Document evj8v6XNkOe6N6NvEBnama3zM

AR226-3333 FLUOROPOLYMERS 81-4 PAGE 10 3. Measurement of Residual C-8 Fluorosurfactant Levels on Various Products - C. S. COPE Following the report from 3M Company that "Fluorad" FC-143 (C-8 APFC) surfactant had been found teratogenic In tests with laboratory animals, we received several company requests for analytical determination of residual C-8 concentrations in various{f|||Xi--------yproducfcs. Included were two types ofBBBBBI^Blflfeltprgguced by F&F'Department's Fairfield, Conn.. plant. offlHimfglass a SBmple cloth manufactured byHfB||------J(sub- Tnitsed by S. J<. Cs-yaaaugh of Research), and two sgmplegg^ TE-9724-JgfirPFA baked powder (prior to extrusion) manufactured a (analysis requested i N l ^ f by P. Thistleton of the Technical Department). r <a As a reference point concerning residual levels of C-8 APFC on our [jU^products, it may be noted that the Customary Advisory Letter issued to our customers on 4-1-81 by the Sales Division stated: "With the exception of aqueous dispersions, there is no significant residual FC-143 in any of the fluoropolymer resins which we sell." This statement is not Inconsistent with the fact that very low levels, of the order of 1-5 ppm, of residual C-8 APFC J^ave been detected (using a method which is not specific for C-8) in the past on miB^ffine powders and granular resins, the main products we sell in dry, bulk Torm in the unextruded condition. (Extruded resins should contain even lower levels of residual C-8.) Analyses of all these samples were performed by boiling the samples in a dilute aqueous solution of phosphoric acid in the presence of a nonvolatile, nonionic detergent ("Triton" X-100). Perfluorocaprylic acid is sufficiently volatile at lOO'C that it steam-distills off under these conditions. The C-8 in the distillate is then analyzed by the chloroform/Azure A (a-blue dye) colori- netric method, sensitive to about 2-3 micrograms. The method is nonspecific for C-8, and other ionic swfactants can, in principle, interfere. In no case among the re .es was a residual C-8 concentration above the range typically found o; felt samples, values of 3.5 ppm C felt XT-7800) were obtained. Fo; Cequivalent to 3.95 ppm based o: fine powder detected. For the F&F I--f--el--t --XT-M72H.1Q10f)abanrdic,0.7weppfomun(dble0a.c3h5edppm solids present). For the TE-9724-J samples, concentrations of 2.8 ppm were found. Results have been trans mitted in writing to those who requested the analyses. The most recent analyses of residual S-8 APFC on fine powder were obtained in March on a set of three samples oqj--------u64 submitted by E. H. Sfcueber of the .Technical Department. Two of these samples were from lots which had performed poorly in wire-coating atMI------------^B--------w|hile the third was a control which performed normally, ^['echnical's request for C-8 analyses was prompted by a desire to test, in the absence of any clear-cut distinctions in the normal characterization data, whether possible differences in residual C-8 level might account for the differences in fabrication behavior. The two unsatisfactory lots, with residual C-8 levels of 4.2 and 2.5 ppm, were found to be somewhat above the control (1.5 ppm) in this respect, but by no means enough so to be considered unusual. Furthermore, tests done years ago showed that Company Sanitized. Does not contain TSCA CB1 TLTJOROPOLYMERS 81-4 PAGE 11 the paste extrusion performance of fine powder was not significantly affected either by removing the last traces of residual dispersing agent (by extraction with a solvent) or hy depositing dispersing agent back onto the dry polymer (by deposition from solution, followed hy evaporation of the solvent) to the (rela tively high) level present on the polymer before drying. We were, therefore, unable to encourage Technical that the differences in wire-coating performance were likely to stem from this source. Samples of the t h r e e -- -- -- --. --^>|6 4 fine powder lots have been submitted to ESL for confirmatory analyses using the GC method developed there which Is specific for C-8. Company Sanitized. Does not contain TSCA CBI FLUOROPOLYMERS 81-4 PAGE 12 4. C-8 Fluorosurfactant^ Analyses on Samples Received from the FEP Area - C. S. COPE Requests were received from Technical Department personnel In the FEP Area for C-8 fluorosurfactant analyses on several polymer and water samples. The polymer samples were taken from theMR^^----gDryer and asso ciated equipment at a time when dryer performance was being questioned. The water samples were taken as part of the program to establish a material balance on C-8 distribution in the FEP Area. In all cases, the C-8 was recovered from the samples by distillation In the presence of dilate aqueous phosphoric acid. Indications are that this treatment is relatively effective in separating out any "in situ" dispersing agent from the C-8 APFC. With the polymer samples, a nonvolatile, nonlonic detergent ("Triton" X-100) was added to promote wetting. Analyses of the distillates from the polymer samples happened to be carried out by titratlon with cetyltriaiethylainmonium bromide, whereas with the distillates from the aqueous samples (run at a later date), the chloroform/Azure A method was employed. Results of the analyses are presented in Table II-CSC. it The measured C-8 concentrations on the polymer sa few surprises, having been established earlier that the I serves to remove most of the C-8 from the (wet) feed enters Press. Since the C-8 removed from the polymer discharges through vent line leading to the water scrubber, a polymer plug in this line would be expected to contain more C-8 than the dry polymer exit the dryer. Dry polymer removed from the rotor of the dryer would be expected to have a C-8 content comparable with that of the dryer product. Two earlier measurements of C-8 level on the dryer product gave values of '6 and 34 ppm, bracketing the value of 17 now obtained on the polymer scraped from the dryer rotor. The concentration of 1950 ppm (0.195%) found on the sample taken ie lid of the screw conveyor now locatedAetween the discharge of the ^----QPress and the entrance to thej------------Dryerwas much higher than at __ first expected, based on previous measurements Press product samples, o i H f which had equivalent C-8 concentrations of 0.042X^Bid0.u66?. The differences are even greater when differences in water content are taken into account (the dispersing agent is expected to be predominantly adsorbed on the polymer, rather than in the aqueous phase). The recent sample contained 70% water (total basis), vs. 24.5Z and 14.8Z, respectively, in the earlier samples, making the relative C-8 concentrations, expressed on a water-free basis, different by about an order of magnitude. We postulate that this difference is attributable to back-leakage of C-8-containing vapor from the dryer during the period when dryer performance was unsatisfactory. The C-8 would be expected to have condensed out on cool surfaces, such as the screw conveyor lid, the top of which was exposed to the ambient atmosphere. IBwany Sanitized. Does not contain TSCA CB1 FLUOROPOLYMERS 81-4 PAGE 13 TABLE II-CSC C-8 FLUOROSUBFACIAHT ANALYSES ON SAMPLES FROM THE FEP ASEA (Submitted by A. R. Behnke and J. R. Evitts, Technical Department) Sample Identification Date Submitted Measured Concentration of Equivalent C-8 APFC. ppm (Based on Initial Weight) Polymer Samples (Analyzed by CTMAB Titratlon) Polymer scraped from rotor of |Dryer 3/30/81 Sample of.plug found in vent line 3/30/81 Dust sample' froa^id of Screw Conveyor f ceding M--------f Dryer (70Z H^O, otal basiSj 4/2/81 17 170 1950 Water Samples (Analyzed by CHCl,/Azure A Method) _JlOO Coagulator Effluent ^Batch 21-4111) 4/27/81 260 WHWy.OO PK (First) Rinse ffater XBatch 21-4111) 4/27/81 15 (0.12Z polymer solids present) Company Sanfthea. Does nol contain TSCA CBI