60th Annual Report on Research 2015

printer friendly

Reports: ND153385-ND1: Borenium Catalysts for Metal-Free Enantioselective Reductions and Petasis Reactions

In the first year of our study we developed Lewis acidic meso-ionic carbene (MIC) borenium ions based on a 1,2,3-triaza-5-ylidene motif that effectively catalyze hydrogenation under exceptionally mild conditions (see previous report).¹ In year two, we examined C–C bond forming reactions and enantioselective reductions. C–C bond forming reactions. Previously synthesized borenium ions A² and B were examined under a wide variety of conditions for the reaction of phenyl borane or vinyl borane with an imine in an attempt to transfer an alkyl or aryl substituent (Scheme 1). However, at ambient temperature no reaction was observed and under forcing conditions at elevated temperature borenium ions A and B lack stability. Since no trace of product was formed, we focused attention on enantioselective reduction chemistry. Enantiopure MICs and enantiopure boranes. One of the challenges with the development of enantiomerically pure MICs is the propensity for deprotonation and concomitant racemization at wing tip groups on nitrogen (See first report). Thus we focused on axially chiral MIC-borenium derivatives as shown in Scheme 2. Known amine 1³ was converted to azide 2 by careful reaction with tert-butyl nitride and trimethylsilylazide. The challenging 1,4-selective Huisgen-cycloaddition of the two sterically encumbered substrates 2 and trimethylsilylacetylene required extensive screening that revealed [Cu(PPh₃)Br] as the most competent catalyst, albeit with extended reaction times. Triazolium 5 was prepared from 3 using Cu-catalyzed N-arylation after silyl-deprotection.⁴ Treatment of 5 with base followed by 9BBN gave boranes 6a/6b in a 10:1 ratio.¹ Hydride abstraction using Ph₃C⁺B(C₆F₅)₄^– gave the borenium ion mixture 7a/b. All reactions up to 5 were high yielding processes amenable to gram scale synthesis.   Borenium-ion mixture 7 was tested in hydrogenation reactions despite the inseparable regio-isomeric mixture (Scheme 3). Low chiral induction for reduction of ketimine 8 and quinoline derivative 10 was observed.   Since NHC–derived borenium ions are known to racemize enantioenriched amines,⁵ we exposed enantio-pure amine (R)-12 to catalytic amounts of achiral borenium ion 7c. This gave only traces of meso-12 by ¹H NMR under conditions similar to catalytic hydrogenations, while B(C₆F₅)₃ showed considerable racemization (Scheme 4). The low degree of enantio induction is therefore likely attributed to poor differentiation of the two prochiral faces of the substrate, or to competing selectivities of 7a and 7b in the hydride delivery step. We therefore targeted MIC-borenium ion 7d, which would give rise to a single isomer (Scheme 5). We also chose to prepare 7e and 7f derived from the chiral 9BBD borane 19⁶ (Scheme 6).   Two routes to the linchpin carboxylic acid 13 were examined: one described by Hoveyda³ and another by Hayashi⁷ (not shown) that starts with cheap, commercial BINOL and catalytic carbonylation chemistry to provide 13. Access to the alkyne 16 via a Sonogashira route in high yield was unsuccessful despite extensive screening. Alkyne 16 was accessed by a Corey-Fuchs reaction sequence (reduction to benzylic alcohol 14, oxidation to aldehyde 15, then Wittig reaction followed by elimination of bromide) and subjected to a base-catalyzed 1,5-selective Huisgen cyclization⁸ with phenylazide in moderate yield. The 1,5-diaryl-1,2,3-triazole 17 was N-arylated to give the 1,3,5-triaryl-1,2,3-triazolium tetrafluoroborate 18. With 18 in hand, future work will be to synthesize the corresponding MIC-borenium 7d. We also plan to explore the use of triazolium salts 5 and 18 as ligand precursors in transition metal compounds. Unfortunately, all attempts at generating and isolating either adducts 6e or 6f of Soderquist's borane 19 failed likely due to excessive steric hindrance of the borane combined with the limited stability of the carbenes under more harsh conditions. Chiral anions. Considering the limited success of both chiral MICs and boranes to support catalytic borenium reduction chemistry, we focused on a chiral anion strategy based on known phosphates and thiophosphates.⁹ Exploring a commercial phosphate and a few synthesized thiophosphates¹⁰ (20, 21a-c; Scheme 7) we found only stoichiometric amounts of reduction products, suggesting that this system fails to activate H₂ and achieve catalytic reduction. Mixing equimolar amounts of MIC-borane 6c and 20 gave rise to the formation of a 4-coordinate boron compound 22 that does not dissociate phosphate sufficiently for H₂ activation even under forcing conditions (heat, elevated pressure). We are currently synthesizing more sterically hindered phosphates with large groups in the 3,3' position of the BINOL-backbone to minimize catalyst deactivation by coordination to the B-centre. Hydrosilylation. As an alternative to hydrogenation, we investigated the hydrosilylation of ketones and imines using MIC-borenium ions and precursor boranes.¹¹ This procedure is simple, and cleavage of H-Si bonds is known to be easier than H₂ activation. We were pleased to find that 7c efficiently catalyzed the reaction of PhSiH₃ and 8 at very low catalyst concentrations, high substrate to catalyst loadings (currently as high as TON 1000) and promising rates (TOF 1000 h^–1) at ambient temperature (Scheme 8)! Importantly, less sterically hindered N-substituents such as benzyl, allyl and methyl resulted also in efficient catalytic hydrosilylation. In addition, a number of silanes are tolerated in this chemistry. Unfortunately, when hydrosilylation reactions with PhSiH₃ and 8 or 10 as substrates were attempted under the same conditions as depicted in Scheme 3, using the 7a/b mixture as the catalyst, the same poor degree of induction was observed. However the high reactivity observed for hydrosilylation will provide us increased opportunities to design effective catalysts that may have better overall enantioselectivities. This work will continue past this granting cycle. In addition, we will continue to explore this exciting silylation chemistry and plan to publish the results in due course.