Skip to main content
NCBI home page
Environment & Health logoLink to Environment & Health
. 2025 Jan 29;3(5):504–514. doi: 10.1021/envhealth.4c00229

Comprehensive Characterization of Organic Chemicals Associated with Urban Particulate Matter in China

Linlin Yao †,‡, Yunhe Guo ‡,§, Yi Wang †,‡, Junya Li ‡,∥, Jiazheng Sun ⊥, Yanna Liu ‡,*, Jianbo Shi †,‡,⊥,#, Guangbo Qu †,‡,#,*, Guibin Jiang †,‡,∥,#
PMCID: PMC12090013  PMID: 40400554

Abstract

graphic file with name eh4c00229_0006.jpg

Atmospheric particulate matter (PM) is considered a health hazard; however, the inadequate identification of the components of PM limits our understanding of its specific toxic pollutants. Herein, by combining three extraction solvents with different polarities (dichloromethane, hexane, and acetonitrile) and three ionization modes (electron ionization and the positive and negative modes of electrospray ionization), we comprehensively analyzed the organic chemicals in the PM2.5 and PM10 samples collected during summer and winter in Beijing. Suspect screening was facilitated by comparison with the mzCloud and the National Institute of Standards and Technology databases for tentatively characterizing chemical identities. Results showed that more compounds were identified in the winter PM2.5/PM10 samples than in the summer samples and that PM2.5 contained a greater number of chemicals than PM10. Based on peak areas of compounds, the predominant pollutants in the winter PM2.5/PM10 samples were phenols, amines, and aromatic compounds; however, significantly high responses of one phenol and two ester compounds were detected in the summer PM2.5/PM10 samples. Based on the Tox21 toxicological database, a total of 60 identified pollutants were associated with 28 biological targets, and ∼50% of the active compounds were phenolic and aromatic compounds. The biological targets most affected by these pollutants were related to metabolic homeostasis, reproduction, and developmental functions. This study underscores the importance of a multiapproach analysis in comprehensively identifying environmental pollutants and highlights the potential health risks posed by PM.

Keywords: atmospheric particulate matter, extraction solvents, ionization modes, suspect screening, seasonal variation

Introduction

Since the 1990s, studies have delineated the potential toxic effects of atmospheric particulate matter (PM).1−7 PM exposure has been positively correlated with multiple diseases such as asthma, lung cancer, and cardiovascular diseases.8−12 In 2013, the World Health Organization International Agency for Research on Cancer classified PM as carcinogenic.13 Fine particles such as PM10 (PM with diameters of <10 μm) and PM2.5 (PM with diameters of <2.5 μm) are particularly concerning in terms of posing health risks owing to their small sizes, which allow them to be easily deposited in the respiratory system and enter the circulatory system of humans.14 PM is a complex mixture comprising organic (polycyclic aromatic hydrocarbons [PAHs],15 organophosphate esters,16 and aliphatic compounds17) and inorganic (black carbon,18,19 heavy metals, and salts20−22) compounds. Considering the versatile toxicity profile of these chemicals, it is important to comprehensively identify the chemicals present in PM2.5/PM10 to enable better risk assessment.

Studies on the chemical characterization of PM2.5/PM10 have primarily focused on certain chemical classes. Such studies usually apply one extraction solvent (e.g., methanol) and one instrumental analytical mode (e.g., electrospray ionization positive [ESI+]) for analysis. For chemical extraction, methods such as Soxhlet, ultrasonic, and accelerated solvent extraction are common. Furthermore, solvents with different polarities are employed for extracting chemicals with specific properties. For example, dichloromethane (DCM) is widely used for extracting (semi)volatile compounds such as polychlorinated pollutants23 and PAHs24,25 owing to their high volatility. Hexane (HEX), a nonpolar solvent, is frequently used to extract fatty acids, alkanes, and steroids from PM2.5/PM10. Acetonitrile (ACN) and methanol (MeOH), two highly polar solvents, are often used for extracting polar compounds (eg., organophosphite antioxidants and organophosphate esters23,26). Lin et al.27 investigated per- and polyfluoroalkyl substances (PFAS, a large group of man-made chemicals with varying polarity) of PM samples collected from Asian countries following MeOH extraction. Conversely, to extract volatile PFASs from PM, DCM, ethyl acetate (EthA), and acetone (ACE)–petroleum ether have been used.28 To expand chemical coverage, it is common to use combined solvents or sequential extraction with different solvents. In a study conducted on PM2.5 samples collected in Vigo, Spain, the mixture of HEX/ACE (1:1, V/V) were employed to extract 50 pollutants, including PAHs, phthalate esters, and organophosphorus flame retardants.24 Cecinato et al.29 used DCM/ACE/MeOH (5:3:2, V/V/V) to extract 16 pharmaceutical substances including 4 parabens, 4 anti-inflammatory drugs, 3 analgesics, and antipyretics in PM2.5/PM10 samples. Currently, various solvents have been used to extract PM samples, but the choice of these solvents has not been adequately explained, and there is a lack of data to support the knowledge of compounds extracted by different solvents.

Target analysis has been widely employed to detect certain pollutants in PM, such as PFASs,27 PAHs,24 organosulfates,30 imidazoles,31 and halogenated flame retardants.15 When focusing only on known chemicals, the toxic effects caused by unknown pollutants might be overlooked, making it difficult to accurately evaluate the scope of health risks posed by environmental pollution. Therefore, high-resolution mass spectrometry (HRMS) based nontarget analysis (NTA) is being used to identify unknown or new contaminants. Consequently, environmental samples undergo comprehensive extraction during pretreatment, and extraction is not limited to known components. For instance, a HEX/DCM mixture (1:1, V/V) was used to extract organic compounds from PM samples for HRMS-based nontarget analysis, and compounds that previously undetected in the air were identified.32 The selection of the extraction solvent during sample pretreatment would affect the analytical results of pollutants. A better understanding of the extraction performance (e.g., effective extraction structures) of different solutions will benefit the selection of extraction solvents, either single or mixtures, for nontarget analysis of environmental samples.

Therefore, to identify the wide range of organics in PM, we collected PM10 and PM2.5 samples from Beijing during summer and winter; performed parallel extractions using HEX, DCM, and can; and analyzed each extract via liquid chromatography (LC) and gas chromatography (GC) coupled HRMS, respectively, using ESI (ESI+ and ESI– modes) and electron ionization (EI, positive mode). By comparison to mzCloud and the National Institute of Standards and Technology (NIST) databases, contaminants in each extraction fraction and under each detection mode were tentatively identified, demonstrating the complexity of chemicals in PM2.5/PM10. We also studied the differences between the chemical composition of PM2.5 and PM10 and that of PM2.5/PM10 collected during the summer and winter. Our study can serve as a data-based reference for method selection in future PM studies and provides a comprehensive understanding of PM pollutants.

Materials and Methods

Chemicals and Reagents

HEX, ACN, DCM, and MeOH were purchased from Thermo Fisher Scientific (Waltham, MA, USA). Formic acid and ammonium acetate were purchased from Merck (Germany, EU).

