Professional Documents
Culture Documents
Mengxi Liu
Controlled Synthesis
and Scanning
Tunneling Microscopy
Study of Graphene
and Graphene-Based
Heterostructures
Springer Theses
The series “Springer Theses” brings together a selection of the very best Ph.D.
theses from around the world and across the physical sciences. Nominated and
endorsed by two recognized specialists, each published volume has been selected
for its scientific excellence and the high impact of its contents for the pertinent field
of research. For greater accessibility to non-specialists, the published versions
include an extended introduction, as well as a foreword by the student’s supervisor
explaining the special relevance of the work for the field. As a whole, the series will
provide a valuable resource both for newcomers to the research fields described,
and for other scientists seeking detailed background information on special
questions. Finally, it provides an accredited documentation of the valuable
contributions made by today’s younger generation of scientists.
Controlled Synthesis
and Scanning Tunneling
Microscopy Study
of Graphene
and Graphene-Based
Heterostructures
Doctoral Thesis accepted by
Peking University, Beijing, China
123
Author Supervisors
Dr. Mengxi Liu Prof. Zhongfan Liu
Center for Nanochemistry Center for Nanochemistry
College of Chemistry and Molecular College of Chemistry and Molecular
Engineering, Peking University Engineering, Peking University
Beijing Beijing
People’s Republic of China People’s Republic of China
Since the Nobel Prize for physics in 2010 was awarded to Prof. Geim and Prof.
Novoselov for their ‘ground-breaking experiments regarding the two-dimensional
material graphene,’ dramatic research progress in graphene has been witnessed,
such as the controllable synthesis, the deep understanding of the chemical and
physical properties, and the broad application prospects. The unique electronic
property of high carrier mobility makes graphene a potential material in future
electronics. However, the Dirac-type band structure of graphene with zero bandgap
limits its application in electronic devices. It is of great importance to develop
techniques to open a gap in the band structure of graphene.
This thesis provides a series of experimental work in fabricating well-designed
nanostructures to modulate the band structure of graphene, involving vertically
coupled bilayer graphene and lateral heterostructure with the ‘white graphene’
hexagonal boron nitride (h-BN). It shows how to synthesize bilayer graphene with
different interlayer rotation angles on Rh substrates via a combined chemical vapor
deposition and segregation process. The obtained bilayer graphene generates dif-
ferent moiré patterns and thus induces van Hove singularities near Fermi level,
which is observed by atomic-resolved scanning tunneling microscopy/spectroscopy.
In addition, it puts forward an efficient way to modulate the bandgap of graphene by
creating a heterostructure with h-BN. It shows how to control the domain shape,
domain size and near 100% zigzag-dominated boundary linking type of
heterostructures. The mechanism of bandgap opening of h-BN-G heterostructure is
discussed based on experimental results and theoretical calculations.
v
Abstract
vii
viii Abstract
(graphene: h-BN = 1:7), which was a challenge for the two analogues linking
together to form a seamless monolayer. By using a two-step growth method,
two kinds of h-BN-graphene in-plane heterostructures, h-BN@G heterostruc-
ture and G@h-BN heterostructure, were synthesized on Ir(111) substrate.
Graphene and h-BN could overcome the differences in layer corrugation and
form in-plane monolayer heterostructure. The grain sizes of graphene and h-BN
could be controlled by tuning the growth temperature.
(2) Atomic Structures and Electronic States on the Interfaces of the
Heterostructures
An atomically sharp zigzag-type boundary has been found to dominate the
patching interface between graphene and h-BN, as evidenced by
high-resolution scanning tunneling microscopy. The formation of zigzag-type
boundaries simply related to the edge of the pre-deposited graphene or h-BN
domains, excluding the effects from growth sequences, domain sizes, domain
shapes, and substrates. Scanning tunneling spectroscopy studies showed that
the graphene and h-BN tended to exhibit their intrinsic electronic structures
near the patching boundary with no doping effects from each other. It suggests
that h-BN behaves as an insulator in the heterostructure and does not change the
intrinsic electronic structure of graphene. However, the existing of h-BN
domains presents important influences on the electronic structure of the
heterostructure. Firstly, the AB-symmetry of graphene lattices was broken
because of the appearance of h-BN domains, leading to the bandgap opening of
graphene. Secondly, the seamless boundaries between graphene and h-BN
domains decreased the electronic scattering of graphene edges, which con-
firmed the high carrier mobility of graphene. The present work offers a deep
insight into the h-BN-graphene heterostructures both geometrically and elec-
tronically together with the graphene energy band engineering.
Keywords Graphene Segregated growth van Hove singularity h-BN-graphene
heterostructure Energy band engineering Scanning tunneling microscopy
Publication List
1. M Liu, Y Zhang, Y Chen, Y Gao, T Gao, D Ma, Q Ji, Y Zhang, C Li, Z Liu.
Thinning Segregated Graphene Layers on High Carbon Solubility Substrates of
Rhodium Foils by Tuning the Quenching Process. ACS Nano 6 (12), 10581–
10589, (2012).
2. M Liu, Y Gao, Y Zhang, Y Zhang, D Ma, Q Ji, T Gao, Y Chen, Z Liu. Single
and Polycrystalline Graphene on Rh (111) Following Different Growth
Mechanisms. Small 9 (8), 1360–1366, (2013).
3. M Liu, Y Li, P Chen, J Sun, D Ma, Q Li, T Gao, Y Gao, Z Cheng, X Qiu, Y
Fang, Y Zhang, Z Liu. Quasi-Freestanding Monolayer Heterostructure of
Graphene and Hexagonal Boron Nitride on Ir (111) with a Zigzag Boundary.
Nano Lett. 14 (11), 6342–6347, (2014).
4. M Liu, Y Zhang, Z Liu. Scanning Tunneling Micoscopy Study of in-Plane
Graphene-Hexagonal Boron Nitride Heterostructures. Acta. Phys. Sin. 64
(7) 078101, (2015).
5. W Yan,* M Liu,* RF Dou, L Meng, L Feng, ZD Chu, Y Zhang, Z Liu, JC Nie,
L He. Angle-Dependent van Hove Singularities in a Slightly Twisted Graphene
Bilayer. Phys. Rev. Lett. 109 (12), 126801, (2012).
6. D Ma, M Liu, T Gao, C Li, J Sun, Y Nie, Q Ji, Y Zhang, X Song, Y Zhang, Z
Liu. High-Quality Monolayer Graphene Synthesis on Pd Foils via the
Suppression of Multilayer Growth at Grain Boundaries. Small 10 (19), 4003–
4011, (2014).
7. Y Chen, Y Hu, M Liu, W Xu, Y Zhang, L Xie, J Zhang. Chiral Structure
Determination of Aligned Single-Walled Carbon Nanotubes on Graphite
Surface. Nano Lett. 13 (11), 5666–5671, (2013).
8. D Ma, Y Zhang, M Liu, Q Ji, T Gao, Y Zhang, Z Liu. Clean Transfer of
Graphene on Pt Foils Mediated by a Carbon Monoxide Intercalation Process.
Nano Res. 6 (9), 671–678, (2013).
9. W Yan, L Meng, M Liu, JB Qiao, ZD Chu, RF Dou, Z Liu, JC Nie, L He.
Angle-Dependent van Hove Singularities and Their Breakdown in Twisted
Graphene Bilayers. arXiv preprint arXiv:1404.4406., (2014).
xi
xii Publication List
10. W Yan, WY He, ZD Chu, M Liu, L Meng, RF Dou, Y Zhang, Z Liu, JC Nie, L
He. Strain And Curvature Induced Evolution of Electronic Band Structures in
Twisted Graphene Bilayer. Nat. Commun. 4, (2013).
11. T Gao, S Xie, Y Gao, M Liu, Y Chen, Y Zhang, Z Liu. Growth And
Atomic-Scale Characterizations of Graphene on Multifaceted Textured Pt Foils
Prepared by Chemical Vapor Deposition. ACS Nano 5 (11), 9194–9201,
(2011).
12. L Meng, WY He, H Zheng, M Liu, H Yan, W Yan, ZD Chu, K Bai, RF Dou, Y
Zhang, Z Liu, JC Nie, L He. Strain-Induced One-Dimensional Landau Level
Quantization in Corrugated Graphene. Phys. Rev. B 87 (20), 205405, (2013).
13. H Yan, ZD Chu, W Yan, M Liu, L Meng, M Yang, Y Fan, J Wang, RF Dou, Y
Zhang, Z Liu, JC Nie, L He. Superlattice Dirac Points And Space-Dependent
Fermi Velocity in a Corrugated Graphene Monolayer. Phys. Rev. B 87 (7),
075405, (2013).
14. Y Gao, Y Zhang, P Chen, Y Li, M Liu, T Gao, D Ma, Y Chen, Z Cheng, X
Qiu, W Duan, Z Liu. Toward Single-layer Uniform Hexagonal Boron Nitride–
Graphene Patchworks with Zigzag Linking Edges. Nano Lett. 13 (7), 3439–
3443, (2013).
15. T Gao, Y Gao, C Chang, Y Chen, M Liu, S Xie, K He, X Ma, Y Zhang, Z Liu.
Atomic-Scale Morphology And Electronic Structure of Manganese Atomic
Layers Underneath Epitaxial Graphene on SiC (0001). ACS Nano 6 (8), 6562–
6568, (2012).
16. H Yan, CC Liu, KK Bai, X Wang, M Liu, W Yan, L Meng, Y Zhang, Z Liu, R
Dou, J Nie, Y Yao, L He. Electronic Structures of Graphene Layers on a Metal
Foil: The Effect of Atomic-Scale Defects. Appl. Phys. Lett. 103 (14), 143120,
(2013).
17. Q Ji, Y Zhang, T Gao, Y Zhang, D Ma, M Liu, Y Chen, X Qiao, PH Tan, M
Kan, J Feng, Q Sun, Z Liu. Epitaxial Monolayer MoS2 on Mica With Novel
Photoluminescence, Nano Lett. 13 (8), 3870–3877, (2013).
18. Y Zhang, Y Zhang, D Ma, Q Ji, W Fang, J Shi, T Gao, M Liu, Y Gao, Y Chen,
L Xu, Z Liu. Mn Atomic Layers Under Inert Covers of Graphene And
Hexagonal Boron Nitride Prepared on Rh (111). Nano Res. 6 (12), 887–896,
1360–1366, (2013).
19. Y Zhang, T Gao, S Xie, B Dai, L Fu, Y Gao, Y Chen, M Liu, Z Liu. Different
Growth Behaviors of Ambient Pressure Chemical Vapor Deposition Graphene
on Ni (111) And Ni Films: A Scanning Tunneling Microscopy Study. Nano
Res. 5 (6), 402–411, (2012).
20. Y Zhang, Y Zhang, Q Ji, J Ju, H Yuan, J Shi, T Gao, D Ma, M Liu, Y Chen, X
Song, HY Hwang, Y Cui, Z Liu. Controlled Growth of High-Quality
Monolayer WS2 Layers on Sapphire And Imaging Its Grain Boundary. ACS
Nano 7 (10), 8963–8971, (2013).
Acknowledgements
The research was supported by the Ministry of Science and Technology of China
(973 program 2011CB921903, 2011CB933003, 2012CB933404, 2012CB921404
and 2013CB932603) and the National Natural Science Foundation of China (NSFC
21603045, 51290272, 21073003, 51222201, 51121091, and 11304053).
The investigations described in this thesis were mostly carried out in the Center
for Nanochemistry, Peking University (PKU). Some investigations were carried out
in National Center for Nanoscience and Technology (NCNST) and Beijing Normal
University (BNU).
I would like to thank the following people who have contributed to this thesis:
• Prof. Zhongfan Liu and Prof. Yanfeng Zhang for supervising and guiding me on
this thesis. Thank you for leading me to the magic world of STM and graphene.
• Prof. Jin Zhang, Prof. Hailin Peng, and Dr. Xie, who give us the nice research
environment in CNC, Peking University.
• Prof. Lin He and Dr. Wei Yan in Beijing Normal University, who carried out the
STS measurement presented in Chap. 3.
• Prof. Yuanchang Li, who carried out the DFT calculations presented in Chap. 4.
• Prof. Xiaohui Qiu and Dr. Pengcheng Chen for offering me the opportunity to
do the STS measurement of h-BN-G interface in National Center for
Nanoscience and Technology of China. Now, I am an Assistant Professor
working with Prof. Qiu in NCNST!
• Dr. Yabo Gao, Dr. Teng Gao, who taught me how to use STM and other
instruments.
• Dr. Jingyu Sun and Dr. Face for sharing attitude on scientific research.
• Dr. Qiucheng Li, Dr. Yue Qi, Dr. Xiebo Zhou, and Dr. Jianping Shi.
• My parents and parents-in-law for their powerful support!
• Dr. Donglin Ma, my soul mate and my husband. Thank you for…all!
xiii
Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 The Properties of Graphene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.1.1 The Structure of Graphene . . . . . . . . . . . . . . . . . . . . . . . . 2
1.1.2 The Band Structure of Graphene . . . . . . . . . . . . . . . . . . . 3
1.2 Synthesis of Graphene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.2.1 The Synthesis Methods of Graphene . . . . . . . . . . . . . . . . . 6
1.2.2 The Synthesis Methods of Bilayer Graphene . . . . . . . . . . . 8
1.3 Bandgap Opening of Graphene . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.3.1 Physical Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.3.2 Chemical Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.4 Research Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2 Controllable Synthesis of Graphene on Rh . . . . . . . . . . . . . . . . . . . . 19
2.1 Experiment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.1.1 Process Engineering for Graphene Growth . . . . . . . . . . . . 20
2.1.2 Factors for Graphene Segregation . . . . . . . . . . . . . . . . . . . 20
2.1.3 Substrate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.1.4 Experimental System . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.2 Controllable Growth of Graphene on Rh Substrates . . . . . . . . . . . 26
2.3 The Mechanism of Graphene Growth on Rh Substrates . . . . . . . . 32
2.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3 STM Study of Twisted Bilayer Graphene . . . . . . . . . . . . . . . . . . . . . 37
3.1 Introduction to STM Study of Graphene . . . . . . . . . . . . . . . . . . . 37
3.1.1 The Working Principle of STM . . . . . . . . . . . . . . . . . . . . 38
3.1.2 The Working Principle of STS . . . . . . . . . . . . . . . . . . . . . 40
3.1.3 Applications of STM/STS Technique in Graphene
Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... 41
xv
xvi Contents
Carbon (C) is a chemical element, which is capable of forming many allotropes due
to the various hybridized state. The well-known carbon allotropes include diamond
(sp3) and graphite (sp2). In recent decades, many new carbon allotropes have been
discovered, such as fullerene (0D), carbon nanotube (1D) and graphene (2D), which
present a series of excellent electrical, optical and mechanical properties and open
the door to carbon nanomaterials [1, 34, 35].
Graphene is a two dimensional materials of sp2 hybridization carbon atoms. The sp2
orbital hybridization is combined of s, px and py orbitals that constitute the r-bond,
while the final pz electron makes up the p-bond. The p-bonds hybridize together to
form the p-band and p-band [3].
Graphene is a single graphitic layer which can be regarded as the basic building
blocks of many carbon allotropes. When the graphene layer forms a sphere, it is
well known as fullerene; when it is rolled up with respect to its axis, it forms a
carbon nanotube; when the graphene layers are stacked and bound by a weak van
der Waals force, it forms bulk graphite (Fig. 1.1a) [5]. Figure 1.1b is an scanning
transmission electron microscopy (STEM) image of graphene, which shows hon-
eycomb atomic lattices with a lattice constant of 0.246 nm. There are two sublat-
tices of the honeycomb lattice, named A and B.
Fig. 1.1 The atomic structure of graphene a Graphene is a 2D building material for carbon
material of all other dimensionalities. It can be wrapped up into 0D buckyballs, rolled into 1D
nanotubes or stacked into 3D graphite. Reproduced with permission of [5]. Copyright [2007],
Rights managed by Nature Publishing Group. b Scanning transmission electron microscope
(STEM) image of graphene. Reproduced with permission of [36]. Copyright [2011], Rights
managed by Nature Publishing Group
1.1 The Properties of Graphene 3
The band structure of graphene depends on the number of layers and stacking orders.
(1) Monolayer graphene
For monolayer graphene, the atomic structure of the real space and the first
Brillouin zone of the reciprocal space are shown in Fig. 1.1. According to the tight
binding calculations, the real space lattice vectors a1, a2 and the reciprocal lattice
vectors b1, b2 of the graphene can be written as:
a pffiffiffi a pffiffiffi
a1 ¼ ð3; 3Þ; a2 ¼ ð3; 3Þ ð1:1Þ
2 2
2p pffiffiffi 2p pffiffiffi
b1 ¼ ð1; 3Þ; b2 ¼ ð1; 3Þ ð1:2Þ
3a 3a
The two points K and K′ at the corners of the graphene Brillouin zone are named
Dirac points. Their positions in momentum space are given by:
2p 2p 2p 2p
K¼ ; pffiffiffi ; K0 ¼ ; pffiffiffi ð1:3Þ
3a 3 3a 3a 3 3a
The full band structure of graphene is shown in Fig. 1.2c, as well as the zoom-in
of the band structure close to one of the Dirac points (K or K′ point in the Brillouin
zone). The conduction band and valence band of graphene meet at the Dirac points
and the energy and momentum of monolayer graphene have a linear dispersion
relationship near the Fermi level. Electrons propagating through graphene’s hon-
eycomb lattice effectively lose their mass, which can be regarded as quasi-particles.
