|
CFSAN/Office of Food Additive Safety
April 2007
Kristina E. Paquette*
The information and conclusions presented in this book chapter do not represent new Agency policy nor do they imply an imminent change in existing policy.
The FDA approved the first materials intended for use as packaging for irradiated foods (polyolefin films, polystyrene, cellophane, vinylidene chloride copolymers, and others) in 1964. Several other materials were approved for this use during the next four years. Since then, only one material, ethylene vinyl acetate copolymer, was added to Title 21 of the Code of Federal Regulations, in 1989. The recent interest in irradiating meat to eliminate pathogens such as E. coli O157:H7 has resulted in several industry submissions to the Agency regarding new packaging materials, as well as the radiation sources, intended for use during the irradiation of prepackaged food. A brief history of FDA regulation of packaging materials irradiated in contact with food, including a discussion of human exposures to radiolysis products formed in irradiated polymers, will be presented. The evaluation of new packaging materials for irradiated foods will be discussed within the context of FDA’s Food Contact Substance Notification Program.
In the 1960s, the FDA approved many packaging materials for use during the irradiation of prepackaged food with only two additional materials receiving approval since (see below for details). Most of the approvals were obtained by the U.S. Army and the U.S. Atomic Energy Commission (AEC) (1). These agencies shared responsibilities under the U.S. Government’s Atoms for Peace program to develop peaceful uses for nuclear technology. The U.S. Army was particularly interested in radiation-sterilization to add to the arsenal of methods for producing shelf-stable foods for the military (2, 3). Why, after 40 years, is there sudden industry interest in obtaining FDA approval for new packaging materials for use during the irradiation of prepackaged food that is intended for consumption by the general public? The answer can be summed up in two words: emerging pathogens.
The following timeline illustrates increasing concern about pathogens and interest in new technologies, such as irradiation, for reducing pathogen levels in meat and poultry:
The list of FDA-approved materials does not adequately cover the expansive number of polymers, adhesives, and colorants that are used in multi-layer, multi-constituent food-packaging materials that offer special properties such as an improved oxygen barrier. In addition, very few of the adjuvants (e.g., antioxidants, plasticizers, and antifogging agents) that are routinely used in today’s materials have been evaluated and approved by the FDA for use in packaging materials that are intended to be irradiated in contact with food.
The Federal Food, Drug, and Cosmetic Act (or the "Act"), Section 409(a), states that the use of a food additive shall conform to a regulation prescribing the conditions under which the additive may safely be used. Section 201(s) of the Act defines a food additive, in part, as “any substance the intended use of which results or may reasonably be expected to result, directly or indirectly, in its becoming a component or otherwise affecting the characteristics of food.” The definition encompasses packaging because packaging components could become a component of food by migrating from the packaging into food. In the past, packaging components have been referred to as “indirect additives” because these substances are not added directly to food for some functional purpose. Section 201(s) also defines any source of radiation intended for use on food as a food additive. Under Section 409 of the Act, as originally established, food additives require premarket approval by the FDA through the submission of a food additive petition and publication of a regulation authorizing their intended use. The requirements for a petition are described in Title 21 of the Code of Federal Regulations (CFR), Part 171.1 (Petitions) (9). FDA’s safety evaluation of a food additive includes a dietary exposure assessment and a toxicological evaluation based on animal feeding studies and other toxicological information.
Recently, Section 309 of the FDA Modernization Act of 1997 (FDAMA) amended Section 409 of the Act to establish a new process, referred to as the food contact notification (FCN) process, as the primary method of authorizing new uses of food additives that are food contact substances (FCS). Section 409(h)(6) defines an FCS as “any substance intended for use as a component of materials used in manufacturing, packing, packaging, transporting, or holding food if such use is not intended to have any technical effect in such food.” The requirements for an FCN are described in 21 CFR 170, Subpart D (Food Additives – Premarket Notifications) (9), and guidance documents are available on FDA’s website (10). Because the safety standard is the same for all food additives, the data and information requirements for FCNs and petitions are comparable.
There are two main differences in the petition and FCN processes. First, in contrast to the petition process, the FCN process will not result in a food additive listing in the CFR authorizing the use for any manufacturer of the FCS. Rather, an FCN for a food contact substance cannot be effective for anyone other than the manufacturer identified in the FCN. FDA maintains a list of effective FCNs on its website (10). Second, under the FCN process, the FDA has 120 days in which to object to an FCN or the FCN becomes effective, and the FCS may be legally marketed for the intended use.
