(48e) Catalytic Hydrodeoxygenation of Lignin Derived Compounds: Synergic Effect Between Pd and Fe

The shortage of energy feedstock is now becoming one of the major obstacles for the sustainable development of our hydrocarbon driven society. Among the various solutions proposed, conversion of renewable resources, especially biomass, to produce hydrocarbon fuels is considered to be the most promising approach. However, it remains challenging to remove oxygen from oxygen-rich biomass derived pyrolysis oil (O% = 10-30 wt%) for production of fuels with low oxygen content (O% < 2 wt%). Catalytic hydrodeoxygenation (HDO) is an effective and atom-economical approach for the oxygen removal of bio-oil. Among the various biomass feedstocks, lignin is the most featured one because of its vast availability, non-edible nature and relatively lower oxygen content. Therefore, the catalytic HDO of lignin derived phenolic compounds (guaiacol, cresol, etc.) to produce aromatics (benzene, toluene, xylene, etc.) became index reactions in the HDO catalysis research.   Traditional sulfided Co-Mo and Ni-Mo catalysts employed in hydrodesulfurization and hydrodenitrogenation in petroleum industry require high H₂ pressure and usually deactivate along with sulfur stripping. Recently, supported base metal catalysts, such like Fe/SiO₂ and Fe/C, were found to be very selective for aromatics production in HDO of phenolics, albeit the low activity. On the other hand, noble metals (e.g. Pt and Pd) were demonstrated to be good additives to enhance the hydrogenation activity of metallic catalyst due to their superior activity in activation of H₂. Our previous report shows that the Pd additive in Fe/C catalyst notably promoted its activity in HDO of guaiacol to produce benzene, suggesting a synergic effect between Pd and Fe.

In this work, a series of Pd-Fe catalysts (Pd% = 0.1-5 wt%) were synthesized by incipient wetness impregnation of Pd precursor (Pd(NH₃)₄(NO₃)₂) onto as-prepared Fe₂O₃ nanoparticles to exclude the potential effect of carbon support and ensure direct interactions between Pd and Fe. The reduced samples featured with Pd entities sitting on the top of reduce Fe particles captured by unreduced iron oxides. TPR results showed that the addition of Pd, even at trace amount (0.1 wt%), will significantly change the redox property of Fe₂O₃, by facilitating the reduction of surface iron oxide to iron metal. Activity test for gas phase HDO of m-cresol showed that the addition of Pd remarkably promoted the activity of Fe catalyst, while the product distributions resembled with that of Fe catalyst, showing high selectivity towards production of toluene. Kinetic data obtained at low conversion level (< 10%) also suggested the synergic effect between Pd and Fe, showing the intrinsic activity of both metals in the bimetallic catalysts almost doubled, compared with their monometallic counterparts. In addition, the fully reduced (reduced at 450 ^oC, confirmed with XRD) catalysts suffered a deactivation with formation of iron carbide, while the partially reduced (reduced at 300 ^oC, confirmed with XRD) catalysts show distinct induction period, their activity increased along with time-on-stream, indicating the in-situ reduction of remaining iron oxide.

The interaction between Pd and Fe was further confirmed in XPS, STEM, XAS and theoretical calculations. A possible mechanism, including Pd facilitated H₂ dissociation and Fe dominated m-cresol activation, is proposed to be responsible for the higher activity and lower ring saturation selectivity in Pd-on-Fe catalysts. The Synergic catalysis derived from Pd-Fe interaction was found in this work, providing a promising strategy for designation of highly active and selective HDO catalysts, and an ideal model system for bimetallic catalysis research.