Document N2qkgvnNJnmB9wKrrQmZYv95D

Jean B. Sweeney Staff Vice President - 3111LFaywronmental, Health and Safety Operations CERTIFIED MAIL April 9, 2009 900 Bush Avenue, ,l~uilding 4?=2E-26 PO Box 33331 St. Paul, MN 55133-3331 651 7785488 0? P 3?3 4n -OW - 3wq NO CBI Document Processing Center EPA East - Room 6428 Attn: Section 8(e) Office ofPollution Prevention and Toxics, U.S. EPA 1200 Pennsylvania Avenue NW Washington, DC 20460-0001 III I I I I I I II I il III If, II il Re: TSCA 8(e) Substantial Risk Notice: Supplemental to Docket No. 8EHQ-0598-373 ; Sulfonate-based and Carboxylate-based Fluorochemicals To whom it may concern: 3M is submitting this notice to supplement its previous submissions on sulfonate and carboxylate-based fluorochemicals . 3M recently received data from a fluorochemical (FC) analytical method validation study in which fish tissues, purchased from Osage Catfisheries, Inc . (Osage Beach, MO), were used as control samples . As detailed in the following table, endogenous levels of various FCs were detected in these tissues . All levels were in the ng/g range and are provided in the enclosed report . whole-body largemouth bass (Micropterus salmoides) perfluorooctane sulfonate (PFOS) perfluorobutanoic acid (PFBA) perfluorodecanoic acid (PFDA) perfluoroundecanoic acid (PFUnA) perfluorododecanoic acid (PFDoA) whole-body channel catfish (Ictalurus punctatus) PFOS PFBA perfluoropentanoic acid (PFPeA) perfluorononanoic acid (PFNA) PFDA PFUnA PFDoA whole-body bluegill sunfish (Lepomis macrochirus) and fillets of rainbow trout (Oncorh nchus m kiss) PFOS PFBA PFUnA While 3M does not believe that any of these data taken alone or cumulatively meet the "substantial risk" reporting threshold, we nevertheless recognize the ongoing work by U.S . EPA to assess fluorochemical exposure pathways . Therefore, we are placing these results in the 8(e) docket as a supplement to previous submissions . If you have any questions or would like any additional information, please contact Deanna Luebker at (651) 737-1374 or djluebker(ir),mmm .com. Sincerely, C~cG~~ Jean B . Sweeney Staff Vice President, 3M Environmental, Health and Safety Operations Enclosure ' CMAMNC CBI 3rY-5,9 3M ENVIRONMENTAL LABORA TORY REPORT NO . E08-02B1 Final Analytical Report Method Validation of ETS-845 "Determination of Fluorochemicals via Protein Precipitation of Fish Tissues (Fillet or Whole Body) and Analysis by High Performance Liquid Chromatography with Tandem Mass Spectrometry" Laboratory Request Number: EOB-0261 Testing Laboratory 3M EHS Operations 3M Environmental Laboratory 3M Center Building 260-5N-17 Maplewood, MN 55144 Requester William Reagen 3M EHS Operations 3M Environmental Laboratory 3M Center Building 260-5N-17 Maplewood, MN 55144 3M EMIIRONMENTAL LABORATORY PAGE 1 OF 25 3M ENVIRONMENTAL LABORATORY REPORT NO. E08-0281 3M Environmental Laboratory 3M Environmental Laboratory Manager: William K. Reagen, Ph,D . 3M Project Coordinator . Cliffton B. Jacoby, Ph.D. 3M Principal Analytical Investigator and Report Author : Michelle D. Malinsky, Ph . D. Analytical Report E08-0269 Method Validation of ETS 8A5 "Determination of Fluorochemicals via Protein Precipitation of Fish Tissues (FilletorWhole Body) and Analysis by High Performance Liquid Chromatography with Tandem Mass Spectrometry" - - - --o-a--. ;;.,.-, - - Report Date: ~ .~ .,.~.~.,..,. March 12, 2.,00,.9.~, ~ .~.._. .. . ., . :- This report summarizes the method validation of 3M Environmental Laboratory analytical method ETS 8-45 "Determination of Fluorochemicals via Protein Precipitation of Fish Tissues (Fillet or Whole Body) and Analysis by High Performance Liquid Chromatography with Tandem Mass Spectrometry" . The method validation protocol was detailed in the general project outline (Attachment 1) and was developed using the 2001 FDA Bioanalytical Method Validation Guidance for Industry as an analytical reference which requires the use of matrix-matched calibration .(1) The extraction procedure employs acetonitrile protein precipitation and incorporates a cryogenic incubation step . Extracts were analyzed using liquid chromatography tandem mass spectrometry (LCIMSIMS) for perfluoroalkane carboxylic acids (PFCAs) ranging from C4 to C12, perfluoroalkane sulfonates (C4, C6, and C8), and perfluorooctanesulfonamide . A comprehensive validation is presented for whole-body homogenates of largemouth bass (MicropFerus salrnoides) and method cross-validatior, results are shown for wholebody channel catfish (1ctalurus punctatus), whole-body bluegill sunfish (Lepomis macractrirus), and rainbow trout fillets (Oncorlrynchus mykiss). All data presented here were generated using fish from a supplier for scientific studies. No environmental samples were Used . For the full method validation, linearity, precision, and accuracy were determined in three separate extraction batches prepared over the course of two separate days which provided inter-day and intraday statistics . Each preparatory batch consisted of the following samples: thirteen point matrixmatched calibration curve ranging from 0.025 nglg to 25 nglg spiked tissue concentrations, four matrix blanks (two with internal standards (ISs) and surrogates, two without !Ss and surrogates), four method (aqueous) blanks (two with ISs and surrogates, two without ISs and surrogates), two aoetonitrile solvent blanks with 1Ss and surrogates, triplicate lab control matrix spikes at three levels, and triplicate lab control matrix spikes of perfluorooctanoate (PFOA) and perfluorooctane sulfonate (PFOS) from 3M eSectrochemicat fluorination (ECF) production lots . On the first extraction day, additional PFOS "dilution" QC samples were also prepared where the PFOS concentration (ppm range) was two to three orders of magnitude higher than the rest of the target ana3ytes (ppb range). Addi6onally, an altemate approach to PFOS quan6tation was explored using a species specific matrix-matched calibration curve prepared using (?,2,3,4-'3C41PFOS . The cross validation procedures for the three other fish species represented an abbreviated approach to the full validation performed for the wholebody largemouth bass. Finally, this study explored the use of a solvent (unextracted) calibration curve for quantitation of the fish extracts . 3M ENVIRONMENTAL LABORATORY "AGE 2 OF 25 3M ENVIRONMENTAL LABORATORY REPORT NO. E08-0261 ' ~'TY . . . ~ 0 A A 6fEIV ~1Fa~OR . : ~p ::Y--: ; :~Stf lw .:" .' . _ . i :~(tY 2.1 Target Analytes . Table 1 below provides pertinent information regarding the target analytes, internal standards, and surrogate compounds invesfigated in this method validation . Table 2 lists the 3M Environmental Laboratory identification numbers for the control tissues used. Table 9 . Target Analyte Summary. Compour+d IYarne Perfluorobutanoicacid Perfluoropentanoicacid Perfluorohexanoic add Perfluorotieptanoic add SA~c'ro~n~ymr PFBA (C4 Acid) PFPeA(C5Acid) PFHxA (C6 Acid) PFHpA (C7 Add) ~~ Purpose Target Target Target Target Perftuorooctanoic acid PFOA (C8Acid) Target Perfluorononanok; acid Pert9uorodecanoicaad Perfluoroundecanoicadd Perfluorododecanoie acid Perfluorobutane sulfonate Perfluorohexane sutfonate Perfluoioodane suffonate PFNA (C$ Acid) PFDA(C1aAcid) PFUnA(C11 Add) PFDaA (C12 Acid) PFBS (C4 suffonate PFHS (C6 Sulfonate PSFuO~S~~()C8 Target Target I Target Target Target Tawt Target Perfluorooctanesulfonamide FOSA (CS Sulfonamide) ~Target Formula C3,FTCOOfj C,F9COOFi CaF COOH CsF COOH C,FsCOOH iCrFisCO07NH,'1 CeFCOOYi CF,oCOOH C,oFt,COOH C FnCOOH tC,FIS031 ItC] ICsFSO3ZNaj fCBF,>SO,TK'J (CaF"SO'7(K'] CoF,7SOzNH, Reference Standard source AtdriCh AffaAesar Pa Aldrich L-PFOA (linear) Wellin gton Labs ECF PFOA (br~ed) 3M Production Lot 332 POk roducts Oakwood Products Oakwood OPakwood 3M Production Lot 2 L-PFHS ( Wellingbon Labs L-PFOS (linear)Wellin gton Labs ECF PFOS (branched) 3M Production Lot 217 ---- -- - - - - 3M (L-5709) 3M EWdIRONMEhJrAL LABORATORY PAGE 3OF 25 3M ENVIRONMENTAL LABORATORY REPORT NO . E08-026i Table 1 Contin ued, Compound Name "C,-Perfluorobutanacadd 'C4-Perfluorooctanoicaoid "CrPerflu o ro de canoicacid ' OTllmmonium Perfluarobutane sulfonate PerfluomheAxnesuffonate 'OrArnmonium Perflmrooctane sulfonate "CrSodium Perfluorooctane suttonate Synonym or Anal cal Acronym --- - pu~~ [1,2,3,4-'3C,jPFBA [1,2,3,d-"C,]PFOA [1,2-"Cs]PFpA ['O,]PFBS COz]PFHS ["eO~,FOS [t,2,3,4"C,;PFOS Surrogate Internal Starxlard Su rrogate Surrogate - Surrogate Intemal Standard Calibration ~ Formula "CF,("CF,)2'3COOH CF,(CFZ),("CFz)z"GOOH CF3( CF2),("CF,) "COOH [C4F,S'O,O'J[NH.`] Reference ~- Staridard Source Wetlingt_on l.abs I Wellington Labs {I Wellington Labs ~RTllnterrrationat [CaF, .,S"OzOjINHj Wellington Labs [CeFnS'0207[NH4`J RTllntemational [CF3(CFz)y("CF2)s"CFzSO3j [Na'] Wellington Labs Table 2. Control Tissue Summary Species Whole-body la emouth bass Whole-body catfish Whole-bod y bl iN Rainbow trout filet 31111 Envirorrmen0l Lab 10 Number TNO&0199-111 TNO&0199-111 TNOS-0203-1I1 TNOS-0196-111 ~-N!e#a~nia~ ~,. a1~a'n:d~R~ 3.1 Tissue Preparation Whole-body largemouth bass (Micropterus salmaides), whole-body channel catfish (fctalurus punefatus), whole-body bluegill sunfish (2_spomis macrochirus), and rainbow trout fillets (Oncorhynchus mykiss) were purchased from Osage Catfisheries, Inc. (Osage Beach, MO). Whole-body and fillet tissues were homogenized frozen in a two-step process. The initial homogenization step occurred at MPI Research (State College, PA). Frozen fish tissue was homogenized with dry ice until a coursemeat consistency was reached. After the initial homogenizaticn, the ground fish was transferred to a polyethylene bag which was placed in the freezer unsealed overnight to allow the residual carbon dioxide to sublime. After sublimation, the bag was sealed and homogenates were kept frozen and shipped to the 3N1 Environmental Laboratory for sample preparation and analysis . Approaching the time of extraction, the frozen course-ground fish tissue was added to a pre-chilled stainless steel bowl of a Robot Coupe RSI 2Y-1 vertical batch processor (Jotiet, IL) and re-homogenized with dry ice untii a powder-like consistency was reached. Again, the tissue homogenates were transferred to a pre-chilled polyethylene bag and the residual dry ice was allowed to sublime overnight. 