hplc analyis of preservatives in cosmetics

Screening procedure for determination of preservatives in commercial skin-care products Larisa Borisova-Jan, Stefan Jönsson, Dick Fransson, Monica Tammela, Monika Johansson 2011-10-06

Postadress/Postal address: P.O. Box 26, SE-751 03 Uppsala, SWEDEN Besöksadress/Visiting address: Dag Hammarskjölds väg 42, Uppsala Telefon/Phone: +46 (0)18 17 46 00 Fax: +46 (0)18 54 85 66 Internet: www.lakemedelsverket.se E-mail: [email protected]

Register Abstract ....................................................................................................................................................... 3 1. Introduction ............................................................................................................................................. 3 2. Experimental ........................................................................................................................................... 5 2.1. Materials ............................................................................................................................................ 5 2.2. Schiff’s test for rapid detection of low levels of aldehydes ................................................................ 6 2.2.1. Spiked sample ........................................................................................................................... 6 2.2.2. Test sample ............................................................................................................................... 6 2.2.3. Analysis ...................................................................................................................................... 6 2.3. HPLC ................................................................................................................................................. 6 2.3.1. Sample preparation .................................................................................................................... 6 2.3.2. Test sample for reanalysis for HPLC method 3 ......................................................................... 7 2.3.3. HPLC method 1: Determination of BNPD, DMDM Hydantoin, IPBC, MDBGN, Kathon and parabens .............................................................................................................................................. 7 2.3.4. HPLC method 2: Determination of diazolidinyl urea, imidazolidinyl urea and Quaternium-15 .. 7 2.3.5. HPLC method 3: Determination of formaldehyde ...................................................................... 8 3. Result and discussion............................................................................................................................ 8 3.1. Schiff’s test for rapid detection of low levels of aldehydes .............................................................. 10 3.2. HPLC method 1: Determination of Bronopol (BNPD), DMDM hydantoin, IPBC, MDBGN, Kathon and parabens in commercial skin creams ............................................................................................. 10 3.3. HPLC method 2: Determination of diazolidinyl urea, imidazolidinyl urea and Quaternium-15 ....... 14 3.4. HPLC method 3: Determination of formaldehyde ........................................................................... 16 3.5. Method validation ............................................................................................................................ 17 3.6. Application ....................................................................................................................................... 18 4. Conclusions .......................................................................................................................................... 20 5. Acknowledgements .............................................................................................................................. 20 6. References ............................................................................................................................................ 20

2

Abstract A convenient screening procedure for the separation, detection and quantification of 10 commonly-used preservatives in skin creams is presented. The developed procedure consists of Schiff’s test for rapid detection of low levels of aldehydes plus three separate HPLC methods with UV detection. A reverse-phase HPLC gradient method run on a Symmetry C18 column was used to simultaneously determine 2-bromo-2-nitropropane-1,3-diol (BNPD, Bronopol ), DMDM hydantoin, iodopropynylbutylcarbamate (IPBC), Kathon, methyldibromoglutaronitrile (MDBGN) and methyl-, ethyl-, propyl-, iso-propyl-, butylparaben. A ZIC-HILIC column was found suitable for detecting and quantifying diazolidinyl urea, imidazolidinyl urea and Quaternium-15. The content of free formaldehyde in the skin creams was determined using the same ZIC-HILIC column with a 2-propanol gradient and post-column derivatization. The developed screening procedure has successfully been applied in a study of the preservative content of 99 commercial skin creams regarding the correct labelling of ingredients as well as their maximum allowed concentration. The preservatives could be quantified to within 10-200% of their maximum allowed concentration. The presented procedure is now implemented as routine screening in our laboratory for the regular control of the preservative content of skin creams available on the Swedish market. Keywords: Cosmetics; Preservatives determination; Formaldehyde releasers; Analytical control; HPLC.

1. Introduction Skin-care products comprise many different formulations (creams, lotions, oils, etc.) but all are called skin creams in this paper. Skin creams usually consist of many components, including preservatives. Consumer health and safety is the main reason for including preservatives as antimicrobial additives in skin cream formulations. Preservatives are often used as a multicomponent mixture to increase the range of microorganisms against which the product is protected. However, preservatives are also important causes of allergic contact dermatitis [1-4]. Since skin-care products are used daily by many people, correct and complete ingredient labelling is important, and the use of preservatives in skin products is strictly regulated. Ingredients in cosmetic products are labelled in accordance with EU legislation [5]. Correct labelling means that people with sensitivities can be aware of any preservatives in a product formulation that could trigger an allergic reaction. Several studies of skin cream preservative content regarding correct ingredient labelling as well as maximum allowed concentration have been performed earlier [6-8]. The separate EU standard [9, 10] or individual methods for each preservative [6-8] were used in these earlier investigations. Their methods were based on highperformance liquid chromatography (HPLC) analysis and performed under different chromatographic conditions. The procedures were also time-consuming because individual test samples were prepared for each studied preservative. The Swedish Medical Products Agency regularly investigates the preservative content of skin creams that are available on the Swedish market. A list of preservatives commonly studied is presented in Table 1.

3

Table 1. Preservatives and maximum allowed concentration, % w/w. Maximum allowed Preservative concentration 2-Bromo-2-nitropropane-1,3-diol (BNPD, Bronopol) 0.1 Diazolidinyl urea 0.5 DMDM hydantoin 0.6 Formaldehyde 0.2 (0.05 a)) Imidazolidinyl urea 0.6 Iodopropynylbutylcarbamate (IPBC) 0.05 c) Kathon (MCI/MI) 0.0015 d) Methyldibromoglutaronitrile (MDBGN) 0.1 Parabens 0.4 (0.8 b)) Quaternium-15 0.2 a) b) c) d)

Presence must be declared on the label if the concentration in the product exceeds 0.05%; Total, when used as a mixture; Mixture (3:1) of MCI and MI with magnesium nitrate and magnesium chloride as stabilizers; Prohibited in cosmetics in EU from 2008.

