J. Cosmet. Sci., 63, 103–117 (March/April 2012)

Simultaneous analysis and monitoring of 16 UV filters in cosmetics by high-performance liquid chromatography DOJUNG KIM, SANGSEOP KIM, SEOL-A KIM, MYOENGSIN CHOI, KYOUNG-JIN KWON, MIJEONG KIM, DONG-SUP KIM, SEUNG-HEE KIM, and BO-KYUNG CHOI, National Institute of Food and Drug Safety Evaluation, Korea Food and Drug Administration, Osong Health Technology Administration Complex, 187 Osongsaenmyeong2(i)-ro, Gangoe-myeon, Cheongwon-gun, Chungcheongbuk-do, Korea 363-951. Accepted for publication September 21, 2011. Synopsis Sixteen UV filters were simultaneously analyzed using the high-performance liquid chromatographic method. They were drometrizole (USAN Drometrizole), 4-methylbenzylidene camphor (USAN Enzacamene), menthyl anthranilate (USAN Menthyl anthranilate), benzophenone-3 (USAN Oxybenzone), benzophenone-8 (USAN Dioxybenzone), butyl methoxydibenzoylmethane (USAN Avobenzone), ethylhexyl triazone (USAN Octyl triazone), octocrylene (USAN Octocrylene), ethylhexyl dimethyl p-aminobenzoic acid (USAN Padimate O), ethylhexyl methoxycinnamate (USAN Octinoxate), p-aminobenzoic acid (USAN Aminobenzoic acid), 2-phenylbenzimidazole-5-sulfonic acid (USAN Ensulizole), isoamyl p-methoxycinnamate (USAN Amiloxate), and recent UV filters such as diethylhexyl butamidotriazone (USAN Iscotrizinol), methylene bis-benzotriazolyl tetramethylbutylphenol (USAN Bisoctrizole), and terephthalylidene dicamphor sulfonic acid (USAN Ecamsule). Separation of the UV filters was carried out in a C18 column with a gradient of methanol-phosphate buffer, and the UV detection was at 300, 320, or 360 nm without any interference. The limits of detection were between 0.08 and 1.94 μg/ml, and the limits of quantitation were between 0.24 and 5.89 μg/ml. The extracting solvent for the UV filters was methanol, except for ethylhexyl triazone and methylene bis-benzotriazolyl tetramethylbutylphenol, which were prepared with tetrahydrofuran. The recoveries from spiked samples were between 94.90% and 116.54%, depending on the matrixes used. The developed method was applied to 23 sunscreens obtained from local markets, and the results were acceptable to their own criteria and to maximum authorized concentrations. Consequently, these results would provide a simple extracting method and a simultaneous determination for various UV filters, which can improve the quality control process as well as the environmental monitoring of sunscreens.

INTRODUCTION

Exposure to UV radiation can harmfully affect the public health by inducing conditions such as skin redness or skin cancer (1,2). As the ozone layers are depleted, protection from hazardous UV radiation, UVA (320 – 400 nm) and UVB (290 – 320 nm), has been a major concern for people (3,4). UVA has a long wavelength that reaches the corium of the skin to initiate skin aging, DNA damage, and immune alteration. UVB is known to 103

104

JOURNAL OF COSMETIC SCIENCE

cause sunburn and to induce malignant melanoma (2). UV filters have been used as an ingredient in sunscreens to protect the skin from UVA and UVB. Most organic UV filters have effective protection against UVB but little effect against UVA, whereas butyl methoxydibenzoylmethane (USAN Avobenzone), terephthalylidene dicamphor sulfonic acid (USAN Ecamsule), and methylene bis-benzotriazolyl tetramethylbutylphenol (USAN Bisoctrizole) are utilized to protect UVA. Inorganic UV filters, such as TiO2 (USAN titanium dioxide) and ZnO (USAN zinc oxide) were combined in sunscreens to protect UVA. To improve protective effects from UV radiation, many cosmetic companies have tried to synthesize new UV filters or to develop certain formulation containing more than one UV filter (5–7). Although UV filters protect skin from UV radiation and help prevent skin damage including skin cancer, they are also concerned about the possibility of causing an allergic reaction to human skin (8). Studies reported that UV filters such as benzophenone-3 (Oxybenzone) and 4-methylbenxylidene camphor (USAN Enzacamene) exert the effects of endocrine disruptors (9,10). To prevent adverse effects of UV filters, many countries have regulated the amount of UV filters that will be used in sunscreens; the maximum authorized concentration is 3–10% for organic UV filters (Table I) and 25% for inorganic UV filters. An easy and accurate analytical method for quantitation of UV filters is required to assure the label claims and maximum authorized concentrations in sunscreens. In some studies, UV filters were investigated using GC-MS or LC-MS, due to their trace levels in environmental or biological samples (11,12). However, in cosmetics, they were generally measured using a HPLC system connected to a UV detector. Although several studies described analytical methods of 11 or 16 organic UV filters using HPLC connected to a diode array detector, they employed different extraction solvents and several steps of extraction according to the cosmetic formulations (13–16). Sunscreens include UV filters as well as other ingredients such as emulsifiers, solubility enhancers, and stability enhancers that are necessary for a cosmetic formulation and that might interfere with reliable analysis. Various chemical properties and physical properties of UV filters also make it difficult for them to be analyzed in sunscreens (17,18). Regarding various formulations of sunscreen such as cream, lotion, or powder, practical methods with a simple preparation are still needed for industry to develop a new product and for administrative bodies to monitor the quality for marketed sunscreens. The purpose of the present study was to develop a simple extraction method and a simultaneous analysis for 16 UV filters, including recent UV filters such as diethylhexyl butamidotriazone (USAN Iscotrizinol), methylene bis-benzotriazolyl tetramethylbutylphenol (USAN Bisoctrizole), and terephthalylidene dicamphor sulfonic acid (USAN Ecamsule) (Figure 1). They were determined using HPLC connected to a C18 column and diode array detector (DAD) at 300, 320, and 360 nm. Method validation was achieved through specificity, linearity, and recovery. Several solvents were examined for extracting UV filters from sunscreens, regardless of the matrixes used. The validated method was applied to various types of sunscreens obtained from local markets and evaluated with regard to the label claims and compliance to their maximum authorized concentrations. EXPERIMENTAL MATERIALS

