Degradation studies of synthetic food colouring matters

The heat and light stabilities of the 16 UK-permitted synthetic food colouring matters were investigated in aqueous solution. A specially constructed light box was used to provide exaggerated lighting conditions with fan-cooling of the samples. Comparison of stability has been made on the basis of the shape of the entire decay curve rather than by simple comparison of loss after a fixed degradation period. Indigo carmine and erythrosine showed little light stability, while ponceau 4R, chocolate brown HT and brown FK all showed fairly mediocre stability. The thermal stabilities of chocolate brown HT, ponceau 4R and, most especially, indigo carmine, all proved poor. A new polarographic peak which appeared during the thermal degradation of red 2G was shown to be due to the deacetylation product, red lOB. Identification was made by polarographic and chromatographic (hplc) comparison with a red lOB standard. A method for the analysis of red 2G and red lOB present in the same solution was developed using differential pulse polarography. This method included the use of tetrapheny1phosphonium chloride to produce a differential shift 1n the E 1/2 values of the respective colouring matters, and so increase the resolution of their polarographic peaks. Hplc analysis was conducted using reversed-phase ion-pair chromatography. A temperature control system was devised for the hplc column (water jacket) which was shown to improve the resolution and reproducibility of retention of the chromatographic peaks. An eluent comprising 63/37/0.25 v/v/w methano1/water/cetrimide, an SAS-Hypersil packing and thermos tatting at 40°C were found to be suitable for most of the thermal degradation solutions. A series of 13 chromatograms of thermally degraded solutions were produced under these conditions for comparative purposes. An unexpectedly large number of decay product chromatographic peaks were observed for some food colours. Whereas thermally degraded ponceau 4R only showed four peaks, most colours showed up to about ten peaks, and a few, e.g. amaranth, green S and sunset yellow FCF showed a dozen or more. Tentative identifications were made of some products from azo colours by reference to standard phenols and aromatic amines. Growth curves were plotted by hplc for a number of the products from the thermal degradation of yellow 2G. Polarographic and/or chromatographic evidence clearly showed that the thermal degradations of yellow 2G and of red 2G were not single step, single pathway processes. The large number of products observed by hplc indicate that this is true of most of the colours. Moreover, thermal degradation pathways need not be the same as photodegradation pathways. This has been shown simply for ponceau 4R by comparison of UV spectra. Qualitative analysis linked to hplc was also investigated. The construction of UV spectra of decay products by repeated chromatography at different detection wavelengths proved to be quite straightforward. It was found that the use of diazotisation as a pre-column reaction for the differentiation of aromatic amines from phenolic analogues was not suitable: the acidic and strongly ionic conditions employed for diazotisation disrupted the chromatographic ion-pair mechanism beyond correction. Use of fluorescamine proved rapid and simple for sulphanilic acid, but no significant reaction occurred in the aase of naphthionic acid. A method was devised for the analysis of phenol-4-sulphonic acid and sulphanilic acid when chromatographically unresolved by comparison of peak absorbance at two wavelengths. Finally, preliminary work was carried out to find how the dependence of retention on eluent composition was affected by the presence of specific functional groups in the analyte. This was seen as of particular application where functional group (e.g. carboxylate) analysis was required for decay products present in low concentration.