Simple and Efficient Analysis of PFAS in Milk, Eggs and Infant Formula

Application Note Food & Beverage Testing Authors Limian Zhao, Matthew Giardina, and Emily Parry Agilent Technologies, Inc. Abstract This application note presents the development and validation of a multiresidue method for the analysis of per- and polyfluoroalkyl substances (PFAS) in infant formula, milk, and eggs. The method uses QuEChERS extraction followed by EMR mixed-mode passthrough cleanup using an Agilent Captiva EMR PFAS Food II cartridge, then LC/MS/MS detection. The method features a simplified and efficient sample preparation, sensitive LC/MS/MS detection, and reliable quantitation using neat standard calibration curves. The novel Captiva EMR PFAS Food II cartridge was developed and optimized specifically for PFAS analysis in foods of animal origin as well as seeded dry foods of plant origin. The method was validated based on the AOAC Standard Method Performance Requirements (SMPR) 2023.003, including method suitability, sensitivity, accuracy, and precision. The method was demonstrated to meet the required limits of quantitation (LOQs), recovery, and repeatability (RSD) for four core PFAS targets—perfluorooctane sulfonic acid (PFOS), perfluorooctanoic acid (PFOA), perfluoronanoic acid (PFNA), and perfluorohexane sulfonic acid (PFHxS)—and 26 remaining PFAS targets in infant formula, milk, and eggs. Determination of 30 Per- and Polyfluoroalkyl Substances in Infant Formula, Milk, and Eggs Using Agilent Captiva EMR PFAS Food II passthrough cleanup and LC/MS/MS detection 2 Introduction Determination of PFAS residues in food has become a topic of rising concern, gaining more attention over the last several years. In April 2023, the European Commission enforced regulations for four PFAS compounds—PFOS, PFOA, PFNA, and PFHxS—in eggs, fish, seafood, meat, and offal.1 In November 2023, AOAC released the SMPR 2023.003 for the analysis of 30 PFAS in produce, beverages, dairy products, eggs, seafood, meat products, and feed.2 Methods based on LC/MS/MS for PFAS analysis have been widely applied for environmental water and soil analyses.3,4 The acidic groups contained in PFAS compounds enable them to be ionized easily and efficiently under negative mode, providing advantages for method sensitivity and selectivity. For food analysis, the sample preparation method plays a critical role for efficient PFAS extraction, removal of matrix co-extractives, and sample concentration or dilution when needed. The large variety and high complexity of food matrices challenge the sample preparation method not only in terms of sample extraction and matrix cleanup efficiency, but also the overall method simplicity, sample processing efficiency, and accommodation of different matrices. Weak anion exchange (WAX) sorbent-based solid phase extraction (SPE) methods have been used widely for PFAS analysis in environmental samples such as water and soil, as well as other matrices. However, the SPE methodology is challenging for sample preparation of complex solid food, as food samples need to be extracted before loading into the cartridge. Also, the typical SPE procedure involving conditioning, equilibrium, loading, washing, and eluting requires a lot of time and solvent. QuEChERS extraction followed by typical dispersive SPE (dSPE) cleanup has been reported for PFAS in food sample preparation.5,6 However, dSPE cleanup does not provide efficient matrix removal for many food matrices, which cannot support the lower LOQ requirement in food. Thus, another cleanup step using WAX SPE is added after dSPE cleanup.5 This causes the method to be time consuming and labor intensive, which significantly impacts sample process productivity. This type of sample cleanup can also result in loss of PFAS targets. Agilent Captiva EMR PFAS Food cartridges were developed and optimized