Design and dynamics of large-scale chromatographic processes

Sheng T Hsu, Purdue University

Abstract

Since large-scale purification columns become more expensive as their size increases, it may be inevitable to use a column of such length that high resolution is not ensured. In these cases, an enhanced separating effect can be obtained as if the column length were increased by re-introducing the eluate band containing the desired component eluted from the column back into the column inlet. A comprehensive study of the design of recycle chromatography has been conducted. A general nonlinear multicomponent rate model including considerations of axial dispersion, external film mass transfer, intra-particle diffusion and Langmuir type isotherm was developed. Ordinary differential equations were set up using finite element method in the bulk phase and orthogonal collocation method in the particle phase. This dynamic problem has been recently solved in an equation-based flow sheeting dynamic simulator, SPEEDUP$\sp{\rm TM}.$ The solving algorithm behind the simulator is the symbolic computation and DAE algorithm originated from Petzold in 1982. The robustness of the chromatography model has been demonstrated. Direct recycle chromatography consists of three units including a mixer, a column and a splitter. It has been optimized in its column length to achieve separation with recycle capacity constrain. The most important parameter in the design of recycle chromatography is the recycle capacity. The optimization of column length and sample load have to be based on the recycle capacity of a given system. Recycle capacity is relatively insensitive to the sample load. Therefore, the design of recycle chromatography can be optimized with column length first. After optimizing the column length, the sample can then be overloaded to the maximum limit within the restriction of recycle capacity. When optimizing the column length, two approaches, namely, either to keep flow velocity constant or to maintain pressure drop constant, can be applied. Displacement effect has been shown to be able to increase recycle capacity, that is, to facilitate the column efficiency. Therefore, higher production can be expected from overloaded condition. Rate model combined with Langmuir isotherm has provided a valuable tool to understand the dynamics of chromatographic process and to aid the scale-up of liquid chromatography. By setting the solution scheme in a processor simulator, it can be combined with other existing models to simulate a more complicated bioprocess; for example, coupling chromatographic columns with bioreactor for in situ product removal.

Degree

Ph.D.

Advisors

Tsao, Purdue University.

Subject Area

Chemical engineering

COinS