CAM Method Tables-03242020
CAM Method Tables-03242020
CAM Method Tables-03242020
01
VALIDATION STATUS: Single-laboratory validation per the Guidelines for the Validation of
Chemical Methods for the FDA FVM Program 2nd Ed.
METHOD SUMMARY/SCOPE:
Analyte(s): Perfluorobutanoic acid, Perfluoropentanoic acid, Perfluorohexanoic acid,
Perfluoroheptanoic acid, Perfluorooctanoic Acid, Perfluorononanoic acid,
Perfluorodecanoic acid, Perfluorobutanesulfonic acid, Perfluoropentanesulfonic acid,
Perfluorohexanesulfonic acid, Perfluoroheptanesulfonic acid, Perfluorooctanesulfonic
acid, Sodium dodecafluoro-3H-4, 8-dioxanonanoate, 2,3,3,3-Tetrafluoro-2-(1,1,2,2,3,3,3-
heptafluoropropoxy) propanoic acid (GenX), Potassium 9-chlorohexadecafluoro-3-
oxanonane-1-sulfonate, 11-chloroeicosafluoro-3-oxaundecane-1-sulfonic acid
The test sample is homogenized and fortified with isotopically labeled surrogates prior to the
addition of water. The PFAS are extracted from the food samples using acetonitrile and formic
acid. Following extraction, a modified QuEChERS extraction technique is performed. The
resulting extract is filtered and fortified with internal standard solution and analyzed using LC-
MS/MS. Some matrices require the extract to be concentrated using nitrogen prior to addition of
the internal standard solution. The PFAS compounds are identified by multiple reaction mode
(MRM) transitions and retention time matching with the calibration standards. Ion ratios, when
available, are used to confirm the identity. In some cases, further clean-up using solid phase
extraction may be required. The concentration of each PFAS is determined using the response
ratio of the PFAS quantitation transition to that of the relevant labeled surrogate standard (SS).
The concentration is calculated by preparing a calibration curve using response ratios versus
concentration ratios for native analytes to that of their labeled-SS. During analysis, quality
1
control samples and method blanks must be analyzed. Analyte response in method blanks must
be subtracted from the sample response prior to final quantitation. After determination of the
concentration from the curve, the concentration must be adjusted for dilution and starting sample
mass.
REVISION HISTORY:
OTHER NOTES:
2
Determination of 16 Perluoroalkyl and Polyfluoroalkyl
Substances (PFAS) in Food using Liquid Chromatography-
Tandem Mass Spectrometry (LC-MS/MS)
Version 2019 (2019)
GLOSSARY
Table of Contents
2019.3 PRINCIPLE
2019.4 REAGENTS
2019.5 STANDARDS
2019.8 APPARATUS/INSTRUMENTATION
2019.9 METHOD
2019.10 CALCULATIONS
2019.12 REFERENCES
3
2019.1 METHOD TITLE: Determination of 16 Per and Polyfluoroalkyl
Substances (PFAS) in Food using Liquid Chromatography-Tandem Mass
Spectrometry (LC-MS/MS)
The method describes a procedure for measuring 16 PFAS in food using LC-MS/MS. The method has
been single laboratory validated in the following food matrices:
• This method should be used by analysts experienced in the use of LC-MS/MS, including but not
limited to operation of the instrumentation and software, data analysis and reporting results.
• Analysts should also be able to identify chromatographic and mass spectrometric interferences
during sample analysis and take necessary actions following validated procedures for their
correction to achieve reliable identification and quantitation.
• The method should be used only by personnel thoroughly trained in the handling and analysis of
samples for the determination of trace contaminants in food and beverage products. PFAS
chemicals are prevalent in all laboratory environments and special care must be taken to
prevent false positives due to accidental and/or routine laboratory contamination.
• Only LC-MS grade solvents should be used unless otherwise noted in the procedure below. All
solvents and complete method blanks should be analyzed on the LC-MS/MS instrument prior to
sample analysis. If PFAS compounds are determined, complete method blank results should be
subtracted from samples. Complete method blanks should be performed and analyzed daily,
preferably in the same instrument sequence as the samples. Sources of potential
contamination during sample preparation include; solvents, syringe filters, centrifuge tubes,
dSPE sorbents, septa, and others.
• A delay column should be used between the mobile phase mixer and sample injector to
temporarily trap any system related interferences, which results in their elution at a later
retention time than the analyte. This eliminates contamination from instrument tubing, mobile
phase solvents, and solvent bottles.
