Hydrogen production from water electrolysis is a key enabling energy storage technology for large scale deployment of intermittent renewable energy sources. Proton Ceramic Electrolysers (PCEs) can produce dry pressurized hydrogen directly from steam, avoiding major parts of cost-driving downstream separation and compression. The development of PCEs has however suffered from limited electrical efficiency due to electronic leakage and poor electrode kinetics. Here, we present the first fully-operational BaZrO3-based tubular PCE, with 10 cm 2 active area and a hydrogen production rate above 15 NmL•min-1. The novel steam anode Ba1-xGd0.8La0.2+xCo2O6-δ (BGLC) exhibits mixed p-type electronic and protonic conduction and low activation energy for water splitting, enabling total polarization resistances below 1 Ω•cm 2 at 600°C and faradaic efficiencies close to 100% at high steam pressures. These tubular PCEs are mechanically robust, tolerate high pressures, allow improved process integration, and offer scale-up modularity. High temperature electrolysers (HTEs) that utilize readily available steam and/or heat (renewable or industrial) as a supplementary energy source provide superior electrical efficiency compared to conventional water electrolysis. 1-4 HTEs developed to date comprise solid oxide electrolysers (SOEs) which utilize oxide ion conducting electrolytes and therefore produce hydrogen on the steam side cathode. The undiluted high pressure oxygen produced on the anode in SOEs presents a safety hazard. Their high operating temperature (typically 800°C)

The electrical properties and the charge carrier mechanism of porous 8 mol% yttria-stabilized zirconia (8YSZ) ceramic samples have been investigated over wide ranges of relative humidity (RH) and temperature (25-400 ºC). The presence of humidity introduces protonic surface conduction, and porous YSZ shows pure protonic conduction below ~150 ºC. H/D isotope effect studies combined with transport number measurements reveal a change in transport mechanism from structural diffusion (Grotthuss type) to vehicular transport when the relative humidity exceeds ~60% RH, coinciding with a change in the adsorbed water layer from an "ice-like" to a "water-like" structure. Similarly, the activation energy for proton transport decreases from 0.43 eV to 0.28 eV when the relative humidity increases from 20% RH to 84% RH, reflecting changes in the enthalpies of formation and migration of charged species with increasing water layer thickness.

This paper develops and demonstrates a model representing radial defect transport through proton-conducting ceramic membranes, such as might be used in shell-and-tube type membrane reactors. The model uses a Nernst–Planck–Poisson (NPP) formulation and is designed to represent both steady-state and transient responses within mixed-conducting membranes with multiple charge-carrying defects. The partial differential equations, representing defect and charge conservation, are solved computationally using the method-of-lines in a differential-algebraic setting. Several example problems are solved and discussed, illustrating important aspects of the model.

This paper derives and demonstrates two models to represent defect transport through mixed-conducting ceramic membranes. The Nernst-Planck-Poisson (NPP) model is more general, but requires the computational solution of a boundary-value problem on a mesh network. The Integral method relies on more assumptions, but defect fluxes can be evaluated analytically based upon material properties and gas-phase composition at the membrane surfaces. Using examples based upon yttrium-doped barium zirconate, the two approaches are compared quantitatively. Under many circumstances the models deliver quantitatively similar results, but for situations where the membrane is exposed to large and/or multiple partial-pressure gradients, the predicted fluxes can deviate. The variations are the result of different physical assumptions and mathematical simplifications that are being used in the two models.