Chem Soc Rev

View Article Online

REVIEW ARTICLE View Journal | View Issue

An atlas of two-dimensional materials†

Published on 14 May 2014. Downloaded by Federal University of Minas Gerais on 07/12/2017 17:17:01.

Cite this: Chem. Soc. Rev., 2014,

Pere Miró,* Martha Audiffred and Thomas Heine*
43, 6537
The discovery of graphene and other two-dimensional (2D) materials together with recent advances in
exfoliation techniques have set the foundations for the manufacturing of single layered sheets from any
layered 3D material. The family of 2D materials encompasses a wide selection of compositions including
almost all the elements of the periodic table. This derives into a rich variety of electronic properties including
metals, semimetals, insulators and semiconductors with direct and indirect band gaps ranging from
ultraviolet to infrared throughout the visible range. Thus, they have the potential to play a fundamental role
in the future of nanoelectronics, optoelectronics and the assembly of novel ultrathin and flexible devices. We
categorize the 2D materials according to their structure, composition and electronic properties. In this
review we distinguish atomically thin materials (graphene, silicene, germanene, and their saturated forms;
hexagonal boron nitride; silicon carbide), rare earth, semimetals, transition metal chalcogenides and halides,
Received 5th March 2014 and finally synthetic organic 2D materials, exemplified by 2D covalent organic frameworks. Our exhaustive
DOI: 10.1039/c4cs00102h data collection presented in this Atlas demonstrates the large diversity of electronic properties, including
band gaps and electron mobilities. The key points of modern computational approaches applied to 2D
www.rsc.org/csr materials are presented with special emphasis to cover their range of application, peculiarities and pitfalls.

1. Introduction the physics community. According to classical physics, 2D

materials are thermodynamically instable at any finite temperature
During the first decades of the 20th century, the existence of due to thermal lattice fluctuations.1–3 This was in agreement with
two-dimensional (2D) materials was a highly debated issue in the decreasing melting temperature observed when the thickness
of thin films was decreased. The development of modern spectro-
Jacobs University Bremen, Center of Functional Nanomaterials (NanoFun), scopy revealed the existence of 3D materials with layered structure,
Campus Ring 1, 28759 Bremen, Germany. E-mail: [email protected], such as graphite or molybdenum disulphide. Even though their
[email protected]
exfoliation down to their two-dimensional monolayers was
† Electronic supplementary information (ESI) available: Cartesian coordinates,
electronic band structures, band gaps and effective masses of electrons and holes. thought to be possible only in the theoretical domain since
See DOI: 10.1039/c4cs00102h Mermin demonstrated that strictly one- and two-dimensional

Pere Miró graduated from Uni- Martha Audiffred completed her

versitat Rovira i Virgili in 2005 PhD under the supervision of
with a BS in Chemistry. He Gabriel Merino in Universidad
completed his PhD under the de Guanajuato in 2013. She is
supervision of Prof. Carles Bo at currently working as a post-
the Institute of Chemical Research doctoral fellow in Prof. Thomas
of Catalonia (ICIQ) in 2010. After a Heine’s group. Her research
postdoc at University of Minnesota interests include synthesis and
under the supervision of Prof. theoretical characterization of
Christopher J. Cramer, he joined low dimensional nanomaterials.
Prof. Thomas Heine’s group to
study 2D layered materials. His
Pere Miró research interests include low Martha Audiffred
dimensionality and porous nano-
materials for nanoelectronics and
catalytical applications.

This journal is © The Royal Society of Chemistry 2014 Chem. Soc. Rev., 2014, 43, 6537--6554 | 6537
 View Article Online

Review Article Chem Soc Rev

Published on 14 May 2014. Downloaded by Federal University of Minas Gerais on 07/12/2017 17:17:01.

Fig. 1 Graphene, the prototype 2D material. (a) Graphite structure (5  5  2 unit cells). (b) Graphene structure (5  5). (c) Dirac cones in graphene
(adopted from ref. 6). (d) Graphene band structure (G–M–K–G). Fermi level has been shifted to 0 eV and depicted with a blue horizontal line.

materials can only exist hypothetically – i.e. if the crystal is within the Brillouin zone) is linear at these points, a unique
described within the harmonic approximation.4 feature that leads to zero effective mass for electrons and holes,
Materials science had a major scientific breakthrough in and thus to very high currents. Due to this linear or conical
2004, when Novoselov and Geim isolated the first single layer dispersion relation, electrons and holes near these six points
2D material, graphene, through the Scotch tape exfoliation of behave like relativistic particles described by the Dirac equa-
graphite (Fig. 1a).5 The importance of this achievement was tion. Therefore, the electrons and holes are called Dirac
sealed in 2010, when the Nobel price was awarded to both fermions and the six corners of the Brillouin zone are called
researchers. the Dirac points. The cones at Dirac points are located at the
Graphene is an atomically thin sp2 carbon layered material high-symmetry K points in the Brillouin zone of the honey-
with a honeycomb lattice that has almost the same crystal comb lattice and have been recently explored towards valley-
energy as diamond. As each graphene carbon has only three tronics applications (Fig. 1c and d). The massless Dirac
bonds instead of four for diamond, the graphene C–C bonds electrons at the Fermi level derive also in the highest known
are about 25% stronger. Thus, it is the most stable material finite-temperature electron mobility. Furthermore, the extre-
known to date. Graphene has a peculiar electronic structure. mely low spin–orbit (SO) interaction in graphene makes it an
It is a semi-metal, characterized by the fact that it does not ideal spin carrier for spintronics.6 Research on electronic
show a band gap, but the density-of-states is zero at the Fermi structure engineering of graphene rapidly emerged to the
level. The Fermi level is crossed by electronic bands near the six development of semiconducting modifications exhibiting a
corners of the two-dimensional hexagonal Brillouin zone (Fig. 1c). (very) low band gap, an important step towards electronic and
The dispersion relation (the change of electron energy of a band optoelectronic applications. However, these devices require
larger band gap semiconductors and dielectric materials as
insulators. Luckily, nature offers a manifold of layered materials
beyond graphene, including metals, semimetals, insulators and
Thomas Heine graduated from TU semiconductors with direct and indirect band gaps ranging from
Dresden and, after postdoctoral ultraviolet to infrared through the visible range.
stages in Bologna, Exeter and The recent progress in exfoliation techniques such as micro-
Dresden became Associated (2008) mechanical cleavage, ion intercalation, and surfactant-assisted
and Full (2011) Professor of ultrasonication has set the foundations for the manufacturing
Theoretical Physics/Theoretical of essentially any given layered bulk material in the monolayer
Materials Science at Jacobs limit.7,8 Layered materials cover an extremely large range of
University Bremen. His research compounds, including clays, layered oxides, chalcogenides,
interests include the development halides, carbides, nitrides, hydrides, hydroxides, phosphates
of methods and software for and phosphonates. Most of these materials are binary layered
materials science, molecular frame- compounds although ternary layered compounds are also
work compounds, 2D inorganic possible (i.e. CuSbS2).9 Layered materials can be further classi-
Thomas Heine materials and theoretical fied based on the nature of the interaction between the layers.
spectroscopy. This interaction is governed by hydrogen bonds, interstitial

6538 | Chem. Soc. Rev., 2014, 43, 6537--6554 This journal is © The Royal Society of Chemistry 2014
 View Article Online

Chem Soc Rev Review Article

cations or London dispersion interactions between the layers. band structures, band gaps and effective masses, are presented
Among the latter, transition metal chalcogenides (TMC) are the in ESI.†
most prominent examples of these materials with MoS2 as its
flagship.10–13 Indeed, the development of MoS2 based transistors
has triggered an immense scientific interest in TMC monolayers 2. Methodology
for novel ultrathin and flexible devices as well as for other
nanoelectronic and optoelectronic applications.14,15 A 2D material can be described as a hybrid between a solid
Besides the immense success of 2D materials science one (in the periodic plane) and a molecule (perpendicular to it).
Published on 14 May 2014. Downloaded by Federal University of Minas Gerais on 07/12/2017 17:17:01.

should not forget the underlying physics. According to all Accordingly, and as in any quantum mechanical system, the
theoretical predictions, 2D materials should not exist above choice of the boundary conditions is crucial to properly
absolute zero.1–4 The study of free standing and supported describe the system properties. Explicit consideration of 2D
graphene revealed the formation of intrinsic ripples. These periodic boundary conditions is possible if local basis functions
findings were also corroborated through theoretical studies on are used. Those are implemented, for example, in CRYSTAL,26
free-standing graphene sheets.16,17 Thus, on the large length scale, SIESTA27,28 and ADF/BAND software.29–31
ripples distort the 2D lattice, making it – strictly speaking – a 3D Modeling a 2D system as finite cluster is not recommended
material and thus resolving the controversy about the existence of since it is experimentally well-known that lateral quantum
these materials beyond absolute zero. Analogously, spontaneous confinement effects arise for particle sizes below 100 nm.132
ripple formation has been observed in free-standing single layer On the contrary, calculations using solid state codes employing
MoS2 sheets, even though the material is not atomically thin and the repeated slab method have become a common approach, as
thus not strictly two-dimensional.18,19 they allow the employment of plane-wave basis functions that
In this review we will discuss the fundamental electronic are known to be computationally very efficient and are hence
properties of single-layer 2D materials with special emphasis widely used in physics. Popular codes include Quantum-
on their range of application, on their peculiarities and on Espresso,32 abinit,33–35 and VASP.36–38 Plane wave basis func-
pitfalls in the theoretical description of their electronic structure. tions span the full 3D simulation box and hence require a
The computational treatment of 2D materials requires special care vacuum layer such that there is no spurious self-interaction
when choosing the boundary conditions since these systems are between the studied system and its periodic images in the
infinite in two dimensions, but finite in the third. Furthermore, direction perpendicular to the 2D lattice plane. This is funda-
the exfoliation down to the monolayer has implications in the final mental since even though the interlayer interaction is weak, it
properties of these materials. The most intriguing paradigm has significant impact on the electronic structure. For example,
change is certainly the appearance of the massless Dirac Fermions band gaps may double from bulk to monolayer, as known for
in graphene.20 Another one is the appearance of the giant spin many transition metal dichalcogenides,39 or the character of
orbit splitting in MoS2.21,22,99 Additionally, due to their 2D the material may change from semiconducting to metallic, as
character, the determination of band structures requires special in graphene40 or palladium sulphide.41 Thus, when using the
attention; as due to a quenched screening of the Coulomb inter- repeated slab method, the convergence of the results must be
action a strong exciton binding energy is present. This has a strong checked with respect to the extension of the vacuum layer, as
impact on quasi-particle approaches such as GW theory.23,24 indicated in Table 1. It is worth to note that plane-wave
In the light of the scarce and not systematic data on the methods scale with the size of the simulation box, therefore,
electronic properties of a representative fraction of single-layer as commonly known from solid state physics, their efficiency is
2D materials in the literature (Fig. 2), we have computed this reduced in case of the calculation of single layered materials.
data at a consistent level of theory (density-functional theory As result, 2D materials are faster processed (per atom and
and approximate quasi-particle theory). These results are pre- compared to their bulk counterparts) if using explicit 2D
sented in this article and compared with previous experimental boundary conditions, due to the reduced complexity of inter-
and theoretical results whenever available. The studied materials actions per atom, while they slow down calculations using 3D
include atomically thin 2D materials, starting with graphene and boundary conditions and plane waves (Table 1).
direct analogues such as hexagonal boron nitride (h-BN), silicon Almost all elements of the periodic table have been incor-
carbide (SiC), silicene and germanene as well as their hydro- porated as main components in 2D materials or as dopants.
genated and halogenated derivatives. Further, we present exfo- Thus, high quality basis sets for a large number of elements are
liated monolayer structures of transition metal chalcogenides required. It is popular to reduce computational complexity and
(TMCs), semimetal chalcogenides (SMCs), and transition metal cost by replacing core electrons by pseudo potentials (PP) or
halides (TMHs) including M 0 X, M2N, MX2, MY2, MY3 (M 0 = Ga effective core potentials (ECPs). It should be reminded that a
and In; M = Ti, Zr, Hf, Nb, Mo, W, Re, Pd, Pt, etc.; X = S, Se and careful validation of those numerical treatments is required, as
Te; Y = Cl, Br and I). Finally, we address an emerging family of they can have a significant effect on computed quantities such as
metal-free layered organic materials with honeycomb lattice, in band gaps and band structures. Furthermore, relativistic effects
particular the recently reported 2D covalent organic frameworks have found to be important even for relatively light-weight
(COFs).25 Only selected examples are presented in the text, elements in several 2D materials. The inclusion of relativistic
but a complete compilation of results, including geometries, effects for core electrons is achieved either by including them