Thus graphene displays remarkable carrier mobility at room temperature, with
reported values in excess of 2 105 cm2 V−1 s−1 [5].
(2) AB-stacked bilayer graphene
Bilayer graphene can exist in two stacking orders: AB-stacking and non-AB
stacking. The AB-stacking order of bilayer graphene is half of the carbon atoms lie
directly over the center of a hexagon in the low graphene sheet, and half of the
carbon atoms lie over carbon atoms of the low graphene sheet (Fig. 1.3a). The band
structure of AB-stacked bilayer graphene is contributed by the coupled two layers
of graphene. The first Brillouin zone in reciprocal space of AB-stacked bilayer
graphene shows a hexagonal shape with two symmetrical valleys. Like monolayer
graphene, the AB-stacked bilayer graphene has a zero bandgap. It is predicted that a
bandgap could be introduced if an electric displacement field was applied to the
bilayer graphene [37]. An experimental result of the tunable bandgap in AB-stacked
bilayer graphene was given in 2009, confirming the prediction [20].
4 1 Introduction
Fig. 1.2 The band structure of monolayer graphene a The real-space atomic structure of
graphene. b The first Brillouin zone of the reciprocal space for graphene. c Electronic dispersion in
the honeycomb lattice. Reproduced with permission of [3]. Copyright [2009], American Physical
Society
Fig. 1.3 The band structure of AB-stacked bilayer graphene a The atomic structure of
AB-stacked bilayer graphene. b The first Brillouin zone in reciprocal space of AB-stacked
bilayer graphene. Reproduced with permission of [3]. Copyright [2009], American Physical
Society
be magnified by the enhanced DOS, resulting in instabilities, which can give rise to
new phases of matter, such as superconductivity, magnetism or charge density
waves (CDW). This implies the possibility of engineering material properties by
bringing EF and the VHS together [38–41]. For most situations, it is difficult to
change the position of VHS in band structure. In this regard, graphene is very
special. Introducing a rotation between stacked bilayer graphene makes is possible
to induce VHSs which are near EF. As shown in Fig. 1.4a and b, the rotation (h)
between two graphene layers causes a shift between the corresponding Dirac points
in momentum space, as a result, the Dirac wave vector of the rotated layer is
Kh = K + △K, where △K = K 2 sin(h/2). The two Dirac cones intersect near
the center of the superlattice Brillouin zone and hybridize (Fig. 1.4c), which results
in two saddle points in the energy dispersion, acting as two symmetric VHSs in
DOS (Fig. 1.4d).
It is important to note that the VHSs can form only in the presence of finite
interlayer coupling. For the case of a detached graphene layer on highly oriented
pyrolytic graphite (HOPG), the VHSs will not form even though moiré patterns
may be visible in STM images [42, 43]. In contrast, for the as-grown twisted bilayer
Fig. 1.4 The band structure of non-AB-stacked bilayer graphene. a The atomic structure in real
space and b the first Brillouin zone in reciprocal space of AB-stacked bilayer graphene. c The
intersection of Dirac point near the center of the superlattice Brillouin zone. d The VHSs near EF
in DOS. Reproduced with permission of [21]. Copyright [2009], Rights managed by Nature
Publishing Group
6 1 Introduction
graphene, both the position of EF and VHSs can be easily controlled by gating and
rotation angles, which provide a powerful toolkit for the manipulation of electronic
structures [44–46].
In recent years, various techniques have been established for synthesizing graphene,
such as, mechanical exfoliation, chemical exfoliation, chemical vapor deposition
(CVD) and so on.
reduction with hydroiodic acid creates “reduced graphene oxide” (r-GO) [48].
Similar to the graphene prepared by chemical exfoliation method, r-GO is low-cost
and easy to be obtained. Because of the abundant oxygenic groups, r-GO has a
good dispersion in solution. Moreover, r-GO is suitable for large-scale production
and being used in transparent conductive film, energy storage and conductive paper.
The shortcomings of r-GO is small-size, uneven thickness and poor conductivity.
(3) Chemical vapor deposition
Chemical vapor deposition (CVD) is a chemical process used to produce high
quality materials. In typical CVD process, the substrate is exposed to one or more
precursors, which react and/or decompose on the substrate surface to produce the
desired materials. Since the R.S. Ruoff group first synthesized a uniform graphene
film of centimeter size on the surface of Cu foil using CVD, it opened a new era of
graphene preparation [49–56]. The most popular carbon source and substrate used
in graphene CVD-preparation process is CH4 and Cu foils.
Since self-limiting growth of graphene on Cu foil, the prepared graphene is
predominantly monolayer and the quality can be comparable to that of mechani-
cally exfoliated graphene. By controlling the growth temperature and H2/CH4 ratio
during the growth process, it is possible to control the domain shape and single-
crystalline domain size of graphene. At present, the maximum domain size of
singlecrystalline graphene grown on Cu foil can reach 1 cm [57]. In addition to Cu
foil, scientists have developed a variety of new substrates suitable for the growth of
graphene. The noble metal Pt foil is suitable to prepare large-size singlecrystalline
graphene at low H2/CH4 ratio; [58] Ni-Mo alloy is suitable to prepare 100%
monolayer graphene [59]; graphene synthesized on SrTiO3 can be directly used for
transport measurement [60]; graphene synthesized on SiO2 surface can be directly
used for electronic devices [61].
(4) Graphene segregation
Segregation process refers to the enrichment of atoms, ions, or molecules on the
surface. Segregation as an important concept in the field of material growth has
been discovered early. Carbon segregation can be traced back to 1965: the low
energy electron diffraction (LEED) spots of Pt(100) singlecrystal show circle pat-
tern after high temperature annealing, instead of the original square pattern, which
is attributed to segregated graphite on surface [62]. Segregated graphene can be
prepared by the use of metals with high carbon solubility such as Ni, Co [63–65].
The segregation growth method is simple and low-cost to prepare large-size gra-
phene film, but the thickness is not easy to control.
Based on the introduction of the above-mentioned methods for the preparation of
graphene, CVD and segregation are the most suitable methods for preparing
large-size, high-quality graphene.
8 1 Introduction
Bilayer graphene has attracted much attention from scientists because of its special
energy band structure and excellent electrical properties. The early preparation of
bilayer graphene mostly used mechanical exfoliation method. However, the exfo-
liated bilayer graphene was hard to be controlled, which was a big challenge for the
deep investigation. The following are two more controllable methods for the
preparation of bilayer graphene.
(1) Top-down method
Since self-limiting growth of graphene on Cu foil, the prepared graphene is
predominantly monolayer. In 2011, Yan et al. [66] reported an approach to syn-
thesize bilayer graphene by a long-range catalytic process on Cu foil. The bilayer
graphene is formed through the combination of an existing monolayer and another
deposited monolayer by a two-step growth scheme in a tube furnace. The mono-
layer graphene serves as the substrate placed downstream at 1000 °C for the fol-
lowing second-step growth. Another piece of fresh copper foil is placed upstream
where the temperature is 1040 °C, as an efficient catalyst to continuously decom-
pose CH4, thus carbon radicals and small graphene fragments are transported
downstream and deposited on the existing monolayer graphene. The coverage of
bilayer graphene can reach 67%. Transmission electron microscope (TEM) and
selected area electron diffraction (SAED) characterizations show that the as-grown
bilayer graphene is AB-stacking, which confirms the epitaxial growth mechanism.
(2) Bottom-up method
As mentioned above, the segregation growth method can synthesize graphene
with different thickness, from monolayer to multi-layer, thus there is a large win-
dow to control the graphene layer and uniformity. The previous literature shows
that 50% of graphene prepared on 300 nm Ni films is bilayer graphene [64, 65].
SAED characterization shows that the as-grown bilayer graphene is
non-AB-stacking. Combined with the mechanism of graphene segregation growth,
there is no epitaxial relationship between the graphene bilayers, and the orientation
of the graphene is only related to the lattice orientation of the nucleated graphene
flakes. Therefore, it is expected to produce large-scale uniform non-AB-stacking
graphene on high carbon solubility metal substrate via segregation process.
A great deal of effort is required to develop techniques for modifying the band
structure of graphene to make it a potential replacement for silicon in future elec-
tronics. The following summarizes the methods for bandgap opening of graphene.
1.3 Bandgap Opening of Graphene 9
Fig. 1.5 External electric field induced bandgap opening of AB-stacked bilayer graphene.
a Optical microscopy image of the electronic device of bilayer graphene (top view). b Illustration
of a cross-sectional side view of the gated device. c Sketch showing how gating of the bilayer
induces top (Dt) and bottom (Db) electrical displacement fields. d Left, the electronic structure of a
pristine bilayer has zero bandgap. Right, upon gating, the displacement fields induces a non-zero
bandgap and a shift of the Fermi energy EF. e Gate-induced absorption spectra for different applied
displacement fields. f Theoretical prediction of the gate induced absorption spectra based on a
tight-binding model where the bandgap value is taken as an adjustable parameter. Reproduced with
permission of [20]. Copyright [2009], Rights managed by Nature Publishing Group
10 1 Introduction
Fig. 1.6 Substrate-induced bandgap opening in epitaxial graphene on SiC. a–d ARPES intensity maps taken on monolayer graphene on 6H-SiC, bilayer
graphene on 4H-SiC, trilayer graphene on 6H-SiC and graphite, respectively e ED and bandgap as a function of sample thickness. Reproduced with permission
of [24]. Copyright [2007], Rights managed by Nature Publishing Group
11
12 1 Introduction
adjacent two holes. In 2010, Duan et al. [12] reported the preparation of GNMs
using conventional block copolymer lithography. The period of GNM can reach
27 nm and the thinnest neck can be 5 nm. The FET based on the as-synthesized
GNM show comparable on/off ratio with that of GNRs.
Compared with the external electric field method and the substrate-induced
method, the quantum confinement method does not depend on the specific electric
field and substrate, thus can be used more widespread. However, the etching pro-
cess for preparing the quantum-confined nanostructures will produce defects and
irregular edges, resulting in a significant decline in carrier mobility, which is not
conducive to the application of graphene in electronic devices. In addition,
according to the tight-binding models and first-principles calculations [70, 71], the
zigzag-edge GNRs have no bandgap, meanwhile only two-thirds of the
armchair-edge GNRs have the possibility to open the bandgap. The situation of
GNMs is more complicated. In short, it is difficult to achieve the perfect model
structure predicted by theoretical calculations, which makes the application of
GNRs and GNM limited.
Fig. 1.7 a, b Monte Carlo simulation of monolayer composed of B, C and N atoms [33].
Reproduced with permission of [33]. Copyright [2011] American Chemical Society
Since graphene and h-BN have similar atomic structures and lattice constants,
the two analogues may form a monolayer atomically. The introduction of h-BN
would change the band structure of graphene. As shown in Fig. 1.7a, b, Chacham
et al. [33] used Monte Carlo to simulate the monolayer structure composed of B, C
and N atoms. They found that the monolayer structure tended to phase separation
and form individual graphene and h-BN domains. With the increase of phase
separation, the bandgap of hybrid structure also shows an increasing trend.
Besides, for the periodical h-BN @ G structure, the bandgap of the graphene can
be opened regardless of the shape and edge of the graphene or h-BN domains
[74–77]. If the number of B and N atoms in the h-BN domain is not equal, it may
produce non-zero magnetic moment and intermediate state [78]. For the G@h-BN
structure, it shows similar properties of graphene quantum dots because the domain
of graphene is localized in the nanometer scale, and the bandgap decreases as the
size of the graphene domain increases. Similarly, when the number of B and N
atoms in the h-BN domain is not equal, the hybrid structure will have a certain
magnetic moment, and its capacity conforms to Lieb’s law [72, 75, 79–81].
In summary, the in-plane heterostructures of graphene-hexagonal boron nitride
can effectively open the bandgap of graphene. In addition, graphene and h-BN can
form a monolayer through perfect atomic patching, thus the edge electron scattering
of graphene is reduced, which ensures the high carrier mobility. For the periodic
zigzag-edged-h-BN@graphene heterostructure, it is predicted that when the band-
gap of the heterostructure is 0.38–1.39 eV, the intrinsic carrier mobility is mainly
distributed at 1.7 103–1.1 105 cm2/V s, and the some special system can
14 1 Introduction
reach 6.6 106 cm2/V s [82]. In addition, the linking boundary between graphene
and h-BN presents special electronic properties, such as magnetic properties,
excellent thermal conduction and interfacial electron reconstruction [83–86].
References
1. Novoselov KS, Geim AK, Morozov SV, Jiang D, Zhang Y, Dubonos SV, Grigorieva IV,
Firsov AA (2004) Electric field effect in atomically thin carbon films. Science 306:666
2. Novoselov KS, Geim AK, Morozov SV, Jiang D, Katsnelson MI IV, Grigorieva SV
Dubonos, Firsov AA (2005) Two-dimensional gas of massless dirac fermions in graphene.
Nature 438:197
3. Castro Neto AH, Guinea F, Peresv NMR, Novoselov KS, Geim AK (2009) The electronic
properties of graphene. Rev Mod Phys 81:109
4. Morozov SV, Novoselov KS, Katsnelson MI, Schedin F, Elias DC, Jaszczak JA, Geim AK
(2008) Giant intrinsic carrier mobilities in graphene and its bilayer. Phys Rev Lett 100:016602
5. Geim AK, Novoselov KS (2007) The rise of graphene. Nat Mater 6:183
6. Novoselov KS, Jiang Z, Zhang Y, Morozov SV, Stormer HL, Zeitler U, Maan JC,
Boebinger GS, Kim P, Geim AK (2007) Room-temperature quantum hall effect in graphene.
Science 315:1379
References 15
7. Zhang Y, Tan YW, Stormer HL, Kim P (2005) Experimental observation of the quantum hall
effect and Berry’s phase in graphene. Nature 438:201
8. Novoselov KS, Fal’ko VI, Colombo L, Gellert PR, Schwab MG, Kim K (2012) A roadmap
for graphene. Nature 490:192
9. Nair RR, Blake P, Grigorenko AN, Novoselov KS, Booth TJ, Stauber T, Peres NMR,
Geim AK (2008) Fine structure constant defines visual transparency of graphene. Science
320:1308
10. Liao L, Lin YC, Bao M, Cheng R, Bai J, Liu Y, Qu Y, Wang KL, Huang Y, Duan X (2010)
High-speed graphene transistors with a self-aligned nanowire gate. Nature 467:305
11. Meric I, Dean CR, Han SJ, Wang L, Jenkins KA, Hone J, Shepard KL (2011) High-frequency
performance of graphene field effect transistors with saturating IV-characteristics, 211.
arXiv:1112.2777
12. Bai J, Zhong X, Jiang S, Huang Y, Duan X (2010) Graphene nanomesh. Nat Nanotechnol
5:190
13. Pan Z, Liu N, Fu L, Liu Z (2011) Wrinkle engineering: a new approach to massive graphene
nanoribbon arrays. J Am Chem Soc 133:17578
14. Han MY, Özyilmaz B, Zhang Y, Kim P (2007) Energy band-gap engineering of graphene
nanoribbons. Phys Rev Lett 98:206805
15. Bae S, Kim H, Lee Y, Xu X, Park JS, Zheng Y, Balakrishnan J, Lei T, Kim HR, Song YI
(2011) Roll-to-roll production of 30-inch graphene films for transparent electrodes. Nat
Nanotechnol 5:574
16. Mak KF, Lui CH, Shan J, Heinz TF (2009) Observation of an electric-field-induced band gap
in bilayer graphene by infrared spectroscopy. Phys Rev Lett 102:256405
17. Oostinga JB, Heersche HB, Liu X, Morpurgo AF, Vandersypen LMK (2008) Gate-induced
insulating state in bilayer graphene devices. Nat Mater 7:151
18. McCann E (2006) Asymmetry gap in the electronic band structure of bilayer graphene. Phys
Rev B 74:161403
19. Ohta T, Bostwick A, Seyller T, Horn K, Rotenberg E (2006) Controlling the electronic
structure of bilayer graphene. Science 313:951
20. Zhang Y, Tang TT, Girit C, Hao Z, Martin MC, Zettl A, Crommie MF, Shen YR, Wang F
(2009) Direct observation of a widely tunable bandgap in bilayer graphene. Nature 459:820
21. Li G, Luican A, Lopes dos Santos JMB, Castro Neto AH, Reina A, Kong J, Andrei EY (2009)
Observation of Van Hove singularities in twisted graphene layers. Nat Phys 6:109
22. Yan W, Liu M, Dou RF, Meng L, Feng L, Chu ZD, Zhang Y, Liu Z, Nie JC, He L (2012)
Angle-dependent Van Hove singularities in a slightly twisted graphene bilayer. Phys Rev Lett
109:249
23. Yan W, Meng L, Liu M, Qiao JB, Chu ZD, Dou RF, Liu Z, Nie JC, Naugle DG, He L (2014)
Angle-dependent Van Hove singularities and their breakdown in twisted graphene bilayers.
