Food Standards Agency
Saturday 4 July 2009
Safer food better business banner
AZ-Directory
What's NewRSS
What is RSS?Listen
Listen to this siteWednesday 1 November 2000
Food Survey Information Sheet
Bisphenol diglycidyl ethers and related substances (Figure 1) may contaminate canned foods in two ways:
There has been considerable work in both continental Europe and the UK to develop and apply methods of analysis for these substances in canned foods. Analytical methods were first developed for BADGE. Evidence was found for migration up to and in some cases exceeding a temporary limit of 1 mg BADGE/kg food set by the European Scientific Committee for Food (SCF; since renamed the Scientific Committee on Food). 1,2 A UK survey published in 1997 found BADGE at levels above this limit in some samples of canned anchovy and sardines. 3 But the levels of BADGE were all less than 1mg/kg in the last part of that survey. 3 Industry had made considerable efforts to reduce the level of BADGE migration from can coatings into food. They took account of the long shelf life of many canned foods and the need to avoid the use of less safe alternatives to BADGE, whilst maintaining the general safety and quality of canned foods.
The possible importance of substances related to BADGE, such as BADGE.HCl and .2HCl, and BFDGE, migrating from can coatings was noted in Swiss work published in 1997. 4 The SCF revised its opinion in 1999 stating that 'In assessing compliance, the restriction [of 1mg/kg] should include the sum of migration of BADGE and its monoepoxy hydrolysis product (BADGE.H2O) when examined in canned foods directly..'. 5 In 2000 the European Commission produced a proposal for a limit of 1 mg/kg for BADGE, BFDGE and their H2O, HCl, 2HCl and H2O.HCl derivatives in canned foods. 6 At present chemical migration from can coatings into food is controlled in Great Britain under the general provisions of the Materials and Articles in Contact with Food Regulations 1987 and of the Food Safety Act 1990. These controls do not state a limit for migration of the above substances from can coatings into food. Similar controls apply in Northern Ireland.
The survey reported here was conducted:
Sampling
Three cans of each sample were purchased, all bearing the same batch number. Samples were purchased from retail outlets mostly in the south of England, as the distribution of the types of samples being tested is similar across the UK. To take account of consumer shopping habits sampling was weighted approximately towards cans from supermarkets (80 per cent), and included 30 per cent of 'own brand' foods. Table 1 provides details of the samples collected.
Much of the work in this survey was on canned fish in oil. Migration of BADGE is most likely to occur into this type of product. BADGE levels are considerably higher in the oil drained from canned fish in oil, than in the fish that is left after the oil is drained off.3 An initial screening exercise was therefore conducted for canned fish in oil samples by measuring BADGE, BFDGE and their reaction products in the drained oil. This initial exercise was conducted by analysing oil samples drained from cans of tuna (13 samples), sardines (17 samples), anchovies (10 samples), mackerel (7 samples), and other fish (5 samples). Analysis of the oil was somewhat easier than that of the total contents of the cans and the results were used to eliminate those samples with no detectable BADGE/BFDGE components. For 15 samples where BADGE or BFDGE components were detected in the oil, second cans were opened and the total contents were analysed. Other foods analysed for BADGE, BFDGE and reaction products were as follows: 11 samples of canned vegetables; 11 samples of canned beverages; 10 samples of canned fish in brine/sauce; 10 samples of canned soup; 5 samples of canned desserts; 4 samples of infant formula; 2 samples of canned fruit; 5 samples of canned pasta; 5 samples of canned meat.
Analysis
Method of analysis of oil from canned fish in oil
The oil was drained from each can and allowed to settle. A 5g sample was removed to a vial and 3ml acetonitrile added. The mixture was shaken vigorously and then allowed to stand. The acetonitrile layer was removed and filtered through a C18 Sep Pak plus cartridge and analysed by HPLC with fluorescence detection. (Details of HPLC conditions used in this work are described in the laboratory report which is available from the library in Nobel House, 17 Smith Square, London, SW1P 3JR; tel. no.: +44 (0) 20 7238 6575).
