in fine particle formation than direct emissions from primary sources. The total CAA concentration in PM10 ranged from 65.8 ng/m3 to 132.5 ng/m3, while total CA concentrations ranged from 15.0 ng/m3 to 23.9 ng/m3. Among the CAA components in both PM10 and PM2.5, hydrophilic amino acids exhibited the highest average concentrations (39.9 ng/m3 and 32.6 ng/m3, respectively), followed by neutral amino acids (28.0 ng/m3 and 28.7 ng/m3) and hydrophobic amino acids (25.0 ng/m3 and 23.6 ng/m3), respectively. The composition of amino acids in PM influences how particulate matter interacts with the environment, contributes to cloud formation, and affects human health. A higher proportion of hydrophilic amino acids may enhance PM’s ability to absorb moisture, potentially affecting its atmospheric deposition and associated health risks [24]. The hydrophilic nature of these compounds may also influence their ability to penetrate respiratory tissues, allowing them to interact with mucus and other bodily fluids, thereby impacting respiratory health [25]. Conversely, hydrophobic components may alter the transport and deposition of PM by adhering to other particles or surfaces, affecting their atmospheric lifetime and interactions with environmental surfaces [26]. Understanding the specific molecular composition of these amino acid components provides valuable insights into the origin, behavior, and potential impacts of PM in the atmosphere.
Concentrations of specific species within each component are summarized in Fig. S1 and S2. As shown in Fig. S1, alanine (Ala) was the most abundant free amino acid (FAA) in PM10, with a concentration of 21.6 ng/m3, accounting for 27% of total FAA. This was followed by asparagine (Asn) (8%), glycine (Gly) (7%), aspartic acid (Asp) (7%), threonine (Thr) (5%), and serine (Ser) (4%). The dominant amines in PM10 were propylamine (PrA), trimethylamine (TMA), and diethylamine (DEA). Similarly, in PM2.5, Ala was again the most abundant FAA (21.0 ng/m3, 27%), followed by Asn, Asp, Gly, and Ser. The dominant amines in PM2.5 were dimethylamine (DMA), followed by PrA and TMA. These findings align with previous studies by Baek et al. (2022) [14], Matsumoto et al. (2021) [10], Barbaro et al. (2011) [17], and Filippo et al. (2014) [15]. Baek et al. (2022) [14] identified Gly (5.45 ng/m3, 51.2%), Ala (1.31 ng/m3, 12.3%), lysine (Lys) (0.85 ng/m3, 8.0%), and glutamic acid (Glu) (0.35 ng/m3, 3.3%) as dominant amino acids in PM2.5 in Seoul. The mass concentration of Ala in Jinju was approximately 20 times higher than in Seoul, showing similar trends to Gly and Thr. Furthermore, aliphatic amines such as TMA, DMA, and methylamine (MA) were the most abundant amines. In Matsumoto et al. (2021) [10], serine (Ser) (51%) and arginine (Arg) (16%) were the dominant amino acids in PM10, while FAA was primarily composed of Ala (7%), Thr (4%), and Gly (3%). The concentration of Ala in Jinju was approximately twice as high as in the Kofu region of Japan, while the Thr ratio was similar. Studies by Barbaro et al. (2011) [17] and Filippo et al. (2014) [15] also reported Gly as the most dominant amino acid, followed by Glu, proline (Pro), Ala, and others. Data collected from Venice, Italy, exhibited similar ratios of Asp (6%), Asn (10%), and Ser (4%), while studies in Rome showed comparable results for Gly (5.33 ng/m3), Asp (3.46 ng/m3), and Ser (4%) during both summer and winter. Gly and Ser are commonly found in both marine and urban environments and are among the most abundant amino acids in fine particulate matter [10]. However, the dominance of Ala observed in this study differs from previous findings and is likely attributable to the agricultural characteristics of the study area. Ala is strongly associated with soil organic matter and microbial biomass, suggesting that local sources, such as agricultural activities and rural soil surfaces, may significantly influence the composition of FAA in Jinju [23].
As shown in Fig. S2, the average concentration of combined amino acids (CAA) in PM10 was highest for aspartic acid (Asp) (17.9 ng/m3, 16%), followed by alanine (Ala), glycine (Gly), asparagine (Asn), and serine (Ser). The major amines detected in PM10 included diethylamine (DEA) (9.0 ng/m3, 8%), dimethylamine (DMA) (5.3 ng/m3, 5%), and propylamine (PrA) (3.6 ng/m3, 3%). In PM2.5, the highest average CAA concentration was observed for Ala (17.5 ng/m3, 17%), followed by Gly, Asp, Asn, and Ser. The primary amines in PM2.5 included DEA (6.3 ng/m3, 6%), PrA (5.1 ng/m3, 5%), and DMA (1.5 ng/m3, 1%). Ala, Gly, Asp, and Asn are α-amino acids that play essential roles in protein biosynthesis, sharing common α-amino and α-carboxyl functional groups. Ala contains a methyl side chain, making it structurally simple and highly abundant in nature. Gly, with a single hydrogen atom as its side chain, is the simplest stable amino acid involved in protein synthesis. Asp features an acidic side chain, allowing it to interact with proteins, and is classified as an acidic amino acid widely involved in biosynthetic pathways. Asn has a carboxamide side chain, classifying it as a neutral amino acid. Filippo et al. (2012) [15] identified Ser, Ala, and Gly as some of the most common amino acids found in proteins, highlighting Ala as one of the most abundant natural amino acids due to its small size and neutral properties. Similarly, Ge et al. (2011) [6] reported that Gly, Ser, Asp, ornithine (Orn), and arginine (Arg) are the five most frequently detected amino acids in clouds. These amino acids can react with glyoxal, the simplest and most abundant α-dicarbonyl compound in the atmosphere, contributing to the formation of secondary organic aerosols (SOA). This suggests that amino acids play a crucial role in cloud condensation nuclei (CCN) activation and atmospheric particle transformation. Furthermore, hydroxyl radicals (•OH) significantly influence the degradation of longer-lived amino acids, serving as a key factor in the atmospheric removal of these compounds. Amino acids with higher photodegradation rates act as small but essential sources of rapidly reactive organic species, affecting atmospheric chemistry and particle aging processes.
