Concurrent Extraction and Analysis of Atmospheric Particulate Matter-bound PBDEs, PAHs, Nitrated PAHs, and OPEs

This study aimed to develop and validate a robust method for the simultaneous extraction of polybrominated diphenyl ethers (PBDEs), polycyclic aromatic hydrocarbons (PAHs), nitrated PAHs (N-PAHs)


INTRODUCTION
Atmospheric particulate matter (PM), composed of tiny airborne particles, plays a crucial role in environmental and public health concerns (Manullang et al., 2023).Elevated levels of atmospheric PM have been linked to compromised air quality, leading to adverse health effects (Yadav et al., 2020;Lin et al., 2022).These particles constitute a complex mixture of organic and inorganic compounds with various sources, including industrial emissions.Among the organic compounds associated with particulate matter, polybrominated diphenyl ethers (PBDEs), polycyclic aromatic hydrocarbons (PAHs), nitrated PAHs (N-PAHs), and organophosphate esters (OPEs) are of particular interest due to their detrimental effects on human health and the environment (Doherty et al., 2019;He et al., 2021;Lao et al., 2022;Akinpelumi et al., 2023;Gomes et al., 2023).
PAHs, derived from incomplete combustion and pyrolysis of organic matter and unburnt petroleum products, are of great concern due to their potential carcinogenic and mutagenic properties (Cheruiyot et al., 2015).These compounds are primarily generated by anthropogenic activities, such as traffic, domestic heating, biomass burning, and industrial processes (Lara et al., 2022).It is worth noting that, despite being well-studied, the full spectrum of PAH derivatives, e.g., N-PAHs, and their carcinogenic properties remain an area of active investigation.In the atmosphere, N-PAHs can form from the photo-oxidation reactions of PAHs (Lee et al., 2022).They can also be formed during combustion processes (Lee et al., 2022).These compounds, characterized by their higher molecular weights and lower vapor pressures, have an elevated tendency to adsorb onto particulate matter compared to their parent PAH compound (Lara et al., 2022).Studies have shown that N-PAHs may pose even higher health risks compared to their parent PAHs, due to their direct mutagenic potency (Zhang et al., 2022).However, these compounds are often underrepresented in air quality monitoring networks.
Polybrominated diphenyl ethers (PBDEs) are widely known as flame retardants and have a history of use in consumer goods (Abbasi et al., 2015).Despite the phase-out of certain formulations, ongoing emissions persist due to the presence of in-use products containing these chemicals (Abbasi et al., 2019).Additionally, these compounds have been shown to form in combustion sources (Altarawneh et al., 2020;Yang et al., 2022).PBDE flame retardants have been replaced by alternative flame retardants, particularly organophosphate esters (OPEs), which have also become increasingly prevalent in the environment (Kung et al., 2022).These compounds have been detected in various environmental samples and even in human biota, raising concerns about their environmental impact and potential health risks (Kung et al., 2022).
As these organic compounds are present in atmospheric PM, the challenges of monitoring their concentrations, variations, and potential effects on public health and the environment have become more evident.Furthermore, quantifying these compounds in trace amounts (ng m -3 and pg m -3 ) within a complex environmental matrix such as atmospheric PM is a demanding endeavor, necessitating the development of reliable, sensitive, and selective analytical methods.
The techniques widely used for solid extraction of semi-volatile and non-volatile compounds include Soxhlet, sonication, pressurized liquid extraction, and microwave extraction.Pressurized liquid extraction and microwave extraction improve on the more traditional extraction techniques, i.e., Soxhlet and sonification by using automation, reducing the amount of solvent used, and quickening the extraction process.Clark et al. (2015) extracted eight groups of compounds, including PAHs, pesticides, and flame retardants, from atmospheric particulate matter using a pressurized liquid extraction technique, achieving acceptable recoveries.Microwave extraction is faster, reducing the extraction times from hours to minutes.It is also more energy efficient, selective, and uses minimal solvent (Nieto et al., 2008).Therefore, microwave extraction is a better technique if green analytical chemistry concepts and extraction time are considered.The technique has been used to simultaneously extract PAHs, organic acids, pesticides, phthalates, and PCBs (Alvarez-Avilés et al., 2007), and PAHs, N-PAHs, and oxygenated PAHs in atmospheric particles (Lara et al., 2022).
The primary objectives of this study were to develop and validate a robust method for simultaneous extraction of PAHs, OPEs, NPAHs, and PBDEs from a single PM sample.Furthermore, we aimed to apply this method to samples collected from and around an industrial park during winter.The accuracy and reliability of our methods were evaluated using Standard Reference Materials (SRMs) and employing quality control measures with deuterated and isotopically labeled standards.Actual atmospheric samples were collected from five locations in and around an industrial park in Taiwan in November and January and analyzed for the four organic contaminants.By addressing these objectives, we aim to establish a reliable and fast extraction method and contribute valuable data to the fields of environmental analysis while also providing insights that may guide future research in air quality management and pollution control.