specifically for PFAS analysis in food, providing comprehensive mix-mode passthrough cleanup. Two types of cartridges (I and II) were designed to cover the large variety of food matrices. The objective of this study was to develop and validate a complete workflow for the determination of 30 PFAS in infant formula, milk, and eggs. The method uses QuEChERS extraction followed by passthrough cleanup with the Captiva EMR PFAS Food II cartridge and detection with the Agilent 6495D triple quadrupole LC/MS. Experimental Chemicals and reagents Native PFAS and isotopically labeled internal standard (ISTD) solutions were purchased from Wellington Laboratories (Guelph, Ontario, Canada). Methanol (MeOH), acetonitrile (ACN), and isopropyl alcohol (IPA) were from VWR (Radnor, PA, USA). Acetic acid (AA) and ammonium acetate were procured from MilliporeSigma (Burlington, MA, USA). Solutions and standards Three native PFAS spiking solutions (I, II, and III) were prepared by diluting the native PFAS solutions with MeOH at concentrations of 200, 20, and 2 ng/mL for 28 PFAS targets, respectively. The exceptions were for PFBA and PFPeA, where the concentrations were a factor of 10 and two times the concentration of the other 28 targets, respectively. The ISTD spiking solution was prepared by diluting the ISTD primary solution with MeOH at a concentration of 100 ng/mL. The native PFAS and isotopic ISTD spiking solutions were used for preparing neat calibration standards at 20, 50, 100, 200, 500, 1,000, 2,000, 5,000, and 10,000 ng/L for native PFAS targets and ISTD concentration of 1,000 ng/L in MeOH. They were also used for matrix prespiked quality control (QC) samples. All standards were stored at 4 °C and used for no more than two weeks. The ACN with 1% AA extraction solvent was prepared by adding 10 mL glacial AA into 990 mL of ACN and storing it at room temperature. LC mobile phase A was 5 mM NH4 OAc in water, and mobile phase B was MeOH. 3 Equipment and material The study was performed using an Agilent 1290 Infinity II LC system consisting of a 1290 Infinity II high-speed pump (G7120A), a 1290 Infinity II multisampler (G7167B), and a 1290 Infinity II multicolumn thermostat (G7116A). The LC system was coupled to an Agilent 6495D LC/TQ equipped with an Agilent Jet Stream iFunnel electrospray ion (ESI) source. Agilent MassHunter Workstation software was used for data acquisition and analysis. Other equipment used for sample preparation included: – Centra CL3R centrifuge (Thermo IEC, MA, USA) – Geno/Grinder (Metuchen, NJ, USA) – Multi Reax test tube shaker (Heidolph, Schwabach, Germany) – Pipettes and repeater (Eppendorf, NY, USA) – Agilent positive pressure manifold 48 processor (PPM-48; part number 5191-4101) – CentriVap and CentriVap Cold Trap (Labconco, MO, USA) – Ultrasonic cleaning bath (VWR, PA, USA) The 1290 Infinity II LC system was modified using an Agilent InfinityLab PFC-free HPLC conversion kit (part number 5004‑0006), including an Agilent InfinityLab PFC delay column, 4.6 x 30 mm (part number 5062-8100). Chromatographic separation was performed using an Agilent ZORBAX RRHD Eclipse Plus C18, 2.1 × 100 mm, 1.8 µm (part number 959758-902) and an Agilent ZORBAX RRHD Eclipse Plus C18 column, 2.1 mm, 1.8 µm, 1,200-bar pressure limit, UHPLC guard (part number 821725-901). Other Agilent consumables used included: – Agilent Bond Elut QuEChERS EN extraction kit, EN 15662 method, buffered salts, ceramic homogenizers (part number 5982-5650CH) – Captiva EMR PFAS Food II cartridges, 6 mL, 750 mg (part number 5610-2232) – Polypropylene (PP) snap caps and vials, 1 mL (part number 5182-0567 and 5182-0542) – PP screw cap style vials and caps, 2 mL (part numbers 5191-8150 and 5191-8151) – Tubes and caps, 50 mL, 50/pk (part number 5610-2049) – Tubes and caps, 15 mL, 100/pk (part number 5610-2039) All the consumables used in the