Aluminum contamination of food during culinary preparation: Case study with aluminum foil and consumers’ preferences

Abstract The aim of the work was to estimate the degree of aluminum leakage from aluminum foil during baking process of selected food/meals. The experiment included 11 different types of food (Atlantic salmon Salmo salar, mackerel Scomber scombrus, duck breasts, cheese Hermelín, tomato, paprika, Carlsbad dumplings, pork roast, pork neck, chicken breasts, and chicken thighs) baked both marinated and not marinated. The aluminum content was measured by AAS and ICP/MS methods. The highest aluminum increase was observed in the samples of marinated Salmo salar (41.86 ± 0.56 mg/kg), Scomber scombrus (49.34 ± 0.44 mg/kg), and duck breast (117.26 ± 1.37 g/kg). The research was also supported by the survey that consisted of 784 respondents with different sociodemographic characteristics. The study clearly showed the occurrence of aluminum contamination of food when it is prepared by baking in aluminum foil. It cannot be concluded that aluminum leakage will occur with each type of food. The aluminum contents found among investigated samples are not alarming, though the increase was measured up to 40 times. On the other hand, revealed aluminum contents can represent a risk for younger/smaller children and for individuals with diagnosed certain ailments.


| INTRODUC TI ON
Aluminum is one of the most abundant elements in earth crust's mass (8% of the earth crust's mass); due to this fact, it can be explained the prevalence of aluminum content in food, though with concentrations below 5 mg/kg (EFSA, 2008;Ertl & Goessler, 2018). Daily aluminum intake depends on food types, and according to studies, following food types contribute the most to aluminum daily intake: cereals, vegetables, fruits, tea, coffee, wheat, and wheat-based products (Bratakos, Lazou, Bratakos, & Lazos, 2012;Fekete, Vandevijvere, Bolle, & Loco, 2013;González-Weller, Gutiérrez, Rubio, Revert, & Hardisson, 2010;Ma et al., 2016). The problem of aluminum intake can be emphasized with the fact that only small amounts of aluminum can be efficiently excreted by human bodies. Meaning that wider population can be exposed to aluminum amounts that cannot be handled by human body (Bassioni, Mohammed, Al Zubaidy, & Kobrsi, 2012).
Aluminum is used as packaging material, and it is approximately calculated that it has produced about 860 000 tons of aluminum foil per year only in the Europe (EAFA, 2016). Aluminum foil is mainly used for food packaging, cosmetics, and chemical products (approximately 75%) (EFSA, 2011). Aluminum foil is broadly used in culinary preparation of different food types due to its easiness to use, disposability, and properties such as that it transfers heat twice quicker than regular metal. It is a common culinary practice in households to wrap food in aluminum foil and baked it. These food preparation and practice can expose consumers to different aluminum amounts in dependence to food types and composition (Bassioni et al., 2012).
This interaction between aluminum foil and food wrapped in it represents a potential hazardous source of aluminum in the human diet.
On the other side, aluminum packaging is very popular due to its following properties: impermeability, freeze proofing, it is inert, especially it possesses a perfect dead fold characteristics and recyclability (Verissimo, Oliveira, & Gomes, 2006).

Aluminum contamination of food represents an important issue
to find relationships between aluminum intake and certain serious illness such as Alzheimer's disease, Parkinson's disease, dialysis encephalopathy, bone disorder, human breast cancer, and it is also considered to be a neurotoxin; aluminum salts can be accumulated by the gut and different human tissues (bones, parathyroid, and brain).
The average human intake of aluminum has been calculated to be from 14 to 105 mg aluminum per week for a 70 kg adult person.
The exposure of child weighing 30 kg is estimated to be from 21 to 69 mg aluminum per week. These findings are stressing out the possibility of reaching aluminum maximum tolerable weekly intake of 1mg/kg of body weight (Center for Food Safety, 2009) or 2 mg/kg of body weight (World Health Organization, 2011) (Stahl, Falk, Taschan, Boschek, & Brunn, 2018). Though, the weekly aluminum intake limit used to be 7 mg/kg of body weight (WHO, 1989).
The aim of the study consisted of two parts. The first was to evaluate the leakage of aluminum to different food types that were baked in aluminum foil. The second was to overview aluminum prevalence/contamination of food and consumers' preferences toward the usage of aluminum foil in food culinary preparation and their awareness of harmfulness. Polička, a.s., Czech Republic), fresh chicken breasts without skin and bone (Cut chicken breasts, quality class A, Tesco Stores ČR a.s., Czech republic), and fresh chicken thighs with skin and bone (Cut chicken thighs, quality class A, Tesco Stores ČR a.s., Czech republic). Deep frozen duck breasts were defrosted (at temperature + 2±2°C/12 hr), half of the samples were left with skin, and the skin from the second half was removed manually. Raw food samples were prepared according to commonly used recipes; marinade was added to experimental samples (the recipe composition is shown in Table 1). Each sample was packed into each of the 5 film types (area 20 × 20 cm), as well as a glass as inert packing material was also used. The baking was done during 40 min at 220ºC in the oven (Professional Ovens GARB-IN, Model: 23 GM UMI).

