Kinetics of CO methanation using a Fe-bearing catalyst from a blast furnace sludge

Abstract

Hydrogenation of CO for methane production was studied using blast furnace sludge (BFS), a Fe-rich residue, as a catalyst. Previously, the raw BFS was subjected to successive leaching stages to reduce some inhibitor compounds. The catalytic runs were carried out in a laboratory scale differential reactor, at 300–350 °C, 1 atm, and variable partial pressures of H2 (10–50 kPa) and CO (0.25–3.0 kPa). Before the reaction, the catalyst was reduced in H2 at 500 °C for 2 h. Product gases were analyzed by gas chromatography. The BFS, reduced (BFS-R), leached (BFS-L) and leached and reduced (BFS-L-R) catalysts were characterized by atomic absorption spectroscopy (AAS), in situ X-ray diffraction (XRD), N2-physisorption at 77 K and thermogravimetric-mass spectrometry analysis (TG-MS). Some selected samples were also analyzed by X-ray photoelectron spectroscopy (XPS). Iron content in the leached sample was 51.7 wt%, present mostly as hematite (Fe2O3) and magnetite (Fe3O4). Carbon was also detected in the BFS-L-R, although its influence on CH4 formation was found to be negligible. When BFS-L-R was used as a catalyst, at 320 °C and H2/CO = 20:1, rate of methane production and selectivity achieved 2.63 μmolCH4/gcat/min and 49.5%, respectively. Experimental results demonstrated that, under the studied conditions, CO hydrogenation towards CH4 proceeds through an H-assisted reaction path. Langmuir-Hinshelwood rate laws for both CH4 and the undesired CO2 production were derived, and parameters were adjusted from the obtained data. The activation energy for methane formation was 93.2 kJ/mol.To evaluate the catalyst resistance to poisoning, a BFS-L sample was treated with SO2 prior to the reaction tests. A first-order deactivation model was consistent with the data. The results of this work demonstrate the feasibility of BFS as a low-cost precursor of a methanation catalyst. Although this is a realistic alternative, further research, both experimental and modeling are still required to optimize operating conditions and to explore different reactor configurations.