Comparison between alumina supported catalytic precursors and their application in thiophene hydrodesulfurization: (NH4)4[NiMo6O24H6]·5H2O/γ-Al2O3 and NiMoOx/γ-Al2O3 conventional systems
Abstract
The effect of the phase composition of alumina supported NiMo catalytic precursors on thiophene hydrodesulfurization (HDS) was investigated. The catalytic precursors were prepared by impregnation of the commercial γ-Al2O3 with solutions of Anderson-type ammonium salts or co-precipitation of ammonium heptamolybdate and nickel nitrate. The precursors were characterized by XRD, BET specific surface area, pore volume and pore size, XPS, elemental analysis, TGA and 27Al MAS NMR. The chemical analyses by ICP showed for the NiMo-AP compounds a clear agreement between experimental and theoretical values according to stoichiometric values (Mo/Ni = 6), while for NiMo-COP deviations were observed (Mo/Ni ∼ 7). The specific surface area and pore volume of NiMo-AP/γ-Al2O3 precursors were greater than those of the NiMo-COP/γ-Al2O3 precursors, 387/325 m2 g−1 vs. 283/265 m2 g−1, and 0.34/0.27 cm3 g−1 vs. 0.21/0.15 cm3 g−1, respectively; whereas the average pore radius for all systems was 12 Å. XRD and XPS analysis confirmed the presence of (NH4)4[NiMo6O24H6]·5H2O and Mo5+/Mo6+ for solids obtained by Anderson-type precursors, whereas NiMo-COP/γ-Al2O3 precursors exhibited Mo6+ from NiMoO4 and MoO3. The NiMo precursor obtained from conventional methods showed a higher amount of sulfur than those synthesized from the Anderson-type phase (6.9 to 4.9 wt%), although this does not mean a highly active sample or optimum sulfided active phase. 27Al solid-state MAS NMR showed higher tetrahedrally coordinated aluminium for the NiMo-COP/γ-Al2O3 catalytic precursors. The catalytic activity was strongly influenced by the type of catalytic precursor and metallic wt%. The activity of the catalysts obtained by the sulfided Anderson-type ammonium salts was greater than the sulfided solids obtained by the conventional method, suggesting that these precursors result in a better active phase with a molar ratio (Ni + Mo)/S = 1.01 (likely “Ni–Mo–S” species), due to lower loss of the Ni promoter into the alumina support (27Al NMR) and the lowest metal–support interaction (TGA). The catalysts obtained the HDS products, butane and cis-butene independent of the precursor type. Furthermore, the catalysts with 15 wt% Mo were more efficient than those obtained with 8 wt% Mo.