The preservatives investigated include the seven that were monitored for patch test sensitivity in Europe as a 10-year overview for the period 1991-2000 [11]. For qualitative and quantitative analyses of preservatives, a large number of different chromatographic methods have been previously reported in the literature. Parabens, the most commonly used preservative, have been determined by several techniques, including HPLC [12-14], LC-MS [15, 16], ion-interaction reversed-phase liquid chromatography [17], micellar liquid chromatography (MLC) [18], GCMS [19], and flow injection analysis [20]. Micellar electrokinetic capillary chromatography (MEKC) was proposed as a method of choice for analysing formulations containing an association of imidazolidinyl urea with parabens [21]. Analyses of diazolidinyl- and imidazolidinyl urea have been made by reverse-phase HPLC using a cyanopropyl-bonded silica column [22] as well as by capillary electrophoresis (CE) [23]. Simultaneous measurement of diazolidinyl urea, urea and allantoin by hydrophilic interaction chromatography (HILIC) has been demonstrated [24]. Quaternium-15 [25], DMDM hydantoin [26], iodopropynylbutylcarbamate (IPBC) [27], and methyldibromoglutaronitrile (MDBGN) [28, 29] have been qualitatively analysed by different chromatographic methods. Several HPLC methods have been suggested for detecting Kathon [30-33] and Bronopol together with its degradation product [35-37]. Simultaneous determination of 21 preservatives including Kathon, Bronopol and parabens was performed by ultraperformance liquid chromatography (UPLC) [38]. Qualitative and quantitative detection of formaldehyde has been widely described in the literature [29-47]. Separation of imidazolidinyl urea and a mixture of parabens was achieved by using a cyanopropyl column and a methanol gradient [48]. However, these compounds were eluted rapidly, which made their quantification in skin creams with polar ingredients very complicated. Methods in the literature mainly address the separation of preservatives characterized by a similar structure. For the separation of preservatives characterized by different hydrophilicities and chemical properties, the use of complex detection systems as well as different sets of conditions and gradient elution was required. Thus, the separation of the 47 preservatives listed in the current EEC Council Directive on cosmetic products was performed by using two isocratic

4

and two revered-phase HPLC gradient systems [49] where the substances of our interest (Kathon and parabens) were characterized by their retention times. Bronopol, parabens, benzoic acid, salicylic acid, sorbic acid, dehydroacetic acid, triclocarban, o-phenyl-phenol, 4-hydroxybenzoic acid, methyl-, ethyl-, propyl-, butyl-, benzyl- benzoates and 4-chloro-m-cresol were simultaneously separated by an ion-interaction reagent (IIR) HPLC-diode-array detection method with a Superspher 100 RP 18 column and 52.59% acetonitrile/3.25 mM hexylammonium phosphate pH 5.35 as mobile phase [50]. However, none of the published methods proposed the simultaneous determination of the mixture of preservatives listed in Table 1. Since these commonly-used preservatives are regularly studied in the Laboratory of the Swedish Medical Products Agency, a reliable, easy and convenient method for their detection and quantification was of great interest. This study presents a convenient screening procedure based on Schiff’s test for rapid detection of aldehydes and three HPLC methods for identifying and quantifying the 10 commonly-used preservatives.

2. Experimental 2.1. Materials Bronopol (BNPD or 2-Bromo-2-nitropropane-1,3-diol), methylparaben (methyl-4hydroxybenzoate), ethylparaben (ethyl-4-hydroxybenzoate), propylparaben (propyl-4hydroxybenzoate), and butylparaben (butyl-4-hydroxybenzoate) were obtained from Fluka (Sigma-Aldrich Sweden AB); iso-propylparaben (iso-propyl-4-hydroxybenzoate) and methyldibromoglutaronitrile (MDBGN) from Alfa Aesar GmbH & Co (Karlsruhe, Germany); diazolidinyl urea, imidazolidinyl urea, 3-iodo-2-propynyl-N-butylcarbamate (IPBC), Quaternium-15, phenoxyethanol and benzophenone from Sigma-Aldrich Sweden AB; and DMDM hydantoin from Chemos Gmbh (Regenstauf, Germany). Kathon was supplied by Rohm & Haas Biocides (The Dow Chemical Company, Buchs, Switzerland). Acetonitrile, 2-propanol and methanol (all HPLC grade) were obtained from Merck; ethanol (spectroscopy pure) was from Kemetyl. Water for preparation of buffer, reagent and sample solutions was produced in-house by the Milli-Q water purification system (ELGA Maxima). The mobile phase was degassed in an ultrasonic bath before use. Formaldehyde solution, 37% in H2O was obtained from Aldrich. Acetic acid, conc., hydrochloric acid (37%), sulphuric acid (95 – 97%), sodium hydroxide and ammonium acetate, anhydrous, were supplied by Merck; sodium sulphite, Na2SO3 · H2O by Sigma-Aldrich. Pentane-2,4-dione and parafuchsin hydrochloride were obtained from Fluka; charcoal, starch, Titrisol c (I2) = 0.05 mol/l (0.1 N) and Titrisol c (Na2S2O3) = 0.1 mol/l (0.1 N) were from Merck (Merck Sharp & Dohme (Sweden) AB, Sollentuna). 1 M NaOH, 1 M HCl and starch solution were prepared according to European Pharmacopeia. Skin creams were chosen from the Swedish market and original samples were delivered directly from the manufacturer or importer. Columns: Symmetry C18, 75x4.6 mm, 3.5 µm; ZIC-HILIC, 100x2.1 mm, 5 µm (SeQuant AB) and guard-columns 5 µm Zorbax SB-C18, 12.5x4.6 mm, 5 µm; and ZIC-HILIC, 20x2.1 mm (SeQuant AB) were supplied by Dalco Chromtech AB (Sollentuna, Sweden).

5

Filters Titan PTFE 0.45 µm were obtained from Dalco Chromtech AB (Sollentuna, Sweden).

2.2. Schiff’s test for rapid detection of low levels of aldehydes Schiff’s reagent was prepared according to the published procedure [51]. 10 ml of conc. hydrochloric acid were added slowly and under constant stirring to an ice-cold mixture of a solution of 1 g parafuchsin dissolved in 600 ml warm water and a solution of 200 g sodium sulphite in 100 ml of water. The resulting solution was transferred to a volumetric flask and diluted to 1 litre with water. After shaking with 0.2 to 0.3 g of decolourising charcoal and filtration, the solution was allowed to stand at room temperature protected from light for 24 hours. Occasionally it could be necessary to add 2 to 3 ml of conc. hydrochloric acid to remove a weak residual pink colour. The reagent was stable for two months at +8°C.

2.2.1. Spiked sample Approximately 1 g of skin cream was accurately weighed in a 10 ml screw-capped plastic tube. 1 ml of a 0.2 mg/ml solution of formaldehyde in 50% aqueous ethanol, 2 ml of Schiff’s reagent and 2 drops of sulphuric acid were then added. The mixture was stirred in a vortex mixer.

2.2.2. Test sample 2 ml of Schiff’s reagent and 2 drops of sulphuric acid were added to approximately 1 g of skin cream accurately weighed in a 10 ml screw-capped plastic tube. The mixture was stirred in a vortex mixer.

