(465b) The Effect of Electrostatic Properties in Binary Protein Ultrafiltration

The effect of electrostatic properties in binary protein ultrafiltration systems were investigated. In this work, bovine serum albumin (BSA, MW 67 kD, pI 4.7) was separated from either hen egg lysozyme (HEL, MW 14 kD, pI 11.0), cytochrome-c(MW 15 kD, pI 10.5) or alpha-lactalbumin (alpha-LA, MW 15 kD, pI 5.2) using 30,000-MWCO polyestersulfone (PES, negatively charged) and composite regenerated cellulose (CRC, relatively neutral charged) membranes. HEL and cytochrome-c were selected because of their similarity in size and electrostatic properties while alpha-LA was selected because of its significant difference in isoelectric point. All proteins were obtained from Sigma Company (St. Louis, MO, USA). HEL, cytochrome-c and alpha-LA will be referred to as the 15 kDa protein below. Experiments were carried out at three different pH values, pH 4.7, pH 7 and pH 10, in order to study the effects of system pH in ultrafiltrations.

The protein concentration of the binary protein mixture solutions were 0.05%(15 kDa protein)/0.05%(BSA). All solutions were made by dissolving measured mass of protein powder into 0.15 M NaCl buffer solution. Sodium azide at a total concentration of 0.02% was also added to protein solution as a preservative. The pH value of the solution was adjusted by HCl or NaOH solution. The solution pH was tested using a pH meter (Model 720A plus, Thermo Orion, Beverly, MA). After preparation, all binary protein solutions were pre-filtered by 0.22 micron membrane to eliminate large aggregates. Before use, membranes were cut from membrane sheet stock, and soaked in deionized water for 1 hour. All experiments were performed in a tangential-flow ultrafiltration system.

Experiments were run in diafiltration mode, with an operating pressure for all experiments of 96 kPa. During experiment, first the initial hydraulic permeability was tested by running corresponding protein-free buffer solution through the membrane. Then the protein were placed into the 500 mL tank and separation experiments were run at cross-flow flow rates 10 ml/min. Permeate flux rates were recorded at various time intervals, and small samples were collected at various times from both the permeate streamline and the retentate. The protein concentrations of the samples were determined later by spectrophotometer. The experiments were stopped after 2 hours. Next, the hydraulic permeabilities of the used membrane were measured. Afterwards the membranes were removed from the ultrafiltration system and a sample of the fouled membranes were cut and then placed in device to test the apparent zeta-potential of the membrane.

The results demonstrated the difference of electrostatic charge on protein separations within the experimental constraints, particularly with regards to solution pH. These findings will be discussed.