Sample Collection

PM2.5 and PM10 samples were collected every 5–7 days in the summer and winter in 2019 using two total suspended particulate high volume air samplers (Laoying 2031, Qingdao, China) placed on the rooftop of a 20m high building in a residential area in urban Beijing, China (116.350044 E, 40.014512 N). The two seasons were chosen because of marked differences in temperature and gaseous emissions that would affect the chemical composition of PM. Sampling date, weather conditions, and PM concentrations were recorded and provided in Table S1 in the Supporting Information (SI). Different size selective inlets were used for each of the two samplers to exclusively allow particulates with diameters of <2.5 and <10 μm to deposit on the filters. Microquartz fiber filters (203 × 254 mm; MK360, Ahlstrom-Munksjo, Falun, Sweden) were used as the collection membrane, and an airflow rate of 1.05 m3 min–1 was used for both samplers for 24 h without interruption. Before collection, each filter sheet was dried (180 °C for 2 h, 550 °C for 6 h, and 50 °C for 24 h), equilibrated in a constant temperature and humidity test chamber for 24 h, and weighed. The membranes were equilibrated again for 24 h after sample collection, weighed, gently wrapped with aluminum foil, and stored at −80 °C. Then, they were placed in both the air samplers for 24 h without initiating the pump after every 4–5 times of sampling to serve as blank membrane samples.

Sample Pretreatment

PM membrane samples collected during July and August, and November and December were, respectively, indicated as “summer” and “winter” samples. Every year, starting from approximately November 10th, Beijing implements citywide heating measures. For each season, multiple membrane samples collected from different days were taken, and 1 or 2 pieces of 1.5 × 1.0 cm2 fragments were cut from each using solvent-rinsed ceramic scissors, resulting in a total of 12 fragments for each size of PM sample (PM2.5 or PM10) per season (Table S1 in Supporting Information, SI). Each of the 12 fragments were subsequently evenly cut into three 0.5 × 1.0 cm2 pieces, which were separately placed into 50 mL centrifuge tubes containing 20 mL of HEX, ACN, and DCM, respectively. The smaller pieces were cut and used to enhance extraction efficiency by increasing the surface area for interaction with the extraction solvent. In total, 12 composite PM2.5/PM10 samples, composing three PM2.5-summer, three PM10-summer, three PM2.5-winter, and three PM10-winter samples, were prepared (Figure 1). One summer blank sample and a winter blank sample were processed using the same method. Ultrasonication was performed to assist chemical extraction due to its accessibility and broad chemical coverage, and ultrasonic extraction with each solvent was repeated twice. For the HEX and ACN extracts, the supernatants were combined and concentrated to 200 μL under gentle nitrogen. For the DCM-extracted samples, the combined supernatants were first concentrated to 2 mL and divided into two 1 mL aliquots, with one being resolubilized in methanol. These extracts were then concentrated to a volume of 200 μL and stored at −80 °C until instrument analysis.

Figure 1.

Figure 1

Open in a new tab

Workflow of sample preparation and analysis.

Instrument Analysis

LC–HRMS Analysis

Instrument analysis was performed using a Quadrupole Exactive Orbitrap HRMS instrument coupled with a Thermo Scientific Vanquish Horizon high-performance (HP)LC system. A Hypersil C18 HPLC reversed-phase column (50 mm length, 2.1 mm internal diameter, and 1.7 μm particle size; Thermo Fisher Scientific) was used for separation. Water (A) and MeOH (B) were used as mobile phase with 0.1% formic acid for positive mode and 2 mM ammonium acetate for negative mode at a flow rate of 0.3 mL min–1. The elution started from 20% methanol, which gradually increased to 50% at 1 min and 100% at 18 min and was maintained for 5 min before returning to 20% B at 23.1 min and maintaining for another 5 min. The Orbitrap analyzer was run in ESI+ and ESI– modes in separate runs. The sheath, aux, and sweep gas flows were 40, 15, and 0 (arbitrary units) in both modes, respectively. Spray voltages were 3.8 and 3.2 kV in the ESI+ and ESI– modes, respectively, while capillary and heater temperatures were 320 and 350 °C, respectively. In the full scan, the automatic gain control was 106 and the maximum injection time was 100 ms, which were 8 × 103 and 50 ms in the MS/MS scan, respectively. The ion isolation window for the precursor ions was 1.5 m/z. Data-dependent acquisition was used for data collection where a full scan was run at a resolving power of 70000 (at m/z 200) over m/z 50–750, while the MS/MS scan was recorded on the top 5 ions at RP 17,500 after applying normalized collision energies of 20, 40, and 60 eV. The injection volume for the MeOH and ACN extracts was 10 μL.

GC–HRMS Analysis

A Quadrupole Exactive Orbitrap HRMS coupled with a 1310 GC and TriPlus RSH autosampler (Thermo Fisher Scientific, Waltham, MA, USA) was used for compound analysis. The compounds were separated using a TraceGOLD TG-5SiLMS semipolar standard column (30 m length, 0.25 mm inner diameter, and 0.25 μm film thickness; Thermo Fisher Scientific). A volume of 1 μL of sample was injected in splitless mode at a constant helium flow of 1.2 mL min–1. The initial temperature of the oven was 70 °C, which was held for 0.5 min, then increased to 80 °C at 10 °C min–1, and then to 320 °C at 6 °C min–1, holding for 8.5 min. The temperature of the transfer line was 280 °C. The EI source was used with 70 eV energy at 310 °C. Mass spectra was performed in full scan range from m/z 50–550 at a resolution of 60000 at m/z 200. A mixed standard solution containing C9–C40 n-alkanes with retention times of 3.30–42.01 min covering the whole analytical time window (0–45 min) was analyzed alongside mask samples for obtaining the Kovats retention index.

Chemical Identification

The obtained LC–HRMS data were processed using Compound Discoverer (CD, version 3.3, Thermo Fisher Scientific, USA) software. The workflow and specific parameter settings are presented in Figure S1 and Table S2 (SI), respectively. A suspect screening was conducted by matching deconvoluted features with the mzCloud database. Only features with matching scores of ≥75 were retained. The GC–HRMS data were processed using Tracefinder (version 5.0, Thermo Fisher Scientific). The full scan data were first deconvoluted into different feature groups (a group of mass spectral peaks with the same retention times and highly similar peak shapes). The boundaries for important parameters in this process were 3, 10000, 95%, and 5 ppm for signal-to-noise threshold, total ion chromatography threshold, ion overlap window, and accurate mass tolerance, respectively.33 The resolved spectra were compared with the NIST spectral library (v2017), and a preliminary filtering was conducted for each feature group to obtain probable candidates for each feature group by filtering using the following parameters: score, ≥ 85; search index, ≥ 750;34 and ΔRI, ≤ 30. If more than one candidate was retained after the preliminary filtering, we would focus on the ones with high score values, and the candidate with the smallest ΔRI value would be assigned to the feature group on the condition that the difference in their score values was within 2. All chemicals were tentatively identified at confidence level 3 (CL = 3), following the criteria described by Schymanski et al.35 The MS/MS spectrum of representative compounds from different classes as shown in Figure S2 (SI).

