Studying Migration of Packaging Components into Food

The potential of food packaging components to migrate into food is an important health concern. Perfecto Paseiro Losada and his group at the University of Santiago de Compostela, in Spain, have been investigating the migration kinetics and actual migration of such compounds into a variety of types of food. They also have been carrying out studies to estimate dietary exposure. Paseiro recently spoke to LCGC about this work.

Laura Bush

You conducted a study on the identification of intentionally added substances (IAS) and non-intentionally added substances (NIAS) in plastic food packaging materials and their migration in food products (1). First, for those unfamiliar with the terms, can you explain the difference between these two types of substances? How do you determine what substances are important to quantify?

Both terms have their origin in the current European legislation on plastic materials for food contact (Regulation 10/2011). Only the substances included in the European Union (EU) list of authorized substances (such as monomers or additives) may be intentionally used for the manufacture of food contact materials, and these substances are commonly known as IAS. An example of an IAS is bisphenol A, used for the synthesis of polycarbonate or erucamide used as a slip additive in polyolefins. In principle, safety of IAS is evaluated before authorization, and they are subject to restrictions, so that their migration into food does not endanger human health.

NIAS are defined as “an impurity in the substances used or a reaction intermediate formed during the production process or a decomposition or reaction product.” NIAS are not in included the EU List, but they may be present in the plastic materials, and must be assessed in accordance with internationally recognized scientific principles of risk assessment.

In the study, you used a non-targeted approach with gas chromatography–mass spectrometry (GC–MS) to identify compounds in the plastic packing materials. Why did you use GC–MS, and why a non-targeted approach? Are you able to detect and quantify the most important compounds using this method?

We focused the research on trying to detect any volatile or semivolatile substances present in plastic packaging samples, and GC–MS is the most appropriated technique. We use two approaches; dynamic headspace sampling and purge-and-trap for volatiles, and splitless injection mode after sample liquid extraction for semivolatiles.

The two approaches complement each other, and they give a very complete view about what volatile and semivolatile substances are present in the packaging, thus obtaining very useful information on what substances could potentially migrate to food. About 100 volatile and semivolatile compounds were detected using the two techniques.

Migration tests were carried out using Tenax and isooctane. Through those tests, 27 compounds were detected, and their relative amounts were estimated against an internal standard.

What performance were you able to achieve with the methods used in the study, in terms of limits of detection and repeatability? How did you verify this?

Method performance was very good, in terms of linearity, recovery, repeatability, and limits of detection and quantification. Recovery in foods (corn snacks, potato snacks, cookies, and cakes) was nearly 100% for most of the selected compounds, with a range between 82.7 and 116.1%, and the relative standard deviation (RSD) derived from the replicate concentrations measured in spiked foodstuffs (n = 6) was less than 9% for most of compounds, with range of 2.22–15.9%. Most of LODs were less than 0.003 mg/L.

What did you find in terms of which compounds migrated into food, and at what levels? Why are these compounds important as related to human health? Are these compounds known to be harmful?

All compounds selected-bis(2-ethylhexyl)phthalate (DEHP), diethyl phthalate (DEP), diisobutyl phthalate (DIBP), dibutyl phthalate (DBP), butylated hydroxytoluene (BHT), acetyl tributyl citrate (ATBC) and benzophenone (BP)-were previously identified in packaging materials. All compounds were found in at least some of the 34 analyzed food samples. ATBC was the most common (in 94% of the samples), and BHT the least common (in 12% of the samples).

The highest concentrations were found were in corn and potato snacks: ATBC (7.09 µg/g), DEP (1.44 µg/g), DEHP (0.57 µg/g), BP (0.2 µg/g), DBP (0.77 µg/g), DIBP (1.51 µg/g) and BHT (6.58 µg/g). DEP and DEHP were the most frequently detected phthalate compounds in the food samples; the specific migration limit (SML) of 0.3 mg/kg established in Regulation 10/2011 was exceeded for DBP in one sample.

Phthalates are a group of chemicals of current concern for human health. They are known to be endocrine disruptors that affect the production of hormones, especially testosterone, and some studies associate them with infertility, obesity, and asthma.

ATBC is a widely used plasticizer to substitute for phthalates. BHT is an antioxidant also used as food additive, and BP is a photoinitiator, but, at the levels found, they do not represent a health concern.

