Published data on migration from laminating adhesives

16.4.1 Aromatic amines

Lawson (1994) examined levels of aromatic amine migration from a number of laminate samples. Specially prepared laminate pouches containing distilled water were boiled for one hour and tested for primary aromatic amine (PAA) migration using a diazotisation procedure. Olive oil migration data at the same temperature was also obtained. Results are shown in Table 16.5 (detection limit 0.3 mg dm-2). These results, if expressed using the conventional surface area to volume ratio of 6 dm2 per kg of food, would range between 3 and 11.4 mg/litre (assuming a density of 1).

Danish aromatic amine survey

In August 2001, following reports in the Danish press about the possible contamination of packaged foods with PAA, the Danish Veterinary and Food Administration instituted a pilot study (Trier and Petersen 2001a) and also a more detailed examination of the potential contamination of food with PAA from laminated packaging (Trier and Petersen 2001b). In the pilot study, no detectable migration of PAA was found from any of the samples analysed. In the more detailed examination, both a quick test (as used in the pilot study) and also standard migration testing using appropriate test simulants were undertaken. The PAA migration from 33 samples was analysed in the more detailed study. In only two samples, the PAA migration was found to be higher than the detection limit of the accredited method of 1.1 mg aniline/kg

Table 16.5 Aromatic amine migration from polyurethane adhesives (Lawson 1994)

Laminate Migration mg dm-2 as MDA

Distilled water

15 mm LLDPE/15 mm HDPE* 15 mm HDPE/15 mm LLDPE* 50 mm Nylon/50 mm LLDPE* 50 mm Nylon/15 mm LLDPE* Retail

0.91 ± 0.21 1.03 ± 0.47 1.10 ± 0.20 0.49 ± 0.13 0.69 ± 0.24

Olive oil

50 mm Nylon/15 mm LLDPE* 50 mm Nylon/50 mm LLDPE*

* Side in contact with simulant food simulant (calculated using the convention of 6 dm2 of food contact material per one kg of packaged foodstuff). A further eight samples showed migration above an 'internal' detection limit of 0.2 mg aniline/kg in acetic acid tests and 0.5 mg aniline/kg in distilled water. The detection of PAA migration was thought to probably reflect that the samples had been taken for analysis shortly after the lamination process when compared to the pilot study. However, in all migration tests performed, the resulting PAA migration was far below the current EU migration limit of 20 mg aniline/kg food simulant (EC 2002).

Co-authors of these studies have recently published a method for the specific determination of 20 primary aromatic amines in aqueous food simulants by liquid chromatography-electrospray ionisation-tandem mass spectrometry (Mortensen et al. 2005).

16.4.2 Migration of polyol components

Lawson et al. (2000) examined the migration of constituents from solventfree adhesives used to bond 12 mm PET film to 45 mm LDPE. The technique of MALDI-MS, a soft ionisation technique capable of looking at sample mixtures over a mass range of 150-500,000 Da without prior separation, was employed. The adhesives studied were based on a solution of mixed isomers of MDI in polymeric MDI with either polyether or polyester-based polyols. Pouch testing of cured laminates with distilled water was undertaken (two hours at 70 °C) with the LDPE surface in contact with the water.

Migration of unreacted polyol components through the polyethylene for the polyether-based laminate was observed. Although an excess of isocyanate was present, diffusion of polyol components into the polyethylene prior to reaction with isocyanate was postulated to explain the migration. Cyclic oligomers from the polyol starting materials were identified as the main migrants from the polyester-based adhesive.

16.4.3 Research for the UK Food Standards Agency

Rapra Technology Ltd has undertaken a detailed investigation into the migration of species from different types of adhesives used with laminated multi-layer materials (Barber et al., 2003). Migration of aromatic amines, BADGE, bisphenol A and polyols were examined. A description of the samples used and a brief summary of some of the results obtained from this analytical work are given below.

Solventless MDI based adhesive systems

Sample 1 - oxygen barrier packaging used for lidding films for meat and cheese

20 mm orientated polypropylene

Laminating adhesive (coating weight =1.60 g/m2)

40 mm LLDPE/Tie/EVOH/Tie/LLDPE coextrusion (food-contact surface).

