Document 2qagno5qExBp0j7ykj775rrwL

.aim o@ Foro"747-1 I-A TECHNICAL REPORT SUMMARY TO: TECHNICAL COMMUNICATIONS CENTER - 201-2CN (important-Ifreportisprintedon bothsidesofpapers,endtwo copiestoTCC.) Division Environmental Laboratory (EE ?z PC) Project Fate of Fluorochemicals Report Title Analytical Methodology and Su@.)port To D. L. Bacon Author(s) Arthur Mendel Notebook Reference 41947, 44191, 46269, 47703, 48277, 49400 SECUIIITY 110, 11 Open (Company Confidential) r Closed (SpecialAuthorization) 3M CHEMICAL REGISTRY KEYWORDS: (Selecttern-tfsrom 3M Thesaurus. Suggest other applicableterms.) CURRENT OBJECTIVE: Progre.-,sReport EE & PC-Div. Fluorochemicals Do 1*1/17/79 D-op-t-gubm-or .0535 ProiectNumber 9970612643 Report Num 008 Employee Number(s) 043939 No. of Pages IncludingCovershost 22 New Chemicals Reported 0 Yes g) No REPORT ABSTRACT: (200-250 words) This abstractinformation isdistributedby the TechnicalCommunications Center to alert3M'ers to O)mpany R&D. ItisCompany confidentialmaterial. nformation Liaiso Iinitiols: v -2- INTRODUCTION This report describes supporting analytical work not detailed in reports on bioassay, die-away, soil absorption-desorption, solubility, and the like on various fluorochemicals. DISCUSSION AND RESULTS The loss of FM 3422 on rotary evaporation of its water solutions was noted previously (1) , and only ten percent of FM 3422 was recovered; see the experimental section of this report for details. In another experiment, a water solution or FM 3422 was allowed to evaporate under ambient conditions with essentially complete loss of FM 3422. Interestingly, similar losses of PCB's in water due to volatilization were reported (2). Note that the solubility of PCB's in water is very similar to that of FM 3422 (around I ppm or less). In the case of PCB'S, volatilization was minimized by working in a cool environment and covering samples with a layer of hexane. The distribution coefficient for I-'M3422 was reported earlier (3), wherein the solubi.lity of FM 3422 in water was not completely defined but based on experimental work, thought to be less than 0.05 ppm. This value was based on water solutions filtered through various size membi-anes and examining the aqueous riltrate with a light beam (Tyndall effect). It is known that membrane filters contain impurities which may interfere by removing a substrate in water or otherwi.,-;(c-oiitlxminatiilgLlicw..LL(@(t4-,5). Ttio solubility of FM 3422 in water using the Veith technique (6) was calculated to be 0.05 ppm, and the distribution coefficient for FM 3422 is thus 330,000 6P pm (in n-octanol) divided by 0.05 ppm (in water) or P = 6.6 x 10 N-alkyl-substituted perfluorooctanesulfonamides have been shown to undergo alkaline and acidio hydrolysis; however, the reaction takes days (7). A pilot stud@,yon the basic ethanolysis of FM 3422 also confirmed thit this substrate is not rapidly degraded (46269-45) and that the salt of perfluorooctanesulfonic -acid was formed (infrared and TLC studies). Accordingl!,, a kinetic study was conducted on FM 3422 using 10% potassium hydroxide in absolute ethanol (to ensure complete solution). The reaction_Sate_2beyed first order kinetics with a r-.t.tceonstant K = 9 x 10 hr and a half life of seventy-seven houi-s. In earlier studies (8), glass containers were shown to be satisfactory for the quantitative recovery of FM 3422 from water in the parts per million (ppm) or less concentration range. In the current work, similar aqueous solutions of FC-95 could be kept in polypropylene, but not in polyetiiylene, polyearb(.)nate,or glass. Aqueous solutions of FC-143 could b(@ kept in polypropylene; polyethylene and glass were less suitablo. These studies were done with radioactive compounds (see exl)erimental). -3- Based on studies to date, it appears that compounds such as FC-95 and FC-143, which have highly ionizable groups adhere to glass (which may be considered to have an ionic surface) but not to plastic (hydrocarbon or nonionic surface). For quantitative recoveries then, such compounds should be stored in plastic containers. On the other hand, nonpolar compounds, such as FM 3422, adhere (dissolve?) to plastic but not to glass, and the latter would be the cont:iiner of choice. Another rationale involves the distribution coefi'icient. Samples which have a high distribution coefficient (i.e., are very lipophilic) would prefer to adhere (dissolve?) to