You carried out a related study of the nontargeted analysis of IAS and NIAS and their migration into food simulants, using purge-and-trap GC–MS for volatile compounds, and extraction with organic solvents followed by GC–MS for semivolatile compounds (2). Why are food simulants used in a study like this? And what food simulants did you choose and why? How do you know that such studies correlate closely with the real-world conditions for packaged foods?

A food simulant is a test medium imitating food; in its behavior, the food simulant mimics migration from food contact materials. Simulants are much simpler analytical matrices than food; their use facilitates identification and quantification of migrants, and more reliable information is obtained about what substances migrate, or may migrate, to food. The nontargeted analysis we carried out would be very difficult to execute in the various food matrices.

Tenax is a food simulant for testing specific migration into dry foods, and isooctane is one of the well-known food simulant substitutes for fatty foods.

What conclusions have you drawn from your studies of the migration of packaging compounds into foods, and the perceived potential harm these of compounds to human consumers?

Many volatile and semivolatile chemicals are present in the analyzed packaging samples, including both IAS and NIAS, some of which are of very high concern (such as 2,4 and 2,6 toluene diisocyanate), although most of them did not migrate to the selected food simulants. In our opinion, the levels found in the studied foods and food simulants were low. However, for a complete estimation of chemical exposure other types of packaging and food must be considered and assessed by experts in risk assessment.

You have also studied the migration of two common components of UV-curing inks-known as photoinitiators-into food simulants (3). Why are these compounds of particular concern in terms of their health risks?

The Rapid Alert System for Food and Feed (RASFF) has reported many cases of different components of UV curing inks in foodstuffs in recent years. Photoinitiators are constituents of many printing inks applied on the non-food-contact side of food packaging. Photoinitiators may reach the food contact side, among other ways, by set-off (such as transfer of wet ink from the substrate film to another surface of the plastic film that comes in contact with the food). During storage these inks may also penetrate, by diffusion, into the internal film layer intended to come into contact with foods, which is usually made of PE. After the internal film layer has come in contact with food, the photoinitiators may migrate into the food.

The main reason for this research was to characterize the process of migration of two common UV ink components from PE into food simulants.

What type of mathematical modeling did you use in the study? Why was this specific approach selected?

We used a model based on Fick’s second law, specifically the solution proposed by Crank for diffusion in a plane sheet from a stirred solution of limited volume. This solution is broadly accepted as a model for the migration of a substance from a plastic layer into a well-mixed liquid.

What did the study reveal about the migration kinetics of these compounds at the four storage temperatures studied? Does the food composition itself determine the migration kinetics?

We determined key parameters of migration (diffusion coefficients and partition coefficients) for 4-methylbenzophenone (a photoinitiator) and ethyl-4-(dimethylamino) benzoate (a coinitiator) from LDPE by fitting the experimental data with the mathematical model based on Fick´s second law. The model may be used to predict the migration process of those migrants. Furthermore, key parameters of the Arrhenius equation (activation energy and pre-exponential factor) were estimated and they can be used to calculate diffusion coefficients at any temperature between 40 ºC and –4 ºC.

We also studied the migration at freezer storage temperature (–18 ºC) into 50% and 90% ethanolic simulants, because very scarce information on migration kinetics at that temperature has been reported. Results showed that migration also happen in a great extension into both simulants.

In this study, foods were not included the composition and physico-chemical properties of each food affect the migration kinetics, especially the partition coefficient (polymer/food). On the other hand, for many foods, the diffusion coefficient of the migrant into the food should be estimated and included in the model, a circumstance that in the case of simulants is not considered significant.

You have also carried out research to estimate dietary exposure to packaging contaminants among the Spanish population from cereal-based foods (4). What foods and compounds did you study, and why? Is packaging compound migration a key health concern, more so than contamination of foods from pesticide residues?

Cereal-based foods are among the most consumed among the population groups studied. Representative food sample pools for each age group were prepared by combining amounts of rice, bread (toasted and not), and alternatives to bread, pasta, and breakfast cereals, according to consumption data obtained from Spanish national dietary survey on children and adolescents (ENALIA).