This journal is © The Royal Society of Chemistry 2014 Chem. Soc. Rev., 2014, 43, 6537--6554 | 6539
 View Article Online

Review Article Chem Soc Rev

Published on 14 May 2014. Downloaded by Federal University of Minas Gerais on 07/12/2017 17:17:01.

Fig. 2 Structure of atomically thin 2D materials: graphene, silicene, germanene (and their derivatives), silicon carbide (SiC), hexagonal boron nitride
(h-BN) and transition metal chalcogenides (TMC) (top). Ultrathin 2D materials such as transition metal chalcogenides (TMC) (centre) and halides (TMH)
(bottom). Unit cells are depicted in red. Colour code: metals in ice blue, halides in green, chalcogenides in yellow, nitrogen in blue, carbon in grey, silicon
and germanium in gold, boron in pink and hydrogen in white.

6540 | Chem. Soc. Rev., 2014, 43, 6537--6554 This journal is © The Royal Society of Chemistry 2014
 View Article Online

Chem Soc Rev Review Article

Table 1 Relative CPU time and electronic band gaps for graphene and MoS2 calculated using explicit 2D periodic boundary conditions (PBC) and the
repeated slab method (3D PBC). For the latter, the interlayer distance has been increased stepwise (0.4 Å) starting from the bulk value. The table shows
the relative computer time with respect to the bulk system (same software, same computer). 2D PBC calculations have been carried out using local basis
functions (ADF/BAND) and plane waves (VASP). All calculations employ the PBE functional. ADF/BAND calculations as specified in the Computational
details. For VASP calculations an energy cutoff of 400 eV was used. Brillouin zone is sampled with a mesh of 20  20  1 (16  16  1) k points for
graphene (MoS2)

Graphene MoS2
a
Relative CPU time Relative CPU timea Band gap (eV)
Interlayer
Published on 14 May 2014. Downloaded by Federal University of Minas Gerais on 07/12/2017 17:17:01.

displacementb (Å) 2D PBC 3D PBC 2D PBC 3D PBC 2D PBC 3D PBC

0.0 1.00 1.00 1.00 1.00 0.89 (indirect) 0.89 (indirect)
0.4 — 1.05 — 1.11 1.22 (indirect) 1.23 (indirect)
0.8 — 1.09 — 1.18 1.46 (indirect) 1.49 (indirect)
1.2 — 1.08 — 1.17 1.64 (indirect) 1.68 (indirect)
1.8 — 1.19 — 1.33 1.77 (indirect) 1.81 (indirect)
2.4 — 1.38 — 1.35 1.82 (direct) 1.89 (indirect)
2.8 — 1.47 — 1.39 1.82 (direct) 1.91 (direct)
3.2 — 1.52 — 1.63 1.82 (direct) 1.91 (direct)
Nc 0.32 — 0.27 — 1.82 (direct) —
a b
Total CPU time relative to bulk calculation. Respect to the bulk interlayer distance. c 2D periodic boundary conditions.

into PPs/ECPs or by treating them explicitly through, for A more rigorous calculation of band gaps is possible using
example, the zeroth order regular approximation (ZORA).42–46 the GW approximation, which is usually applied on top of DFT
Spin Orbit (SO) effects are surprisingly large for single layered calculations.23,24 While it is known to produce excellent results
2D materials due to the break of the inversion symmetry from in 3D solids, surprisingly, this approach overestimates the
bulk/bilayer to the monolayer. For example, the SO induced band gap in 2D materials. The reason for this behavior has
splitting of the valence band in MoS2 is ca. 150 meV at the K been identified as the reduced Coulomb screening in 2D
point which is marking the valence band maximum.22 Density- systems, which results in high exciton binding energies. For
functional theory (DFT) has become the working horse of example, PBE calculations on MoS2 monolayers yield a band
modern solid state physics as a method with a reasonable gap of 1.8 eV, in very good agreement with the experimentally
balance between accuracy and computational cost.47 However, determined value.39 On the contrary, the GW approximation
some points need to be raised before studying 2D materials yields a significantly larger band gap of ca. 3 eV. If the GW
using this method. First, on a general basis, DFT is not adequate calculations are corrected using the Bethe–Salpether equation
to describe systems exhibiting strong electron correlations, as it (BSE), an exciton binding energy of approximately 1 eV is
tends to over-delocalize the electron density. The use of hybrid obtained. The experimentally observed band gap is obtained
DFT functionals, which incorporate a fraction of the exact if the GW value is corrected by the exciton binding energy.54–56
exchange interaction, improves this situation, but only solves An alternative approach is the use of the GLLB-SC model, which
this problem partially.48 Therefore, a commonly adopted strategy predicts band gaps of comparable quality compared to those
is the comparison of band structures obtained by pure and determined at the GW level of theory, but a much lower
hybrid GGA functionals, as for example PBE and PBE0.49–51 computational cost.57 However, this model also includes the
On one hand, band structures obtained at the DFT level of strong exciton binding energy and thus overestimates band
theory are typically of excellent quality, allowing the determina- gaps for 2D materials.
tion of effective masses of holes (h+) and electrons (e ), or As in solids, the calculation of the Brillouin zone of the
the parameterization of effective Hamiltonians for multiscale system reflects the ideal crystal, a perfectly planar sheet without
transport simulations.52 On the other hand, DFT is known to deformations or defects. While those calculations are state-of-
incorrectly describe band gaps, a consequence that unoccupied the-art and accurately yield most results on the electronic
Kohn–Sham orbitals are rather a mathematical result with a lack structure of these systems, they do not reflect the intrinsic
of a clear physical meaning. However, experience has proven that rippling of the monolayers. Ripples can significantly alter the
DFT yields a reasonable spectrum of the conduction bands when band gap of these systems, for example, ripples in MoS2 are
using an orbital basis with sufficient quality, but they are shifted able to reduce its band gap by B400 meV.19 Thus, the rippling
to lower energies. In consequence, DFT band gaps are typically of 2D systems will unlikely change the character of the systems,
strongly underestimated, with notable extreme cases such as but it does have a quantitative effect. As the consideration of
Mott insulators, where DFT predicts metallic character. Hybrid rippling is computationally demanding and leads to more difficult
DFT functionals such as HSE06 or PBE0 improve in general the analysis of the results due to the absence of an interpretable band
predicted band gap when compared with experimentally determined structure, it is neglected in most calculations, including those in
ones.48,53 However, hybrid functionals are computationally expensive this article.
compared with GGA ones, especially if used together with plane Another important issue is the presence of defects in the
waves as basis functions. materials, such as impurities, point defects and grain boundaries.

This journal is © The Royal Society of Chemistry 2014 Chem. Soc. Rev., 2014, 43, 6537--6554 | 6541
 View Article Online

Review Article Chem Soc Rev

In 2D materials, defects play a much stronger role than in 3D been isolated in free-standing form, they are highly debated in
materials, as there is no third dimension that may act as the nanoelectronics community.
stabilizing factor, e.g. to allow the transmission of a current
or to stabilize the structure. Hence, calculations of the perfect 3.1. Graphene, silicene and germanene
crystal only reflect the threshold value of a hypothetical, perfect Even though Boehm coined the term graphene already in
layer. Defects act as scattering centres and typically reduce the 1962,63 the electronic structure of a graphite monolayer has
electron transmission. Point defects can strongly influence been studied since 1947.64 Surprisingly, it was not until 2004
the electronic properties and even introduce an anisotropy into that Geim’s group was able to isolate a monolayer graphene
Published on 14 May 2014. Downloaded by Federal University of Minas Gerais on 07/12/2017 17:17:01.

the system.58 The role of the various defects is not yet completely sheet.5 Afterwards, graphene has been the focus of extensive
understood and requires more detailed studies both in experi- theoretical and experimental studies. Due to its extraordinary
ment and in theory. physical properties such as its high carrier mobility and high
Thus, the calculation of the band structure yields important electrical and thermal conductivity, graphene is considered as a
information, such as the band gap and its character. Other very promising material for nanoelectronics, even being a
quantities that are crucial for the prediction of the performance candidate to replace silicon in future electronic devices. Never-
of the materials in electronic applications are the electron theless, graphene is a semimetal without band gap, which
transmission, or quantum conductance. The most popular methods makes it not suitable for using it as electronic switch (Fig. 3,
to calculate this quantity are the non-equilibrium Green’s function left). Band gap engineering is thus required to open a band gap
(NEGF) method in conjunction with the Landauer–Büttiker in graphene without compromising any of its other properties.
formula59–61 for small-scale models, and the semi-classical The most common techniques and methods reported so far
Boltzmann transport equation (BTE), that has been applied for are doping, functionalization, defective or hybrid structures,
MoS2 and other 2D materials.52,62 The latter one is particularly substrate induced band gap opening, and quantum confine-
useful for the prediction of room temperature device perfor- ment. To date, a large manifold of potential applications of
mances as it allows the treatment of electron–phonon coupling graphene are under development, and many more have been
via a perturbation potential. proposed. These include flexible display screens, electric circuits,
and solar cells, as well as medical, chemical, and industrial
processes that could be improved by the use of new graphene
3. Atomically thin two-dimensional materials.65,66
materials Due to the fascinating properties and extensive applications
of graphene, scientists started to question whether the other
In this section, we discuss graphene and materials that are elements in main group IV, i.e. Si, Ge, Sn, and Pb, can form
directly related to this prototype 2D system. Those include stable layered structures. In 2010, Wen et al. theoretically
isoelectronic hexagonal boron nitride (h-BN), 2D materials of explored main group IV structures going from 1D to 2D to
the main group IV elements Si and Ge (silicene and germa- 3D.67 They found that the graphene layers of Si, Ge, Sn, and Pb
nene), and atomically thin transition metal chalcogenides essentially collapse to five-coordinated structures with bonds
(TMCs). The most interesting feature about these materials is between the layers. The reason is that p-bonding in main group
the significant change in properties compared with their bulk IV, being fundamentally important in all the graphene structures,
counterparts. Even though some of these systems have not yet is only possible for carbon, but not for Si, Ge, Sn, or Pb. Carbon

Fig. 3 Band structures of graphene (left), silicene (centre) and germanene (right). Valence and conduction bands are highlighted in blue and red
respectively. The Fermi levels (horizontal green dashed lines) have been shifted to 0 eV. All band structures at PBE-D3 level of theory.