Toxicity Assessment

The tentatively identified compounds were matched with the Tox21 program database (https://tripod.nih.gov/tox/) to filter chemicals that either activated or deactivated the 60 toxic targets. The activity test data for biological targets of each identified compound were retrieved from the “SAMPLE_DATA_TYPE,” and the test results, including inactive, active agonist, inconclusive agonist, active antagonist, and inconclusive antagonist, were collected from “ASSAY_OUTCOME.” A compound was considered to have an effect on biological targets if at least one of the results (active agonist, inconclusive agonist, active antagonist, and inconclusive antagonist) indicated so. If all biological test results in “ASSAY_OUTCOME” indicated inactivity, then the compound was deemed to have no effect on biological targets.

Quality Control and Quality Assurances

To avoid procedure interference and false positive results, solvent blanks (processed together with PM2.5/PM10 samples) and blank membrane samples were used. The positive and negative modes of the Orbitrap mass spectrometer were calibrated using a mixture of references before injection.

Results and Discussion

Number of Tentatively Identified Compounds via Different Extraction and Analytical Methods

Combing chemicals detected in three different extracts and analyzed under three different ionization modes, a total of 282 compounds were tentatively identified in PM2.5/PM10 samples. When considering the number of identified compounds in different fractions, ACN revealed slightly more compounds than DCM (108 vs 103, under both ESI+ and ESI– modes), while DCM fractions showed an obvious higher number of compounds than HEX (88 vs 39, under the EI mode) (Figure 2). Collectively, when combing chemicals detected in each fraction together, DCM identified the highest number of chemicals (i.e., 51 + 52 + 88 = 144), followed by ACN (i.e., 52 + 56 = 108), and HEX presented the least (i.e., 39). When comparing between different detection modes, the ESI mode presented more compounds (i.e., 101) than EI (i.e., 88) in the DCM fraction (Figure 2). For both DCM and ACN fractions, the number of identifiable chemicals were comparable for both ESI+ (i.e., 51 vs 52) and ESI– (i.e., 52 vs 56) modes. Although the above numbers are affected by factors such as the number of chemicals extracted, the number of chemicals responsive to a specific ionization mode, and the number of chemicals identifiable using available databases, the observed differences clearly demonstrated the differences between applying different extraction solvents and detection modes.

Figure 2.

Figure 2

Open in a new tab

Number of identified compounds in different extraction solvents using ESI and EI sources.

Type of Tentatively Identified Compounds Vary with Extraction and Identification Methods

The identified compounds from the PM2.5/PM10 samples were classified into 16 classes based on the rule of organic functional groups by Norbert Haider (checkmol):36 alcohol, alkane, alkene, amide, amine, aromatic compound, carbonyl compound, carboxylic acid derivative, ester, ether, halogen derivative, ketone, N-heterocyclic compound, O-heterocyclic compound, phenol, and others (Figure 3A and Table S3 and S4 in the SI). Notably, the shown compound was identified from at least one PM2.5/PM10 sample, and there were duplicate compounds among the different methods.

Figure 3.

Figure 3

Open in a new tab

(A) Types of compounds identified and duplicated obtained through different extraction and detection methods. Numbers of identified and duplicate compounds: (B) extracted by DCM and ACN using the ESI ion source; (C) extracted by DCM and HEX using the EI ion source; (D) detected by ESI+ and ESI- modes using ACN as the extraction solvent; (E) detected by ESI+, ESI-, and EI modes using DCM as the extraction solvent.

Using the ion source of ESI, a total of 12 and 13 classes of compounds were extracted using DCM and ACN, respectively (Table S3 and Table S4, SI). Most classes of compounds were extracted using DCM and ACN simultaneously, except for carbonyl compounds, which were exclusively detected in the ACN-extracted fraction. More N-heterocyclic compounds were extracted using ACN (Figure 3A) than DCM. Based on the detection method of EI, the comparison between DCM and HEX showed that DCM extracted more classes of compounds and more compounds in each class than HEX (Table S3 and Table S4, SI). The predominant compounds extracted using the different solvents that were detected with the same ion source were comparable (the top four columns presented in Figure 3A).

Although ESI– detected more compounds than ESI+ in the DCM- and ACN-extracted fractions, the number of compound classes exhibited the opposite trend (Table S4, SI). A total of 10 and 9 classes of compounds were detected in the DCM-extracted fractions via ESI+ and ESI–, respectively, while 12 and 8 classes were detected in the ACN-extracted fractions via ESI+ and ESI–, respectively. In addition, based on the extraction method of DCM, the ion source of EI detected more classes of compounds than either ESI+ or ESI–; however, it detected the same number of compound classes as ESI. The lower five columns presented in Figure 3A show that the predominantly detected compound classes via ESI+, ESI–, and EI are different. Amine and carboxylic acid derivatives were predominantly identified via ESI+ and ESI–, respectively, while aromatic compounds were predominantly detected via EI. Overall, in terms of compound classes, the ion source and detection mode exhibited a greater influence than the extraction solutions on the number and type of compound classes.

Duplicate compounds were identified between the different extraction and detection methods, and the number of overlapped compounds extracted using DCM and ACN during detection via ESI was 45 (Figure 3B). Most of these compounds were phenol (12 compounds), followed by carboxylic acid derivative (8 compounds), amine (6 compounds), and the other 8 classes. By contrast, there were only fibe compounds simultaneously extracted via DCM and HEX under EI detection (Figure 3C), and four of them were aromatic compounds. In terms of the different detection methods, only two compounds were duplicated between the detection modes of ESI+ and ESI–, a phenol and a N-heterocyclic compound (Figure 3D and 3E). One compound belonging to the ester class was detected under the ESI+ and EI modes (Figure 3E).

Overall, the compound information that can be obtained from environmental samples mostly depends on the extraction method and detection capability. Compared with the single extraction solvent, a mixed extraction solution is usually used during sample pretreatment to collect as many compounds as possible. Besides, the limited overlap of compounds identified by three analytical methods suggested a multiapproach analysis when applying NTA strategy. Qiao et al.37 identified over eight hundred compounds in air samples after extraction with mixed solvent HEX/DCM (1:1, V/V) and analysis with two-dimensional GC with time-of-flight mass spectrometry. In a previous study, groundwater samples were analyzed by the combined method of GC–MS and LC–HRMS to screen for potential organic micropollutants from more than 1300 compounds, and 233 pollutants were finally identified and determined.38

Comparison of Extraction Efficiency and Detection Response of Overlapped Compounds

The difference (fold change) between the methods was calculated based on the response values of the mass spectrum (Figure 4). The duplicate compounds between the two methods were used for the calculation here, and the response values were obtained by the detection of the same PM2.5/PM10 sample.

Figure 4.

Figure 4

Open in a new tab

Difference in response values on the mass spectrum of duplicate compounds between: (A) different detection modes using the same extraction solvent; (B) extraction solvents DCM and HEX using the EI detection mode; (C) extraction solvents DCM and ACN using the ESI+ detection mode; and (D) using the ESI- detection mode (each column represents a single compound).

