Tirapelle, M; Besenhard, M; Mazzei, Luca; Zhou, Jinsheng; Hartzell, Scott; Sorensen, Eva; (2023) Towards a digital twin for analytical HPLC. In: (Proceedings) Foundations of Computer Aided Process Operations / Chemical Process Control. (In press).

Preview

Text Article.pdf - Accepted Version Download (270kB) | Preview

Abstract

Digital twins for industrial process development are quickly gaining popularity in the pharmaceutical industry as an effective alternative to expensive and time-consuming physical experiments. This work describes the digital model element of a digital twin of High-Performance Liquid Chromatography (HPLC). The model is based on a mechanistic model implemented in gPROMS ModelBuilder and integrated into the MATLAB environment. Unlike other models reported in the literature, our model comprises a more accurate prediction of the injection profile and can predict the elution behaviour for a wide range of HPLC conditions given a reduced number of experiments. The model is compared against experimental data performed to separate a mixture of eight small drug molecules on a C18 column, in gradient elution mode, and under nine different operative conditions (i.e. 3 temperatures × 3 solvent gradient). We will show that by considering only two isotherm parameters for each molecule, the digital model can accurately predict the retention behaviour of the eight analytes. Furthermore, it facilitates HPLC in-silico method development, showcased here via method time minimization through a dynamic solvent strength gradient. The proposed model is intended to be integrated into a digital twin architecture for offline decision support and real-time optimization.

Download activity - last month

Export as CSVXMLJSON

Download activity - last 12 months

Export as XMLCSVJSON

Downloads by country - last 12 months

Export as XMLCSVJSON

Archive Staff Only

View Item