2.2.3. Analysis The colours of the spiked and test samples were compared. A pink or mauve tint observed in the test sample was considered a positive test result (i.e. aldehydes were present). Positive samples were further analysed by HPLC method 2 and HPLC method 3.

2.3. HPLC HPLC samples were placed in a defined sequence: blank solution (50% aqueous methanol or 50% aqueous ethanol) followed by three standard solutions, one more blank and then three tests samples. This sequence was repeated. The blank/standards/blank combination was thus run between every third test sample (i.e. blank, standards 1-3, blank, test, test, test, blank, standards 1-3, blank, test etc. …).

2.3.1. Sample preparation Approximately 2 g of skin cream was accurately weighed into a 50 ml screw-capped plastic tube and 20.0 ml 50% aqueous methanol (HPLC method 1) or 50% aqueous ethanol (HPLC methods

6

2 and 3) was added. The mixture was stirred in a vortex mixer, ultrasonicated for 30 min and then centrifuged for 30 min (9000 rpm). The solution was filtered through a 0.45 µm membrane filter directly into the HPLC vial.

2.3.2. Test sample for reanalysis for HPLC method 3 If the concentration of formaldehyde in the studied skin cream was found to be higher than the maximum allowed concentration, the cream was reanalysed using freshly prepared samples. Approximately 2 g was accurately weighed into a 50 ml screw-capped plastic tube and 20.0 ml of 50% aqueous ethanol was added. The suspension was shaken with slid-glass beads and then filtered into the HPLC vial. Samples were analysed immediately after preparation.

2.3.3. HPLC method 1: Determination of BNPD, DMDM Hydantoin, IPBC, MDBGN, Kathon and parabens HPLC was performed using a HP-1100 Agilent system (Agilent Technologies Sweden AB, Kista, Sweden) equipped with a quaternary pump and an Agilent 1200 Diode Array detector. Empower.2 software was used for data acquisition and analysis. The chromatographic column was a Symmetry C18, 75x4.6 mm, 3.5 µm with a Zorbax SB-C18, 12.5x4.6 mm, 5 µm guard-column. Mobile phase gradient elution was used. Solvent A was 0.1% o-phosphoric acid in water; solvent B was methanol and solvent C was 0.1% o-phosphoric acid in acetonitrile. The following gradient was used: 1% B was constant during analysis; 99% A was constant for 10 min, decreased to 49% in 30 min and then to 10% in 1 min, after which it was constant for 9 min. Equilibration before the next injection was 16 min. The mobile flow rate was 1 ml/min, the injection volume 20 µl, and the column temperature ambient. Detection wavelengths are shown in Table 2.

2.3.4. HPLC method 2: Determination of diazolidinyl urea, imidazolidinyl urea and Quaternium-15 HPLC was performed using a Waters Alliance 2695 Separation Module (Waters Sverige AB, Sollentuna, Sweden) equipped with a quaternary pump and a Waters 2996 Photodiode Array detector. Empower.2 software was used for data acquisition and analysis. The chromatographic column was ZIC-HILIC, 100x2.1 mm, 5 µm, with a ZIC-HILIC, 20x2.1 mm, 5 µm guard-column. Mobile phase gradient elution was used. Solvent A was 0.1% acetic acid in water and solvent B was 0.1% acetic acid in acetonitrile. The following gradient was used: 93% B was constant for 5 min, decreasing to 75% in 15 min and then to 60% in 2 min, after which it was constant for 5 min. Equilibration before the next injection was 8 min. The mobile flow rate was 0.2 ml/min, the injection volume 5 µl, and the column temperature ambient. Detection wavelengths are shown in Table 2.

7

2.3.5. HPLC method 3: Determination of formaldehyde HPLC was performed using a Waters Alliance 2695 Separation Module (Waters Sverige AB, Sollentuna, Sweden) equipped with a Waters 2996 Photodiode Array detector, a quaternary pump that delivers mobile phase to the column and a reagent pump that adds the post-column reagent to the column eluant. Mobile phase gradient elution was used. Solvent A was 0.1% acetic acid in water and solvent B was 2-propanol. The following gradient was used: 99% B was constant for 5 min, then decreased to 40% in 15 min. Equilibration before the next injection was 15 min. The mobile flow rate was 0.2 ml/min, the injection volume 5 µl, and the column temperature ambient. Detection wavelength was 420 nm. The derivatization procedure was used as previously described [7, 9] with some slight modifications. The reaction coil was a knitted reactor coil with a total volume of about 0.8 ml (15 m x 0.25 mm ID, a photoreactor coil from Supelco is suitable). The coil was immersed in a polyethylene glycol bath heated to 80°C (a magnetic stirrer equipped with a heater plate was used). The reaction vessel was protected from light. The flow rate for the post-column reagent was 0.1 ml/min. The post-column reagent was 62.5 g of ammonium acetate, 7.5 ml of acetic acid and 5 ml of pentane-2,4-dione diluted to 1000 ml with water. It was kept protected from light for a maximum of three days at room temperature, flow rate: 0.1 ml/min.

3. Result and discussion The selected preservatives (Table 1) belong to different classes of compounds and are often present in skin-care products as a multi-component mixture. Initially, our goal was to develop one HPLC method for the simultaneous separation, determination and quantification of all substances listed in Table 1. A number of different columns and gradients intended to reveal conditions for detecting the selected preservatives in one run were therefore tested, but without a successful outcome. Instead, a screening procedure based on Schiff’s test and three separate HPLC methods with UV detection has been developed. The investigated preservatives have different UV absorption and this was utilized for quantification. Peak areas were measured at different wavelengths (Table 2) by the use of a photodiode array (PDA) detector. The UV spectra and the specific retention time were used to identify the analytes.

8

Table 2. Quantification of preservatives. Preservative UV detection, nm Methylparaben 290 Ethylparaben 290 Propylparaben 290 iso-Propylparaben 290 Butylparaben 290 DMDM hydantoin 210 BNPD 250 IPBC 250 MDBGN 250 Kathon 274 Diazolidinyl urea 201 Imidazolidinyl urea 201 Quaternium-15 210 Formaldehyde 420

Used method 1 1 1 1 1 1 1 1 1 1 2 2 2 3

The following scheme (Fig. 1) was used in the screening procedure.

Skin creams

HPLC method 1

HPLC method 2

positive results

Schiff’s test

HPLC method 3

negative results Results

Fig. 1. Screening procedure for identification and quantification of preservatives in skin creams.