3.2 Sample Extraction Polyethylene centrifuge tubes (5Q mL) were chilled in a cooler with dry ice for approximately thirty minutes prior to weighing homogenate aliquots to prevent the tissue powder from thawing and congealing during the weighing process. A 0.5 g sub-sample of the fish tissue homogenate was accurately weighed in a pre-chilled 5 mL polyethylene centrifuge tube. Tissue aliquots were then spiked with internal standard (IS) to produce an approximate concentration of 1 nglg. Samples designated as matrix-matched calibration standards were additionally spiked with the target and surrogate analytes at appropriate levels to produce the desired tissue concentrations . Samples 3M ENVIRONMENTAL LABORATORY PAGE 4 OF 25 3M ENVIRONMENTAL LABORATORY REPORT NO. EQ8-026i intended for laboratory QC (matrix blanks and laboratory control matrix spikes) were spiked with surrogates at 1 ngtg and the other target analytes at appropriate levels. After spiking, the tfssue homogenates were allowed to sit for 30 minutes before 5 mL of acetonitrie was added. The samples were then thoroughly homogenized with the acetonitrile using an Omni Prep mulG-place homogentzer with disposable plastic probes (15,000 rpm for 2 minutes) . The acetoridrileJtissue extracts were then placed in a freezer at -20C for at least one hour. Upon removatfrom the freezer, sample extracts were centrifuged at -5C for 20 minutes at 3000 rpm. If any delays occurred, samples were returned to the 20C freezer and then re-centrifuged priorto resumption of preparation and analysis. After centrifugation, 1 mL of clarified supernatant was transferred to a 2 mL autovial spiked with 10 pL of 10% formic acid . Autovials were then capped and vortex mixed . 3.3 Analysis Analysis for the suite of PFCs listed above was performed using high performance liquid chromatography-tandem mass spectrometry (HPLCIMSIMS). For this investigation, samples were analyzed using an Agiient 1100 series (Palo Alto, CA) HPLC system interfaced to either a PE SCIEX API 4000 triple quadrupale or SCIEX API 4000 Q-Trap mass spectrometer (Foster City, CA) . Both instruments were equipped with a SCIEX Turbo V ion-spray interface operating in the negative an MS/MS mode using multiple reaction monitoring (MRM). AnalystT" 1 .4.2 software was used for all data colleebon and reduction. Table 3 lists the MRM transitions monitored for each analyte. A Thermo Scientific PRISM RP guard column (2 .1rnm X 50 mm ; 51i particle size) was placed irt_line after the purge valve and before the sample injection port to trap any PFC contaminants coming from the HPLC instrument and/or the mobie phases . This sufficiently separated the elution of the "system" PFC peaks from those present in the sample extracts . Table 3. MRM Transition Summary . C d PFBA (C4 Acid) PFPeA (C5 Acid) PFHxA (C6 Acid) PFHpA (C7 Acid) f PFOA (C8 Acid) PFNA(C9Acid) PFDA (C10 Acid) PFUnA (C11 Acid) PFDoA (C12 Acid) ~ PFBS C4 SuMonate Descri tia+ Target Target Target Target Target Target Target Target Target Ta et roMRM Transition(s I 213>169 263>219 313>269 313>119 363>319 363>169 413>369 413>219 413>169 463>419 463>219 463>1B9 513>469 513>269 513>219 563>519 563>269 563>219 613>569 613>319 613>169 299>80 299>99 Drvefl Tme ( 100 100 100 100 50 50 50 so 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 3M ENVIRONMENTAL LABORATORY PAGE 5 OF 25 3M ENVIRONMENTAL LABORATORY REPORTNO . E08-026i Table 3 Continued. Corn d Anaf 2=2.00n rnMRM Transidon s Dwell rime ms PFHS (C6 Sulfortate) Target I 399>80 50 399>99 50 499>80 50 PFdS (CS Sultanate) FOSA CS Sulfonamide 499>99 50 Target 499>130 50 ~ Target ^. ~ 498>78 50 [1,2,3,4 ='C,]t'FBA Surrogate (Small PFCAs : C4-C6) ` 217>172 100 (1,2-"C,JPFbA Surrogate (Large PFCAs : C7-C12) 515>470 50 ~0, PFBS Surrog ate Sulfonates, FOSA 303>84 50 (1,2,3,4-"C .)PFOA Internal Standard (AN PFCAs) 417>372 50 'O2 PFOS Intemal Standard Sulfonates, FOSA 503>84 50 (1) Individual transitions vwere summed to produce a "total ion chromatogram" (T1C) . The TICs were used for quanttation . Analysis of the small acids (C4 through C6) was performed :Ising a Thermo Scientific PRISM RP column (2 .1 mm x 50 mm ; 5,u particle size) held at 300 with the rnobile phase system consisting of 5 mM ammonium acetate with 0.01 % acetic acid in water (A) and methanol (B). The gradient used to elute the anafytes of interest is presented in TaCEe 4. This flow rate was held at 300 [tL min-' throughout the run and a 20 ftL injection volume was used. The outlet of the ana;ytical column was directed to a column switching valve where the first three minutes of the run were diverted to waste. After three minutes, the valve was switched to direct the effluent to the mass spectrometer fnr analysis. The large acids (C7 through C12), the sulfonates, and FO SAwere analyzed using a Thermc Scientific BetasiIT'`' C,B analytical column (2 .1mm x 100 mm ; 5tt parUcle size) held at 30C with the mobile phase system consisting of 2 mM ammonium acetate in water (A) and acetonitrile (8). The gradient used is also provided in Table 4. The flow rate was held at 400 trL min' throughout the run and a 25 [IL injection volume was used . Samples were injected onto a'Vllaters (Milford, MA) Oasis HLB on-line extraction column (20 mm x 3.0 mm, 25 tt particle size) with the outlet directed to a column switching valve where the first five minutes were diverted to waste. After five minutes, the valve was switched and the effluent was directed to ft analytical column for separation of the target anafytes and analysis by the mass spectrometer. Table 4. LC Gradient Parameters . Sma:! A: cids C4-C6 -- La rge ik:itfs- C; 7-C12 Suf& nates, FOSA Step Total rime (min) %A 0 0.0 ~ 90 1 3 .0 90 2 3 .5 30 3 9 .0 5 .0 4 15 .0 5 .0 5 151 90 6 190 fin row rfia+e %B SW (Mtn) %A %a 10 0 0.0 10 1 3 .0 70 2 3 .5 95 3 15 .0 95 4 17_0 10 5 17 .1 10 6 20 .0 97 3.0 97 3.0 75 25 10 90 10 90 97 3.0 97 3.0 A 5 mM ammonium acetate with 0.01% acetic acid (aq) A 2 n,hll ammonium acetate (aq) B Methanol S :,cetonitrile 3M ENYIRONMENTAL LABORATORY PAGE 6 OF 2~1 3MENVJRONMENTAL L4$0RATORY REPORT NO. E08-02fi1 To manage the large number of transitions required for the large acids, sulfonates, and FOSA, two separate injections/analyses were acquired. Inclusion of all transitions in one chromatographic run made it challenging to get a sufficient number of scans across the chromatographic peak while maintaining a long enough dwell time to achieve the sensitivity needed for accurate and reproducible quantitation . The first injection acquired the MRM transitions forthe C8-C12 acids (17 transitions) and the second injection acquired the MIRM transitions needed for the anlaysis of C7 aad, the sulfonates, and FOSA (13 transitions) . Chromatographic conditions were identical for both injections. For analytes where more than one transition was acquired, a total ion chromatograxn (TIC) which summed the respective transitions for that anatyte was used analyze the data . A table summarizing the extraction and analysis dates is provided in the supplemental information . a.M... ; .~ IF ~., . . . .. . ,_..., ._ . . .,., ...._.. .. . ." .. . . Y.O Irk. 4.1 Calibration Species-specific matrix-matched calibration standards were prepared by spiking known amounts of target analytes, surrogates, and internal standards into individual 0.5 g aliquots of fish tissue homogenate . Each spiked fish aliquot was then extracted using the procedures outlined in Section 3:2, A total of thirteen spiked standards ranging from 0.025 nglg to 25 nglg (nominal) were prepared. A quadratic, 1lx weighted, calibration curve was used to fit the data for each anatyte. Internal standard quanfitation was used . The data was not forced-through zero during the fitting process. The accuracy of each curve point was verified by back-calculating the concentration using the area count ratio of the target analyte to the internal standard . Each extracted calibration standard used to generate the final calibration curve met the method calibration accuracy re uirement of 100125%, except the LOQ standard (10030lo) . The coefficients of determination () were greater than 0.990 for al analytes. As required by ETS 8-45, a minimum of six calibration points were used to generate the final calibration curve. 4.2 Method of Standard Addition Some target analytes, most notably PFBA and PFOS, had significant levels present in the control matrix. This resulted in the exclusion of several low-level points from the final calibration curve. When this occurred, the method of standard addition was used to determine the endogenous amount in the matrix, which was then used to correct the calibration and QC spiked concentrations . The two matrix blanks that were spiked with IS were included in the final calibration curve with concentration values assigned as the determined endogenous level. The calibration curve, with the adjusted concentrations, was then used to quantitate the QC samples. 4.3 Limit of Quantitation-(E.OQ) The LOQ as defined in ETS 5-45 is the lowest non-zero calibration standard in the curve in which the area counts of the target analyte are at least twice those of the matrix blank(s). The limit of quantihation for each anaiyte varied from extraction date and instrument batch. If the percent relative standard deviation (RSD) of the matrix blanks area counts was less than 30%, then the average area counts was used to determine the LOQ. If the RSD was greater than 30%, then the matrix blank with the largest area counts was used for the t-OQ determination- The resulting LOQs are provided in the Data Summary and Discussion section . 4.4 System Suitability Five replicate injections of the solvent calibration standard were analyzed at the beginning of the analytical sequence to demonstrate overall system suitability . Method criteria states that system suitability injections shall produce a RSD of less than 7% for the ratio of target analyte area counts to internal standard area counts and an RSD of less than 2% forthe retention time . In general, method 3M ENVIRONMENTAL LABORATORY PAGE 7 OF 25 3M ENVIRONMENTAL LAEiOF2ATORY RF_PORT NO E08-0281 acceptance criteria were met for both area counts and retention times . Method deviations have been issued for the instances of non-compliant system suitabiliaes and are documented in the Supplemental information . 