Article pubs.acs.org/accounts
Solution-Based Processing of Monodisperse Two-Dimensional Nanomaterials Joohoon Kang,† Vinod K. Sangwan,† Joshua D. Wood,†,# and Mark C. Hersam*,†,‡,§,∥,⊥ †
Department of Materials Science and Engineering, Northwestern University, Evanston, Illinois 60208, United States Graduate Program in Applied Physics, Northwestern University, Evanston, Illinois 60208, United States § Department of Chemistry, Northwestern University, Evanston, Illinois 60208, United States ∥ Department of Medicine, Northwestern University, Evanston, Illinois 60208, United States ⊥ Department of Electrical Engineering and Computer Science, Northwestern University, Evanston, Illinois 60208, United States ‡
CONSPECTUS: Exfoliation of single-layer graphene from bulk graphite and the subsequent discovery of exotic physics and emergent phenomena in the atomically thin limit has motivated the isolation of other two-dimensional (2D) layered nanomaterials. Early work on isolated 2D nanomaterial flakes has revealed a broad range of unique physical and chemical properties with potential utility in diverse applications. For example, the electronic and optical properties of 2D nanomaterials depend strongly on atomic-scale variations in thickness, enabling enhanced performance in optoelectronic technologies such as light emitters, photodetectors, and photovoltaics. Much of the initial research on 2D nanomaterials has relied on micromechanical exfoliation, which yields high-quality 2D nanomaterial flakes that are suitable for fundamental studies but possesses limited scalability for real-world applications. In an effort to overcome this limitation, solution-processing methods for isolating large quantities of 2D nanomaterials have emerged. Importantly, solution processing results in 2D nanomaterial dispersions that are amenable to roll-to-roll fabrication methods that underlie lost-cost manufacturing of thin-film transistors, transparent conductors, energy storage devices, and solar cells. Despite these advantages, solution-based exfoliation methods typically lack control over the lateral size and thickness of the resulting 2D nanomaterial flakes, resulting in polydisperse dispersions with heterogeneous properties. Therefore, post-exfoliation separation techniques are needed to achieve 2D nanomaterial dispersions with monodispersity in lateral size, thickness, and properties. In this Account, we survey the latest developments in solution-based separation methods that aim to produce monodisperse dispersions and thin films of emerging 2D nanomaterials such as graphene, boron nitride, transition metal dichalcogenides, and black phosphorus. First, we motivate the need for precise thickness control in 2D nanomaterials by reviewing thickness-dependent physical properties. Then we present a succinct survey of solution-based exfoliation methods that yield 2D nanomaterial dispersions in organic solvents and aqueous media. The Account subsequently focuses on separation methods, including a critical analysis of their relative strengths and weaknesses for 2D nanomaterials with different buoyant densities, van der Waals interactions, and chemical reactivities. Specifically, we evaluate sedimentation-based density gradient ultracentrifugation (sDGU) and isopycnic DGU (iDGU) for post-exfoliation 2D nanomaterial dispersion separation. The comparative advantages of sedimentation and isopycnic methods are presented in both aqueous and nonaqueous media for 2D nanomaterials with varying degrees of chemical reactivity. Finally, we survey methods for forming homogeneous thin films from 2D nanomaterial dispersions and emerging technologies that are likely to benefit from these structures. Overall, this Account provides not only an overview of the present state-of-the-art but also a forward-looking vision for the field of solution-processed monodisperse 2D nanomaterials.