6542 | Chem. Soc. Rev., 2014, 43, 6537--6554 This journal is © The Royal Society of Chemistry 2014
 View Article Online

Chem Soc Rev Review Article

effortlessly forms sp2 hybrid orbitals, the same is not possible was predicted to be 1.55 meV.77 Another interesting property of
for the other elements because of the poor p-type overlap silicene is its predicted quantum spin Hall effect in an accessible
between neighbouring p orbitals at the distance imposed by temperature regime. This property makes silicene particularly
normal s bonding. Therefore, the graphene-like sheets of Si, Ge, interesting for applications as spin Hall effect devices.77
Sn and Pb are very unlikely to have an independent existence. Germanene is the germanium analogue of silicene, where
To be consistent with the commonly used nomenclature, the atoms of silicon are replaced by germanium. Up to now, there
silicene is the word to name the silicon analogue of graphene. is no experimental evidence of this system. As in free-standing
In the past couple of years, the electronic and structural properties silicene, it is found that there is no band gap at the Fermi level,
Published on 14 May 2014. Downloaded by Federal University of Minas Gerais on 07/12/2017 17:17:01.

of silicene have been predicted theoretically. Among these studies, indicating metallic properties (Fig. 3, right). The zero band
Cahangirov et al. reported that silicon and germanium can have gap observed in germanene originates from the high-buckling
stable, 2D, slightly corrugated (b-type), honeycomb structures distance between the two atomic layers.68
which are more stable than their corresponding planar-layer type Fig. 3 shows the electronic band structure of graphene, silicene
counterparts (Fig. 2).68 Despite this buckling, the free-standing and germanene, the latter two in their buckled configurations. All
silicene and germanene structures have enough symmetry to the structures present similar features in the electronic band
preserve the feature of linearly crossing bands around the Fermi structure; the bands cross at the Fermi level, located in the Dirac
level. This makes electrons of silicene behave as mass-less Dirac (K) points of the reciprocal lattice. The linear dispersion around the
Fermions as in graphene (Fig. 3, centre). Dirac points results from the honeycomb structure. Interestingly,
Due to its unique bonding behaviour, bulk Si cannot form a this linear dispersion is preserved even when the hybridization of
layered phase like graphite. However, experiments of surface- the atoms changes (C is in sp2 hybridization, while isoelectronic Si
assisted epitaxial growth show the presence of nanoribbons and and Ge prefer sp3).
2D monolayers of silicene on Ag(110)69 and Ag(111),70–73 respec-
tively. The associated scanning tunnelling microscopy (STM) images 3.2. Boron nitride
revealed hexagons in a honeycomb structure similar to those of Boron nitride exhibits various crystalline polymorphs analogous to
graphene. Although free-standing silicene is expected to have a zero carbon, including diamond-like cubic BN, graphite-like hexagonal
band gap, a tiny gap can be opened in epitaxial silicene, due to the boron nitride (h-BN), onion-like fullerenes, etc.79 Within these
symmetry-breaking induced by the interaction with the silver sub- polymorphs, h-BN is thermodynamically the most stable allotrope
strate. Vogt et al. provided evidence for the successful synthesis of and, due to its 2D nature, has attracted enormous attention.
epitaxial silicene sheets on an Ag(111) substrate, by analysing both This material is an isoelectronic analogue of graphene,
the structural and electronic properties through the combination of composed of alternating boron and nitrogen atoms in a honeycomb
STM and angular-resolved photo-emission spectroscopy (ARPES) in arrangement. sp2-bonded h-BN shows strong covalent bonds within
conjunction with DFT calculations.73 Later on, theory predicted that the plane and weak bonds with van-der-Waals forces between
Dirac electrons are absent near the Fermi energy in all stable different planes. It is worth mentioning that, in contrast to graphite,
silicene on Ag(111) structures due to the buckling of the silicene bulk h-BN favours AA 0 (eclipsed) stacking.80 Regardless of
monolayer and hybridization between Si and Ag orbitals.74 the geometrical similarities, h-BN shows strikingly different
Recently, silicene has been reported to also grow on ZrB2 and electronic properties from those of graphene. The former exhibits
Ir(111) substrates.75,76 Due to the higher mass of Si compared to a direct K - K band gap between 4.69 eV (PBE) and 7.92 eV (with
C, silicene has a stronger SO splitting compared to graphene, it the GLLB-SC quasi-particle correction) (Table 2), characteristic for

Table 2 Selected properties of atomically thin materials. mhh and me stand for the effective mass of the hole, and the electron, respectively. Band gap
and effective masses in eV and m0 (rest mass of electron), respectively78

Band gap Effective massa

Material Isolated PBE-BJ-D3 GLLB-SC Transition mh me
Graphene Yes 0.00 (Dirac point) 0.00 (Dirac point) K — —
Graphane Yes 3.56 (direct) 7.00 (direct) C-C 0.249 0.983
Fluorographene Yes 3.29 (direct) 5.16 (direct) C-C 0.338 0.466

b-Silicene Yes 0.00 (Dirac point) 0.00 (Dirac point) K — —

Silicane No 2.26 (indirect) 3.56 (indirect) C-M 0.128 3.838
Fluorosilicene No 0.66 (direct) 1.66 (direct) C-C 0.128 0.235

b-Germanene No 0.00 (Dirac point) 0.00 (Dirac point) K — —

Germanane Yes 1.16 (direct) 1.84 (direct) C-C 0.061 0.063
Fluorogermanene No Metallic 0.39 (direct) —b 0.017 0.017

h-BN Yes 4.69 (direct) 7.92 (direct) K-K 0.792 1.175

SiC Yes 2.55 (direct) 3.63 (direct) C-C 0.549 0.645

a b
From GLLB-SC band structure. Only x direction. In GLLB-SC the transition is C - C.

This journal is © The Royal Society of Chemistry 2014 Chem. Soc. Rev., 2014, 43, 6537--6554 | 6543
 View Article Online

Review Article Chem Soc Rev

Published on 14 May 2014. Downloaded by Federal University of Minas Gerais on 07/12/2017 17:17:01.

Fig. 4 Band structures of hexagonal boron nitride and silicon carbide (top); atomically thin group XII transition metal chalcogenides (bottom). Valence
and conduction bands are highlighted in blue and red respectively. The Fermi levels (horizontal green dashed lines) have been shifted to 0 eV. All band
structures at PBE-D3 level of theory.

an insulating white material. The narrowing of the sp2 p bands common formula MX2, which are not atomically thin. Instead,
(due to the difference in electronegativity) compared with graphene they are arranged in triatomic layers with a metal sheet sand-
is responsible for the loss of conductivity yielding an insulator wiched between two chalcogenide sheets and will be discussed
instead of a semimetal (Fig. 4, top left). in the next Section. However, during the last years significant
advances have been achieved towards the synthesis of atomically
3.3. Silicon carbide thin TMCs. Mono- and multi-layered cadmium sulphides and
Naturally, silicon carbide (SiC) occurs in the form of cubic, selenides have been recently synthesized using solvothermal and
hexagonal and rhombohedral structures, where Si and C favour colloidal techniques.84–86 Since the synthesis of these materials
sp3 hybridization. However, a planar structure of SiC with sp2- was performed in the presence of a surfactant (e.g. long chain
hybridized bonds, thus resembling graphene, was theoretically alkylamines), no free-standing monolayers were isolated. In a
predicted to have high structural stability.81 The graphene-like different approach, Tusche et al.87 and Weirum et al.88 revealed
SiC consists of alternating Si and C atoms, where each Si atom that the deposition of thin layers of zinc oxide on top of a metal
has three C atoms as its nearest neighbours and vice versa, with a substrate leads to the formation of a honeycomb layered
Si–C bond length of 1.79 Å.82,83 Due to the Si–C ionicity, the planar structure. These layered materials are less stable than their
SiC system is a semiconductor with a direct K - K band gap of respective wurtzite bulk ground state phases, however, they
about 2.55 eV (Fig. 4, top right), which increases to 3.63 eV when could be stabilized by the dispersive interactions between the
computed with GLLB-SC as shown in Table 2. Both valence and stacked layers. These advances are highly promising towards
conduction bands have predominantly Si and C p-orbital character. the synthesis of free-standing atomically thin group XII TMCs
The upper valence band consists of one p band which arises from from non-layered materials.
the 2pz and 3pz orbitals, extending above and below the SiC layer Exfoliated atomically thin TMCs present a honeycomb lattice,
plane, and two s bands involving the three C 2s, 2px, 2py and three however, depending on their composition the sheets are flat
Si 3s, 3px, 3py orbitals, which form the Si–C s bonds. (a-type) or slightly corrugated (b-type). Light metal–chalcogenide
combinations usually present the former structure (e.g. ZnS,
3.4. Transition metal chalcogenides ZnSe and CdS), while heavier combinations prefer the latter
Many transition metal chalcogenides (TMCs) present non- one (e.g. CdSe). DFT calculations on atomically thin zinc and
layered structures such as zinc blende or wurtzite, while layered cadmium chalcogenides predict a semiconductor nature with a
TMCs are commonly restricted to metals in groups IV–VI and X. direct band gap at the G point (Fig. 4, bottom). Our band gaps of
Typical 2D TMCs are transition metal dichalcogenides with a zinc and cadmium sulphides and selenides obtained at the PBE

6544 | Chem. Soc. Rev., 2014, 43, 6537--6554 This journal is © The Royal Society of Chemistry 2014
 View Article Online

Chem Soc Rev Review Article

Table 3 Selected properties of atomically thin materials transition metal chalcogenides at PBE-BJ-D3 and GLLB-SC levels of theory. mlh, mhh and me
stand for the effective mass of the light hole, the heavy hole, and the electron, respectively (for explanation of the effective masses, see Scheme 1).
Relative stability, band gap and effective masses in meV/formula, eV and m0 (rest mass of electron), respectively78

Band gap Effective massa

Material Rel. stability PBE-BJ-D3 GLLB-SC Transition mlh mhh me
ZnS Flat (a) 0.0 2.58 (direct) 4.50 (direct) C - C 0.134 — 0.187
Corrugated (b) 1.5 2.57 (direct) 4.48 (direct) C - C 0.176 0.178 0.187
ZnSe Flat (a) 0.0 1.91 (indirect) 3.37 (direct) K Cb 0.107 0.529 0.126
Published on 14 May 2014. Downloaded by Federal University of Minas Gerais on 07/12/2017 17:17:01.