| Food samples
After cooling, the samples were homogenized, and the homogenates were vacuum packed and frozen. Analyzes were performed sequentially from thawed homogenate.

| Aluminum foils
Five foils present on the market were used in the research. The foils differ by purchase price, labeling (can be used for barbecue, general use in the kitchen). The following aluminum foils were used: foil

| Physical-chemical analysis
The laboratory analysis included determination of pH, salt, fat, protein, and dry matter content. pH was measured using the WTW GmbH, Germany. The salt (NaCl) content was determined by titration with silver nitrate-Mohr's method (standard operating procedure, potassium chromate indicator). The lipid content was determined quantitatively (ČSN ISO 1443:1973 using Soxtec 2055 (FOSS Tecator, Höganäs, Sweden). The protein content (ČSN ISO 937:1978) was determined as the amount of organically bound nitrogen using the analyzer Kjeltec 2300 (FOSS Tecator, Höganäs, Sweden). The dry matter/moisture content was determined gravimetrically according to the Czech National Standard (ČSN ISO 1442:1997 by drying the sample to a constant weight at + 103 ± 2ºC (Binder FD 53, Germany). The dry matter/moisture values of foodstuffs were used to convert the results for the aluminum content determined in dry matter by the protocol methods (see below) to the wet mass (Tables 2, 3, 4).

| Aluminum content analysis
Aluminum contents were measured by two methods that included inductively coupled plasma mass spectrometry (ICP-MS) and atomic absorption spectroscopy (AAS).
TA B L E 1 Weights (g) of foodstuffs and marinade ingredients used in the research Samples Foodstuffs weight (g)

| ICP/MS protocol
Prior to the microwave digestion, the homogenized food samples

| X-ray generator spectroscopy
The profile of each used aluminum foil was measured by using the portable X-Ray Analyzer NITON XL3t GOLDD+ (Thermo Fisher Scientific

| RE SULTS AND D ISCUSS I ON
In this study, the real conditions of food culinary treatment by consumers in households were simulated to monitor the level of aluminum transfer from different aluminum foils into a food whose method of preparation, consumption size, or weight (e.g., one piece of tomato, peppers, one slice of pork roast, and one chicken thigh) is common for culinary preparation. These modeled experiment conditions created a completely different set of eleven foods that varied in many parameters. There were differences in geometric shape, in surface size, surface character (smooth, rough), in origin (plant, animal), weight, thickness and technological processing of food (with/without skin or bone), physical conditions (fresh, frozen). Evaluated samples varied also according to use marinades.
The variety of shape had an effect, for example, on the tightness of the foil's adhesion to the food, and differences in the food weight affect the dispersion/dilution of the amount of aluminum transferred from the foil to the homogenized matrix. The study showed that some of the food matrices analyzed in our study are riskier in terms of aluminum transmission from the packaging foil. Samples of salmon, mackerel, and duck contained more aluminum in the homogenized matrix, although they represent different foods in terms of weight, nutritional composition, and method of preparation (Table 1 and 6).