study were tested and verified for acceptable PFAS cleanliness. LC/MS/MS instrument conditions The LC binary pump conditions are listed in Table 1 and the multisampler program is listed in Table 2. The column temperature was set at 55 ± 0.8 °C. MS data were acquired in negative ion mode with a constant fragmentor setting of 166 V. The ESI source settings were: drying gas at 150 °C, 18 L/min; sheath gas at 390 °C, 12 L/min; nebulizer gas at 15 psi; capillary voltage at 2,500 V; and nozzle voltage at 0 V. Parameter Setting Mobile Phase A 5 mM NH4 OAc in water Mobile Phase B MeOH Gradient Time (min) %A %B Flow (mL/min) 0.00 98.00 2.00 0.400 2.00 98.00 2.00 0.400 2.50 45.00 55.00 0.400 6.50 30.00 70.00 0.400 8.00 20.00 80.00 0.460 14.20 0.00 100.00 0.460 17.00 0.00 100.00 0.400 17.10 98.00 2.00 – Post Time 3.0 min Table 1. LC pump conditions for LC/MS/MS. Parameter Setting Injection Program Draw 10 µL water Draw 10 µL Wash needle Draw 50 µL water Mix 10 µL from air five times Inject Multiwash Step Solvent Time (s) Seat Backflush Needle Wash 1 IPA 10 Enabled Enabled 2 ACN 10 Enabled Enabled 3 Water 10 Enabled Enabled Starting Conditions Water NA Enabled Enabled Table 2. LC multisampler program for LC/MS/MS. Sample preparation Infant formula, milk, and egg samples were purchased from local grocery stores. Fresh milk and infant formula samples were used directly for extraction. Fresh eggs were broken and mixed thoroughly in a polypropylene bottle before extraction. For sample preparation of infant formula, a 5 g sample was used for extraction; for milk and egg samples, a 10 g sample was used for extraction. The native PFAS and ISTD spiking solutions were added to the QC samples appropriately, and only ISTD was added to matrix blanks. The samples were vortexed for 10 to 15 seconds after spiking. The samples were then ready for the procedure, which is described in Figure 1. 4 Compound RT (min) Precursor Ion (m/z) Product Ion (m/z) Collision Energy (V) Collision Cell Accelerator (V) iFunnel Mode PFBA 4.78 213 169 8 2 Standard PFPeA 5.29 263 219 4 2 Standard PFHxA 5.93 313 269 119 8 24 2 Standard PFHpA 6.72 363 319 169 8 16 2 Standard PFOA 7.6 413 369 219 8 16 2 Standard PFNA 8.51 463 419 219 169 8 16 20 2 Standard PFDA 9.3 513 469 269 219 12 16 20 2 Standard PFUnDA 9.88 563 519 319 269 12 20 20 2 Standard PFDoA 10.35 613 569 319 269 8 20 24 2 Standard PFTrDA 10.77 663 619 319 169 12 20 32 2 Standard PFTeDA 11.17 712.9 669 219 169 12 28 32 2 Standard PFBS 5.39 298.9 99 80 34 36 2 Standard PFPeS 5.99 348.9 99 80 40 44 2 Standard PFHxS 6.76 398.9 99 80 40 56 2 Standard PFHpS 7.63 448.9 99 80 42 50 2 Standard PFOS 8.50 498.9 99 79.9 50 54 2 Standard PFNS 9.29 548.9 99 80 52 56 2 Standard PFDS 9.86 598.9 99 80 56 60 2 Standard PFUnDS 10.31 648.9 99 79.8 56 76 2 Standard PFDoS 10.73 698.9 99 80 62 67 2 Standard PFTrDS 11.13 748.9 98.9 79.6 64 80 4 Standard PFOSA 10.0 497.9 169 78 48 36 36 110 3 Standard 9Cl-PF3ONS 9.03 530.9 350.9 83 28 32 3 Standard Table 3. MS acquisition conditions (on an Agilent 6495D LC/MS system) for PFAS targets, ISTDs, and cholic acid interference monitoring. Compound RT (min) Precursor Ion (m/z) Product Ion (m/z) Collision Energy (V) Collision Cell Accelerator (V) iFunnel Mode 11Cl-PF3OUdS 10.14 630.9 450.9 83 36 32 2 Standard HFPO-DA 6.15 285 185 169 119 20 4 32 5 Fragile DONA 6.83 377 251 85 8 32 5 Standard 4:2 FTS 5.87 327 307 81 80 20 36 42 2 Standard 6:2 FTS 7.55 427 407 81 80 30 32 58 2 Standard 8:2 FTS 9.29 527 507 81 80 30 46 50 4 Standard 10:2 FTS 10.35 627 606.9 81 80 34 42 54 4 Standard 13C2 -4:2 FTS 5.87 329 309 24 2 Standard 13C2 -6:2 FTS 7.55 429 409 28 2 Standard 13C2 -8:2 FTS 9.29 529 509 28 4 Standard 13C2 -PFDoA 10.35 615 570 12 2 Standard 13C2 -PFTeDA 11.17 715 670 12 2 Standard 13C3 -HFPO-DA 6.15 287 169 4 5 Standard 13C3 -PFBS 5.39 302 80 44 2 Standard 13C3 -PFHxS 6.76 402 80 48 