Chemicals
All reagents, materials, and analytical standards were purchased from commercially available

Standard Reference Materials (SRMs)
The accuracy and precision of the extraction and analytical method were evaluated using Standard Reference Materials (SRMs).SRM 2585 "Organic Contaminants in House Dust" was used for the analysis of PBDEs, PAHs, and OPEs, while SRM 2975 "Diesel Particulate Matter" was used for the analysis of N-PAHs.100 mg of SRM 2585 and 10 mg of SRM 2975 were measured and extracted via microwave extraction, as shown in Fig. 1 and described in Section 2.4.After extraction and cleanup, gas chromatography and triple quadrupole tandem mass (Agilent 7000D) (GC-MS/MS) were used for the analysis of PAHs, N-PAHs, and OPEs and high-resolution gas chromatography (6890N) and high-resolution mass spectrometry (Micromass AutoSpec-Ultima) (HRGC-HRMS) for PBDE analysis.Identification of congeners was performed using the retention time (± 0.05 min) and the ions and their ratios (± 20%).The quantification of the congeners was carried out using

Sampling of Atmospheric Total Particulate Matter (PM)
Atmospheric particulate matter (PM) was collected from five designated locations (A1-A5) surrounding Liuying Technology Industrial Park in Tainan, Taiwan, during two periods in winter (24-26 November 2021 and 5-7 January 2022), as shown in Table 3.The primary industries within the industrial park include, electrical and electronic equipment, machinery manufacturing, and metal manufacturing.Sites A1 and A2 were situated in residential areas, whereas A3 and A4 were located within an elementary school and junior high school, respectively.Site A5 was specifically located within the industrial park.Additionally, meteorological parameters including, temperature, relative humidity, wind speed, and wind direction, during sampling were collected.The sampling followed the U.S. EPA, 40 CFR Appendix B to Part 50 -Reference Method for the Determination of Suspended Particulate Matter in the Atmosphere (High-Volume Method).The PM samples were collected continuously at a flow rate of 1.1 m 3 min -1 to 1.7 m 3 min -1 for 24 hours on a glass fiber filter (20 cm × 25 cm) using a high-volume air sampler.Prior to sampling, the glass fiber filters were pretreated by rolling them into clean microwave tubes and adding 10 mL of n-hexane, 10 mL of acetone, and 5 mL of dichloromethane, and extracting them at 120°C for 30 min.After sampling, North the filters were covered with aluminum foil and placed in a desiccator.The filters were weighed periodically until a constant weight was attained.

Microwave Extraction, Cleanup, and Fractionation
The glass filter fiber (GFF) and SRM samples were extracted with 10 mL n-hexane, 10 mL acetone, and 5 mL dichloromethane via microwave extraction.Before extraction internal standards of PBDEs, OPEs, PAHs, and N-PAHs were added.The temperature was then ramped up from room temperature to 75°C in 5 min and held for 5 min.Then the temperature was increased to 120°C in 10 minutes and held for 20 min.After the extraction, the extract was filtered and concentrated to near dryness using a stream of pure nitrogen, then reconstituted in n-hexane before cleanup and fractionation.
The columns packed with silica gel (4.3 cc × 3) and anhydrous sodium sulfate (2.8 cc × 1) were prewashed sequentially with acetone, dichloromethane, and n-hexane.The sample extract was added to the column with two additional 3 mL n-hexane rinses.Then the column was eluted with two aliquots of 10 mL n-hexane aliquots to obtain the first fraction of PBDEs.The second fraction (PAHs and PBDEs) was eluted with two 10 mL dichloromethane:n-hexane (1:4) and then two additional 5 mL n-hexane aliquots.A third fraction (N-PAHs) was eluted with three aliquots of 10 mL dichloromethane, while a fourth fraction (OPEs) was eluted with two aliquots of 10 mL dichloromethane:acetone (1:2) aliquots.Fractions 2, 3, and 4 were concentrated and analyzed for PAHs, N-PAHs, and OPEs using GC-MS/MS, respectively.
After PAH analysis, the first and second fractions were combined and purified using acidic silica gel: alumina tandem columns.The purification columns, with acidic silica gel (8.6 cc × 1), alumina (12 cc × 1), and anhydrous sodium sulfate (2 cc × 2), were prewashed with 20 mL n-hexane.The extracts were added to the column with three 5 mL n-hexane rinses and then the column was flushed with three additional 10 mL n-hexane aliquots.The extract was discarded, and the acidic silica part was removed.The alumina column was eluted with 10 mL n-hexane, and the extract was retained.Finally, the column was flushed with 25 mL of dichloromethane: n-hexane (1:1), and the extract was collected.Both extracts were concentrated and analyzed for PBDE using HRGC-HRMS.

