No abstract

No abstract

Gas separation behavior of permeators containing asymmetric cellulose acetate hollow fibers with the dense skin on the outside, facing the feed gas, has been investigated experimentally and theoretically using a sweep gas technique. Two systems were studied: C0,-N,, and O,-N, (air). If the membrane structure is assumed to be symmetric or homogeneous, the model complies with the present data much better, compared to the assumption that the membrane is asymmetric.Separation was carried out with the high-pressure feed outside the fibers as well as inside the fibers. In both cases the data were well predicted by the homogeneous model. No concentration polarization effects were found when the feed flowed inside the fibers. The inherent membrane separation capability appeared practically the same whether the feed was inside or outside the fibers. Internal pressurization did not damage the membrane performance for the low pressure range used.These findings enable the operation of asymmetric hollow-fiber permeators with the feed inside the fiber lumen, which gives better separation in high stage cut situations. IntroductionLarge-scale membrane gas separation devices invariably use asymmetric membranes to achieve the desired combination of high-permeation flux and selectivity (Bollinger et al., 1982;Mazur and Chan, 1982;Schell and Houston, 1982;Yamashiro et al., 1984;Pan, 1983 Pan, , 1986. Understanding the separation behavior of such membranes is therefore of considerable importance. If the asymmetric membrane is modeled as a combination of (1) an ultrathin dense skin wherein lies all the permeation resistance, and (2) a very porous support layer that offers no permeation resistance but creates a crossflow pattern in the memibrane, then the permeation equations will be different from those for a homogeneous (symmetric) membrane (Pan 1983(Pan , 1986). Other modeling efforts involving asymmetric membranes have followed the same idealized assumption (Sengupta and Sirkar, 1986, 1987;Perrin and Stern, 1986).In a recent investigation (Sengupta and Sirkar, 1987), it was found that the behavioral pattern of a supposedly asymmetric membrane did not quite follow the ideal asymmetric membrane Correspondence concerning this paper should bc addressed to K.K. Sirkar. M. Sidhoum, A. Sengupta, K. K. SirkarDepartment of Chemistry and Chemical Engineering Stevens Institute of Technology Castle Point, Hoboken, NJ 07030 model described above. In the above study, the experimental performance of an asymmetric cellulose acetate (CA) hollowfiber permeator was compared with the theoretical predictions based on an asymmetric membrane model, and also with the theoretical predictions based on the usual homogeneous membrane model. In most cases it was not possible to conclude unequivocally that the asymmetric membrane model fitted the data better. This was observed for binary (C0,-N,) as well as ternary (He-C0,-N,) systems, for pressures up to 5 atm, and over a wide range (up to 60%) of stage cuts. The pure-component permeabilities, measured in ...

The intrabead diffusion coefficients of acetophenone and phenethyl alcohol were measured at 30 degrees C in the triphasic immobilized yeast-water-hexane system. Saccharomyces cerevisiae cells were deactivated with hydrochloric acid and entrapped in calcium-alginate beads. Measurements of dry cell loss during deactivation, shrinkage of the beads during deactivation and the final porosity of the beads were made for various cell loadings. Final concentrations of wet cells in the beads ranged from approximately 0.25 to 0.30 g/mL. Mass transfer in the hexane phase, external to the beads, was eliminated experimentally. The estimated error of 5% to 10% in the diffusion coefficients is within the experimental error associated with the bead method. The effect of significant sampling volumes on the diffusivities was estimated theoretically and accounted for experimentally. The intrabead concentration of acetophenone and phenethyl alcohol was 150 to 800 ppm. The deactivated cells were shown to be impervious to acetophenone so that the measured diffusivities are extracellular parameters. The cell volume fraction in the beads ranged from 0.70 to 0.90, significantly higher than previously reported data. The effective diffusion coefficients conform to the random pore model. No diffusional interaction between acetophenone and phenethyl alcohol was observed. The addition of 2 vol% ethanol or methanol slightly increased the diffusivities. The thermodynamic partition coefficients were measured in the bead-free water-organic system and found to be an order of magnitude lower than the values calculated from the analysis of the diffusion data for the organic-bead system, suggesting that bead-free equilibrium data cannot be used in triphasic systems. (c) 1994 John Wiley & Sons, Inc.

No abstract