4.5 Continuing Calibration During the course of the analytical sequence, several centinuing calibration verification samples (CCVs) were analyzed to confirm that the instrument response from initial calibration curve was still in control. In general, CCV injections produced recoveries within 100%{25%, which met method criteria . Non-compliant CCV recoveries were documented in method deviations in the raw data package and . are provided in the Supplemental Information . 4.6 Blanks Five types of blanks were prepared and analyzed with the samples~ rnatrix blanks (two with IS and surrogate, two without IS and surrogate), aqueous method blanks (two with IS and surrogate, two without IS and surrogate), acidified acetonitrile solvent blanks, acetcmitri}e blanks with internal standards and surrogates, and straight ace.tenitrile blanks (no acid., IS, or surrogate) . Each blank result was reviewed and used to evaluate method performance to determine the LOQ foreach anatyte. Surrogate recoveries of spiked blanks are provided in the Supplemental Information . 4.7 Lab Control Spikes (LCSs) Triplicate lab control spikes at three different levels were prepared each extraction day. For PFHS, PFOS, and PFOA, the standard reference materia! used for the LCS spikes was the linear isomer. Separate ECF spikes of branched PFOS and PFOA were prepared to evaluate any potential bias from quantitation against a linear standard. For Day 1 only, triplicate PFOS dilution QC samples were prepared where PFOS was spiked at ppm levels whereas the rest of the analytes were at low ppb levels. Table 5 provides the approximate spike levels for the prepared LCS samples. The lab control spikes were prepared to evaluate method accuracy and precision . LCS recoveries will be presented and discussed in the following section. Table 5. Validation QC Spike Levels . QC Sample Description t"PFBA Apppprox~mate S fke Concentration n ) All Other Target I21PFOS PFOA Anal es Surro ates LCS Low 4 4 0.3 0.3 1 LCS Mid 12 12 1 .5 1 .5 1 LCS High 20 20 8 8 1 E-CF LCS ~ NA 5 NA 1 ' PFOS Dilutian QC 12 400Q 1 .5 1 .5 4 (1) Initial screening of the largemouth bass control matrix indicated that PFBA and PFOS were at levels higher than the rest of the target analytes . Spike levels were adjusted accordingly based on the endogenous level so that the low level spike was approximately twice that of the endogenous level. The spike concentration listed reflects the amount of analyte spiked into the tissue and does not account for the endogenous level- (2) PFOS dilubon QC samples were only prepared on Day 1 . 3M ENVIRONMENTAL LABORATORY PAGE 8 OF 25 3W ENVIRONMENTAL LABORATORY REPORT NO. E08-026 f DISCYlSSlan: .-... .. ., .;. _. ..:. RAE 11 :E ]i ?"E6'~dVqFR.7s,13P ".f" ' k" 5.1 Whole body Largemouth Bass Validation Results 5.1 .1 Accuracy and Precision Table 6 below displays the accuracy (average percent recovery) and precision (%RSD) results for the three separate levels of LCSs (N=3) prepared in the whoie-body largemouth bass control matrix for each preparation batch of the method validation . Internal standard, species-specific matrix-matched calibration was used for quan6tation. Table 7 provides the average batch accuracy and precision when all three spiCe levels are considered collectively (N=9) as well as the LOCI, inter-day, and intra-day statistics. Additionally, the validation results are summarized graphically in Figure 1 . For sKnpNcity, the results of the ECF QC spikes are provided separately (N=3) for a given batch. In general, almost all of the analytes demonstrated excellent accuracy and precision with average percent recoveries within arbitrary method acceptance criteria of 100*34% and %RSDs less than 2010 when individual batch, inter-day, intra-day, and all data collectively were considered . With the exception of the small PFCAs(f'FBA, PFPeA,'and PFHxA), all analytes demonstrated average accuracies within 140t15% with qoRSDs less than 10%. Endogenous levels were detected in the control matrix for the following analytes: PFBA, PFWpA (Day 2 Analyst B only), PFDA, PFUnA PFDoA, and PFOS . For these analytes, the method of standard addition was performed to determine the endogenous concentration. The resulting concentration was then used to correct the spiked amount for the calibration standards and QC samples. 3M ENVIRONMENTAL LABORATORY PAGE 9 OF 25 4 ~ow. mo Q ~ O W y z W M 8 00 N V Q1 ~ N (O U~- CV CN ~- (D n n, o O N V N r U) r v') N '- O q fi N u0 yl 44 +1 44 H 41 ~ i1 +i H ,H i~ 44 'H -N 0 o m ~ ~s a a~ $ J C v s C 07 t~ N ~ N O ~ 1D ^J f0 ~ 4 t7 m ~D, .b ri ui o~ v ~i cv o ~ CN ~ cv M v Iri :-: ~n ~. fl il ~ip1 ~ipf H +1 .FI +1 -4 * il * 1 44 44 ii Q V y n~ O 6 O O ~ ~ O ODp ~ p ~ ~y ~ N p r ep U : Q~ Q~ Q ~1 C_7 3 r m d h m t OD r- o t~ a ac n ~"" r o t'? v N c7 N JW ,= In 0 ii ~}+ Y~7 N Q? vfii CV-01 4 +i !1 r+'I +1 r ii Iiti v i1 (+~1j ri d d ~ ~ N ~ N T fO O ~ C (O (h O ~ Y . O 01 O ~ ' oi o a h S T 8 m _ a R M N N W o o~ a> ~y r ~fl tn p~ ~ ) r ,C r N [M r r' (V r ~ r (N N Y) M N il t1 fi il ii "f1 +1 ii "p1 ai ii ii ii ,~ +{ t1 J`' t0 O ~ 01 tD 1., t~ ~ p KY Q~ T " {V N C C D ro ar U . Q tti0 f' 1` N ~ ~ p t~! W ~ O ~ 'V M q; 6? N ri N N c"! N ~i ~"i r N N 3 ~ +1 -t1 +1 * ii .}~ +l .H * * .H +1 * .H t1 af ~ V W .~ IO ~ e~ M p p Ir p tp N O1 .UC O O fA d a cca T U 3 s -H 44 4.1 +1 ^ 44 44 +1 +1 .4 +1 +1 44 ~ +., .3 1~, C7 u7 1, t- N LO P V, v, cV c0" ~ r~ Ih C7 W N ~ N Q u) v7 0 W C' O~ 4> ~ M ~A ~ h ,- b 4 ~ m 00 ~ aa `s r ~ ~r ~ S N N ~y O N O N C7 ~ ~ to 4 ~(7 4 p O U b n I~ [7 V h O ~ C G~ ~ ~:. Sf1 ~ y= ~,'1 N y^ ~ t") c] C7 'C) O ai i1 41 F1 41 f4 .H i1 t1 +1 f1 'H i1 i1 + n Q CD C> b C> Q . E N N 10 (O M N q a0 V '- r- a0 c~ W iP ~ n 1~ ! 19 b V 6 (V m =V Q N LfJ fw M) u;~ O"+ N R ~. +i ,i -H + +I ~1 1-I i1 +1 i1 il i; -H ~ -H - m C to tl W 1l ~ ~ O N O t0 N Q1 C9 CO b ^ 0 K7 P O o~ ~ O t C C7 P O Q T ~ ~ r c_ m ~ a +4+1 '+,d ^ r` * +, ao +1 .- 44 o`.1 u+, ~ ~ (,n; u-)H uN*) nIo? c +417 ~ J r r ~~ ~ ~ r g ~ ~ m ~ ~+' IwN v .t T ~u m a 4 7 ; 0 Q U N J2 _Q1 [ O Cip ( ? O LL Ta~ R"_ Lm5 re a 02 d .c c LaL" U- LL mLL s N e a, a. a ,aL ua, Ca .r,.i .o, . a ua. ,a. pu. 1~`!. TC`-! :21 p. 11 0LL 0 w Q a z w No . (L) If the notice includes a health and safety study. concerning the new chemical substance, the submitter must also answer the questionb in Sec. 720.90(b)(2). See below. 720.90 (b) (2) (i) Would disclosure of the chemical identity disclose processes used in the manufacture or processing of a chemical substance or mixture? Describe how this would occur. Yes. Disclosure of the chemical identity would enable a chemist to identify feedstock chemicals and the manufacturing process . Because of the type of chemical, its identity would also enable a chemist to identify processing and use information . (ii) Would disclosure of the chemical identity disclose the portion of a mixture comprised by any of the substances in the mixture? Describe how this would occur . The notified substance is not a mixture . (iii) Do you assert that disclosure of the chemical identity is not necessary to interpret any of the health and safety studies you have submitted? If so, explain how a less specific identity would be sufficient to interpret the studies . The health and safety studies stand by themselves . The generic name and the toxicology studies should enable any toxicologist to comment on the safety of the chemical . c ca a~ y 0 G Ch ca CL 4 c 0 a a r- r- N cmsa t~3 . x N O C LD G ? (D aN W c c 3 Accuracy {ldRecovery} t Precision (~IaRSD) a ro.~ tn~ ccny oo~ a0 Ncs ot~- ac+a PFBA PFf'eA PFNxA F"FHpA PFCA PFNA PFDA PFl1nA PFnaA PFBS PFNS PFC7S FOSA [1,2,3,4-13C4]-PFBA [1,2,-13C2]-PFDA [18t721-PFBS a 0 c! o0 o CU ~ ~ `C "G K jiy ~ ~ t~y 1V -" :~U;, W d) d: CO > T p. 14 ,,,yyy CA T O' 4~ D a il y O dS d C N 41 n ii -n 1_u U! ~ '" T+ S' ~ ~ O Y ~ D +' " =~s Ip o OD ~ N Q `~ " ~y aa o o V tT ~ N -W" ~ ,P ~ N N W Iv ~ W ~ ~n }} M 1F VF O f+ Fb H W FF IF M VF K ~ N' 11 f(, ~ [7 A ~A 47 V ~ fT V co O) OD (7~ cu W v y, y~ Si d CD - d A h1 00 (T ~ CT A O ~ ~ A ao 7 ' r d ~ N r ~ o p 0 0 0 o m o o m tn tr A 'ic~' o i6 0 0 p o rrt ~+ D ? (pI~I N (~IWI ~~ tp Jfl tW0 Q O O (W7~ Cp Wfl ~ t~J N A N . " 07 d ] i_l 3 -W V -+ A A U7 ~ ~ (p ~ A pNp ~ N A 'O v v' ~ ;ar ~~8 1+ N S_' m N Oo V ip ~ O W S v O p ~ (f~ Ui A If FF W FF F+ IF e+ FF ~ ~+ ~ FF !F 4 EF (T N A W A fD W ~ v A V ~ A Ol W W ~D A A f0 ~. M O ~~ N (D N W A ~! A 1 14T ~, fi ~,~` 47 tD V <D 60 a w r tn w N ;a c y C a7 7 N C Q 11 0 0 0 -p ~ ~ p .op ,o _a v, ~ c Gi '~~n~ u p v 'N r~ Oo l7~ O~ S N Q W N ~ tD p SWWO. 3~ ~I w N A W C; 7 f~ r A w n rT O (R~ 0 J ~ n ~ O.`G`"~tpO ,~ 3 'm~ ;, 1 C N CA a1 (] ?1 O 00 p O ~ ~ w O r+ ~ P ~ .I th rr K rr W 't, O IT 1+ rr 1+ ,~ O V 1+ p Q> ~ ~ A If Z !p ~.Z p ~~0 n v, n~ m a ~ a m ca a W w ~ ; t+. ~ W ~ ~St m d U7 A O y 1J~ m d~ N `J ~`I N tQ ~+ O rl V P-~ (C_ ~ ~. ~C 50- a D 01 c a. Km $a N ~D mH& ~ ~ d w $~e d~ ~ O O tb .,Wj W lfl 01 3'm9 r>v. A cu 0 co !p !Q ~ N ' ~ +Y A i~7~ G1 + ~ ~ m S N w N W ~ .Na N l0 ~, ~ Ff Ih 1~ Iq . FF 1~ M W 1+ H' Ff fF FF 14- 11" C C1 W f71 W V A, tS1 00 O7 Q7 (p A ~p ~ y A ~ N u. 0) rl A 41 (Q A) (7i LT Ih) ~.j CO K3 (D 'K c m r tD ~ f0 1 p r W ._ s f0 ~ .-.. ( ~ 5U0f O Q O p O ~. W U N +' ~ ~ 1+ Ft ~+ tt ~ IF W FF 1* HF H N. It FF FF -1 tD fj~ tD -" ~ Af O V OO Q~ fQ tD N 6~ m ni :a1 T 2k!! A ~1 fT A ta ~1 J -~ 1 ~p fT 1f' Ih 1+ 4F 1~ M H 1f H Ft Fh M IF 11IA f A lA Oo fD A fT ra CT t~ rl A m ?~ ~ -+ c7T (n 0 f3~ Oi Oo O lh ~ Syp;: ~ tS ~o ~" ,C z G b O "1 om Tb Op -~ < >f , ;V? -~ . 