Instrumental Analysis
High-resolution gas chromatography and high-resolution mass spectrometry were used to analyze 14 PBDE congeners.A DB-5HT (15 m × 0.25 mm × 0.1 µm) column was utilized for the separation of the congeners.Helium gas, at a flow rate of 1 mL min -1 , was used as the carrier gas.The injection temperature was 270°C, with the temperature program as follows: the initial temperature of 100°C was held for 1 min, then ramped up to 180°C at a rate of 20°C min -1 , where it was held for 2 min.Subsequently, the temperature was ramped up to 250°C at 8°C min -1 and then further increased to 315°C at 20°C min -1 and held for 4 min.Electron ionization, at a temperature of 250°C, was employed as the ionization source.The resolution power was 10000, and the selected ion monitoring (SIM) mode was adopted for the analysis.
Chemical analysis of 16 PAH, 13 N-PAH, and 11 OPE congeners was carried out using GC-MS/MS.Separation was accomplished using a DB-17MS (30 m × 0.25 mm × 0.25 µm) column.Helium gas was used as the carrier gas with a flow rate set at 1.2 mL min -1 , and the injection temperature was maintained at 280°C.For PAH analysis, the temperature program was as follows: initially, the temperature was held at 60°C for 0.5 min, then ramped up to 265°C at a rate of 15°C min -1 , followed by a further increase to 305°C at a rate of 5°C min -1 and held at that temperature for 10 min.Electron ionization (EI), at a temperature of 280°C, served as the ionization source, and Multiple Reaction Monitoring (MRM) was employed as the monitoring mode.For N-PAH analysis, the oven temperature was initially held at 120°C for 1 min, then the temperature was raised to 185°C at 30°C min -1 .Subsequently, the temperature was further increased at a rate of 8°C min -1 until reaching 310°C, where it was held for 5 minutes.The ionization source and monitoring mode were consistent with the ones chosen for PAH analysis.Finally, the temperature program for OPE analysis was as follows: the temperature was maintained at 45°C for 3 min, then ramped up to 260°C at a rate of 25°C min -1 , followed by a further increase to 310°C temperature at a rate of 10°C min -1 , where it was held for 8 min.Electron ionization, at a temperature of 250°C, was the ionization source and MRM was used as the monitoring mode.

Quality Control and Assurance
Five blank GFF samples were spiked with known amounts of target analytes near the mid-point of all calibration curves: PBDE addition concentrations were 20, 40, 100 pg mL -1 , PAH concentration was 200 ng mL -1 , N-PAH concentration was 200 ng mL -1 , and OPE concentration was 200 ng mL -1 .Additionally, known amounts of internal standards were spiked prior to extraction and analysis.
In this study, the limit of detection (LOD) for PBDE was defined as the point at which the signal-tonoise ratio exceeded 3.This rigorous criterion ensured the reliable identification of PBDE compounds in our analytical measurements.The method detection limit (MDL), calculated as three times the standard deviation of the instrument's response to the blank, was used to determine the lowest concentrations of PAHs, N-PAHs, and OPEs.The U.S. EPA Definition and Procedure for the Determination of the Method Detection Limit was used to estimate MDLs.In brief, seven sets of pre-cleaned GFF aliquots, used as procedural blanks, were prepared and taken through the analytical method.The MDL was calculated by multiplying the standard deviation of the concentrations measured in each replicated by the student's t-test value for a total of five samples.These welldefined MDL criteria were crucial in ensuring the accuracy and sensitivity of our analytical methods throughout our study.

Reproducibility Studies and SRM
In this section, we present the method validation and reproducibility studies for the detection of atmospheric PM-bound PBDEs, PAHs, N-PAHs, and OPEs.Commercially available standard reference materials, SRM2585 and SRM2975, were employed for method validation.SRM2585 was used for PBDEs, PAHs, and OPEs validation, while SRM2975 was used for NPAHs.Four SRM samples were processed according to our procedures, and the results were compared to the reported values to assess the accuracy and precision of the methods.Table 1 summarizes the relative time, ion pairs, quantitation standards, and QC/QA analysis parameters for the detection of PM-bound PBDEs, while Table 2 provides similar parameters for PM-bound PAHs, N-PAHs, and OPEs.Notably, the detected values for PBDE, PAH, and OPE congeners were below 30% of the reported values, indicating the accuracy and precision of the methodology (Tables 1 and 2).However, for 3-NitroPhe, 1-NitroPyr, and 6-NitroChr, the recovery rates were slightly above 30%.These findings suggest that our technique is generally reliable and accurate for majority of the target compounds.