A reference list of materials currently permitted for use during the irradiation of prepackaged food is given in Table I. With the exception of polystyrene (PS) foam trays, all the materials in Table I are listed in 21 CFR 179.45 (Packaging materials for use during the irradiation of prepackaged food) (9). The PS foam trays were reviewed under FDA’s Threshold of Regulation policy, which exempts certain food additives from a food additive regulation listing when the use results in a dietary concentration (DC) of less than 0.5 ppb (see 21 CFR 170.39 (9)).
In addition to §179.45, two other sections of the CFR are applicable to the irradiation of prepackaged food: §179.26 (Ionizing radiation for the treatment of food) and §179.25(c) (General provisions for food irradiation). Section 179.26 lists the radiation sources that may be used on food, the specific foods that may be irradiated, the conditions under which those foods may be irradiated, and the labeling that is required on irradiated foods. Section 179.25(c) inextricably links the packaging materials listed in §179.45 with the conditions of use described in §179.26, meaning that no other packaging materials (including adjuvants) are permitted for prepackaging food that will be irradiated. The finished packaging material and all adjuvants must meet any specifications and limitations of the applicable regulations in order to be marketed in the U.S. for food contact.
Although the vast majority of food contact substances are evaluated via the FCN process, it is still possible that an FCS might require evaluation via the petition process, especially if its dietary concentration is exceptionally high (on the order of 1 ppm or higher; see §170.100(c)(1)) (9). The FCN process can take much less time than the petition process because of the additional time needed to prepare and publish a regulation in response to a petition. The Threshold of Regulation policy (described above) is another route to FDA approval of new FCSs.
| Year | Regulation | Material | Requester | Max. Dose (kGy) |
|---|---|---|---|---|
| 1964 | §179.45(b) | Nitrocellulose-coated cellophane | AEC | 10 |
| Glassine paper | AEC | 10 | ||
| Wax-coated paperboard | AEC | 10 | ||
| Polyolefin filma | AEC | 10 | ||
| Polystyrene filma | AEC | 10 | ||
| Rubber hydrochloride filma | AEC | 10 | ||
| Vinylidene chloride-vinyl chloride copolymer filma | AEC | 10 | ||
| 1965 | §179.45(b) | Vinylidene chloride copolymer-coated cellophane | AEC | 10 |
| §179.45(d) | Vegetable parchments | U.S. Army | 60 | |
| 1967 | §179.45(b) | Kraft paper to contain only flour | U.S. Army | 0.5 |
| §179.45(d) | Polyethylene filma | U.S. Army | 60 | |
| Polyethylene terephthalate (PET) filma | U.S. Army | 60 | ||
| Nylon 6 filma | U.S. Army | 60 | ||
| Vinyl chloride-vinyl acetate copolymer filma | U.S. Army | 60 | ||
| 1968 | §179.45(b) | Optional adjuvants for polyolefin films plus optional vinylidene chloride copolymer coating | AEC | 10 |
| PET film plus optional adjuvants, vinylidene chloride copolymer and polyethylene coatings | AEC | 10 | ||
| Nylon 11 | AEC | 10 | ||
| 1989 | §179.45(c) | Ethylene-vinyl acetate copolymers | Cryovac | 30 |
| 1996 | Threshold of Regulation submission | Polystyrene foam tray | Amoco | 7.2 |
aPlus limited optional adjuvants. | ||||
In order to approve NCFST’s petition regarding the equivalency of the three radiation sources that may be used on prepackaged food (see the timeline above), it was necessary to reevaluate the dietary exposure to RPs formed in the packaging materials currently listed in §179.45 for the following reasons:
The first step in evaluating exposure to RPs from packaging materials was to identify and quantify parameters that depict the conditions under which prepackaged food would be irradiated and stored. Based on an extensive review of the literature, the following six parameters were determined to be relevant to RP formation in polymers:
Absorbed Dose
Irradiation leads to two competing reactions in polymers: chain scission, which leads to the formation of low-molecular-weight RPs, and crosslinking, which can lead to a decrease in residual oligomers (13, 14, 15). An increasing absorbed dose can lead to crosslinking up to an optimum point. If the dose is increased beyond that point, chain scission becomes dominant. In the absence of crosslinking, which occurs only in an O2-free atmosphere (see below), concentrations of RPs generally increase linearly with absorbed dose within limited dose ranges that include the ranges needed for irradiating foods (13, 14, 15, 16).
