Synthesis, structure and applications of graphene-based 2D heterostructures

Pablo Solís-Fernández a, Mark Bissett b and Hiroki Ago *acd
aGlobal Innovation Center (GIC), Kyushu University, Fukuoka 816-8580, Japan. E-mail:
bNational Graphene Institute, School of Materials, University of Manchester, Manchester M13-9PL, UK
cInterdisciplinary Graduate School of Engineering Sciences, Kyushu University, Fukuoka 816-8580, Japan
dPRESTO, Japan Science and Technology Agency (JST), Saitama 332-0012, Japan

Received 1st March 2017

First published on 10th July 2017

With the profuse amount of two-dimensional (2D) materials discovered and the improvements in their synthesis and handling, the field of 2D heterostructures has gained increased interest in recent years. Such heterostructures not only overcome the inherent limitations of each of the materials, but also allow the realization of novel properties by their proper combination. The physical and mechanical properties of graphene mean it has a prominent place in the area of 2D heterostructures. In this review, we will discuss the evolution and current state in the synthesis and applications of graphene-based 2D heterostructures. In addition to stacked and in-plane heterostructures with other 2D materials and their potential applications, we will also cover heterostructures realized with lower dimensionality materials, along with intercalation in few-layer graphene as a special case of a heterostructure. Finally, graphene heterostructures produced using liquid phase exfoliation techniques and their applications to energy storage will be reviewed.

image file: c7cs00160f-p1.tif

Pablo Solís-Fernández

Pablo Solís-Fernández graduated in Physics from the University of Oviedo (Spain), and obtained his PhD in 2011 for his research on the surface modification of graphite and graphene related materials in the National Carbon Institute (INCAR, Spain). He joined the group of Prof. Ago in Kyushu University (Japan) as a postdoctoral researcher, and was awarded a JSPS fellowship in 2013. His current research interests are focused on the synthesis and characterization of graphene and other 2D materials.

image file: c7cs00160f-p2.tif

Mark Bissett

Dr Mark A. Bissett received his PhD in Nanotechnology from Flinders University (Adelaide, Australia) in 2011. In 2012 he joined the Institute for Materials Chemistry and Engineering in Kyushu University (Japan) as a postdoctoral researcher and in 2013 was appointed as a Research Assistant Professor. In 2014 he joined the University of Manchester in the School of Chemistry as a Research Associate before moving to the School of Materials at the beginning of 2016. In November 2016 he was appointed as a Lecturer in Nanomaterials within the School of Materials. His research interests include the synthesis, functionalisation and characterisation of carbon nanotubes and two-dimensional materials, such as graphene and transition metal dichalcogenides, and their integration into devices such as photovoltaics, flexible composites, and electrochemical energy storage.

image file: c7cs00160f-p3.tif

Hiroki Ago

Hiroki Ago received his PhD from Kyoto University in 1997 for experimental and theoretical research on the electronic structure of functional materials. After spending one year and a half at the Cavendish Laboratory, Cambridge University supported by JSPS, he became a researcher at Nanocarbon Research Center of National Institute for Advanced Industrial Science and Technology (AIST) at Tsukuba. In 2003, he joined the Institute for Materials Chemistry and Engineering, Kyushu University as an associate professor. In 2016, he became a full professor at Global Innovation Center, Kyushu University. His current research focuses on the controlled synthesis of graphene and related low-dimensional materials as well as experimental investigation of their physical and chemical properties.

1. General introduction

Graphene is an atomically thin single layer of carbon atoms covalently bound in a honeycomb lattice, and the basic building block of graphite. This structure means that graphene has some exceptional characteristics, including mechanical strength and flexibility, high charge carrier mobility, high optical transparency and a work function that can be tuned by electrostatic or chemical doping.1 However, it was not until 2004 that a method was developed to exfoliate graphite into single graphene layers by simply using adhesive tape.2 Although obtained using this simple method, this work demonstrated changes in the resistivity by a factor of 100 by electrostatic doping at low temperatures, and carrier mobilities up to 104 cm2 V−1 s−1. Within a few weeks, another group independently reported the growth of ultrathin graphite films on SiC, providing a method for the large-scale growth of graphene.3 These groundbreaking studies set the focus on graphene, the first 2D material ever obtained. Very soon other elaborate methods were developed to produce graphene. These include the already mentioned epitaxial growth on SiC; chemical vapor deposition (CVD) on several metal catalysts using a wide range of gaseous, liquid and solid precursors;4 exfoliation and reduction of graphite oxide;5 and several liquid phase exfoliation techniques.6

However, the limitations of graphene soon became evident. Among these, probably the most important issue is the lack of a bandgap in its electronic structure.2 This has prevented the efficient use of graphene for several electronic and optoelectronic applications. As a consequence of these limitations, interest in new families of 2D layered materials has emerged in recent years.7 The most remarkable of these, given its technological importance, is hexagonal boron nitride (h-BN),7,8 a layered insulator with a structure similar to that of graphene, as well as the family of transition metal dichalcogenides (TMDs) from group VI,7,9 which are semiconductors with bandgaps between 1 and 2 eV. Interestingly, 2D heterostructures can be made by combining graphene with one or more of these materials, as illustrated in Fig. 1. Such heterostructures not only serve to overcome the inherent limitations of each of the materials,10 but novel properties can be realized by their proper combination. For the sake of brevity, from now on the term heterostructure in this text will refer to a heterostructure of graphene with other materials, unless otherwise specified.

image file: c7cs00160f-f1.tif
Fig. 1 Schematic overview of the different kinds of graphene heterostructures introduced in this work (top), along with the methods usually employed to fabricate them.

In this review, we will describe the evolution and the current state in the synthesis and applications of graphene-based heterostructures. After a brief introduction in Section 2 to some of the already synthesized 2D materials, Section 3 will cover the existing methods to synthesize thin-film heterostructures of graphene with other 2D materials, along with their properties and potential applications. Heterostructures with lower dimension materials and those obtained by intercalation in bi- to few-layer graphene will also be addressed in this section. Section 4 is devoted to heterostructures processed in liquid solutions, which we will denote as hybrid composites. As will be shown, the importance of liquid processed composites resides in the large throughput obtained, along with their convenience for applications related with energy production and storage.

2. Brief introduction to non-graphene 2D materials

As general overviews to graphene can be found in different articles of this volume, here we will briefly introduce only other 2D materials that are commonly employed in heterostructures with graphene. Less than one year after their groundbreaking work on the mechanical exfoliation of single-layer graphene,2 the same group reported the stability of many other layered 2D crystals at room temperature.7 Some of the materials thinned down to single or few layers in this pioneering work included h-BN, some TMDs and Bi2Sr2CaCu2Ox, a high temperature bulk superconductor. As in the case of graphene, the obtained 2D crystals are the building blocks of layered materials held together by weak van der Waals interactions. The methods to obtain isolated single layers have greatly evolved since these first attempts, both for top-down exfoliation techniques and for bottom-up approaches. Currently there exist a wide range of techniques to produce 2D materials, the choice of which depends on the specific material, and on the required applications. The most common synthesis methods include bottom-up approaches, such as CVD, chemical vapor transport (CVT) or other physical deposition methods (molecular beam epitaxy (MBE), magnetron sputtering or physical vapor deposition),11,12 and colloidal growth.13 Top-down approaches other than mechanical exfoliation are also widely extended, commonly involving exfoliation in some liquid medium and sometimes assisted with chemical reactions.14 Interestingly, most of these mentioned methods can be adapted to the production of heterostructures with graphene.

Simultaneous to the evolution of synthesis techniques, the amount of discovered and produced 2D materials has significantly increased during the last decade,14–16 with the new materials covering an extensive spectrum of electronic properties that vastly expand the potential applications from those expected for the original graphene. Among the family of 2D materials there are insulators, such as h-BN;8 semiconductors covering a wide range of band gaps, such as MoS29 or phosphorene;17 zero-gap semiconductors such as graphene; metals, such as NbS218 or the unstable 1T phase of some TMDs;19 and superconductors, such as NbSe2,20 FeSe21 or Bi2Sr2CaCu2O8+x.22 Molecular nanosheets are also an emerging new class of 2D materials with great potential in heterostructures,23,24 which will be briefly introduced in Section 3.2.5 of this review. Structures of some of the most representative 2D materials are shown in Fig. 2. Among the different 2D materials, there are truly atomically thin ones, as is the case of graphene or h-BN, and those whose structure comprises a few atomic layers, such as the case of TMDs, or with a crumpled structure, as in the case of phosphorene. However, in either case the estimated thickness is below or around 1 nm. Although non-layered 2D materials, such as WO3, PbS or ultrathin metal films, are out of the scope of this review, recently they have been attracting increasing interest due to their wide range of applications.25

image file: c7cs00160f-f2.tif
Fig. 2 Chemical (top) and band (bottom) structures of some of the most representative 2D materials. (a) Graphene, (b) h-BN, and (c) MoS2. Band structures adapted from ref. 16 with permission from The Royal Society of Chemistry.

Owing to their technological importance, h-BN and the family of TMDs are among the most prominent and studied layered materials. Boron nitride can present several crystalline forms, the most stable being the hexagonal boron nitride, often abbreviated as h-BN. h-BN possesses a hexagonal lattice analogous to that of graphene, although with boron and nitrogen atoms, respectively, occupying the two inequivalent sublattices (Fig. 2b). Owing to the ionic nature of the bonding of the B and N atoms, h-BN is an insulator with a large band gap of ∼5.97 eV, in contrast to the gapless nature of graphene. The lattice constants of graphene and h-BN are also quite similar, with a lattice mismatch of only ∼1.8%. As we shall see in the following sections, this mismatch induces some interesting changes in the electronic structure of graphene when it is in the proximity of h-BN.

The family of TMDs presents the general form MX2, where M is a transition metal, commonly from groups IV–VI, such as W and Mo, and X is an element from the group of the chalcogens, usually S and Se, and to a less amount Te. These materials are composed of a 3-atom layer of the form X–M–X, with a plane of hexagonally packed metal atoms separating two planes of chalcogen atoms. The bonds formed between the metal and the chalcogen are predominantly covalent. Depending on the geometrical coordination of the metal atoms, the structures of the TMD present two different polymorphs known as 2H (trigonal prismatic) or 1T (octahedral). The electronic properties of TMDs cover a wide range from metals to insulators. From here we will mostly focus on the group VI TMDs (MoS2, WS2, MoSe2, and WSe2) for which 2H is the thermodynamically stable coordination (Fig. 2c).26 In the 2H configuration these TMDs are semiconductors, covering a range of gaps between 1 and 2 eV, while they are metallic in the less stable 1T polymorph. Although the bandgap is indirect in the case of the bulk materials, it becomes direct when they are thinned down to single layers. The direct bandgap, along with their high absorption coefficients (>107 m−1),27 make TMDs interesting materials for optoelectronic applications.

3. Graphene-based heterostructures

3.1. Thin film graphene heterostructures

As mentioned in the Introduction, the extraordinary properties of graphene have some drawbacks, in most cases related with the absence of a bandgap.2 Considering the wide range of existing 2D materials, and with some of them exhibiting complementary properties, their integration into heterostructures is interesting and promising.10 Heterostructures can vastly increase the potential of such materials, bringing unique functionalities for such thin materials. The properties of the different materials in the heterostructures can be combined to produce all-2D devices with new and interesting properties and applications, including novel electronic and optoelectronic devices.28 Among the most significant properties that graphene can contribute for heterostructures are its high conductivity and carrier mobility, along with its transparency and mechanical flexibility.29

Given the layered nature of most of the 2D materials, the most evident and feasible way to construct such heterostructures is by artificially stacking layers of the different materials.10,28 van der Waals 2D layered materials present a high anisotropy between the in-plane and out-of-plane directions, with strong covalent bonds within the layer but weak van der Waals interactions between the layers. As already seen, this weak inter-layer interaction is essential to isolate single layers by exfoliation,7 but also permits the stacking of completely different materials without restrictions arising from the mismatch in their crystal lattices.28 This allows the fabrication of virtually any kind of vertical heterostructures by mechanically piling up the different layers in the desired sequence.10 As we will see in the following sections, the fast advance in the processing of 2D materials has provided an easy way to produce quite complex vertical heterostructures with diverse functionalities. Routes for synthesizing such heterostructures by avoiding mechanical stacking are being pursued at the same time, in order to achieve the high throughputs required for their mass production.28

Lateral, in-plane heterostructures in which one-dimensional junctions are created between the different materials (Fig. 1) are also interesting.30–32 These kinds of heterostructures open the possibility to obtain ultimately thin devices. However, the development in this case has been slower than for vertical heterostructures. There are two main reasons for this. The first and foremost is the lack of a straightforward method to produce such lateral heterostructures, similar to that of mechanical stacking for vertical heterostructures. Thus, all of the existing methods rely on the bottom-up synthesis of at least one of the materials. The second reason arises from the necessity to realize in situ processing techniques, in which one of the materials starts to grow from the edge of the other. This can result in damage of the original material, requiring the development of synthesis techniques compatible for all the involved materials. Nonetheless, as will be seen in the corresponding section, some interesting advances have been realized in the last few years and a few in-plane graphene-based heterostructures have been reported. Although in most cases the heterostructures present a small overlap in the junction, some cases of truly 1D stitching have been observed at the interface between graphene and h-BN.33–35

3.2. Stacked van der Waals graphene heterostructures

3.2.1. Synthesis methods: mechanical stacking vs. direct growth. Mechanical stacking of 2D layered materials was first demonstrated in 2010 by Dean et al. for the production of graphene FET devices supported by h-BN.8 The process started with the mechanical exfoliation of a thin film of h-BN onto the target substrate, along with the exfoliation of graphene on a PMMA thin membrane. Graphene was then carefully aligned with h-BN, and both were brought into contact with the aid of a micromanipulator (Fig. 3a). By repeating this procedure it is possible to stack an arbitrary number of additional layers of different materials,36–41 thus increasing the complexity and the functionality of the potential heterostructures that can be assembled (Fig. 3b). However, this method is likely to introduce some contamination to the interface of the stacked layers, due to the processes involved during the transfer of the different layers.42 The presence of these contaminants has a negative effect on the quality of the heterostructure.41 Almost pristine interfaces could be realized after annealing the heterostructure at moderate temperatures (300 °C), which induces the aggregation of most of the contaminants into small delimited regions of the interface.41,43 However, even these low temperatures may result in permanent damage to certain heterostructures.44 The methods for mechanically stacking 2D materials have evolved through the years by exploiting the van der Waals interactions existing between them, resulting in different dry transfer methods. In the dry transfer methods the first exfoliated flake is employed to progressively pick up the rest of the exfoliated layers that will form the final heterostructures (Fig. 3c).44–47 This method results in cleaner interfaces between the stacked layers by avoiding direct contact of the interface surfaces with polymers or solvents. It should still be noted that these surfaces are in general exposed to ambient conditions during the exfoliation and stacking, leading to the unavoidable presence of water, hydrocarbons and other kinds of contamination.48 Transfer in an inert atmosphere can however be realized in the case of materials unstable in air, such as NbSe2.49 Currently employed stacking techniques allow the realization of highly oriented heterostructures with misalignments of the lattices below 2°, with enough precision to fabricate artificial stacks of Bernal-stacked bilayer graphene.50 Moreover, van der Waals interactions between the layers can also assist in increasing the degree of alignment of the heterostructures, by adjusting the relative angle to the most stable configuration.51 In this sense, recent studies have demonstrated the self-alignment of large flakes of dry-transferred graphene on h-BN by annealing at temperatures over 100 °C in Ar or Ar/H2 (Fig. 3d).51,52 The misaligned angle ends up being below ∼0.7° after annealing, producing highly aligned heterostructures.51 Although 0° is the most stable configuration, flakes initially twisted by ∼20° tend to rotate to 30° during annealing.52 The thermal self-alignment is driven by the minimization of interaction energy between the layers, so it is expected to be also present for other different layers, and in fact has been successfully tested for graphene-on-graphene stacks.51,53 Solely relying on the van der Waals interactions between the different 2D layers, the dry-stacking methods can result in a low transfer yield. The transfer process can be improved by exploiting the different thermal coefficients of the polymers employed for the transfer, and those of the 2D layers.43 Recently, a new technique has been developed to exfoliate and transfer large-area single-layer TMDs onto several substrates.54 A thin film of Au is evaporated on top of the bulk TMD and then peeled using a thermal release tape. The large affinity between the Au film and the chalcogen ensures the peeling of only the top layer of the TMD, allowing large, isolated single layers with lateral sizes up to a few hundred microns to be obtained. The optical and electronic quality of these samples rival those obtained using other exfoliation techniques, while the size and homogeneity obtained are far superior.54
image file: c7cs00160f-f3.tif
Fig. 3 (a) Schematic of the alignment and stacking of a graphene flake on h-BN with the aid of a micromanipulator. (b) Sectional STEM image and schematic of a heterostructure fabricated by the sequential stacking of graphene and h-BN layers. (c) Schematics for the polymer-free stacking of 2D materials aided by van der Waals interactions. In this case, graphene is picked up by an h-BN flake, and then both are transferred onto another h-BN flake to encapsulate the graphene. (d) AFM image showing the rotation of a graphene flake on h-BN after annealing at 400 °C. (a) Adapted by permission from Macmillan Publishers Ltd: Nat. Nanotechnol., (ref. 8), copyright 2010. (b) Reprinted by permission from Macmillan Publishers Ltd: Nat. Mater., (ref. 41), copyright 2012. (c) From ref. 45. Reprinted with permission from AAAS. (d) Reprinted figure with permission from ref. 52. Copyright 2016 by the American Physical Society.

The use of layers exfoliated from the bulk materials make the mechanical stacking methods appropriate for fast prototyping of layered heterostructures without the need to develop methods to directly grow 2D versions of the materials. Another advantage of these approaches is that a lattice match between the two materials forming the interface is not required, and thus completely different materials can be piled up together without any restriction. Moreover, control of the relative rotation between the different layers can be attained with a high precision,50 allowing the fine tuning of the properties of the heterostructures. Consequently, these methods have been widely used for fundamental studies to produce a vast amount of different heterostructures, such as several kinds of TMDs with graphene, including MoS227,44,55–62 and WS2;27,44,63 the mentioned case of h-BN and graphene;8,36–38,40,41,44,45,63–67 mixed heterostructures containing graphene, h-BN and TMDs;65,66 and graphene heterostructures with other materials.44,57

Although quite convenient from the point of view of basic research, mechanical stacking methods are not yet suitable for the production of graphene heterostructures at an industrial scale. Efficient fabrication methods that can be easily scaled-up to mass production are thus required. Methods for the direct growth of graphene heterostructures have been developed, aimed to increase the productivity. At the same time, direct growth methods can provide much cleaner interfaces than those obtained in mechanically stacked samples, especially when two-step growth approaches are employed. Direct bottom-up growth methods have already been reported for the synthesis of h-BN on top of graphene68,69 and of graphene on h-BN;70–84 of several TMDs on graphene, including MoS2,69,85–96 WS2,97–100 MoSe2,101 WSe2,69,90,102,103 NbS218 and NbSe2;104 of successive stacks of different TMDs on graphene;105 and of other materials such as GaSe,106,107 non-transition metal chalcogenides,108–110 single layers of metal oxides such as ZnO,111 or perovskites112 on graphene. Most of these methods to synthesize heterostructures have evolved along with the possibility to synthesize isolated layers of the respective materials using CVD or related methods. In the following sections, some of these methods will be covered in more detail.

3.2.2. h-BN/graphene heterostructures. Apart from the methods based on mechanical stacking, several procedures have been developed for the growth of h-BN on graphene. A few methods have been reported for the direct synthesis of h-BN on graphene.68,69 Liu et al. reported the two-step synthesis of h-BN on graphene.68 First, graphene was grown on Cu foil using CVD at 950 °C using hexane as the carbon feedstock. The h-BN was then grown on the graphene using a second CVD with ammonia borane as the precursor. The thermal decomposition of ammonia borane at 1000 °C results in a film of h-BN almost homogeneously covering the graphene. However, Raman spectroscopy reveals that the graphene is damaged during the second CVD.

The synthesis of graphene on h-BN is more widely studied than the opposite, as h-BN is better able to resist the second CVD growth compared to graphene, and such a structure is more useful from the point of view of applications.70–74,76–84 In early attempts to grow single-layer graphene/h-BN heterostructures, graphene was grown on an epitaxial h-BN film on Ni(111) by exposure to large amounts of benzene at 800 °C.70 More recent studies show the CVD growth of multilayer graphene on h-BN flakes71 and of single-layer graphene on CVD-grown h-BN,74,75 both conducted at 1000 °C by using CH4 as the carbon feedstock (Fig. 4a and b). The growth of graphene using CVD using CH4 usually requires a catalytic surface to decompose the CH4 molecules, with the growth being self-limiting as the graphene starts to cover the catalyst. The mechanisms for the growth of graphene on h-BN are not completely clear yet, but most probably involve some catalytic transparency of thin h-BN layers when supported on a surface such as Cu.75,84 Nevertheless, the growth procedures should require certain adjustments with respect to the growth on catalytic substrates, such as increases in the feedstock partial pressures, processing temperatures and growth times.77 Molecular beam epitaxy (MBE) growth of graphene on exfoliated h-BN has also been demonstrated at temperatures of 930 °C, although the quality of graphene is not as good as in the case of CVD methods.72 The temperature of the CVD synthesis can be decreased with plasma enhanced CVD (PE-CVD), which also makes the use of a catalyst to decompose the CH4 molecules unnecessary. The use of PE-CVD allows the growth of graphene on exfoliated h-BN using CH4 as the feedstock at temperatures as low as 500 °C.78 However, these procedures usually result in small graphene grains with sizes well below 1 μm. Graphene can also be synthesized on h-BN by LPCVD using liquid precursors.77,83 By using h-BN/Cu(111)77 and h-BN/Rh(111)83 as the growth substrates, Roth et al. demonstrated that graphene can be grown with 3-pentanone at ∼830 °C and ∼880 °C, respectively, using LPCVD. In the case of h-BN/Rh(111), a two-step LPCVD procedure was used to first decouple the h-BN from the substrate, resulting in a flatter surface.83 In both cases, the graphene is completely aligned with the underlying h-BN, even though a lattice mismatch exists between them (∼1.8%) (Fig. 4c and d).77,83 Graphene/h-BN heterostructures have also been obtained by the confined growth of graphene in the space between Ni(111) and a CVD-grown disoriented h-BN layer (Fig. 4e).82 The growth of the graphene in this case is done by LPCVD at a relatively low temperature (540 °C), using ethylene (C2H4) as the carbon feedstock. Song et al. have recently developed a procedure for the growth of patterned graphene on h-BN by using PMMA seeds as nucleation points for the graphene and benzoic acid as the precursor (Fig. 4f–h).84 The growth of patterned heterostructures can be interesting for upscaling the growth of 2D-based electronic circuits.

image file: c7cs00160f-f4.tif
Fig. 4 (a) Schematics of the growth of single-layer graphene on h-BN using CVD. (b) STM image of the graphene grown on h-BN, showing a Moiré pattern with a period of 0.55 nm. Size of the area is 10 × 10 nm2. Inset is the FFT of the image. (c and d) LEED patterns taken from a layer of h-BN on Cu(111) (c) and after the growth of graphene (d). (e) Schematics representing the growth of graphene in the space between Ni(111) and a disoriented h-BN layer. (f) Schematics of the growth of patterned graphene/h-BN heterostacks on Cu foils from PMMA seeds. (g and h) SEM images showing patterned domains. (a and b) Reproduced with permission from ref. 74. Copyright 2013, John Wiley and Sons. (c and d) Reprinted with permission from ref. 77. Copyright 2013 American Chemical Society. (e) Reprinted with permission from ref. 82. Copyright 2015 American Chemical Society. (f–h) Reprinted with permission from ref. 84. Copyright 2016 American Chemical Society.

The use of h-BN as a substrate for graphene originates from the limitations arising from using commonly employed substrates, such as SiO2. These limitations include the surface roughness of the substrate, the presence of charge traps and surface optical phonons, and other inhomogeneities.8,113,114 All of these have a negative impact on the performance of graphene-based electronic devices. The benefits of h-BN as a substrate compared to SiO2 include its chemical inertness, and its ultra-flat surface free from dangling bonds and of charge impurities.8 Recent LEEM based studies have determined that for misoriented lattices there is no electronic coupling between graphene and h-BN.115 Thus, the band structure of graphene in the proximity of the Dirac point is not perturbed. As already mentioned, mechanically exfoliated h-BN7 was first employed as a substrate for graphene in 2010.8 Graphene on h-BN has been proved to be much flatter than on SiO2 (Fig. 5a), with a surface roughness similar to those measured for graphene on other atomically flat substrates such as mica or of freshly cleaved graphite, and without surface charge inhomogeneities (see Fig. 5b and c).8,113,116 This is accompanied by an improvement of the electronic properties of graphene, with the mobility increasing roughly an order of magnitude and almost no doping induced from the substrate, as shown in Fig. 5d.8,64,116–118 Taking a step forward, the electronic properties of graphene can be further improved by encapsulating it between h-BN layers, attaining mobilities over 105 cm2 V−1 s−1 at room temperature.36,119 It should be noted that the mobilities achieved on SiO2 rarely exceed the values of 2 × 104 cm2 V−1 s−1.2,8 By optimizing the transfer process of the layers, mobilities up to 1.5 × 105 cm2 V−1 s−1 have been realized on exfoliated graphene at room temperature, and of ∼5 × 105 cm2 V−1 s−1 at temperatures below 20 K.44 In the case of CVD graphene, carrier mobilities have been reported as large as 3.5 × 105 cm2 V−1 s−1 (1.6 K), and of 3 × 106 (1.8 K) after optimization of the transfer.67 Graphene encapsulated in h-BN layers shows ballistic transport at the micrometer scale, for distances of ∼1 μm at 200 K and of more than 28 μm at temperatures below 2 K.36,44,45,67 Furthermore, encapsulation with h-BN layers protects the graphene against the environment. Thus, devices show similar behavior in air and under vacuum,119 while the chemical stability of h-BN preserves the integrity of graphene under extreme environments, such as oxidizing atmospheres at high temperatures.120 Encapsulation with 2D materials can also be used as a general way to protect other 2D materials under harsh conditions, as shown by the decreased damage in graphene-encapsulated 2D materials exposed to the high energy electron beams of TEM.23,121,122 As we will cover in the following sections, h-BN has also been proved to be an efficient ultrathin dielectric gate36,123 and a capable thin tunnel barrier in the fabrication of vertical 2D heterostructures, with the barrier thickness going down to a single layer.40,124

image file: c7cs00160f-f5.tif
Fig. 5 (a) Surface roughness of graphene on SiO2 and on h-BN with different thicknesses. (b and c) Charge density for graphene on h-BN (b) and graphene on SiO2 (c). (d) Transfer characteristics of a single-layer graphene FET on h-BN (red) and on SiO2 (blue). (a) Adapted by permission from Macmillan Publishers Ltd: Nat. Nanotechnol., (ref. 8), copyright 2010. (b and c) Adapted with permission from ref. 117. Copyright 2011 American Chemical Society. (d) Reprinted with permission from ref. 118.

