Journal of Chromatography A,1049(2004)
127–130
韩国网站大全>jiggyDetermination of sixteen UVfilters in suncare formulations by
high-performance liquid chromatography
D.J.Schakel∗,D.Kalsbeek,K.Boer
Food and Consumer Product Safety Authority,Inspectorate for Health Protection and Veterinary Public Health(VWA/KvW),
Rearch Lab Non Food Chemical,Paterswoldweg1,9726BA Groningen,The Netherlands
Received22June2004;received in revid form9July2004;accepted28July2004
Abstract
Reverd-pha liquid chromatography(RP-LC)with UV detection was developed for the simultaneous determination of16organic UVfilters worldwide authorid in suncare products.Thefilters determined were:4-aminobenzoic acid,homosalate,benzophenone-3,2-phenylbenzimidazole-5-sulfonic acid,terephthalidene dicamphor sulfonic acid,4-tert-butyl-4 -methoxy-dibenzoylmethane,octocrylene,2-ethylhexyl-4-methoxycinnamate,isoamyl-p-methoxycinna
mate,ethylhexyltriazone,drometrizole trisiloxane,diethylhexyl butamido triazone, 3-(4-methylbenzyliden)camphor,2-ethylhexylsalicylate,2-ethylhexyl-4-dimethylaminobenzoate and benzophenone-4.A C18stationary pha and a gradient of ethanol-aqueous acetate buffer containing0.2mM of EDTA,was ud with aflow-rate of1.0ml/min.UV detection was carried out at313and360nm.The analysis required32min and the limits of detection were between30and4130mg/kg in the original suncare product.Tween80was ud to break down the different emulsions in order to procure a proper extraction of the UVfilters.The method was validated for UVfilters in three matrices,oil,water-in-oil emulsion and oil-in-water emulsion.Recoveries from spiked samples were86–113%depending on the matrix ud.
©2004Elvier B.V.All rights rerved.
Keywords:Suncare products;UVfilters;Tween80;EDTA
1.Introduction
The growing publicity about the damaging effects of UV radiation,the hole in the ozone layer and protection against UV radiation receive great public interest.The protection consists predominantly of the ability of a sunscreen tofilter out UVB rays(290–320nm).UVB rays are responsible for sunburn.S
uncare products do in some cas claim a protec-tion against UV A(320–400nm).Exposure to UV A caus skin ageing.In contrast to UVB,UV A does not cau sun-burn.
留学A list of approved UVfilters and their maximum allowed concentrations in commercial products have been drawn up by the regulatory authority in Europe in Annex VII of Direc-tive76/768/EEC(Table1).
∗Corresponding author.Tel.:+31505886000;fax:+31505886100.
E-mail address:durk.schakel@vwa.nl(D.J.Schakel).
Many methods are reported to quantify UVfilters in cos-metics.UVfilters are easy to determine with liquid chro-matography(LC)in combination on different types of sta-tionary pha and with a great variety of mobile phas[1–4]. Isocratic as well as gradient elution has been ud.Often the methods are appropriate to determine four to six UVfilters. In some cas different mobile pha-column combination are ud within a method.Due to the similar structure of some of the UVfilters the baline paration gave difficul-ties,even when gradient elution was ud.The most difficult UVfilters to parate are BMDBM,EMC,ED-PABA,ES and HMS.HMS especially is problematic becau this UV filter prents two peaks corresponding two isomeric forms [2].Together with BMDBM and ES migration takes place in the same time window and even the spectra are similar.In t
he cas the method is only for identification purpos[1].
