Lawrence Berkeley National Laboratory

The reduction of pure and Sm-doped ceria in hydrogen has been studied by synchrotron-based in situ X-ray diffraction to eventually prove or disprove the presence of crystalline cerium hydride (CeH _x ) phases and the succession of potential structural phase (trans)formations of reduced cerium oxide phases during heating-cooling cycles up to 1273 K. Despite a recent report on the existence of bulk and surface CeH _x phases during reductive treatment of pure CeO₂ in H₂, structural analysis by Rietveld refinement as well as additional ¹H-NMR spectroscopy did not reveal the presence of any crystalline CeH _x phase. Rather, a sequence of phase transformations during the re-cooling process in H₂ has been observed. In both samples, the reduced/defective fluorite lattice undergoes at first a transformation into a bixbyite-type lattice with a formal stoichiometry Ce_0.58 ³⁺Ce_0.42 ⁴⁺O_1.71 and Sm_0.15 ³⁺Ce_0.39 ³⁺Ce_0.46 ⁴⁺O_1.73, before a transformation into rhombohedral Ce₇O₁₂ takes place in pure CeO₂. This phase is clearly absent for the Sm-doped material. Finally, a triclinic Ce₁₁O₂₀ phase appears for both materials, which can be recovered to room temperature, and on which a phase mixture of bixbyite Ce_0.66 ³⁺Ce_0.34 ⁴⁺O_1.67, rh-Ce_0.60 ³⁺Ce_0.40 ⁴⁺O_1.70 and tri-Ce_0.48 ³⁺Ce_0.52 ⁴⁺O_1.76 (for pure CeO₂) or bixbyite Sm_0.15 ³⁺Ce_0.47 ³⁺Ce_0.38 ⁴⁺O_1.69 and tri-Sm_0.15 ³⁺Ce_0.31 ³⁺Ce_0.54 ⁴⁺O_1.77 (for Sm-doped CeO₂) prevails. The absence of the rhombohedral phase indicates that Sm doping leads to the stabilization of the bixbyite phase over the rhombohedral one at this particular oxygen vacancy concentration. It is worth noting that recent work proves that hydrogen is indeed incorporated within the structures during the heat treatments, but under the chosen experimental conditions it has apparently no effect on the salient structural principles during reduction.