Document mB1kL24KQvbKK0KLnMgzokwEd

;u?\c applied cHiiMTsrnv division Cor.ai saion on rood Contaminants % w7// - [> l/ R&S 135181 / Project Ko. l: Collection of information on * Food Contaminants derived from Food Packaging Part 1: Vinyl Chloride in the Packaging Material 1) Introduction During 1973,increasing concern vas expressed about possible carcinogenic and other effects of vinyl chloride (VC). The practice 1 implications include a) the health of industrial workers exposed to VC during manufacture and processing of PVC, b) the migration of VC, whose presence in PVC has been unequivocally demonstrated, into the atmosphere during processing, c) the migration of VC into food and drink packaged in PVC containers. A careful literature search over the last years revealed that most interest has been given to occupational hazards (points a and b) whereas data on non-occupational exposure (point c) are relatively scarce. According to the objective of the project this report, deals exclusively with VC in packaging materials and its migration into the packed foodstuffs. /... R&S 135182 i 2 Host of the data given in this report emerged iron investigations performed in the author's laboratory and have not been published up till nov. To enable a reasonable interpretation of the data the methods for analysis are described in detail. 2) VC in the packaging material In a technical report of WHO (l) the content; of VC in food packaging material was summarized as follovs: a) bottles generally contain lass than 100 ppm and a level below 30 PPm is now more usually observed, b) rigid film generally contains less than 10 ppm and a level of less than 1 ppm is often observed, c) flexible film generally contains less than 1 ppm and a level below 0,2 ppm is usually observed. According to our analyses we observed three periods depending on the effort the manufacturer have undertaken to decrease the VC content. 2.1. PVC-filras c.f former production charges (l. period) In the investigaticn of 80 PVC-filcs of former production charges (up till approx. June 197^*) the frequency distribution of the VCcontent was found to be as shown in the following table. As can be seen from this table, the VC-content mostly varied between 200 and ^*00 ppm. These PVC-filns were made of raw materials with high VC- concentrations. Sometimes VC-concentrations of more than 1000 ppra were found in powdered PYC. / VC-co.it t o f P'v'C-filar s f former product ion oh a roe; VC-content | [pp=] U 50- :-c'-2` o r-'o- 300 300-400 > 4co f i I r.i s 12 15 23 24 c 2.2. Transitional stage (2. period) Immediately after discussions about the possibility of monomeric c VC being detrimental to health had started, considerable efforts were made on the part of the PVC-processing industry to reduce the VC-content of PVC-filtas drastically. o In Fig. l'ths reduction of the VC-content during processing is illustrated, starting from powdered PVC with a VC-contcnt of - over 1100 pen. By hot nixing at different temperatures (50, 75 or 100 C), cold-mixing, extrusion and in the film-making process the VC-content of the final product can be reduced to 85, 30 and 15 ppm respectively. This confirms the finding of the Verband Kunststoifcraeugende Industrie e.V. (2) that increased processing temperatures reduce the VC-content of FVC. Uurir.g the transitional stage PVC-films were made, containing generally less than jO ppx VC and only in exceptional cases up to IOO ppm VC. At the sane time successful attempts were made by she PVC-xar.u lecturing industry to reduce the VC-cor.tent in tne PVC starting materials. As can be seen from the next table, it was now possible for the processing industry to use powdered FVC containing less than p ppa VC. R&S 135183 t VC -cn r.tent of PYC-powders 2.3* PVC-films wi th a lov VC-content (3. period) After all the difficulties had been surmounted, it was now possible to make PYC-films with a low VC-content. Of the charges produced between December 1974 and Kay 1975 a total number of 776 films were analysed. Vith a few exceptions the VC-concentrations were less than 5 PPra. VC-content 01 recently produced PVC-films VC-content [ppnj Number of films < 0. 1 3SO 0.1-1.0 308- 1*1-2.0 53 2.1-5-0 32 5.1-10 1 3 Prom these values it is obvious that the PVC manufacturing and processing industries have made great efforts to reduce the VCcontent of their products and that the production of PVC-films with VC-concentrations ^10 ppm is now technically feasible. 3- Migration of VC into simulants and packed poods 3*1. Simulants To investigate the relationship between the VC-concentration in the packaging material and the VC-content in the packed product, migration tests were carried out using the synthetic fat simulant 1 . ] IIB 3C7 ' . HS 307 is a mixture of synthetic triglycerides. The `) 1.11 j07 in cor: tent and defined composition is delivered by: * hATKC, tit- s oil chitft fiir r.n tu rwi s senschaf 11 i eh-techni schc Dienste mbit 2 Kuril'*:'"'-! ^0. ii.vhri .'