We have designed transition metal - rare earth oxide (TM-REO) systems using computation, synthesis and characterization to catalytically reform tars in model syngas effluents. Naphthalene was the initial model “tar” compound, but we also studied propane reforming, because propane was more amenable to computational (DFT) reaction study. We set as targets catalysts stable for at least a week, with the primary reforming products as CO and H2.
Materials were synthesized by adapting templated sol-gel methods. They were characterized by before- and after XRD, BET, and TPO, and select materials by XANES/XAFS, Raman and magnetic susceptibility. Feeds for reaction studies were (typical) 9-21% H2O, 45-55% CO (in some cases CO2 used), 0.6-4% CH4, 26-31% H2, 1-4% N2, 0.3-3% hydrocarbon, 0-40 ppmv H2S, at 110 kPa, GHSV = 6000-40000, temperatures 903-1073 K.
By DFT, we predicted ease of oxygen vacancy formation and established a relationship between vacancies and reforming reaction rates. Catalysts based on these predictions, such as Mn0.4/Ce/Zr/Al and Fe/Ce3/La/Al (molar ratios, except for Al) proved active for tar reforming, with good sulfur tolerance, reasonable tolerance for coke, and without excessive CH4 or CO2 production. By all these measures, these materials proved superior to typical Ni-based high temperature reforming catalysts.
We further examined reforming with simpler single phase TM-REOs using propane as a model hydrocarbon. Propane was chosen because it was facile for consideration in DFT studies. DFT computations (using the p(2 x 2) unit cell expansion of sulfided M-doped CeO2 (1 1 1)) predicted that doping of CeO2 with these TMs reduces the energies of vacacy regeneration and CO desorption; these are the energetically difficult steps.
In experimental reaction studies, we examined the effects of varying amounts of water, CO2 and H2S on the propane reforming. Results were:
While Fe1/Ce3 was the most active simple TM-REO for C3 reforming, all of the doped CeO2 catalysts were more active than the typical Ni-based high temperature (which also accumulates 2.4 times more coke by TPO). The absence of a transition metal dopant leads to higher C2 (less CO + H2) yield. Note that it is possible to find a practical temperature where even these simpler materials are somewhat sulfur tolerant, showing that the doped REO oxysulfides can retain some reforming activity. All of the characterization studies of the used materials suggest that the TMs remain (mostly) doped in the REO phase, even after >100 h reaction.
It is therefore possible to design a syngas cleanup system operating solely at temperatures between the gasifier and the high temperature water-gas shift reactor. This could greatly improve heat integration (and so, operating costs) of syngas cleanup processes. For a typical syngas process (e.g., to ethanol), the gasifier effluent cleanup costs represent at least 60% of the total costs, including the collection and transport of the biomass source.
See more of this Group/Topical: Catalysis and Reaction Engineering Division