Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

Model Calibration and Optimization of a Protein Purification Process

Al-Kaisy, Sadiq LU (2015) KET920 20151
Chemical Engineering (M.Sc.Eng.)
Abstract
Four different chromatography models were calibrated to describe the separation of a ternary protein mixture consisting of lysozyme, cytochrome C and ribonuclease A in an ion-exchange chromatography column. The models are based on the same column model, the kinetic dispersive model. Protein adsorption was described by four different adsorption models, the Langmuir model with mobile phase modulators (MPM), the steric mass action (SMA) model, the self-association (SAS) model and the generalized Langmuir (GL) model. The models were calibrated against two kind of experiments, multi-component gradient experiments at low column load and single-component gradient experiments at high column load. The models were also validated against a... (More)
Four different chromatography models were calibrated to describe the separation of a ternary protein mixture consisting of lysozyme, cytochrome C and ribonuclease A in an ion-exchange chromatography column. The models are based on the same column model, the kinetic dispersive model. Protein adsorption was described by four different adsorption models, the Langmuir model with mobile phase modulators (MPM), the steric mass action (SMA) model, the self-association (SAS) model and the generalized Langmuir (GL) model. The models were calibrated against two kind of experiments, multi-component gradient experiments at low column load and single-component gradient experiments at high column load. The models were also validated against a multi-component validation experiment. All the models, especially the Langmuir MPM model, fit the experimental profiles at low column load very well. At high column load only the SAS model and GL model could capture the behavior of the experimental profiles, but even these two models did not fit the experimental profiles so well. The thesis was concluded with an optimization of the protein purification process. Three different objective functions were optimized, productivity, yield and normalized earnings. Optimization was performed with regard to two decision variables, the variables correspond to the amount of proteins loaded and the slope of the salt gradient, and one purity constraint. Maximum productivity was obtained at high column load and steep salt gradient. Maximum yield was obtained at low column load and flat salt gradient. (Less)
Popular Abstract
In this thesis four different models were calibrated to describe the separation of a ternary protein mixture in an ion-exchange chromatography column. The thesis was concluded with an optimization of the separation process.
Please use this url to cite or link to this publication:
author
Al-Kaisy, Sadiq LU
supervisor
organization
course
KET920 20151
year
type
H2 - Master's Degree (Two Years)
subject
keywords
Model calibration, Protein purification, Ion-exchange chromatography, Optimization, chemical engineering, kemiteknik
language
English
id
5470040
date added to LUP
2015-06-10 16:08:41
date last changed
2015-06-10 16:08:41
@misc{5470040,
  abstract     = {{Four different chromatography models were calibrated to describe the separation of a ternary protein mixture consisting of lysozyme, cytochrome C and ribonuclease A in an ion-exchange chromatography column. The models are based on the same column model, the kinetic dispersive model. Protein adsorption was described by four different adsorption models, the Langmuir model with mobile phase modulators (MPM), the steric mass action (SMA) model, the self-association (SAS) model and the generalized Langmuir (GL) model. The models were calibrated against two kind of experiments, multi-component gradient experiments at low column load and single-component gradient experiments at high column load. The models were also validated against a multi-component validation experiment. All the models, especially the Langmuir MPM model, fit the experimental profiles at low column load very well. At high column load only the SAS model and GL model could capture the behavior of the experimental profiles, but even these two models did not fit the experimental profiles so well. The thesis was concluded with an optimization of the protein purification process. Three different objective functions were optimized, productivity, yield and normalized earnings. Optimization was performed with regard to two decision variables, the variables correspond to the amount of proteins loaded and the slope of the salt gradient, and one purity constraint. Maximum productivity was obtained at high column load and steep salt gradient. Maximum yield was obtained at low column load and flat salt gradient.}},
  author       = {{Al-Kaisy, Sadiq}},
  language     = {{eng}},
  note         = {{Student Paper}},
  title        = {{Model Calibration and Optimization of a Protein Purification Process}},
  year         = {{2015}},
}