Because fresh or frozen poultry, fresh meat, and frozen meat may be irradiated to doses up to 3, 4.5, and 7.0 kGy, respectively, and because only a few foods of limited consumption may be irradiated to higher doses, [2] 10 kGy was selected as a conservative value for use in exposure estimates for polymer RPs that form in foods irradiated in their final packaging.
Atmosphere
In the presence of oxygen or air (21% O2), polymer chain scission leads to the formation of oxidative degradation products, which are primarily oxygenated volatile and semi-volatile organic compounds such as aldehydes, ketones, and carboxylic acids (13, 15, 17, 18, 19, 20, 21, 22). Crosslinking dominates under vacuum or an inert atmosphere. In air, increasing dose leads to higher concentrations and a wider variety of measurable RPs (21). For example, for low-density polyethylene (LDPE) irradiated to 20 kGy with an e-beam source at room temperature, the levels of oxygenated volatile and semi-volatile organic compounds are about one order of magnitude higher in polymer samples irradiated in air than in those irradiated in a vacuum, while the levels of hydrocarbons are the same in the presence or absence of O2 (17).
In the U.S., all commercial facilities that irradiate food and other bulk materials such as medical supplies are currently irradiating in air (23, 24, 25). If the food is packaged in a material that contains an oxygen barrier and the interior is purged of oxygen, then oxygenated RPs are not likely to form in the packaging layers inside that barrier or in the food. However, if the interior is not completely purged of oxygen, RPs may form in the inner layers of the packaging material and migrate to the food.
Dose Rate
In the presence of air, for a given dose, the low dose rates typical for gamma sources can lead to levels of RPs in polymers that are higher than levels generated at the higher dose rates typical for X-ray and e-beam sources; [3] the latter, for a given dose, also result in the formation of fewer types of detected RPs (13, 15, 17, 19, 20, 22, 26). Exposure data based on studies of migrants from polymers irradiated by gamma sources can therefore be considered conservative for migrants formed by any source of radiation.
The difference between the levels of RPs generated by gamma and e-beam sources is generally not great at doses below 20 kGy (20). For LDPE irradiated to 20 kGy in air at room temperature, the levels of RPs induced by gamma radiation exceed those induced by e-beam radiation only by about a factor of two (17). Therefore, when gamma irradiation data are unavailable, it is reasonable to use the levels of RPs generated by e-beam or X-ray sources to estimate exposures at low doses (<20 kGy), particularly considering the uncertainties involved in the exposure calculations (see below).
Temperature
In general, the temperature of a polymer during irradiation does not have an effect on RP concentrations when the polymer is irradiated below its glass transition temperature (Tg). But, as the temperature is increased above the Tg, the concentrations of RPs can increase significantly (17). For LDPE irradiated to 20 kGy with an e-beam source in air at room temperature, the concentrations of volatile and semi-volatile RPs have been shown to increase by about a factor of three between -75o C (Tg = -78o C) and 0o C, at which point they level off (17).
Because fresh or frozen poultry and meat are expected to contribute significantly to the total daily diet among irradiated prepackaged foods, the FDA has generally assumed that half of all foods irradiated in their final packaging will be treated at a temperature £ 4o C (fresh) and the other half at -18o C or below (frozen), the temperatures recommended in ASTM Standard F1356 for irradiation of poultry and meat (27).
Time after Irradiation
After a polymer has been irradiated in air, RP concentrations increase for some time and then level off. This behavior indicates that radiation-induced peroxy radicals become trapped in the polymer, where they continue to react with the polymer and generate RPs until they have all reacted (16, 20, 28). For polypropylene (PP) irradiated to 10 kGy with an e-beam source in air at room temperature, the concentrations of volatile (low molecular weight) RPs formed primarily from the polymer have been shown to level off after about 15 days, while the concentrations of less volatile (higher molecular weight) degradation products of Irganox and Irgafos antioxidants used in the polymer can increase by a factor of 2 to 5 during a period of 1 to 60 days after irradiation, at which point they level off (16, 28). These facts indicate that, after irradiation, PP degradation reaches a steady state more rapidly than antioxidant degradation.
Exposure estimates for species that migrate from packaging to food are based on the concentrations of migrants (e.g., RPs) in food or food simulants determined under time and temperature conditions that reflect processing and extended storage prior to consumption (12). In the absence of market data, it is not possible to estimate how much irradiated poultry or meat is consumed immediately after purchase and how much is stored by consumers in their freezers for some time. Therefore, the FDA has assumed, for the purposes of this exposure evaluation, that fresh poultry and meat are maintained at 4o C for 3 days after irradiation, frozen at -18o C for 6 months (180 days), and then thawed at 4o C for 1 day prior to preparation and consumption. The FDA has also assumed that frozen poultry and meat are maintained at ‑18o C for 6 months and then thawed in the packaging at 4o C for 1 day. These assumptions should not lead to overly exaggerated exposure estimates because the rate of migration (diffusion) of RPs is greatly reduced at freezing temperatures compared to the rate at room temperature.