Quantitative results were obtained by constructing a calibration curve and reading concentrations present in the samples by interpolation. The calibration standards were prepared by fortifying sunflower oil with BADGE, BADGE.2HCl, BFDGE and BFDGE.2HCl at five levels and extraction with acetonitrile as described below.
Method for analysis of total contents of canned foods
The entire contents of a can were homogenised using a food mixer. 20g sub-samples were taken and each blended with 20ml n-heptane and 20ml acetonitrile. After filtering into a Buchner flask through a GF/C filter, the residue and heptane layer were returned to the sample jar for a further extraction with acetonitrile. The filtrate was treated with anhydrous sodium sulphate and decanted to a measuring cylinder. The second extraction was filtered and treated in the same way, a note being made of the final volume. About 2ml of the extract was then pumped through a 0.2 micron syringe filter and injected for analysis by HPLC with fluorescence detection.
Quantitative results were obtained using the HPLC results by comparison against external standards prepared in acetonitrile. For each analytical run, a duplicate sample was fortified with a known quantity of BADGE, BFDGE, BADGE.2HCl and BFDGE.2HCl and then processed in exactly the same fashion as described above. The results were used to estimate the method recovery and check that the method was 'in control'. A solvent blank was also prepared and run alongside the samples.
For the canned samples with a low fat content, for example vegetables, soup, desserts and pasta, the heptane was omitted from the above procedure.
For the analysis of infant formula, 10g portions were extracted with 20ml acetonitrile by shaking vigorously in a 40ml vial for 1 minute. The layers were allowed to settle. The acetonitrile portion was then filtered through a 0.2 micron filter and was ready for HPLC analysis.
Calibration curves were constructed for the analytes available at a high purity by injection of external standards dissolved in acetonitrile. Extract concentrations were obtained from the graphs by interpolation and the levels in the food (in mg/kg) were calculated by multiplying by the final volume of extract and dividing by the mass of food taken. For those analytes where pure standards were not available the same fluorescence response for BFDGE or BADGE was assumed.
An estimate of the repeatability (95 per cent probability level) and recovery of the analytical method was made by conducting six replicate analyses on the homogenised contents of a can of tuna (A9) which had been 'spiked' with known quantities of BADGE, BFDGE, BADGE.2HCl and BFDGE.2HCl.
Method for determination of total BADGE/BFDGE components by hydrolysis
An alternative procedure developed by TNO (Zeist) was evaluated for confirmation and involves hydrolysing the derivatives to the bis-glycols which are then determined by HPLC. At the time the method was evaluated only very scant detail was available and subsequent improvements have been made by TNO using a buffer instead of water to effect hydrolysis.
To determine the total hydrolysis products of BADGE and BFDGE, 1ml of the extract prepared as above was pipetted into a 50ml glass Schott bottle and 2ml water added. The bottle was sealed and placed in the oven at 100¿C for 28 days. After cooling, the sample was filtered through a 0.2 micron syringe filter and analysed by HPLC. Shorter times were found to give incomplete hydrolysis of the chlorohydrins, particularly BADGE.2HCl.
Quantitative values for the appropriate bis-glycol were obtained by interpolating from calibration graphs of peak areas of diluted bis-glycol standards in water against concentration. The recoveries were estimated by fortification of extracts with BADGE/BFDGE and their chlorohydrins. BADGE, BADGE.2HCl, BFDGE and BFDGE.2HCl standards in acetonitrile were also subjected to the hydrolysis procedure to ensure stoichiometric conversion to the bis-glycol.
Quality assurance
The following procedures were used for internal quality control:
The laboratory carrying out the work, Pira International, Leatherhead, Surrey, took part in FAPAS trial Series XII: Round 8 No. 1208, November 1998, relating to the determination of BADGE (as Lab No. 11) and achieved an acceptable z-score of - 0.2.
Reporting
The detection limits were calculated from the levels of BADGE/BFDGE moieties equivalent to three times the standard deviation of noise on analysis. The limits of detection were 0.03 mg/kg for BADGE and each related compound, and 0.04 mg/kg for BFDGE and each related compound. The limits of quantification (ten times the standard deviation of analytical noise) were 0.1 mg/kg for BADGE and each related compound, and 0.13 mg/kg for BFDGE and each related compound. Results were not corrected for recovery because 'spiked' samples were used only for internal quality control and sufficiently pure standards were not available for all of the substances being studied. However variablility of results was encouragingly small, given the complexity of analysing for this group of substances.