Apart from the aforementioned increase of the carrier mobility of graphene, h-BN has been demonstrated to be an ideal substrate to realize some of the most exotic properties of graphene. These include fractional quantum Hall states and broken symmetries,125–127 long-distance spin transport over lengths of 20 μm128 and quantum spin Hall states,129 among many others. Furthermore, specific effects dependent on the stacking angle arise from the small lattice mismatch between the graphene and the h-BN. The presence of h-BN induces a periodic potential characterized by a Moiré pattern, which can be imaged using STM and AFM (Fig. 6a–c).52,78,130 For a relative orientation (ϕ) between the h-BN and the graphene, the wavelength of the formed superlattice is given by image file: c7cs00160f-t1.tif, where a is the lattice constant of graphene and δ the lattice mismatch with the h-BN.130 The superlattice is oriented by an angle (θ) with respect to the graphene, given by the expression image file: c7cs00160f-t2.tif. As reflected in the expression, θ depends on the relative orientation of the graphene and the h-BN (ϕ), being 0° when both materials are oriented.130 The presence of this periodic potential can significantly alter the electronic properties of graphene. One of the effects of the small lattice mismatch is the possibility to open a bandgap in the graphene.131 Transport measurements indicate that the periodic potential induced by the lattice mismatch of h-BN and graphene can open a measurable bandgap at the main Dirac point of the latter of ∼27 meV when both are highly aligned (Fig. 6d and e).132 The magnitude of such bandgaps decreases as the lattices misalign, being absent in highly twisted heterostructures.116,132 Recently, an even larger bandgap (∼160 meV) has been observed using ARPES for aligned graphene on h-BN.133 The bandgap arises from an asymmetry between the two sublattices of graphene when it is placed near the h-BN.116 However, the lattice mismatch between h-BN and graphene reinstates the symmetry on a spatial average, and hence no bandgap should be observed. In this sense, a gap has not been observed in other studies by STS,116,130 or in ARPES for aligned graphene grown on h-BN/Cu(111)77 and on h-BN decoupled from Rh(111).83 The experimental discrepancies probably arise from the strain-induced increase of the level of commensuration observed for small misalignment angles, which leads to a global disappearance of the symmetry on the graphene sublattices.134 These commensurate areas have been observed to disappear when graphene is encapsulated in h-BN, leading to the vanishing of the bandgap (see Fig. 6f).134 On the other hand, recent experiments have shown that the commensuration level can be increased by low temperature annealing in Ar/H2 due to an increase of alignment.51

image file: c7cs00160f-f6.tif
Fig. 6 (a–c) Moiré patterns observed by STM for graphene on h-BN and dependence on the relative orientation between both. The relative orientations are 21° (a), 7° (b) and 0° (c). (d) Measured bandgaps for graphene deposited on h-BN as a function of their relative orientation. λ is the wavelength of the Moiré pattern. (e) Schematics of the band structure of graphene (top) and of graphene on top of h-BN (bottom), showing the opened band gap (Δ) and the appearance of secondary Dirac points. (f) Longitudinal resistivity at different temperatures of graphene supported on h-BN before and after (inset) encapsulation with h-BN. After the encapsulation the resistivity does not increase at lower temperatures, evidencing the disappearance of the gap. (g) STS dI/dV curves of graphene on h-BN with two different Moiré wavelengths (9.0 nm, black curve; 13.4 nm, red curve). The arrows indicate the position of the SDCs. (h) Energy difference between the SDCs and the original Dirac point with respect to the Moiré wavelength. (i and j) ARPES measurements of graphene on aligned h-BN. (i) Band gap of the Dirac point, observed after doping the sample with rubidium so that the gap appears in the ARPES. (j) Band gap at the SDCs. (a–c) Adapted with permission from ref. 117. Copyright 2011 American Chemical Society. (d and e) From ref. 132. Reprinted with permission from AAAS. (f) Reprinted by permission from Macmillan Publishers Ltd: Nat. Phys., (ref. 134), copyright 2014. (g and h) Reprinted by permission from Macmillan Publishers Ltd: Nat. Phys., (ref. 130), copyright 2012. (i and j) Adapted by permission from Macmillan Publishers Ltd: Nat. Phys., (ref. 133), copyright 2016.

Another effect of the superlattice periodic potential caused by the h-BN is the emergence of different sets of Dirac cones. A set of first-generation Dirac cones (FDC) appears at the same energy level of the original Dirac cone in the center of the six nearest superlattice Brillouin zones (SBZs). These have been reported using ARPES both on aligned and misaligned graphene/h-BN.133,135 A set of secondary Dirac cones (SDCs) also appear in highly aligned heterostructures on the edges of the SBZ, both on the valence and conduction bands of graphene (Fig. 6e). The presence of the SDCs has been observed using STS (Fig. 6g),130 transport measurements52,78,132,136 and ARPES.133 The energy difference of the SDCs with respect to the original Dirac cone increases with the misalignment of the graphene and h-BN (Fig. 6g and h).52,130 This, along with weaker interaction between graphene and h-BN at large misorientation angles makes it difficult to observe the SDCs for misaligned samples.135 ARPES measurements have determined the existence of a bandgap of ∼100 meV for the SDCs in highly aligned heterostructures, similar to that observed for the original Dirac cone (Fig. 6i and j).133 The modification of the band structure of graphene with superlattices is interesting from the viewpoint of applications, as it allows the tuning of the electronic properties of the heterostructures. For example, anisotropic graphene heterostructures can be realized, as the additional Dirac points induce an anisotropy in the velocities of the charge carriers in graphene.130 The presence of the satellite Dirac points have been observed to cause Fabry–Pérot resonances, indicating the possibility to confine electrons by engineering the periodic potential caused by the h-BN.137

3.2.3. TMD/graphene heterostructures. Heterostacks of graphene and TMDs can be obtained using the general mechanical stacking methods covered in Section,44,55–63 Here, we will briefly introduce other synthesis methods and some of the most relevant kinds of heterostructures that have been realized to date. Several TMDs have been successfully grown on graphene.18,69,85–105 Apart from the mechanical stacking, the most widespread methods to synthesize graphene/TMD heterostructures rely on the CVD synthesis of the TMD over the graphene. Usual temperatures to grow high-quality graphene using CVD (>1000 °C) can damage the TMD. This, along with the need of a metal catalyst surface to grow graphene, makes this approach more prevalent than the growth of the graphene on the TMD. The most studied TMD to date is MoS2, and hence the literature of MoS2 grown on graphene outnumbers that of other TMDs.69,85–96 One of the earliest reported attempts to grow few layer MoS2 on graphene was by Shi et al.,85 using ammonium thiomolybdate ((NH4)2MoS4) deposited on the graphene (Fig. 7a). The growth was done by the sequential decomposition of (NH4)2MoS4 at 400 °C in the presence of hydrogen. However, this approach is complicated and requires several steps. Besides, the control of the number of layers is not assured, while the MoS2 is composed of randomly oriented small crystals. Using copper hexadecafluorophthalocyanine (F16CuPc) as a seed promoter, Ling et al. demonstrated that it is possible to grow MoS2 on exfoliated graphene using APCVD at a relatively low temperature of 650 °C (Fig. 7b).86 By using higher temperatures for the graphene substrate (∼960 °C), the growth of MoS2 on graphene using APCVD without promoters has been demonstrated (Fig. 7c and d).87,94 Given that the graphene is obtained using CVD, it should be noted that large-scale heterostructures are also possible. The orientation of the MoS2 coincided in this case with that of the graphene template, indicating heteroepitaxial growth (Fig. 7d). This method has also been used on graphene nanoribbons, allowing the growth of nanoribbon heterostacks of MoS2/graphene with both partial and total MoS2 coverages.88 The use of low pressures during the CVD also allows the rotationally commensurate growth of MoS2 on epitaxial graphene (EG) at 800 °C (Fig. 7e and f).93 Control of the orientation is essential, as the electronic structure of TMDs/graphene depends on the relative orientation,138 as in the case of graphene/h-BN.
image file: c7cs00160f-f7.tif
Fig. 7 (a) AFM phase image of MoS2 grains grown on graphene/Cu. (b) Optical image of MoS2 grown on exfoliated graphene/SiO2 by using a seed promoter. (c and d) SEM image (c) and selected area diffraction (d) of MoS2 growth on graphene by CVD. Bottom of (d) shows an atomic model schematic of MoS2 grown on graphene determined by the SAED. (e and f) AFM image of MoS2 grown on EG (e) and the grain orientation extracted from it (f). (g) SEM image of WS2 flakes grown on EG. (h) SEM image of an NbS2 grain grown on graphene by CVD. (a) Adapted with permission from ref. 85. Copyright 2012 American Chemical Society. (b) Reprinted with permission from ref. 86. Copyright 2014 American Chemical Society. (c and d) Adapted with permission from ref. 87 and 94. Copyright 2015 and 2016 American Chemical Society. (e and f) Reprinted with permission from ref. 93. Copyright 2016 American Chemical Society. (g) Adapted with permission from ref. 98. (h) Reproduced from ref. 18 with permission from The Royal Society of Chemistry.

For the CVD growth of MoS2 on graphene, the most used precursors are S and MoO3, although Cl-based compounds can also be used as the Mo feedstock.139 MoO3 is located on the hot area of the furnace and usually close to the substrate, while S is placed upstream on a colder area (<200 °C), and both are carried to the growth area by a stream of either Ar or Ar/H2. Thus, the precursors can be changed for the growth of different TMDs, e.g. using WO3 and Se instead of MoO3 and S, respectively, to obtain WSe2. In this way, WS2 has been epitaxially grown on CVD graphene as well as on EG on SiC using WO3 and S as precursors (Fig. 7g).98,99 Kim et al. developed a slightly modified approach in which a thin film of WO3 deposited on SiO2 is placed in direct contact with the graphene during the growth process.99 By using Se, the CVD growth of MoSe2101 and of WSe269,102 on graphene has also been demonstrated. Direct growth of a thin film of NbS2 has been also achieved on graphene using NbCl5 and S as precursors (Fig. 7h).18 In this case, graphene proved to be a critical factor to favor the lateral growth of the NbS2 nanosheets.

In the previously mentioned studies the TMD is grown either on exfoliated graphene or on already synthesized graphene in a different process. It is interesting to grow the heterostructures sequentially, without the need to remove the sample from the processing chamber between the growth of the different materials. In this sense, Shi et al. reported the production of MoS2–graphene heterostructures in a two-step CVD.91 First, graphene is grown on Au using CH4 at ambient pressure, and then a subsequent step was conducted at low pressure for the growth of MoS2 from MoO3 and S (Fig. 8a). However, the MoS2 grains are randomly oriented with respect to the underlying graphene, as opposed to other studies (Fig. 8b). The two-step CVD growth of heterostructures is still interesting in terms of processing at large scales and to obtain cleaner heterostructure interfaces. Although scarce, other methods apart from CVD have been reported for the growth of TMDs on graphene, such as MBE,92,96 sulfurization of deposited metal89,95 and metal–organic CVD (MOCVD).103

image file: c7cs00160f-f8.tif
Fig. 8 (a) Schematics for the two-step CVD growth of MoS2 on graphene. (b) SEM image of the as-grown MoS2/graphene on Au using the two-step CVD. (c–e) Determination of the grain structure of graphene by growing oriented MoS2 grains on it. SEM images showing a large region containing several domains (c) and enlarged images of some grains (d and e). The numbers represent the relative orientations of the grains. (f) Transfer curve of a graphene FET supported on MoS2 and encapsulated with an h-BN layer serving also as a dielectric gate. The inset shows an image of the device. (a and b) Reproduced with permission from ref. 91. Copyright 2015, John Wiley and Sons. (c–e) Adapted with permission from ref. 94. Copyright 2016 American Chemical Society. (f) Adapted with permission from ref. 44. Copyright 2014 American Chemical Society.

The possibility to control the orientation of TMD grains grown on graphene87,93 allows to manipulate and restrict the kind of grain boundaries formed in the TMD (GBs).140 This result is interesting to obtain more reproducible heterostructures, while it also opens the opportunity for the tuning of their electronic properties in a similar way as for graphene/h-BN.140,141 The oriented growth of TMDs has also been used to visualize the crystal orientation and grain structure of the underlying graphene (Fig. 8c).94 This is done by comparing the relative orientation of MoS2 grains across the surface, which allows the mapping of the crystal structure of the underlying polycrystalline graphene given the existing epitaxial relation (Fig. 8d and e). As the MoS2 nucleation is promoted at the grain boundaries of the graphene, this technique also allows their direct visualization.

An improvement in the stability in air of TMDs directly grown on graphene has been observed without the need of encapsulation.99 This is attributed to the strong interlayer interaction and to the higher crystallinity of the TMD crystals synthesized on the graphene. On the other hand, an enhancement of the electronic properties of graphene has been observed when using some TMDs as the substrate, although the performance is below that attained for h-BN substrates.59,142 Some other reports indicate that the structural defects on the crystal structure of the TMDs can have a negative impact on the electronic properties of the supported graphene.143 When enclosed between h-BN and a TMD (either MoS2 or WS2), graphene presented high carrier mobilities up to 6 × 104 cm2 V−1 s−1 (Fig. 8f).44 However, encapsulation between flat layered oxides (such as mica, BSCCO or V2O5) resulted in graphene mobilities of only ∼103 cm2 V−1 s−1.

3.2.4. Applications.
(a) Optoelectronics. Among the many applications of graphene-based heterostructures, the most interesting ones are those in the field of optoelectronics. Optoelectronic devices are electronic devices with the potential to interact with electromagnetic radiation, mainly in the visible region of the spectrum. These devices are mostly used to control, detect or generate light. Graphene can interact with light with a broad range of wavelengths, with a wavelength independent adsorption of ∼2.3% of the incident light in the visible and infrared range for a single layer.29 This, along with its high response speed and carrier mobility, and its flexibility, are interesting properties for optoelectronic applications. However, the lack of a bandgap also translates into optoelectronic devices with very poor characteristics due to the short lifetime of exciton separation.29 On the other hand, most TMDs are direct-bandgap semiconductors when thinned down to single layers, also exhibiting large optical absorptions (>107 m−1).27 Moreover, the interaction of graphene with the TMDs can be used to modify their electronic and optoelectronic properties,144,145 which opens new ways to realize optoelectronic devices. In the following sections we will show some recent results of graphene heterostructures with application in optoelectronics. In most of these heterostructrures, the TMD is used as the optical active material.
(b) Broadband ultrafast photodetection and light harvesting. Photodetectors are devices designed to convert optical signals to electric signals. This is generally conducted in three steps, with the light generating charge carriers, which then need to be separated and carried to each electrode as a photocurrent. The efficiency of a photodetector is usually expressed by the external quantum efficiency (EQE) (sometimes expressed as η), a parameter that represents the fraction of incident photons that are converted into photocarriers. Another commonly employed parameter is the responsivity, R, a measure of the electrical output per optical input and with units of A W−1. The responsivity and the EQE are related by the expression image file: c7cs00160f-t3.tif, with e being the elementary charge and the energy of the incident photons.

Stacks of CVD-grown graphene with TMDs have been extensively studied to produce highly sensitive photodetectors. One of the approaches to make graphene/TMD photodetectors is to use the TMD as a photogate, with the photocurrent flowing through the graphene. In this way, Roy et al. fabricated photodetectors by stacking single-layer graphene on few-layer MoS2 (5 to 10 layers) (Fig. 9a).55 The transfer characteristics of the graphene are influenced by the underlying MoS2, which screens the gate for gate voltages (VG) larger than the MoS2 threshold voltage (VT) (Fig. 9a). Upon illumination for VG < VT, a gate-dependent photocurrent (IP) is observed in the graphene due to the injection of photoelectrons from the MoS2. At low illumination powers (∼1–6 mW m−2) these devices show high responsivities of 5 × 108 A W−1 and 1010 A W−1 at RT and at 130 K, respectively, with a calculated external quantum efficiency (EQE) of ∼32% at 130 K.55 A higher recombination induced by the increase of electron–hole pairs results in the decrease of responsivity with increasing illuminating power, to values of ∼6 × 107 A W−1 at 130 K for a power of 400 mW m−2. The responsivity also decreases for thinner MoS2 flakes,55 with similar studies showing maximum responsivities for single-layer MoS2 of ∼107 A W−1 for 10 mW m−2[thin space (1/6-em)]60 and of 10 A W−1 for a power of 2.2 W m−2.146 The trapped holes in the MoS2 also induce a slower response of the devices, with values over 103 s.55 As we shall see later, this persistent photocurrent can be used to produce memory devices (see Fig. 9b).55

image file: c7cs00160f-f9.tif
Fig. 9 (a) Transfer curve of graphene on MoS2 with (red) and without (black) illumination. Inset is a schematic of the device. (b) Dependence of the photocurrent on the gate voltage for the device shown in (a). Shaded regions correspond to illumination periods. (c) Schematics of a photodetector composed of WS2 encapsulated with graphene electrodes and h-BN. (d) Photocurrent maps taken on the device shown in (c) for different gate voltages and without applied bias. Signal is only observed in the region where the layers overlap. (e) Schematic of the MoS2/h-BN/graphene photodetector with the h-BN tunnel barrier to decrease the dark current. (f and g) Schematic of a single-gated graphene/MoS2 with a metal electrode (f), and laser power dependence of the EQE for different excitation wavelengths (g). (h) Schematic of a photodetector composed of graphene encapsulated in h-BN, and coupled to a silicon waveguide. Right bottom inset: Dependence of the response time of the graphene autocorrelator to the excitation power. (a and b) Adapted by permission from Macmillan Publishers Ltd: Nat. Nanotechnol., (ref. 55), copyright 2013. (c and d) From ref. 27. Adapted with permission from AAAS. (e) Reprinted with permission from ref. 150. Copyright 2017 American Chemical Society. (f and g) Reprinted by permission from Macmillan Publishers Ltd: Nat. Nanotechnol., (ref. 58), copyright 2013. (h) Adapted with permission from ref. 151. Copyright 2015 American Chemical Society.

Another kind of photodetector relies on the encapsulation of one or several TMDs with graphene27,58 (Fig. 9c). In this case, the TMD is used as the absorption layer, while the graphene layers act as the top and bottom electrodes, owing to its transparency and good electronic properties. If a bias is applied between the two graphene layers or they are doped differently (either chemically or by an electric gating), the photocarriers are extracted from the TMD as a photocurrent. The generated photocurrent is observed only when the areas in which all the layers overlap are illuminated (Fig. 9d).27 It is worth noting that in these vertical devices the direction of the photocurrent is the same as that of the built-in electric field generated by the gate voltage or by doping the graphene layers. Thus it is possible to modulate the photocurrent with the gate, and hence these devices can be operated at zero bias voltage (Fig. 9d).27,58 This greatly reduces the dark current that originates from biasing the device, decreasing the power consumption of the devices. The transparency of the graphene electrodes allows for it to cover the active areas of the device. Thus, the collection of the photocarriers can be done vertically instead of the lateral collection occurring with lateral metal contacts. The charge extraction is then faster, which decreases the number of carriers lost due to recombination. This can be seen in the responsivity increase reported for MoS2/WSe2 stacks with (∼10 mA W−1 at 920 W cm−2) compared to without (∼2 mA W−1 at 100 W cm−2) graphene electrodes.147 This fast extraction, along with the high mobility of graphene, also results in ultrafast photodetectors. Response times of 5.5 ps have been reported on graphene/WSe2/graphene photodetectors by Massicotte et al.,148 which is comparable to that of graphene. These devices present IQE values exceeding 70% in thin WSe2 devices (3 layers) and low dark currents, although the EQE is low (7.3%) due to the low optical absorption. Consequently, the EQE can be increased for thicker TMDs (up to 50 nm), with reported values up to 27% (MoS2),58 33% (WS2)27 and 34% (MoS2/WSe2 junction).147 In the case of thin TMDs, responsivities of 2.5 A W−1 and 3.5 A W−1 have been demonstrated in the case of single- and double-layer WS2, respectively.149 Interestingly, these values were obtained under ambient conditions for a large illumination power (∼104 W cm−2). Introducing a relatively thick h-BN tunnel barrier between the MoS2 and graphene has recently proved to be an effective method to decrease the dark current in photodetectors (Fig. 9e).150 Photodetectors with responsivities as high as 180 A W−1 were fabricated by using a 7 nm thick h-BN layer, with a relatively fast response time of ∼0.25 s.

Configurations in which one of the graphene electrodes is replaced by a metal have also been tested. In these configurations, the TMD makes a Schottky contact with graphene and an almost ohmic contact with the metal (Fig. 9f and g).58 An increment of the photocurrent and of the EQE can be observed in this configuration compared with the conventional one, which in the case of MoS2 implies an increase from 27% to 55%.

Recently, ultrafast photodetection has also been demonstrated on h-BN/graphene/h-BN heterostacks on top of a photonic integrated circuit, and with a top electrolyte gate (Fig. 9h).151 The large mobility attained by encapsulating the graphene in h-BN (up to 60[thin space (1/6-em)]000 cm2 V−1 s−1) resulted in response times as short as 3 ps. The use of the integrated waveguide and the top gate allows responsivities up to 0.36 A W−1, which is a 3-fold increase with respect to previous graphene photodetectors integrated in a waveguide.151

The optical properties of the graphene heterostructures can be also applied for light harvesting, transforming the incident light into an electric current. Heterostructures containing TMDs have become an attractive alternative to traditional semiconductors, owing to their bandgaps in the visible range of the electromagnetic spectrum, their high optical absorption coefficients, and the relative accessibility and abundance of their compounds. The efficiency of ultrathin graphene-based photovoltaic devices cannot directly compete with that of much thicker devices based on conventional semiconductors (e.g. 1 μm thick GaAs based solar cells),152 but the power densities can be 3 orders of magnitude higher than in conventional solar cells.152 Some recent attempts have shown a photoelectric conversion efficiency of 3.3% under solar simulator (AM 1.5) illumination for few-layer graphene/WS2 heterostructures with an Al bottom electrode.153 By inserting a MoS2 film into a graphene/Si Schottky junction solar cell, Tsuboi et al. have shown conversion efficiencies up to ∼11% under AM 1.5 illumination.154

(c) Light emission. Graphene heterostructures are also being studied for light emitting diode (LED) applications. Such LED devices operate by the recombination of electrons and holes in the junction of the different layers of the heterostructure. These typically contain a tunnel junction, which in most of the cases is made of h-BN. The main role of graphene in these cases is to inject electrons and holes into the active junction area. Withers et al. demonstrated the possibility to produce highly efficient proof-of-concept LEDs composed of mechanically stacked 2D layers (Fig. 10a).155 The band structure of such LEDs is tailored by inserting a TMD layer between two graphene electrodes. Two additional layers of h-BN serve as tunnel barriers between the graphene and the TMD, with the latter acting as a quantum well (QW). When a large enough bias is applied between the graphene layers, the holes and electrons are tunneled into the QW (Fig. 10b). For h-BN thicknesses of 2 layers or more, the lifetime of the injected carriers is long enough to form excitons that then recombine emitting photons (Fig. 10c and d). The wavelength of the emitted photons can be tuned by selecting a specific TMD. Also, several QWs can be piled up to increase the efficiency of the LEDs. The EQE achieved at low temperatures (below 150 K) is ∼1% for a single QW and up to ∼8.4% for devices with four QWs, which is comparable to the values of organic LEDs. However, the efficiencies for MoS2 or MoSe2 QWs decrease when increasing the temperature. The efficiency at room temperature can be increased by using WSe2, whose electroluminescence shows an opposed temperature dependence to that of Mo-based TMDs due to the opposite orientation of the spin in the valence and conduction bands on WSe2, which makes the lowest-energy exciton dark (Fig. 10e).156 This allows the production of LEDs with efficiencies of ∼5% for a single QW at room temperature and with current densities up to 1000 A cm−2, a 250-fold increase compared to MoS2 QWs.156 Interestingly, the presence of point defects on the TMD can be exploited to further tune the properties of the 2D LEDs. The defects act as single-photon sources,157 emitting spatially localized electroluminescence tuned by adjusting the vertical bias on the heterostructure (Fig. 10f).157–159 The electroluminescence from the defect-bound excitons present narrow emission lines, located at energies below those of the intrinsic excitons.158
image file: c7cs00160f-f10.tif
Fig. 10 (a) False colored STEM image of a LED composed of a heterostructure with a WS2 QW separated by h-BN layers from the graphene electrodes. (b) Band structure of a LED composed of a single QW for zero (top), intermediate (middle) and high (bottom) applied bias. (c and d) Optical images of a working LED composed of a single QW in the off (c) and on (d) states. (e) Electroluminescence of heterostructure LEDs with WSe2 (top) and MoSe2 (bottom) QWs measured at several temperatures, evidencing the increased electroluminescence with the temperature for the heterostructures with the WSe2 QW. (f) Localized electroluminescence from defects on WSe2 on a LED heterostructure. (a–d) Adapted by permission from Macmillan Publishers Ltd: Nat. Mater., (ref. 155), copyright 2015. (e) Reprinted with permission from ref. 156. Copyright 2015 American Chemical Society. (f) Reprinted with permission from ref. 158. Copyright 2016 American Chemical Society.

