Validation of cross-flow ultrafiltration for sampling of colloidal particles from aquatic systems

The use of cross-flow ultrafiltration (CFF) for the isolation of colloids from the bulk aquatic matrices has been increasing over the last decade; however, a thorough validation of each CFF system is needed before its application to natural samples. Laboratory experiments have been conducted to validate a 1-kDa Millipore Pellicon 2 cartridge type CFF system, using a range of molecular probes spiked in natural waters. Using the separation data of these probes, key validation parameters have been calculated, including retention coefficient (RC), concentration factor (cf), colloid concentration, recovery, and mass balance. The permeation behaviour of these molecules has also been studied using a simple permeation model. Results show that good retention (>80%) of high molecular weight (HMW, >1 kDa) molecules and low retention (<20%) of low molecular weight (LMW, <1 kDa) molecules can be achieved at high cf values or over long timescales. In addition, permeation modelling and the calculation of colloid concentration demonstrate the same findings, again recommending high cf values or long timescales for ultrafiltration to minimise the retention of LMW molecules. The retention characteristics of 3-kDa dextran change slightly with salinity. Mass balance calculations show good recoveries for all of the molecules tested. The apparent membrane molecular cut-off of the CFF system has been estimated to be between 2.1 and 2.5 kDa, which is slightly greater than the manufacturer's nominal MW cut-off of 1 kDa.