| dc.description.abstract | Environmental concerns and depleting availability of fossil resources, are raising interest regarding energy production from renewable sources. Biomass has the potential for being a CO2 neutral feedstock for fuel production. To avoid competing with food supply, the utilization of lignocellulosic materials are of high relevance. Degradation of lignocellulose via fast pyrolysis produces a liquid product often referred to as bio-oil. Unprocessed bio-oil has some undesired properties such as high oxygen content and low heating value, and further upgrading is therefore necessary to be able to compete with petroleum. Catalytic hydrodeoxygenation (HDO) is an upgrading reaction which selectively removes oxygen from oxygenated hydrocarbons by reacting with H2over a catalyst.\newline \newline In this project, five bimetallic, Al2O3 supported phosphide catalysts were tested for upgrading of synthetic bio-oil. The goal was to determine if the incorporation of a second metal into a MoP-catalyst could enhance HDO activity. Therefore the monometallic MoP/Al2O3 catalyst was synthesized as a reference. Bio-oil was simulated by iso-eugenol, guaiacol, phenol, furfural, acetol and acetic acid. The catalysts were synthesized by a coprecipitation method called the Pechini method, calcined and reduced. The bimetallic phosphides all consisted of molybdenum, while varying the second metal. Cobalt (Co), nickel (Ni), iron(Fe), copper (Cu) and manganese (Mn) were used as second metals in the individual catalysts. The molar ratios of the three active components, molybdenum, phosphorous and second metal were 1:1:1. They were supported on 90 wt\% of Al203. Three additional versions of MoCoP/Al203 were prepared with 20, 30 and 50 wt\% active material respectively. Also the unsupported versions of the catalysts were synthesized. \newline \newline
XRD confirmed the presence of the bimetallic phosphide phase in the catalysts containing Co, Ni and Fe as the second metal, with no support. Mn was not reduced by the reduction procedure, hence it did not form any bimetallic phase with molybdenum. The Cu based catalyst was reduced, but no bimetallic phosphide phase could be observed. XRD of the supported catalysts gave only diffraction peaks from the Al2O3 support, indicating the active material was highly dispersed on the surface. XRD analysis of 30, 50 wt\% MoCoP supported on Al2O3 showed peaks corresponding to phases from interaction between the support and the active material. The bimetallic phosphide phase may also have been present.
XRF analysis of the supported catalysts confirmed the presence of the desired oxides and the support, in addition some minor impurities were detected. N2-adsorption of the fresh calcined catalysts and the spent catalysts showed that there was a decrease in surface area after the reaction. TGA of the spent catalysts suggests coke was formed on the surface during the experiments. A decrease in catalytic activity was observed by the declining gas production over time. Coke formation was likely the main reason for the deactivation. \newline \newline
Activity testing with the different supported catalysts were performed at 400 C, pH2 of 10 bar, with a 0.04 ml/min flow of bio oil and 2 grams of catalyst for 6 hours. The experiment with the manganese catalyst resulted in bed plugging and was therefore not further studied. Two liquid phases (oily and aqueous) and gas were produced. The oil product constituted of 100 different components. In the gas phase CO2, methane, ethane, ethene, propane, acetylene, benzene, n-pentane were identified as gas phase products. 100 \% conversion of furfural, acetol, acetic acid and iso-eugenol, and >90\% conversion of guaiacol was obtained for all catalysts. The phenol conversion ranged between 60\%-75\% in the sequence Cu | en |