Contact with Food Simulants
A “realistic” testing scenario for determining the concentrations of polymer RPs in food would involve irradiating the polymer while it is in contact with a food or food simulant and then conducting a migration study on the same irradiated sample. However, from the point of view of analyzing the RPs and distinguishing those generated in the polymer from those generated in the food simulant, a more “practical” approach involves irradiating the polymer alone, analyzing the polymer for RPs, and calculating exposure estimates using migration modeling or the assumption of 100% migration to food (12). These results can be further refined by irradiating a second sample of the polymer alone and then placing it in contact with an appropriate food simulant in order to conduct a migration study using the appropriate testing protocols. The food simulant could then be analyzed for the RPs already identified in the first polymer sample.
One study, which involved aspects of both the realistic and practical testing scenarios, reported that the concentrations of RPs produced in polymers differ by at most a factor of 2 in polymers irradiated in contact with air on one side and with aqueous/acidic food simulants on the other side, compared with polymers irradiated in air alone (16). [4] The somewhat higher concentrations in polymers irradiated in contact with food simulants are likely due to the direct contact of the polymer with liquids comprised of oxygenated species. For example, the concentration of acetone in PP film was 2.6 mg/kg when the polymer was irradiated in air alone, 3.7 mg/kg when the polymer was irradiated in contact with water, 2.9 mg/kg when the polymer was irradiated in contact with 15% ethanol, and 4.2 mg/kg when the polymer was irradiated in contact with 3% acetic acid (16). [5]
Based on these results, it may be concluded that RP migration values derived from the “practical” scenario are not likely to yield significantly different exposure estimates from those derived from the “realistic” scenario, particularly considering the conservatisms already built into the exposure estimates (see below).
Based on a comprehensive collection of recent literature, several organic RPs were identified and quantified in seven of the major polymers listed in §179.45, which made it possible to calculate exposures to the RPs using the relevant parameters described above. The polymers, RPs, and dietary concentrations (DC) are summarized in Table II. In general, the test polymers had been irradiated to 10-50 kGy with gamma or e-beam sources in air at room temperature in the absence of food simulants, and the polymers were analyzed within one day of irradiation. Concentrations of RPs obtained at 20-50 kGy were extrapolated to 10 kGy, assuming a linear relationship between concentration and dose (see the “Absorbed Dose” section above).
Because RPs are expected constituents of packaging resulting from its conditions of use, the DCs were compared to 0.5 ppb, the DC that FDA equates to negligible risk for a substance that has not been shown to be a carcinogen in humans or animals and for which there is no reason, based on the chemical structure of the substance, to suspect that it is a carcinogen. It should not be construed that substances whose DCs exceed 0.5 ppb are unsafe nor that substances of DC ≤ 0.5 ppb are exempt from FDA approval. All food additives, including their related constituents, must be evaluated by the FDA for safety on a case-by-case basis, regardless of their exposures, and may be deemed safe at DCs well above 0.5 ppb.