Sample 2 - lidding film packaging for meat or cheese

12 mm PET(PVDC coated on inner surface) Laminating adhesive (coating weight = 1.65 g/m2) 38 mm LLDPE (food-contact surface).

Sample 3 - used for packaging rice and cereals in vertical form fill and seal packs

12 mm PET

Laminating adhesive ~2 g/m2 70 mm LLDPE (food-contact surface).

The laminating adhesive for Samples 1-3 was a 100:25 mix of an isocyanate terminated polyurethane prepolymer based upon polypropylene glycol and MDI (amount of free monomer approximately 25%) and a tri-functional polypropylene glycol with a molecular weight of approximately 450.

Sample 4 - packaging for washed lettuce 15 mm OPP

Laminating adhesive (two-component solvent-free laminating adhesive: 100 parts of MDI based prepolymer, 35-50% free monomer) 20 mm CPP (food contact surface).

Sample 5 - packaging for smoked fish

12 mm PET PVDC coating Ink

Laminating adhesive (solvent-based MDI-based polyether polyurethane 2.5 g/m2 containing 1-5% 4,4' MDI) 70 mm polyethylene (food contact surface).

Sample 6 - packaging for sweets

19 mm polypropylene Ink

Laminating adhesive - water based epoxy amine, 2.5 g/m2

(100 parts polyurethane resin including some (<2%) aminoethylethanolamine,

5 parts epoxy resin, average MW <700)


20 mm polypropylene

Laminating adhesive 2.5 g/m2 (solvent-based two-part MDI polyester polyurethane), resin solution 1 containing 1-5% 4,4' MDI, solution 2 including 3-aminopropyltriethoxysilane 38 mm polyethylene (food contact surface).

Solventless epoxy amine laminates

Sample 7 - packaging for crispbread Clay coated paper

Laminating adhesive - epoxy amine (polyurethane backbone) 20 mm polypropylene (food contact surface).

Sample 8 - packaging for wrapped sweets

25 mm polypropylene Ink

Laminating adhesive (water based epoxy amine) 2.5 g/m2

100 parts polyurethane resin including some (<2%) aminoethylethanolamine

5 parts epoxy resin, average MW <700

50 mm polyethylene (food contact surface).

IPDI based solventless adhesives

Sample 9 PET/adhesive/PA/adhesive/PP (microwaveable pouch)

12 mm PET Ink

Laminating adhesive (IPDI based aliphatic) 4 g/m2 Silicon oxide coating 15 mm polyamide

Laminating adhesive (IPDI based aliphatic) 4 g/m2 70 mm polypropylene (food contact surface).

Migration of MDA from Samples 1-6

Migration was studied using the food simulant 3% acetic acid for two hours at 70 °C. Acetic acid is generally considered to be the worst-case simulant for primary aromatic amines. A standard photometric procedure involving diazotisation in hydrochloric acid solution, coupling with N-(1-naphthyl)-ethylenediamine dihydrochloride, concentration using solid phase extraction columns and colorimetric determination at 550 nm was first used to measure migration. Test pouches were prepared of dimensions 200 mm x 100 mm filled with 100 ml of test simulant prior to sealing with minimal air space. Known deficiencies of this colorimetric method include possible interference from other extracted components and variation in the maximum wavelength of absorption of different amine complexes in comparison with the 550 nm stated in the method (wavelength for the aniline standards). The 2,4' and 2,2' MDA isomers also give a lower percentage response than the more reactive 4,4' isomer. Later in the studies the technique of liquid chromatography-mass spectroscopy (LC-MS) was employed. Levels of extracted MDA depended on the length of time following lamination and the sample storage condition and environment as demonstrated for Samples 3 and 4 in Tables 16.6 and 16.7.

