In-vial Micro-matrix-solid Phase Dispersion for the Analysis of Fragrance Allergens, Preservatives, Plasticizers, and Musks in Cosmetics

Fragrance allergens, preservatives, plasticizers, and synthetic musks are usually present in cosmetic and personal care products formulations and many of them are subjected to use restrictions or labeling requirements. Matrix solid-phase dispersion (MSPD) is a very suitable analytical technique for the extraction of these compounds providing a simple, low cost sample preparation, and the possibility of performing both extraction and clean-up in one step, reducing possible contamination and analyte losses. This extraction technique has been successfully applied to many cosmetics ingredients allowing obtaining quantitative recoveries. A new very simple micro-MSPD procedure performing the disruption step in a vial is proposed for the gas chromatography-mass spectrometry (GC-MS) analysis of 66 chemicals usually present in cosmetics and personal care products. The method was validated showing general recoveries between 80% and 110%, relative standard deviation (RSD) values lower than 15%, and limits of detection (LODs) below 30 ng·g −1. The validated method was applied to a broad range of cosmetics and personal care products, including several products intended for baby care.


Introduction
Fragrances and preservatives are common ingredients in cosmetics and personal care products.Fragrances provide nice and attractive scents and preservatives are used to prevent microbial growth because the aqueous nature of many personal care products is an optimal medium for microbial growth.European legislation [1] requires the monitoring of 26 volatile compounds, the so-called potentially allergen substances (PAS) or fragrance allergens.Their presence must be indicated in the list of ingredients when their concentrations exceed 0.01% for rinse-off products, and 0.001% for leave-on products.Of these 26 substances, 24 are chemically defined volatile compounds whereas the other two are natural moss extracts.One of these 24 fragrance allergens, lyral ® , was recently proposed to be transferred to the Annex III (list of substances which cosmetic products must not contain except subject to restrictions) to Annex II (list of substances prohibited in cosmetic products).Also, pinene and methyleugenol were included in the referred study; pinene is proposed to be labelled when its concentration exceeds 0.01% for rinse-off products, and 0.001% for leave-on products, whereas methyleugenol has been banned in cosmetics and personal care products for some years, and now it is included in Annex III.
Parabens are the most frequently used preservatives (their maximum concentration in cosmetics and personal care products is 0.4% for a single ester and 0.8% for mixture of esters).Its extended use is due to their broad antimicrobial spectrum and low cost [2,3].Although these compounds are not mutagenic agents, recent studies have reported that certain parabens have been associated with genotoxicity, allergies and may also act as antiandrogens [4][5][6].In recent years, another preservative, phenoxyethanol, is increasing its use as substitute of parabens.According to the European regulation [1], the maximum concentration permitted for this compound is 1% regardless of its use.However, a recent study reported by the France National Agency for Security of Medicaments (ANSM) proposed not using phenoxyethanol in products intended for children under 3 years and to reduce the maximum permitted concentration (0.4%) in other personal care products [7].Triclosan (2,4,4′-trichloro-2′hydroxydiphenyl ether) and the bromine-containing preservative bronidox, are also preservatives present in personal care products.Their maximum permitted concentrations according European legislation is 0.3% and 0.1%, respectively.IPBC (iodopropynyl butylcarbamate) is not permitted in products for children under 3 years of age, except in bath products, shower gels and shampoo.The antioxidants butylated hidroxyanisole (BHA) and butylated hydroxytoluene (BHT) can be used without restrictions.
Synthetic musks are other chemical compounds usually present in personal care products under the term "fragrance" or "parfum".Synthetic musks are used as an alternative for natural musks.The European regulation has forbidden the use of three nitromusks: musk ambrette, musk moskene and musk tibetene due to their bioaccumulative properties [8].Another two nitromusks (musk ketone and musk xylene) are allowed with restrictions [1].
In order to guarantee product safety, the development of analytical methods is mandatory in cosmetic quality control.In this way, several analytical methods to determine fragrance allergens, preservatives, plasticizers, and/or musks in cosmetics and personal care products have been reported.A summary of the more recent extraction and analysis techniques for the analysis of these compounds in different cosmetic matrices can be found in recent reviews [10][11][12][13].
Matrix solid-phase dispersion (MSPD) is a very suitable analytical technique for the extraction of contaminants in environmental and other matrices [14] as well as to determine fragrances, preservatives, plasticizers and musks in cosmetic samples.This technique is primarily used because of its flexibility and selectivity providing efficient and low cost extractions; the possibility of performing extraction and clean-up in one step is one of their main advantages [15][16][17][18][19][20][21].Also, its miniaturizing allows reducing the amount of sample, reagents and solvents required.MSPD combines different aspects of several analytical techniques, performing sample disruption while dispersing the components of the sample on and into a solid support, thereby generating a chromatographic material that possesses a particular character for the extraction of compounds from the dispersed sample [15].This extraction technique allowed obtaining quantitative recoveries for many cosmetic ingredients [16,19,20,22].For very volatile compounds such as pinene and limonene, that are easily lost during extraction processes [23], MSPD can constitute a good alternative to lower analyte losses [20].
The aim of the present study is to compare the performance of two micro-MSPD procedures, performing the sample disruption in mortar and also in vial, for the gas chromatography-mass spectrometry (GC-MS) analysis of 66 compounds including fragrance allergens, preservatives, plasticizers, and musks, usually present in cosmetics and personal care products.All these families of compounds are subjected to restrictions according international regulation.
Individual stock solutions were prepared in acetone, isooctane or methanol.Further dilutions and mixtures were prepared in acetone or ethyl acetate.Solutions were stored in amber glass vials at −20 °C.All solvents and reagents were of analytical grade.
Metallic, glass, and ceramic materials; sorbents (Florisil and sodium sulphate anhydrous) and the glass wool for laboratory use (Sigma-Aldrich) were baked at 230 °C for 12 h before use to eliminate possible phthalate contamination.All materials were allowed to cool down wrapped with aluminum foil and Florisil and sodium sulphate anhydrous in desiccator.
Samples of cosmetics and personal care products from national and international brands were obtained from local sources.They included leave-on and rinse-off products such as shampoo, shower gel, body milk, sunblock, among others, including products intended for babies.Until their analysis, samples were kept in their original containers at room temperature.