| Polymer | RP |
Conc. in Polymer (mg/kg)a, b |
Ref. |
Conc. in food (µg/kg)c |
DC (ppb) |
|---|---|---|---|---|---|
| Polystyrene (PS) (density 1.06 g/cm3) |
1-phenylethanol |
3 | (29) |
8.2 |
0.41d |
acetophenone |
18 | (29) |
fresh: 7.8 |
0.39e |
|
benzene |
1 |
(29) |
2.7 |
0.14d |
|
fresh: 0.53 |
0.02e |
||||
froz.: 0.36 |
|||||
phenylacetaldehyde |
3 | (29) |
8.2 |
0.41d |
|
benzaldehyde |
18 | (29) |
fresh: 8.4 |
0.42e |
|
phenol |
5 | (29) |
fresh: 2.5 |
0.12e |
|
benzoic acid |
4 | (29) |
fresh: 1.7 |
0.09e |
|
unidentified carboxylic acid a |
2.7 | (30) |
7.4 |
0.37d |
|
unidentified carboxylic acid b |
2.7 | (30) |
7.4 |
0.37d |
|
| Poly(ethylene terephthalate) (PET) (density 1.4 g/cm3) |
diisopropyl ether |
0.8 |
(30) |
2.89 |
0.14d |
fresh: 0.11 |
0.006e |
||||
formic acid |
0.297 | (31) |
1.0 |
0.05d |
|
acetic acid |
0.369 | (31) |
1.3 |
0.06d |
|
1,3-dioxolane |
0.384 | (31) |
1.4 |
0.07d |
|
2-methyl-1,3-dioxolane |
3.7 | (31) |
fresh: 0.55 |
0.03e |
|
acetone |
0.086 | (31) |
0.30 |
0.02d |
|
| Low-Density Polyethylene (LDPE) (density 0.92 g/cm3) |
acetic acid |
8.5 | (17) |
8.5 |
1.0d, f |
propionic acid |
5.1 | (17) |
12 |
0.6d, f |
|
n-butyric acid |
1.0 | (17) |
2.4 |
0.12d |
|
n-valeric acid |
0.4 | (17) |
0.95 |
0.05d |
|
butanoic acid vinylester or |
1.68 | (30) |
4.0 |
0.20d |
|
1,3-di-tert-butylbenzene |
1.7 | (30) |
4.0 |
0.20d |
|
2,4-di-tert-butylphenol |
30 | (32) |
71 |
3.6d, f |
|
2,6-di-tert-butyl-p-benzoquinone from Irganox 1010, 1076 |
4 | (32) |
9.5 |
0.47d |
|
| Polypropylene (PP) (density 0.90 g/cm3) |
2,4-pentanedione |
2.4 |
(16) |
5.6 |
0.22d |
1-dodecene |
1.4 |
(16) |
1.4 |
0.13d |
|
acetone |
2.6 |
(16) |
6.0 |
0.24d |
|
2-pentanone |
0.75 |
(16) |
1.7 |
0.07d |
|
4-hydroxy-4-methyl-2- |
1.9 |
(30) |
4.4 |
0.18d |
|
3-methyl-2-butanone (?) |
1.5 |
(30) |
3.5 |
0.14d |
|
acetic anhydride |
7.4 |
(30) |
17 |
0.69d, f |
|
3-methylbutanoic acid |
2.0 |
(30) |
4.6 |
0.19d |
|
acetic acid-(1-ethylhexyl)-ester |
0.7 |
(30) |
1.6 |
0.07d |
|
octanoic acid |
1.8 |
(30) |
4.2 |
0.17d |
|
3-methyl-4-methylene-hexane-2-one |
0.9 |
(30) |
2.1 |
0.08d |
|
2,5-cyclohexadiene-1,4-dione |
2.1 |
(30) |
4.9 |
0.20d |
|
hexadecanol or octadecanol |
2.0 |
(30) |
4.6 |
0.19d |
|
4-methyl-2,3-pentanedione (?) |
1.1 |
(30) |
2.6 |
0.10d |
|
1,3-di-tert-butylbenzene from Irgafos 168 |
17 |
(16) |
39 |
1.6d, f |
|
2,4-di-tert-butylphenol |
75 |
(32) |
174 |
7.0d, f |
|
16g |
(33) |
28 |
1.1h |
||
1,3-di-tert-butyl-2-hydroxybenzene from Irgafos 168 |
14 |
(16) |
33 |
1.3d, i |
|
2,6-di-tert-butyl-p-benzoquinone from Irganox 1010, 1076 |
14 |
(28) |
33 |
1.3d, i |
|
| Ethylene-Vinyl Acetate Copolymers (EVA) (density 0.94 g/cm3) |
acetaldehyde |
-- |
(34) |
1600 |
32j |
n-propyl acetate |
-- |
(34) |
570 |
11j |
|
3-methylhexane |
-- |
(34) |
1000 |
20j |
|
n-heptane |
-- |
(34) |
430 |
8.6j |
|
n-octane |
-- |
(34) |
67 |
1.3j |
|
| Nylon 6 (density 1.1 g/cm3) |
butanamide |
2 |
(35) |
5.7 |
0.11d |
pentanamide |
85 |
(35) |
fresh: 42 froz.: 29 |
0.71e |
|
| Poly(vinyl chloride) (PVC) (density 1.3 g/cm3) |
4-hydroxy-4-methyl-2-pentanone |
6.2 |
(30) |
21 |
1.0d, i |
5-hexen-2-one |
3.8 |
(30) |
13 |
0.64d, i |
|
1-ethoxy-2-heptanone |
7.1 |
(30) |
24 |
1.2d, i |
|
methoxyacetaldehyde diethyl acetal |
15 |
(30) |
50 |
2.5d, i |
|
diethoxy acetic acid ethylester |
4 |
(30) |
13 |
0.67d, i |
|
3-methylheptyl acetate |
2.4 |
(30) |
8.0 |
0.40d |
|
diethyl adipate |
8.3 |
(30) |
28 |
1.4d, i |
|
nonanoic acid ethylester |
2.4 |
(30) |
8.0 |
0.40d |
|
unidentified n-alkane acid ethylester a |
34.5 |
(30) |
116 |
5.8d, i |
|
unidentified n-alkane acid ethylester b |
50.3 |
(30) |
169 |
8.4d, i |
|