Drometrizole (2440-22-4), menthyl anthranilate (134-09-8), benzophenone-3 (205-031-5), benzophenone-8 (131-53-3), octocrylene (6197-30-4), ethylhexyl dimethyl p-aminobenzoic

2

6197-30-4 92761-26-7

2440-22-4 131-53-3 36861-47-9 150-13-0 70356-09-1 88122-99-0 134-09-8 21245-02-3 27503-81-7 205-031-5 154702-15-5 5466-77-3 71617-10-2 103597-45-1

CAS # 225.25 244.24 254.37 137.14 310.38 823.07 275.39 277.40 274.30 228.25 765.98 290.40 248.32 658.88 361.48 562.69

C24H27NO2 C28H34O8S2

Molecular weight

C13H11N3O C114H12O4 C18H22O C7H7NO2 C20H22O3 C48H66N6O6 C17H25NO2 C17H27NO2 C13H10N2O3S C14H12O3 C44H59N7O5 C18H26O3 C15H20O3 C41H50N6O2

Molecular formula

10% 10%

—2 3% 4% 5% 5% 5% 5% 8% 8% 10% 10% 10% 10% 10%

Max. authorized concentration (%)1

Maximum authorized concentration of UV filters is listed in Directive 76/768/EEC and in Notification No. 2009-52 of the Korea Food and Drug Administration. Maximum authorized concentration of Drometrizole is 7% in Korea.

OC TDS

Octocrylene Ecamsule

1

DT BZ-8 MBC PABA BMDM EHT MA ED-PABA PBS BZ-3 DBT EMC IMC MBBT

Drometrizole Dioxybenzone Enzacamene Aminobenzoic acid Avobenzone Octyl triazone Menthyl anthranilate Padimate O Ensulizole Oxybenzone Iscotrizinol Octinoxate Amiloxate Bisoctrizole

Drometrizole Benzophenone-8 4-Methylbenzylidene camphor p-Aminobenzoic acid Butyl methoxydibenzoylmethane Ethylhexyl triazone Menthyl anthranilate Ethylhexyl dimethyl p-aminobenzoic acid 2-Phenylbenzimidazole-5-sulfonic acid Benzophenone-3 Diethylhexyl butamidotriazone Ethylhexyl methoxycinnamate Isoamyl p-methoxycinnamate Methylene bis-benzotriazolyl tetramethylbutylphenol Octocrylene Terephthalylidene dicamphor sulfonic acid

Abbreviation

USAN name

UV filter

Table I Characteristics and Maximum Authorized Concentrations of UV Filters in the Present Study

MONITORING OF UV FILTERS BY HPLC 105

106

JOURNAL OF COSMETIC SCIENCE

Figure 1. Structures of 16 UV filters: (1) PABA; (2) 2-Phenylbenzimidazole-5-sulfonic acid; (3) Terephthalylidene dicamphor sulfonic acid; (4) Benzophenone-8; (5) Benzophenone-3; (6) Isoamyl p-methoxycinnamate; (7) Drometrizole; (8) 4-Methylbenzylidene camphor; (9) Octocrylene; (10) Menthyl anthranilate; (11) Ethylhexyl dimethyl PABA; (12) Butyl methoxydibenzoylmethane; (13) Ethylhexyl methoxycinnamate; (14) Diethylhexyl butamidotriazone; (15) Ethylhexyl triazone; (16) Methylene bis-benzotriazolyl tetramethylbutylphenol.