1. INTRODUCTION Layered bulk materials possess strong in-plane covalent bonds but weak out-of-plane bonds coupled by van der Waals interactions.1 Breaking these weak bonds allows the exfoliation of two-dimensional (2D) nanomaterials with thicknesses down to the atomically thin limit. More than a decade after the successful isolation of single-layer graphene from bulk graphite,2 researchers have shifted their attention to related 2D nanomaterials such as hexagonal boron nitride (h-BN),3,4 transition metal dichalcogenides (TMDCs),5−13 silicates,14 borophene,15 and black phosphorus (BP).16−20 This growing family of 2D nanomaterials © 2017 American Chemical Society
exhibit a broad spectrum of properties that underlie fundamental studies and show significant promise for applied technologies including electronics, optoelectronics, energy storage, photovoltaics, and biomedical applications.1,6,11,16,21,22 Micromechanical exfoliation of layered bulk materials by the so-called scotch-tape method has enabled the production of 2D nanomaterial flakes for intrinsic property characterization and prototype device development.1,2,8,11,22 However, this method Received: December 24, 2016 Published: February 27, 2017 943
DOI: 10.1021/acs.accounts.6b00643 Acc. Chem. Res. 2017, 50, 943−951
Article
Accounts of Chemical Research
Figure 1. 2D nanomaterial lattice structures and layer-dependent band structure. (a) Lattice structure of monolayer graphene and electron band structures for monolayer, bilayer, and trilayer graphene. (b) Lattice structure of monolayer black phosphorus (phosphorene) and calculated band gap as a function of layer number. Adapted with permission from ref 61. Copyright 2014 American Physical Society. (c) Lattice structure of monolayer transition metal dichalcogenides (TMDCs) and evolution of MoS2 band structure with layer number. Adapted from ref 11. Copyright 2010 American Chemical Society. (d) Hexagonal boron nitride (h-BN) lattice structure and electronic band structure.
massless carriers with exceptionally high mobility in excess of 100 000 cm2 V−1 s−1 under ideal conditions.2 In contrast, bilayer and trilayer graphene have parabolic dispersion relations and band gaps that can be induced and tuned by an external electric field (Figure 1a).37,38 Monolayer TMDCs are three atoms thick with the chemical formula MX2 (M = Mo, W; X = S, Se, Te), in which the atoms are arranged in trigonal-prismatic coordination lacking inversion symmetry (Figure 1c).39 To date, the most studied TMDC is MoS2, which has an indirect band gap of 1.2 eV in the bulk. As the thickness of MoS2 is decreased, its band gap increases monotonically with a transition from an indirect band gap in two or more layers to a direct band gap in monolayer MoS2 (Figure 1c). This qualitative change in band structure implies that the photoluminescence (PL) yield is enhanced in monolayer MoS2 by ∼104 times relative to bilayer MoS2.11 Similar PL enhancements have been observed for WS2, MoSe2, and WSe2 in the monolayer limit.6,8,12 Reduced screening in monolayer TMDCs results in large exciton binding energies (∼0.5 eV), while large spin−orbit coupling leads to valley polarization.40 BP is an elemental 2D semiconductor that is similarly undergoing intense study for field-effect transistors, photodetectors, and heterojunctions (Figure 1b).8,10,18 BP shows a strong variation in band gap with thickness (from ∼0.3 eV in the bulk to ∼2 eV in the monolayer limit) and possesses a direct band gap for all thicknesses.16,22 Because of its buckled atomic structure, BP exhibits anisotropic optical and electronic properties.18,41 Unlike graphene and TMDCs, BP is highly reactive chemically and degrades under ambient conditions, thus necessitating specialized processing conditions to isolate nanosheets.16,17 Following exfoliation, BP can be handled and studied under ambient conditions if it is suitability encapsulated19 or passivated.42 Encapsulated BP has been utilized for field-effect transistors with reported mobilities of up to ∼1000 cm2 V−1 s−1.43 With a large band gap of ∼6 eV, h-BN is a 2D electrical insulator that has shown promise as an ultrathin gate dielectric in field-effect transistors (Figure 1d).3 Because of its chemical inertness and absence of surface states, h-BN has been effective at mitigating the effects of charged impurities and acts as a robust encapsulation layer.44 h-BN also possesses superlative in-plane thermal conductivity, which holds promise for thermal management applications.45 In the 2D limit, h-BN has been effective in tuning the tunneling characteristics in van der Waals heterojunctions involving graphene, TMDCs, and BP.46