plastic compared to glass (so use glass containers), while samples which have a low distribution coefficient (i.e. are very hydrophilic) would prefer to adhere to glass compared to plastic (so use plastic containers). The use of the proper aquaria (glass or plastic) and the most suitable containei-s for shipping water samples and fish back to the Environmental Laboratory --For'farther analyses was made known to Bionomics Laboratory persoiinel before they conducted critical life stage studies on FC-95, FC-143, and F.M 3422. Aluminum foillined caps, used to seal bottles containing aqueous solutions of FM 3422, were also examined before use for possible contaminants that might. interfere with subsequent GC analyses; no interferences were observed. Since a simple, rapid method for quantitation of FC-95 is still not available, the methylene blue procedure for methylene blue active substances (MBAS, (10)) was used to quantitate FC-95. A I i.ne,,xr range from 014 to 0.06 ppm was noted on sc-milog paper (Figure 1). Of course, MBAS, such as linear alkyl sulfonates (detergents), if present, will give a positive interference. TLC and GC were also performed on die-away experiments with radiolabeled FC-95 and FC-143. Die-away samples were spotted directly on E. Merck silica gel TLC plates which were then developed and visualized by autoradiography. A comparison with controls showed no new compounds were formed respectively. Samples including controls were methylated and gas chromatograpped to see if any new volatile materials were generated but none were observed in either case. GC of the samples prior to methylation also indicated that no new compounds were formed in each case. Chiou and co-workers (9) published an article which described an empirical equation relating experimental n-octanol/water distribution coefficients to a(lueous solubilities of many organic materials. Their correlation, whic@ coverid more than eight orders of magnitude in solubility (from 10 to 10 ppm) -.xnlsix orders or magnitude in distribution coefficient (from 10 to 10 ), may allow an assessment of distribution coefficient from solubility with a predicted error of less than one order of magnitude. Furthermore, these workers observed a correlition between the bioconcentration factors in 100 ui i 0 T 1:1 R A N s m s s 0 N '0 !j ix -4- t iLt L 1 i-k I"C-93 wi.Lli Me Lhy 1 1.u(@ I- 0.1 ().2 0.3 0.4 0.5 FC-95 1 -5- rainbow trout and the aqueous solubilities for some stable organic compounds. It was of interest to apply their empirical equations to solubility, bio concentration factor, and distribution coefficient data obtained in the Environmental Laboratory on FM 3422, FC-95, and FC-143. Results are -summarized in Figures 2 and 3 (11). Soil thin-layer chromatography (soil TLC) of pesticides appears to offer a simrile, yet reliable, method of evaluating relative mobility in soils, even though a universal "standard" soil has not been designated. The advantages of soil TLC include rapidity, reproducibility, and low equipment costs. In the current study, a soil-coated TLC plate was spotted with radiolabeled FM 3422, FC-95, and PC-143, and the ascending waterdeveloped plate was visualized by autoradiography. Results show that none of the samples appear to migrate (R =0). Soil TLC in the descending development mode will be done lo simulate the leaching of materials through soil. To date, no evidence has been noted for the biodegradation of FM 3422. It was Lhought that elect rochem i.-caloxidition (voltammetry) might be successful. If so, the potential required would be an indication of the ease or difficulty of oxidation of this alcohol to the aldehyde and/or acid. Furthermore, it was hoped that voltammetry (i.e., oxidation potential).might be a good general physical-chemical predictor of the biodegradation of a candidate. In this work, FM 3422 in acetonitrile was not oxidized by electro- chemical procedures using a platinum electrode. Under the same conditions, n-bu-t:i.n(-w)-l.xsnot oxi(li-ycdoil.lic@r(Cill, B34247). It appears that this method is probably riot a good index for the "oxidation potential" of a candida'L-eorganic. 7 10 :r DDT 6 10 Relationship Figure 2 of Partition Coefficients to Water of Organic Compounds Solubility 2,4,5,2,4'05#,-PCB cC".2.4.5.2,5',-PCB o Chloropyrifos 5 F.Nl.3322 10 4 10 io3 10 0 10 10 -3 -2 10 Niethyl Chloropyrifos Parathion .2.4-D FC-95 o TOLU Benzi-.ne io- 10 SOlv.bility in water 10.;, Plitn PhenoxyaeL@ 1 umoles/1), log scale Figire 3 Relationship o-r Biocon cent ration Factor to Wat.