Phys Rev B 90:5758
24. Zhou SY, Gweon GH, Fedorov AV, First PN, De Heer WA, Lee DH, Guinea F, Castro
Neto AH, Lanzara A (2007) Substrate-induced bandgap opening in epitaxial graphene. Nat
Mater 6:770
25. Wang M, Fu L, Gan L, Zhang C, Rümmeli M, Bachmatiuk A, Huang K, Fang Y, Liu Z
(2013) CVD growth of large area smooth-edged graphene nanomesh by nanosphere
lithography. Sci Rep 3:142
26. Liao L, Peng H, Liu Z (2014) Chemistry makes graphene beyond graphene. J Am Chem Soc
136:12194
27. Elias DC, Nair RR, Mohiuddin TMG, Morozov SV, Blake P, Halsall MP, Ferrari AC,
Boukhvalov DW, Katsnelson MI, Geim AK, Novoselov KS (2008) Control of graphene’s
properties by reversible hydrogenation: evidence for graphane. Science 323:610
28. Balog R, Jørgensen B, Nilsson L, Andersen M, Rienks E, Bianchi M, Fanetti M,
Laegsgaard E, Baraldi A, Lizzit S (2010) Bandgap opening in graphene induced by patterned
hydrogen adsorption. Nat Mater 9:315
16 1 Introduction
52. Wood JD, Schmucker SW, Lyons AS, Pop E, Lyding JW (2011) Effects of polycrystalline Cu
substrate on graphene growth by chemical vapor deposition. Nano Lett 11:4547
53. Han GH, Güneş F, Bae JJ, Kim ES, Chae SJ, Shin HJ, Choi JY, Pribat D, Lee YH (2011)
Influence of copper morphology in forming nucleation seeds for graphene growth. Nano Lett
11:4144
54. Li X, Magnuson CW, Venugopal A, Tromp RM, Hannon JB, Vogel EM, Colombo L,
Ruoff RS (2011) Large-area graphene single crystals grown by low-pressure chemical vapor
deposition of methane on copper. J Am Chem Soc 133:2816
55. Sun Z, Yan Z, Yao J, Beitler E, Zhu Y, Tour JM (2010) Growth of graphene from solid
carbon sources. Nature 468:549
56. Wofford JM, Nie S, McCarty KF, Bartelt NC, Dubon OD (2010) ‘Graphene islands on Cu
foils: the interplay between shape, orientation, and defects. Nano Lett 10:4890
57. Hao Y, Bharathi MS, Wang L, Liu Y, Chen H, Nie S, Wang X, Chou H, Tan C, Fallahazad B
(2013) The role of surface oxygen in the growth of large single-crystal graphene on copper.
Science 342:720
58. Gao L, Ren W, Xu H, Jin L, Wang Z, Ma T, Ma LP, Zhang Z, Fu Q, Peng LM (2012)
Repeated growth and bubbling transfer of graphene with millimetre-size single-crystal grains
using platinum. Nat Commun 3:23
59. Dai B, Fu L, Zou Z, Wang M, Xu H, Wang S, Liu Z (2011) Rational design of a binary metal
alloy for chemical vapour deposition growth of uniform single-layer graphene. Nat Commun
2:193
60. Sun J, Gao T, Song X, Zhao Y, Lin Y, Wang H, Ma D, Chen Y, Xiang W, Wang J, Zhang Y,
Liu Z (2014) Direct growth of high-quality graphene on high-j dielectric SrTiO3 substrates.
J Am Chem Soc 136:6574
61. Chen J, Guo Y, Jiang L, Xu Z, Huang L, Xue Y, Geng D, Wu B, Hu W, Yu G (2014)
Near-equilibrium chemical vapor deposition of high-quality single-crystal graphene directly
on various dielectric substrates. Adv Mater 26:1348
62. Hagstrom S, Lyon HB, Somorjai GA (1965) Surface structures on the clean platinum
(100) surface. Phys Rev Lett 15:491
63. Reina A, Jia X, Ho J, Nezich D, Son H, Bulovic V, Dresselhaus MS, Kong J (2008) Large
area few-layer graphene films on arbitrary substrates by chemical vapor deposition. Nano Lett
9:30
64. Kim KS, Zhao Y, Jang H, Lee SY, Kim JM, Kim KS, Ahh JH, Kim P, Choi JY, Hong BH
(2009) Large-scale pattern growth of graphene films for stretchable transparent electrodes.
Nature 457:706
65. Liu N, Fu L, Dai B, Yan K, Liu X, Zhao R, Zhang Y, Liu Z (2010) Universal segregation
growth approach to wafer-size graphene from non-noble metals. Nano Lett 11:297
66. Yan K, Peng H, Zhou Y, Li H, Liu Z (2011) Formation of bilayer bernal graphene:
layer-by-layer epitaxy via chemical vapor deposition. Nano Lett 11:1106
67. Li X, Wang X, Zhang L, Lee S, Dai H (2008) Chemically derived, ultrasmooth graphene
nanoribbon semiconductors. Science 319:1229
68. Jiao L, Zhang L, Wang X, Diankov G, Dai H (2009) Narrow graphene nanoribbons from
carbon nanotubes. Nature 458:877
69. Kosynkin DV, Higginbotham AL, Sinitskii A, Lomeda JR, Dimiev A, Price BK, Tour JM
(2009) Longitudinal unzipping of carbon nanotubes to form graphene nanoribbons. Nature
458:872
70. Nakada K, Fujita M, Dresselhaus G, Dresselhaus MS (1996) Edge state in graphene ribbons:
nanometer size effect and edge shape dependence. Phys Rev B: Condens Matter 54:17954
71. Son YW, Cohen ML, Louie SG (2006) Energy gaps in graphene nanoribbons. Phys Rev Lett
97:089901
72. Park H, Wadehra A, Wilkins JW, Castro Neto AH (2012) Magnetic states and optical
properties of single-layer carbon-doped hexagonal boron nitride. Appl Phys Lett 100:253115
18 1 Introduction
73. Giovannetti G, Khomyakov PA, Brocks G, Kelly PJ, van den Brink J (2007)
Substrate-induced band gap in graphene on hexagonal boron nitride: Ab initio density
functional calculations. Phys Rev B 76:073103
74. Seol G, Guo J (2011) Bandgap opening in boron nitride confined armchair graphene
nanoribbon. Appl Phys Lett 98:143107
75. Zhao R, Wang J, Yang M, Liu Z, Liu Z (2012) BN-embedded graphene with a ubiquitous gap
opening. J Phys Chem C 116:21098
76. Shinde PP, Kumar V (2011) Direct band gap opening in graphene by BN doping: Ab initio
calculations. Phys Rev B, Condens Matter 84:9682
77. Fan X (2012) Band gap opening of graphene by doping small boron nitride domains.
Nanoscale 4:2157
78. Xu B, Lu YH, Feng YP, Lin JY (2010) Density functional theory study of BN-doped
graphene superlattice: role of geometrical shape and size. J Appl Phys 108:073711
79. Bhowmick S, Singh AK, Yakobson BI (2011) Quantum dots and nanoroads of graphene
embedded in hexagonal boron nitride. J Phys Chem C 115:9889
80. Kan M, Zhou J, Wang Q, Sun Q, Jena P (2011) Tuning the band gap and magnetic properties
of BN sheets impregnated with graphene flakes. Phys Rev B 84:4327
81. Li J, Shenoy VB (2011) Graphene quantum dots embedded in hexagonal boron nitride sheets.
Appl Phys Lett 98:013105
82. Wang J, Zhao R, Liu Z, Liu Z (2013) Widely tunable carrier mobility of boron
nitride-embedded graphene. Small 9:1373
83. Jun S, Li X, Meng F, Ciobanu CV (2011) Elastic properties of edges in BN and SiC
nanoribbons and of boundaries in C-BN superlattices: a density functional theory study. Phys
Rev B: Condens Matter 83:1127
84. Jiang JW, Wang JS, Wang BS (2011) Minimum thermal conductance in graphene and boron
nitride superlattice. Appl Phys Lett 99:043109
85. Ramasubramaniam A, Naveh D (2011) Carrier-induced antiferromagnet of graphene islands
embedded in hexagonal boron nitride. Phys Rev B 84:173
86. Pruneda JM (2010) Origin of half-semimetallicity induced at interfaces of C-BN
heterostructures. Phys Rev B 81:2149
Chapter 2
Controllable Synthesis of Graphene on Rh
Compared with monolayer graphene, bilayer graphene displays even more complex
electronic band structures and intriguing properties. Recent studies reveal that the
low-energy band structure of bilayer graphene is extremely sensitive to the stacking
order. Two low-energy VHSs, which originate from the two saddle points in the
band structure, were observed in the twisted graphene bilayer as two pronounced
peaks in the DOS. The VHSs will induce novel physical properties, such as,
superconductivity and magnetism. Therefore, the preparation of large area
non-AB-stacked bilayer graphene is an efficient way to modify the energy band
structure near Fermi level. Combined with the preparation methods of graphene
introduced in Chap. 1, especially the growth method of bilayer graphene, I choose
segregation growth as a method for preparing non-AB stacking bilayer graphene.
The first section introduces the process engineering of graphene growth and the
factors that affect the segregation of graphene. I designed the carrier gas system,
heating system, cooling rate control system and pressure control system of graphene
growth and chose the Rh foil as a substrate for the growth of non-AB-stacked
graphene.
The second section mainly introduces the growth of graphene on Rh substrate.
Because Rh is a metal with a large carbon solubility, the low pressure(LP) CVD
system cannot provide enough carbon source for the growth of graphene. Thus I
selected the atmospheric pressure (AP) CVD system. By changing the cooling rate
after the growth of graphene, monolayer, bilayer, or multilayer graphene can be
obtained on Rh substrate. With the same growth system and method, I can control
the preparation of monolayer and bilayer graphene on Rh(111) substrate. It is
indicated that in addition to the grain boundary of the substrate, the terraces and
step edges are also the graphene segregation channel. In addition, wrinkles were
observed on graphene grown on Rh substrates, which are originated from the
thermal expansion mismatch between graphene and Rh.
In the third section, I summarize the growth characteristics of graphene on Rh
substrate. Thus, I propose that the thickness of the graphene prepared in this system
has a great relationship with the cooling rate after growth. The advantages and
© Springer Nature Singapore Pte Ltd. 2018 19
M. Liu, Controlled Synthesis and Scanning Tunneling Microscopy Study
of Graphene and Graphene-Based Heterostructures,
Springer Theses, https://doi.org/10.1007/978-981-10-5181-4_2
20 2 Controllable Synthesis of Graphene on Rh
2.1 Experiment
I choose the metal with high carbon solubility as the substrate to grow graphene,
and control the segregation process of graphene by CVD system to obtain the
large-scale bilayer non-AB-stacked graphene. In order to control the growth of
graphene, the basic steps of graphene growth are needed to be understood. Yan
et al. [1] summarized the previous work on graphene synthesized by CVD method
and proposed the process engineering of graphene growth.
Figure 2.1a is a typical CVD system for graphene growth, consisting of carrier gas
part and growth part. Carrier gases include carbon precursors (CH4, C2H4, benzene,
etc.), carrier gases (H2, Ar) and doped precursors (N2, BH3). A surrounding resis-
tance wire is used to heat the sample in quartz tube at temperature up to 1200 °C.
A typical CVD process consists of four main elementary steps: (A) adsorption
and catalytic decomposition of precursor gas, (B) diffusion and dissolution of
decomposed carbon species into bulk metal, (C) segregation of dissolved carbon
atoms onto the metal surface, and finally, (D) surface nucleation and growth of
graphene. Absence or enhancement of each elementary step would lead to signif-
icant changes in the whole growth process. Metals with certain carbon solubility,
such as nickel and cobalt, involve all four elementary steps in a typical CVD
process. Copper with negligible carbon solubility provides another platform for
process engineering, where both carbon dissolution and segregation steps are
negligible in the CVD process. Carbon atoms decomposed from precursors diffuse
on the surface and build up the thermodynamically stable honeycomb graphene
lattice. As a result, graphene growth on copper is self-limited, and formation of
multilayer graphene is generally prohibited. For most cases, the two growth paths
are coexisting.
Fig. 2.1 The process engineering of graphene. a Typical CVD growth process of graphene. b,
c Graphene growth process on different metal substrates: the red arrow and blue arrow represent
the graphene growth process on metals with high carbon solubility and low carbon solubility,
respectively. Reproduced with permission of [1]. Copyright [2013], American Chemical Society
becomes the most important factor for the study of graphene segregation growth.
According to the process engineering of graphene growth, the result of segregated
graphene is mainly related to the two steps of B and C. There are many factors to
control graphene segregation process, including the basic properties of the substrate
(e.g., carbon solubility, thickness and crystallization, etc.) and control parameters of
graphene growth (e.g., growth temperature and cooling rate). I will briefly describe
the impact of these factors on the graphene segregation process:
(1) The carbon solubility and thickness of metal substrates
The segregation growth of graphene conforms to the equation:
graphene on Fe surface is about 40, and 15% of the substrate is bare. Monolayer
graphene on Co and Ni accounted for 50 and 70%, respectively. Consequently, with
the decrease of carbon solubility of metal substrate, the homogeneity of segregated
graphene is improved remarkably [2]. For the same metal substrate, increased
thickness of the substrate leads to increasing in the amount of dissolved carbon
atoms. As a result, the segregated carbon increases. For example, graphene layer
grown on Co substrate with thickness of 300 nm is very inhomogeneous and
monolayer graphene is rarely observed. With the decrease of the thickness of Co
substrate, the homogeneity of graphene increases, and the area of monolayer
increases. The monolayer graphene grown on Co substrate with 100 nm thickness
accounts for more than 70% [3]. However, the thickness of the substrate cannot be
reduced indefinitely. If the thickness of the substrate layer is too thin, the metal
substrate tends to be agglomerated into nanoparticles at high growth temperature at
1000 °C.
(2) The crystallization of metal substrate
From the kinetic point of view, carbon atoms preferentially diffuse from the
grain boundary of substrate to form graphene films [4, 5]. Reducing the quantity of
grain boundary can effectively prevent the segregated carbon atoms and improve
the film uniformity. For example, the grain boundary between the different crystal
faces of the polycrystalline Pd substrate is the main channel for the segregation of
carbon atoms. However, due to the high carbon solubility of Pd, the graphene on
grain boundary is much thicker than that on Pd crystal facet. As shown in Fig. 2.2,
the grid-like graphite-coated graphene structure is formed on Pd foils. The
long-term annealing of the Pd foil at 1050 °C can optimize the Pd crystal plane and
reduce the number of grain boundaries. With the increasing of the optimization
time, the thickness of segregated graphene at the grain boundary gradually
decreases, and the homogeneity of graphene increases. What’s more, uniform
monolayer graphene can be obtained on Pd single crystal substrate, which has no
grain boundary.
(3) Growth temperature and cooling rate
The growth temperature for graphene segregation affects the balance between the
carbon dissolution into substrate bulk and carbon segregation onto substrate sur-
face. When the equilibrium is reached, the different temperatures correspond to the
different states of the metal on the metal surface [6, 7]. In the case of Ni: the carbon
atoms begin to form sp2-hybrid fragments on the Ni surface, and small graphite
domains can be obtained at 850 °C; at 1000 °C, high coverage of the graphite
aggregates are obtained; at 1100 °C, monolayer graphene can be achieved; when
the temperature is higher than 1150 °C, the amount of carbon atoms are dissolved
in Ni substrates, thus graphene cannot be formed [2]. Consequently, 1100 °C is the
optimum temperature for the segregated graphene on Ni substrates (Fig. 2.3).
The effect of cooling rate and growth temperature on graphene segregation is
associated. When the equilibrium of carbon dissolution-segregation is achieved, the
2.1 Experiment 23
Fig. 2.2 Annealing pretreatment of Pd foils for the subsequent growth of graphene. a–c Optical
images of Pd foils after annealing pre-treatment for 1 (a), 3 (b), and 5 (c) hours. d–f Corresponding
SEM images of the subsequently grown graphene on the pretreated Pd foils. g–i Corresponding
optical images of the graphene samples after being transferred onto SiO2/Si substrates.
j–l Schematics of the pre-treatment effect on the reduction of the grain boundaries and on the
suppression of the excessive carbon segregation. Reproduced with permission of [4]. Copyright
[2014] Wiley–VCH Verlag GmbH & Co. KGaA, Weinheim
cooling process will break the balance. During the slow cooling process, as the
temperature decreases, the carbon solubility of the metal substrate decreases,
causing the excessive carbon atoms to segregate on the surface to form thick
graphene layers or graphite. The rapid cooling process can maintain the equilibrium
state of the system at high temperature, resulting in the uniform graphene layers at
the growth temperature.
2.1.3 Substrate
Fig. 2.3 Temperature effect on graphene segregation. a–d Layer distribution of graphene on
300 nm SiO2/Si grown from 200 nm Ni at temperatures of 850, 1000, 1100, and >1150 °C,
respectively. e Schematic illustration of various equilibrium states corresponding to temperature
change during graphene segregation. Reproduced with permission of [2]. Copyright [2011],
American Chemical Society
substrates for further applications. (3) Alloys, such as, Cu–Ni alloy [19–21]. This
kind of alloy combines metals with high carbon solubility and low carbon solu-
bility, thus the control of thickness of graphene can be significantly improved. The
disadvantage of the alloy system is that the surface roughness is high and the
quality of the prepared graphene is poor.
It is mentioned in process engineering of graphene growth that the four basic
steps are all related to the metal substrates, meanwhile most of the factors that affect
graphene segregation are also related to the intrinsic properties of the metal sub-
strate. Therefore, it is essential to select a suitable substrate for the preparation of
segregated graphene. The most important factor is the carbon solubility of metal
substrate. In addition to the most commonly used Cu and Ni substrates, the metals
of VIIIB and IB also have high carbon solubilities (up to 7 at.% at 1000 °C) and
high catalytic activity for graphene growth (Fig. 2.4).