To evaluate this ability, we selected the compounds that can be detected using the two ion sources or ion modes and calculated the fold changes in the response values. According to the results showed in Figure 4A, there are two zwitterions (4-Methylumbelliferone (CAS: 90–33–5) and 4-Indolecarbaldehyde (CAS: 1074–86–8)) that were detected through the positive as well as negative mode of ESI (fold change calculated by ESI+/ ESI−), while a compound belonging to the class of N-heterocyclic compound exhibited considerably higher response under ESI+ (fold change = 2.1 and 2.5 under extraction via DCM and ACN, respectively), and another phenol compound was detected to have higher response under ESI– (fold change = 0.2 and 0.6 under extraction via DCM and ACN, respectively). One ester compound (1,2-Benzenedicarboxylic acid, bis(2-methylpropyl) ester, CAS: 84–69–5) was detected through both the ESI+ and EI following the extraction of DCM, and its response was higher on EI with a fold change of ESI+/EI = 0.5. Overall, the detection of the same compound is different with different types of ion sources. Of the 282 compounds tentatively identified in our study, there are other zwitterions, such as 8-Hydroxyquinoline, 6-Methyl-2-pyridinemethanol, and 3-Methylhippuric acid, however, they were only detected under the single mode of ESI+ or ESI- (Table S3, SI). The extraction solvents, matrix effects, compound concentration, and mass response may be responsible for their detection only in a specific ion mode.

To evaluate the extraction efficiency of the different solutions during pretreatment, we calculate the response value of compounds that were obtained from the same ion source. When detecting with the ions source of EI, the response of an ester compound (1,2-Benzenedicarboxylic acid, bis(2-methylpropyl) ester, CAS: 84–69–5) was especially higher when extracted using DCM than HEX, with the value of the fold change of DCM/HEX being >300, indicating a highly efficient extraction of this compound via DCM, while another four aromatic compounds were more efficiently extracted via HEX (the value of fold change was from 0.06 to 0.6; Figure 4B). Using detection mode of ESI+, the extraction efficiencies between DCM and ACN were compared. Among the 20 compounds that were simultaneously extracted using two solutions, 6 were better extracted via DCM, showing higher responses on chromatograms (fold change of DCM/ACN was >1), while the others were better extracted via ACN (fold change of DCM/CAN was <1; Figure 4C). Under the detection mode of ESI–, the response values of 12 compounds were higher following DCM extraction (fold change of DCM/ACN was >1), and that of the other 10 compounds were higher following ACN extraction (fold change of DCM/CAN was <1; Figure 4D). Among the 42 compounds that were extracted via DCM and ACN (detected using ESI+ and ESI−), the fold change values of 32 compounds ranged between 0.6 and 1.4.

According to the results of this study, the extraction efficiencies of the same compound via DCM and ACN were comparable; however, there was significant difference between the extraction efficiencies of DCM and HEX, by considering the deviation between values of fold change and 1. When focusing on the several compounds in the same class (i.e., several columns grouped within one type of compound in Figure 4), the fold change values were not consistently >1 or <1, suggesting that the most efficient extraction solution for compounds in one class can be different.

However, it is important to note that, while ultrasonication is considered an exhaustive extraction method, minimizing significant chemical loss, matrix effects arising from varied coeluting chemicals across different samples can also influence detection responses. However, these matrix effects are both chemically and sample-specific. Following the preliminary chemical identification as performed in this study, evaluating matrix effects by spiking a wide range of chemical standards into each sample would provide a more accurate assessment of matrix effects and enable precise concentration measurements, thus facilitating more accurate method selection.

Difference in the Tentatively Identified Compounds Between the PM2.5/PM10 Samples

The PM2.5 and PM10 samples collected in the summer and winter were pretreated and analyzed using different methods in this study to identify their constituent pollutants. Different pretreatment and detection methods identified more compounds in the winter PM2.5/PM10 samples than those in the summer PM2.5/PM10 samples, while more compounds were identified in the PM2.5 samples than those in the PM10 samples (Table S5 and Table S6, SI). The total number of compounds identified in PM2.5/PM10 samples from summer and winter is 181 and 254 respectively. Excluding duplicate compounds between samples from the two seasons, there are 28 and 101 compounds specifically identified in the summer and winter, respectively. Overall, 139 and 147 compounds were detected in the summer PM10 and PM2.5 samples, while 225 and 241 compounds were detected in the winter PM10 and PM2.5 samples, respectively. Even when considering the potential differences in the compounds identified using the various methods, the results showed that more compounds were detected in the winter samples than in the summer samples (Figure S3, SI). However, the diversity of the extraction efficiency and the mass response of the different compounds resulted in a difference in the total response on the chromatograms of the PM2.5/PM10 samples (Figure 5).

Figure 5.

Figure 5

Open in a new tab

Comprehensive analysis of PM10 and PM2.5 samples collected in the summer and winter through different combinations of extraction and ionization methods.

Except for the pretreatment and detection method of DCM–EI, the compound response on the chromatograms of the winter PM2.5/PM10 samples was higher than that of the summer PM2.5/PM10 samples (Figure 5). Based on the same extraction and pretreatment method, the phenol and amine compounds (DCM–ESI and ACN–ESI) and aromatic compounds (HEX–EI) were significantly detected in the winter PM2.5/PM10 samples compared with that in the summer PM2.5/PM10 samples, indicating the special occurrence of these compounds in PM in the winter. The number of identified compounds belonging to these three classes of compounds in the winter PM samples was more than that in the summer PM samples. The peak areas of the simultaneously detected compounds belonging to the phenol and aromatic compound classes were larger in the winter PM2.5/PM10 samples than in the summer PM2.5/PM10 samples, indicating a higher abundance of such compounds in winter PM. In terms of the amine compounds, the chromatogram response of only one simultaneously identified pollutant (CAS: 52080–58–7) was higher in the summer PM2.5/PM10 samples compared with that in the winter PM2.5/PM10 samples, and the total peak area of this class of pollutants was still smaller in the summer PM2.5/PM10 samples, suggesting variations in the distribution levels of different compounds belonging to the same class.

Phenol compounds were also detected in the summer PM2.5 sample through the method of DCM–EI with a significantly high response on chromatograms. It was predominately contributed by the compound of 2,5-ditert-butylphenol (CAS: 5875–45–6), which was not identified in the same sample using the other three methods. Except for this phenol compound, the higher response of summer PM2.5/PM10 samples than that of winter PM2.5/PM10 samples obtained by the method of DCM–EI mostly originated from an ester compound (phthalic acid, 2-chloropropyl propyl ester). It was only detected in the summer PM2.5/PM10 samples and using DCM–EI and accounted for 80% and 78% of the chromatogram response of all the identified ester compounds in the samples of S_PM10 and S_PM2.5, respectively (Figure 5). In addition to the two compounds, we found that the response of 1,2-benzenedicarboxylic acid, bis(2-methylpropyl) ester (in the class of ester), obtained via DCM–EI was present in an order of magnitude higher in the summer PM2.5/PM10 samples than in the winter PM2.5/PM10 samples. When the three compounds (one phenol compound of 2,5-ditert-butylphenol and two ester compounds of phthalic acid, 2-chloropropyl propyl ester, and 1,2-benzenedicarboxylic acid, bis(2-methylpropyl) ester) are excluded, the total peak area of the summer PM2.5/PM10 samples was smaller than that of the winter PM2.5/PM10 samples, consistent with the tendency revealed by the other three methods.

