Résumé
Fine-quality dark chocolates (70% cocoa content) have a dark brown color [1] that is partially influenced by phenolic compounds [2]. However, some of those chocolates have a light brown color, which is a challenge for manufacturers that opens new marketing strategies.The aim of this work is to evaluate the phenolic profiles of dark chocolates having a black and brown color to reveal discriminating compounds. Sixteen fine chocolate samples, containing 70% cocoa, provided by Valrhona from thirteen different Theobroma cacao clones and years (2019 and 2020) were selected according to their dark black or light brown color. Non-targeted metabolomics approach based on UHPLC−HRMS/MS experiments, univariate and multivariate statistical analysis, as well as Feature-Based Molecular Networking analysis (FBMN) [3] was conducted. The analysis of differentially accumulated metabolites between brown and black chocolates showed an overaccumulation of 27 metabolites in black chocolates. Among them glycosylated flavanols and small glycosylated A-type procyanidins (dimers and trimers) were highly representative. On the other hand, 50 overaccumulated metabolites were found for brown chocolates. Among them, 27 larger B-type procyanidins, from trimers to nonamers, were present. C-glycosylated or oxydized dimers or trimers and dehydrodicatechins B also participated in the discrimination of the two sets of samples. Phenolic and color profiles were more related to genetic factors, i.e. the studied cocoa clones. However, environmental conditions may also have influenced the black chocolate color. This study provides new insights on the phenolic profiles of black and brown chocolates that may be useful to better understand the color variations of dark chocolates. Moreover, owing to their different phenolic profiles, these two types of chocolates may have different potential functional properties for human health.Keywords : Theobroma cacao; dark chocolates; metabolomics; Feature-Based Molecular Networking; phenolic compounds; discriminating compounds Figure 1: Chocolate samples (a), type of instrumental analysis (b), multivariate (c), univariate (d) statistical analysis, and FBMN analysis (e) to reveal discriminating compounds of black and brown chocolates1. Beegum P.P, S. et al. J. Texture Stud. 2022, 53, 870–8822. Żyżelewicz, D.et al. Eur. Food Res. Technol. 2014, 238, 549–5636.3. Nothias, L.-F.; et al. Nat. Methods 2020, 17, 905–908