.on t r.iiin 1~*V. R&S 135185 fat simulant HD 3CV permits the Geternination of the migration of p.'.CAiu'i n g material cooponent. s into fatty food tinder standard _ :i 1 and with a nigh degree of accuracy. The technique applied has been described in detail by K. Figge and J. Kor.h (3). The table below shova the VC raigration values for a PVC-filci with a VC-content of 60 ppm at different temperatures (20C and 40C) as a function of rime. Fron these values it can be seen that during storage at 20C the VC content in. the sinulant increased contir.ously. At 40 C, however, the migration reaches a maximum, after which the VC content slowly decreases again. VC-mi cirat ion f ron a PVC-film cont aininq 60 ppm VC into H3 30? Storage t erape rature LCJ VC -oig ra tior. ptg/d 10 20 4.0 80 days days days days after 160 days VC-content of the film after l60 days Fppa] 20 5 8 15 29 45 40 IS 29 60 57 51 14 0.7 2 In the calculation of the migration values (ug/rira ) the following fsciors vre taken into account: contact surface between PVC and simulant, weight of sinulant and VC-content in sinulant after storage. In the conversion of the migration values into the VC-contcnt in the packed product, the PVC-contact surface must be taken into account. Applying the rule of thumb that 1 kg is packed in 6 2 dm , the VC- content in the packed product stored at 20C for 80 days works out at 0.l3 ppm. The following table and Fig. 2 show migration values for PVC-filns contair.g different VC-concentrat i ons ( 10 days, 40 C and 20 cays, 4C-dC). f ma-- VC-mig rntion into K3 30 7 at o 4C C : VC-con'-tn: of the FYC-film f? P^J 1 r- , JVC-migra tion U:y/dm after IO days 20 clays 1093 340 520 833 300 420 799 230 390 571 170 280 / 438 130 240 60 - 18 29 r Especially from Fig, 2 it can be seen that a linear relationship exists between the VC content of the packaging material and the VC content in the fat simulant. R&S 135186 For a given VC-limit for the packed product the maximum permissible VC-conccntration in the packaging material can be calculated from this relationship. Putting the life of a PVC-packed product at 8 veeks plus a margin of 4 weeks, the value for 30 days/20C (s. table) can be taken as a basis* Assuming the maximum permissable VC- concentration in the packed product to be 1 ppm, as recommended by Kunststi! f f koisiaiisie-r: des 'dundesgesundhei tsamtes in Berlin, it can be calculated that uhe packaging material must not contain more than 300 ppm. If the VC-linit for the packed product is, for instance, 0.05 ppm, the maximum permissible VC concentration in the packaging material is 15 ppm. Naturally, these values are only approximate values, a3 other factors, e.g. the cor.t en t s/'su rf ace ratio, the wall thickness and hence the absolute VC-content in the packaging material also play a role. The results reported in this chapter have been published by V.R. Eckert (4). 7 R&S 135187 3-2. Food stuffs Only few data are available on the content of VC in food. After up to three year's storage in niniature PVC bottles, gin and whisky contained 0,57 and 0,62 ppm of VC respectively ( l).. This led US Food and Drug Administration to the proposal to ban the use of PVC bottles for alcoholic beverages. The concentration of VC in orange squash and cooking oil have been found to be in the range of 0,01 to 0,08 ppm and 0,01 to 0,04 ppm respectively. The mioration of VC from PVC bottles is dependent on the VItC content of the bottle, nuairtL. storaae (l). the temperature and the duration of 3*3- Hair care products Ve examined various hair cosmetics and the corresponding PVC packaging materials: Product Shampoo A Shampoo B Shampoo C Shampoo D Set ting lotion 100 cc Setting lotion 19 cc Setting lotion V Setting lotion Y Setting lotion 5 Fon lotion Hair conditioner A Hair conditioner B Storage period at room temperatu re 30 weeks 30 weeks 30 weeks 5 months 30 weeks 30 weeks 30 weeks 15 weeks 3 months 3 months 30 weeks 30 weeks VC-contcnt (ppm) Containsr Filled product 51 44 15 162* 0. l4 O.56 0.08 0.6 156 2.8 32 0.25 <1 43 l4l* 26* 31 18 < 0.05 1.0 1.2 0.2 0.24 0.17 j during filling operation The measured values show a clear relationship between high VCcontent of the packaging and high VC-content in the packed product. Alcohol-based products were observed to absorb particularly large quantities of VC. 3.4. Rinsing_agents The VC-contents of different packed products were between C.3 and , 2.5 ppm. It should be considered that rinsing agents are used in dilutions of 1 : 100 with water - corresponding to VC-values of max. O.O3 ppm. C. Product Storag e period (weeks) VC-conter.t (ppm) Household cleaner Dish washing fluid A 750 nl Dish washing fluid B 750 "I Rinsing fluid for automatic machines ( 35 0.4 19 1.1 24 0.3 17 0.3 31 2.2 31 2.5 21 1.0 29 1.0 29 1.3 20 1.6 20 2.0 