Development of a manual method for the determination of mineral oil in foods and paperboard.

So far the majority of the measurements of mineral oil saturated hydrocarbons (MOSH) and mineral oil aromatic hydrocarbons (MOAH) were obtained from on-line high performance liquid chromatography-gas chromatography-flame ionization detection (on-line HPLC-GC-FID). Since this technique is not available in many laboratories, an alternative method with more easily available tools has been developed. Preseparation on a small conventional liquid chromatographic column was optimized to achieve robust separation between the MOSH and the MOAH, but also to keep out the wax esters from the MOAH fraction. This was achieved by mixing a small portion of silica gel with silver nitrate into highly activated silica gel and by adding toluene into the eluent for the MOAH. Toluene was also added to the MOSH fraction to facilitate reconcentration and to serve as a keeper preventing loss of volatiles during solvent evaporation. A 50 µl volume was injected on-column into GC-FID to achieve a detection limit for MOSH and MOAH below 1 mg/kg in most foods.

Evidence for cosmetics as a source of mineral oil contamination in women.

BACKGROUND: There is strong evidence that mineral oil hydrocarbons are the greatest contaminant of the human body, amounting to approximately 1 g per person. Possible routes of contamination include air inhalation, food intake, and dermal absorption. The present study aims to identify the most relevant sources of mineral oil contamination. METHODS: One hundred forty-two women undergoing elective cesarean section were enrolled. A specimen of subcutaneous fat was removed prior to wound closure. On days 4 and 20 postpartum, milk samples were collected from the women. Fat and milk samples were analyzed for mineral oil saturated hydrocarbons (MOSH). All women completed a questionnaire on personal data, nutrition habits, and use of cosmetics. MOSH concentrations in fat tissue were compared with data from the questionnaire and with MOSH concentrations in corresponding milk samples. RESULTS: The predominant predictor for MOSH contamination of fat tissue was age (p<0.001). Furthermore, body mass index (p=0.001), country of main residence (p=0.03), number of previous childbirths (p=0.029), use of sun creams in the present pregnancy (p=0.002), and use of hand creams and lipsticks in daily life (p=0.011 and p=0.007, respectively) were significant independent determinants. No association was found with nutritional habits. A strong correlation was seen between MOSH concentration in fat tissue (median 52.5 mg/kg) and in the corresponding milk fat sample from day 4 (median 30 mg/kg) (p<0.001) and day 20 (median 10 mg/kg) (p=0.028). CONCLUSIONS: The increase in MOSH concentration in human fat tissue with age suggests an accumulation over time. Cosmetics might be a relevant source of the contamination.

Aromatic hydrocarbons of mineral oil origin in foods: method for determining the total concentration and first results.

An online normal phase high-performance liquid chromatography (HPLC)-gas chromatography (GC)-flame ionization detection (FID) method was developed for the determination of the total concentration of the aromatic hydrocarbons of mineral oil origin with up to at least five rings in edible oils and other foods. For some samples, the olefins in the food matrix were epoxidized to increase their polarity and remove them from the fraction of the aromatic hydrocarbons. This reaction was carefully optimized, because also some aromatics tend to react. To reach a detection limit of around 1 mg kg(-1) in edible oils, an off-line enrichment was introduced. Some foods contained elevated concentrations of white paraffin oils (free of aromatics), but the majority of the mineral oils detected in foods were of technical grade with 20-30% aromatic hydrocarbons. Many foods contained mineral aromatic hydrocarbons in excess of 1 mg kg(-1).

Search for a more adequate test to predict the long-term migration from the PVC gaskets of metal lids into oily foods in glass jars.

As shown previously, the conventional testing procedure for simulating long-term migration from the gaskets of metal closures into oily foods does not adequately reflect reality. It appears to be impossible to accelerate migration to the extent that the situation at the end of the shelf life of a product can be anticipated in a few days or weeks. Therefore, we investigated whether long-term migration could be extrapolated from migration rates determined for new lids. Jars were kept in the normal upright position. Since heat treatment may have a strong temporary impact, migration during the initial heating for pasteurization or sterilization and storage at ambient temperature were determined using different lids. Commercial products were recalled from sales points throughout Europe to determine the real migration over extended periods of time and for jars with differing histories. This migration was compared with data from the short-term testing to investigate whether an empirical relationship could be derived. The results show that the short-term test enables the comparison of lids and plasticizers in the initial phase of migration, but that long-term extrapolation presupposes more complex kinetic modeling. The results also demonstrate that the legal relevance of "official" testing methods should be reconsidered to avoid conflict when food contact materials comply with migration limits in the test but not in actual application.

Activated aluminum oxide selectively retaining long chain n-alkanes. Part I, description of the retention properties.