3.5. Correlation with Amino Acids and Amines
Fig. 2 (a), (c) presents a heatmap illustrating the correlation between particulate matter (PM10, PM2.5) and amino acid classifications (hydrophobic, hydrophilic, and neutral). In addition, the correlation with neutral CAA species—those showing weak associations with PM10—is examined in Fig. 2(b), while the correlation with hydrophilic CAA species—weakly correlated with PM2.5—is presented in Fig. 2(d). Prior to the correlation analysis, outliers were removed using z-scores (±3) to minimize potential distortion of the results. Subsequently, the Shapiro–Wilk test was conducted to assess the normality of each variable. Most variables yielded p-values below 0.05, indicating deviations from a normal distribution. Therefore, to more accurately evaluate potential nonlinear relationships among variables, the non-parametric Spearman rank correlation analysis was employed. In PM10, hydrophilic FAA (0.31) and neutral CAA (0.36) exhibited weak correlations. Similarly, in PM2.5, weak correlations were observed for neutral FAA (0.32), hydrophilic CAA (0.37), and neutral CAA (0.34). These findings suggest that primary pollutants, such as FAA originating from emission sources, are more closely associated with hydrophilic amino acids. In contrast, secondary pollutants, such as CAA, which are formed through processes like photochemical reactions, appear to be more strongly influenced by neutral amino acids in PM10 and hydrophilic amino acids in PM2.5. Mandalakis et al. (2011) [22] reported that CAA exhibited significant contributions from neutral amino acids (49%), followed by hydrophobic (33%) and hydrophilic (18%) amino acids. This is consistent with the present study, supporting the role of neutral amino acids in CAA formation. However, differences in findings may arise due to the exclusion of cysteine (Cys), cystine (Cyst), γ-aminobutyric acid (GABA), hydroxyproline (Hyp), and ornithine (Orn) in the classification of hydrophilic amino acids in Mandalakis et al. (2011) [22], potentially influencing the observed variation. In PM10, the correlation between neutral CAA and neutral FAA was relatively high at 0.40 compared to other factors. In PM2.5, the correlation between hydrophilic CAA and hydrophilic FAA was the highest at 0.53. For specific neutral CAA compounds, a weak correlation was observed between PM10 and proline (Pro) (0.22), as well as between serine (Ser) and Gly (0.31) and Thr (0.36). Among individual hydrophilic CAA compounds, PM2.5 showed weak correlations with aspartic acid (Asp) (0.25), arginine (Arg) (0.21) and cystine(Cys) (0.20). Notably, a weak correlation was observed between Hyp and Lys, Cys and Orn, suggesting their potential role in secondary organic aerosol formation.
4. Conclusions
From November 2022 to January 2023, the average PM10 concentration in Jinju, Gyeongsangnam-do, was 39 μg/m3 (39% of the 24-hour standard of 100 μg/m3), while PM2.5 averaged 27 μg/m3 (76% of the 35 μg/m3 standard). Among anions, SO42−, NO3−, and Cl− were the most abundant, while NH4+, Na+, and K+ were the dominant cations, reflecting emissions from roadways and industrial sources. Concentrations decreased with distance from urban centers.
The average FAA concentrations in PM10 and PM2.5 were 54.8 ng/m3 and 52.7 ng/m3, respectively, while FA averaged 25.9 ng/m3 and 27.8 ng/m3. Hydrophobic amino acids dominated FAA, with a hydrophobic-to-hydrophilic ratio of 1.5, indicating contributions from biologically derived soil components like Ala, Val, Leu, and Ile. In contrast, CAA was primarily composed of hydrophilic amino acids, followed by neutral and hydrophobic types. Key FAA components included Ala, Asn, Gly, Asp, Thr, and Ser, with prominent amines such as PrA, TMA, and DEA. Gly and Ser were abundant across various environments, while Ala, Val, Leu, and Ile were linked to soil organic matter and microbial biomass. For CAA, Asp, Ala, Gly, Asn, and Ser were prevalent, alongside amines like DEA, DMA, and PrA. These amino acids, particularly Gly, Ser, Asp, Orn, and Arg, contribute to secondary organic aerosol (SOA) formation and cloud condensation nuclei activation.
FAA mass concentrations showed no strong correlation with PM but were closely associated with hydrophobic amino acids. In contrast, CAA displayed strong correlations with both hydrophobic and neutral amino acids, emphasizing its role in secondary pollutant formation via photochemical reactions. Within FAA, strong correlations were observed for Ala, Val, Asp, Asn, Gly, and amines like DEA, DMA, and PrA. CAA showed additional correlations with Ile, Leu, Phe (hydrophobic), Glu, Gln (hydrophilic), and Pro, His (neutral), further highlighting its role in PM formation through photochemical synthesis. These findings enhance our understanding of the distribution of organic compounds, ions, amino acids, and amines in Jinju’s particulate matter, emphasizing the contribution of secondary pollutants to PM formation.
PDF Links
PubReader
Full text via DOI
Download Citation 
Supplement
Print 
