Partial hydrolysis of whey proteins by enzymes immobilized on an inert support can either change or evidence functional properties of the produced peptides, thereby increasing their applications. The hydrolysis of sweet cheese whey proteins by alcalase, which is multipoint-immobilized on agarose gel, is studied here. A Michaelis-Menten model that takes into account competitive inhibition by the product was fitted to experimental data. The influence of pH on the kinetic parameters in the range 6.0 to 11.0 was assessed, at 50ºC. Initial reaction-rate assays in a pHstat at different concentrations of substrate were used to estimate kinetic and Michaelis-Menten parameters, k and K M. Experimental data from long-term batch assays were used to quantify the inhibition parameter, K I. The fitting of the model to the experimental data was accurate in the entire pH range.

At present, direct on-line measurements of key bioprocess variables as biomass, substrate and product concentrations is a difficult task. Many of the available hardware sensors are either expensive or lack reliability and robustness. To overcome this problem, indirect estimation techniques have been studied during the last decade. Inference algorithms rely either on phenomenological or on empirical models. Recently, hybrid models that combine these two approaches have received great attention. In this work, a hybrid neural network algorithm was applied to a fermentative process. Mass balance equations were coupled to a feedforward neural network (FNN). The FNN was used to estimate cellular growth and product formation rates, which are inserted into the mass balance equations. On-line data of cephalosporin C fed-batch fermentation were used. The measured variables employed by the inference algorithm were the contents of CO2 and O2 in the effluent gas. The fairly good results obtained encourage further studies to use this approach in the development of process control algorithms.

This work presents a kinetic study of the side reactions of the ampicillin enzymatic synthesis, from phenylglycine methyl ester and 6-aminopenicillanic acid using penicillin G acylase immobilized on agarose. A Michaelis-Menten model with competitive inhibition was fitted to initial rates of ester and antibiotic hydrolysis, at pH 6.5 and 25ºC. Inherent kinetic parameters were estimated for low enzymatic loads, to assure that diffusional resistance was not important. It was observed that ampicillin inhibits the hydrolysis of PGME, but the inhibitory effect of the ester on ampicillin hydrolysis was almost negligible. The obtained parameters were: k cat1= 0.025 mM/UI min, Km1 = 155.4mM, K AE = 16.18mM, k cat2= 4.67x10-3 mM/UI min, Km2 = 11.47, K EA = 0.68 mM. Parameter values are in the range reported in the literature, except for Km1, which is much higher. The large confidence interval for this parameter denotes that the model presents low sensitivity with respect to it.

In this work, a bidisperse model is built to represent the hydrolysis of maltose using immobilized glucoamylase. The experimental set is a mixed-batch reactor, maintained at 30ºC, with pectin gel spherical particles that contain enzyme immobilized in macroporous silica. The possibility of substrate adsorption on the pectin gel is also studied because this phenomenon may result in smaller values of diffusivity. Equilibrium assays are then performed for different substrates (maltose, lactose and glucose) at different temperatures and pHs. These assays show that adsorption on the pectin gel is not important for the three dextrins analysed. The bidisperse model presents a good fit with the experimental data, when using previously-estimated kinetic and mass transfer parameters (Gonçalves et al., 1997). This result shows that the methodology used (wrapping the silica in pectin gel) is appropriate for experimental studies with silica, since it allows a higher degree of agitation without causing shearing

The production of penicillin G by Penicillium chrysogenum IFO 8644 was simulated employing a feedforward neural network with three layers. The neural network training procedure used an algorithm combining two procedures: random search and backpropagation. The results of this approach were very promising, and it was observed that the neural network was able to accurately describe the nonlinear behavior of the process. Besides, the results showed that this technique can be successfully applied to control process algorithms due to its long processing time and its flexibility in the incorporation of new data

In this work the effects of diffusion on the hydrolysis of maltose using glucoamylase immobilized on macroporous silica at 30o C are studied. Two different kinds of experimental assays are carried out: one using free enzyme and the other using an enzyme covalently linked to macroporous silica, preactivated with glutaraldehyde. Free enzyme assays are used to estimate the kinetic parameters for hydrolysis of maltose at 30ºC, according to a Michaelis-Menten equation, with inhibition by the product. Runs with the immobilized enzyme enable us to estimate dextrin diffusivities in silica, assuming that the inherent kinetic parameters are equal to the intrinsic ones. The values of kinetic and mass transfer parameters are: K3 = 0.90x10-5 g/U.s, Km = 0.61x10-3 g/ml, Ki = 1.23x10-1 g/ml, DMs = 0.50x10-6 cm2 /s and DGs = 0.55x10-6 cm2 /s