MONITORING OF UV FILTERS BY HPLC

107

acid (21245-02-3), ethylhexyl methoxycinnamate (5466-77-3), p-aminobenzoic acid (15013-0), 2-phenylbenzimidazole-5-sulfonic acid (27503-81-7), isoamyl p-methoxycinnamate (71617-10-2), and Tween 80 were purchased from Sigma-Aldrich Korea (Seoul, Korea). We obtained 4-methylbenzylidene camphor (36861-47-9) from Alfa Aesar Korea (Seoul, Korea), butyl methoxydibenzoylmethane (70356-09-1) from Wako (Osaka, Japan), ethylhexyl triazone (88122-99-0) and methylene bis-benzotriazolyl tetramethylbutylphenol (103597-45-1) from BASF Korea (Seoul, Korea), diethylhexyl butamidotriazone (154702-15-5) from 3V Sigma (Georgetown, USA), and terephthalylidene dicamphor sulfonic acid (92761-26-7) from L’Oreal Korea (Seoul, Korea). Methanol (MeOH), acetonitrile, water, ethanol (EtOH), isopropanol, dimethylformamide (DMF), acetone, and tetrahydrofuran (THF) were obtained as HPLC grade from Burdick & Jackson (Morristown, NJ). For monitoring, 23 sunscreens were purchased from local markets. Cream and lotion matrixes for recovery study were generously given by LG Household & Health Care (Korea), and their formulas are described in Table II.

Table II Formula of Cream and Lotion Matrixes Used for Recovery Study Cream matrix

Lotion matrix

INCI Name

(%)

INCI Name

(%)

Water Cyclopentasiloxane Glycerin Hydrogenated polydecene Cetearyl alcohol Butylene glycol Triethylhexanoin Pentaerythirityl tetraethylhexanoate Butyrospermum park II (shea butter) Mineral oil, microcrystalline wax, paraffin PEG-40 stearate Sorbitan stearate Portulaca oleracea extract Glyceryl stearate Glyceryl stearate, PEG-100 stearate Stearic acid Beeswax Dimethicone Methyl paraben Fragrance Triethanolamine Propyl paraben Ethyl paraben Tocopheryl acetate Carbomer Xanthan gum Trisodium EDTA

64.90 6.46 6.46 4.62 2.95 2.77 1.85 1.85 1.38 0.92 0.92 0.92 0.92 0.46 0.46 0.46 0.46 0.28 0.23 0.18 0.14 0.09 0.09 0.09 0.06 0.04 0.02

Water Glycerin Hydrogenated polydecene Dipropylene gylcol Octyldodecyl myristate Isocetyl myristate Polyglyceryl-3 methylglucose distearate Cetearyl alcohol G1yceryl stearate, PEG-100 stearate Glyceryl stearate Cetyl ethylhexanoate Dimethicone Methyl paraben Triethanolamine Carbomer Fragrance Propyl paraben Ethyl paraben Tocopheryl acetate Trisodium EDTA

77.73 4.69 4.69 2.82 1.88 1.41 1.41 1.13 1.13 0.94 0.94 0.38 0.19 0.13 0.13 0.11 0.09 0.09 0.09 0.02

108

JOURNAL OF COSMETIC SCIENCE

PREPARATION OF STANDARD SOLUTIONS

Stock solutions of 16 UV filters were prepared using different organic solvents due to their having various properties. Menthyl anthranilate (MA), 4-methylbenzylidene camphor (MBC), benzophenone-3 (BZ-3), benzophenone-8 (BZ-8), butyl methoxydibenzoylmethane (BMDM), octocrylene (OC), ethylhexyl dimethyl p-aminobenzoic acid (ED-PABA), ethylhexyl methoxycinnamate (EMC), p-aminobenzoic acid (PABA), isoamyl p-methoxycinnamate (IMC), diethylhexyl butamidotriazone (DBT), and terephthalylidene dicamphor sulfonic acid (TDS) were prepared as 10000 ppm in MeOH and ethylhexyl triazone (EHT) as 10000 ppm in THF. Drometrizole (DT) was dissolved as 1000 ppm in MeOH, 2-phenylbenzimidazole-5-sulfonic acid (PBS) as 1000 ppm in the mixture of MeOH and 12% acetic acid (1:1), and methylene bis-benzotriazolyl tetramethylbutylphenol (MBBT) as 1000 ppm in THF. The stock solutions were combined and diluted to become 1–50 ppm in MeOH. VALIDATION STUDY

The linearity of UV filters was examined in the range of 1–50 ppm, except for 5–50 ppm for MBBT. The stability of stock solutions was determined to compare the correlation coefficients of the calibration curves one, three, and six days after their preparation. The intraday and interday variation for the determination of the 16 UV filters was carried out at concentrations of 10, 25, and 40 ppm. For the recovery study, the standard solution was spiked into cream and lotion matrixes that contained only basic ingredients for cosmetics. The spiked samples were weighed as 1 g and added to MeOH to become concentrations of 15, 20, and 25 ppm. Each spiked sample was homogenized under sonification and filtered through a 0.2-μm PTFE Acrodisc syringe filter (Port Washington, NY). Another standard solution and spiked sample set was prepared for EHT and MBBT, followed by the same procedure except for using THF instead of MeOH. The precision of the determination of the UV filters spiked in cream and lotion matrixes was tested at concentration levels of 10, 25, and 40 ppm. Each performance was carried out in triplicate. Additionally, other solvents such as 5% DMF in MeOH, EtOH, 2% Tween 80 in EtOH, isopropanol, and acetone were applied in choosing a suitable solvent for extracting UV filters from spiked samples. SAMPLE PREPARATION

