Measuring PFAS in a variety of food samples at different price points
Karl Oetjen, Simon Roberts, Igor Zakharevich
SCIEX, USA
Abstract
Recent studies have shown that per- and polyfluoroalkyl substances (PFAS) accumulate in the food chain and are found in a wide variety of food products, including fish, meat, dairy and vegetables.1 As a result, there is a growing concern about the potential health risks associated with consuming PFAS-contaminated foods. As the public becomes more aware of these compounds, they might make purchases that they believe would decrease the amount of potential exposure to PFAS. Many people may assume that more expensive alternatives might represent a healthier or less PFAS-contaminated option.
Introduction
This study explored the differences in PFAS concentrations in several food types at different price points. Eight different food products were selected to represent a wide range of food types, including ground beef, cake, salmon, tuna, potato chips, seltzer, beer and ice cream. For each type of food, low-, mid- and high-priced options were purchased for analysis. The samples were analyzed using the SCIEX 7500 system using a large volume injection, PFAS compounds were detected in food in the low parts per trillion (ppt) or picogram per gram (pg/g) depending on the analyte.
Figure 1. The foods sampled for this study ranged from solid, protein-rich foods, like fish and ground beef to liquid beverages like beer and seltzer and highly processed foods like chips and cake. The study sought to determine PFAS levels across very different matrices and whether price plays a factor in the distribution.
Methods
Sample preparation
Eight different food commodities were selected to represent a wide range of food types. For each type of food, a low-, mid- and high-priced option was purchased for analysis. A 5.0 g of homogenized sample was fortified with isotopically labeled standards, then diluted with water (5 mL for fruits and vegetables, 15 mL for dry foods, 25 mL for powders and 0 mL for liquids). Buffer salts consisting of 6 g of MgSO4 and 1.5 g of sodium acetate were added along with 10 mL of acetonitrile, followed by vortexing and separation by centrifugation (Figure 2). 1.5 mL of the supernatant was placed a 2 mL centrifuge tube containing 150 mg MgSO4, 50 mg PSA and 50 mg GCB and vortexed vigorously before centrifugation. From here, 1 mL of the supernatant was removed and blow down under a gentle stream of nitrogen before being reconstituted to 200 µL using 80:20 (v/v), methanol/water.
Liquid chromatography
The analysis of PFAS by LC-MS/MS poses several unique challenges. In particular, the ubiquitous presence of these compounds in indoor environments complicates their removal from blanks and analytical systems (Figure 3). PFAS contamination can lead to reduced method sensitivity, contaminated blanks, inaccurate quantitation results and the need for system cleaning or sample reinjection. Several strategies were employed to address this including incorporating a delay column and substituting fluoropolymer components with PFAS-free LC kits employing PEEK and stainless-steel alternatives are integral components of a typical PFAS method. In this study, a large volume injection of 20 μL was used to avoid potential sources of background PFAS contamination from sample preparation steps like evaporation while maintaining method sensitivity. A major concern with large volume injections of samples reconstituted with a high concentration of organic solvent is peak splitting. No peak splitting was observed with early eluting compounds like perfluorobutane sulfonate (PFBS), maintaining an asymmetry factor of 1.0 due to the use of a mixing column referred to as an AnaCondA (Figure 4).
Figure 2. QuEChERS extraction of eight different types of food matrices.
Figure 3. Extracted ion chromatograms (XICs) of perfluorooctanoic acid (PFOA) and perfluorooctane sulfonic acid (PFOS) in a 0.5 ppt standard compared to extraction blanks in which PFOA and PFOS were not detected.
Figure 4. PFBS peak shape using large volume injection at 5 ppt. Asymmetry remained 1.0 across all calibrators.
Chromatographic separation was carried out using the ExionLC 2.0 system. The chromatography column used was a Phenomenex Luna Omega PS C18 (3 µm, 100 x 3 mm Part No.: 00D-4758-Y0) maintained at 40°C and the delay column was a Phenomenex Luna Omega PS C18 (5 μm, 50 x 4.6 mm Part No.: 00B-4753-E0). Mobile phases A and B were water with 10mM ammonium acetate and 60:40 (v/v), acetonitrile/methanol with 10mM ammonium acetate respectively. At a constant flow rate of 0.8 mL/min solvent B was ramped using a linear gradient from 5% to 35% from 0.5 to 1 minutes. Solvent B was then ramped again to 50% by 4 minutes where it was brought to 80% at 7 minutes. Finally, solvent B was ramped to 95% at 10 minutes and held for 3 minutes.
Mass spectrometry
The SCIEX 7500 system was used in negative polarity multiple reaction monitoring (MRM) mode to detect the targeted panel of PFAS compounds (Table 1). Method settings include curtain gas set to 45 psi, ion source gas 1 set to 35 psi, source gas 2 set to 70 psi, ion spray voltage set to -1250 V and ion source temperature set to 300°C. Using the OptiFlow Pro ion source on the SCIEX 7500 system, the source and gas settings were optimized for PFAS. In particular, the optimal source temperature was observed to change across different PFAS classes. For example, the perfluoroalkane sulfonic acids (PFSAs) tend to ionize better with higher temperatures, whereas the perfluoroalkyl carboxylic acids (PFCAs) tend to ionize at lower optimal temperatures. A temperature of 300°C was observed to yield the best overall results across the entire PFAS panel.