2 Standard 13C4 -PFBA 4.78 217 172 8 2 Standard 13C4 -PFHpA 6.72 367 322 8 2 Standard 13C5 -PFHxA 5.93 318 273 8 2 Standard 13C5 -PFPeA 5.29 268 223 4 2 Standard 13C6 -PFDA 9.3 519 474 8 2 Standard 13C7 -PFUnDA 9.88 570 525 8 2 Standard 13C8 -PFOA 7.6 421 376 8 2 Standard 13C8 -PFOS 8.52 507 80 54 2 Standard 13C8 -PFOSA 10 506 78 36 3 Standard 13C9 -PFNA 8.51 472 427 8 2 Standard TUDCA 6.8 498 124 80 53 80 4 Standard TCDCA 8.6 498 124 80 65 80 4 Standard TDCA 9.0 498 124 80 69 80 4 Standard 5 Method performance evaluation The novel passthrough cleanup using Captiva EMR PFAS Food II cartridges was evaluated in terms of matrix removal, target recovery, and repeatability during sample cleanup with the cartridge. The entire method was then validated, which included a calibration study, method LOQ determination, and recovery accuracy and precision. Due to the different requirements of the target LOQs, five prespiked QC-level samples were prepared in replicates of four or five at each level. In addition, the matrix blanks were prepared in replicates of five to seven for quantitation of the targets in the matrix control sample. This is important for accuracy evaluation, as the contribution from the matrix for some PFAS is unavoidable. Table 4 shows the matrix zero blanks and prespiked QCs with PFAS standard and ISTD spiking. Depending on the different concentration factors introduced through sample preparation, the actual spiking levels in the matrices varied. Infant Formula Milk Eggs Sample Size (g) 5 10 10 Concentration Factor 5x 10x 10x Matrix Spiked Samples Spiking Concentration (µg/kg) STD* ISTD STD* ISTD STD* ISTD Zero – 0.2 – 0.1 – 0.1 PR-QC 1 0.01 0.2 0.01 0.1 0.01 0.1 PR-QC 2 0.02 0.2 0.02 0.1 0.02 0.1 PR-QC 3 0.1 0.2 0.1 0.1 0.1 0.1 PR-QC 4 0.4 0.2 0.2 0.1 0.2 0.1 PR-QC 5 1.0 0.2 0.5 0.1 0.5 0.1 * Concentrations only indicate generic concentration of 28 PFAS targets. Concentrations of PFBA and PFPeA were 10x and 2x the generic concentrations, respectively. Table 4. Matrix-matched QC and matrix-zero samples in group II food matrices. Figure 1. Sample preparation procedure for PFAS analysis in infant formula, milk, and eggs. Cap and shake the sample on a Geno/Grinder at 1,500 rpm for 5 minutes. Transfer 5.4 mL supernatant to another 15 mL tube and mix with 0.6 mL water. Optional: Prewash the Agilent Captiva EMR PFAS Food II cartridges with 5 mL of 1:1 ACN/MeOH with 1% AA. Transfer 5 mL of supernatant mixture into MR PFAS Food II 6 mL cartridges. Elute by gravity and apply 10 psi for 2 minutes at the end to completely dry the sorbent bed. Vortex for 2 minutes, sonicate for 5 to 10 minutes, and centrifuge for 2 minutes. Weigh 5 to 10 g of homogenized sample into a 50 mL tube, spike ISTD and STD appropriately. For infant formula, add 10 mL of water. Vortex for 10 to 15 minutes. Add 10 mL of ACN with 1% acetic acid (AA). Vortex for 20 seconds to mix. Add QuEChERS EN extraction salt and two ceramic homogenizers. Equilibrate cartridge with 0.8 mL of corresponding sample. Centrifuge tubes at 5,000 rpm for 5 minutes. Collect eluent and dry at 50 °C in the CentriVap (or TurboVap). Reconstitute the dried sample with 450 µL of 80:20 MeOH:water. Elute by gravity and apply 9 to 12 psi at the end to completely dry the cartridge. Discard the eluent and replace with prelabelled collection tubes. 