Search

Osmotic water flow through membranes with uniform defined pores was measured for a variety of macromolecular solutes. Water flow increased linearly with applied hydrostatic pressure, allowing the effective osmotic pressure of the solutes to be estimated by extrapolation. Reflection coefficients for each solute-membrane combination were calculated and correlated with the ratio of solute size to pore size. For the same mean molecular size, proteins were found to have larger reflection coefficients than dextrans. Molecular rigidity may play a role in this difference in behavior.

A graphical method of determining the Michaelis-Menten constant free of the external mass transfer resistance for a packed bed immobilized enzyme system was illustrated with examples from 3 different enzyme reactions. The intercept at the ordinate obtained by the straight line extrapolation of data points in the plot of apparent Km value vs. the reciprocal of superficial velocity in column allowed an easy calculation of Km free of external mass transfer resistance. An asymptotic value of apparent Km value at infinite zero superficial velocity was ascribed to the fact that the mass transfer coefficient kL, approached a definite value at this condition.

The whole cell ofHumicola spp. ATCC 20620 with rifamycin oxidase activity was immobilized by copolymerization with acrylamide. The whole cell was defatted by treatment with acetone to reduce the diffusional resistance through the cell membrane. The recovery of enzyme activity after the immobilization step was about 50%. The acetone-defatted cell showed the maximum activity at pH 7.5 for both free and the immobilized forms. No appreciable activity loss could be detected when stored at 4 °C and pH 7.8 for one month, while the half life at 40 °C and pH 8 was decreased to about 8 days. The apparent Km values of rifamycin oxidase for the free and immobilized acetonedefatted cells were 0.3mM and 0.6mM, respectively. The enzyme demonstrated substrate inhibition, but the degree of substrate inhibition was different between two forms of the enzyme preparation. A complete substrate inhibition was observed for the immobilized cell, whereas the enzyme activity was partially inhibited at high substrate concentration in the acetone-defatted cells.

A theoretical model equation was derived to find the correlation between the conversion and the amount of immobilized penicillin amidase in column. The theoretical values of the conversion were predicted form this correlation and compared with experimental results. It was observed in a column reactor that the pH drop along the column path was linear versus the enzyme loading and that the enzyme activity was also linearly dependent on pH up to 8.0. In order to diminish the effect of pH drop, a continuous two-stage plug-flow reactor (PFR) with pH adjustment between the two columns was used was used in the experiments, and two- and three-stage PFRs were simulated by computer. In the case of the two-stage PFR, the maximum productivity was demonstrated experimentally and theoretically as well. when an equal amount of the immobilized enzyme was packed in both columns. It was also predicted in the tree-stage PFR system that the optimal distributions of enzyme loading in three columns were found to be 1:1:1. It was demonstrated that the increased number of reactors in series could enhance the level of the maximum productivity with a given amount of enzyme loading.

The effect of external mass transfer resistance on the overall reaction rate of the immobilized whole cell penicillin amidase of E. coli in a recirculation batch reactor was investigated. The internal diffusional resistance was found negligible as indicated by the value of effectiveness factor, 0.95. The local environmental change in a column due to the pH drop was successfully overcome by employing buffer solution. The reaction rate was measured by pH-stat method and was found to follow the simple Michaelis-Menten law at the initial stage of the reaction. The values of the net reaction rate experimentally determined were used to calculate the substrate concentration at the external surface of the catalyst pellet and then to calculate the mass transfer coefficient, k(L), at various flow rates and substrate concentrations. The correlation proposed by Chilton and Colburn represented adequately the experimental data. The linear change of log j(D) at low log N(Re) with negative slope was ascribed to the fact that the external mass transfer approached the state of pure diffusion in the limit of zero superficial velocity.