Document evJq9kpn95jvGgbEZ3km2Dk39

.;ij Johns-Manvie EXHIBIT 1 T' R. S. Lamar - R&D Ctr From V. E. Wolkodoff - R&D Ctr Date: August 28, 1973 Copies: Subject. F IL L E R MIGRATION FROM FOOD WRAP PAPERS On Tuesday, August 28, 1973, I spoke to Mr. G. Higgenbottom of the FDA in Washington, D.C. He and Mr. A. Spiher had discussed possible methods which could be used to determine the amount of filler migration from a food wrap paper to the foodstuff contained therein. It is necessary to devise a suitable method, since there is no standard method in existence. The lack of an approved method was determined by Mr. A. Holtz, Packaging Chemist for the FDA. The method they suggest is as follows: v V 1. Construct a small container from the paper in question and partially fill with ordinary salt (NaCl). We agreed that reagent grade salt would be used since it would not contain the insoluble additives normally used in table salt. 2. Shake or tumble the salt-filled container for 24 hours. The intensity of shaking was not specified, but they will accept anything we feel is reasonable. 3. Remove the salt from the container and dissolve in a suitable quantity of water. 4. Filter the salt solution through a Millipore membrane (0.8 urn pore diameter), and examine for "asbestos" fibers by the normal (NIOSH) methods. In addition to the above, the FDA would also like to have a determination of the limit of detection. This would be d e termined by adding known increments of "asbestos" or "asbestos"containing talc to salt and following the same analytical procedure. V. E. Wolkodoff < V APPENDIX I I I Figure 1 E 4 1 4 -2 JOHNS-MAMILLE RESEARCH AND ENGINEERING CENTER ReportNo. E 4 1 4 -2 Page 14 SCUFF TEST (Cont'd) 6. Empty the salt from the test jars into 900 ml (0.45 micron filtered) of water which is being stirred by a magnetic stirrer. Remove the test paper lining and tap it over the beaker in order to remove any salt clinging to it. Then use an additional 100 ml of filtered water to rinse the jar and cap. Pour this into the beaker also. Allow the mixture to stir approximately 5 minutes in order to dissolve all the salt. 7. Preweigh two 0.8 micron millipore filters for each sample. When the salt is dissolved, pour the liquid through the filter apparatus in which both filters are present, one directly on top of the other, (the bottom filter is a control). Use an additional 200 ml of filtered water to rinse out the beaker and pour this through the filter. (A wash bottle provides the most effective method of rinsing.) Then use an additional 300 ml of filtered water to wash the filters. This will help wash away any minute salt particles that are clinging to the filters. When washing is completed, let both filters dry completely. 8. When the filters are dry, weigh them. These samples are now available for microscopic examination. D. Calculations 1. (For 1 Sample) Filter No. 1 (Before) = X gms Filter No. 1 (After) = X + (residue) gms (Residue) gins Filter No. 1A (Before) = Y gms Filter No. 1A (After) = Y Z gms (Z) gms Filter No. 1A acts as a control. It might change in weight by the amount of (Z) gms. This might very well be an increase in weight if all the salt was not washed out of the filters. But whatever weight is found in No. 1A, it must be subtracted from the apparent residue weight, because approximately the same change in weight will occur to the top filter as did to the control filter. 2. (Residue) gms - ( Z) gms = Actual Residue (gms) JOHHS-MAHVILLE RESEARCH AND ENGINEERING CENTER E ReportNo. E 4 1 4 -2 Page 13 APPENDIX II SCUFF TEST A. Equipment 1. L.A.B. type RVH vibration test stand or equivalent. 2. Millipore filtration apparatus, 0.45 and 0.80 micron filters (vacuum filtration). 3. Fisher Magnetic Stirrer. 4. 8 ounce flint glass, squat form wide mouth bottles. 5. Several 1500 ml beakers (carefully cleaned). 6. Mettler Balance or equivalent.. B. Reagents 1. Distilled water 2. Sodium chloride, CP 3. Test paper C. Test Method 1. Cut selected test paper to the dimensions of 2 3/16-inch x 8 1/2-inch. Curl the paper and place it inside a clean, dirt and fiber free 1/2 pint jar. Line the paper on the inside surface (there will be a slight overlap); mark this overlap with a few pencil dots; remove the paper and tape the outside of the overlap; then re-insert inside the jar. 2. Weigh out 200 gm of the salt and place it in the lined jar. Cap the jar. 3. Place on vibrator which has been set to run at 19 cps, 1/16-inch amplitude, (1/8-inch full excursion), for 24 hours. 