Gardini, Diego1; Mortensen, Peter Mølgaard2; Carvalho, Hudson W. P.6; Damsgaard, Christian Danvad1; Grunwalst, Jan-Dierk6; Jensen, Peter Arendt2; Jensen, Anker Degn2; Wagner, Jakob Birkedal1
1 Center for Electron Nanoscopy, Technical University of Denmark2 Department of Chemical and Biochemical Engineering, Technical University of Denmark3 Department of Physics, Technical University of Denmark4 Experimental Surface and Nanomaterials Physics, Department of Physics, Technical University of Denmark5 CHEC Research Centre, Department of Chemical and Biochemical Engineering, Technical University of Denmark6 Karlsruhe Institute of Technology
Hydrodeoxygenation (HDO) is proposed as an efficient way to remove oxygen in bio-oil, improving its quality as a more sustainable alternative to conventional fuels in terms of CO2 neutrality and relative short production cycle . Ni and Ni-MoS2 nanoparticles supported on ZrO2 show potential as high-pressure (100 bar) catalysts for purification of bio-oil by HDO. However, the catalysts deactivate in presence of sulfur, chlorine and potassium species, which are all naturally occurring in real bio-oil. The deactivation mechanisms of the Ni/ZrO2 have been investigated through scanning transmission electron microscopy (STEM), energy dispersive X-ray spectroscopy (EDX), scanning electron microscopy (SEM) and X-ray diffraction (XRD). Catalytic testing has been performed using guaiacol in 1-octanol acting as a model compound for bio-oil. Addition of sulphur (0.3 vol% octanethiol) in the feed resulted in permanent deactivation of the catalyst by formation of a catalytically inactive Ni-S phase, as suggested by the very similar spatial distribution of nickel and sulphur signals in STEM-EDX elemental maps (Figure 1) and confirmed by XRD and X-ray absorption spectroscopy (XAS) techniques. Deactivation by chlorine (0.3 vol% chlorooctane) co-feeding was found to be reversible, as the catalyst could regain close to its initial deoxygenation activity upon restoration of a clean feed. SEM-EDX investigations excluded the presence of chlorine species; however, XRD analysis revealed sintering of nickel nanoparticles (Figure 2). Impregnating KCl and KNO3 on two different batches of catalysts decreased permanently their deoxygenation activity, suggesting the adsorption of potassium at low coordinated nickel sites . The high mobility of potassium under the electron beam  prevented the spatial distribution study of this element through STEM-EDX. Moreover, nickel sintering was observed in the KCl poisoned sample and was ascribed once again to the formation of mobile Ni-Cl species upon reaction of HCl with surface oxides . Furthermore, environmental transmission electron microscopy (ETEM) has been used in order to investigate the oxidation of Ni-MoS2/ZrO2 catalyst active phase as a function of different HDO reaction conditions and using methanol as a model molecule for bio-oil.
Microscopy and Microanalysis, 2014, Vol 20, Issue Suppl. 3, p. 458-459