• Due to the extreme low concentrations of detection required for this analysis, choice of MS/MS
instrumentation is critical. Our analysis has been performed using Sciex 6500 and 6500 plus
instrumentation platforms. We have not fully evaluated any Orbitrap MS systems, but
preliminary investigations have not demonstrated adequate lower levels of quantitation (LLOQ)
for these systems.
• The analyte 11Cl-PF3OUdS exhibits known issues with recovery in certain matrices, which may
reduce the confidence in this result in certain food types.
2019.3 PRINCIPLE
The test sample is homogenized and fortified with isotopically labeled surrogates prior to the addition of
water. The PFAS are extracted from the food samples using acetonitrile and formic acid. Following
4
extraction, a modified QuEChERS extraction technique is performed. The resulting extract is filtered and
fortified with internal standard solution and analyzed using LC-MS/MS. Some matrices require the
extract to be concentrated using nitrogen prior to addition of the internal standard solution. The PFAS
compounds are identified by multiple reaction mode (MRM) transitions and retention time matching
with the calibration standards. Ion ratios, when available, are used to confirm the identity. In some
cases, further clean-up using solid phase extraction may be required. The concentration of each PFAS is
determined using the response ratio of the PFAS quantitation transition to that of the relevant labeled
surrogate standard (SS). The concentration is calculated by preparing a calibration curve using response
ratios versus concentration ratios for native analytes to that of their labeled-SS. During analysis, quality
control samples and method blanks must be analyzed. Analyte response in method blanks must be
subtracted from the sample response prior to final quantitation. After determination of the
concentration from the curve, the concentration must be adjusted for dilution and starting sample mass.
2019.4 REAGENTS
The use of trade names in this method constitutes neither endorsement nor recommendation by the
U.S. Food and Drug Administration (FDA). Equivalent performance may be achievable using apparatus
and materials other than those cited here.
• Formic acid, reagent grade >95%-- (Sigma Aldrich St. Louis, MO)
• LC/MS grade Optima water (Fisher Scientific, Hampton, NH)
• LC/MS grade Optima acetonitrile (Fisher Scientific, Hampton, NH)
• LC/MS grade Optima methanol (Fisher Scientific, Hampton, NH)
• Acetic acid, ammonium salt, 98% for analysis (Acros Organic, Geel, Belgium)
• Original QuEChERS extraction salt ECMSSCFS-MP with 6000 mg MgSO4 and 1500 mg NaCl (UCT,
Bristol, PA)
• QuEChERS dSPE ECMPSCB-MP with 900 mg MgSO4, 300 mg PSA, 150 mg graphitized carbon
black (UCT, Bristol, PA) or ECMPSCB15-CT prefilled units
• Ammonium hydroxide, certified ACS Plus 14.8N (Fisher Scientific, Hampton, NH)
2019.5 STANDARDS
• Isotopically labeled PFAS analytical standards (Wellington laboratories, Guelph, ON, Canada)
• Native PFAS analytical standards (Wellington laboratories, Guelph, ON, Canada)
2019.6.1 Prepare native PFAS stock solution at 1 µg/mL and 0.01 µg/mL
• Add 0.2 mL of each 50 µg/mL PFAS analytical standard (16 native compounds in Table 1) to 6.8
mL methanol. In the resulting solution, each compound has a concentration of 1 µg/mL in
methanol. Individual PFAS 50 µg/mL methanol standards were purchased from Wellington, but
other sources are acceptable.
5
• Add 0.1 mL of 1 µg/mL stock solution to 9.9 mL of methanol to produce a 0.01 µg/mL stock
solution.
2019.6.2 Prepare isotopically labeled PFAS surrogate stock solution (SS) at 1 µg/mL
• Add 0.2 mL of each 50 µg/mL analytical standard (7 isotopically labeled PFAS in Table 1) to 8.6
mL methanol. Individually labeled PFAS 50 µg/mL methanol standards were purchased from
Wellington but other sources are acceptable. This stock solution was used for both sample
analysis and calibration curve preparation.
• The stock solution of the 1 ng/mL calibration standard was used as the CCV standard (Table 2).
• Add 6 mL of a 14.8 N ammonium hydroxide solution to 1000 mL volumetric flask and fill to
volume with acetonitrile
6
Table 1. PFAS native, surrogate, and internal standard compounds
Calibration standards are prepared at concentrations of 0.01, 0.05, 0.10, 0.50, 1.0, 5.0, 10, 25, and 50
ng/mL according to the table below.
7
Table 2. Calibration standard preparation
*This curve was initially created to match surrogate and internal standard concentrations (10 ng/mL) in
the final extracts of samples that require N2 concentration. If this method is going to be primarily used
for samples that are not concentrated, a modification may be preferred to match the surrogate and
internal standard concentrations (1 ng/mL) in these extracts.