Percentage of respondents
Aluminum contents in the samples of salmon and mackerel samples (both marinated and not marinated) determined by ICP-MS and AAS are shown in Figure 1.
The aluminum contents (shown in Figure 1)  week of storage (Ranau, Oehlenschläger, et al., 2001b). The authors monitored aluminum content during 2 years of storage and found a constant significant aluminum increase during the whole period of storage (by 1,321%). In our study, aluminum content increased by 20 times after marinating and heating of salmon and mackerel samples ( but the leakage is occurring also during can shelf life, since the prevalence of aluminum content in raw fish is under 1 mg/kg wet mass. That is the reason why aluminum content in raw seafood is considered as not a significant risk for consumers (Ranau, Oehlenschläger, et al., 2001b).

Aluminum contents in duck breasts wrapped in aluminum foil
and baked are shown in Figure 2 and  Figure 3).
According to these results, we have come to the new conclusion that, in addition to the known facts previously published, skin as a natural barrier has a major effect on aluminum transmission. We assume that all three skin layers (epidermis, dermis, and subcutis) have barrier properties and their components such as protein fibers, especially keratin, collagen, and elastin have direct responsibility for the aluminum's impermeability into the inner layers of the food.
Similar behavior can be assumed in matrices with different intramuscular tissue content, that is, aluminum will be present in larger quantities in deeper layers of fish meat (which have a very low content of stroma proteins) compared to mammalian or poultry meat.
It was found that during baking of food wrapped in aluminum foil at lower temperatures (<160ºC), the leakage of aluminum is at a lower rate than when the process of baking is done at temperatures over 220ºC (Turhan, 2006). Previous studies indicate that cooking temperature is more influencing aluminum leaching than cooking time, due to the changes of oxide layer from an amorphous to a crystalline structure. Oppositely, from our results it had been stated that aluminum foil is resistant to corrosion in the pH range of 4-8.5 (Lamberti & Escher, 2007). The occurrence of aluminum ions in food and especially secondary food contamination with aluminum represents the risk for consumers also since elements such as zinc, magnesium, and iron are essential for organism, and no scientific studies have indicated that aluminum plays an important role for a living organism (Schafer & Seifert, 2006;Stahl et al., 2018). Oppositely, aluminum intake represents a health risk, affecting hazardously nervous system, bones, and hemopoietic system (Becaria, Campbell, & Bondy, 2002).
The compositions of aluminum foils used in the research are presented with principal component analysis (PCA) (Figure 3). As it can be seen in Figures 3 and 4, groups are formed out of used aluminum foils compositions. This finding is showing how aluminum foils that are present on the market distinguished between themselves statistically significant (p < .05). The inclusion of other metals in aluminum foils affects its corrosion. Magnesium is often incorporated in aluminum foil due to the mechanical strength improvement, though higher magnesium content leads to lower stability of the alloy against weak acids (Lamberti & Escher, 2007). Though, older literature data stressed out that the leakage of aluminum from aluminum foil is almost neglectable in following food types and products: starch, sugar, egg, powder, coffee, chocolate, and tea packed in bare aluminum foil and undergone treatments such as storing at room temperature, in the refrigerator, freezing, and heating (Kunze, 1976;Lamberti & Escher, 2007;Servus, 1989). The neglectable leakage of aluminum was also found in the studies that included acidic beverages in coated aluminum cans and white wines in aluminum cans (Bloeck, Kreis, & Stanek, 1986;Dürr & Bloeck, 1987).
The significant (p < .05) differences between marinated and not marinated food samples, included in the research, are confirming findings that aluminum dissolution is dependent on pH, same as on temperature (Bi, 1996). Bi (1996) explained that aluminum leaching is happening according to the following formula: Aluminum leaching from aluminum foil represents an existent problem/issue, especially because aluminum leaching can be increased during culinary preparation, people may lower acidity or add salt, known agents that increase the transfer level of aluminum from aluminum foil. Oppositely, sugar added to food during heating in aluminum foil reduces aluminum leaching due to the formation of kind of coating (Joshi, Toma, Medora, & O'Connor, 2003;Verissimo et al., 2006).
Aluminum contents in other investigated food samples (hermelin, tomato, paprika, karlovarsky knedlik, pork roast, pork neck, chicken breasts, and chicken thighs) are present in Table 4. The highest concentration of aluminum was measured in marinated pork roast (6.90 ± 0.08 mg/kg wet mass). These samples were marinated and wrapped in foil E, the foil on which labeling was written that it can be used for the baking. These concentrations can be considered low, but they are higher in comparison with the results of Ertl and Goessler, (2018), and their aluminum levels in pork and fish fillets samples were under 1 mg/kg. Though, they found increment up to 4 times compared with raw samples. The lower aluminum contents in the study of Ertl and Goessler, (2018), probably, occurred due to lower baking temperatures (180ºC) than temperatures applied in our The hazardous side of aluminum foil used for food preparation and storage was also emphasized by the finding that aluminum content in food was increased during 1 to 3 days of storage in aluminum foil. The aluminum content was higher than 60 mg/kg and more than 20 mg/kg in ham and cheese samples stored for 3 days in aluminum foil, respectively (Ertl & Goessler, 2018).
Chemical composition of investigated samples is shown in Table 6. PCA analysis is emphasizing that there is no significant difference (p < .05) between samples' chemical compositions (Figure 3).
Found aluminum contents are not alarming, but the estimated dietary exposure to aluminum in various countries is between 1.3 mg/ day to 13 mg/day, while in European countries it is from 1.6 to 13 mg/day (Bratakos et al., 2012). These data are indicating that dietary aluminum exposure is from 0.2 to 1.5 mg/per kg of body mass in a 60 kg adult (EFSA, 2008).

