Simultaneous determination of cosmetics ingredients in nail products by fast gas chromatography with tandem mass spectrometry

Journal of Chromatography A, 1446 (2016) 134–140

Contents lists available at ScienceDirect

Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma

Simultaneous determination of cosmetics ingredients in nail products by fast gas chromatography with tandem mass spectrometry Wanlong Zhou ∗ , Perry G. Wang, James B. Wittenberg, Diego Rua, Alexander J. Krynitsky U.S. Food and Drug Administration, Office of Regulatory Science, CFSAN/HFS-717, 5100 Paint Branch Parkway, College Park, MD 20740-3835, USA

a r t i c l e

i n f o

Article history: Received 3 February 2016 Received in revised form 31 March 2016 Accepted 1 April 2016 Available online 2 April 2016 Keywords: Nail products Cosmetics GC–MS/MS Toluene N-methylpyrrolidone 2,4-Dihydroxybenzophenone Diethylene glycol dimethacrylate

a b s t r a c t A rapid and sensitive gas chromatography with tandem mass spectrometry (GC–MS/MS) method has been developed and validated to quantitatively determine cosmetic ingredients, such as toluene, N-methylpyrrolidone, 2,4-dihydroxybenzophenone (benzophenone-1, BP-1), and diethylene glycol dimethacrylate, in nail products. In this procedure, test portions were extracted with acetone, followed by vortexing, sonication, centrifugation, and filtration. During the extraction procedure, BP-1 was derivatized making it amenable to GC–MS analysis, using N,O- bis(trimethylsilyl) trifluoroacetamide. The four ingredients were quantified by GC–MS/MS in an electron ionization mode. Four corresponding stable isotopically labeled analogues were selected as internal standards, which were added at the beginning of the sample preparation to correct for recoveries and matrix effects. The validated method was used to screen 34 commercial nail products for these four cosmetic ingredients. The most common ingredients detected in the nail products were toluene and BP-1. Toluene was detected in 26 products and ranged from 1.36 to 173,000 ␮g/g. BP-1 ranged from 18.3 to 2,370 ␮g/g in 10 products. Published by Elsevier B.V.

1. Introduction Nail products for home and salon use are considered cosmetics and regulated by the Food and Drug Administration (FDA) under the Federal Food, Drug, and Cosmetic Act [1]. Although nail products are not subject to premarket approval by FDA, cosmetics must not be adulterated or misbranded. This means that they must be safe for consumers under labeled or customary conditions of use, and they must be properly labeled. Although consumers are told to carefully read the label to make sure that they know what ingredients are present in the products they will use, recent investigations by FDA and other government agencies have found inconsistencies between ingredients declared on the label and those actually contained in the product [2]. In order to prevent misbranding and adulteration, analytical methods capable of detecting and quantifying compounds of interest are needed. We therefore have developed a method to simultaneously quantitate four cosmetic ingredients, which are commonly used in nail products and may be potentially harmful [3–7]. These ingredients are toluene, N-methylpyrrolidone (also known as N-methyl-2-pyrrolidone, NMP), 2,4-dihydroxybenzophenone

∗ Corresponding author. E-mail address: [email protected] (W. Zhou). http://dx.doi.org/10.1016/j.chroma.2016.04.003 0021-9673/Published by Elsevier B.V.

(benzophenone-1, BP-1) and diethylene glycol dimethacrylate (DEGDMA). The structures of these four compounds are shown in Fig. 1. Toluene is widely used as a solvent to help nail polish apply smoothly and adhere evenly to nails. NMP, a 5-membered lactam structure, is a clear-yellow liquid miscible with water and other common solvents such as ethyl acetate, chloroform, and benzene. Due to its usefulness as an organic solvent, NMP has been used as a solvent and surfactant in the manufacturing of cosmetic products [8]. BP-1 is used as an ultraviolet filter in nail polish to prevent discoloration when the polish is exposed to sunlight or other forms of ultraviolet light [9]. Methacrylate ester monomers are used as artificial nail builders in nail enhancement products. These monomers undergo rapid polymerization to form a hard material on the nail that is then shaped. While ethyl methacrylate is the primary monomer used in nail enhancement products, other methacrylate esters, such as DEGDMA, are also used [6]. There are many methods, such as gas chromatography-flame ionization detector (GC-FID) [10,11], gas chromatography–mass spectrometry (GC–MS) [12–14], GC–MS/MS [15–17], liquid chromatography (LC) [18,19], LC–MS [20] and capillary electrophoresis (CE) [21–23] to determine individual analytes: toluene [10–12], NMP [13], DEGDMA [14,19,20], BP-1 [15–17,18,21,23] in different matrices including nail polish [10], cosmetic products [18,21–23], water [15,16], Biocef (sandoz) tablets [13], dental

W. Zhou et al. / J. Chromatogr. A 1446 (2016) 134–140

135

Fig. 1. Structures of the analytes.

