Introduction: There is a theoretical foundation that suggests that temperature directly correlates with separability and resolution in all forms of chromatography. We researched the topic further, both in fundamental fluid mechanics and heat transfer as well as the potential for improving separation efficiency in large-scale analytical chromatography setups such as Liquid Chromatography tandem Mass Spectrometry (LC-MS) by increasing the temperature of such systems, and derived a robust heat-transfer model for a potential heating system which could raise a chromatography column's temperature profile uniformly or linearly.
Materials and
Methods: Our understanding of separation efficiency is based primarily on kinetic parameters, diffusion parameters, and the van Deemter equation defining the theoretical plate height in a specified chromatographic run. After researching this basis, we then developed a model for a multiple-input, multiple-output heating and sensing array along a column and defined its heat transfer properties basing in Newton and Fourier's laws in state-space form. Finally, we developed a control mechanism, and after suspecting various difficulties with a more straightforward Proportional Integral Derivative (PID) controller, we finalized on a Linear Quadratic Gaussian (LQG) regulator.
Results, Conclusions, and Discussions: Our model responded very effectively and was able to achieve temperature gradients within a range from room-temperature up to 80˚C with relatively fast response times of under 1 minute. The regulation mechanism achieved critical damping without overshoot, which is important in cases where we heat close to the stability point of a compound, and this was after incorporating variable noise as well. We also noticed a key interplay between our achievable response time, the capacity of our device to heat the system, and the base geometry and measurements of our system and insulation.
Acknowledgements (Optional): Timothy Lannin was an extremely helpful and guiding mentor as I did this research, and we've pitched this project now to senior engineering capstone teams to help us build the physical prototype and test our model.