The edible portion of the food sample was collected and homogenized using an IKA tube mill
with a disposable 100 mL polypropylene grinding chamber. Samples were ground at 5000 rpm
for approximately 2 minutes. The minimum sample size for analysis is 5 grams. Sample
composites from different brands of the same product is acceptable.
2019.8 APPARATUS/INSTRUMENTATION
Equipment:
• IKA tube mill 100 control (IKA Works Inc, Wilmington, NC)
• Digital pulse mixer/vortexer (Glas-Col, Terre Haute, IN) capable of 1500 rpm with pulse
• Sorvall legend XTR centrifuge (Thermo Fisher Scientific, Waltham, MA)
• Nitrogen evaporation system (Organomation, Berlin, MA)
• Nexera X2 (Shimadzu, Kyoto, Japan) with binary pump, degasser, autosampler, and
thermostatted column compartment
• A Sciex 6500 plus QTRAP hybrid triple quadrupole/linear ion trap mass spectrometer with an
electrospray ESI ion source (Sciex, Toronto, ON Canada)
• Analyst® Software version 1.6.3
8
Supplies:
2019.9 METHOD
QuEChERS (Quick, easy, cheap, effective, rugged, safe) is used for the extraction of PFAS from foods.
Due to the high variability of the sample matrix, sample preparation steps are based on food type.
yes -take 5 mL of
Milk 5.0 mL 5 10
extract to 0.5 mL
Cheese 1.0 g 5 10 no
Other Dairy 5.0 g 5 10 no
Meat 5.0 g 5 10 no
9
• Add amount of sample based on Table 1 and commodity type to a 50 mL polypropylene (PP)
centrifuge tube
• Add 10 µL of 1 µg/mL isotopically labeled surrogate standard solution to the sample.
• Add amount of LC/MS grade Optima water based on Table 3 to the 50 mL PP conical centrifuge
tube
• Add 10 mL acetonitrile to the 50 mL PP conical centrifuge tube
• Add 150 µL formic acid to the 50 mL PP conical centrifuge tube
• Shake vigorously for 1 minute
• Add QuEChERS salt packet (Original extraction salt ECMSSCFS-MP from UCT with 6000 mg
MgSO4 and 1500 mg NaCl)
• Place on Glas-Col shaker at 1500 rpm with pulse set to 70 for 5 minutes
• Centrifuge for 5 minutes at 10000 rcf
• Add supernatant to 15 mL PP conical centrifuge tube with dSPE sorbent (ECMPSCB-MP from UCT
with 900 mg MgSO4, 300 mg PSA, 150 mg graphitized carbon black)
• Vortex/shake for 2 minutes
• Centrifuge 5 minutes at 10000 rcf
• Filter 5 mL of the extract with a 0.2 µm nylon syringe filter and transfer to a 15 mL conical
centrifuge tube
• For samples that do not require nitrogen concentration:
o Add 5 µL of 1 µg/mL d5-N-EtFOSAA internal standard solution to the 5 mL extract to give
a final concentration of 1 ng/mL. Surrogates will also have a final concentration of 1
ng/mL in the final extract.
• For samples that require nitrogen concentration:
o Concentrate to near dryness with nitrogen and reconstitute to 0.5 mL with methanol.
o Add 5 µL of the 1 µg/mL d5-N-EtFOSAA internal standard solution to give a final
concentration of 10 ng/mL in solution. Surrogates will also have a final concentration of
10 ng/mL in the final extract.
• Briefly vortex/shake.
• Transfer 100 µL to a Thomson nano filter vial with 0.2 µm nylon® filter and a PP screw cap (Sun
Sri) to run using LC-MS/MS
2019.9.2 Clean-up of extract using weak anion exchange solid-phase extraction (SPE) column
Due to the complexity of food samples and the possibility of matrix interferences, any samples with a
positive detection above the method detection limit for any compound was run through an additional
SPE step.
• Take 1 mL of filtered QuEChERS extract and dilute to ~ 15 mL with LC Optima water in a clean 15
mL PP conical centrifuge tube
• Condition a Strata™-XL-AW 100 µm column with 9 mL of 0.3% ammonium hydroxide in
acetonitrile
• Add sample to column and let pass through
• Add 5 mL of LC Optima water to wash column
• Let column dry 1 minute
• Add 4 mL of 0.3% ammonium hydroxide in acetonitrile to elute analytes into a clean 15 mL PP
conical centrifuge tube
• Blow to near dryness
10
• Reconstitute to 1 mL with methanol and transfer to a clean PP nano filter vial
All samples were analyzed using a liquid chromatograph (Nexera X2, (Shimadzu, Kyoto, Japan)). The
MS/MS data was acquired using scheduled MRM with an AB Sciex 6500 plus QTRAP.
