Choc-a-bloc with flavour enhancers

All chocoholics yearn for the intense sensation of pleasure to be gained by placing a piece of milk chocolate in the mouth. Apparently, it is not only due to the guilt some feel by eating this luxurious food, but also due to the release of endorphins in the brain. These are natural hormones that generate feelings of pleasure and well-being and probably contribute to the "warm glow" reported by susceptible consumers.

There are other chemicals among the 300 or so in chocolate that might also stimulate cravings. Neurotransmitters like phenylethylamine, which can trigger the pleasure centres of the brain, and the amino acid tryptophan, which helps the brain to produce serotonin that can produce feelings of elation, are two candidates. Theobromine is a smooth muscle relaxant that acts in the lungs and caffeine is the well-known central nervous system stimulant that we know from coffee and tea.

However, some scientists argue that the amounts of these chemicals found in chocolate are far too low to have any noticeable effect. Many of them are found in other foods that do not have the same appeal as chocolate.

Natural chocolate has a bitter taste, the darker the more bitter. So, many producers add a range of flavour enhancers to modify the taste in pure chocolate and chocolate-based products to make them more palatable. These additives include maltol and ethyl maltol, which mask bitter-tasting compounds and heighten the creamy and rich tones. Vanillin, vanillic acid and ethyl vanillin all add that creamy, milky vanilla flavour. In fact, 75% of the market for vanillin as a flavour is taken up by the chocolate and ice cream industries.

For consumers, taste modification of chocolate is a good thing but it is not welcomed by food analysts because it adds to the number of compounds that need to be analysed. The various additives are structurally diverse and there are many published methods for their measurement in chocolate and chocolate-containing products.

For a panel of nine common additives (theobromine, maltol, ethyl maltol, vanillin, vanillic acid, ethyl vanillin, caffeine, (+)-catechin and (-)-epicatechin) HPLC with UV detection is the most widespread published method, although amperometric, electrochemical and mass spectrometric detection have also been employed. The methods come with an assortment of sample preparation methods that include methanolic extraction, defatting and solid-phase extraction.

Despite all of the available methods, none apply to the universal analysis of all nine additives. This failing has been addressed by two chemists from the RJ Reynolds Tobacco Company in Winston-Salem, NC, who undertook to develop a simple, inexpensive method that could be easily adopted in many labs. Charles Risner and Melissa Kiser chose HPLC since it has been employed successfully for the target analytes and is a common piece of lab kit.

The team used a C18 column with a gradient of methanol in aqueous acetic acid, which separated all of the analytes within 24 minutes at a flow rate of 500 µL/min. They would have preferred to use water as the mobile phase but it induced rapid phase collapse in the columns, resulting in loss of resolution within about 10 injections.

For a universal method, some compromises have to be made. Hence, UV detection was conducted at 273 nm, not the optimum wavelength for all of the analytes, but one which gave adequate responses for them all. Under these conditions, the detection and quantification limits were 0.03-0.11 and 0.10-0.35 µg/mL for standard solutions of the analytes, with linear measurement ranges up to 20 µg/mL.

The optimised method was tested on a NIST standard reference baking chocolate (SRM 2384) in the first reported single procedure to measure theobromine, catechin, caffeine and epicatechin in this standard without using extensive sample preparation or mass spectrometric detection. The contents agreed well with the certified values and recoveries were 99.6-104.3%.

Next, real chocolate products bought in a local supermarket were tested: dark and milk chocolate, chocolate syrup, chocolate milk, a cocoa mix and a chocolate candy bar. Extraction could not have been simpler. Portions of each product were vortex-mixed with water, the chocolate bars being melted first, and the filtered extracts were analysed.

Maltol and ethyl maltol were not detected but the remaining additives were found in most of the products at varying levels. Peaks attributed to vanillic acid and ethyl vanillic acid were also detected in chocolate milk and chocolate syrup, respectively. They were attributed to the oxidation of vanillin and ethyl vanillin even though the products were stored in a refrigerator as recommended.

The analysis of a set of four proprietary artificial cocoa/chocolate flavours presented some difficulties due to the high contents of some of the flavour enhancers. For instance, vanillin was 300-fold more concentrated than in the consumer products. The large HPLC peaks masked the areas in the chromatograms where catechin and epicatechin would appear. Changes in the concentration of the extract which might have helped only served to cause column overload and introduce sample carryover between injections.

Overall, however, especially for commercial chocolate products as opposed to concentrated flavour enhancers, the universal method proved simple to follow and gave good results. It could be adopted in most food safety and control labs since they generally hold HPLC systems with UV detectors.