Among these transition metals, Rh has the smallest carbon solubility, between
Cu and Ni, and the growth of graphene on Rh(111) singlecrystal in UHV system
has been studied deeply. Monolayer graphene can be synthesized on Rh(111)
substrate at 600 °C, and the lattice orientation of graphene is consistent with that of
Rh(111) facet, which means the graphene shows a singlecrystalline feature [22–24].
Consequently, Rh is expected to be a good substrate for the growth of large-scale
non-AB stacked bilayer graphene.
2.1 Experiment 25
Fig. 2.4 Comparison of transition metals of VIIIB and IB on melting point, carbon solubility and
the lattice mismatch with graphene
Figure 2.5 is a schematic diagram for the growth system. The sample is placed in a
quartz tube in furnace; the mixed gases are carried through the pipeline into the
quartz tube for the growth of graphene; after the growth, the sample is rapidly
removed from the heating zone by magnet to complete the rapid cooling process.
The detailed growth process of graphene is shown in Fig. 2.6. First, the Rh foil
is heated to 1000 °C in a hydrogen and argon atmosphere for 40 min. In this
process, the organic impurities adsorbed on the surface of the metal substrate can be
removed, and the Rh crystal facet can be optimized. Then the CH4 is introduced
into the quartz tube for the growth of graphene. After that, the system is cooled
down to room temperature. It is worth noting that in the process of cooling, if a
slow cooling method is used, it is necessary to keep the CH4 on. If the CH4 is cut off
during the slow cooling process, the graphene can be easily etched by H2. In
opposite, there is no need for continuous access to CH4 during the fast cooling
process.
Chemical etching is the most commonly used method for graphene transferring
[25]. However, Rh is insoluble in most acid, alkali or salt solutions, thus graphene
on Rh cannot be transferred by chemical etching method. Graphene grown on noble
metal substrate can be transferred by using the “bubble transfer method”. The
bubble transfer method is proposed by the Cheng et al. in 2012 [26], that is, the
graphene and substrate is separated by the interfacial bubbles generated from
electrolyzed water. The graphene can be transferred onto SiO2/Si or other substrates
for further characterization and applications.
The Rh substrate used in the experiment was purchased from Sigma-Aldrich, with a
thickness of 0.025 mm and a purity of 99.9%, the picture of which is shown in
Fig. 2.7a. After the Rh foil was annealing at 1000 °C, X-ray diffraction (Fig. 2.7b)
shows that there are only two facets, Rh(111) and Rh(200).
In order to minimize the amount of carbon source introduced into the system and
control the thickness of the segregated graphene, we first tried the low pressure
(LP) CVD system, where the tail gas was connected to a rotary pump so that the
sample in the quartz tube was at low pressure.
When the carrier gas is CH4:H2:Ar = 20:10:50 (here means the flow of CH4, H2
and Ar are 20, 10, 50 sccm, respectively, where sccm is short for the standard cubic
per minute abbreviation), the result is shown in Fig. 2.8. It is observed that the
coverage of graphene is very low. After the graphene is transferred onto SiO2/Si
substrate, we find that the graphene sheets are not uniform, and the OM image
shows that the thickness of graphene sheet is more than 10 layers. These results
2.2 Controllable Growth of Graphene on Rh Substrates 27
Fig. 2.7 a Picture and b XRD characterization of Rh foils after annealing at 1000 °C
reveal that the LP system cannot provide sufficient carbon source for the growth of
large-scale uniform graphene.
Therefore, we choose the atmospheric pressure (AP) CVD system for graphene
growth, that is, the tail gas and the rotary pump are isolated, so that the sample in
the quartz tube is at atmospheric pressure. The carrier gases are CH4:H2:
Ar = 10:300:100 and growth temperature is 1000 °C for 5 min. After growth, the
sample was rapidly cooled to room temperature at a rate of 150 °C/min. As shown
in Fig. 2.9, the coverage of graphene is more than 80% and the SEM contrast is
very uniform. We then transfer the graphene sample into the UHV chamber and
degas at 400 °C for further STM characterizations. STM image shows that the Rh
foil surface forms straight array of steps with graphene wrinkles across the steps.
The width of wrinkle is *30 nm, height *3 nm and the length can reach several
hundred nanometers. The atomic resolution of the graphene honeycomb lattices can
be obtained by high-resolution STM, as shown in Fig. 2.9c. The AFM image of
28 2 Controllable Synthesis of Graphene on Rh
Fig. 2.9 Monolayer graphene synthesized on Rh foils through a fast cooling process after growth.
a SEM image of graphene. b STM image of graphene on Rh foils. c Atomic resolved STM image
of graphene showing honeycomb lattices. d–f AFM images of graphene on Rh foils, d, e and
f corresponding to height, amplitude and phase channels, respectively
as-grown graphene on the Rh foil shows a highly corrugated surface. AFM height
and amplitude images are difficult to distinguish the graphene domains.
Interestingly, the AFM phase image shows different contrasts of graphene and the
bare Rh foils.
We used the bubble method to transfer the graphene to SiO2/Si substrate for
further study. The graphene on the SiO2/Si substrate shows a light purple contrast in
OM image, which is corresponds to the contrast of monolayer graphene (Fig. 2.10).
The Raman spectra of graphene (the wavelength of the laser is 633 nm) shows that
the G peak and 2D peak of graphene appear at *1584 and *2650 cm−1,
respectively, meanwhile the 2D/G ratio is *1.2 and the full width at half maximum
(fwhm) of 2D peak is *30 cm−1, consistent with the characteristics of monolayer
graphene. In addition, there is no D peak at *1350 cm−1, indicating a high quality
of graphene. In order to further reconfirm the thickness of graphene, we transfer the
graphene film onto the TEM grid to do the high-resolution measurement.
Figure 2.10c is a large-scale image of TEM, and its edge fold is enlarged to show
that the graphene is monolayer. The selected area electronic diffraction (SAED) on
graphene shows only one set of six symmetric diffraction spots, and the second
order diffraction intensity is weaker than that of the first order diffraction, again
confirming that the graphene is monolayer.
2.2 Controllable Growth of Graphene on Rh Substrates 29
The experimental results reveal that graphene tends to form a uniform monolayer
on Rh foils with the coverage over 80% at 1000 °C. It means that the dissolution of
the carbon atoms at this temperature is greater than segregation. In order to obtain a
full monolayer or bilayer graphene, it is necessary to slow down the cooling rate to
apply sufficient time for carbon atoms to segregate on Rh surface.
We slow down the cooling rate to 15 °C/min, and set the precursor ratio (CH4:
H2:Ar = 5:50:850 sccm), while the growth temperature and time remain unchan-
ged. The growth results are displayed in Fig. 2.11. SEM images of the samples
show that the Rh foils are fully covered with graphene, and the white wires cor-
respond to the graphene wrinkles. The angles between the graphene wrinkles
(Fig. 2.11a, b) are 60° or 120°, forming a triangular or hexagonal shape, meanwhile
Fig. 2.11 SEM images of graphene grown on Rh foils through a slow cooling process
30 2 Controllable Synthesis of Graphene on Rh
Fig. 2.12 Characterizations of graphene growth on Rh foils under slow cooling process
(15 °C/min). a OM image of graphene after being transferred onto SiO2/Si. b Raman spectra of
graphene sample in (a) showing multilayer (nL) films along with a little portion of 1L and 2L
regions, under a Raman excitation wavelength of 633 nm. c–f TEM characterization of graphene
grown with a slow cooling process. TEM images of the folded regions showing 6L graphene
the angles between the graphene wrinkles in Fig. 2.11c is mostly 90°, forming a
rectangular structure. The formation of wrinkles is due to the thermal expansion
mismatch between the graphene and the substrate in the cooling process and the
wrinkles prefer to form along the grain boundaries of the substrates. Therefore, we
can infer that Fig. 2.11a, b, c represent the graphene grown on Rh(111) and
Rh(200) facets, respectively (Fig. 2.12).
We transferred graphene to the SiO2/Si substrate by bubble method to confirm
the thickness of graphene. OM image shows that graphene exhibits various con-
trasts from light purple to purple, indicating that the thickness of graphene is less
than 10 layers, but the uniformity is poor. The Raman spectra measured randomly
(the wavelength was 633 nm) indicate the characteristic spectra of graphene with
different thickness. Such as, characteristics of single-layer graphene shows
I(2D/G) = 1.6, and fwhm of 2D peak *30 cm−1; bilayer graphene I(2D/G) = 0.8,
fwhm of 2D peak *45 cm−1; few layer graphene I(2D/G) < 1, fwhm of 2D
peak >70 cm−1. The graphene film was transferred to the TEM grid for further
confirmation of the thickness, and the high-resolution TEM image of the graphene
folded edge proves that the thickness of graphene is 6 layers. SAED image shows
multiple sets of six symmetric diffraction spots with different rotation angles,
indicating non-AB stacked feature of graphene films.
2.2 Controllable Growth of Graphene on Rh Substrates 31
Fig. 2.13 Bilayer graphene synthesized on Rh foils a The SEM image, b high-resolution TEM
image and c Raman spectra of bilayer graphene
The slow cooling process provides a large amount of carbon source to segregate
onto the Rh surface to form few layer graphene films. In order to reduce the
segregation of carbon atoms and improve the homogeneity of graphene, we change
the ratio of the precursor to CH4:H2:Ar = 3:300:850. Reducing the flow of methane
while increasing the flow of hydrogen is for both diluting the methane concentration
and increasing the etching process. The SEM image shown in Fig. 2.13a reveals
that there are only dark gray and light gray contrast in graphene region. According
to the work mechanism of SEM, graphene in dark gray area is thicker than light
gray. The high-resolution TEM image of the folded edge of graphene layer confirm
that the thickness of graphene is bilayer. The graphene is transferred to the SiO2/Si
substrate and the Raman spectra are measured in different contrast regions. The
results show that the light gray area has the characteristic of the monolayer gra-
phene and the dark gray region has the characteristic of bilayer graphene.
By controlling the growth time from 2 to 5 min, the coverage of bilayer gra-
phene can be controlled from 50 to 95% (Fig. 2.14a–d). In addition, we also use
Rh(111) single crystal as a substrate to synthesize bilayer graphene. The growth
parameters on the Rh(111) substrate are set to CH4:H2:Ar = 20:50:200, at 1000 °C
for 10 min and the cooling rate is 15 °C/min.
The growth results are shown in Fig. 2.14e–h, and the proportion of bilayer
graphene is about 60%. It is found from the gradually magnified SEM images that
the segregated graphene on Rh(111) substrate has two characteristics: no winkle is
observed on graphene; bilayer graphene has no specific nucleation sites and domain
shapes. Compared with the Rh foil substrate, there is no grain boundary on Rh(111)
substrate. It is generally believed that grain boundaries are the main route of gra-
phene segregation. This work shows that in the absence of grain boundary of the
metal substrate, the metal steps and the terraces are the main path for graphene
segregation.
32 2 Controllable Synthesis of Graphene on Rh
Fig. 2.14 Bilayer graphene synthesized on Rh foils and Rh(111) substrates a–d SEM images of
bilayer graphene on Rh foils. e–h SEM images of bilayer graphene on Rh(111)
Fig. 2.15 Schematic drawing showing the growth mechanism of graphene on Rh foils. a CH4
decomposition at 1000 °C, followed with carbon dissolution into Rh foils. b–e Side-view and 3D
images of carbon atoms segregating from bulk to Rh surfaces throughout the quenching processes
under fast and slow cooling, forming monolayer and multilayer graphene, respectively
rapid cooling or slow cooling process is applied. Comparing the growth of graphene
on the three noble metal substrates, we can obtain monolayer, bilayer and multi-
layer graphene on the Rh substrate in a controllable way, and the control window of
the graphene layer is relatively wide. Consequently, Rh is the best substrate for the
preparation of segregated bilayer graphene.
2.4 Conclusion
In this chapter, I summarized the factors that control the thickness of segregated
graphene, and selected the Rh (with the appropriate carbon solubility) as substrate,
to synthesize segregated graphene using CVD growth system.
Monolayer graphene is obtained with a cooling rate of 150 °C/min after the
growth process, and the monolayer feature is confirmed by TEM, OM and Raman
measurements. Combined with the control factors of graphene segregation growth,
few layer graphene is obtained by reducing the cooling rate to 15 °C/min.
Furthermore, bilayer graphene is obtained by reducing the ratio of CH4/H2 in gas
precursors. Using the same growth system and method, I can also obtain bilayer
graphene on the Rh(111) substrate, which indicates that in addition to the grain
boundary of the substrate, the crystal terraces and steps are also segregated channel
for carbon atoms.
34 2 Controllable Synthesis of Graphene on Rh
References
1. Yan K, Fu L, Peng H, Liu Z (2013) Designed CVD growth of graphene via process
engineering. Acc Chem Res 46:2263
2. Liu N, Fu L, Dai B, Yan K, Liu X, Zhao R, Zhang Y, Liu Z (2010) Universal segregation
growth approach to wafer-size graphene from non-noble metals. Nano Lett 11:297
3. Ramón ME, Gupta A, Corbet C, Ferrer DA, Movva HCP, Carpenter G, Colombo L,
Bourianoff G, Doczy M, Akinwande D (2011) Cmos-compatible synthesis of large-area,
high-mobility graphene by chemical vapor deposition of acetylene on cobalt thin films. ACS
Nano 5:7198
4. Ma D, Liu M, Gao T, Li C, Sun J, Nie Y, Ji Q, Zhang Y, Song X, Zhang Y, Liu Z (2014)
High-quality monolayer graphene synthesis on pd foils via the suppression of multilayer
growth at grain boundaries. Small 10:4003
5. Zhang Y, Gomez L, Ishikawa FN, Madaria A, Ryu K, Wang C, Badmaev A, Zhou C (2010)
Comparison of graphene growth on single-crystalline and polycrystalline Ni by chemical
vapor deposition. J Phys Chem Lett 1:3101
6. Shelton JC, Patil HR, Blakely JM (1974) Equilibrium segregation of carbon to a nickel
(111) surface: a surface phase transition. Surf Sci 43:493
7. Odahara G, Otani S, Oshima C, Suzuki M, Yasue T, Koshikawa T (2011) In-situ observation
of graphene growth on Ni(111). Surf Sci 605:1095
8. Han GH, Güneş F, Bae JJ, Kim ES, Chae SJ, Shin HJ, Choi JY, Pribat D, Lee YH (2011)
Influence of copper morphology in forming nucleation seeds for graphene growth. Nano Lett
11:4144
9. Reina A, Thiele S, Jia X, Bhaviripudi S, Dresselhaus MS, Schaefer JA, Kong J (2009) Growth
of large-area single-and bi-layer graphene by controlled carbon precipitation on polycrys-
talline Ni surfaces. Nano Res 2:509
10. Sicot M, Bouvron S, Zander O, Rüdiger U, Dedkov YS, Fonin M (2010) Nucleation and
growth of nickel nanoclusters on graphene Moiré on Rh(111). Appl Phys Lett 96:093115
11. Xue Y, Wu B, Guo Y, Huang L, Jiang L, Chen J, Geng D, Liu Y, Hu W, Yu G (2011)
Synthesis of large-area, few-layer graphene on iron foil by chemical vapor deposition. Nano
Res 4:1208
12. Ago H, Ito Y, Mizuta N, Yoshida K, Hu B, Orofeo CM, Tsuji M, Ikeda K, Mizuno S (2010)
Epitaxial chemical vapor deposition growth of single-layer graphene over cobalt film
crystallized on sapphire. ACS Nano 4:7407
13. Kim E, An H, Jang H, Cho WJ, Lee N, Lee WG, Jung J (2011) Growth of few-layer graphene
on a thin cobalt film on a Si/SiO2 substrate. Chem Vap Deposition 17:9
14. Sutter E, Albrecht P, Sutter P (2009) Graphene growth on polycrystalline Ru thin films. Appl
Phys Lett 95:133109
15. Sutter E, Acharya DP, Sadowski JT, Sutter P (2009) Scanning tunneling microscopy on
epitaxial bilayer graphene on ruthenium (0001). Appl Phys Lett 94:133101
16. Sutter P, Hybertsen MS, Sadowski JT, Sutter E (2009) Electronic structure of few-layer
epitaxial graphene on Ru (0001). Nano Lett 9:2654
17. Que Y, Xiao W, Fei X, Chen H, Huang L, Du SX, Gao HJ (2014) Epitaxial growth of
large-area bilayer graphene on Ru (0001). Appl Phys Lett 104:093110
References 35
18. Nie S, Walter AL, Bartelt NC, Starodub E, Bostwick A, Rotenberg E, McCarty KF (2011)
Growth from below: graphene bilayers on Ir (111). ACS Nano 5:2298
19. Liu X, Fu L, Liu N, Gao T, Zhang Y, Liao L, Liu Z (2011) Segregation growth of graphene
on Cu–Ni Alloy for precise layer control. J Phys Chem C 115:11976
20. Chen S, Cai W, Piner RD, Suk JW, Wu Y, Ren Y, Kang J, Ruoff RS (2011) Synthesis and
characterization of large-area graphene and graphite films on commercial Cu–Ni alloy foils.