Estimated Limit of Detection and Method Detection Limit
The limit of detection (LOD) and method detection limits (MDLs) for the target compounds are presented in Tables 1 and 2. The LOD for PBDEs ranged from 0.31-6.43pg, highlighting the discernable differences in detectability among various PBDE congeners.The variability may stem from differences in molecular structures and chemical properties among the congeners.Notably, higher brominated PBDEs had higher LODs.The extremely low LODs achieved by the HRGC/HRMS instill confidence in analyzing very low PBDE concentrations.The MDL for PAHs, N-PAHs, and OPEs ranged from 0.73-6.15ng, 0.066-0.512ng, and 0.67-13.6ng, respectively.While PAHs and OPEs were measured in nanograms, N-PAHs were in the sub-nanogram range, which signifies the higher sensitivity required for N-PAH analysis due to their relatively low concentrations in the environment.The MDLs in this study were much lower compared to previous studies that used GC-MS (Clark et al., 2015), which is attributed to the high sensitivity of GC-MS/MS.

Analysis of Actual Atmospheric PM Samples
The developed extraction and analysis methods were tested on real atmospheric PM samples collected in and around an industrial park in Tainan, Taiwan.Fig. 2 shows the concentrations from the five locations in and around the industrial park.The concentrations in the first day of sampling were in the order of: A3 (67.4 µg m -3 ) > A1 (63.5 µg m -3 ) > A5 (61.4 µg m -3 ) > A4 (47.6 µg m -3 ) > A2 (39.1 µg m -3 ), while in the second day of sampling they were: A1 (68.9 µg m -3 ) > A3 (65.2 µg m -3 ) > A2 (45.1 µg m -3 ) > A4 (43.5 µg m -3 ) > A5 (41.0 µg m -3 ).The concentrations in A1 and A2 were higher in the second than the first day, while the concentrations in A3 to A5 were higher in the first than the second day.Atmospheric PM originates from a variety of sources, including vehicle emissions, cooking activities, construction work, wind-blown dust, tire wear, and the formation of secondary organic aerosols through the oxidation of volatile organic compounds in the atmosphere (Mukherjee and Agrawal, 2017;Hopke et al., 2020;Le et al., 2023).

Atmospheric Total PM-bound PBDE, PAHs, N-PAHs, and OPEs
The mass concentrations of PM-bound compounds during the two sampling periods are shown in Fig. 3. On average, OPEs (432-2815 pg m -3 , mean = 1088 pg m -3 ) had the highest concentrations followed by PAHs (503-1774 pg m -3 , mean = 1075 pg m -3 ), N-PAHs (39.1-277 pg m -3 , mean = 125 pg m -3 ), and finally PBDEs (5.07-13.6 pg m -3 , mean = 9.19 pg m -3 ).Clark et al. (2015) also found OPEs bound to atmospheric particulate matter were in higher concentrations than PAHs and PBDEs.In a review of OPEs in the atmosphere, Kung et al. (2022) concluded that OPEs tend to partition mostly in the particulate phase.In addition, OPEs are prevalent compounds in the environment and are considered high production value chemicals by the Organization of Economic Co-operation and Development (OECD) (Kung et al., 2022).
For the spatial variation, only PAHs had a similar trend as PM, i.e., higher PM concentrations lead to higher PAH concentrations and vice versa.The toxic PAH concentrations (Fig. S1) also had a similar spatial concentration trend as PM, with concentrations ranging from 18-294 pg BaP-TEQ m -3 (mean = 120 pg BaP-TEQ m -3 ).