Aluminum oxide activated by heating to 350-400 degrees C retains n-alkanes with more than about 20 carbon atoms, whereas iso-alkanes largely pass the column non-retained. Retention of n-alkanes is strong with n-pentane or n-hexane as mobile phase, but weak or negligible with cyclohexane or iso-octane. It is strongly reduced with increasing column temperature. Even small amounts of polar components, such as modifiers or impurities in the mobile phase, cause the retention of n-alkanes to irreversibly collapse. Since n-alkanes are not more polar than iso-alkanes and long chain n-alkanes not more polar than those of shorter chains, retention by a mechanism based on steric properties is assumed. The sensitivity to deactivation by polar components indicates that polar components and n-alkanes are retained by the same sites. The capacity for retaining n-alkanes is low, with the effect that the retention of n-alkanes depends on the load with retained paraffins. These retention properties are useful for the pre-separation of hydrocarbons in the context of the analysis of mineral oil paraffins in foodstuffs and tissue, where plant n-alkanes, typically ranging from C(23) to C(33), may severely disturb the analysis (subject of Part II).

Activated aluminum oxide selectively retaining long chain n-alkanes: Part II. Integration into an on-line high performance liquid chromatography-liquid chromatography-gas chromatography-flame ionization detection method to remove plant paraffins for the determination of mineral paraffins in foods and environmental samples.

Aluminum oxide activated by heating to 300-400 degrees C retains n-alkanes with more than about 20 carbon atoms, whereas iso-alkanes largely pass non-retained (with characteristics described in more detail in Part I). This property is useful for the analysis of mineral oil contamination of foods and other matrices: it enables the removal of plant n-alkanes, typically ranging from C(23) to C(33), when they disturb the analysis of mineral paraffins (usually almost exclusively consisting of iso-alkanes). An on-line HPLC-LC-GC-FID method is proposed in which a first silica gel HPLC column isolates the paraffins from the bulk of edible oils or extracts and is backflushed with dichloromethane. In a second separation step, a 10 cm x 2 mm i.d. column packed with activated aluminum oxide separates the long chain n-alkanes from the fraction of the iso-alkanes which is transferred to GC-FID by the on-column interface and the retention gap technique. The retained n-alkanes are removed by flushing with iso-octane.

Contamination of grape seed oil with mineral oil paraffins.

The contamination of 11 commercial grape seed oils with paraffins of mineral oil origin was analyzed by online-coupled HPLC-HPLC-GC-FID and ranged from 43 to 247 mg kg(-1). The analysis of the marc and seeds indicated that the contamination is primarily from the peels. Since superficial extraction of the seeds with hexane removed most of the mineral paraffins, the contamination of the seeds is largely on the surface, perhaps transferred from the peels during storage of the marc. Mechanical purification of the seeds combined with washing with hexane reduced the contamination of the oil by a factor of about 10. The refining process removed 30% of the mineral paraffins, primarily the more volatile components. Oil obtained from the seeds of fresh grapes, including grapes not having undergone any phytochemical treatment, contained clearly less mineral paraffins (up to 14 mg kg(-1)), and the peels were less contaminated, suggesting an environmental background contamination. To this an additional contamination might be added by a treatment of the grapes used for wine making.

Mineral oil paraffins in human body fat and milk.

Paraffins of mineral oil origin (mineral paraffins) were analyzed in tissue fat collected from 144 volunteers with Caesarean sections as well as in milk fat from days 4 and 20 after birth of the same women living in Austria. In the tissue samples, the composition of the mineral paraffins was largely identical and consisted of an unresolved mixture of iso- and cycloalkanes, in gas chromatographic retention times ranging from n-C(17) to n-C(32) and centered at n-C(23)/C(24). Since the mineral oil products we are exposed to range from much smaller to much higher molecular mass and may contain prominent n-alkanes, the contaminants in the tissue fat must be a residue from selective uptake, elimination by evaporation and metabolic degradation. Concentrations varied between 15 and 360 mg/kg fat, with an average of 60.7 mg/kg and a median of 52.5 mg/kg. Mineral paraffins might be the largest contaminant of our body, widely amounting to 1g per person and reaching 10 g in extreme cases. If food were the main source, exposure data would suggest the mineral paraffins being accumulated over many years or even lifetime. The milk samples of day 4 contained virtually the same mixture of mineral paraffins as the tissue fat at concentrations between 10 and 355 mg/kg (average, 44.6 mg/kg; median, 30 mg/kg). The fats from the day 20 milks contained <5-285 mg/kg mineral paraffins (average, 21.7; median, 10mg/kg), whereby almost all elevated concentrations were linked with a modified composition, suggesting a new source, such as the use of breast salves. The contamination of the milk fat with mineral paraffins seems to decrease more rapidly than for other organic contaminants, and the transfer of mineral paraffins to the baby amounts to only around 1% of that in the body of the mother.

Assessment of epoxidized soy bean oil (ESBO) migrating into foods: comparison with ESBO-like epoxy fatty acids in our normal diet.