6 Results and discussion EMR mixed-mode passthrough cleanup The Captiva EMR PFAS Food cartridges provide comprehensive matrix removal after traditional QuEChERS extraction through a mixed-mode passthrough cleanup. It is a simple and efficient procedure to remove matrix interferences including carbohydrates, organic acids, pigments, fats and lipids, and other hydrophobic and hydrophilic matrix coextractives. The Captiva EMR PFAS Food I cartridges contain less sorbent with a simpler formula and are recommended for fresh and processed fresh foods of plant origin, such as fruits and vegetables, baby food, and juices. The EMR PFAS Food II cartridges contain more sorbent with a more complex formulation, and are recommended for fresh and processed fresh and dry foods of animal origin, such as milk, eggs, meat, fish, and infant formula, as well as some foods of plant origin like dry seed feed and food, and oils. Compared to a traditional dSPE cleanup used after QuEChERS extraction, EMR mixed-mode passthrough cleanup provided significant improvement on PFAS recovery and reproducibility. The PFAS recovery using the EMR passthrough cleanup was evaluated in representative food sample extracts including grape, baby food, infant formula, tuna, eggs, and soybean crude extract after QuEChERS extraction, and was compared to typical dSPE cleanup. Captiva EMR PFAS Food I cartridges were used for baby food and grape extract cleanup, and Captiva EMR PFAS Food II cartridges were used for soybean, infant formula, tuna and egg extract cleanup. Figure 2 shows the comparison results based on the average recovery of each target in each food matrix, demonstrating significant improvement in recovery using EMR mixed-mode passthrough cleanup compared to dSPE cleanup. The matrix removal during sample cleanup was also evaluated using a GC/MS full scan and an LC/Q-TOF total ion chromatogram (TIC) scan, as shown in the chromatogram comparison in Figures 3 and 4, respectively. The results demonstrate significant improvement in matrix removal using EMR mixed-mode passthrough cleanup. Figure 2. PFAS recovery comparison for food extract cleanup methods using either EMR mixed-mode passthrough cleanup or traditional dSPE cleanup. 7 Figure 3. Food matrix removal comparison between EMR mixed-mode passthrough cleanup versus traditional dSPE cleanups using GC/MS full scan for (A) infant formula sample and (B) egg sample for matrix blanks with (1) no cleanup, (2) EMR passthrough cleanup, (3) dSPE 1 cleanup, and (4) dSPE 2 cleanup. 0 0.5 1.0 1.5 2.0 2.5 3.0 Infant formula control 0 0.5 1.0 1.5 2.0 2.5 3.0 Infant formula EMR passthrough cleanup matrix blank 0 0.5 1.0 1.5 2.0 2.5 3.0 Infant formula dSPE 1 cleanup matrix blank 0 0.5 1.0 1.5 2.0 2.5 3.0 Infant formula dSPE 2 cleanup matrix blank 3 5 7 9 11 13 15 17 19 21 23 25 27 29 0 1 2 3 4 5 6 Egg control 0 1 2 3 4 5 6 Egg EMR passthrough cleanup matrix blank 0 1 2 3 4 5 6 Egg dSPE 1 cleanup matrix blank 0 1 2 3 4 5 6 Egg dSPE 2 cleanup matrix blank 3 5 7 9 11 13 15 17 19 21 23 25 27 29 ×108 ×108 ×108 ×108 ×108 ×108 ×108 ×108 Counts Counts Counts Counts Counts Counts Counts Counts Acquisition time (min) Acquisition time (min) A-1 B-1 A-2 B-2 A-3 B-3 A-4 B-4 Figure 4. Food matrix removal comparison between EMR mixed-mode passthrough cleanup versus traditional dSPE cleanups using LC/Q-TOF TIC (+) scan for infant formula sample (A) and egg sample (B). 0 1 2 3 4 5 6 7 8 LC/Q-TOF scan comparison for infant formula 0 1 2 3 4 5 6 7 LC/Q-TOF scan comparison for eggs 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 EMR mixed-mode passthrough cleanup dSPE 1 cleanup dSPE 2 cleanup Egg control without cleanup ×108 Counts ×108 Counts Acquisition time (min) A B EMR mixed-mode passthrough cleanup dSPE 1 cleanup dSPE 2 cleanup Infant formula control without cleanup 8 Besides the improvement on PFAS target recovery and matrix removal, another important feature provided by EMR mixed‑mode passthrough cleanup is the increased sample volume recovery. Sample volume recovery is usually not a concern in other common food safety analyses such as pesticide and vet drug analyses; however, it can be critical for PFAS analysis in food since the required LOQs are in the low- to mid-ppt level. The ultralow LOQs require the use of a postconcentration step to boost the method sensitivity. It is common to apply a 5 to 10 times postconcentration factor after