(@rSolubi-lity ol' Organic Compounds 5 10 to 0 -10 0 0 2,4,2 4 -PCB xachlorobenezene .0ri 4J 10'. 4-) r. 0 C) 10 0 Diphenyl Ether, Biphenyl p Dichlorobenzene 10 io-2 io-i cci 4 1 10 10 2 10 3 4 10 SOLUBILITY IN WATER uMOLES/1, log scale EXPERIMENTAL Volatility Studies on FM 3422 A. Rotary Evaporation under Reduced Pressure (41947-38,-39,-40) Water,-saturated with FM 3422 conducted in aquaria (see 42669-17), was used for samples, and a 5 ppm reference of FM 3422 in methanol was also prepared. GC analyses indicated the water thus saturated had a 5 ppm concentration of FM 3422 (note that this is high, based on the Veith technique used later. This high value may be due to supersaturation and/or micelle formation because these liquids had a large Tyndall effect, and the solutions were not filtered), A 25-ml water sample above was concentrated on a rotary evaporator under water aspirator pressure. The residue was taken up in octanol and gas chromatographed. Results showed that only 0.5 ppm of FM 3422 remained. A replicate experiment indicated that only 0.6 ppm of FM 3422 remained. Thus about ninety percent of this alcohol was lost due to volatilization. B. Evaporation under Ambient Conditions (44191-43, -44; 46269-32) A saturated a(lueous solut-.Lonof FM 3422 prepared and determined by the Veith technique (6; see 44191-43 and references therein) was s'-..owtno have a concentration of 0.05 ppm. Foi-the evaporition study, 100 ml (vol. pipet) of this solu(.ionwis tr,-tnsrorr(,t,cd)a tii(iIts contents allowed to evaporate (ca. 1.5 weeks). The beaker was rinsed with 2 ml (vol. pipet) of methanol, and an aliquot was g:ts chromatog-raphed. Results indicated,that 0.007 ppm of FM 3422 remained. Reaction of FM 3422 with Alcoholic Potassium Hydroxide (48277-29, -31, -39; 46269-4, -5): An alcoholic potassium hydroxide solution was prepared in a 500-ml round-bottomed flisk (containing two joints, -one a T 24/40 female and the other a t 10/30 female for aliquot withdrawal*, and containing a teflon-coat(!dmagnetic s irrer bar, and an air-cooled condenser topped witli an Ascariteh protection tube) by dissolving 11.6 g of 86.2% pc)tassium hydroxide (10 g if 100% KOH) in 200 ml of absolute alcohol. This solution was then maintained at 50-530 by an external oil bath for the duration of the experiment (504 hours). One milliliter aliquots were withdrawn from this solution with a volumetric pipet through the normally closed t 10/30 entry port. This aliquot was (juantitativelytransferred to a 25-ml volumetric flask containing tbout 10 ml of absolute alcohol, one drop of phenolphthalein solution and three drops of concentrated nitric acid. The purpos(.@was to stop any further reicti(.)nand to neutralize the base whicli proved to adversely affect the gas chromatographic -9- column used subsequently. The volumetric flask was diluted to the mark with absolute alcohol, and an aliquot of this solution was subjected to gas chromatographic analysis on the HewlettPackard Model 5813 gas chromatograph using electron capture detection. To the above alcoholic KOH solution was added 40 mg of sublimed FM 3422, and immediately thereafter (solution of the FM 3422 required ab-out I minute) a 1-T@ilaliquot was withdrawn and designated as time zero. Aliquots were removed and analyzed at times indicated in Table 1 and illustrated in Figure 4. A first-order reaction rate is shown by the following kinetic treatment (12): ln(a-x) = -kt + In a or lnlo(a-x) = '-k t + In a :@30 3 10 Since for any experiment, a is a constant, the above is an equation for a straight line. When plotting ln(a-x) versus t, lnioa will k be the y intercept (found to be 2.2) with the slope equal to /2.303) and found to be -0.004. Table 1 All@aline Alcoholysis of FM 3422 t (hrs._l 1 4 7 2,1 .48 72 96 168 216 264 336 384 432 504 C (Ppm)* 169 178 176 165 135 ill 88 73 44.4 27.5 20.3 7.3 0.6 1.2 In 10@l 2.22 2.25 2.24 2.22 2.13 2.04 1.94 1.86 1.65 1.44 1.310.86 0.78 0.04 This number rel)resents the average of two duplicate. Poor electroni(- integration - no results: replicates done in A plot of In C versus time gives a slope of -0.004; 1() Since k = -2.303 (slope) then k = -2.303 x -0.004 9 x 10 - 3 hr -1 half life = t 1 - In 2 k- 0.693 9xlo-.j 77 hrs. 175 150CF 0m N 3 c 4 2 2 100 (ppm, 