Elimination of non-uniform, extra-device flow effects in membrane adsorbers

Abstract

Commercial use of membrane adsorbers in the biotechnology industry is increasing. Here the system time lag created by membrane adsorber peripherals and the membrane adsorber flow distribution headers has been modeled using an anion exchange membrane and bovine serum albumin (BSA). The system time lag was modeled as a zero order and first order time lag. The zero and first order time lags have been removed from the breakthrough curve. The method used does not involve fitting a mathematical expression to the breakthrough curve. Further no assumptions are made regarding the shape of the breakthrough curve in the absence of the time lag. The method has been used to calculate the Langmuir isotherm parameters. The membrane capacity was found to be twice as large as the capacity determined after removal of the time lag. The Langmuir constant was five times as large for the system without accounting for the time lag. Errors in fitting isotherm parameters can significantly impact frontal analysis and membrane adsorber scale-up. The Langmuir isotherm calculated under dynamic conditions with the system time lag removed, was in agreement with the static adsorption isotherm.