sample cleanup using a dry-and-reconstitute step. As a result, the sample volume becomes important to achieve a high concentration factor and consistent reconstitution. Usually, the dSPE cleanup only provides ~ 50% sample volume recovery, which means the cleaning of 5 mL sample extract can only generate ~ 2.5 mL cleaned sample volume. However, the EMR mixed-mode cleanup volume recovery is > 90%, which means the cleaning of 5 mL sample extract delivers ~ 4.5 mL of sample. This large volume provides easy postconcentration and consistent sample reconstitution. Sample preparation procedure The use of EMR mixed-mode passthrough cleanup simplifies the entire sample preparation procedure with fewer steps, which saves time, effort, and consumables. The newly developed method includes two major processes: QuEChERS extraction and EMR passthrough cleanup, while the traditional method includes three major processes: QuEChERS extraction, dSPE cleanup, and WAX SPE extraction.5 Figure 5 shows a comparison of the two sample preparation method procedures. The WAX SPE step used in the traditional method was added to further clean the sample extract after dSPE cleanup. However, the SPE method is challenging to implement with the previous sample extraction and dSPE cleanup steps. The crude organic (ACN) extract needs to be switched in a solution containing at least 90% water before loading on the cartridge. This can be done by either drying and reconstituting in a highly aqueous solution, or by direct dilution with water, which either significantly increases sample loading time or the preparation time for loading in the WAX cartridges. The typical SPE procedure involving conditioning, equilibrium, loading, washing, and eluting also requires more time and uses more solvent. Given the same sample quantity for preparation, the traditional method can take up to triple the time of the new method. Also, fewer solvents and consumables are used in the new method compared to the traditional. Collectively, these benefits of the new method can improve overall lab productivity. Figure 5. Sample preparation procedure comparison for PFAS in food analysis using the newly developed method (A) versus the traditional method (B). QuEChERS extraction + EMR mixed-mode passthrough cleanup Weighing the sample (5 to 10 g) Optional: Dry sample hydration and equilibrium QuEChERS extraction EMR mixed-mode passthrough cleanup Sample postconcentration by drying and reconstitution QuEChERS extraction + dSPE cleanup + WAX SPE extraction Weighing the sample (5 g) Optional: Dry sample hydration and equilibrium QuEChERS extraction Collect 1 mL extract and dilute 10x with water Sample postconcentration by drying and reconstitution dSPE cleanup WAX SPE extraction and cleaning A B 9 Entire method validation The new method was validated for the determination of 30 PFAS targets in infant formula, milk, and eggs following the AOAC SMPR guidance. The requirements for PFAS target LOQs for the tested food matrices are listed in Table 5. Food Matrix LOQ (µg/kg) PFHxS, PFOA, PFNA, and PFOS PFBA and PFPeA Other PFAS Infant Formula ≤ 0.01 ≤ 1 ≤ 0.1 Eggs ≤ 0.3 ≤ 3 ≤ 3 Milk ≤ 0.01 ≤ 1 ≤ 0.1 Table 5. AOAC SMPR requirements for LOQs in infant formula, milk, and eggs. Method LOQs The three food matrices evaluated in the study all showed positive detection in matrix blanks. As a result, matrix background correction was necessary and was used in method validation for target recovery calculation. Matrix blanks were prepared in five to seven replicates, then the lowest method reportable LOQs were calculated according to the following equation: LOQcal = 10 × SDMBs Where: – LOQcal is the method's lowest reportable limit of quantitation (LOQ) – SDMBs is the standard deviation (SD) of detected incurred targets from five to seven replicates of matrix blanks The method LOQs were then