Although 1,2-benzenedicarboxylic acid, bis(2-methylpropyl) ester, was also detected using DCM–ESI and HEX–EI, DCM–EI showed the highest response on the chromatograms. Results in Figure 4A and B (the one ester compound in the green column indicates 1,2-benzenedicarboxylic acid, bis(2-methylpropyl) ester) also indicate that the efficiency of extraction and ionization was higher using DCM and EI, respectively, than when using HEX and ESI. In addition, only the combination of DCM extraction and EI detection identified the two compound of phthalic acid, 2-chloropropyl propyl ester and 2,5-ditert-butylphenol, which were the major reason underlying the higher response in the summer PM2.5/PM10 samples than that in the winter PM2.5/PM10 samples. Notably, a higher response on the chromatograms does not always represent a higher abundance, owing to the variation of ionization intensity of the different compounds. Nonetheless, the study results indicate the selectivity of the pretreatment extraction solution and detection ion source in extracting and identifying certain compounds and the importance of utilizing different methods for the comprehensive analysis of environmental samples with no target.

The organic pollutants present in the PM2.5/PM10 samples have previously been reported using the NTA method; however, the identified compounds were limited, owing to the single selection of the pretreatment method. The pollutants alkylbenzenes, cycloalkanes, biphenyls, PAHs, per- and polyfluoroalkyl substances, phenols, alcohols, amines, ketones, phthalates, and esters have been previously detected in PM2.5/PM10 samples by different studies.39−44 Herein, more than 15 classes of pollutants were identified in the PM2.5/PM10 samples based on the multiple pretreatment and analysis methods. Furthermore, some pollutants with extremely high chromatogram responses (e.g., the phenol compound of 2,5-ditert-butylphenol, ester compound of phthalic acid, 2-chloropropyl propyl ester, and 1,2-benzenedicarboxylic acid, bis(2-methylpropyl) ester) were screened only through the specific combination of extraction solvent and analytical source. These results further demonstrate the potential to overlook pollutants, either known or unknown, when applying a single method to the NTA of environmental samples.

In summary, the compounds identified in PM2.5 were more than those in PM10 obtained from both summer and winter, possibly due to the smaller size of PM2.5, since finer particles have a larger surface area relative to volume, allowing more compounds to be adsorbed and more active atmospheric reactions to occur. Although the response of the compounds simultaneously detected in the PM2.5/PM10 samples was not always higher in winter based on the comparison of the MS response of the compounds simultaneously identified in the summer and winter PM2.5/PM10 samples, the number of identified compounds in the winter PM2.5/PM10 samples were more than that in the summer PM2.5/PM10 samples. It can be concluded that PM2.5/PM10 pollution is more serious during winter owing to more heating and less wet precipitation in China.45 A previous study reported that more PM2.5 was collected in winter than in summer over the same sampling period,46 which may indicate a greater accumulation of air pollutants on winter PM2.5/PM10 samples than summer PM2.5/PM10 samples. Combustion heating activity in winter was thought to be a major contributor to the severe accumulation of pollutants in PM2.5/PM10, and previous studies mostly reported PAHs pollution. For example, it was found that the contamination of aromatic compound PAHs in PM2.5/PM10 was more serious in winter than in summer24 and was highly associated with the coal combustion for heating during winter in Beijing.15 In this study, using multiple extraction and NTA detection methods, 73 more compounds were specifically identified in the winter than in the summer. In addition to aromatic compound, most of these compounds were carboxylic acid derivative, N-heterocyclic compound, and phenol (Table S6, SI), providing more compound information for tracing potential pollutant sources in future studies. It is worth noting that several compounds were identified exclusively or predominantly in summer PM2.5/PM10 samples in this study, further emphasizing sources of air pollution other than seasonal coal-fired processes. For example, 2,5-ditert-butylphenol is usually used as a stabilizer in the polymer industry to extend the material usage, and 1,2-benzenedicarboxylic acid, bis(2-methylpropyl) ester, is widely used as a plasticizer in the plastics, rubber, and paint industries.

Potential Toxicity Induced by Compounds Tentatively Identified in the PM2.5/PM10 Samples

The potential toxic effects of the identified pollutants in the PM2.5/PM10 samples were determined according to public data provided by the Toxicology in the 21st Century (Tox21) Consortium. Among the 282 identified compounds, a total of 60 compounds were screened to be active agonists or antagonists of the 28 toxic targets (Table S7, SI). These compounds, which could potentially induce biological toxic effects, were mainly aromatic (14) and phenolic (12) compounds, accounting for 43% of the total toxic compounds, and were the predominant pollutants, both in number and abundance, in the winter PM2.5/PM10 samples compared with that in the summer PM2.5/PM10 samples (Figure S4, SI). Benzo[a]pyrene (CAS: 50–32–8) and benzo[k]fluoranthene (CAS: 207–08–9) belonging to the class of aromatic compounds affected the most targets (17) among the 28 detected biological targets (Table S8, SI), suggesting that special attention should be paid to these compounds when analyzing PM2.5/PM10 samples. Particulate PAHs have been reported to be significantly associated with PM2.5-induced inflammatory response and oxidative stress, implying their contribution to the PM related health effects.5 Comparing the number of active compounds among 28 targets, more compounds could induce activating or inhibiting effects on the AR-MDA antagonist (tox21-ar-mda-kb2-antagonist-p2), mitochondrial toxicity (tox21-mitotox-p1), AhR (tox21-ahr-p1), and RAR antagonist (tox21-rar-antagonist-p2), which are associated with metabolic homeostasis, reproduction, and developmental functions. The activation of AhR was considered a key event that triggered the suppression of cardiomyocyte differentiation, subsequently leading to adverse effects on cardiac structure and function.47

Although PM can affect health, the relation between the pollutants in PM and adverse health outcomes is poorly understood, with the main limitation being that the unknown or toxic pollutants have yet to be adequately identified. Effect-direct analysis for screening toxic components has recently been developed to promote environmental health research. It relies on nontarget analysis to identify specific compounds from the toxic component. Thus, multiple pretreatment and analysis methods are necessary to enhance the possibility of identifying unknown toxic pollutants.