materials [14,19,20]. However, no method has been reported to simultaneously determine these four compounds in cosmetic nail products. The published methods suffered poor recoveries and matrix interferences in complex matrices [12,13]. Time-consuming standard addition was required to quantify BP-1in the untreated water samples [15]. Furthermore, the published methods lack the confirmation of the identity of analytes [10,11,16,18,19]. To minimize the interferences from complex matrices of the nail products, a suitable sample preparation procedure without a cleanup step, including derivatization of BP-1, has been developed and optimized in this study. A GC–MS/MS method, which is more selective and sensitive than previously published procedures, has been developed and fully validated to determine these four analytes in nail products. Stable isotopically labeled analogues of the four native compounds were used as internal standards to compensate for matrix effects and correct any recovery issues. Confirmation of the identity of analytes was performed using corresponding confirmatory selected reaction monitoring (SRM) transitions for each analyte. To the best of our knowledge, this is the first approach to simultaneously quantify toluene, NMP, BP-1, and DEGDMA in nail products. This validated analytical method has been used in a survey to screen 34 commercial nail products. This limited survey will allow the FDA to assess the prevalence of the 4 mentioned analytes in nail products and to determine if additional sampling is warranted. 2. Experimental 2.1. Chemicals Toluene, toluene-d8 , NMP, BP-1, N,O-bis(trimethylsilyl) trifluoroacetamide (BSTFA, derivatization reagent for BP-1), and capillary GC grade acetone were purchased from Sigma-Aldrich (St. Louis, MO, USA). DEGDMA, DEGDMA-d8 , and BP-1-13 C6 were obtained from Toronto Research Chemicals (Toronto, ON, Canada). NMP-d9 was purchased from Cambridge Isotope Laboratories, Inc. (Tewksbury, MA, USA). All of the chemicals were used without further purification. Thirty-four nail products were purchased via the internet. 2.2. Preparation of standard solutions The primary stock solutions of each native analyte and stable isotopically labeled analogue were separately prepared in volumetric flasks at approximately 1000 ␮g/mL in ethyl acetate. The stock solutions were transferred into amber narrow mouth bottles and stored at −18 ◦ C. Working stock 1 (a mixture of four standards, 200 ␮g/mL for each analyte) was prepared by dilution of the primary stock solutions in acetone. Working stock 2 (a mixture of four standards, 10 ␮g/mL for each analyte) was prepared by dilution of Working Stock 1 in acetone. Standard calibration solutions (10 mL in volumetric flasks) were prepared by dilution in acetone using working stocks 1 and 2. The concentration of the calibration standards ranged from 0.100 to 50.0 ␮g/mL for toluene and BP-1, and 0.100 to 5.00 ␮g/mL for NMP and DEGDMA, respectively. For toluene and BP-1, the concentrations for standards 1–7 were 0.100, 0.250, 0.500, 1.00, 5.00, 10.0, and 50.0 ␮g/mL. For NMP and

DEGDMA, the concentrations for the standards 1–7 were 0.100, 0.250, 0.500, 0.750, 1.00, 2.50, 5.00 ␮g/mL. 1 mL of internal standards (a mixture of four standards, 10.0 ␮g/mL for each analyte) was added to each of the standard solutions at a constant concentration of 1.00 ␮g/mL for each analyte. 1 mL of BSTFA (derivatization reagent for BP-1) was added to each of the standard solutions at constant concentration of 10.0% (v;v). BP-1was allowed to react with BSTFA at room temperature (∼25 ◦ C) for 30 min to form the BP-1 derivative. To help the derivatization of BP-1, the standard solutions were shaken well for three 60-s. The standards were transferred into amber GC–MS vials with crimp top caps and stored at −18 ◦ C. 2.3. Sample preparation Approximately 100 ␮L of a nail product was transferred into a 15-mL centrifuge tube in triplicate and accurately weighed (Approximately 100 mg). To each of the tubes 700 ␮L of acetone, 100 ␮L of internal standards (a mixture of four standards, 10.0 ␮g/mL for each analyte), and 100 ␮L of BSTFA (derivatization reagent for BP-1) were added, which resulted in the final volume of 1.0 mL. The tubes were capped tightly during the sample preparation to minimize the loss of acetone, toluene and NMP. The tubes were first vortexed for approximately 3 min to rinse the walls and disperse the sample, and then sonicated for approximately 20 min. The tubes were again vortexed for 3 min and centrifuged at 11,000 revolutions per min for 10 min using an Eppendorf 5804 centrifuge (Hamburg, Germany). The supernatant was filtered through a 0.2 ␮m PTFE filter (Pall Life Sciences, Port Washington, NY) directly into a GC/MS sample vial, and 1.0 ␮L of the filtered extract was injected into the GC–MS/MS. Since nail products had various formula and content of ingredients, each product was first screened to determine whether the sample needed to be diluted to fit into the calibration range or whether a significant interference existed. If a sample did need to be diluted, acetone, the ISs, and BSTFA were used for dilution without sonication, centrifugation and filtration. BP-1was allowed to react with BSTFA at room temperature (∼25 ◦ C) for 30 min to form the BP-1 derivative. The diluted solutions were shaken well for three 60-s. For those samples with significant interferences (e.g., signal to noise ratio of the internal standard was less than 20) or that were difficult to filter, a smaller sample volume (50 ␮L) and more extraction solvent (1.95 mL acetone, 250 ␮L of ISs and 250 ␮L of BSTFA, with a total volume of 2.5 mL) were used to extract the samples with the same procedures as above. 2.4. Instrumentation The study was carried out on an Agilent 7890A GC System coupled to an Agilent 7000 Triple Quad MS (Santa Clara, CA, USA). The data were acquired and processed using MassHunter GC–MS Acquisition and Quantitative Analysis, respectively. Different columns were tested: Restek Rtx® 5 amine (30 m × 0.25 mm × 0.5 ␮m, Bellefonte, PA, USA) and Phenomenex ZB-SemiVolatiles (30 m × 0.25 mm × 0.25 ␮m, Torrance, CA, USA). The ZB-SemiVolatiles column was selected for this study since it resulted in overall best peak shapes for these four analytes. The

