Development of novel palladium catalysed N-arylation and diastereoselective Suzuki reactions for the large-scale syntheses of pharmaceutical intermediates

Abstract

Palladium and copper catalysed N-arylation reactions for the synthesis of 3-bromo-2,5-difluoroaniline as an early stage intermediate for the preparation of KRAS inhibitor AZD4625

Use of palladium catalysis for the large-scale manufacturing of pharmaceuticals is both environmentally and economically challenging[1]. 3-Bromo-2,5-difluoroaniline.HCl 4 was identified as a key early intermediate for the synthesis of AZD4625, a KRASG12c inhibitor under investigation for the treatment of solid tumours. Aniline 4 was initially prepared via Pd(0) catalysed amination of dibromo-difluoro arene 2 with benzophenone imine 1, using high palladium catalyst loadings of 1.5 mol%. As significant quantities of AZD4625 were required for early-stage clinical trials, the use of the original process to prepare 1500 kg of aniline 4 was predicted to require 60 kg of Pd(dba)2 at an unacceptable cost of ~$1M. Consequently, the first research project in this thesis describes how new conditions were developed that allowed ~1400 kg of aryl amine 4 to be prepared using only 0.5 mol% Pd(dba)2 catalyst, corresponding to an overall saving of ~$820,000 (Scheme 1). Kinetic analysis of the Pd(0)-catalysed reaction revealed that the rate of the base-mediated imine deprotonation step by the heterogenous K3PO4 was close to the turnover limiting step of the reaction, with mass transfer limitations meaning that high loadings of K3PO4 were required. Further investigations revealed that catalyst poisoning/inhibition was occurring, most likely from a dimeric species found to be the resting state for the catalytic cycle. The catalyst decomposition via an unknown pathway prevented catalyst loading levels being reduced further.

Alternative copper catalysed conditions for this N-arylation reaction were also developed that provided a potentially cheaper but less efficient process. The copper conditions might be useful if the costs of the palladium catalysed process became prohibitive (Scheme 2). The potential of both N-arylation processes were compared, with screening results indicating that the palladium catalysed conditions had broader applicability for N-arylation of simple aromatics, whilst copper catalysts afforded some advantages for N-arylation of some drug-like haloaryl substrates.

A diastereoselective Suzuki reaction for synthesis of the biologically active atropisomer of a key late stage biaryl intermediate used to prepare KRAS inhibitor AZD4747

AZD4747 is a KRASG12c inhibitor under investigation for the treatment of secondary brain cancers. Its structure contains a sterically hindered tetrasubstituted biaryl motif which has restricted rotation, giving rise to atropisomers. This biaryl functionaility was originally installed using a Suzuki-Miyaura cross coupling (SMC) reaction of (R)-7 with boronic acid 8, using an expensive palladium catalyst at high loading to form a ~45:55 mixture of diastereomers (R,Sa)-10 and (R,Ra)-11. The highest yield in this Suzuki reaction was found to occur prior to complete consumption of starting material (R)-7, with extensive decomposition of (R,Sa)-10/(R,Ra)-11 occurring when extended reaction times were used to drive the cross-coupling reaction to completion. The second research project described in this thesis reports investigations into this diastereoselctive Suzuki cross-coupling reaction, that determined which side reactions of (R,Sa)-10/11 were occurring and elucidated the structures of side products that were formed. These investigations led to new conditions being devised that provided better, more reproducible yields of (R,Sa)-10/(R,Ra)-11 and used more readily available Pd(dba)2 as the catalyst precursor. The ~45:55 mixture of diastereomeric atropisomers (R,Sa)10/(R,Ra)-11 formed in this cross-coupling reaction could be separated using a classical salt resolution process to provide the biologically active atropisomer (R,Sa)-10 in high dr, although in <50% overall yield. Investigations into using asymmetric Suzuki cross-coupling conditions to selectively produce (R,Sa)-10 as a major diastereomer were successful, with use of a bespoke (R,R)-ByrnePhos 9 as a ligand in the cross-coupling reaction, resulting in an increase in diastereoselectivity to produce the desired (R,Sa)-10 atropoisomer in 72:38 dr (Scheme 3). O-Protection of the phenol group of the parent boronic acid 8 gave further diastereoselectivity improvements, with the cross-coupling reaction of an O-benzyl protected version of 8 giving a 80:20 dr for formation of the desired biaryl atropoisomer (R,Sa)-10. Notably, a mismatched effect was observed when (R,R)-ByrnePhos 9 was used as an enantiomeric ligand in the cross-coupling reactions of MOM/Bn protected versions of boronic acid 8, producing the undesired atropisomer (R,Ra)-11 in >5:95 dr.

Date of Award	19 Feb 2025
Original language	English
Awarding Institution	University of Bath
Sponsors	AstraZeneca UK Ltd
Supervisor	Steven Bull (Supervisor) & James Taylor (Supervisor)