It is widely accepted that vanadium is unsuitable as a membrane material due to an extreme susceptibility to hydrogen embrittlement. Consequently, the focus of R&D effort towards hydrogen-selective, vanadium-based metal membranes in recent years has been the development of robust alloys with improved embrittlement resistance. What the literature hasn’t really addressed, however, is whether vanadium's shortcomings can be overcome through the implementation of suitable controls. This work attempts to address this question by closely examining V-H phase equilibrium and undertaking practical demonstrations of Pd-coated vanadium membranes in a tubular geometry.

Membranes were prepared from a 500 mm-long tubular 99.9% V substrate, coated on each surface with Pd catalyst layers. This single tube was sectioned for several permeation and hydrogen absorption tests. An examination of the V-H phase diagram show that hydride phase transitions (α→β, β→γ) and corresponding miscibility gaps can readily be avoided using appropriate operating procedures. Hydrogen permeating testing showed these membranes exhibit very high permeability (initially exceeding 3.0 × 10−7 mol m−1 s−1 Pa−0.5 at 320 °C and above) which allows the use of thick-walled (~ 0.25 mm), self-supporting, pinhole-free vanadium tubes as the membrane substrate. These membranes also exhibited robustness, with mechanical integrity being maintained through multiple thermal and hydride cycles and over several hundred hours of testing.

This work shows that the main natural advantages of vanadium (low cost relative to Pd and very high permeability which affords the use of self-supporting, defect-free substrates) can still be exploited if used in conjunction with appropriate geometry and operating procedures.