Contribution of response surface design to the development of glycerolysis systems catalyzed by commercial immobilized lipases

Two commercial immobilized lipases ("Lipozyme(R) 1M" and "Novozym(R) 435") were tested as biocatalysts for the glycerolysis of olive residue oil in n-hexane aimed at the production of monoglycerides (MG) and diglycerides (DG). A central composite rotatable design (CCRD) was followed to model and optimize glycerolysis as a function of both the amount of biocatalyst (L) and of the molar ratio glycerol/triglycerides (Gly/TG). For both biocatalysts, the production of free fatty acids (FFA) was described by second order models. In terms of MG and DG production, as well as of TG conversion, the best fits were obtained with first-order models. The highest MG productions were in the range 43-45% (w/w, on the basis of total fat) for both biocatalysts tested at a (Gly/TG) ratio of one. In the case of "Novozym 435", the lowest load used (12%, w/w) gave the best results, in contrast with "Lipozyme IM" with which a concentration of about 26% (w/w) was necessary to obtain the highest production. Under these conditions, the amount of FFA produced was about 2% and 10% (w/w), respectively, for "Novozym 435" and "Lipozyme IM" catalyzed systems. Considering both FFA production and lipase loading, "Novozym 435" was shown to be a better biocatalyst for the glycerolysis of olive residue oil in n-hexane, aimed at the production of MG, than "Lipozyme IM". (C) 2001 Elsevier Science B.V. All rights reserved.