4. After 24 hours of running, remove jar(s) from vibrator. Any run should also be accompanied by a control blank of a jar simply filled with 200 gms of salt. 5. Filter some distilled water through 0.45 micron millipore filters. Approximately 1500 ml will be needed for each sample. Also, use this filtered water to rinse out all containers and equipment with which the sample liquid will come in contact. It is important to keep all equipment dust and fiber free. JOHNS-MANVILLE RESEARCH AND ENGINEERING CENTER E Report No. E 4 1 4 -2 Page 12 APPENDIX I Table IVB CELLULOSE AND TALC ON MILLIPORE FROM 19 CPS FREQUENCY SERIES RS 73242-2 BROWN PAPER Time 0 - Salt Blank 2 hours 4 hours 6 hours 12 hours 24 hours 48 hours 72 hours Estimated No. of Cellulose Fibers Per Membrane 400 600 1,100 2,500 >3,000-*>3,000-*>3,000-*>3,000-*- Estimated No. of Talc Particles Per Membrane 70012 2,0Q02 5,0002 12,0002 17,0003 38,0003 68,0003 7 5 ,0003 1. The cellulose fiber population was tremendous on these membranes. Their inter-twining and matting made counting extremely difficult. By extrapolating, their count could easily reach as high as 20,000 per membrane. V. E. Wolkodoff 2. Based on ash 3. Based on dissolved millipore membrane JOHNS-MANVILIE RESEARCH AND ENGINEERING CENTER Report No. E 4 1 4 -2 Page 11 APPENDIX I Table IVA SCUFF RATE OF SAMPLE RS 73242-2 (Determined Over a Cycle of 72 Hours) Total Average Insoluble Residue Due Time_______ Filter Residue______ Salt Residue**_______ to Paper Abrasion 2 hours 0.00236 gms * 0.00216 gms 0.00020 gms 4 hours 0.00242 gms * 0.00216 gms 0.00026 gms 6 hours 0.00445 gms * 0.00216 gms 0.00229 gms 12 hours 0.00266 gms 0.00216 gms 0.0005 gms 24 hours 0.00300 gms 0.00216 gms 0.00084 gms 48 hours 0.00359 gms 0.00216 gms 0.00143 gms 72 hours 0.00364 gms 0.00216 gms 0.00148 gms Average residue ** Average from 7 salt blanks HJOHNS-MANVILLE RESEARCH Report No. E 4 1 4 - 2 AND ENGINEERING CENTER Page 10 APPENDIX I Table IIIB APPROXIMATE NUMBER OF TALC PARTICLES PER 35 MM DIAMETER RESIDUE ON MILLIPORE MEMBRANE BY PETROGRAPHY AT 50OX - RESIDUES FROM 19 CPS - 24 HOUR TEST Sample Actual Talc Content Number of Number of Number of of Paper Talc Particles Chrysotile Fibers Tremolite Fibers RS 73242-1 0.35% 29,000 2 0 668 KS 73242-2 RS 73242-3 1.0 % ** 52,500 2 2,500 0 0 2,002 0 RS 73242-4 RS 73242-5 0.05% ** 7,500 3 Present 0 0 668 0 AS 73242-6 0.60% 35,000 0 2,804 RS 73242-7 RS 73242-8 RS 73248-6 1 J!S 73248-7 1 0.25% ** ** ** 50,000 10,000 Too waxy to read Too waxy to read 0 0 - 2,937 0 - Predominant constituent appears to be minute blebs of crystalline wax ranging from 0.5 to 5 microns in size. Very small percentage of carbonate and cellulose-type fiber. The average size of these talc platelets is approximately 1 micron x 1 micron x 1/2 micron. Using this average particle size and a density of 2.8 gms/cm^ for talc, 29,000 particles would weigh approximately 2 x 10~*mg, and 52,500 particles would weigh approximately 3.7 x 10"^mg. Profuse amount of doubly terminated minute acicular crystals (2 to 10 microns in length) which are anisotropic, but unidentified. These acicular crystals are not retained after ashing. Two blanks were examined to try to establish background level of talc. CELITE Mo. 73248-7 (salt-washed water) was studied in some detail by employing 20 fields of view at 50OX. Minute particles of carbonate-like material predominate and may constitute as much as one million particles per 35 mm diameter residue. Certain quantities of talc platelets were detected on these blanks for no discernible reason, except possibly due to contamination. The calculated results for the entire residue are: No. of talc particles No. of chrysotile fibers 8,0000 (platelets, few shreds) No. of tremolite fibers 0 Cellulose-like fibers trace JOHNS-MANVILLE RESEARCH AND ENGINEERING CENTER ReportNo. E 4 1 4 -2 Page 9 APPENDIX I Table IIIA QUANTITATIVE RESULTS OF SCUFF TEST RUN AT 19 CPS - 24 HOURS Sample RS 73242-1 RS 73242 2 RS 73242-3 RS 73242-4 RS 73242-5 RS 73242-6 RS 73242-7 RS 73242-8 RS 73248-6 RS 73242-7 Residue Due To Scuffed Paper (gms) 0.00151 0.00444 0.00148 0.00356 0.00133 0.00353 0.00508 0.00293 0.00685 0.00798 Loss per sq in Loss per sq cm of Paper of Paper (gms/sq in)_________ (gms/sq cm) 0.83 X 10"4 1.28 X 10"5 2.44 X 10-4 3.78 X 10"5 0.81 X 10"4 1.26 X 