The extraction consists of breaking the usually very com-plex emulsions.Besides water-in-oil(W/O)and oil-in-water
0021-9673/$–e front matter©2004Elvier B.V.All rights rerved. doi:10.1016/j.chroma.2004.07.068
128 D.J.Schakel et al./J.Chromatogr.A1049(2004)127–130
Table1
Maximum allowed concentrations stated in Directive76/768/EEC
UVfilter Concentration
(%)
4-Aminobenzoic acid(PABA)5 Homosalate(HMS)10 Benzophenone-3(BENZ-3)10
2-Phenylbenzimidazole-5-sulfonic acid(PBSA)8 Terephthalidene dicamphor sulfonic acid(TDSA)10
4-tert-Butyl-4 -methoxy-dibenzoylmethane(BMDBM)5 Octocrylene(OC)10
2-Ethylhexyl-4-methoxycinnamate(EMC)10
Isoamyl-p-methoxycinnamate(IMC)10 Ethylhexyltriazone(ET)5 Drometrizole trisiloxane(DTS)15 Diethylhexyl butamido triazone(DBT)10
3-(4-Methylbenzyliden)camphor(MBC)4
valentines2-Ethylhexylsalicylate(ES)5妥当的读音
2-Ethylhexyl-4-dimethylaminobenzoate(ED-PABA)8 Benzophenone-4(BENZ-4)5
(O/W)emulsions ternary systems are found often.In most cas extraction of cosmetics is performed with ethanol or methanol,if necessary in combination with low pH and/or high temperature(60◦C).It ems that vigorous shaking is necessary to break the emulsion.If not,the result will not be reproducible.Another way to break an emulsion system is to add a surfactant.In cosmetic microbiology the u of surfac-tants is well accepted in tho cas the product is immiscible with water.Tween80is preferred most of the time.
In the prent study extraction took place in a waterbath at60◦C,followed by vigorous shaking and ultrasonication at ambient temperature with ethanol as extraction solvent and Tween80for breaking down the emulsion.Ethylene di-aminotetraacetic acid(EDTA)was ud as a modifier of the mobile pha to perform a HPLC paration of sixteen UV filters,using a C18stationary pha and ethanol-aqueous ac-etate buffer mobile pha with gradient elution.UV detection was carried out at313and360nm.The method was tested for robustness and validated for12UVfilters.
2.Experimental
2.1.Reagents
reeds4-Aminobenzoic acid,benzophenone-3,2-phenylbenzi-midazole-5-sulfonic acid and2-ethylhexyl-4-methoxycin-namate were obtained from Across(Landsmeer,The Netherlands),homosalate,4-tert-butyl-4 -methoxy-diben-zoylmethane,octocrylene,3-(4-methylbenzyliden)camp-hor,2-ethylhexylsalicylate and2-ethylhexyl-4-dimethyl-aminobenzoate from Merck obtained supplied by Boom (Meppel,The Netherlands),terephthalidene dicamphor sul-fonic acid and drometrizole trisiloxane from L’Oreal (Paris,France),isoamyl-p-methoxycinnamate from Geyer (Friederichsthal,Germany),ethylhexyltriazone from BASF (Ludwigshafen,Germany),diethylhexyl butamido triazone Table2
Gradient time table
Time(min)Solvent A(%)Solvent B(%) 06040
46040
57525
187525
191000
5202
271000
286040
Solvent A:ethanol(containing80.0mg EDTA dipotassium salt dihydrate dissolved in5ml water per litre),solvent B:buffer(aqueous solution contain-ing56.7mg sodiumacetate and80.0mg EDTA dipotassium salt dihydrate per litre,pH adjusted to2.5with glacial acetic acid).
from Beiersdorf(Hamburg,Germany)and benzophenone-4from ICN(Zoetermeer,The Netherlands),we
re ud to prepare the standards.The solvents ud,ethanol96%and glacial acetic acid from Merck,were supplied by Boom.Poly-oxyethylene sorbitan monooleate from Merck,ud to disrupt the emulsion system,was supplied by Boom.Ethylenedi-aminetetraacetic acid dipotassium salt dihydrate from Merck, ud as modifier for the mobile pha,was supplied by Boom.
2.2.Chromatography
An Agilent liquid chromatographic system equipped with a binary pump,an injector with variable loop and a DAD was ud.The above system was controlled using ChemStation software(Agilent Technologies).A5-m LiChrospher100 RP-C18column(125mm×4.6mm)was ud.The mobile pha was a gradient of ethanol-aqueous acetate buffer con-taining EDTA.LC was performed at28◦C with gradient elu-tion at1.0ml/min as described in Table2.Injection volume was20l.UV absorption was done at313and360nm.The run time was32min.Fig.1shows chromatograms obtained under the
conditions.
