Optimization of bitterness in chocolate through roasting with analysis of related changes in important bitter compounds
Chocolate is made from the fermented, dried, and roasted seeds of the Theobroma cacao tree, an important agricultural food crop which contains bioactive flavonoid polyphenols with beneficial health effects. Such effects include improvement of antioxidant status, positive impacts on cardiovascular health and endocrine system function, association with cancer prevention, LDL cholesterol reduction, and reduction of obesity and related conditions. However, products which have the highest levels of cacao flavonoids of all eating-chocolate, such as high-cacao-percentage dark chocolate, are known to be quite bitter, a taste modality that is not readily appreciated by humans. Though the complex causes of bitterness in cacao are still not completely understood, it has long been known that the methylxanthines theobromine and caffeine impart bitterness, as do certain flavan-3-ols, sometimes called catechins, which are a class of the aforementioned healthy bioactive polyphenolic flavonoids, also found in tea. Yet, what else is known of bitterness in cacao is sparse and even contradictory. Work on cacao bitterness has described the importance of cyclic dipeptides called 2,5-diketopiperazines (DKPs), while suggesting some form of interaction between theobromine and DKPs as well. Yet these earlier assertions have only been confirmed with mixed results by others, in part due to the incredible complexity of bitterness in roasted cacao, which has been said to require further sensory evaluation. More recent work on bitterness in cacao suggested for the first time that a DKP called cyclo(Pro-Val) is the most important bitter compound. However, even while seeming to confirm the importance of previously known important bitter compound classes, this research was based upon only a single cacao sample from a single origin of cacao, and with an undefined roasting treatment, even though previous work had noted that differences in DKP formation are dependent upon roast profile. Additionally, sensory work was based in part on recombinants of bitter compounds in aqueous solution, allowing for potentially biased estimation of the contribution of the different compounds to finished chocolate bitterness, since the varying kinetics of dissolution of the diverse bitter compounds from low-moisture, high-fat cacao matrix into saliva were not considered, nor were interactions with aroma compounds present in chocolate. Therefore, much was still to be learned about the variation in bitter-compound composition in cacao and related sensory characteristics, within and between different cacao origins and across different roast profiles. This fact, combined with a growing desire for healthy, functional versions of foods such as chocolate makes research into the impact of cacao roasting on consumer perceptions of bitterness and overall liking in chocolate, and the underlying chemical changes, all the more timely. This research project has resulted in findings covering a significant range of chocolate topics. First of all, a new efficient method for extraction and analysis of important bitter compounds in cacao and chocolate was developed. A custom response-surface methodology (RSM)-based design for the roasting treatments, with emphasis on I-optimality for minimizing prediction variance,was created. Chemical and sensory analysis of the roasted chocolate treatments were carried out, followed by in-depth data analysis and interpretation in the context of current chocolate science. Specifically, the aqueous 70% N,N-Dimethylformamide solvent system and HPLC method developed for fast and efficient extraction, followed by analysis, of important bitter compounds from three different chemical classes (i.e., methylxanthines, flavan-3-ols, and 2,5-diketopiperazines) simultaneously, functioned successfully, resulting in acceptable standard curves,% RSD values, and% recovery values. As for quantitative chemical findings, our work generally supports previous studies as regards changes in chemical concentrations during roasting. However, even with the large number of roasting treatments (i.e. 24, or 8 for each of 3 origins) across a reasonably large experimental region, we did not confirm the presence of concentrations of cyclo(Pro-Val) similar to that of previous research. As for sensory evaluation findings, we discovered that reduction of bitterness, sourness, and astringency are all correlated with increased liking in our chocolates. We also noted that consumers appear to have a preference for increased cocoa intensity. Roast profiles that minimize and maximize these characteristics respectively can vary by origin, but temperature and time combinations such as 20 minutes/171[degrees]C, 80 minutes/135[degrees]C, and 54 minutes/151[degrees]C were generally effective, whereas, raw and lightly roasted treatments (i.e., 0 minutes at 24[degrees]C, 11 minutes at 105[degrees]C, or 55 minutes at 64[degrees]C) were not, resulting in the lowest liking ratings. As with any complex food system, caveats do exist. Additions of sugar, salt, and other ingredients would likely introduce significant effects relevant to overall sensory characteristics and consumer liking, and intensity of various other aroma profiles not yet analyzed could do the same (e.g., floral, fruity, nutty). One additional sensory finding is that we can now say that perception of chocolate aroma is likely to play a large role in the perception of taste modalities (i.e., bitterness, sourness, sweetness), and astringency, as well as liking in chocolate. Finally, regarding the relationship of bitter chemical concentrations in the treatments, and consumer bitterness perception thereof, while the analysis is somewhat complicated by the stability of theobromine and caffeine during roasting, we can say that we have little evidence to suggest that theobromine concentration is strongly correlated to bitterness in chocolate. There is far more evidence that caffeine may play a role in the increase of bitterness in cacao, though the magnitude of its importance is not yet known, and to better understand the impact of both theobromine and caffeine, study of many more origins will be required. As for epicatechin and procyanidin B2, as already known, they are quite well correlated, and of all the chemicals we studied, they were, as a pair, the most correlated with changes in bitterness in our data across all treatments, including all three origins. Given that epicatechin has previously been shown to be a more important contributor to bitterness than higher molecular weight procyanidins (e.g., procyanidin B2), the overall importance of epicatechin could be the greatest of all the compounds that we studied. In contrast, catechin and cyclo(Pro-Val), do not appear to be particularly important for changes in bitterness. More specifically, we have found no evidence supporting the claim that the DKP cyclo(Pro-Val) is the most important bitter compound in cacao or chocolate. This does raise additional questions about the importance of diketopiperazines (DKPs) as a class as they relate to bitterness in chocolate.