136

W. Zhou et al. / J. Chromatogr. A 1446 (2016) 134–140

Table 1 Time segments and SRM scan parameters. Segment Time (min)

Analytes (CAS #)

2.50

Toluene (108-88-3)

5.02

Toluene-d8

4.93

NMP (872-50-4)

9.71

NMP-d9

9.67

DEGDMA (2358-84-1)

12.25

DEGDMA-d8

12.24

BP-1 Deri (131-56-6* )

13.85

BP-1-13 C6 -Deri

13.85

8.00

11.50

13.00

RT (min)

Delta EMV, V

Precursor Ion

Product Ion

Dwell Time, ms

Collision Energy, V

Use

600 600 600 600 600 600 600 600 600 600 600 600 600 600 600 600 600

91.0 92.0 98.0 98.0 99.0 98.0 108.0 106.0 113.0 113.0 143.0 117.0 117.0 343.0 343.0 349.0 349.0

65.0 65.0 70.0 54.0 44.0 70.0 50.0 78.0 69.0 41.0 113.0 69.0 41.0 73.0 271.0 73.0 277.0

50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50

23 34 23 25 15 10 15 10 10 10 10 10 10 35 35 35 35

quantitation confirmation quantitation IS confirmation IS quantitation confirmation quantitation IS confirmation IS quantitation confirmation-1 confirmation-2 quantitation IS confirmation IS quantitation confirmation quantitation IS confirmation IS

Note: RT = Retention time; IS = Internal standard. * CAS # of BP-1.

oven temperature was held at 40 ◦ C for 2 min, then ramped at 5 ◦ C/min to 65 ◦ C, ramped at 50 ◦ C/min to 300 ◦ C, and held at this temperature for 6 min. The total run time was 18 min. The injector temperature was fixed at 250 ◦ C and a split mode was used with helium as a carrier gas at a split ratio of 5:1. The constant pressure of 3.50 PSI and a positive electron impact (EI) mode were applied. The MS source temperature was set at 230 ◦ C. Selected reaction monitoring (SRM) mode was used for acquisition. The injection volume was 1.0 ␮L. Transitions, time segments (dynamic selected reaction monitoring), and scan parameters are summarized in Table 1.

2.5. Method validation 2.5.1. Linearity, intraday, and interday instrument precision studies The peak area ratios of the native analogues to the corresponding stable isotopically labeled internal standards were used for quantitation. The response of internal standards (S/N) was used to estimate the matrix effects in this study. To reduce errors in the lower end of a calibration curve, a weighting factor of 1/x2 was applied for calibration, where x represents the concentration of analytes. Intraday instrument precision studies were performed by sequentially injecting the calibration standard solutions in triplicate within a single day. The linear calibration curve was separately performed for each replicate. For interday instrument precision studies, a set of standard solutions was placed in the autosampler (at room temperature, ∼25 ◦ C) and injected on days 1–3, 7, 14, and 21. The caps were replaced after each injection to avoid evaporation. The calibration curve was performed daily. Linearity, accuracy, and precision were determined for intraday and interday instrument precision studies, respectively. For a freezer stability study, a set of standards was stored in a freezer (−18 ◦ C) and analyzed on days 1, 8, 14, 21, 35, and 110. The caps were replaced after each injection.

2.5.2. Extraction efficiency and recovery studies One nail product that formed a cake after centrifugation was selected to perform the extraction efficiency study. The product was weighed into a centrifuge tube to which 1000 ng of each analyte was added. The tube was placed on a bench for 60 min at room temperature. The tube was consecutively extracted three times using the above procedure. The three individual extracts were injected into the GC–MS/MS for quantitation.