has severe scalability limitations that have motivated the development of alternative strategies for mass production. While direct growth by chemical vapor deposition and related thin-film deposition schemes is a promising route for wafer-scale electronic applications, solution-based processing has nearly limitless scalability and compatibility with roll-to-roll additive manufacturing. 4−7,9,14,16,17,21,23−33 Specifically, liquid-phase exfoliation4−7,9,13,14,16,17,21,23−28,31−33 is promising for industrial-scale production of 2D nanomaterials. However, the direct products of liquid-phase exfoliation methods are 2D nanomaterials with a broad distribution of lateral sizes and thicknesses. Since the electronic band structure of many 2D nanomaterials depends sensitively on thickness (Figure 1), the structural polydispersity of solution-processed 2D nanomaterials implies heterogeneity in properties that limit downstream applications. Therefore, solution-based separation methods are desirable to improve the structural homogeneity and thus realize the full potential of 2D nanomaterials in large-volume applications. Since 2D nanomaterials have sizes and shapes similar to those of many biological macromolecules, biochemical separation methods such as density gradient ultracentrifugation (DGU) have the potential to improve the monodispersity of 2D nanomaterial dispersions.34 In this Account, we review two forms of DGU, namely, sedimentation-based DGU (sDGU) and isopycnic DGU (iDGU), for lateral size and thickness separation of 2D nanomaterial dispersions. The specific advantages and challenges for each method are considered for a range of 2D nanomaterials including graphene, TMDCs, h-BN, and BP. In addition, methods for forming thin films from 2D nanomaterial dispersions are delineated as a stepping stone toward highperformance applications. It should be noted that the production and applications of unsorted 2D nanomaterial dispersions have been reviewed previously35,36 and thus will not be addressed in detail here. Instead, we focus on the specific processing challenges, emerging applications, and future potential of monodisperse 2D nanomaterial dispersions.
2. LAYER-DEPENDENT PROPERTIES OF 2D NANOMATERIALS The properties of ultrathin 2D nanomaterials are fundamentally distinct from those of their bulk counterparts because of strong quantum confinement, reduced screening, and surface effects. For example, monolayer graphene is a gapless semiconductor with linear energy dispersion, resulting in effectively 944
DOI: 10.1021/acs.accounts.6b00643 Acc. Chem. Res. 2017, 50, 943−951
Article
Accounts of Chemical Research
Figure 2. 2D nanomaterial isolation by organic solvent exfoliation. (a) Photograph of MoS2, WS2, and h-BN exfoliated in organic solvents. Adapted with permission from ref 24. Copyright 2011 American Association for the Advancement of Science. (b) Photograph of black phosphorus exfoliated in N-methyl-2-pyrrolidone (NMP). Adapted from ref 17. Copyright 2015 American Chemical Society. (c, d) Graphene concentrations in various organic solvents based on (c) boiling point and (d) dispersive Hansen parameter. Adapted from ref 31. Copyright 2011 American Chemical Society. (e) Lateral size and (f) thickness histograms for graphene in different organic solvents. Adapted with permission from ref 27. Copyright 2015 Macmillan Publishers Ltd.
Figure 3. 2D nanomaterials isolation by aqueous surfactant-assisted exfoliation. (a) Schematic illustration of aqueous graphene exfoliation by ultrasonication with the small-molecule surfactant sodium cholate. (b) Photograph of the resulting graphene dispersion. Adapted from ref 25. Copyright 2009 American Chemical Society. (c) Photograph of 2D nanomaterial aqueous surfactant dispersions. Adapted with permission from ref 9. Copyright 2011 John Wiley and Sons. (d) Large-volume graphene aqueous surfactant dispersions. Adapted with permission from ref 32. Copyright 2014 Macmillan Publishers Ltd. (e) Graphite powder ball milling with dry ice yields edge-carboxylated graphene (ECG) nanosheets. Adapted from ref 29. (f) Thickness distributions for h-BN, MoS2, and WS2 following aqueous surfactant-assisted exfoliation. Adapted with permission from ref 26. Copyright 2014 The Royal Society of Chemistry.