Micro-Matrix Solid-Phase Dispersion (MSPD)
Cosmetic samples (0.1 g) were exactly weighted into a 10-mL glass vial and spiked with 25 µL of each surrogate solution (10 µg•mL −1 ) containing benzyl alcohol-d 7 , MeP-d 4 , PrP-d 4 and DEHP-d 4 .Then, the sample was gently blended with 0.2 g of a drying agent (anhydrous Na 2 SO 4 ), and 0.4 g of the dispersing sorbent (Florisil), into the vial or in a porcelain mortar, using a glass rod or a porcelain pestle, respectively, until a homogeneous mixture was obtained (ca. 5 min).The mixture was transferred into a glass Pasteur pipette (approximately 150 mm), with a small amount of glass wool at the bottom, containing 0.1 g of Florisil (to obtain a further degree of fractionation and sample clean-up).Finally, a small amount of glass wool was placed on top of the sample before compression with a spatula.Elution with ethyl acetate was made by gravity flow, collecting the extract into a 1 mL volumetric flask.Then, 12.5 µL of PCB-30 internal standard solution (1 µg•mL −1 ) was added.The micro-MSPD extracts diluted when necessary were directly analyzed by GC-MS.Fortified samples were spiked with 20 µL of the corresponding acetone solution of the target compounds to get the desired final concentration and submitted to the same process described above.The optimization of the experimental conditions (amount of sample, solvent, dispersant and volume elution) has been described elsewhere [19,20].Figure 1 illustrates the described micro-MSPD process.

Gas Chromatography-Mass Spectrometry (GC-MS) Analysis
The analysis was performed using an Agilent 7890A (GC)-Agilent 5975C inert MSD with triple axis detector and an Agilent 7693 autosampler from Agilent Technologies (Palo Alto, CA, USA).The temperatures of the transfer line, the quadrupole and the ion source were set at 290, 150 and 230 °C, respectively.Electronic impact (EI) was used as ionization technique.The system was operated by Agilent MSD ChemStation E.02.00.493 software.