Brand names were reported as this survey was carried out in accordance with guidelines for reporting survey results published in the Food Safety Information Bulletin in September 1997. The absence of a particular brand from the Tables in this Information Sheet means only that the brand was not included in the survey.
Back to topOut of 52 samples of canned fish in oil, 13 samples were identified as having quantifiable levels of BADGE, BFDGE and/or reaction products in the drained oil (Table 2). Where one or more of these substances was detected in quantifiable amounts in drained oil, a second replicate can was opened and the substances were measured in the total homogenised contents of the can. One sample (A11) known not to contain detectable BADGE or BFDGE, and two samples (A8 and A12) containing detectable but unquantifiable levels of BADGE were also included. Analysis of the total homogenised contents of these canned fish in oil products showed that three samples contained detectable levels of BADGE and/or its reaction products, and seven contained detectable levels of BFDGE reaction products (Table 3a). Two samples of fish in oil contained BADGE (Table 3a). BADGE was not detected in any other homogenised samples in the survey.
In general, the qualitative findings for samples of homogenised fish in oil (Table 3a) were in line with what was expected from the drained oil analysis (Table 2). However, there were some anomalies, particularly for samples E1, E3 and E5, where BFDGE was found in the drained oil but was not detected in the homogenised fish. To check on this a further replicate can of each sample with the same batch number was opened and the drained oil re-extracted and re-analysed. BFDGE was not detected in this extract, but a peak with a different retention time was seen with the fluorescence detector.
Five out of ten samples of canned fish in aqueous media contained detectable levels of BFDGE reaction products; no BADGE reaction products were detected (Table 3b). BADGE reaction products were detected in four samples of canned vegetables; BFDGE reaction products were found in six samples of canned vegetables (Table 3c). BFDGE products were found in one sample of canned soup and one sample of canned pasta (Table 3d & Table 3e). BADGE, BFDGE and their products were not detected in samples of canned beverages, desserts, infant formula, fruit or meat products (Table 3e, Table 3f & Table 3g).
A comparison was made between the analytical method used in this work and a method involving hydrolysis. The two methods gave similar results for BFDGE reaction products (Table 4a). The hydrolytic method gave slightly higher results for BADGE and reaction products (Table 4b).
NOGE (novolac glycidyl ethers) appeared to be present in one sample of canned fish in oil (A6). NOGE is made by reaction of novolac resins with epichlorohydrin, which was found not to migrate in a previous survey. 7 Novolac resins are complex mixtures made from phenol and formaldehyde. The substance present in sample A6 interfered with the determination of BADGE.2HCl and was tentatively identified as 3-ring NOGE. The extract was reinjected using isocratic HPLC conditions. Under these conditions both 3- and 4-ring NOGE could be seen. No evidence for the presence of 3-ring NOGE was found in any of the other canned food samples.
Back to topUK consumers continue to be protected. Levels of BADGE in canned foods are still within the limit of 1 mg/kg set by the SCF. Indeed the incidence of BADGE residues has declined. BADGE was detected in 2 per cent of samples (2/115) in the survey reported here compared to 36 per cent of samples (58/163) obtained in 1995-97 in the previous UK survey. 3 Combined levels of BADGE and BADGE.H2O in canned foods were also below 1 mg/kg.
This survey was of products on the UK market. It provided encouraging evidence in relation to the European Commission (EC) proposal for a limit of 1 mg/kg for BADGE, BFDGE and their H2O, HCl, 2HCl and H2O.HCl derivatives in canned foods. 6 Ninety-six per cent (110/115) of samples in the survey contained less than the draft limit in the EC proposal of 1 mg/kg for bisphenol diglycidyl ethers and related substances in canned foods.