milling.29,30 Since chemical exfoliation has been reviewed previously7,47,48 and results in high defect densities and/or structural transformations that compromise monodispersity, we will not cover this method here. Instead, this section will provide an overview of liquid-phase exfoliation in organic solvents or
3. SOLUTION-BASED EXFOLIATION Solution-based exfoliation methods can be split into three major categories: (1) chemical exfoliation7,47,48 and (2) organic solvent or (3) aqueous surfactant-assisted exfoliation via ultrasonication,4,6,7,9,13,14,16,17,21,24−26,28,31 shear mixing,32,33 or ball 945
DOI: 10.1021/acs.accounts.6b00643 Acc. Chem. Res. 2017, 50, 943−951
Article
Accounts of Chemical Research
resulting dispersions possess intrinsic 2D nanomaterials that are amenable to post-exfoliation separation methods (Figure 3f).26
aqueous solutions, which serves as the starting point for subsequent 2D nanomaterial separation and purification methods. 3.1. Organic Solvent Exfoliation
4. CENTRIFUGAL SEPARATION METHODS As discussed above, the 2D nanomaterial dispersions produced by solution-based exfoliation are heterogeneous in lateral size and thickness. To date, the most effective strategies for performing post-exfoliation separation by lateral size and thickness employ ultracentrifugation, namely, sedimentation-based density gradient ultracentrifugation (sDGU) and isopycnic density gradient ultracentrifugation (iDGU).34 In general, sDGU is more effective for achieving homogeneity in lateral size, whereas iDGU enables precise thickness sorting. More details on these methods and their suitability for different classes of 2D nanomaterials are provided below.
Organic solvent exfoliation produces 2D nanomaterial dispersions without covalent modification, ion intercalation, or surfactant stabilization. In this case, the van der Waals interactions in layered bulk materials are overcome by the energy provided by ultrasonication. For example, exfoliation of monolayer graphene from graphite requires an energy of >2 eV per 1 nm2 of graphene surface.49 During ultrasonication, collapsing cavitation bubbles yield intense tensile and shear stress fields that exfoliate and fragment bulk crystals. Figure 2a shows stable dispersions of MoS2, WS2, and h-BN exfoliated in organic solvents.24 Similarly, stable dispersions of BP have been realized without chemical degradation by employing anhydrous organic solvents (Figure 2b).17 Solution thermodynamics predicts stable dispersions upon minimization of the net energetic cost of mixing a solute in a solvent. Consequently, efficient exfoliation is achieved by selection of proper solvent and solute solubility parameters such as solvent boiling point (∼200 °C) and surface tension (∼40 mJ/m2).31 On the basis of these considerations, 2D nanomaterials are well-dispersed in high-boiling-point solvents (e.g., N-methyl-2pyrrolidone (NMP), dimethylformamide (DMF), and N-cyclohexyl-2-pyrrolidone (CHP)) (Figure 2c).17,31 Moreover, the Hansen solubility parameter, which is a function of the dispersive, polar, and hydrogen-bonding components of the cohesive material energy density, provides additional guidance on appropriate solvents that promote 2D nanomaterial exfoliation (Figure 2d).28,50 After ultrasonication, the 2D nanomaterial dispersions are centrifuged to remove unexfoliated bulk crystals. Figure 2e,f shows representative lateral size and thickness histograms of graphene dispersions prepared by ultrasonication in different organic solvents.27 Although organic solvent exfoliation eliminates dispersion additives and minimizes chemical degradation, the exfoliation yield is relatively low compared with that of aqueous surfactant-assisted exfoliation, which limits the scalability.19,20 In addition, organic solvents often present downstream processing challenges for post-exfoliation separation and thin-film assembly methods.16
4.1. Sedimentation-Based Density Gradient Ultracentrifugation
In sDGU, nanosheets are separated by exploiting the sizedependent sedimentation rates of 2D nanomaterials during ultracentrifugation. The sedimentation coefficient depends on several parameters, including the nanomaterial mass, shape, and buoyant density.16,52 Among these parameters, mass usually dominates, causing larger nanosheets to sediment more quickly in the presence of the centrifugal field. In practice, the centrifugation time needs to be carefully chosen to achieve a desired lateral size separation. Alternatively, multiple centrifugation steps can be employed for more subtle separation targets such as ultrathin WS2 nanosheets (Figure 4a).23 While frequently performed for
3.2. Aqueous Surfactant-Assisted Exfoliation