Fig.1.Chromatograms of16UVfilters.Conditions as mentioned in the text.
D.J.Schakel et al./J.Chromatogr.A1049(2004)127–130129
2.3.Standard solutions
Stock standard solutions of the UVfilters were prepared daily.The individual solutions of all UVfilters except for PBSA,TDSA and ET were prepared in ethanol and their concentration was about4mg/ml.Stock solution of PBSA with concentration of4mg/ml was prepared by dissolving 200mg in2ml2M sodium hydroxide and then diluted to 50ml with ethanol.Stock solution of TDSA with concentra-tion of4mg/ml was prepared by dissolving200mg in2ml 25%acetic acid and then diluted to50ml with ethanol.Stock solution of ET with concentration of4mg/ml was prepared by dissolving200mg in2ml ethyl acetate and then diluted to50ml with ethanol.
Keeping the allowed concentrations of the UVfilters in mind,different concentration ranges were prepared for the working standard solutions.If for any reason TDSA is not prent in the working standard solutions check the pH be-ing7.0due to the instability of HMS,OC and ES at high pH.Two working standard solutions were prepared becau PABA,PBSA,TDSA and BENZ-4migrated in the same time window.The concentration ranges for the parate UVfilters are typically between10and2
00mg/l.
2.4.Sample preparation
Commercial samples were purchad in local shops.Sam-ples of0.5g were dissolved in25ml ethanol in the prence of0.5ml Tween80.Extraction was performed in a water-bath at60◦C for10min followed by vigorous shaking during 30s and ultrasonication at ambient temperature for10min. Each step homogenisation was performed.After cooling to ambient temperature the solution was transferred to a50ml volumetricflask and diluted with ethanol.This solution was diluted10times in ethanol by transferring1ml in a10ml vol-umetricflask.If PABA is prent in the sample,another dilu-tion was performed with ethanol–acetate buffer(60:40,v/v) parallel with the ethanol dilution.Prepared solutions were then injected into the HPLC system.
3.Results and discussion
3.1.Robustness
Robustness of the method was determined with an internal procedure[5].With this procedure a number of potential criti-cal parameters in the analytical method were varied in order to test whether
the results remained constant.The experiments were performed for an O/W-emulsion and a W/O-emulsion.
Labels showed that BENZ-3,BMDBM,EMC and MBC were most ud as UVfilters in sun protection products.EMC alone or the combination of EMC with BENZ-3and BMDBM was the most udfilter in sun protection products.Therefore, the robustness test included only thefilters,extended with PABA,PBSA and ET due to the prence of the UVfilters in the investigated matrix or due to deviating molecule structure.
It emed that the column temperature is a critical param-eter.Experimental work showed that the temperature of the column should be28±1◦C in order to keep the resolution optimid.
3.2.Mobile pha
The method was developed using a gradient elution as described in Table2without using EDTA.During analysing commercial samples the performance of the column changed. The cosmetic matrix altered the properties of the station-ary pha of the chromatographic column.The chromato-graphic behaviour of BMDBM changed.Due to the tailing peak of BMDBM quantitation of BMDBM and ES,which elutes in the same time window,became difficult.Sometimes compounds that are of no
interest can interact with residual silanols.Retention times can shift and tailing can occur[6].
A chelating reagent as EDTA can beflushed through the col-umn to overcome this problem.Inactivation of metaloproteins could be realid with0.1–1mM EDTA.
Addition of EDTA resulted in a good chromatographic peak shape.The higher the concentration of EDTA,the better the peak shape of BMDBM became;moreover,resolution of HMS,BMDBM and ES was influenced positively.Since a high concentration of EDTA will contaminate the pump,due to the insolubility of EDTA in ethanol,0.2mM EDTA was chon.
pisa
3.3.Analytical performance
The LOD was defined as the analyte concentration that gives a signal equal to3y b,where y b is the baline noi at the retention time the analyte was expected to migrate.Similarly, the LOQ was defined as9y b.Bad on the above-mentioned equations,the calculated LOQ values of the original sample are calculated(Table3).