Set up the LC-MS/MS method with the following parameters and monitor for the transitions using the
information in the table below.
Table 4. MS/MS Conditions for the Monitored Transitions on a 6500 plus QTRAP
Internal Standard
Retention DP EP CE
ID Q1 mass (m/z) Q3 mass (m/z) CXP (volts)
Time (min) (volts) (volts) (volts)
NN EtFOSAAa 21.2 589 419 -50 -10 -30 -20
NN EtFOSAA 21.2 589 219 -50 -10 -38 -20
Surrogates
M3 PFBAa 3.8 216 172 -17 -8 -12 -14
MPFHxA a
10.5 315 270 -13 -10 -14 -12
13C PFOA a
15.9 421 172 -19 -5 -25 -7
13C PFOA 15.9 421 376 -36 -8 -13 -20
M3 PFBSa 7.7 302 99 -85 -6 -36 -8
M3 PFBS 7.7 302 80 -88 -6 -73 -9
MPFHxSa 13.6 403 103 -60 -10 -81 -15
MPFHxS 13.6 403 169 -60 -10 -42 -15
13C PFOSa 18.1 507 80 -100 -5 -125 -15
13C PFOS 18.1 507 99 -100 -5 -100 -15
M3 HFPOa 11.6 332 287 -10 -8 -9 -13
M3 HFPO 11.6 332 169 -10 -8 -17 -11
11
Table 4. MS/MS Conditions for the Monitored Transitions on a 6500 plus QTRAP (cont.)
Natives
Retention DP EP CE
ID Q1 mass (m/z) Q3 mass (m/z) CXP (volts)
Time (min) (volts) (volts) (volts)
12
Table 5. Gradient Profile for the LC Conditions
Time
Concentration of B
(min)
0.01 40%
1 40%
25 90%
25.1 40%
26.1 40%
• Curtain gas: 40 au
• Collisionally activated dissociation (CAD) gas: medium
• Ion spray voltage: -4500 V
• Source temperature: 350 °C
• Gas 1 pressure: 50 au
• Gas 2 pressure: 50 au
• Injection volume: 5 µL
• Column temperature: 40 °C
• Flow rate: 0.30 mL/min
For every 6 samples analyzed, a CCV standard (typically 1 ng/mL) is run to check for accuracy. The
accuracy of the calculated concentration of the CCV should be statistically evaluated, which can typically
be within 70-120 % of the original value. If the accuracy falls outside this range, the calibration curve is
rerun, and any test samples run since the last successful CCV are remeasured.
2019.10 CALCULATIONS
0.01 𝑛𝑛𝑛𝑛
∗ 0.5 𝑚𝑚𝑚𝑚 = 0.005 𝑛𝑛𝑛𝑛
𝑚𝑚𝑚𝑚
13
• The amount (ng) in 0.5 mL is also the same amount (ng) in 5 mL of extract since this solution was
concentrated from 5 mL to 0.5 mL. Because the total extract was 10 mL, the amount (ng) in the
10 mL extract is equal to twice that of the 5 mL extract.
Example calculation for concentration measured on LC-MS/MS to concentration in 5 grams of food with
a final extract of 10 mL:
0.01 𝑛𝑛𝑛𝑛
∗ 0.5 𝑚𝑚𝑚𝑚 = 0.005 𝑛𝑛𝑛𝑛
𝑚𝑚𝑚𝑚
• Since there are 0.005 ng in 0.5 mL of extract, there would be 0.1 ng in the total 10 mL extract
10 𝑚𝑚𝑚𝑚
0.005 𝑛𝑛𝑛𝑛 ∗ = 0.1 𝑛𝑛𝑛𝑛
0.5 𝑚𝑚𝑚𝑚
An example chromatogram is included below of a calibration standard with native and labeled PFAS
concentrations at 10 ng/mL.
14
Analyst software is used to prepare a linear standard curve where x is the concentration ratio
(analyte/SS) and y is the instrument response ratio (analyte/SS) with 1/x weighting. Surrogates and their
internal standard pairs are listed in Table 6, which are used to calculate absolute recoveries of the
surrogate standards over the entire extraction method. Surrogates and their native analyte pairs are
also listed in Table 6 with their curve fit. Adjustments are made in Analyst for the internal standard
concentration. The calibration curve has surrogate and internal standard concentrations of 10 ng/mL as
well as samples concentrated with nitrogen. Samples that are not concentrated with nitrogen have
surrogate and internal standard concentrations of 1 ng/mL.
