Abstract
Hydrogen selective membranes are interesting candidates to be applied in hydrogen and energy technology. In water gas shift membrane reactors, for example, the membranes can be used to achieve CO2 capture in pre-combustion decarbonisation cycles for large-scale power generation or hydrogen production. The current commercial availability of hydrogen selective palladium membranes and their utilization, however, are limited due to an unfavorable cost-performance combination. These commercially available membranes are available in the form of tubes or foils, which are relatively thick (20 μm or more). The hydrogen flux, being inversely proportional to the thickness of the membrane, is therefore too low for most applications. For practical use, it is therefore necessary to develop membranes with a reduced thickness of the Pd layer. An affordable, robust and selective hydrogen-separating membrane could significantly reduce this cost. SINTEF has developed a technique for the manufacturing of palladium-based hydrogen separation membranes based on a two step process allowing a reduction in the membrane thickness making palladium membranes economically viable. First, a defect free Pd/Ag alloy membrane is prepared by magnetron sputtering onto the ‘perfect surface’ of a silicon wafer. In a second step the membrane is removed from the wafer and transferred to a porous stainless steel support. This allows the preparation of very thin (approximately 2-3 µm) defect free high-flux membranes supported on macroporous substrates, which can be operated at elevated pressures. In the present work the hydrogen permeation and stability of these tubular palladium alloy composite membranes have been investigated at elevated temperatures and pressures. In our analysis we differentiate between dilution of hydrogen by other gas components, hydrogen depletion along the membrane length, concentration polarization adjacent to the membrane surface, and effects due to surface adsorption on the hydro