Problems related to packaging inks

13.5.1 Unevaluated substances and degradation products

Unfortunately, the inventory list included in the draft TD 1 to the CoE resolution on packaging inks is incomplete. It appears that many of the numerous substances used in inks for food packaging have not been properly evaluated as regards their toxicological properties. Industry should ensure that there is no detectable migration into food of these unevaluated substances. Very little is known about degradation of substances during printing.

13.5.2 Migration

Prints are almost always applied to the outer surface of a food packaging and are not intended to make direct contact with food. However, low molecular substances in ink migrate (permeate) easily through the packaging material into the food. Only a few packaging materials, such as glass and aluminium foils, are barriers to all substances in the ink. Fibrous materials and most kinds of plastics do not act as barriers to migrants. In particular solvents travel easily through packages made of paper, board or plastics. In the case of polyethylene laminated board, the plastic film acts as a barrier to water, but not to fat-soluble substances. Some recent works on migration of ink components into food are mentioned below. Further references can be found in a comprehensive Norwegian study on colourants prepared by Brede et al. in 2003 (see section 13.7.3 below).

Johns et al} have published a report on benzophenone in printed cardboard for food stored frozen. This substance was found at 0.4-3.0 mg/dm2 in four board samples out of seven. The authors suggest that UV-cured inks had been used with benzophenone as an initiator in the printing of these boards. Three of the corresponding foodstuffs contained benzophenone at a level of 0.6-2.9 mg/kg, though there was a polyethylene layer applied to the board. Model ink substances were added to the carton, which then was in direct contact with foodstuffs stored at -20 °C for a year. It was found that transfer of the volatile substances could be considerable even at low temperatures.

Aurela et al.2 studied migration of alkylbenzenes used as solvents in offset inks. Alkylbenzenes migrated from a hamburger collar into a roll at the level of 2 mg/kg of food. In a risk assessment report published by the European Commission,3 a NOAL (no observed adverse effect level) value of 50 mg/kg body weight per day was used. Thus, the migration from the printed hamburger collar is quite acceptable.

13.5.3 Set-off and curing

Invisible set-off is not easily detected at the printing house and some printers are not even aware of its existence. A study on plasticisers by Aurela and Ohra-aho4 showed that the set-off phenomenon considerably increased migration into food when the substrate was a high barrier material and migration through the substrate was low. In order to avoid set-off, emphasis should be put on drying of the print. Inks containing a high proportion of solvents, such as gravure and flexographic inks, are typically dried by evaporating the solvent in an air flow, for example when plastic films are printed. Inks for sheet-fed offset, dry by oxidation and/or polymerisation. As regards corrugated board, drying is no problem as the high mass of fibrous material readily absorbs the solvent. The time required for absorption depends also on the capillary structure of the substrate surface.

Set-off can be avoided by the use of anti-set-off compounds. These are spray powders that reduce the frictional contact between the sheets. Materials like silica and starch, which have a particle diameter slightly greater than the printed ink film thickness, can be used. Slow setting problems may be reduced by the use of less solvent or by a higher ratio of resin to oil. Varnishing with a formula that utilises polymerisation induced by radiation is another way to avoid set-off. The varnish covers the printed surface and direct contact between the layer of print and the inner food contact surface of the substrate is avoided.

UV-radiation is used to achieve rapid curing and to avoid set-off effects in offset printing. The binders in UV-inks are highly reactive acrylate monomers and oligomers to which photoinitiators are added. These initiators start the cross-linking reaction in response to UV-radiation. Photoinitiators are low molecular weight substances, such as benzophenone. Dry prints still contain residues of photoinitiators or their degradation products. The initiators and their degradation products may cause an undesirable odour or migrate into the packed food. The low molecular products can be removed by careful airing. Photoinitiators are intensily studied in order to avoid problems. In a recent study, Bedolla5 has shown that on flooding the curing environment with nitrogen, the amount of photoinitiator can be significantly reduced without loss in curing efficiency. Electron beam (EB) curing might also become more common. Its advantage is no need for any initiator, but high costs might restrict its use. Cationically cured UV inks are based on cycloaliphatic epoxy resins. They contain photoinitiators that break down into acids. However, they are unsuitable for offset printing for technical reasons. This type of ink is often used when printing labels.

