Kinetics of hydrogen release from ammonia borane and role of the support for supported Ru catalysts in methanol solvent

Abstract

Ruthenium catalysts are promising for the catalytic release of H₂ from ammonia borane (AB). The reaction kinetics and the role of the support in the release of H₂ from AB in methanol solvent were studied using a series of Ru nanoparticle catalysts on various supports (SiO₂, carbon, γ-Al₂O₃, and TiO₂). Catalysts were synthesized using incipient wetness impregnation and solution deposition (SD) methods and characterized using N₂ physisorption, X-ray diffraction, inductively coupled plasma-optical emission spectroscopy, CO diffuse reflectance infrared Fourier transform spectroscopy, and transmission electron microscopy. The reaction kinetics were examined by measuring initial rates using a semi-batch reactor with a constant flow through the headspace analyzed by online mass spectrometry. The Ru/TiO₂-SD and Ru/C catalysts had the highest measured initial rates compared to the Ru/SiO₂ and Ru/Al₂O₃. Measured apparent activation energies were relatively similar between all catalysts. Regression of initial rates as a function of AB concentration was done for two different rate equations derived from two different proposed mechanisms for AB methanolysis. Both mechanisms were able to sufficiently describe trends in H₂ formation rate versus concentration at low AB concentrations. However, determination of which of the two proposed mechanisms best captures the trends in the data trends was inconclusive due to the lack of a statistically significant difference between degree-of-fit of the two models. Stability tests showed that Ru/C was less prone to deactivation over repeated use compared to Ru supported on the oxides. This study presents a comprehensive examination of the kinetics of AB methanolysis on different supported Ru catalysts; analysis of reaction mechanisms on all catalysts revealed similar apparent first order rate constants from either proposed kinetic models. This study demonstrates that while the support has relatively little influence on the measured reaction rates, carbon may be a preferred support given the decreased deactivation observed relative to the other supports tested.