Nano Lett 11:3519
21. Wu Y,Chou H, Ji H, Wu Q, Chen S, Jiang W, Hao Y, Kang J, Ren Y, Piner RD (2012)
Growth mechanism and controlled synthesis of AB-stacked bilayer graphene on Cu–Ni alloy
foils. ACS Nano 6:7731
22. Liu M, Gao Y, Zhang Y, Zhang Y, Ma D, Ji Q, Gao T, Chen Y, Liu Z (2013) Single and
polycrystalline graphene on Rh(111) following different growth mechanisms. Small 9:1360
23. Voloshina EN, Dedkov YS, Torbrügge S, Thissen A, Fonin M (2012) Graphene on Rh(111):
scanning tunneling and atomic force microscopies studies. Appl Phys Lett 100:241606
24. Wang B, Caffio M, Bromley C, Früchtl H, Schaub R (2010) Coupling epitaxy, chemical
bonding, and work function at the local scale in transition metal-supported graphene. ACS
Nano 4:5773
25. Reina A, Son H, Jiao L, Fan B, Dresselhaus MS, Liu Z, Kong J (2008) Transferring and
identification of single-and few-layer graphene on arbitrary substrates. J Phys Chem C
112:17741
26. Gao L, Ren W, Xu H, Jin L, Wang Z, Ma T, Ma LP, Zhang Z, Fu Q, Peng LM (2012)
Repeated growth and bubbling transfer of graphene with millimetre-size single-crystal grains
using platinum. Nat Commun 3:23
Chapter 3
STM Study of Twisted Bilayer Graphene
In the 1980s, the invention of scanning tunneling microscope (STM) opened the
door to observe the world from the atomic scale. For STM, good resolution is
considered to be 0.1 nm lateral resolution and 0.01 nm (10 pm) depth resolution.
With this resolution, individual atoms within materials are routinely imaged and
manipulated.
In the first section, I briefly introduce the invention, working principles and
working modes of STM. Then I introduce the application of STM in graphene
research.
The second section mainly introduces the study of graphene synthesized on Rh
substrate using STM, including the apparent morphology and stacking order. It is
indicated that bilayer graphene on Rh are twisted and show various moiré patterns.
The third section introduces the STS study on the moiré patterns of twisted
bilayer graphene. Two new states near the Fermi level are observed, which are
known as van Hove singularities (VHSs). It is interestingly to be found that the
energy difference DE corresponding to the two VHSs is linearly related to the twist
angle between the double graphene layers.
The fourth section introduces the formation of graphene wrinkles on the Rh
substrate. The winkles on monolayer graphene and multilayer graphene are
observed by AFM and STM, and the strain-induced evolution of electronic band
structure in graphene is investigated on winkles.
The scanning tunneling microscopy (STM) was invented by Binnig and Rohrer
(Fig. 3.1a) in 1981, who earned the Nobel Prize in Physics in 1986 [1]. With atomic
resolution, individual atoms within materials are routinely imaged and manipulated.
As a two-dimensional material with one-atomic-layer thickness, the atomic struc-
tures, defects, domain boundaries of graphene have a great influence on the
© Springer Nature Singapore Pte Ltd. 2018 37
M. Liu, Controlled Synthesis and Scanning Tunneling Microscopy Study
of Graphene and Graphene-Based Heterostructures,
Springer Theses, https://doi.org/10.1007/978-981-10-5181-4_3
38 3 STM Study of Twisted Bilayer Graphene
Fig. 3.1 The working principle of STM. a The inventors of STM: G. Binnig and H. Rohrer.
b The quantum tunneling. c The schematic view of STM
chemical and physical properties. STM is a powerful means to study the surface
atomic structure and electronic structure. The following will briefly introduce the
working principle of STM and the application on graphene research.
STM is based on the concept of quantum tunneling (Fig. 3.1b). When a conducting
tip (W or Pt/Ir) is brought very near to the surface, a bias applied between the two
can allow electrons to tunnel through the vacuum between them. The tunneling
current is a function of bias (V), tip-sample distance (S) and the local density of
states (Ф) of the sample and tip:
1
I ðsÞ ¼ KVexpðlU2 SÞ ð3:1Þ
It can be seen that the tunnel current and the tip-sample distance are exponen-
tially related, thus the tunnel current is very sensitive to the change of distance.
STM is mainly composed of four parts, including scanning tip, three-dimensional
(3D) piezoelectric scanner, coarse sample-to-tip control and computer, and vibra-
tion isolation system [2].
(1) Scanning tip
The scanning tip is made of tungsten or platinum-iridium. Tungsten tips are
usually made by electrochemical etching, and platinum-iridium tips by mechanical
shearing.
3.1 Introduction to STM Study of Graphene 39
Fig. 3.2 Two typical working modes of STM. a Constant-current mode. b Constant-height mode
40 3 STM Study of Twisted Bilayer Graphene
where qs(E) and qt(E) are the LDOS of sample and tip, respectively and the tun-
neling current is a convolution of the DOS of the tip and the sample. Generally, the
DOS of tip does not change during the experiment, which can be regarded as a
constant. Equation (3.2) reduces to:
ZeV
I/ qs ðEÞ dE ð3:3Þ
0
3.1 Introduction to STM Study of Graphene 41
dI
/ qs ðr0 ; EÞ ð3:4Þ
dV
where qs(r0, E) is the spatial distribution of the DOS on the sample surface. Thus,
the tunneling conductance is directly proportional to the sample DOS.
STM has high resolution and can be used to investigate the surface morphology,
superlattices, defect states and the corresponding electronic states of graphene.
I will briefly introduce the application of STM/STS on graphene.
(1) The atomic structure and surface morphology of the graphene
In 2007, Flynn et al. [3] reported the STM study of graphene on SiO2/Si sub-
strates. In the monolayer graphene region, STM image shows a honeycomb-like
six-symmetric atomic lattice, that is, the intrinsic seamless atomic structure of
graphene. In the multilayer graphene region, the graphene maintains the stacking
order in the HOPG: three of the carbon atoms (in a hexagonal ring) in the first
graphene layer are located directly above the carbon atoms of the second layer
(labeled A-type carbon atom), and the other three carbon atoms are located above
the hole of second layer (labeled as B-type carbon atom). The AB-stacking order
breaks the six-symmetry of the graphene lattice, which results in the hexagonal dot
arrays in STM image. This is an easy way to determine the thickness of graphene
using STM.
Since the successful preparation of graphene on Cu foils by CVD method
reported by Ruoff et al. [4] in 2009, the synthesis of graphene on polycrystalline
metal substrates has attracted great interest. Polycrystalline metal substrates are
consisted of different facets, which have a certain effect on the growth and prop-
erties of graphene. Zhang et al. [5] reported the study of graphene grown on Pt foils
using STM. Pt(110), (111), (100) and (311) planes were formed on Pt foil after
annealing at 1050 °C. The graphene on Pt(110) and Pt(100) facets shows stripe
moiré patterns, meanwhile the graphene on Pt(111) surface shows honeycomb
lattices with no moiré patterns. The STM characterization of graphene on Pt foils
reveals that the graphene lattices are continuous and seamless.
42 3 STM Study of Twisted Bilayer Graphene
d
D¼ ð3:5Þ
2 sin h2
ð1 þ dÞd
k ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð3:6Þ
2ð1 þ dÞð1 cos hÞ þ d2
When h = 0°, the moiré period reaches a maximum of 13.9 nm. If there is a twist
between the graphene and h-BN, the moiré period will decrease sharply as the twist
angle increases. The theoretical calculations predict that this periodic potential field
will induce the formation of superlattice Dirac cones in band structure of graphene.
(3) The defects and defect states of graphene
Since the one-atom-thick nature, the defective structures in graphene would
dramatically modulate its chemical and physical properties. STM is a powerful tool
to be used to study the defects and defect states of graphene. The graphene defects
mainly include point defects and line defects.
Substitutional doping with heteroatoms is one of the most feasible methods to
open the bandgap and tune the Fermi level of graphene. Zhao et al. [8] probed the
doping effect of N atoms in graphene. As an electron-rich element compared to C,
single N dopant appeals as a triangular shape consisting with three bright spots as
revealed by the close-up STM image, and the electronic perturbation caused by a N
dopant is localized near the N atom and a decay length of *7 Å. Vacancies in
graphene could also induce novel properties, such as, magnetism [9]. Single
vacancies present as triangular protrusions in STM image which is due to the
electronic scattering effects.
The line defects of graphene usually refer to the grain boundary (GB) formed
between two adjacent graphene domains. Lahiri et al. [10] reported the
one-dimensional (1D) line defect in graphene using STM. The unit of the defects is
proposed to consist with one octagon and a pair of pentagons. The calculated DOS
3.1 Introduction to STM Study of Graphene 43
indicates a metallic character, and the decay width of the nanowire was estimated
*0.8 nm.
In summary, STM/STS can not only observe the atomic structure, superlattice
structure and defects of graphene surface from atomic scale, but also can study the
electronic structure of graphene and the new electronic states induced by structural
modulation. Thus, STM and STS plays an important role for studying the surface
structures and physical properties of graphene.
There are different ways to determine the stacking order of bilayer graphene, such
as Raman spectroscopy and TEM. However, most methods need to transfer gra-
phene onto target substrates, which will cause damage to graphene. STM is an
in situ characterization method and can easily determine the stacking order of
bilayer graphene.
As shown in Fig. 3.3, bilayer graphene has two stacking orders: AB stacking
(Fig. 3.3a) and non-AB stacking (Fig. 3.3c). AB stacking bilayer graphene shows
hexagonal close-packed structure in STM image, meanwhile non-AB stacking
bilayer graphene shows moiré patterns in STM image.
The graphene sample on the Rh foil was welded on the STM sample stage and
transferred into the vacuum chamber. The water and impurities adsorbed on the
surface were removed by annealing sample at 400 °C for 30 min.
The large-scale STM image of bilayer graphene on the Rh foil is shown in
Fig. 3.4. The uniform moiré patterns indicate that the graphene sample is non-AB
Fig. 3.3 Different stacking order of bilayer graphene a, c Schematic diagrams of AB-stacking and
non-AB stacking bilayer graphene, respectively. b, d The STM images of bilayer graphene with
AB-stacking and non-AB stacking orders, respectively
44 3 STM Study of Twisted Bilayer Graphene
stacking. The moiré period and the twisted angle between graphene layers is 3.8 nm
and 3.7°, respectively, which is calculated by using Eq. (3.5).
A series of STM characterization of several graphene samples is carried out, and
the results of which are shown in Fig. 3.5. Figure 3.5a is the atomic-resolution
STM image of monolayer graphene on Rh foils, and the line profile in Fig. 3.5g
confirms the lattice constant of 0.246 nm. More graphene moiré patterns listed in
Figure b–f show the periods of 1.2, 2.0, 3.8, 4.6 and 6.4 nm, corresponding to the
twisted angles of 12°, 6.9°, 3.7°, 3.0° and 2.2°, respectively.
From the STM images and the corresponding line profiles of graphene moiré
patterns, it can be found that, the atomic structure and moiré patterns can be
Fig. 3.5 Graphene moiré patterns on Rh substrates. a STM image of monolayer graphene on Rh
foils. b–f STM images of moiré patterns on twisted bilayer graphene on Rh foils. g Line profiles
along the yellow arrows labeled in (a–f)
3.2 STM Study of Bilayer Graphene 45
identified simultaneously when the moiré period is smaller than 3 nm. When the
moiré period is larger than 3 nm, the graphene appears to a large corrugation which
makes the atomic structure difficult to be obtained.
The above results show that the bilayer graphene prepared on the Rh substrate
tends to be stacked in the non-AB way, and there is no fixed twist angle between the
two layers of graphene. At present, the range of twist angles can span from 1° to
12°. This provides a good experimental sample for the subsequent study of the
interlayer coupling-induced novel electronic states on graphene moiré patterns.
Fig. 3.6 Coexistence of monolayer and bilayer graphene on Rh substrates. a–c Stepwise
magnified images of the coexistence region of monolayer and bilayer graphene showing a
highlight linking boundary. d, e Another coexistence region with a mild linking boundary. f The
sketch map of the linking of monolayer and bilayer graphene on Rh substrates
Fig. 3.7 Different linking type of twisted bilayer graphene on Rh substrates. a Line-defect linked
and b point-defect linked graphene moiré patterns. c Schematic diagram of linking boundary of
twisted bilayer graphene
means that the orientation of graphene lattices in the two regions are similar. Thus
the boundary consists of a series of point defects. According to the literature [11],
this type of boundary will induce new state at the Fermi level, which may induce
new properties of graphene, such as magnetism.
In addition, the coexistence of bilayer and trilayer graphene can also be observed
on Rh substrate (Fig. 3.8). The top left corner of Fig. 3.8a is characterized with
uniform moiré patterns (marked as moiré A). The period of moiré A is 1.3 nm,
3.2 STM Study of Bilayer Graphene 47
Fig. 3.8 Coexistence of bilayer and trilayer graphene on Rh substrates. a Large-scale STM image
of coexistence of bilayer and trilayer graphene. b–d Magnified atomic-resolved STM images of
graphene shown in (a). e The FFT of (b). f Magified image of (e)
which corresponds to the twisted angle of 5.4°. Meanwhile, in the bottom left
corner, a nested two sets of graphene moiré patterns appeared (marked as moiré B).
The moiré pattern with small period is consistent with the moiré A, while the period
of large moiré pattern is *3 nm, corresponding to a twisted angle of 2.3°. These
results reveal a twisted trilayer graphene. The close-up STM image of Fig. 3.8d
shows that on the interfacial region of the two type moiré patterns, the first layer
graphene is atomic continuous.
Electronic instabilities at the crossing of the Fermi energy with an VHSs in the DOS
often lead to new phases of matter such as superconductivity, magnetism or density
waves [12–14]. This implies the possibility of engineering material properties by
bringing EF and the VHSs together. However, in most materials this condition is
difficult to control [6].
Graphene, as a typical 2D material, has linearly dispersing bands at the K (K0) point
in the Brillouin zone, the so-called Dirac points, and a DOS that is linear and
48 3 STM Study of Twisted Bilayer Graphene
vanishes at Dirac point. However, in the case of single-layer graphene, the singu-
larity is too far from the Fermi energy and hence difficult to reach with standard
doping and gating techniques. Theoretical calculation predicted that the band
structure of twisted graphene will split at the point K, forming a saddle-like
structure, which will generate VHSs.
The twisted bilayer graphene with different periods can be synthesized on Rh
substrates by segregation mechanism, which provide an ideal platform to study the
twist-induced VHSs on graphene. For a twisted bilayer graphene, the Dirac points
of the two layers no longer coincide and the zero-energy states occur at k = −△K/2
in the first layer and k = △K/2 in the second layer. The displaced Dirac cones cross
at energies ±ħvF△K/2 and two saddle points are unavoidable along the intersection
of the two cones when there is a finite interlayer hopping. Figure 3.9 show the STM
image and STS spectra of a twisted bilayer graphene with a twisted angle *1.1°.
Although the peak heights and degree of asymmetry of the spectra depend on their
positions in the moiré pattern, all STS spectra show two peaks near Fermi level. The
tunneling spectrum gives direct access to the local DOS of the surface at the
position of the STM tip. The two peaks in the tunneling spectra are attributed to the
two VHSs in DOS, which originate from the two saddle points of the band
structure, as shown in Fig. 3.9f.
Fig. 3.9 VHSs in twisted bilayer graphene on Rh substrates. a–c A moiré pattern of twisted
bilayer graphene with a period of 4.9 nm. d A typical STM image of a graphene moiré pattern with
the twisted angle of 1.1°. e STS curves recorded on bright and dark regions of the moiré pattern at
positions indicated in panel (d). The spectra show two peaks attributed to two VHSs. f Electronic
band structure of the twisted graphene bilayer with a finite interlayer coupling
3.3 Angle-Dependent VHSs in Twisted Bilayer Graphene 49
Figure 3.10 shows six typical STM images of twisted bilayer graphene with dif-
ferent twisted angles, which provide an ideal platform to study the twisted
angle-dependent VHSs. Figure 3.10g shows eight typical STS curves recorded on
twisted bilayer graphene with different twisted angles. In order to ascertain the
reproducibility of the results, several tens of STS spectra on different positions of
the graphene moirés are recorded. Although the peak heights and peak positions of
the spectra vary slightly, the main features of these dI/dV–V curves are almost
completely reproducible. Obviously, the two VHSs show a strong angle-dependent
energy difference. Additionally, the positions of the two VHSs are not always
symmetric around the Fermi level, suggesting charge transfer between the graphene
and the substrate. The magnitude of the charge transfer should mainly depend on
the coupling strength between the sublayer graphene and the substrate, which varies
in different samples.
Figure 3.11 summarizes the energy difference of the two VHSs △EVHS as a
function of the twisted angles. Except at 1.3°, △EVHS increase linearly with the sine
of the twisted angle. Our experimental result reveals that the energy difference of
the two VHSs follows △EVHS * ħvF△K between 1.0° and 3.0°, where vF *
1.1 106 m/s. It suggests that the displaced Dirac cones of a slightly twisted
graphene bilayer cross and two saddle points are formed along the intersection of
the two cones at energies about ±ħvF△K/2. This indicates that the rotation angle
between graphene sheets does not result in a significant reduction of the Fermi
velocity.