To monitor the concentration of UV filters in commercial sunscreens, 1 g of each sunscreen was weighed and diluted with MeOH or THF to adjust the concentration in the range of 1–50 ppm. The mixture was homogenized under sonification and filtered through a 0.2-μm PTFE syringe filter. In the case of sunscreens containing EHT and MBBT, the sample solution was prepared using THF. The prepared standard solutions, spiked samples, and sunscreen samples were analyzed in an Agilent 1200 series HPLC system (Santa Clara, CA) with an injection volume of 10 μl. Each sample was contained in an amber vial to protect the light in ambient condition. HPLC CONDITIONS

Sixteen UV filters in the present studies were analyzed using HPLC connected to a diode array detector (DAD). Chromatographic separation of the UV filters was carried out with

MONITORING OF UV FILTERS BY HPLC

109

a Cap Cell PAK C18 MG2 (4.6 mm × 250 mm, 4.6 μm of particle size, Shiseido, Japan) maintained at a constant temperature of 40°C. The mobile phase was followed by a gradient of MeOH-sodium hydrogen phosphate (pH 4.0; 0.01 M) in the volume proportions described in Table III. The UV filters were simultaneously determined at 300, 320, or 360 nm, and the detection wavelengths are summarized in Table IV. All analytical results were calculated from triplicate data. RESULTS AND DISCUSSION CHROMATOGRAPHIC ANALYSIS

Twelve UV filters (MA, MBC, BZ-3, BZ-8, BMDM, OC, ED-PABA, EMC, PABA, IMC, DBT, and TDS) were readily soluble in MeOH to prepare a stock solution of 10000 ppm. EHT and MBBT were dissolved in THF, and PBS was prepared in the mixture of MeOH and 12% acetic acid (1:1, v/v) due to their solubility and stability. DT was prepared in MeOH as a stock solution 1000 ppm due to low solubility. Sixteen UV filters were separated in a C18 column with a mobile phase of MeOH and phosphate buffer. The mobile Table III Gradient Profile for Analysis of 16 UV Filters Time (min)

0.01 M NaH2PO4 (pH 4.0) (%)

MeOH (%)

Flow rate (ml/min)

0 40 40.1 70

70 0 0 0

30 100 100 100

1.0 1.0 1.2 1.2

Table IV Results of Linearity, LOD, and LOQ for UV Filters No.

UV filter

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

PABA PBS TDS BZ-8 BZ-3 IMC DT MBC OC MA ED-PABA BMDM EMC DBT EHT MBBT

tR (min) Wavelength (nm) 4.48 7.77 15.37 26.57 30.82 36.26 37.06 37.14 39.22 39.86 40.18 40.58 40.66 44.90 52.35 65.47

300 300 360 320 320 320 360 320 320 360 320 360 320 300 300 360

Regression

R2

y = 49.10 χ + 0.76 y = 61.45 χ + 2.98 y = 65.10 χ − 0.01 y = 16.56 χ + 1.08 y = 18.06 χ + 0.18 y = 92.91 χ − 0.13 y = 27.08 χ + 2.17 y = 29.35 χ − 3.08 y = 63.73 χ − 0.77 y = 10.27 χ − 0.39 y = 71.23 χ − 12.59 y = 43.58 χ − 8.87 y = 52.04 χ + 5.65 y = 56.29 χ + 2.62 y = 59.70 χ + 1.68 y = 18.25 χ + 5.87

0.9998 0.9998 0.9997 0.9998 0.9999 0.9999 0.9998 0.9998 0.9998 0.9997 0.9996 0.9998 0.9998 0.9998 0.9998 0.9978

LOD (ppm) LOQ (ppm) 0.28 0.21 0.08 0.25 0.15 0.21 0.15 0.21 0.23 0.25 1.66 0.58 0.20 0.20 0.18 1.94

0.85 0.64 0.24 0.76 0.47 0.63 0.47 0.64 0.71 0.75 5.02 1.75 0.62 0.62 0.55 5.89

110

JOURNAL OF COSMETIC SCIENCE

phase was a gradient that achieved an efficient resolution of the 16 UV filters, and the flow rate increased by 1 ml/min to 1.2 ml/min at 40 min to diminish the run time. The chromatograms and spectra of the UV filters are displayed in Figure 1. Acidic UV filters such as PABA, PBS, and TDS were relatively quickly eluted from the column, and EHT and MBBT were eluted over 50 min (Figure 2). Each peak of UV filters was detected at 300, 320, or 360 nm in view of its intensity (Table IV). In the case of BMDM and EMC, two peaks were shown in the close retention time, but they were determined at 360 nm and 320 nm with high intensity, respectively. VALIDATION STUDY

