Published in 2009. The following is a summary of the stated research mentioned above. The content summarized here, including the figures and tables, all belong to the researchers (unless otherwise indicated). The summary attempts to stay as close to the original paper as much as possible with some adjustments in regards to jargon, length, or to focus on bean to bar aspects.

Introduction

Previous studies suggest that there may be a relationship between cocoa pulp flavour and the presence of fruity and/or floral flavours in fine cocoa genotypes. In this study, the fruit of two fruit species were added to Amelonado cocoa beans during fermentation to see whether the flavours of the fruit would impact the flavour of the beans. The two fruits used were Theobroma grandiflorum (Cupuaçu, a close relative to cacao), and Anona muricate. These fruits were added in the meddle of the fermentation mass 1 or two days after the start of fermentation. A study by the amateur public in Montpellier, France showed that they preferred chocolates that had acquired flavours from T. grandiflorum and A. muricata above the control chocolate.

Soursop (top) and cupuaçu (bottom) are two tropical fruits found in Brazil. These fruits were added to cacao beans during fermentation. Cupuaçu is closely related to cacao, as it is part of the same genus Theobroma.

Organoleptic analysis revealed some presence of flavour attributes from the fruits in the fermented beans. This suggests that the cotyledon absorbed flavours from the pulps during the fermentation process. Compared to the control, the astringency of the chocolates obtained increased. This also suggests that there is evidence that the cocoa cotyledons are capable of also absorbing the fruity flavours from the cocoa mucilage surrounding the beans during fermentation. As well, the findings here suggest that sensory quality of the nibs, cocoa liquors, and chocolate can be modified by adding aromatic substances during fermentation. As far as known to the authors, these results are the first publish evidence that the cocoa bean flavour environment (mucilage around the bean) can affect the flavours in the cocoa liquors and chocolate.

Cocoa flavour has been related to genetic and environmental factors (Clapperton 1992, Lockwood and Eskes 1996, Sukha and Butler 2006). Clapperton observed significant varitations among cocoa gentotypes, and significant parent-offspring correltions for cocoa flavour, astringency, fruity flavour, and viscosity of the chocolate. Other studies suggested no effect of the pollen donor or of the cocoa fermentation environment.

However, more recent work on the impact of pollen donors (Sukha, 2008), saw a small effect on astringency, nutty, and other flavours (but not on acidity, cocoa flavour, fruity or floral flavour).

Eskes (2007) has reported significant differences which exist in cocoa pulp flavours among genetically different cocoa origins, and that human preference for cocoa pulp had a positive correlation to the presence of fruity and or floral flavours in the pulp. As well, the genotypes and pulp preferred by humans appeared to be the types known for their fruity or floral flavours in chocolates made with those beans. This is indirect evidence that the flavour environment can affect the final flavour of the chocolate.

In this study, the direct effect of the cocoa bean flavour environment on the flavour profile of cocoa nibs, cocoa liquors, and chocolates was assessed by adding the aromatic pulp of fruits to Forastero cocoa (Amelonado) cocoa beans during fermentation.

Materials and Methods

Fruit pulps were obtained from two fruits:

T. grandiflorum (cupuaçu) - Pulp is sweet, tart, creamy, and has a strong aromatic fresh fruit flavour which is very present throughout.

Anona muricata (sour sop, guanabana) - Pulp is more watery, slightly tart, with more “green” notes at the beginning with a delicate flavour coming through towards the end.

Both of these fruits are common in Brazil to flavour juices or ice-creams. The cocoa beans used were of the Amelonado type (“Comun” variety) and grown on the farm “Boa Sorte” in urucuca, South Bahia, Brazil. The cocoa is consisted bulk cacao because it has a strong cocoa flavour, but not much else including no fruity flavour.

Fermentations were conducted in January of 2007. The mass was turned 48, 72, 96, and 120 hours after initiation of the fermentation process and total fermentation duration was 144 hours (6 days). The micro fermentations were conducted by placing 1kg of wet cocoa beans either alone or mixed with an equal quantity of aromatic fruit pulp of one fruit or the other. The fruit was added in polystyrene netted bags in the middle of the mass.