9

All selected skin creams were analysed using HPLC method 1 for the determination of Bronopol, DMDM hydantoin, IPBC, MDBGN, Kathon and parabens (reversed-phase HPLC gradient system). All creams were also tested with Schiff’s test to rapidly detect low levels of aldehydes [51]. Creams with a positive result from Schiff’s test were analysed using HPLC method 2 (hydrophilic interaction (HILIC) - HPLC system with UV diode array detection) for the determination of diazolidinyl urea, imidazolidinyl urea and Quaternium-15, and HPLC method 3 (HILIC-HPLC system with post-column derivatization) for the quantification of formaldehyde. The preservatives in the tested samples were identified by comparing their HPLC retention times and UV-spectra with the retention times and UV-spectra of the analytes prepared in standard solutions and analysed under the same conditions. Previously described sample preparation methods [52] prior to HPLC analysis of the analytes were adapted and 50% aqueous methanol (HPLC method 1) or 50% aqueous ethanol (HPLC method 2 and 3) were used as extraction solvent. However, during this method development, it was concluded that the results obtained were independent of whether methanol or ethanol was used to prepare the sample. Detailed descriptions of all methods in the used screening procedure are presented below.

3.1. Schiff’s test for rapid detection of low levels of aldehydes Schiff’s test for rapid detection of low levels of aldehydes is well known and was performed in accordance with the literature description [51]. The spiked sample containing 0.2 mg/ml of formaldehyde and the test sample were prepared (see Experimental). The spiked sample was introduced into the test procedure to facilitate the interpretation of the colour shift since beside aldehydes, creams may also contain other ingredients that can give a coloured solution. The colour formed in the test sample was visually compared with the colour of the spiked sample. In our experiments, the colour of the spiked sample was always observed as mauve. The test result was considered positive if a pink or mauve tint was observed in the test sample. Samples with positive test results were further analysed by HPLC method 2 for determination of diazolidinyl urea, imidazolidinyl urea and Quaternium-15, and HPLC method 3 for determination of formaldehyde.

3.2. HPLC method 1: Determination of Bronopol (BNPD), DMDM hydantoin, IPBC, MDBGN, Kathon and parabens in commercial skin creams The separation of Bronopol (BNPD), DMDM hydantoin, IPBC, MDBGN, Kathon and parabens was performed on a Symmetry column C18, 75x4.6 mm, 3.5 µm with a gradient containing 0.1 % o-phosphoric acid in water/ methanol/ 0.1 % o-phosphoric acid in acetonitrile and UV-detection at different wavelengths (see Table 2). Phenoxyethanol and benzophenone were also tested under these conditions since both are commonly used in skin-care products and it was therefore important to separate them from the analytes. The chromatogram of the mixture of 12 preservatives at 210 nm (Fig. 2) shows the excellent separation of all tested preservatives under the described conditions.

10

0,20

0,10

benzophenone

butylparaben

ethylparaben

iso-propylparaben propylparaben IPBC

0,30

MDBGN

0,40

bronopol

AU

0,50

MCI (Kathon)

0,60

DMDM degradation product DMDM

0,70

methylparaben

phenoxyethanol

0,80

0,00 5,00

10,00

15,00

20,00

25,00

30,00

35,00

40,00

Minutes

Fig. 2. Separation of preservatives at 210 nm by HPLC method 1.

The identities of the preservatives in the prepared test samples were verified by comparing the retention time and the UV spectrum for each detected preservative with the retention time and UV spectrum in the chromatogram of the standard solution. It has been shown [35, 36] that Bronopol (BNPD) slowly degrades in water solutions; after storage at room temperature for 24 h, its concentration decreases by about 20%. Degradation of DMDM hydantoin was also observed in the standard solution and in the studied skin creams during the analysis. The peaks of the degradation product of Bronopol (BNPD) and the degradation product of DMDM hydantoin have approximately the same retention times and this may disturb the quantification of the analytes. The following three separate standard solutions were therefore prepared in 50% aqueous methanol and used for the screening: • Standard solution 1: Parabens (0.4 mg/ml of each paraben). • Standard solution 2: Bronopol (BNPD) (0.1 mg/ml); MDBGN (0.1 mg/ml); IPBC (0.05 mg/ml); Kathon (0.0015 mg/ml). • Standard solution 3: DMDM hydantoin (0.6 mg/ml). The standard solutions were kept at -20°C. The concentrations of the preservatives in the standard solutions correspond to the maximum allowed concentrations of the preservatives in the cosmetic products (see Table 1). The response of the parabens detected at 256 nm (absorbance maximum) was outside the linear range of the detector at the concentration of the standard solution of parabens (Standard solution 1). Quantification of the parabens was therefore performed at 290 nm. A chromatogram of a sample extracted from a commercial cream at 290 nm is shown in Fig. 3. The parabens in the mixture were identified by comparing the retention time and the UV spectrum (Fig. 4) for each detected preservative with the equivalent data in the standard solution chromatogram.

11

Me-paraben

0,110 0,100 0,090 0,080 0,070 0,060

0,040 0,030 0,020

Pr-paraben

Et-paraben

0,050

0,010

Bu-paraben

AU

0,000 0,00

5,00

10,00

15,00

20,00

25,00

30,00

35,00

40,00

45,00

50,00

Minutes

Fig. 3. Chromatogram of extract from a commercial cream containing a mixture of parabens at 290 nm, HPLC method 1. Me-paraben 200,00

nm 250,00

300,00

Et-paraben 200,00 350,00

nm 250,00

25,60 256,0

300,00

iPr-paraben 200,00 350,00

nm 250,00

300,00

nm 250,00

256,0

300,00

350,00

37,71

33,98

29,82 256,0

Bu-paraben 200,00 350,00

256,0

336,0

Fig. 4. UV spectra of parabens at 200 to 350 nm.

Quantification of DMDM hydantoin was performed by using the combined peak areas of the preservative and the degradation product (Fig. 2). The quantification of Bronopol (BNPD) was handled in the same way (Fig 5). Its degradation product was identified by using the retention time and UV spectrum (Fig. 6). Kathon was quantified using only the peak for methylchloroisothiazolinone (MCI in Fig. 2).

12

Bronopol Bronopol degradation product

0,025

0,020

AU

0,015

0,010

0,005

0,000

-0,005 0,00

5,00

10,00

15,00

20,00

25,00

30,00

35,00

40,00

45,00

50,00

Minutes

Fig. 5. Chromatogram of extract from a commercial cream containing Bronopol (BNPD) at 250 nm, HPLC method 1.