The limit of detection (LOD) was defined as the concentration of analyte that was given from 3.3σ/s, where σ is the standard deviation of the y intercepts and s is a slope of regression. Likewise, the limit of quantitation (LOQ) was defined as 10σ/s. The LOD and LOQ of the UV filters, summarized in Table IV, accounted for 0.08–1.94 ppm and 0.24–5.89 ppm, respectively. ED-PABA and MBBT were relatively higher in LOD and LOQ than the others. The linearity showed up to 50 ppm with R2 >0.999, except for MBBT, showing 0.998 (Table IV). Intraday and interday variations were represented as % relative standard deviation (RSD), showing under 5% (Tables V, VI). Prior to the recovery test, several solvents, such as MeOH, EtOH, isopropanol, acetone, THF, 2% Tween in EtOH, and 5% DMF in MeOH, were examined to choose a suitable solvent for extracting the cream and lotion matrixes spiked with the standard solution of 20 ppm (data not shown here). With EtOH, 2% Tween in EtOH, isopropanol, and acetone, there were split peaks displayed in the chromatograms of PABA, PBS, and TDS. The chromatogram of THF extract also showed split peaks of PBS and TDS. From the extract with 5% DMF in MeOH, there was no split peak in the chromatogram, but the recoveries of PABA and PBS were considerably low. According to the chromatograms and recovery test, MeOH was revealed as a suitable solvent for the UV filters except for EHT and MBBT, which were practically extracted in THF. According to these results, MeOH was used as a solvent for 14 UV filters and THF for EHT and MBBT. The recoveries of the UV filters spiked in the cream matrix were 94.98–116.54% (Table VII). The recovery at low concentration of 15 ppm presented slight variation with 1.99–8.02% of RSD, which might be easy to be interfered with by diverse ingredients in the cream matrix at low concentrations. The recoveries of the UV filters from the spiked lotion samples were 94.90–116.21%, and the RSD was under 2.7%, which indicated a suitable degree of precision (Table VII). The recovery of MBBT from both cream and lotion matrixes was slightly higher than that of the other UV filters, and accounted for 108.09–116.54% and 109.99–116.27%, respectively. For the 16 UV filters, the precision of recovery from the spiked samples revealed a good result of under 0.5% of RSD (Table VIII). STABILITY STUDY

The UV filters in standard solution were generally stable at ambient conditions for a week (Table IX). However, MA, BMDM, and EMC gradually decreased R2 as time went by, which agreed with the photostability study that showed that BMDM and EMC were steadily degraded by exposure to UV radiation. The linearity of MBBT noticeably deteriorated in our study, while MBBT was reported to show

MONITORING OF UV FILTERS BY HPLC

111

Figure 2. (A) Chromatograms at 300 (top), 320 (middle), and 360 nm (bottom) and (B) spectra of 16 UV filters: (1) PABA; (2) 2-Phenylbenzimidazole-5-sulfonic acid; (3) Terephthalylidene dicamphor sulfonic acid; (4) Benzophenone-8; (5) Benzophenone-3; (6) Isoamyl p-methoxycinnamate; (7) Drometrizole; (8) 4-Methylbenzylidene camphor; (9) Octocrylene; (10) Menthyl anthranilate; (11) Ethylhexyl dimethyl PABA; (12) Butyl methoxydibenzoylmethane; (13) Ethylhexyl methoxycinnamate; (14) Diethylhexyl butamidotriazone; (15) Ethylhexyl triazone; (16) Methylene bis-benzotriazolyl tetramethylbutylphenol.

112

JOURNAL OF COSMETIC SCIENCE Table V Results of Intraday Validation for 16 UV Filters 10 ppm

25 ppm

40 ppm

UV filters

Area

SD

%RSD

Area

SD

%RSD

Area

SD

%RSD

PABA PBS TDS BZ-8 BZ-3 IMC DT MBC OC MA ED-PABA BMDM EMC DBT EHT MBBT

484.90 609.10 646.26 164.62 178.38 916.54 269.24 284.92 628.05 99.12 697.46 421.99 518.77 557.22 589.83 182.34

6.30 5.96 6.59 1.41 1.33 4.93 1.62 2.05 4.96 0.96 23.48 8.07 4.14 1.82 4.20 1.97

1.30 0.98 1.02 0.86 0.75 0.54 0.60 0.72 0.79 0.97 3.37 1.91 0.80 0.33 0.71 1.08

1240.70 1558.61 1649.81 420.19 456.09 2341.11 687.07 734.35 1601.65 259.56 1764.53 1091.09 1322.50 1427.30 1511.53 475.01

15.98 13.33 13.77 3.66 3.86 13.27 2.61 7.19 15.02 2.68 43.68 14.86 10.67 10.61 12.19 5.54

1.29 0.86 0.83 0.87 0.85 0.57 0.38 0.98 0.94 1.03 2.48 1.36 0.81 0.74 0.81 1.17

1971.09 2468.97 2611.00 665.53 724.05 3732.52 1088.72 1177.87 2563.91 412.15 2803.01 1738.72 2093.24 2260.39 2396.14 722.00

20.69 16.65 17.93 4.28 4.45 9.03 2.18 8.83 9.01 1.99 86.88 25.32 11.23 11.48 18.80 26.20

1.05 0.67 0.69 0.64 0.62 0.24 0.20 0.75 0.35 0.48 3.10 1.46 0.54 0.51 0.78 3.55

comparable photostability (7,17,19). This inconsistent finding might result from the poor solubility of MBBT in either the organic or aqueous phase (17). Although the mixture of stock solutions was stable at ambient conditions for a couple of days, the standard solution has to be freshly prepared when concentrations of UV filters in Table VI Results of Interday Validation for 16 UV Filters 10 ppm