Control: One kg of cacao beans (fermented 24 hours) with no fruit added, in a bag, and added to the mass.

The idea of the experiment set up. Bags of the 5 different treatments used slightly fermented cacao beans from a larger fermentation mass, placed in bags according to treatment, and then those bags were placed back into the larger fermentation mass until fermentation was complete. Drawing by Bean To Bar World.

Treatment 2: One kg of cocoa beans (fermented 24 hours) mixed with 1kg of cupuaçu, and then added into a bag and into the mass.

Treatment 3: One kg of cocoa beans (fermented 48 hours) mixed with 1kg of cupuaçu, and then added into a bag and into the mass.

Treatment 4: One kg of cocoa beans (fermented 24 hours) mixed with 1kg of soursop, and then added into a bag and into the mass.

Treatment 5: One kg of cocoa beans (fermented 48 hours) mixed with 1kg of soursop, and then added into a bag and into the mass.

The beans were then sun- dried for 7 days to attain 8-10% moisture. They were sent to CIRAD, Montpellier, France in May 2007 where nib flavour was assessed by the first and third authors. Cocoa liquors and chocolates (65% and 35%) were prepared by Guittard Chocolate Co. in California, USA in Sept. Oct 2007, and the chocolates sent to Montpellier in Nov. 2007.

Results

Evaluation of cocoa nibs

Cocoa nibs were evaluated by sniffing and tasting, with a focus on fruity flavours. Treatment 1 (control) tasted similar to nibs of the normal Amelonado type beans, which is low in acidity, quite bitter, with no fruity notes.

Fruit flavours were detected in all other treatments (2-5), with treatments 2 & 3 having a fresh-fruit flavour resembling the cupuaçu pulp that was contained within it. Treatments 3 and 5 used beans that were fermenting longer than in treatments 2 and 4, and also appeared to have fruity flavours that were more easily detectible.

Evaluation of Cocoa Liquors

Note A cocoa liquor has nothing to do with alcohol. It is a term used to describe unrefined and unfinished chocolate, where the nibs are ground into a liquid, but not fully refined into a chocolate yet. Also keep in mind that the panelists here were not familiar with the specific flavour characteristics of the added pulps.

Treatments 2 and 4 (24 hour with cupuaçu and sour sop respectively) had an altered flavour profile consistent with the aromatic pulps of those fruits. They also appear to alter aspects of the fermentation, with lower acidity and astringency detected as well. Liquors of Treatments 2 and 4 also had a reduction in the woody notes present in the control. As well, treatment two had a brown fruit flavour, while treatment 4 had a fresh floral note to it.

Treatments 3 and 5 (48 hour with cupuaçu and sour sop respectively) also had an altered flavour profile, but the result was less than favorable. The acidity was heightened and became more of a “dirty”, astringent, metallic, and vinegar-like flavour. They also lost completely the woody notes that are present in the control.

Evaluation Of Chocolate

The chocolates also had much variation between the different treatments. Treatment one (control) had the most intense cocoa flavour notes compared to the other 4 treatments. It also had a low-intensity but complex fruity flavour. Treatments 2 and 3 (both cupuaçu at 24 hour and 48 hour) had much stronger fruity notes resembling cupuaçu, and had flavours also described as “yellow prune/cherry like”. Treatments 4 and 5 (both sour sop at 24 hour and 48 hour) had typical sour sop flavours within the chocolate, and also described as “ripe yellow fruit” flavours.

The fruity flavours were more intense in the chocolates made from the micro-fermentations initiated 48 hours after the beginning of fermentation (3 and 5), than the 24 hour micro fermentations. However, the 48-hour batches also had higher astringency, and higher sour notes for treatment 3 (cupuaçu at 48 hours).

Chocolate Preference Of An Amateur Public

Amateur public panelists tasted the same 5 chocolate samples of the various 5 treatments. Generally, they preferred the chocolates made from cocoa beans of treatment 2 (cupuaçu at 24 hours) and treatment 5 (soursop at 48 hours) compared to the control. People who were familiar with cupuaçu and sour sop were able to distinguish those flavours within the chocolate. A minority of the group preferred the low-intensity complex fruity flavour of the control.