10-5 1.95 X 10"4 3.03 X 10"5 0.73 X 10"4 1.13 X 10"5 1.94 X 10"4 3.01 X 10-5 2.79 X 10"4 4.32 X 10"5 1.61 X 10-4 2.48 X 10"5 3.76 X 10-4 5.82 X 10-5 4.38 X 10"4 6.79 X 10~5 JOHNS-MANVILLE RESEARCH AND ENGINEERING CENTER ReportNo. E 4 1 4 -2 Page 8 APPENDIX I Table II TALC IDENTIFICATION OF PAPER SAMPLES BY X-RAY DETERMINATION* Sample Source Reported Talc Content % Talc Determined By X-Ray Diffraction RS 73242-1 0.67 percent 0.35% RS 73242-2 RS 73242-3 1.02 percent None 1.0 % ** RS 73242-4 RS 73242-5 Unknown Unknown 0.05% ** RS 73242-6 1-1.5 percent 0.60% RS 73242-7 RS 73242-8 RS 73248-6 RS 73248-7 1-1.5 percent None Unknown Unknown 0.25% ** ** ** * Paper was ashed for X-ray determination. ** Below detectable limits JOHNS-MANVILLE RESEARCH AND ENGINEERING CENTER H ReportNo. E414-2 Page 7 Identification No. RS 73242-1 RS 73242-2 RS 73242-3 RS 73242-4 RS 73242-5 RS 73242-6 RS 73242-7 RS 73242-8 RS 73248-6 RS 73248-7 APPENDIX I Table I DESCRIPTION OF PAPER SAMPLES TESTED Source Description Longview Paper Company Brown kraft paper reportedly con taining 6.7-pound CYCLO-SORB per 1000-pound pulp. Longview Paper Company Ditto except containing 10.2-pound CYCLO-SORB per 1000-pound pulp. Longview Paper Company Ditto except no talc. Brown Company 40-pound MG bleached bag stock coated with 5-pound clay latex, used for cookie bags. Talc content not stated. Brown Company 30-pound white DOK, wide food wrap use. Talc content not stated. Brown Company 25-pound Royal white waxing (this sample not waxed). Wide food contact uses (cookies, sugar, etc.). Talc content not stated. Brown Company 36 1/2-pound Spec, white locker waxed one side. ("Wax" not apparent). Used for freezer locker paper. Talc content not stated. Longview Paper Company 30-pound No. 801 wrapping paper. No talc. Brown Company 31-pound Royal white food wrap paper (waxed). Talc content not stated. Brown Company Printed and waxed paper for wrapping butter. Talc content not stated. ' JOHNS-MANVILLE R E S E K H AND ENGINEERING CENTER Repo^^o. E414-2 Page 6 2. It is essential to perform additional testing to determine if during commercial shipment of food and drugs, the vibration and movement encountered is sufficient to result in the partial breakdown of the paper packaging. This is necessary as the tests described herein indicate that there is no migration of tremolite or any of the ingredients of the paper unless there is a partial breakdown of the integrity of the paper itself. This test has not established whether paper breakdown actually occurs under commercial conditions. JOHNS-MANTILLE R E S E C O AND ENGINEERING CENTER E Re E 4 1 4 -2 Page 5 CONCLUSIONS Inspection of the results obtained from the Scuff Tests lead to the following conclusions: 1. At a frequency of 10 cps or a frequency equivalent to that of a freight train having a flat spot on its wheels traveling almost 60 mph, no evidence of paper breakdown was detectable when the samples were run for 24 hours, and no migration at all was detected. 2. As reported before, visible movement of the salt relative to the movement of the jar was not in evidence until approximately 16 cps. 3. Paper breakdown due to the abrasive action of the salt is detectable when run at a frequency of 19 cps and 1/8-inch full excursion for extended periods of time. But let it be emphasized that this frequency was settled upon only to ensure visible salt movement within the container, and at the time this frequency was selected, there was no reason to believe that such a high frequency rate was necessary to duplicate the vibration and move ment that actually occurs during the commercial shipment of food or drugs. It is, therefore, very possible that the frequency rate of 19 cps resulted in test conditions far more stringent than those which occur in actuality during commercial shipment. 4. In all cases, the material scuffed off and recovered as residue was due to the partial destruction of the paper lining itself, and is not due to the selective migration of individual materials such as cellulose, talc, tremolite, etc., out of the paper. The examined residues were identified to contain all the materials of the original paper sample in the same proportions, as best as can be determined optically. 