Chemicals

Table 1 lists the analytes quantified in this study, comprising seven PFAS from the PFSA group, 11 from the PFCA group, and six PFAS precursors. PFAS standards were obtained from AccuStandard (New Haven, CT, USA). Internal standards, namely perfluoro-n-[1,2,3,4-C] octanoic acid (C-PFOA, 50 ± 2.5 µg/mL) and sodium perfluoro-1-[1,2,3,4-C] octanesulfonate (C-PFOS, 50 ± 2.5 µg/mL), were purchased from Wellington Laboratories (Canada). HPLC-grade ammonium acetate (CH₃COONH₄), MeOH, and acetonitrile (ACN) were purchased from Merck (Germany). Solid-phase extraction was performed using Supelclean ENVI-Carb SPE cartridges (Supelco, USA).

Method validation

Instrument analysis

All measurements were performed using UPLC-MS/MS (ExionLC AC series, Triple Quad 3500, Sciex, USA). Separations were performed on a Kinetex C18 column (100 mm × 2.1 mm, 1.7 μm). The mobile phase gradient started with 10:90 of 20 mmol/L CHCOONH aqueous solution: acetonitrile (v/v) for 0.5 min, then increased to 95% in 10 min, held for 3 min, and finally returned to the initial composition. The injection volume, column temperature, and flow rate were 10 μL, 45 °C, and 0.2 mL/min, respectively3.

Sample collection

The developed method was used to analyze 24 PFAS in TSP samples collected in Ho Chi Minh City, Vietnam. The sampling site was located on the rooftop of the 6th floor of the G Block at the University of Science’s Linh Trung Campus. The surrounding area is influenced by various anthropogenic sources, including cooking, burning, traffic emissions, and nearby industrial activities. TSP samples were collected on pre-treated quartz filters (Advantec, Japan), using a high-volume air sampler (Kimoto Electric Co., Model 120H, Japan) operating at a flow rate of 600 L/min over a 24-hour sampling period. Before use, quartz filters were heated at 400 °C for 8 hours to remove impurities, cooled, and stabilized in a desiccator at 25 °C with a relative humidity of 30%–40%. After sampling, the filters were wrapped with aluminum foil to avoid photochemical reactions and determine the mass after stabilizing, and stored at ˗20 °C22. In total, four samples were collected from October 10 to 13, 2022, and three samples were collected from March 15 to 17, 2023.

Mass spectrometry parameters

PFAS are acidic compounds that readily lose a proton (H⁺) because of their strong acidic functional groups, such as -COOH and -SO₃H. The loss of a proton forms the stable [M-H]⁻ ion, which is easily detected in the negative ESI mode. The stability of PFAS is due to electron-withdrawing fluorine atoms, which delocalize the negative charge, making the ionization process efficient. Thus, PFAS are well-suited for detection by ESI, commonly used in UPLC-MS/MS analysis (Voogt and Saez, 2006). Four basic parameters (DP, EP, CE, and CXP) directly affect the PFAS signals. DP and EP depend on the precursor ions, whereas CE and CXP depend on the product ions. The EP and CXP values had little impact on the signals and were set at ˗10 V3. Table 1 lists the optimal DP and CE values for each analyte, together with the precursor ions, product ions, and the corresponding transitions. The results obtained are similar to those in the literature and published standards.

The retention times of the 24 PFAS showed a positive correlation with increasing carbon chain length. In the PFSA group, compounds with longer carbon chains (e.g., PFDS and PFOS) exhibited longer retention times than those with shorter chains (e.g., PFBS and PFPeS). Similar results were observed in the PFCA group. Furthermore, variations in chemical structure may also influence retention times. PFSAs generally have longer retention times than PFCAs of the same carbon chain length.

