Design of liquid chromatography systems for capture and purification
Our objective is to understand the fundamentals of how the different variables act on the separation, and to develop systematic methods for designing different types of chromatography systems. Among the cases we investigated in the last five years, the following two are selected to be discussed in the thesis: (1) affinity chromatography for target component capture; (2) ligand-assisted elution chromatography for multi-component purification. Affinity chromatography has been broadly applied in pharmaceutical industry for recovering a target from a complex mixture. When a feed solution is loaded onto a column, the target solute is selectively captured from the mixture by the sorbent. The challenge in designing a capture process is to maximize sorbent productivity, or minimize column volume, while satisfying yield, loading time and pressure limit requirements. A general design method based on dimensionless groups has been developed for Langmuir isotherm systems. This method requires only the values of intrinsic parameters, which can be estimated from a small number of bench-scale experiments. Given feed conditions, loading time, and desired yield, the minimum column volume for capture and the maximum operating velocity can be determined readily without simulations. If the design requirements or material properties are varied, their effects on the column volume can be visualized graphically. This method has been verified by Protein A chromatography data for antibody purification, and is applicable to a wide variety of capture systems and production scales. The general method was applied to design the TiO2-packed columns for Mo-99 recovery from uranium fission products. Mo-99 is the parent of Tc-99m, which is used annually in 25 million diagnostic procedures worldwide. The minimum column volume and maximum operating velocity for capture were found rapidly using the general design method. Sensitivity analysis was performed to find the safety factor that needs to be considered in response to fluctuations in operating conditions and variations in material properties. Adaptive designs were also carried out for varied material properties, design constraints, and feed conditions. The optimal stripping conditions for Mo recovery were found based on rate-model simulations. The capacities of multiple adsorption sites of TiO 2 were first estimated from frontal tests with various acids and bases. Appropriate models for Mo adsorption and desorption were then established based on the estimated capacities to simulate the stripping processes. The simulated Mo peaks and the experimental data agreed closely with each other, indicating the models and parameters were validated. Further simulations were then performed to find the optimal conditions for achieving high product concentration and short stripping time. When multiple target solutes need to be recovered and purified, an elution chromatography process is commonly used to achieve the goal. In some cases, however, the target solutes have very similar properties (e.g. enantiomers and lanthanides), and the sorbent may therefore have no selectivity for them. To achieve separation, a ligand which can form complexes with the solutes with different equilibrium constants is added into the mobile phase. If the complexed solutes adsorb, the overall selectivity approaches the sorbent selectivity for the complexed solutes at a high ligand concentration. If the complexed solutes do not adsorb, the overall selectivity is approximately the ratio of the sorbent selectivity for the free solutes to the ligand selectivity. Therefore, it will be favorable for the separation if the sorbent and the ligand have opposite affinity sequences for the solutes. A good resolution can be obtained only when the overall selectivity is high and the complexation has comparable strength as the adsorption of free solutes. The mechanisms we understood from the ligand-assisted elution chromatography were used to develop the separation of lanthanides (Ln's), which are critical materials in many high-tech products. Current production of Ln's relies on multiple sequential and parallel solvent extraction steps, which require large amounts of harsh solvents and are environmentally hazardous. In this study, a ligand which can complex with Ln's with significantly different equilibrium constants was added to the mobile phase of chromatography to increase the overall selectivity. The elution process was demonstrated by separations of Pr, Nd and Sm on TiO2 using ethylenediaminetetraacetic acid (EDTA) as the ligand. Both purity and yield for each component were 95% or higher in linear gradient elution and step-wise elution. The ligand-assisted elution chromatography process is simpler than the conventional solvent extraction, because a series of Ln's products can be obtained from a single set of chromatography system. Moreover, this process has less impact on the environment, because the ligand is generally recognized as safe and can be mostly recycled. For large-scale production, the separation of Ln's can be carried out in a continuous mode with step-wise elution to save ligand and to increase sorbent productivity. (Abstract shortened by ProQuest.)
Wang, Purdue University.
Chemical engineering|Materials science
Off-Campus Purdue Users:
To access this dissertation, please log in to our