5. Although only limited testing has thus far been performed, it is apparent that attempts to quantify test results at these exceedingly low levels of concentration is difficult. It is equally difficult to generates reproduceable results. RECOMMENDATIONS It is recommended that: 1. A determination must be made as to whether the test conditions described herein do, in fact, duplicate conditions during actual commercial shipment of food and drugs. JOHNS-MAHVILLE R E S E jJA AND ENGINEERING CENTER Reoor Page E 4 14-2 4 However, one must recognize that it is possible that the salt and glass jar were moving at different frequencies which were not apparent to the eye, and it is conceivable that the frequency rate of 10 cps may more nearly duplicate vibration and possible scuffing during commercial shipment. If this is so, based on the data from the 10 cps test, no degradation would be anticipated under actual shipping conditions. ' A fourth test was then initiated to determine at what rate the scuffing occurred at the higher frequency of 19 cps. Again, sample RS 73242-2 was selected for observation because of its performance ,in Test 2. Samples were removed from the vibriitor at 2, 4, 6, 12 , 24 , 4 3 and 72 ..ours. Duplicate samples were run for the 2, 4, and 5-hour tests in order to serve as a check for the minute amount of residue expected. It must be emphasized that these results only reflect the scuffing characteristics of this particular paper sample and may not be representative of the other papers. This sample demonstrated that the rate at which the integrity of the paper is broken down is linearly related to time, up to 48 hours. The quantitative results are located in Appendix I, Table IVA and a plot of the paper residue weight versus time is represented in Appendix III, Figure 1. Microscopic and petrographic examination of the residues revealed that the residue contained all of the components that the original paper contained and so far as can be determined optically, in the same proportions. This leads to the conclusion that the material scuffed off and recovered as residue was due to the partial destruc tion of the integrity of the paper lining itself, and cannot be attributed simply to the migration of any select materials such as talc, cellulose, fillers, etc., out of the paper. This conclusion is supported by the fact that approximately the same amount of material was recovered from scuffed papers which did not contain tremolitic talc as those which did. Comments on Test Procedure Although only limited testing has thus far been performed, it is apparent that attempts to quantify test results at these exceedingly low levels of concentration is difficult. It is equally difficult to generate reproduceable results. A case in point involves the use of technical grade salt which has a maximum insoluble impurities limit of 0.005 percent. In using 200 grams of salt, it is possible to have between 0-10 milligrams of unwanted impurities contaminate the residue deposit from each test. When it is considered that only a few milligrams of abraded residue is recovered at the conditons the test was run, it is obvious that a deviation or error of well over 100 percent is possible. But let it be emphasized that although a weight difference was found in some samples run at identical conditions, this had no effect on the relative proportions of materials identified microscopically. JOHNS-MANVILIE RES AND ENGINEERING CEM '<> E414-2 d Page 3 Therefore, the vibrator was set at 19 cps (resulting in an acceleration of 2.5 "G's" at the amplitude peaks). The contents of the jar were -then removed and the entire contents dissolved in prefiltered water and the liquid was filtered through a 0.8 micron millipore filter. The residue left on the filter was then available for weighr determination and microscopic examination. A complete description of the Scuff Test procedure is described in Appendix II. Test Program The initial Scuff Test run consisted of the evaluation of paper samples PS 73242-1 through RS 73242-8. These samples were run at a frequency of 19 cps for 24 hours. The 24-hour time period was specified by FDA. (See Exhibit 1) In addition, a control blank consisting of an unlined jar containing salt was tested along with the paper samples. The results of this initial test confirmed the occurrence of paper scuffing at these conditions. Minute amounts of scuffed cellulose fibers were visibly detectable in each of the recovered residues. These residues were then examined microscopically. Since it was then determined that the initial test resulted