(d) Increasing the efficiency of optoelectronic devices. Being extremely thin, the interaction of light with graphene 2D heterostructures is lower than in bulk 3D devices. Methods to increase the amount of absorbed or emitted light by 2D heterostructures without increasing their thickness are of great importance. The coupling of 2D heterostructures with photonic integrated circuits, such as photonic crystal cavities or waveguides (Fig. 9h), can greatly enhance the light adsorption151,160 or emission.161 In this sense, integration of a LED with a photonic crystal cavity has shown a 4-fold enhancement of the electroluminescence, producing a linearly polarized emission along the cavity mode.161 The interaction with light has also been improved through the use of plasmonic nanoparticles. An example of this is the observed enhancement in the light absorption induced by Au nanoparticles (5–10 nm) deposited on graphene/WS2/graphene heterostructures, which result in an increase of the photocurrent by a factor of 10.27
(e) Memory devices. One of the strategies to produce memory devices is based on using a floating gate structure. The basic mechanism relies on modulating the conductivity of a channel by adjusting the charge stored in an electrically isolated gate. Bertolazzi et al. realized a nonvolatile cell from a graphene/MoS2 heterostructure.56 The device is composed of a FET with a MoS2 channel and graphene contacts. A floating gate is placed on top of the channel, consisting of few-layer graphene electrically isolated from the rest of the device by a HfO2 dielectric layer (Fig. 11a). The polarity of the voltage applied to the floating gate electrode will mean that electrons tunnel from the MoS2 channel into the floating gate (positive voltage, program state), or in the reverse direction (negative voltage, erase state). This will induce a shift of the threshold voltage in the transfer characteristics, with ratios of current for the erase/program states of ∼104 (Fig. 11b and c). As the induced shift depends on the amount of charge stored on the floating gate, multiple levels could be used to store information instead of just two. The charge retention of the memory proved to be enough for a few hours, with the device retaining an erase/program ratio of ∼103 within 1 h (Fig. 11c). Independently, Choi et al. realized a similar structure but using a h-BN barrier.65 The h-BN barrier allowed for similar erase/program ratios that are maintained without significant losses for at least 1400 s. The performance of such memory devices is comparable or even better than that of organic memory devices.56,65 More recently, a new concept for a two-terminal floating memory device without a gate electrode has been proposed.162 It consists of a MoS2 channel, a graphene floating gate and an h-BN barrier (Fig. 11d). By applying drain voltages larger than the operational ones, the graphene can be charged or discharged by tunneling through the h-BN from the top electrodes. Erase/program ratios as high as 109 can be obtained by adjusting the h-BN thickness, with retention times over 104 s (Fig. 11e). The absence of the gate in these devices also allows easy production of flexible devices.
image file: c7cs00160f-f11.tif
Fig. 11 (a) Schematics of a floating-gate memory device, with a graphene electrode, a MoS2 channel and a graphene floating-gate. (b) Transfer characteristics of the device in (a) collected in two different sweep directions for the gate. The hysteresis is a consequence of the charge accumulated in the graphene floating-gate. (c) Time dependence of the drain currents in the program and erase states. (d) Schematics of a two-terminal memory device consisting of a MoS2 channel and a graphene floating gate separated by h-BN. (e) Time dependence of the program and erase states in the device shown in (d). (f) Transfer characteristics of a memory device based on the light-induced doping of graphene on h-BN. The red trace is the pristine device, while the subsequent traces were collected after photodoping with different applied gates. (a–c) Reprinted with permission from ref. 56. Copyright 2013 American Chemical Society. (d and e) Reprinted with permission from ref. 162. (f) Reprinted by permission from Macmillan Publishers Ltd: Nat. Nanotechnol. (ref. 163), copyright 2014.

By taking advantage of the light sensitivity of some 2D materials, light programmable memory devices can be also realized. Roy et al. demonstrated the operation of one of these memory devices, consisting of a graphene FET on top of a multilayer MoS2, assembled using exfoliation and stacking techniques (Fig. 9a).55 These devices revealed sensitivity to light, which induces variations in the drain-source currents. These variations do not completely disappear with the light source, and instead present logarithmic relaxation times for periods of ∼100 s (Fig. 9b). Similar kinds of programmable memory devices have been studied, consisting of graphene FETs supported on h-BN.163,164 In this case, illumination of the graphene/h-BN heterostructure under an applied gate voltage induces electrostatic doping of the graphene due to defect charge migration on the h-BN (Fig. 11f).163

(f) Graphene as an electrode for FETs and logic gates. Given its outstanding electronic properties, graphene can act as an ideal contact to other 2D materials. Thus, as we have seen, in many heterostructures graphene plays the role of an electrode. An example of this is the use of graphene in highly sensitive MoS2-based NO2 gas sensors.165 The ability to tune the work function of graphene with a gate results in the possibility to adjust the Schottky barrier with other materials.166–168 This allows the formation of almost ohmic contacts of graphene with TMDs, as opposed to the case of junctions of TMDs with metals. Based on this, graphene can be employed to substitute metal electrodes for conventional devices with other 2D materials, such as TMD-channel FETs and in more complex logic gates.166,167 In these devices, graphene is used as an ohmic contact to the TMD channel (e.g. MoS2), and also as an electric contact for the top gate. This allows the fabrication of electronic circuits entirely using 2D materials. The devices with graphene contacts outperformed those with conventional metal contacts, with a tenfold increase in the measured carrier mobility.166 Interestingly, the actual processing techniques for 2D materials allow for the fabrication of all-2D circuits with these geometries at large scales. This has been demonstrated by making devices including isolated FETs and transistors integrated into logic gates.166 Techniques to realize 1D contacts can also be applied to such heterostructures, which is expected to improve the performances.45,169
(g) Vertical and tunnel FETs. The lack of a band gap in graphene results in FETs with low on/off ratios, rendering them inadequate for implementation in integrated circuits. To overcome this limitation, graphene can be stacked with other 2D materials to fabricate ultrathin vertical FETs (VFETs), in which the current flows in the vertical direction through the different layers. The VFET structure proposed by Britnell et al. is composed of a vertical heterostructure of two graphene electrodes separated by a thin h-BN insulating barrier (Fig. 12a and b).40 The device was prepared by the stacking of mechanically exfoliated flakes, which are then encapsulated in h-BN. A tunnel current is observed when applying a bias voltage between the two graphene layers. This current can be modulated by shifting the graphene bands with a gate voltage. The weak screening of the gate produced by the bottom graphene layer results in the band structure of both layers being mainly aligned, while the bias can control the difference in the chemical potential. The band structure of graphene is crucial for the VFET to work efficiently, as it allows for large shifts of the bands with the applied gate voltage. Given the small thickness of the VFET, the transit time is expected to be of only a few femtoseconds. An on/off ratio of 50 was obtained at RT, higher than those usually achieved for planar graphene FETs at RT (<10). The performance of the VFET can be increased by using a TMD barrier instead of h-BN. Due to the smaller bandgap of the TMD, the changes in the Fermi level of the graphene induced by the back gate are now of the order of the barrier height. The gate can then switch the device between an off state governed by tunneling, and an on state governed by thermionic transport.63 This results in a further increase of the on/off ratios, with reported values at RT of 104 for an MoS2 barrier (6 layer thickness),40 and of 106 for WS2 (4–5 layer thickness).63 By using graphene as the conducting channel and MoS2 as the barrier for the contact with electrodes, devices with on/off ratios up to 100 can be obtained that maintain most of the graphene mobility.170
image file: c7cs00160f-f12.tif
Fig. 12 (a) Schematics of a graphene field-effect tunnel transistor (top) and band structure for zero gate and zero bias between the two graphene layers (top), for an applied gate (middle) and for applied bias and gate (bottom). (b) Tunneling current of a device such as that of (a) for different gate voltages. (c) Bias dependence of the current density of a tunnel VFET for large applied bias, and for different applied gates (+40, 0 and −40 V for the red, green and blue curves respectively). (d) Schematics of a RF oscillator fabricated with a resonant VFET, and resonant frequency dependence on the inductance of the LC circuit. (e and f) Schematics (e) and IV characteristics of a graphene/MoSe2/Ti vertical heterostructure. An on/off ratio of 105 is achieved by tuning the gate voltage from +50 to −50 V. (g) Output and gain of an inverter device by combining n-type (graphene/MoS2) and p-type (WSe2) vertical FETs. (a and b) From ref. 40. Reprinted with permission from AAAS. (c and d) Adapted by permission from Macmillan Publishers Ltd: Nat. Nanotechnol., (ref. 172), copyright 2014. (e and f) Reprinted from ref. 177, with the permission of AIP Publishing. (g) Reproduced with permission from ref. 178. Copyright 2016, John Wiley and Sons.

By increasing the applied bias voltages, the tunnel current in VFETs with an h-BN barrier have been observed to achieve a relative maximum (Fig. 12c).171–174 This maximum is followed by a region showing negative differential conductance (NDC), with peak to valley ratios of ∼4 for temperatures <10 K (Fig. 12c). This behavior is a due to resonant tunneling occurring due to the control of the energy shift of the bands of the graphene layers with the gate and bias voltages. The performance of the resonant tunnel VFET can be increased by aligning the two graphene layers, resulting in tunneling with energy and momentum conservation and with NDC persisting at RT.50,172 One of the important uses of devices with NDC is the fabrication of high-frequency electronic oscillators. A schematic of a high-frequency oscillator made with the resonant VFET coupled to an inductor–capacitor circuit can be seen in the inset of Fig. 12d.172 The circuit is able to generate oscillations in the MHz range (Fig. 12d), although the short transition times in the femtosecond range potentially allow it to reach the THz regime. Graphene can also be used as the electrode of TMD-based resonant tunnel diodes. TMD resonant tunnel diodes have been realized by the sequential CVD growth of WSe2 and MoS2 or MoSe2 on epitaxial graphene.105 Such TMD resonant tunnel diodes achieved a NDC of 2.2 at RT.

There exists another VFET configuration in which a TMD is vertically contacted by a graphene electrode, while the other contact is a metal electrode (Fig. 12e). Such VFETs do not rely on tunnel currents, but on the Schottky contact between the graphene and the TMD. When a metal and a semiconductor are brought into contact, either an ohmic contact or a Schottky barrier is formed. In the latter case, a rectifying behavior can be observed. In the case of graphene–semiconductor junctions, the low density of states of graphene allows the tuning of the height of the Schottky barrier by the application of a gate voltage.166,167,175 As already mentioned, the contact can then be tuned from almost ohmic to rectifying, for low and high Schottky barriers respectively. Such kinds of 3-terminal variable barrier devices, sometimes known as barristors, allow a high modulation of the vertical current. By adjusting the thickness of the TMD, VFETs with on/off ratios at RT of ∼1500 (∼30 nm MoS2 and Ti contact),57 3 × 104 (∼11 nm WSe2 and Pt/Au contact),176 and 105 (∼30 nm MoSe2 and Au/Ti contact) have been realized (Fig. 12e and f).177 As opposed to the tunnel VFETs, here the TMD layer acts as a semiconducting channel instead of as a barrier. Thus, the obtained current densities are in general superior to those of tunnel-based VFETs, with values that can go up to 3.5 × 105 A cm−2.57 By replacing the back gate employed in these devices with an ion-gel gate, the voltage operation of the VFETs can be lowered to values below 3 V.178 This result is essential for operation at low power consumption. Integration of several of these VFETs into logic gates can be easily realized. As such, logical inverters with gains up to 2 have been demonstrated by combining a p-type (e.g. WSe2 or Bi2Sr2Co2O8) and an n-type (e.g. MoS2) VFET in series (Fig. 12g).57,178

(h) Transparent and flexible electronics. Owing to their optical, mechanical and electrical properties, 2D heterostructures of graphene are an interesting option for integration into transparent and flexible electronic devices. Here we will briefly describe some examples of 2D devices mounted on flexible substrates that have been already developed, although more detailed discussions can be found in ref. 179 for 2D materials in general, and in ref. 180 for graphene. Different kinds of photodetectors have been demonstrated on flexible substrates, such as PET.27,146 However their performances are not better than on SiO2, which is partially due to a performance increase on SiO2 due to the interference of light.27 However, heterostructures on PET can endure bending tests for up to 1000 cycles without any sign of fatigue (Fig. 13a and b).146 LEDs based on graphene/h-BN/TMD/h-BN/graphene heterostructures have been also realized on PET, without any evident loss in the performance under a uniaxial strain of ∼1%.155 FETs with a MoS2 channel, graphene gate and h-BN dielectric gate were demonstrated on flexible substrates.66 These devices showed mobility (27 cm2 V−1 s−1) only slightly lower than on a rigid SiO2 substrate (45 cm2 V−1 s−1). The performance was unaltered under uniaxial strains of up to 1.5%, while for larger strains the performance decreases due to damage of the bulk metal electrodes. Flexible floating-gate memory devices have been demonstrated on PET and PI/PDMS (Fig. 13c).162 The erase/program ratio on these memory devices decreased from 109 on the rigid substrate to ∼104, but was able to work for strains up to 19%. As the last example, a transparent tunnel VFET on PET has been realized, performing similar to that on a rigid substrate under strains of ∼4%.63 Overall, most of the 2D heterostructures presented in the previous sections can be integrated into flexible substrates, which is interesting for emerging fields in wearable electronics. However, further advances are yet required with respect to the fabrication or even the direct synthesis of the flexible heterostructures using low temperature growth techniques.
image file: c7cs00160f-f13.tif
Fig. 13 (a) SEM image of graphene/MoS2 phototransistors on flexible PET. Inset: Photograph of the sample. (b) Time traces of the drain current with light on and off measured before (top) and after (bottom) a bending test of 1000 cycles. (c) Flexible memory device composed of graphene heterostructures supported on PET, and strain dependence of the on and off currents. (a and b) Reproduced with permission from ref. 146. Copyright 2014, John Wiley and Sons. (c) Adapted with permission from ref. 162.

(i) Spintronics and superconductivity. Graphene is an interesting material for spintronics, with spin diffusion lengths and lifetimes at RT rivaling those of other materials. However, the spin–orbit coupling in graphene is very weak, as opposed to the case of certain TMDs. Interestingly, the spin–orbit coupling in graphene can be enhanced by the presence of nearby TMD layers.181–183 This enhancement enables the electric-field modulation of spin currents on graphene, while its electronic quality is still preserved. These studies are expected to pave the way for future high quality graphene-based spintronic circuits. Graphene has also been employed to realize superconductive states on NbSe2,49,104 which is highly unstable in air. Encapsulation by graphene (top) and h-BN (bottom) protected NbSe2 from ambient degradation, allowing a superconductive state with Tc ∼ 2 K to be achieved.49
3.2.5. Vertical heterostructures with other 2D materials: synthesis and applications. Some reports have been published on vertical heterostructures of graphene with other types of metal chalcogenides, mainly GaSe. GaSe is a layered semiconductor with a bandgap of ∼2.1 eV for single layer thickness. This gap is direct independent of the number of layers, making it a promising material for optoelectronic applications. GaSe has been grown on graphene using vapor-phase deposition at temperatures of 750 °C and at low pressures106 and using MBE at 400 °C.107 An epitaxial relation has been observed between the GaSe and the graphene, although the relative orientation observed depends on the growth conditions. The GaSe shows misalignments of ∼10° in the vapor-phase deposition growth,106 while no misalignment or rotations of 30° were observed in the case of the MBE synthesis growth.107 Photodetectors have been realized by drop casting a solution of GaSe nanosheets on graphene.184 Such photodetectors show responsivities between 102 and 3 × 105 A W−1 for illumination powers of 102 and 10−2 mW cm−2 respectively (Fig. 14a). However, photodetectors prepared by mechanical stacking of exfoliated GaSe thin layers showed lower efficiencies than those for WS2 photodetectors.27
image file: c7cs00160f-f14.tif
Fig. 14 (a) Schematics of a graphene/GaSe FET, and photoresponsivity dependence on the laser power intensity. (b) SEM image of In2Se3 flakes grown on graphene. (c) Optical image of Bi2Se3 grown on graphene. (d) STM image of COFBTA-PDA on graphene/Cu. (e) Schematic for laser writing of p–n junctions on TIPS-pentacene/graphene, and mapping of the Raman G band wavelength after writing “UI”. (f) STM image of an assembled film of phthalocyanine on graphene/Cu(100), showing a boundary of different assembled regions. (g) False colored SEM image of a perovskite grown on graphene. (h) IV curves with and without external illumination, of a device consisting of a perovskite encapsulated in two graphene electrodes. (a) Adapted with permission from ref. 184. (b) Reprinted with permission from ref. 109. Copyright 2013 American Chemical Society. (c) Reprinted with permission from ref. 108. Copyright 2010 American Chemical Society. (d) Adapted with permission from ref. 187. Copyright 2014, John Wiley and Sons. (e) Adapted with permission from ref. 191. Copyright 2014 American Chemical Society. (f) Reprinted with permission from ref. 194. Copyright 2013 American Chemical Society. (g) Reproduced with permission from ref. 112. Copyright 2015, John Wiley and Sons. (h) Adapted with permission from ref. 196. Copyright 2016 American Chemical Society.

Non-transition metal chalcogenides have also been synthesized on graphene. The epitaxial growth of single to few-layers of semiconducting α-In2Se3 on graphene using physical vapor deposition has been achieved using In2Se3 powder as a precursor (Fig. 14b).109 The orientation of In2Se3 crystals is also determined by that of the underlying graphene, reflecting the epitaxial growth. The study of α-In2Se3 heterostructures may be interesting for optoelectronic applications, as it shows a higher photosensitivity than MoS2.109 Bi2Se3 and Bi2Te2Se, both topological insulators, are other examples of non-transition metal chalcogenides that have been synthesized on graphene (Fig. 14c).108,110 The thin layers of Bi2Se3 show defined orientations on the graphene, indicating epitaxial growth.108 Vertical tunnel devices have been fabricated by mechanically stacking a Bi2Se3 flake on graphene, with the tunnel transport occurring through the interface at low temperatures.185 Spintronics devices have been demonstrated by heterostructures of graphene with another topological insulator, Bi2Te2Se.110 A spin polarized current can be efficiently injected into the graphene, owing to the strong spin–orbit coupling with Bi2Te2Se. In general, thin layers of non-transition metal chalcogenides are currently far less explored than in the case of TMDs, although their exotic properties open new possibilities for their integration in heterostructures.

Another interesting approach is the integration of graphene with organic materials or metal–organic hybrids. Covalent organic frameworks (COFs) are porous periodic frameworks composed of covalently bonded light-weight elements.16 COFs can be semiconducting and photosensitive, and can exhibit relatively large mobilities. Their integration in practical devices is however difficult, as usual synthesis processes produce bulk powder without control over the pore structure. However, 2D COFs have been synthesized on graphene, showing improved crystallinities with respect to their respective bulk phase (Fig. 14d).186,187 Given their porous nature, the study of 3D heterostructures of graphene and COFs is applicable for applications such as gas absorption and storage. However, 2D heterostructures with COFs remain largely unexplored. Graphene has also been used to protect the 2D sheets of conjugated polymers from high energy electron beams by encapsulation.23 Recently, it has been shown that carbon based nanomembranes can be used to fabricate graphene vertical heterostructures, providing an easy and flexible method for the functionalization of graphene and the engineering of its electronic properties.188 Heterostructures with pentacene thin films, another organic material, have also been studied. Pentacene is a flexible p-type semiconductor, with a low cost of fabrication. Therefore, it is widely used on organic thin-film transistors, although it presents a low mobility (∼1 cm2 V−1 s−1). So far, barristors composed of vertical heterostructures of graphene and pentacene have been reported with on/off ratios of ∼104.189,190 Integration of such devices into flexible substrates have shown high stability for strains up to 4%.190 Recently, Seo et al. have demonstrated a method to produce p–n junctions on graphene by selective doping with 6,13-bis(triisopropylsilylethynyl) (TIPS) pentacene.191 As-deposited TIPS-pentacene does not induce a large doping in graphene FETs. However, the FET characteristics showed n-type doping after laser irradiation, due to the oxidation of the TIPS-pentacene (Fig. 14e). The level of doping can be adjusted by the irradiation time or the intensity. By spatially controlling the exposure to the laser, p–n junctions with arbitrary shapes can be patterned on the graphene.191 Self-assembly is also used to produce thin films of organic molecules on graphene, such as perylene-3,4,9,10-tetracarboxylic dianhydride (PTCDA),192 oleylamine193 or phthalocyanine (Fig. 14f).194 Such molecules present ordered patterns and relatively strong interactions with the graphene,194 and can be employed to modify its electronic properties.193 By adjusting the temperature of the graphene during the epitaxial deposition of certain molecules, it is possible to obtain ultrathin organic heterostructures with precise control over the number of layers and the crystallinity of the deposited molecules.195 By the successive growth of layers of different organic molecules, ultrathin vertical p–n junctions can be easily produced, showing a good rectifying behavior and with a photoresponsivity of ∼0.37 mA W−1.195

Perovskite materials are those that share the structure of CaTiO3. Due to their remarkable optical properties, layered inorganic–organic perovskites have attracted recent interest for their application in photovoltaic applications.112 Thus, studies on their integration with other 2D materials, especially graphene, have been rising for the last few years.112,196 Niu et al. have demonstrated the direct growth of a perovskite (methylammonium lead halide)/graphene heterostructure (Fig. 14g).112 Their method consists of the PVD growth of PbI2 nanoplatelets on graphene, followed by conversion into perovskite by the reaction with CH3NH3I. Cheng et al. have devised a similar method in which PbI2 crystals are mechanically exfoliated and deposited on graphene.196 Encapsulation with 2D materials is also essential to protect the perovskite, which is unstable under ambient conditions.196 Patterning the 2D substrate before the growth of the perovskite allows patterned heterostructures to be obtained.112 Vertical heterostructures of perovskite encapsulated by two graphene electrodes show photocurrents on the overlapping areas, with responsivities of ∼950 A W−1 and response times of a few tens of ms (Fig. 14h).196

3.3. In-plane heterostructures with 2D materials

While vertical heterostructures of 2D materials can be prepared by physically stacking the different layers, the fabrication of in-plane heterostructures with truly 1D contacts requires more advanced synthesis methods. Advances on the synthesis procedures have allowed the development of such in-plane heterostructures of graphene with other 2D materials. The combination of metal, semiconductor and insulator 2D materials as in-plane heterostructures is interesting for the development of continuous, single-atom thick, in-plane integrated circuits with ultimately thin contacts.31

Probably one of the first examples of the synthesis of such in-plane heterostructures is that of h-BN and graphene in-plane hybrid sheets simultaneously grown during a single CVD process with CH4 and ammonia borane as precursors.197 The resulting sheets consisted of randomly located patches of both h-BN and graphene, and with thicknesses from a single to a few layers. This method produces hybrid sheets with a fixed ratio of graphene and h-BN, but does not provide control over the size or shape of the domains and their interfaces. To obtain controlled in-plane heterostructures, two-step growing processes were developed in which the materials are grown at different stages. In-plane heterostructures of graphene and h-BN with controlled shape and sharp interfaces can thus be produced by patterning a CVD-grown graphene layer and then selectively growing the h-BN using a second CVD,30 or in the opposite order (Fig. 15a).32,33 The direct growth of elaborate 2D graphene-based devices, such as multi-terminal transistors or band pass filters, is possible using this approach. Even though there is a slight mismatch in the crystal lattice of graphene and h-BN (∼1.7%), STM and HR-TEM studies indicated that it is possible to achieve seamless stitching at the atomic scale between both (Fig. 15b).33–35 The two lattices are then oriented, with the junctions corresponding predominantly to zigzag edges. TEM inspections also showed evidence of hybridization, with bonding occurring between the N atoms of the h-BN and the C atoms of the graphene.35 These findings agree with DFT calculations, which show that zigzag junctions are more energetically favorable than armchair junctions, while there is no large energy difference between C–N and C–B bonds and hence the kind of hybridization responds to the edge termination of the h-BN layer.33,198,199 In-plane hybrid sheets can also be obtained on dielectric SiC(0001) surfaces using two-step growth processes.34 A h-BN layer is firstly grown using CVD on the SiC, while a subsequent annealing step at high temperatures introduces some graphene regions. The orientation of the hybrid sheet obtained in this way is completely determined by that of the underlying SiC. In-plane graphene/h-BN junctions have been recently reported to grow in a TEM chamber (Fig. 15b).35 Graphene grows from the edge of h-BN layers supported on graphene by the supply of carbon atoms from residual hydrocarbons in the chamber. Graphene islands can also be grown on holes created in the h-BN by an electron beam, opening the possibility to produce quantum confined heterostructures with high control.

image file: c7cs00160f-f15.tif
Fig. 15 (a) SEM images of patterned h-BN/graphene in-plane heterostructures. Graphene circles (left, scale 50 μm) and lines (right, scale 10 μm) on h-BN. (b) Lateral growth of graphene from an N-terminated edge of h-BN in a TEM chamber. Left and right sides correspond to before and after the graphene growth. (c) SEM images of patterned in-plane MoS2/graphene heterostructures. The inset shows the results for larger Mo pressure, resulting in some MoS2 grains growing also on top of the graphene. (d and e) False colored DF-TEM of polycrystalline MoS2 grown from an edge of graphene (d), and the corresponding diffraction pattern showing the single crystal nature of the graphene (e). (f and g) In-plane logic circuits made from stitched graphene and MoS2. (g) Logic inverter from two n-type graphene/MoS2 lateral transistors. (a) Adapted by permission from Macmillan Publishers Ltd: Nat. Nanotechnol., (ref. 32), copyright 2013. (b) Adapted with permission from ref. 35. (c) Reprinted with permission from ref. 202. Copyright 2016 American Chemical Society. (d, e and g) Adapted by permission from Macmillan Publishers Ltd: Nat. Nanotechnol., (ref. 201), copyright 2016. (f) Adapted with permission from ref. 200. Copyright 2016, John Wiley and Sons.