-
Corrugated (b) 3.4 2.01 (direct) 3.48 (direct) C - C 0.114 0.141 0.139

CdS Flat (a) 0.0 1.72 (direct) 3.23 (direct) C - C 0.157 0.723 0.167
Corrugated (b) 4.9 1.65 (direct) 3.14 (direct) C - C 0.157 0.751 0.165
CdSe Flat (a) 9.6 1.20 (indirect) 2.34 (direct) K - Cb 0.096 0.274 0.111
Corrugated (b) 0.0 1.30 (direct) 2.47 (direct) C - C 0.094 0.141 0.127
a b
From GLLB-SC band structure. Only x direction. In GLLB-SC the transition is C - C.

level of theory are in good agreement with previous theoretical Analogously, the electron effective mass (me) is obtained by
studies (Table 3). Besides, recent quasi-particle calculations fitting the minimum of the conduction band. Effective masses
(GW0) suggested significantly larger band gaps (4.27 and 3.53 eV are related to charge carrier mobilities and thus, consequently,
for CdS and CdSe, respectively), but they are subject to the exciton to the electric and thermal conductivities of the materials. The
binding energy and thus probably overestimate the band effective masses for atomically thin TMCs are presented in
gaps.89,90 Finally, the GLLB-SC model yields results between PBE Table 3.
and GW0 values. In all cases, both a and b types are nearly
degenerate with an energy difference of a few meV and no overall
changes are observed in their band structure. 4. Ultrathin 2D materials
The experimentally available band gaps for atomically thin
In this section, we discuss 2D materials that are more than one
ZnS, ZnSe, CdS and CdSe layers are 3.9, 3.5, 2.9 and 2.7 eV,
atom thick (ultrathin materials). We limit our study to three
respectively. However, these layers were synthesized in
families of ultrathin 2D materials: hydrogenated and fluorinated
presence of alkylamine surfactants, which induce a major
graphene, silicene and germanene derivatives; rare earth, semimetal
corrugation in the layered structure.84 In consequence, these
and transition metal chalcogenides; and rare earth and transition
band gaps are not directly comparable with the calculated ones
metal halides. Some of the presented systems have not yet been
where the surfactants are absent and corrugation is inexistent
isolated in their bulk and/or free-standing form, however, the
or minimal.
theoretical prediction of their properties is the vanguard for the
In these materials, two bands converge at the valence band
experimental growth and isolation of the most promising materials.
maximum (the C-point). The flatter band is known as heavy-hole
band, while the steeper one is the light-hole band (Scheme 1). 4.1. Graphane, silicane and germanane
The heavy and light hole effective masses (mlh and mhh, respec-
tively) of a semiconductor is obtained by fitting the area in the k Chemical functionalization, especially hydrogenation and
space around these bands maximum to a parabola. The effective fluorination, are being used to tailor the electronic properties
mass reflects the inverse of the band curvature, in consequence of graphene-like materials. Here we describe briefly the latest
the lower the effective mass the larger the band curvature is. studies on these saturated structures. In 2013, Bianco et al.
reported the synthesis of stable, single-layered germanane
using topotactic deintercalation. This involves a structural
change to a crystalline solid and the final lattice is related to
the original material by one or more crystallographically
equivalent orientations.91 A large crystal of b-CaGe2 was con-
verted into a layered GeH when placed in aqueous HCl at 40 1C.
b-CaGe2 has alternating planes of covalently bonded germa-
nium atom layers separated by ionically bonded interstitial
calcium. The bonding type in germanane appears to have
mixed sp2 and sp3 hybridization. Ultrathin germanane has a
remarkable resistance to oxidative degradation and has been
found to be stable over five months in air. Germanane, unlike
silicane, does not require a substrate to be stable. The atomic
structure of germanane is slightly corrugated rather than flat
Scheme 1 Heavy- and light-hole bands for ZnS. The Fermi level (blue line) like graphane or graphene (Fig. 2). Calculations do not show the
has been shifted to 0 eV. existence of a Dirac cone, but germanane still has a surprisingly

This journal is © The Royal Society of Chemistry 2014 Chem. Soc. Rev., 2014, 43, 6537--6554 | 6545
 View Article Online

Review Article Chem Soc Rev

high electron mobility, which is much higher than that of its bulk task, as GGA provides a band gap of 3.29 eV, hybrid (HSE06)
material. A theoretical analysis showed that germanane has a 4.9 eV, while many-electron approaches based on GW approxi-
direct band gap of 1.53 eV, making it a potential material for solar mation give a band gap of 8.1 eV (twice as large).94 Therefore, it
cells.92 Contrary to graphane, the hydrogenation in germanane is necessary to include the exciton binding energy to obtain the
is not a reversible process. It becomes an amorphous material correct band gap value. Band structures of fluorinated and
above 75 1C. hydrogenated graphene and its analogues are given in Fig. 5.
Fluorinated silicene was theoretically studied by Ding
4.2. Fluorographene, fluorosilicene and fluorogermanene et al.,95 who reported that it has a direct band gap which values
Published on 14 May 2014. Downloaded by Federal University of Minas Gerais on 07/12/2017 17:17:01.

can be tuned by strain in the lattice. The values of the

Fluorographene is the fluorinated derivative of graphene,
formation energies for these structures are negative, indicating
namely, it is a 2D carbon monolayer of sp3 hybridized carbons,
that the hydrogenation/fluorination of silicene is an exothermic
with each carbon atom bound to one fluorine atom. Similar to
reaction and the corresponding H-/F-silicene is stable.
other fluorocarbons, fluorographene is highly insulating and
has a high thermal and chemical stability. However, it can be
transformed back into graphene by a reaction with potassium 4.3. Transition metal chalcogenides and semimetal
iodide under high temperature. chalcogenides
Fluorographenes are expected to present similar structural Transition metal chalcogenides (TMCs) form a family of materials
and electronic characteristics as graphane. However, since with a general formula MnXm, where M is a transition metal and X
fluorine has a much larger electronegativity than hydrogen, is always a chalcogenide element. Layered TMCs have been
the charge transfer between graphene and F atoms in fluoro- reported with elements of group IV (Ti, Zr and Hf), group VI
graphene is distinctive from that in graphane. The single-layer (Mo and W), group X (Pd and Pt) as well as with Ta, Re and Nb.96
fluorographene exhibits a wide band gap semiconducting These materials always present a MX2 stoichiometry and are
behaviour with an optical gap of ca. 3 eV; which is in good composed of 2D sheets stacked on top of one another. Each sheet
agreement with the recent experimentally measured one of is three atoms thick, with a metal atom in the middle that is
around 3.8 eV.93 Nevertheless, first principle calculations show strongly bonded to chalcogenide atoms located above and
again that the estimation of the band gap can be a challenging below (Fig. 2). The intralayer metal–chalcogen bond is

Fig. 5 Band structures of hydrogenated (top) and fluorinated (bottom) derivatives of graphene (left), silicene (centre) and germanene (right). Valence and
conduction bands are highlighted in blue and red respectively. The Fermi levels (horizontal green dashed lines) have been shifted to 0 eV. All band
structures at PBE-D3 level of theory.

6546 | Chem. Soc. Rev., 2014, 43, 6537--6554 This journal is © The Royal Society of Chemistry 2014
 View Article Online

Chem Soc Rev Review Article

predominantly covalent in nature, while the sheets are held Jeong et al. studied the chemical reactivity of the edge and basal
together by weak interlayer interactions (London dispersion). planes and their use in hybrid materials (e.g. TiS2–TiO2) with
This facilitates the shearing of TMC layers, which has led to the enhanced solar energy uptake and facilitated electron-transfer
initial applications of these materials as high performance properties.117 The noble TMC class, by contrast, has been known
lubricants.97 Layered TMCs only occur in two polytypes, in both for almost a century and was theoretically studied recently, but
the metal centre either has a trigonal prismatic (H = hexagonal their exfoliation and application as 2D materials has yet to be
or rhombohedral) or octahedral (T = trigonal prismatic or pursued.41,118
octahedral) environment. The electronic structure of bulk layered TMCs has been
Published on 14 May 2014. Downloaded by Federal University of Minas Gerais on 07/12/2017 17:17:01.