50- Ell I 1-i*9LIre. 4 Alkaline Ethano YSIS of FM 3422 -FT 0 50 100 150 200 250 300 350 400 Time (Hrs.) NOI.LV&iocl mt mow. mom N3v3l3lo N3dVd HONI tl3d 01 X Di HaVefEl N313Z-L31CI 01-0t'C 'ON PM 3422 Sorption ()nto Soil (46269-50) Sorption-desorption soil studies were conducted by S. K. Welsh, and the results are given in Table 2. Soil TLC (48277-30) Various types of soil were ground and sievedand that fraction of 262 micron size was saved for TLC studies (42669-42). An intimate mixture of 150 g of 262 micron size greenhouse soil (sandy loam) with 80 ml of water was well slurried and spread onto 10 x 20 em glass plates using an in-house designed spreader (designed by J. W. Belisle, A. Mendel, and G. Guthrie). The plates were allowed to air dry. A plate was spotted 2 em from the bottom with radiolabeled-FC-95,-FC-143, and-FM 3422. The samples were those used earlier for water solubility studies. The plate was air-dried and developed 100 mm from the original spot in the ascending mode with water in a filter paper-lined TLC developing tank previously equilibrated with water for one day. The development required nearly two hours and the air-dried plate was then autoradj()graphed (13 ). results ,howed t.iiitFC-1,13 and FM 3422 did not leave the orii' .-in(R =0) The spot, due to FC-95, was extremely faint for visualizati@n b*ut appears not to have migrated either (Rf=o). Preliminary Studies on Samples from Bionomics Labs A. Examination of Foil-lined Caps for Interferences (46269-50) Four foil-lined caps which Bionomics Labs uses to seal bottles containing samples were screwed onto four clean bottles containing 25 ml of deionized water. The bottles were mechanically shaken for 4 hours,and the combined .100ml of water was extracted with ethyl acetate. This extract was concentrated and chromatographed. No apparent interference for FM 3422 was noted. This extract was then spiked with FM 3422 to confirm no interference at the retention time of FM 3422. B. Examination oC Standard s and Controls Prepared at Bionomics Lab Personnel from Bionomics Labs prepared standard solutions of FM 3422 in water, then submitted one-liter samples of ,these standards and controls to us for analyses. The respective sample was extracted with ethyl acetate and the organic extract was concentrated to 10 ml. An aliquot was used then for GC analyses, and the results are summarized in Table 3 (Notebook 46269-50). Sample 32A 32B 56A 56B 100A IOOB BKA BKB Day 2/22/78 2/22/78 2/22/78 2/22/78 2/22/78 2/22/78 2/22/78 2/22/78 Table 2 Sorption-Desorption Studies on FM 3422* FM 3422 (12pm) .0.01 0.01 0.01 0.22 0.22; 0.20 O.'L9; 0.21 0.48; 0.510.49 Day 2/23/78 2/23/78 2/23/78 2/23/78 2/23/78 2/23/78 2/23/78 2/23/78 FM 3422 (ppm) <0.002 0.002 0.003 0.005 0.014; 0.019 0.010; 0.011 0.044; 0.035 0.029; 0.029 -Day 2/24/ 2/24/ 2/24/ 2/24/ 2/24/ 2/24/ 2/24/ 2/24/ *See interoffice memo of S. K. Welsh to D. L. Bacon, FC Project, Soil Adsorption, Sept. 22, 1977 for methodology. Table 3 Analyse s fox: FM 3422 (in ppb) in Bionomics Samples Sample I.D. A5609 A5610 A5611 A5612 A5613 A5614 A5615 A5616 A5617 A5618 A5619 A5620 A5621 Theoretical Conc. (ppb) 20 20 10 10 5 5 2.5 2.5 1.25 1.25 Control A Control B Solvent Control A Determined 20.8; 19.8 19.8; 20.6 10.2; 9.9 9.7; 10.1 5.2; 5.2 4.8; 5.1 2.5;.2.6 1.2; 1.3 (a) (a) <0.1 <0.1. <0.1 A5622 Solvent :@:k Control *B <0.1 A5670 A5670 1.25 1.25 1.1; 1.0(b) 1.0; 1.6(b) (a) received broken in shipment; (b) replaces original samples A5617; A5618 C. Examination ol'Water Samples for Possible Intc-rferences of FC-143 (46269-53) As done in Pai-t B above, seven samples of water, received from Bionomics Labs were extracted with ethyl acetate, followed by diazomethane methylation and gas chromatography to determine if the water and the various containers holding samples may present interferences for the analysis of FC-143. All samples appeared to be free of interferences in the GC region for FC-143 analysis. Studies on FC-95: To Determine if FC-95 is Sorbed On/In Containers (47703-20, -22, -24) Three 0.20-ml samples of 14 C FC-95(aqueous solutions saturated by the Veith technique (42669-30)) were transferred to scintillation vials to which 15-ml Aquaso,R was added in each case. These control samples were labeled C-1, C-2, and C-3. Ten milliliters of the 1,4C FC-95 saturat(,d solution was transferred to e.