For the recovery study, known amounts of analytes were added to a sample at three spiked levels before extraction. The same sample preparation as above was performed to extract the spiked samples. The concentrations of the three spiked levels corresponded to the lower, medium, and upper limits of quantitation. The recoveries were calculated by the following formula: Recovery(%) = [

Total Amount − Total in Sample ] × 100 Spiked Amount

where the Total Amount is the amount determined after spiking; Amount in Sample is the amount incurred from sample before spiking and Spiked Amount is the known spiked amount. 3. Results and discussion 3.1. Optimization for derivatization of BP-1 BP-1 is polar and not amenable to GC–MS analysis [15], therefore derivatization was necessary to make it amenable to GC–MS analysis. The following derivatization agents were tested: BSTFA, BSTFA containing 1% trimethylchlorosilane (TMCS), trimethylphenylammonium hydroxide solution, N,O-bis(trimethylsilyl)acetamide, and N-tert-butyldimethylsilyl-N-methyltrifluoroacetamide with 1% tert-butyldimethylchlorosilane. Although all derivatization agents resulted in more or less interference with toluene or NMP, BSTFA was selected for this study because it delivered minimal interference and adequate derivatization efficiency (yield) was achieved. The derivatization efficiency was tested by varying temperature (room temperature, 45 ◦ C, and 60 ◦ C), time (15 min, 30 min, and 60 min), solvents (ethyl acetate, acetonitrile, dichloromethane, methyl tert-butyl ether, dimethyl sulfoxide and acetone), and BSTFA amount (1% and 10%). The results showed that temperature, time, and BSTFA amount had little influence on the derivatization efficiency, and acetone had minimum interference to toluene. Nail products may contain different amounts of alcohols, which can react with BSTFA. Even though 1% BSTFA was enough for the derivatization of BP-1 in pure solvents, 10% BSTFA was used for preparing standards and samples to ensure the complete derivatization of BP-1 at room temperature. 3.2. Optimization of mass spectrometer and GC conditions The mass spectrometer was operated in electron ionization (EI) mode. All MS parameters were optimized by injecting standard

W. Zhou et al. / J. Chromatogr. A 1446 (2016) 134–140

137

×104 Toluene

2 1 ×104 8

NMP

Counts

4

×104 6

DEGDMA

3 ×104 8 BP-1_Deri

4 0

4

6

8 10 Acquisition Time (min)

12

14

Fig. 2. Representative chromatograms of toluene, NMP, DEGDMA and BP-1 standards at concentration of 1.00 ␮g/mL for each analyte. Transitions: m/z 91 → 65 for toluene, m/z 99 → 44 for NMP, m/z 113 → 69, for DEGDMA, and 343 → 73 for BP-1 derivative. GC conditions are the same as those described in the Experiment section.

solutions. Precursor ions and product ions for each analyte were determined by full scan and product scan, respectively. One quantitation and one confirmation SRM transition ions were selected for each analyte, except DEGDMA for which one quantitation and two confirmation transitions (m/z 113 → 41 and 143 → 113) were selected because no molecular ion (m/z 242) was observed [14]. Based on “Guidelines for the Validation of Chemical Methods for the FDA Foods Program [24], confirmation of identity for each analyte must be performed. Guidance for Industry Mass Spectrometry for Confirmation of the Identity of Animal Drug Residues (CVM Guidance 118) [25] was used to confirm the identity of each analyte. The allowed variability in ion ratios (Intensity of confirmation ion/Intensity of quantitation ion × 100%) between samples and standard solutions for the confirmation depends on how many structurally-specific ions (transitions) of analyte to be monitored. If two ions (transitions) are monitored, the tolerance window is ±10% absolute unit. If three or more ions (transitions) are monitored, the tolerance window is ±20% absolute unit. All transitions are summarized in Table 1. The optimized delta EMV (electron multiplier voltage), dwell time, and collision energy for each analyte are also listed in Table 1. Both the Resetk Rtx® 5 amine column and the Phenomenex ZB-SemiVolatiles column were tested. The Phenomenex ZBSemiVolatiles column was finally selected because it provided overall better peak shapes for all four analytes. It was observed that even neat solvents and derivatization reagents contained impurities that interfered with toluene. To avoid the interference, low initial temperature (40 ◦ C) and slow ramp rate (5 ◦ C/min) were applied to elute toluene. After toluene was eluted, a much faster ramp rate (50 ◦ C/min) was applied to reduce the total run time to 18 min. In this study, the GC–MS/MS run time was divided into four segments. A representative chromatogram is presented in Fig. 2. 3.3. Optimization for sample preparation procedure To save extraction solvent and expensive labeled internal standards, 1 mL of extraction solution including ISs and BSTFA (derivatization reagent) was used for sample preparation. As mentioned in Section 3.1, the derivatization of BP-1 was conducted at room temperature for 30 min. The derivatization and extraction were combined into one step to simplify the sample preparation. BFSTA was added prior to the extraction. Based on the instrument