Recently, TMCs have emerged as a graphene alternative, as previously studied by means of ab initio calculations using the
they display unique chemical and physical properties that are plane wave approach as well as by employing local basis
absent or difficult to obtain in other 2D materials. For exam- functions. Among them, we selected zirconium, niobium,
ple, in graphene, complex band gap engineering is required molybdenum, and palladium disulphides as representative
towards its applications as transistors. Various TMC represen- materials for both T and H polytypes. The metals in these
tatives, however, are inherent semiconductors.98 Furthermore, materials have formally an oxidation state of IV, leading to
TMCs present a wide variety of electronic properties including Zr4+(d0), Nb4+(d1) Mo4+(d2), and Pd4+(d6) metal centres. The
metals, semimetals, insulators and semiconductors with crystal field theory predicts a small d orbital splitting for
direct and indirect band gaps and a range of electron and transition metals in presence of chalcogenide ligands. This
hole mobilities depending on their composition. Conse- allows a qualitative prediction of the electronic nature of
quently, these materials have a variety of highly desirable these materials without quantum mechanical calculations.
characteristics affecting charge transport, magnetism, the However, the d orbital splitting for H and T polytypes differs
intercalation of ions and small molecules, and their catalytic and is well established. For the former, this splitting derives in
and optical properties. three orbital groups of increasing energy (dz2; dxy and dx2–y2;
The most prominent TMCs are MoS2, WS2, MoSe2 and WSe2 dxz and dyz), while on the latter a traditional octahedral field
(sometimes referred to collectively as MoWSeS) that have been orbital splitting is observed (dxy, dxz and dyz; dz2 and dx2–y2).
widely studied for applications in electronics since they are The band structures of the representative materials are
semiconductors. The band gaps of these materials can be shown in Fig. 6. Trigonal and hexagonal zirconium disul-
‘‘easily’’ tuned by stacking confinement and strong electric phides present always a semiconductor band structure
fields, as recently shown by Zibouche et al.99 For example, bulk with band gaps of ca. 1 eV, however, the trigonal polytype is
MoS2 has an indirect band gap of 0.9 eV, while at the monolayer the energetically most favourable one. Molybdenum disul-
limit it switches to a direct band gap of 1.8 eV.22 The transition phide is a metal in T polytype, but a direct band gap semi-
from indirect to direct bandgap has a huge implication when conductor (1.82 eV) in the H form, which is the most stable
studying photoluminescence as the latter increases when and common one. On the contrary, niobium disulphides are
decreasing the layer thickness.100–102 Other features that make always metallic, independently on the polytype due to the
MoWSeS interesting for applications in nanoelectronics are unpaired electron in the metal centre. Finally, the palladium
chemical stability, the absence of dangling bonds (except at the disulphide trigonal polytype is an indirect band gap semicon-
sheet edges and corners), and thermal stability up to E1100 1C. ductor with a C - 12M transition, while the hexagonal polytype
Radisavljevic et al. have recently demonstrated that single-layer is a metal.
MoS2 can be used to fabricate transistors with high electron Semimetal chalcogenides (SMCs) are a family of layered
mobility and high current on/off ratios.14 As similar manufacturing materials similar to TMCs, but including a semimetal instead
principles as in silicon semiconductor electronics are feasible, of transition metals. Their most common stoichiometry is
including gating and contacting, complex devices suitable M2X2 in X–M–M–X tetraatomic thick layers (Fig. 2). Gallium
for building integrated circuits are possible. Indeed, logic and indium sulphides and selenides are among the most
operations and integrated circuits and nonvolatile memory prominent materials in this family. They are predicted to be
cells based on single-layer MoS2 and MoS2/graphene have been semiconductors with indirect 12M to C band gaps of 2.57, 2.05,
recently demonstrated.14,92,103–108 Moreover, the high stiffness 2.09, and 1.70 eV for GaS, GaSe, InS, and InSe, respectively,
and breaking strength of MoWSeS materials has shown their within the PBE level of theory (Fig. 7). Recent theoretical studies
compatibility for the use in flexible electronics.19,109–112 These pointed out that possible direct and indirect transitions
studies are part of the vanguard of a rapidly emerging field, 2D are only slightly different in energy. This difference is small
materials beyond graphene. And, while devices based on enough to allow the switching between direct and indirect
MoWSeS materials are blooming,104,113 basic research on other semiconductors for GaS and GaSe by thermal energy.119
layered TMCs is still in early stages. Furthermore, the band gaps of SMC monolayers can be tuned
Two examples of semiconducting TMCs beyond MoWSeS are by mechanical deformation, making these materials potential
those based on the early transition metals (e.g. TiS2, ZrS2, HfS2. . .) candidates for novel nanodevices. The GLLB-SC model predicts
and the noble metals (PdS2, PtS2. . .). Members of the former class significantly larger band gaps, which lay in the insulator range in
have recently been explored as nanodiscs due to a large anisotropy many cases, however, due to the excitonic effects, the GLLB-SC
between their in-plane and out-of-plane growth.114–116 In addition, band gaps are overestimated (Table 4).

This journal is © The Royal Society of Chemistry 2014 Chem. Soc. Rev., 2014, 43, 6537--6554 | 6547
 View Article Online

Review Article Chem Soc Rev

Published on 14 May 2014. Downloaded by Federal University of Minas Gerais on 07/12/2017 17:17:01.

Fig. 6 Band structures of H (top) and T (centre) polytypes of MoS2, ZrS2, NbS2 and PdS2 transition metal chalcogenides (left to right respectively).
Valence and conduction bands are highlighted in blue and red respectively. The Fermi levels (horizontal green dashed lines) have been shifted to 0 eV.
All band structures at PBE-D3 level of theory.

Fig. 7 Band structures of gallium (left) and indium (right) chalcogenides. Valence and conduction bands are highlighted in blue and red respectively.
The Fermi levels (horizontal green dashed lines) have been shifted to 0 eV. All band structures at PBE-D3 level of theory.

in both cases. Analogously to TMC materials, the 2D TMH sheets

4.4. Transition metal halides are triatomic (halide–metal–halide) and stacked on top of each
Transition metal halides (TMHs) encompass a family of materials other (Fig. 2). The interlayer interaction is also mainly dominated
with the general formula MnYm, where M is a transition metal and by weak interlayer interactions (London dispersion).
X is always a halogen element. Only few 3D layered TMH materials Initially, we want to focus on the electronic structure of TMH
have been observed experimentally, such as cadmium, vanadium monolayers with MY2 stoichiometry from which we selected
and molybdenum halides.120 However, their exfoliation down to calcium, manganese, iron and nickel chlorides and bromides as
the monolayer still remains unachieved. Two main stoichiometries TMH representative materials. This derives formally in Ca2+(d0),
(MY2 and MY3) have been observed in experimental TMHs, with Mn2+(d5), Fe2+(d6) and Ni2+(d8) metal centres. The crystal field
the metal centre in a trigonal prismatic (octahedral) environment theory predicts a small d orbital splitting for transition metals in

6548 | Chem. Soc. Rev., 2014, 43, 6537--6554 This journal is © The Royal Society of Chemistry 2014
 View Article Online

Chem Soc Rev Review Article

Table 4 Selected properties of transition metal chalcogenides and semimetal chalcogenides at PBE-BJ-D3 and GLLB-SC levels of theory. mlh, mhh and
me stand for the effective mass of the light hole, the heavy hole, and the electron, respectively. Band gap, spin orbit splitting and effective masses in eV,
meV and m0 (rest mass of electron), respectively78

Band gap Effective massa

Material PBE-BJ-D3 (scalar) PBE-BJ-D3 (SO) SO splitting GLLB-SC Transition mlh mhh me
ZrS2 T 1.10 (indirect) 1.06 (indirect) 80 2.30 (indirect) C-M 0.218 0.221 1.869
b
H 0.92 (indirect) 0.90 (indirect) 20 2.03 (indirect) — 0.726 13.881
MoS2 T Metallic Metallic — Metallic — — — —
Published on 14 May 2014. Downloaded by Federal University of Minas Gerais on 07/12/2017 17:17:01.

H 1.82 (direct) 1.74 (direct) 150 2.51 (direct) K-K — 0.557 0.463

NbS2 T Metallic Metallic — Metallic — — — —

H Metallic Metallic — Metallic — — — —
PdS2 T 1.17 (indirect) 1.15 (indirect) 80 1.73 (indirect) C - 12M 0.725 0.638 0.407
H Metallic Metallic — Metallic — — — —
1
GaS — 2.57 (indirect) 2.56 (indirect) — 4.12 (indirect) 2 M - C — 2.755 2.020
1
GaSe — 2.05 (indirect) 2.02 (indirect) — 3.38 (indirect) 2 M - C — 2.007 0.173
1
InS — 2.09 (indirect) 2.09 (indirect) — 3.58 (indirect) 2 M - C — 3.758 0.313
1
InSe — 1.70 (indirect) 1.66 (indirect) — 2.93 (indirect) 2 M - C — 3.228 0.213
a b
From GLLB-SC band structure. Only x direction. Check Fig. 6. Computational details are given in Section 7.

Fig. 8 Band structures of iron (left) and molybdenum (right) transition metal chlorides and bromides (MY3 stoichiometry). Valence and conduction bands are
highlighted in blue and red respectively. The Fermi levels (horizontal green dashed lines) have been shifted to 0 eV. All band structures at PBE-D3 level of theory.

presence of ligands at the beginning of the spectrochemical series Finally we present TMH monolayers with MY3 stoichiometry.
(such as halogens). This makes it difficult to predict the electronic The unit cells of these materials are similar to the MY2 ones, where
nature of these materials with the exception of calcium halides, one of every three metal centres has been removed. In conse-
which are expected to be insulators due to the completely empty quence, these materials present a slightly larger d orbital splitting
4s and 3d bands. DFT calculations on TMHs predict wide varieties and their metal centres are in oxidation state III. On one hand,
of electronic structures and confirm the small d orbital splitting iron halides are metallic since the metal electrons (5) are localized
(e.g. Mn in manganese halides is predicted to be high spin; in the three t2g orbitals (dxy, dxz and dyz), while the eg orbitals lay
5 electrons in 5d orbitals) (see page S54, ESI†). Manganese halides higher in energy. On the other hand, their molybdenum analogues
are high spin small band gap semiconductors (0.38 and 0.18 eV for are semiconductors with partially filled t2g orbitals that lead to
MnCl2 and MnBr2, respectively); iron halides are always metallic band gaps of 0.80 and 0.56 eV in MoCl3 and MoBr3, respectively, at
due to the partial filling of the metal d orbitals (6 electrons in PBE level of theory (Fig. 8 and Table 5).
5 orbitals); finally nickel halides are high spin semiconductors
with band gaps of 1.06 and 0.64 eV, respectively, at the PBE level of
theory. Furthermore, the band gap in semiconducting TMH 5. Synthetic 2D organic frameworks
materials decreases when going down in the halogen group, in
agreement with the smaller crystal field splitting (I o Br o Cl o F) 5.1. Layered covalent organic frameworks
predicted in the spectrochemical series.121 Thus, these trends Covalent organic frameworks (COFs) are crystalline porous
confirm the d nature of the frontier bands above and under the materials formed from the covalent bonding of light elements
Fermi energy. Analogously with TMC monolayers, calculations (H, B, C, N and O). COFs have low mass densities, possess high
with GLLB-SC model largely overestimate TMH band gaps. thermal stabilities, and provide permanent porosity. COFs can

This journal is © The Royal Society of Chemistry 2014 Chem. Soc. Rev., 2014, 43, 6537--6554 | 6549
 View Article Online

Review Article Chem Soc Rev

Table 5 Selected properties of MY2 and MY3 transition metal halides at PBE-BJ-D3 level of theory. mlh, mhh and me stand for the effective mass of the
light hole, the heavy hole, and the electron, respectively. Band gap and effective masses in eV and m0 (rest mass of electron), respectively78

Band gap Effective massa

Material PBE-BJ-D3 GLLB-SC Transition mlh mhh me
CaCl2 5.97 (indirect) 9.92 (indirect) C-M — 2.156 1.698
CaBr2 5.13 (indirect) 10.32 (indirect) C-M 0.621 2.224 0.384
MnCl2 0.37 (direct) —b M - Mc —c,d 1.079c,d 2.582c,d
MnBr2 0.18 (indirect) —b C - Mc 0.314 0.326c,d 2.862c,d
Published on 14 May 2014. Downloaded by Federal University of Minas Gerais on 07/12/2017 17:17:01.

