Influence of raw material and processing on aroma in chocolate

Abstract

There is no chocolate without cocoa, the main raw material for its production. The quality of the raw material is very important to obtain good quality chocolate as some ‘imperfections’ in it cannot be removed from the chocolate even with processing. Ghana is an important cocoa-growing country, supplying more than 20% of the world’s cocoa. In spite of Ghana’s cocoa beans being a standard for the Forastero variety, variation exist in the exported beans due to differences in the geographic areas of the cocoa growing regions within Ghana, as well as fermentation method/degree. There is therefore the possibility of obtaining cocoa beans of better quality with proper separation. Aside the primary processes of fermentation/drying of cocoa, the chocolate production processes of roasting and conching are also important determinants of the aroma/flavor of the final product. These processes together determine the aroma/flavor compounds that define the aroma/flavor characteristics of chocolate. The general objective of this PhD thesis was to obtain more knowledge about the factors determining the aroma of chocolate and specifically, to enable production of chocolate with a better quality from Ghanaian cocoa beans. The aroma compounds in dark chocolate produced from heap-fermented and tray-fermented Ghanaian cocoa beans were isolated by an optimized Dynamic Headspace Sampling Method and identified by Gas Chromatography-Mass Spectrometry (GC-MS). Odor-active compounds in the chocolates were detected by Gas Chromatography-Olfactometry (GC-O) and identified by GC-MS. The intensities of flavor characteristics of ‘heap’ and ‘tray’ chocolates were determined by a Quantitative Descriptive Analysis (QDA) using a trained sensory panel. Differences were found between the types but mainly the levels of aroma compounds in the two chocolates. There was also observed differences between the most important aroma compounds determined by their SNIF (Surface of Nasal Impact Frequency) value for ‘heap’ and ‘tray’ chocolates. Peak areas of isoamylacetate, linalool and methyl phenylacetate were significantly different at p<0.05 for the two types of chocolate. Phenylacetaldehyde with a bitter/green/grassy odor and a peak described as having a chocolate/cocoa/roasted odor and identified as a non-separable mixture of 2- and 3-methylbutanal seemed important to the aroma of both chocolates, based on their SNIF values. However, Linalool (fuity/sweet/flowery) had a SNIF value in tray chocolate which is almost twice that of heap chocolate. ‘Tray’ chocolate was generally perceived to have a more fresh fruit, sweet and yoghurt flavor than ‘heap’ chocolate. These flavor attributes were also positively correlated but negatively correlated with bitter and astringent. Observed differences in the levels of some key odorants identified probably account for observed sensory differences in ‘tray’ and ‘heap’ chocolates. The effect of fermentation duration on the aroma of four dark chocolate produced from 2, 3, 4 and 4 days tray-fermented Ghanaian cocoa beans was also investigated. Results showed a general increase in the number of aroma compounds with increasing fermentation duration. The 3 day-fermented sample, however, did not follow this trend as it had higher GC-MS peak areas in a number of compounds than the other three samples. More odor-active pyrazines were detected in the 5 day-fermented sample than in the other three samples. Increased days of fermentation also resulted in the reduction of undesirable attributes such as sourness, bitterness and astringency, whilst fruity, flowery and sweet flavors were related to 4 and 5 days-fermented samples. 3-day fermented samples seemed to be the worst chocolate as it was rated higher in all the undesirable attributes than even the 2-day fermented samples. Due to the unexpected results from the 3 day-fermented samples, the experiment is being repeated to confirm the results. In the meantime, however, it can be concluded that at least 4 days of tray fermentation is required to obtain chocolate which is less bitter, sour and astringent but sweet, flowery and fruity in flavor. Additionally, the combined effect of cocoa fermentation method, as well as roasting and conching conditions on the aroma of ‘heap’ and ‘tray’ chocolates have been investigated in this thesis. A completely randomized design was used to study the two factors, roasting and conching conditions for the two fermentation methods, heap and tray. Three roasting conditions: 100oC/100 min, 120oC/45 min, 150oC/30 min and three conching durations, all at 80 oC: 6 h, 8 h and 10 h were used in combinations to produce ‘heap’ and ‘tray’ chocolates. Aroma volatiles in the chocolates were influenced by all three processing factors. A PCA (Principal Component Analysis) plot based on sixteen most important compounds detected by all judges in at least one sample during GC-O explained 81% of the variation in the samples with two PCs. The levels of most important odorants, among them 2- and 3-methylbutanal (cocoa, roasted), 1,2/3-butanediol and benzyl acetate (sweet, flowery) were generally higher at high roasting temperature (150oC) but decreased significantly (p<0.05) with increased conching duration from 6-10 h. On the other hand, the level of 5-methyl-2-phenyl-2-hexenal generally increased drastically with short/medium conching duration but reduced drastically with further conching in both ‘heap’ and ‘tray’ chocolates indicating its formation at the short/medium durations of conching. Since the dataset was reduced to only represent the volatiles that were most important from a sensory point of view, it was expected that the heap and tray fermented samples in general will have different sensory quality and that this difference would become more pronounced with samples roasted at the higher temperatures (120 or 150 °C) and conched for rather short time (6-8 hours) or not conched at all. To investigate this, a QDA was carried out on six selected samples. Eight of the sixteen flavor attributes used to describe and quantify the intensity of attributes in the samples were significantly different (p<0.05). Unroasted samples were rated higher in astringency than roasted samples whilst unconched samples were rated higher in fruity and banana attributes than conched samples. Multivariate data analytical tools, PCA and PLS were used to investigate quantitative descriptive analysis and GC-O data and also to relate the two sets of data. PLS1 models based on single sensory attributes performed better than PLS2 models based on all sixteen sensory attributes. Ethyl-3-methylbutanoate (fruity, flowery); 2,5-dimethylpyrazine (popcorn), dihydro-2(3H)-furanone (sweet); linalool oxide (sweet, flowery); benzaldehyde (earthy, nutty) and 2/3-methylbutanal (cocoa, roasted) efficiently modeled fruitiness. It was also possible to model the taste sensory attribute astringent from the aroma compounds 5-methyl-2-phenyl-2-hexenal (sweet, roasted cocoa), ethyl-3-methylbutanoate and pentyl acetate (green, cucumber). Since fruity flavor was higher in unconched samples and astringent mouth-feel higher in unroasted samples, it may be possible to predict conching and roasting in dark chocolate from the concentrations of the above mentioned important aroma compounds. The results of this investigation are presented in three papers submitted to international journals, three conference manuscripts (two of which have been published) and one manuscript under preparation