Survey of Chemical Substances in Consumer Products, No. 86 2007 – Survey and risk assessment of chemical substances in deodorants

Survey and risk assessment of chemical substances in deodorants

3 Analysis

3.1 Materials
3.2 Analysis

3.1 Materials

The fragrance substance standards were obtained through various sources as described in an earlier report (11). Triclosan-standard was purchased from Sigma-Aldrich, Denmark. All other chemicals used were of pro analysi or HPLC quality.

3.2 Analysis

3.2.1 Sample preparation

3.2.1.1 Fragrance substances

Aerosol spray-products were opened as described before (11) to remove and measure the amount of propellant. A portion of the samples, without propellant, was transferred into vials for the analysis by gas chromatography-mass spectrometry (GC-MS).

Deostick and roll-on products were treated as follows: To approximately 2 g sample, weighed exactly in a dark bottle, a few boiling chips and 8 ml methanol were added and the flask was closed with a screw cap. The mixture was mixed gently and warmed thereafter at 60°C for 5 min to dissolve the fatty substances (heating of homogenous liquid products was not required). The sample solution/suspension was then cooled to room temperature (20°C). A 20 cm (length) x 1.8 cm (diameter) glass column was packed with wet silica gel (in methanol) to 7 cm. The cooled sample solution was quantitatively transferred on to the silica-gel column and that was eluted with 20 ml methanol. The first 5 ml of the eluate was discarded, and thereafter the eluate was collected in a 25 ml measuring flask. The flask was filled up to the mark with methanol. The fragrance extract thus obtained was transferred into GC vials and analysed within 24 hours. Duplicate analysis was performed on each sample.

3.2.1.2 Triclosan

To approximately 1 g homogenous sample weighed in a 100 ml dark bottle, 0.25 ml sulfuric acid (4M) and 10 ml methanol were added. The mixture was shaken for 15 min at 60°C and then filtered through Whatman No. 2 filter paper and collected in a 25 ml measuring flask. The measuring flask was filled with methanol up to the mark. For each sample, two extracts were made. The sample extracts were transferred into vials and analysed by high performance liquid chromatography (HPLC).

3.2.2 Analysis of fragrance substances

Analysis of fragrance substances in the sample extracts as well as in the undiluted samples was performed by GC-MS, as described before (12). Each sample as well as each calibration standard solution (2-100 ppm) were analysed in duplicate. Identification and determination was performed in selective ion mode (SIM). Repeatability of the analytical method was evaluated for 10 determinations of two mixtures of fragrance substances (10 ppm and 25 ppm of each substance except farnesol). Repeatability test of farnesol (10 ppm and 25 ppm) determination was performed separately, because this in itself is a mixture of three substances. Oakmoss/treemoss in the samples was identified by the presence of evernic acid ethyl ester. Recovery of the fragrance substances, except oakmoss/treemoss, was investigated by the analysis of two products, which were spiked with 10 ppm of each of the target fragrance substance. Only qualitative analysis of oakmoss/treemoss in the deodorants is performed.

The detection limit of each substance was ca. 1 ppm and the limit of quantification of the fragrance substances was ca. 2 ppm. Recovery of all fragrance substances was 80-115% and relative standard deviation (repeatability) of the method for all substances was 8-12%.

3.2.3 Analysis of triclosan

Each sample extract as well as all calibration solutions were analysed in duplicate by HPLC as described before (13). Analyses of several dilutions of triclosan solution (5-300 ppm) were performed to generate calibration curve of the substance. Repeatability of the analytical method was determined by the analysis of two solutions of triclosan (30 ppm and 120 ppm), and the recovery of triclosan in the products was investigated by the analysis of two products spiked to concentration levels 60 ppm and 120 ppm.

Identification of triclosan in the samples was performed by the comparison of HPLC-retention time and UV-spectrum of the HPLC-peak of standard triclosan with those of the samples analysed under the same conditions as the standard triclosan. The content of triclosan was determined by the use of calibration curve of the standard triclosan. The recovery of triclosan from the spiked samples was ca. 98%, and the relative standard deviation of the analytical method was less than 5%.