In-plane heterostructures of graphene with some TMDs have been recently realized, although studies are still scarce compared to the case of h-BN.200–203 In all of them, patterned heterostructures are obtained by pre-patterning the first material (either the graphene or the TMD) and then growing the second material using CVD (Fig. 15c). A perfect planar stitching of the layers has not been realized so far, with TEM measurements showing an overlapping in the junction with sizes going from a few nm to some tens of nm.200,201 Evidence of weak van der Waals interactions in the overlapped region instead of covalent bonding between the two materials has been reported, owing to the large lattice mismatch between the two lattices.200 Consequently, no relation between the orientation of the two lattices has been reported, indicating that the edges simply act as nucleation centers but the growth is not epitaxial (Fig. 15d and e).200,201 However, good ohmic contacts between the graphene and the TMD have been reported, making this a perfect method to integrate graphene as an electrode for 2D devices.201,202 This, along with the scalability of the employed methods for the CVD synthesis, allows for the fabrication of high quality devices at large scales, including FETs, diodes, and NAND, NOR and inverter gates (Fig. 15f and g).200,201

3.4. Heterostructures with lower dimensionality materials

In this section we will discuss heterostructures of graphene with lower dimensionality materials. Carbon nanotubes (CNTs) are one-dimensional structures composed of carbon atoms, and with a structure similar to that of graphene. Graphene heterostructures with CNTs have been prepared, either by deposition of aligned CNT arrays on graphene,204 or using the CVD synthesis of graphene on a Cu substrate covered by functionalized CNTs (Fig. 16a).205 In these structures, the CNTs are connected to the graphene sheet either by covalent or π–π bonds. The presence of the CNTs provides reinforcement to the graphene, and hence the name “rebar” graphene is sometimes given to the heterostructure, from “reinforcement bar”. The rebar graphene is strong enough to endure free-standing handling, with the possibility to realize polymer free transfer from Cu, or to use it as a holder for TEM specimens (inset of Fig. 16a). Further reinforcement can be attained by performing a second CVD with Fe particles deposited on the rebar structure.206Rebar graphene with h-BN nanotubes has been also demonstrated, showing mechanical properties similar to those of the CNT rebar.207 Other than for reinforcement, heterostructures of graphene with CNTs have been applied to electronic and optoelectronic devices. Broadband photodetectors composed of graphene and single-walled CNT (SWNT) films showed responsivities of 120 A W−1, with expected values of ∼1000 A W−1 for lower illumination powers (Fig. 16b).208 The fabrication consists of coating a SiO2/Si substrate with a thin SWNT network film and then transferring the graphene and fabricating the electrodes. As graphene retains its mobility almost intact, the response time of these photodetectors can be as low as ∼100 μs. Owing to their flexibility and transparency, graphene heterostructures with SWCNTs have also been used as electrodes in CNT-based flexible electronics.209 The heterostructure was used for making the drain, source and gate electrodes of FET devices, for which the channel is a SWNT network. The whole device showed on/off ratios up to 105 and mobilities of 40 cm2 V−1 s−1, which were stable for strains up to 20%.
image file: c7cs00160f-f16.tif
Fig. 16 (a) TEM image of a graphene/CNT heterostructure (rebar). Top inset: Schematic of the heterostructure. Bottom inset: Handling of a free-standing heterostructure. (b) Responsivity of a graphene/CNT heterostructure for different light wavelengths. (c) Illumination power dependence of responsivity for a graphene transistor with a PbS deposited film. (d) Responsivity of a graphene device with deposited PbS before and after a bending test of 1000 cycles. (e) Schematics of FeCl3 intercalation in double- (top) and triple-layer (bottom) graphene. (f) Raman spectra of pristine graphene (dotted lines) and that intercalated with FeCl3 (continuous lines) for different flake thicknesses. (g) Sheet resistance versus optical transmittance of 4L and 5L graphene intercalated with FeCl3, compared with those of ITO and CNTs. (h) Sheet resistance versus number of layers for few-layer pristine graphene and after different stages of intercalation with Li. (i) Sheet resistance versus optical transmittance at 550 nm of Li-intercalated few-layer graphene compared with values for CNTs, ITO and acid-doped graphene. (a) Adapted by permission from Macmillan Publishers Ltd: Nat. Commun., (ref. 204), copyright 2013. (b) Reprinted with permission from ref. 208. (c) Adapted by permission from Macmillan Publishers Ltd: Nat. Nanotechnol., (ref. 210), copyright 2012. (d) Reproduced with permission from ref. 212. Copyright 2012, John Wiley and Sons. (e) Adapted with permission from ref. 221. Copyright 2010, John Wiley and Sons. (f) Adapted with permission from ref. 222. Copyright 2011 American Chemical Society. (g) Reproduced with permission from ref. 223. Copyright 2012, John Wiley and Sons. (h and i) Reprinted by permission from Macmillan Publishers Ltd: Nat. Commun., (ref. 226), copyright 2014.

Heterostructures of graphene with zero-dimensionality materials have also been realized, posing special interest for the increase of sensitivity of photoactive graphene devices. The responsivity of a purely graphene photodetector is usually ∼10 mA W−1. By depositing a thin film (∼80 nm) of PbS quantum dots (QDs) on bilayer graphene, photoresponsivities up to 107 A W−1 have been reported by Konstantatos et al. (Fig. 16c).210 This ranks among the highest sensitivities reported for graphene based photodetectors. The improvement in the responsivity is due to generation of photocarriers in the PbS QDs that are transferred to the graphene, along with the high carrier mobility of the graphene channel. The response times are however comparatively long, in the scale of seconds. The attained EQE is also only 25%, although it can be increased to values of 80% by using a top gate to direct the generated photocarriers to the graphene channel in a more efficient way.211 These devices can also be realized on flexible PET substrates and still maintain their efficiency after bending tests of 1000 cycles (Fig. 16d).212 Apart from these, there exist theoretical and practical studies about the possibility to assemble more exotic heterostructures.213,214

3.5. Intercalation

Intercalation of foreign species in the space between layers is a procedure employed for a wide range of layered materials and intercalants. A recent comprehensive review on the intercalation on 2D materials can be found in ref. 215. Here, we will introduce some of the most pertinent results on the intercalation of double- and few-layer graphene supported on a substrate. Intercalation of atomic or molecular species in the interlayer spaces of graphite, resulting in graphite-intercalation compounds (GICs), is widely used to modify its physical and electrical properties. The space existing between the layers can also be used as a nanoreactor to produce reactions that require extreme pressures.216 The van der Waals interaction between the layers provides localized areas of high pressure, which are able to induce chemical reactions between intercalated molecules. Intercalation of FLG has also led to the realization of magnetic217,218 and superconducting219,220 2D states for FeCl3 and Ca as an intercalant, respectively.

Intercalation of FeCl3 in BLG and FLG has been widely investigated as a way to decrease the sheet resistance while keeping a high optical transmittance (Fig. 16e–g). It is generally conducted using chemical vapor transport, with the graphene and the FeCl3 kept in separated zones of a glass ampoule sealed in a vacuum.217,221 A differential of temperatures is then applied to the ampoule, with the graphene area being in the hot zone (∼340–350 °C), while FeCl3 is maintained at a lower temperature (∼290–320 °C). The intercalation usually requires a few hours to complete (∼10 h), after which the ampoule can be brought to room temperature. By using this setup, Zhan et al. demonstrated FeCl3 intercalation into exfoliated graphene flakes with thicknesses of 2, 3 and 4 layers.221 The intercalation can be homogeneous and shows high stability in air, with no evidence of degradation occurring for as long as 1 year in the best reported cases.221–223 Intercalation is stable in air even under high humidity conditions for a few weeks (relative humidity up to 100%), or at temperatures of 150 °C for periods of hours.224 The charge transfer of the graphene can be monitored using Raman spectroscopy, which shows a stiffening of the G and 2D bands indicating a p-type doping of the graphene layers (Fig. 16f).221,222 Calculations from Raman results provide an estimated shift of the Fermi level of ∼0.9 eV, corresponding to a hole density of ∼5.8 × 1014 cm−2 for the stage-1 intercalated samples.222 These values are on the same order of magnitude as those obtained from conductivity measurements (∼5.8 × 1014 cm−2).223 After the intercalation, the 2D band of few-layer graphene can be fitted by a single Lorenztian,222 indicating a decoupling between the layers induced by the increase of the inter-layer distance. This decoupling has been corroborated by observing the Shubnikov–de Haas oscillations of the longitudinal resistance in a perpendicular magnetic field of intercalated BLG,217 while partial decoupling has been observed in the case of FLG due to unintentional deintercalation in the two upper layers of the stack during the process.223 The different environments to which the layers are exposed for thicknesses of 3 and more layers is evidenced by the splitting of the G band into two modes (Fig. 16f).221,222 The mode at lower wavelengths (G1) arises from the surface layers (top and bottom), while the mode at higher wavelengths (G2) corresponds to inner layers surrounded by FeCl3 on both sides (Fig. 16e).221 By decreasing the intercalation time, selective intercalation of FeCl3 between some of the layers in FLG has been recently reported,225 which could result in the opening of a bandgap owing to the vertical electric field caused by asymmetrical doping.

In terms of electronic properties, the hole doping caused by the FeCl3 intercalation on few-layer graphene leads to a decrease of the sheet resistance. Pristine, single-layer graphene has a high optical transmittance (∼97.7%), but its sheet resistance is high (usually over 1000 Ω sq−1 for undoped single layer graphene) for the use of graphene as a transparent electrode. The sheet resistance in FLG is still a few hundreds of Ω sq−1. After intercalation with FeCl3, FLG shows a decrease of the sheet resistance to values as low as 8.8 Ω sq−1 at room temperature for a thickness of 5 layers.223 Interestingly, the intercalation does not severely decrease the optical transmittance of the FLG, which retains a value of 84% at 550 nm compared with 88.5% for 5 pristine layers. These values can compete with those of ITO, with transparencies of 90% and sheet resistances between 10 to 50 Ω sq−1 (Fig. 16g). The relatively low temperatures employed for the intercalation and the inherent mechanical properties of FLG allow for their use as flexible and transparent substrates, which provides a clear advantage compared with the rigid ITO. Overall, FeCl3-intercalated FLG is an interesting alternative material to the commonly employed ITO for diverse optoelectronic applications.

Other materials have been intercalated into substrate supported FLG, one of the most promising being Li. In situ measurements on electrochemical intercalation of Li in FLG have recently been demonstrated to n-dope the FLG and decrease its sheet resistance (Fig. 16h).226 Remarkably, the intercalation of Li also increases the optical transmittance of the FLG. On relatively thick flakes of 19 layers, optical transmittances up to 91.7% have been attained, with sheet resistances as low as 3 Ω sq−1 (Fig. 16i). The small size of a Li atom make it also convenient as an intermediate for the intercalation of larger atoms, such as Ca.219,220 A detailed description of graphene heterostructures for Li-ion batteries will be provided in Section 4.

4. Solution-processed heterostructures

4.1. Scalable exfoliation of 2D materials

Layered 2D materials can be produced using a variety of techniques as discussed in the previous sections, and while the materials created using both micromechanical exfoliation and CVD are of high quality and large crystal size, these production methods lack scalability and remain costly, preventing their use for many applications. Wet chemical processes, such as liquid phase ultrasonication and electrochemical intercalation, can currently produce large volumes of dispersed monolayers and few layer flakes of various 2D materials, cheaply and with the potential for scale-up.14,227–233 These liquid-based exfoliation techniques allow the easy creation of thin films, coatings, or additives consisting of a large number of nanosheets, which is ideal in applications such as energy generation and storage. These different scalable liquid phase exfoliation techniques can be visualised schematically in Fig. 17.
image file: c7cs00160f-f17.tif
Fig. 17 Overview of liquid exfoliation techniques. (a) Crystal structure of a bulk layered 2D material (MoS2) along with a photograph showing the bulk MoS2 powder. (b) Schematic showing two common liquid exfoliation techniques: direct ultrasonication in solvent and ion intercalation. Both of these techniques are easily scalable and able to produce large amounts of exfoliated dispersed 2D materials. (c) Crystal structure of solvent-stabilised exfoliated MoS2 along with a photograph showing a MoS2 dispersion prepared by ultrasonication in NMP.

The first large-scale method of liquid exfoliation of 2D materials was for graphene oxide (GO), where graphite is first treated with strong oxidising agents such as sulphuric acid and potassium permanganate, which increases the hydrophilicity of the graphite. This increased hydrophilicity allows the graphite layers to be easily intercalated by water and separated by mild ultrasonic treatment, producing flakes down to monolayer thickness and hundreds of nanometres in lateral dimension in concentrations up to 1 mg ml−1.234 Intercalation of ionic species, typically by using electrochemical methods, into a variety of layered species beyond simply graphite leads to increased interlayer spacing and the subsequently weakened interlayer coupling allows the material to be easily exfoliated.229 The resulting dispersion can also be stabilized by surfactants. Intercalation can also produce large quantities of monolayer flakes with dimensions of a few hundred nanometres. Ultrasonic treatment in specially chosen solvents can lead to direct exfoliation of many 2D materials.235–237 The ultrasonic sound waves produce cavitation bubbles, causing the 2D material to break up into individual nanosheets. Each of these techniques is able to produce large volumes of exfoliated nanosheet dispersions which can subsequently be processed to form a variety of devices. Although the quality of such exfoliated materials is typically lower than that produced using mechanical cleavage or CVD, it is sufficient for a wide range of applications, particularly when different 2D materials are combined to form hybrid nanocomposites.

4.2. Formation of graphene-based hybrid composites

There has been enormous effort focused on exploiting the many beneficial properties of 2D materials including graphene; however, typically what is desired for the best device performance is a variety of properties from different materials. This has led to the exciting area of hybrid composite research, where multiple 2D materials are mixed together to synergistically improve the performance over any single component. This typically involves the addition of highly conductive graphene derivatives to mixtures of other layered materials to increase the conductivity, while simultaneously altering the morphology and surface area of the resultant hybrid. The most common example of these hybrid composites, due to its applicability for electrical devices, is the TMD/graphene hybrid composite. Such composites of TMDs and graphene allow the conductivity of graphene to be exploited, while maintaining the unique chemistry of the TMD.238–240 Of these TMD/graphene composites studied so far the most common, due to its relative abundance and ease of production, is the MoS2/graphene hybrid composite, which can act as a proxy for the performance of many similar TMDs. However, different TMDs, as well as other layered materials, have also gained interest for use in graphene based composites and these will be discussed in further detail later.

These hybrid composites can be produced using several methods. The simplest is to exfoliate each of the individual 2D materials separately and then mix the dispersions together. This can be achieved by producing exfoliated dispersions of a selected material and then generating a film either by filtration onto a membrane or drop-casting directly onto the desired substrate. An example of this mixing procedure to form a MoS2/graphene composite is shown schematically in Fig. 18a and an SEM image showing the resultant layered composite structure is shown in Fig. 18b.241 This procedure has the benefit of being very simple and easily scalable. However, the interaction between each material typically occurs only when the solvent is removed from the dispersion, causing the mixture of flakes to form a composite film. By varying the relative ratio of each component it is possible to tune the properties of the laminar structure that is produced using this technique, to optimise performance for the desired application.242–245

image file: c7cs00160f-f18.tif
Fig. 18 Comparison of hybrid composite synthesis procedures. (a) Schematic illustration showing the formation of a graphene nanosheet (GNS) and MoS2 composite by first mixing with ultrasonication and then forming a film by vacuum filtration. (b) SEM image showing a cross-section of the resultant MoS2–graphene hybrid composite with a laminar structure. (a and b) Modified from Hu et al.241 (c) Schematic illustration showing the formation of a MoS2/GO hybrid composite aerogel through hydrothermal self-assembly. (d) SEM image showing the morphology of the resulting MoS2/GO porous aerogel composite. The inset shows a photograph of the composite structure, the ratio of MoS2[thin space (1/6-em)]:[thin space (1/6-em)]GO is 80[thin space (1/6-em)]:[thin space (1/6-em)]20. (c and d) Adapted with permission from ref. 250. Copyright 2016 American Chemical Society.

Hydrothermal synthesis is another process that is widely used to produce these hybrid nanocomposites. The hydrothermal growth process typically involves the addition of an aqueous dispersion of chemical precursors in a sealed vessel, which is then heated (∼200 °C) in an autoclave to form the desired crystalline material.246–249 Solvothermal growth is similar except that instead of water, a non-aqueous solvent is used. This in situ growth process allows for a close interaction between the different materials, with crystals of MoS2 being able to be grown directly on the surface of a graphene supporting layer to form the hybrid composites. Fig. 18c shows schematically the process to hydrothermally produce a GO/MoS2 hybrid composite structure.250 This was achieved by first taking Li-intercalation exfoliated MoS2 and mixing by ultrasonication with a GO dispersion, before treating at 180 °C for 24 h. This was then chemically reduced by N2H4·H2O at 90 °C for 12 h followed by freeze-drying to produce a porous structure. Due to the high temperatures that the dispersions are exposed to, the interaction between the MoS2 and GO is strengthened, leading to good structural stability and fast charge transport. This freeze-drying process also allows for the formation of a highly porous 3D aerogel structure, which maximises the available surface area. The formation of such a porous structure avoids the common drawback of solution exfoliated processes, which is the restacking of the exfoliated material that reduces the available surface area and the overall performance of subsequent devices. Fig. 18d shows an SEM image illustrating this high porosity of the resulting MoS2/graphene composite structure, where the ratio of MoS2[thin space (1/6-em)]:[thin space (1/6-em)]graphene is 80[thin space (1/6-em)]:[thin space (1/6-em)]20. This simple and scalable procedure allows the formation of a unique 3D structure where the relatively larger graphene sheets are decorated with smaller MoS2 nanoflakes. This nanostructured morphology, with its high surface area conductive network of exfoliated material, is ideal for many applications, but in particular for electrochemical energy storage.251

4.3. Hybrid composite supercapacitors

2D materials are excellent candidates for electrochemical energy storage due to their highly available specific surface area and versatile electronic structure. Electrochemical energy storage devices, such as lithium ion batteries or supercapacitors, are widely used in portable electronic devices, electric transportation, and even grid scale systems. The application of graphene in energy storage technology has been investigated widely due to its many excellent properties such as mechanical strength, high specific surface area (2675 m2 g−1) and high electrical conductivity.252–254 Similarly-structured layered TMDs, particularly MoS2, have so far attracted less interest but are gaining increased popularity for integration into energy generation and storage devices due to similarities with graphene.255,256 Despite the increasing interest in the electrochemical properties of MoS2 the fundamental charge storage processes are still not well established, with some disagreement between reports of its electron transfer characteristics and pseudocapacitance.257–259 Hybrid composites consisting of multiple 2D materials are gaining increased interest for applications in energy storage and generation. This is due to the beneficial synergy that occurs when the high conductivity of graphene can be used to enhance the intrinsic charge storage properties of another material, such as MoS2. Examples of such hybrid composites include graphene/metal oxides243,260–262 as well as graphene/TMD composites which will be discussed in further detail here.242,244,245,250,263–273

Electrochemical supercapacitors, sometimes referred to as ultracapacitors, have attracted increasing attention for their rapid charging capabilities, very long cycle lifetimes and wide operating temperature range.274 The charge storage mechanism in supercapacitors consists of a combination of electric-double layer capacitance (EDLC), where electrolyte ions accumulate at the electrode interface, as well as pseudocapacitive ion absorbance or intercalation where rapid redox processes can occur. As no chemical reactions are typically occurring during the charge/discharge of the device, supercapacitors can operate on very short time scales (i.e. high power density) with very long cycle lifetimes. Supercapacitor devices consist of two electrodes (either symmetrical or asymmetrical) separated from each other and filled with an electrolyte solution. An ideal electrode material is one that has both high conductivity and surface area to maximise the amount of charge stored in the double layer at the interface. This makes exfoliated graphene an ideal electrode material, with several reviews previously discussing the benefits of graphene for supercapacitor applications.254,275,276

The formation of hybrid composites of graphene with other 2D materials allows the optimisation of their performance by tuning the conductivity and morphology of the composite structure. Fig. 19 shows an example of a graphene/MoS2 composite supercapacitor device created using solution based exfoliation processes.245 Commercially available graphene powder was first ultrasonicated to produce a stable dispersion, and a dispersion of MoS2 was similarly produced by solution exfoliation (seen in Fig. 19a). These dispersions were then mixed and filtered to create thin (∼500 nm) flexible films as electrodes in a symmetrical supercapacitor arrangement using an aqueous electrolyte. Fig. 19b–d shows a schematic of the coin-cell architecture that was used to test the composite films, and demonstrates the flexibility of the thin film electrodes. The use of this coin-cell structure allows for a real-world investigation of the performance for possible commercial applications.277 When tested using an aqueous electrolyte (Na2SO4) it was found that the graphene by itself exhibits moderate performance and behaved as an ideal double-layer capacitor, as established in the literature.254 On the other hand, MoS2 suffered from high internal resistance due to its semiconducting nature and so produced a lower capacitance. However, as seen in Fig. 19e, when the graphene and MoS2 dispersions were combined in a 1[thin space (1/6-em)]:[thin space (1/6-em)]3 ratio the performance was significantly enhanced over either individual component. This was attributed to a combination of physical interaction between the two different materials, as the smaller MoS2 flakes could coat and prevent restacking of the larger graphene sheets, as well as the increased conductivity that occurs due to the network of graphene flakes and the ion adsorption pseudocapacitance that occurs due to MoS2.245 It has also been observed that with repeated charge/discharge cycling the measured capacitance can increase over time, due to the repeated intercalation/de-intercalation of the electrolyte ions (Na+) into the composite film, leading to increased active surface areas with time.245,278,279

image file: c7cs00160f-f19.tif
Fig. 19 Examples of MoS2/graphene hybrid composite supercapacitors. (a) Photograph showing a dispersion of exfoliated MoS2 flakes along with a schematic (b) showing the exploded view of an assembled symmetrical coin-cell supercapacitor. Photographs (c and d) show the MoS2/graphene composite electrodes formed after filtration onto a supporting polymer membrane. (e) Comparison of the electrochemical performance of the single component MoS2, graphene, and the MoS2/graphene composite illustrating the increased capacitance. Low (f) and high magnification (g) SEM images showing the morphology of a MoS2/graphene aerogel structure. (h) Comparison of the supercapacitor charge/discharge behaviour again demonstrating the enhanced performance of the MoS2/graphene composite. (a–e) Adapted with permission from ref. 245. (f–h) Adapted from ref. 270, Copyright 2016, with permission from Elsevier.

Maximising the electrode surface area and preventing restacking by creating a porous 3D structure is another route to increase the device performance. Fig. 19f and g shows the morphology of a MoS2/graphene aerogel structure produced using hydrothermal synthesis, illustrating the high available surface area. Aerogel structures are ideal for energy storage applications as they provide increased active surface area by forming a large interconnected network of 2D materials.269,280,281 This aerogel structure allows the high specific surface areas of many 2D materials to be more properly exploited, but also suffers from issues of long-term stability.282 The relative charge/discharge behaviour of this MoS2/graphene aerogel electrode, when tested in a supercapacitor device (Fig. 19h), showed that the composite electrode outperformed either of the single component devices. This 3D structured electrode demonstrated both high capacity and stability with extended cycling due to the flexible nature of the graphene, showing that 3D hybrid composite structures are an ideal electrode morphology for electrochemical energy storage.270

Naturally occurring MoS2 exists in the stable 2H phase (Fig. 2c), which possesses a trigonal prismatic lattice arrangement and is semiconducting, as discussed previously. This can be distorted, typically through charge injection by intercalation from lithium ions, to form the metallic 1T phase which has an octahedral arrangement.283,284 For energy storage applications, where the conductivity of the electrode material is vitally important, it is beneficial to increase the percentage of the 1T polymorph present in the electrode material. The synthesis of a hybrid composite aerogel structure consisting of a mixture of MoS2 in both the metallic 1T and semiconducting 2H phase, combined with reduced graphene oxide (rGO), is shown in Fig. 20.268 This mixture of crystal lattice structures is important as the metallic 1T phase has already been shown to provide improved electrochemical charge storage properties due to a combination of its conductivity and the ability of the exfoliated and restacked MoS2 sheets to intercalate various electrolyte ions.285 The 1T/2H-MoS2/graphene composite structure was produced by forming MoS2in situ through the reduction of phosphomolybdic acid hydrate (H3PMo12O40·xH2O) in the presence of a sulphur precursor (L-cysteine). These precursors were combined with a GO dispersion and underwent hydrothermal treatment (180 °C for 12 h), followed by freeze-drying to make the aerogel shown in Fig. 20a. In this work the in situ growth of MoS2 directly onto GO produces a strong interaction between the two materials, allowing for not only a robust 3D structure to be formed but also beneficial electrochemical performance. By maximising the interaction between the different materials in the heterostructure composite the charge transfer between the two is also maximised, leading to beneficial performance. These 3D networks of 2D materials are ideal for supercapacitor applications and Fig. 20b compares the relative energy storage performance of the rGO aerogel by itself to the composite structure; clearly the composite performance is far superior to the single component. Hybrid composites made by simply mixing commercial MoS2 powders with rGO were also produced (rGO–MoS2-p, green curve) and compared to the in situ co-synthesised composite (rGO–MoS2-co, blue), as well as MoO2/graphene composites co-synthesised without the sulphur precursor (rGO–MoO2, red). From the results shown in Fig. 20b the superior performance of the in situ grown MoS2 can clearly be seen, illustrating the increased capacitance that can be achieved due to the close interaction between the two different materials. The high stability of this in situ produced composite structure was also demonstrated by measuring the change in capacitance with repeated cycling with very little degradation in the performance over 50[thin space (1/6-em)]000 cycles, better than either the rGO or MoS2 single component electrodes which reduced dramatically after 13[thin space (1/6-em)]000 cycles.

image file: c7cs00160f-f20.tif
Fig. 20 Hybrid composites of mixed 1T–2H phase MoS2/graphene for supercapacitors. (a) SEM image showing the morphology of a rGO–MoS2 composite that was formed using hydrothermal co-synthesis. The MoS2 is a mixture of both 1T and 2H phases. (b) Comparison of the electrochemical capacitance for each of the single component electrodes as well as the composite material demonstrating the enhanced performance of the co-synthesised composite which has the mixed phase structure. (c) Schematic diagram showing a MoS2/graphene heterostructure formed via hydrothermal synthesis illustrating the phase transition that occurs for the MoS2 directly in contact with the rGO. The charge storage mechanism is also illustrated where the MoS2 hybrid undergoes reversible proton (H+) adsorption. (d) Cross sectional HAADF-STEM image showing the interface between the MoS2 and rGO layers, where the layer directly in contact shows a 1T arrangement while those away from the interface are 2H. (e) Comparison of the electrochemical capacitance for the composite compared to the bulk and exfoliated pure MoS2 illustrating the enhanced capacitance of the hybrid structure. (a and b) Adapted with permission from ref. 268. Copyright 2016 American Chemical Society. (c–e) Adapted with permission from ref. 286. Copyright 2016, John Wiley and Sons.

The interaction between different 2D materials present in a hybrid composite can have an enormous effect on the electrochemical energy storage performance and requires further investigation. It has been shown that the interaction between rGO and MoS2 in a hybrid composite can even lead to a phase transition of the MoS2 lattice, shown schematically in Fig. 20c, from the 2H to the 1T phase.286 Mahmood et al. demonstrated that when MoS2, produced by ultrasonication of bulk MoS2 powder, was mixed with a dispersion of GO and then heated to 200 °C for 12 h under vacuum before filtering to form a composite structure, a unique interaction occurred. This procedure also thermally reduces the GO layers forming rGO. The hybridisation between the MoS2 and rGO nanosheets during the high temperature treatment forms a unique interlayer coupling, where the negative charge was transferred from the GO into the MoS2 lattice leading to the observed phase transition. Evidence of this phase transition can be seen in the TEM image shown in Fig. 20d. The MoS2 lattice in direct contact with the rGO at the interface shows an octahedral orientation, indicative of the 1T phase, while layers that are not in contact show the typical 2H trigonal prismatic lattice structure. The transfer of the negative charge from the GO into the closest MoS2 layer destabilises the 2H crystal lattice and favours the octahedral coordination.287 Typically the 1T phase is metastable, reverting to the lower energy 2H phase over time. However, DFT calculations indicated that the interaction between the 1T-MoS2 and graphene are enhanced, providing a higher energy barrier for the 1T-to-2H transition allowing the unstable 1T phase to exist at the interface.286 As discussed previously the presence of this 1T-MoS2 phase can greatly improve the energy storage properties of a supercapacitor device, and so the presence of this mixed phase is beneficial for energy storage applications.285 This increased performance of the 1T phase is related to the charge storage mechanisms present, which are a combination of proton intercalation in between layers of MoS2 and adsorption of the protons onto the surface of the layers. Fig. 20e demonstrates the enhanced electrochemical charge storage behaviour that occurs from producing a rGO/MoS2 composite (2D RM Hybrid) when compared to only bulk or exfoliated MoS2 (2D MoS2), in agreement with the previously shown examples. The high density of the exposed redox active edge and pore sites on the composite surface provides a large number of active sites for charge storage, in the form of proton adsorption (MoS2 + xH+ + xe ↔ HxMoS2), to occur.286 The peaks present in the curve in Fig. 20e indicate that some Faradaic reactions are occurring, and these have been attributed to the adsorption/desorption of protons, as the electrolyte used was H2SO4, at the interface.

