Magnetic and dendritic catalysts

Acc Chem Res. 2015 Jul 21;48(7):1871-80. doi: 10.1021/acs.accounts.5b00039. Epub 2015 Jun 22.

Abstract

The recovery and reuse of catalysts is a major challenge in the development of sustainable chemical processes. Two methods at the frontier between homogeneous and heterogeneous catalysis have recently emerged for addressing this problem: loading the catalyst onto a dendrimer or onto a magnetic nanoparticle. In this Account, we describe representative examples of these two methods, primarily from our research group, and compare them. We then describe new chemistry that combines the benefits of these two methods of catalysis. Classic dendritic catalysis has involved either attaching the catalyst covalently at the branch termini or within the dendrimer core. We have used chelating pyridyltriazole ligands to insolubilize catalysts at the termini of dendrimers, providing an efficient, recyclable heterogeneous catalysts. With the addition of dendritic unimolecular micelles olefin metathesis reactions catalyzed by commercial Grubbs-type ruthenium-benzylidene complexes in water required unusually low amounts of catalyst. When such dendritic micelles include intradendritic ligands, both the micellar effect and ligand acceleration promote faster catalysis in water. With these types of catalysts, we could carry out azide alkyne cycloaddition ("click") chemistry with only ppm amounts of CuSO4·5H2O and sodium ascorbate under ambient conditions. Alternatively we can attach catalysts to the surface of superparamagnetic iron oxide nanoparticles (SPIONs), essentially magnetite (Fe3O4) or maghemite (γ-Fe2O3), offering the opportunity to recover the catalysts using magnets. Taking advantage of the merits of both of these strategies, we and others have developed a new generation of recyclable catalysts: dendritic magnetically recoverable catalysts. In particular, some of our catalysts with a γ-Fe2O3@SiO2 core and 1,2,3-triazole tethers and loaded with Pd nanoparticles generate strong positive dendritic effects with respect to ligand loading, catalyst loading, catalytic activity and recyclability. In other words, the dendritic catalysts were more efficient and more stable than their nondendritic γ-Fe2O3@SiO2 analogues. The bulk at the dendritic periphery helps to localize the metal nanoparticles at the SPION core surface, which confers these advantages. We could also use sonification as a remarkably simple and efficient method to impregnate the SPIONs with dendrimer-encapsulated PdNPs. Catalysis within the hydrophobic dendrimer pockets that include ligands leads to rapid turnover with or without a γ-Fe2O3@SiO2 core. In addition, catalytically active metal nanoparticles are more robust when they are loaded onto the surface of a γ-Fe2O3@SiO2 dendritic core. Herein, we illustrate this chemistry with examples including olefin metathesis, click chemistry, cross carbon-carbon bond forming reactions, and selective alcohol oxidation.