Bronopol degradation product

Bronopol 200,00 220,00

240,00

260,00

nm 280,00

300,00

320,00

200,00 340,00

220,00

240,00

260,00

nm 280,00

300,00

320,00

6,22 201,0

340,00

7,12

Fig. 6. UV spectra of Bronopol and Bronopol degradation product at 200 to 350 nm.

All analytes were eluted from the column within 40 minutes with a linear gradient of 0 to 50% acetonitrile. A step gradient to 89% acetonitrile was then used to wash out hydrophobic substances in the creams. The duration of this wash step was 10 minutes and the column was then equilibrated for 16 minutes before the next injection. This washing and equilibrating procedure was found to be necessary to obtain a stable retention time for Bronopol (BNPD), the first eluting analyte (see Fig. 5). Other ingredients in the creams (e.g. water, fat, odours, coloured and surface active agents) were not identified or quantified. Due to the washing and equilibrating procedure described above, they did not disturb the identification and quantification of the analytes.

13

3.3. HPLC method 2: Determination of diazolidinyl urea, imidazolidinyl urea and Quaternium-15 Hydrophilic interaction liquid chromatography (HILIC) was applied for the determination and quantification of diazolidinyl urea, imidazolidinyl urea and Quaternium-15 in all skin creams with positive results from the Schiff’s test. Analyses were performed on a ZIC-HILIC column, 100x2.1 mm, 5 µm with a 0.1% acetic acid in water/0.1% acetic acid in acetonitrile gradient. Diazolidinyl urea and imidazolidinyl urea were characterized as a mixture of several compounds [23]. In the previous study, the specific finger-print was used to detect and quantify these preservatives [6]. During this study, it was observed that consecutive injections of the same solution gave different patterns of peaks (Fig. 7). After 2 hours, the degradation of diazolidinyl urea had reached a steady state and the pattern of peaks was consistent. The solutions were then stable throughout the study. To reach equilibrium in the mixture, the calibration curve solutions (see 3.5 Method validation) and standard solutions containing diazolidinyl urea were therefore kept at room temperature for about 2 hours prior to analysis. The standard solutions were kept in HPLC vials at -20°C following the full preparation procedure. 1,00

0,90

0,80

0,70

AU

0,60

0,50

Diazolidinyl urea, afrer 2 hours

0,40

0,30

0,20

Diazolidinyl urea, initial chromatogr

0,10 13,00

14,00

15,00

16,00

17,00

18,00 19,00 Minutes

20,00

21,00

22,00

23,00

24,00

Fig. 7. HPLC chromatograms of diazolidinyl urea at 201 nm, HPLC method 2.

The following three standard solutions were prepared in 50% aqueous ethanol and used for the screening: • Standard solution 1: diazolidinyl urea (0.5 mg/ml). • Standard solution 2: imidazolidinyl urea (0.6 mg/ml). • Standard solution 3: Quaternium-15 (0.2 mg/ml). Chromatograms of the diazolidinyl urea, imidazolidinyl urea and Quaternium-15 standard solutions are presented in Fig. 8.

14

0,60

0,55

Quaternium-15 0,50

0,45

0,40

AU

0,35

0,30

Diazolidinyl urea

0,25

0,20

0,15

0,10

Imidazolidinyl urea

0,05

13,00

14,00

15,00

16,00

17,00

18,00 19,00 Minutes

20,00

21,00

22,00

23,00

24,00

Fig. 8. HPLC chromatograms of diazolidinyl urea and imidazolidinyl urea at 201 nm and Quaternium-15 at 210 nm, HPLC method 2.

Since diazolidinyl and imidazolidinyl urea exist in their degraded forms in skin creams, the following procedure was used to determine those preservatives. The total area for the major peaks in the diazolidinyl urea chromatogram where the retention times and UV spectra corresponded to those of the standard solution was used to identify and quantify the analyte in the commercial products (group of major peaks with retention times 17 to 22 minutes, see Fig. 8). The same procedure was applied to identify and quantify imidazolidinyl urea (group of major peaks with retention times 16 to 22 minutes, see Fig. 8). The appearance of the specific pattern of peaks for diazolidinyl or imidazolidinyl urea in the chromatogram was mandatory for identification since the UV spectra of these analytes are identical (see Fig. 9 and Fig. 10). _ 200,00 220,00

240,00

,

nm 260,00 280,00

300,00

320,00

340,00

Diazolidinyl urea

Fig. 9. UV spectrum of diazolidinyl urea at 200 to 350 nm.

15

_ 200,00 220,00

240,00

260,00

nm 280,00

300,00

320,00

340,00

Imidazolidinyl urea

Fig. 10. UV spectrum of imidazolidinyl urea at 200 to 350 nm.

Diazolidinyl urea, imidazolidinyl urea and Quaternium-15 were eluted from the column between 15 and 22 minutes with a gradient from 93 to 60 % acetonitrile. Other constituents of the creams’ formulations were eluted in the beginning of the chromatogram. An equilibrating step was necessary before the next injection to reach a stable baseline at the start of analysis. For determining lower concentrations of Quaternium-15, this method needs to be improved. Its limit of detection is currently 0.2 mg/ml, which corresponds to the maximum allowed concentration in skin-care products (0.2%).

3.4. HPLC method 3: Determination of formaldehyde The determination of formaldehyde in the presence of formaldehyde-releasing preservatives (diazolidinyl urea, imidazolidinyl urea, Quaternium-15, BNPD and DMDM hydantoin) was performed by hydrophilic interaction liquid chromatography (HILIC) with post-column derivatization according to the procedure based on the official EEC method [10]. Free formaldehyde is reacted with pentane-2,4-dione in the presence of ammonium acetate to form fluorescent 3,5-diacetyl-1,4-dihydroluitidine. The derivative formed was detected at 420 nm. The selected HPLC gradient system consisted of a ZIC-HILIC, 100x2.1 mm, 5 µm column and a 2-propanol/water mobile phase. This method was used to analyse all skin creams with positive results from the Schiff’s test, and can be applied as a more selective procedure for determining formaldehyde. The release of formaldehyde on the decomposition or degradation of formaldehyde-releasing preservatives (diazolidinyl urea, imidazolidinyl urea, Quaternium-15, BNPD and DMDM hydantoin) depends on many factors such as temperature, pH, light, etc. There is a risk that the concentration of formaldehyde in prepared sample solutions will increase with time. During screening of the skin creams for formaldehyde, the contents of the prepared samples were analysed within 20 hours of sample preparation. If the concentration of formaldehyde in the sample was found to be higher than the maximum allowed concentration in the cream, the cream was reanalysed using freshly prepared samples. Formaldehyde elutes at 5.8 minutes and a washing step was necessary to elute all other cream ingredients from the column before injecting the next sample.