4.4. Hybrid composite batteries

4.4.1. Lithium ion batteries. Lithium ion batteries (LIB) are another key energy storage technology in which composites of 2D materials can produce large improvements in performance. LIBs are rechargeable energy storage devices which function by shuttling lithium ions between a negative electrode (anode) and a positive electrode (cathode). Unlike supercapacitors, which are of interest due to rapid charge/discharge times, LIBs are able to provide a sustained current output over a long time (i.e. high energy density) but suffer from fatigue with repeated use. The typical anode in commercial LIBs consists of a graphite electrode, however graphite anodes suffer from a low theoretical capacity (372 mA h g−1) as well as susceptibility to structural damage with continued lithium intercalation/de-intercalation and repeated charge/discharge cycles.288 Graphene has already begun to attract attention for use in LIBs due to the increased performance that occurs when reduced to nanoscale dimensions, which produces increased capacity (600–1000 mA h g−1) and mechanical flexibility to resist deformation.289,290 Interest has also been extended to TMDs, where higher capacities (∼670 mA h g−1) have been measured.291–293 Sulphide containing materials are of particular interest due to the difference in the lithium storage mechanism; unlike graphite where the ions are simply intercalated between the carbon layers, the majority of intercalated lithium is stored through a conversion reaction with the sulphide, forming Li2S. This allows between 2–6 electrons per single transition metal compared to only a single electron for a simple intercalation reaction.294 As with supercapacitors there is also increasing attention focused on improving the performance (up to 1675 mA h g−1) through the creation of graphene/MoS2 hybrid composites.250,263,264,266,295 By creating these hybrid composites of nanoscale 2D materials it is possible to achieve the desired high capacity along with long-term cycling stability. As with the other nanostructured composites containing graphene, the carbon can provide a highly conductive network providing a conductive pathway within the composite, along with physical prevention of material re-agglomeration and restacking.

One method for achieving the maximum surface area, while preserving the close interaction between the composite materials, is through direct growth, typically via hydrothermal chemistry, of MoS2 onto a graphene support.267–272 By maximising the connectivity between the different materials that make up the composite it is possible to ensure rapid charge transfer kinetics as well as structural stability to compensate for the volume changes that occur when the lithium ions are intercalated. By creating a graphene aerogel structure that acts as both a highly conductive and flexible backbone as well as a porous structural support which can be ‘decorated’ with MoS2 it is possible to create a highly efficient LIB electrode. This has been shown by Lee et al., where chemically synthesised MoS2 is added to a dispersion of GO before hydrothermal treatment (80 °C for 6 h) to form a hydrogel, which is then dried and compressed to form an aerogel paper, shown schematically in Fig. 21a.269 The MoS2 chemical synthesis process, instead of forming stacked laminar sheets leads to the growth of ‘nanoflowers’ and the introduction of these into the GO structure leads to the formation of pores that are ideal for lithium intercalation. The electrochemical performance of this hybrid structure, which can be seen in Fig. 21b, shows the high capacity of the electrodes even at high current density and illustrates again how the composite material (MoS2 Graphene Aerogel Paper, MGAP) outperformed both the pure MoS2 as well as the simple mixture of MoS2 and GO without the hydrothermal treatment (MrGO) at a low current density. As the current density was increased the other composites produce higher capacitance, however, with repeated cycling this was found to decrease significantly, while the MGAP composite retained its performance. SEM images showing the morphology of the compressed aerogel paper electrode are shown in Fig. 21c and d, and a photograph illustrating the flexibility of the paper electrode material in Fig. 21e, indicating its applicability for flexible energy storage devices.

image file: c7cs00160f-f21.tif
Fig. 21 MoS2/graphene hybrid composites for lithium ion batteries. (a) Schematic illustration showing the procedure for the formation of a highly porous MoS2/GO aerogel structure, along with the lithium intercalation pathways demonstrated. (b) Charge storage capacity for the single component and MoS2/GO hybrid aerogel structure with increasing current densities. (c and d) Cross-section SEM images showing the morphology of the compressed MoS2/GO aerogel electrode, illustrating the laminar structure which is ideal for ion intercalation. (e) Photograph demonstrating the flexibility of the MoS2/GO hybrid composite electrode, indicating its suitability for flexible energy storage. (f) Schematic illustration showing the synthetic procedure to form a perpendicularly aligned MoS2/rGO hybrid composite structure. (g) Comparison of the cycling performance (current density of 0.1 A g−1) of the single MoS2 and rGO electrodes to the hybrid composite material. Both single component electrodes suffer from poor stability and low capacity, while the composite has both high capacity and stability. (a–e) Modified from ref. 269 with permission from The Royal Society of Chemistry. (f and g) Adapted with permission from ref. 271. Copyright 2016 American Chemical Society.

In the case of GO there is typically a negative surface charge due to the presence of oxygen functional groups on the surface and this prevents the use of typical molybdenum oxide ions (MoO42−) being used to couple to the surface due to electrostatic repulsion. To overcome this and ensure that MoS2 can be easily directly grown onto GO while remaining in an aqueous dispersion several researchers have attempted to functionalise the GO using cationic surfactants to neutralise its surface charge.247,296 As the surface charge on GO is strongly affected by the pH it is also possible to tune this charge by altering the acidity of the dispersion. By carefully controlling the pH it was demonstrated by Teng et al. that it is possible to selectively grow MoS2 nanosheets on GO using hydrothermal synthesis, leading to vertically aligned MoS2 sheets for use in LIBs.271 In this work, shown schematically in Fig. 21f, the MoS2 nanosheets are directly coupled to the oxygen atoms present on the surface of the GO by hydrothermal synthesis. This leads to perpendicular alignment of the MoS2 which helps to prevent restacking of the composite structure, thereby maximising the available surface area, as well as ensuring high structural stability during the cycling process. The GO backbone structure provides a conductive pathway for charge and lowers the diffusion distance required for the electrolyte ions. Similar to previously discussed work (Section 4.3), the MoS2 sheets here are significantly smaller in diameter (100 nm) than the GO sheets, and this small size is preferential due to the shorter diffusion lengths required for the lithium ions as well as promoting the electrode/electrolyte interface. The cycling performance of this MoS2/graphene electrode is shown in Fig. 21g, along with comparison to pure MoS2 and rGO electrodes. The initial capacity of the composite electrode is higher than either of the singular component electrodes (1160 mA h g−1), and during initial cycling the composite and pure MoS2 electrodes undergo an increase in capacity. This is attributed to an increase in the available surface sites with repeated cycling, and matches what has been observed previously for supercapacitor devices. After ∼20 cycles the pure MoS2 electrode undergoes a rapid decline in capacity, due to the combination of its low conductivity and inability to withstand the repeated volume changes. The composite however displays negligible capacity loss over 150 cycles, indicating the greatly enhanced stability and capacity of an in situ grown MoS2/graphene composite.

4.4.2. Sodium ion batteries. In addition to the well-studied MoS2/graphene Li-ion systems there has also been a recent increase in interest in sodium ion battery (SIB) systems using these hybrid composites.244,267,297 This increased interest is driven by the relative abundance combined with worldwide distribution of sodium reserves. Sodium intercalation chemistry is also very similar to lithium, making it simple to transition to as an alternative material. However, the radius of Na+ ions is ∼55% larger than Li+, which means that electrode materials must be able to withstand much greater physical deformation during charge/discharge cycles. An example of one MoS2/GO composite paper film for use in SIBs, where bulk MoS2 powder was first exfoliated using ultrasonication in acidic media and combined with a GO dispersion, is shown in Fig. 22a and b.244 This procedure, similar to the others discussed, produces a laminar structure where the large GO sheets are interleaved with smaller MoS2 flakes, providing a robust electrode architecture ideal for reversible storage of the large sodium ions, as shown schematically in Fig. 22c. The incorporation of GO into this composite was found to have a non-linear effect on the measured conductivity of the resultant composite due to the established percolation theory that describes the formation of a single conductive pathway that can form when a conductive additive is inserted into an insulating matrix.244,298 The 1st and 2nd cycles of sodiation and de-sodiation are shown in Fig. 22d for a 60% MoS2/graphene composite reaching a capacity of 230 mA h g−1 and a Coulombic efficiency of 99%. The high irreversible capacity observed during the first cycle is attributed to the formation of the solid electrolyte interphase (SEI) layer typically observed in such battery systems. As the ratio of MoS2 exceeded 40% the electrochemical behaviour was dominated by the sodium interaction with MoS2 over that of the GO, indicating the ratio of each material in these composites is of importance. The mechanical strength of these composites was also investigated, as 2D materials are of great interest for flexible energy storage due to their flexibility,299 and the composite was found to be able to withstand 2–4 MPa of uniaxial tension. The electrochemical intercalation and de-intercalation of sodium ions that occur during charge/discharge cycles also lead to phase transitions of MoS2, as seen previously in the case of lithium, due to charge transfer from the sodium ions, as shown schematically in Fig. 22e. This has been investigated, on the atomic scale, to better understand this transition and was shown to occur in discrete stages where depending on the degree of intercalation the transition is reversible.300 During the charge cycle the sodium ions first intercalate into alternating layers of MoS2, and as all the layers are intercalated the large degree of strain on the lattice induces a glide of the sulphur atoms along the interlayer atomic plane causing the phase transition from 2H-to-1T.287,300 The partially intercalated MoS2 layers, with distinct combinations of both 1T and 2H regions, are shown in Fig. 22f, along with a cross-section in Fig. 22g showing the intercalated sodium ions (blue circles). Properly understanding the structural changes that occur within the MoS2 layers when intercalation occurs is important for the design and optimisation of composite devices for use in energy storage.
image file: c7cs00160f-f22.tif
Fig. 22 MoS2/graphene hybrid composites for sodium ion batteries. (a) Schematic diagram illustrating the synthesis of MoS2/rGO composite paper. (b) Photograph of a large area composite paper electrode formed by vacuum filtration. (c) Plot demonstrating the process of intercalation and de-intercalation of sodium ions into the composite structure. (d) Graph showing the electrochemical performance of the MoS2/rGO (60% MoS2) composite for sodium intercalation/de-intercalation over the first two cycles. (e) Schematic showing the phase transition that occurs in 2H-MoS2 when sodium ions (Na+) are intercalated, forming 1T-MoS2. (f) High-angle annular dark-field (HAADF) image showing the different domains of 1T- and 2H-MoS2 formed after sodium intercalation. (g) Annular bright field (ABF) cross-section image showing the sodium ions (blue circles) intercalated between the layers of MoS2. (a–d) Adapted with permission from ref. 244. Copyright 2014 American Chemical Society. (e–g) Adapted with permission from ref. 300. Copyright 2014 American Chemical Society.

4.5. Other TMDs for energy applications

As discussed above so far the vast majority of hybrid composites that are present in the literature are combinations of MoS2 and graphene derivatives (graphene, GO and rGO). However, other TMDs have also been investigated for use in hybrid composites for applications in energy storage. Each of the most commonly used TMDs has previously been compared for their energy storage performance and due to differences in density, electronic structure, toxicity, as well as ease of production each are suited to different applications.256,301 An example of the different capacitive performance for several TMDs, without the use of any carbon additives is shown in Fig. 23a–c.256 The different TMDs investigated were MoS2, WS2, TiS2, and MoSe2 (crystal structures shown in Fig. 23a) and each was prepared by ultrasonication. A photograph of the dispersions is shown in Fig. 23b. From the comparison of the capacitive performance (Fig. 23c) the best performing material was TiS2, which can be attributed to its metallic electronic structure. The use of TiS2 compared to other TMDs has already been demonstrated to provide improved performance for energy storage applications.302 This demonstrates that depending on the particular application careful thought must be given to which TMD will work best. Hybrid graphene-based composites of different TMDs have also been recently investigated, with WS2/rGO248 as well as MoSe2/rGO303 composites each produced from hydrothermal synthesis showing improved capacitive performance (Fig. 23d and e). The superior performance of WS2/graphene304 and MoSe2/graphene305 composites for LIB applications is also shown in Fig. 23f and g, showing that as well as increased energy storage these composites provide improvements in cycle stability and device lifetimes. Each of these examples shows the universality of the improved performance that occurs when hybrid graphene based composites are compared to single component materials. The addition of a highly conductive yet flexible network is ideal for these applications. The use of in situ growth processes has also been shown to provide improved performance for different material combinations as the strengthened interaction between the different materials leads to increased charge transfer and mechanical robustness.
image file: c7cs00160f-f23.tif
Fig. 23 Hybrid composites of other TMDs. (a) Structure of the four most popular TMDs investigated for energy storage applications along with a photograph (b) of ultrasonication exfoliated dispersions of these materials. (c) Comparison of the electrochemical charge storage performance for each of these TMDs without any additives. Plots demonstrating the enhanced capacitance of a WS2/rGO (d) and MoSe2/rGO (e) hybrid composite compared to the single components. Plot showing the increased capacity and increased cycle lifetime of MoS2/graphene and WS2/graphene composites (f), and a MoSe2/rGO composite (g) for use in lithium ion batteries. (a–c) Adapted with permission from ref. 256. (d) Adapted with permission from ref. 248. Copyright 2013 American Chemical Society. (e) Adapted from ref. 303 with permission from The Royal Society of Chemistry. (f) Adapted from ref. 304, Copyright 2015, with permission from Elsevier. (g) Adapted from ref. 305 with permission from The Royal Society of Chemistry.

4.6. Phosphorene/graphene hybrid composites

Recently some groups have begun to investigate graphene based composites with more exotic layered 2D materials, and one example of this that shows particular promise for energy storage applications is graphene/phosphorene composites. Phosphorene, which is exfoliated black phosphorous (BP), has already been demonstrated to be suitable for the same scalable liquid based exfoliation techniques that are applicable for TMDs.306–308 Phosphorene possesses high p-type carrier mobility (>1000 cm2 V−1 s−1) and a tuneable direct band gap in the range of 0.3–2 eV making it of great interest for many electronic applications.308 The layered structure of phosphorene can be seen in Fig. 24a along with an SEM image of bulk BP before exfoliation. By dispersing this BP in suitable solvents, such as n-methylpyrrolidone (NMP), it is possible to create stable dispersions (shown in Fig. 24b) of few-layer BP, which is often referred to as phosphorene. TEM images of the structure of these exfoliated few-layer phosphorene flakes (Fig. 24c–f) show the ordered lattice array.309 Typically phosphorene is considered highly unstable in air and degrades quickly; however, the process of solution exfoliation in NMP leads to a coating of residual solvent molecules which protect the phosphorene flakes from oxidation.308 This allows the solution exfoliated phosphorene flakes to be used in a much larger range of applications, including when exposed to aqueous solutions. This solution exfoliated phosphorene can also be combined with GO dispersions and used to create hybrid composite papers through vacuum filtration, such as those seen in Fig. 24g.310 To produce these papers, dispersions of GO and solution exfoliated BP were combined and then filtered to produce a thin, flexible membrane. This membrane was subsequently thermally reduced (300 °C for 1 h) to produce an rGO/phosphorene composite (PG) and also compared to a membrane that underwent spark plasma sintering (300 °C, 10 MPa) (PG-SPS). The cross-section SEM images (Fig. 24h and i) show that the PG-SPS sample is significantly denser than the thermally reduced sample (PG), attributed to a reduction in oxygen and water content. These phosphorene/graphene composites were then characterised for use in LIBs and the measured specific capacity, with increasing current densities, for both membranes is shown in Fig. 24j. The PG-SPS electrode displayed a high reversible specific capacity (1302 mA h g−1) and superior cycle lifetime illustrating the enhanced performance of these phosphorene based composites.
image file: c7cs00160f-f24.tif
Fig. 24 Phosphorene/graphene hybrid composites for energy storage. (a) Schematic and SEM image of the layered structure of black phosphorous. (b) Photograph showing a stable dispersion of ultrasonication exfoliated phosphorene. (c–f) TEM images showing the exfoliated structure and atomic lattice of exfoliated phosphorene. Scale bars: 100 nm (c), 500 nm (d), 2 nm (e), and 1 nm (f). (g) Photograph showing a flexible GO/phosphorene hybrid composite paper electrode. (h and i) SEM cross-section images showing the structure of the phosphorene/graphene composite (PG) paper, and that after spark plasma sintering (PG-SPS). (j) Comparison of the lithium-ion rate capability for the PG and PG-SPS composite electrodes with different current densities. (k) Schematic showing the process of sodiation/desodiation into a graphene/phosphorene composite. (l) Comparison of the charge storage capacity of hybrid composites containing different graphene (C) and phosphorene (P) ratios with repeated cycling. (a–f) Adapted with permission from ref. 309. (g–j) Adapted with permission from ref. 310. Copyright 2016, John Wiley and Sons. (k and l) Adapted by permission from Macmillan Publishers Ltd: Nat. Nanotechnol., ref. 311, copyright 2015.

Phosphorene/graphene composites can also be used for SIBs and a schematic showing the process of sodium intercalation (sodiation) and de-intercalation (desodiation), where the sodium ions attach directly to the phosphorene flakes is shown in Fig. 24k. Phosphorus can react with sodium and lithium to form Na3P and Li3P, respectively, with a high theoretical specific capacity of 2596 mA h g−1, which is significantly larger than other SIB materials currently available.311 Phosphorene/graphene composites created by mixing ultrasonicated dispersions of both graphene and phosphorene in NMP were characterised during electrochemical sodium intercalation by Sun et al.311 During sodiation there was a significant volume expansion (∼160%) and to ensure long cycle lifetimes the use of smaller phosphorene flakes relative to the size of the graphene, similar to the results seen from the TMDs discussed previously, was suggested to be key. This provides an elastic ‘buffer’ space in between the smaller phosphorene flakes that are sandwiched between larger graphene flakes, this also minimises the diffusion length for both electrons and sodium ions. The relative electrochemical performance for different ratio graphene/phosphorene composites is shown in Fig. 24l, where the ratio between the graphene and phosphorene is presented as a carbon/phosphorous (C/P) ratio. The charge/discharge cycling showed that composites with a C/P ratio of 2.78[thin space (1/6-em)]:[thin space (1/6-em)]1 (blue line) and 3.46[thin space (1/6-em)]:[thin space (1/6-em)]1 (green line) maintained the highest capacity after 100 cycles. As the graphene by itself is electrochemically inactive for sodium the lower C/P ratios also provide the highest measured capacities. This work and the previous example both demonstrate that hybrid composites of phosphorene/graphene are promising candidates for future studies into energy storage. The graphene provides elastic, highly conductive buffer layers that can allow for large volume expansion while maintaining a preferential pathway for electrons generated during charge/discharge allowing the capacity of the phosphorene to be exploited.

4.7. Hybrid composites for catalytic hydrogen evolution

As well as energy storage technologies, such as supercapacitors and batteries, hybrid composites of 2D materials are ideal for catalytic energy production, such as catalytic hydrogen evolution (HER) reactions.312 Electrocatalytic production of hydrogen, through splitting of water, is a vitally important component of renewable energy. An ideal catalyst for the HER should minimise the required overpotential required for hydrogen production, and the best materials for this currently are the platinum group metals. The volcano plot showing the relative catalytic performance of a variety of noble metals, as well as TMDs and graphene/TMD composites is shown in Fig. 25a, where ΔGH and j0 are the Gibbs free energy of hydrogen evolution and current density respectively. The performance of the TMDs can be seen to approach that of the far more expensive noble metals (Pt and Pd). MoS2 in particular has been investigated for the catalytic HER previously due to calculations and experimental results which show that the active sites on the MoS2 are the edges while the basal plane remains catalytically inert.239,313–315 Thus, exfoliated nanosheets of MoS2 are an ideal catalyst for this application due to the high density of edge sites.
image file: c7cs00160f-f25.tif
Fig. 25 MoS2/graphene composites for catalytic HER. (a) ‘Volcano’ plot showing the relative catalytic performance of the noble metals (Pt, Pd) compared to the TMDs (MoS2, WS2), and hybrid composites (MoS2/graphene). (b) 3D illustration of the sandwich of vertically grown MoS2 (vMoS2), monolayer graphene and deposited metal nanoparticles. (c) Comparison of HER activity demonstrating the improved performance of the vMoS2/Gr/Pd sandwich composite over the other configurations. (d) Plot comparing the capacitance of the sandwich structures after transfer to a supporting substrate, the composite vMoS2/Gr/Pd structure again displays the highest capacitance and cycling stability. eMoS2 stands for liquid-phase exfoliated MoS2. (a) Reproduced with permission from ref. 312. Copyright 2016, John Wiley and Sons. (b–d) Modified with permission from ref. 317.

Hybrid composites consisting of nanostructured MoS2 supported on conductive graphene based materials provide performance benefits due to the utilization of the high specific surface area combined with electrical conductance of the supporting material. Such hybrid composites where rGO sheets have been decorated with MoS2 created through a one-step solvothermal reaction in dimethylformamide have been demonstrated previously.316 During this process a precursor ((NH4)2MoS4) was reduced to form MoS2 on the surface of the GO, while reducing the GO to form rGO by reacting with hydrazine. This procedure forms rGO sheets which are uniformly decorated with MoS2 nanoparticles, and it was observed that the MoS2 selectively formed in contact with the rGO sheets, with minimal formation of nanoparticles freely in solution. The mechanism behind this selective growth was attributed to the interaction between the oxygen functional groups present on the GO surface, in agreement with the previously discussed work,271 providing control over the heterostructure morphology. The catalytic performance of the individual components was poor while the hybrid composite displayed excellent catalytic activity, approaching that of the theoretical maximum of platinum. This high catalytic performance was attributed to the strong electronic coupling between the small MoS2 nanosheets and rGO sheets, which maximises the availability of catalytically active edge sites, in agreement with previously observed increases when 2D materials are combined to form a synergistic composite structure.

The combination of metal nanoparticles into a composite TMD/graphene structure can also be used to further improve the catalytic efficiency, as demonstrated by Toth et al., and shown in Fig. 25b–d.317 To measure the catalytic activity and fundamental electrochemical properties of a graphene/MoS2 heterostructure a ‘sandwich’ was created at the interface between two immiscible liquids. This liquid/liquid interface has been widely used in electrochemistry to investigate the properties of a material free from the interference that may be caused by interaction with a solid substrate.318,319 This interface also allows for the formation of films of material to be studied at a highly reproducible, molecularly sharp, and defect free focal plane and this has been demonstrated for a wide range of particles with different shapes and aspect ratios.320 By creating a range of different composite structures, shown schematically in Fig. 25b, using CVD grown graphene as a supporting layer combined with either vertically aligned CVD grown MoS2 (vMoS2) or ultrasonication liquid exfoliated MoS2 (eMoS2) it was possible to investigate the behaviour of these composites at the liquid/liquid interface. This MoS2/graphene sandwich could then be decorated with metal nanoparticles (Pd, Pt, Au or Ag) that were grown directly onto the heterostructure at the liquid interface by reducing metal salts present in the electrolyte solutions. This asymmetrical decoration, or Janus functionalization, can be used to tune the properties of different 2D materials and change their properties.321 The catalytic performance of these ‘sandwiches’ is shown in Fig. 25c where the catalytic performance of the MoS2/graphene composite is again greatly increased over the single component films, and this can be enhanced further by the deposition of these metal nanoparticles. The most active composite structure was the vertically aligned CVD grown MoS2, due to the high density of exposed edge sites, combined with graphene and palladium nanoparticles (vMoS2/Gr/Pd, shown in green). The capacitance of these composite ‘sandwich’ films was also measured, after transferring them onto a supporting substrate, shown in Fig. 25d and e, again the vMoS2/Gr/Pd composite exhibited superior performance. This increased capacitance was attributed to a high available surface area, and the ability to provide large amounts of active surface sites for the electrolyte ions (Li+) to adsorb/intercalate into the MoS2 surface. This work demonstrates a unique architecture to characterise the fundamental electrochemical behaviour of MoS2/graphene heterostructure composites for both HER and capacitive energy storage.

5. Summary

In this review, we have overviewed the research on graphene-based heterostructures which are prepared either using dry methods, such as mechanical stacking and CVD, or using wet methods, such as liquid-phase exfoliation and mixing processes. These heterostructures are composed of graphene with other 2D materials or with lower dimensionality materials, which are stacked, connected or intercalated in the graphene material. Integration into the heterostructure gives us a new opportunity to bring out new functionalities due to synergetic effects which can be enhanced by van der Waals interactions. The integration also allows the drawbacks of graphene, such as the lack of a bandgap, and of other 2D materials, such as the low mobility of TMDs, to be overcome. Even though research on such heterostructured materials mainly started at the beginning of this decade, advances have been quite fast, relying on the accumulated knowledge on the synthesis and handling of 2D materials. These advances have allowed the fabrication of almost any kind of heterostructure at the laboratory scale, with prototypes showing functionalities that in some cases approximate or even exceed those of current technologies. Also, there is still space for new 2D layered materials to emerge, which will definitely enrich the field of heterostructures. Finally, tailored physical properties and designed geometrical structures may be obtained by appropriate selection of the preparation methods and materials. With further development of the production methods of graphene and other 2D materials as well as their heterostructures, advances are expected to develop these unique materials into practical applications in a wide variety of fields, including electronics, photonics, sensing technologies and production and storage of energy.


This work was supported by the PRESTO-JST and JSPS KAKENHI Grant Numbers JP15H03530, JP15K13304, and JP16H00917.