Exfoliation in aqueous media requires amphiphilic surfactants to stabilize the resulting 2D nanomaterial dispersions. During ultrasonication, the hydrophobic headgroup of the amphiphilic surfactant interacts with the nanosheet surface, while the hydrophilic tail interacts strongly with the surrounding water. Reaggregation of the exfoliated 2D nanomaterials is then mitigated by electrostatic or steric repulsion from ionic or nonionic surfactants, respectively (Figure 3a).25,51 Figure 3b,c shows aqueous surfactant dispersions of graphene, TMDCs, and h-BN.9,25 Shear mixing32,33 and ball milling29,30 present scalable alternatives to ultrasonication for surfactant-assisted exfoliation (Figure 3d,e). Shear mixing provides high shear rates (>104 s−1) that have allowed graphene to be efficiently exfoliated up to the hundred-liter scale (concentrations up to 0.07 mg/mL with production rates exceeding 100 g/h in 10 m3).32 In ball milling, shear forces and mechanical collisions drive exfoliation. Additional components can be introduced to the ball milling apparatus to further promote exfoliation, such as the use of dry ice in the ball milling of graphite (Figure 3e).29 Although aqueous surfactant-assisted exfoliation produces thicker and broader nanosheet size distributions than chemical exfoliation, the
Figure 4. Separation by sedimentation-based ultracentrifugation. (a) Schematic illustration of monolayer-enriched WS2 solutions after multiple sedimentation-based centrifugation steps. Adapted from ref 23. Copyright 2016 American Chemical Society. (b) Photograph and size distribution histograms of black phosphorus (BP) following sedimentation-based centrifugation in NMP at different speeds. Adapted from ref 17. Copyright 2015 American Chemical Society. (c) Finer size separation of BP nanosheets in deoxygenated aqueous surfactant solutions through sedimentation-based density gradient ultracentrifugation (sDGU). The histogram shows the lateral area distribution from each fraction. Adapted from ref 16. 946
DOI: 10.1021/acs.accounts.6b00643 Acc. Chem. Res. 2017, 50, 943−951
Article
Accounts of Chemical Research
Figure 5. Separation by isopycnic density gradient ultracentrifugation (iDGU). (a) Layer-by-layer graphene separation by iDGU. Atomic force microscopy images and cross-sectional line profiles reveal nanosheets of monolayer and bilayer thickness. Adapted from ref 25. Copyright 2009 American Chemical Society. (b) Layer-by-layer h-BN separation by iDGU. Adapted from ref 4. Copyright 2015 American Chemical Society. (c−f) Layer-by-layer separation of (c) MoS2, (d) WS2, (e) MoSe2, and (f) WSe2 by iDGU with polymeric surfactants that reduce the effective nanomaterial buoyant density in aqueous solution. Adapted with permission from ref 6. Copyright 2014 Macmillan Publishers Ltd. (g) Layer-by-layer separation of high-density ReS2 by iDGU using a mixture of iodixanol and CsCl to increase the buoyant density of the gradient medium. Adapted from ref 13. Copyright 2016 American Chemical Society. (h−j) Thickness histograms of (h) graphene, (i), h-BN, and (j) MoS2, indicating improved thickness monodispersity after iDGU.
nanosheet−surfactant complexes, thus guiding the development of iDGU conditions that achieve precise thickness separation.4,6,13,25 Graphene was the first 2D nanomaterial sorted by iDGU.25 In this case, the graphene nanosheets were encapsulated by the planar, ionic, small-molecule surfactant sodium cholate, which was previously employed for carbon nanotube iDGU separations.53,54 Atomic force microscopy of the iDGU-sorted graphene nanosheets revealed tight thickness distributions, including highly homogeneous populations of monolayer and bilayer graphene.25 Similarly, h-BN has been successfully sorted by thickness using iDGU because of the similarity in the bulk buoyant densities of h-BN and graphene (Figure 5b).4 In contrast, TMDCs possess significantly higher buoyant densities that require modifications to the original iDGU process that was developed for graphene. Since the overall buoyant density of a nanomaterial−surfactant complex in aqueous solution is dictated by the nanomaterial density, the surfactant encapsulation layer, and the affiliated hydration layer, modification of the surfactant chemistry allows the buoyant density of a given nanomaterial to be tuned. For example, by using the bulkier amphiphilic block copolymer surfactant Pluronic F68, the hydration layer thickness can be significantly increased, thus reducing the buoyant density of the nanomaterial−surfactant complex (Figure 5c−f).6 In this manner, layer-by-layer thickness sorting has been achieved for TMDCs using iDGU. However, the commonly used density-gradient medium iodixanol has a finite maximum buoyant density that precludes the use of iDGU for the highest-density nanomaterial−surfactant complexes. In these cases, higher-density gradient media have recently been
