Comparison of the Catalytic Activity of Surface-Immobilized Copper Complexes with Phosphonate Anchoring Groups for Atom Transfer Radical Cyclizations and Additions

Abstract

Covalent immobilization of molecular catalysts onto metal oxide surfaces through linker groups is a common strategy for heterogenizing homogeneous catalysts with the expectation that the immobilized catalyst will have properties similar to those of its molecular counterpart. However, the catalytic properties of the immobilized species are often quite different compared to their soluble counterparts in ways that are difficult to predict. This phenomenon is poorly understood and could be due to a variety of factors, including steric shielding of the complex by the surface, changes to the coordination sphere upon immobilization, or a lack of conformational flexibility of the immobilized complexes. Here, we tested the effect of surface immobilization on the catalytic activity and selectivity of atom transfer radical additions and cyclizations. In this study, we varied the proximity of the phosphonate anchoring group to the Cu center by attachment at varying positions of chelating nitrogen ligands such as 1,10-phenanthroline (phen), tris(pyridylmethyl)amine, and 2,9-dimethyl-1,10-phenanthroline as ligand scaffolds. Catalytic testing revealed that in cases where the anchoring group is remote from the catalytic center, as is the case for Cu(phen), the immobilized catalyst functions overall slightly better than its homogeneous counterpart (resulting in higher yields). However, for complexes in which the linker group is close to the active center, the catalytic performance of the immobilized complex was generally worse when immobilized than when in solution (decreased yield upon immobilization). Potential explanations of these observations are discussed. This study very clearly demonstrates the highly complex nature of immobilized catalysts and highlights the need for more in-depth comparisons between immobilized and soluble organometallic catalysts.