FeCl2 Metallic Metallic — — — —

FeBr2 Metallic Metallic — — — —
NiCl2 1.06 (indirect) —b C - Mc 0.490c,d 0.549c,d 2.447c,d
NiBr2 0.64 (indirect) —b C - Mc 0.286c,d 1.154c,d 1.988c,d

FeCl3 Metallic Metallic — — — —

FeBr3 Metallic Metallic — — — —
1
MoCl3 0.80 (indirect) 1.20 (indirect) 2M - C 1.748 8.239 3.291
1
MoBr3 0.56 (indirect) 0.87 (indirect) 2M - C 0.382 42.664 3.591
a b
From GLLB-SC band structure. Only x direction. GLLB-SC not implemented for unrestricted calculations. c From BPE-BJ-D3 band structure.
d
Holes spin up, electrons spin down.

be known to crystallize in either 2D or 3D forms. If connectors used Schiff-base coupling on solid surfaces. They mixed different
and linkers are planar, layered COFs can be obtained (most of linkers, principally benzene-1,3,5-tricarbaldehyde, with diamine in
the boronic-acid-derived bulk COFs). Indeed, the first COFs that octanoic acid and let them polymerize on a surface while heating
have been reported are layered materials. Their structures have moderately. Using this approach, Xu et al. prepared four well-
been denoted by the Yaghi group as COF-n, n being an integer ordered honeycomb Schiff-base surface COFs on a HOPG surface.127
labelling the material in the historical order of their synthesis.25 The lone pair electrons of nitrogen make the Schiff-base group a
More recently, Banerjee and co-workers have extended the good ligand for coordination, thus these Schiff-base surface COFs
range of connectors and linkers e.g. by using boroxine and can be potentially used in the field of chemical sensing and catalysis.
1,3,5-triformylphloroglucinol.122–126 A large number of phenyl- 2D COFs are predicted to be stable in organic media, but exfoliation
based molecules have been used as linkers such as phenyl, techniques to obtain single layers are still a hot topic. First Covalent
meta-terphenyl and hexadecahydropyrene, leading to a wide Organic Nanosheets (CONs), exfoliated by micromechanical means,
variety of 2D COFs (Fig. 9). As another alternative, Xu et al. have have been reported recently by the Banerjee group.123

Fig. 9 Selected examples of layered covalent organic frameworks (COFs): COF-1 (a), COF reported by Xu and co-workers (b), COF incorporating
in-plane hydrogen bridges reported by Banerjee and co-workers. (c) COF-5 (d). Colour code: boron in pink, oxygen in red, nitrogen in blue, carbon in
grey and hydrogen in white. Unit cell is depicted in black.

6550 | Chem. Soc. Rev., 2014, 43, 6537--6554 This journal is © The Royal Society of Chemistry 2014
 View Article Online

Chem Soc Rev Review Article

The electronic structure of 2D COFs has been predicted to be important problem is the interaction of the 2D materials with
insulating or semiconducting. The band gap is typically very substrates and other factors present in their environment. The
similar to the smallest HOMO–LUMO gap of the constituting enormous progress in the field of 2D materials suggests that many
molecules, confirming the concept of reticular chemistry.128 of those challenges will be tackled in the not too distant future. In
For example, single layer phenyl/boroxine, meta-terphenyl/boroxine, this vein it is interesting to note that special issues on 2D materials
hexadecahydropyrene/boroxine, biphenyl/boroxine and hexadeca- beyond graphene are planned for 2014, one in Accounts of Chemical
hydropyrene COFs have gaps larger than 2 eV, respectively, at PBE Research and another one in Chemical Society Reviews.
level of theory and using the G point approximation. This restricts
Published on 14 May 2014. Downloaded by Federal University of Minas Gerais on 07/12/2017 17:17:01.

their application in nanoelectronics. However, their porous nature

opens new possibilities towards the development of 2D nanoporous
7. Computational details
organic membranes (nanosieves). The membrane selectivity could All DFT calculations were performed using the Amsterdam Density
be tuned with the pore size as well as with the proper linker Functional (ADF2013-BAND) package.29–31 We used the local
functionalization. VWN exchange–correlation potential with nonlocal Perdew–Burke–
Ernzerhof exchange–correlation correction and empirical D3
treatment of London dispersion interactions (PBE-BJ-D3).50,51,130
6. Conclusions and perspectives A triple-z plus two-polarization function basis set was used for all
atoms. Relativistic corrections were introduced by scalar relativistic
The discovery of 2D materials has started a new era of materials zeroth order regular approximation (ZORA).42–46 Both k-space and
science. New materials, atomically thin and mechanically, thermally accuracy were set to 5. All band structures have been evaluated on
and electronically stable, with a large variety of electronic properties the PBE-BJ-D3 optimized structures including spin–orbit correc-
are available and they can be assembled in ultrathin flexible devices. tions. Additional calculations using the M06-L meta-GGA
The manufacturing of new devices requires the detailed under- exchange–correlation functional and the GLLB-SC model57,131 have
standing of the properties of 2D materials, which is supported by been performed. The former ones are given in the ESI,† while the
the electronic properties presented in this work. Band gaps and latter ones are included in the tables of the main text.
effective electron and hole masses allow the estimation of charge
and spin mobilities for basic device simulations. The band struc-
tures, given in the ESI,† allow the parameterization of tight-binding 8. The atlas of 2D materials
models that are useful for more elaborate electron transport calcula-
For all the materials discussed in this article and additional
tions and device simulations. Our results show that all classes of
ones, Cartesian coordinates, electronic band structures, band
materials that are known from the 3D world have counterparts in
gaps and effective masses of electrons and holes are available
2D. We find metals and insulators, semiconductors with a large
as part of the ESI.†
variety of band gaps and with different band features. In addition,
2D materials exhibit properties that are unknown from the bulk, for 8.1. Atomically thin 2D materials
example massless Dirac electrons.
DFT works very well for predicting the structure of 2D Graphene, graphane, fluorographene, chlorographene, silicene,
materials. It should be reminded that – if interlayer interactions silicane, fluorosilicene, germanene, germanane, fluorogermanene,
are investigated – presently available DFT functionals need to chlorogermanene, silicon carbide, boron nitride, a-ZnO, a-ZnS,
be augmented by a scheme that corrects for the missing a-ZnSe, a-ZnTe, a-CdO, a-CdS, a-CdSe, a-CdTe, b-ZnS, b-ZnSe,
London dispersion interactions. b-ZnTe, b-CdO, b-CdS, b-CdSe and b-CdTe.
DFT yields excellent band structures, however, the position
8.2. Ultrathin 2D materials
of the conduction bands, and hence the band gaps, tend to be
underestimated. The quenched Coulomb screening in 2D GaS, GaSe, InS, InSe, HfS2, HfSe2, HfTe2, MoS2, MoSe2, MoTe2,
materials causes strong exciton binding energies of B1 eV. NbS2, NbSe2, NbTe2, NiS2, NiSe2, NiTe2, PdS2, PdSe2, PdTe2, PtS2,
This energy is included in the electronic band gaps predicted by PtSe2, PtTe2, ReS2, ReSe2, ReTe2, TaS2, TaSe2, TaTe2, TiS2, TiSe2,
quasiparticle theories such as the GW approximation or the TiTe2, WS2, WSe2, WTe2, ZrS2, ZrSe2, ZrTe2, CoCl2, CoBr2, FeCl2,
GLLB-SC model. In some cases, e.g. in the MoWSeS TMCs, this FeBr2, FeI2, HfCl2, HfBr2, HfI2, MnCl2, MnBr2, MnI2, MoCl2,
leads to an error cancellation for DFT functionals that perform MoBr2, MoI2, NbCl2, NbBr2, NbI2, NiCl2, NiBr2, TaCl2, TaBr2,
very well for these systems. TaI2, TiCl2, TiBr2, TiI2, VCl2, VBr2, VI2, WCl2, WBr2, WI2, ZrCl2,
In 2D materials, the ultimate frontier of miniaturization has ZrBr2, ZrI2, AsCl3, CrCl3, CrBr3, CrI3, FeCl3, FeBr3, MoCl3,MoBr3,
been reached – at least in one dimension. The calculations of SbCl3, ScCl3, ScBr3, TiCl3, TiBr3, VCl3, VBr3, YCl3 and ZrCl3.
properties of 2D infinite layers are state-of-the-art. However,
calculations beyond that limit are still demanding. Besides the Acknowledgements
obvious challenges such as the inclusion of structural and
stoichiometric defects and doping, a remaining issue is lateral This work was supported by the German Research Council
quantum confinement, where electronic and optical properties (Deutsche Forschungsgemeinschaft, HE 3543/18-1), The European
depend on the lateral size of the 2D structures.114,129 Another Commission (FP7-PEOPLE-2009-IAPP QUASINANO, GA 251149 and

This journal is © The Royal Society of Chemistry 2014 Chem. Soc. Rev., 2014, 43, 6537--6554 | 6551
 View Article Online

Review Article Chem Soc Rev

FP7-PEOPLE-2012-ITN MoWSeS, GA 317451), The Office of Naval 23 W. G. Aulbur, L. Jonsson and J. W. Wilkins, Solid State
Research Global (N62909-13-1-N222). We thank Dr Agnieszka Kuc Phys., 2000, 54, 1–218.
and Mr Mahdi Ghorbani-Asl for critically reading and commenting 24 L. Hedin, Phys. Rev., 1965, 139, A796–A823.
the manuscript; and Mr Wenqing Liu for the help with the plane 25 A. P. Cote, A. I. Benin, N. W. Ockwig, M. O’Keeffe,
wave calculations. A. J. Matzger and O. M. Yaghi, Science, 2005, 310, 1166–1170.
26 R. Dovesi, R. Orlando, B. Civalleri, C. Roetti, V. R. Saunders
and C. M. Zicovich-Wilson, Z. Kristallogr., 2005, 220, 571–573.
27 J. M. Soler, E. Artacho, J. D. Gale, A. Garcia, J. Junquera,
Notes and references
Published on 14 May 2014. Downloaded by Federal University of Minas Gerais on 07/12/2017 17:17:01.