Optimization of the ultrasonic extraction process

(1) Optimizing solvent volume

MeOH was chosen as the extraction solvent because of its high efficiency in dissolving PFAS, stemming from its ability to interact with the non-polar C-F chains and enhance desorption from particulate surfaces. MeOH also exhibits superior wetting properties compared to other organic solvents, facilitating deeper penetration into the particulate matrix and increasing the extraction yield. Hence, MeOH is widely used as the standard solvent in PFAS extraction from atmospheric PM, offering an optimal balance of polarity, low viscosity, and compatibility with SPE clean-up and UPLC-MS/MS analysis1, 6, 11, 12.

The recovery efficiencies obtained from testing extraction volumes of 4, 5, and 6 mL met the Association of Official Analytical Collaboration (AOAC) International criteria for acceptable recovery (typically 70% – 120%), indicating that each tested volume provided reliable extraction performance. Among them, a 4 mL volume was selected as the optimal choice, as it balanced high recovery with reduced solvent consumption and shorter processing time through the SPE column. This MeOH volume has been mentioned in previous studies, as it ensures effective analyte extraction from the sample matrix while improving the overall efficiency of the analytical workflow6.

(2) Optimizing extraction time

The results indicate that ultrasonic extraction achieves high efficiency within the first 10 minutes, suggesting that a 10-minute duration per extraction is sufficient to maximize analyte recovery from the sample matrix. Furthermore, performing three successive extractions ensures complete transfer of analytes into the solvent. A 10-minute extraction time also agrees with protocols adopted in previous studies, supporting its reliability and effectiveness6.

(3) Optimizing extraction temperature

The recovery efficiency and %RSD values obtained at different extraction temperatures met the AOAC standards for all analytes. Among the tested conditions, ultrasonic extraction at 40 °C and 50 °C yielded the highest extraction efficiencies. Theoretically, increasing the extraction temperature enhances analyte solubility and transfer into the solvent phase. However, to prevent excessive heating and reduce energy consumption and processing time in the ultrasonic bath, 40 °C was selected as the optimal extraction temperature. This temperature has been used in previous studies, supporting its applicability and effectiveness6.

LOQs and application to PFAS determination in TSP samples

This study presents the first method for analyzing PFAS in TSP in Vietnam. false presents the PFAS concentrations in seven TSP samples collected from Ho Chi Minh City, along with the corresponding LOQs and quality control recoveries. The proposed method achieved low LOQs, ranging from 0.01 to 0.14 ng/mL, comparable to or lower than those reported in previous studies, enabling the sensitive detection of PFAS at trace levels. Notably, LOQs for PFPeA, PFHpA, PFOA, PFBS, PFHxS, PFHpS, PFOS, PFDA, PFDoA, PFPeS, and PFTeA were lower than or equal to those in previous studies, highlighting the improved sensitivity of our method for these analytes. Although some compounds, including PFBA, PFHxA, PFNA, PFDS, and PFOSA, exhibited slightly higher LOQs, the values remained within acceptable ranges for environmental trace analysis. The ultrasound-assisted extraction using MeOH (4 mL, 40 °C, 10 min, in triplicate) yielded high recovery efficiencies (95.5%–114.4%) while reducing extraction time and solvent consumption. Moreover, the method employs commonly available ENVI-Carb SPE cartridges and UPLC-MS/MS instrumentation, making it practical and well-suited to the facilities and conditions of most laboratories.

“<LOD” indicates concentrations below the limit of detection; “ND” means not detected.

Based on the analysis, 22 of 24 PFAS were detected in the samples, except for PFTrDA and 8:2FTS. Furthermore, 12 of 24 PFAS were found across all samples: PFBA, PFPeA, PFHxA, PFHpA, PFOA, PFHxS, PFHpS, PFOS, PFUnA, PFDoA, 6-2FTS, PFTeA. Among these, PFHpA exhibited the highest concentration, followed by PFBA. This result is in line with the findings of previous studies on PFAS in airborne dust in Japan (6,27). The concentration trend observed was PFHpA > PFBA > PFHxA > PFPeA > PFOA, with PFNS showing a markedly lower concentration. This distribution highlights the varying prevalence of different PFAS in the TSP samples.