GC-MS Performance
The chromatographic conditions were optimized to achieve an efficient separation of 66 target compounds frequently used in cosmetics and personal care products: 26 fragrance allergens, 13 preservatives, 15 plasticizers (phthalates and adipates) and 12 musks.For GC-MS analysis, the mass spectra detector (MSD) was operated in the selected ion monitoring (SIM) mode, monitoring three ions per compound.Table 2 shows the quantification and identification ions, and the retention time of the compounds.Chromatograms of a standard solution containing 200 ng•mL −1 of target compounds (DIHP, 400 ng•mL −1 ) are shown in Figure 2.   The GC-MS method performance parameters for the 66 target compounds are summarized in Table 3. Regarding the instrumental linearity, the method exhibited a direct proportional relationship between the amount of each analyte and the chromatographic response.Calibration standards in ethyl acetate were prepared covering a concentration range from 10 to 1000 ng•mL −1 (anise alcohol, cinnamyl alcohol, amylcinnamyl alcohol, triclosan, and musk ketone, 20-1000 ng•mL −1 ; IPBC, 50-1000 ng•mL −1 ; di-iso-heptyl-phthalate (DIHP), 100-4000 ng•mL −1 ; and farnesol, 250-1000 ng•mL −1 ).Correlation coefficients R ≥ 0.9915 were generally obtained.Method precision was studied within-a-day (n = 3) and among-days (n = 6) at 250 ng•mL −1 (other concentration levels, 50, 500 and 1000 ng•mL −1 were calculated, data not shown).Relative standard deviation (RSD) values ranged from 1.7% to 9.5% for intra-day analysis, and between 1.8 and 10% for inter-day analysis.Instrumental detection limits (IDLs) were in all cases calculated as the concentration giving a signal-to-noise of three (S/N = 3) since none of the target compounds were detected in the solvent chromatographic blanks and they were at the low ng•mL −1 with values in general below 6 ng•mL −1 (farnesol, IPBC, and DIHP, 70 ng•mL −1 , 10 ng•mL −1 , and 24 ng•mL −1 , respectively).The phthalate DIHP is complex mixtures of isomers, and the chromatographic signal is composed of several chromatographic peaks.

Analytical Method Performance
Complete method quality parameters were evaluated using real cosmetic samples and the results are shown in Tables 4 and 5.In this way, recovery studies were carried out by applying the optimized method to two samples spiked at three levels of concentration: 2, 10 and 20 µg•g −1 .These samples are a regenerating cream (leave-on) and a shampoo (rinse-off); they were selected for recoveries studies since the leave-on sample was labeled as perfume-free and preservative-free, and the rinse-off sample was almost free of the target compounds (only contained MeP, BHT, and PrP).In any case, previous analyses of the samples showed the presence of some of the target compounds, and these initial concentrations were taken into account to calculate the recoveries.Recoveries were higher than 90% for the most of the studied compounds (see Tables 4 and 5 for leave-on and rinse-off samples, respectively) regardless of using vial or mortar for the MSPD disruption step.In the case of the most volatile compounds, pinene recovery was 70% and 35%, for leave-on and rinse-of samples, respectively; and for limonene, recovery presented an average value of 75% employing a vial, whereas lower recoveries were obtained employing a mortar.Recovery study was extended to three other cosmetic matrices (shampoo, sunblock product, body milk) that were fortified at 10 µg•g −1 .Results are presented in Table 6, and demonstrate the quantitative recovery of the compounds.Precision was evaluated attaining RSD values generally lower than 10% (see also Tables 4 and 5).
Figure 3 shows a comparison of the results obtained using vial or mortar for the micro-MSPD for a real leave-on sample containing 23 target analytes (hands cream).Obtained responses are equivalent employing mortar or vial for the disruption step, excluding pinene and limonene for which responses were higher using vial.Limits of detection (LODs) were calculated as the compound concentration giving a signal-to-noise ratio of three (S/N = 3).As shown in Table 3, LOD values for the fragrance allergens ranged from 0.0118 to 0.0604 µg•g −1 (excluding farnesol, 0.700 µg•g −1 ), for preservatives, these values were between 0.0053 and 0.0595 µg•g −1 (excluding IPBC) and for plasticizers and musks LODs values ranged from 0.0026 to 0.1200 (excluding DIHP).
Therefore, the proposed micro-MSPD method using a vial instead of a mortar for the disruption and dispersion step can be considered suitable for the determination of fragrance allergens, preservatives, musks, and plasticizers in cosmetic and personal care products.It is highly recommended to decrease losses of most volatile fragrances such as pinene and limonene during sample preparation.For these compounds, the increase of temperature in the mortar disruption step is unfavorable for their quantitative extraction, whereas for in-vial disruption the generated heat is lower, and the most volatile compounds can be extracted lossless; also, in-vial disruption reduces extraction steps, providing a quicker extraction procedure.

Application to Real Samples
Finally, the validated method was applied to the analysis of 18 real cosmetic and personal care products, including five rinse-off (shower gel, shampoos and baby liquid soap) and 13 leave-on (sunblock, after sun, body milk, hands cream, deodorants, among others) products, which represent a wide variety of personal care products.Results are shown in Table 7. Forty-eight of the 66 targets were found in the samples, with a minimum of 12 and a maximum of 28 compounds in each sample, at global concentrations ranging from 0.043% to 1.6%.It is worthy to note that the sample containing more targets is a baby body care lotion (sample S16).