3M EIWIRONMENTAL LABORATORY REPORT NO E08-0261 5.1 .2 ECF PFOA and PFOS QC Spikes Table 8 provides the accuracy (percent recovery) and precision (%RSD) results of the 3M ECF QC spikes of PFOS and PFOA quantitated against the linear reference standard . For PFOS, the mean recovery of all three batches collectively was 95 .9lo with a RSD of 3.4% demonstrating that no measurable bias could be attributed to the use of the hnear reference material, Altematively, the average PFOA recovery of the ECF spike was consistently approximately 87% for all three batches indicating that a potential bias may exist when branched PFOA isomers are quantitated using a linear reference material. Potential explanations of the observed decrease in recovery may be differences in the response factor from the linear to the branched isomers or co-extracted matrix components may suppress the signal of the branched isomers relative to the linear . Even though the PFOA demonstrated some decrease in recovery, the results were still within 10015%. Table 8. Results Summary : ECF (branched} PFOA and PFOS QC Spikes Quantitated Against a Linear Reference Standard . A t! Percent Rem 2 Predslo,n aRSD Anal PFOS PFOA 5.1 .3 Da y 1 t A j Da 2 Anal A Day 2 E 97-6 2.0 87 .4t4.0 95.7 2 .6 L- 87 .031 .1 T 94 .5 3 5.3 86 .634.1 PFOS Dilution QC Spikes Mter 96.6 2 .3 87.232 .7 kiba.Da 95 .1 t 3.8 86.Bt2.7 A1f Batches 95-93-4 7.0 13.0 On Day 1, additional OC spikes identified as "PFOS dilution QC" were also prepared . These OC samples were prepared with PFOS at approximate levels of 4000 ngtg with all other analytes at 1 .5 nglg, with the exception of PFBA at 12 nglg . The undiluted extract was analyzed for all anaiytes except PFOS . A 1:1000 acetonitrile dilution of the final extract was prepared for the PFOS analysis . The PFOS analysis of the diluted extract was perfonned using external standard calibration as the interinal standard spiked into the tissue prior to extraction was diluted below the limit of detection. Spiking these QC samples with IS at a level that could be diluted into a usable range was not feasible as the isotopicafly labeted standards come from the vendor already in solution at a concentration of 50 pglmL. Table 9 lists the results of the PFOS dilution QC . With the exception of PFBA and PFPeA, the average recoveries were within 100+10% with RSDs less than 15%. The internal standard for PFBS, PFHS, FOSA and ['802]f'FBS surrogate was switched from ['80?.]PFOS to [1 .2,3,4-'3C,JPFOA for these samples only . When ['802]PFOS was used as the internal standard, recoveries ranged from approximately 65-75Io because the internal standard area counts increased by about 32% in these samples. The IS signal increase was attributed to the co-elubon of the ['eOz}PFOS sViked at 1 nglg with the unlabeled PFOS spiked at 4000 nglg . The combination of "S and a single' O in the unlabeled PFOS would also produce the same 503 m/z parent ion as the ['802]PFOS . Based on the natural abundance of 3"S (4_4%) and "0 (0 .2%), the probability of both 3t and a single "30 present in the unlabeled PFOS spike is 0.0088% .(2) V1hth the PFOS spiked at 4000 nglg, this contribution would be equivalent to approximately 0.35 nglg or 35% of the spiked IS and explains the observed increase in IS area counts . When the IS was switched to [1,2,3,4'3C4]PFOA, which does not co-elute with PFOS, the recoveries improved dramatically demonstrating that the extraction procedures were not responsible for the low recoveries when the ['a02]PFOS IS was used. The average recovery of the post-extraction dilutions for PFOS was lower than the other analytes at 84 .49'0. Possible explanations include dilution of matrix components atong with the PFOS that alter the signal enhancement against the matrix-matched curve or simple variability in dilution technique. 3M ENVIRONMENTAL LABORATORY PAGE 13 OF 25 3M ENVIRONMENTAL LABORATORY REPORT NO. E08-026J Table 9. PFOS Dilution QC Results . Analyte PFBA PFPeA PFHxA PFHpA PFOA PFNA PFDA PFUnA PFDOA OPFSS mPFHS c3}PFOS OFOSA '3C.-PF6A 13C7.PFDA M'80TPFBS tAccutacy Precision 129 t 7.4 144 t 9.6 115 t 12 93.8f 6.8 93.2 7.9 106 t 13 105 5.3 96 .4 f 6.7 l06 t 10 99 .1 1 .9 91 .7 t 7.6 84.4 t 12 96.4 15.0 120 6,2 11712 98 .0 t 6 .1 (1) Accuracy= average percent recovery . Precision = %RSD. (N=3) (2) 11,2,3, 4"C,JPFOA was used as the internal standard . (3) PFOS e*ncentration determined by external standani c2libratian of 1 :1000 dilution of the final extract . 5.1 .4 Altemate PFOS Qtrantitatiorf In a separate preparation batch, a whole-body largernouth bass matrix-r?matched calibration curve was prepared using [1,2,3,4-"C4]PFOS as the calibrant, ['1"O.JPFBS as the internal standard, and ['eOZ]PFHS as the surrogate. Thirteen calibration standards ranging from 0.025 nglg to 25 nglg spiked tissue concentrations were generated along with four rrlatrtx blanks (Wvo with IS and surrogate and two without IS and surrogate) . Triplicate lab control mat'ix spikes were al-so prepared by spiking unlabeled linear PFOS at approximate levels of 4 nglg (low), 12 ng :'g (mild), and 20 ngfg (high) along with triplicate ECF PFOS spikes at 5 ngtg . Consistent with previous va4dation QC samples, the surrogate was spiked at an approximate level of 1 ngJg for all sampie3. The ttlree ;1,2,3,4-'~C4]PFOS transitions (503>131, 503>99, 503>80) as well as the corresponding unlabeied PFOS transitions (499>130, 49S>`39, and 499,-80) were monitored. The calibration curve was s;Ons.rtlc~ed using the total ion chromatogram (summation) of the three [1,2,3,4-'3C4]PFOS tlansitiol?s . The three unlabeled PFOS transitions were also summed and the total area counts were plugged into the resulting regression equation for (1,2,3,4-"C4]PFOS . The average endogenous PFOS concentration in the control matrix was calculated far the thirteen matrix-matched calibration standards and the two matrix blanks spiked with 15 . The mean endogenous PFOS concentration in the whole-body largemouth bass control matrix was 2.76 nglg t 10!o(RSD) (N=15) . This concentration was then used to correct the spiked values of the QC spikes. Table 10 provides the accuracy and precision data of the QC spikes . All QC resutts met method acceptance criteria for accuracy (100*30/4) and precision (RS0<24%). 3M ENVIRONMENTAL LABORATORY PAGE,,' nF 25 3M ENVIRONMENTAL LABORATORY REPORT NO. E08-0261 Table 10. QC Results for Alternate PFOS Quantitation using [1,2,3,4-"C4]PFOS as a Calibrant. ---- ----- Accurac A % Reco i Precision SD Low !1W HigA ECF Batch Average w-_0 r~r PFOS (Linear) 107 t 5 .8 1143 .8 1115.1 0 'OZPFHSsurr 111 t1 .8 117t2.5 112t4.4 (1) OZPFHS surrogate was spiked at approxirnately 1 rglg for all QC levels . 105 3 2 .4 112t4.8 109t 5.0 113t3.8 5.2 MuItiSpecies Cross Validation The method presented here was subsequently cross-validated for three additional fish species: wholebody channel catfish (lctaJurus purctatus), whole-body bluegill sunfish (r_epomis macrochirus), and rainbow trout fillets (Oncortiynchus mykiss) using species-specific matrix-matched calibration curves . The cross-validation proeedures represented an abbreviated approach to the full validation performed for the whole-body largemouth bass. The cross-validation for the three additional species included species-specific matrix-matched calibration with triplicate lab control matrix spikes at three levels to assess accuracy and precision. Additional ECF spikes of. PI=OS and PFOA were prepared and analyzed against predominantly linear reference materials. PFOS dilution QC and inter- and intr-a-day evaluations were not included . The results from the multi-species cross-validation are summarized below (whole-body catfish: Table 11, whole-body bluegill : Table 12, and rainbow trout fillet: Table '[3) . The average percent recovery and %RSD for all spike levels combined is presented giraphically in Figure 2. For all three species, the average percent recovery (all levels combined) was within 100+-300/c with a %RSD less than 20% for all analytes, with the exception of PFBA for rainbow trout fillet Rainbow trout filet PFBA results were not reported because several of the calibration curve pants did not meet method acceptance criteria once the concentrations were adjusted for endogenous levels . Rainbow trout cross validatlon samples were re-extracted and analyzed for the smat PFCAs to see if a better calibration curve could be achieved for PFBA. For the re-prepared trout samples, the area counts of the [1,2,3,4-'3C4]PFOA intemal standard in the t_CS samples dropped by approximately 30% when czmpared to the IS area counts of the curve resulting in high recoveries . No results .were reported from the reanalysis; however, external standard q3uantitation of the samples produced acceptable recoveries in general. The variability of the [1,2,3,4 ' C4lPFOA IS during the small acid analysis was observed throughout the cross validation analyses and was mom pronounced as the PRISM analytical column aged . This variability was not observed during the analysis of the large PF.CAs using the Betasil C1$ column, therefore, the signal degradation and variability over the course of an instrument batch is largely attributed to the IS's lack of robustness on the analytical column and not to extraction issues . Analysis of the smaller PFCAs may benefit by selecting a different isotopically labeled internal standard with a smaller chain length, more representative of the target analytes . As observed with whole-body largemouth bass, the ECF spikes of PFOS did not exhibit any measurable bias when quantitated against a linear reference standard for the three additional species (average recoveries were greater than 95%) . The ECF PFOA LCSs exhibited lower recoveries for the whole body catfish and bluegill species, 90 .9% 0.82% and 83 .0% 3.0%, respectively; however, the rainbow trout fillets produced an average recovery of 96.4% 3.0%. This may suggest that a coextracted matrix component present in whole-body tissues, but not in fillets, is affecting the PFOA isomer response. Because quantitation of the ECF materials against a linear reference standard produced variable recoveries in the different species and tissue types, it is recommended that this QC component be evaluated at a minimum each time this method is applied to a new species and ideally with every preparation batch. 3M ENVIRONMENTAL LABORATORY PAGE 15 OF 25 3M ENVIRONMENTAL LASGRA TORY REPORT NO. E080261 Table 11 . Method Cross-Validation Results for Whole-Body Catfish. ACcurda Avera %Rec t PrecJston %RSA Mal e Low Mid N! All Levels CombJrred PFBA I PFPeA PFHxA PFHpA PFOA PFNA PFDA PFU(u4 PFDaA PFBS PFHS PFOS FOSA ECF PFOA ECF PFOS 1211 .0 102 t 3.3 113t3.7 1092.1 t'11 26 t 15 101t1,7 107 3 .1 ")124 t 14 1164 .4 94.310 94 .4 t 7 .7 97 .6 t 3.8 118t2.9 NA NA 1202.1 108 :t 3 .0 117t2.8 1.14 t 1 .6 111 84 11434.2 1093.4 1206.3 1126 .9 99 .9 # 1 .4 106 13-3 1032-9 1173.7 NA NA 123 t 4.6 115 t 3 .5 111 t5 .0 110 3 .2 102 * 1 .2 113t3.2 96.0 11 .4 4,06 t 3 8 104 t 4 .2 96.112 .2 95 .1 t 3.2 98 .7 t 3.1 105t3.1 NA NA 121 2.9 1095 .9 113t4.2 111 Z9 113 t 13 109t6.3 1()4 :t 6.3 117 t 11 111 6 .5 96,5 5.7 98 .6 7 .3 99 .7 3.7 113t6.4 90.9 f 0.82 99 .1 t 2 .1 [1,2,3,4"'C41-PFBA [1,2,-"C,]-PFDA 116 11 .1 109t 2 .6 102 t 2.9 107 8.8 114 12 .7 111 5 .7 111 62 109 15 .6 ' 'e FBS 101 t 6.2 (1) One ZCS exceeded 130% recovery . 