Assessing cocoa aroma quality by multiple analytical approaches

• MS-fingerprinting appeared to be a reliable and rapid classification technique based on the aroma properties.
• MS-fingerprinting was capable to identify the Criollo cocoa and classify the cocoa according to their extent of processing.
• The clustering was confirmed by aroma precursor measurements and HS-SPME-GC‐–MS.
• Several markers for the bean origin/quality were identified.

In this study the quality of 26 fermented cocoa beans from 4 different countries (Vietnam, Indonesia, Peru and Ghana) is assessed by means of both a compositional analysis (amino acids, reducing sugars, polyphenols and lactic acid) and profiling of their volatile aroma composition (Mass Spectrometry (MS)-fingerprinting and Headspace-solid Phase Micro-extraction-Gas Chromatography–Mass Spectrometry (HS-SPME-GC–MS)) following high roasting processing (30 min at 150 °C). Compared with the labour-intensive measurement of cocoa precursors, MS-fingerprinting on the cocoa headspaces appeared to be a very powerful and fast classification technique. Different cocoa groups were researched having unique organoleptic characteristics which were affected by fermentation, roasting conditions and geographical origin. A clear separation of fine flavour cocoa (Criollo variety), well fermented/roasted cocoa and low quality cocoa could be made. Additionally, the clustering was confirmed by more conventional HS-SPME-GC–MS aroma analyses on the cocoa samples. Several markers for the bean origin/quality were identified. Roasting degrees were also calculated based on typical pyrazine ratios.