Optimal alloy mixtures exist. It’s approximately 60-40 for Pd-Cu. “With gold, at about 5%-to-20% alloy, you get about 200% more hydrogen permeability than with just palladium,” Ma notes. A 23% silver alloy gives about 70% more hydrogen permeability compared to a pure Pd membrane, he adds.
Figure 1. Hydrogen diffuses through membrane and porous stainless steel support tube to its annular space. Source: Worcester Polytechnic.
Meanwhile, John Falconer, chair of Colorado’s chemical and biological engineering department, has shown that absorbed molecules can improve the performance of the zeolites such as MFI, whose pore size approximates the size of many organic molecules. “Recently, we’ve found that when zeolites absorb certain molecules in the pores, that makes the zeolite crystals swell.” This shrinks the pore size, dramatically raising selectivity. “We think it’s a big find,” he says. “If you look at branched C6s, a molecule that only goes through the defects — such as 2,2-dimethylbutane [C6H14] — is too large to go through the zeolites at significant rate.” However, by adding 1% to 2% of hexane to the gas, “we see the flux decrease through the defects by one to two orders of magnitude for some membranes.”
A lower cost option
However, some see costs remaining prohibitive for membranes made from pure inorganics and metallics. “They have good performance, but they cost about 500 to 1,000 more times per square meter [than polymerics],” says Koros, who also is editor-in-chief of the Journal of Membrane Science. “The problem [potential end-users say] is: ‘They have high performance, but if it’s so expensive, I can’t do that.’” He suggests the solution is obvious: “Get production that’s less expensive. But you can’t do that if there’s no demand.”
The potential demand certainly exists. Commercial plants would require thousands of square meters, notes Tsapatis, who’s currently collaborating with Pall Corp. as well as Koros. But Tsapatsis asks, “Who’s going to make those thousands of square meters? And how?” Those are the primary obstacles now, he believes. Techniques used in the laboratory don’t suit largescale production, he notes.
Mixed-matrix membranes, which combine zeolites with polymers, promise to address both production and cost issues, hopes Koros. He and his colleagues focus on hollow-fiber membranes. This should allow using the production technology for hollow-fiber membranes, he believes. “Hollow fibers are spun like textile fibers. The same spinning machines could be used for zeolite-polymeric membranes,” he explains.
His group aims to marry the performance of the zeolite and the cost of the polymer. “You can’t get performance of zeolite with the cost of polymer — but you can get close,” Koros notes. How near depends on selectivity, which is very significant to process efficiency, he says. “You might be able to get as close as a factor of two.”
Mixed-matrix membranes still must overcome “very serious obstacles,” he admits. The biggest is building new membranes around old technology. “Do that and thermal separation problems arise.”
The mixed-matrix product represents a practical plat-form, Koros emphasizes. It can be implemented relatively quickly and provides, at lower cost, many of the advan-tages of pure inorganic membranes, he says. However, it’ll be three to five more years before commercial production of zeolite-polymeric membranes begins, he predicts.