Figure 1. Layer deposited on alumina support is about 100-nm thick (inset shows close-up of organic and inorganic bonds and pores); only water molecules pass through sieve. Source: ECN.
|
Potential applications include removing water from solvents and biofuels, desalinating water, and separating hydrogen from gas mixtures.
The new membrane, which ECN is developing in collaboration with the University of Twente, Enschede, and the University of Amsterdam, overcomes the hydrothermal limitations of nanoporous silica (SiO2), which in the presence of water can degrade at temperatures as low as 60°C. Replacing siloxane bonds with hydrolytically stable Si–Cx- Hy-Si links enables the membrane to handle separations at temperatures up to at least 150°C, explain the researchers. Use of polymer materials such as polyvinyl alcohol is limited to around 85°C, they add.
The membrane, which typically has pores in the 2–4 Å range and a narrow pore-size distribution, is deposited as an about-100-nm layer on a tubular mesoporous alumina support (Figure 1). It’s possible to adjust the pore-size distribution to suit particular potential applications; that’s an area where the University of Twente is heading efforts, says Vente.
During 18 months of continuous testing, the water content of permeate from the pervaporation of a 95-wt% n-butanol/5- wt% water mixture at 150°C remained at 98-wt%. The lack of a substantial performance decline over the period, coupled with the selectivity, which is high enough for commercial applications, point up potential of the membrane for large-scale industrial use, say the researchers.
“These tests were performed on a number of membranes, each with a surface area of about 40 cm2, i.e., one tube of 10 cm length,” says Vente.“Currently, we have some 10 membranes running under various conditions in order to determine the application window,” he adds. ECN , which can produce membranes up to 1 m2 and has a skid-mounted test unit (Figure 2), plans to work with companies in pilot testing at that scale.
ECN’s role at this point, says Vente, is defining which application makes most sense as the first industrial-scale use of the membrane, and to address the issues involved in scaling up to large sizes and ensuring reliable operation.
“We anticipate that it will be easier to make on a large scale and that the success rate will be much higher than with pure silica membranes,” says Vente. “Per square meter, these membranes will be more expensive than polymer membranes. We are, however, looking at applications where polymers are not an option. Further, the separation performance of these newly developed membranes is much superior!”
The biggest obstacle remaining isn’t technical, says Vente. It’s “industrial willingness and bravery — someone who dares to give it a try and is willing to take the risk.”