Oligomers in plastics packaging. 
Part 1: Migration tests for vinyl chloride tetramer  

Food Additives and Contaminants v.13, n.3, 307-314. 1996

Laurence Castle,^1§ David Price,² and John V. Dawkins²

¹ Ministry of Agriculture, Fisheries and Food, CSL Food Science Laboratory, Norwich Research Park, Colney, Norwich NR4 7UQ, UK;
² Department of Chemistry, Loughborough University of Technology, Loughborough, Leicestershire LE11 3TU, UK
^§ To whom correspondence should be addressed.
   (Received 30 June 1995; revised 12 October 1995; accepted 16 October 1995)

Vinyl chloride (VC) tetramer has been studied as a representative oligomer that has the potential for migration from plastics packaging. Poly(vinyl chloride) (PVC) bottles for retail beverages were analysed by a process of dissolution followed by gas chromatography. Tetramer levels ranged from 70 to 190 mg/kg in the plastic. When these bottles were tested for migration into the simulants distilled water, 3% acetic acid, 15% ethanol and olive oil, no tetramer migration was detected at a limit of 5-10 µg/kg. Since, of the low molecular weight oligomers, the tetramer had the highest concentration in the PVC plastics, it is concluded that the other VC oligomers of higher molecular weight, would not migrate above this limit of detection either.

Keywords : PVC, packaging, composition, oligomers, tetramer, migration

Introduction

The possibility of migration of monomers, additives and low molecular weight oligomers from plastics packaging into foodstuffs is well recognized (Shepherd 1982). Poly(vinyl chloride) (PVC) is used in many countries as a food packaging material although in some its use seems to have declined over recent years because of its perceived environmental impact. It is a versatile material that can be formulated for many different packaging applications (Jenkins and Harrington 1991). Since vinyl chloride monomer (VCM) is highly toxic (Viola 1970, Maltoni and Lefemine 1974) monomer levels in PVC for food contact applications are closely controlled. Both the residual monomer content of the polymer and migration levels to foods or food simulants are regulated (EC, 1980, 1981). In addition to residual VCM, however, PVC can contain low molecular weight oligomeric species with the potential to migrate (Dawkins et al. 1991). The total migration from unplasticized PVC (`overall migration') is typically around 1 mg/dm² when sheets are tested with a fatty food simulant, such as olive oil, or isooctane (de Kruijf and Rijk 1988). It is possible that a large part of this migration could comprise vinyl chloride (VC) oligomers.

We have reported recently on the isolation and characterization of the two major VC tetramers. These were found to be isomeric forms of the structure shown in figure 1 (Dawkins et al. 1995). These oligomers do not share the RCH=CHCl structural feature of the VC monomer and therefore do not raise an immediate alert with respect to toxicity by analogy with the monomer. Nevertheless, as organo-chlorine compounds, the migration of VC oligomers is of interest. The purpose of this work was to measure VC tetramer concentrations in food contact grade PVCs and thereafter estimate migration levels of the tetramer and other small oligomers from the plastics.

A standard of VC tetramer (in two isomeric forms) was obtained from a mass polymerized PVC using diethyl ether extraction followed by a multi-stage chromatographic isolation as described by Dawkins et al. (1995). The standard was dissolved in hexane (1 mg/ml) and stored at 4°C in the dark.

Hexane and tetrahydrofuran were glass distilled grade from Rathburns (Walkerburn, UK). Olive oil of a grade suitable for migration testing was obtained from Pira International (Leatherhead, UK). Acetic acid (Aristar grade) and hexachlorobuta-1,3-diene were supplied by BDH Ltd (Poole, UK) and ethanol (absolute alcohol) was from Hayman Ltd (Waitham, UK).

PVC base resins were supplied in powder form by Atochem (masspolymerized), BASF (suspension process) and Solvay (process undeclared). PVC bottles fabricated from mass-polymerized plastic were supplied by Norsk Hydro. Retail samples of beverages in PVC bottles were identified in store from the characteristic injection moulding mark for PVC (Anon. 1992) and were purchased locally.

The approach to polymer analysis was to dissolve the polymer in solvent, divide the solution into two portions, and quantify the tetramer by a single-point standard addition. Thus, a sample of plastic (0.10 g) was dissolved in tetrahydrofuran (THF, 4 ml) by shaking in a sealed vial for 2 h. Approximately half the solution (2.0 g) was removed to a fresh vial and to this was added VC tetramer standard (4 µg). The mass of the remaining portion was also recorded. Hexane (8 ml) was added to each of the two portions to precipitate the high molecular weight polymer. Each of the samples was then filtered (pore size 4 µg) and the solvent removed using a stream of nitrogen at ambient temperature. The residue was re-dissolved in hexane (025 ml) and the solution placed in a sealed vial for gas chromatographic analysis with electron capture detection (GC-ECD).

