(208g) Digital Design and Evaluation of Separation Alternatives for the Green Manufacturing of Lomustine | AIChE

(208g) Digital Design and Evaluation of Separation Alternatives for the Green Manufacturing of Lomustine


Casas Orozco, D. - Presenter, Purdue University
Laky, D., Purdue University
Sundarkumar, V., Purdue University
Thompson, D. H., Purdue University
Reklaitis, G., Purdue University
Nagy, Z., Purdue
Murbach De Oliveira, G., Purdue University
Mackey, J., Purdue University
Manufacturing of active pharmaceutical ingredients (APIs) typically involves a series of chemical synthesis and separation steps, which together comprise the so-called drug substance (DS) process. For chemical synthesis, batch or continuous flow reactors are used, whereas cooling/antisolvent crystallization is typically employed to recover pure or nearly pure APIs. Often, there are additional intermediate purification/separation steps performed for various purposes. For instance, undesired impurities, precursors or catalysts that may negatively impact subsequent synthesis or purification steps must be removed prior to those affected operations. Also, concentrations of certain toxic species/solvents are required to be reduced to an acceptable level in the final product formulation.

The unit operations involved in these separation/purification steps typically rely on phase equilibrium, with vapor-liquid equilibrium (VLE) - and liquid-liquid equilibrium (LLE) - based equipment most frequently used [1], [2]. These operations have been long studied, and both first-principles and reduced-order models are available for process design and analysis. With the availability of such models and their efficient numerical implementation, systematic in-silico evaluation and comparison of such separation trains can be performed, and a thorough techno-economic analysis of their design and operating conditions via optimization or enumeration techniques can be achieved.

As a way to make unit operation models for pharmaceutical manufacturing readily accessible, we recently developed the Python-based library PharmaPy [3], which has been successfully used for digital design and analysis of a series of pharmaceutical process case studies [4]–[6]. The object-oriented architecture of PharmaPy has made it possible to develop and maintain a diverse set of DS unit operations in different processing modes, viz. batch, semi-batch and continuous operation. For the present work, one-stage flash batch/semi-batch/continuous vaporization and multi-stage, steady state and dynamic distillation models are evaluated. Moreover, single and multiple-stage LLE models are introduced as part of the PharmaPy library, with differing levels of complexity depending on how the coexisting liquid phases are modelled (detailed thermodynamic description or empirical partition coefficients). With the large number of possible unit operation types, operating modes, and operating conditions it is evident that a digital framework for systematic simulation, analysis and optimization of such alternatives is an essential process engineering tool.

In the present work, digital design of the synthesis-solvent switch section for lomustine production is developed using the PharmaPy library. Lomustine is a high-value drug for the treatment of brain cancer, which has seen a significant price increment in the recent years, creating important patient access problems [7]. The most recent efforts of our team have been in developing synthesis routes that lead to efficient continuous manufacturing of lomustine, by replacing traditional solvents by more environmentally-friendly ones in both the chemical synthesis and the separation/purification steps. The case studies presented show that effective lomustine crystallization requires that solvent switch must be performed after the synthesis steps using either VLE or LLE equipment.

In this work, different end-to-end batch, hybrid and continuous flowsheets were systematically enumerated and analyzed in terms of process economy and green chemistry metrics. To this end, a logical system first generates all the possible reactor – separation technology combinations between continuous plug-flow reactors (PFRs) or batch reactors and their corresponding VLE or LLE batch or continuous separation equipment. Then, PharmaPy flowsheet objects of each alternative process pathway are created and their design and operating conditions determined via a simulation/optimization framework. Significant improvements in terms of process feasibility for downstream processing are achieved when continuous staged distillation was used instead of flash separation, where the inherent thermodynamic limitations of one-stage VLE separations are overcome. LLE also proves to be an attractive option to VLE separation given the negligible lomustine solubility in the extraction solvent used. In terms of economics, continuous processing exhibits the lowest capital investment but at the cost of losses of process materials during start-up times. Thus, campaign length again is shown to be an important decision variable in the digital design.


[1] K. P. Cole et al., “Kilogram-scale prexasertib monolactate monohydrate synthesis under continuous-flow CGMP conditions,” Science (80-. )., vol. 356, no. 6343, pp. 1144–1150, Jun. 2017.

[2] S. Diab, N. Mytis, A. G. Boudouvis, and D. I. Gerogiorgis, “Process modelling, design and technoeconomic Liquid–Liquid Extraction (LLE) optimisation for comparative evaluation of batch vs. continuous pharmaceutical manufacturing of atropine,” Comput. Chem. Eng., vol. 124, pp. 28–42, 2019.

[3] D. Casas-Orozco et al., “PharmaPy: An object-oriented tool for the development of hybrid pharmaceutical flowsheets,” Comput. Chem. Eng., vol. 153, p. 107408, Oct. 2021.

[4] D. Casas-Orozco et al., “Digital Design of a Lomustine Manufacturing Process Using PharmaPy,” 2021 AIChE Annual Meeting. Boston, p. 2, 2021. Paper 317f

[5] D. J. Laky, D. Casas-Orozco, C. D. Laird, G. V. Reklaitis, V. Wang, and Z. K. Nagy, “Determination of probabilistic design spaces in the hybrid manufacture of an active pharmaceutical ingredient using the Python-based framework PharmaPy,” in 14th International Symposium on Process Systems Engineering, 2022, p. 2.

[6] I. Hur, D. Casas-Orozco, and Z. K. Nagy, “Dynamic Flowsheet Simulation and Application of Soft Sensors on an Intensified and Integrated Purification Step for Pharmaceutical Upstream Manufacturing,” 2021 AIChE Annual Meeting. Boston, 2021. Paper 117f

[7] Z. Jaman, T. J. P. Sobreira, A. Mufti, C. R. Ferreira, R. G. Cooks, and D. H. Thompson, “Rapid On-Demand Synthesis of Lomustine under Continuous Flow Conditions,” Org. Process Res. Dev., vol. 23, no. 3, pp. 334–341, 2019.