decided based on the lowest validated QC spiking level that was equal to or above the lowest reportable LOQs. Table 6 shows the calculated lowest reportable LOQcal and validated method LOQval for each target in each matrix. For the core PFAS targets PFHxS, PFNA, and PFOS, the validated method LOQs were demonstrated to be below or equal to the required LOQs in all three tested matrices. The validated method LOQ for PFOA was below or equal to the required LOQs listed in milk and eggs. In infant formula, it was higher than the required LOQs due to matrix positive occurrence. For other PFAS targets, the validated method LOQs were demonstrated to be below or equal to the required LOQs in all three matrices, except higher LOQs for 6:2 FTS in milk due to positive occurrence. The PFOS LOQ in eggs was higher than other core PFAS targets due to the significant high interference of TCDCA at qualifier transition 498.9 to 79.9, which resulted in the failure of qualifier ratio identification at lower levels. However, the validated method LOQ (0.1 µg/kg) was still below the required 0.3 µg/kg LOQ in eggs. Figure 6 shows the chromatograms of matrix blanks and validated method LOQs for the core PFAS targets in infant formula, milk, and eggs. Target Infant Formula Milk Eggs LOQcal LOQval LOQcal LOQval LOQcal LOQval PFBA NA 0.1 NA 0.1 0.147 0.2 PFPeA 0.005 0.02 0.003 0.02 0.011 0.02 PFBS 0.002 0.01 0.001 0.01 NA 0.01 4:2 FTS NA 0.01 NA 0.01 NA 0.01 PFPeS NA 0.02 NA 0.02 0.005 0.01 PFHxA NA 0.01 NA 0.01 NA 0.01 HFPO-DA NA 0.01 NA 0.01 NA 0.01 PFHpA 0.001 0.01 0.002 0.01 0.001 0.01 PFHxS* 0.004 0.01 0.002 0.01 0.001 0.01 DONA 0.001 0.01 0.045 0.1 NA 0.01 6:2 FTS 0.007 0.01 0.322 0.5 0.006 0.01 PFOA* 0.016 0.02 0.004 0.01 0.006 0.01 PFHpS NA 0.01 NA 0.01 NA 0.01 PFNA* 0.005 0.01 0.002 0.01 0.003 0.01 PFOS* 0.002 0.01 0.001 0.01 0.003 0.1 9Cl-PF3ONS NA 0.01 NA 0.01 NA 0.01 8:2 FTS NA 0.01 NA 0.01 NA 0.01 PFNS 0.006 0.01 0.002 0.01 0.001 0.01 PFDA NA 0.01 NA 0.01 NA 0.01 PFDS 0.002 0.01 0.002 0.01 0.001 0.01 PFUnDA NA 0.01 NA 0.01 NA 0.01 PFOSA 0.001 0.01 0.001 0.01 NA 0.01 11Cl-PF3OUdS NA 0.01 NA 0.01 NA 0.01 PFUnDS NA 0.01 NA 0.01 NA 0.01 PFDoA 0.005 0.01 0.001 0.01 NA 0.01 10:2 FTS NA 0.01 NA 0.01 NA 0.01 PFDoS NA 0.01 NA 0.01 NA 0.01 PFTrDA 0.002 0.01 NA 0.01 NA 0.01 PFTrDS 0.004 0.01 NA 0.01 NA 0.01 PFTeDA NA 0.01 NA 0.01 NA 0.01 * Core PFAS targets Red text indicates the LOQval level is above the required LOQ level due to matrix impact. Table 6. Method lowest reportable calculated LOQ (LOQcal ) and validated LOQ (LOQval ) for 30 PFAS targets in infant formula, milk, and eggs. 10 Figure 6. Infant formula (top), milk (middle), and egg (bottom) matrix blanks and LOQ level chromatograms for the core PFAS targets: PFHxS, PFOA, PFNA, and PFOS. LOQ levels in each matrix are listed in Table 6. PFHxS PFOA PFNA PFOS Milk matrix blank Milk LOQ Infant formula matrix blank Infant formula LOQ Eggs matrix blank Eggs LOQ Method calibration The use of 18 PFAS isotopically labeled ISTDs allows the same standard calibration curve to be used for PFAS quantitation in different food matrix samples. Therefore, a matrix-matched calibration curve is not needed for each food matrix. This significantly increases sample testing productivity, saving time and cost of labor and materials, and improving sample analysis consistency. The calibration curve range was decided based on the required LOQs in the food matrices, the concentration factor introduced through sample preparation, and the instrument method sensitivity. Due to the higher detection levels required for infant formula, milk, and eggs, a calibration set range from 20 to 10,000 ng/L was used. The results confirmed a 500x calibration curve dynamic range with correlation coefficient R2 > 0.99 for all 30 PFAS targets. 