  1. A. C. Ferrari, F. Bonaccorso, V. Fal'ko, K. S. Novoselov, S. Roche, P. Bøggild, S. Borini, F. H. L. Koppens, V. Palermo, N. Pugno, J. A. Garrido, R. Sordan, A. Bianco, L. Ballerini, M. Prato, E. Lidorikis, J. Kivioja, C. Marinelli, T. Ryhänen, A. Morpurgo, J. N. Coleman, V. Nicolosi, L. Colombo, A. Fert, M. Garcia-Hernandez, A. Bachtold, G. F. Schneider, F. Guinea, C. Dekker, M. Barbone, Z. Sun, C. Galiotis, A. N. Grigorenko, G. Konstantatos, A. Kis, M. Katsnelson, L. Vandersypen, A. Loiseau, V. Morandi, D. Neumaier, E. Treossi, V. Pellegrini, M. Polini, A. Tredicucci, G. M. Williams, B. H. Hong, J.-H. Ahn, J. M. Kim, H. Zirath, B. J. van Wees, H. van der Zant, L. Occhipinti, A. D. Matteo, I. A. Kinloch, T. Seyller, E. Quesnel, X. Feng, K. Teo, N. Rupesinghe, P. Hakonen, S. R. T. Neil, Q. Tannock, T. Löfwander and J. Kinaret, Nanoscale, 2015, 7, 4598–4810 RSC .
  2. K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, Y. Zhang, S. V. Dubonos, I. V. Grigorieva and A. A. Firsov, Science, 2004, 306, 666–669 CrossRef CAS PubMed .
  3. C. Berger, Z. Song, T. Li, X. Li, A. Y. Ogbazghi, R. Feng, Z. Dai, A. N. Marchenkov, E. H. Conrad, P. N. First and W. A. de Heer, J. Phys. Chem. B, 2004, 108, 19912–19916 CrossRef CAS .
  4. Y. Zhang, L. Zhang and C. Zhou, Acc. Chem. Res., 2013, 46, 2329–2339 CrossRef CAS PubMed .
  5. S. Stankovich, R. D. Piner, X. Chen, N. Wu, S. T. Nguyen and R. S. Ruoff, J. Mater. Chem., 2006, 16, 155 RSC .
  6. J. I. Paredes, S. Villar-Rodil, P. Solís-Fernández, M. J. Fernández-Merino, L. Guardia, A. Martínez-Alonso and J. M. D. Tascón, J. Alloys Compd., 2012, 536, S450–S455 CrossRef CAS .
  7. K. S. Novoselov, D. Jiang, F. Schedin, T. J. Booth, V. V. Khotkevich, S. V. Morozov and A. K. Geim, Proc. Natl. Acad. Sci. U. S. A., 2005, 102, 10451–10453 CrossRef CAS PubMed .
  8. C. R. Dean, A. F. Young, I. Meric, C. Lee, L. Wang, S. Sorgenfrei, K. Watanabe, T. Taniguchi, P. Kim, K. L. Shepard and J. Hone, Nat. Nanotechnol., 2010, 5, 722–726 CrossRef CAS PubMed .
  9. B. Radisavljevic, A. Radenovic, J. Brivio, V. Giacometti and A. Kis, Nat. Nanotechnol., 2011, 6, 147–150 CrossRef CAS PubMed .
  10. A. K. Geim and I. V. Grigorieva, Nature, 2013, 499, 419–425 CrossRef CAS PubMed .
  11. R. Kitaura, Y. Miyata, R. Xiang, J. Hone, J. Kong, R. S. Ruoff and S. Maruyama, J. Phys. Soc. Jpn., 2015, 84, 121013 CrossRef .
  12. G. R. Bhimanapati, Z. Lin, V. Meunier, Y. Jung, J. Cha, S. Das, D. Xiao, Y. Son, M. S. Strano, V. R. Cooper, L. Liang, S. G. Louie, E. Ringe, W. Zhou, S. S. Kim, R. R. Naik, B. G. Sumpter, H. Terrones, F. Xia, Y. Wang, J. Zhu, D. Akinwande, N. Alem, J. A. Schuller, R. E. Schaak, M. Terrones and J. A. Robinson, ACS Nano, 2015, 9, 11509–11539 CrossRef CAS PubMed .
  13. M. Nasilowski, B. Mahler, E. Lhuillier, S. Ithurria and B. Dubertret, Chem. Rev., 2016, 116, 10934–10982 CrossRef CAS PubMed .
  14. V. Nicolosi, M. Chhowalla, M. G. Kanatzidis, M. S. Strano and J. N. Coleman, Science, 2013, 340, 1226419 CrossRef .
  15. M. Xu, T. Liang, M. Shi and H. Chen, Chem. Rev., 2013, 113, 3766–3798 CrossRef CAS PubMed .
  16. P. Miró, M. Audiffred and T. Heine, Chem. Soc. Rev., 2014, 43, 6537–6554 RSC .
  17. H. Liu, A. T. Neal, Z. Zhu, Z. Luo, X. Xu, D. Tománek and P. D. Ye, ACS Nano, 2014, 8, 4033–4041 CrossRef CAS PubMed .
  18. W. Ge, K. Kawahara, M. Tsuji and H. Ago, Nanoscale, 2013, 5, 5773–5778 RSC .
  19. A. Ambrosi, Z. Sofer and M. Pumera, Chem. Commun., 2015, 51, 8450–8453 RSC .
  20. D. K. Efetov, L. Wang, C. Handschin, K. B. Efetov, J. Shuang, R. Cava, T. Taniguchi, K. Watanabe, J. Hone, C. R. Dean and P. Kim, Nat. Phys., 2015, 12, 328–332 CrossRef .
  21. C.-L. Song, Y.-L. Wang, Y.-P. Jiang, Z. Li, L. Wang, K. He, X. Chen, X.-C. Ma and Q.-K. Xue, Phys. Rev. B: Condens. Matter Mater. Phys., 2011, 84, 20503 CrossRef .
  22. D. Jiang, T. Hu, L. You, Q. Li, A. Li, H. Wang, G. Mu, Z. Chen, H. Zhang, G. Yu, J. Zhu, Q. Sun, C. Lin, H. Xiao, X. Xie and M. Jiang, Nat. Commun., 2014, 5, 5708 CrossRef CAS PubMed .
  23. H. Sahabudeen, H. Qi, B. A. Glatz, D. Tranca, R. Dong, Y. Hou, T. Zhang, C. Kuttner, T. Lehnert, G. Seifert, U. Kaiser, A. Fery, Z. Zheng and X. Feng, Nat. Commun., 2016, 7, 13461 CrossRef CAS PubMed .
  24. A. Turchanin and A. Gölzhäuser, Adv. Mater., 2016, 28, 6075–6103 CrossRef CAS PubMed .
  25. C. Tan and H. Zhang, Nat. Commun., 2015, 6, 7873 CrossRef CAS PubMed .
  26. K.-A. N. Duerloo, Y. Li and E. J. Reed, Nat. Commun., 2014, 5, 4214 CAS .
  27. L. Britnell, R. M. Ribeiro, A. Eckmann, R. Jalil, B. D. Belle, A. Mishchenko, Y.-J. Kim, R. V. Gorbachev, T. Georgiou, S. V. Morozov, A. N. Grigorenko, A. K. Geim, C. Casiraghi, A. H. C. Neto and K. S. Novoselov, Science, 2013, 340, 1311–1314 CrossRef CAS PubMed .
  28. Y. Liu, N. O. Weiss, X. Duan, H.-C. Cheng, Y. Huang and X. Duan, Nat. Rev. Mater., 2016, 1, 16042 CrossRef CAS .
  29. F. Bonaccorso, Z. Sun, T. Hasan and A. C. Ferrari, Nat. Photonics, 2010, 4, 611–622 CrossRef CAS .
  30. M. P. Levendorf, C.-J. Kim, L. Brown, P. Y. Huang, R. W. Havener, D. A. Muller and J. Park, Nature, 2012, 488, 627–632 CrossRef CAS PubMed .
  31. G. Fiori, A. Betti, S. Bruzzone and G. Iannaccone, ACS Nano, 2012, 6, 2642–2648 CrossRef CAS PubMed .
  32. Z. Liu, L. Ma, G. Shi, W. Zhou, Y. Gong, S. Lei, X. Yang, J. Zhang, J. Yu, K. P. Hackenberg, A. Babakhani, J.-C. Idrobo, R. Vajtai, J. Lou and P. M. Ajayan, Nat. Nanotechnol., 2013, 8, 119–124 CrossRef CAS PubMed .
  33. Y. Gao, Y. Zhang, P. Chen, Y. Li, M. Liu, T. Gao, D. Ma, Y. Chen, Z. Cheng, X. Qiu, W. Duan and Z. Liu, Nano Lett., 2013, 13, 3439–3443 CrossRef CAS PubMed .
  34. H.-C. Shin, Y. Jang, T.-H. Kim, J.-H. Lee, D.-H. Oh, S. J. Ahn, J. H. Lee, Y. Moon, J.-H. Park, S. J. Yoo, C.-Y. Park, D. Whang, C.-W. Yang and J. R. Ahn, J. Am. Chem. Soc., 2015, 137, 6897–6905 CrossRef CAS PubMed .
  35. Z. Liu, L. H. G. Tizei, Y. Sato, Y.-C. Lin, C.-H. Yeh, P.-W. Chiu, M. Terauchi, S. Iijima and K. Suenaga, Small, 2016, 12, 252–259 CrossRef CAS PubMed .
  36. A. S. Mayorov, R. V. Gorbachev, S. V. Morozov, L. Britnell, R. Jalil, L. A. Ponomarenko, P. Blake, K. S. Novoselov, K. Watanabe, T. Taniguchi and A. K. Geim, Nano Lett., 2011, 11, 2396–2399 CrossRef CAS PubMed .
  37. L. A. Ponomarenko, A. K. Geim, A. A. Zhukov, R. Jalil, S. V. Morozov, K. S. Novoselov, I. V. Grigorieva, E. H. Hill, V. V. Cheianov, V. I. Fal'ko, K. Watanabe, T. Taniguchi and R. V. Gorbachev, Nat. Phys., 2011, 7, 958–961 CrossRef CAS .
  38. H. Wang, T. Taychatanapat, A. Hsu, K. Watanabe, T. Taniguchi, P. Jarillo-Herrero and T. Palacios, IEEE Electron Device Lett., 2011, 32, 1209–1211 CrossRef CAS .
  39. C. T. Nottbohm, A. Turchanin, A. Beyer, R. Stosch and A. Gölzhäuser, Small, 2011, 7, 874–883 CrossRef CAS PubMed .
  40. L. Britnell, R. V. Gorbachev, R. Jalil, B. D. Belle, F. Schedin, A. Mishchenko, T. Georgiou, M. I. Katsnelson, L. Eaves, S. V. Morozov, N. M. R. Peres, J. Leist, A. K. Geim, K. S. Novoselov and L. A. Ponomarenko, Science, 2012, 335, 947–950 CrossRef CAS PubMed .
  41. S. J. Haigh, A. Gholinia, R. Jalil, S. Romani, L. Britnell, D. C. Elias, K. S. Novoselov, L. A. Ponomarenko, A. K. Geim and R. Gorbachev, Nat. Mater., 2012, 11, 764–767 CrossRef CAS PubMed .
  42. Y.-C. Lin, C.-C. Lu, C.-H. Yeh, C. Jin, K. Suenaga and P.-W. Chiu, Nano Lett., 2012, 12, 414–419 CrossRef CAS PubMed .
  43. T. Uwanno, Y. Hattori, T. Taniguchi, K. Watanabe and K. Nagashio, 2D Mater., 2015, 2, 41002 CrossRef .
  44. A. V. Kretinin, Y. Cao, J. S. Tu, G. L. Yu, R. Jalil, K. S. Novoselov, S. J. Haigh, A. Gholinia, A. Mishchenko, M. Lozada, T. Georgiou, C. R. Woods, F. Withers, P. Blake, G. Eda, A. Wirsig, C. Hucho, K. Watanabe, T. Taniguchi, A. K. Geim and R. V. Gorbachev, Nano Lett., 2014, 14, 3270–3276 CrossRef CAS PubMed .
  45. L. Wang, I. Meric, P. Y. Huang, Q. Gao, Y. Gao, H. Tran, T. Taniguchi, K. Watanabe, L. M. Campos, D. A. Muller, J. Guo, P. Kim, J. Hone, K. L. Shepard and C. R. Dean, Science, 2013, 342, 614–617 CrossRef CAS PubMed .
  46. A. Castellanos-Gomez, M. Buscema, R. Molenaar, V. Singh, L. Janssen, H. S. J. van der Zant and G. A. Steele, 2D Mater., 2014, 1, 11002 CrossRef .
  47. P. J. Zomer, M. H. D. Guimarães, J. C. Brant, N. Tombros and B. J. van Wees, Appl. Phys. Lett., 2014, 105, 13101 CrossRef .
  48. M. H. Gass, U. Bangert, A. L. Bleloch, P. Wang, R. R. Nair and A. K. Geim, Nat. Nanotechnol., 2008, 3, 676–681 CrossRef CAS PubMed .
  49. Y. Cao, A. Mishchenko, G. L. Yu, E. Khestanova, A. P. Rooney, E. Prestat, A. V. Kretinin, P. Blake, M. B. Shalom, C. Woods, J. Chapman, G. Balakrishnan, I. V. Grigorieva, K. S. Novoselov, B. A. Piot, M. Potemski, K. Watanabe, T. Taniguchi, S. J. Haigh, A. K. Geim and R. V. Gorbachev, Nano Lett., 2015, 15, 4914–4921 CrossRef CAS PubMed .
  50. K. Kim, M. Yankowitz, B. Fallahazad, S. Kang, H. C. P. Movva, S. Huang, S. Larentis, C. M. Corbet, T. Taniguchi, K. Watanabe, S. K. Banerjee, B. J. LeRoy and E. Tutuc, Nano Lett., 2016, 16, 1989–1995 CrossRef CAS PubMed .
  51. C. R. Woods, F. Withers, M. J. Zhu, Y. Cao, G. Yu, A. Kozikov, M. B. Shalom, S. V. Morozov, M. M. van Wijk, A. Fasolino, M. I. Katsnelson, K. Watanabe, T. Taniguchi, A. K. Geim, A. Mishchenko and K. S. Novoselov, Nat. Commun., 2016, 7, 10800 CrossRef CAS PubMed .
  52. D. Wang, G. Chen, C. Li, M. Cheng, W. Yang, S. Wu, G. Xie, J. Zhang, J. Zhao, X. Lu, P. Chen, G. Wang, J. Meng, J. Tang, R. Yang, C. He, D. Liu, D. Shi, K. Watanabe, T. Taniguchi, J. Feng, Y. Zhang and G. Zhang, Phys. Rev. Lett., 2016, 116, 126101 CrossRef PubMed .
  53. M. Zhu, D. Ghazaryan, S.-K. Son, C. R. Woods, A. Misra, L. He, T. Taniguchi, K. Watanabe, K. S. Novoselov, Y. Cao and A. Mishchenko, 2D Mater., 2017, 4, 11013 CrossRef .
  54. S. B. Desai, S. R. Madhvapathy, M. Amani, D. Kiriya, M. Hettick, M. Tosun, Y. Zhou, M. Dubey, J. W. Ager, D. Chrzan and A. Javey, Adv. Mater., 2016, 28, 4053–4058 CrossRef CAS PubMed .
  55. K. Roy, M. Padmanabhan, S. Goswami, T. P. Sai, G. Ramalingam, S. Raghavan and A. Ghosh, Nat. Nanotechnol., 2013, 8, 826–830 CrossRef CAS PubMed .
  56. S. Bertolazzi, D. Krasnozhon and A. Kis, ACS Nano, 2013, 7, 3246–3252 CrossRef CAS PubMed .
  57. W. J. Yu, Z. Li, H. Zhou, Y. Chen, Y. Wang, Y. Huang and X. Duan, Nat. Mater., 2013, 12, 246–252 CrossRef CAS PubMed .
  58. W. J. Yu, Y. Liu, H. Zhou, A. Yin, Z. Li, Y. Huang and X. Duan, Nat. Nanotechnol., 2013, 8, 952–958 CrossRef CAS PubMed .
  59. S. Larentis, J. R. Tolsma, B. Fallahazad, D. C. Dillen, K. Kim, A. H. MacDonald and E. Tutuc, Nano Lett., 2014, 14, 2039–2045 CrossRef CAS PubMed .
  60. W. Zhang, C.-P. Chuu, J.-K. Huang, C.-H. Chen, M.-L. Tsai, Y.-H. Chang, C.-T. Liang, Y.-Z. Chen, Y.-L. Chueh, J.-H. He, M.-Y. Chou and L.-J. Li, Sci. Rep., 2014, 4, 3826 CrossRef PubMed .
  61. P. T. K. Loan, W. Zhang, C.-T. Lin, K.-H. Wei, L.-J. Li and C.-H. Chen, Adv. Mater., 2014, 26, 4838–4844 CrossRef CAS PubMed .
  62. D. Pierucci, H. Henck, C. H. Naylor, H. Sediri, E. Lhuillier, A. Balan, J. E. Rault, Y. J. Dappe, F. Bertran, P. L. Fèvre, A. T. C. Johnson and A. Ouerghi, Sci. Rep., 2016, 6, 26656 CrossRef CAS PubMed .
  63. T. Georgiou, R. Jalil, B. D. Belle, L. Britnell, R. V. Gorbachev, S. V. Morozov, Y.-J. Kim, A. Gholinia, S. J. Haigh, O. Makarovsky, L. Eaves, L. A. Ponomarenko, A. K. Geim, K. S. Novoselov and A. Mishchenko, Nat. Nanotechnol., 2013, 8, 100–103 CrossRef CAS PubMed .
  64. P. J. Zomer, S. P. Dash, N. Tombros and B. J. van Wees, Appl. Phys. Lett., 2011, 99, 232104 CrossRef .
  65. M. S. Choi, G.-H. Lee, Y.-J. Yu, D.-Y. Lee, S. Hwan Lee, P. Kim, J. Hone and W. Jong Yoo, Nat. Commun., 2013, 4, 1624 CrossRef PubMed .
  66. G.-H. Lee, Y.-J. Yu, X. Cui, N. Petrone, C.-H. Lee, M. S. Choi, D.-Y. Lee, C. Lee, W. J. Yoo, K. Watanabe, T. Taniguchi, C. Nuckolls, P. Kim and J. Hone, ACS Nano, 2013, 7, 7931–7936 CrossRef CAS PubMed .
  67. L. Banszerus, M. Schmitz, S. Engels, M. Goldsche, K. Watanabe, T. Taniguchi, B. Beschoten and C. Stampfer, Nano Lett., 2016, 16, 1387–1391 CrossRef CAS PubMed .
  68. Z. Liu, L. Song, S. Zhao, J. Huang, L. Ma, J. Zhang, J. Lou and P. M. Ajayan, Nano Lett., 2011, 11, 2032–2037 CrossRef CAS PubMed .
  69. Y.-C. Lin, N. Lu, N. Perea-Lopez, J. Li, Z. Lin, X. Peng, C. H. Lee, C. Sun, L. Calderin, P. N. Browning, M. S. Bresnehan, M. J. Kim, T. S. Mayer, M. Terrones and J. A. Robinson, ACS Nano, 2014, 8, 3715–3723 CrossRef CAS PubMed .
  70. C. Oshima, A. Itoh, E. Rokuta, T. Tanaka, K. Yamashita and T. Sakurai, Solid State Commun., 2000, 116, 37–40 CrossRef CAS .
  71. X. Ding, G. Ding, X. Xie, F. Huang and M. Jiang, Carbon, 2011, 49, 2522–2525 CrossRef CAS .
  72. J. M. Garcia, U. Wurstbauer, A. Levy, L. N. Pfeiffer, A. Pinczuk, A. S. Plaut, L. Wang, C. R. Dean, R. Buizza, A. M. Van Der Zande, J. Hone, K. Watanabe and T. Taniguchi, Solid State Commun., 2012, 152, 975–978 CrossRef CAS .
  73. S. Tang, G. Ding, X. Xie, J. Chen, C. Wang, X. Ding, F. Huang, W. Lu and M. Jiang, Carbon, 2012, 50, 329–331 CrossRef CAS .
  74. M. Wang, S. K. Jang, W.-J. Jang, M. Kim, S.-Y. Park, S.-W. Kim, S.-J. Kahng, J.-Y. Choi, R. S. Ruoff, Y. J. Song and S. Lee, Adv. Mater., 2013, 25, 2746–2752 CrossRef CAS PubMed .
  75. M. Wang, M. Kim, D. Odkhuu, N. Park, J. Lee, W.-J. Jang, S.-J. Kahng, R. S. Ruoff, Y. J. Song and S. Lee, ACS Nano, 2014, 8, 5478–5483 CrossRef CAS PubMed .
  76. S. M. Kim, A. Hsu, P. T. Araujo, Y.-H. Lee, T. Palacios, M. Dresselhaus, J.-C. Idrobo, K. K. Kim and J. Kong, Nano Lett., 2013, 13, 933–941 CrossRef CAS PubMed .
  77. S. Roth, F. Matsui, T. Greber and J. Osterwalder, Nano Lett., 2013, 13, 2668–2675 CrossRef CAS PubMed .
  78. W. Yang, G. Chen, Z. Shi, C.-C. Liu, L. Zhang, G. Xie, M. Cheng, D. Wang, R. Yang, D. Shi, K. Watanabe, T. Taniguchi, Y. Yao, Y. Zhang and G. Zhang, Nat. Mater., 2013, 12, 792–797 CrossRef CAS PubMed .
  79. S. Tang, H. Wang, Y. Zhang, A. Li, H. Xie, X. Liu, L. Liu, T. Li, F. Huang, X. Xie and M. Jiang, Sci. Rep., 2013, 3, 2666 CrossRef PubMed .
  80. J. H. Meng, X. W. Zhang, H. L. Wang, X. B. Ren, C. H. Jin, Z. G. Yin, X. Liu and H. Liu, Nanoscale, 2015, 7, 16046–16053 RSC .
  81. C. Zhang, S. Zhao, C. Jin, A. L. Koh, Y. Zhou, W. Xu, Q. Li, Q. Xiong, H. Peng and Z. Liu, Nat. Commun., 2015, 6, 6519 CrossRef CAS PubMed .
  82. Y. Yang, Q. Fu, H. Li, M. Wei, J. Xiao, W. Wei and X. Bao, ACS Nano, 2015, 9, 11589–11598 CrossRef CAS PubMed .
  83. S. Roth, T. Greber and J. Osterwalder, ACS Nano, 2016, 10, 11187–11195 CrossRef CAS PubMed .
  84. X. Song, T. Gao, Y. Nie, J. Zhuang, J. Sun, D. Ma, J. Shi, Y. Lin, F. Ding, Y. Zhang and Z. Liu, Nano Lett., 2016, 16, 6109–6116 CrossRef CAS PubMed .
  85. Y. Shi, W. Zhou, A.-Y. Lu, W. Fang, Y.-H. Lee, A. L. Hsu, S. M. Kim, K. K. Kim, H. Y. Yang, L.-J. Li, J.-C. Idrobo and J. Kong, Nano Lett., 2012, 12, 2784–2791 CrossRef CAS PubMed .
  86. X. Ling, Y.-H. Lee, Y. Lin, W. Fang, L. Yu, M. S. Dresselhaus and J. Kong, Nano Lett., 2014, 14, 464–472 CrossRef CAS PubMed .
  87. H. Ago, H. Endo, P. Solís-Fernández, R. Takizawa, Y. Ohta, Y. Fujita, K. Yamamoto and M. Tsuji, ACS Appl. Mater. Interfaces, 2015, 7, 5265–5273 CAS .
  88. R. M. Yunus, H. Endo, M. Tsuji and H. Ago, Phys. Chem. Chem. Phys., 2015, 17, 25210–25215 RSC .
  89. J. A. Miwa, M. Dendzik, S. S. Grønborg, M. Bianchi, J. V. Lauritsen, P. Hofmann and S. Ulstrup, ACS Nano, 2015, 9, 6502–6510 CrossRef CAS PubMed .
  90. A. Azizi, S. Eichfeld, G. Geschwind, K. Zhang, B. Jiang, D. Mukherjee, L. Hossain, A. F. Piasecki, B. Kabius, J. A. Robinson and N. Alem, ACS Nano, 2015, 9, 4882–4890 CrossRef CAS PubMed .
  91. J. Shi, M. Liu, J. Wen, X. Ren, X. Zhou, Q. Ji, D. Ma, Y. Zhang, C. Jin, H. Chen, S. Deng, N. Xu, Z. Liu and Y. Zhang, Adv. Mater., 2015, 27, 7086–7092 CrossRef CAS PubMed .
  92. W. Wan, X. Li, X. Li, B. Xu, L. Zhan, Z. Zhao, P. Zhang, S. Q. Wu, Z. Zhu, H. Huang, Y. Zhou and W. Cai, RSC Adv., 2015, 6, 323–330 RSC .
  93. X. Liu, I. Balla, H. Bergeron, G. P. Campbell, M. J. Bedzyk and M. C. Hersam, ACS Nano, 2016, 10, 1067–1075 CrossRef CAS PubMed .
  94. H. Ago, S. Fukamachi, H. Endo, P. Solís-Fernández, R. M. Yunus, Y. Uchida, V. Panchal, O. Kazakova and M. Tsuji, ACS Nano, 2016, 10, 3233–3240 CrossRef CAS PubMed .
  95. S. Ulstrup, A. G. Čabo, J. A. Miwa, J. M. Riley, S. S. Grønborg, J. C. Johannsen, C. Cacho, O. Alexander, R. T. Chapman, E. Springate, M. Bianchi, M. Dendzik, J. V. Lauritsen, P. D. C. King and P. Hofmann, ACS Nano, 2016, 10, 6315–6322 CrossRef CAS PubMed .
  96. W. Wan, L. Zhan, B. Xu, F. Zhao, Z. Zhu, Y. Zhou, Z. Yang, T. Shih and W. Cai, Small, 2017, 13, 1603549 CrossRef PubMed .
  97. G. V. Bianco, M. Losurdo, M. M. Giangregorio, A. Sacchetti, P. Prete, N. Lovergine, P. Capezzuto and G. Bruno, RSC Adv., 2015, 5, 98700–98708 RSC .
  98. A. Rossi, H. Büch, C. D. Rienzo, V. Miseikis, D. Convertino, A. Al-Temimy, V. Voliani, M. Gemmi, V. Piazza and C. Coletti, 2D Mater., 2016, 3, 31013 CrossRef .
  99. S.-Y. Kim, J. Kwak, J. H. Kim, J.-U. Lee, Y. Jo, S. Y. Kim, H. Cheong, Z. Lee and S.-Y. Kwon, 2D Mater., 2017, 4, 11007 CrossRef .
