3 Discussion

Development of an analysis method for quantification of colophonium components in cosmetic products

A new method for detection and quantification of the major components, AbA and DeA, in unmodified colophonium (rosin) was developed. The method also includes the possibility of detection and quantification of one of the major oxidation products, 7-O-DeA, obtained at the oxidative degradation of colophonium in contact with air. The final method involves an HPLC method with UV-DAD able to separate the resin acids. The method is fast, robust, simple, without using toxic chemicals and has a high specificity compared to ordinary UV methods. High specificity is needed due to the complexity of the cosmetics to be analyzed. This is obtained by a DAD which allows specificity without using mass spectrometry. Compared to GC methods used no derivatisation is needed.

According to the EU legislation on dangerous substances – Annex 1 (Directive 67/548 /EEC) a content of >1% colophonium in a product must be declared and the product must be labeled with risk phrase R 43 (“May cause sensitisation by skin contact”). However colophonium should always be listed on the list of ingredients if it is used in a cosmetic product according to the Cosmetic Directive (76/768/EØF) .

However, colophonium is a mixture of many compounds. Quantifying a mixture used as an ingredient of a product cannot be achieved by any analytical means. This applies to any natural extracts used as ingredient of a compounded consumer product. To overcome the impossibility of quantifying a complex ingredient in a product, the quantification of tracers (= defined substances) specific of the complex natural sources is feasible. Therefore, quantification must be based on chemical analyses of specific major compounds. For colophonium we have chosen to base our quantification on its two major resin acids. However, it is not possible to give the exact content of colophonium in a compounded product based on analysis of the acids, since there is a difference in the acid content in different types of colophonium due to variations in extraction, handling, storage and manufacturing. This is especially true for the content of DeA. For AbA the content can vary between 30-50% for gum rosin and 35-40 % for tall oil rosin. For DeA the figures can be 5-10% in gum rosin and around 30 % in tall oil rosin. Furthermore, at the wavelength used for detection of DeA (220 nm) the risk for interference with constituents from the sample matrixes is high. Thus, the detected amount of AbA should be used for quantification of colophonium in most cases.

A low level of AbA can indicate a low level of colophonium. However, it is important to check that the low level is not due to an oxidative decomposition of AbA resulting in highly allergenic compounds. To avoid such a misinterpretation, we also analyzed for DeA, a much more stable compound than AbA, and for 7-O-DeA as a tracer for oxidation compounds of AbA. High values of these tracers together with a low level of AbA indicate that AbA has decomposed to a great extent and that the content of colophonium is higher than estimated based on the AbA analysis and further that the presence of allergenic oxidation product might be high.

In general, two major types of analyses for colophonium components are described in literature. Traditionally, GC-FID methods with derivatisation of the resin acids to allow a good separation have been used e. g. in the pulp industry. Later on, LC methods with UV, fluorescence, and mass spectroscopic methods for detection have been developed. Most published methods have been developed for the investigation of specific products that might contain colophonium (See Appendix A). The developed method is not restricted to a specific group of consumer products in. e. cosmetics. Rather, it can be used universally. However, there could be some limitations of the methods of sample preparations for certain products, where the components are stronger bound in the material compared to cosmetic products..

In the present investigation the SPE phase used for sample preparation is a mixture of lipophilic and weak anion exchanging sites, which seems to suite the resin acids perfectly. Other lipophilic compounds can be washed out from the column by using methanol, while our compounds are retained until acid is added. The method used gives very clean extracts according to the LC analyses.

Samples from different products were analyzed. Only one wax strip for hair removal had significant concentrations of colophonium. For the other products (lip gloss, foundation, wax strips for face) there were traces of AbA and DeA present, however, below detection limits. According to the declaration of the products only the two types of wax strips were declared containing colophonium.