Here wc could only examine PVC-bottles of another delivery: Values between 60 and 46y ppm V'C vere observed. R&S 135188 4. Methods A highly-sensitive method was required for the identification of YC, Ve decided to use the headspace technique and developed the following procedure in close accordance with the methods pubiished by H,, Puschmann (5)- 4.1. Principle of the method For the investigation of PVC-products 10% solutions are prepared in closed injection flasks. The solvent used is N,N-dime thy1 acetamide (DMA) to which about 20 ppm of diethyl ether have been added c c u L.. as internal standard. These solutions are therraostated at 50C to establish the equilibrium. The subsequent gas chromatographic analysis of the gas phase is carried out with the GLC-automat Kultifrakt F 40 (Perkin-Elmer). Packed goods are investigated I analogously, whereby the ratio test material to solvent must be adapted for each packed good. 4.2. Reagents N', N-dimethyl acetamide, a.g., Merck diethyl ether, a.g., Merck vinyl chloride, 99*9%, Baker Chemicals Standard solution: Weigh 70 g of dimethyl acetamide into a 100-ml injection flask and add l40 mg of diethyl ether. Weigh exactly to 0.1 rag. Dilute this solution ~ with dimethyl acetamide until the standard solution contains a defined amount of 2ug ether / ml. Calibration solutions: Starting from a 0,2% solution of diethyl ether in DMA, calibration solutions with defined amounts of VC and ether are prepared by introduction of VC and subsequent dilution with DMA. ID to co cn 03 co 4*3. Preparation of samples Weigh 200 mg of PVC exactly to 0.1 mg into a 23-nl injection- flask (accessories to F 40) and provide with a magnetic rod. Add 2 ml of standard solution with pipette and close flask immediately. Dissolve sample on a heatable magnetic stirrer while stirring vigorously. Finally the injection flask may be placed in a water bath at about 60~70C. As soon as all polymer parts are dissolved, equilibrate the flask for 30 nin at $0C in the water bath of the r 40. For the investigation of simulants or packed goods weigh 2 g of sample into injection flask and add 0*5 ml of standard solution. /... 1o R&S 135190 Add oagnetic rod and close test tube immediately. Dissolve or homogenize sample in the warmth on a magnetic stirrer. Transfer sample to thermostate of the F 40. After 30 rain the analysis can be performed. 4.4. Gas chromatographic conditions The gas chromatographic .analysis is carried out with the Multifrakt F 40. Up to 30 samples are placed in the sample changer therraostate o at 50 c* After the equilibration time of 30 nun, the analysis cycle is started. Instrument conditions on F 40: water thermostate dosing line dosing time time of analysis blowing time re-entrance time column column temperature carrier gas retention times 50c 150C 5s 3 rain (VC in PVC) 0,5 min 0,5 min high-grade steel tube, 4 m in length, 2 mm int. diara., packed with 15% of Ucon LB 550"1 on Chroaosorb W-HP (100-120 raesh) 60C 20 ml N^/min 3.45 rain for VC 5-20 for diethyl ether The chorraatograras are evaluated quantitatively in a connected 3352-B data system (Hewlett-Packard). Under the conditions given the cycle time is 9 tain. After 22 analyses, the automatic course of analysis must be interrupted and the column temperature increased to l80C to heat out the DMA. After this the oven temperature again is adjusted to 60 C and the automatic course .of analysis restarted for further samples. The tine of interruption due to heating-out is approx. 45 min. This R&S 135191 - 11 time table of capacity only applies to the deternination of VC in PVC. Tn the determination of VC in packed goods longer times of analysis (up to 20 min) rust be taken into consideration since some of these products contain easily volatile components which may influence the course of analysisconsiderablv. 4.5* determination of the calibration factors For the determination of the calibration factors volumes of 2 ml of the calibration solutions are placed in injections flasks and closed. The flasks are equilibrated in the F 40 and analysed under the conditions described above. The calibration factor F is calculated according to ast .. cvc F CST . AVC A ST = peak area ether AVC = peak area VC Q ST = concentration ether (internal standard) CVC = concentration VC The instrument-dependent calibration factor was determined to F = 0.8p By addition of 200 mg of VC-free PVC to the calibration solutions it was ascertained that dissolved PVC does not influence the factor, The calibration factor-only applies to the quantitative determination of VC in PVC. For the determination of VC in simulants or packed goods nev calibration factors must be determined for each substrate because of the varying solubility of VC, Moreover, it has to be ascertained that the samples under investigation do not show peaks interfering with the evaluation. This can be done by the analysis of samples which have not been in contact with PVC. / i2- 4.6. Calculation of the results The calculation is carried out according to V . F * CST ppra VC = The concentration for the standard C the amount of the sample weighed in. is 8iven in ppm, based on 4.7* Reproducibility of the method c In order to check the reproducibility of the method, two series of measurements were carried out. In each case we analysed 12 PVCbottles of the same batch. 