precision data in Section 3.4.1, the recovery results in Section 3.4.3, and precision data for nail products in Section 3.5, the combination of derivitization of BP-1 with the extraction was suitable for GC–MS/MS analysis and no further clean-up was needed. 3.4. Method validation 3.4.1. Linearity, intraday and interday instrument precision Seven-point calibration curves were obtained for each analyte in concentrations ranging from 0.100 to 50.0 ␮g/mL for toluene and BP-1, and 0.100 to 5.00 ␮g/mL for NMP and DEGDMA, respectively. The results of intraday instrument precision (within a single day) are summarized in Table 2 (data for DEGDMA and BP-1 not shown). The slope and the coefficient of determination (r2 ) were very consistent for three consecutive replicates. The RSD values of the slope were between 0.38% and 5.82% and the r2 values were ≥0.99 for the four analytes. The accuracies of the standard solutions for the three replicates were between 85% and 111% (data not shown) and the RSD values were between 0.2% and 7.0% for the four analytes. Based on the results of the interday instrument precision at ∼25 ◦ C (over the course of twenty one days, data not shown), the four analytes at room temperature were stable for at least 21 days. This can also be considered for autosampler stability. The freezer stability results at −18 ◦ C (data not shown) showed that the RSDs of the slopes ranged from 2.3% to 9.0% for the four analytes over 110 days. Therefore, the standards are stable at −18 ◦ C for at least 110 days. 3.4.2. Carryover A calibration curve with a broad range (0.100–50.0 ␮g/mL for toluene and BP-1) was created for samples containing high amount of toluene and BP-1. A significant carryover for BP-1 (∼0.5%) was observed when the intraday precision study was performed. It was found that injecting three blank extraction solvent (acetone) was needed to minimize this carryover after injecting standards and samples with high concentrations of BP-1 (≥10.0 ␮g/mL). 3.4.3. Extraction efficiency and recovery The extraction efficiency results are listed in Table 3. The firsttime extraction efficiencies reached 98.3% for BP-1 and 100.0% for other three analytes (data for NMP and DEGDMA not shown). The second-time extraction efficiencies were 1.7% for BP-1 and 0.0%

138

W. Zhou et al. / J. Chromatogr. A 1446 (2016) 134–140

Table 2 Intraday method validation/instrument precision for toluene and NMP. Set

Toluene

1 2 3 Mean RSD%

NMP

Slope

r2

Intercept

Slope

r2

Intercept

1.173 1.169 1.178 1.173 0.38

0.9998 0.9996 0.9998 0.9997 0.01

−4.593E-03 −5.599E-03 −1.441E-03 −3.878E-03 −55.9

1.904 1.844 1.831 1.860 2.09

0.9980 0.9966 0.9992 0.9979 0.13

−7.976E-03 −8.501E-04 −3.700E-03 −4.175E-3 −85.9

Table 3 Extraction efficiency for toluene and BP-1 (n = 3). Extraction

1st 2nd 3rd

Toluene

BP-1

Determined Amt. ␮g

RSD, %

Extraction Efficiency

Determined Amt. ␮g

RSD, %

Extraction Efficiency

1.36 ND ND

3.10 NA NA

100 NA NA

52.9 0.901 ND

2.60 13.6 NA

98.3 1.70 NA

Note: The actual amount incurred from sample before spiking was 0.246 ␮g for toluene and 45.3 ␮g for BP-1, respectively. Table 4 Recovery results (n = 3). Spiked Level

Toluene

NMP

Spike ␮g/mL

Meas. ␮g/mL

Rec. %

RSD, %

Spike ␮g/mL

Meas. ␮g/mL

Rec. %

RSD, %

Low Middle High

0.251 0.984 5.09

0.249 1.06 5.01

99.2 108 98.3

2.90 5.09 5.29

0.243 0.973 5.32

0.260 1.03 5.28

107 106 99.3

1.00 2.26 1.31

Spiked Level

DEGDMA

Low Middle High

BP-1

Spike ␮g/mL

Meas. ␮g/mL

Rec. %

RSD, %

Spike ␮g/mL

Meas. ␮g/mL

Rec. %

RSD, %

0.257 1.07 5.51

0.280 1.11 5.63

109 104 102

1.76 2.80 0.620

0.247 1.010 5.02

0.258 1.02 5.06

105 101 101

0.830 1.13 0.690

Note: Meas. = Measured, Rec. = Recovery. Different amount of analytes was determined before spiking (0.0094 ␮g/mg for toluene, 0.00063 ␮g/mg for NMP, 0.0014 ␮g/mg for DEGDMA, and 0.00077 ␮g/mg for BP-1). The amount (concentration) incurred from sample was substrate based on the sample weight when calculating the recovery as shown in the formula in Section 2.4.

