DOI: 10.1002/chem.201404896
Communication
& Asymmetric Catalysis
Rhodium-Catalyzed Arylative Cyclization for the Enantioselective Synthesis of (Trifluoromethyl)cyclobutanols Thomas Johnson, Ken-Loon Choo, and Mark Lautens*[a] Abstract: A rhodium-catalyzed cyclization of 1-(trifluoromethyl)-4-alkyn-1-ones with arylboronic acids is described to yield a novel class of small rings: (trifluoromethyl)cyclobutanols bearing an exocyclic double bond. The use of a rhodium/chiral diene complex allowed the reaction to proceed under mild conditions, often with high enantioselectivity. An X-ray crystal structure was obtained confirming the formation of the four-membered ring products.
Cyclobutanes are drawing attention as core structures principally due to their strained nature. Although less commonly encountered than their smaller cyclopropane siblings, they show comparable ring strain (26.5 kcal mol 1 vs. 27.5 kcal mol 1, respectively).[1] This property makes them particularly suitable for ring-opening and ring-expansion reactions to elaborate more complex molecules.[2a] Indeed, four-membered rings can serve both as synthetic intermediates, and as substrates in various metal-catalyzed reactions.[2] They are also found in a variety of natural products.[3] In addition, cyclobutanes are sought after in medicinal chemistry as atypical scaffolds, allowing the coverage of unexplored chemical space.[4] Although different methods exist for the synthesis of four-membered rings,[5] most notably [2+2] cycloadditions with ketenes and photochemical [2+2] cycloadditions between alkenes, there are fewer asymmetric variants and their scope is relatively limited.[6] Methods to synthesize new classes of cyclobutanes in an asymmetric fashion are thus desirable. As part of our ongoing research on rhodium-catalyzed reactions, we sought to develop an enantioselective synthesis of functionalized cyclobutanes. The strategy described herein relies on the arylative cyclization of alkynones (or alkynals) with boronic acids (Scheme 1). This reaction, first reported in 2005 by Hayashi[7d] and Murakami,[7e] has since been developed for the synthesis of various cyclic molecules.[7] Enantioselective variants using chiral diene or phosphine ligands have also been reported.[7b, d, 15e] In the process, a substrate possessing appropriately located carbonyl and alkyne functional groups reacts with a boronic acid under Rh catalysis to generate prod[a] T. Johnson, K.-L. Choo, Prof. Dr. M. Lautens Department of Chemistry University of Toronto 80 St. George Street, Toronto, ON, M5S 3H6 (Canada) E-mail: [email protected] Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/chem.201404896. Chem. Eur. J. 2014, 20, 14194 – 14197
Scheme 1. Catalytic cycle of the rhodium-catalyzed arylative cyclization of alkynones with boronic acids R1 = H, alkyl, aryl, CF3.[7d]
ucts bearing an exocyclic alkene. Two new carbon–carbon bonds and one chiral center are formed in the process, which typically affords five- or six-membered rings (Scheme 1). To the best of our knowledge, there are two reports on the use of such cyclizations for the formation of four-membered rings, neither being enantioselective.[7c, 8] In the course of our initial investigations, we discovered that trifluoromethyl ketone 1 a reacted with phenylboronic acid to form cyclobutanol 2 a in 56 % yield under conditions previously reported by our group (entry 1, Table 1).[9] The process generates a tertiary trifluoromethylcarbinol, a motif found in different therapeutic agents.[10] With the racemic version established, a ligand screen was performed in order to achieve an enantioselective reaction. We turned our attention to chiral dienes, which have been shown to promote a variety of rhodium-catalyzed reactions, including cyclization of alkynones with boronic acids.[7d, 9] A range of chiral dienes have been reported with different substitution patterns and symmetry elements (i.e., most often C1 or C2).[11] While L1, L2, and L3 performed poorly, L4[9] gave the desired product in 60 % yield and with a promising 74 % ee. A solvent screen with different alcohols was undertaken, but did not lead to any improvement in yield; the sideproduct arising from protodemetalation of the first intermediate (see Scheme 1) being always generated in 15–20 % yield. Ligand L4 belongs to a class of ligands that can easily be tuned by modification of the aromatic ring via Suzuki crosscoupling from a common vinyl triflate intermediate.[12] This flexibility allowed us to further explore ligands similar to L4 of differing electronic and steric characters (Table 2).