Conclusions

Herein, different pretreatment solvents and analytical methods were used to comprehensively identify the constituent chemicals of PM2.5 and PM10 samples collected in the summer and winter in Beijing. A total of 282 compounds were tentatively identified in the PM2.5/PM10 samples. More compounds were identified in winter PM2.5/PM10 than in summer, and in PM2.5 samples than in PM10 samples. Phenol, amine, and aromatic compounds were predominantly detected in the winter PM2.5/PM10 samples, and 26 of the 60 compounds that exert toxic effects on selected biological targets based on cell experiments belonged to phenol and aromatic classes. One phenol compound (i.e., 2,5-ditert-butylphenol) and two esters (i.e., phthalic acid, 2-chloropropyl propyl ester and 1,2-benzenedicarboxylic acid, bis(2-methylpropyl) ester) were detected in significantly higher abundance in the summer PM2.5/PM10 samples than in the winter PM2.5/PM10 samples, providing a potential clue for future research on tracing the sources of PM2.5/PM10 pollution in the summer. The identified compounds in the same PM2.5/PM10 samples varied highly when different extraction solvents and analytical ion sources were used, suggesting that a single pretreatment and analytical method during an NTA study were insufficient for obtaining a comprehensive picture of the pollutants comprising the environmental sample. The compounds reported in this study were all tentatively identified, without quantitative data using high-purity standards at this stage, which may limit the understanding of the occurrence and characteristics of contaminants in PM. However, we provide a preliminary evaluation of the comprehensive chemical coverage and explores differences across extraction methods, detection techniques, and seasons. This informs the future selection and combination of pretreatment and detection methods, as well as the choice of seasonal PM samples for various research purposes.

Acknowledgments

This work was supported by the National Key Research and Development Program of China (Grant No. 2023YFA0915100), the National Natural Science Foundation of China (Grant No. 22325606, 22193050, 42377386, 21906166, 21527901), the Scientific Instrument Developing Project of the Chinese Academy of Sciences (Grant No. GSZXKYZB2024012), and Strategy Priority Research Program of the Chinese Academy of Sciences (Grant No. XDB0750100, XDB0750300).

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/envhealth.4c00229.

  • Details on analytical workflow (Figure S1), MS/MS spectrum of representative compounds (Figure S2), three-dimensional scatter plots of tentatively identified compounds (Figure S3), hierarchical cluster analysis results (Figure S4), parameter setting of Compound Discoverer (Table S2), number of identified compounds (Table S4 and Table S5), biological targets and effective compounds (Table S7 and Table S8) (PDF)

  • PM10 and PM2.5 membrane samples and weather conditions (Table S1), identified compounds in PM10 and PM2.5 samples (Table S3 and Table S6) (XLSX)

The authors declare no competing financial interest.

Supplementary Material

eh4c00229_si_001.pdf (1.7MB, pdf)
eh4c00229_si_002.xlsx (55.6KB, xlsx)

