Please use this identifier to cite or link to this item: http://scholarbank.nus.edu.sg/handle/10635/79731
Title: ADVANCED CONTROL STRATEGIES FOR PHARMACEUTICAL ANTISOLVENT CRYSTALLIZATION PROCESSES
Authors: VAMSI KRISHNA KAMARAJU
Keywords: Antisolvent crystallization, Concentration control, Adaptive C-control, Neighboring extremal, Model Predictive Control, Parametric uncertainty
Issue Date: 23-Jan-2014
Source: VAMSI KRISHNA KAMARAJU (2014-01-23). ADVANCED CONTROL STRATEGIES FOR PHARMACEUTICAL ANTISOLVENT CRYSTALLIZATION PROCESSES. ScholarBank@NUS Repository.
Abstract: During the manufacture of Active Pharmaceutical Ingredients (APIs), the crystal size distribution (CSD) obtained at the end of the crystallization process has significant effect on the efficiency of other downstream operations like filtration, drying, milling, and so forth. Besides, the product CSD also affects the efficacy of the formulated drug. Hence, during the manufacture of various APIs, tighter control for robust and optimal operation of crystallization processes is of grave importance. In this thesis, a laboratory-scale isothermal, seeded, antisolvent semi-batch crystallization process for paracetamol in acetone-water mixture is considered via simulation studies, where water acts as the antisolvent. Owing to the advancements in sensor technology and introduction of Process Analytical Technology (PAT) tools, direct design approaches like concentration control (C-control), which uses concentration or supersaturation measurements feedback, are developed for the control of crystallization processes. Although these strategies were found to be more robust due to their closed loop nature, they are often operated in a sub-optimal fashion and often fail in scenarios where there are shifts in solubility data. To alleviate this shortcoming, an integrated approach based on a newly developed data-based modeling framework is proposed to enable the determination of constant relative supersaturation setpoint for optimal operation of C-control strategy. Simulation results show that the integration of the developed framework in conjunction with the existing C-control strategy gives improved operation compared to its conventional counterpart. Recognizing the advantages of integrating additional PAT tools for particle counts measurements, a new approach for adaptive C-control strategy that updates the relative supersaturation setpoint based on real-time optimization scheme using the proposed modeling framework is developed. The developed adaptive C-control strategy produces significantly better performance than the direct design C-control strategy. On the other hand, understanding the capabilities of measurement based optimization schemes as alternative approaches for real-time optimal control that do not require the implementation of rigorous online re-optimization techniques, the necessary conditions of optimality (NCO) tracking based control for semi-batch antisolvent crystallization processes is developed in this thesis. The NCO tracking control has displayed better robustness towards shifts in solubility data compared to direct design C-control strategy. Furthermore, in order to address the shortcomings of the traditional design of neighboring extremal (NE) controller that tracks the interior arcs during the implementation of NCO-tracking based control, a reformulated NE control feedback law is developed. Simulation results show that NCO tracking control using the reformulated NE controller design minimizes the loss in optimality to a much greater extent.
URI: http://scholarbank.nus.edu.sg/handle/10635/79731
Appears in Collections:Ph.D Theses (Open)

Show full item record
Files in This Item:
File Description SizeFormatAccess SettingsVersion 
KamarajuVK.pdf1.8 MBAdobe PDF

OPEN

NoneView/Download

Page view(s)

114
checked on Feb 17, 2018

Download(s)

99
checked on Feb 17, 2018

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.