14194
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Communication Table 1. Evaluation of ligand classes for the arylative cyclization of 1 a with phenylboronic acid.[a]
Entry 1 2 3 4 5
Rh source[b]
Ligand
[{Rh(cod)Cl}2] [{Rh(coe)2Cl}2] [{Rh(coe)2Cl}2] [{Rh(coe)2Cl}2] [{Rh(coe)2Cl}2]
– L1 L2 L3 L4
Yield [%][c] [e]
56
Communication
& Asymmetric Catalysis
Rhodium-Catalyzed Arylative Cyclization for the Enantioselective Synthesis of (Trifluoromethyl)cyclobutanols Thomas Johnson, Ken-Loon Choo, and Mark Lautens*[a] Abstract: A rhodium-catalyzed cyclization of 1-(trifluoromethyl)-4-alkyn-1-ones with arylboronic acids is described to yield a novel class of small rings: (trifluoromethyl)cyclobutanols bearing an exocyclic double bond. The use of a rhodium/chiral diene complex allowed the reaction to proceed under mild conditions, often with high enantioselectivity. An X-ray crystal structure was obtained confirming the formation of the four-membered ring products.
Cyclobutanes are drawing attention as core structures principally due to their strained nature. Although less commonly encountered than their smaller cyclopropane siblings, they show comparable ring strain (26.5 kcal mol 1 vs. 27.5 kcal mol 1, respectively).[1] This property makes them particularly suitable for ring-opening and ring-expansion reactions to elaborate more complex molecules.[2a] Indeed, four-membered rings can serve both as synthetic intermediates, and as substrates in various metal-catalyzed reactions.[2] They are also found in a variety of natural products.[3] In addition, cyclobutanes are sought after in medicinal chemistry as atypical scaffolds, allowing the coverage of unexplored chemical space.[4] Although different methods exist for the synthesis of four-membered rings,[5] most notably [2+2] cycloadditions with ketenes and photochemical [2+2] cycloadditions between alkenes, there are fewer asymmetric variants and their scope is relatively limited.[6] Methods to synthesize new classes of cyclobutanes in an asymmetric fashion are thus desirable. As part of our ongoing research on rhodium-catalyzed reactions, we sought to develop an enantioselective synthesis of functionalized cyclobutanes. The strategy described herein relies on the arylative cyclization of alkynones (or alkynals) with boronic acids (Scheme 1). This reaction, first reported in 2005 by Hayashi[7d] and Murakami,[7e] has since been developed for the synthesis of various cyclic molecules.[7] Enantioselective variants using chiral diene or phosphine ligands have also been reported.[7b, d, 15e] In the process, a substrate possessing appropriately located carbonyl and alkyne functional groups reacts with a boronic acid under Rh catalysis to generate prod[a] T. Johnson, K.-L. Choo, Prof. Dr. M. Lautens Department of Chemistry University of Toronto 80 St. George Street, Toronto, ON, M5S 3H6 (Canada) E-mail: [email protected] Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/chem.201404896. Chem. Eur. J. 2014, 20, 14194 – 14197
Scheme 1. Catalytic cycle of the rhodium-catalyzed arylative cyclization of alkynones with boronic acids R1 = H, alkyl, aryl, CF3.[7d]
ucts bearing an exocyclic alkene. Two new carbon–carbon bonds and one chiral center are formed in the process, which typically affords five- or six-membered rings (Scheme 1). To the best of our knowledge, there are two reports on the use of such cyclizations for the formation of four-membered rings, neither being enantioselective.[7c, 8] In the course of our initial investigations, we discovered that trifluoromethyl ketone 1 a reacted with phenylboronic acid to form cyclobutanol 2 a in 56 % yield under conditions previously reported by our group (entry 1, Table 1).[9] The process generates a tertiary trifluoromethylcarbinol, a motif found in different therapeutic agents.[10] With the racemic version established, a ligand screen was performed in order to achieve an enantioselective reaction. We turned our attention to chiral dienes, which have been shown to promote a variety of rhodium-catalyzed reactions, including cyclization of alkynones with boronic acids.[7d, 9] A range of chiral dienes have been reported with different substitution patterns and symmetry elements (i.e., most often C1 or C2).[11] While L1, L2, and L3 performed poorly, L4[9] gave the desired product in 60 % yield and with a promising 74 % ee. A solvent screen with different alcohols was undertaken, but did not lead to any improvement in yield; the sideproduct arising from protodemetalation of the first intermediate (see Scheme 1) being always generated in 15–20 % yield. Ligand L4 belongs to a class of ligands that can easily be tuned by modification of the aromatic ring via Suzuki crosscoupling from a common vinyl triflate intermediate.[12] This flexibility allowed us to further explore ligands similar to L4 of differing electronic and steric characters (Table 2).
14194
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Communication Table 1. Evaluation of ligand classes for the arylative cyclization of 1 a with phenylboronic acid.[a]
Entry 1 2 3 4 5
Rh source[b]
Ligand
[{Rh(cod)Cl}2] [{Rh(coe)2Cl}2] [{Rh(coe)2Cl}2] [{Rh(coe)2Cl}2] [{Rh(coe)2Cl}2]
– L1 L2 L3 L4
Yield [%][c] [e]
56
