Application of multi-block analysis and mixture design with process variable for development of chocolate cake containing yacon (Smallanthus sonchifolius) and maca (Lepidium meyenii)

Abstract

Background: In this study, a chocolate cake formulation was developed with partial substitution of wheat flour by yacon and maca flour. A simplex-centroid design was applied to determine the proportions of the three flours, and the amount of water was included as a process variable at three distinct levels. According to the overall acceptability of the cakes, the tasters were separated into two groups using k-means.

Results: After segmentation, regression models were constructed for overall acceptability of each group; R²_adjusted values of 92.5% for group 1 and 98.9% for group 2 were obtained. Using the sequential simplex method an optimized formulation was determined for group 1 (0.49 kg_wheat kg^-1_{total flour} , 0.37 kg_yacon kg^-1_{total flour} , 0.14 kg_maca kg^-1_{total flour} and 140.0 mL of water) and another for group 2 (0.35 kg_wheat kg^-1_{total flour} , 0.65 kg_yacon kg^-1_{total flour} and 120.0 mL of water). In addition to these formulations, a third formulation was proposed with a greater maca proportion (0.32 kg_maca kg^-1_{total flour} ), which does not significantly alter the overall acceptability of both groups. The three optimized formulations and two control formulations were evaluated through free-choice profiling. The data were evaluated using the multi-block method common components and specific weights analysis (CCSWA).

Conclusion: It was observed that a greater proportion of maca intensified brownness and burnt aroma and taste, whereas a larger proportion of yacon produced a better appearance, softness, sweetness and chocolate flavor. © 2017 Society of Chemical Industry.

Keywords: common components and specific weights analysis; constrained optimization; free-choice profiling; multi-block analysis; simplex-centroid design with process variable; taster segmentation.