16

3.5. Method validation Linear calibration curves were obtained for each analyte in the range studied; see Table 3. Calibration curves were constructed from five calibration curve solutions for each analyte. Solutions were prepared in 50% aqueous methanol for the following analytes and concentration ranges: 0.04 mg/ml to 0.8 mg/ml for parabens; 0.01 mg/ml to 0.2 mg/ml for BNPD; 0.005 mg/ml to 0.1 mg/ml for IPBC; 0.01 mg/ml to 0.2 mg/ml for MDBGN; 0.00015 mg/ml to 0.003 mg/ml for Kathon and 0.06 mg/ml to 1.2 mg/ml for DMDM hydantoin. Solutions were prepared in 50% aqueous ethanol for the following analytes and concentration ranges: 0.05 mg/ml to 1.0 mg/ml for diazolidinyl urea; 0.06 mg/ml to 1.2 mg/ml for imidazolidinyl urea; 0.02 mg/ml to 0.4 mg/ml for Quaternium-15 and 0.02 mg/ml to 0.4 mg/ml for formaldehyde. The range of the calibration curves corresponded to 10 to 200% of the maximum allowed concentration of the preservatives in the skin creams. The reporting limits of the preservatives were set at 1/10 of the highest allowed concentrations (see Table 1) except for Quaternium-15, where the limit was set to the maximum allowed concentration (since it was not possible to detect lower concentrations with this screening procedure). The main issue of interest was to find a suitable procedure to assess whether or not the concentrations of preservatives in the skin-care products are higher than the maximum authorized levels. Hence, only the possibility to detect the preservatives at concentrations close to and higher than the maximum allowed was mandatory during the development of the screening procedure. The repeatability of the methods was determined by consistent injections of the solutions at two different concentrations: the maximum allowed for the different preservatives (see Table 1) and the lowest point of the calibration curves, corresponding to the reporting limits of the preservatives. Results are presented in Table 3. The reproducibility of the HPLC methods was evaluated during the study of the commercial skin creams. All selected skin creams were analysed according to the screening procedure. Products where the detected preservatives were not declared in the labelling or products found to contain higher-than-allowed concentrations of preservatives were re-analysed. Samples for re-analysis were prepared in duplicate. Results from the initial screening analyses were compared with the results from the re-analyses. (Samples for the screening and re-analysis were prepared and analysed on different days). Of the 99 creams studied, 15 creams were re-analysed and the relative standard deviation (RSD) calculated from the analyte peak areas was < 1% (n=3). Ingredients of the skin-cream formulations such as fat and various active agents were eluted from the chromatographic systems by careful washing and equilibration during all HPLC analyses. The absence of interfering peaks was ascertained by running blank samples (50% aqueous methanol or 50% aqueous ethanol). No carry-over effect was observed in the chromatographic systems after injection of calibration curve solutions containing the highest concentrations. Adequate selectivity of the HPLC methods was observed during the development of the HPLC methods and the study of commercial skin creams. The ability to accurately and specifically measure the analyte in the presence of other components that may be included in the sample matrix has been well demonstrated in Figures 2, 3, 5 and 8.

17

Table 3. Linear range, correlation coefficient and repeatability. RSD (%) maximum allowed Correlation Linear range, concentration (see Preservative coefficient (mg/ml) Table 1) (R2) Area RT (n=6) (n=6) Methylparaben 0.04-0.8 0.9998 0.01 0.07 Ethylparaben 0.04-0.8 0.9998 0.02 0.06 Propylparaben 0.04-0.8 0.9997 0.02 0.04 iso0.04-0.8 0.9998 0.02 0.07 Propylparaben Butylparaben 0.04-0.8 0.9997 0.02 0.07 DMDM hydantoin 0.06-1.2 0.9997 0.09 0.44 BNPD 0.01-0.2 0.9984 0.08 0.50 IPBC 0.005-0.1 0.9999 0.03 0.47 MDBGN 0.01-0.2 0.9999 0.05 0.20 Kathon 0.000150.9998 0.26 0.45 0.003 Diazolidinyl urea 0.05-1.0 0.9992 0.09 1.03 Imidazolidinyl 0.06-1.2 0.9992 0.10 1.11 urea Quaternium-15* 0.2-0.4 0.9755 0.07 0.90 Formaldehyde 0.02-0.4 0.9987 0.12 1.56

LOQ Area (n=3) 1.1 1.2 1.2 1.1 0.5 0.5 4.2 9.4 3.3 13.1 2.0 5.1 0.6

RT = retention time RSD = relative standard deviation LOQ = reporting limit (1/10 of the maximum allowed concentration for each preservative) *= reporting limit and maximum allowed concentration were the same

Recoveries in the extraction procedures for the preservatives from the skin creams were estimated from literature data to be between 80 and 90%. Our data indicates that recoveries down to 60% might be obtained for some products, depending on the composition of the skin cream. This means that the concentrations measured might be underestimated, but if so, underestimated in favour of the producer.

3.6. Application The developed screening procedure was initially successfully utilized for routine quality control analysis of commercial batches of 99 different skin creams. The combination of Schiff’s test and HPLC methods helps avoid numerous unnecessary analyses. In this study, 56 creams of 99 tested showed a positive result for aldehyde content with Schiff’s test. All these 56 products were analysed with HPLC method 2 to determine diazolidinyl urea, imidazolidinyl urea and Quaternium-15, and with HPLC method 3 for quantifying formaldehyde. Analysis by HPLC methods 2 and 3 was unnecessary for 46 creams with a negative Schiff’s test result. However, two samples whose labelling claimed

18

formaldehyde releaser did not produce a positive reaction with Schiff’s test. When included in HPLC analysis methods 2 and 3, neither showed any evidence of formaldehyde releaser or formaldehyde.

Table 4. Skin cream analysis results. Number of skin creams with preservative content content content higher Preservative not declared according according than maximum but present to labelling to analysis allowed Parabens 65 73 8 2 IPBC 3 3 MBGN Kathon 3 1 Bronopol (BNPD) 3 3 Diazolidinyl urea 10 10 1 Imidazolidinyl urea 4 5 1 DMDM hydantoin 1 1 Quaternium-15 1 a) Formaldehyde 18 20 1 4 a)

skin creams labelled to contain the formaldehyde releaser.

