The goal of every experimenter is to obtain a design that gives maximum information. Similarly, the performance of a design is measured by the amount of information it contains. This paper investigates mixture experiments in the second-degree Kronecker model. The parameter subspace of interest in this study is maximal parameter subsystem which is subspace of the full parameter space. Previous studies in this area have not been able to show how a design can be improved based on the same parameter subspace. This paper attempts to show an improvement of such designs by first obtaining a proper coefficient matrix. Optimal designs of mixture experiments are derived by employing the Kronecker model approach and applying the various optimality criteria. Results of A- and D-optimal designs for

 ingredients are given. The results obtained are higher than those presented in the previous studies. Finally, the efficiencies of these designs are then calculated.

mixture experiments, Kronecker product, optimal designs, weighted centroid designs, optimality criteria, moment and information matrices, efficiency.