25 ppm

40 ppm

UV filters

Area

SD

%RSD

Area

SD

%RSD

Area

SD

%RSD

PABA PBS TDS BZ-8 BZ-3 IMC DT MBC OC MA ED-PABA BMDM EMC DBT EHT MBBT

478.23 602.43 639.59 161.29 175.05 909.88 262.57 278.26 621.38 101.12 690.38 415.32 512.11 550.55 583.17 177.67

7.32 7.87 7.27 4.53 4.53 6.77 10.14 10.40 6.95 4.34 30.01 5.62 8.04 11.44 8.66 6.35

1.53 1.31 1.14 2.59 2.59 0.74 3.86 3.74 1.12 4.29 4.34 1.35 1.57 2.08 1.49 3.57

1234.03 1551.95 1643.15 413.52 450.76 2334.45 677.07 724.35 1594.98 262.90 1827.87 1084.42 1315.83 1420.63 1504.86 468.34

17.15 14.72 15.29 11.46 9.43 18.91 19.36 17.06 16.71 6.98 85.40 16.59 14.69 15.36 15.14 12.57

1.39 0.95 0.93 2.77 2.09 0.81 2.86 2.36 1.05 2.65 4.67 1.53 1.12 1.08 1.01 2.68

1964.42 2462.31 2604.33 717.38 717.38 3725.85 1085.38 1171.21 2557.24 408.82 2863.01 1732.05 2086.57 2253.72 2389.47 721.13

20.63 17.68 19.15 11.57 11.57 14.33 7.64 11.18 14.69 6.17 126.87 25.50 14.18 12.29 17.85 30.59

1.39 0.72 0.74 1.61 1.61 0.38 0.70 0.95 0.57 1.51 4.43 1.47 0.68 0.55 0.75 4.24

MONITORING OF UV FILTERS BY HPLC

113

Table VII Results of Recovery (%) Study for 16 UV Filters Spiked into Cream and Lotion Matrixes Cream matrix UV filters PABA PBS TDS BZ-8 BZ-3 IMC DT MBC OC MA ED-PABA BMDM EMC DBT EHT MBBT

15 ppm

20 ppm

Lotion matrix 25 ppm

15 ppm

20 ppm

101.49 ± 7.75 98.14 ± 0.87 103.64 ± 0.65 106.32 ± 0.66 101.92 ± 0.77 98.71 ± 1.96 103.04 ± 2.04 98.23 ± 2.18 99.07 ± 1.14 105.66 ± 0.20 101.10 ± 7.85 97.35 ± 0.70 102.13 ± 0.73 105.57 ± 0.32 101.42 ± 0.50 99.70 ± 7.61 96.52 ± 0.80 101.82 ± 0.59 105.24 ± 0.65 100.79 ± 0.69 98.96 ± 7.69 95.96 ± 0.90 101.72 ± 0.61 104.34 ± 0.61 99.90 ± 0.73 99.68 ± 7.68 96.62 ± 0.88 101.93 ± 0.58 103.79 ± 0.60 99.36 ± 0.75 96.47 ± 3.66 97.97 ± 2.80 97.10 ± 1.90 94.90 ± 0.54 97.35 ± 0.73 101.69 ± 8.15 97.15 ± 0.88 104.00 ± 0.86 105.43 ± 0.78 99.16 ± 0.83 101.17 ± 7.73 97.82 ± 0.88 102.86 ± 0.59 105.28 ± 0.67 100.86 ± 0.61 101.80 ± 7.09 99.02 ± 0.69 104.23 ± 0.50 104.30 ± 0.59 100.06 ± 0.78 99.07 ± 7.76 95.88 ± 0.90 100.79 ± 0.59 102.75 ± 0.41 98.08 ± 0.81 97.95 ± 7.53 94.98 ± 0.73 99.47 ± 0.50 100.82 ± 0.35 96.70 ± 0.76 98.69 ± 7.68 95.47 ± 0.77 100.50 ± 0.60 102.04 ± 0.52 97.41 ± 0.75 100.56 ± 7.79 97.62 ± 0.70 102.07 ± 0.66 104.60 ± 0.48 100.13 ± 0.73 100.63 ± 1.47 104.94 ± 2.70 96.74 ± 0.51 98.39 ± 0.83 101.01 ± 1.35 108.09 ± 4.10 116.54 ± 7.89 114.42 ± 5.47 109.99 ± 1.31 116.27 ± 0.33

25 ppm 100.97 ± 2.55 96.25 ± 1.61 100.68 ± 2.43 99.95 ± 2.46 99.08 ± 2.44 98.73 ± 2.39 95.64 ± 0.29 100.04 ± 2.61 100.22 ± 2.39 99.52 ± 2.27 97.55 ± 2.33 95.03 ± 2.27 96.81 ± 2.40 99.45 ± 2.37 95.67 ± 2.30 116.21 ± 0.45

Table VIII Repeatability for UV Filters Spiked into Cream and Lotion Matrixes Cream matrix 15 ppm Area DT MBC MA BZ-3 BZ-8 BMDD EHT OC ED-PABA EMC PABA PBS IMC DBT MBBT TDS

