Going against years of chemical practice, process chemists at the pharmaceutical company Merck have shown that enamine substrates for the enantioselective synthesis of ß-amino acids need not be protected at the amino group during asymmetric hydrogenation.
The findings dramatically reduce the number of steps required in the manufacture of enantiopure ß-amino acids from enamine substrates. Merck has practiced the reaction at the hundred-kilogram scale to prepare an active pharmaceutical ingredient for a drug in Phase III clinical trials.
Chemists in Mercks departments of process research and analytical researchincluding Yi Hsiao, Nelo R. Rivera, Thorsten Rosner, Shane W. Krska, Yongkui Sun, and Joseph D. Armstrong IIIdiscovered the reaction and developed it for commercial-scale application in collaboration with chemists Felix Spindler and Christophe Malan of Solvias AG, based in Basel, Switzerland. The team disclosed its findings last month in presentations before the Organic Chemistry Division of the ACS national meeting in Philadelphia and in a journal communication [J. Am. Chem. Soc., 126, 9918 (2004)].
Asymmetric hydrogenation of acyl-protected enamine substrates to make b-amino acids is widely practiced. According to Hsiao, who is a research fellow in process research at Merck, Nobel Laureate Ryoji Noyori carried out the first reaction using an acyl-protected enamine 14 years ago. Since then, everyone just automatically put on the acyl group, but so far no manufacturer has been able to scale up that reaction because of the difficulty in preparing the substrate, he tells C&EN.
According to the accepted mechanism for acyl-protected enamines, the carbonyl group of the acyl moiety contributes one set of electrons to the bidentate binding to metal catalyst that leads to asymmetric induction. But the need to protect and then deprotect functional groups is a drag on commercial-scale synthesis. Not only does it add steps, but it also poses the risk that the sometimes harsh conditions of deprotection could destroy the stereochemistry of chiral molecules, explains Armstrong, who is a director of process research at Merck. People have been trying to use unprotected enamines for a long time, but conversions and enantiomeric excesses were very, very low, he adds.
What has led to Mercks success is the ligand called Josiphos. This ferrocenyl phosphine ligand was discovered by Spindler and Antonio Togni in what was then the central research laboratories of Ciba-Geigy in the course of developing a commercial route to the herbicide (S)-metolachlor (C&EN, June 14, page 47). Mechanistic studies by the Merck chemists indicate that bidentate binding to the metal catalyst may also occur with unprotected enamines. In this case, however, the p or nonbonding electrons come from a carbon-nitrogen double bond formed when the enamine tautomerizes to an imine.
The Merck chemists began developing an enantioselective synthesis of ß-amino acids from enamines in January 2002. The target was a preclinical candidate being developed for the treatment of type 2 diabetes. As the candidate progressed in the development pipeline, it was denoted MK-0431.
In June 2002, the Merck chemists discovered that asymmetric hydrogenation to b-amino acids works with bare enamines. Two months later, they began collaborating with Solvias. And by years end, an optimized manufacturing process was in place. To date, the route has yielded more than 1 ton of MK-0431, which currently is in Phase III clinical trials.
Time was of the essence. Although the Merck team could have figured out all the chemistry by itself, it turned to Solvias to accelerate development. They were the experts with this ligand, and they had access to many other ligands of that type that we wanted to screen. No other company has those ligands. That was the key driver for going with Solvias, Armstrong says.
We also knew that they could manufacture this ligand on a large scale. That was critical, Armstrong continues. Solvias has a proven track record of manufacturing highly enantiopure ligands. Thats something that a lot of people say they can do, but not many people are actually doing it.
Working in concert, the Merck and Solvias teams optimized the reaction using the ligands in the Solvias library. But Spindler believed that Solvias could design a better ligand than what was already available.
Spindler has been active in the business of asymmetric catalytic hydrogenations for 20 years. Catalyst selection being more of an art than a science, his intuition for what will work is finely honed. Given the substrates that Merck wanted to try, he quickly saw that a further electronically tuned Josiphos would be even more effective.
The result of that hunch is a new ligand that has two 4-trifluoromethylphenyl groups instead of the two phenyl groups in the regular Josiphos. Reactions with this ligand are more robust, requiring lower loadings of catalyst, Spindler tells C&EN. In addition, the ligand enables the use of enamine ester substrates as well as enamine amides. Josiphos is efficient only with enamine amides. The ability to use enamine esters or amides gives great flexibility in the kinds of molecules that Merck can prepare, he adds.
To prepare MK-0431, Merck hydrogenates the corresponding enamine amide with the regular Josiphos catalyst. Reactions are carried out in 160-kg batches at a catalyst loading of 0.3 mol %. Conversion is 95%, and the crude product is better than 93.5% enantiopure.
The best part of the project was that the Merck team had already developed two other commercial routes to MK-0431, Armstrong says. It was risky to pursue the asymmetric hydrogenation of the enamine amide because it had never been done before. I think we did the right thing, which is, Lets take on a reaction that is unknown and unprecedented and bring it to fruition.