GC-MS analysis of allergens in plant oils meant to cosmetics

Article received on 20/06/2006 accepted on 20/12/2006 Abstract: Cutaneous allergy occurs mainly as a result of the use of domestic products and cosmetics. Some fragrances, present in these products, may contain compounds that are responsible for allergy (allergens). The European Council offered a Directive limiting the level of 26 allergens found in cosmetics. GC-MS technique was used to determine the retention times of 25 allergens, determine detection and quantification limits and make calibration with standard solution of each allergen in concentrations ranging from 10 to 200 mgL (21 allergens) and 50 to 200 mgL (4 allergens). Quantification was performed by the use of 2 internal standards (tetradecane and hexadecane). Seven oils issued from plants were studied by GC-MS. For all of them, the concentration of potential allergens was lower than their minimum detectable level. The alcoholic solution of extracts issued from different samples of oil did not demonstrate the presence of any quantifiable allergen, even when was concentrated 25 times. GC-MS could be a useful technique in the identification and, if necessary, quantification of allergen in ingredients meant to cosmetics.


Introduction
Allergy falls under three main categories: 1) respiratory such as pollens (trees and herbs) and hair of animals (cats); 2) alimentary such as fruits (strawberry), shellfish, fish, eggs, pharmaceuticals (antibiotics) and 3) cutaneous such as bites of insects (bees and wasps) and industrial products such as domestic products and cosmetics.Cutaneous allergy is accompanied by eczema, hives or oedema of Quincke.The treatment includes the use of antihistamines, corticosteroids and desensitisation as well.The most common reaction to fragrance materials is the allergic contact dermatitis [1].The increased use of perfumes, essential oils and plant extracts in cosmetics and deodorants, has involved an increased incidence of allergy [2][3][4][5][6][7][8][9][10].In addition, domestic products, containing fragrances, have increased eczema of the hands [11][12][13].Essential oils are volatile oily substances immiscible with water and usually have a density value lower than one.They have a complex chemical composition (mono-and sesqui-terpenes, oxidised substances and aromatics) [14].In most cases, the essential oils can be obtained by aqueous distillation, squeezing out from Citrus epicarps or dry distillation.Some essential oils can be modified.Redistilled essential oils are deprived of one or several constituents.Oil and plant extracts may contain ingredients showing, a pharmacological activity inconsistent with a cosmetic use, or potential risk of toxicity to the consumer.The European Council presented a Directive relating to cosmetic products.Twenty six fragrances have been considered as cutaneous contact allergens.The European Union has decided that they should be kept below 100 ppm in products that are designed to be applied to the body and then rinsed off, such as soaps, and must be below 10 ppm in products that are designed to remain on the body throughout the day, such as perfumes.This Directive can be of great value for the consumers capable of reacting to these components and for helping the dermatologists in the diagnosis of cutaneous allergenic reactions [15].The allergens may be present in natural essential oils or obtained by synthetic ways.Allergenic activity could be due to oxidised derivatives, such as peroxides, as in case of limonene and linalool.On the contrary, phenols present in natural essential oils (lavender), have antioxidant properties and can protect from allergenic reactions [16][17][18][19][20][21][22][23][24].Several publications have described the use of Gas Chromatography-Mass Spectrometry (GC-MS) technique in the qualitative and quantitative analysis of many compounds present in essential oils and perfumes [25][26][27][28][29][30][31][32].The detection and quantification limits for each allergen were determined by GC-MS technique.The regression equations of each allergen standard used in calibration were presented.It also was important to search and determine the allergens in several plant oils.

Experimental Allergens
The allergens and their CAS No are listed in table 1.All standards were purer than 95%.To this list methyl-2-octynoate was added.In contrary, the last two components of the list (oak moss and tree moss extracts) were not injected into GC, because they are non volatile.

Sample preparation
Pure essential oils were not tested, because GC studies were described in several references [26][27][28][29][30]. Seven oils issued from plants and produced by S.A. Huiles Bertin (60330 -Lagny-Le-Sec, France) were studied: virgin oil of Sisymbrium, virgin oil of Hazel Nuts, virgin oil of Sweet Almonds, virgin oil of Souchet, virgin oil of Cashew Nuts, virgin oil of Borage, castor oil.2.5 to 5g of vegetable oils was stirred with 10.0 mL alcohol (96%) or hexane during 15 minutes.Only the alcoholic (or hexanic) solution was injected into GC.Sometimes, this solution was concentrated 10 or 25 times, by heating at 75 °C, for 15 min, under nitrogen sweeping.

Gas chromatography conditions
Tests were done on GC-MS instrument, Varian 3400 -Saturn 4D ion trap.A capillary column (30 m × 0.32 mm ID × 1 lm DB5 film with 5% phenylmethyl silicon).Initial temperature was held for 1 min at 60 °C, then heated at 3 °C min -1 until 200 °C.Then the temperature was held at 200 °C for 30 min.Temperatures of the injector and detector were 250 and 285 °C, respectively.The carrier gas used was helium and supplied at 138 kPa head pressure.A 1 lL sample was injected in splitless mode.

Detection limit
The concentration of allergen at which the peak was equal or higher than 3 times the background noise was retained (Detection Limit or DL).

Qualitative analysis
Identification was made according to retention times and mass spectra.Three MS libraries were used: NIST, Wiley and TR.

Quantification limit
The concentration of allergen at which the peak was equal or higher than 10 times the background noise was retained (Quantification Limit or QL).