| Survey
Humans are exposed to aluminum mainly through diet, meaning that consumers' sociodemographic attributes, eating habits, same as culinary preparation play the important role in aluminum daily intake (Yokel, 2012). Aluminum intake through daily diet is obviously influenced by consumers' preferences. Respondents' preferences toward the use of aluminum foil for food preparation are shown in Figure 4.
The issue can be overviewed through the results gained by the survey (Figure 4). The majority of respondents (72.9%; n = 467) answered that they do not check the labeling of aluminum foil, meaning they do not read the information that usually food should not be heated wrapped in aluminum foil, especially sour and salty food. Respondents' current status influenced significantly (p < .05) the checking of aluminum foil labeling (unemployed respondents and an employee of state checked aluminum foil less often) (Figure 4).
The majority of respondents (67.5%; n = 464) stated that they had not been looking for the information about the hazardous side of aluminum content in food,and they (56.4; n = 409) did not know how the intake of aluminum can adversely affect human organisms ( Figure 4). Chi-square test showed that women, married respondents, respondents under 20 years, and older than 30 ages are statistically significant (p < .05) more aware of aluminum use hazardous sides, in food preparation, while education level did not affect respondents' preferences toward the use of aluminum foil.
The important aspect of the survey is how respondents are informed about hazardous sides of aluminum usage. It was found that statistically significant (p < .05) that women, married respondents, and respondents aging from 31 to 50 years are better informed about hazardous sides of aluminum usage (Figure 4).

| CON CLUS ION
The study is showing leakage of aluminum to food that was baked in aluminum foil, and the research is indicating that excessive consumption of food prepared by baking in aluminum foil can carry a health risk. Although the measured level of aluminum contamination in food used in the experiment is not alarming, it is probably hard to achieve 2 mg/kg of body mass weekly limit, but due to hazardous side of aluminum intake it can have high potential risk to people with certain aliments (especially people suffering from chronic renal F I G U R E 4 Respondents' preferences concerning the use of aluminum foil failure) and smaller children. Beside these statements, aluminum absorption is well influenced by the presence of elements as iron, calcium, zinc, same as additive such as citrate (Al Juhaiman, 2010;Verrissimo et al., 2006). Our data will certainly serve as useful source and confirmation of the hazardous side of aluminum foil use during food culinary preparation, same as the confirmation of not so predictable aluminum contamination in different food types. The survey also confirmed that consumers are not enough informed about hazardous side of aluminum foil usage.
In conclusion, we can say the target of the study was met, and we have gained new up-to-date information on aluminum levels in a wide range of foods that was not been available in this research area.
Certainly that this pilot experiment will be used for more detailed case studies in future.

ACK N OWLED G M ENT
The research was done by financial supported Internal Grant Agency IGA VFU Brno 223/2018/FVHE.

CO N FLI C T S O F I NTE R E S T
The authors declare no conflict of interest or relationship, financial or otherwise.

E TH I C A L A PPROVA L
This study does not involve any human testing.