Four samples of orange or lemon cordial were analysed for VC tetramer. A sample of the cordial (100 ml, not diluted with water) was poured from the retail bottle and extracted with hexane (1 x 50 ml). The solvent was removed under reduced pressure and the residue taken up in hexane (0.25 ml) for GC-ECD analysis.

Strips of plastic (six at 110 x 20 mm) were cut from the wall of PVC bottles. Retail purchases were first emptied and rinsed with distilled water to clean. The specimens were then tested with the simulants distilled water, 3% w/v acetic acid in water or 15°/a v/v ethanol in water (150 ml) by immersion or filling as appropriate. After the exposure period, the simulant was poured into a round bottomed flask and the solvent evaporated under reduced pressure. The residue was taken up in hexane (1.0 ml) and the solution divided into two equal portions for standard addition. To one portion was added tetramer standard (4 µg). The portions were then analysed by GC-ECD.

Since it was anticipated that the analysis of olive oil for VC tetramer would have a relatively poor limit of detection, exposure of PVC samples to olive oil used a high ratio of contact area to oil mass. The wall of a plastic bottle was cut down its axis to give. a portion of wall along with part of the base and shoulder. This portion of plastic formed a `boat' when laid down. Two further pieces were cut from the bottle to nest closely inside the first. The `boat' was then filled with olive oil (9-15 ml) taking care that all surfaces were wetted to give a total area in contact of 7-10 dm² (depending on bottle size). After the exposure period, a portion of the oil (0.10 g) was diluted with hexane (0.10 ml) and analysed by GC-ECD. Further portions of oil were taken for quantification by standard addition.

GC-ECD analysis for plastics and stimulant samples was performed using a Carlo Erba 4160 chromatograph fitted with a CPSil 5CB capillary column (Chrompack), of dimensions 17 m x 0.25 mm with a 0.25 µm film thickness. The carrier gas was argon/methane at 1 ml/min. Injections (0.5 µl) were made in splitless mode using a Carlo Erba A200S autosampler and with the injection block held at 220°C. The column oven was programmed from 70°C (held 1 min) rising at 20°C/min to 110°C, at 4°C/min to 160°C, and finally at 50°C/min to 180°C to clean. A Carlo Erba ECD 400 electron capture detector was used, held at 280°C. Quantification was by integrated peak areas (Spectra Physics Chromjet).

Tetramer standard (4 µg) was added to olive oil (0.1 g) in a clear glass vial with a PTFE-faced septum closure. Stability at 40, 150 and 175°C was studied. For tests at 40°C, two sets of storage conditions were used; the first in total darkness and the second in subdued light stored by the glass window of an incubator under normal laboratory lighting. Tests at 150 and 175°C were conducted in subdued light. After the storage period, a chromatographic internal standard (hexachlorobuta-1,3-diene, 35 µg). was added and the oil then diluted by the addition of hexane (0.10 ml). Analysis was by GC-MS (Hewlett Packard model 5971) with a 17 m x 0.25 mm x 0-12 µm CPSil 5CB column (Chrompack) programmed from 70°C (held for 1 min) then raised at 20°C/min to 110°C then at 4°C/min to 160°C. Injections (1 µl) were made splitless at 280°C using an HP A7673 autosampler. Ions monitored were m/z 105 (tetramer) and 225 (HCBD). These ions were selected as the least prone to oil-derived interferences. The peak area ratio of tetramer to internal standard was compared directly with that for samples prepared freshly, to reveal the extent of any loss of tetramer.

The samples analysed

The major use of unplasticized PVC in food contact applications is bottles used for mineral water and fruit juices and analysis focused on these articles. Two samples of PVC base resin and one set of PVC bottles supplied unfilled by the fabricator, were analysed also. Plasticized PVC as used in thin wrapping films or in gaskets and seals, was not examined since it was anticipated (and subsequently found) that oligomers would be at trace levels only and the potential for migration from these thin materials would be lower than for the heavier PVC bottles.

The approach taken to determine levels of VC tetramer in samples of PVC was to dissolve the plastic in THF and then precipitate the high molecular weight fraction by the addition of a poor solvent, hexane. The tetramer isomers (figure 1) co-eluted under the GC conditions used and were summed. To establish the sensitivity and linearity of this approach, a series of solutions of mass-polymerized PVC was spiked with tetramer at levels from 0 to 40 mg/kg and then analysed. The resulting calibration line was rectilinear (r = 0.9882) with an intercept at the y-axis equal to 90 mg/kg---an early indication of the tetramer levels expected. These results demonstrated the method to be capable of measuring down to 10 mg/kg in the plastic with ease.

