Document x1209k40EQmQOXvZRY95mJODE

P-1 fH2^(o-3 m Du Pont de Nemours (Nederland) BV Dordrecht W orte To: C L. Guris - Geneva FromiR.A. Brandenburg - Dordrecht cc. T.J B. Schroots F. Marcoz M v d Nood C. de Heer H. Benjamins Dordrecht, 30 January 1997 RESU LTS OF DU PO NTS ANALYSES O F C-8 IN HOSTAFLON TUBES R e f: My summary of 6-dec-96 meeting in Geneva about C-8 in PTFE for food contact applications Dear Chris. Here are the results of our C -8 analyses on Hoslaflon lubes The conclusion is that if we agree extraction depth is less than 3 mm, only minor toxicity testing is required. I am looking lorward to the results ot Hoechst and ICI. Samples provided by Dr. Mitterberger of Dyneon: A 1 kg of Hoslaflon TF 2025 tubes 12 mm diam, 1 mm thickness B 1 kg of same tubes, shredded by Dyneon, ca 5X5X1 mm C samples A, cut in rings (1-2 mm) and further shredded by DuPont laboratory, ca 1 mm3. Extractions EID751385 1. In chloroforrrVwater B2-1 50 grams ol B refluxed lor 2 hrs in a mixlure of 40 ml CHCI3 and 25 ml H 20 C 2 -1 50 grams of C refluxed for 2 hrs in a mixlure of 40 ml CH CI3 and 25 ml H20 2. In ethanol 99.8 % B2-2 2 0 0 grams of B refluxed for 2 hrs in 250 ml elhanol, pH > 10 (by adding NaOH) C 2 -2 192 grams of C refluxed for 2 hrs in 250 ml elhanol, pH > 10 B2-8 as B2-2, refluxed for 8 hrs in 250 ml elhanol, pH > 10 C2-B as C2 2, refluxed for 8 hrs in 250 ml ethanol. pH > 10 P-2 EID751386 Analyses 1. Methylene Blue method short description To the 25 ml water layer was added: 5 ml methylene blue -1 ml 2 N H 2 S 0 4 This was extracted 3 times in 15 ml CHCI3. The 45 ml extract was completed to 50 ml with CHCI3, and extinction at 640 nm was measured on a UV-Vis spectrophotometer, and compared with a calibration tine. 2. G C method short description (Du Pont ret. W W -3690) The 250 ml ethanol was evaporated on a steam plate. This took appr. 48 hours. The residue was esterified with 10 ml. methanol containing 1.5 mol H 2 S 0 4 per titer methanol at appr. 54 C during 2 hours. To this mixture was added 10 ml hexane and 20 ml of a stock solution ot 180 g/l NaCI solution in water. After shaking and phase separation 5 microliler ol the hexane layer was injected on a G C capillary column with ECO detector. (50 m. 0.53 mm C P S i 5CB fused silica). Peak area was calculated into concentration by comparison to the peak area of 1 ppm of an esterified standard solulion of C-8. W e used a 1 ppm C 9 solution as an internal standard to correct lor variations in injection volume. Results B2-1 C2-1 120 ppm wt 410 ppm wt CHCI3 Dyneon shred, Me-blue CHCI3 DuPont shred, Me-Wue B2-2 C2 2 B2-8 C2-B .068 ppm wt .097 ppm wt .065 ppm wt 106 ppm wt EthOH Dyneon shred, GC EthOH DuPont shred, GC EthOH Dyneon shred, G C EthOH DuPont shred, GC P-4 Discussion 1. As w e discussed before, Ihe higher numbers found with the methylene blue method are less reliable, due to: - extinction values 0 033 and 0.066 w ere too low lor reliable result - there may have been other anionic surfactants in the laboratory glassware, which significantly contribute to the blue color in this low concentration range. 2. W e found little difference between 2 hours and 8 hours extraction. This suggests that 2 hours of extraction should be sufficient 3. The finer shred made af the DuPont laboratory gave results that are roughly 50 % higher than Dyneon shred. EID751388 4. In the food extraction tests, 6 dm2 of PTFE surface is exposed to 1 kg of the extraction medium. If we assume the effective extraction depth is d mm, this corresponds with 0.06*d dm3 = 0 1296 * d kg of PTFE exposed to I kg of the extraction medium. W e hope to find less than 50 microgram of C -8 per kg of extraction medium, so that only minor toxidly testing would be required. So the 0 .1 296 * d kg of PTFE should contain less than 50 microgram of C -8. In other words the C-8 concentration in the PTFE should be less than 386/d microgram per kg. If w e take d = 1 mm, the concentration should not exceed 386 ppb. W e find appr. 100 ppb. As you can see. d is rather important. If w e decide that effective extraction depth is 4 mm, the concentrafion should not exceed 96 ppb. compared to actual 100 ppb. Estimation of extraction depth There is som e indication that extraction depth is less than 1 mm. The Dyneon shredded tube pieces were fiat pieces of 1 m m thickness (= tube wall thickness), and appr. 5 x 5 mm size. Apparently even Ihough the smallest size is 1 mm, and both fiat sides of Ihe flat pieces are exposed to the extraction medium, this is not enough to lake out all C 8. as evidenced by the tact that cutting into smaller pieces gave 50 % higher numbers. Although what follows is speculative, here is a calculation of Ihe effect of cutting: Suppose all large pieces are exactly 5 x 5 x 1 mm, and the small ones are 1 x 1 x 1 mm. Then each large piece can be cut into 25 smalt ones. Each small piece has a surface area of 6 mm2, so 25 of them h ave a surface area of 150 mm2. This should be compared with the surface area of the large piece from which they were cut: 2 x 25 mm2 + 4 x 5 mm 2 = 70 mm2. So culling inlo small pieces raises the surface area by a factor 2, and the analytical result by a factor 1 5 If the extraction depth is d mm, you can show that lor 25 1 mm3 cubes, the extracted volume = 25'(6d - 12dA2 -*8dA3} mm3 = 150 d 3 0 0 dA2 + 200 dA3 For a 1 x 5 x 5 piece the extracted volume is 70d - 44 d*2 + 8dA3. Let us assume that the factor 1 5 increase in C-8 result is fully caused by the increased extracted volume. Than w e can say that: 150 d - 300 dA2 + 200 dA3 = 1.5 x 170 d - 44dA2 + 8dA3J. d can be solved by algebra, and w e find d = 0 2387 mm. A l this extraction depth the increase in extracted volume of 1 mm 3 pieces is exactly 1.5 limes as high as the extracted volume of the sam e weight of 5 x 5 x 1 mm3 pieces. P-6 EID751390