Improving dextran tests for ultrafiltration membranes: Effect of device format

Although dextran retention tests have become the de facto standard in membrane characterization and qualification, the effects of device configuration and operational parameters on the measured results are still poorly understood. This paper examines the underlying membrane and bulk mass transport phenomena governing dextran sieving coefficients and calculated membrane rejection values. Particular emphasis is placed on the effects of device configuration and operating conditions, including a comparison of sieving curves in a stirred cell and tangential flow (Minitan) system. Model calculations examine the effects of solute diffusion through the membrane, a phenomenon that becomes important at the low filtrate flux typically used to minimize concentration polarization. Simulations were also per-formed to evaluate the pressure and flow profiles within the Minitan device and their effect on the measured dextran sieving coefficients. Experimental data are presented to support the model calculations. Results indicate that stirred cells provide a more robust and accurate test because of better transmembrane pressure and flux uniformity, smaller influence of concentration polarization, and absence of back filtration effects. (c) 2007 Elsevier B.V. All rights reserved.