(504c) Effect of Sulfur on Surface Structures of Palladium Alloys

A major limitation of modern fuel cell technology is the low tolerance of fuel cell catalysts to even parts per million levels of carbon monoxide and sulfur-containing gases.  For this reason, highly selective separation of the hydrogen fuel from reformate gas streams is necessary to obtain a useable feed gas; this requirement can be especially demanding for high-sulfur fuels.  Perhaps the greatest challenge is that any hydrogen purification technology must itself be tolerant of reactive gases.  A particularly promising method for hydrogen purification is the use of palladium-alloy thin film membranes.  Although these membranes are highly efficient for exclusion of contaminant gases, sulfur compounds can block surface sites necessary for hydrogen adsorption, sometimes drastically lowering the overall permeation rate.  Our research has set out to gain a fundamental understanding of how the composition of H₂-permeable metal membranes can be tuned to enhance resistance to sulfides present in low-quality reformate streams.  

We have created an array of metal alloy membranes using dual e-beam evaporation.  This technique allowed very accurate control of alloy composition and film thickness.  Initial SEM analysis for all of the membranes showed very smooth films.  The membranes were heated under 100 sccm flow of nitrogen to 300C.  Once the temperature was stabilized, the nitrogen was reduced to 50 sccm, and a 50 sccm flow of hydrogen and a 500 ppm level of hydrogen sulfide was introduced.  The temperature was raised incrementally by 50C, and held at each temperature for approximately 30 minutes.  A maximum temperature of 450C was reached, and the total exposure time to the H₂S was around 3 hours.  Permeation data was taken, however due to the lack of a good seal, no conclusions could be reached using it.

The SEM imaging after the sulfur exposure revealed interesting results.  There was little change in the pure palladium membrane.  The pure copper, on the other hand, showed extreme restructuring.  There were multiple layers of crystal like structures.  Most crystals were around 0.5 mm, with some larger ones over 1 mm.  The sulfur had a great enough effect on the copper, that the membrane had actually turned color from copper to a grayish blue, thus leading us to believe that the crystals are copper sulfide.

Each of the alloys exhibited this surface restructuring effect after the sulfur exposure.  The extent of the crystal formation seems to correspond to the percentage of other metal in the palladium alloys.  The copper-palladium and silver-palladium alloys have the same type of restructuring.  However, the gold-palladium alloys exhibit a slightly different structure.  The structures were more round in shape and much less crystalline. 

We have also performed density functional theory (DFT) calculations to aid in interpretation of the experimental results.  Calculations reveal that on the metals that suffer from surface restructuring, H₂S adsorption is weak, while on Pd, it is strong.  H₂S adsorption is also affected by lattice strain, which can induce reconstruction of the Pd surface.  Additionally, lattice strain greatly affects the repulsion of adsorbed H atoms by adsorbed S atoms.