Chitin extraction from crab shells was studied experimentally and optimized aiming to obtain chitosan with predefined deacetylation degree and molecular mass. To find out the optimum operating conditions that ensure the obtaining of a chitosan with highest deacetylation degree and specific molecular mass four parameters were varied: the concentration of NaOH and the temperature for deproteinization step, respectively HCl concentration and the number of acidic treatments for the demineralization stage. The experiment was carried on following Taguchi orthogonal array L9, and the best combination of factors was found using the desirability function approach. The optimization results showed that 5% NaOH concentration and low temperatures lead to a chitosan with high deacetylation degree. High molecular mass chitosan is obtained when a single step acidic treatment is used, while a chitosan with low molar mass is obtained for multiple acid contacts and higher HCl concentration.

Shrimp waste is a valuable source for chitin extraction and consequently for chitosan preparation. In the process of obtaining chitosan, a determining step is the chitin deacetylation. The main characteristic of chitosan is the degree of deacetylation, which must be as high as possible. The molar mass is another important parameter that defines its utilizations, and according to these, high or low molar masses are required. The present study is an attempt to optimize the deacetylation step to obtain chitosan with a high degree of deacetylation and high or low molar mass. The study was carried out based on experimental data obtained in the frame of a central composite design where three working parameters were considered: NaOH concentration, liquid:solid ratio, and process duration. The regression models defined for the degree of deacetylation (DD) and for the mean molar mass (MM) of chitosan powders were used in the formulation of optimization problems. The objectives considered were simultaneous maximum DD and maximum/minimum MM for the final chitosan samples. For these purposes, multiobjective optimization problems were formulated and solved using genetic algorithms implemented in Matlab^®. The multiple optimal solutions represented by trade-offs between the two objectives are presented for each case.