Plantwide technoeconomic analysis and separation solvent selection for continuous pharmaceutical manufacturing: Ibuprofen, artemisinin, and diphenhydramine

Continuous pharmaceutical manufacturing (CPM) has been recognized as a new production paradigm capable of achieving lower costs, higher yields and productivity, enhanced heat transfer and mixing, smaller equipment, and more economical energy and material usage. These potential benefits show significant advantages over the currently implemented batch methods which suffer from significant intermediate storage, high quantities of waste and intensive labour requirements. However, CPM has yet to be widely adopted by the pharmaceutical industry due to significant investments in current batch infrastructures. In order to facilitate this transition, the benefits of producing active pharmaceutical ingredients (APIs) in continuous mode must be demonstrated.

Process modeling and simulation is a valid approach for screening candidate CPM processes prior to significant financial investments in experimental and pilot plant studies. Theoretical modeling and simulation of candidate CPM process for the production of globally marketed APIs allows screening of continuous flowsheets to highlight the material reduction and cost savings benefits achievable with respect to current batch manufacturing methods.

Herein, we develop process models for 100 kg per annum CPM plants of three globally marketed APIs: the popular analgaesic, ibuprofen (IBU), the important antimalarial, artemisinin (ART) and the first-generation antihistamine, diphenhydramine (DPH). Continuous flow syntheses for all three APIs have been demonstrated on the labscale in the literature, with subsequent batchwise purification processes. In this work, we investigate potential continuous separation options to develop fully integrated end-to-end continuous processes for these APIs prior to downstream processing. Continuous liquid-liquid extraction (LLE) is investigated for the CPM of IBU and DPH, whilst continuous antisolvent cooling crystallisation is investigated for the CPM of ART.

Systematic separation solvent selection is conducted based upon the conceptual modeling of these continuous processes. This requires thermodynamic modeling of multicomponent mixture phase equilibria via the popular UNIFAC and NRTL models. Candidate separation solvent lists for IBU (n-hexane, n-heptane, cyclohexane, methylcyclohexane, toluene, and isooctane as LLE solvents), ART (ethanol, ethyl acetate, acetonitrile, and methyl ethyl ketone as crystallization antisolvents), and DPH (n-hexane, n-heptane, cyclohexane, methylcyclohexane, dichloromethane, and chloroform as LLE solvents) are evaluated based upon attainable API recoveries, material efficiencies, regulatory classification and total costs as a function of operating temperature and solvent usage. Capital (CapEx) and operating (OpEx) expenditures and total costs (expressed as the net present value, NPV) are compared for all continuous separation options relative to the batch separation method to determine the best CPM route. Total costs savings are achievable for all API cases when implementing CPM routes, as well as benefits of material efficiency and reduced process complexity. The sensitivity of the total costs is investigated by considering the effect of varying interest rate on NPV, return on investment and payback period are also considered. The efforts presented in this work demonstrate the potential of all three APIs for CPM implementation and the advantages of process modeling and simulation for CPM investigation.