150
C. Sun et al.
demonstrates that the graphene-based membranes are really practicable by gradually
overcoming the challenges toward the industrial applications. All of these state-ofthe-art works offered a proof that the NPG membranes can serve as a novel and
high-efficient membrane for water purification, along with a high water permeance
and high ionic rejection rate thanks to its atomic thickness and excellent ionic sieving
effects. We hope that the ongoing efforts on the graphene-based membranes can not
only make the 2D separation membranes feasible but also facilitate the applications
of graphene sheets in energy harvesting and storage, bioprocessing and so on.
(b) Simulation works
The simulation works were conducted relatively considerably comparing to the
experimental works, especially by using the MD method. The first simulation work
was conducted by Cohen-Tanugi and Grossman [1] in 2012, who first demonstrated
the concept of NPG membranes for water purification. They demonstrated that the
saltwater can be effectively purified by the single-layer NPG sheet and the desalination performance was dependent on the pore size, pore functionalization and
driving pressure. Meanwhile, the NPG membranes presented water permeability
of several orders of magnitude higher than those of conventional RO water desalination membranes. For the salt rejection rate, the small pores usually exhibited a
higher rate, but it presented a decreasing trend at high pressures; the rejection rate
of hydrogenated pores can be higher comparing to the hydroxylated pores. Subsequently, Konatham et al. [48] found that the NPGs with pores of diameter ≈7.5
Å can effectively reject ions and that the ion rejection rates can be enhanced by
the carboxyl groups on the pore rims (Fig. 3). Guerrero-Avilésa and Orellana [49]
identified a water permeation flux through the hydroxylated pores agreeing with the
experimental measurements and showed that the O-passivation on the hydroxylated
pores can enhance the water transport rate owing to the formed hydrogen bonds with
the water molecules. Kommu et al. [50] showed a higher salt rejection rate and intermediate water permeance of the NPG membranes with N-functionalized pores while
a lower salt rejection rate and a higher water permeance of the NPG membranes with
OH-functionalized pores.
After the demonstration of NPG water purification membranes by these pioneering
MD simulation works, many simulation studies were performed for more detailed
investigations. The effects of chemical functionalization, pore size and other factors
on the desalination performance were revealed. Ebrahimi [51] revealed that the curvature of graphene sheets and the shape of iso-surfaces of density affected the rejection
rates of ions. Wang et al. [52] displayed that the chemical functionalization on the
pore rim can greatly affect the ion rejection rate and that the potential of mean
forces can well explain the hierarchy of water permeation rates through the NPG
membranes. Chen and Yang [53] indicated that the NPG membranes with pores
partially functionalized by hydroxyl groups showed a high salt rejection rate. The
desalination performance of NPG membranes was studied by Chogani et al. [54] by
considering the membrane flexibility and the result showed that the water permeation
flux increased while the salt rejection rate decreased with the increase of the applied
C. Sun et al.
demonstrates that the graphene-based membranes are really practicable by gradually
overcoming the challenges toward the industrial applications. All of these state-ofthe-art works offered a proof that the NPG membranes can serve as a novel and
high-efficient membrane for water purification, along with a high water permeance
and high ionic rejection rate thanks to its atomic thickness and excellent ionic sieving
effects. We hope that the ongoing efforts on the graphene-based membranes can not
only make the 2D separation membranes feasible but also facilitate the applications
of graphene sheets in energy harvesting and storage, bioprocessing and so on.
(b) Simulation works
The simulation works were conducted relatively considerably comparing to the
experimental works, especially by using the MD method. The first simulation work
was conducted by Cohen-Tanugi and Grossman [1] in 2012, who first demonstrated
the concept of NPG membranes for water purification. They demonstrated that the
saltwater can be effectively purified by the single-layer NPG sheet and the desalination performance was dependent on the pore size, pore functionalization and
driving pressure. Meanwhile, the NPG membranes presented water permeability
of several orders of magnitude higher than those of conventional RO water desalination membranes. For the salt rejection rate, the small pores usually exhibited a
higher rate, but it presented a decreasing trend at high pressures; the rejection rate
of hydrogenated pores can be higher comparing to the hydroxylated pores. Subsequently, Konatham et al. [48] found that the NPGs with pores of diameter ≈7.5
Å can effectively reject ions and that the ion rejection rates can be enhanced by
the carboxyl groups on the pore rims (Fig. 3). Guerrero-Avilésa and Orellana [49]
identified a water permeation flux through the hydroxylated pores agreeing with the
experimental measurements and showed that the O-passivation on the hydroxylated
pores can enhance the water transport rate owing to the formed hydrogen bonds with
the water molecules. Kommu et al. [50] showed a higher salt rejection rate and intermediate water permeance of the NPG membranes with N-functionalized pores while
a lower salt rejection rate and a higher water permeance of the NPG membranes with
OH-functionalized pores.
After the demonstration of NPG water purification membranes by these pioneering
MD simulation works, many simulation studies were performed for more detailed
investigations. The effects of chemical functionalization, pore size and other factors
on the desalination performance were revealed. Ebrahimi [51] revealed that the curvature of graphene sheets and the shape of iso-surfaces of density affected the rejection
rates of ions. Wang et al. [52] displayed that the chemical functionalization on the
pore rim can greatly affect the ion rejection rate and that the potential of mean
forces can well explain the hierarchy of water permeation rates through the NPG
membranes. Chen and Yang [53] indicated that the NPG membranes with pores
partially functionalized by hydroxyl groups showed a high salt rejection rate. The
desalination performance of NPG membranes was studied by Chogani et al. [54] by
considering the membrane flexibility and the result showed that the water permeation
flux increased while the salt rejection rate decreased with the increase of the applied