Process modelling, simulation and technoeconomic optimisation for continuous pharmaceutical manufacturing of (S)-warfarin

Abstract

Continuous pharmaceutical manufacturing (CPM) has the potential to attain several technical and operational economic benefits compared to the currently prevalent batch production paradigm. Despite the variety of demonstrated continuous flow syntheses of active pharmaceutical ingredients (APIs), the limited number of cost-effective continuous separations is a bottleneck to end-to-end CPM. Establishing promising APIs for integrated CPM is paramount. (S)-Warfarin is an anticoagulant API whose continuous flow synthesis features a single reaction with high enantiomeric selectivity followed by liquid-liquid extraction (LLE). This work describes the steady-state process modelling and technoeconomic optimisation for the upstream CPM of (S)-warfarin, implementing reactor design and LLE solvent comparison for purification. Ethyl acetate, isopropyl acetate, isobutyl acetate, 1-heptanol, 1-octanol and heptane are candidate LLE solvents. Reported reaction conversions and computed LLE efficiencies allow mass balance calculation and total cost estimation to establish promising LLE solvents. The nonlinear optimisation problem is formulated for total cost minimisation. Liquid-liquid phase equilibria, API phase compositions and solubilities for LLE design are implemented via surrogate polynomials based on extensive UNIFAC modelling; API recovery rates are calculated via detailed mass transfer correlations. The methodology used here screens optimum process configurations to achieve a technoeconomically optimal design for a continuous (S)-warfarin manufacturing plant.

Process modelling, design and technoeconomic Liquid–Liquid Extraction (LLE) optimisation for comparative evaluation of batch vs. continuous pharmaceutical manufacturing of atropine

“…Implementation of combined experimental and modelling approaches towards integrated LLE design in the literature for pharmaceutical purifications and separations demonstrate the utility of theoretical methods in establishing optimal design and operating parameters (Drageset and Bjørsvik, 2016;Monbaliu et al, 2016;Weeranoppanant et al, 2017). Mathematical optimisation can be used to identify cost optimal process designs for pharmaceutical manufacturing campaigns (Gross and Roosen, 1998;Patrascu and Barton, 2018); plantwide modelling and optimisation with a focus on optimal continuous separation process design towards total cost minimisation have been implemented by our group for numerous APIs (Jolliffe and Gerogiorgis, 2017a,b;Diab and Gerogiorgis, 2018b). Elucidating cost-optimal designs for atropine CPM will further aid process development for this societally-important API.…”

Design Space Identification and Visualization for Continuous Pharmaceutical Manufacturing

“…Several candidate separation solvents are compared for continuous LLE: ethyl acetate (EtOAc), isopropyl acetate (iPrOAc) and isobutyl acetate (iBuOAc). The CPM flowsheet for warfarin is shown in Figure 14 [68]. The continuous synthesis of (S)-warfarin was demonstrated by Porta et al (2015), featuring the nucleophilic addition of 4-hydroxy-coumarin to benzalacetone in the presence of TFA and a chiral amine catalyst in 1,4-dioxane [67].…”

“…Several candidate separation solvents are compared for continuous LLE: ethyl acetate (EtOAc), isopropyl acetate (iPrOAc) and isobutyl acetate (iBuOAc). The CPM flowsheet for warfarin is shown in Figure 14 [68].…”

Design Space Investigation for Development of Continuous Flow Syntheses of Active Pharmaceutical Ingredients

“…Warfarin (anticoagulant): one reaction followed by continuous LLE with EtOAc, isopropyl acetate (iPrOAc) or isobutyl acetate (iBuOAc) (Diab and Gerogiorgis, 2018).…”

“…For ibuprofen, the different separation options (LLE solvent = {nHex, PhMe}) give similar results and thus the LLE solvent with the lower environmental/EHS impact (i.e., PhMe) is preferable (Jolliffe and Gerogiorgis, 2016). Similarly, for warfarin and atropine, each considered LLE solvent performs similarly, however, each also have similar EHS characteristics; solvent selection should thus be informed by subsequent crystallisation process design (Diab and Gerogiorgis, 2018;. For artemisinin and diphenhydramine, plantwide technoeconomic performance varies more drastically with separation solvent choice; for artemisinin, EtOH as antisolvent allows for lower costs and is more environmentally friendly that EtOAc (Jolliffe and Gerogiorgis, 2016), and for diphenhydramine, nHep has both poorer EHS characteristics than either CyHex or MeCyHex as well as incurring higher costs (Diab and Gerogiorgis, 2017).…”