Discussion

Effect of pulp flavour environment

The addition of the pulp had a profound impact on all treatments 2-5. Those who were familiar with cupuacu and soursop were able to recognize these flavours within the chocolate samples, suggesting the flavours of the fruit migrate into the cocoa bean during fermentation. The cupuacu had a stronger impact in regards to flavour than did the soursop, which correlates with the intensity levels of the fresh fruit themselves, with the cupacu having a more intense and lingering flavour than the soursop.

The low-intensity fruity flavours of the control was a surprise, since these beans don’t normally have such flavours. This suggests that perhaps since the control and all the treatments were contained within the larger fermentation mass, that there may have been some exchange of fluids from the bags containing the fruit into the surrounding cacao fermentation mass and control bag. This reinforces the idea of the flavours within the fermentation mass imparting into the cocoa beans themselves.

Comparison of liquor and chocolate flavour profiles

The evaluation of the liquors showed a presence of fruity notes in treatments 2 and 4 (24 hour treatments) but not in the control or 48 hour treatments (1, 3, 5). However, in the final chocolate, fruity flavours were detected for all 5 treatments, and most strongly expressed in treatments 2 (24 hour cupuacu) and 5 (48 hour soursop). Why would the final chocolates have a fruiter flavour than the liquors? It’s hypothesized that the presence of sugar in the chocolate enhances the expression of fruity flavours perceived. If this is true, than fruity flavours may be more noticeable in chocolates than in cocoa liquors.

Effect of the timing of the micro-fermentations

The slightly stronger fruity flavours of treatments 3 and 5, where beans were mixed with fruit after initial 48 hours of fermentation, might be related to the germination process of cocoa beans (Rohsius et al. 2006). During the first two days of fermentation, the cocoa beans initiate germination before they are killed due to heat and acidity levels. It is suggested the speed of uptake of soluble compounds (such as those from the fruits) by the cotyledon is possibly affected by the opening of the testa as they begin to germinate.

Although fruity notes increased in treatments 2-5, the cocoa flavour intensity was reduced, and acidity and astringency increased (especially for treatments 3 and 5 at 48 hours). The 48 hour treatments seem to interfere more with the normal cocoa bean fermentation process than the 24 hour treatments. This effect on the 48 hour treatments may have been caused by the transfer of astringency from the fruit pulps to the cocoa beans. It may also be a by a modification in the fermentation conditions caused by adding these fresh fruit pulps at 48 hours versus 24 hours. The added pulps would have increased the availability of fermentable sugars, and therefore increasing acidity around and within the cocoa beans. This likely reduces the activity of polyphenoloxidase and proteolytic enzymes involved in developing flavour precursors (Biehl et al. 1985; Lee et al. 1991). High astringency and low cocoa flavour has often been associated with fermentation experiments (Clapperton, 1992).

Prospects for applications

The results of this study suggest the fruity traits of pulp added during fermentation can migrate to the beans. This means that the cacao fruit of the pulp around the cacao beans themselves can have a direct impact on the flavour of the cotyledon. Therefore, it would be possible to select cocoa varieties based on the flavour of their pulp, which vary widely among various cocoa genotypes (Eskes et al. 2007). This may lead to an improvement in the selection efficiency for cocoa traits. This may also lead to farmers identifying new flavours in cocoa according to the flavours of the fruits they find within some of their trees. This may also allow them to raise the economic value of their cocoa beans if finding economic ways to improve the flavour of the beans during fermentation.

These findings require a more rigorous investigation, especially in regards to any negative impacts of the fruit additions (such as in treatments 3 and 5). Once proven effective and consistent, the addition of aromatic substances to the cocoa beans during fermentation could open up new opportunities for developing cocoa liquors, chocolates, or other cocoa-based products with a diverse array of unique flavours.

Evidence For The Effect Of The Cocoa Bean Flavour Environment During Fermentation On The Final Flavour Profile Of Cocoa