aqueous dispersions, sedimentation-based centrifugal separation has also been achieved in organic solvents (Figure 4b).17 In these first two examples, the sedimentation occurred in a constantdensity medium. However, a narrower distribution in lateral sizes has been realized by performing sedimentation in a linear density gradient medium (i.e., sDGU). For example, deoxygenated aqueous dispersions of BP have been separated by sDGU using a linear iodixanol density gradient, thereby allowing the largest lateral area and thinnest BP flakes to be isolated. The quality of the resulting sDGU-separated BP was verified by charge transport measurements that found field-effect mobilities comparable to those of mechanically exfoliated BP transistors (Figure 4c).16 4.2. Isopycnic Density Gradient Ultracentrifugation
In iDGU, a 2D nanomaterial dispersion is introduced into a density gradient that is designed to match the size-dependent buoyant density distribution of the desired nanosheets. During ultracentrifugation, each nanosheet sediments through the gradient to its respective isopycnic point (i.e., the point where the buoyant density of the nanosheet matches that of the density gradient medium). Following iDGU, visual bands are evident in the centrifuge tube, thus revealing subpopulations of nanosheets with similar buoyant densities. It should be noted that iDGU often requires longer processing times than sDGU since nanosheets decelerate as they approach their respective isopycnic points. Since the buoyant density depends strongly on the thickness and is nearly independent of lateral size for a 2D nanomaterial, iDGU achieves exceptionally high monodispersity with respect to thickness. Geometrical models provide quantitatively precise predictions of the buoyant densities of 947
DOI: 10.1021/acs.accounts.6b00643 Acc. Chem. Res. 2017, 50, 943−951
Article
Accounts of Chemical Research
Figure 6. Graphene assembly methods and applications. (a) Transfer characteristics of field-effect transistors produced from reduced graphene oxide nanosheets (scale bar = 20 μm). Adapted from ref 55. Copyright 2010 American Chemical Society. (b) Transparency−conductivity plot for graphene and competing materials. The dashed region represents the theoretical limits of graphene, and the black crosses represent the graphene film. Other symbols show competing materials, including indium tin oxide (black), carbon nanotubes (red), metal grating (green), and silver nanowires (blue). Adapted from ref 56. Copyright 2010 American Chemical Society. (c) Schematic illustration of hybrid graphene/MnO2 nanostructured textiles. (d) Current−voltage characteristics of the resulting electrochemical capacitors. Adapted from ref 57. Copyright 2011 American Chemical Society. (e) 3D-printed graphene scaffolds for biomedical applications. Adapted from ref 59. Copyright 2015 American Chemical Society.
graphene dispersions can also be inkjet-printed into conducting structures58 or 3D-printed into scaffolds for biomedical applications (Figure 6e).58,59 With desirable electronic properties, high optical absorption coefficients, and mobile defects,60 TMDC nanosheets are promising for photodetectors and nonvolatile memory applications. In particular, Langmuir−Blodgett assembly has been used to create pinhole-free MoS2 films that show promising photoresponsivity (∼10−4 A/W) (Figure 7a,b).5 Adhesive polymer (e.g., polydimethylsiloxane, PDMS) stamping also provides a scalable transfer scheme for solution-processed 2D nanomaterials that has been utilized for high-performance BP-based field-effect transistors (Figure 7c,d).16 For ultrathin films, layer-by-layer assembly employs sequential adsorption of oppositely charged polymers and nanomaterials. This method is particularly effective for forming ultrathin h-BN gate dielectrics in field-effect transistors (Figure 7e,f).4,14
developed and demonstrated for 2D nanomaterials. Specifically, the high-density TMDC ReS2 has been separated by thickness by increasing the buoyant density of iodixanol through the addition of CsCl (Figure 5g).13 In summary, Figure 5h−j shows the thickness histograms for graphene, h-BN, and MoS2 following sorting by iDGU,4,6,25 which evidently results in narrower thickness distributions compared with the as-exfoliated 2D nanomaterial dispersions in Figure 3f.26