for other three analytes, respectively. The third-time extraction efficiencies were 0.0% for all four analytes. Evidently, one-time extraction was sufficient. As shown in Table 4, extraction recoveries ranged from 98% to 109% for the four analytes at three spiking levels of low, medium, and high concentrations. 3.4.4. Limits of quantitation The instrumental limit of quantitation (LOQ) in neat solution (acetone) was the lowest concentration on the calibration curve (0.100 ␮g/mL for all four analytes) that can be quantified within 100 ± 20% accuracy and 20% precision [26]. The LOQ for real sample was determined based on the lowest amount that was spiked into a nail product and could be quantified within 20% accuracy and precision [27,28]. Since 100 ␮L of samples (∼100 mg) were weighed and 1 mL extraction solution was added, the LOQ for real samples was determined and calculated as follows. The LOQs were confirmed by using the spiked samples. LOQ = 0.100 (g/mL) × 1.0 (mL)/100 (mg) = 0.001 (g/mg) = 1.0 (g/g)

3.4.5. Identification and confirmation of the analytes For all positive peaks (>LOQ) from the nail products, the CVM guidance 118 [25] was applied to confirm the identity of each analyte in a sample. In this situation, the acceptability range of retention times (RT) should not exceed 2% when compared to the

average retention times found in the calibration standards for each of the analytes. For toluene, NMP, and BP-1, the corresponding peak area ratio (confirmation transition/quantitation transition) should agree within ±10% absolute unit when compared to the average values found in the standards. For the first and second confirmation peak of DEGDMA, the peak area ratio of confirmation transition/quantitation transition should agree within ±20% absolute unit because three SRM transitions were used to monitor DEGDMA. A confirmed positive result for DEGDMA was described below. The retention time difference between the samples (product OCAC126, 12.24 min) and standards (12.24 min) was 0.00 min, i.e. 0.0% (≤±2%). The difference of ion ratio 1 between the samples (139%) and standards (126%) was 13% (≤±20%). The difference of ion ratio 2 between the samples (2.9%) and standards (2.8%) was 0.1% (≤±20%). Both the retention time and the ion ratios of the samples met the criteria. For product OCAC-129, the difference of retention time between the samples and standards was −0.01 min, i.e., −0.08% (≤±2%). The difference of ion ratio 1 was 104% (≥±20%). The difference of ion ratio 2 was 6.0% (≤±20%). Although the retention time and ion ratio 2 of the sample met the criteria, the ion ratio 1 did not. The result for product OCAC-129 was unconfirmed so it was not positive (i.e., the sample did not contain DEGDMA). 3.5. Nail product survey The validated method has been successfully applied to analyze thirty-four nail products. The results are summarized in Table 5.

W. Zhou et al. / J. Chromatogr. A 1446 (2016) 134–140

139

Table 5 Results for the nail products (n = 3). Prod #

OCAC-123 OCAC-167 OCAC-125 OCAC-164 OCAC-166 OCAC-163 OCAC-131 OCAC-155 OCAC-127 OCAC-159 OCAC-165 OCAC-121 OCAC-126 OCAC-122 OCAC-153 OCAC-156 OCAC-129 OCAC-154 OCAC-169 OCAC-162 OCAC-132 OCAC-174 OCAC-175 OCAC-161 OCAC-157 OCAC-168 OCAC-130 OCAC-124 OCAC-158 OCAC-160 OCAC-170 OCAC-171 OCAC-172 OCAC-173

Type of Prod.

Nail Art Base coat Nail Art Base coat Base coat Top coat Nail Art Thinner Nail Art Top coat Base coat Base coat Gel Remover Remover Thinner Nail Art Nail Art Thinner Top coat Gel Nail Art Nail Art Top coat Remover Thinner Gel Nail Art Top coat Top coat Remover Remover Remover Remover

Toluene

NMP

DEGDMA

BP-1

␮g/g

RSD, %

␮g/g

RSD, %

␮g/g

RSD, %

␮g/g

RSD, %

173,000 140,000 107,000 95,000 89,600 29,300 857 194 159 129 52.9 17.2 14.4 11.5 10.6 9.48 5.89 5.01 4.65 4.21 3.85 2.50 2.45 1.63 1.50 1.36 ND ND ND ND ND ND ND ND

6.09 1.95 12.9 1.52 2.18 1.39 0.920 0.660 3.22 8.29 4.53 2.45 1.35 1.36 3.87 11.3 10.9 1.12 0.810 3.17 2.02 7.59 13.5 8.12 1.91 14.4 NA NA NA NA NA NA NA NA

ND 2.80 ND ND ND ND ND 2.73 ND ND ND ND 102 ND ND 4.16 ND ND ND ND 17.2 ND ND ND ND ND ND ND ND ND ND ND ND ND

NA 6.97 NA NA NA NA NA 4.49 NA NA NA NA 0.590 NA NA 8.45 NA NA NA NA 1.12 NA NA NA NA NA NA NA NA NA NA NA NA NA

ND ND ND ND ND ND ND ND 21.8 ND ND ND 35.1 1.61 ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND

NA NA NA NA NA NA NA NA 3.08 NA NA NA 3.78 3.39 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA

2,370 ND 18.3 733 86.5 ND 70. 9 ND ND ND 1,940 ND ND ND ND ND 503 1,090 ND ND 176 ND ND ND ND ND 224 ND ND ND ND ND ND ND

0.790 NA 1.97 1.14 0.650 NA 2.26 NA NA NA 0.080 NA NA NA NA NA 2.47 8.58 NA NA 1.02 NA NA NA NA NA 0.32 NA NA NA NA NA NA NA

Note: ND = Not determined (Less than the LOQ), NA = Not applicable. For all positive samples, ion ratios for the analytes were checked to avoid unconfirmed positive results.