The chemicals selected comprise a wide range of substances, all of them previously identified in food packaging, mainly plasticizers such as phthalates (dibutyl phthalate, diisobutyl phthalate, diethyl phthalate, and benzyl butyl phthalate). Other substances include citrates (ATBC), adipates (DEHA), UV stabilizers (octocrylene), and slip agents (erucamide).

Chemical migration from packaging to foods is an inevitable process. The key point is to ensure that the amounts of substances that migrate to food do not endanger human health. When this principle is not achieved, then there is a health concern. The important thing is that the type and amount of chemical migrants from packaging, pesticides, or other chemicals in foods do not generate a health concern.

What were the analysis conditions of the GC and LC methods used in the study?

For GC, a ZB-5MS (30 m × 0.25 mm x 0.25-μm) column, splitless injection mode and oven temperature from 40 to 300 ºC. For MS, full scan mode (m/z range of 35–500). For HPLC, a Kinetex biphenyl column (100 mm × 3 mm x 2.6-μm) at 30 ºC with a mobile phase composed of methanol and water, both containing 0.1% (v/v) formic acid, gradient elution method from 30% water and 70% methanol to 100% methanol was used. For MS/MS, positive ESI mode and for each compound precursor ion was selected and two product ions, one for quantification and other for qualification purposes.

How did you optimize the sample preparation or extraction procedures for both the GC–MS and the LC–MS/MS methods?

In comparative extraction studies, acetonitrile got better recoveries for all analytes, and it was selected as the extraction solvent; the extraction time and solvent concentration steps also were optimized, especially the latter to avoid irreproducible results and loss of some analytes.

What further research would you propose in the field of migration of packaging chemicals into foods? Would you propose a healthier form of packaging and would food companies be receptive to changing their packaging methods?

From an analytical point of view, it is necessary to develop methodologies that facilitate the detection, identification, and quantification of any substance that migrates to food, especially for non-volatile compounds. There is still a lot of research to be done.

Many chemical substances of unknown or variable composition, complex reaction products, and biological materials (UVCB substance) are used for the manufacture of food contact materials (plastics, coatings, inks, adhesives, paper and paperboard, etc.), among them many resins containing prepolymers with reactive oligomers with a MW less than 1000 Da, therefore they are chemical hazards of concern.

In the framework of EU plastics regulation, prepolymers are authorized generically if they are used as starting substances and are synthesized from monomers already included in the EU list. These substances have not been evaluated before authorization, and they are not included as other not-listed substances (such as NIAS or aids to polymerization) to be assessed in accordance with internationally recognized scientific principles on risk assessment and they may migrate to foods without specific restrictions.

The positive identification, quantification, and safety assessment of hundreds, probably thousands, of unknown substances is a huge challenge in this field, because for most of them there are no in-standard MS-libraries and analytical standards are not available.

Polymeric materials used in food contact packaging have solved many food safety problems of the past, but they have also generated new problems related to the migration of chemical substances to food. At present, there is no realistic alternative to the use of polymeric materials in food packaging. Updated legislation is needed that does not generically authorize substances that have not been previously evaluated; we also need enforcement to ensure compliance with the legislation.

References

1. V. García Ibarra, A. Rodríguez Bernaldo de Quirós, P. Paseiro Losada, and R. Sendón, Anal. Bioanal. Chem. 410, 3789–3803 (2018). https://doi.org/10.1007/s00216-018-1058-y
2. V. García Ibarra, A.Rodríguez Bernaldo de Quirós, P. Paseiro Losada, and R. Sendón, Food Pack. Shelf Life, 21, 100325 (2019).
3. M.A. Lago, R. Sendón, J. Bustos, M.T. Nieto, P. Paseiro Losada, and A. Rodríguez-Bernaldo de Quirós, Molecules 24, 3607 (2019). doi:10.3390/molecules24193607
4. V. García Ibarra, R. Sendón, J. Bustos, P. Paseiro Losada, and . Rodríguez-Bernaldo de Quirós, Food Chem. Toxicol. 128, 180–192 (2019). https://doi.org/10.1016/j.fct.2019.04.003

Perfecto Paseiro Losada is a Professor of Analytical Chemistry, Nutrition, and Bromatology within the Faculty of Pharmacy at the Universidade de Santiago de Compostela, in Santiago de Compostela, Galicia, Spain.