1st series of measurements (Batch l) single values: 7.0 7.0 6.6 6.8 7.0 6.9 70 7.0 7.1 7.2 70 6.6 ppra VC mean value: 7.0 _+ 0.23 ppm VC 2nd series of measurements (Batch 2) single values: 0.62 O.58 0.57 0 O 0.58 0.57 0.55 O.57 O.58 0.56 0.61 0,60 ppra VC mean value: O.58 + 0.022 ppra VC 4.S. Reference analyses Basing on two examples it shall be shown, in how far the analytical results of different laboratories are in agreement. -Example 1: From Van den Bergh + Jurgens Ltd., Purfleet, Great Britain, we obtained PVC-bottles ex I.C.I. The results of analysis are I.C.I.: our laboratory : 1.5 ppm VC 1.4 ppm VC /... R&S 135192 13 R&S 135193 Examp]e 2: Two batches of PVC powder ex Chomische Herke Hu Is, Gernany (CtH) were analysed at CIP.I and In our Laboratory. The following values were obtained: Batch 1 2 Content of VC (ppm) CVH our laboratory 8l 93 41 37 Since VC is an easily volatile substance, these.differences in case of PVC powders can be explained already by different periods of sample preparation. Detection limits The detection limit of the method for VC in PVC-powders and PVCpackaging materials is o.l ppm. In simulants and packed goods the detection limits depend on the substrate. VC-contents as low as 0.01 ppm can still be identified accurately in simulants. Conelu sions On 26 March 19?4 a group of European toxicologists discussed the available toxicological and migration data on VC, with special reference to its carcinogenic potential. The meeting was held at the National Institute of Public Health, Bilthoven, The Netherlands. As far as the problem of VC in packaging material is concerned the following conclusions were reached and agreed unaniraously: "At the present time there is no need to recommend that PVC should be banned as a f ood-wrapp.i ng material. The basis of this conclusion is that relatively high concentrations of VC (by inhalation) are needed to produce a carcinogenic effect. In comparison,the exposure of man from food intake is probably much lower. The further quantitation of this factor will be possible when oral studies on VC have been completed. Furthermore, it was noted that even less / k toxicoloyicaL data worsevailablu on pcsr.ibl c alternative plastics for food wrappings. It was thought desirable and possible that PVC used for food and drink packaging should contain less than 20 ppm VC monomer. In effect this represents the establishment of a food grade of PVC, to ensure very lov levels of contamination.*' (6) In their 57th meeting the "Kunststoff-Konmission" of the "Bundesgesundheitsarat" in Germany declared that there is no evidence of any health hazard after consumption of food stuffs packed in FYC containers. In September 197^ 90 days feeding trials, undertaken by the Dutch Zentralinstitut fur Ernahrungsforschung (CIVO-TNO) have been finished, whereby considerably higher amounts of VC have been applied orally than ever can be expected in food stuffs having had a contact with PVC~. No toxicity was observed. Therefore the oral uptake of traces of VC in food stuffs has to be considered in another way than uptake of VC via inhalation (2). The "Kunststoffkocnissior." has no objections if, as a precaution, t^j content of VC in food stuffs is limited to 1 ppn. The statement bases upon the feeding trials and the analytical results and may be changed after obtaining the results of long-term feeding experiments which have already been a.sked for (2). /\ Hamburg, 29 July 1975 G. Billek, W.R. Eckert u. C. Roloff Unilever Forschungsgesellschaft abH Hamburg 50, Behringstrafie 15^ R&S 135194 * * , ! L. C O 66 . ?.' :'oorroonnoo mi.'s 1. Interr:;;cioiul Agency for Research on Cancer, Internal Technical Report No. 7k/005, "Report of a Working Group on Vir.yl Chloride", Lyon, June 197-*. 2. Verband Kunststoffeneuoende Industrie e.V. "VC/PVC : Halnahnen ;un Gesundhaitsschuti", Frankfurt/K,, November I??*'*- 3* K. Figge and J. Koch, Fd. Cosset. Toxicol. IX.. 975 (1973). ^ Bo ^ Co Ol <D CJi V.R. Eckert, fettc. seifen. ans t ri chtai t tel, 4' in press. 5 H. Pucchnann, "The analytical determination of the content of residual monomers in PVC and food simulants" lecture on the occasion of the PVC discussion meeting, Darmstadt, 2Gth/27th September, 197^* 6. G.J. Esch and M.J. van Leg ten, Information Bulletin B,1.3.it.A. J/}, 2^6 (l97;i). 96VSC1- S * U Fig. 2 Migration of VC into fat simulant H3 3Q7 ( 6 din2 packaging area / kg pro due t ) VC in simulant ^ 6 ts e t