-Leho-r the follow- ing containers: Polypropylene (PP), polyethylene (PE), poly- carbonate (PC), and glass (G). Each container was tightly capped and allowed to stand at room temperature for one week, at which time the solutions were transferred to glass jars. The various containers were rinsed twice with 2-ml water and the rinses added to the respective jar. By pipet, 0.2 ml of the solution was trans- ferred from the j:i.irnto a scintillation vial. This was re eated two times to genei-ate three samples from each jar. wlts added to each scintillation vial and then'counted (13). These vials were labeled PP-1, PP-2, etc. Each of the saved containers from above was rinsed a first and a second time with 4 ml of methanol, and each rinse was transferred to a separate, clean vial . By pipet, 1 ml of the respective methanol rinse above was transferred to each of three scintillation vials containing AquasolR and then counted. The samples were labeled using the notation above, but the letter F for first rinse and S for the second rinse was added to the d6signa- tion, e.g., PPF-1 indicates polypropylene container, first methanol rinse of the first sample,while GS-2 indicates a glass container, second methanol rinse of the second sample. The results are sum- marized in Table 1: Polypropylene was the -ri.rst(@ontii.nerof choice followed by polyethylene. Studies on FC-143: To Determine if P(@-143 is Sorbed On/In Container!r;(47703-30, -37; 49400-51) This experiment was performed essentially like that for FC-95,with the exception that the polycarbonate containers were not investigated. Various solvents attack polyearbonate (e.g., ethyl acetate). Control samples C-1, C-2, and C-3, as well as the 500-ppm 14C FC-143 solution were transferred to polypropylene (PP)-, polyethylene (PE)-, and glass (G)-containers. Each container was tightly capped and allowed -15- Table 4 Sorption Studies on FC-95 (Labeled) Part 1: Scintillation Counting on Samples Sample dpm Conc. mg/l FC-95 pg absolute C-1 15,163 266 C-2 16,058 282 C-3 15,735 276 2660 2820 2760 PP-1 11,294 198 PP-2 10,523 184 PP-3 11,585 203 PE-1 llp278 198 PE-2 11,808 207 PE-3 11,994 210 PC-l 11,493 201 PC-2 11,558 203 PC-3 11)138 195 2770 2580 2840 2770 2900 2940 '2810 2840 2730 G-1 11,840 208 G-2 11,252 197 G-3 11,760 206 2910 2760 2880 Part 2: ScintilliLion Countir;gon Rinses. PPF-1 PPF-2 PPF-3 15.9 16.8 7.22 0.06 0.06 0 .ox@, 0.24 0.24 0.08 PPS-1 PPS-2 PPS-3 5.80 <Background <Background 0.02 0.00 0.00 0.10 0.00 0.00 Avg. Total Avg. pg 2750 2730 2870 2790 2850 @0.19 0.03. 0.27 PEF-1 PEF-2 PEF-3 PES-1 PES-2 PES-3 50.5 60.5 40.5 26.1 4.66 31.0 0.18 0.21 0.14 0.09 0.02 0.11 0.72 0.84 0.56 0.45 0.10 0.55 Avg. Total 0.71 0.37 1.08 Table 4 (continued) -16- Sample PCF-1 PCF-2 PCF-3 PCS-L PCS-2 PCS-3 dpm 1829 1865 1958 270.3 294.0 254.0 Conc. mg/l FC-95 6.41 6.54 6.86 0.95 1.03 0.89 GF-1 GF-2 GF-3 GS-1 GS-2 GS-3 140.9 140.6 146.1 126.8 118.6 125.5 0.49 0.49 0.51 0.44 0.42 0.44 pg absolute 25.6 26.2 27.4 4.75 5.15 4.45 Avg. Total. Avg. pg 26.4 4.78 31.18 1.96 1.96 2.04 2.20 2.10 2.20 Avg. Total 1.99 2.17 4.16 -17- to stand at room temperature for one week, at which time the solutions were transferred to glass jars. The various containers were rinsed twice with 2 ml of water and the rinses added to the respective jar. By pipet, 0.2 ml of the solution was transferred from the jar into a scintillation vial. This was repeated two times to generate three samples from each jar. Aquaso,R was added to each vial and then counted. These vials were labeled PP-1, PP-2, etc. Each of the saved containers from above was rinsed a first and a second time with 4-ml methanol, and each rinse was transferred to a separate, clean vial (in the case for the glass container, 3 ml rather than 1 ml of methanol was used for the first transfer, thus GF-2 contained 1 ml of rinse and GF-3 had to be eliminated). By pipet, 1 ml of the respective methanol rinse above was transferred to each of three scintillation vials containing Aquaso,R and then counted. The samples were labeled with the notation mentioned earlier. The results are given in Table 5. Polypropylene, the container of choice, was used in sx-bsequent experiments. Solubility of FC-95 in Water (47703-15) The Veith technique (6) for saturating water with organics was conducted using radiolabeled FC-95 as described in Notebooks 287103-5 and 42669-30. Three 0.25-ml samples were withdrawn from the saturator at 0, 1, 2, 4, 6, 24, and 48 hours. The samples