The Food Standards Agency will be pressing for practical and effective methods of control to be used in developing and applying the EC proposal, so that consumers are equally well-protected in the whole of the European Union. 6 For example BFDGE was not found in fish samples in this survey but it was detected in a recently-reported survey of canned fish in Europe, at levels up to and exceeding 1 mg/kg. 8 The EC proposal includes a wide range of substances including NOGE, which was studied in the work reported here but with an equivocal result. 6 More work is needed on both toxicology and analytical methodology for NOGE. It will also be important to agree a practical limit above which it is certain that samples exceed the 1mg/kg limit. From the work here on analytical methods a figure of 1.2 mg/kg would seem to be appropriate.
Back to top
1
. European Commission, Directorate-General III Industry.
Opinion on bisphenol A diglycidyl ether (BADGE)
. Document CS/PM/2812-Final, 7/6/96.
2
. C. Simoneau, A. Theobald, P. Hannaert, P. Roncari, A. Roncari, T. Rudolph and E. Anklam. Monitoring of bisphenol-A-diglycidyl-ether (BADGE) in canned fish in oil.
Food Additives and Contaminants
1999,
16
, 189-195.
3
. Survey of BADGE epoxy monomer in canned foods.
Food Surveillance Information Sheet no.
125, 1997
4
. M. Biedermann, M. Bronz, K. Grob, H. Gfeller and J. P. Schmid. BADGE and its accompanying compounds in canned oily foods: further results.
Mitt. Gebiete Lebensm. Hyg.
1997, 88, 277-292.
5
. European Commission, Directorate-General XXIV, Consumer Policy and Consumer Health Protection.
Opinion on Bisphenol A diglycidyl ether (BADGE)
. Document SCF/CS/PM 3243 Final, 6/4/99.
6
. European Commission, Health and Consumer Protection Directorate-General.
Working Document concerning the use of BADGE, BFDGE and NOGE in materials and articles intended to come into contact with foodstuffs
. Document SANCO/2000/2874-Rev.5b, 24/10/00.
7
. Survey of chemical migration from can coatings into food and beverages - 2. epichlorohydrin.
Food Surveillance Information Sheet no. 170
, 1999
8
. A. Theobald, C. Simoneau, P. Hannaert, P. Roncari, A. Roncari, T. Rudolph and E. Anklam. Monitoring of bisphenol-F-diglycidyl-ether (BFDGE) in fish canned in oil.
Food Additives and Contaminants
2000,
17
, 881-887.
Dr David Watson
Food Standards Agency
Chemical Safety and Toxicology Division
Email: : david.watson@foodstandards.gsi.gov.uk
A copy of the full report of this survey has been placed in the Food Standards Agency library. If you wish to consult a copy please contact the library for an appointment giving at least 24 hours notice or alternatively copies can be obtained from the library; a charge will be made to cover photocopying and postage.
Back to topDownload pdf
(pdf 9KB) 09/00 - Table 1aDownload excel
(Excel 17KB) 09/00 - Table 1bDownload excel
(Excel 18KB) 09/00 - Table 1cDownload excel
(Excel 17KB) 09/00 - Table 1dDownload excel
(Excel 21KB) 09/00 - Table 1eDownload excel
(Excel 15KB) 09/00 - Table 1fDownload excel
(Excel 19KB) 09/00 - Table 1iDownload excel
(Excel 16KB) 09/00 - Table 1jDownload excel
(Excel 16KB) 09/00 - Table 1kDownload excel
(Excel 17KB) 09/00 - Table 2aDownload excel
(Excel 35KB) 09/00 - Table 2bDownload excel
(Excel 18KB) 09/00 - Table 3aDownload excel
(Excel 19KB) 09/00 - Table 3bDownload excel
(Excel 58KB) 09/00 - Table 3cDownload excel
(Excel 17KB) 09/00 - Table 3dDownload excel
(Excel 209KB) 09/00 - Table 3eDownload excel
(Excel 71KB) 09/00 - Table 3fDownload excel
(Excel 17KB) 09/00 - Table 3gDownload excel
(Excel 21KB) 09/00 - Table 4aDownload excel
(Excel 15KB) 09/00 - Table 4bDownload excel
(Excel 33KB) (External) Get Adobe Acrobat reader You may need the free Acrobat Reader to view a pdfFind out what our other sites have to offer