There were no significant interference peaks for any of the four analytes. The most common ingredients detected in nail products were toluene and BP-1. Toluene was detected in 26 products and ranged from 1.36 to 173,000 ␮g/g. BP-1 ranged from 18.3 to 2370 ␮g/g in 10 products. NMP ranged from 2.73 to 102 ␮g/g in 5 products. DEGDMA was only found in 3 products and ranged from 1.61 to 35.1 ␮g/g. Other studies showed that the amount of toluene in nail products ranged from none detected to 17.7% [29], from none detected to 22.1% [30], and approximately 20% found [10]. The amount of BP-1 in cosmetic products was from none detected to 5.65% [18]. BP-1 was not detected in any cosmetic products [23]. No NMP and DEGDMA were reported in nail/cosmetic products. For all positive samples, both the retention time and the ion ratios were checked as discussed above. All analytes reported in Table 5 were within the identification criteria. Unconfirmed peaks were found for both NMP and DEGDMA. It was observed that both retention time and ion ratio 2 of a few ‘positive’ DEGDMA samples met the criteria, but the ion ratio 1 did not meet the criteria.

4. Conclusion Nail products are widely used in salons and homes. To the best of our knowledge, this is the first fully validated GC–MS/MS method to simultaneously determine four different cosmetics ingredients in nail products using dSRM mode and combine extraction and derivatization into one step without clean up. In addition, individual isotopically labeled analogues were used as internal standards to compensate for the loss during the sample preparation and instrument fluctuation. The method has a short run time (18 min) with

high sensitivity (i.e., picogram on column). The method has been successfully applied to many nail products including base coat, top coat, thinner, gels, remover, and nail art (polish) products. Acknowledgements The authors wish to thank Drs. Gregory O. Noonan, Linda M. Katz, Nakissa Sadrieh, Stanley R. Milstein, Shaun A. MacMahon, Jeanne I. Rader, Wayne G. Wamer, Hardy J. Chou, Patricia A. Hansen for helpful discussions. References [1] (accessed 6.06.14) http://www.fda.gov/cosmetics/productsingredients/ products/ucm127068.htm. [2] (accessed 22.07.15) https://www.osha.gov/SLTC/formaldehyde/hazard alert. pdf. [3] T. Quach, R. Gunier, A. Tran, J. Von Behren, P.A. Doan-Billings, K.D. Nguyen, L. Okahara, B.Y.B. Lui, M. Nguyen, J. Huynh, P. Reynolds, Characterizing workplace exposures in vietnamese women working in California nail salons, Am. J. Public Health 101 (S1) (2011) S271–S276. [4] B.K. Schindler, S. Koslitz, S. Meier, V.N. Belov, H.M. Koch, T. Weiss, T. Bruning, H.U. Kafferlein, Quantification of four major metabolites of embryotoxic N-methyl and N-ethyl-2-pyrrolidone in human urine by cooled-injection gas chromatography and isotope dilution mass spectrometry, Anal. Chem. 84 (2012) 3787–3794. [5] M.A. Park, K.A. Hwang, H.R. Lee, B.R. Yi, E.B. Jeung, K.C. Choi, Benzophenone-1 stimulated the growth of BG-1 ovarian cancer cells by cell cycle regulation via an estrogen receptor alpha-mediated signaling pathway in cellular and xenograft mouse models, Toxicology 305 (2013) 41–48. [6] US Patent: US4260701, Formulation for self-curing artificial fingernails containing methoxyethyl methacrylate (1981). [7] A. Escobar, T.A. Yamarik, F.A. Andersen, Final report of the safety assessment of methacrylate ester monomers used in nail enhancement products, Int. J. Toxicol. 24 (S5) (2005) 53–100.