were placed in scintillation vials to which 15 ml of AquasolR was added, and counts were tal,.enon the Nuclear Chicago scintillation counter (Agrichem). The solubility results are tabulated below: Hr. mg/l FC-95 0 0 1 226 2 238 4 282 6 280 24 292 48 291 Avg. (Hrs. 4-48) 286 Quantitation of FC-95 by the Methylene Blue Method (47703-44) FC-95 was quantitated by the methylene blue method for methylene blueactive substances as detailed in Standard Methods for the Examination of Water and Wastewater (10) . Known concentrations of FC-95 (100 ml, see below) were added to each of five separatory funnels followed by 25 ml of methylene blue reagent and 10 ml of etilorororm. The funnels were shaken for one-half minute and the contents were allowed to separate. The desired organic phase was withdrawn through glass wool and collected. The percent transmission of the organic phase was determined after the Spectronic 20 instrument was first calibrated (652 nm) with chloroform. The concentration versus percent transmission is shown iiiTable 6 and illustrated in Figure 1. Tat-le 5 Sorption Studies c-n FC-143 (Labeled) Part 1: Scintillation Counting on Samples Sample dpm Cone. mgll FC-143 pg absolute C-1 34,456 534 C-2 34,521 535 34,599 536 5340 5350 5360 PP-1 24,829 385 PP.-@-2 24,560 381 PP-3 24,190 375 5390 5334 5250 PE-1 23,664 367 PE-2 24,078 373 PE-3 23,409 362 G-1 24,962 387 G-2 25,546 396 G-3 25,002 388 5138 5222 5068 .5418 5544 5432 Avg. pg 5350 5325 5143 5465 Part 2: PPF-1 PPF-2 PPF-3 PPS-1 PPS-2 PPS-3 Scintillation 293 270 296 21.5 20.9 25.3 Counting on Rinses 0.91 0.84 0.92 0.07 0.06 0.08 3.6 3.4 3.7 0.35 0.30 0.40 Avg. Total 3.6 0.35 3.95 PEF-I PEF-2 PEF-3 PES-1 PES-2 PES-3 16,936 17,268 17,430 859 849 802 52.5 53-- 5 54.'0 2.7 2.6 2.5 210 214 216 13.5 13.0 12.5 Avg. Total 213 13.0 226 GF-1 GF-2 GS-1 GS-2 GS-3 13,530 4,414 124 129 126 14.0 13.'t7 0.3f,' 0.40 0.39 56.0 54.8 1.9 2.0 2.0 Avg. Total 55.4 1.0 57.4 -19- Table 6 FC-95 by Methylene Blue Concentration (ppm) 0.0 0.06 0.12 0.20 0.40 of FC-95 Transmission 100 76 56 40 17 Note that FC-95 can be quantitated in this range; however, the limitations are that any methylene blue-active substances such as anionic surfactants or other sulfonic acids will present a positive interference. GC Calibration and Recovery Studies on FC-143 (46269-34, -54) The purpose of this work was to prepare absolute standards of FC-143 and to evaluate recovery from water. An internal standard of,perfluorodecanoic acid (obtained from Dr. Jon Belisle, CRL) was prepared by weighing and dissolving 0.100 g of this acid into 100 ml of hexane (1000 ppm). A standard of 1000 ppm of FC-143 (lot 83) was made by weighing and dissolving 0.100 g of FC-143 in 100 ml of methanol. For acidification c)f aqueous solutions, I g of p-toluene sulfonic acid (PTSA) in 100 ml of methanol (104 ppm) was prepared, and 2-3 drops of this solution was added to extractants prior to diazomethane methylation. Sariples prepared for GC was as follows: In a 20-ml vial was placed 5 rpl benzene and in a second vial was placed 5 ml of 107o ether in hexane. Each vial was spiked with 20 pg of the FC-113 and with 20 11.-o1f the internal standard. Three drops of PTSA were added to each vial followed by ethereal diazomethane. After 15 minutes of reaction time, excess diazomethane was removed (N2 purge) and the samples were diluted to 10 ml.(volumetric flask) followed by GC. Results showed that benzene was superior to ether@hexane as a solvent (larger integrator response). Furthermore, the electron capture response to FC-143 (as the methyl ester) was different from that to the perfluorodecanoic acid (as the methyl ester). Accordingly, various concentrations of b@6-ththese acids were made, follow(,d by methylation and gas chromatography. An average response ]'actor was thus obtained. Hen, 0.1 and I ppm standards of FC-143 were prepared in witer. The water sample, containing the internal standard acid, was benzene extracted, concentrated, acidified (PTSA) , and methylated in the usual manner. GC of the samples showed recoveries of 93% and 100.4k for the 0.1 ppm solution and 90.96/'aond 99.4% for the 1 ppm solution. -20- TLC OF Biodegradation St'udies on Radioactive FC-95, FC-143, and FM 3422 (47703; 6-140 21P,25-27, -35; 39-43, 46) Details of the biodegradation studies on labeled FC-95, FC-143, FM 3422, and cohtrols are described in E. A. Reiner's report (14). In the