The incitement for decreasing the modern society's dependency on fossil based fuel and energy is both environmentally and politically driven. Development of biofuels could be part of the future solution. The combination of ash pyrolysis and catalytic upgrading of the produced bio-oil has been identied as a prospective route to bio-fuels. The upgrading is most favorably done by hydrodeoxygenation (HDO), producing bio-fuels at a quality equivalent to conventional fossil fuels. The topic of this Ph.D. thesis has been the development of active and stable catalysts for this reaction. In the search for new HDO catalyst 23 catalytic systems were screened in a batch reactor at 100 bar H2 and 275 C with phenol as bio-oil model compound, comparing dierent combinations of active phase and support. Among the tested catalysts, reduced metal catalysts displayed the highest activity, with the apparent order of activity for some of the most active catalysts being: Ni/ZrO2 > Ru/C > Ni/Al2O3 > Ni/SiO2 Pd/C > Pt/C. Catalysts should be eective in both hydrogenation and deoxygenation to eectively convert phenol at the applied conditions. Ni/SiO2 was further investigated by varying the Ni particle size, showing that deoxygenation was linked to low coordinated sites (e.g. steps and corners), favored at small nickel particles. Hydrogenation was also inuenced by particle size, where the reaction was hindered at small particles due to competitive adsorption of alcohols on the nickel sites. However, the support was found to have the largest eect on the hydrogenation rate. A good support should have high availability of Lewis acid sites. Overall Ni/ZrO2 prepared with small (<10 nm) nickel particles was found as a very prospective catalyst for HDO, displaying both good hydrogenation and deoxygenation activity. With Ni/ZrO2 identied as a prospective HDO catalyst, the stability and resistance toward bio-oil impurities was investigated in a high pressure gas and liquid continuous ow setup with bio-oil model compounds and compared to the more traditional hydrotreating catalysts Mo2C/ZrO2 and Ni-MoS2/ZrO2. The investigated bio-oil impurities included water (up to 30 wt% in bio-oil), chlorine, potassium and sulfur. Of the tested catalysts, Ni/ZrO2 displayed the highest activity, requiring only 250 C for performing HDO of aromatic compounds, followed by Ni-MoS2/ZrO2 (requiring 280 C), and then Mo2C/ZrO2, which required temperatures >320 C. Ni/ZrO2 and Ni-MoS2/ZrO2 displayed good stability for HDO of pure model compounds over periods of more than 100 h. However, Ni-MoS2/ZrO2 required a signicant co-feed of sulfur (as dimethyl disulde or H2S) to remain stable, as oxidation of the sulde phase took place in the absence of sulfur. Part of the sulfur was III incorporated in the liquid product as thiols if the residence time was not suciently high. Mo2C/ZrO2 displayed poor stability during HDO of a mixture of phenol and 1-octanol at 300 C, loosing roughly 50% activity over 74 h of operation. The loss of activity was found to be due to oxidation of the catalyst by the water formed from the HDO reaction. This was evidenced in a separate experiment co-feeding water, where all activity was lost over only 12 h of operation, and further veried by thermodynamic calculations showing the anity for oxidation of Mo2C with water at the given operating conditions. Ni-MoS2/ZrO2 was also found to be unstable in the presence of water if not co-fed with sucient sulfur. The best stability of the catalyst was obtained when co-feeding sulfur in a H2O/H2S ratio of 9.4, but in this case still 36% loss of activity occurred over 94 h of operation. This loss of activity was probably linked to the loss of edge sulfur atoms during exposure to water. Potassium was found to severely deactivate both Ni/ZrO2 and Ni-MoS2/ZrO2, as the activity of the catalyst decreased by 88% and 94%, respectively, when impregnated with potassium (in a stoichiometric ratio between active metal and potassium) compared to the un-poisoned cases. Chlorine was inhibiting both catalysts but the activity could be regained when removing it from the feed. Sulfur was found as the overall worst poison for Ni/ZrO2, as this transformed the nickel to a NiSx phase leading to fast and complete loss of activity. This is in contrast to the Ni-MoS2/ZrO2 catalyst where sulfur was a requirement to maintain the activity. Overall, Ni-MoS2/ZrO2 displayed the best resistance, but the somewhat high temperature requirement and heavy dependency of a sulfur source are drawbacks. Ni/ZrO2 was in contrast more easily poisoned by especially sulfur, but is an attractive catalyst as it catalyzes HDO at relative low temperatures. However, in general a lot of work is still needed in catalyst development.
Main Research Area:
Jensen, Peter Arendt, Grunwaldt, Jan-Dierk, Jensen, Anker Degn
Technical University of Denmark, Department of Chemical and Biochemical Engineering, 2013