aConcentrations determined at 20-50 kGy were extrapolated to 10 kGy, assuming a linear relationship between concentration and dose. Concentrations reported for unirradiated control samples were subtracted from those reported for irradiated test samples. bOnly the highest concentration reported for each RP in the literature is included in this table. cAssuming a food mass-to-polymer surface area ratio of 10 g/in2 (see text). d100% migration calculation. eModeled migration (see text). fMigration models failed to describe migration below 100% from thin films made of polymers that yield fast diffusion coefficients. gMigration to 10% ethanol food simulant expressed as mg/kg polymer tested. hMeasured migration value into 10% ethanol after 10 d at 40o C. iMigration modeling not possible due to lack of diffusion coefficients for PVC films. jMeasured migration value into 95% ethanol after 1 d at room temperature. |
|||||
Calculation Methods
Because the RPs were analyzed in the polymers rather than in food simulants, DCs were initially calculated by assuming 100% migration of the RPs to food. The following parameters were used in this calculation: a polymer film thickness of 40 mm (0.004 cm, 1.57 mils), a density typical of each polymer (see Table II), a food mass-to-polymer surface-area ratio of 10 g/in2 (1.55 g/cm2), and a CF of 0.05. FDA’s default CF of 0.05 was used because 1) only polymer films are regulated in §179.45 – the CFs given in (12) for the various polymers do not distinguish between rigid containers and films, and 2) only a small fraction of food packaged in contact with a given polymer is expected to be irradiated. For PP and Nylon, the CFs given in (12) were used (0.04 and 0.02, respectively). Because the packaging materials listed in §179.45 are not restricted by food type, a total fT value of 1 was assumed. A sample calculation follows for a level of 3 mg/kg 1-phenylethanol (PhE) in PS film (density: 1.06 g/cm3, thickness: 40 mm or 0.004 cm):
| ( | 3x10-6g PhE | )( | 1.06 g PS | )( | 0.004 cm | )( | 6.45 cm2 | )( | 1 in2 | )( | 0.05 CF | ) | = 0.41 ppb DC |
| g PS | cm3 | 1 in2 | 10 g food |
For cases in which the DCs from 100% migration calculations exceeded 0.5 ppb, migration modeling based on Fick’s law of diffusion was used to calculate a more realistic exposure (36). The Piringer model was used to calculate diffusion coefficients for use in the migration model (37). The time and temperature conditions for fresh poultry and meat described in the “Time after Irradiation” section above were used as inputs for the model. Then, to calculate the DC, the food mass-to-polymer surface-area ratio, CFs, and fT described above were applied to the modeled migration values.
If the DC exceeded 0.5 ppb for fresh poultry and meat, the FDA assumed that half the products will be irradiated fresh and that half will be irradiated frozen (see the “Temperature” section above). This assumption led to two terms in the DC calculation, each of which was dominated by the 6 months at -18o C time and temperature condition. Because the two terms were practically equal, this calculation did not reduce any exposure estimates for fresh products to £ 0.5 ppb DC. Nevertheless, this calculation did yield the most realistic exposure estimates possible for the RPs (see, for example, the Polystyrene and Nylon 6 entries in Table II).
Exposure information from Table II on the 58 RPs quantified in seven major polymers is summarized in Table III. From this table, it is evident that the majority of the RPs (31) met the 0.5 ppb DC limit based on a 100% migration calculation and that five more met the limit based on migration modeling. For the remaining 22 RPs that did not meet the 0.5 ppb DC limit, it should be noted that the 10-kGy dose used in the exposure estimates is conservative because fresh or frozen poultry and meat, which may be irradiated to 3 to 7 kGy, are expected to contribute significantly to the total daily diet among irradiated prepackaged foods. In addition, the test polymers had been irradiated at room temperature. If the polyolefins (Tg << 20o C) had been irradiated at the intended refrigerated or frozen use temperatures, the DCs of their RPs likely would have been lower (see the “Temperature” section above).
