70
A. S. Kazemi and M. A. Abdol
of wells on the wafer. Mechanical exfoliation of graphite was then used to deposit
suspended graphene sheets over the wells and pores were introduced by ultravioletinduced oxidative etching in only 5 µm wide standing graphene membranes. The
membranes were tested up to 200 kPa and successfully used as molecular sieves
[27]. Later, O’Hern et al. [28] fabricated a millimeter-scale (25 mm
2 ) monolayer
CVD graphene membrane with intrinsic pores instead of extrinsically generated
ones and transferred it onto a polycarbonate membrane (PCTE). Natural defects
were formed during the fabrication process with 1–15 nm in size. The rejection
percentage for potassium chloride (KCl), tetramethylammonium chloride, Allura
red, and thiuram disulfide were 46, 71, 23, and 17%, respectively. Two years later, the
same group reported ionic selectivity in macroscale monolayer graphene membranes
[166]. This time, they introduced sub-nanometre pores in graphene by ion bombardment followed by chemical etching. Although there was significant leakage through
defects, the mem-branes displayed some K
+ /Cl
− selectivity. Further etching caused
the membrane to permit selective transport of KCl over a larger organic molecule
(~1.0 nm size). In the same year, Celebi et al. [29] stacked two layers of micrometersized graphene on silicon support and formed an almost impermeable layer. Ordered
arrays of pores were machined by FIB onto bilayer graphene with diameters of
8–1000 nm. Water transport measured across the membrane under pressures up to
200 kPa exhibited water flow rates five to seven folds greater than commercial ultrafiltration mem-branes. It is noteworthy that there are no reports of creating aligned
sub-nanometre pores in multilayer materials [20]. In an attempt to overcome leakage
across large intrinsic tears generated in the transferred graphene sheets, O’Hern et al.
[33], plugged the gaps with inorganic materials in 2015. For this, they activated the
graphene sheet in a KOH solution and deposited a 3.5 nm layer of Hafnia (HfO 2 ) on
the sheet to fill the intrinsic defects, followed by intrusion of nylon-6,6 plugs within
the pores by interfacial polymerization. Sub-nanometer pores with an average size
of 0.5 nm and density of 3.8 × 10
13 pore/cm
2 were then created by ion irradiation and etch-ing on the graphene sheet. The designed membranes demonstrated
water permeance of ~1.41 L m
−2 h
−1 bar
−1 , in the same order as RO commercial
membranes. However, the membranes exhibited different rejections under osmotically driven flow at 70, 90 and 83% for MgSO 4 (0.86 nm size), Allura red (~1 nm
size), dextran (~3.7 nm size), respectively. Additionally, the membrane exhibited
a negative rejection ratio toward NaCl (0.716 nm). Although the membrane this
team produced was comparatively large-sized (cm-scale), it was only tested using
an osmotic pressure-driven configuration. Higher pressures were avoided since the
sealed defects were still prone to failure in RO-driven configuration [204]. Later in
2017, Qin et al. [205] used an innovative liquid-casting method to transfer 63 cm
2
graphene nano-porous membrane onto a porous polymer substrate with the minimal
defect. Nanopores were generated on the graphene surface using similar method by
O’Hern et al. [33]. The resulting membrane was not able to reject salt ions (0% NaCl
rejection) but showed flux a few times higher than typical polymeric membranes.
Back in 2015, Surwade et al. [31] demonstrated water/ion selectivity by nanoporous monolayer CVD transferred graphene onto 5 µm apertures of SiN in gravitydriven configuration up to a pressure of 17 kPa. Sub-nanometre pores were etched
A. S. Kazemi and M. A. Abdol
of wells on the wafer. Mechanical exfoliation of graphite was then used to deposit
suspended graphene sheets over the wells and pores were introduced by ultravioletinduced oxidative etching in only 5 µm wide standing graphene membranes. The
membranes were tested up to 200 kPa and successfully used as molecular sieves
[27]. Later, O’Hern et al. [28] fabricated a millimeter-scale (25 mm
2 ) monolayer
CVD graphene membrane with intrinsic pores instead of extrinsically generated
ones and transferred it onto a polycarbonate membrane (PCTE). Natural defects
were formed during the fabrication process with 1–15 nm in size. The rejection
percentage for potassium chloride (KCl), tetramethylammonium chloride, Allura
red, and thiuram disulfide were 46, 71, 23, and 17%, respectively. Two years later, the
same group reported ionic selectivity in macroscale monolayer graphene membranes
[166]. This time, they introduced sub-nanometre pores in graphene by ion bombardment followed by chemical etching. Although there was significant leakage through
defects, the mem-branes displayed some K
+ /Cl
− selectivity. Further etching caused
the membrane to permit selective transport of KCl over a larger organic molecule
(~1.0 nm size). In the same year, Celebi et al. [29] stacked two layers of micrometersized graphene on silicon support and formed an almost impermeable layer. Ordered
arrays of pores were machined by FIB onto bilayer graphene with diameters of
8–1000 nm. Water transport measured across the membrane under pressures up to
200 kPa exhibited water flow rates five to seven folds greater than commercial ultrafiltration mem-branes. It is noteworthy that there are no reports of creating aligned
sub-nanometre pores in multilayer materials [20]. In an attempt to overcome leakage
across large intrinsic tears generated in the transferred graphene sheets, O’Hern et al.
[33], plugged the gaps with inorganic materials in 2015. For this, they activated the
graphene sheet in a KOH solution and deposited a 3.5 nm layer of Hafnia (HfO 2 ) on
the sheet to fill the intrinsic defects, followed by intrusion of nylon-6,6 plugs within
the pores by interfacial polymerization. Sub-nanometer pores with an average size
of 0.5 nm and density of 3.8 × 10
13 pore/cm
2 were then created by ion irradiation and etch-ing on the graphene sheet. The designed membranes demonstrated
water permeance of ~1.41 L m
−2 h
−1 bar
−1 , in the same order as RO commercial
membranes. However, the membranes exhibited different rejections under osmotically driven flow at 70, 90 and 83% for MgSO 4 (0.86 nm size), Allura red (~1 nm
size), dextran (~3.7 nm size), respectively. Additionally, the membrane exhibited
a negative rejection ratio toward NaCl (0.716 nm). Although the membrane this
team produced was comparatively large-sized (cm-scale), it was only tested using
an osmotic pressure-driven configuration. Higher pressures were avoided since the
sealed defects were still prone to failure in RO-driven configuration [204]. Later in
2017, Qin et al. [205] used an innovative liquid-casting method to transfer 63 cm
2
graphene nano-porous membrane onto a porous polymer substrate with the minimal
defect. Nanopores were generated on the graphene surface using similar method by
O’Hern et al. [33]. The resulting membrane was not able to reject salt ions (0% NaCl
rejection) but showed flux a few times higher than typical polymeric membranes.
Back in 2015, Surwade et al. [31] demonstrated water/ion selectivity by nanoporous monolayer CVD transferred graphene onto 5 µm apertures of SiN in gravitydriven configuration up to a pressure of 17 kPa. Sub-nanometre pores were etched