P. Ordejon and D. Sanchez-Portal, J. Phys.: Condens. Matter,

1 L. D. Landau and E. M. Lifshitz, Statistical Physics, Part I & 2002, 14, 2745–2779.
II, Pergamon Press, Oxford, 1980. 28 D. Sanchez-Portal, P. Ordejon and E. Canadell, Principles
2 R. E. Peierls, Ann. Inst. Henri Poincare, 1935, 5, 177–222. and Applications of Density in Inorganic Chemistry II, 2004,
3 R. E. Peierls, Helv. Phys. Acta, 1934, 7, 81–83. vol. 113, 103–170.
4 N. D. Mermin, Phys. Rev., 1968, 176, 250–254. 29 G. T. Velde and E. J. Baerends, Phys. Rev. B: Condens. Matter
5 K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, Mater. Phys., 1991, 44, 7888–7903.
Y. Zhang, S. V. Dubonos, I. V. Grigorieva and A. A. Firsov, 30 G. Wiesenekker and E. J. Baerends, J. Phys.: Condens.
Science, 2004, 306, 666–669. Matter, 1991, 3, 6721–6742.
6 C. N. R. Rao, A. K. Sood, K. S. Subrahmanyam and 31 BAND2013, SCM, Theoretical Chemistry, Vrije Universiteit,
A. Govindaraj, Angew. Chem., Int. Ed., 2009, 48, 7752–7777. Amsterdam, The Netherlands.
7 V. Nicolosi, M. Chhowalla, M. G. Kanatzidis, M. S. Strano 32 P. Giannozzi, S. Baroni, N. Bonini, M. Calandra, R. Car,
and J. N. Coleman, Science, 2013, 340, 1226419. C. Cavazzoni, D. Ceresoli, G. L. Chiarotti, M. Cococcioni,
8 J. N. Coleman, M. Lotya, A. O’Neill, S. D. Bergin, P. J. King, I. Dabo, A. Dal Corso, S. de Gironcoli, S. Fabris, G. Fratesi,
U. Khan, K. Young, A. Gaucher, S. De, R. J. Smith, I. V. R. Gebauer, U. Gerstmann, C. Gougoussis, A. Kokalj,
Shvets, s. K. Arora, G. Stanton, H.-Y. Kim, K. Lee, G. T. Kim, M. Lazzeri, L. Martin-Samos, N. Marzari, F. Mauri,
G. S. Duesberg, T. Hallam, J. J. Boland, J. J. Wang, J. F. R. Mazzarello, S. Paolini, A. Pasquarello, L. Paulatto,
Donegan, J. C. Grunlan, G. Moriarty, A. Shmeliov, R. J. Nicholls, C. Sbraccia, S. Scandolo, G. Sclauzero, A. P. Seitsonen,
J. M. Perkins, E. M. Grieveson, K. Theuwissen, D. W. McComb, A. Smogunov, P. Umari and R. M. Wentzcovitch, J. Phys.:
P. D. Nellist and V. Nicolosi, Science, 2011, 331, 568–571. Condens. Matter, 2009, 21, 395502.
9 K. Ramasamy, H. Sims, W. H. Butler and A. Gupta, J. Am. 33 X. Gonze, J. M. Beuken, R. Caracas, F. Detraux, M. Fuchs,
Chem. Soc., 2014, 136, 1587–1598. G. M. Rignanese, L. Sindic, M. Verstraete, G. Zerah, F. Jollet,
10 R. F. Frindt and A. D. Yoffe, Proc. R. Soc. London, Ser. A, M. Torrent, A. Roy, M. Mikami, P. Ghosez, J. Y. Raty and
1963, 273, 69–83. D. C. Allan, Comput. Mater. Sci., 2002, 25, 478–492.
11 L. F. Mattheiss, Phys. Rev. Lett., 1973, 30, 784–787. 34 X. Gonze, G. M. Rignanese, M. Verstraete, J. M. Beuken,
12 L. F. Mattheiss, Phys. Rev. B: Solid State, 1973, 8, 3719–3740. Y. Pouillon, R. Caracas, F. Jollet, M. Torrent, G. Zerah,
13 P. Joensen, R. F. Frindt and S. R. Morrison, Mater. Res. M. Mikami, P. Ghosez, M. Veithen, J. Y. Raty, V. Olevano,
Bull., 1986, 21, 457–461. F. Bruneval, L. Reining, R. Godby, G. Onida, D. R. Hamann
14 B. Radisavljevic, A. Radenovic, J. Brivio, V. Giacometti and and D. C. Allan, Z. Kristallogr., 2005, 220, 558–562.
A. Kis, Nat. Nanotechnol., 2011, 6, 147–150. 35 X. Gonze, B. Amadon, P. M. Anglade, J. M. Beuken, F. Bottin,
15 D. Jariwala, V. K. Sangwan, L. J. Lauhon, T. J. Marks and P. Boulanger, F. Bruneval, D. Caliste, R. Caracas, M. Cote,
M. C. Hersam, ACS Nano, 2014, 8, 1102–1120. T. Deutsch, L. Genovese, P. Ghosez, M. Giantomassi,
16 J. C. Meyer, A. K. Geim, M. I. Katsnelson, K. S. Novoselov, S. Goedecker, D. R. Hamann, P. Hermet, F. Jollet, G. Jomard,
T. J. Booth and S. Roth, Nature, 2007, 446, 60–63. S. Leroux, M. Mancini, S. Mazevet, M. J. T. Oliveira, G. Onida,
17 A. Fasolino, J. H. Los and M. I. Katsnelson, Nat. Mater., Y. Pouillon, T. Rangel, G. M. Rignanese, D. Sangalli, R. Shaltaf,
2007, 6, 858–861. M. Torrent, M. J. Verstraete, G. Zerah and J. W. Zwanziger,
18 J. Brivio, D. T. L. Alexander and A. Kis, Nano Lett., 2011, 11, Comput. Phys. Commun., 2009, 180, 2582–2615.
5148–5153. 36 G. Kresse and J. Hafner, Phys. Rev. B: Condens. Matter
19 P. Miro, M. Ghorbani-Asl and T. Heine, Adv. Mater., 2013, Mater. Phys., 1993, 47, 558–561.
25, 5473–5475. 37 G. Kresse and J. Hafner, Phys. Rev. B: Condens. Matter
20 K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, Mater. Phys., 1994, 49, 14251–14269.
M. I. Katsnelson, I. V. Grigorieva, S. V. Dubonos and 38 G. Kresse and J. Furthmüller, Phys. Rev. B: Condens. Matter
A. A. Firsov, Nature, 2005, 438, 197–200. Mater. Phys., 1996, 54, 11169–11186.
21 L. F. Sun, J. X. Yan, D. Zhan, L. Liu, H. L. Hu, H. Li, 39 A. Kuc, N. Zibouche and T. Heine, Phys. Rev. B: Condens.
B. K. Tay, J. L. Kuo, C. C. Huang, D. W. Hewak, P. S. Lee Matter Mater. Phys., 2011, 83, 245213.
and Z. X. Shen, Phys. Rev. Lett., 2013, 111, 126801. 40 Y. B. Zhang, T. T. Tang, C. Girit, Z. Hao, M. C. Martin,
22 Z. Y. Zhu, Y. C. Cheng and U. Schwingenschlogl, Phys. Rev. A. Zettl, M. F. Crommie, Y. R. Shen and F. Wang, Nature,
B: Condens. Matter Mater. Phys., 2011, 84, 153402. 2009, 459, 820–823.

6552 | Chem. Soc. Rev., 2014, 43, 6537--6554 This journal is © The Royal Society of Chemistry 2014
 View Article Online

Chem Soc Rev Review Article

41 P. Miró, M. Ghorbani-Asl and T. Heine, Angew. Chem., Int. 69 B. Aufray, A. Kara, S. Vizzini, H. Oughaddou, C. Leandri,
Ed., 2014, 53, 3015–3018. B. Ealet and G. Le Lay, Appl. Phys. Lett., 2010, 96, 183102.
42 E. vanLenthe, R. vanLeeuwen, E. J. Baerends and 70 B. J. Feng, Z. J. Ding, S. Meng, Y. G. Yao, X. Y. He, P. Cheng,
J. G. Snijders, Int. J. Quantum Chem., 1996, 57, 281–293. L. Chen and K. H. Wu, Nano Lett., 2012, 12, 3507–3511.
43 E. vanLenthe, J. G. Snijders and E. J. Baerends, J. Chem. 71 H. Jamgotchian, Y. Colignon, N. Hamzaoui, B. Ealet,
Phys., 1996, 105, 6505–6516. J. Y. Hoarau, B. Aufray and J. P. Biberian, J. Phys.: Condens.
44 E. Vanlenthe, E. J. Baerends and J. G. Snijders, J. Chem. Matter, 2012, 24, 172001.
Phys., 1994, 101, 9783–9792. 72 J. F. Dienstmaier, D. D. Medina, M. Dogru, P. Knochel,
Published on 14 May 2014. Downloaded by Federal University of Minas Gerais on 07/12/2017 17:17:01.

45 E. Vanlenthe, E. J. Baerends and J. G. Snijders, J. Chem. T. Bein, W. M. Heckl and M. Lackinger, ACS Nano, 2012, 6,
Phys., 1993, 99, 4597–4610. 7234–7242.
46 E. van Lenthe, A. Ehlers and E. J. Baerends, J. Chem. Phys., 73 P. Vogt, P. De Padova, C. Quaresima, J. Avila,
1999, 110, 8943–8953. E. Frantzeskakis, M. C. Asensio, A. Resta, B. Ealet and
47 W. Koch and M. C. Holthausen, A Chemist’s Guide to G. Le Lay, Phys. Rev. Lett., 2012, 108, 155501.
Density Functional Theory, Wiley-VCH, 2001. 74 S. Cahangirov, M. Audiffred, P. Z. Tang, A. Iacomino,
48 W. Q. Li, C. F. J. Walther, A. Kuc and T. Heine, J. Chem. W. H. Duan, G. Merino and A. Rubio, Phys. Rev. B: Condens.
Theory Comput., 2013, 9, 2950–2958. Matter Mater. Phys., 2013, 88, 035432.
49 C. Adamo and V. Barone, J. Chem. Phys., 1999, 110, 75 A. Fleurence, R. Friedlein, T. Ozaki, H. Kawai, Y. Wang and
6158–6170. Y. Yamada-Takamura, Phys. Rev. Lett., 2012, 108, 245501.
50 J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett., 76 L. Meng, Y. L. Wang, L. Z. Zhang, S. X. Du, R. T. Wu,
1996, 77, 3865–3868. L. F. Li, Y. Zhang, G. Li, H. T. Zhou, W. A. Hofer and
51 J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett., H. J. Gao, Nano Lett., 2013, 13, 685–690.
1997, 78, 1396. 77 L. Liu, Y. P. Feng and Z. X. Shen, Phys. Rev. B: Condens.
52 Y. Yoon, K. Ganapathi and S. Salahuddin, Nano Lett., 2011, Matter Mater. Phys., 2003, 68, 104102.
11, 3768–3773. 78 Effective masses calculated assuming a parabolic behaviour
53 J. Heyd and G. E. Scuseria, J. Chem. Phys., 2004, 121, of top and bottom of valence and conduction bands,
1187–1192. respectively.
54 A. Ramasubramaniam, Phys. Rev. B: Condens. Matter Mater. 79 A. Pakdel, Y. Bando and D. Golberg, Chem. Soc. Rev., 2014,
Phys., 2012, 86, 115409. 43, 934–959.
55 H. L. Shi, H. Pan, Y. W. Zhang and B. I. Yakobson, Phys. 80 G. Constantinescu, A. Kuc and T. Heine, Phys. Rev. Lett.,
Rev. B: Condens. Matter Mater. Phys., 2013, 87, 155304. 2013, 111, 036104.
56 D. Y. Qiu, F. H. da Jornada and S. G. Louie, Phys. Rev. Lett., 81 H. Sahin, S. Cahangirov, M. Topsakal, E. Bekaroglu,
2013, 111, 216805. E. Akturk, R. T. Senger and S. Ciraci, Phys. Rev. B: Condens.
57 M. Kuisma, J. Ojanen, J. Enkovaara and T. T. Rantala, Phys. Matter Mater. Phys., 2009, 80, 155453.
Rev. B: Condens. Matter Mater. Phys., 2010, 82, 115106. 82 J. C. Garcia, D. B. de Lima, L. V. C. Assali and J. o. F. Justo,
58 M. Ghorbani-Asl, A. N. Enyashin, A. Kuc, T. Heine and J. Phys. Chem. C, 2011, 115, 13242–13246.
G. Seifert, Phys. Rev. B: Condens. Matter Mater. Phys., 2013, 83 P. Gori, O. Pulci, M. Marsili and F. Bechstedt, Appl. Phys.
245440–245441. Lett., 2012, 100, 261906.
59 S. Datta, Quantum Transport: Atom to Transistor, Cambridge 84 X. Y. Huang and J. Li, J. Am. Chem. Soc., 2007, 129,
University Press, New York, 2nd edn, 2005. 3157–3162.
60 D. S. Fisher and P. A. Lee, Phys. Rev. B: Condens. Matter 85 J. S. Son, X. D. Wen, J. Joo, J. Chae, S. I. Baek, K. Park,
Mater. Phys., 1981, 23, 6851–6854. J. H. Kim, K. An, J. H. Yu, S. G. Kwon, S. H. Choi,
61 M. P. L. Sancho, J. M. L. Sancho and J. Rubio, J. Phys. F: Z. W. Wang, Y. W. Kim, Y. Kuk, R. Hoffmann and
Met. Phys., 1985, 15, 851–858. T. Hyeon, Angew. Chem., Int. Ed., 2009, 48, 6861–6864.
62 A. Sengupta, D. Saha, T. A. Niehaus and S. Mahapatra, 86 S. Ithurria, M. D. Tessier, B. Mahler, R. P. S. M. Lobo,
2014, arXiv:1401.4553. B. Dubertret and A. Efros, Nat. Mater., 2011, 10, 936–941.
63 H. P. Boehm, A. Clauss, G. O. Fischer and U. Hofmann, Z. 87 C. Tusche, H. L. Meyerheim and J. Kirschner, Phys. Rev.
Anorg. Chem., 1962, 316, 119–127. Lett., 2007, 99, 026102.
64 P. R. Wallace, Phys. Rev., 1947, 71, 622–634. 88 G. Weirum, G. Barcaro, A. Fortunelli, F. Weber,
65 A. K. Geim and K. S. Novoselov, Nat. Mater., 2007, 6, 183–191. R. Schennach, S. Surnev and F. P. Netzer, J. Phys. Chem.
66 Q. Peng, J. Crean, A. K. Dearden, C. Huang, X. D. Wen, C, 2010, 114, 15432–15439.
S. P. A. Bordas and S. De, Mod. Phys. Lett. B, 2013, 89 J. Zhou, J. S. Huang, B. G. Sumpter, P. R. C. Kent,
27, 1330017. H. Terrones and S. C. Smith, J. Phys. Chem. C, 2013,
67 X. D. Wen, T. J. Cahill and R. Hoffmann, Chem. – Eur. J., 117, 26858.
2010, 16, 6555–6566. 90 J. Zhou, J. S. Huang, B. G. Sumpter, P. R. C. Kent,
68 S. Cahangirov, M. Topsakal, E. Akturk, H. Sahin and H. Terrones and S. C. Smith, J. Phys. Chem. C, 2013, 117,
S. Ciraci, Phys. Rev. Lett., 2009, 102, 236804. 25817–25825.