Table 16.6 PAA results obtained on Sample 3 (colorimetric method)

Test material


PAA migration

PAA migration

PAA migration


mg/100 ml/4

mg aniline/kg

mg 4,4' MDA/kg


dm2 as aniline hydrochloride (as 4,4' MDA)

(6 dm2/kg)

(6 dm2/kg)

4 days after


3.1 (2.4)




10 layers into roll

4 days,


3.3 (2.5)



10 layers into roll


7 days,


1.8 (1.2)



10 layers into roll

10 days,


2.2 (1.6)



1 layer into roll

14 days, outer


2.0 (1.4)




14 days,


1.6 (1.1)



2 layers into roll

14 days,


1.9 (1.3)



10 layers into roll

23 days,


2.0 (1.5)



10 layers into roll

Table 16.7 PAA results obtained on Sample 4 (colorimetric method)

Test material Maximum PAA migration colour mg/100 ml/4

absorbance dm2 as aniline hydrochloride (as 4,4' MDA)

PAA migration mg aniline/kg (6 dm2/kg)


10 layers into roll


10 layers into roll

Note that Sample 3 had a 70 mm LLDPE layer between the acetic acid and the laminating adhesive. Sample 4 had a 20 mm layer of cast polypropylene between the simulant and adhesive. Both Samples 3 and 4 were provided to Rapra on a roll, which was stored at ambient temperature in the laboratory prior to removal of material for testing.

The primary aromatic amine (PAA) migration was determined using both aniline hydrochloride and 4,4' MDA standards. Results are expressed as mg/ 100 ml/4 dm2 as aniline hydrochloride, as mg/100 ml/4 dm2 as MDA (using the MDA calibration), as mg aniline/kg (using the 6 dm2/kg convention) and as mg MDA/kg (using the 6 dm2/kg convention and the MDA colorimetric data). Data can be viewed against the EU migration limit for PAA of 20 mg/ kg expressed as aniline (EC 2002).

Examination of MDA migration using LC-MS

The following conditions were employed for the examination of spiked solutions and sample extracts.

Instrument: Agilent 1100 LC/MSD model SL

Column: Aqua 3 mm C18 125A, 150 x 2.0 mm, (Phenomenex)

Flow rate: 0.5 ml/minute

Mobile phase A: Acetonitrile

B: 5 mM ammonium acetate in distilled water Timetable: Time %B

(minutes) 0 92 5 89 20 10 MSD Electrospray +ve*

SIM ion 199

Fragmentor voltage: 80 V Gas temperature: 350 °C Drying gas: 10 l/min Nebuliser pressure: 40 psig Capillary voltage 4000 V

A cation exchange clean-up of the method was employed in the examination of samples. Amines were eluted from the clean-up column using 70 vol% 0.1M sodium citrate buffer (pH 2.5) + 30 vol% methanol. This solution was then injected into the LC-MS. A slow drop-off in MS response was observed using this procedure. In later studies, delaying the introduction of the LC mobile phase into the mass spectrometer for three minutes (i.e. not introducing any ionic non-retained components) overcame this problem.

Levels of specific migration of the different MDA isomers from Sample 3 are detailed in Table 16.8. For Samples 1 and 2 and Samples 5 and 6, when tested several months after lamination, levels of MDA migration, calculated on the basis of 6 dm2 of film being in contact with 1 kg of food, were <1 mg/kg.

*Better ionisation of amines was found using electrospray (+ve ion) compared to APCI ionisation.

Chemical migration from multi-layer packaging into food 365 Table 16.8 MDA migration from Sample 3 (two hours at 70 °C)

mg/litre mg/litre mg/litre isomers mg/100

ml/4 dm2

Examination of extractables from Samples 1 and 2 into food simulants using GC-MS

In addition to examining aromatic amine migration from Samples 1 and 2 using the colorimetric procedure, two-hour pouch extracts at 70 °C were also examined by GC-MS. Water and 3% acetic acid test solutions from pouch tests undertaken three days after lamination were examined by GC-MS. The pH of 50 ml of test solutions was adjusted to alkaline by adding 0.1 molar sodium hydroxide solution. Resulting solutions were shaken with 5 ml of dichloromethane in a separating funnel and the layers allowed to separate. The dichloromethane layer was then transferred to a GC-MS vial and examined by GC-MS under the following conditions.

Instrument: Agilent 6890 GC with autosampler and 5973

mass-selective detector Column: SGE BPX50 (50% phenyl polydimethylsiloxane)

Column length: 30.0 metres

Nominal diameter: 250 mm

Film thickness: 0.25 mm

Temperature programme: 40 °C for two minutes and then 20 °C per minute to 300 °C (15 minutes run time) Ionisation: 70 eV electron impact.