Results
Detecting PFAS in foods of different prices
Most of the sampled food types contained detectable levels of at least one PFAS species. Only the beer samples were found to contain no detectable PFAS at any of the price points. Figure 1 summarizes the total summed PFAS levels found in the food samples according to each type. The protein-rich animal products like salmon, tuna and beef demonstrated the highest PFAS contamination while the seltzer exhibited the lowest amount of PFAS contamination.
Next, the question of variations in contamination among differently priced options was investigated. Figures 6-13 display the measured concentrations of individual PFAS species in each sample and compare total PFAS concentrations across different price points. The European Union Reference Laboratory has required limits of quantitation (LOQs) for several foods including meat and fish.2 The LOQs for PFOS, PFOS, PFNA, and PFHxS were set at ≤ 0.1 µg/kg.2 All three tuna samples and one salmon sample exhibited concentrations above this LOQ with the highest concentration of 0.156 µg/kg observed in the middle-priced tuna sample. The lowest amount of PFOS was observed in the cheapest ground beef sample at a concentration of 0.01 µg/kg (Figure 5).
An interference was observed in all meat and fish samples when monitoring the 499 > 80 MRM transition for PFOS, which was resolved by using the 499 > 99 transition instead. PFOA was not found in any of the ground beef samples, but was found in all three tuna samples and one of the salmon samples at concentration ranges of 0.002–0.028 µg/kg. Finally, PFNA was observed in all the tuna samples and one of the salmon samples at concentrations ranging from 0.001–0.162 µg/kg, with the cheap tuna containing the highest concentration. PFHxS was not detected in any of the meat and fish samples.
Figure 5. Example chromatograms in food samples. Lowest detection of PFOS, PFOA and PFNA in fish and meat samples.
Table 1. PFAS compounds that were monitored.
Figure 6. PFAS in cupcakes at different price points. PFOA, PFOS, and PFPeS were detected in the cupcake samples. The cheap cupcakes had the greatest total PFAS concentration, but only PFOA and PFPeS were detected, with the latter detected at about 50 ppt. Both the middle-priced and most expensive cakes contained mostly PFOS at about 10 ppt.
Figure 7. PFAS in ice cream at different price points. Only PFOS was detected in the ice cream samples. The cheapest ice cream had no detectable PFAS, the middle-priced ice cream had < 5 ppt PFOS and the most expensive ice cream had the highest PFOS concentration of about 20 ppt.
Figure 8. PFAS in tuna at different price points. The tuna had the highest total PFAS compared to the other food types. The cheapest tuna contained the greatest amount of PFAS compounds, however, all samples contained PFASs.
Figure 9. PFAS in ground beef at different price points. Four PFAS species were detected across the beef samples, but PFPeS was measured at the highest concentration in the middle-priced sample at nearly 900 ppt. The cheapest beef only contained PFOS at < 25 ppt. The most expensive beef was found to contain mostly PFOS, but still with a total PFAS of < 100 ppt.
Figure 10. PFAS in salmon at different price points. Several PFAS species were detected in each of the salmon samples. The middle-priced salmon had the greatest total PFAS concentration as well as the most diverse PFAS distribution. PFUdA was the predominant compound detected in the middle-priced sample.
Figure 11. PFAS in beer of different price points. PFAS was not detected in any of the beer samples.
Figure 12. PFAS in seltzer at different price points. Only PFOS and PFOA were detected in the seltzer samples; the cheap seltzer had no detectable PFAS, the middle-priced seltzer had only PFOA at <10 ppt and the most expensive seltzer contained both PFOS and PFOA. The sum of both species was <16 ppt.
Figure 13. PFAS in chips at different price points. Only PFOA and PFDoA were detected in the chips samples; the cheap chips had only PFDoA, the middle-priced chips had only PFOA at <10 ppt and a sum of both species at 15ppt. The most expensive chips had the lowest PFDoA with a sum of both species at <10 ppt.
Conclusion
Diet represents one possible route of human exposure to PFAS. In response to the growing awareness and concern about the environmental persistence and potential toxicity of PFAS, experts are questioned about the exposure risks to these chemicals and how to avoid them. This study aimed to not only demonstrate the capability of the SCIEX 7500 system to measure PFAS at very low levels in diverse and complex food matrices, but also to explore the differences in contamination levels of several food types at different price tiers. While it may be common to presume that more expensive food options might contain less PFAS, there was no readily observable trend between cost and total measured PFAS to support that assumption. As expected, the animal-based food products exhibited the highest total PFAS concentrations with most samples containing 3 or more PFAS compounds at quantifiable levels. In contrast, the lowest total PFAS values were found in seltzer, while no PFAS was detected in beer.
References
- K. L., Saab, N., Silva, P., Curtzwiler, G., Steketee, A. (2021). Risk assessment of per-and polyfluoroalkyl substances (PFAS) in food: Symposium proceedings. Trends in Food Science 38; Technology, 116, 1203– 1211.
- European Union Reference Laboratory. (2022). Guidance Document on Analytical Parameters for the Determination of Per-and Polyfluoroalkyl Substances (PFAS) in Food and Feed Version 1.2. https://eurlpops.eu/core-working