286.12 466.10 97.68 252.85 222.37 742.39 855.65 199.36 594.61 366.13 750.58 726.14 601.31 636.08 895.84 278.22

SD

20 ppm Area

SD

Lotion matrix 25 ppm Area

SD

15 ppm Area

SD

20 ppm Area

SD

0.30 387.50 0.61 476.42 0.15 274.78 0.30 380.18 0.20 0.37 572.61 0.93 772.14 1.32 458.43 0.72 587.15 0.80 0.22 123.00 0.46 162.51 0.35 95.33 0.24 124.40 0.14 0.19 317.29 0.55 421.10 0.67 254.42 0.24 330.71 0.34 0.14 278.32 0.17 368.95 0.21 224.06 0.17 291.35 0.43 0.41 934.14 1.05 1232.23 2.36 726.61 1.15 953.81 1.43 0.96 1201.83 2.73 1399.76 1.97 828.84 0.95 1157.96 2.45 0.06 248.77 0.39 328.31 0.31 197.43 0.18 257.65 1.27 0.13 742.13 0.65 981.79 2.37 585.68 0.53 761.45 1.34 0.30 457.38 0.45 605.95 0.56 359.64 0.41 467.40 0.31 0.54 940.09 0.57 1246.49 0.56 750.68 0.68 978.19 1.21 0.21 1003.63 0.66 1145.22 0.99 705.89 0.41 1008.05 0.98 0.24 753.15 0.56 998.28 0.66 596.60 0.43 775.84 0.52 0.57 791.23 0.77 1047.98 0.58 632.48 0.62 825.07 0.68 0.89 1123.17 0.75 1478.59 1.05 885.72 0.57 1154.94 0.88 0.79 402.67 0.67 488.45 1.07 276.65 1.07 388.13 1.22

25 ppm Area

SD

464.21 748.21 156.05 413.66 364.42 1185.67 1411.88 321.96 954.65 586.86 1222.55 1147.44 972.62 1046.02 1457.26 478.60

0.48 0.63 0.39 0.34 0.23 1.48 1.67 0.14 0.91 0.79 0.93 0.03 0.77 0.56 0.82 0.67

114

JOURNAL OF COSMETIC SCIENCE Table IX Results of Stability Test for 16 UV Filters Correlation coefficient (R2) UV filters

1 day

3 day

6 day

PABA PBS TDS BZ-8 BZ-3 IMC DT MBC OC MA ED-PABA BMDM EMC DBT EHT MBBT

0.9997 0.9996 0.9996 0.9996 0.9996 0.9997 0.9996 0.9997 0.9997 0.9993 0.9996 0.9996 0.9996 0.9996 0.9996 0.9989

0.9994 0.9994 0.9993 0.9985 0.9994 0.9994 0.9993 0.9994 0.9993 0.9974 0.9994 0.9994 0.9993 0.9993 0.9992 0.9456

0.9994 0.9994 0.9994 0.9994 0.9984 0.9995 0.9993 0.9992 0.9994 0.9946 0.9994 0.9930 0.9915 0.9994 0.9993 0.8525

sunscreen samples are measured. The mixture of stock solutions was readily precipitated at 4°C. ANALYTICAL APPLICATION

The method was applied to 23 commercial sunscreens to represent various types, such as cream, lotion, makeup, powder, and oil. The sunscreens included one to five UV filters, of which BMDM and EMC were predominantly contained in the products. However, a sunscreens containing DT, BZ-8, or PABA was not commercially available in a local market. The sunscreen samples were homogenized in MeOH under sonification to analyze in HPLC (Figure 3). For the recovery of EHT and MBBT, the sunscreens were additionally prepared in THF. The recoveries of UV filters accounted for 90.12–125.15%, compared with the label claims for each sunscreen (Table X). These results fulfilled the criteria for sunscreens that are controlled to contain more than 90% of the label claim but not to exceed the maximum authorized concentration listed in Table I.

CONCLUSIONS

Consequently, simultaneous determination of 16 UV filters was performed using HPLC detected at 300, 320, and 360 nm without any analytical predicament. The mobile phase was used as a gradient, with MeOH and phosphate buffer adjusted to pH 4.0. The LODs of the UV filters were 0.08–1.94 μg/ml, and the LOQs were 0.24–5.89 μg/ml. The linearity displayed up to 50 ppm. According to the properties of UV filters, MeOH or THF were utilized for sample preparation. MeOH was used as an extracting solvent for DT,

MONITORING OF UV FILTERS BY HPLC

115

Figure 3. (A) Chromatograms of sunscreen sample. (B) Blank matrix at 300 (top), 320 (middle), and 360 nm (bottom). The sample (lotion 3) contains five UV filters: PBS (2), BZ-3 (5), MBC (8), BMDM (12), and DBT (14).