Bad on structure and frequency of UVfilters in407 commercial samples,the accuracy of the method is tested for 12UVfilters.The recovery was studied by standard addi-tion of the UVfilters at th
ree different levels.Addition was performed in duplicate at half the legal limit,the legal limit and one and a half the legal limit mentioned in Table1.The study included a W/O-emulsion,an O/W-emulsion and an oil sample.The six recoveries for an UVfilter were checked for outliers with the single and double Grubbs test[7].From the remaining data,the mean recovery of each UVfilter was calculated(Table3).
The recoveries,which varied between86and113%,do not comply with the limits according to the AOAC[8]. For the maximum allowed concentrations stated in Direc-tive76/768/EEC(Table1)the recovery should be between 98and102%.Also the maximum allowed relative standard deviation(RSD)is in some cas too high.In accordance with Pocklington[9],the RSD should not exceed2.5%.Due to the complexity of the determination of cosmetics,a moderately large imprecision of the results must be accepted.
130 D.J.Schakel et al./J.Chromatogr.A1049(2004)127–130
Table3
Recovery study for12UVfilters
UVfilter LOQ(mg/kg)W/O-emulsion O/W-emulsion Oil
Mean recovery(%)RSD(%)Mean recovery(%)RSD(%)Mean recovery(%)RSD(%) PABA2490111.63.4110.55.8113.03.7 HMS42094.62.993.21.196.32.3 BENZ-311096.61.097.73.096.42.1 PBSA4130103.31.1103.32.6102.52.4 BMDBM70105.47.185.912.4106.84.2
OC12093.61.893.01.495.00.9 EMC7096.95.593.85.393.43.4
ET152099.71.098.61.794.22.4 DTS129095.71.091.92.888.83.1 DBT105098.65.398.96.099.52.9 MBC4090.85.394.45.392.92.2
ES42097.31.796.21.597.81.6 TDSA70
IMC30
ED-PABA50
BENZ-470
4.Conclusions
From the prent study it can be concluded that with the propod method16UVfilters can be determined in all kind of suncare products.The u of Tween80as an afforcement to ethanol in combination with heat and ul-trasonication is a good alternative to break the emulsions that are prent in cosmetic products.EDTA as mobile pha modifier makes it possible to create a good chromato-graphic peak for BMDBM.As a conquence the resolu-tion of HMS,BMDBM and ES improved.The migration of PABA,PBSA,TDSA and BENZ-4in the same time win-dow is no problem.From407different commercial sam-ples can be concluded that combinations of the UVfilters rarely occurred.In one sample PBSA and TDSA were both prent.
At prence of PABA the sample should also be diluted with ethanol–acetate buffer to determine PABA.It emed that peak splitting occurred for PABA when diluted with ethanol.Of the407commercial suncare products only one contained PABA.
Recoveries from spiked samples were86–113%depend-ing on the ud matrix.Due to the complex matrix of cosmet-ics a moderate great deviation in the results must be accepted. References
[1]S.C.Rastogi,G.H.Jenn,J.Chromatogr.A828(1998)311.
[2]A.Chisvert,M.C.Pascual-Marti,A.Salvador,J.Chromatogr.A921
广州新东方英语学校(2001)207.
[3]P.Schneider,A.Bringhen,H.Gonzenbach,Drug Cosmet.Ind.159
(1996)32.
[4]U.Hauri,B.L¨u tolf,C.Hohl,Mitteilungen der Lebensmittel Hygeine
94(2003)80.
[5]A.Stern,H.A.P.M.Jann,Quality Assurance Handbook,Food and
Consumer Product Safety Authority,Inspectorate for Health Protection and Veterinary Public Health,2002,KAM05-WV104.
[6]R.E.Majors,LC-GC,July2003.
[7]I.Franke,Quality Assurance Handbook,Food and Consumer Product
成都雅思培训Safety Authority,Inspectorate for Health Protection and Veterinary Public Health,2003,KAM05-WV007.
[8]AOAC Peer-Verified Methods Program,Manual on Policies and
Procedures,1995.
[9]W.D.Pocklington,Pure Appl.Chem.62(1990)149.