Study results are briefly presented in Table 4. For 12 products, the ingredient labelling was incomplete and 7 skin creams contained a preservative above the allowed concentration. Skin creams that gave concentrations of preservatives close to or above the maximum allowed were reanalysed. The same strategy was applied to products where preservatives not declared on the labelling were found. The screening procedure is now used in the Laboratory of the Swedish Medical Products Agency to routinely analyse 10 commercial skin creams per month. With a few samples, it was not necessary to perform Schiff’s test as the gain in time this achieved was quite small. Results of the sample analyses for 2009 are presented by the Department of Cosmetics of the Swedish Medical Products Agency on the homepage www.lakemedelsverket.se in a separate report. This screening procedure makes it feasible to obtain results for all preservatives in a short period of time. Only one sample preparation needs to be performed since the choice of solvent does not affect the results and the samples can thus be analysed by all three HPLC methods in parallel. This makes it possible for an approved authority to take measures against products while the tested batches are still on the market.

19

4. Conclusions The developed screening procedure for the identification and quantification of commonlyused preservatives in skin creams can be advantageously used in the control of skin-care products. The procedure permits the detection of preservatives belonging to different classes of compounds using one simple sample preparation and standard HPLC equipment.

5. Acknowledgements The authors would like to thank Ann-Marie Hesselgren, Monika Tydén and Kerstin Rönnqvist for valuable technical assistance.

6. References 1.

2.

3.

4. 5.

6. 7.

8. 9.

10. 11.

De Groot A C, Flyvholm M, Lensen G, Menné T, Coenraads P. Formaldehydereleasers: relationship to formaldehyde contact allergy. Contact allergy to formaldehyde and inventory of formaldehyde-releasers. Contact Dermatitis, 2009: 61: 63-85. Dyrgaard Lundov M, Moesby L, Zachariae C, Duus Johansen J. Contamination versus preservation of cosmetics: a review on legislation, usage, infections, and contact allergy. Contact Dermatitis, 2009: 60: 70-78. Cashman A L, Warshaw E M, Parabens: A Review of Epidemiology, Structure, Allergenicity, and Hormonal Properties. Dermatitis (formerly American Journal of Contact Dermatitis), 2005 Vol 16: 2: 57-66. Aalto-Korte K, Kuuliala O, Suuronen K, Alanko K. Occupational contact allergy to formaldehyde and formaldehyde releasers. Contact Dermatitis, 2008, 59:5: 280-289. Council Directive 76/768/EEC of 27th July on the approximation of the laws of Member States relating to cosmetic products, Official journal of the European Communities, OJ L 262, 27.9.1976, p.169. Gruvenberger B, Bruze M, Tammela M. Preservatives in Moisturizers on the Swedish Market. Acta Derm. Venereol., 1998: 78: 52-56. Rastogi S C, Jensen G H, Petersen M R, Worsøe I M, Christoffersen C. Preservatives in skin creams. Analytical Chemical Control of Chemical Substances and Chemical Preparations. Ministry of Environment and Energy. National Environmental Research Institute, Denmark. Technical Report No. 297. 1999. Rastogi S C. Analytical control of preservatives labelling on skin creams. Contact Dermatitis, 2000: 43: 339-343. Second Commissions Directive 82/434/EEC of 14 July 1982 on the approximation of the laws of the Member states relating to methods of analysis necessary for checking the composition of cosmetic products. Seventh Commissions Directive 96/45/EC of 2 July 1996 relating to methods of analysis necessary for checking the compounds of cosmetic products. Wilkinson J D Shaw S, Andersen K E, Brandao F M, Bruynzeel D P, Bruze M, Camarasa J M G, Diepgen T L, Ducombs G, Forsch P J, Goossens A, Lachappelle J-M, Lahti A, Menné T, Seidenari S, Tosti A, Wahlberg J E. Monitoring levels of preservative sensitivity in Europe. A 10-year overview (1991-2000). Contact Dermatitis, 2002: 46: 207-210.

20

12.

13.

14.

15.

16.

17.

18.

19. 20.

21.

22.

23.

24.

25.

26. 27.

Saad B, Bari F, Idiris Saleh M, Ahmad K, Talib M K M. Simultaneous determination of preservatives (benzoic acid, sorbic acid, methylparaben and propylpraben) in foodstuffs using high-performance liquid chromatography. J. Chromatogr. A, 2005: 1073: 393397. Hájková R, Solich P, Pospišilová, Šicha J. Simultaneous determination of methylpraben, propylparaben, sodium diclofenac and its degradation product in a topical emulgel by reversed-phase liquid chromatography. Anal. Chim. Acta, 2002: 467: 91-96. Zhang Q, Lian M, Liu L, Cui H. High-performance liquid chromatography assay of parabens in wash-off cosmetic products and foods using chemiluminescence detection. Anal. Chim. Acta, 2005: 537: 31-39. Lee M-R, Lin C-Y, Li Z-G, Tsai T-F. Simultaneous analysis of antioxidants and preservatives in cosmetics by supercritical fluid extraction combined with liquid chromatography-mass spectrometry. J. Chromatogr. A, 2006: 112: 244-251. Pedrouzo M, Borrull F, Marcé R M, Pocurull E. Ultra-high-performance liquid chromatography-tandem mass spectrometry for determining the presence of eleven personal care products in surface and wastewaters. J. Chromatogr. A, 2009: 1216: 69947000. Marengo E, Gianotti V, Angioi S, Gennaro M C. Optimization by experimental design and artificial neural networks of the ion-interaction reversed-phase liquid chromatographic separation of twenty cosmetic preservatives. J. Chromatogr. A, 2004: 1029: 57-65. Memon N, Iqbal Bhanger M, Khuhawer M Y. Determination of preservatives in cosmetics and food samples by micellar liquid chromatography. J. Separation Sci., 2005: 28: 635-638. Tsai T-F, Lee M-R. Determination of Antioxidants and Preservatives in Cosmetics by SPME Combined with GC-MS. Chromatographia, 2005: 67: 425-431. García-Jiménez J F, Valencia M C, Capitán-Vallvey L F. Simultaneous determination of antioxidant, preservatives and sweetener additives in food and cosmetics by flow injection analysis coupled to a monolithic column. Anal. Chim. Acta, 2007: 594: 226233. Baalbaki B, Blanchin M-D, Fabre H. Validation of a micellar electrokinetic capillary chromatography method for the determination of imidurea, methyl and propylparabens in a pharmaceutical ointment. Anal. Chim. Acta, 2002: 463: 15-20. Williams R O III, Mahaguna V, Sriwongjanya M. Determination of diazolidinyl urea in a topical cream by high-performance liquid chromatography. J. Chromatogr. B, 1997: 696: 303-306. Vig Lehmann S, Hoech U, Breinholdt J, Olsen C E, Kreilgaard. Characterization and chemistry of imidazolidinyl urea and diazolidinyl urea. Contact Dermatitis, 2006: 54: 50-58. Doi T, Kajimura K, Takatori S, Fukui N, Taguchi S, Iwagami S. Simultaneous measurement of diazolidinyl urea, urea, and allantion in cosmetic samples by hydrophilic interaction chromatography. J. Chromatogr. B, 209: 877: 1005-1010. Summers W R, Characterization of Azoniaadamantande-Based Preservatives by Combined Reversed-Phase Cation-Exchange Chromatography with Suppressed Conductivity Detection. Anal. Chem., 1992: 64: 1096-1099. Ottersonn R J, Howerton L M. Determining DMDM Hydantoin in Personal Cleansing Products. Allured’s Cosmetics & Toiletries magazine, 1999: 10: 55-61. Harunobu N, Ichiro M, Naoko M, Mikiya K. Determination of Organoidodine Antimicrobial Ingredients in Commercially Available Antimicrobial/Deodorant Agents. J. Health Sci., 2000: 6: 459-466.