References

  1. Jin L.; Xie J.; Wong C. K. C.; Chan S. K. Y.; Abbaszade G.; Schnelle-Kreis J.; Zimmermann R.; Li J.; Zhang G.; Fu P.; Li X. Contributions of City-Specific Fine Particulate Matter (PM(2.5)) to Differential in Vitro Oxidative Stress and Toxicity Implications between Beijing and Guangzhou of China. Environ. Sci. Technol. 2019, 53 (5), 2881–2891. 10.1021/acs.est.9b00449. [DOI] [PubMed] [Google Scholar]
  2. Salgueiro-González N.; López de Alda M. J.; Muniategui-Lorenzo S.; Prada-Rodríguez D.; Barceló D. Analysis and Occurrence of Endocrine-Disrupting Chemicals in Airborne Particles. TrAC-Trend. Anal. Chem. 2015, 66, 45–52. 10.1016/j.trac.2014.11.006. [DOI] [Google Scholar]
  3. Wang Y.; Xiong L.; Tang M. Toxicity of Inhaled Particulate Matter on the Central Nervous System: Neuroinflammation, Neuropsychological Effects and Neurodegenerative Disease. J. Appl. Toxicol. 2017, 37 (6), 644–667. 10.1002/jat.3451. [DOI] [PubMed] [Google Scholar]
  4. You R.; Ho Y. S.; Chang R. C. The Pathogenic Effects of Particulate Matter on Neurodegeneration: A Review. J. Biomed. Sci. 2022, 29 (1), 15. 10.1186/s12929-022-00799-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Zhou Q.; Chen J.; Zhang J.; Zhou F.; Zhao J.; Wei X.; Zheng K.; Wu J.; Li B.; Pan B. Toxicity and Endocrine-Disrupting Potential of PM(2.5): Association with Particulate Polycyclic Aromatic Hydrocarbons, Phthalate Esters, and Heavy Metals. Environ. Pollut. 2022, 292, 118349. 10.1016/j.envpol.2021.118349. [DOI] [PubMed] [Google Scholar]
  6. Jia Z. L.; Zhu C. Y.; Rajendran R. S.; Xia Q.; Liu K. C.; Zhang Y. Impact of Airborne Total Suspended Particles (TSP) and Fine Particulate Matter (PM(2.5))-Induced Developmental Toxicity in Zebrafish (Danio rerio) Embryos. J. Appl. Toxicol. 2022, 42 (10), 1585–1602. 10.1002/jat.4325. [DOI] [PubMed] [Google Scholar]
  7. Zhang Y.; Li S.; Li J.; Han L.; He Q.; Wang R.; Wang X.; Liu K. Developmental Toxicity Induced by PM2.5 Through Endoplasmic Reticulum Stress and Autophagy Pathway in Zebrafish Embryos. Chemosphere 2018, 197, 611–621. 10.1016/j.chemosphere.2018.01.092. [DOI] [PubMed] [Google Scholar]
  8. Kim K. H.; Kabir E.; Kabir S. A Review on the Human Health Impact of Airborne Particulate Matter. Environ. Int. 2015, 74, 136–143. 10.1016/j.envint.2014.10.005. [DOI] [PubMed] [Google Scholar]
  9. Pražnikar Z.; Pražnikar J. The Effects of Particulate Matter Air Pollution on Respiratory Health and on the Cardiovascular System. Slovenian J. Public Health 2012, 51 (3), 190–199. 10.2478/v10152-012-0022-z. [DOI] [Google Scholar]
  10. Saravia J.; Lee G. I.; Lomnicki S.; Dellinger B.; Cormier S. A. Particulate Matter Containing Environmentally Persistent Free Radicals and Adverse Infant Respiratory Health Effects: A Review. J. Biochem. Mol. Toxicol. 2013, 27 (1), 56–68. 10.1002/jbt.21465. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Sun H. Y.; Chen X.; Huang W. Z.; Wei J.; Yang X. L.; Shan A. Q.; Zhang L. W.; Zhang H. L.; He J. Y.; Pan C. J.; Li J. J.; Wu J.; Wang T.; Chen J.; Guo Y. M.; Tong S. L.; Dong G. H.; Tang N. J. Association between Long-Term Exposure to PM2.5 Inorganic Chemical Compositions and Cardiopulmonary Mortality: A 22-Year Cohort Study in Northern China. Environment & Health 2024, 2, 530–540. 10.1021/envhealth.4c00020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Chen Y.; Ngan H. L.; Song Y. Y.; Qi Z. H.; Zhao L. F.; Dong C.; Li R. J.; Li Y. B.; Yang Z.; Cai Z. W. Chronic Real-Ambient PM2.5 Exposure Exacerbates Cardiovascular Risk via Amplifying Liver Injury in Mice Fed with a High-Fat and High-Cholesterol Diet. Environment & Health 2024, 2, 221–232. 10.1021/envhealth.3c00168. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. New WHO Global Air Quality Guidelines Aim to Save Millions of Lives From Air Pollution. https://www.who.int/news/item/22-09-2021-new-who-global-air-quality-guidelines-aim-to-save-millions-of-lives-from-air-pollution (accessed 2024–08–20).
  14. Organization, G. W. H. WHO Global Air Quality Guidelines. Particulate Matter (PM2.5 and PM10), Ozone, Nitrogen Dioxide, Sulfur Dioxide and Carbon Monoxide, 2021. [PubMed] [Google Scholar]
  15. Wang T.; Tian M.; Ding N.; Yan X.; Chen S. J.; Mo Y. Z.; Yang W. Q.; Bi X. H.; Wang X. M.; Mai B. X. Semivolatile Organic Compounds (SOCs) in Fine Particulate Matter (PM(2.5)) During Clear, Fog, and Haze Episodes in Winter in Beijing, China. Environ. Sci. Technol. 2018, 52 (9), 5199–5207. 10.1021/acs.est.7b06650. [DOI] [PubMed] [Google Scholar]
  16. Yang F.; Ding J.; Huang W.; Xie W.; Liu W. Particle Size-Specific Distributions and Preliminary Exposure Assessments of Organophosphate Flame Retardants in Office Air Particulate Matter. Environ. Sci. Technol. 2014, 48 (1), 63–70. 10.1021/es403186z. [DOI] [PubMed] [Google Scholar]
  17. Gonçalves C.; Figueiredo B. R.; Alves C. A.; Cardoso A. A.; da Silva R.; Kanzawa S. H.; Vicente A. M. Chemical Characterisation of Total Suspended Particulate Matter From A Remote Area in Amazonia. Atmos. Res. 2016, 182, 102–113. 10.1016/j.atmosres.2016.07.027. [DOI] [Google Scholar]
  18. Crenn V.; Fronval I.; Petitprez D.; Riffault V. Fine Particles Sampled at An Urban Background Site and An Industrialized Coastal Site In Northern France - Part 1: Seasonal Variations and Chemical Characterization. Sci. Total Environ. 2017, 578, 203–218. 10.1016/j.scitotenv.2015.11.165. [DOI] [PubMed] [Google Scholar]
  19. Ljungman P. L. S.; Andersson N.; Stockfelt L.; Andersson E. M.; Nilsson Sommar J.; Eneroth K.; Gidhagen L.; Johansson C.; Lager A.; Leander K.; Molnar P.; Pedersen N. L.; Rizzuto D.; Rosengren A.; Segersson D.; Wennberg P.; Barregard L.; Forsberg B.; Sallsten G.; Bellander T.; Pershagen G. Long-Term Exposure to Particulate Air Pollution, Black Carbon, and Their Source Components in Relation to Ischemic Heart Disease and Stroke. Environ. Health Perspect. 2019, 127 (10), 107012. 10.1289/EHP4757. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Maring T.; Kumar S.; Jha A. K.; Kumar N.; Pandey S. P. Airborne Particulate Matter and Associated Heavy Metals: A Review. Macromol. Symp. 2023, 407 (1), 202100487. 10.1002/masy.202100487. [DOI] [Google Scholar]
  21. Rabha S.; Subramanyam K. S. V.; Sawant S. S.; Saikia B. K. Rare-Earth Elements and Heavy Metals in Atmospheric Particulate Matter in An Urban Area. ACS Earth Space Chem. 2022, 6 (7), 1725–1732. 10.1021/acsearthspacechem.2c00009. [DOI] [Google Scholar]
  22. Soleimani M.; Amini N.; Sadeghian B.; Wang D.; Fang L. Heavy Metals and Their Source Identification in Particulate Matter (PM(2.5)) in Isfahan City, Iran. J. Environ. Sci. (China) 2018, 72, 166–175. 10.1016/j.jes.2018.01.002. [DOI] [PubMed] [Google Scholar]
  23. Chen Y. J.; Zhang Y.; Chen Y.; Lu Y.; Li R.; Dong C.; Qi Z.; Liu G.; Chen Z. F.; Cai Z. GC-MS/MS Analysis for Source Identification of Emerging POPs in PM(2.5). Ecotoxicol. Environ. Saf. 2020, 193, 110368. 10.1016/j.ecoenv.2020.110368. [DOI] [PubMed] [Google Scholar]
  24. Lara S.; Villanueva F.; Martin P.; Salgado S.; Moreno A.; Sanchez-Verdu P. Investigation of PAHs, Nitrated PAHs and Oxygenated PAHs in PM(10) Urban Aerosols. A Comprehensive Data Analysis. Chemosphere 2022, 294, 133745. 10.1016/j.chemosphere.2022.133745. [DOI] [PubMed] [Google Scholar]