98,7 3.4 101113 100 t 3.8 E= endogenous . The endogenous ooncenntration listed +s the lOQ forthe controlmatrix s. tudied NA = Not appficabe. LOQ +'rr ) E=2 :29 E=0 376 0 103 0 .0486 0 .265 E=0.332 t==0.0784 E=0_ 142 E=0.0376 0.103 0 .0468 E=0.166 0.268 'VA NA 0 .0244 0 .0245 0.0229 3M ENVIRONME,'fTAL LABORATORY PAGE 16 OF 25 3M ENVIRONMENTAL LABORATORY REPORTNO. E08-0261 Table 12. Method Cross-Validation Results for Whole-Body Bluegill . Acewae A %Rec t AYecTsTon RSD AnaWe LOW Aqd All Levofa C"blited LOO PFBA PFPeA PFHxA PFHQA PFOA PFNA PFDA PFUnA PFDoA PFBS PFHS PFOS FOSA ECF PFOA ECF PFOS [1,2,3,4'"C4}-PFBA [1,2,-"Cj-PFDA 121 t 11 (1)NA (')NA 104 :k 5.3 93 .5 8.9 112 9.4 93 .5 t 8. 0 99.1 2 .6 103 12 112t3.8 93.8 19 .9 99,7 .t 0.93 122t8.7 NA NA 110 11 92 .2 t 4 .0 1113.7 109 17.0 133 t 3.2 100 t 2.2 99.8 7.4 111 3.0 97.2 t 8.9 107 :k 13 1163.8 105 7.0 101 t 2.1 98 .30-85 114t3.9 NA NA 99,2 5 .4 100 t 5 .8 1057.5 100 4.5 109 t 3.9 95.8 * 2.5 '84 .0 10,46 1196.8 f 2.1 1180,60 .60 N7.1 t 3.9 098.18.9 99 .0 :t 2 .3 94 .5 1 .6 96 .2 t 3.5 106t8 .8 NA NA 93 .37 .8 093 .6i 5.2 112 t 9.4 104 t 7 .1 121 t 11 1004 .6 93.5 9.4 108 8.5 91 .7 t 10 99.1311 10710 105 t 6 .8 96.5 6.2 98.0 24 114t8 .9 83 .0 t 3.0 98.1 t 0.46 101 10 95 .7 t 5.9 E=1 .45 0.359 0 .505 0 .251 0-255 0 .259 0 .248 E=0.382 0.251 0.258 0 .0977 E=1 .99 0.105 NA NA 0.0247 0.051 "O, PFBS 98 .8 3 .0 98 .0 3 .4 99.0 3 .2 98 .6 2.8 0.0231 (1) Spike concentration less than the resultant LOO, recoveries not r. epor( ted (2) Bad instrument injection for one of the replicates, data not generated. Precision evaluated as percent relative difkensnce (N=2). NA= Not apppcable. 3M ENVIRONMENTAL LABORATORY PAGE 17 OF 25 3M ENWRONMENTAL LABORATORY REPORT N0 . E08-026f Table 93 . Method Cross-Validation Results for Rainbow Trout Fillets. Accuracy (Average 96Recovery) Precision SD Ar,aJ Row Mid Hi AfJLevefs Combined LOO n PFBA NR , NR NR NR NR PFPeA PFHxA PFHpA PFOA PFNA PFDA PFUnA PFDoA P FBS PFHS PFQS FOSA ECF PFOA ECF PFOS (1)131 t 19 124 t 7.8 109 0.27 105 t 4 .0 110z2.2 102 11 .6 108 t 7 .5 108 16 .9 102 ;t 6.7 99 .0 t 3.3 103 t 1 .7 124 11 .5 NA NA "'127t67 122 t 1 .8 110 4,5 108 5 .6 117t2.1 102 * 2.8 112136 112 1 .8 106 i4.9 102 t 3 .5 102 t3.1i 121 t 5 .7 NA NA n)131 t 14 115 f 12 104 .5 .9 103 16 .5 114t4.9 100 31 911 :+ 1 .1 109 4 .5 93 .2 1 G 93.8 111 .4 104 i 3.8 '.24 ::-9,9 NA tit. 129 13 120 t 7.8 108 t 4 .4 106 t 5_2 '114 3 .9 101 2 .5 110 4 .5 110 4 .5 101 t 10 98 .1 t 5 .9 10312.8 123 t 5 .8 96 .4 t 3 0 96 .5 2_5 0.252 0.252 0.0243 0.0501 E=0.110 0.0487 0 .0495 0 .0493 0 .0249 0 .0234 E=0.287 0 .104 NA NA [1,2,3,4et',}PFBA 113 12 114 t 2 .2 107 : 2 1 111 7 .0 0 .0246 [1,2, "C,J-PFDA 105 t 3 .2 107 2 .6 113 41 108 4.5 0 .0247 I, ['e07]-PFBS 103 t 2 .7 1033.2 96 .8t 13 101 7.1 0 .0231 (1) One or more of the LCS replicates produced a recovery 9teater than 130 NR= Not reported. The calibration curve once adjusted for the endogenous concentration did not meet method acceptance criteria forseveral pants. NA = Not applicable. 3M FRMRONMENTAL LABORATORY PAGE 18 OF 25 3M ENVIRONMENTAL LABORATORY REPORT NO. E08-026i 5.3 Solvent Curve Analysis . For each of the four species studied here, triplicate matrix blanks along with triplicate laborat-vj ma-0 ix spikes at three levels were prepared in a separate preparation batch and analyzed against an acetonitrile solvent (un-extracted) calibration curve that was acid adjusted in a similar fashion to the sample extracts . The results from the solvent (un-extracted) calibration are presented in Table 14 and graphically in Figure 3 for all the target analytes and surrogates except forthe small PFCAs (PFE3A, PFPeA, PFHxA, and [1,2,3,4 -3C,)PFBA. Results for the small acids were not reported as the [1,2,3,4 -'3Ca1PF0A IS signal rosponse was significantly suppressed when compared to IS signal in the soivent curve. Furthermore, the small acid target analyte signals were also suppressed when compared to the solvent response, but at different percentages than the IS . The combined target analyte and IS signal suppression observed in the extracted samples for the small acids resulted in recoveries that greatly exceeded method acceptance criteria of 10030% for PFBA, PFPeA, and [1,2,3,4 -'3C4JPFE3A. Acceptable results were observed for PFHxA as it was the largest acid analyzed and structurally most resembles the C8 internal standard. For simplicity, none of the PFHxA results are reported. Samples were prepared and analyzed twice for the small acids using newer PRISM columns to see if improved results could be achieved . Varying degrees of IS signal suppression relative to the solvent curve were observed during both analyses . The [1,2,3,4 -'3C,IPFOA IS signal response did not demonstrate near the level of suppression in the large acid and sulfonate analyses for most species which indicates that the low IS response was not an extraction efficiency issue . These results again suggest that the PRISM analytical column may not be suitable for analysis of the larger PFCAs and selection of a different internal standard for the smaller acids may improve the analysis. Although the ISs for the large PFCAs and sulfonates exhibited some suppression in some species, the target analyte signal was suppressed by approximately the same amount. This produced a "self-correcting' result and most recoveries for the target analytes were within 10030% (Table 14). However, it should be noted that the average overall recovery of FOSA in the whole-body bfuegill samples was less than 40% when quantitated against the solvent curve. This result emphasizes the critlcality of inclusion of appropriate QC samples to verify the method applicability and quantitation approach for each species for each analyte. 3M EIV4IRONMENTAL LABORATORY PACE 20 OF 25 71 6c' Accuracy Precision 0 G C1 K? O O U O Q O 47 tD .r 0 c PFHpA m PFOA su PFNA wro c PFDA c PFUnA 3 PFDaA PFBS , PFH S PFOS FOSA ,r9,2~-13C2]-PFI3A (18021-PFEiS m i -, n ro~ m z c-r_ a ~ o no y h "- p. 23 ^N ~~, v~ ~s vW ~op vc v-n 'M0 ~-Q 'rpo ~gn2 r(Do -n a w A ~W D ~ ~ D D n ~t ~~03D~ n A ~30 La ~o_ 0 ag o ~ ~ 0 cs '~ ~ ~ m n - C 0 ~N o O _ ~Np q N f N z M * 4 ~+ 1+ ~ ~S tt ~ ~a Z'.O A A A iJ ~ li+ v tD A ~ (n ~ Of c7~ ~ .-~ a N O ~ a o Cn g w ~ C R' ~m -'Wa~ 0* 2 0) 8 r c~ N ~ m ~c ~ ` ~ o0 0~ oo ao ` 13 A l' 1 'pC 3 fAX GNNI 40 . c G a 3 N !F ~ Q Co Q`~, C W p p ~ <n d ~ N ' O Z 00 CA (A ~4 z ~ 1} FF H M ~ }~ FF VF ~ ~ ~J ~] ,ZI N A7 ~ M ~ W N p A PQ G.1 a a2L ~g a~ ,g 213 a 9; ~ ~ m A ~"--- Rig O Q L m d O . 'SR N t~ FF ~ O) Ql K ~ V fJ Fr * j. ~ z ~ ".-. A W flf ~ - 1+ 1F H W W FF ~ tT N OD CA O~ A iJ N :rr ~ A ~1 t0 i~ y, ,y Z~'1' b 3 W N Ll N N tD to w 10 w (D fO tD P 10 2 o 087 ~ 2. jp cOo O ~ OAP~ -J -A-i ca W "L Z 2 CJ~ *~1 ~ ~N Of ~ A y tt ~ F+ 1 ~ ~I .Z1 rZ1 Ou V IJ A W W t' A N Q1 z. W <D t3~ ~ C tSi g 1~T, z W z oy?A ~O op p. 24 FQS ;3,2,13C2]~PFCbR [18021-PFE! S 0 0 " 11 , -aa rs~ tn r> u~ o~ :u tf! cG c~ w r n-< t n 5. r~7 m wrr O a7 0 : C`h r.i --" -< 3M ENVIRONMENTAL LABORATORY REPORtNO. E08-0281 II I .1~ ,~ 1'~~t~~ r !! .Rw-s~sw~I~~! .r ". . ~.:~:~. J*' C~ ~~~r. ~ !! ~I It : ._s, +:~n+.,'~4"4~., i_:10 W -~ * ~1 1 The method presented here produces excellent accuracy (average percent recovery) and precision (%RSD) when species-specific, matrix-matched calibration is used for the C7 through C12 perfluorocarboxylic acids, PFBS, PFHS, PFOS, and FOSA . In general, the method used here produces acceptable results (accuracy 10030%: precision <20% RSD) for the small PFCAs (C4-C6); however, it is hypothesized that improved results can be achieved if a different internal standard is selected for the small PFCA analysis in future studies. Analysis against a solvent curve produced more varied results than the species-speffic matrixmatched calibration . The solvent curve analysis may be an acceptable approach for screening level analyses or when a suitable control matrix is not available. If the solvent curve calibration is used for generating quantitative values, enough QC elements (lab control and lab matrix spikes) should be included to verify that matrix effects are not artificially biasing the sample results . When quantitated against a linear reference standard,.QC spikes of 3M ECF PFOS and PFOA produced recoveries within method acceptance criteria of 100t30%, although ft recoveries for the ECF PFOA exhibited a small, but measurable, bias in the whole body largemouth bass, catfish, and bluegill tissues (average recoveries ranging from 83.0-90.9%). No bias was observed for the ECF PFOS (all species) and PFOA in rainbow trout fillets with average recoveries greater than 95%. J 8 .0 .ii-0 ;~ .- r. .i ./_~ ~y :.lZa 1: All remaining sample and associated project data (hardcopy and electronic) will be archived according to 3M Environmental Laboratory standard operating procedures . 3M ENVIRONMENTAL LABORATORY PAGE 23 OF 25 p. 26 3M ENVIRONMENTAL LABORATORY REPORTNO . c0$-0257 Mtelle D. Malinsky, Ph .D f ~J31'-~Jael pecialist, Principal Analytical Investigaker Date 102 ~~.~r/~.~C=d'f' William K, Reagen, Ph.D., Environmental Laboratory Management Date Ciiffton ll.liakby, Ph .D ., /3M Technical Reviewer, Project Coordinator Date 'The 3M Environmental Laboratory's Quality Assurance Unit has audited the data and report for this project. ~-13-a~ Date 3P4 tWRCIWENTAL LABORATORY PACE; 24 OF 2t~ p. 27 3M ENVIRONMENTAL LABORATORY REPORT NO. E08-026'r 9,1 3M Environmental Laboratory Michelle D. Malinsky, Ph.D Research Specialist 9 .2 Pace Lab Ops Jonathan Stieege _ ~-,-:;-- ,j' .! l\I'~'.~3 ?>:'L'.` : t~ . . . 1~iFA0-:ri: 7= ~_- .. "s. .w. .' . :_ .~ . '7: " !1 . !.~, ' ". . R 4F . .4 ~: ~+j .y a' ..i . : ' :i .s: 4~. .. .a,.~ ...".: . . ~ .r. r . .+" . .. . ~ .,. .~, ~ . ... . . ~ . . . .: .-~ : ._ ~~~~~_. . .. .. .. .. . w " -i . .. . ._.,< . [1] Guidance for Industry : Bioanalytical Method Validation, U.S . Department of Health and Human Services, Food and Drug Administration, Center far Drug Evaluation and Research (CDER), Center for Veterinary Medicine (CVM), May 2001 . [2] F. W. McLafferty, Interpretation of Mass Spectra, Third Edition 1980, University Science Books. ['~achments X1K 4 _....... .. .x_. ._^.. .. . . :. : _ . .. ... .: .-.p" _ ~ T"7-,1-."-. ._ _ ~ . . ._, . .~. . _ ._#..a.:o ."~ ..c.r "T+r~ ...a..,,.. ...~ ~ _ ,p---~. .-.... < .,,. .. - .~_. . . . . ::. ~ ". .. : ._ . . . .. ... _. . . _ . .. . . . . . . _ '~ .. ._.,._ ..._~ _~. t..__.. , . . . ._ . .._.._ ... .>_._ . ...._ .. .... ..~_._ ..