Table 6. Analytes with calibration curve fit and surrogates used as the internal standard
Surrogates
M3 PFBA a
NN EtFOSAAa mean response factor none
MPFHxA a
NN EtFOSAA a
mean response factor none
13C PFOA a
NN EtFOSAA a
mean response factor none
13C PFOA NN EtFOSAAa mean response factor none
M3 PFBS a
NN EtFOSAA a
mean response factor none
M3 PFBS NN EtFOSAA a
mean response factor none
MPFHxS a
NN EtFOSAAa mean response factor none
MPFHxS NN EtFOSAA a
mean response factor none
13C PFOS a
NN EtFOSAA a
mean response factor none
13C PFOS NN EtFOSAA a
mean response factor none
M3 HFPO a
NN EtFOSAA a
mean response factor none
M3 HFPO NN EtFOSAA a
mean response factor none
15
Table 6. Analytes with calibration curve fit and surrogates used as the internal standard (cont.)
Natives
PFBA a
M3 PFBA a
Linear 1/x
PFPeA a
MPFHxA a
Linear 1/x
PFHxA a
MPFHxA a
Linear 1/x
PFHpA a
MPFHxA a
Linear 1/x
PFOA a
13C PFOA a
Linear 1/x
PFOA 13C PFOA a
Linear 1/x
PFNA a
13C PFOA a
Linear 1/x
PFNA 13C PFOAa Linear 1/x
PFDA a
13C PFOA a
Linear 1/x
PFDA 13C PFOA a
Linear 1/x
PFBS a
M3 PFBSa Linear 1/x
PFPeS a
MPFHxS a
Linear 1/x
PFPeS MPFHxS a
Linear 1/x
PFHxS a
MPFHxSa Linear 1/x
PFHxS MPFHxS a
Linear 1/x
PFHpS a
MPFHxS a
Linear 1/x
PFHpS MPFHxS a
Linear 1/x
PFOS a
13C PFOS a
Linear 1/x
PFOS 13C PFOS a
Linear 1/x
NaDONA a
13C PFOA a
Linear 1/x
NaDONA 13C PFOA a
Linear 1/x
HFPO-DA a
M3 HFPO a
Linear 1/x
HFPO-DA M3 HFPO a
Linear 1/x
9Cl-PF3ONS a
MPFHxSa Linear 1/x
11Cl-PF3OUdS a
MPFHxS a
Linear 1/x
11Cl-PF3OUdS MPFHxS a
Linear 1/x
a
Primary MRM transition used for quantitation
Single lab validation. A level 2 validation was conducted under the Guidelines for the Validation of
Chemical Methods for the FDA FVM Program 2nd Ed. A total of 4 different types of foods and beverages
were evaluated. These include produce, milk, fish, and bread. The method was validated at 6
concentrations (0.05, 0.15, 0.5, 1.5, 2, 5 ng/mL) in 4 food matrices. Acceptable recovery ranges for
these compounds based on the FDA guidelines for the validation of chemical methods is 40-120% for
concentrations spiked at 1 ng/mL. All compounds were within the acceptable range except for 11Cl-
16
PF3OUdS in bread samples which were on the lower side at 26-42% recovery. Raw data may be
examined by contacting the study director.
Table 7. Single Lab validation recovery ranges. “Not validated” indicates that for this compound and
matrix, the compound fell below the acceptable recovery range.
Method detection limits were calculated by performing 7 low-level spikes at 0.25 ng/mL for produce,
fish, and bread and at 0.075 ng/mL for milk. The standard deviation of the replicates was multiplied by
3.14 (t-value for seven replicates where 1-α =0.99). The MDL is defined as the statistically calculated
minimum concentration that can be measured with 99% confidence that the reported value is greater
than 0. This procedure is published in the Code of Federal Regulations, see references.
Table 8. Method detection limits in ng/kg. *Note that for PFBA, no MDL could be calculated for cheese
due to an interference.
2019.12 REFERENCES
FDA Guidelines for the Validation of Chemical Methods for the FDA Foods
Program;
https://www.fda.gov/food/laboratory-methods-food/foods-program-
methods-validation-processes-and-guidelines.
( https://www.fda.gov/media/81810/download)
17
Definition and procedure for the determination of the method detection limit-revision 1.11. Code of
Federal Regulations. 40 CFR Appendix B to Part 136. Washington (DC).
https://www.govinfo.gov/app/details/CFR-2011-title40-vol23/CFR-2011-title40-vol23-part136-appB
18