13.5.4 Sensory contamination, odour and taint (off-flavour)* Several substances in the ink can have a noticeable smell or they can penetrate through the substrate and cause taint problems to the packed food. Food producers should therefore regularly check the deliveries of packages before accepting the material, in particular when a new grade is introduced. Solvents often contain malodorous substances. Vegetable oils used in offset printing have a tendency to oxidise. The unsaturated fatty acids undergo auto-oxidation reactions resulting in hexanal and other volatile compounds. The reaction is catalysed by the presence of heavy metals, such as iron or copper.

Binders, mainly alkyd resins, in inks drying by oxidation may also be a source of hexanal and other aldehydes. In a study by Landy et al.6 on the odour of paper-based packaging materials printed by the offset process 13 odorous substances were identified, ten of them being aldehydes and ketones resulting from oxidation of printing ink resins. Mineral oils are high boiling point, aliphatic hydrocarbons. Some mineral oils contain aromatic compounds in order to increase their dissolution power. The aromatic compounds can easily diffuse through a fibrous or plastic material migrating into the foodstuff. Aromatic substances like toluene and xylene should be avoided when food packages are printed. Hydrogenation of aromatic hydrocarbons leads to ringlike aliphatic substances. These have a high dissolution capacity, but less smell and lower toxicity compared to aromatic solvents.

Low-odour inks for food packaging are based on odour and taint-free substances, for example, aromatic free solvents and maleic resins. So called semi-drying oils are also used. These are slow to oxidise and therefore do not develop malodorous aldehydes as they dry. However, some smell may be noticed after a long period of time as oxidation proceeds. other solvents, such as tung oil, develop quite some odour during drying but if properly

* The vocabulary varies. Off-flavour is an atypical flavour often associated with deterioration of the product. Taint is often used as a synonym for off-flavour.

aired they are free of odour when oxidation is complete. Water-based inks are intensely studied by ink manufacturers in order to avoid sensory problems and some ink producers market so called low-migration and low-odour offset inks intended for food packaging. These are based on non-oxidative curing.

13.5.5 Print on food contact surfaces

Publications dealing with substances on food contact surfaces are scarce though the use of direct contact prints seems to increase, at least in the UK. Bradley7 has investigated the substances used in this kind of print on plastics. The results were reassuring. None of the substances detected had the potential to migrate into food above their specific migration limits (SMLs). For substances not having an SML, 10 mg per kg of food was set as a 'level of interest'. By worst case calculation it was found that only one substance, 2-ethylhexyl acrylate exceeded this limit and only by a small margin. The author states that practical tests would probably have resulted in migration below the level of interest. She recommends that inks intended for direct contact with foodstuffs should be controlled as coatings rather than as packaging inks.

13.5.6 Conclusion

Quality and safety risks can be reduced only by selecting harmless raw materials for the ink, and an appropriate printing process that ensures sufficient curing.

13.6 Testing 13.6.1 General remarks

The ink to be tested should be printed on the same substrate and by the same procedure that is intended for the final product. The ink should not be tested as such because its composition may change during the printing process. Test pieces of a printed material should be selected so that all components of the ink are represented at the same percentage as in the original material. When possible, migration testing should be carried out using the same kind of food that is intended to come into contact with the printed material. If this is not practical, certain food simulants should be used. In situations where migration limits exist for the finished product, only the migration level and not the origin of the migrated substance will affect the compliance decision. Consequently, there is no need to differentiate between migration caused by transfer through the substrate and by set-off.

13.6.2 Sensory testing

Sensory analysis is a subjective sensation by the sense organs. Assessors perform sensory analyses. For a packaging material, transfer of taint is the critical sensory property from a legal point of view. Odour is less significant, but equally important as a marketing consideration. A panel of trained assessors evaluates the sensory properties. The number of assessors is based on the sensitivity desired for the test; fewer assessors result in less statistical reliability. A well-trained panel will give consistent results.

