Abstract
A European interlaboratory study was conducted to validate an analytical procedure for the detection and quantification of cocoa butter equivalents in cocoa butter and plain chocolate. In principle, the fat obtained from plain chocolate according to the Soxhlet principle is separated by high-resolution capillary gas chromatography into triacylglycerol fractions according to their acyl-C-numbers, and within a given number, also according to unsaturation. The presence of cocoa butter equivalents is detected by linear regression analysis applied to the relative proportions of the 3 main triacylglycerol fractions of the fat analyzed. The amount of the cocoa butter equivalent admixture is estimated by partial least-squares regression analysis applied to the relative proportions of the 5 main triacylglycerols. Cocoa butter equivalent admixtures were detected down to a level of 2% related to the fat phase, corresponding to 0.6% in chocolate (assumed fat content of chocolate, 30%), without false-positive or -negative results. By using a quantification model based on partial least-squares regression analysis, the predicted cocoa butter equivalent amounts were in close agreement with the actual values. The applied model performed well at the level of the statutory limit of 5% cocoa butter equivalent addition to chocolate with a prediction error of 0.6%, assuming a chocolate fat content of 30%.
NATIVE STATE OF METALS IN NON-DIGESTED TISSUES BY PARTIAL LEAST SQUARES REGRESSION ANALYSIS OF VISIBLE AND NEAR-INFRARED SPECTRA
