Nano-Porous Graphene as Free-Standing Membranes
73
from compaction [210]. They also showed that if wrinkles on single-layer graphene
membranes were isolated using supports with small pores, the sheet could sustain
ultrahigh pressure. Schmidt et al. [211] fabricated nanomesh by patterning largearea monolayer graphene with nanometer-scale pores with 10 nm pitch and <4 nm
pore diameter by the direct helium ion beam milling requiring no post-patterning
process. Akbari et al. [25] fabricated ultra-clean single and bilayer CVD graphene
membranes with diameters up to 500 µm and 750 µm, respectively, using inverted
floating method (IFM) followed by thermal annealing in vacuum. By measuring the
dynamic mechanical properties of the membranes, they observed a reduction of the
effective intrinsic stress in the graphene membranes with annealing.
5 Research Gaps
Although a lot of research has been conducted on the physics and chemistry of nanoporous graphene membranes in the last decade, the incorporation of these materials
into RO systems at an industrial scale is still far from reality. There are aspects that
have not been fully understood or limitations that have not been tackled and analyses that are yet to be done both theoretically and experimentally to overcome all
the existing challenges in this field. Being part of a much larger family of 2D materials, any advancement in understanding the fundamental science at the atomic level
across the nano-porous graphene membranes or progress in fabrication techniques
will develop similar characteristics in other 2D materials. The research gaps in the
study of nano-porous graphene membranes can be categorized into two main parts:
Computational and experimental.
Computational gaps: In general, the computation suffers from limitations on
simulations due to computational cost, regarding both spatial cell size and time scales.
MD simulations as the main computational tool in capturing the overall essence of
water and ion dynamics across the membranes have the following constraints [137]:
uncertainties in the applied force fields, quasi-long range convergence of the bulk
to the pore, resistance to “normal” bulk flow and Newtonian mechanics that govern
dynamics of the species. While quantum mechanics needs to be incorporated in
the simulations to grasp a more legitimate picture of the interactions between atomic
features in these studies, larger spatial cell size and lower time scales are necessary to
allow an envision of the bulk of the system. Due to these limitations, many realistic
aspects of transport phenomena are ignored or simplified that hinders a genuine
understanding and envisage of these materials. Another important constraint due to
computational costs is considering the pristine form of the membranes as the basis
of computational studies, while a large area of free-standing CVD graphene with
all its intrinsic defects, cannot be simulated realistically. Even with a very small
cell size of pristine graphene and well-defined pore shape, the applied pressure and
salinity of the solution in the simulations are assumed much higher than conventional
RO systems to avoid additional computational costs. Fouling, as the major obstacle
for the operation of the current RO membranes, is almost ignored in computational
73
from compaction [210]. They also showed that if wrinkles on single-layer graphene
membranes were isolated using supports with small pores, the sheet could sustain
ultrahigh pressure. Schmidt et al. [211] fabricated nanomesh by patterning largearea monolayer graphene with nanometer-scale pores with 10 nm pitch and <4 nm
pore diameter by the direct helium ion beam milling requiring no post-patterning
process. Akbari et al. [25] fabricated ultra-clean single and bilayer CVD graphene
membranes with diameters up to 500 µm and 750 µm, respectively, using inverted
floating method (IFM) followed by thermal annealing in vacuum. By measuring the
dynamic mechanical properties of the membranes, they observed a reduction of the
effective intrinsic stress in the graphene membranes with annealing.
5 Research Gaps
Although a lot of research has been conducted on the physics and chemistry of nanoporous graphene membranes in the last decade, the incorporation of these materials
into RO systems at an industrial scale is still far from reality. There are aspects that
have not been fully understood or limitations that have not been tackled and analyses that are yet to be done both theoretically and experimentally to overcome all
the existing challenges in this field. Being part of a much larger family of 2D materials, any advancement in understanding the fundamental science at the atomic level
across the nano-porous graphene membranes or progress in fabrication techniques
will develop similar characteristics in other 2D materials. The research gaps in the
study of nano-porous graphene membranes can be categorized into two main parts:
Computational and experimental.
Computational gaps: In general, the computation suffers from limitations on
simulations due to computational cost, regarding both spatial cell size and time scales.
MD simulations as the main computational tool in capturing the overall essence of
water and ion dynamics across the membranes have the following constraints [137]:
uncertainties in the applied force fields, quasi-long range convergence of the bulk
to the pore, resistance to “normal” bulk flow and Newtonian mechanics that govern
dynamics of the species. While quantum mechanics needs to be incorporated in
the simulations to grasp a more legitimate picture of the interactions between atomic
features in these studies, larger spatial cell size and lower time scales are necessary to
allow an envision of the bulk of the system. Due to these limitations, many realistic
aspects of transport phenomena are ignored or simplified that hinders a genuine
understanding and envisage of these materials. Another important constraint due to
computational costs is considering the pristine form of the membranes as the basis
of computational studies, while a large area of free-standing CVD graphene with
all its intrinsic defects, cannot be simulated realistically. Even with a very small
cell size of pristine graphene and well-defined pore shape, the applied pressure and
salinity of the solution in the simulations are assumed much higher than conventional
RO systems to avoid additional computational costs. Fouling, as the major obstacle
for the operation of the current RO membranes, is almost ignored in computational