For evaluation of the odour of a printed material, paper, board or plastic, test pieces are stored in jars for a certain time. The odour of the air is then evaluated by the panellists. The intensity of the odour is rated on a scale such as 0 (no odour) to 4 (strong odour). For evaluation of taint, test pieces are stored with a test food in a jar. (One version of this test is known as the Robinson test.) The taint of the food is then evaluated and rated by the panelists e.g., as a triangle test or a multicomparison test. In the triangle test the assessors should select the odd test food out of three test food portions. One has been stored together with the material to be tested and two are reference samples, or vice versa. In the multicomparison test, the assessors receive one known control sample that holds the value of 0. The assessors score the intensities of the taint of the analysis sample compared to the control sample. The scale indicating the strength of taint is usually from 0 (no taint) to 4 (strong taint). Figure 13.1 shows experimental set-ups for odour and taint tests.

Though there are no specific international standards for assessing odour and taint caused by the ink, a general standard for assessing modifications to the flavour of foodstuffs due to packaging has been issued by ISO. There is also a specific standard for paper and board and some standards on sensory analysis in general. See section 13.7.4 below.

Conventional sensory tests are time consuming and require a well-motivated panel. Much work has therefore been done to replace panel evaluations by instrumental methods. An instrument called an electronic nose seems soon to be applicable for routine analysis within the food and the packaging industry. It is based on a chemical sensor array. An investigation by van Deventer8 on volatile chemicals from inks on plastic films has shown that the quartz sensors of an electronic nose system were able to discriminate between packages with different levels of retained solvents. In a project partly funded by the European Union, the Parfum-Project, an instrument based on an array of eight quartz crystal microbalances (QMB) and eight metal oxide (MO) sensors was investigated. Frank et al.9 showed that there was a very satisfying correlation between the prediction of the 'nose' and human sensory assessments as regards retained solvents in printed wrapping foils. This project has been followed by another European project, ESCAPE (Electronic Sensor System for the Characterisation of Packing Emissions) with the intention to develop a rapid instrumentation system that combines sampling with sensor arrays for monitoring of residual solvents and unwanted odours of packaging materials.

Fig. 13.1 Experimental set-ups for (a) odour and (b) taint tests of a packaging material.

Fig. 13.1 Experimental set-ups for (a) odour and (b) taint tests of a packaging material.

Instruments based on sensor arrays will become tools for practical application and should enable rapid screening of printed materials directly connected to the production process. Figure 13.2 shows a commercial electronic nose and a human panellist.

Fig. 13.2 Two kinds of 'noses' at work. (a) The electronic nose can smell tens of samples during the night, while (b) a human nose fatigues more easily.

Fig. 13.2 Two kinds of 'noses' at work. (a) The electronic nose can smell tens of samples during the night, while (b) a human nose fatigues more easily.

13.6.3 Migration testing

Migration tests are designed to measure the quantity of substances that are likely to transfer from a packaging material to the food. In principle, testing for compliance with established specific migration limits should be carried out using the conditions established in actual EU controls on migration testing. It is, however, extremely difficult in many cases to achieve the limits of detection required by legislation as a limit might be as low as 10 mg of substance in 1 kg of food. Analytical means are still under development. For practical reasons certain specified food simulants are generally used to replace actual foodstuffs.

In the migration test the specimen to be tested is brought into single-sided contact with a food simulant. The conditions (time/temperature) are chosen depending on the final use of the printed material. After exposure the simulant is analysed or extracted, and the extract is analysed for example by gas chromatography and mass spectrometry (GC/MS). The test was originally developed for plastic materials, but it can be applied also to other printed or unprinted materials. In the case of paper and board, the standard liquid simulants are not suitable, as they penetrate into the material. They are therefore replaced by modified polyphenylene oxide (Tenax) according to a European standard EN14338. Figure13.3 shows the experimental set-up for migration test using Tenax.