in the partial breakdown of the integrity of the paper, the entire test was rerun for the purpose of determining the exact weight of each of the residue deposits. By knowing the percentage of talc contained in the tested papers, this would allow the calculation of the theoretical weight cf talc contained in the residue and thus the determination of whether the talc found in the residue was migrating in a greater proportion than was contained in the original paper. The purpose of this procedure was to determine whether or not there is any selective migration of talc and tremolite. Quantitative results of this test are reported in Appendix I, Table IIIA and IIIB. It was then decided to select one of the paper samples which produced the largest cellulose fiber deposit from Test 2 (4.4 milligrams) and run it at a lower frequency to see what effect this change would produce. Sample RS 73242-2 was selected and it was run at 10 cps for 24 hours. As mentioned before, no visible shaking of the salt was in evidence at these conditions. The results of this test confirmed that when the salt contents were not visibly "shaking" because of lower frequency conditions, there was no breakdown of the integrity of the paper and no migration. In duplicate test samples, the identical results were obtained. JOHNS-MANVILLE RESEARCH AND ENGINEERING CENTER Report Page E 41 4 -2 2 Confirmation of Talc Content Since migration of tremolite from food packaging paper to food products in contact with such paper is the thrust of this study, positive identification of the presence or absence of tremolitic talc is important. Talc per se can be positively identified by both its characteristic X-ray diffraction (XRD) patterns and petrographic microscope examination. However, attempts to positively identify the talc in any of the paper samples as received proved unsuccessful by either of these methods; and, it was not until the paper samples were low-temperature (400C) ashed that talc was detected and identified by these methods. Both the Longview Paper Company and Brown Company have advised us that their sole source of supply for talc is Johns-Manville. Both companys' purchases of talc are restricted to J-M "CYCLO-SORB", a grade of talc containing approximately five (5%) percent tremolite by weight. The actual talc concentrations of the paper samples, as determined by X-ray diffraction, are located in Table II. Scuff Test The philosophy behind the "Scuff Test" was to attempt to simulate actual shipping and handling conditions of food-containing packages. It is assumed that during shipping, particularly rail shipping, the most severe and prolonged shaking or vibrating of the packages will be incurred. During this time, it is possible that the vibra ting contents could have an abrasive effect on their containers, thus possibly causing some of the package interior to be scuffed off and mixed with the contents. Therefore, a test that could simulate these conditions would permit examination of the material scuffed off and determination of what, if any, materials (i.e., cellulose, talc, tremolite, etc.) have been dislodged from the paper fiber. J-M was directed by FDA to use salt as the medium for the "Scuff Test" since its sharp edged cubic crystals would act as a more abrasive material than any other dry food substance or drug. (See Exhibit 1) Paper containers were simulated by lining the inside of 8 ounce squat glass jars with test paper. Vibration settings on a shake tester were set at its maximum amplitude of 1/16-inch (1/8-inch full excursion) and minimum frequency of 8 cycles per second. It was soon found though, that these conditions did not produce any visible vibration of the salt. The contents simply rode up and down with the jars. It was not until the frequency was raised to approximately 15 cps that visible movement was in evidence. This fact aroused some question as to how much, if any, prolonged shaking actually does occur to packaged foods during transport. It was calculated that a freightcar, having a flat spot on one pair of wheels, traveling at 40 mph would only vibrate with a frequency in the range of 6.8-7.5 cps; and at 60 mph, 10.2-11.2 cps, depending on wheel diameter. However, it was decided to run the initial test at severe enough conditions to cause detectable movement of the salt so that microscopic analysis of the salt would be meaningfuL. JOHNS-MANVILLE RESEARCH