21

28.

29.

30. 31. 32.

33. 34.

35. 36.

37.

38.

39.

40. 41.

42.

43.

44.

45.

Rastogi S C, Sys S, Johansen. Comparison of high-performance liquid chromatographic methods for the determination of 1,2-dibromo-2,4-dicyanobutane in cosmetic products. J. Chromatogr. A, 14995: 692: 53-57. Rastogi S C, Zachariae C, Johansen J D, Devantier C, Menné T. Determination of methyldibromoglutaronitrile in cosmetic products by high-performance liquid chromatography with electrochemical detection. Method validation. J. Chromatogr. A, 2004: 1031: 325-317. Bettero A, Galiano F, Daolio S, Benassi C A. The characterization of isothiazolinone preservatives in cosmetics. J. Pharm. Biomed. Anal., 1985: 6: 581-587. Gruvberger B, Persson K, Björkner B, Bruze M, Dahlquist I, Fregert S. Demonstration of Kathon® GC in some commercial products. Contact Dermatitis, 1986: 15: 24-27. De Kruijf N, Schouten A, Rijk M A H, Pranoto-Soetardhi L A. Determination of preservatives in cosmetic products. High-performance liquid chromatographic identification. J. Chromatogr., 1989: 469: 317-328. Rastogi S C. Kathon GC and cosmetic products. Contact Dermatitis, 1990: 22: 155-160. Sible V S. Quantitative analysis of a biocide in silicone emulsions using high performance liquid chromatography. J. Liquid Chromatogr. & related technologies, 1996: 9: 1353-1367. Wang H, Provan G J, Helliwell K. Determination of bronopol and its degradation products by HPLC. J. Pharm. Biomed. Anal., 2002: 29: 387-392. Lian H Z, Zhang W B, Miao J, Jiang Q, Mao L, Zong L. Li L, Wu X X, Cheng R M. A study on the stability of bronopol in bronopol lotion by ion-paired reversed-phase high performance liquid chromatography J. Pharm. Biomed. Anal., 1997: 15: 667-671. Kajimura K, Tamagi T, Yamamoto T, Iwagami S. The Release of Formaldehyde upon Decomposition of 2-bromo-2-nitropan-1,3-diol (Bronopol). J. Health Sci., 2008: 4: 488492. Wu T, Wang C, Wang X, Ma Q. Simultaneous determination of 21 preservatives in cosmetics by ultra performance liquid chromatography. Int. J. Cosmetic Sci., 2008: 30: 367-372. Karlberg A-T, Skare L, Lindberg I, Nyhammar E. A method for quantification of formaldehyde in the presence of formaldehyde donors in skin-care products. Contact Dermatitis, 1998: 38: 20-28. Benassi C A, Semenzato A, Bettero A. High-performance liquid chromatographic determination of free formaldehyde in cosmetics. J. Chromatogr., 1989:464: 387-393. Zurek G, Karst U. Liquid chromatography – mass spectrometry method for the determination of aldehydes derivatized by the Hantzsch reaction. J. Chromatogr. A, 1999: 864: 191-197. Summers W R. Characterization of Formaldehyde and Formaldehyde-Releasing Preservatives by Combined Reversed-Phase Cation-Exchange High-Performance Liquid Chromatography with Postcolumn Derivatizaiton Using Nash’s Reagent. Anal. Chem., 1990: 62: 1397-1402. Reche F, Garrigós M C, Sánchez A, Jiménez A. Simultaneous supercritical fluid derivatization and extraction of formaldehyde by the Hantzsch reaction. J. Chromatogr. A, 2000: 896: 51-59. Michels J J. Improved measurement of formaldehyde in water-soluble polymers by high-performance liquid chromatography coupled with post-column reaction detection. J. Chromatogr. A, 2001: 914: 123-129. Vander Heyden Y, Nguyen Minh Nguyet A, Detaevernerier M R, Massart D L, PlaizierVercammen J. Simultaneous determination of ketoconazole and formaldehyde in a shampoo: liquid chromatography method development and validation. J. Chromatogr. A, 2002: 958: 191-201.

22

46.

47.

48.

49.

50.

51. 52.

Rivero R T, Topiwala V. Quantitative determination of formaldehyde in cosmetics using combined headspace-solid-phase microextraction-gas chromatography. J. Chromatogr. A, 2004: 1029: 217-222. Emeis D, Anker W, Wittern K-P. Quantitative 13C NMR Spectroscopic Studies on the Equilibrium of Formaldehyde with Its Releasing Cosmetic Preservatives. Anal. Chem., 2007: 5: 2096-2100. Sottofattori E, Anzaldi M, Balbi A, Tonello G. Simultaneous HPLC determination of multiple components in a commercial cosmetic cream. J. Pharm. Biomed. Anal., 1998: 18: 213-217. De Kruijf N, Schouten M A, Rijk A H, Pranoto-Soetardhi L A. Determination of preservatives in cosmetic products. High-performance liquid chromatographic identification. J. Chromatogr., 1989: 469, 317-328. Marengo E, Gennaro M C, Gianotti V. A Simplex-Optimized Chromatographic Separation of Fourteen Cosmetic Preservatives: Analysis of Commercial Products. J. Chromatogr. Sci., 2001: 39: 339-344. “The quantitative analysis of drugs”. D.C. Garratt, third edition, 1964, p.250. Report No. 6, Medical Products Agency, 1998, ISSN 1401-5463, ISBN 91-97111-8-9.

23