When designing a filter, engineers must always consider the trade-off between “permeability” and “selectivity.” Essentially, this means prioritizing either quantity or quality. When filtering water, for example, the more permeable the filter, the faster water can flow through it, but the fewer contaminants it will catch. Conversely, a perfectly selective filter might let only pure water through, but only a few drops at a time.
This trade-off is rooted in the fact that controlling the size of a filter’s pores gets increasingly difficult as they approach the size of the contaminants they are meant to catch. Current manufacturing techniques are not precise enough to make each pore uniformly match the size needed to filter their target molecules, so must settle for erring on the side of permittivity or selectivity to various degrees.
Now, researchers from Penn Engineering have demonstrated a way to achieve this kind of uniformity at the nanoscale, and their membranes are poised to tackle this long-standing permeability-selectivity trade-off. They can produce thin-film membranes with one-nanometer-wide pores, which provide performance that rivals or exceeds currently available nanofiltration membranes.
Their study was led by Chinedum Osuji, Eduardo D. Glandt Presidential Professor in the Department of Chemical and Biomolecular Engineering, and Yizhou Zhang, a postdoctoral researcher in his lab.
It was published in the journal ACS Nano.
With this sort of filtration performance,” Osuji says, “you can think of point-of-use water filtration, treatment of wastewater coming out of industrial facilities, water softening for household use, pervaporation or membrane distillation used in the production of milk and orange juice, and things of that nature.”
At the heart of Osuji’s technology is a method for getting surfactants, closely related to the surfactants used in household detergents, to self-assemble into long cylinders that pack into a hexagonal array within an aqueous medium. This assembly, called a liquid crystal mesophase, is crosslinked with UV light to lock the array into place.
Once set, the arrays of cylinders form a thin membrane, with the long edges of the cylinders running parallel to the plane of the film. Given the regular hexagonal stacking, the spaces between each cylinder act as uniform one-nanometer-wide pores; small enough to let water easily pass while blocking a variety of larger solutes.
In addition to microscopic and X-ray structural analyses of their membranes, the researchers demonstrated their abilities by filtering various molecular dyes and polymers out of water. They also showed how the filtration properties of the membrane could be fine-tuned to a given application, by filtering whiskey to varying degrees of clarity. Because the pore size of the membranes can be controlled by the chemical composition of the liquid crystal mesophase, the researchers were able to design filters that would catch none, some or all of the different organic molecules that give whiskey its color.
This demonstration has already inspired a potential start-up: Osuji’s membranes were featured in this year’s Y-Prize, a business plan competition where students propose applications for Penn Engineering technologies. The winning team, LiberTech, aims to use these membranes to filter beer and wine, removing their alcohol while preserving their color and taste. Existing dealcoholization processes suffer from the permeability-selectivity tradeoff and are often energy inefficient as a result. Osuji’s technology opens the possibility of more efficient dealcoholization overall, and economically viable operation for even small producers, allowing small wineries and microbreweries to offer non-alcoholic versions of their products.
“These membranes are potentially a game-changer for nanofiltration applications due to the pore size uniformity, their compatibility with scalable fabrication techniques, and the fact that pore size can be adjusted precisely in steps as small as 0.1 nanometer,” says Zhang.
“Another interesting aspect of the membranes,” says Osuji, “is that they can filter aqueous as well as organic liquids, as the crosslinking provides stability in a wide range of conditions. That means the membranes are of interest for solvent purification processes as well as organic nanofiltration.”
Osuji’s lab will continue to develop their membranes, with an eye toward making them compatible for mass-production techniques and integrating them with antifouling agents, which would prevent clogs.
Current and former Osuji lab members Ruiqi Dong, Uri R. Gabinet, Ryan Poling-Skutvik, Na Kyung Kim, Changyeon Lee, Omar Q. Imran and Xunda Feng also contributed to the research.
The research was partially supported by the National Science Foundation through grants PFI:AIR-TT IIP-1640375, CBET1703494, and DMR-1945966. Facilities use was supported by the Singh Center for Nanotechnology at the University of Pennsylvania, and the Dual Source and Environmental X-ray scattering facility operated by the Laboratory for Research on the Structure of Matter at the University of Pennsylvania, which is funded in part by NSF MRSEC grant 17-20530. The equipment purchase was made possible by an NSF MRI grant (17-25969), an Army Research Office grant (W911NF-17-1-0282), and the University of Pennsylvania.