Fig. 3.10 VHSs on different graphene moiré patterns. a–f STM images of twisted bilayer
graphene with six different twisted angles marked on the images and the corresponding moiré
periods are 14.1, 10.8, 8.8, 7.4, 5.8, 4.9 nm, respectively. g STS curves taken on twisted bilayer
graphene with different twisted angles
50 3 STM Study of Twisted Bilayer Graphene
Fig. 3.11 The energy difference of the two VHSs △EVHS as a function of the twisted angles. The
solid black circles are the average experimentally measured values obtained from several tens of
STS curves for each twisted angle and the error bars in energy represent the minimum and
maximum observed energies of △EVHS. The red dashed line is plotted according △EVHS = ħ vF
△K, where vF * 1.1 106 m/s, △K = 2K sin (h/2)
Fig. 3.12 Graphene wrinkles on Rh substrates a–c AFM images of wrinkles formed on few layer
graphene and the corresponding line profile. d–f AFM images of wrinkles formed on monolayer
graphene and the corresponding line profile
The wrinkles on the multilayer graphene and monolayer graphene exhibit distinct
characteristics. Figure 3.12a, b are the AFM height and phase images of the
wrinkles on multilayer graphene, which show bright-strip feature. The wrinkle is
about 80 nm wide and about 35 nm high. Differently, the AFM height image of
monolayer graphene on Rh foil only shows the high roughness of the substrate, and
the height information of the graphene is completely concealed. The AFM phase
image can distinguish monolayer graphene and bare Rh foil substrates, as well as
graphene wrinkles. The line profile of wrinkle on monolayer graphene shows that
the width is about 60 nm and the height is about 2 nm. The above comparison
reveals that the wrinkles on multilayer graphene has high height and regular shape,
meanwhile the wrinkles formed on monolayer graphene has low wrinkle height and
irregular shape. This is due to formation mechanism of wrinkles on graphene during
the cooling process after growth. For the multilayer graphene, the wrinkles tend to
form at the domain edge or the step edge of Rh substrate to release the interfacial
strain. As reported in literature [15], with the increased graphene thickness, the
small graphene wrinkles are combined into large winkles, thus the graphene wrinkle
density decreases but height increases. The merging of wrinkles dose not occurred
on monolayer graphene because the strain can be easily released on the graphene
edges.
52 3 STM Study of Twisted Bilayer Graphene
It is well established that strain and geometry could affect the band structure of
graphene dramatically. Figure 3.13a shows a typical STM image of a graphene
wrinkle at the step edge of the Rh surface. Graphene moiré patterns with identical
period of 2.8 nm are observed on both the right and left terraces flanking the
wrinkle, as shown in Fig. 3.13b, which is attributed to a twisted angle of *5.1°
between the adjacent graphene layers. The average height and width (peak width at
half-height) of the wrinkle are about 7.9 and 8.2 nm, respectively.
Figure 3.13d, e show ten dI/dV-V curves recorded at different positions along
the line profile in Fig. 3.13c. Curves 1–7, 9 and 10 are measured on the flat region
of the twisted bilayer graphene on the Rh terraces. Curve 8, which shows distinct
characteristics comparing with that of the other curves, is measured on the strained
wrinkle.
The two pronounced peaks near Fermi level in curves 1–7, 9 and 10 are
attributed to the two VHSs on twisted bilayer graphene, and the energy difference of
the two peaks decreases when the experimental position approaches the strained
wrinkle. For the curve 1, the energy difference of the two VHSs (△EVHS) is
*0.59 eV. For the curve 7, the value of △EVHS decreases to 0.38 eV. The curve 8
shows sharp peaks which are attributed to the landau levels (LLs) of the strained
graphene wrinkle. The large strain stress and out-of-plane distortions of the gra-
phene wrinkle can result in a large pseudo-magnetic field on its electronic
Fig. 3.13 Strain and curvature induced evolution of electronic band structure in graphene
wrinkle. a Large-area STM image of a twisted bilayer graphene with a wrinkle along a step of the
Rh foil. Moiré pattern with identical period appears on both the right and left terraces flanking the
wrinkle. b Atomic-resolution STM image of the twisted graphene bilayer in the red frame of panel
(a). It shows a moiré pattern with the period of 2.8 nm and twisted angle of *5.1°. c A line profile
along the black curve in panel (a). The average height and width (peak width at half-height) of the
wrinkle are about 7.9 and 8.2 nm, respectively. STS measurements at different positions along the
line profile show quite different characteristics, as shown in (d) and (e). The spectra have been
vertically offset for clarity. Curves 1, 2, 3, 4, 5, 6, 7, 9 and 10 are measured on the flat twisted
bilayer graphene. The two peaks flanking the Dirac point, as pointed out by the arrows, are
attributed to the two VHSs. The sharp peaks in curve 8 are attributed to the Landau quantization of
the strained bilayer graphene in a large pseudo-magnetic field
3.4 STM/STS Study of Graphene Wrinkles 53
3.5 Conclusion
In this chapter, STM is used to study the growth characteristics and electronic
structure of bilayer graphene prepared on Rh substrates. It is revealed that the
bilayer graphene is non-AB stacking and shows moiré patterns in STM images. The
moiré period (D) and the twisted angle (h) follows D = d/2 sin (h/2), that is, the
moiré period decreases as the twisted angle increases. There is no preferred twist
angle between bilayer graphene, and the range can be over 1°–12°. VHSs are
observed on the graphene moiré patterns using STS, and the energy difference
between the two peaks are coincided with △EVHS = ħvF△K. Furthermore, the
△EVHS which is angle-dependent, increases with the increase of the twist angle
between the coupled graphene layers.
In addition, the bilayer graphene on the Rh substrate also has a high density of
wrinkles, which induce the evolution of local electronic properties of the twisted
bilayer graphene. The △EVHS decreases with increasing lattice deformation and the
states condensed into well-defined LLs, which mimic the quantization of massive
chiral fermions in a magnetic field of about 100 T, along a graphene wrinkle.
References
1. Binnig G, Rohrer H (1983) Scanning tunneling microscopy. IBM J Res Dev 126:236
2. Chen CJ (1993) Introduction to scanning tunneling microscopy (Vol. 4). Oxford University
Press on Demand
3. Stolyarova E, Flynn GW (2007) High-resolution scanning tunneling microscopy imaging of
mesoscopic graphene sheets on an insulating surface. Proc Natl Acad Sci 104:9209
4. Li X, Cai W, An J, Kim S, Nah J, Yang D, Piner R, Velamakanni A, Jung I, Tutuc E,
Banerjee SK, Colombo L, Ruoff RS (2009) Large-area synthesis of high-quality and uniform
graphene films on copper foils. Science 324:1312
5. Gao T, Xie S, Gao Y, Liu M, Chen Y, Zhang Y, Liu Z (2011) Growth and atomic-scale
characterizations of graphene on multifaceted textured pt foils prepared by chemical vapor
deposition. ACS Nano 5:9194
6. Li G, Luican A, Lopes dos Santos JMB, Castro Neto AH, Reina A, Kong J, Andrei EY (2009)
Observation of Van Hove singularities in twisted graphene layers. Nat Phys 6:109
7. Yankowitz M, Xue J, Cormode D, Sanchez-Yamagishi JD, Watanabe K, Taniguchi T,
Jarillo-Herrero P, Jacquod P, LeRoy BJ (2012) Emergence of superlattice dirac points in
graphene on hexagonal boron nitride. Nat Phys 8:382
8. Zhao L, He R, Rim KT, Schiros T, Kim KS, Zhou H, Gutiérrez C, Chockalingam SP,
Arguello CJ, Pálová L (2011) Visualizing individual nitrogen dopants in monolayer graphene.
Science 333:999
54 3 STM Study of Twisted Bilayer Graphene
9. Rutter GM, Crain JN, Guisinger NP, Li T, First PN, Stroscio JA (2007) Scattering and
interference in epitaxial graphene. Science 317:219
10. Lahiri J, Lin Y, Bozkurt P, Oleynik II, Batzill M (2010) An extended defect in graphene as a
metallic wire. Nat Nanotechnol 5:326
11. Červenka J, Katsnelson MI, Flipse CFJ (2009) Room-temperature ferromagnetism in graphite
driven by two-dimensional networks of point defects. Nat Phys 5:840
12. Dewar MJS (1987) A new mechanism for superconductivity. Angew Chem, Int Ed Engl
26:1273
13. Gonzalez J (2008) Kohn-Luttinger superconductivity in graphene. Phys Rev B 78:205431
14. Fleck M, Ole AM, Hedin L (1997) Magnetic phases near the Van Hove singularity in S- and
D-band Hubbard models. Phys Rev B 56:3159
15. Chae SJ, Güneş F, Kim KK, Kim ES, Han GH, Kim SM, Shin HJ, Yoon SM, Choi JY,
Park MH (2009) Synthesis of large-area graphene layers on poly-nickel substrate by chemical
vapor deposition: wrinkle formation. Adv Mater 21:2328
Chapter 4
Controlled Synthesis of in-Plane h-BN-G
Heterostructures
The theoretical study of the hybrid materials consisting of B, N and C have been
studied from the 1980s. In addition to bandgap opening of graphene, the hybrid
materials also have many novel physical properties.
(1) Ultrahigh carrier mobilities
The carrier mobility of h-BN-G heterostructures has been calculated by DFT and
Boltzmann transport equations. For h-BN nanostructures (with zigzag edge)
embedded by graphene, the carrier mobility is related to the bandgap and the
content of carbon. The intrinsic carrier mobility is mainly distributed at 1.7 103–
1.1 105 cm2/Vs, while the bandgap of heterostructures are 0.38–1.39 eV. With
the increase of the content of C in heterostructures, the carrier mobility increases
exponentially, the values of which is mainly distributed in the range of 1.7 103–
7.9 104 cm2/Vs [1]. Thus, the h-BN-G heterostructure is an excellent nanoma-
terial which can simultaneously show high carrier mobility and non-zero bandgap.
(2) Carrier-induced antiferromagnetism
DFT calculations predict that graphene islands (with zigzag edge) embedded by
h-BN show intrinsic antiferromagnetic property. The magnetic moment of each
graphene island is derived from the extreme imbalance of two sublattices of gra-
phene. This method of trapping carbon atoms in h-BN lattices is of great signifi-
cance for the design of monomolecular magnets [2].
(3) Minimum thermal conductance in h-BN-G superlattice
The minimum thermal conductance of h-BN-G superlattice is revealed with
supercell size (ds) far below the phonon mean free path. The minimum value is
reached at a constant ratio of ds/L 5%, where L is the thickness of the super-
lattice. With the increase of ds, the localization of the confined mode is enhanced
while the number of confined modes decreases, which directly results in the min-
imum thermal conductance of h-BN-G heterostructures [3].
4.1 Introduction of in-Plane h-BN-G Heterostructures 57
Fig. 4.1 Detailed characterizations of the h-BNC hybrid films. a Photograph of a transparent
h-BNC film on a quartz slide after transferred. b AFM image of the uniform h-BNC hybrids
transferred onto SiO2/Si substrates. c High-resolution TEM image showing the section view of the
h-BNC films. d Moiré patterns of a bilayer h-BNC films characterized by HRTEM. e EELS
spectrum of the h-BNC films showing three K-shell excitation edges of B, C, and N, respectively.
f Atomic model of the h-BNC film showing hybridized h-BN and graphene domains. g Optical
image of a four-terminal h-BNC ribbon FET. h I–V curves of as-grown h-BNC with different
percentages in carbon measured at RT. i An R–T curve for a typical h-BNC ribbon with a width of
5 µm and a length of 11 µm. Reproduced with permission of [6]. Copyright [2010], Nature
Publishing Group
4.2 Experiment
Previous STM studies have separately examined pristine graphene and h-BN films
synthesized on various transition metals. Depending on the lattice mismatch,
symmetry of the supporting surface and interfacial interaction strength, a variety of
STM morphologies were achieved. In 2012, Sutter et al. [10] reported the mono-
layer h-BN-G heterostructure synthesis on Ru(0001) via a sequential growth of
graphene and h-BN in an ultrahigh vacuum (UHV) system. H-BN was found to be
preferentially nucleated from the edges of pre-deposited graphene domains, toward
the formation of a complete monolayer. Interestingly, narrow mixed phase con-
taining B, N and C was observed by the nano-Auger electron spectroscopy
(AES) line scan, which located at the interface region between pure graphene and
h-BN domains.
According to the early theoretical calculations, the B-C-N alloys were consid-
ered to behave as semiconductors with tunable band gap. To achieve delicate
control of the alloy formation, Loh et al. [11] reported a systematic study of the
phase transition from graphene to h-BC2N alloys through in situ B, N substitution
reactions. The substitution reaction occurred from a nearly complete monolayer
graphene on Ru(0001), and borazine molecules incorporated into the graphene
lattice by the following reaction:
Ru
G þ B3 N3 H3 ) BCx N þ CHx "
Fig. 4.4 Formation and characterization of h-BC2N alloys on Ru(0001). a Schematic illustration
of the substitutional doping of G/Ru(0001) gradually by borazine leading to the formation of
brick-and-mortar patterns and then domain-wise h-BN-G mosaic films. b STM image of the mixed
B-C-N alloy domains enclosed by segregated GNR networks, as a result of the substitution
reaction of G/Ru(0001) with a 10–20 Langmuir (L) borazine dosage at 900 K. c STM image of the
domain-wise h-BN-G mosaic film demonstrating complete phase segregation into h-BN and
graphene domains at 900 K. d, e High-resolution STM image of the brick-and-mortar patterns and
the corresponding magnified view of the square-enclosed region in (d) to show the atomic
structure of GNRs. f High-resolution STM image of the h-BC2N alloy, with the bright spots
representing the carbon atoms in dark BN matrices. g The statistic of C atom and BN atoms in
h-BCN alloys. Reproduced with permission from Ref. [11] Copyright (2013), Nature Publishing
Group
are also achieved. By combining STM/STS, Loh et al. [11] revealed the atomic
structures of the h-BCxN alloys with bright and dark contrasts in STM images. The
bright spots were explained to originate from enhanced local density of states
(LDOS), presumably due to carbon doping. By calculating the ratio of bright spots
(carbon) and dark regions (BN), Loh et al. derived a stoichiometric composition for
h-BC2N, which is consistent with related theoretical predictions (Fig. 4.4).
Density functional theory (DFT) calculations indicate that graphene on Rh(111)
displays weaker chemical interaction than that of graphene on Ru(0001), which
may largely prevent the element intermixing at the interface of h-BN-G. Note that,
the growth behavior and the atomic-scale structures of graphene and h-BN on Rh
(111) were systematically investigated with the aid of STM, respectively. Based on
these, the previously work of our group chose Rh(111) as a substrate for the
synthesis of in-plane h-BN-G heterostructure [12] (Fig. 4.5).
By using a two-step growth process, h-BN-G heterostructures were synthesized
on Rh(111) under UHV conditions. Two essential preconditions are proposed for
the seamless linking of the two analogues: (i) both materials follow the epitaxial
growth mechanism on Rh(111), in parallel with uniform lattice orientations; (ii) the
superstructures of graphene/Rh(111) and h-BN/Rh(111) demonstrate similar moiré
4.2 Experiment 63
Fig. 4.5 h-BN-G heterostructures evolved with an atomically sharp interface on Rh(111).
a Schematic diagram of the two-step growth process of the h-BN-G heterostructures on Rh(111).
b Large-scale STM image of the linking interface. c Line profiles along the blue arrows in (b),
showing the similar corrugations of the two analogues. d, e Atomically-resolved STM images of
an armchair linking boundary. f, g Atomically-resolved STM images of a zigzag linking boundary.
h Experimental statistics of the interface type in h-BN-G heterostructures. i DFT calculations of
binding energies for h-BN-G heterostructures on Rh(111) with armchair, N-C zigzag and B-C
zigzag interfaces. Reproduced with permission from Ref. [12]. Copyright (2013), American
Chemical Society
periods (*3.0, *3.2 nm) and surface roughness (*0.118, *0.124 nm), respec-
tively. In this regard, graphene and h-BN domains are patched together on Rh(111)
with the evolution of a sharp interface, which is significantly different from that on
Ru(0001) with the formation of an intermixing phase.
The seamless interfaces linking graphene and h-BN are identified by high res-
olution STM, and both zigzag and armchair type interface are noticed. Based on
statistical analysis, the ratio between zigzag and armchair type interfaces is *77.6
versus *22.4%. In order to understand the energetic origin of the preferred zigzag
type interface, DFT calculations are also carried out. The zigzag interfaces with
either B–C or N–C links show larger binding energies than that of armchair ones,
revealing that the zigzag interfaces are energetically more favorable. Notably, the
DFT calculation results agree well with the experimental statistics.
In this regard, the h-BN-G heterostructure on Rh(111) provided a perfect plat-
form for investigating the atomic structure, the continuity, as well as the interface
type between graphene and h-BN.
Consequently, corrugated graphene films can be formed on Ru(0001) and Rh
(111), which are characterized with landscapes of humps and valleys. This is
64 4 Controlled Synthesis of in-Plane h-BN-G Heterostructures
Fig. 4.6 The band structure of graphene on Ru(0001). a, c In situ ARPES characterization of
graphene on Ru(0001) a before and c after O2 intercalation. b, d The schematic of the structures
corresponding to (a) and (c). Reproduced with permission from Ref. [13]. Copyright (2010),
American Chemical Society
mainly mediated by the strong interfacial chemical bonding and large lattice mis-
match between graphene and substrates. Meanwhile, the electronic property of
graphene is greatly influenced by the strong p–d orbital hybridization, leading to
complete disruption of the graphene p-band near the Fermi energy, as observed by
angle-resolved photoemission spectroscopy (ARPES) in Fig. 4.6a.