Congener profiles
The congener profiles for the four target compounds are shown in Fig. S2.For PBDEs, the highly brominated congener, BDE 209, was the main congener followed by BDE 206 and 207 for all locations and periods.Several studies have also reported the predominance of BDE 209 in the particle phase and total concentrations (gas + particle phases) (Yang et al., 2013;Qiao et al., 2020;Su et al., 2022).BDE 209 is the main ingredient in deca-PBDE flame retardants (Li et al., 2023) but has also been reported to form during the combustion of municipal waste (Lin et al., 2014;Yang et al., 2022) and recycling of electrical and electronic wastes (Zennegg et al., 2014).High-molecularweight PAHs (> 4 rings) were predominant for all locations and periods, accounting for 88.9%-93.9% of the total PAHs.The main congeners were BghiP, IcdP, BbF, BaP, Fla, and Pyr.High-molecularweight PAHs usually partition in the particulate phase due to their hydrophobicity (Cheruiyot et al., 2015).Furthermore, the sampling was carried out during winter when temperatures were relatively low, enhancing the partitioning of the PAHs on the particulate phase (He et al., 2014).These results are similar to other studies on atmospheric PM-bound PAHs (He et al., 2014;Clark et al., 2015;Shen et al., 2019).For N-PAHs, 2-NitroFla was the main congener for all locations and periods, contributing to 34.5%-81.6%,followed by 7-NitoBaA (2.69%-38.1%)and 1-NitroPyr (6.34%-18.1%).9-NitroAnt was only present on the first day, accounting for 8.36%-17.7%.TCPP was the main OPE congener in A1 to A4. TCPP has been also reported as the main congener bound to particulate matter (Clark et al., 2015).Aromatic OPEs (TCP, and TPhP) and long-chain alkyl TEHP were also predominant.TiBP and TBEP were found in locations A4 and A5.

Emission sources based on diagnostic ratios
The PAH diagnostic ratios were calculated and used to assess the sources and origin of PAH contamination in the atmosphere (Fig. 4).The ∑LMW/∑HMW and COMB/∑PAHs for all the locations and periods were < 1 and ~1, respectively, indicating that combustion was the PAH source   Tobiszewski and Namieśnik, 2012;Kozak et al., 2017).The Fla/(Fla+Pyr), BaA/(BaA+Chr), and InP/(InP+BghiP) values were within 0.4-0.5, 0.2-0.35,and 0.2-0.5, respectively, suggesting that petroleum combustion was the probable source of PAHs (Tobiszewski and Namieśnik, 2012;Jiao et al., 2017;Kozak et al., 2017).Furthermore, the BaP/(Chr+BaP) values for all the locations and periods were > 0.35, suggesting that the PAHs originated from vehicular emissions (Tobiszewski and Namieśnik, 2012).For all locations and periods, except for the second sampling days in A1 and A3, the Fla+(Fla+Pyr) values were > 0.5, indicating the PAH source came from diesel combustion.Therefore, we can conclude that the PM-bound PAHs might have originated from vehicular emissions.

CONCLUSIONS
This study provides a reliable method for simultaneous extraction and analysis of atmospheric PM-bound PBDEs, PAHs, N-PAHs, and OPEs.The validation process included the use of standard reference materials (SRMs), which demonstrated the accuracy and precision of the analytical techniques employed.The results indicated that the majority of target compounds exhibited excellent recovery rates (within 30%), with minor deviations observed for specific congeners.Overall, these findings affirm the reliability and accuracy of the developed methodology for the analysis of the selected compounds.Additionally, the reproducibility of the method was assessed through replicates, consistently yielding robust results and reinforcing the reliability of the analytical procedures.The recovery rates for the four compounds ranged from 82.2% ± 8.7% to 115% ± 17.4%.The practical application of the method was demonstrated by analyzing actual atmospheric PM samples collected in and around an industrial park in Tainan, Taiwan.The results highlighted the spatial variation of PM-bound compounds, with some evidence suggesting a potential pollution source in the vicinity of the industrial park.The congener profiles of the target compounds revealed important insights into their distribution and partitioning in atmospheric PM.High-molecularweight PAHs, BDE 209, and TCPP were the main congeners.Finally, diagnostic ratios of PAHs were used to assess the likely sources of PAH contamination in the atmosphere, with the results suggesting combustion, petroleum combustion, and vehicular emissions as the primary sources of PM-bound PAHs.

Fig. 1 .
Fig. 1.The flow chart of the sample cleanup process.

Fig. 2 .
Fig. 2. Atmospheric PM concentrations in the different sampling sites on the first and second days.

Fig. 3 .
Fig. 3. Atmospheric PM-bound PAH, N-PAH, PBDE, and OPE concentrations in the different sampling sites on the first and second days.

Fig. 4 .
Fig. 4. Diagnostic ratios for the atmospheric PM-bound PAHs for the first and second days.

Table 1 .
Relative time (RT), ion pairs, quantitation standards, and HRGC/HRMS QC/QA analysis parameters for the detection of PM-bound PBDEs.

Table 2 .
Relative time (RT), precursor and product ion pairs, quantitation standards, and GC-MS/MS QC/QA analysis parameters for the detection of PM-bound PAHs, N-PAHs, and OPEs.
a SRM 2975 was used.

Table 3 .
Coordinates and sampling information of the sampling sites.