| Polymer | Number of Radiolysis Products | ||||
|---|---|---|---|---|---|
| DC ≤ 0.5 ppb | DC > 0.5 ppb |
Total Quantified |
|||
| 100% Migration | Modeled | ||||
| PS | 5 | 4 | 9 | ||
| PET | 5 | 1 | 6 | ||
| LDPE | 5 | 3 | 8 | ||
| PP | 13 | 5 | 18 | ||
| EVA copolymers | 5 | 5 | |||
| Nylon 6 | 1 | 1 | 2 | ||
| PVC | 2 | 8 | 10 | ||
| Total: | 31 | 5 | 22 | 58 | |
It should be noted that, although none of the exposures to RPs from PET exceeded 0.5 ppb DC, five of the RPs were determined in test samples that had been irradiated in the absence of O2 (31), which could result in underestimates of exposure if food packaged in PET film were irradiated without an oxygen barrier between the PET and the ambient air of the radiation chamber (see the “Atmosphere” section above). It is not likely that the presence of O2 would cause the exposures to the identified RPs to exceed 0.5 ppb DC because 1) exposure estimates based on an absorbed dose of 10 kGy are conservative for foods that are irradiated in their final packaging, and 2) the five exposure estimates are approximately one order of magnitude below 0.5 ppb DC (see Table II). Only through further testing would it be possible to identify and quantify additional RPs that might form when PET is irradiated in the presence of O2.
Polymer Adjuvants
Thirteen of the 22 RPs with DCs > 0.5 ppb were from polymer adjuvants that are not listed in §179.45. 2,4-Di-tert-butylphenol (2,4-DTBP), which was identified in irradiated LDPE and PP, is a breakdown product of Irgafos 168, an antioxidant often added to polyolefins (32). Three additional breakdown products of Irgafos 168 and a breakdown product of the antioxidants Irganox 1010 and 1076 were also identified in PP (see Table II). These breakdown products also form during photochemical and thermal oxidation of packaging materials containing hindered phenol antioxidants (28). However, the kinetics of their formation is much more rapid during irradiation, i.e., irradiation is comparable to accelerated ageing (28). The FDA has typically not been concerned with the breakdown products of these antioxidants because they have not been observed under conventional food-contact conditions.
Although basic polyolefins listed in §177.1520 (Olefin polymers) are regulated for use as films under §179.45(b)(4) and §179.45(d)(2)(i), the optional adjuvants listed in §177.1520(b) and the antioxidants and stabilizers listed in subparagraph (b) of §178.2010 (Antioxidants and/or stabilizers for polymers) such as Irgafos 168, Irganox 1010, and Irganox 1076 are not (9). Therefore, polyolefins containing the adjuvants listed in §177.1520(b) or §178.2010(b) are not currently permitted for prepackaging food that will be irradiated and would need to be evaluated for safety via FDA’s food contact notification process.
In addition to the exposure calculated assuming 100% migration of 2,4-DTBP from irradiated PP to food (7.0 ppb DC), actual migration data in the literature made it possible to obtain a more realistic exposure (33). In the study, a 10% ethanol food simulant was sealed inside a PP pouch and irradiated to 10 kGy with an e-beam source in air at room temperature. The pouch was then maintained at 40o C for 10 days prior to analysis. The resulting DC (1.1 ppb) is about a factor of 6 less than the 100% migration calculation noted above.
All eight of the RPs from PVC whose DCs exceeded 0.5 ppb are RPs of the plasticizer (probably di(2-ethylhexyl)adipate, based on the presence of diethyl adipate) rather than of the PVC itself. PVC is regulated only as an optional adjuvant for polyolefin films and PET films in §179.45. PVC films per se are not currently permitted for prepackaging food that will be irradiated. Therefore, the likely exposure to the RPs from regulated plasticized PVC will be well below 0.5 ppb DC.
EVA Copolymers
The migration data for the five RPs from EVA copolymers whose DCs exceed 0.5 ppb are from Food Additive Petition 7B3968, which resulted in the listing of EVA copolymers in §179.45 (34). EVA copolymer pouches were filled with 95% ethanol food simulant and irradiated to 30 kGy with a gamma source in air at room temperature. The food simulants were maintained at room temperature until they were analyzed one day after irradiation. In order to calculate the DC, migration values were extrapolated to 10 kGy, a food mass-to-polymer surface area of 10 g/in2 was assumed, and a CF of 0.02 (the CF for EVA copolymers (12)) was applied. These values are slightly lower than the exposures originally calculated for the petition, which were based on the 30‑kGy dose and an additional assumption that 40% of food packaged in EVA copolymers would be irradiated. Although the DCs for these RPs exceed 0.5 ppb DC, they were individually evaluated by the FDA and determined to be safe.
