Kinetic Relevance of Surface Reactions and Lattice Diffusion in the Dynamics of Ce–Zr Oxides Reduction–Oxidation Cycles

Abstract

Reduction–oxidation cycles in oxides are ubiquitous in oxygen storage and transport, chemical looping processes, and fuel cells. O-atom addition and removal are mediated by coupling reactions of oxidants and reductants at surfaces with diffusion of O-atoms within oxide crystals, with either or both processes as limiting steps. CeO2–ZrO2 solid solutions (CZO) are ubiquitous in practice. They are used here to illustrate general experimental strategies and reaction–diffusion formalisms for nonideal systems that enable assessments of the kinetic relevance of the steps that mediate O-atom addition and removal in these materials; these experiments are described within the context of models that describe the driving forces for reaction and diffusion rigorously in terms of oxygen chemical potentials (μO). These strategies assess the rate consequences of varying the fluid phase redox potential, through changes in the identity and pressures of the reactants and products used in redox cycles (O2; CO/CO2; H2/H2O; N2O/N2), of introducing dispersed metal nanoparticles that capture and react lattice O-atoms in CZO using CO or H2, and of imposing intervening dwells without reaction within redox cycles. O-removal rates depend on reductant pressures, even when CO/CO2 and H2/H2O ratios are chosen to maintain the same surface μO if surface reactions were quasi-equilibrated. These data, taken together with significant rate enhancements in O-removal when Pt nanoparticles are present at CZO crystal surfaces and with similar rates before and after inert dwells, demonstrate that reduction rates by both CO and H2 are limited by surface reactions without the presence of consequential spatial gradients in μO within CZO crystals. In contrast, O-addition rates to partially reduced CZO crystals are similar for N2O and O2 reactants and are not affected by the presence of Pt nanoparticles; O-addition rates are significantly higher after intervening inert dwells during CZO oxidation, indicative of spatial gradients in μO, which relax during nonreactive periods. These methods and models, illustrated here for CZO redox cycles at conditions relevant to oxygen storage practice, allow systematic assessments of the kinetic relevance of lattice diffusion and surface reactions for systems that use solids for the reversible storage and release of atoms, irrespective of the identity of the solids or the atoms (e.g., O, H, N, and S).