(734c) Rational Design of Supported Ni Inverse Catalysts for Hydrogenolysis
The group of Hartwig first reported low temperature hydrogenolysis of aryl alkyl ethers using homogeneous Ni catalysts. Their observation of small 2 nm Ni colloids at the termination of the reaction suggested possible role of heterogeneous catalysis . Their initial study using homogeneous catalysts was followed with heterogeneous catalysts prepared by supporting Ni on activated carbon using Ni(COD)2 as the Ni precursor . The average Ni particle size after decomposition of the ligand from the precursor was about 2-6 nm. The catalyst was active for selective hydrogenolysis of a broad range of aryl ethers in the presence of a stochiometric amount of base. They also observed that commercial Ni/Al2O3/SiO2 catalysts can accomplish the reaction under the same condition but at much higher Ni loading. They inferred from their studies that large Ni particles were not as active as smaller Ni particles.
Recent interesting reports from the group of Zhang et al.  showed that supported Au nanoparticles did not alter the catalytic activity and selectivity using supports such as ZrO2, Al2O3, SiO2, activated carbon, and rutile TiO2 for guaiacol hydrodeoxygenation. However remarkable changes in guaiacol hydrodeoxygenation activity were observed when Au nanoparticles were deposited on anatase titania. The activity was affect by both the Au and anatase particle size, although the selectivity appeared not to be affected.
In view of these examples of the importance of controlled synthesis of supported metal particles for hydrodeoxygenation reaction, we investigated supported Ni inverse catalysts for the dependence on the presence of stochiometric amount of added base by tuning the basicity of the metal oxide overcoating the Ni particles. The Ni inverse catalysts were synthesized by depositing Ni(COD)2 onto silica presilylated with (3-aminopropyl)triethoxysilane. The amine was introduced to deter surface mobility of the Ni(0) atoms once their protective COD ligands were removed upon heating. The Ni particle size could be varied with different degree of silylation. Deposition of the metal oxide overcoat was conducted after desorption of COD ligands from Ni(COD)2 but before oxidative removal of the propylamine groups from the silica surface so as to enhance metal oxide-metal interaction. Thus, starting from the same parent supported Ni catalysts, different metal oxides could be deposited without changing the Ni particle size distribution. Effective deposition of titania on Ni/SiO2 was evidenced from XPS that showed highly attenuated Ni signal after small amounts of titania deposition (Ti:Ni =2 mole ratio). The Ni signal reference to Ti was as expected for samples without the coating and also quite intense when the coating was removed by ion bombardment, and the order was: Ni/SiO2 > Ti-Ni/SiO2 with ion beam bombarded > Ti-Ni/SiO2. The catalysts with and without titania deposition also showed different TPR profiles with the titania decorated catalyst exhibiting a low temperature reduction peak absent for Ni/SiO2. Different reactions were conducted to investigate the catalytic activities of the various catalysts. The probe reaction of 1-phenylethanol hydrogenolysis, conducted at 140 °C and ambient pressure by bubbling a stream of N2/H2 (50:50) gas through a disperser into the reactor, showed surprised formation of ethylbenzene and acetophenone. The presence of acetophenone, an oxidized product, in the presence of H2 gas indicated that its formation proceeded via hydrogen transfer from one molecule of the alcohol to another to yield acetophenone (the oxidized product) and ethylbenzene (the reduced product). The result suggests that this catalyst is a good candidate for reactions involving hydrogen transfer.
 A. G. Sergeev, and A.F. Hartwig, Science, 2011, 332, 439-443.
 F. Gao, J.D. Webb and J. F. Hartwig, Angew. Chemie, Int., 2016, 55, 1474-1478.
 J. Mao, J. Zhou, Z. Xia, Z. Wang, Z. Xu, W. Xu, P. Yang, K. Liu, X. Guo and Z.C. Zhang, ACS Catal, 2017, 7, 695-705.