Quantitative analysis
It was based on internal standardisation, using tetradecane and hexadecane as internal standards.Five standard solutions, included in the range of 10 to 200 mg L -1 of each allergen, were used in calibration.Intervals of calibration were 50 to 200 mg L -1 for benzyl alcohol, anisyl alcohol, cinnamyl alcohol and coumarin.

Quantification repeatability
Ten tests were done showing a relative standard deviation lower than 2.5% [28].

Determination of detection and quantification limits of each allergen
First, a mixture of 25 allergen standards (mother solution), each about 1 to 2 g L -1 was studied.The chromatogram is shown in figure 1.

Fundamentals of allergen identification
The following steps were performed: 1) In the chromatogram of unknown plant oil, the peaks with RT close to those of allergen standards was searched.
2) If the peak height was 3 times lower than the background noise, the peak was considered absent.
3) If the height of the peak was 3 times higher than the background noise, the peak was considered present.4) Identification was done when the theoretical RT given by TR library corresponded to the observed RT of the unknown peak, and we compared mass spectra of the unknown peak with those of allergen standards, and after with those standards present in libraries.5) If a doubt arose about the peak, we re-injected the unknown plant oil enriched with the expected standard.Identification will be confirmed if only one peak appeared without a shoulder.6) If the height of the peak was 10 times higher than the background noise, quantification would be possible.

Quantification of the allergens present in the samples
The calibration of each allergen was done with the use of two internal standards: tetradecane and hexadecane (table 4).The choice of internal standard depends on the presence of interfering peaks.The retention times (sec) of several saturated hydrocarbons were tested and found to be 520, 790, 1390, 1967, 2230 and 2490 for nonane, decane, dodecane, tetradecane, pentadecane and hexadecane, respectively.A linear correlation was observed, Y = 284.39X-2037.8,Y = retention time, X = number of carbons, R = 0.9998 and p < 0.001.In this paper the choice of tetradecane and hexadecane was based on the best results obtained by analysing different samples.They appeared in retention times that were fairly different from other peaks present in all the samples.

Results of studied samples
Our chromatographic procedure seemed to be slightly longer than that described before [30], but our conditions showed good resolution for about 200 terpenoid compounds determined in general essential oil studies.
From the 25 studied standard allergens, 19 showed a retained DL lower than 13 mg L -1 and 5 were between 30 and 50 mg L -1 .These variations are well explained by the form of the peaks.A narrow and high peak gave a value lower than 13 mg L -1 , in contrary, a broad and low height peak gave a higher value of DL.GC-MS is considered as a good technique for the determination of volatile substances.Results were obtained with good repeatability.To avoid a possible volatilisation or degradation of allergens, ethanolic solutions were prepared and injected quickly.Solutions were not kept more than one day at 4 °C.Ethanol is a good solvent for extraction.It dissolves several fractions of vegetable oils, but sometimes hexane will be prefered.For a good extraction, the solvent should be added in excess (2-5 times) more than the volume of the sample.So, to increase the sensitivity, it was necessary to concentrate this alcoholic (or hexanic) extract, 10 to 25 times.But in this case, losses were observed due to volatilisation.In each case, standard solutions were prepared and concentrated in the same conditions as that the studied samples.Allergen loss percentage was then determined.Limonene, benzyl alcohol, linalool, citronellol, citral, cinnamaldehyde, anisyl alcohol and cinnamyl alcohol lose 55% (75%), 15.5% (23%), 10% (23%), 13% (19%), 20% (31%), 20% (20.5%), 26% (33%) and 10% (20.5%) when were concentrate 10 times (25 times), respectively.All the others lost less than 10%.Concentration of 10 and 25 times can improve the determination of DL.

Result expression
Results expression would be done according to the following model: 5.0434 g of Borage oil were stirred with 10.0 mL of alcohol (96%) during 15 min.After centrifugation, the alcoholic volume was 9.0 mL.Then, 10 times concentrated solution was injected into GC.Limonene was considered present (3 times the background noise) and corresponded to the Minimum Detectable Level (MDL), in this concentrated solution (VC = 0.90 mL).DL * VC = 2.1 * 0.90 * 10 -3 = 1.89 lg of limonene in the sample.
If we take into account the initial mass of plant oil (5.0434 g) and the loss of 55% of limonene in 10 times concentration, corrected MDL will be: (1.89 * 100) / 5.0434 * (100 -55) = 0.83 lg of limonene per 1 g of Borage oil.All tested oils had potential allergen level lower than DL.
The fragrance mix (FM) is the main screening chemicals used to test for fragrance allergy.FM is composed of 1% of the following 8 chemicals: cinnamal, cinnamyl alcohol, eugenol, isoeugenol, hydroxycitronellal, geraniol, oak moss absolute and amylcinnamaldehyde.Dermatologists have used FM mixture as a standard to test cutaneous allergy.The composition of this mixture should be extended to include other allergens [33][34][35][36][37].The GC-MS technique, described here, could be applied in allergen determination, in plant oils meant to cosmetics.Evaluation of allergens in cosmetics will be less easier.

Figure 1 .
Figure 1.The chromatogram of the 25 allergen standard mixture.

Table 1 .
List of Allergens from the European Directive 2003/15/CE.

Table 2 .
Retention times (RT) determination of the allergen standards.
* was not cited in the EC Directive.Standard solutions were prepared in absolute ethanol in a concentration about 1g L -1 .SD = standard deviation; RSD = relative standard deviation.OCL VOL.N°2 MARS-AVRIL 2007 111

Table 3 .
Retained detection limit (DL) and quantification limit (QL) for each allergen in mixture.

Table 4 .
The calibration data for compounds listed as skin-sensitising ingredients.