Layered nanosheet membranes for ultrafast molecule separation
(Nanowerk Spotlight) The atomically thin, porous graphene membranes represent a new class of ideal molecular sieves for nanofiltration, where transport occurs through pores which have a thickness and diameter on the atomic scale. These characteristics make graphene an ideal material for creating a separation membrane because it is durable and yet doesn’t require a lot of energy to push molecules through it.
"Although tens of novel 2D layered materials are found, the separation membranes made of them are rather scarce, except recently for MoS2 and graphene oxide nanosheets," Xinsheng Peng, a Professor in the State Key Laboratory of Silicon Materials, Department of Materials Science and Engineering at Zhejiang University, tells Nanowerk. "Like graphene and its derivatives, layered transition-metal dichalcogenides have also desirable mechanical properties and could be assembled into lamellar thin films. Therefore, they are expected to be used to construct novel high-performance lamellar separation membranes."
In new work, Peng and his collaborators have developed a new separation membrane with 2D layered transition metal dichalcogenides (tungsten disulfide) for size-selective separation of small molecules of about 3 nm. As they reported in ACS Nano ("Ultrafast Molecule Separation through Layered WS2 Nanosheet Membranes"), as-prepared WS2 membranes exhibit 5 times higher water permeance than graphene oxide membranes with similar rejection.
Schematics of the nanostrand-channeled WS2 membrane. (Reprinted with permission by American Chemical Society)
The team assembled their separation membrane from chemically exfoliated WS20 nanosheets by filtration. As prepared, this 300-500 nm thick membrane demonstrates a water permeance of 450 L/m2•h•bar with over 90% rejection for 3 nm molecules (Evans Blue). To further improve the water permeance, they employed ultrathin metal hydroxide nanostrands to create more fluidic channels while keeping the rejection rate of specific molecules unchanged. This more than doubled the membrane performance to 930 L/m2•h•bar.
Peng points out that a well calibrated thickness is crucial for a highly efficient separation membrane to balance water flux and rejection rate: "A too thick membrane has low water flux despite high rejection rate, while a thinner membrane usually presents higher flux but worse rejection and suffers mechanical problems."
When testing their membranes under pressure, the team found that the as-prepared WS2 membrane linearly depends on pressure, as was expected. The nanostrand-channeled WS2 membrane however displays a rather different pressure-dependent water flux.
"At lower pressure range, similar to the as-prepared membranes, the water flux increases linearly with external pressure," Peng describes the results. "However, at 0.3 MPa, we observed a transition of water flux with respect to pressure. The flux at the external pressure above 0.3 MPa is fitted with a straight line with larger slope. The transition implies a geometry evolution of the nanochannels during the pressure loading on the channeled membranes beyond 0.3 MPa."
The team speculated that the larger water flux at higher pressure may be attributed to the formation of new fluidic channels.
"Our pressure loading-unloading tests suggests that the channels arising from ultrathin nanostrands are cracked between 0.3 and 0.4 MPa," explains Peng. "These cracks produce new fluidic nanochannels that further results in water flux 4 times that of the as-prepared WS2 membrane without degradation of the rejection performance."
He notes, though, that the ratio of WS2 suspension and ultrathin nanostrands needs to be carefully adjusted. An excess of nanostrands will result in their overlapping, which produces larger channels in the membranes, leading to worse separation performance.
Overall, the results suggest that WS2 membranes hold promising potential for use in applications for ultrafast small organic molecule separation for water purification.
"The development of more 2D-layered materials will also expand the family of 2D-layered material separation membranes," concludes Peng. "Due to their individual unique surface states in combination with different preparation strategies, these novel membranes will exhibit different water permeation behavior and separation performances. In our opinion, the challenge likely comes from how to model the new 2D-layered materials in a proper way."