Fragrance Allergens
Twenty-two of the 26 fragrance allergens were found in the analyzed samples.Linalool was detected in 83% of the samples at concentration values up to 0.1%.Also limonene, coumarin, benzyl alcohol, and benzyl salicylate were found in many samples at concentrations below 800 µg•g −1 .Other fragrance allergens were detected in 2-12 samples.It is remarkable the presence of farnesol at high concentrations (>0.2%) in two leave-on samples (S12 and S13).Regarding the number of compounds per sample, three leave-on samples (S7, S16 and S18) contained 15 target allergens.In the other samples, the number of compounds was 3-12.

Preservatives
Phenoxyethanol was the most frequent found preservative (83% of the samples) at concentration values higher than 0.1% (1000 µg•g −1 ) in nine samples.In sample S11, phenoxyethanol concentration (1.5%) surpassed the maximum concentration permitted by European regulation (see Table 1).In the case of parabens, six of the seven targets were found in the analyzed samples.The most common was PrP (67%) at 0.1% in two leave-on samples (S9 and S10).Other parabens, MeP, EtP, BuP, iBuP, and iPrP, were found in 11, seven, five, four and one samples, respectively.Triclosan was detected in two samples, at very high concentration in a deodorant (S13), reaching the limit established by the European legislation (0.3%).BHT, IPBC, and BHA were detected in 13, three and one sample, respectively.The highest number of preservatives was found in S5 and S6, with eight targets.It should be noted that S5 is a care cream intended for babies.A hand cream (S11) does not comply with European restrictions regarding PhEtOH (>1%) and total paraben concentration (> 0.8%).

Plasticizers
DEP was found in all analyzed samples at concentration levels below 432 µg•g −1 , except in the deodorant S14 (1539 µg•g −1 ).Two banned phthalates (DBP and DEHP) were detected in 13 and 11 samples, respectively, at concentrations between 0.05 and 4.8 µg•g −1 .It should be noted that sample S11 (a hand cream) contained a DBP concentration >0.1%.In the other personal care products, the number of plasticizers was 1-6, highlighting the presence of DEHA at very high concentration (1.6%) in make-up (leave-on sample, S17).

Musks
Galaxolide was found in 72% of the samples at concentrations below 625 µg•g −1 .Celestolide and ambrettolide were found in 17% of the samples at concentration levels between 0.11 and 35 µg•g −1 .The restricted musks tonalide, ketone, xylene and phantolide were detected in at least one sample (tonalide in five samples); at concentrations fulfilling the EU limits, with the exception of musk ketone, found at >0.042% in sample S16, a baby moisturizing lotion.

Conclusions
A micro-MSPD-GC-MS method has been proposed for the determination of four families of compounds extensively used in cosmetics and personal care product formulations: fragrance allergens, preservatives, plasticizers, and synthetic musks.This study included 66 chemicals subjected to restrictions according European legislation.We compared the performance of two micro-MSPD procedures for the extraction of the targets, performing the sample disruption in mortar and in vial.The proposed in-vial method allows analytes extraction in less than 5 min, providing a quick and low cost extraction procedure with lower losses of the more volatile compounds.The method was validated showing satisfactory linearity, sensitivity, accuracy, and precision, with recoveries higher than 90% and RSD values below 10%.Finally, the method was applied to real cosmetic samples including different matrices to demonstrate the method performance.Forty-eight of the 66 targets were detected in the analyzed samples.Several compounds were present at concentrations higher than 0.1%, and two of the samples did not comply with

Figure 3 .
Figure 3. Comparative results between in vial or mortar µ-MSPD for a leave-on sample (hands cream).

Table 1 .
Target compounds: chemical names, suppliers, purity, CAS and European restrictions.

Table 2 .
Retention time, quantification and identification ions.

Table 3 .
Quality parameters of the method.

Table 4 .
Recoveries of fragrance allergens, preservatives, plasticizers and musks in a leave-on sample (regenerating cream) fortified at three concentration levels, analyzed by the proposed method µMSPD-GC-MS.

Table 5 .
Recoveries of fragrance allergens, preservatives, plasticizers and musks in a rinse-off sample (shampoo) fortified at three concentration levels, analyzed by the proposed method µMSPD-GC-MS.

Table 6 .
Cont.See initial concentration in Table7.n.c: not calculated since initial sample concentration is higher than the spiked level. a