SnAP Reagents for the Preparation of Functionalized, Saturated N-Heterocycles

Abstract

Saturated N-heterocycles, for example, piperidines, piperazines, or morpholines can be found with increasing frequency in small bioactive molecules, despite their limited commercial availability and challenging synthesis. Enormous efforts have been made on the synthesis of cyclic amines, but a direct extension of cross-coupling methods to include saturated N- heterocycles remains elusive. Strategies are often specific to a single target, effective for narrow classes of N-heterocycles or substitution patterns, and require multiple protection-deprotection steps. In 2013, our group disclosed SnAP (tin (Sn) amine protocol) reagents for the one-step transformation of aldehydes into unprotected thiomorpholines. This copper mediated cyclization takes place under mild conditions, affords unprotected, saturated N-heterocycles in one-step and tolerates a wide range of functional groups. To evaluate whether SnAP reagents are a practical, versatile approach to saturated functionalized N-heterocycles, we designed new reagents to access 5–9 membered unprotected secondary amines. SnAP reagents are combined with aldehydes or ketones in the presence of molecular sieves to afford the corresponding imine which is cyclized subsequently with stoichiometric Cu(OTf)2. More than 20 SnAP reagents were prepared, which proved to be general to access functionalized, unprotected,sat.N-heterocycles.IncollaborationwithSigma-Aldrich,routesviable to large-scale preparation of the air- and moisture-stable SnAP reagents were designed and reagents were brought to market. Mechanistic investigations involving radical clocks and stereoconvergent cyclization of an enantiopure SnAP reagent suggest a radical-based process initiated by a copper-mediated oxidation of the C–Sn bond to form a stabilized nucleophilic carbon radical driven by formation of the thermodynamically strong Sn–O bond. Coordination of the unprotected N- heterocycles to Cu(II) may lead to catalyst inhibition rendering this process stoichiometric in Cu(II). Screening of solvents, ligand classes (phosphines, bisoxazolines, etc.) and additives allowed for the identification of ligand-accelerated conditions that operate with catalytic amounts of copper. An excess of HFIP was found to be crucial for catalyst turn over. We attribute this to its strong hydrogen bond-donor character, complexing the N- heterocyclic products, thereby promoting turnover of the Lewis acidic copper catalyst. The observation of significant enantioinduction solicited further studies on ligand optimization. Therefore, a chemoinformatic approach in collaboration with the Denmark group (University of Illinois, Urbana-Champaign, USA) involving rounds of calculation, ligand synthesis, and screening, is currently ongoing. Looking at 15’936 bisoxazoline ligands, a subset of 30 candidates, which represent the greatest chemical diversity, was selected and synthesized. Empirical data (TON, enantiomeric ratio) were recorded and computational analysis to suggest ligands that are predicted to be superior for the given transformation are ongoing. Nevertheless, initial results indicate that bisoxazoline ligands are valuable to minimize side reactions as protodestannylation, improve catalyst turn-over, and the enantiomeric ratio. --> Saturated N-heterocycles, for example, piperidines, piperazines, or morpholines can be found with increasing frequency in small bioactive molecules, despite their limited commercial availability and challenging synthesis. Enormous efforts have been made on the synthesis of cyclic amines, but a direct extension of cross-coupling methods to include saturated N- heterocycles remains elusive. Strategies are often specific to a single target, effective for narrow classes of N-heterocycles or substitution patterns, and require multiple protection-deprotection steps.In 2013, our group disclosed SnAP (tin (Sn) amine protocol) reagents for the one-step transformation of aldehydes into unprotected thiomorpholines. This copper mediated cyclization takes place under mild conditions, affords unprotected, saturated N-heterocycles in one-step and tolerates a wide range of functional groups. To evaluate whether SnAP reagents are a practical, versatile approach to saturated functionalized N-heterocycles, we designed new reagents to access 5–9 membered unprotected secondary amines. SnAP reagents are combined with aldehydes or ketones in the presence of molecular sieves to afford the corresponding imine which is cyclized subsequently with stoichiometric Cu(OTf)2.More than 20 SnAP reagents were prepared, which proved to be general to access functionalized, unprotected,sat.N-heterocycles.IncollaborationwithSigma-Aldrich,routesviable to large-scale preparation of the air- and moisture-stable SnAP reagents were designed and reagents were brought to market.Mechanistic investigations involving radical clocks and stereoconvergent cyclization of an enantiopure SnAP reagent suggest a radical-based process initiated by a copper-mediated oxidation of the C–Sn bond to form a stabilized nucleophilic carbon radical driven by formation of the thermodynamically strong Sn–O bond. Coordination of the unprotected N- heterocycles to Cu(II) may lead to catalyst inhibition rendering this process stoichiometric in Cu(II).Screening of solvents, ligand classes (phosphines, bisoxazolines, etc.) and additives allowed for the identification of ligand-accelerated conditions that operate with catalytic amounts of copper. An excess of HFIP was found to be crucial for catalyst turn over. We attribute this to its strong hydrogen bond-donor character, complexing the N- heterocyclic products, thereby promoting turnover of the Lewis acidic copper catalyst.The observation of significant enantioinduction solicited further studies on ligand optimization. Therefore, a chemoinformatic approach in collaboration with the Denmark group (University of Illinois, Urbana-Champaign, USA) involving rounds of calculation, ligand synthesis, and screening, is currently ongoing. Looking at 15’936 bisoxazoline ligands, a subset of 30 candidates, which represent the greatest chemical diversity, was selected and synthesized. Empirical data (TON, enantiomeric ratio) were recorded and computational analysis to suggest ligands that are predicted to be superior for the given transformation are ongoing. Nevertheless, initial results indicate that bisoxazoline ligands are valuable to minimize side reactions as protodestannylation, improve catalyst turn-over, and the enantiomeric ratio.