A predictive model is developed to describe the salt effects on the adsorption equilibrium of protein to Cibacron Blue-modified agarose gel (Sepharose CL-6B). This model assumes that, with the addition of salt, a fraction of dye-ligand molecules will lodge to the surface of the agarose gel resulting from the induced strong hydrophobic interaction between the dye ligands and agarose surface. This leads to a decrease in the dye ligands accessible to protein adsorption and consequently decreases the adsorption capacity of protein. The effect of salt on the dye-ligand lodging is presented by the equilibrium between salt and the dye ligands. Combined with the basic concept of steric-mass action theory, which considers both the multipoint nature and the macromolecule steric shielding of protein adsorption, an implicit isotherm of the protein adsorption equilibrium on Cibacron Blue 3GA-modified Sepharose CL-6B is formulated, involving salt concentration as a variable. Good agreement between experimental data and the predicted adsorption isotherms for single-component protein systems has been demonstrated, and fairly good matching between the predicted and experimental data for the binary adsorption of bovine serum albumin and bovine hemoglobin is obtained under most conditions. This model is expected to be useful in the design and optimization of the salt-gradient elution process of dye-ligand affinity chromatography.

A two-state protein model is proposed to describe the salt effects on protein adsorption equilibrium on hydrophobic media. This model assumes that protein molecules exist in two equilibrium states in a salt solution, that is, hydrated and dehydrated states, and only the dehydrated-state protein can bind to hydrophobic ligands. In terms of the two-state protein hypothesis and the steric mass-action theory, protein adsorption equilibrium on hydrophobic media is formulated by a five-parameter equation. The model is demonstrated with the adsorption of bovine serum albumin to Phenyl Sepharose gels as a model system. The effects of salt type (sodium chloride, sodium sulfate and ammonium sulfate) on the model parameters are discussed. Then, the model formulism is simplified in terms of the small magnitude of the protein dehydration equilibrium constant in the model. This simplification has returned the model derived on the basis of the two-state protein hypothesis to its original mechanism of salt effects on the hydrophobic adsorption of protein. This simplified model also creates satisfactory prediction of protein adsorption isotherms.

The ion-exchange adsorption kinetics of bovine serum albumin (BSA) and g-globulin to an anion exchanger, DEAE Spherodex M, has been studied by batch adsorption experiments. Various diffusion models, that is, pore diffusion, surface diffusion, homogeneous diffusion and parallel diffusion models, are analyzed for their suitabilities to depict the adsorption kinetics. Protein diffusivities are estimated by matching the models with the experimental data. The dependence of the diffusivities on initial protein concentration is observed and discussed. The adsorption isotherm of BSA is nearly rectangular, so there is little surface diffusion. As a result, the surface and homogenous diffusion models do not fit to the kinetic data of BSA adsorption. The adsorption isotherm of g-globulin is less favorable, and the surface diffusion contributes greatly to the mass transport. Consequently, both the surface and homogenous diffusion models fit to the kinetic data of g-globulin well. The adsorption kinetics of BSA and g-globulin can be very well fitted by parallel diffusion model, because the model reflects correctly the intraparticle mass transfer mechanism. In addition, for both the favorably bound proteins, the pore diffusion model fits the adsorption kinetics reasonably well. The results here indicate that the pore diffusion model can be used as a good approximate to depict protein adsorption kinetics for protein adsorption systems from rectangular to linear isotherms.

Novel dense composite adsorbents for expanded bed adsorption of protein have been fabricated by coating 4% agarose gel onto Nd-Fe-B alloy powder by a water-in-oil emulsification method. Two composite matrices, namely Nd-Fe-B alloy-densified agarose (NFBA) gels with different size distributions and densities, NFBA-S (50-165mm, 1.88g/ml) and NFBA-L (140-300mm, 2.04g/ml), were produced. Lysozyme was used as a model protein to test the adsorption capacity and kinetics for the NFBA gels modified by Cibacron blue 3GA (CB-NFBA gels). Liquid-phase dispersion behavior in the expanded beds was examined by measurements of residence time distributions, and compared with that of Streamline SP (Amersham-Pharmacia Biotech, Sweden). The dependence of axial mixing in the expanded beds on flow velocity, bed expansion degree, settled bed height, and viscosity of liquid phase was investigated. Breakthrough curves of lysozyme in the expanded beds of the CB-NFBA gels were also examined. The dynamic binding capacity at 5% breakthrough was 23.3mg/ml matrix for the CB-NFBA-S gels, and 16.7mg/ml matrix for the CB-NFBA-L, at a flow velocity of 220cm/h. The results indicate that the NFBA gels are promising for expanded bed adsorption of proteins.

The adsorption equilibria and kinetics of bovine serum albumin(BSA)and lysozyme to two Cibacron Blue 3GA(CB)modified agarose gels, that is, 6% agarose-coated steel (6AS)and Sepharose CL-6B, in 0.01 kmol•my- tris-HCl buffer(pH7.5)were studied. Effects of aqueous-phase ionic strength on both the adsorption equilibrium and uptake rate of the proteins were significant and distinctly different between BSA and lysozyme. The adsorption of lysozyme decreased monotonically with increasing ionic strength. Ionic strengths, however, maximized BSA adsorption capacities of the two CB-modified agarose gels in the tris-HCl buffer (about 0.2kmol•m-3 for CB-6AS and 0.05kmol•m-3 for CB-Sepharose), when the pore-size difference of the two matrixes and electrostatic interactions between the BSA and CB molecules of like charge were considered. The effective diffusivity of BSA, derived from a pore-diffusion model, to both the CB-modified agarose gels increased significantly with the increasing ionic strength at the ionic strength range of 0.01 to 0.3kmol•m-3, due to the electrostatic interactions between the BSA and CB molecules of like charge. In contrast, the effective diffusiities of lysozyme to CB-Sepharose in the buffer containing 0.1 and 0.3kmol/m-3 NaCl were nearly the same.