present work, the nineteen samples each from sludge, water, and ethyl acetate extracts generated from biodegradation studies were applied directly to plates (20x2O cm, E. Merck GF 254 silica gel). The plates were developed in ethyl acetate then examined first by ultraviolet light (long and short wavelength) and then by autoradiography. The plates were exposed to Kodak x-ray film for one week, and the film was then developed and examined. Comparison with control samples indicatecino significant differences. In a second part of this experiment, a portion of the ethyl acetate extract from the various samples was methylated with diazomethane to convert any nonvolatile components to GC volatile species. Both the unmethylated and methylated samples were then analyzed by gas chromatography. No differences were noted between these and control samples. GC conditions are detailed in Notebook 47703-43. Determination of the Solubility_pf 1,4-MethylFOSE Alcohol FM 3925 (49277-8; 46269@5@; 42969-4) About 20 g of FM 3925 Lot 505 was melted (steam bath) to obtain a homogeneous sample. The cooled, solid material, 200 mg, was dis- solved in about 200 ml of reagent grade acetone, and this solution was added to precleaned 2 mm diameter solid glass beads contained in a beaker. Beads were precleaned successively with heptane, chloroform, acetone, water, methanol, and then air dried. The FM 3925 coated glass beads were placed in a glass coluwn (350xl8 mm), and supported by glass wool (pretreated with R Glastreat The end rubber stoppers w(@re connected to Bev-Line V tubing and ELperistaltic pump...wasused to circulate deionized water through the column. The entire setup was conducted in th8 dark in a temperature-controlled walk-in environmental room (18+1 ). Prior to the start of this experiment, water was pumped f5r 5 hours through the setup, and this water was discarded. The purpose of this was to remove any impurities that would be as soluble as, or moie soluble than, FM 3925. Fresh w;tter was then pl aced into the system'and at intervals (see Table 3), 100 ml (volumetric' pipet) of water was withdrawn from the main 2-liter storage chamber. Each of the 100 ml samples was quantitatively transferred to a separatory funnel, and 10 ml of saturated aqueous sodium chloride was added followed by 15 ml of ethyl acetate. The mixture was extracted and the separated organic phase was quantitatively transferred to a 10-ml volumetric flask. The aqueous phase was reextracted with 4 ml of ethyl acetate, and the organic phase was added to the 10-ml volumetric flask whose volume was finally brought to the mark with ethyl acetate. An aliquot was used l'orgas chromatography using an electron capture detector aBd a six-foot (1.8 meters) column of Carbowax CW20M run at 180 The results are reported in Table 3. -21- Table 3 Sol ubility of FM 3925 in Water Time of Water Circulation (hrs.) 1 2 4 6 8 24 FM 3925 (ppm) 1.72; 1.90; 2.10; 2.33; 2.34; 2.17; 1.82 1.87 2.17 2.30 2.35 2.14 Avg. (Hrs. 4-24) = 2.3 ppm The The the are solubility of FM 3925 is forty-six times that retention tim(@ or FM 3925 is close to ihat of gas chromatographic patterns (see Figure 2 in essentially the same. of FM 3422. FM 3422, and reference 1) REFERENCES (1) "Analytical Methodology on FM 3422,11 A. Mendel's progress report to D. L. Bacon, dated 11/15/77, p. 8. (2) D. B. Easty and B. A. Wabers, Analytical Letters, 10, 857 (1977). (3) Interoffice cot-respondence of G. A. Vraspir to D. L. Bacon, March 26, 1976, entitled "Gas Chromatographic Analyses of FM 3422." The distribution coefficient is defined as the ratio of the concentration of compound soluble in n-octano I phase to the (-.oncentrationof all species in the aqueous * phase at a given pH (usually 7). (4) A. Otsuki and K. Fuwa, Talanta, 24, 584 (1977). (5) G. T. Wallace, Jr., I. S. Fletcher, and R. A. Duce, J. Environ. Sci. ll(,,-iltAh1,2 493 (1977). (6) G. D. Veith and V. M. Comstock, J. Fish. Res. Board Can., 32, 1849 (1975). (7) R. F. Heine and R. R. Burford, C. G. Klaus, J. D. LaZerte, J. W. Sargent. (Unpublished) (8) A. Mendel report on Fate of Fluorochemicals Nov. 15, 1977. to D. L. Bacon, (9) C. T.,Chiou, V. H. Freed,'D. W. Schmedding and R. L. Kohnert, Environmental Science and Technology, 11, 475 (1977). -22- (10) "Standard Methods for the Examination of Water and Wastewater, M. C. Rand, A. E. Greenberg, and M. J. Taras, Eds., Washington, DC, 14th