A new high-pressure liquid chromatography (HPLC) method has been developed to estimate the percentage of cocoa-bean shells in cocoa products by determining the concentration of fatty acid tryptamides in these products. This new method is expected to yield more precise information than the previously used “Blue Value” method developed by FINCKE in 1963; the latter is a summation method and is therefore not highly specific.

Within the scope of this research project, statistical data was collected on the tryptamide concentrations in cocoa products. The tryptamide concentrations of 454 cocoa products were measured. Fatty acid tryptamide (FAT) is found in all cacao products (Table 1) – ranging from cocoa nibs over cocoa shells, cocoa powder and cocoa liquor to cocoa butter. In all of these products, the ratio of lignoceric acid tryptamide (LAT) to behenic acid tryptamide (BAT) is almost constant. About twice as much LAT as BAT is formed in cocoa. However, by far the highest concentration of FAT was measured in the cocoa shells; the concentrations measured in the shells were higher than the concentrations measured in the nibs by an average factor of 16. As a result, HPLC analysis can be used to estimate the percentage of shells in cocoa products.

Table 1: Tryptamide concentrations of cocoa products (HPLC analysis)

cocoa products	No. of samples	FAT (mg/kg)	FAT (mg/kg of ffdcc)
cocoa nibs	186	19,3	45,1
cocoa shells	171	313,8	-
african cocoa powder	24	8,4	10,7
american cocoa powder	5	18,4	23,5
cocoa liquor	15	34,7	77,8
cocoa nibs + liquor	201	20,4	47,5
pressed cocoa butter	28	35,6	-
expelled cocoa butter	2	97,5	-

-: No data provided since not relevant – ffdcc: fat free dry cocoa content

The processing step of roasting has no effect on the tryptamide concentration. It was noted only that alkalization affected the tryptamide concentration of cocoa. A slight drop in the FAT concentration is observed after alkalization.

Studies exploring the effect of bean size on the FAT concentration have not yet yielded any clear-cut results. Tryptamide concentration rises as bean size increases; however, this increase is not proportional as one might expect. Only in african nibs samples can a small rise of the FAT concentration be observed as bean size increases. In shells, in contrast, more pronounced fluctuations are observed.

Investigations carried out with cocoa beans from different regions have not revealed any major differences between the various cocoa-producing countries. Hence, a calculation basis can be derived from the total cocoa nibs and shells examined.

Following the statistical analysis, it was possible to group the tryptamide concentrations of cocoa nibs and cocoa liquor into one basic entity. A question that has not yet been definitively settled, however, is why the concentrations of all the samples of cocoa liquor examined were in the upper concentration range whereas most of the samples of cocoa nibs exhibited tryptamide concentrations of 19 mg/kg. These intriguing findings should be subjected to further investigation to determine conclusively whether cocoa nibs and cocoa liquor can be grouped into one basic entity.

Following the analysis of the semi-finished cocoa products, a relationship can be established between the tryptamide concentrations of the individual cocoa products. The tryptamide concentrations of cocoa butter and press cake are interdependent; higher concentrations of tryptamide in cocoa butter are associated with lower concentrations in press cake and vice versa. If the FAT concentration of a cocoa liquor consisting of cocoa butter and powder is calculated on the basis of the percentages of cocoa butter and powder it contains, this calculated FAT concentration is found to be comparable to the concentrations measured in the liquor or nibs.

Moreover, a comparison of the results obtained by HPLC and by the “Blue Value” Method, respectively, reveals a distinct relationship between the two. Various effects due to the different matrixes occur, however, causing the discrepancy between the results obtained with the two methods to be greater or less. In general, the “Blue Value” Method can be expected to yield higher tryptamide concentrations since it is a summation method. Cocoa nibs and liquor are the exception here; the FAT concentrations measured in these samples with the “Blue Value” Method are lower than those obtained by HPLC.

This research project has shown that fatty acid tryptamides are suitable indicators for ascertaining the percentage of shells in cocoa products. Owing to the natural fluctuations in the tryptamide concentration in cocoa and the analytical error limits, however, this method results only in an estimation. Investigations conducted on both a laboratory and an industrial scale to evaluate the feasibility of applying this method have shown that the percentage of shell in cocoa nibs can be determined with sufficient precision by calculation. It should thus be possible to use the calculation basis developed here for routine analysis in industry. However, a specific method still has to be developed for determining the percentage of shells; this is an interesting subject for further investigation.

Information on the concentrations of fatty acid tryptamide in cocoa butter is also of interest for the quality management of cocoa butter even though this is only one of several components making up the whole picture. Since FAT is fat-soluble, it goes into the cocoa butter after pressing. For this reason, the FAT concentration yields information on the type of extraction and the force used for pressing. The tryptamide concentrations found in expelled butter are about twice those found in pressed butter. A quick estimation can be made, on the basis of the parameter FAT, as to whether a particular cocoa butter was made by pressing or expelling. The use of the limit value for pressed cocoa butter proposed here should be the subject of further discussion since it is only informative for starting products. In finished products containing additional cocoa butter, it is difficult, if not impossible, to make such a statement. This aspect of quality analysis should also continue to be followed closely and, if necessary, investigated with additional indicator substances.

Literature:
1) Münch M, Schieberle P (1999) Z Lebensm Unters Forsch A 208 • 39-46
2) Münch M, Schieberle P, Janßen K, Raters M, Matissek R (2000) Süsswaren 43(9) • 28-31
3) Janßen K, Matissek R (2001) Eur Food Res Technol (in preparation)

Authors: K. Janßen, M. Raters, R. Matissek

Institute: LCI - Food Chemistry Institute, Cologne, Germany