No MDA was detected in the water or 3% acetic acid extract solutions from either Sample 1 or Sample 2 in the three-day samples using reverse searching (looking for ions at m/z 198, 106 and 182). However, two species could be identified in the extracts.

Bisphenol A (chemically similar to MDA but having -OH groups rather than amine groups) was identified in the water and 3% acetic extracts from Sample 2. Bisphenol A was not detected in the extracts from Sample 1. Infrared transmission spectra recorded through Samples 1 and 2 showed that the MDI isocyanate was slower to react in Sample 2 and so the bisphenol A could possibly relate to the laminating adhesive. As a result, LC-MS tests were undertaken later in the research, on extracts from Sample 3 (laminated with the same adhesive) specifically looking for bisphenol A - none was detected. Tri-ethyl phosphate, thought to be used as a cure catalyst, was identified in the extracts from both Sample 1 and Sample 2.

Examination of migration of BADGE and its derivatives from Samples 7 and 8

LC-MS was used for the determination of BADGE and its derivatives. The chemical structure of BADGE is shown in Fig. 16.1. The following LC-MS conditions were employed:

Instrument: Injection volume: Flow rate: Pump sequence

Agilent 1100 LC/MSD model SL 5 ml

0.5 ml/min

A: 50/50 methanol/water (v/v) (some initial runs 75/25 methanol water) B: 50/50 ethyl acetate/acetonitrile (v/v) 90% to 0% A, via linear gradient over 25 minutes held for five minutes, reverting to 90% A at 30 minutes. Total run time 30 minutes

Phenomenex Aqua 3 mm C18, 125 A, 150 x 2.00 mm 50 °C

SIM m/z 211 (fragmentor voltage of 290 volts) Gas temperature 350 °C Vaporiser 400 °C

Drying gas 4.0 l/min

Nebuliser pressure 35 psig

Using a high fragmentor voltage of 290 V, BADGE and its derivatives were detected using specific ion monitoring at m/z 211 (Fig. 16.2). Some differences in sensitivity were observed as shown in Fig. 16.3.

Samples 7 and 8 laminated with epoxy amine adhesives were examined for BADGE migration. In order to overcome matrix ionisation effects,

Column: Column temp: APCI -ve: Spray chamber:

Fig. 16.1 Chemical structure of BADGE.
Fig. 16.2 BADGE ion examined (exact mass 211.12).

22500 20000 17500 15000 12500 10000 7500 5000 2500 0

MSD4, TIC, MS File (H0703\08030016.D) APCl, Neg, SIM MSD4, TIC, MS File (H0703\08030017.D) APCl, Neg, SIM MSD4, TIC, MS File (H0703\08030018.D) APCl, Neg, SIM MSD4, TIC, MS File (H0703\08030019.D) APCl, Neg, SIM MSD4, TIC, MS File (H0703\08030020.D) APCl, Neg, SIM MSD4, TIC, MS File (H0703\08030021.D) APCl, Neg, SIM

Frag: 290, Frag: 290, Frag: 290, Frag: 290, Frag: 290, Frag: 290,

5 10 15 20 25 min

Fig. 16.3 BADGE derivatives at 0.5 mg/kg in acetonitrile showing the difference in response for the various derivatives (SIM 211 ion).

fro r sa

CD p

quantifications were made against standards prepared by spiking each food simulant type. BADGE and its derivatives were extracted from olive oil test simulant with acetonitrile. Aqueous simulants were diluted 1:1 with acetonitrile prior to examination.

Some BADGE migration into olive oil was found. Hydrolysed BADGE was found in the aqueous simulants. Levels of migration are detailed in Table 16.9. BADGE HCl and BADGE 2HCl were not detected.

Examination of polyol migration from test samples Samples 5, 6, 8 and 9 in the form of pouches were subjected to ten-day testing at 40 °C with the food simulants distilled water, 3% acetic acid, 10% ethanol and olive oil. The three aqueous simulant extracts (1 ml of each) were diluted 1:1 with acetonitrile prior to analysis. Olive oil extracts (5 g) were shaken and extracted with 3 ml of acetonitrile prior to examination, using the same conditions. No evidence was found from the total ion traces to suggest migration of any unreacted polyols under the ten days at 40 °C test condition.

0 0

Post a comment