Papilloud and Baudraz10 studied the official European simulants A, B and D, which are to be used in migration tests. Simulant A, water, mimics liquid foods; simulant B, 3% acetic acid, mimics acidic liquids; and simulant D, olive oil, mimics fatty food but is usually replaced by iso-octane or 95% ethanol for practical reasons. Transfer into these simulants of some major acrylates and photoinitiators used in UV-printing was investigated. The printed materials were kept in one-sided contact with the simulants in migration cells made of stainless steel. The aqueous solutions were concentrated by solid-phase extraction, the organic solutions by evaporation. The determination of the migrants was made by GC/MS. Reproducibility and accuracy were good both for the acrylates and the photoinitiators at levels above 10 mg/l. However, the limits of detection were not quite satisfactory, as most real samples would have levels of migrants too low to be detected by present means.

In another publication by the same authors11 a method was developed for quantification of migrants from UV cationic printed materials, in aqueous simulants. Quantitative analyses were performed by HPLC. The limits of detection were 2.7-17 mg/l for the investigated substances, except for the epoxy monomer. The corresponding limits of quantification were 8.9-58 mg/l. These figures show the need to develop analytical means further.


Petri dish

Extract of Tenax is analysed with GC/MS


Fig. 13.3 Experimental set-up for a migration test using Tenax as a food simulant according to EN14338.

13.6.4 Testing of materials for use at elevated temperatures

For migration testing of printed materials intended for use at elevated temperatures only Tenax is used as food simulant. Testing should take into account possible degradation products formed at elevated temperatures. Prior to testing, the sample should be preheated in a closed container.

13.6.5 Worst-case calculation and modelling

Migration testing can be replaced by so-called worst-case calculation. The amount of the actual substance applied on the surface of the substrate must be known. It is then assumed that 100% of the substance migrates into the food to be packed. An example of the calculation is presented in the CoE document on test conditions for packaging inks (TD 3). Mathematical modelling or simulation can be used to predict the migration of a substance and to verify compliance with migration limits. Though in migration testing the result represents the sum of a substance transferred through a substrate and the possible set-off, this is not the case when using diffusion models for predicting migration. The transfer resulting from set-off must be estimated separately. Some recognised diffusion models are available: see Chapter 8.

13.6.6 Investigation of set-off

For investigation of the tendency for set-off of a specific substance, a freshly printed surface of a material is brought into contact with an unprinted sheet of the same material under pressure. The unprinted sheet should then be analysed, for example, by extraction and GC/MS. A method to identify spots of invisible set-off of inks and lacquers on the food-contact surface of a food packaging material has been developed by Bradley et al.12 The authors use optical means to excite and observe luminiscence from invisible set-off. In their model experiments with several resins applied on different substrates they have achieved a level of detection of 20 mg per cm2 of sheet surface. This approach might gain wider use in the future.

13.6.7 Analysis of inks and prints

Inks and prints are analysed in order to ensure safety. This includes testing for harmful substances with potential to migrate into food. Environmental pressure may result in a demand for testing, for example, vegetable oils versus petroleum distillates. There might also be a need to determine the content of aromatic compounds in the ink in order to avoid odour and taint problems. For chemical analysis of prints and determination of ink components a number of methods are available, such as pyrolysis, infra-red spectrometry, gas chromatography and mass spectrometry. Volatile compounds are usually analysed by a headspace technique. The progress in chemical analysis is so rapid that any method may be considered obsolete after a limited number of years. It is therefore recommended to search the literature in order to find an appropriate method.

13.6.8 Olfactometric analysis

In a recent study by Landy et al.5 on odorous substances in paper-based packaging materials, the main volatile substances that caused odour of an offset printed product were identified by olfactometry. In this technique solvent-free extraction using microfibres was successfully applied. The extract was introduced in a gas chromatograph connected to a sniffing port. The odour was then evaluated by a trained assessor.

13.6.9 International standards for testing prints and inks

There are no specific international standards for packaging inks dealing with the determination of ink components or their migration. Some guidance may be found in the standards dealing with substances in paper and board and in plastics, referred to in section 13.7.5.

0 0

Post a comment