You are using an older browser that does not support current Web standards. Although this site is viewable in all browsers, it will look much better in a browser that supports Web standards.

Pitt HomeFind PeopleContact Us

Chemical and Petroleum Engineering Department

Reactor and Process Engineering Laboratory (RAPEL)

Completed Research

Evaluation of Inorganic, Hydrogen Membranes at Elevated Pressures and Temperatures

Bryan David Morreale, MS, 2001

(Thesis: University of Pittsburgh ETD)

 

A membrane reactor with the capability of simultaneous high-pressure, high-temperature hydrogen permeation testing has been fabricated at the National Energy Technology Laboratory (NETL) in conjunction with the University of Pittsburgh and Parsons Project Services Incorporated. The design of the hydrogen membrane testing (HMT) unit allows for hydrogen membrane evaluation at pressures and temperatures up to 3100 kPa and 1173 K, respectively. The permeability experiments conducted with the NETL's HMT unit include membrane materials, such as palladium, tantalum, Inconel 600 and a cermet membrane fabricated by Argonne National Laboratories (ANL).

The hydrogen permeability of palladium was found to be proportional to the following range of the partial pressure drop of hydrogen across the membrane, ΔP0.57-0.74. The exponent range of 0.57 to 0.74 is indicative of the mass transport of hydrogen being influenced more so by bulk diffusion (n=0.5) rather than surface dissociation (n=1). This exponential range is in good agreement with the only prior study of palladium permeability at elevated pressures, n=0.68 (Hurlbert and Konecny, 1961). The permeability data collected for the palladium membrane can be described by an Arrhenius-type expression where 9.29 × 10-7 mol H2 m-1s-1Pa-n and 37.5 kJ mol-1, representing the pre-exponential factor and the activation energy of permeation, respectively.

The palladium permeability was also determined by assuming the partial pressure of hydrogen across the membrane to the power of 0.5, ΔP0.5. The permeability results appear to be in excellent agreement with the previously published correlations which assumed an exponent of 0.5 and were developed over lower temperature and/or pressure ranges than this study. The permeability of palladium obtained using this assumption can be described by an Arrhenius-type expression where 3.40x10-7 mol H2 m-1s-1Pa-0.5 and 17.29 kJ mol-1, for the pre-exponential factor and the activation energy of permeation, respectively.

Preliminary tests have also been carried out with an ion-transport cermet membrane. The cermet membrane studied (ANL-1), composed of 60vol% BaCe0.8Y0.2O3 and 40vol% nickel powder was fabricated by Argonne National Laboratory, and was tested under conditions similar to those used for the palladium membranes. The cermet membrane experiments permeability was approximately one order of magnitude lower than of the palladium results with partial pressure exponent values close to 0.5. Assuming the membrane's rate limiting step being diffusion, the ANL-1 can be described by an Arrhenius type expression where 1.57x10-6 mol H2 m-1s-1Pa-0.5 and 44.06 kJ mol-1 for the pre-exponential factor and the activation energy of permeation, respectively.

The first known permeability experiment of tantalum can be described by an Arrhenius type expression where 1.10x10-7 mol H2 m-1s-1Pa-0.5 and 16.63 kJ mol-1 represent the pre-exponential factor and the activation energy of permeation, respectively assuming a partial pressure exponent of 0.5. It should be noted however, that the partial exponential results varied from 0.44 to 0.68, which indicates that diffusion may not be the rate-limiting step in the hydrogen transport mechanism of tantalum. The tantalum membranes of this studied were chemically etched to reduce the surface oxides prior to testing. Surface analysis via x-ray photoelectron spectroscopy of the chemically etched tantalum membrane resulted in a negligible oxygen concentration after about 25 angstroms into the membrane, to a maximum concentration of 35% at the surface.

Inconel 600 was also studied to determine the influence of the material of construction on the permeation results of the other tested materials. The permeability associated with the Inconel 600 sample was estimated to be negligible because of its permeability results being two orders of magnitude lower than that of palladium. The Arrhenius pre-exponential factor and the activation energy of permeation of the Inconel 600 membrane were 1.57 × 10-7 mol H2 m-1s-1Pa-0.5 and 58.70 kJ mol-1, respectivley.

Top