  100. C. Lan, C. Li, S. Wang, T. He, Z. Zhou, D. Wei, H. Guo, H. Yang and Y. Liu, J. Mater. Chem. C, 2017, 5, 1494–1500 RSC .
  101. G. W. Shim, K. Yoo, S.-B. Seo, J. Shin, D. Y. Jung, I.-S. Kang, C. W. Ahn, B. J. Cho and S.-Y. Choi, ACS Nano, 2014, 8, 6655–6662 CrossRef CAS PubMed .
  102. Y.-C. Lin, C.-Y. S. Chang, R. K. Ghosh, J. Li, H. Zhu, R. Addou, B. Diaconescu, T. Ohta, X. Peng, N. Lu, M. J. Kim, J. T. Robinson, R. M. Wallace, T. S. Mayer, S. Datta, L.-J. Li and J. A. Robinson, Nano Lett., 2014, 14, 6936–6941 CrossRef CAS PubMed .
  103. Y.-C. Lin, J. Li, S. C. de la Barrera, S. M. Eichfeld, Y. Nie, R. Addou, P. C. Mende, R. M. Wallace, K. Cho, R. M. Feenstra and J. A. Robinson, Nanoscale, 2016, 8, 8947–8954 RSC .
  104. M. M. Ugeda, A. J. Bradley, Y. Zhang, S. Onishi, Y. Chen, W. Ruan, C. Ojeda-Aristizabal, H. Ryu, M. T. Edmonds, H.-Z. Tsai, A. Riss, S.-K. Mo, D. Lee, A. Zettl, Z. Hussain, Z.-X. Shen and M. F. Crommie, Nat. Phys., 2016, 12, 92–97 CrossRef CAS .
  105. Y.-C. Lin, R. K. Ghosh, R. Addou, N. Lu, S. M. Eichfeld, H. Zhu, M.-Y. Li, X. Peng, M. J. Kim, L.-J. Li, R. M. Wallace, S. Datta and J. A. Robinson, Nat. Commun., 2015, 6, 7311 CrossRef CAS PubMed .
  106. X. Li, L. Basile, B. Huang, C. Ma, J. Lee, I. V. Vlassiouk, A. A. Puretzky, M.-W. Lin, M. Yoon, M. Chi, J. C. Idrobo, C. M. Rouleau, B. G. Sumpter, D. B. Geohegan and K. Xiao, ACS Nano, 2015, 9, 8078–8088 CrossRef CAS PubMed .
  107. Z. Ben Aziza, H. Henck, D. Pierucci, M. G. Silly, E. Lhuillier, G. Patriarche, F. Sirotti, M. Eddrief and A. Ouerghi, ACS Nano, 2016, 10, 9679–9686 CrossRef CAS PubMed .
  108. W. Dang, H. Peng, H. Li, P. Wang and Z. Liu, Nano Lett., 2010, 10, 2870–2876 CrossRef CAS PubMed .
  109. M. Lin, D. Wu, Y. Zhou, W. Huang, W. Jiang, W. Zheng, S. Zhao, C. Jin, Y. Guo, H. Peng and Z. Liu, J. Am. Chem. Soc., 2013, 135, 13274–13277 CrossRef CAS PubMed .
  110. K. Vaklinova, A. Hoyer, M. Burghard and K. Kern, Nano Lett., 2016, 16, 2595–2602 CrossRef CAS PubMed .
  111. H.-K. Hong, J. Jo, D. Hwang, J. Lee, N. Y. Kim, S. Son, J. H. Kim, M.-J. Jin, Y. C. Jun, R. Erni, S. K. Kwak, J.-W. Yoo and Z. Lee, Nano Lett., 2017, 17, 120–127 CrossRef CAS PubMed .
  112. L. Niu, X. Liu, C. Cong, C. Wu, D. Wu, T. R. Chang, H. Wang, Q. Zeng, J. Zhou, X. Wang, W. Fu, P. Yu, Q. Fu, S. Najmaei, Z. Zhang, B. I. Yakobson, B. K. Tay, W. Zhou, H. T. Jeng, H. Lin, T. C. Sum, C. Jin, H. He, T. Yu and Z. Liu, Adv. Mater., 2015, 27, 7800–7808 CrossRef CAS PubMed .
  113. C. H. Lui, L. Liu, K. F. Mak, G. W. Flynn and T. F. Heinz, Nature, 2009, 462, 339–341 CrossRef CAS PubMed .
  114. Y. Zhang, V. W. Brar, C. Girit, A. Zettl and M. F. Crommie, Nat. Phys., 2009, 5, 722–726 CrossRef CAS .
  115. J. Jobst, A. J. H. van der Torren, E. E. Krasovskii, J. Balgley, C. R. Dean, R. M. Tromp and S. J. van der Molen, Nat. Commun., 2016, 7, 13621 CrossRef CAS PubMed .
  116. J. Xue, J. Sanchez-Yamagishi, D. Bulmash, P. Jacquod, A. Deshpande, K. Watanabe, T. Taniguchi, P. Jarillo-Herrero and B. J. LeRoy, Nat. Mater., 2011, 10, 282–285 CrossRef CAS PubMed .
  117. R. Decker, Y. Wang, V. W. Brar, W. Regan, H.-Z. Tsai, Q. Wu, W. Gannett, A. Zettl and M. F. Crommie, Nano Lett., 2011, 11, 2291–2295 CrossRef CAS PubMed .
  118. S. M. Kim, A. Hsu, M. H. Park, S. H. Chae, S. J. Yun, J. S. Lee, D.-H. Cho, W. Fang, C. Lee, T. Palacios, M. Dresselhaus, K. K. Kim, Y. H. Lee and J. Kong, Nat. Commun., 2015, 6, 8662 CrossRef CAS PubMed .
  119. D. H. Tien, J.-Y. Park, K. B. Kim, N. Lee, T. Choi, P. Kim, T. Taniguchi, K. Watanabe and Y. Seo, ACS Appl. Mater. Interfaces, 2016, 8, 3072–3078 CAS .
  120. Z. Liu, Y. Gong, W. Zhou, L. Ma, J. Yu, J. C. Idrobo, J. Jung, A. H. MacDonald, R. Vajtai, J. Lou and P. M. Ajayan, Nat. Commun., 2013, 4, 2541 Search PubMed .
  121. R. Zan, Q. M. Ramasse, R. Jalil, T. Georgiou, U. Bangert and K. S. Novoselov, ACS Nano, 2013, 7, 10167–10174 CrossRef CAS PubMed .
  122. G. Algara-Siller, S. Kurasch, M. Sedighi, O. Lehtinen and U. Kaiser, Appl. Phys. Lett., 2013, 103, 203107 CrossRef .
  123. A. M. Goossens, S. C. M. Driessen, T. A. Baart, K. Watanabe, T. Taniguchi and L. M. K. Vandersypen, Nano Lett., 2012, 12, 4656–4660 CrossRef CAS PubMed .
  124. L. Britnell, R. V. Gorbachev, R. Jalil, B. D. Belle, F. Schedin, M. I. Katsnelson, L. Eaves, S. V. Morozov, A. S. Mayorov, N. M. R. Peres, A. H. Castro Neto, J. Leist, A. K. Geim, L. A. Ponomarenko and K. S. Novoselov, Nano Lett., 2012, 12, 1707–1710 CrossRef CAS PubMed .
  125. C. R. Dean, A. F. Young, P. Cadden-Zimansky, L. Wang, H. Ren, K. Watanabe, T. Taniguchi, P. Kim, J. Hone and K. L. Shepard, Nat. Phys., 2011, 7, 693–696 CrossRef CAS .
  126. A. F. Young, C. R. Dean, L. Wang, H. Ren, P. Cadden-Zimansky, K. Watanabe, T. Taniguchi, J. Hone, K. L. Shepard and P. Kim, Nat. Phys., 2012, 8, 550–556 CrossRef CAS .
  127. F. Amet, A. J. Bestwick, J. R. Williams, L. Balicas, K. Watanabe, T. Taniguchi and D. Goldhaber-Gordon, Nat. Commun., 2015, 6, 5838 CrossRef CAS PubMed .
  128. P. J. Zomer, M. H. D. Guimarães, N. Tombros and B. J. van Wees, Phys. Rev. B: Condens. Matter Mater. Phys., 2012, 86, 161416 CrossRef .
  129. A. F. Young, J. D. Sanchez-Yamagishi, B. Hunt, S. H. Choi, K. Watanabe, T. Taniguchi, R. C. Ashoori and P. Jarillo-Herrero, Nature, 2014, 505, 528–532 CrossRef CAS PubMed .
  130. M. Yankowitz, J. Xue, D. Cormode, J. D. Sanchez-Yamagishi, K. Watanabe, T. Taniguchi, P. Jarillo-Herrero, P. Jacquod and B. J. LeRoy, Nat. Phys., 2012, 8, 382–386 CrossRef CAS .
  131. C.-H. Park, L. Yang, Y.-W. Son, M. L. Cohen and S. G. Louie, Phys. Rev. Lett., 2008, 101, 126804 CrossRef PubMed .
  132. B. Hunt, J. D. Sanchez-Yamagishi, A. F. Young, M. Yankowitz, B. J. LeRoy, K. Watanabe, T. Taniguchi, P. Moon, M. Koshino, P. Jarillo-Herrero and R. C. Ashoori, Science, 2013, 340, 1427–1430 CrossRef CAS PubMed .
  133. E. Wang, X. Lu, S. Ding, W. Yao, M. Yan, G. Wan, K. Deng, S. Wang, G. Chen, L. Ma, J. Jung, A. V. Fedorov, Y. Zhang, G. Zhang and S. Zhou, Nat. Phys., 2016, 12, 1111–1115 CrossRef CAS .
  134. C. R. Woods, L. Britnell, A. Eckmann, R. S. Ma, J. C. Lu, H. M. Guo, X. Lin, G. L. Yu, Y. Cao, R. V. Gorbachev, A. V. Kretinin, J. Park, L. A. Ponomarenko, M. I. Katsnelson, Y. N. Gornostyrev, K. Watanabe, T. Taniguchi, C. Casiraghi, H.-J. Gao, A. K. Geim and K. S. Novoselov, Nat. Phys., 2014, 10, 451–456 CrossRef CAS .
  135. E. Wang, G. Chen, G. Wan, X. Lu, C. Chen, J. Avila, A. V. Fedorov, G. Zhang, M. C. Asensio, Y. Zhang and S. Zhou, J. Phys.: Condens. Matter, 2016, 28, 444002 CrossRef PubMed .
  136. L. A. Ponomarenko, R. V. Gorbachev, G. L. Yu, D. C. Elias, R. Jalil, A. A. Patel, A. Mishchenko, A. S. Mayorov, C. R. Woods, J. R. Wallbank, M. Mucha-Kruczynski, B. A. Piot, M. Potemski, I. V. Grigorieva, K. S. Novoselov, F. Guinea, V. I. Fal'ko and A. K. Geim, Nature, 2013, 497, 594–597 CrossRef CAS PubMed .
  137. C. Handschin, P. Makk, P. Rickhaus, M.-H. Liu, K. Watanabe, T. Taniguchi, K. Richter and C. Schönenberger, Nano Lett., 2017, 17, 328–333 CrossRef CAS PubMed .
  138. D. Pierucci, H. Henck, J. Avila, A. Balan, C. H. Naylor, G. Patriarche, Y. J. Dappe, M. G. Silly, F. Sirotti, A. T. C. Johnson, M. C. Asensio and A. Ouerghi, Nano Lett., 2016, 16, 4054–4061 CrossRef CAS PubMed .
  139. K. M. McCreary, A. T. Hanbicki, J. T. Robinson, E. Cobas, J. C. Culbertson, A. L. Friedman, G. G. Jernigan and B. T. Jonker, Adv. Funct. Mater., 2014, 24, 6449–6454 CrossRef CAS .
  140. X. Liu, I. Balla, H. Bergeron and M. C. Hersam, J. Phys. Chem. C, 2016, 120, 20798–20805 CAS .
  141. H. Coy Diaz, J. Avila, C. Chen, R. Addou, M. C. Asensio and M. Batzill, Nano Lett., 2015, 15, 1135–1140 CrossRef CAS PubMed .
  142. J. Y. Tan, A. Avsar, J. Balakrishnan, G. K. W. Koon, T. Taychatanapat, E. C. T. O'Farrell, K. Watanabe, T. Taniguchi, G. Eda, A. H. C. Neto and B. Özyilmaz, Appl. Phys. Lett., 2014, 104, 183504 CrossRef .
  143. M. Yankowitz, S. Larentis, K. Kim, J. Xue, D. McKenzie, S. Huang, M. Paggen, M. N. Ali, R. J. Cava, E. Tutuc and B. J. LeRoy, Nano Lett., 2015, 15, 1925–1929 CrossRef CAS PubMed .
  144. H. Tan, Y. Fan, Y. Rong, B. Porter, C. S. Lau, Y. Zhou, Z. He, S. Wang, H. Bhaskaran and J. H. Warner, ACS Appl. Mater. Interfaces, 2016, 8, 1644–1652 CAS .
  145. C. E. Giusca, I. Rungger, V. Panchal, C. Melios, Z. Lin, Y.-C. Lin, E. Kahn, A. L. Elías, J. A. Robinson, M. Terrones and O. Kazakova, ACS Nano, 2016, 10, 7840–7846 CrossRef CAS PubMed .
  146. H. Xu, J. Wu, Q. Feng, N. Mao, C. Wang and J. Zhang, Small, 2014, 10, 2300–2306 CrossRef CAS PubMed .
  147. C.-H. Lee, G.-H. Lee, A. M. van der Zande, W. Chen, Y. Li, M. Han, X. Cui, G. Arefe, C. Nuckolls, T. F. Heinz, J. Guo, J. Hone and P. Kim, Nat. Nanotechnol., 2014, 9, 676–681 CrossRef CAS PubMed .
  148. M. Massicotte, P. Schmidt, F. Vialla, K. G. Schädler, A. Reserbat-Plantey, K. Watanabe, T. Taniguchi, K. J. Tielrooij and F. H. L. Koppens, Nat. Nanotechnol., 2016, 11, 42–46 CrossRef CAS PubMed .
  149. H. Tan, Y. Fan, Y. Zhou, Q. Chen, W. Xu and J. H. Warner, ACS Nano, 2016, 10, 7866–7873 CrossRef CAS PubMed .
  150. Q. A. Vu, J. H. Lee, V. L. Nguyen, Y. S. Shin, S. C. Lim, K. Lee, J. Heo, S. Park, K. Kim, Y. H. Lee and W. J. Yu, Nano Lett., 2017, 17, 453–459 CrossRef CAS PubMed .
  151. R.-J. Shiue, Y. Gao, Y. Wang, C. Peng, A. D. Robertson, D. K. Efetov, S. Assefa, F. H. L. Koppens, J. Hone and D. Englund, Nano Lett., 2015, 15, 7288–7293 CrossRef CAS PubMed .
  152. M. Bernardi, M. Palummo and J. C. Grossman, Nano Lett., 2013, 13, 3664–3670 CrossRef CAS PubMed .
  153. M. Shanmugam, R. Jacobs-Gedrim, E. S. Song and B. Yu, Nanoscale, 2014, 6, 12682–12689 RSC .
  154. Y. Tsuboi, F. Wang, D. Kozawa, K. Funahashi, S. Mouri, Y. Miyauchi, T. Takenobu and K. Matsuda, Nanoscale, 2015, 7, 14476–14482 RSC .
  155. F. Withers, O. Del Pozo-Zamudio, A. Mishchenko, A. P. Rooney, A. Gholinia, K. Watanabe, T. Taniguchi, S. J. Haigh, A. K. Geim, A. I. Tartakovskii and K. S. Novoselov, Nat. Mater., 2015, 14, 301–306 CrossRef CAS PubMed .
  156. F. Withers, O. Del Pozo-Zamudio, S. Schwarz, S. Dufferwiel, P. M. Walker, T. Godde, A. P. Rooney, A. Gholinia, C. R. Woods, P. Blake, S. J. Haigh, K. Watanabe, T. Taniguchi, I. L. Aleiner, A. K. Geim, V. I. Fal'ko, A. I. Tartakovskii and K. S. Novoselov, Nano Lett., 2015, 15, 8223–8228 CrossRef CAS PubMed .
  157. C. Palacios-Berraquero, M. Barbone, D. M. Kara, X. Chen, I. Goykhman, D. Yoon, A. K. Ott, J. Beitner, K. Watanabe, T. Taniguchi, A. C. Ferrari and M. Atatüre, Nat. Commun., 2016, 7, 12978 CrossRef CAS PubMed .
  158. G. Clark, J. R. Schaibley, J. Ross, T. Taniguchi, K. Watanabe, J. R. Hendrickson, S. Mou, W. Yao and X. Xu, Nano Lett., 2016, 16, 3944–3948 CrossRef CAS PubMed .
  159. S. Schwarz, A. Kozikov, F. Withers, J. K. Maguire, A. P. Foster, S. Dufferwiel, L. Hague, M. N. Makhonin, L. R. Wilson, A. K. Geim, K. S. Novoselov and A. I. Tartakovskii, 2D Mater., 2016, 3, 25038 CrossRef .
  160. Y. Gao, R.-J. Shiue, X. Gan, L. Li, C. Peng, I. Meric, L. Wang, A. Szep, D. Walker, J. Hone and D. Englund, Nano Lett., 2015, 15, 2001–2005 CrossRef CAS PubMed .
  161. C.-H. Liu, G. Clark, T. Fryett, S. Wu, J. Zheng, F. Hatami, X. Xu and A. Majumdar, Nano Lett., 2017, 17, 200–205 CrossRef CAS PubMed .
  162. Q. A. Vu, Y. S. Shin, Y. R. Kim, V. L. Nguyen, W. T. Kang, H. Kim, D. H. Luong, I. M. Lee, K. Lee, D.-S. Ko, J. Heo, S. Park, Y. H. Lee and W. J. Yu, Nat. Commun., 2016, 7, 12725 CrossRef CAS PubMed .
  163. L. Ju, J. Velasco Jr, E. Huang, S. Kahn, C. Nosiglia, H.-Z. Tsai, W. Yang, T. Taniguchi, K. Watanabe, Y. Zhang, G. Zhang, M. Crommie, A. Zettl and F. Wang, Nat. Nanotechnol., 2014, 9, 348–352 CrossRef CAS PubMed .
  164. J. Velasco, L. Ju, D. Wong, S. Kahn, J. Lee, H.-Z. Tsai, C. Germany, S. Wickenburg, J. Lu, T. Taniguchi, K. Watanabe, A. Zettl, F. Wang and M. F. Crommie, Nano Lett., 2016, 16, 1620–1625 CrossRef CAS PubMed .
  165. B. Cho, J. Yoon, S. K. Lim, A. R. Kim, D.-H. Kim, S.-G. Park, J.-D. Kwon, Y.-J. Lee, K.-H. Lee, B. H. Lee, H. C. Ko and M. G. Hahm, ACS Appl. Mater. Interfaces, 2015, 7, 16775–16780 CAS .
  166. L. Yu, Y.-H. Lee, X. Ling, E. J. G. Santos, Y. C. Shin, Y. Lin, M. Dubey, E. Kaxiras, J. Kong, H. Wang and T. Palacios, Nano Lett., 2014, 14, 3055–3063 CrossRef CAS PubMed .
  167. T. Roy, M. Tosun, J. S. Kang, A. B. Sachid, S. B. Desai, M. Hettick, C. C. Hu and A. Javey, ACS Nano, 2014, 8, 6259–6264 CrossRef CAS PubMed .
  168. H.-J. Chuang, X. Tan, N. J. Ghimire, M. M. Perera, B. Chamlagain, M. M.-C. Cheng, J. Yan, D. Mandrus, D. Tománek and Z. Zhou, Nano Lett., 2014, 14, 3594–3601 CrossRef CAS PubMed .
  169. X. Cui, G.-H. Lee, Y. D. Kim, G. Arefe, P. Y. Huang, C.-H. Lee, D. A. Chenet, X. Zhang, L. Wang, F. Ye, F. Pizzocchero, B. S. Jessen, K. Watanabe, T. Taniguchi, D. A. Muller, T. Low, P. Kim and J. Hone, Nat. Nanotechnol., 2015, 10, 534–540 CrossRef CAS PubMed .
  170. C.-J. Shih, Q. H. Wang, Y. Son, Z. Jin, D. Blankschtein and M. S. Strano, ACS Nano, 2014, 8, 5790–5798 CrossRef CAS PubMed .
  171. L. Britnell, R. V. Gorbachev, A. K. Geim, L. A. Ponomarenko, A. Mishchenko, M. T. Greenaway, T. M. Fromhold, K. S. Novoselov and L. Eaves, Nat. Commun., 2013, 4, 1794 CrossRef CAS PubMed .
  172. A. Mishchenko, J. S. Tu, Y. Cao, R. V. Gorbachev, J. R. Wallbank, M. T. Greenaway, V. E. Morozov, S. V. Morozov, M. J. Zhu, S. L. Wong, F. Withers, C. R. Woods, Y.-J. Kim, K. Watanabe, T. Taniguchi, E. E. Vdovin, O. Makarovsky, T. M. Fromhold, V. I. Fal'ko, A. K. Geim, L. Eaves and K. S. Novoselov, Nat. Nanotechnol., 2014, 9, 808–813 CrossRef CAS PubMed .
  173. S. Kang, B. Fallahazad, K. Lee, H. Movva, K. Kim, C. M. Corbet, T. Taniguchi, K. Watanabe, L. Colombo, L. F. Register, E. Tutuc and S. K. Banerjee, IEEE Electron Device Lett., 2015, 36, 405–407 CrossRef CAS .
  174. L. Brey, Phys. Rev. Appl., 2014, 2, 14003 CrossRef .
  175. H. Tian, Z. Tan, C. Wu, X. Wang, M. A. Mohammad, D. Xie, Y. Yang, J. Wang, L.-J. Li, J. Xu and T.-L. Ren, Sci. Rep., 2014, 4, 5951 CrossRef CAS PubMed .
  176. J. Shim, H. S. Kim, Y. S. Shim, D.-H. Kang, H.-Y. Park, J. Lee, J. Jeon, S. J. Jung, Y. J. Song, W.-S. Jung, J. Lee, S. Park, J. Kim, S. Lee, Y.-H. Kim and J.-H. Park, Adv. Mater., 2016, 28, 5293–5299 CrossRef CAS PubMed .
  177. Y. Sata, R. Moriya, S. Morikawa, N. Yabuki, S. Masubuchi and T. Machida, Appl. Phys. Lett., 2015, 107, 23109 CrossRef .
  178. Y. Choi, J. Kang, D. Jariwala, M. S. Kang, T. J. Marks, M. C. Hersam and J. H. Cho, Adv. Mater., 2016, 28, 3742–3748 CrossRef CAS PubMed .
  179. D. Akinwande, N. Petrone and J. Hone, Nat. Commun., 2014, 5, 5678 CrossRef CAS PubMed .
  180. H. Jang, Y. J. Park, X. Chen, T. Das, M.-S. Kim and J.-H. Ahn, Adv. Mater., 2016, 28, 4184–4202 CrossRef CAS PubMed .
  181. A. Avsar, J. Y. Tan, T. Taychatanapat, J. Balakrishnan, G. K. W. Koon, Y. Yeo, J. Lahiri, A. Carvalho, A. S. Rodin, E. C. T. O'Farrell, G. Eda, A. H. C. Neto and B. Özyilmaz, Nat. Commun., 2014, 5, 4875 CrossRef CAS PubMed .
  182. Z. Wang, D.-K. Ki, H. Chen, H. Berger, A. H. MacDonald and A. F. Morpurgo, Nat. Commun., 2015, 6, 8339 CrossRef CAS PubMed .
  183. Z. Wang, D.-K. Ki, J. Y. Khoo, D. Mauro, H. Berger, L. S. Levitov and A. F. Morpurgo, Phys. Rev. X, 2016, 6, 41020 Search PubMed .
  184. R. Lu, J. Liu, H. Luo, V. Chikan and J. Z. Wu, Sci. Rep., 2016, 6, 19161 CrossRef CAS PubMed .
  185. L. Zhang, Y. Yan, H.-C. Wu, D. Yu and Z.-M. Liao, ACS Nano, 2016, 10, 3816–3822 CrossRef CAS PubMed .
  186. J. W. Colson, A. R. Woll, A. Mukherjee, M. P. Levendorf, E. L. Spitler, V. B. Shields, M. G. Spencer, J. Park and W. R. Dichtel, Science, 2011, 332, 228–231 CrossRef CAS PubMed .
  187. L. Xu, X. Zhou, W. Q. Tian, T. Gao, Y. F. Zhang, S. Lei and Z. F. Liu, Angew. Chem., Int. Ed., 2014, 53, 9564–9568 CrossRef CAS PubMed .
  188. M. Woszczyna, A. Winter, M. Grothe, A. Willunat, S. Wundrack, R. Stosch, T. Weimann, F. Ahlers and A. Turchanin, Adv. Mater., 2014, 26, 4831–4837 CrossRef CAS PubMed .
  189. C. Ojeda-Aristizabal, W. Bao and M. S. Fuhrer, Phys. Rev. B: Condens. Matter Mater. Phys., 2013, 88, 35435 CrossRef .
  190. G. Oh, J.-S. Kim, J. H. Jeon, E. Won, J. W. Son, D. H. Lee, C. K. Kim, J. Jang, T. Lee and B. H. Park, ACS Nano, 2015, 9, 7515–7522 CrossRef CAS PubMed .
  191. B. H. Seo, J. Youn and M. Shim, ACS Nano, 2014, 8, 8831–8836 CrossRef CAS PubMed .
  192. Q. H. Wang and M. C. Hersam, Nat. Chem., 2009, 1, 206–211 CrossRef CAS PubMed .
  193. B. Li, A. V. Klekachev, M. Cantoro, C. Huyghebaert, A. Stesmans, I. Asselberghs, S. D. Gendt and S. D. Feyter, Nanoscale, 2013, 5, 9640–9644 RSC .
  194. Y. Ogawa, T. Niu, S. L. Wong, M. Tsuji, A. T. S. Wee, W. Chen and H. Ago, J. Phys. Chem. C, 2013, 117, 21849–21855 CAS .
  195. B. Wu, Y. Zhao, H. Nan, Z. Yang, Y. Zhang, H. Zhao, D. He, Z. Jiang, X. Liu, Y. Li, Y. Shi, Z. Ni, J. Wang, J.-B. Xu and X. Wang, Nano Lett., 2016, 16, 3754–3759 CrossRef CAS PubMed .
  196. H.-C. Cheng, G. Wang, D. Li, Q. He, A. Yin, Y. Liu, H. Wu, M. Ding, Y. Huang and X. Duan, Nano Lett., 2016, 16, 367–373 CrossRef CAS PubMed .