As reported by Near-edge X-ray absorption fine structure (NEXAFS) and
photoelectron spectroscopy (PES) results, 5d transition metals (TMs, such as Pt and
Ir) usually coupled weakly with graphene than 4d TMs (such as Rh and Ru), as
sorted by Pt(111) < Ir(111) < Rh(111) < Ru(0001). For the very weakly coupled
substrate of Pt(111), the adlayer-substrate interaction is not strong enough to induce
a uniform stacking of graphene with Pt(111). Polycrystalline graphene with various
oriented lattices and different moiré superstructures were usually evolved on Pt
(111), as observed from both STM images and low energy electron diffraction
(LEED) analysis. In contrast, graphene/Ir(111) and h-BN/Ir(111) prefer to present
uniform moiré patterns, and tend to show their intrinsic energy bands, respectively
[14]. It is thus reasonable to select the singlecrystalline Ir(111) substrate for the
synthesis and further investigation of the intrinsic electronic properties of the
in-plane h-BN-G heterostructures.
In order to identify graphene and h-BN on Ir(111), it is necessary to prepare the two
materials separately and study their specific surface structure. In this section, I will
introduce the preparation and STM characterization of graphene on Ir(111). The
preparation of h-BN on Ir(111) will be described in detail in the next Sect. 4.4.
4.3 Controlled Growth of Graphene on Ir(111) 65
The growth of graphene follows the UHV-CVD process: first, the substrate tem-
perature is raised to 900 K; then, the ethylene is introduced into the UHV chamber
for the growth of graphene. In this process, the ethylene is cracked on the metal
surface and the graphene lattice is formed. After graphene growth, the sample is
held at growth temperature for 15–30 min, which is conducive to self-healing of
graphene defects.
The factors governing the growth of graphene in UHV-CVD systems include:
(1) Precursor molecules: Ethylene is a molecular precursor commonly used in the
growth of graphene in UHV, because it is a flat molecular structure easily to
adsorb on metal substrate. Comparing with methane which is commonly used in
atmospheric pressure CVD system, ethylene can adsorb and crack at low tem-
perature and form graphene lattice, which is suitable for the growth of graphene
in UHV-CVD system with extremely low pressure and low temperature.
(2) Growth Temperature: The growth of graphene on Ir(111) in UHV system can
be realized at relatively lower temperature and wide temperature range (600–
1300 °C). When the growth temperature is low, the nucleation density
increases, but the orientation order is poor. Thus it is easy to synthesize a
variety of graphene domains with different orientations. When the growth
temperature is high, the nucleation density decreases, as follows the highly
ordered graphene film will be synthesized. At the same time, the growth
temperature also has a great influence on the growth speed. When the substrate
temperature is high, the migration rate of the carbon atoms on metal surface
increases, thus the effective reaction and the growth rate increase.
(3) Carbon Source Supplying: The supplying of carbon source during
UHV-CVD growth process is judged by the change of pressures of UHV
chamber. The increase of carbon source supplying will result in the raise of the
nucleation density and growth rate of graphene simultaneously.
(4) Relative Position of the Carbon Source and the Substrate: In UHV growth
system, the precursor molecules are led into the UHV chamber through the
stainless steel tube, meanwhile they are simultaneously led out by the molecular
pump. As a result, the gas partial pressure distributes nonuniformly in UHV
chamber. When the stainless steel tube is perpendicular to the sample, the
maximum gas partial pressure is obtained. When growing sub-monolayer
graphene or h-BN, it is necessary to leave the sample at a proper distance or
back to the stainless steel tube.
Following the growth step sketched in Fig. 4.7, the growth time is set to 3 min to
obtain sub-monolayer graphene on the Ir(111) substrate. When the growth process
66 4 Controlled Synthesis of in-Plane h-BN-G Heterostructures
is completed, the sample was sent to the analysis chamber by a magnetic transfer
rod in UHV system for further STM characterization. The STM tips used in this
work are Pt–Ir tip, and the STM images are obtained at constant-current mode at
room temperature.
Large-scale STM images (Fig. 4.8) shows that graphene nucleated along the Ir
(111) step. Because the height of Ir(111) atomic step is similar to that of monolayer
graphene, the domain region of sub-monolayer graphene is not easily to be
Fig. 4.8 Sub-monolayer graphene on Ir(111) a STM height image of large-scale sub-monolayer
graphene and the corresponding b current image. c Small range of sub-monolayer graphene STM
height image and the corresponding d current image
4.3 Controlled Growth of Graphene on Ir(111) 67
Fig. 4.9 STM images of a large-scale monolayer graphene on Ir(111) and corresponding
b magnified image of the area marked by rectangle in (a)
recognized when the scanning size is large. However, the current of the STM is
very sensitive to the different conductivity of the material. Thus the current phase
images could act as a good identification for graphene domains on Ir(111). The
zoom-in the STM images (Fig. 4.8c, d) shows that the domain shape of the gra-
phene tends to hexagons. Due to the relatively low growth temperature (900 K), the
graphene domains show different orientations.
Following the above growth process with 5 10−7 mbar of ethylene for
15 min, monolayer graphene can be obtained on Ir(111) at 1000 K as shown in
Fig. 4.9. The graphene domain can grow across the dense Ir(111) step and patch
seamlessly to form a monolayer structure. The magnified image shows that the
moiré patterns of graphene are uniform and highly ordered without any defects,
which indicates the high crystallization of monolayer graphene.
The STM images in Fig. 4.10 show that graphene on Ir(111) is featured by a
periodic superstructure with a period of 2.5 nm, named moiré patterns, which is
attributed to the lattice mismatch of 10 10C–C/9 9 Ir(111) as inferred from the
atomically resolved STM image. The moiré patterns detected by STM are usually
attributed to two factors: the one is the height of carbon atoms in different occu-
pations; the other one is electronic state modulated by the substrate periodically.
Since the electron doping effect of Ir(111) on graphene is relatively weak, the
electronic state of graphene exhibits an intrinsic state. The height measured by STM
on graphene/Ir(111) can be approximately expressed as the apparent corrugation of
graphene. As shown in Fig. 4.10b, the moiré corrugation of the graphene on Ir(111)
is about 0.04 nm. Compared with graphene on Rh(111) (0.16 nm) and Ru(0001)
(0.15 nm), graphene on Ir(111) shows less undulation, which confirms the weak
interaction between Ir(111) and graphene.
68 4 Controlled Synthesis of in-Plane h-BN-G Heterostructures
Fig. 4.10 Moiré structure of graphene on Ir(111). a High-resolution STM image of graphene
moiré patterns on Ir(111). b Line profile recorded along the dotted blue line marked in (a). c–
e Different STM contrasts of graphene moiré patterns on Ir(111). f High-resolution nc-AFM image
of graphene moiré patterns on Ir(111)
In different tunnel junctions, the graphene moiré patterns on the Ir(111) substrate
will exhibit different STM contrasts, for example, hexagons (c: VT = −0.002 V,
IT = 58.085 nA), nanomeshes (d: VT = −0.006 V, IT = 2.003 nA) and triangles (e:
VT = −0.002 V, IT = 40.953 nA). However, regardless of the changes in the
contrast of the picture, the superlattice of the moiré unit cells remain unchanged (as
shown in the blue parallelogram). Figure 4.10f is a high-resolution nc-AFM image
of graphene on Ir(111) showing the uniform moiré structures. It is noted that the
imaged dots are not the carbon atoms of graphene, but the pores of the graphene
honeycomb lattice. This phenomenon is due to the different force detected by tip on
carbon-site and pore-site [15]. DFT simulations of graphene moiré patterns on the Ir
(111) substrate shows that the highest point of the graphene is about 0.46 nm and
the in-plane corrugation is *0.02 nm (Fig. 4.11). In other words, monolayer
graphene on the Ir(111) substrate is almost a two-dimensional planar structure,
which agrees well with the experimental results.
Consequently, submonolayer or monolayer graphene on Ir(111) substrate can be
controllable synthesized by modulate the growth temperature, precursor supplying
and growth time. The graphene tends to epitaxially grow aligned with the lattice of
Ir(111) to form moiré patterns with period of 2.5 nm. Due to the weak interaction
between Ir(111) and graphene, graphene show small in-plane corrugation.
4.4 Controlled Growth of h-BN on Ir(111) 69
Fig. 4.11 DFT calculation results of graphene superlattice structure a Theoretical calculation
model b High-resolution STM image of graphene superlattice on Ir(111) c, d DFT Calculations
results of the physical undulation and line profile of the graphene moiré patterns
Following the growth process described above, the submonolayer h-BN can be
synthesized on Ir(111) surface in a 1 10−7 mbar precursor atmosphere for 5 min.
As shown in Fig. 4.14a, the h-BN islands tend to form a triangular domain shape.
As mentioned in Sect. 4.3, graphene islands tend to form a hexagonal domain
shape, thus it can be speculated that this result is attributed to the different lattice
symmetry of h-BN and graphene. The magnified STM image of Fig. 4.14b shows
that the h-BN moiré patterns in one domain show uniform orientation, but there are
a small number of defective moiré lattices, such as clover-shaped moirés. If increase
the precursor pressure to 5 10−7 mbar and growth time to 20 min, monolayer
h-BN could be obtained on Ir(111) surface, as shown in Fig. 4.14c.
The LEED measurement can identify the lattice constants and orientations of the
one or two layers on surface. The LEED image of monolayer h-BN on Ir(111)
shows two sets of diffraction points. The six points of the outer cycle correspond to
h-BN, and the three diffraction points of the inner cycle correspond to Ir(111)
surface. The lattice orientation of h-BN is highly consistent with Ir(111) lattice,
which confirms the uniform orientation and high quality of h-BN.
4.4 Controlled Growth of h-BN on Ir(111) 71
As shown in Fig. 4.15, the moiré period of h-BN on Ir(111) is about 3.2 nm,
and the formation of moiré patterns is due to lattice mismatch of (13 13)
BN/(12 12) Ir(111). Compared with graphene moirés, the h-BN moiré patterns
on Ir(111) show larger in-plane corrugations of *0.14 nm (Fig. 4.15b). In different
tunnel junctions, h-BN on the Ir(111) exhibits various STM contrasts, for example,
nanomeshes (c: VT = −0.002 V, IT = 43.639 nA), rectangles (d: VT = −0.002 V,
IT = 40.952 nA) and nanobubbles (e: VT = −0.002 V, IT = 40.952 nA).
Regardless of the different STM contrasts, the unit cells of the h-BN moiré patterns
remain unchanged (as shown by the blue parallelogram in the Fig. 4.15c–e).
Figure 4.15f is a high-resolution nc-AFM image of h-BN moiré patterns. The
honeycomb lattices of h-BN can be clearly observed and the B-site and N-site show
slight contrast difference due to the different interactions with tip.
The DFT simulation of the superlattices of h-BN on the Ir(111) substrate
(Fig. 4.18a, b, d) shows that the highest point of the h-BN is about 0.38 nm from Ir
(111) surface. The in-plane corrugation is 0.14 nm, which is consistent with the
experimental results. In the context of this chapter, it is mentioned that reason for
graphene and h-BN on the Rh(111) substrate can form a good in-plane atomic layer
is that the two materials have similar moiré periods and in-plane corrugations on the
Rh substrate. However, the experimental results and theoretical calculations of the
moiré structures of monolayer graphene and h-BN on the Ir(111) substrate show
that the two materials have extremely different moiré periods and corrugations on
72 4 Controlled Synthesis of in-Plane h-BN-G Heterostructures
Fig. 4.15 Moiré patterns of h-BN on Ir(111) a High-resolution STM image of h-BN moiré
patterns on Ir(111) b Line profile recorded along the dotted blue line marked in (a). c–e Different
STM contrasts of h-BN moiré patterns on Ir(111). f High-resolution nc-AFM image of h-BN moiré
patterns on Ir(111)
the Ir(111). This result means that graphene and h-BN on the Ir(111) need to
overcome the great height difference to patch together to form in-plane
heterostructures, which remains huge challenge (Fig. 4.16).
Fig. 4.16 DFT calculation results of h-BN superlattice structure a Theoretical calculation model
b High-resolution STM image of h-BN superlattice on Ir(111) c, d DFT Calculations results of the
physical undulation and line profile of the h-BN moiré patterns
4.5 Controlled Growth of h-BN-G Heterostructures on Ir(111) 73
I used the two-step growth method to prepare the h-BN-G heterostructures in UHV
chamber. firstly, submonolayer graphene is synthesized on Ir(111) substrates by
UHV-CVD method as shown in Fig. 4.17; after graphene growth, the sample is
transferred to the analysis chamber by a magnetic transmission rod for STM
characterization; then the sample is transferred back to the preparation chamber, and
h-BN is synthesized on bare Ir(111). After that, the sample is sent to the analysis
chamber again for STM characterization. The whole process is carried out in UHV
system.
As shown in Fig. 4.18a, the heterostructure of graphene embedded by h-BN
(named as G @ h-BN) can be formed by first step growth of submonolayer gra-
phene on Ir(111), with following the second step growth of h-BN. The STM image
of Fig. 4.18b shows that the submonolayer graphene domains demonstrate hexagon
shapes, which is consistent with the symmetry of graphene lattices. The inset is the
magnified STM image showing atomic-resolved structure of graphene. Two blue
arrows correspond to the zigzag direction of graphene lattices. Interestingly, we find
that the directions of the two arrows are parallel to the graphene domain edges, that
is, the graphene domain is terminated by the zigzag-type edge. This finding is of
great significance to the study of the atomic structure of the boundary in the next
chapter. After the second step of h-BN patching growth, as shown in Fig. 4.18c, d,
graphene and hexagonal boron nitride show different moiré patterns with periods of
2.5 and 3.2 nm, respectively. It is also noted that the corrugation of graphene is less
than that of h-BN. These two phenomena agree well with theoretical calculations.
More importantly, the domain boundary of graphene and h-BN is very clear, and
the graphene domain still maintains the hexagonal shape. This shows that the
growth of h-BN in the second step does not etch the original graphene domain or
the formation of mixed phase.
Similarly, exchanging the growth order of the two-step can also obtain the
heterostructure of h-BN embedded by graphene (called h-BN@G) (see Fig. 4.19).
The STM image shows that the sub-monolayer h-BN domain on Ir(111) is
Fig. 4.17 Schematic diagram of UHV-CVD two-step growth method for preparing h-BN-G
heterostructures
74 4 Controlled Synthesis of in-Plane h-BN-G Heterostructures
Fig. 4.18 Two-step growth method to synthesize monolayer in-plane h-BN-G heterostructures on
Ir(111) a sketch map of G @ h-BN heterostructure growth b STM image of submonolayer
graphene on Ir(111) (the inset is the atomic structure of the graphene domain) c, d The STM
images of monolayer in-plane h-BN-G heterostructures after the second growth of h-BN
triangular (Fig. 4.19b), which is consistent with the triple symmetry of h-BN lat-
tices. The inset of Fig. 4.19b is the magnified atomic resolution of the h-BN
domain. We can determine that the triangular h-BN domain is also terminated with
zigzag-type edges using the above-mentioned method. The h-BN @ G
Fig. 4.19 Two-step growth method to synthesize monolayer in-plane h-BN-G heterostructures on
Ir(111) a sketch map of h-BN @ G heterostructure growth b STM image of submonolayer h-BN
on Ir(111) (the inset is the atomic structure of the h-BN domain) c, d The STM images of
monolayer in-plane h-BN-G heterostructures after the second growth of graphene
4.5 Controlled Growth of h-BN-G Heterostructures on Ir(111) 75
4.6 Conclusion
This chapter first briefly introduces that graphene and h-BN can form in-plane
heterostructures, which have many excellent physical properties, such as, bandgap
opening, ultra-high carrier mobility, antiferromagnetic and half-semimetallic char-
acteristics. These physical properties are usually related to the domain shape,
domain size, the atomic structure and bonding type in the linking boundaries.
Therefore, it is important to investigate the atomic structure and electronic structure
of the linking boundaries of h-BN-G heterostructures.
The h-BN-G heterostructures synthesized on the noble metal surface in the UHV
system are an excellent model system, which excludes external disturbances such as
etching and impurities. Ru and Rh, which are usually used for graphene growth in
UHV system, have strong electron doping effect on the band structure of graphene,
which leads to the downward displacement of the p band of graphene away from
the Fermi level.