~ 1 .1 General Project Outline 3M ENVIRONMENTAL LABORATORY PAGE 25 OF 25 Vm Environmental Health & Safety Operations, Environmental Laboratory Amended General Project Outline To: Bill Reagen, 3M EHS&Opns ; Environmental Lab From: cc: Michelle Malinsky, 3M EHS&Opns; Environmental !.ao Cliff Jacoby, 3M EHS&Opns : Environmental Lab Date : Subject: January 14, 2009 Method Validation - Determination of Ftuorochernicals via Protein Precipitation of Fish 1-issues (Fillet and Whole Body) and analysis by High Performance Liquid Chromatography with Tandem Mass Spectrometry 1 General Project Information William Reagen 3M EHS C?pns - Environmental Laboratory Project Requester 2B0-5N-1? St. Paul, MN 55144-1000 651-733-9739 -----~~--- --- wkrea enp_m._m...~m.Y..c_o_r_n. -._ .---------------------- ._...__.W._ ---- Cliff B. ,lacoby 3M EHS&Opns, Environmental ilabcratory Project Coordinator 260-5N-17 St . Paul, MN 5r,144-1000 ` 651-733-2533 ' --^- - ~ cb_iac_G_bV_2Qmrrrm._corn_-4 _-a~T----- - i Michelle D, Malinsky I 3M EHS&Opns, Environmental Laboratory Principal Analytical Investigator 2S6t0. -P5a-u4l,- 1M7 N 55~~-144-100n0 I 651-733-9859 --- m_m_alinsky@rror;irn .com __ -- Contract Fac il itylL a boratory -- 3M Environmental Health and Safety Operations, - I Enviro nmental Laboratorles ~ Lab Request Number --- Digit Department Number Ea8-0261 -------_ -f - ------ ----- 832202 -~- i ~, ~i Project ScheduielTest Dates . Starting August, 2008 A1l verbal and written correspondence will be directed to Cliff Jacoby. 3M Envtronmertt&1 Laboratory GPO E08-U261 Page 9 of 10 2 Background Information and Project Objective(s) The purpose of this project is to validate method ETS-08-045 . The validation of this method will generally follow the FDA Guidance for Industry ; Bioanalytical Method Validation, May 2001 . Method ETS-08-045 is used to extract perfluorochemicals (PFCs) from fish fillet and whole body sample homogenates, followed by quantitation by LClMSIMS methodology. In general, this method involves the homogenization of fish fillet or whole body fish samples with acetonitrile at a 1 :10 fish mass to solvent volume ratio (i.e. 0.5 g of tissue to 5 mL acetonitrile) . Note : all fish fillets and whole body samples will be pre-homogenized before weighing out aliquots for extraction . The centrifuge tubes containing the protein precipitated fish tissue and solvent are then placed in a freezer for at least one hour. After removal from the freezer, samples are centrifuged at -5C to pelletize the fish solids . Then, a measured volume of the aoetonitrite supernatant is removed to an autovial and acidified with a known volume of 10lo formic acid (necessary for the analysis of small acids) . PFCs in the acetonitrile protein precipitate extract are analyzed using LCIMSfMS where the mass transitions appropriate for the analytes of interest are monitored. Stable isotopes of various analytes will be incorporated, into the method for use as internal standards and surrogates to allow for matrix internal standard quantitation and monitoring of method performance via surrogate recoveries. The method validation proposed here will explore PFC quantitation in fish tissues via the following approaches: (1) internal standard calibration against a matrix-matched calibration curve. (If quantifiable levels of target analytes are present in the control matrix, method of standard addition will be used to determine the endogenous concentration.) (2) intemal standard quantitation of endogenous PFC using a matrix-matched calibration curve of the target analyte's isotopicaliy labeled counterpart . (This approach will be performed for quantitation of PFOS only using a matrix-matched calibration curve of (1,2,3,a'3C]PFOS .) (3) internal standard calibration against a solvent (unextracted) calibration curve . The target quantitative range of this method validation will be 0.2 nglg (0.2 ppb) to 10 nglg (10 ppb), with selected QC samples spiked at levels up to 4000 nglg of fish tissue for PFOS. However, if the endogenous concentrations of the test matrices are sufficiently low, quantitation down to 0.025 nglg will be attempted . The validation of this method will proceed utlizing whole body homogenates from largemouth bass samples purchased from Osage Catfisheries (Osage Beach, MO), a supplier of fish for sclenfific testing purposes. These samples have been sent to MPI (State College, PA) for initial processing, then returned to the 3M Environmental Lab for further processing, storage and utilization . Following the full validation of this method far whole body largemouth bass homogenates, tissues from other species will be cross validated using abbreviated procedures. The planned species cross validation includes fillet and/or whole body hornogenates of channel catfish, bfuegill, and rainbow trout. The application of this method for additional sample types, ie . shellfsh, may be appropriate, providing sufficient quality control components are included in the analysis of those sample types. This project will not include any type of shellfish as part of the validation . This validation project does not address the process of the initial whole body or fillet homogenization, .only the preparation and analysis of individual aliquots of the homogenized whole body or fillet. 3M Environmental Laboratory GPO E08-028t Page 2 of 10 3 Project Schedule 'The project is scheduled to start in August 2008. The validation of this method utilizing largemouth bass is projected to take 1 month. Cross validation of this method to other species will depend on the number of species analyzed . 4 Test Parameters The goal of Ns project is to validate analytical method ETS-S-05 for whole body largemouth bass hornogenates . This method can be used to quantitate the levels of perfluorochemical anafytes in both fish fillet and whole body fish homogenates. This method utilizes solvent extraction of the PFCs via protein precipitation, followed by t-CIMSIMS analysis . The analytes to include in this method validation are the C4 to C perf[uorocarboxylic acids, PFBS, PFHS, PFOS and FOSA. A list of the formulae and molecular weights of these analytes can be found in Table 1 . The stable isotope labeled compounds [1,2,3,4-13C,a]PFBA,-[1,2,3,4-'3C4]PFOA, [1,2 -'3C,]PFDA, ['BOZ]PFOS, [5gO21PFHS, and [j802]PFBS will be spiked for potential use as internal standards (IS) and surrogates. The sample spike level for each IS and surrogate will be approximately 1 .0 nglg (1 .0 ppb). The reference materials to be used in this project are all commercially available . This method validation will incorporate the commercially available linear forms of PFOA, PFHS, and PFOS . However, select quality control samples spiked with 3M electrochemical tluorination (ECF) production lots of PFOA and PFOS will be analyzed with the intent to show that these branched materials may be accurately quantitated against the linear standard . The target quantitation range of this validation for all analytes is 0.2 nglg (2iu0 ppt) to ~10 nglg (10 ppb) for each analyte, or as defined/limited by the endogenous levels of any of these anafytes in the blank fish tissues. If the endogenaus levels of any analytes are below the target level of 0.2 ng1g, the actual LOQ will be the lowest standard point of the curve. In this case, the next target LOQ will be 0.025 ng/g (25 ppt) . fYie ability of this method to quantitate high levels of PFOS in the presence of lower levels of other PFC analytes will be evaluated via dilution QC. The dilution QC saIriples Will be spiked with PFOS at levels up to 4000 nglg (4 ppm) in the presence of the other analytes spiked at lower levels . Because PFOS may be present in the endogenous fish samples at levels above 0.2 nglg . a separate analysis will be performed where a matrix calibration curve of [1,2,3,r+-"C,]PFOS will be prepared using ['eOz]PFBS as an internal standard and ['B02]PFHS as the surrogate. The purpose of this analysis will be to investigate if a lower quantitation limit of PFOS can be achieved using a calibration curve prepared from a stable isotopically substituted analog of PFOS . Area counts of the unlabeled PFOS transitions will be monitored and entered into the calibration equation of [1,2,3,4 13Cq]PFOS to quantify the endogenous and spiked levels . 3M F_nvironmejval Laboratory GPO F-069-0261 Fa99 3 of 10 Table 1 . Target Analytes . Name Perfluorobutanoic acid Synonym or Acronym PFBA (C4 Acid) Formula C3F7COOH Molecular - Weight (Anionic Form 213 [M-H]" Perfluoropentanoic acid PFPeA (C5 Acid) CaFsCOOH 263 [M-H]" Perfluorohexanoic acid PFHxA (C6 Acid) CSFIICOOH 313 (M-Hr Perfluoroheptanoic acid PFHpA (C7 Acid) C6F,3COOH 363 (M-H]~ . Perfluorooctanoic acid PFOA (C8 Acid) [ C7F,5COOH 413 (M-HI' Perfluorononanoic acid PFNA (C9 Acid) CBF,7COOH 463 (M-F!j" Perfluorodecanoic acid ~ PFDA (C10 Acid) CgF19COOH 513 [M-H1' Perfluoroundecanoic acid PFUnA (C11 Acid)~ C,oF2,COOH 563 (M-H]- Perfluorododecanoic acid ~ PFDoA (C12 Acid) - --F- - .: ' Perfluorobutane sulfonic acid PFBS (C4 Sulfonate) C FZ3COOH CdF9S03H 613 (M-H]' 299 [M-Hj" Perfluorohexane suifonic acid PFHS (C6 Sulfonate) C6F13S03H 399 [M-H]' ~, Perfluorooctane sulfonic acid PFOS (C8 Sulfonate) CeFS03H 499 [M-H]~ ' Perfuorooctanesulfonamide FOSA (C8 Sulfonamide) CeFSOzNH2 498 (M-H]- I,, '3C4-Perfluorobutanoic acid 13 C4-Perfluorooctanoic acid 13C2-Perfluorodecanoic acid [1,2,3,4-'3Ca]PFBA [1,2,3,4-"C ]PFOA f1,2-'3C2]PFDA '3CF3('3CF2)z'3COOH CF3(CFZ)3('3CFZ)3'3COOH CF3(CF2)7('3CFZ)'aCOOH 217 IM-H]417 [M-HI- ' 515 [M-H)- 1802-Ammonium Perfluorobutane sulfonate ~ 1802-Ammonium Perfluorohexane suffonate '$02-Ammonium Perfluorooctane sulfonate '3C4-Sodium Perfluorooctane sulfonate --- - - i- +"Oz]PFBS ['eOZ1PFHS u ['BOZ]PFOS [ 1,2,3,4-'3C al PFOS C4F9S'8020' NH4' 303 [M-H]j ,, ~CF3(CF2)3('3CFz)3'CFzS03N-a'~ ' [CeF13S'8020jNH4` [C8F"S'802OrNH4" 403 [M-H]- 503 [M-H ]~ - 503 [ M-H I I ---- Table 2 lists the mass transitions (MRMs) typically monitored for these anafytes . Multiple MRM transitions can be summed for an individual analyte to improve the sensitivity for that anatyte. However, Care should be taken to use a reasonable number of transitions within any one time period, as the inclusion of a large number of transitions can affect the precision and accuracy of all results obtained during that time period . For this reason, it is recommended that the analyte list be split between two or more injections to optimize the overall performance of the method when analysis of all the listed analytes is required . Division of the analyte list also allows the analyst to vary chromatographic conditions to separate matrix interferences from the target analyte 3M Environmental Laboratory GPO EO&0261 Page 4 of 10 when present, which often cannot be done for simultaneous analysis of all corrnpounds in a single cui. For this validation, three separate injections of each sample, QC, blank, etc- are planned. -Table 3 provides a sumrnaiy of the analytes to be monitored in each injection and a briefdescription of the chromatographic conditions . Table 2 Mass Transitions of Analytes . r Compound PFBA (C4 Acid) 49 i Q3 213 ~ ~ 169 PFPeA (C5 Acid) -~~ ----263 -- Y ------ --- 219 - PFHxA (C6 Acid) 313 - -~T - 269, 't 19 PFHpA (C7 Acid) 363 i - 319, 169 PFOA (C8 Acid) 413 - E 369, 2'19, 169 J I ~V PFNA ( C9 Ac id ) 463 j 419, 219 , 169 ---PFDA (C10 Acid) 513 --------4 -`- 469, 269, 219-- PFUnA (Cl 1 Acid) 563 519, 269, 219 - PFDoA (C12 Acid) 613 -- -- ~ ~' 669, 3% 169 PFBS (C4 Sulfonate) PFHS (C6 Sulfonate) ~ 299 -- 399 99.80 ~ 9y"~80' PFOS (C8 Sulfonate) FOSA (C8 Sulfonamide) ~53C4-PFBA ~~ ----- ~~ '3C4-PFOA ---------13CrPFDA - _ ,a02-PFBS 499 498 217 -- ~ T-- 417 l ---- - -- t --- 515 I 303-- ---------- 130, 99, 80" ---i 78 i - 172 I 3'2 ---- -- 470 t34 'a02-PFHS ` - 403 -~-~ - 84 --~ ',a02-PFOS 11Co-PFOS 103 I 8 4 ~ 503 -"----_-- .