ABSTRACT
Scalable alternate end-game strategies for the synthesis of the anti-COVID drug molecule Nirmatrelvir (1, PF-07321332) have been described. The first involves a direct synthesis of 1 via amidation of the carboxylic acid 7 (suitably activated as a mixed anhydride with either pivaloyl chloride or T3P) with the amino-nitrile 10·HCl. T3P was found to be a more practical choice since the reagent promoted efficient and concomitant dehydration of the amide impurity 9 (derived from the amino-amide contaminant 8·HCl invariably present in 10·HCl) to 1. This observation allowed for the development of the second strategy, namely a continuous flow synthesis of 1 from 9 mediated by T3P. Under optimized conditions, this conversion could be achieved within 30 min in flow as opposed to 12–16 h in a traditional batch process. The final API had quality attributes comparable to those obtained in conventional flask processes.