  25. Zhang J.; Yang L.; Mellouki A.; Chen J.; Chen X.; Gao Y.; Jiang P.; Li Y.; Yu H.; Wang W. Atmospheric PAHs, NPAHs, and OPAHs at An Urban, Mountainous, and Marine Sites in Northern China: Molecular Composition, Sources, and Ageing. Atmos. Environ. 2018, 173, 256–264. 10.1016/j.atmosenv.2017.11.002. [DOI] [Google Scholar]
  26. Liu R.; Mabury S. A. Unexpectedly High Concentrations of A Newly Identified Organophosphate Ester, Tris(2,4-di- tert-butylphenyl) Phosphate, in Indoor Dust from Canada. Environ. Sci. Technol. 2018, 52 (17), 9677–9683. 10.1021/acs.est.8b03061. [DOI] [PubMed] [Google Scholar]
  27. Lin H.; Taniyasu S.; Yamazaki E.; Wei S.; Wang X.; Gai N.; Kim J. H.; Eun H.; Lam P. K. S.; Yamashita N. Per- and Polyfluoroalkyl Substances in the Air Particles of Asia: Levels, Seasonality, and Size-Dependent Distribution. Environ. Sci. Technol. 2020, 54 (22), 14182–14191. 10.1021/acs.est.0c03387. [DOI] [PubMed] [Google Scholar]
  28. Wang Q.; Ruan Y.; Lin H.; Lam P. K. S. Review on Perfluoroalkyl and Polyfluoroalkyl Substances (PFASs) in the Chinese Atmospheric Environment. Sci. Total Environ. 2020, 737, 139804. 10.1016/j.scitotenv.2020.139804. [DOI] [PubMed] [Google Scholar]
  29. Cecinato A.; Romagnoli P.; Perilli M.; Balducci C. Pharmaceutical Substances in Ambient Particulates: A Preliminary Assessment. Chemosphere 2017, 183, 62–68. 10.1016/j.chemosphere.2017.05.100. [DOI] [PubMed] [Google Scholar]
  30. Wang K.; Zhang Y.; Huang R. J.; Wang M.; Ni H.; Kampf C. J.; Cheng Y.; Bilde M.; Glasius M.; Hoffmann T. Molecular Characterization and Source Identification of Atmospheric Particulate Organosulfates Using Ultrahigh Resolution Mass Spectrometry. Environ. Sci. Technol. 2019, 53 (11), 6192–6202. 10.1021/acs.est.9b02628. [DOI] [PubMed] [Google Scholar]
  31. Gao K.; Zhang Y.; Liu Y.; Yang M.; Zhu T. Screening of Imidazoles in Atmospheric Aerosol Particles Using A Hybrid Targeted and Untargeted Method Based on Ultra-Performance Liquid Chromatography-Quadrupole Time-of-Flight Mass Spectrometry. Anal. Chim. Acta 2021, 1163, 338516. 10.1016/j.aca.2021.338516. [DOI] [PubMed] [Google Scholar]
  32. Qiao L.; Gao L.; Liu Y.; Huang D.; Li D.; Zheng M. Recognition and Health Impacts of Organic Pollutants with Significantly Different Proportions in the Gas Phase and Size-Fractionated Particulate Phase in Ambient Air. Environ. Sci. Technol. 2022, 56 (11), 7153–7162. 10.1021/acs.est.1c08829. [DOI] [PubMed] [Google Scholar]
  33. Wise C. F.; Hammel S. C.; Herkert N. J.; Ospina M.; Calafat A. M.; Breen M.; Stapleton H. M. Comparative Assessment of Pesticide Exposures in Domestic Dogs and Their Owners Using Silicone Passive Samplers and Biomonitoring. Environ. Sci. Technol. 2022, 56 (2), 1149–1161. 10.1021/acs.est.1c06819. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Travis S. C.; Kordas K.; Aga D. S. Optimized Workflow for Unknown Screening Using Gas Chromatography High-Resolution Mass Spectrometry Expands Identification of Contaminants In Silicone Personal Passive Samplers. Rapid Commun. Mass Spectrom. 2021, 35 (8), e9048 10.1002/rcm.9048. [DOI] [PubMed] [Google Scholar]
  35. Schymanski E. L.; Jeon J.; Gulde R.; Fenner K.; Ruff M.; Singer H. P.; Hollender J. Identifying Small Molecules via High Resolution Mass Spectrometry: Communicating Confidence. Environ. Sci. Technol. 2014, 48 (4), 2097–2098. 10.1021/es5002105. [DOI] [PubMed] [Google Scholar]
  36. checkmol/matchmol Homepage. https://homepage.univie.ac.at/norbert.haider/cheminf/cmmm.html#howtoobtain (accessed 2024–09–05).
  37. Qiao L.; Gao L.; Huang D.; Liu Y.; Xu C.; Li D.; Zheng M. Screening of ToxCast Chemicals Responsible for Human Adverse Outcomes with Exposure to Ambient Air. Environ. Sci. Technol. 2022, 56 (11), 7288–7297. 10.1021/acs.est.1c06890. [DOI] [PubMed] [Google Scholar]
  38. Li X.; Tian T.; Shang X.; Zhang R.; Xie H.; Wang X.; Wang H.; Xie Q.; Chen J.; Kadokami K. Occurrence and Health Risks of Organic Micro-Pollutants and Metals in Groundwater of Chinese Rural Areas. Environ. Health Perspect. 2020, 128 (10), 107010. 10.1289/EHP6483. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Lin Y.; Yang J.; Fu Q.; Ruan T.; Jiang G. Exploring the Occurrence and Temporal Variation of ToxCast Chemicals in Fine Particulate Matter Using Suspect Screening Strategy. Environ. Sci. Technol. 2019, 53 (10), 5687–5696. 10.1021/acs.est.9b01197. [DOI] [PubMed] [Google Scholar]
  40. Ning C.; Gao Y.; Yu H.; Zhang H.; Geng N.; Cao R.; Chen J. FT-ICR Mass Spectrometry for Molecular Characterization of Water-Insoluble Organic Compounds in Winter Atmospheric Fine Particulate Matters. J. Environ. Sci. (China) 2022, 111, 51–60. 10.1016/j.jes.2020.12.017. [DOI] [PubMed] [Google Scholar]
  41. Wan Y. B.; Xing C.; Wang X. Y.; Yang Z. M.; Huang X. P.; Ge X. L.; Du L.; Wang Q. Q.; Yu H. Nontarget Tandem High-Resolution Mass Spectrometry Analysis of Functionalized Organic Compounds in Atmospherically Relevant Samples. Environ. Sci. Tech Let 2022, 9 (12), 1022–1029. 10.1021/acs.estlett.2c00788. [DOI] [Google Scholar]
  42. Wang X. K.; Hayeck N.; Brüggemann M.; Yao L.; Chen H. F.; Zhang C.; Emmelin C.; Chen J. M.; George C.; Wang L. Chemical Characteristics of Organic Aerosols in Shanghai: A Study by Ultrahigh-Performance Liquid Chromatography Coupled With Orbitrap Mass Spectrometry. J. Geophys Res-Atmos 2017, 122 (21), 11703–11722. 10.1002/2017JD026930. [DOI] [Google Scholar]
  43. Xu C.; Gao L.; Lyu Y.; Qiao L.; Huang D.; Liu Y.; Li D.; Zheng M. Molecular Characteristics, Sources and Environmental Risk of Aromatic Compounds in Particulate Matter During COVID-2019: Nontarget Screening by Ultra-High Resolution Mass Spectrometry and Comprehensive Two-Dimensional Gas Chromatography. Environ. Int. 2022, 167, 107421. 10.1016/j.envint.2022.107421. [DOI] [PubMed] [Google Scholar]
  44. Yu N.; Guo H.; Yang J.; Jin L.; Wang X.; Shi W.; Zhang X.; Yu H.; Wei S. Non-Target and Suspect Screening of Per- and Polyfluoroalkyl Substances in Airborne Particulate Matter in China. Environ. Sci. Technol. 2018, 52 (15), 8205–8214. 10.1021/acs.est.8b02492. [DOI] [PubMed] [Google Scholar]
  45. Lv D.; Chen Y.; Zhu T.; Li T.; Shen F.; Li X.; Mehmood T. The Pollution Characteristics of PM10 and PM2.5 During Summer and Winter in Beijing, Suning and Islamabad. Atmos. Pollut. Res. 2019, 10 (4), 1159–1164. 10.1016/j.apr.2019.01.021. [DOI] [Google Scholar]
  46. Zhang X.; Diao Z.; Ma H.; Xie X.; Wang Y.; Liu X.; Yuan X.; Zhu F. Multi-class Organic Pollutants in PM(2.5) in Mixed Area of Shanghai: Levels, Sources and Health Risk Assessment. Sci. Total Environ. 2023, 903, 166352. 10.1016/j.scitotenv.2023.166352. [DOI] [PubMed] [Google Scholar]
  47. Liang C.; Ding R. Y.; Sun Q. L.; Liu S. Q.; Sun Z. W.; Duan J. C. An Overview of Adverse Outcome Pathway Links between PM2.5 Exposure and Cardiac Developmental Toxicity. Environment & Health 2024, 2, 105–113. 10.1021/envhealth.3c00143. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

eh4c00229_si_001.pdf (1.7MB, pdf)
eh4c00229_si_002.xlsx (55.6KB, xlsx)

Articles from Environment & Health are provided here courtesy of American Chemical Society

RESOURCES