11 Method accuracy and precision Method recovery and repeatability were validated in infant formula, milk, and eggs. The acceptance criteria for eggs is 80 to 120% recovery and for infant formula and milk is 65 to 135% recovery for PFOS, PFOA, PFHxS, and PFNA.5 For other PFAS targets in the three matrices, the acceptance recovery criteria is also 65 to 135% for targets with corresponding isotopic ISTD, and 40 to 140% for targets without corresponding ISTD. The repeatability (RSD%) acceptance is ≤ 20% for core PFAS in eggs, ≤ 25% for core PFAS in milk and infant formula, and ≤ 25% for all other PFAS targets with corresponding isotopic ISTD, and ≤ 30% for other PFAS targets without corresponding isotopic ISTD in all three matrices. The final method validation results included three QC levels in each matrix, including LOQ, mid, and high QC levels. The validated method LOQs are listed in Table 6. The mid-level QCs are reported at 5 to 10 times the LOQ, and the high-level QCs are reported at 20 to 50 times the LOQ. There is one exception for 6:2 FTS in milk, where only one level at 0.5 µg/kg was reportable due to significant high positive detection in sample matrix control. Figure 7 shows the method validation recovery and repeatability (RSD) summary for PFAS analysis in infant formula, milk, and eggs. Overall, the method delivered acceptable recovery and repeatability results for all 30 targets in tested food matrices that meet the acceptance requirements. The core PFAS targets all generated acceptable recovery (80 to 120%) and RSD (< 20%) for all spiking levels in all matrices. Targets with corresponding isotopically labeled ISTD generated better quantitation results than targets without corresponding isotopically labeled ISTD. Figure 7. Method validation recovery (Rec) and repeatability (RSD%) summary for PFAS analysis in infant formula, milk, and eggs. www.agilent.com DE90981738 This information is subject to change without notice. © Agilent Technologies, Inc. 2024 Printed in the USA, July 2, 2024 5994-7366EN Conclusion A simplified, rapid, and reliable method using QuEChERS extraction followed by mixed-mode passthrough cleanup with the Agilent Captiva EMR PFAS Food II cartridge and LC/MS/MS detection was developed and validated for 30 PFAS targets in infant formula, milk, and eggs. The novel cleanup method demonstrated a significant improvement over traditional dSPE cleanup in terms of matrix removal, PFAS recovery, and sample volume recovery. This method is also simpler, saving time and effort, and thus improves overall lab productivity. The entire method was validated with acceptance criteria, and method performance was shown to meet the requirements described in AOAC SMPR 2023.003. References 1. EUR-Lex (2023) Consolidated text: Commission Regulation (EU) 2023/915 of 25 April 2023 on Maximum Levels for Certain Contaminants on Food and Repealing Regulation (EC) No 1881/2006. 2. AOAC (2023) Standard Method Performance Requirements (SMPRs) for Per- and Polyfluoroalkyl Substances (PFAS) in Produce, Beverages, Dairy Products, Eggs, Seafood, Meat Products, and Feed (AOAC SMPR 2023.003) 3. EPA method 533 Determination of Per- and Polyfluoroalkyl Substances in Drinking Water by Isotope Dilution Anion Exchange Solid Phase Extraction and Liquid Chromatography/Tandem Mass Spectrometry (EPA 533:2019). 4. EPA method 1633 Analysis of Per- and Polyfluoroalkyl Substances (PFAS) in Aqueous, Solid, Biosolids, and Tissue Samples by LC-MS/MS (EPA 1633:2024). 5. Genualdi, S.; Young, W.; Peprah, E.; et al. Analyte and Matrix Method Extension of Per- and Polyfluoroalkyl Substances in Food and Feed, Anal. and Bioanal. Chem. 2024, 416, 627–633. doi: 10.1007/s00216-023-04833-1. 6. Hwang, S. H.; Ryu, S. Y.; Seo, D.; et al. Development of a Method Using QuEChERS and LC-MS/MS for Analysis of Per- and Polyfluoroalkyl Substances in Rice Matrix, Food Chem. 2024, 445, 138687. doi: 10.1016/j. foodchem.2024.138687.