Triplicate bottles fabricated from mass-polymerized PVC were sub-sampled by taking portions from the top of the neck and from the base. The mean of the six determinations was 118 mg/kg with a coefficient of variation of 29%. This indicated that any bottle inhomogeneity coupled with variability of the analytical method, was not greater than 30% which was considered acceptable for establishing typical compositional levels. This meant that for the sampling of retail bottles, a portion of plastic cut from the neck of the bottle (with the advantage of no prolonged contact with the food) would be representative of the bottle as a whole.

Levels of tetramer in retail packaging
The analysis of 13 retail PVC bottles found tetramer levels in the range 70-190 mg/kg (table 1). Two of the resins supplied by manufacturers were at the extremes of this range, at 80 and 210 mg/kg (table 1). The approximate three-fold range in tetramer content of retail articles was not unexpected. It is known that the residual oligomer content of PVC will depend on three factors; (i) the original polymerization process to form the base resin (Billmeyer 1984), (ii) vacuum stripping to reduce VC monomer levels (Burgess 1982) which will also remove low molecular weight oligomers, and (iii) any subsequent thermal processing of the base resin to form the finished material or article. At around 100-200 mg/kg, the VC tetramer is at a similar concentration to, for example, antioxidants and other additives in polyolefins. It was considered therefore that migration studies were warranted.

Table 1. Tetramer content of PVC packaging samples.

               Description of                   Tetramer content a
                 the plastic                         of the
Sample         used in packaging                 plastic (mg/kg)
  1       Whole orange cordial brand A                  82
  2       Whole orange cordial brand B                  73
  3       Diet orange cordial brand A                   85
  4       Orange drink brand C                         110
  5       Whole lemon cordial brand A                  160
  6       Diet lemon cordial brand A                   140
  7       Orange, lemon and pineapple cordial brand G   90
  8       Mineral water brand D                        120
  9       Mineral water brand D                        140
 10       Mineral water brand D                        100
 11       Mineral water brand E                        190
 12       Mineral water brand E                        110
 13       Mineral water brand E                        130
 14       Base resin 1 (suspension)                    210
 15       Base resin 2 (process not known)              80
 16       Base resin 3 (mass-polymerized) b             90
 17       Bottles supplied by fabricator c             120
          (mass-polymerized)

a Measured by single-point standard addition. 
b Measured by four-point standard addition. 
c Mean of six determinations. 
Note: A cordial is a sweetened drink containing a fruit flavouring, 
      usually requiring dilution with water before consumption.

Tetramer migration into retail beverages

Four samples of orange and lemon drinks were purchased in PVC bottles that contained 80-110 mg/kg tetramer in the plastic (table 1). When the drink was analysed no tetramer was detected (table 2). Drink spiked with tetramer at 30 µg/kg showed a recovery of 70-120% and therefore it may be concluded that migration had not occurred above the 10 µg/kg limit of detection.

Migration testing using simulants

To test PVC articles for general food use, the four EC recommended simulants were employed (EC 1985). When tested using 3% acetic acid or 15% ethanol over 7 days at 40°C, migration was below the limit of detection (5 µg/kg table 2).

Table 2. Migration levels.

                                        Tetramer
                     Conditions         migration
   Simulant           of test            (pg/kg) 
Distilled water    7 days at 40°C         < 5
Distilled water    1 h at 70°C            < 5
3% acetic acid     7 days at 40°C         < 5
15% ethanol        7 days at 40°C         < 5
Olive oil          7 days at 40°C         < 10
Retail drinks      Several weeks at       < 10
                   ambient temperature

                           Tetramer survival (%) 
Storage conditions       Isomer 1        Isomer 2
150°C/30 min               110              90
175°C/120 min              110             110
40~C/10 days: Light         70              90
              Dark          60              80

Similarly, when a mineral water beverage bottle was tested using distilled water at the severe conditions of 70°C for 1 h, no migration was observed (table 2). PVC bottles have in the past been widely used for cooking oils but PET and glass are now commonly used for this application-in the UK at least. Nevertheless the PVC bottles were tested using olive oil since this simulant can be expected to be the most aggressive towards both the plastic and the lipophilic tetramer. Since olive oil cannot be evaporated to a small volume for analysis and so detection limits are inferior to the aqueous simulants, the migration experiments with oil were designed to give a high plastic ratio of 7-10 dm² for each 9-15 g of oil, in contrast to the conventional ratio of 6 dm²/kg (Rossi 1987). This gave a lower effective limit of detection when calculated for the standard area to simulant ratio. Figure 2 shows representative chromatograms for the analysis of exposed olive oil before and after the standard addition of tetramer. For all samples studied no migration was observed above the limit of detection of 10 µg/kg (table 2).

