Flow chemistry investigations into the synthesis of donepezil, sevoflurane and Suzuki-Miyaura cross couplings

Abstract

This dissertation investigates the implementation of flow chemistry in the synthesis of active pharmaceutical ingredients (API) and translation of organic transformations, aiming to address challenges such as low process efficiency and scalability that are associated with batch manufacturing. In this research, we developed a three-step flow-based synthetic pathway for the anti-Alzheimer’s disease agent donepezil. The first step involved a selective reduction of ethyl-1-benzylpiperidine-4-carboxylate to 1-benzylpiperidine-4-carboxylate. The reduction was optimized using design of experiments (DoE) to afford full conversion of the carboxylate to form 95% of 1-benzylpiperidine-4-carbaldehyde in the product mixture with minimal over-reduction to the unwanted (1-benzylpiperidine-4-yl)methanol (5%) at -60 °C in just 2 min, a huge improvement from the batch counterpart which required 150 min at -78 °C for the same outcome. Thereafter, an aldol condensation to afford (E)-2-[(1-benzylpiperidine-4-yl)methylene]-5,6-dimethoxy-2,3-dihydro-1H-inden-1-one was optimized systematically using DoE to afford the corresponding alkene with a quantified yield of 95%. A final hydrogenation step was optimized by carefully controlling the residence time over a Pd/C packed catalyst, ensuring 90% quantified yield of donepezil after only 13.4 min while the batch hydrogenation required the use of a poisoning catalyst and a long reaction time of 240 min to achieve close to quantitative yield of the donepezil. The sequential integration of the three steps enabled the staggered continuous production of donepezil with a throughput of approximately 2.50 g h-1 and an overall isolated yield of 50%. Our research group has also been involved in the fabrication of reactors to mitigate flow chemistry challenges such as handling solids. As such, we demonstrated the application of a bespoke screw reactor for handling different complexities of solids formed during in-line HCl salt formations of amines and the recrystallizations of several α,β-unsaturated carbonyl compounds. Overall, twenty-one salt formations and nine recrystallizations were effectively conducted and of all the reactions conducted, less than 3% resulted in blockages. We also demonstrated the use of diazonium salts as coupling partners in the Suzuki-Miyaura cross coupling reaction via a heterogeneous flow set-up using 10% Pd/C as the catalyst. The 4-methylbenzenediazonium tetrafluoroborate salt and the catalyst were effectively introduced into the system as a slurry hence avoiding the dissolution of the diazonium salt in the inhibiting acetonitrile solvent. The yield of the reaction was promoted by higher equivalences of the diazonium salt (4.00 eq.), higher temperatures (80 °C) and residence times closer to 9.5 min to quantify a maximum yield of 4-methyl-1,1’-biphenyl at 76%. Lastly, we demonstrated initial proof of concept for the continuous photochemical chlorination of 1,1,1,3,3,3-hexafluoro-2-methoxypropane, an intermediate of sevoflurane, an anaesthetic. The chlorine gas used in the reaction was generated in situ from the reaction of aqueous NaOCl and HCl with yields ranging between 71 - 79%. The photochemical reaction was promoted by increasing the amount of chlorine gas at 80 °C but this was also accompanied by an increase in unidentified by-products.