GRAS Substances
Acetic acid and propionic acid are two RPs from LDPE. These two acids are affirmed as Generally Recognized as Safe (GRAS) for direct addition to food in §184.1005 and §184.1081, respectively. Thus, they have been deemed safe at DCs much greater than 0.5 ppb (9). Both acids can have a significant effect on the organoleptic properties of food, so their concentrations are self-limiting at levels far below their good manufacturing practice use levels described in the regulations.
The acetic anhydride RP from PP is likely to hydrolyze to acetic acid in food. In addition, it was necessary to use a 100% migration calculation for acetic anhydride because the simple migration model, coupled with the Piringer model for calculating diffusion coefficients, indicates that 100% migration will occur from thin PP films. Simple migration models assume that an infinitely thick plane of material is in contact with the food (i.e., an infinite source of the migrant) and therefore do not depend on the polymer thickness (36, 37). The models generally fail in describing migration below 100% from thin films made of polymers that yield fast diffusion coefficients (e.g., polyolefins). However, the models successfully predicted migration from films made with polymers that yield very slow diffusion coefficients (e.g., PS, Nylon, and PET).
Pentanamide from Nylon 6
The DC of pentanamide from Nylon 6 was calculated via migration modeling to be 0.7 ppb, which exceeds 0.5 ppb by a very small amount. As is discussed above, the 10-kGy dose selected for this exposure evaluation is conservative for foods that are irradiated in their final packaging.
Exposures to 58 RPs from seven polymer types have been evaluated, based on a survey of the available literature. Many more RPs have been identified in five of the polymer types but have not been quantified (up to 63 in LDPE (21, 38, 39), 73 in PP (21, 39), 14 in plasticized PVC (40), 10 in Nylon 6 (35), and 10 in EVA copolymers (34)). However, the compounds that have been quantified tended to have the highest GC peak areas, i.e., the highest concentrations in the polymer. Because the RPs are primarily aldehydes, ketones, and carboxylic acids, it can be assumed that the GC response factors for the quantified and unquantified compounds are similar. Therefore, in most instances, the concentrations of the unquantified compounds should not exceed those of the quantified compounds. Quantitative data on RPs formed in the remaining materials listed in §179.45 are not available. However, fresh and frozen poultry and meat are most likely to be prepackaged in the polymeric films discussed above (except plasticized PVC) or multilaminates comprised of them.
Based on a comprehensive review of recent literature articles in which RPs were quantified in major polymers, it has been determined that irradiation of the most commonly used materials listed in §179.45, under the conditions typical for foods, results in exposures to many RPs that are below 0.5 ppb DC. These RPs and others with higher exposures have been evaluated and determined to be safe. Although RPs from the currently regulated materials are not of concern, new polymers might yield RPs of toxicological concern. In addition, the literature data have shown that RPs from unregulated polymer adjuvants such as antioxidants and plasticizers are of potential concern due to their high concentrations. Polymeric materials and adjuvants that are not listed in Table I must be approved by the FDA for use during the irradiation of prepackaged food.
[1] The Gray (Gy) is a unit of radiation-absorbed dose that equals the amount of energy absorbed per unit mass of a material during irradiation (1 Joule/kg). 10 kGy = 1 Megarad (Mrad), a previous unit of absorbed dose.
[2] Dry enzyme preparations may be irradiated to 10 kGy, dry spices to 30 kGy, and frozen, packaged meats used solely in NASA space flight programs to 44 kGy (see §179.26).
[3] Due to the lower efficiency of machines in generating X-radiation compared to e-beam radiation and due to the large mass of material that would be required to absorb the more penetrating X-rays, the dose rate for X-ray sources is lower than that for e-beam sources. The sources of radiation, listed in order of increasing dose rate, are: gamma << X-ray < e-beam.
[4] In these experiments, the RPs were analyzed in the polymer, not the food simulant.
[5] These values were extrapolated from 50 kGy to 10 kGy. The article did not provide data to show whether any of the reported differences among the tested samples were statistically significant.
*Kristina E. Paquette, Ph.D., is a Review Chemist in the Division of Food Contact Notifications; Office of Food Additive Safety; Center for Food Safety and Applied Nutrition; US Food and Drug Administration; 5100 Paint Branch Parkway; College Park, MD 20740.
