Polymer Grafted Nonwoven Membranes for Bioseparations

Polybutylene terephthalate (PBT) nonwovens can be readily grafted with glycidyl methacrylate (GMA) via UV induced radical polymerization to create uniform and conformal polymer brush networks around each fiber that can be chemically modified to function as anion or cation exchangers. As ion exchangers these systems are capable of achieving equilibrium protein binding capacities hundreds of times what is possible from monolayer adsorption on the membranes. Capacities, as high as, 1000 mg of protein per gram of material were observed for the ion exchange capture of bovine serum albumin (BSA) and human immunoglobulin G (hIgG) for these systems. However, diffusion limitations in the grafted layer result in very long time scales to reach these large binding capacities. The rates of protein adsorption were determined to be a function of the thickness of the grafted layer. A high surface area islands-in-the-sea (I/S) PBT nonwoven (specific surface area = 2.5 m2/g) was investigated for polyGMA grafting and protein capture. Compared to commercially available PBT nonwovens (specific surface area = 0.9 m2/g) it was determined that polyGMA grafting was capable of being distributed in thinner layers at a given degree of grafting. This resulted in a significantly higher instantaneous amount of protein bound and a faster observed rate of adsorption. The comparatively faster rates of protein adsorption are due to the thinner grafting layers of the I/S nonwovens.;To overcome issues with polyGMA grafting by UV-light, PBT nonwovens were successfully grafted with polyGMA using a heat. When functionalized as ion exchangers, protein binding capacities, as high as, 200 mg/g were achieved for PBT nonwovens grafted using this thermal initiated grafting method. Compared to UV grafted polyGMA PBT nonwovens, the rates of protein adsorption are several times faster for the heat grafted PBT nonwoven. However, ion exchange polyGMA UV grafted PBT nonwovens are capable of binding between 5 and 7 times more protein for similar degrees of grafting. It was determined that the molar binding capacity of the heat grafted nonwovens significantly decreased as the molecular weight of the target increased. The heat grafted polyGMA layer is highly cross-linked and denser than the UV grafted polyGMA layer resulting in size exclusion effects for protein capture. On the other hand, the UV grafted polyGMA layer is a network of brushes that can accommodate a large amount of protein but also suffers from diffusion limitations in the graft layer.;PolyGMA grafted PBT nonwovens functionalized as ion exchangers were evaluated for their performance under flow conditions. It was determined that the UV grafted polyGMA brushes swell and shrink as the ionic strength of the solution changes. At low ionic strength the grafted layers swell significantly, reducing the membrane permeability and increasing the pressure drop of the nonwoven. At high ionic strength the flow permeability increases and the pressure drop is greatly reduced. Ion exchange membranes grafted using heat do not exhibit this behavior and have better flow properties than ion exchange membranes grafted with UV-light. The cation exchange polyGMA PBT nonwovens grafted using UV- light and heat exhibited dynamic binding capacities of 16 mg/ml and 20 mg/ml respectively at residence times from 1-5 minutes. The dynamic binding capacity is significantly lower than the equilibrium or static binding capacity of the membranes due to pore blockage occurring when the membranes are packed into the column for dynamic studies. The use of rigid PET nonwoven spacers increased the effective porosity of the columns, resulting in an increase in dynamic binding capacity for IgG of 24 mg/ml for the UV-light grafted materials and 35 mg/ml for the heat grafted materials.