Carbon molecular sieves separate xylenes

RESEARCHERS in the US have developed a new carbon-based molecular sieve membrane that can separate similar organic isomers in a low energy process.

Conventionally, mixtures of alkyl aromatics like xylenes are separated using energy-intensive refining processes including crystallisation and adsorption with distillation. It is estimated that these separation processes use the equivalent energy produced by 20 average-sized power stations globally. The researchers from Georgia Institute of Technology and ExxonMobil have instead developed membranes for a new “organic solvent reverse osmosis” (OSRO) process which uses pressure to separate hydrocarbon mixtures, and does not require a phase change of the mixture. The researchers believe it is the first time reverse osmosis with carbon membranes has been used to separate hydrocarbon mixtures.

The researchers, led by Ryan Lively, an assistant professor in Georgia Tech’s School of Chemical and Biomolecular Engineering, use hollow polymer fibres, around 200 µm in diameter, with a pore size of less than 1 nm. The researchers treat the fibres using a cross-linking, which enables the fibres to hold their shape through a pyrolysis process, which converts them to pure carbon. The carbon fibres are bundled together into modules.

Lively and his colleagues tested the membrane with mixtures of para-xylene and ortho-xylene, the size of which differs by just 0.1 nm. At room temperature, a 50:50 mixture of para-xylene and ortho-xylene was converted to an 85:15 mixture.

“These molecules have incredibly similar sizes and properties, but the membranes can tell them apart,” said 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.”

The researchers now plan to test the membranes with more difficult separations, and determine whether any of the many organic compounds found in industrial separation mixtures could foul the membrane. They also need to test the membranes’ long-term resilience.

“Because we are starting with commercially-available polymers and we are using commercial-type equipment, we can see a clear line-of-sight to commercialisation with this technology,' said ExxonMobil research associate Benjamin McCool. “It's a big advantage that the membranes are being spun on a hollow-fibre line similar to that currently used in the industry. The time horizon to make this happen and the cost of production could be highly advantaged over other inorganic systems or more exotic materials like graphene.”

Science DOI: 10.1126/science.aaf1343