140

W. Zhou et al. / J. Chromatogr. A 1446 (2016) 134–140

[8] (accessed 29.05.14) http://ec.europa.eu/health/scientific committees/ consumer safety/docs/sccs o 050.pdf. [9] J. Toedt, D. Koza, K. Cleef-Toedt, Chemical Composition of Everyday Products, Greenwood Publishing Group, 2005, pp. 50. [10] E.L. Sainio, K. Engstrom, M.L. HenriksEckerman, L. Kanerva, Allergenic ingredients in nail polishes, Contact Dermat. 37 (1997) 155–162. [11] M.L. HenriksEckerman, Sampling efficiency of Amberlite XAD-2 with glassfiber prefilter and ctivated charcoal as back-up section for collection of mixtures of organic air pollutants at the ␮g/m3 level, Chemosphere 1 (7) (1990) 889–904. [12] S.A. Zhang, T.B. Zhao, H.W. Wang, J. Wang, Comparison of SPE-TD-GC-FID with UPLC-PDA and GC–MS methods for analysis of benzene, toluene and xylene isomers in solid-liquid mixing paints, Chromatographia 74 (2011) 163–169. [13] G. Laus, M. Andre, G. Bentivoglio, H. Schottenberger, Ionic liquids as superior solvents for headspace gas chromatography of residual solvents with very low vapor pressure, relevant for pharmaceutical final dosage forms, J. Chromatogr. A 1216 (2009) 6020–6023. [14] N. Gjos, K. Urdal, I.E. Ruyter, I.J. Sjovik, Identification of methacrylates and acrylates in dental materials by mass spectrometry, Anal. Chim. Acta 149 (1983) 87–99. [15] N. Negreira, I. Rodriguez, M. Ramil, E. Rubi, R. Cela, Sensitive determination of salicylate and benzophenone type UV filters in water samples using solid-phase microextraction, derivatization and gas chromatography tandem mass spectrometry, Anal. Chim. Acta 638 (2009) 36–44. [16] Y. Zhang, H.K. Lee, Determination of ultraviolet filters in water samples by vortex-assisted dispersive liquid–liquid microextraction followed by gas chromatography–mass spectrometry, J. Chromatogr. A 1249 (2012) 25–31. [17] C. Sánchez-Brunete, E. Miguel, B. Albero, J.L. Tadeo, Analysis of salicylate and benzophenone-type UV filters in soils and sediments by simultaneous extraction cleanup and gas chromatography–mass spectrometry, J. Chromatogr. A 1218 (2011) 4291–4298. [18] L.H. Wang, Simultaneous determination of seven sunscreen benzophenones in cosmetic products by high-performance liquid chromatography, Chromatographia 50 (9–10) (1999) 565–570. [19] I.E. Ruyter, I.J. Sjovik, Composition of dental resin and composite materials, Acta Odontol. Scand. 39 (3) (1981) 133–146.

[20] W. Spahl, H. Budzikiewicz, Qualitative analysis of dental resin composites by gas and liquid chromatography/mass spectrometry, Fresenius J. Anal. Chem. 350 (1994) 684–691. [21] L.J. Juang, B.S. Wang, H.M. Tai, W.J. Hung, M.H. Huang, Simultaneous identification of eight sunscreen compounds in cosmetic products using high-performance liquid chromatography and capillary electrophoresis, J. Food Drug Anal. 16 (6) (2008) 22–28. [22] Y.W. Wu, Y.Y. Jiang, J.F. Liu, K. Xiong, Cloud point extraction combined with micellar electrokinetic capillary chromatography determination of benzophenones in cosmetic matrix, Electrophoresis 29 (S4) (2008) 819–826. [23] S.P. Wang, W.T. Lee, Determination of benzophenones in a cosmetic matrix by supercritical fluid extraction and capillary electrophoresis, J. Chromatogr. A 987 (2003) 269–275. [24] (accessed 22.07.15) http://www.fda.gov/downloads/ScienceResearch/ FieldScience/UCM298730.pdf. [25] (accessed 22.07.15) http://www.fda.gov/downloads/AnimalVeterinary/ GuidanceComplianceEnforcement/GuidanceforIndustry/ucm052658.pdf. [26] (accessed 29.05.14) http://www.fda.gov/downloads/Drugs/Guidances/ ucm070107.pdf. [27] Y.H. Park, K.M. Joo, B.Y. Woo, E.D. Son, S.Y. Byun, H.J. Shin, K.W. Lee, Y.H. Park, K.M. Lim, Oral and topical pharmacokinetic studies of a novel TRPV1 antagonist, PAC-14028 in rats and minipigs using liquid chromatography/tandem mass spectrometric method, J. Pharm. Biomed. Anal. 61 (2012) 8–14. [28] W. Zhou, P.G. Wang, O.A. Ogunsola, M.E.K. Kraeling, Rapid determination of hexapeptides by hydrophilic interaction LC–MS/MS for in vitro skin-penetration studies, Bioanalysis 5 (11) (2013) 1353–1362. [29] (accessed 6.03.14) https://www.dtsc.ca.gov/PollutionPrevention/upload/ NailSalon Final.pdf. [30] A. Gonzalvez, S. Garrigues, S. Armenta, M. de la Guardia, Headspace-liquid phase microextraction for attenuated total reflection infrared determination of volatile organic compounds at trace leves, Anal. Chem 82 (2010) 3045–3051.