--_~------ 131, 99, 80 ------ " The MRMs of 399 to 80 and 99 have been documented in literature to result in interferenc.es in some biological tissues, arising from the presence of 5-Nregnan-3,20-diol-3-sulfate isomers[2] . "''The MRM of 499 to 80 for PFOS has been documented in literature to result in interferences in sorne biological tissues, arising from the presence of taurodeoxycholate +sorners (bile sait)[2). ;3M F: ;,4roa ;mental Laboratory GPO E08-0261 Page 5 of iC Table 3. Analysis Summary . Method Transitions Monitored ~ Chromatographic Conditions Small Acids (6 transitions) L 213>169 (PFBA) 217>172 ((1,2,3,4-'3Ca)PFBA surrogate) 263>219 (PFPeA) 313>269, 313>119 (PFHxA) 417>372 ([1,2,3,4-"C41PFOA intemal standard) Analytical Column : Prism RP (2 x 50 mm, 5pm particle size) Mobile Phases A: 5 mM ammonium acetate in 0.01% acetic acid B: Methanol PFHpA, Sulfonates, and FOSA (13 transitions) '"363>319, 363>169 (PFHpA) i 299>99, 299>80 (PFBS) 399>99 , 399>80 (PFHS) 499>99,499>80, 499>130 (PFOS) 498>78 (FOSA) 304>84 (['802]PFBS surrogate) 5p4>84 ([180JPFOS internal standard ~ for sulfonates, FOSA) Extraction Pro-Column: Waters HLB Online Column (3xQOmm, 25pm particle size) Analytical Column : Betasil C18 (2 .1x100 mm, 5Wn particle size Mobile Phases A: 2mM ammonium acetate B: Acetonitrile , 417>372 ([1,2,3,4-'3C4jPFOA internal standard for PFHPA) Large Acids ('17 transitions) 413>369,413>219,413>169 (PFOA) 463>419,463>219,463>169 (PFNA) ~ 513>469, 513>269, 513>219 (PFDA) ~ 583>519 , 563>269, 563>219 (PFUnA) 613>569, 613>319,613>169 (PFDoA) 417>372 ([1,2,3,4-"C41PF0A internal standard) Extraction Pro-Column: Waters HLB Online Column (3x20mm, 25pm particle size) Analytical Column : Betasil C18 (2.1x1Q0 mm, 5pm particle size Mobile Phases A: 2mM ammonium acetate $~ Acetonitrile 515>470 ([1,2-'3C2]PFDAsurrogate) 1,2- CJPFDA surrogate recovery for PFHpA will be recorded from the large acid analysis . The ((1,2C~PFDA transition will not be monitored in the sulfonate analysis. 5 Method Validation The individual components of this method validation are listed below. It is anticipated that all of these aspects will be evaluated, addressed and discussed in the method validation report. 5.1 Species and Sample Types The method validation will include analysis of whole body homogenates of largemouth bass (Micmpterus salmoides) . Method cross validation may include whole body and/or filet homogenates from the following species : Channel catfish (lctalurus punctatus), bluegill (Lepornis macrochirus), and rainbow trout (Oncorhynchus mykiss) . 3m Environmenta! Laboratory GPO ED8-0281 Page 6 o110 5.2 Analytes The analytes to be evaluated in this validation are PFBA; PFPeA, PFI1xA, PFHpA linear and branched PFOA, PFIVA, PFDA, PFLSnA, PFDoA, PFBS, linear PFHS, linear and branched PFOS and FOSA (Ca to C,Z PFCAs, Ca, Cs and Cij sulfonates, and FOSA) . 5.3 Internal Standards and Surrogates The isotopically substituted compounds [1,2,3,4'3C,jPFBA, [1,2,3,4-"C4jPFOA, [1,2-'3C2]PFDA, [y$Oz]PFOS, ['80z]PFHS, and ['BO~PFBS will be used as internal standards and surrogates . Internal standards, surrogates, and calibration spikes {where appropriate} will be spiked into individual aliquots of homogenized fish tissue prior to extraction . The surrogate concentration in extracted calibration standards wilt be at the same level as the other target anatytes, while the internal standard concentration will be 1 nglg for each individual standard (i .e . a multi-level surrogate calibration curve will be generated for accurate quantification of surrogate recovery). QC samples, appropriate blanks, and 'samples" will have surrogate spiked at 1 ngJg . 0 [1,2,3,4-'3Ca1PFOA will be used as the internal standard for the quantitation of all the target carboxylic acids. " f 1,2,3,4-"C4]PFBA will be used as a surrogate to estimate the recoveries of the f;4-C6 carboxylic acids " [1,2-''Q]PFDA will be used as a surrogate to estimate the recoveries of the C7-12 carboxylic acids. " ['$OJPFOS will be used as the internal standard for the quantitation of the target sulfonates and FOSA, " ['802]PFBS will be used as a surrogate to estimate the recoveries of the target sulfonates and FOSA. " For the separate PFOS analysis, [1,2,3,4-'sCdjPFUS will be used to construct the calibration curve with ['aOZ]PFBS serving as the internal standard and [''J2;PFHS as the surrogate. 5.4 Endogenous Levels of Target Qnalytes in Control Timmes The endogenous levels of the target analytes will be determined in thf; +a+fiole body largemouth bass in triplicate (x3) by quantitation against a solvent curve and by the method of standard addition . These data can be collected as part of the evaluation of Linearity, Precision and Accuracy (Section 5.5) . The endogenous levels of each analyte in the tissue wil; be evaluated. These: values, if significant, may need to be accounted for in the rest of this validation project, and in fuure quantitative studies. The endogenous values will be considered specific to the validation tissue used for the study (ie. date of sample receipt, lot#). 5.5 Linearity, Precision and Accuracy The intended quantitative range for this method is 0.25 ngig ;0 .25 ppb) to iJ ngig (10 ppb) comprised of at least six standard points in the final calibration curve If the endogenous level of the target analyte is significant, then the endogenous level of that analyte in that sample will be the L0{? . If the endegAnous level of the.anaiyte is sufficiently low, then the lowest calibration standard satisfying the method criteria for LOQ will be the LOQ (area counts > two times the area counts of the blank samples and accuracy -of 10f--l-3Q%). Quantitation down to 0.025 nglg will be attempted when endogenous levels are sufficiently low. Using whole body largemouth bass homogenates, three curves will be prepared over the range of 0.025 nglg (25 ppt) to 25 nglg (25 ppb) . The final LOO for each analyte wili iargQly depend on the endogenous amount detected in the control tsssue, and for some anafytes will be signifrcantly greater than 0.025 nglg . Each curve point will include internal standards and surrogates . (IS concentration will be constant at 1 nglg, surrogate concentrations will be at the same levels as the other target analytes .l Each run will also include duplicates of the following control samples: 0 control matrix blanks with the ISs and surrogates control matrix blanks without the ISs and surroqates, . aqueous blanks with the iSs and surrc-x,taies, s aqueous blanks without the ISs and surrogates 3Nf Fnvironmenfhl : .abc :rati7ry GPO E08-0261 Page 7 of f 0 " acid adjusted acetonitrile extraction solvent bJanks with~tiieI5 and surrogates " acid adjusted acetonitrile extraction solvent blanks without the ISs and surrogates. The full vaGdation of whole-body bass will include the following preparations to assess linearity, precision, and accuracy. " Duplicate matrix-matched curves prepared on a single day to evaluate the intra-day precision and accuracy. (Two separate analysts may prepare the curves ; however, each analyst must have his or her own set of accompanying QCs as outlined in Section 5.6.) . - A single matrix-matched curve prepared on a different day to evaluate the inter-day precsion and , - a=uracy. For each target analyte and surrogate, the data wtl be reduced according to ETS 8-45 using the following approaches: (1) internal standard calibration against a matrix-matched calibration curve using the method of standard addition when appropriate. (2) internal standard quantitation of PFOS using a matrix-matched calibration curve of (1,2,3,4-'3C4IPFOS for a separate single preparation batch 5.6 Quality Control Samples Each day an extracted curve is prepared, four levels of QC. samples should be prepared for the whole body bass . An additional fifth level of QC will be included on the extraction day in which the single extracted curve is prepared for inter-day comparisons. For PFOA, PFHS, and PFOS, the standard curves will use the linear reference materials. Aspects of the QC. samples will utilize both the linear and 3M ECF production tot branched reference materials. Each OC sample will be analyzed against the relevant species-specific matrix matched curve. As discussed previously, samples will include appropriate internal standards and surrogates spiked at 1 nglg . Internal standard quantitation wit be performed. The QC spike levels should be at different concentrations than the curve points . The first four QC levels need to be prepared each time an extracted curve is generated. Dilution QC need to only .be prepared once during the validation . (1) The low-level spiked QC should be approximately 200% of the LLOQ or approximately 0 .3 to 0.4 nglg ifthe LLOQ is ~5 0.5 nglg. If-the LLOQ is >0.5 nglg, then the low-level spikes will be adjusted accordingly for the affected analytes. (PFOA, PFHS, and PFOS in these QC samples will be the linear reference materials). (2) The mid-level spiked OC should be in the middle part of the curve range . (PFOA, PFHS, and PFOS in these QC samples