Document x5xRz7k5kvGN81eEy5yQGYMO0

Introduction Union Carbide Corporation has been a manufacturer of solution polymerized vinyl chloride copolymer resins for more than 45 years. The solution polymerization process, used exclu sively by Union Carbide, by its nature represents the most facile way, known technically, to remove residual vinyl chloride monomer. The manufacture, distribution and use of these products complies with all existing regulations and Union Carbide plans to demonstrate compliance with any future regulations. The solution polymerized copolymers are listed by chemical identity in Regulation 175.300 for resinous and polymeric coatings and may lawfully be used as food contact surfaces applied to metallic substrates under this regulation. On September 3, 1975, the FDA published a notice of two proposed rules relating to use of vinyl chloride polymers that come into contact with food. The first proposal called for amend ment of existing FDA regulations by adding vinyl chloride monomer to the list of substances prohibited from use in, or in contact with human food. The second proposal contained a paragraph that speci fically permits the continued use of vinyl chloride homopolymers and copolymers in coatings, gaskets, cap liners, flexible tubing and plasticized film. Neither of these proposals by the FDA were ever put into effect. The proposals did result in studies sub mitted to the FDA that demonstrate that rigid vinyl chloride homopolymers containing very low residual monomer levels can be ucc 036635 2. produced and that at these levels the monomer may not migrate to food. At this time. Union Carbide initiated an extensive testing program to demonstrate to ourselves, the appropriate government regulatory agencies and the industry that the solution polymeri zation process, used exclusively by Union Carbide, results in the lowest residual monomer level in the industry and that these products may be safely processed and used in industry and in food contact applications that will meet the criteria of any new regulations. Analytical Testing ^ The initial objective of this work therefore was to develop the capability of measuring vinyl chloride monomer levels in the low ppb range using the headspace/GC method proposed by FDA analysts.^ It was early recognized there are many potential inter ferences when measuring KVCM in the ppb range. The peak attributed to vinyl chloride must be confirmed by a subsequent GC-MS analysis which is not always feasible when processing large numbers of samples. Vinyl chloride copolymers appear to contain many more interfering species than homopolymers. In addition, interferences are observed in most lota of helium sparged DMAC and more inter fering species are introduced when the resins are dissolved in the solvents typically used during processing by the coatings industry. The use of the Hall electrolytic detector to replace the FID detector in the FDA method appears to be an effective way to resolve non-chlorine containing interfering species. The results in Table I show the differences obtained using the FDA method with FID detection and the Hall detector. The interfering species in ucc 036638 3. the VAGH sample was subsequently identified by GC-MS as isobutylene. TABLE I RVCM - COMPARISON FID VS HECD RVCM CONTENT - PPB FDA METHOD VYHH B-7661 VAGH A-4608 20% VYHH Solution in MEK FDA Referee Sample FDA Results FID 110/ 130 160 420 2/ 1 6 HECD* 4, 4, <2/ <2 5 4 -- IN V * *Hall Electrolytic Conductivity Detector. Good agreement between FDA analyses and UCC analyses were obtained on FDA submitted referee samples. These plasticized PVC homopolymer samples were tested by UCC using the FDA method separately using FID detection or Hall electrolytic detection. Results are shown in Table II. TABLE II Lab Detector RVCM REFEREE SAMPLE TESTS RVCM Content - ppb__________ FDA FlS 6 ________UCC FID 1/2 HECD 4 UCC 036637 the VAGH sample was subsequently identified by GC-MS as isobutylene. TABLE I RVCM - COMPARISON FID VS HECD RVCM CONTENT - PPB FDA METHOD VYHH B-7661 VAGH A-4608 20% VYHH Solution in MEK FDA Referee Sample FDA Results FID 110, 130 160 420 2,1 6 HECD* 4, 4, <2, <2 7, 7, <2 5 4 *Hall Electrolytic Conductivity Detector. Good agreement between FDA analyses and UCC analyses were obtained on FDA submitted referee samples. These plasticiz d PVC homopolymer samples were tested by UCC using the FDA method separately using FID detection or Hall electrolytic detection. Results are shown in Table II. TABLE II Lab Detector RVCM REFEREE SAMPLE TESTS RVCM Content - ppb FDA FID 6 ________UCC________ FID HECD 1, 2 4 An evaluation of the precision of the