This journal is © The Royal Society of Chemistry 2014 Chem. Soc. Rev., 2014, 43, 6537--6554 | 6553
 View Article Online

Review Article Chem Soc Rev

91 E. Bianco, S. Butler, S. S. Jiang, O. D. Restrepo, W. Windl 112 Z. M. Wang, MoS2 Materials, Physics, and Devices, Springer,
and J. E. Goldberger, ACS Nano, 2013, 7, 4414–4421. 2014.
92 K. J. Koski and Y. Cui, ACS Nano, 2013, 7, 3739–3743. 113 O. Lopez-Sanchez, D. Lembke, M. Kayci, A. Radenovic and
93 R. R. Nair, W. C. Ren, R. Jalil, I. Riaz, V. G. Kravets, L. Britnell, A. Kis, Nat. Nanotechnol., 2013, 8, 497–501.
P. Blake, F. Schedin, A. S. Mayorov, S. J. Yuan, M. I. Katsnelson, 114 J. H. Han, S. Lee and J. Cheon, Chem. Soc. Rev., 2013, 42,
H. M. Cheng, W. Strupinski, L. G. Bulusheva, A. V. Okotrub, 2581–2591.
I. V. Grigorieva, A. N. Grigorenko, K. S. Novoselov and 115 S. Jeong, D. Yoo, J. T. Jang, M. Kim and J. Cheon, J. Am.
A. K. Geim, Small, 2010, 6, 2877–2884. Chem. Soc., 2012, 134, 18233–18236.
Published on 14 May 2014. Downloaded by Federal University of Minas Gerais on 07/12/2017 17:17:01.

94 D. K. Samarakoon, Z. F. Chen, C. Nicolas and X. Q. Wang, 116 J. T. Jang, S. Jeong, J. W. Seo, M. C. Kim, E. Sim, Y. Oh,
Small, 2011, 7, 965–969. S. Nam, B. Park and J. Cheon, J. Am. Chem. Soc., 2011, 133,
95 Y. Ding and Y. L. Wang, Appl. Phys. Lett., 2012, 100, 083102. 7636–7639.
96 M. Chhowalla, H. S. Shin, G. Eda, L. J. Li, K. P. Loh and 117 S. Jeong, J. H. Han, J. T. Jang, J. W. Seo, J. G. Kim and
H. Zhang, Nat. Chem., 2013, 5, 263–275. J. Cheon, J. Am. Chem. Soc., 2011, 133, 14500–14503.
97 F. L. Claus, Solid Lubricants and Self-Lubricating Solids, 118 S. Furuseth, K. Selte and A. Kjekshus, Acta Chem. Scand.,
Academic Press, New York, 1972. 1965, 19, 257.
98 C. Ataca, H. Sahin and S. Ciraci, J. Phys. Chem. C, 2012, 116, 119 Y. D. Ma, Y. Dai, M. Guo, L. Yu and B. B. Huang,
8983–8999. Phys. Chem. Chem. Phys., 2013, 15, 7098–7105.
99 N. Zibouche, P. Philipsen, A. Kuc and T. Heine, Phys. Chem. 120 H. Hillebrecht, P. J. Schmidt, H. W. Rotter, G. Thiele,
Chem. Phys., 2014, DOI: 10.1039/C4CP00966E. P. Zonnchen, H. Bengel, H. J. Cantow, S. N. Magonov
100 A. Splendiani, L. Sun, Y. Zhang, T. Li, J. Kim, C.-Y. Chim, and M. H. Whangbo, J. Alloys Compd., 1997, 246, 70–79.
G. Galli and F. Wang, Nano Lett., 2010, 10, 1271–1275. 121 D. Shriver and P. Atkins, Inorganic Chemistry, W. H.
101 G. Eda, H. Yamaguchi, D. Voiry, T. Fujita, M. Chen and Freeman, 2009.
M. Chhowalla, Nano Lett., 2011, 11, 5111–5116. 122 S. Kandambeth, D. B. Shinde, M. K. Panda, B. Lukose,
102 H. R. Gutiérrez, N. Perea-López, A. L. Elı́as, A. Berkdemir, T. Heine and R. Banerjee, Angew. Chem., Int. Ed., 2013, 52,
B. Wang, R. Lv, F. López-Urı́as, V. H. Crespi, H. Terrones 13052–13056.
and M. Terrones, Nano Lett., 2012, 13, 3447–3454. 123 S. Chandra, S. Kandambeth, B. P. Biswal, B. Lukose,
103 S. Bertolazzi, D. Krasnozhon and A. Kis, ACS Nano, 2013, 7, S. M. Kunjir, M. Chaudhary, R. Babarao, T. Heine and
3246–3252. R. Banerjee, J. Am. Chem. Soc., 2013, 135, 17853–17861.
104 B. Radisavljevic and A. Kis, Nat. Mater., 2013, 12, 815–820. 124 B. P. Biswal, S. Chandra, S. Kandambeth, B. Lukose, T. Heine
105 H. Y. Chang, S. X. Yang, J. H. Lee, L. Tao, W. S. Hwang, and R. Banerjeet, J. Am. Chem. Soc., 2013, 135, 5328–5331.
D. Jena, N. S. Lu and D. Akinwande, ACS Nano, 2013, 7, 125 S. Kandambeth, A. Mallick, B. Lukose, M. V. Mane,
5446–5452. T. Heine and R. Banerjee, J. Am. Chem. Soc., 2012, 134,
106 G. H. Lee, Y. J. Yu, X. Cui, N. Petrone, C. H. Lee, M. S. Choi, 19524–19527.
D. Y. Lee, C. Lee, W. J. Yoo, K. Watanabe, T. Taniguchi, 126 B. Lukose, A. Kuc and T. Heine, Chem. – Eur. J., 2011, 17,
C. Nuckolls, P. Kim and J. Hone, ACS Nano, 2013, 7, 2388–2392.
7931–7936. 127 L. R. Xu, X. Zhou, Y. X. Yu, W. Q. Tian, J. Ma and S. B. Lei,
107 J. Yoon, W. Park, G. Y. Bae, Y. Kim, H. S. Jang, Y. Hyun, ACS Nano, 2013, 7, 8066–8073.
S. K. Lim, Y. H. Kahng, W. K. Hong, B. H. Lee and H. C. Ko, 128 B. Lukose, A. Kuc, J. Frenzel and T. Heine, Beilstein
Small, 2013, 9, 3295–3300. J. Nanotechnol., 2010, 1, 60–70.
108 N. Scheuschner, O. Ochedowski, M. Schleberger and 129 J. V. Lauritsen, J. Kibsgaard, S. Helveg, H. Topsoe,
J. Maultzsch, Phys. Status Solidi B, 2012, 249, 2644–2647. B. S. Clausen, E. Laegsgaard and F. Besenbacher, Nat.
109 S. Bertolazzi, J. Brivio and A. Kis, ACS Nano, 2011, 5, Nanotechnol., 2007, 2, 53–58.
9703–9709. 130 S. H. Vosko, L. Wilk and M. Nusair, Can. J. Phys., 1980, 58,
110 M. Ghorbani-Asl, S. Borini, A. Kuc and T. Heine, Phys. Rev. 1200–1211.
B: Condens. Matter Mater. Phys., 2013, 87, 235434. 131 Y. Zhao and D. G. Truhlar, J. Chem. Phys., 2006, 125, 194101.
111 K. Roy, M. Padmanabhan, S. Goswami, T. P. Sai, 132 C. Bouet, M. D. Tessier, S. Ithurria, B. Mahler, B. Nadal and
G. Ramalingam, S. Raghavan and A. Ghosh, Nat. Nanotechnol., B. Dubertret, Chem. Mater., 2013, 24, 1262–1271, DOI:
2013, 8, 826–830. 10.1021/cm303786a.

6554 | Chem. Soc. Rev., 2014, 43, 6537--6554 This journal is © The Royal Society of Chemistry 2014