  197. L. Ci, L. Song, C. Jin, D. Jariwala, D. Wu, Y. Li, A. Srivastava, Z. F. Wang, K. Storr, L. Balicas, F. Liu and P. M. Ajayan, Nat. Mater., 2010, 9, 430–435 CrossRef CAS PubMed .
  198. J. Jung, Z. Qiao, Q. Niu and A. H. MacDonald, Nano Lett., 2012, 12, 2936–2940 CrossRef CAS PubMed .
  199. Q. Sun, Y. Dai, Y. Ma, W. Wei and B. Huang, RSC Adv., 2015, 5, 33037–33043 RSC .
  200. X. Ling, Y. Lin, Q. Ma, Z. Wang, Y. Song, L. Yu, S. Huang, W. Fang, X. Zhang, A. L. Hsu, Y. Bie, Y.-H. Lee, Y. Zhu, L. Wu, J. Li, P. Jarillo-Herrero, M. Dresselhaus, T. Palacios and J. Kong, Adv. Mater., 2016, 28, 2322–2329 CrossRef CAS PubMed .
  201. M. Zhao, Y. Ye, Y. Han, Y. Xia, H. Zhu, S. Wang, Y. Wang, D. A. Muller and X. Zhang, Nat. Nanotechnol., 2016, 11, 954–959 CrossRef CAS PubMed .
  202. M. H. D. Guimarães, H. Gao, Y. Han, K. Kang, S. Xie, C.-J. Kim, D. A. Muller, D. C. Ralph and J. Park, ACS Nano, 2016, 10, 6392–6399 CrossRef PubMed .
  203. X. Chen, Y. Park, T. Das, H. Jang, J.-B. Lee and J.-H. Ahn, Nanoscale, 2016, 8, 15181–15188 RSC .
  204. X. Lin, P. Liu, Y. Wei, Q. Li, J. Wang, Y. Wu, C. Feng, L. Zhang, S. Fan and K. Jiang, Nat. Commun., 2013, 4, 2920 Search PubMed .
  205. Z. Yan, Z. Peng, G. Casillas, J. Lin, C. Xiang, H. Zhou, Y. Yang, G. Ruan, A.-R. O. Raji, E. L. G. Samuel, R. H. Hauge, M. J. Yacaman and J. M. Tour, ACS Nano, 2014, 8, 5061–5068 CrossRef CAS PubMed .
  206. X. Li, J. Sha, S.-K. Lee, Y. Li, Y. Ji, Y. Zhao and J. M. Tour, ACS Nano, 2016, 10, 7307–7313 CrossRef CAS PubMed .
  207. Y. Li, Z. Peng, E. Larios, G. Wang, J. Lin, Z. Yan, F. Ruiz-Zepeda, M. José-Yacamán and J. M. Tour, ACS Nano, 2015, 9, 532–538 CrossRef CAS PubMed .
  208. Y. Liu, F. Wang, X. Wang, X. Wang, E. Flahaut, X. Liu, Y. Li, X. Wang, Y. Xu, Y. Shi and R. Zhang, Nat. Commun., 2015, 6, 8589 CrossRef CAS PubMed .
  209. S. H. Chae, W. J. Yu, J. J. Bae, D. L. Duong, D. Perello, H. Y. Jeong, Q. H. Ta, T. H. Ly, Q. A. Vu, M. Yun, X. Duan and Y. H. Lee, Nat. Mater., 2013, 12, 403–409 CrossRef CAS PubMed .
  210. G. Konstantatos, M. Badioli, L. Gaudreau, J. Osmond, M. Bernechea, F. P. G. de Arquer, F. Gatti and F. H. L. Koppens, Nat. Nanotechnol., 2012, 7, 363–368 CrossRef CAS PubMed .
  211. I. Nikitskiy, S. Goossens, D. Kufer, T. Lasanta, G. Navickaite, F. H. L. Koppens and G. Konstantatos, Nat. Commun., 2016, 7, 11954 CrossRef CAS PubMed .
  212. Z. Sun, Z. Liu, J. Li, G. Tai, S.-P. Lau and F. Yan, Adv. Mater., 2012, 24, 5878–5883 CrossRef CAS PubMed .
  213. L. A. Chernozatonskii, A. G. Kvashnin and P. B. Sorokin, Nanotechnology, 2016, 27, 365201 CrossRef PubMed .
  214. Z. Zheng, X. Zhang, C. Neumann, D. Emmrich, A. Winter, H. Vieker, W. Liu, M. Lensen, A. Gölzhäuser and A. Turchanin, Nanoscale, 2015, 7, 13393–13397 RSC .
  215. J. Wan, S. D. Lacey, J. Dai, W. Bao, M. S. Fuhrer and L. Hu, Chem. Soc. Rev., 2016, 45, 6742–6765 RSC .
  216. K. S. Vasu, E. Prestat, J. Abraham, J. Dix, R. J. Kashtiban, J. Beheshtian, J. Sloan, P. Carbone, M. Neek-Amal, S. J. Haigh, A. K. Geim and R. R. Nair, Nat. Commun., 2016, 7, 12168 CrossRef CAS PubMed .
  217. N. Kim, K. S. Kim, N. Jung, L. Brus and P. Kim, Nano Lett., 2011, 11, 860–865 CrossRef CAS PubMed .
  218. T. H. Bointon, I. Khrapach, R. Yakimova, A. V. Shytov, M. F. Craciun and S. Russo, Nano Lett., 2014, 14, 1751–1755 CrossRef CAS PubMed .
  219. K. Kanetani, K. Sugawara, T. Sato, R. Shimizu, K. Iwaya, T. Hitosugi and T. Takahashi, Proc. Natl. Acad. Sci. U. S. A., 2012, 109, 19610–19613 CrossRef CAS PubMed .
  220. S. Ichinokura, K. Sugawara, A. Takayama, T. Takahashi and S. Hasegawa, ACS Nano, 2016, 10, 2761–2765 CrossRef CAS PubMed .
  221. D. Zhan, L. Sun, Z. H. Ni, L. Liu, X. F. Fan, Y. Wang, T. Yu, Y. M. Lam, W. Huang and Z. X. Shen, Adv. Funct. Mater., 2010, 20, 3504–3509 CrossRef CAS .
  222. W. Zhao, P. H. Tan, J. Liu and A. C. Ferrari, J. Am. Chem. Soc., 2011, 133, 5941–5946 CrossRef CAS PubMed .
  223. I. Khrapach, F. Withers, T. H. Bointon, D. K. Polyushkin, W. L. Barnes, S. Russo and M. F. Craciun, Adv. Mater., 2012, 24, 2844–2849 CrossRef CAS PubMed .
  224. D. J. Wehenkel, T. H. Bointon, T. Booth, P. Bøggild, M. F. Craciun and S. Russo, Sci. Rep., 2015, 5, 7609 CrossRef CAS PubMed .
  225. D. Zhan, J. X. Yan, Z. H. Ni, L. Sun, L. F. Lai, L. Liu, X. Y. Liu and Z. X. Shen, Small, 2015, 11, 1177–1182 CrossRef CAS PubMed .
  226. W. Bao, J. Wan, X. Han, X. Cai, H. Zhu, D. Kim, D. Ma, Y. Xu, J. N. Munday, H. D. Drew, M. S. Fuhrer and L. Hu, Nat. Commun., 2014, 5, 4224 CAS .
  227. R. J. Smith, P. J. King, M. Lotya, C. Wirtz, U. Khan, S. De, A. O'Neill, G. S. Duesberg, J. C. Grunlan, G. Moriarty, J. Chen, J. Wang, A. I. Minett, V. Nicolosi and J. N. Coleman, Adv. Mater., 2011, 23, 3944–3948 CrossRef CAS PubMed .
  228. J. Zheng, H. Zhang, S. Dong, Y. Liu, C. Tai Nai, H. Suk Shin, H. Young Jeong, B. Liu and K. Ping Loh, Nat. Commun., 2014, 5, 2995 Search PubMed .
  229. Y. Jung, Y. Zhou and J. J. Cha, Inorg. Chem. Front., 2016, 3, 452–463 RSC .
  230. Z. Y. Xia, S. Pezzini, E. Treossi, G. Giambastiani, F. Corticelli, V. Morandi, A. Zanelli, V. Bellani and V. Palermo, Adv. Funct. Mater., 2013, 23, 4684–4693 CrossRef CAS .
  231. A. M. Abdelkader, A. J. Cooper, R. A. W. Dryfe and I. A. Kinloch, Nanoscale, 2015, 7, 6944–6956 RSC .
  232. A. Ambrosi and M. Pumera, Chem. – Eur. J., 2016, 22, 153–159 CrossRef CAS PubMed .
  233. C. Backes, T. M. Higgins, A. Kelly, C. Boland, A. Harvey, D. Hanlon and J. N. Coleman, Chem. Mater., 2017, 29, 243–255 CrossRef CAS .
  234. D. R. Dreyer, S. Park, C. W. Bielawski and R. S. Ruoff, Chem. Soc. Rev., 2010, 39, 228–240 RSC .
  235. Y. Hernandez, V. Nicolosi, M. Lotya, F. M. Blighe, Z. Sun, S. De, I. T. McGovern, B. Holland, M. Byrne, Y. K. Gun'Ko, J. J. Boland, P. Niraj, G. Duesberg, S. Krishnamurthy, R. Goodhue, J. Hutchison, V. Scardaci, A. C. Ferrari and J. N. Coleman, Nat. Nanotechnol., 2008, 3, 563–568 CrossRef CAS PubMed .
  236. J. N. Coleman, M. Lotya, A. O'Neill, S. D. Bergin, P. J. King, U. Khan, K. Young, A. Gaucher, S. De, R. J. Smith, I. V. Shvets, S. K. Arora, G. Stanton, H.-Y. Kim, K. Lee, G. T. Kim, G. S. Duesberg, T. Hallam, J. J. Boland, J. J. Wang, J. F. Donegan, J. C. Grunlan, G. Moriarty, A. Shmeliov, R. J. Nicholls, J. M. Perkins, E. M. Grieveson, K. Theuwissen, D. W. McComb, P. D. Nellist and V. Nicolosi, Science, 2011, 331, 568–571 CrossRef CAS PubMed .
  237. L. Niu, J. N. Coleman, H. Zhang, H. Shin, M. Chhowalla and Z. Zheng, Small, 2016, 12, 272–293 CrossRef CAS PubMed .
  238. Q. H. Wang, K. Kalantar-Zadeh, A. Kis, J. N. Coleman and M. S. Strano, Nat. Nanotechnol., 2012, 7, 699–712 CrossRef CAS PubMed .
  239. M. Chhowalla, H. S. Shin, G. Eda, L.-J. Li, K. P. Loh and H. Zhang, Nat. Chem., 2013, 5, 263–275 CrossRef PubMed .
  240. C. Tan and H. Zhang, Chem. Soc. Rev., 2015, 44, 2713–2731 RSC .
  241. Y. Hu, X. Li, A. Lushington, M. Cai, D. Geng, M. N. Banis, R. Li and X. Sun, ECS J. Solid State Sci. Technol., 2013, 2, M3034–M3039 CrossRef CAS .
  242. K.-J. Huang, L. Wang, Y.-J. Liu, Y.-M. Liu, H.-B. Wang, T. Gan and L.-L. Wang, Int. J. Hydrogen Energy, 2013, 38, 14027–14034 CrossRef CAS .
  243. L. Peng, X. Peng, B. Liu, C. Wu, Y. Xie and G. Yu, Nano Lett., 2013, 13, 2151–2157 CrossRef CAS PubMed .
  244. L. David, R. Bhandavat and G. Singh, ACS Nano, 2014, 8, 1759–1770 CrossRef CAS PubMed .
  245. M. A. Bissett, I. A. Kinloch and R. A. W. Dryfe, ACS Appl. Mater. Interfaces, 2015, 7, 17388–17398 CAS .
  246. X. Feng, Z. Yan, N. Chen, Y. Zhang, Y. Ma, X. Liu, Q. Fan, L. Wang and W. Huang, J. Mater. Chem. A, 2013, 1, 12818–12825 CAS .
  247. G. Huang, T. Chen, W. Chen, Z. Wang, K. Chang, L. Ma, F. Huang, D. Chen and J. Y. Lee, Small, 2013, 9, 3693–3703 CrossRef CAS PubMed .
  248. S. Ratha and C. S. Rout, ACS Appl. Mater. Interfaces, 2013, 5, 11427–11433 CAS .
  249. A. Ramadoss, T. Kim, G.-S. Kim and S. J. Kim, New J. Chem., 2014, 38, 2379–2385 RSC .
  250. L. Jiang, B. Lin, X. Li, X. Song, H. Xia, L. Li and H. Zeng, ACS Appl. Mater. Interfaces, 2016, 8, 2680–2687 CAS .
  251. D. R. Rolison, J. W. Long, J. C. Lytle, A. E. Fischer, C. P. Rhodes, T. M. McEvoy, M. E. Bourg and A. M. Lubers, Chem. Soc. Rev., 2009, 38, 226–252 RSC .
  252. M. Pumera, Energy Environ. Sci., 2011, 4, 668–674 CAS .
  253. D. Chen, L. Tang and J. Li, Chem. Soc. Rev., 2010, 39, 3157–3180 RSC .
  254. R. Raccichini, A. Varzi, S. Passerini and B. Scrosati, Nat. Mater., 2015, 14, 271–279 CrossRef CAS PubMed .
  255. C. N. R. Rao, K. Gopalakrishnan and U. Maitra, ACS Appl. Mater. Interfaces, 2015, 7, 7809–7832 CAS .
  256. M. A. Bissett, S. D. Worrall, I. A. Kinloch and R. A. W. Dryfe, Electrochim. Acta, 2016, 201, 30–37 CrossRef CAS .
  257. M. Velický, M. A. Bissett, P. S. Toth, H. V. Patten, S. D. Worrall, A. N. J. Rodgers, E. W. Hill, I. A. Kinloch, K. S. Novoselov, T. Georgiou, L. Britnell and R. A. W. Dryfe, Phys. Chem. Chem. Phys., 2015, 17, 17844–17853 RSC .
  258. M. Velický, M. A. Bissett, C. R. Woods, P. S. Toth, T. Georgiou, I. A. Kinloch, K. S. Novoselov and R. A. W. Dryfe, Nano Lett., 2016, 16, 2023–2032 CrossRef PubMed .
  259. B. Zhang, X. Ji, K. Xu, C. Chen, X. Xiong, J. Xiong, Y. Yao, L. Miao and J. Jiang, Electrochim. Acta, 2016, 217, 1–8 CrossRef CAS .
  260. S. Chen, J. Zhu, X. Wu, Q. Han and X. Wang, ACS Nano, 2010, 4, 2822–2830 CrossRef CAS PubMed .
  261. Z. Fan, J. Yan, T. Wei, L. Zhi, G. Ning, T. Li and F. Wei, Adv. Funct. Mater., 2011, 21, 2366–2375 CrossRef CAS .
  262. M. Zhi, C. Xiang, J. Li, M. Li and N. Wu, Nanoscale, 2013, 5, 72–88 RSC .
  263. K. Chang and W. Chen, ACS Nano, 2011, 5, 4720–4728 CrossRef CAS PubMed .
  264. Z. Wang, T. Chen, W. Chen, K. Chang, L. Ma, G. Huang, D. Chen and J. Y. Lee, J. Mater. Chem. A, 2013, 1, 2202–2210 CAS .
  265. E. G. da Silveira Firmiano, A. C. Rabelo, C. J. Dalmaschio, A. N. Pinheiro, E. C. Pereira, W. H. Schreiner and E. R. Leite, Adv. Energy Mater., 2014, 4, 1301380 CrossRef .
  266. H. Li, K. Yu, H. Fu, B. Guo, X. Lei and Z. Zhu, J. Phys. Chem. C, 2015, 119, 7959–7968 CAS .
  267. X. Xie, Z. Ao, D. Su, J. Zhang and G. Wang, Adv. Funct. Mater., 2015, 25, 1393–1403 CrossRef CAS .
  268. A. Gigot, M. Fontana, M. Serrapede, M. Castellino, S. Bianco, M. Armandi, B. Bonelli, C. F. Pirri, E. Tresso and P. Rivolo, ACS Appl. Mater. Interfaces, 2016, 8, 32842–32852 CAS .
  269. W. S. V. Lee, E. Peng, T. A. J. Loh, X. Huang and J. M. Xue, Nanoscale, 2016, 8, 8042–8047 RSC .
  270. T. Sun, Z. Li, X. Liu, L. Ma, J. Wang and S. Yang, J. Power Sources, 2016, 331, 180–188 CrossRef CAS .
  271. Y. Teng, H. Zhao, Z. Zhang, Z. Li, Q. Xia, Y. Zhang, L. Zhao, X. Du, Z. Du, P. Lv and K. Świerczek, ACS Nano, 2016, 10, 8526–8535 CrossRef CAS PubMed .
  272. R. Thangappan, S. Kalaiselvam, A. Elayaperumal, R. Jayavel, M. Arivanandhan, R. Karthikeyan and Y. Hayakawa, Dalton Trans., 2016, 45, 2637–2646 RSC .
  273. N. Savjani, E. A. Lewis, M. A. Bissett, J. R. Brent, R. A. W. Dryfe, S. J. Haigh and P. O'Brien, Chem. Mater., 2016, 28, 657–664 CrossRef CAS .
  274. P. Simon and Y. Gogotsi, Nat. Mater., 2008, 7, 845–854 CrossRef CAS PubMed .
  275. J. Chen, C. Li and G. Shi, J. Phys. Chem. Lett., 2013, 4, 1244–1253 CrossRef CAS PubMed .
  276. G. Xiong, C. Meng, R. G. Reifenberger, P. P. Irazoqui and T. S. Fisher, Electroanalysis, 2014, 26, 30–51 CrossRef CAS .
  277. M. D. Stoller and R. S. Ruoff, Energy Environ. Sci., 2010, 3, 1294–1301 CAS .
  278. Q. Cheng, J. Tang, J. Ma, H. Zhang, N. Shinya and L.-C. Qin, Carbon, 2011, 49, 2917–2925 CrossRef CAS .
  279. M. Beidaghi and C. Wang, Adv. Funct. Mater., 2012, 22, 4501–4510 CrossRef CAS .
  280. M. A. Worsley, S. J. Shin, M. D. Merrill, J. Lenhardt, A. J. Nelson, L. Y. Woo, A. E. Gash, T. F. Baumann and C. A. Orme, ACS Nano, 2015, 9, 4698–4705 CrossRef CAS PubMed .
  281. Y. Lin, F. Liu, G. Casano, R. Bhavsar, I. A. Kinloch and B. Derby, Adv. Mater., 2016, 28, 7993–8000 CrossRef CAS PubMed .
  282. J. Biener, M. Stadermann, M. Suss, M. A. Worsley, M. M. Biener, K. A. Rose and T. F. Baumann, Energy Environ. Sci., 2011, 4, 656–667 CAS .
  283. M. A. Py and R. R. Haering, Can. J. Phys., 1983, 61, 76–84 CrossRef CAS .
  284. G. Eda, H. Yamaguchi, D. Voiry, T. Fujita, M. Chen and M. Chhowalla, Nano Lett., 2011, 11, 5111–5116 CrossRef CAS PubMed .
  285. M. Acerce, D. Voiry and M. Chhowalla, Nat. Nanotechnol., 2015, 10, 313–318 CrossRef CAS PubMed .
  286. Q. Mahmood, S. K. Park, K. D. Kwon, S.-J. Chang, J.-Y. Hong, G. Shen, Y. M. Jung, T. J. Park, S. W. Khang, W. S. Kim, J. Kong and H. S. Park, Adv. Energy Mater., 2016, 6, 1501115 CrossRef .
  287. Y.-C. Lin, D. O. Dumcenco, Y.-S. Huang and K. Suenaga, Nat. Nanotechnol., 2014, 9, 391–396 CrossRef CAS PubMed .
  288. J. M. Tarascon and M. Armand, Nature, 2001, 414, 359–367 CrossRef CAS PubMed .
  289. G. Wang, X. Shen, J. Yao and J. Park, Carbon, 2009, 47, 2049–2053 CrossRef CAS .
  290. G. Kucinskis, G. Bajars and J. Kleperis, J. Power Sources, 2013, 240, 66–79 CrossRef CAS .
  291. C. Feng, J. Ma, H. Li, R. Zeng, Z. Guo and H. Liu, Mater. Res. Bull., 2009, 44, 1811–1815 CrossRef CAS .
  292. X. Rui, H. Tan and Q. Yan, Nanoscale, 2014, 6, 9889–9924 RSC .
  293. X. Hu, W. Zhang, X. Liu, Y. Mei and Y. Huang, Chem. Soc. Rev., 2015, 44, 2376–2404 RSC .
  294. J. Cabana, L. Monconduit, D. Larcher and M. R. Palacín, Adv. Mater., 2010, 22, E170–E192 CrossRef CAS PubMed .
  295. T. Stephenson, Z. Li, B. Olsen and D. Mitlin, Energy Environ. Sci., 2014, 7, 209–231 CAS .
  296. Z. Wang, L. Ma, W. Chen, G. Huang, D. Chen, L. Wang and J. Y. Lee, RSC Adv., 2013, 3, 21675–21684 RSC .
  297. Y.-X. Wang, S.-L. Chou, D. Wexler, H.-K. Liu and S.-X. Dou, Chem. – Eur. J., 2014, 20, 9607–9612 CrossRef CAS PubMed .
  298. G. Cunningham, M. Lotya, N. McEvoy, G. S. Duesberg, P. van der Schoot and J. N. Coleman, Nanoscale, 2012, 4, 6260–6264 RSC .
  299. X. Wang, X. Lu, B. Liu, D. Chen, Y. Tong and G. Shen, Adv. Mater., 2014, 26, 4763–4782 CrossRef CAS PubMed .
  300. X. Wang, X. Shen, Z. Wang, R. Yu and L. Chen, ACS Nano, 2014, 8, 11394–11400 CrossRef CAS PubMed .
  301. C. C. Mayorga-Martinez, A. Ambrosi, A. Y. S. Eng, Z. Sofer and M. Pumera, Electrochem. Commun., 2015, 56, 24–28 CrossRef CAS .
  302. G. A. Muller, J. B. Cook, H.-S. Kim, S. H. Tolbert and B. Dunn, Nano Lett., 2015, 15, 1911–1917 CrossRef CAS PubMed .
  303. S. K. Balasingam, J. S. Lee and Y. Jun, Dalton Trans., 2016, 45, 9646–9653 RSC .
  304. D. H. Youn, C. Jo, J. Y. Kim, J. Lee and J. S. Lee, J. Power Sources, 2015, 295, 228–234 CrossRef CAS .
  305. Z. Luo, J. Zhou, L. Wang, G. Fang, A. Pan and S. Liang, J. Mater. Chem. A, 2016, 4, 15302–15308 CAS .
  306. J. R. Brent, N. Savjani, E. A. Lewis, S. J. Haigh, D. J. Lewis and P. O'Brien, Chem. Commun., 2014, 50, 13338–13341 RSC .
  307. J. Kang, J. D. Wood, S. A. Wells, J.-H. Lee, X. Liu, K.-S. Chen and M. C. Hersam, ACS Nano, 2015, 9, 3596–3604 CrossRef CAS PubMed .
  308. E. A. Lewis, J. R. Brent, B. Derby, S. J. Haigh and D. J. Lewis, Chem. Commun., 2017, 53, 1445–1458 RSC .
  309. D. Hanlon, C. Backes, E. Doherty, C. S. Cucinotta, N. C. Berner, C. Boland, K. Lee, A. Harvey, P. Lynch, Z. Gholamvand, S. Zhang, K. Wang, G. Moynihan, A. Pokle, Q. M. Ramasse, N. McEvoy, W. J. Blau, J. Wang, G. Abellan, F. Hauke, A. Hirsch, S. Sanvito, D. D. O'Regan, G. S. Duesberg, V. Nicolosi and J. N. Coleman, Nat. Commun., 2015, 6, 8563 CrossRef CAS PubMed .
  310. Y. Zhang, H. Wang, Z. Luo, H. T. Tan, B. Li, S. Sun, Z. Li, Y. Zong, Z. J. Xu, Y. Yang, K. A. Khor and Q. Yan, Adv. Energy Mater., 2016, 6, 1600453 CrossRef .
  311. J. Sun, H.-W. Lee, M. Pasta, H. Yuan, G. Zheng, Y. Sun, Y. Li and Y. Cui, Nat. Nanotechnol., 2015, 10, 980–985 CrossRef CAS PubMed .
  312. D. Voiry, J. Yang and M. Chhowalla, Adv. Mater., 2016, 28, 6197–6206 CrossRef CAS PubMed .
  313. T. F. Jaramillo, K. P. Jørgensen, J. Bonde, J. H. Nielsen, S. Horch and I. Chorkendorff, Science, 2007, 317, 100–102 CrossRef CAS PubMed .
  314. H. Wang, Q. Zhang, H. Yao, Z. Liang, H.-W. Lee, P.-C. Hsu, G. Zheng and Y. Cui, Nano Lett., 2014, 14, 7138–7144 CrossRef CAS PubMed .
  315. Y.-R. An, X.-L. Fan, Z.-F. Luo and W.-M. Lau, Nano Lett., 2017, 17, 368–376 CrossRef CAS PubMed .
  316. Y. Li, H. Wang, L. Xie, Y. Liang, G. Hong and H. Dai, J. Am. Chem. Soc., 2011, 133, 7296–7299 CrossRef CAS PubMed .
  317. P. S. Toth, M. Velický, M. A. Bissett, T. J. A. Slater, N. Savjani, A. K. Rabiu, A. M. Rakowski, J. R. Brent, S. J. Haigh, P. O'Brien and R. A. W. Dryfe, Adv. Mater., 2016, 28, 8256–8264 CrossRef CAS PubMed .
  318. F. Reymond, D. Fermín, H. J. Lee and H. H. Girault, Electrochim. Acta, 2000, 45, 2647–2662 CrossRef CAS .
  319. F. Reymond and H. H. Girault, Encyclopedia of Analytical Chemistry, John Wiley & Sons, Ltd, 2006 Search PubMed .
  320. S. G. Booth and R. A. W. Dryfe, J. Phys. Chem. C, 2015, 119, 23295–23309 CAS .
  321. M. A. Bissett, Y. Takesaki, M. Tsuji and H. Ago, RSC Adv., 2014, 4, 52215–52219 RSC .

This journal is © The Royal Society of Chemistry 2017