MODELING AS A PREDICTIVE TOOL FOR PERFORMANCE GAINS IN LIQUID CHROMATOGRAPHY

Date

2023-7-25

Author

Şencan, Simge

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

75
views

94
downloads

Cite This

The essential purpose of this research is to investigate and compare peak characterization and parameter estimation methods applied to experimental data to establish their effectiveness for case studies in analytical chromatography. Chromatography is a powerful separation method used to separate complex molecules. For separations, the chromatographic efficiency is a crucial parameter affecting resolution. The efficiency can be affected by many experimental variables, and is the consequence of the phenomenon of band broadening, also called dispersion. The theoretical plate number is one of the important measures of band broadening and calculating this parameter correctly is a crucial step in estimating chromatographic efficiency accurately. Robust and accurate results can be obtained for Gaussian peaks with the graphical theoretical plate number calculation methods. However, the accuracy of this calculation method remains poor for peaks that have tailing or fronting since it does not consider the non-Gaussian shape. In this thesis study, experimental chromatograms and their associated peaks were obtained for a range of settings and variables, and the raw data was used to obtain chromatographic parameters using MATLAB codes designed to implement moment analysis. These parameters, such as peak variance, theoretical plate number and plate height, provide illuminating information in terms of separation efficiency. The results from different theoretical plate number calculation methods were compared in van Deemter curve for different molecules, columns, and temperatures. Thus, accurate ways of efficiency calculation for both Gaussian and non-Gaussian peaks were investigated. Although accurate and consistent results can be obtained for Gaussian peaks when graphical methods are used, it has been observed that the results for non-Gaussian peaks, that is, peaks with tailing, deviate from reality. In the case of using graphical methods, we cannot see the effect of tailing or fronting in the results because the calculation is made by treating non-ideal peaks as if they are ideal. Also, a database for chromatographic parameters of different molecules in various conditions was created and by using this database an open- source prediction tool that can give results quickly with a few required parameters were developed. It is aimed to obtain a useful open-source tool that can predict the analysis result without performing chromatographic analysis, with further development of the database and MATLAB code as a future work.

Subject Keywords

Chromatographic Efficiency, Peak Analysis, van Deemter, non-Gaussian peaks, Kromatografik Verim, Pik analizi, Gauss olmayan pikler

URI

https://hdl.handle.net/11511/105100

Collections

Graduate School of Natural and Applied Sciences, Thesis