(79e) From Hydrodesulfurization to Hydrodeoxygenation: What Are the Similarities At the Atomic-Scale?

Grabow, L. C., University of Houston
Baek, B., University of Houston
Kasiraju, S., University of Houston

Fast pyrolysis of biomass is a promising low-cost technology that produces bio-oil suitable for use as transportation fuel after an appropriate upgrade step. The upgrade is necessary to increase the heating value, lower the viscosity and improve the long-term stability, and can be achieved by reducing the oxygen content through hydrotreatment over heterogeneous catalysts. While hydrodeoxygenation (HDO) of bio-oil is an emerging technology, hydrotreating for the removal of sulfur from petroleum products (HDS) has matured over decades and detailed knowledge regarding the catalyst structure, active sites and reaction mechanism are available. Both hydrotreating processes show similarities on the macroscopic scale and the identification of atomic-scale catalytic reactivity descriptors may speed up the discovery of novel HDO catalysts from existing knowledge of HDS chemistry.

We have investigated the HDO mechanism of acetaldehyde, a surrogate molecule for the over 400 different oxygenated species in biomass-derived pyrolysis oil, on the Ru(0001) RuO2(110) and the RuO2/TiO2(110) overlayer surfaces using Density Functional Theory (DFT) calculations. Future efforts will focus on HDO of furan to make a direct comparison with results of similar studies of thiophene HDS on CoMoS.1,2 The surface structure of metal-sulfides and metal-oxides under reaction conditions and the role of surface vacancies as active sites during both, HDS and HDO, are discussed. DFT results indicate that acetaldehyde can be selectively deoxygenated on vacancy sites of RuO2(110), but metallic Ru favors C-C bond over C-O bond scission, which leads to the unwanted decarboxylation reaction and the formation of carbon deposits.

Given the higher C-O bond scission selectivity on the metal-oxide, we intend to screen more metal-oxides for their HDO activity and selectivity. Therefore, we are developing a simplified model for the description of the surface free energy of rutile(110) surfaces that will allow the prediction of surface phase diagrams and equilibrium O-vacancy concentrations by calculating just a few parameters. In combination with a more detailed understanding of the HDO mechanism on metal-oxide catalysts and by drawing parallels to the related and well-studied HDS catalysis, new and efficient materials for HDO can be developed faster and ultimately lead to an increased utilization of biofuels.


(1)     Moses, P. G.; Hinnemann, B.; Topsøe, H.; Nørskov, J. K. Journal of Catalysis 2009, 268, 201–208.
(2)     Moses, P. G.; Hinnemann, B.; Topsøe, H.; Nørskov, J. K. Journal of Catalysis 2007, 248, 188–203.