AND ENGINEERING CENTER Repm*o. E414-2 17i Page 1 Introduction DISCUSSION FDA on September 28, 1973, published a Notice of Proposed Rule Making which would restrict the use of talc in food, drugs, and food and drug packaging materials. The proposed rule would pre clude all such uses for talc, if the talc contained more than specified numbers of chrysotile and tremolite fibers. Talc is widely used as a pitch control agent in the initial manu facture of paper pulp from wood, and also in the actual manufacturing of paper. Talc so used, becomes an integral part of finished paper produced from this pulp. The finished paper, which may also contain various other special purpose fillers and pigments such as clay, whiting, and Ti02 , etc., is used for the full spectrum of paper applications, including food and drug packaging. This study was undertaken to determine the extent to which the tremolite fibers contained in talc used in the manufacture of paper for food and drug packaging applications might be expected to "migrate" into food or drugs to which it would come in contact. Tremolite Migration (or Scuffing) Studies Basis Migration as used here is appropriate only in the sense of overall determination of loss of paper fiber, talc and other fillers tightly held in a homogeneous sheet of paper. None of the types of paper commonly used for food contact applications has loose surface fibers or powdery material which will come loose on simple contact, hence the term "migration" as used in that sense does not apply. Mechanical action or "scuffing" of some sort is required to dislodge talc or other filler materials, and incidentally, cellulose paper fibers, from finished paper. With direction from FDA, a test method was devised which does scuff or abrade material from a paper sheet simulating the action a cylindrical table salt carton might receive during commercial shipment for an extended period of time. For simplicity, this is referred to as the Scuff Test in this report. A description of the directions from FDA for a test method is set forth in Exhibit I attached hereto. Paper Samples Ten samples of paper made for various food packaging applications were obtained from trade sources. Four of these were brown kraft, two with tremolitic talc and two without; four were white locker and counter papers, again two with tremolitic talc and two without; and two waxed papers purported to have talc. A more complete listing is given in Appendix I, Tables I and II. JOHNS-MANVILLE RESEARCH AMD ENGINEERING CENTER EXHIBIT I Report No. E414-2 Date December 12, 1973 Title: PRELIMINARY STUDIES OF FOOD CONTAMINATION BY CONTACT WITH PAPER SUMMARY A study was undertaken to determine the extent to which the tremolite fibers contained in talc used in the manufacture of paper for food and drug packaging applications might be expected to "migrate" into food or drugs with which it would come in contact. For these studies, a test method and procedure were worked out which established that the migration of tremolite contained in talc used in food packaging paper occurs only after partial destruction of the paper after being in contact with food. Without some destruction of the integrity of the paper, no migration of tremolite occurs. . When partial destruction of the integrity of the paper occurred, the resulting migration had the composition of the paper insofar as could be measured. Partial destruction of the integrity of the paper and migration of the component elements of the paper occurred both with talc-containing and non-talc containing papers. In those paper samples containing tremolitic talc, the residue in the food also contained a similar proportion of talc and tremolite fiber. Since talc usually constitutes one percent or less of the total paper, talc migration is so low as to be barely measurable and tremolite fiber levels could be three orders of magnitude less. The test conditions used here were arrived at by discussions with the Food and Drug Administration. Additional testing must be undertaken to determine whether or not the test conditions used herein actually, in fact, duplicate the conditions and what takes place when food is shipped commercially, so as to determine whether in commercial shipping the partial destruction of the integrity of food packaging paper occurs. Contents: Summary, Discussion, Conclusions, Recommendations, and Appendix I, II, and III. Reported by :