A carbon-based molecular sieve membrane combined with a reverse osmosis process could significantly reduce energy needed to separate alkyl aromatics hydrocarbon molecules, says a research team from the Georgia Institute of Technology (Georgia Tech), Atlanta, Georgia, and ExxonMobil, Irving, Texas.
“We see this as a potentially disruptive technology in the way we separate xylenes and similar organic compounds,” notes Benjamin McCool, an advanced research associate at ExxonMobil Corporate Strategic Research in Annandale, N.J., who worked on the project. “If we can make this work on an industrial scale, it could dramatically reduce the energy required by these separation processes.”
The researchers created the material using hollow polymer fibers (Figure 2) that were treated via cross-linking before being converted to carbon through pyrolysis. The fibers, bundled together into modules, have pore sizes less than one nanometer that can be adjusted during the fabrication process. The membranes are then used in a new organic solvent reverse osmosis (OSRO) process that applies pressure to induce the separation without changing the chemical mixture. A recent article in Science contains more details.
“We take a scalable platform based on polymeric membranes and then turn those materials into inorganic molecular sieves,” explains Ryan Lively, an assistant professor in Georgia Tech’s School of Chemical & Biomolecular Engineering who collaborated with ExxonMobile on the project. “Our membranes are mechanically robust and they can withstand the process conditions required by OSRO. They maintain advantageous mechanical properties and membrane performance as they are converted to carbon fiber.”
The team used the OSRO process to separate mixtures of para-xylene and ortho-xylene, molecules whose sizes differ by one-tenth of a nanometer. Applying pressure at room temperature, the membrane converted the 50:50 mixture to an 85:15 mixture at a high flux relative to zeolite membranes.
“These molecules have incredibly similar sizes and properties, but the membranes can tell them apart,” notes Lively. “This bulk cut of the mixture greatly enhances the concentration with a very low energy input. This mixture could then be fed into a conventional thermal process for finishing, which would reduce the total energy input dramatically.”
In industrial use, the membranes would be bundled together in modules that would be used in chemical facilities. “In practice, you would get as many modules as you needed for a particular application, and if the need increased, you could simply add more modules,” Lively adds. “It would be totally scalable.”
Lively says the next step involves extending the OSRO platform to other challenging separations. “Indeed, we think the membrane materials can be tailored towards specific separation challenges, and this is something we are actively pursuing.”
Tailoring the membranes for other separations involves a variety of factors — the three most important being the polymer precursor to the carbon material, the crosslinking material used for the polymer, and the pyrolysis conditions used to create the carbon, explains Lively. “This is a rich space to explore, but we and others have showed the possibility of tuning carbon structures using these approaches,” he adds.
The team also is studying methods to scale these materials up to larger devices. “The work described in the Science article is a proof of concept for the OSRO approach. We see a line of sight to commercialization of these materials; however, this is just the beginning … so it may be some time before these are used in practice,” says Lively.
The technology still faces a number of challenges before it can be considered for commercialization and industrial use. “The membranes used in the process will need to be tested under more challenging conditions, as industrial mixtures normally contain multiple organic compounds and may include materials that can foul membrane systems. The researchers must also learn to make the material consistently and demonstrate that it can withstand long-term industrial use,” explains McCool.