PDA method using the HECD at very low monomer levels was completed. Five samples of an aged, well characterized blend of VYHH (vinyl chloride-vinyl acetate copolymer) were analyzed over a period of five days. Results are shown in Table III. TABLE III PRECISION-MODIFIED FDA METHOD Individual Values Separate Analyses Mean Standard Deviation RVCM - ppb 1, 1, 2, 1, 1 1, 2 0.45 F-45 Method The F-45 method was used to test large numbers of pro duction samples. This method was adapted and used because of the much lower analysis time required. Results, shown in Table IV, agree very well with data obtained using the FDA method with electrolytic conductivity detection. The method is a headspace technique using an automatic analyzer (Perkin-Elmer F-45) coupled with an electrolytic conduc tivity (Hall) detector. The chromatographic column contains porous polymer (Tenax GC) and is backflushed to eliminate high boiling components. Polymers are dissolved in high boiling, interferencefree solvent (dimethylacetamide) at concentrations of 7-29% depending on solubility considerations. Volatiles are freed from the resin solutions by heating at 80C. Two types of standards are used: i ir.n (J3B639 J t 5. commercially prepared VCM in nitrogen and monomer solutions in tetrahydrofuran prepared at this site. VCM can be measured quantitatively in the 10 to 8000 ppb range by electronic integration routinely. A level of 4 ppb can be achieved using maximum sample size, an amplifying filter, and manual measurement. This procedure employs an automated headspace analyzer and a flame ionization detector. It had originally been developed by Dow Chemical. Similar methods have been described in the current literature; the most pertinent references are: 1) H. Hackenberg, A. P. Schmidt, "Gas Chromatographic Headspace Analysis", Heyden and Son, Ltd., London 1977, pp. 59-62. 2) "The Determination of Vinyl Chloride-A Plant Manual", Chemical Industries Association, Ltd., London 1977, Standard Methods *5 and #8. Production Resin An analysis of samples of production resins was made to determine: 1. Residual monomer content as produced. 2. Monomer content variability. 3. Residual monomer content as received by UCC customers. The 'as produced' products were measured on Blend samples. Production variability data was determined from several Blend samples. The 'as received' monomer contents were determined on samples taken from bagged resin (50 lb. bags) stored in the laboratory from 2-4 weeks. We estimate the typical time from production through warehousing and the distribution system to the customer- is six weeks. 036M0 6. The corresponding blend samples, taken before bagging, are shown. The results shown in Table IV are typical of UCC solvent vinyl resins. It should be noted that the data shown are results to date; testing on all products will be completed as the production schedule permits. Particle Size The rate of diffusion of a monomer from a polymer increases approximately as the square of the film thickness decreases. In the Union Carbide solution polymerization process, the dry resin is recovered from solution by precipitation with an alcohol^ water mixture. The precipitate thus produced is subsequently dried in hot air. This leads to a porous particle that should result in rapid monomer loss. This theory seems to be confirmed by the observed decrease in monomer content as the products are scored in bags. Particle size and particle size distribution typical of the Union Carbide products is shown in Table V. T. B. Gibb, G. S. Peacock UCC 036641 7 Resin VYHH VYHH VYHH VMCA n n n if VAGH If H VAGD n IV TABLE IV UNION CARBIDE SOLVENT VINYL RESINS .Sample Blend Blend Bagged 4 weeks Blend rt tt Bagged 4 weeks Blend Blend Bagged 2 weeks Blend Blend Bagged 4 weeks Date 1979 8-23 8-23 9-24 8-9 8-13 8-15 8-23 9-24 - - - Blend 7920 7915 7915 140 138 141 142 142 4357 4608 4608 4980 275 275 RVCM - ppb Method p-45 FDA 38 65 <16, 6 8 tf<16, <16, 10 6 <16, <16 39 f <16, <16, <16 - - - 32 7, 7 6 < 2 <10 <10 <10 TABLE V TYPICAL n! VINYL COPOLYMER PARTICLE SIZE AMD DISTRIBUTION VYHH - Shift Composite - 12/2/79 Mesh Microns Wt. % Through Screen 20 841 100 40 420 99 60 250 95 80 177 88 100 149 80 140 105 65 200 270 74 53 46 45 325 44 19 400 Median 37 80 9 036643 ) 9. References (1) J. Lawrence Dennison, Charles V. Breder, Timothy McNeal, Roger C. Snyder, John A. Roach, and James A. Sphon; "Headspace Sampling and Gas-Solid Chromatographic Deter mination and Confirmation of < 1 ppb Vinyl Chloride Residues in Polyvinyl Chloride Food Packaging"; Journal of the Association of Official Analytical Chemists; Vol. 61, No. 4, 1978. X /Id UCC 036644