Abstract
Novel signal-processing protocols were used to extend the in situ sensitivity of ultrasonic frequency-domain reflectometry (UFDR) for real-time monitoring of microfiltration (MF) membrane fouling during protein purification. Different commercial membrane materials, with a nominal pore size of 0.2 µm, were challenged using bovine serum albumin (BSA) and amylase as model proteins. Fouling induced by these proteins was observed in flat-sheet membrane filtration cells operating in a laminar cross-flow regime. The detection of membrane-associated proteins using UFDR was determined by applying rigorous statistical methodology to reflection spectra of ultrasonic signals obtained during membrane fouling. Data suggest that the total power reflected from membrane surfaces changes in response to protein fouling at concentrations as low as 14 μg/cm2, and results indicate that ultrasonic spectra can be leveraged to detect and monitor protein fouling on commercial MF membranes.