Edition, 1976, pp. 600-603. (11) Taken from A. N. Welter's oral presentation to Commercial Chemicals personnel, Oct. 23, 1978; minutes of meeting, Oct. 31, 1978. (12) C. F. Prutton and S. H. Maron, "Fundamental Principles of Physical Chemistry," Macmillan, 1951, Chapter 19. (13) Agrichem Laboratory personnel are acknowledged for their help in autoridiography and scintillation studies. (14) E. A. Reiner, "Biodegradation Studies of Fluorocarbons III," July 19, 1978. AM/den Attachedarecomments on the3M TechnicalReport "AnalyticaMlethodology and Support. ArthurMendel, Project9970612643 Fateof FluorochemicalsR,eport No. 8,Jan.17,1979" made by ProfessorStephenA. Boyd, Michigan StateUniversityd,atedMay 19,1993. Review of TechnicalReport Summary AnalyticalMethodology and Support Discussionand Results The octanol-wateprartitiocnoefficienotfFM 3422 was reportedas 6.6 x 101.There isa strong correlatiobnetweentheoctanolwaterpartitiocnoefficienatnd thesolubilitiynwater fororganic solutegsivenby theequation(Chiouetal.,1982,Environ.Sci.Technol.16:4-10). logK,. = 0.862log S,,+ 0.710 This relationshwiopuld predicta supercooledliquidsolubilitoyf ca..47ppb FM 3422 based on theK.. valueof 6.6x Job . For theK,. and S,,valuestobe compatiblerequirethateitherFM 3422 isa liquidatroom temperatureor have a low meltingpoint(< 1000C). Ifthem.p. is > 100*Cthentheactualsolubilitsyhouldbe lower(- lox)thanthepredictedsupercooledliquid solubilitiy.,e.a,bout5 ppb insteadof50 ppb. Ifthem.p. of MY 3422 is> 100 to 15(rC then eitherthemeasuredK,,.or S isprobablywrong. The relationshirpeferredto in thereportsis by thesame group (Chiouetal.)and I thinkthisisa usefulexercise.The lineplottedinFig. 2 looksa littldeifferentthan the equationabove. The main cautionarynote is thatthese relationshipprsedictthesupercooledliquidsolubilitoyf solids.Ifthe solidhas a high m.p. (> 100 to150*C)theactualwatersolubilimtayy be substantial(l-y lox)lower. When working withtheseequations,you need toplotthesupercooledliquidsolubilitnyotthemeasured aqueous solubilitoyf the solid.At any rate,itisa good estimateto giveyou an ideaof where the experimentavlalue shouldfall.Italsoemphasizesthe importanceand utilitoyf obtaining accurateS,.a,nd K..,values. Similarestimatesofbioconcentratiofnactors(BCF) can.be obtainedfrom empiricalrelationships betweenBCF and eitherK., or S,,,with each species(e.g.,trout,guppie)yieldingitsown relationshipA. unifiedrelationshicpan be obtainedby normalizingtheBCF of a speciesby its lipidcontent,analogoustonormalizingthesoilsorptioncoefficienKt, tothesoilorganiccarbon contenttogiveK.. Tle normalizedfactorB,CFI, islinearwith Sw and K,. on a logscalefor differenstpeciesof aquaticorganisms(e.g.z,ooplankton,guppies,rainbowtrout,catfish)T.he equationis: log BCF, 0.899logK,,.+ 0.623 (from Chiou, 1985,Environ.Sci.Technol.19:57-62).This equationhas the advantaeg of predictinBgCF valuesfordifferenstpecies,as longas you know theirlipidcontent. As mentionedearliert,he soilTLC platesare of limitedvalue. They didn'tdistinguisthhe mobilitieosf FC 95 and FC 143 versusFM 3422 despitethefactthattheformer two compounds have much lower soilsorptioncoefficienthsan the latter.Hence, theassay seems insensitive. Experimental The watersolubilithyereisreportedas 5 ppm or 100 timesthe0.05 ppm valuediscussedabove. This has to be determinedaccurately.Micelleformationmay occurand resultin thishigh apparentsolubilittyh;ecompound likelhyassurfactanqtualitiejsudgingfrom itsstructureY.ou shouldknow ifit forms micelles.This is easilydone by measuringthe criticamlicelle concentratio(nCMC). There area varietyof methodsforthisdeterminationS.orptionof this compound by soilmay be quitedifferenatbove and below the CMC. Water solubilidteyterminationsT.he differencoef43 timesinthewatersolubilitioefsFM 3925 are FM 3422 seems ratherlargeconsiderintghehigh degreeof structuraslimilarity. Recommendation 1. Obtainan accuratewatersolubilitfyorFM 3422 and 3925*. 2. Obtainan accurateoctanol-wateprartitiocnoefficienftorFM 3422 and 3925*. 3. Determineif FM 3422 and 3925 form micellesi,.e.,measure the criticamlicelle concentrations. *A key referencedescribingmethodology fordeterminingwater solubilitaynd octanol-water partitiocnoefficienits: C.T. Chiou and D. W. Schmedding. 1980. Measurement and interpretatioofnoctanol-water partitiocnoefficienatnd water solubiliotfyorganicchemicals.A'ssociatioonfOfficiaAlnalytical Chemists. Washington,D.C.