(397g) Predicting Hydrogen Permeance in Composite Pd-Porous Support Membranes

Depending on the type of support, the hydrogen permeance of composite Pd or Pd alloy membranes can be affected differently by the mass transfer resistance in the support. Therefore, it is difficult to compare the hydrogen permeance of different composite Pd membranes by only considering their Pd layer thickness. For instance, composite Pd-porous ceramic membranes are in general thin, but their permeance is sometimes low due possibly to the mass transfer resistance in the porous ceramic support. A model was developed to predict and investigate how mass transfer within the porous support affected the hydrogen permeance, the activation energy for the hydrogen permeance and the hydrogen pressure exponent factor, usually called the n-value, in composite Pd-porous support membranes. The transport of hydrogen through the composite membrane was modeled in two steps: (1) hydrogen transport through the Pd layer following Sievert's law and (2) hydrogen transport through the porous media based on the dusty gas model. At steady state, both hydrogen fluxes were equal, and one could solve for the pressure at the Pd layer-support interface, which was the only unknown. By solving the equation for various pressures at the high pressure side (and keeping the low pressure side equal to 1 bar), it was then possible to predict the hydrogen flux at various pressures. Fitting the hydrogen flux with a non-linear equation led to the hydrogen permeance and the n-value. Ultimately, calculating the hydrogen permeance at various temperatures one can estimate how mass transfer affects the activation energy for the hydrogen permeation. A dimension less number ξ which compared the resistance in the Pd layer with the resistance in the porous support (R_Pd/R_support), was defined to predict the presence of mass transfer resistance within the porous support. Higher values of ξ generally signified negligible mass transfer resistance in the support. It was interesting to note that even when ξ was higher than 100, the hydrogen permeance of the composite Pd membrane was only equal to 93% of the hydrogen permeance of the Pd layer. When ξ values were lower than 10, mass transfer resistance within the support leads to the decrease of the hydrogen permeance to 60% of the free-standing Pd layer and the activation energy for hydrogen permeance decreases from 114.9 kJ mol^-1 to 9 kJ mol^-1. Also, due to the mass transfer limitations the n-value could be higher than 0.5. When n-values are higher than 0.5, the deviations are generally attributed to surface kinetic limitations in the literature. However, in many cases these deviations can, at least in part, be caused by the mass transfer in the porous support. Experimental data were used to validate the model. For instance, the experimental hydrogen permeance at 500ºC of a Pd-Porous Metal (PM) 5.6 µm thick membrane was determined to be 42.3 m³/m²-h-bar^0.5, while a value of 93 m³/m²-h-bar^0.5 was expected according to its thickness. Knowing the mass transfer properties of the bare support and the thickness of the Pd layer (5.6µm), a hydrogen permeance of 55.7 m³/m²-h-bar^0.5 was predicted with the model, which was closer to the experimental hydrogen permeance value of 42.3 m³/m²-h-bar^0.5.