Molybdenum Disulfide and Tungsten Disulfide as Novel …
205
When it comes to the experimental realization of 2D MoS 2 membranes, the layerstacked scheme is the current feasible option [12, 59, 61] due to the challenges
concerning scalability and fabrication of large areas of MoS 2 monolayers [59] and the
generation of nanopores with homogeneous size distribution. Nevertheless, 2D layerstacked MoS 2 with few ~7 nm thick layers has been recently tested with promising
water permeability of up to 320 L·m
−2 ·h
−1 ·bar
−1 , and stable high ionic sieving capability (> 99%) [12]. This type of membrane allows for water flow in between their
grain boundaries and gaps. As for MoS 2 , a 300 nm layer-stacked membrane of WS 2
exhibited an even higher water permeance of 730 L·m
−2 ·h
−1 ·bar
−1 [15] at the cost of
rejecting 90% 3 nm Evans blue molecules—a typical solution procedure to evaluate
rejection rates. In terms of RO saltwater purposes, it is very important for a membrane
to be able to reject 99.5% of salt at standard test conditions [62, 63].
To emulate the natural hydrophilicity present in both MoS 2 and WS 2 nanopores, it
is possible to add chemical functional groups in nanoporous graphene to then adjust
the pore’s chemistry. Using classical MD simulations Cohen-Tanugi and Grossman
[64] investigated water flux through hydrogenated (bonded with H) and hydroxylated (bonded with H and OH) graphene nanopores. They reported permeabilities
two to three orders of magnitude higher than commercial RO membranes at the
same salt rejection rate for some nanopore sizes. By maintaining pore size and using
hydroxyl groups, they also discovered water flux enhancement when compared to the
hydrogenated case. In addition, Risplendi et al. [65] used quantum (DFT) and classical (MD) simulations to show that functionalized graphene nanopores can reject
even neutral solutes such as boric acid (H 3 BO 3 ). In order to guide future membrane
designs, it is very important to understand the challenges involved in the commercialization of RO membranes such as chemical and thermal sensitivity, rapid fouling,
and cleaning.
An experimental work of Surwade et al. [72] demonstrated the possibility of
using this kind of nanoporous graphene monolayer as a desalination membrane. The
nanopores, produced through exposure to oxygen plasma, confirmed the indication of
previous theoretical studies and presented a permeability of ~252 L·m
−2 ·h
−1 ·bar
−1 —
assuming a nanopore density of 10
12 cm
−2 and sizes of 1 nm in diameter. Test
conditions have further confirmed that graphene oxide membranes remain with
excellent water permeance, separation efficiency, chemical and mechanical stability
in water, acid, and basic solutions even after months [73]. However, despite the
attractive potential improvements in the manufacturing process, a cost-effective
graphene-based desalination device and other 2D membranes are still uncertain [66].
The option of layer-stacked graphene membranes is attractive comparing the
industrial scale challenges related to the fabrication of large-area monolayer graphene
with controlled pore density and size, a process which is intrinsically stochastic [67].
Although the water transport mechanism is a bit different in stacked GO nanosheets,
experimental works have demonstrated the possibility of controlling the interlayer
spacing in GO membranes and use it as a water transport channel with salt exclusion up to 97% [74], a performance that is comparable to a typical forward osmosis
membrane. On the other hand, GO needs some stabilization strategy (embedded in
epoxy, for example) once it can disintegrate in aqueous solutions [68].
205
When it comes to the experimental realization of 2D MoS 2 membranes, the layerstacked scheme is the current feasible option [12, 59, 61] due to the challenges
concerning scalability and fabrication of large areas of MoS 2 monolayers [59] and the
generation of nanopores with homogeneous size distribution. Nevertheless, 2D layerstacked MoS 2 with few ~7 nm thick layers has been recently tested with promising
water permeability of up to 320 L·m
−2 ·h
−1 ·bar
−1 , and stable high ionic sieving capability (> 99%) [12]. This type of membrane allows for water flow in between their
grain boundaries and gaps. As for MoS 2 , a 300 nm layer-stacked membrane of WS 2
exhibited an even higher water permeance of 730 L·m
−2 ·h
−1 ·bar
−1 [15] at the cost of
rejecting 90% 3 nm Evans blue molecules—a typical solution procedure to evaluate
rejection rates. In terms of RO saltwater purposes, it is very important for a membrane
to be able to reject 99.5% of salt at standard test conditions [62, 63].
To emulate the natural hydrophilicity present in both MoS 2 and WS 2 nanopores, it
is possible to add chemical functional groups in nanoporous graphene to then adjust
the pore’s chemistry. Using classical MD simulations Cohen-Tanugi and Grossman
[64] investigated water flux through hydrogenated (bonded with H) and hydroxylated (bonded with H and OH) graphene nanopores. They reported permeabilities
two to three orders of magnitude higher than commercial RO membranes at the
same salt rejection rate for some nanopore sizes. By maintaining pore size and using
hydroxyl groups, they also discovered water flux enhancement when compared to the
hydrogenated case. In addition, Risplendi et al. [65] used quantum (DFT) and classical (MD) simulations to show that functionalized graphene nanopores can reject
even neutral solutes such as boric acid (H 3 BO 3 ). In order to guide future membrane
designs, it is very important to understand the challenges involved in the commercialization of RO membranes such as chemical and thermal sensitivity, rapid fouling,
and cleaning.
An experimental work of Surwade et al. [72] demonstrated the possibility of
using this kind of nanoporous graphene monolayer as a desalination membrane. The
nanopores, produced through exposure to oxygen plasma, confirmed the indication of
previous theoretical studies and presented a permeability of ~252 L·m
−2 ·h
−1 ·bar
−1 —
assuming a nanopore density of 10
12 cm
−2 and sizes of 1 nm in diameter. Test
conditions have further confirmed that graphene oxide membranes remain with
excellent water permeance, separation efficiency, chemical and mechanical stability
in water, acid, and basic solutions even after months [73]. However, despite the
attractive potential improvements in the manufacturing process, a cost-effective
graphene-based desalination device and other 2D membranes are still uncertain [66].
The option of layer-stacked graphene membranes is attractive comparing the
industrial scale challenges related to the fabrication of large-area monolayer graphene
with controlled pore density and size, a process which is intrinsically stochastic [67].
Although the water transport mechanism is a bit different in stacked GO nanosheets,
experimental works have demonstrated the possibility of controlling the interlayer
spacing in GO membranes and use it as a water transport channel with salt exclusion up to 97% [74], a performance that is comparable to a typical forward osmosis
membrane. On the other hand, GO needs some stabilization strategy (embedded in
epoxy, for example) once it can disintegrate in aqueous solutions [68].