Reactive stability of promising scalable doped ceria materials for thermochemical two-step CO2 dissociation

Metal-doped ceria (Ce1−xMxO2−δ) is an attractive redox-active material for thermo/electrochemical synthesis of renewable fuels due to its high mixed ionic/electronic conductivity and variable valence (Ce4+/Ce3+) and oxygen nonstoichiometry (δ) at high temperatures. Previously, we have investigated all 26 potentially tetravalent dopants for efficient thermochemical splitting of CO2. Here, we fine-tune the dopant activity (x = 0.10 Zr4+, 0.10 Hf4+, 0.07 Ta5+, and 0.05 Nb5+) of all thermally stable ceria materials with an oxygen exchange capacity (OEC) surpassing that of pristine ceria (CeO2−δ), and we employ thermogravimetric analysis to evaluate long-term stability of their OEC over 50 consecutive redox cycles. Each cycle swings between 40 min ceria oxidation with approximately 500 mbar CO2 at 1000 °C and 90 min ceria reduction in about 0.01 mbar O2 at 1500 °C. Along with analyses of phase purity and stability (PXRD), of composition and dopant concentration (EDX and ICP-MS), and of sintering via SEM, the cycling results show long-term stable OEC and kinetics of the oxygen exchange for Zr-, Hf-, and Nb-doped ceria, despite their distinctly sintered particle surfaces. This attractive performance is rationalized by characterizing oxidation states and oxygen vacancies and by excluding surface carbonation through Raman and FT-IR spectroscopy. Furthermore, we find that introducing stable oxygen vacancies in Ce0.95Hf0.05O2−δ by doping with additional 5% lower-valent Li+, Mg2+, Ca2+, Y3+, and Er3+ does not significantly accelerate the oxygen exchange kinetics. From this first comprehensive long-term stability study of systematically optimized ceria, we propose ceria co-doped with permutations of Hf4+, Zr4+, and Nb5+, yielding an optimal average dopant radius of 0.8 A, as the benchmark redox material for thermochemical production of solar fuels.