Characterization of pore structure of a strong anion-exchange membrane adsorbent under different buffer and salt concentration conditions.

The quantitative characterization of pore structure of Sartobind Q, a strongly basic membrane anion exchanger that is formed by cross-linked cellulose support and a hydrogel layer on its pore surface, was made combining the results obtained by several experimental techniques: liquid impregnation, batch size-exclusion, inverse size-exclusion chromatography, and permeability. Mercury intrusion and nitrogen sorption porosimetry were carried out for a dry cellulose support membrane in order to get additional information for building a model of the bimodal pore structure. The model incorporated the distribution of the total pore volume between transport and gel-layer pores and the partitioning of solutes of different molecular weights was expressed through the cylindrical pore model for the transport pores and random plane model for the gel layer. The effect of composition of liquid phase on the pore structure was investigated in redistilled water, phosphate and Tris-HCl buffers containing up to 1M NaCl. Evident differences in the bimodal pore structure were observed here when both the specific volume and size of the hydrogel layer pores significantly decreased with the ionic strength of liquid phase.

Characterization of pore structure of chromatographic adsorbents employed in separation of monoclonal antibodies using size-exclusion techniques.

Structural properties of commercial chromatographic adsorbents designated for separation of monoclonal antibodies were investigated using size-exclusion techniques. A batch technique provided the specific pore volumes distributed among small, medium and large pores. Inverse size-exclusion chromatography yielded the partition coefficients of dextran solute probes in medium pores. These data were fitted with one- and two-parametric models corresponding to different pore-size distribution functions and the suitability of the individual models for the characterization of the examined materials was then assessed. The reliability of estimation of the parameters of log-normal distribution of cylindrical pores was investigated. Large uncertainties and a strong correlation of the mean pore diameter and width of distribution function were observed.