5. EMERGING ASSEMBLY METHODS AND APPLICATIONS 2D nanomaterial dispersions must be assembled into uniform thin films for electronic applications and highly porous thin films for sensing and energy storage applications. The simplest method to form thin films is by spin-coating the nanosheet dispersion. Although this method has produced highperformance graphene-based field-effect transistors (Figure 6a) and transparent conductors (Figure 6b), it suffers from residual surfactant/solvent contamination and spatial nonuniformity.55,56 Vacuum filtration avoids both of these limitations,25 where nanosheet thin films are formed by filtering the dispersion through a porous membrane, after which they are transferred onto the desired substrates. This method has also been used to produce graphene transparent conductors that compete well with commercial indium tin oxide.21 On the other hand, sensing and energy storage applications require porous and nanostructured films. For example, high-performance supercapacitors have been demonstrated from graphene/MnO2 nanostructured textiles (Figure 6c,d).57 Highly concentrated
6. CONCLUSIONS AND OUTLOOK Since the properties of 2D nanomaterials strongly depend on lateral size and thickness, applications of solution-processed 2D nanomaterials are limited by the structural polydispersity that exists immediately following exfoliation. In this Account, we have surveyed recent progress in post-exfoliation separation methods that improve the monodispersity of graphene, h-BN, TMDC, and BP dispersions. Preceding separation, organic solvent or surfactant-assisted aqueous exfoliation methods are selected on the basis of interlayer van der Waals interactions, solubility parameters, and chemical stability of the targeted 2D nanomaterial. 948
DOI: 10.1021/acs.accounts.6b00643 Acc. Chem. Res. 2017, 50, 943−951
Accounts of Chemical Research
■
Article
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. ORCID
Vinod K. Sangwan: 0000-0002-5623-5285 Mark C. Hersam: 0000-0003-4120-1426 Present Address #
J.D.W.: MicroLink Devices, Inc., Niles, IL 60714, USA.
Notes
The authors declare no competing financial interest. Biographies Joohoon Kang is a Ph.D. candidate in the Department of Materials Science and Engineering at Northwestern University under the supervision of Professor Mark C. Hersam. He received his B.S. in Metallurgical System Engineering in 2009 and his M.S. in Materials Science and Engineering in 2011 from Yonsei University in Korea. His research interests include solution-based production, postsynthetic processing, and electronic and optoelectronic applications of 2D nanomaterials. Vinod K. Sangwan is a postdoctoral researcher with Professor Mark C. Hersam in the Department of Materials Science and Engineering at Northwestern University. He received his B.Tech. in Engineering Physics from the Indian Institute of Technology, Mumbai in 2002 and his Ph.D. in Physics from the University of Maryland, College Park in 2009. His current research interests include the fundamental study of defects, charge transport, and ultrafast carrier dynamics in nanomaterials and their heterojunctions.
Figure 7. 2D nanomaterial assembly methods and applications. (a) Langmuir−Blodgett film of MoS2 floating on water. (b) Current− voltage characteristics of an illuminated phototransistor based on a MoS2 Langmuir−Blodgett film. Adapted with permission from ref 5. Copyright 2013 The Royal Society of Chemistry. (c) False-color scanning electron microscopy image of BP nanosheets transferred via polydimethylsiloxane stamping. (d) Transfer curve of the resulting BP field-effect transistor. Adapted from ref 16. (e) Layer-by-layer assembly of hexagonal boron nitride nanosheets. (f) Schematic illustration of the resulting graphene transistors. Adapted from ref 4. Copyright 2015 American Chemical Society.
Joshua D. Wood is a senior process engineer at MicroLink Devices, Inc. He received his B.S. in Computer Engineering from Valparaiso University in 2008, his M.S. in Electrical and Computer Engineering (ECE) from the University of Illinois at Urbana−Champaign (UIUC) in 2009, and his Ph.D. in ECE from UIUC in 2013. He was a postdoctoral fellow with Professor Mark C. Hersam from 2013 to 2016. His research interests include nanomaterials, surface science, nanofabrication, and photovoltaics.
Subsequently, sedimentation-based centrifugal methods are best suited for lateral size separations, whereas isopycnic DGU allows for precise separation by thickness and layer number. By combination of these 2D nanomaterials with thin-film deposition methods, a wide range of high-performance applications have been demonstrated. Despite the significant progress to date, several challenges remain for the cost-effective mass production of electronic-grade 2D nanomaterials. First, while the small lateral size (