— — 98.65 — — — 105.48 — — 105.84 109.14 97.17 — — — — — — — — — — —

Lotion 1 Lotion 2 Lotion 3 Lotion 4 Lotion 5 Lotion 6 Lotion 7 Lotion 8 Lotion 9 Cream 1 Cream 2 Cream 3 Cream 4 Cream 5 Cream 6 Cream 7 Cream 8 Cream 9 Cream 10 Stick Powder Makeup Oil

— — — — — — 94.35 90.12 — — — — — 90.53 96.34 — — — — — — 92.37 95.01

TDS

— — — — — 91.23 96.42 — 93.80 — 92.42 — — 94.54 — — 90.86 — — — — — —

BZ-3 — — 96.59 — — — — — — — — — — — — — — 95.69 — — — — —

IMC — — — — 98.16 100.95 — — — — 91.86 — 106.30 — — — — 97.18 — — — — —

MBC — — 101.95 — — — — — — 97.46 — — — — 95.09 — — 100.91 97.25 93.17 — — —

OC

* Recoveries were calculated on the basis of label claims for each sunscreen.

PBS

Sunscreens — — — — — — — — 90.58 — — — — — — — — — — — — — —

MA — — — — — — — — — — — — — — — 100.35 — — — — — — —

ED-PABA

UV filters

— 105.79 125.15 — 104.73 — — — — 105.52 100.45 — — — 111.64 — — 110.58 99.14 — — — —

BMDM

Table X Recovery (%) of UV filters in Commercial Sunscreens*

— 93.00 106.39 99.35 — 98.11 104.74 102.97 107.23 — — 98.26 103.62 102.20 — 91.77 102.93 95.06 — 98.75 95.97 98.03 105.34

EMC 92.69 106.78 — — — — — — — 106.19 97.93 101.50 — — — — — — — — — — —

DBT — 100.55 — — — — — — — — — 93.45 — — — — — — — — — — —

EHT

— — — 98.16 94.79 100.64 — — — — — — 100.67 — — — — — 101.57 — 98.94 — —

MBBT

116 JOURNAL OF COSMETIC SCIENCE

MONITORING OF UV FILTERS BY HPLC

117

MA, MBC, BZ-3, BZ-8, BMDM, OC, ED-PABA, EMC, PABA, IMC, DBT, PBS, and TDS, and THF was used for EHT and MBBT. The recoveries from the spiked UV filters were 94.90–116.54%, depending on the matrixes used. The method was applied to 23 commercial sunscreen samples to determine the contents of 16 UV filters, which showed acceptable levels compared with label claims and maximum authorized concentrations. Consequently, the method provides assessable and efficient determination of UV filters, including DBT, TDS, and MBBT, that are recently registered in the EU or USA.

ACKNOWLEDGMENT

This study was supported by a grant (09101 KFDA 251) from the Korea Food & Drug Administration.

REFERENCES (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19)

C. Tuchinda, H. W. Lim, U. Osterwalder, and A. Rougier, Dermatol. Clin., 24, 105–117 (2006). E. C. De Fabo, F. P. Noonan, T. Fears, and G. Merlino, Cancer Res., 64, 6372–6376 (2004). P. Kullavanijaya and H. W. Lim, J. Am. Acad. Dermatol., 52, 937–958 (2005). D. Sehedic, A. Hardy-Boismartel, C. Couteau, and L. J. Coiffard, Arch. Dermatol. Res., 301, 603–608 (2009). A. Fourtanier, D. Moyal, and S. Seite, Photodermatol. Photoimmunol. Photomed., 24, 164–174 (2008). S. Scalia, R. Tursilli, and V. lannuccelli, J. Pharm. Biomed. Anal., 44, 29–34 (2007). L. R. Gaspar and P. M. Maia Compos, Int. J. Pharm., 307, 123–128 (2006). P. P. Agin, K. Ruble, S. J. Hermansky, and T. J. McCarthy, Photodermatol. Photoimmunol. Photomed., 24, 211–217 (2008). S. Durrer, C. Ehnes, M. Fuetsch, K. Maerkel, M. Schlumpf, and W. Lichtensteiger, Environ. Health Perspect., 115(Suppl. 1), 42–50 (2007). J. Ashby, H. Tinwell, J. Plautz, K. Twomey, and P. A. Lefevre, Regul. Toxicol. Pharmacol., 34, 287–291 (2001). A. Zenker, H. Schmutz, and K. Fent, J. Chromatogr. A, 1202, 64–74 (2008). L. Dimosthenis, L. Giokas, A. Salvador, and A. Chisvert, Trends Anal. Chem., 26, 360–374 (2007). D. De Orsi, G. Giannini, L. Gagliardi, R. Porrà, S. Berri, A. Bolasco, I. Carpani, and D. Tonelli, Chromatographia, 64, 509–515 (2006). D. J. Schakel, D. Kalsbeek, and K. Boer, J. Chromatogr. A, 1049, 127–130 (2004). A. Chisvert and A. Salvador, J. Chromatogr. A, 977, 277–280 (2002). A. Chisvert, M. C. Pascual-Marti, and A. Salvador, Fresenius’ J. Anal. Chem., 369, 638–641 (2001). C. Couteau, A. Faure, J. Fortin, E. Paparis, and L. J. Coiffard, J. Pharm. Biomed. Anal., 44, 270–273 (2007). H. R. Buser, M. D. Muller, M. E. Balmer, T. Poiger, and I. J. Buerge, Environ. Sci. Technol., 39, 3013– 3019 (2005). E. Chatelain and B. Gabard, Photochem. Photobiol., 74, 401–406 (2001).