will be the linear reference materials.) (3) The high-level spiked QC should be greater thao .the mid-level spike but still within 80% of the ULOQ . (PF(UA, PFHS, and PFOS in th~'se QC samples will be the linear reference materials.) (4) Separate triplicate mid-level QCs at approximately 5 nglg wig be prepared that contain the 3M ECF PFOA and PFOS only (no other target analytes). These samples will only be quantitated for PFOA and PFOS (i,e . endogenous levels and impurities of other analytes will not be evaluated) . (5) The dilution QC should be spiked with PFOS at a level of approximately 4000 nglg (4 ppm). The spike levels of the other analytes should be approximately 1 nglg. These QCs will be analyzed without dilution for all analytes excluding PFOS and then post-extraction dilutions will be performed to bring the PFOS levels within calibration range. These high-level QCs are intended to show that low levels of PFCs can be accurately quantitated in the presence of high levels of PFOS . Note: the quantitation of the high levels of PFOS will be done via external standard calibration only . Spiking the internal standard at a comparable ppm level is not practical given that the laboratory's labeled PFOS reference material is only available as a 50 ppm solution, The dilution QC only needs to be prepared once during the validation . Care should be taken to sun the dilution QC near the end of the run after any low-level samples. Additionally, several blanks should be analyzed after the dilution QC to verify that PFOS instrument carryover from the high level sample is not present. 3M Environmental Laboratory GPO E08,0261 Page 6 of 10 Forthe alternate PFoS analysis using the [1,2,3,4-'3Ca]PF05 calibration cune, a minimum of triplicate LCSs of non-labeled PFOS at a level 2,204/o the endogenous concentration will be included . 5.7 Solvent Curve Analysis In a separate preparation batch; triplicate matrix blanks (with IS and surrogate) and triplicate lab centroi spikes at the low, mid, and high levels will be prepared and analyzed against an acid-adjusted solvent curve in acetonitrile . The preparation and analysis batch will include the whole body largemouth bass and all the species/tissue types selected for the cross-validation discussed below. Internal standard quantitation will be performed. 6 Cross Validation of Other Species Cross validation procedures of this method to other tissue matrices will depeno on the results obtained a;, part of this formal validation of whole body largemouth bass . The nine additional matrices proposed for cross validation include the following: largemouth bass fillet, whole body catfish, catfish fillet, whole body bluegill, bluegill fillet, whole body carp, carp fillet, whole body rainbow truuf, ana rainbow trout fillet, The priority and inclusion/exclusion of some matrices may be reassessed after the full vaiidatio:l of the whole body bass . The cross validation procedures will include all preparation, analysis, and quantitation aspects of the full validation described in Sections 5 with the following exceptions : (1) Only a single species-specific, matrix-matched ca!ibration curve will be prepared or a single day (no intra-day or inter-day comparisons will be performed) . (2) Dilution QC of PFOS will not be included. (3) Alternate PFQS quantitation via calibration with [1,2,3,4-"C,JI?FOS will not be included . 7 Method Validation Acceptance Criteria The applicability and acceptability of the method towards each species and tissue type will be determined by evaluation of the results from this project. The aspects to be evaluated and their proposed acceptance criteria are listed below. (1) Precision - The precision of each analyte in each control tissue will be a /oRSD of s t20%. (2) Accuracy - The accuracy of each analyte in each control tissue will be 100t25% (10030% at the LLOQ) . (3) Correlation Coefficient - The correlation coefficient (r) of each analyle in each control tissue will be 4.995. (4) QC Samples - The precision and accuracy of the QC samples will be 10030% for each analyte and the %,R,SD of [_ 2fl% ,for each analy{e, ,resrnentively, Based on the results of the method validation results, the acceptance criteria of this method may be modified based on the analytical needs of a project. Because the intent of the method validation is to determine what acceptance criteria is plausible for this method, method deviations will not be issued for QC samples not meeting the arbitrary requirements listed above. 8 Attachments None 3N1 Envirnnmente! Laboratory GPO E(18-02G1 Page 9 of a0 9 References 1 . FDA Guidaree for Industry, Bioanalytical Method Validation, May 2001. 2. Benskin, J. P.. Bataineh, M, Martin, J.W. . Anal. Chem 2007, 79, 6455-5464. 10 Revisions The original GPO stated that method ETS 8-49 would be vaiidated. While wri0ng the validation report, management decided that the method validated was sufficiently different from ETS 8-49 and requested that a new method number be assigned (ETS 8-45). - The original GPO included four different aspects of quantitation (internal standard quantitation with a matrixmatched calibration curve, external standard quantitation against a matrix-matched curve, internal standard quantitation against a solvent curve and external standard quantitation against a solvent curve). As the project progressed, it became apparent that the amount of instrument analysis time and data reduction needed to fulfill the original proposal was not going to be feasible in the time allotted . After discussions with management, it was decided to eliminate all external standard and solvent curve quantitation requirements . After analysis was completed for the three selected species for cross-vakdation, management reconsidered the need for internal standard solvent curve analysis, The samples outlined in Section 5.7 were then prepared . PFOS dilution QC was intended to be 10,000 nglg (10 ppm). Due to a calculation error, the PFOS dilution QC were prepared at 4,000 nglg . It was decided that this concentration was sufficiently high . The GPO was updated to reflect the concentration prepared . The original GPO detailed proposed analyses for NIST SRMs of Lake Michigan and Lake Superior trout fillets. Analyses of these samples was moved to a separate project with its own GPO (E07-0295) . 3M Envfronrrrenfal Leboratory GPO EOa-0261 Page 10 of 10 3M ENVIRONMENTAL LABORATORY REPORT NO . E'fJ8-0261 SUPPLEMENTAL INFORMATION Supplemental Information Method Validation of ETS-845 "Determination of Ruorochemicals via Protein Precipitation of Fish Tissues (Fillet or Whole Body) and Analysis by High Performance Liquid Chromatography with Tandem Mass Spectrometry" Laboratory Request Number: E08-0261 Testing Laboratory 3M EHS Operations 3M Environmental Laboratory 3M Center Building 260-5N-17 Maplewood, MN 55144 Requester William Reagen 3M EHS Operations 3M Environmental Laboratory 3M Center Building 26Q-5N-17 Maplewood, MN 55144 3W1 IEhfJ(RONiVSEN"iAL LABORATORY PAGE 1 OF 5 >. Z;0 o q ,7~O 1w0 ~ g ,~ 2 Z42 4 a O tn J~ ~ t0 a _N M ~ ~p T~ W c0 G~ O_ C~D fi N N ~ N ~ 0~ ~p ~Cp O ppp~~~ ~Q1 pf O O ~ O ~~ O O O C co w ~ O G 0 O W ~ O W co cc 13 9 ~ o0 0~ o v ~ o v o o w 75 ` I c m a_i a~ m a_i a> ! ,d o~ C u a_~ o _ ~ ~ o 01 . a~ 2 o o b o 25 ~~ 8 a ~ o Z5 a a os b ~ o o 0 - --~ --- - ar R E w E=7 G N C l4 (tj~/j~ A fA r W D i 4; 1 i?r i n`. : h ~r~ a ! w a> tRr 0 , C y O N O ~ 'O a_ 3 D a ~I '~O a ~~ ~ ~ L e n ~ E r- E n Lm E Q U ~ fn CJ ~ (n U ~ VJ ad a3 ~ n ol . O O - A 3~2 `~ N ~~a V ~ ~ ~ a ~ ~ ~ ~ ~a a a l~ r~ ,a E n E r, E E a U ~ U~ U ~ V) C) ~ ~l7 UJ T . ~p T OoW C co 13 o ~ "" OD C7 N 07 i m a a < N N ITy N N m t~ 0 0 Q C ~ ,ro F tp p ~ m ~ !D ~0 ~ _ j Q t-_i o 0 0 ~O V m CeC5~5 a (D `9~ ~ a o Oo1 :~wa Qo Szz J z LLI a " ~ o O ~o c0`~+ oo 00 00 00 ~a' ~o o~oooooo I Q a ' '(( /~ C qj a p I d ~ tb W Q m ~E U ,S dFS tEo Un -~j': vWEJ I o o ~c o m co C3 a a c z c 0 d , R h 1 ll ; a o oW g~ CQ oz'O wz a o 4 w I c a _ d e a 0 Z ~ ~ io 1.`7 qi ~ '-~y O N Q ~ u . CI lL m rn n ~o a o O _ O O O Vi ~ O 2 UI'! a m _m N ~ y ` _m _m C ,_St 0o N GY C . R Q G .R1 a G m 0 m O 0 OT U G U O o $o p G. L7 ~ 72 N ~ y E 0 13 LM .0 -9 a L. o d C Q ~V cn U ~ U C a O 2 .. r . .c U T w T Yi 1A N _ O 9 ' U d `q$ c~o' ~ nb a a~ 2 F d cR ro b 0 3 C~7 . E N ay N 4 3M ENVIRONMENTAL LABOR4 TORY REPORT NO . E(78-0261 SCIPPLEMFNTAL fNFORMATIO(V 4 - -- .--- surrogaffnecoverres -in . sptKetx .:.-----.---~,.,_.__. . . "fable 4. Surrogate Recoveries : Small Acids. 1 .2,3,4-"C PFSA Su i+e Recoveries in 'ked Blank Samples (Sirraft Adds) Batch Qesc ' ffon Alairix Blank R 9 Rep 2 Aqueous R 9 plank 2 WB Largemouth Bass Day 1 ~ 112 112 89 .6 89 .9 WB Largemouth Bass Day 2A ~ 101 102 113 104 VJB Largemouth Bass Day 2B 104 101 98.8 95 .4 WB Catfish 108 105 85.7 86 .6 UVB Bluegill 131 107 132 110 Rainbo wTrout Fillet 124 114 140 156 Acetor+idile BJarsk R 'f R 2 78.6 80-9 110 101 92 .2 86.4 83 .2 84 .0 95 .5 98 .8 136 120 Table 5. Surrogate Recoveries: Large Acids. T 2-PC,M" S P Batch Desc " Son Rowverfres in Spiked Blank Sam Matrix Blank R 9 R2 ~ Acids Aqueous 81ank R 1 2' Acetonlfalle Blank R 1 R 2 WB Largemouth Bass Day 1 WB Largemouth Bass Day 2A WB Largemouth Bass Day 25 WB Catfish WB BIue9EU ~ Rainbow Trout Fillet 121 109 90 .3 106 110 104 103 100 103 105 104 100 92.6 82.6 109 106 104 103 131 -T 132 101 106 99 .4 106 102 103 114 ~ 120 109 , 114 121 95.0 99 .2 96-8 115 1-08 I Table 6. Surrogate Recoveries: Sulfonates, FOSA . 0 PFBS Su ate Recoveries in S Red Blank Sam s ulfonafes, FOS Matrix Blank Aqueous Blank T Aectw8 iff tffae nk ' R 1 Re 2 ~ R l R 2 R 1 ft 2 WB Largemouth Bass Day 1 97.5 99 .0 96 .6 106 94.7 90 .4 WB Largemouth Bass Day 2A 98-3 106 104 95 .5 103 96 .1 ~ WB Largemouth Bass Day 2B 104 102 97 .4 103 I 91 .9 90 .6 ' WB Catfish 102 99 .2 103 94 .9 101 93 .5 W8 Bluegill 97.5 101 95-9 98 .9 88.5 90 .7 Rainbow Trout Fillet 92.4 114 107 98 .3 104 98.8 Vii : NVII'tONMENTAL LABORATORY PAGE 5 OF 6 3M ENVfRONINENTAL LABORATORY REPORT NO . E08-0261 SUPPLEMENTAL INFORMATION Table 7. Surrogate Recoveries: Solvent Curve Analysis . 8vfvent Curve Large Acids Large Acids V11BB1*gillR6pn9p C7 Acid, SuIFonaOes, FOSA C7 Acid, S utfonaQes, FOSA WB BIuegiR ReRrep _ R t 89 .6 99 .3 106 110 Aqueous Blank RY R3 101 100 106 NA 98 .3 97.1 109 -NA -_ R 4 108 NA 105 NA Acetwrihile Blank R 9 R 109 98 .2 I 101 10.0 98 .4 92 .5 110 101 3M ENVIRONMENTAL LABORATORY PAGE 6 OF 6