There is anecdotal evidence that vinyl chloride is not stable in foods or simulants following spiking (Biltcliffe and Wood 1982). The issue of monomer stability in simulants has been studied systematically more recently (Philo et al. 1994). The stability of the VC tetramer was tested here in olive oil only, since the detection limit was the worst for this simulant and also, as discussed above, the oil was considered to be the most severe test simulant towards PVC. The results in table 3 demonstrate that each tetramer isomer is stable enough to draw the conclusion that the failure to detect tetramer in migration tests was because of a lack of migration and not because of instability in the simulant.

Figure 2. Typical chromatograms from the analysis of olive oil for VC tetramer.
(a) Blank olive oil. (b) Olive oil spiked with tetramer equivalent to 55 µg/kg migration under standard conditions of test (see text). Tetramer (mixed isomers) at 15.1 min. y-axis; ECD response. x-axis, GC retention time in minutes.

Assessment of oligomers other than the tetramer

From a knowledge of the concentration of oligomers, other than the tetramer, in typical PVCs, and from basic migration theory (Pugh 1994), it is possible to estimate the potential for these other oligomers to migrate using the tetramer as a guide. The liquid chromatographic, GC-MS and ¹H-nuclear magnetic resonance spectroscopic analysis by Forrest (1988) indicated that concentrations of VC dimers and trimers in PVC are at least 10-fold lower than the tetramer. Therefore, although these smaller oligomers will migrate more readily than the tetramer (Pugh 1994), their low concentration in the plastic means they are unlikely to migrate above 10 µg/kg. For oligomers larger than the tetramer, i.e. the pentamer, hexamer, etc., concentrations in the plastic are comparable to, but always lower than, the tetramer (For-rest 1988). Since these higher molecular weight oligomers will migrate less readily than the tetramer they also do not have the potential to migrate above 10 µg/kg under the conditions of the test used here.

References

Anon., 1992, Plastics recycling. Packaging Week, 12 February, 20-21.

Billmeyer F. W., 1984, Textbook of Polymer Science, 3rd edition (London: Wiley).

Biltcliffe, D. O., and Wood, R., 1982, Determination of vinyl chloride in foods: collaborative study. Journal of the Association of Public Analysts, 20, 55-65.

Burgess, R. H., 1982, The toxicity of vinyl chloride and its removal from PVC. Manufacture and Processing of PVC, edited by R. H. Burgess (London, UK: Applied Science), pp. 101-122.

Dawkins, J. V., Forrest, M. J., and Shepherd, M. J., 1991, Multi-step procedure for the separation of vinyl chloride oligomers. Journal of Chromatography, 550, 539-547.

Dawkins, J. V., Moody, C. J., Price, D., Castle, L., and Howarth, O. W., 1995, Isolation and structural characterisation of tetrameric oligomers of vinyl chloride. Macromolecules, 28, 29852987.

EC, 1980, Commission Directive 80/766/EEC laying down the Community method of analysis for the official control of the vinyl chloride monomer level in materials and articles which are intended to come into contact with foodstuffs. Official Journal of the European Communities, L213, 42-46.

EC, 1981, Commission Directive 81/432/EEC laying down the Community method of analysis for the official control of the vinyl chloride release by materials and articles into foodstuffs. Official Journal of the European Communities, L167, 6-11.

EC, 1985, Council Directive 85/572/EEC of 19 December 1985 laying down the list of simulants to be used for testing migration of constituents of plastics materials and articles intended to come into contact with foodstuffs. Official Journal of the European Communities, 1372, 14-21.

Forrest, M. J., 1988, Characterisation of vinyl chloride oligomers. PhD thesis, Loughborough University of Technology, UK.

Jenkins, W. A., and Harrington, J. P., 1991, Packaging Foods with Plastics (Lancaster, PA, USA: Technomic).

de Kruijf, N., and Rim, M. A. H., 1988, Isooctane as fatty food simulant: possibilities and limitations. Food Additives and Contaminants, 5, 467-483.

Maltoni, C., and Lefemine, G., 1974, Carcinogenicity bioassays of vinyl chloride Part 1. Research plan and early results. Environmental Research, 7, 387-405.

Philo, M. R., Jickells, S. M., Damant, A. D., and CASTLE, L., 1994, The stability of plastics monomers in food simulating liquids under European Union migration test conditions. Journal of Agricultural and Food Chemistry, 42, 1497-1501.

Pugh, S. R., 1994, Modelling chemical migration from packaging into food. Food Science Reviews, Vol. 2, Chemical Migration from Food Packaging, edited by D. H. Watson and M. N. Meah. (Chichester UK: Ellis Horwood), 137-155.

Rossi, L., 1987, Activities of the Commission of the European Communities concerning materials and articles intended to come into contact with foodstuffs-a review. Food Additives and Contaminants, 5, 21-31.

Shepherd, M. J., 1982, The trace contamination of foods by migration from plastics packaging: a review. Food Chemistry, 8, 129-145.

Viola, P. L., 1970, Pathology of vinyl chloride. Medicina del Lavoro, 61, 174-180.