Epoxidized soy bean oil (ESBO) was found to be toxic for rats, but the toxic constituent is unknown. It became an issue as the migration from the gaskets in the lids for jars into oily foods regularly far exceeds the European legal limit (overall migration limit and specific migration limit derived from the tolerable daily intake (TDI)). In the context of risk management it was of interest to determine the epoxidized fatty acids of ESBO in those foods of our normal diet which are expected to contain the highest concentrations, i.e., oxidized edible oils (including degraded frying oils), fried foods, bakery ware and roasted meat. The contribution of epoxy oleic acid from ESBO to our diet turned out to be negligible. If this acid were the toxic component in ESBO, the toxicological assessment would primarily be a warning regarding oxidized fats and oils. The contribution of diepoxy linoleic acid from ESBO might be similar to the exposure from oxidized fats and oils of our diet, whereas the intake of triepoxy linolenic acid from ESBO exceeds that from normal food by around two orders of magnitude. Hence use of an epoxidized edible oil virtually free of linolenic acid would be inconspicuous in our diet.

Compositional GC-FID analysis of the additives to PVC, focusing on the gaskets of lids for glass jars.

A gas chromatographic (FID) method is described which aims at the quantitative compositional analysis of the additives in plasticized PVC, particularly the plastisols used as gaskets for lids of glass jars. An extract of the PVC is analysed directly as well as after transesterification to ethyl esters. Transesterification enables the analysis of epoxidized soya bean and linseed oil (ESBO and ELO) as well as polyadipates. For most other additives, the shifts in the chromatogram resulting from transesterification is used to confirm the identifications made by direct analysis. In the gaskets of 69 lids from the European market used for packaging oily foods, a broad variety of plastisol compositions was found, many or possibly all of which do not comply with legal requirements. In 62% of these lids, ESBO was the principal plasticizer, whereas in 25% a phthalate had been used.

Injector-internal thermal desorption from edible oils. Part 2: chromatographic optimization for the analysis of migrants from food packaging material.

Injector-internal thermal desorption from edible oil or fat is a convenient sample preparation technique for the analysis of solutes in lipids or extracts from fatty foods. The injector temperature is selected to vaporize the solutes of interest while minimizing evaporation of the bulk material of the oil. This technique has been in routine use for pesticides for some time. Now its potential is explored for migrants from food contact materials, such as packaging, into simulant D (olive oil) or fatty/oily food, which means extending the range of application towards less volatile compounds. The performance for high boiling components was investigated for diisodecyl phthalate (DIDP) and diundecyl phthalate (DUP). Since the injector temperature needs to be as high as 260degreesC, some bulk material of the oil enters the column and must be removed after every analysis. This is achieved by a coated precolumn backflushed towards the end of each analysis. Desorption of the solutes is particularly efficient in the initial phase, when a thin sample film is spread on the liner wall, and is largely determined by the diffusion speed in the oil after the latter has contracted to droplets. An increased carrier gas flow rate during the splitless period supports the transfer into the column. It is concluded that the technique is attractive for migrant analysis, with DUP being at the upper limit of the boiling point.

Injector-internal thermal desorption from edible oils. Part 1: visual experiments on sample deposition on the liner wall.

Injector-internal thermal desorption from edible oils or fats enables the analysis of a wide range of compounds in oils or extracts of fatty food without prior removal of the sample matrix. The oil or fat is deposited onto the wall of the injector liner. The solutes of interest are evaporated, leaving behind the sample matrix. The injector is kept at a temperature volatilizing the solutes of interest, but minimizing evaporation of the bulk material of the oil. This technique was optimized regarding sample deposition on the liner wall (Part 1) and desorption of high boiling compounds, such as migrants from food packaging materials into simulant D (olive oil) or fatty food (Part 2). The sample liquid should be transferred to the liner wall and spread to a thin film in order to facilitate the release of high boiling components. Visual experiments with perylene-containing solutions showed that the oil must be diluted to reduce the viscosity (separation from the needle tip). The oil concentration should not exceed 20% in order to rule out that squirting sample liquid drops to the bottom of the vaporizing chamber. Further dilution to about 10% oil improves spreading of the liquid to a thin film. A rather high boiling solvent should be used, such as n-butyl acetate, to prevent thermospray at the needle exit and violent evaporation from the liner wall with sputtering liquid. Using a 5-mm ID liner, 5-10-microL injections of 10-20% oil solutions were at the upper limit.

Epoxidized soy bean oil migrating from the gaskets of lids into food packed in glass jars. Analysis by on-line liquid chromatography-gas chromatography.

The migration of epoxidized soy bean oil (ESBO) from the gasket in the lids of glass jars into foods, particularly those rich in edible oil, often far exceeds the legal limit (60 mg/kg). ESBO was determined through a methyl ester isomer of diepoxy linoleic acid. Transesterification occurred directly in the homogenized food. From the extracted methyl esters, the diepoxy components were isolated by normal-phase LC and transferred on-line to gas chromatography with flame ionization detection using the on-column interface in the concurrent solvent evaporation mode. The method involves verification elements to ensure the reliability of the results for every sample analyzed. The detection limit is 2-5 mg/kg, depending on the food. Uncertainty of the procedure is below 10%.