76 4 Controlled Synthesis of in-Plane h-BN-G Heterostructures
References
1. Wang J, Zhao R, Yang M, Liu Z, Liu Z (2013) Inverse relationship between carrier mobility
and bandgap in graphene. J Chem Phys 138:084701
2. Ramasubramaniam A, Naveh D (2011) Carrier-induced antiferromagnet of graphene islands
embedded in hexagonal boron nitride. Phys Rev B 84:173
3. Jiang JW, Wang JS, Wang BS (2011) Minimum thermal conductance in graphene and boron
nitride superlattice. Appl Phys Lett 99:043109
4. Pruneda JM (2010) Origin of half-semimetallicity induced at interfaces of C-BN
heterostructures. Phys Rev B 81:2149
5. Kaner RB, Kouvetakis J, Warble CE, Sattler ML, Bartlett N (1987) Boron-carbon-nitrogen
materials of graphite-like structure. Mater Res Bull 22:399
6. Ci L, Song L, Jin C, Jariwala D, Wu D, Li Y, Srivastava A, Wang ZF, Storr K, Balicas L
(2010) atomic layers of hybridized boron nitride and graphene domains. Nat Mater 9:430
7. Levendorf MP, Kim CJ, Brown L, Huang PY, Havener RW, Muller DA, Park J (2012)
Graphene and boron nitride lateral heterostructures for atomically thin circuitry. Nature
488:627
8. Liu Z, Ma L, Shi G, Zhou W, Gong Y, Lei S, Yang X, Zhang J, Yu J, Hackenberg KP (2013)
In-plane heterostructures of graphene and hexagonal boron nitride with controlled domain
sizes. Nat Nanotechnol 8:119
9. Liu L, Park J, Siegel DA, McCarty KF, Clark KW, Deng W, Basile L, Carlos Idrobo J, Li AP,
Gu G (2014) Heteroepitaxial growth of two-dimensional hexagonal boron nitride templated
by graphene edges. Science 343:163
10. Sutter P, Cortes R, Lahiri J, Sutter E (2012) Interface formation in monolayer graphene-boron
nitride heterostructures. Nano Lett 12:4869
11. Lu J, Zhang K, Liu XF, Zhang H, Sum TC, Castro Neto AH, Loh KP (2013) Order–disorder
transition in a two-dimensional boron–carbon–nitride alloy. Nat Commun 4:2681
12. Gao Y, Zhang Y, Chen P, Li Y, Liu M, Gao T, Ma D, Chen Y, Cheng Z, Qiu X (2013)
Toward single-layer uniform hexagonal boron nitride-graphene patchworks with zigzag
linking edges. Nano Lett 13:3439
13. Sutter P, Sadowski JT, Sutter EA (2010) Chemistry under cover: tuning metal-graphene
interaction by reactive intercalation. J Am Chem Soc 132:8175
14. Usachov D, Fedorov A, Vilkov O, Adamchuk VK, Yashina LV, Bondarenko L, Saranin AA,
Grüneis A, Vyalikh DV (2012) Experimental and computational insight into the properties of
the lattice-mismatched structures: monolayers of h-BN and graphene on Ir(111). Phys Rev B
86:119
15. Boneschanscher MP, van der Lit J, Sun Z, Swart I, Liljeroth P, Vanmaekelbergh D (2012)
Quantitative atomic resolution force imaging on epitaxial graphene with reactive and
nonreactive AFM probes. ACS Nano 6:10216
Chapter 5
Atomic and Electronic Structures
of h-BN-G Interfaces
boundaries. The dotted green lines marked in Fig. 5.1b, d located on the boundaries
are parallel to the vector directions of graphene and h-BN domains, which means that
the boundaries are zigzag type.
In addition to the heterostructures with regular domain shapes, some areas of the
graphene domains, for example, are not as experimental design to form a perfect
hexagonal structure due to uncontrollable factors in growth process. Figure 5.2 is a
heterostructure with irregular domain shape. The graphene domains nucleated along
the Ir(111) step edges and spliced together due to the dense steps. After the growth
of graphene, h-BN filled on the bare Ir(111) surface to form in-plane heterostruc-
tures. I investigate the two regions marked by the yellow and green boxes in
Fig. 5.2b to study the boundary types of irregular domain regions. As shown in
Fig. 5.2c, d, the atomic resolved STM images reveal that the boundary types in both
regions are zigzag types, which are the same as those in the heterogeneous structure
with regular domain shapes.
An atomic resolved STM image (Fig. 5.3a) is performed on a relatively
large-scale heterostructure with irregularly shaped domains, and the image was
subjected to Fourier transformation to obtain the inverted spatial diffraction points.
As shown in Fig. 5.3b, there is only one set of six symmetric diffraction points
5.1 Atomic Structure of h-BN-G Interface 79
Fig. 5.2 Interface of h-BN-G heterostructure with irregular shape a Large-scale STM image and
b the corresponding magnified STM image of the area marked by blue box in (a). c Atomic
resolved STM images of the area marked by yellow and green boxes, respectively
Fig. 5.3 The atomic continuity of the irregular-shaped the h-BN-G heterostructures on Ir(111)
a An irregular boundary on the interface of h-BN-G heterostructures. b Sketch map of the
formation of irregular-shaped domains and boundaries. c The corresponding FFT of (a). d,
e Atomic-resolved STM images of fence-shaped boundaries between h-BN and graphene domains
80 5 Atomic and Electronic Structures of h-BN-G Interfaces
located *1/2.1 nm−1 from the center. It is revealed that even if the domain shape is
irregular, graphene and h-BN patched seamlessly on the linking boundaries towards
forming a single layer of heterostructures. The origin of linking boundaries with
regular and irregular shape is modeled in Fig. 5.3c. In addition, we observed
amount of boundaries with fence-shaped structures (Fig. 5.3d, e). The progressive
magnified STM images show that the graphene and h-BN domains are staggered
with each other. It is noticed that from the high resolution atomic images, that
atomic type of the fence-shaped boundaries are zigzag type.
As reported in previous study of graphene and h-BN grown in UHV system, the
origin of this fence-shaped boundary depends on the growth process with unit in
moiré superstructure. Frenken et al. [1] have reported the growth of h-BN on Rh
(111) at 978 K and the in situ observation of the growth behavior by
high-temperature STM. The initial hexagonal boron nitride domain is an asym-
metric hexagonal structure with three long sides and three short sides, corre-
sponding to low energy termination edges and high energy termination edges,
respectively. When the precursor continues to be led in, the domain grows into a
triangular shape, which is consistent with the lattice symmetry of h-BN. During the
growth process, it was surprisingly found that the growth of h-BN is in the form of
moiré superstructures and one unit contains 169 N atoms and 169 B atoms. The
new h-BN superstructure tends to begin to grow on straight edges and can grow
simultaneously in two opposite directions along the edges. When the growth
temperature is low, h-BN nucleation density is relatively high and the growth rate is
slow, thus the adjacent superlattices cannot be linked into a straight edge, which
results in the formation of the fence-shaped edges. After the sequential patching of
graphene, the edges remain unchanged, that is, the formation of fence-shaped
boundary between graphene and h-BN.
In addition, I select some high-resolution STM images on linking boundaries
between graphene and h-BN as shown in Fig. 5.4. The atomic resolved images
Fig. 5.4 Randomly selected atomic-resolved STM images of h-BN-G boundaries showing
zigzag-type
5.1 Atomic Structure of h-BN-G Interface 81
show that the tested linking boundaries are all zigzag type. Consequently, I find that
the linking boundaries of most regions are zigzag type. Only one armchair-type
boundary with the length of about 20 nm is observed, as shown in Figure 5.5. The
atomic model at the linking boundary shows that the connection of the
armchair-type boundary is the alternately B–C bond and N–C bond.
In conclusion, over 99% of the h-BN-G linking boundaries on the Ir(111) is
zigzag type, which is not affected by the growth process, the domain size and the
domain shape, and the selectivity is higher than the h-BN-G heterostructure on Rh
(111) (zigzag type over 78%)
It is mentioned in Chap. 4 that after the second growth step of the h-BN, the shape
of the graphene domains does not change. Even if the order of growth changes, the
results are the same. Therefore, the linking type of the monolayer h-BN-G
heterostructures depends on the edge type of submonolayer graphene or h-BN
grown in the first step.
In the case of the growth of submonolayer graphene, the termination type of the
edge is controlled by both the kinetic and thermodynamic factors.
82 5 Atomic and Electronic Structures of h-BN-G Interfaces
(a) Kinetic factor. During the growth of graphene, a new lattice is formed at the
edge of the domain to enlarge the domain. The length of the new lattice at the
armchair-type edge is 1.23 Å, while the length of the new lattice at zigzag-type
edge is 2.13 Å. Therefore, the formation energy of the new graphene lattice at the
armchair edge is lower than that of zigzag edge. The calculated results by Ding
et al. [2] show that the new nucleation formation of graphene lattice at the zigzag
edge and the armchair edge is 2.24 and 1.60 eV, respectively, which is consistent
with the above experimental results. Due to the higher nucleation potential, the
growth rate of the zigzag edge is less than that of the armchair edge. The growth
rates of different types of graphene edges are listed as follows:
where AC, ZZ, S19, SL represent armchair-type edge, zigzag-type edge, the edge of
19° (compared to graphene lattice vector) and other types edges. According to
KWC theory, [3] the faster the growth rate of the edge type in the growth will
gradually become less, and ultimately, the slowest growth rate to survive the edge.
Thus, from the kinetic point of view, submonolayer graphene tends to be terminated
by zigzag-type edges.
(b) Thermodynamic factors. Theoretical calculations suggest that the
zigzag-type edge and armchair-type edge of the freestanding graphene have large
forming energies of 13.46 and 10.09 eV, respectively. The armchair-type edge is
more stable than the zigzag edge because the two unsaturated carbon atoms prefer
to form carbon-carbon triple bonds. However, when there is a metal substrate, the
formation energy of the two type edges can be decreased by nearly 50%. For
example, the formation energy of the zigzag and armchair edges of graphene on the
Ni(111) substrate is 6.95 and 7.14 eV, respectively. If the edge is connected to the
Ni(111) step, the formation of the two edges can be reduced to 3.45 and 4.79 eV,
respectively [4]. In addition, the calculated results on various metal substrates show
that the formation energy of the zigzag edge is lower than that of armchair edge, [5]
that is, the zigzag edge of the graphene is more stable on metal surface.
In addition to the theoretical results, the high-resolution STM can observe the
domain edge structure of graphene and h-BN to determine the terminating the
domain edge type. Figure 5.6 shows atomic-resolution STM images at the edge of
the graphene domains on the Ir(111) substrate. The lattice orientation of graphene
edge shows that the two domains are both terminated by zigzag edges.
It is noted that impurity particles are easily adsorbed on bare Ir(111) substrate,
especially on the edge of graphene domains. Furthermore, the STM tip needs to be
approached to the sample surface during high-resolution scanning. Thus the STM
tip is susceptible to be disturbed by adsorbed impurity particles, and it is difficult to
obtain the atomic-resolution images of the graphene edges. In the Chap. 4, we have
mentioned a simple method for judging the edge types of submonolayer graphene
and h-BN. As shown in Fig. 5.7, I selected a magnified area in a graphene domain
to get the atomic lattice orientation of this domain, and compared the direction of
graphene lattice with graphene domain edges to determine the edge type of this
5.2 The Formation Mechanism of the Zigzag-Type Boundaries 83
Fig. 5.6 The edge type of submonolayer graphene observed by STM. a, b STM images of
different graphene domains both terminated by zigzag-type edges
Fig. 5.7 Method for determining edge type of graphene and h-BN domains on Ir(111)
a, b Hexagonal graphene domains and the atomic-resolved STM images of the domains.
c, d Triangular h-BN domains and the atomic-resolved STM images of the domains
edges, which coincides with the fact shown in the previous section that the linking
boundaries of h-BN-G heterostructures prefer to be zigzag type.
In order to track the formation of linking boundaries, we capture the initial growth
of h-BN from the edge of the graphene domain. As shown in Fig. 5.8, h-BN tends
to nucleate along the edge of the hexagonal graphene domain, and the width of the
h-BN coincides with the period of the h-BN moiré pattern on Ir(111). The
high-resolution STM image shows that the h-BN lattices are heteroepitaxial along
the graphene edge to form a zigzag boundary. Figure 5.8b shows that there are a
few defects in graphene domains, which will be self-repaired during the
high-temperature annealing process after growth.
Since the STM image cannot identify whether the zigzag boundary is an N–C
linkage or a B–C linkage, we use the DFT calculation to simulate the in-plane
h-BN-G heterostructure with two types of zigzag boundaries on Ir(111). As shown
in Fig. 5.9, both N–C and B–C zigzag linkages can be formed on Ir(111). On the
interface, the graphene lattice is pulled down closing to the metal substrate by
h-BN. The height of the B–C linkage and the N–C linkage to substrate is 2.16 and
2.58 Å, respectively, and the B–C linkage is more stable. The bond lengths of B–C
and N–C linkage range from 1.59 ± 0.01 and 1.41 ± 0.01 Å, and the adjacent C-C
bonds are 1.47 ± 0.03 and 1.49 ± 0.03 Å, respectively. The small changes in the
length of the chemical bond have little effect on the entire two-dimensional
structures.
Fig. 5.8 The initial process of the heterepitaxial growth of h-BN on graphene domain edge
5.3 Electronic Structure of h-BN-G Interface 85
Fig. 5.9 The DFT calculation of h-BN-G heterostructure on Ir(111) with both of B–C and N–C
zigzag boundaries
In order to determine the electronic doping effect of Ir(111) on graphene and h-BN,
it is necessary to make dI/dV measurements of monolayer graphene and h-BN
respectively on Ir(111). The dI/dV spectra were measured at 77 K by recording the
output of a lock-in system with the manually disabled feedback loop. A modulation
signal of 5 mV, 952 Hz was selected under a tunneling condition of 1 V, 20 pA.
Figure 5.10a, b is the STM image of the graphene on Ir(111) and the corresponding
dI/dV curve, which shows the Dirac cones of intrinsic graphene at 0 eV.
Figure 5.10c, d is the STM image of h-BN on Ir(111) and the corresponding dI/dV
curve, which shows a broad bandgap insulator characteristic, agreeing well with the
theoretical value of 5.9 eV.
Figure 5.11a is an STM image of a monolayer h-BN-G heterostructure formed
by a hexagonal graphene domain surrounded by h-BN, with a graphene domain size
of about 40 nm. Sequential line scan dI/dV data were obtained across a zigzag
boundary from graphene to h-BN regions, as shown in Fig. 5.11b. Each spectrum
was labeled with a letter corresponding to the spatial position in Fig. 5.11a. It can
be clearly seen that at the location A far from the boundary the spectrum appears to
be a V-shape with the Dirac point locating at *0 eV, which is typical of an
Fig. 5.10 The intrinsic electronic states of graphene and h-BN on Ir(111) a STM image of
graphene on Ir(111) and b the corresponding dI/dV curve. c STM image of h-BN on Ir(111) and
d the corresponding dI/dV curve
5.3 Electronic Structure of h-BN-G Interface 87
Fig. 5.11 Electronic structure on the interface of h-BN-G heterostructures a STM image of a
typical h-BN-G interface on Ir(111) b A series of dI/dV curve across the h-BN-G interface
Due to the difference of apparent height and in-plane undulations of graphene and
h-BN on Ir(111) substrate, it is not conductive to measure the intrinsic interface
state on the boundary. Thus we just measured the electronic state of regions 3 nm
away from the linking boundary. Liljeroth et al. [8] reported that the h-BN-G/Au/Ir
(111) sandwich structure was formed by intercalating 5–10 layers of Au between
the Ir(111) and h-BN-G heterostructures, further weakening the interaction between
adlayer and the substrate. The Au atoms are well packed on the Ir(111) surface to
form the perfect Au(111) facet which displays herringbone reconstruction.
The high resolution STM images of the zigzag-type h-BN-G hetero-boundary
show that the canopy-like structure appears on the boundary, which is assigned to
the B–C bonded zigzag boundary. A state near the Fermi level appeared in dI/dV
curves measured on the boundary under different tip condition, and disappeared
when the tip moved from the boundary to the graphene or h-BN domains. The
graphene domain near the boundary shows only a set of sublattices. The breakage
of this lattice symmetry is due to the absence of intervalley scattering. The location
of the interface state is close to 0 eV, which is very similar to the edge state of
88 5 Atomic and Electronic Structures of h-BN-G Interfaces
References
1. Dong G, Fourré EB, Tabak FC, Frenken JWM (2010) How boron nitride forms a regular
nanomesh on Rh(111). Phys Rev Lett 104:096102
2. Ma T, Ren W, Zhang X, Liu Z, Gao Y, Yin LC, Ma XL, Ding F, Cheng HM (2013)
Edge-controlled growth and kinetics of single-crystal graphene domains by chemical vapor
deposition. Proc Natl Acad Sci 110:20386
3. Sekerka RF (2005) Equilibrium and growth shapes of crystals: how do they differ and why
should we care? Cryst Res Technol 40:291
4. Gao J, Yip J, Zhao J, Yakobson BI, Ding F (2011) Graphene nucleation on transition metal
surface: structure transformation and role of the metal step edge. J Am Chem Soc 133:5009
References 89
5. Shu H, Chen X, Tao X, Ding F (2012) Edge structural stability and kinetics of graphene
chemical vapor deposition growth. ACS Nano 6:3243
6. Liu L, Park J, Siegel DA, McCarty KF, Clark KW, Deng W, Basile L, Carlos Idrobo J, Li AP,
Gu G (2014) Heteroepitaxial growth of two-dimensional hexagonal boron nitride templated by
graphene edges. Science 343:163
7. Phark S, Borme J, León Vanegas A, Corbetta M, Sander D, Kirschner J (2012) Scanning
tunneling spectroscopy of epitaxial graphene nanoisland on Ir(111). Nanoscale Res Lett 7:255
8. Drost R, Uppstu A, Schulz F, Hämäläinen SK, Ervasti M, Harju A, Liljeroth P (2014)
Electronic states at the graphene-hexagonal boron nitride zigzag interface. Nano Lett 14:5128
Chapter 6
Summary
We used the APCVD growth system to carry out the segregation growth of gra-
phene by using Rh as substrates. By controlling the growth conditions and the
cooling rate after growth, we can controllably obtain monolayer graphene, bilayer
graphene and multilayer graphene on Rh foils. Wrinkles can be observed on gra-
phene synthesized on Rh foils, which is due to the mismatch of coefficient of
thermal expansion between graphene and Rh substrates. The height of wrinkles on
monolayer graphene and multilayer graphene is *3 and *10 nm, respectively,
and the wrinkles have specific orientations, which are closely related to the lattice
orientation of the Rh facets. We can also controllable synthesize bilayer graphene
on Rh(111) substrate. However, there is no wrinkles observed, which is indicated
that graphene wrinkles are originated on Rh grain boundaries and they are the main
channel of stress release during the cooling process. Based on the growth charac-
teristics of graphene, We propose the combined CVD and segregated growth
mechanism of graphene on Rh substrates. The main path of carbon segregation is
Rh grain boundaries, meanwhile the Rh atom steps and terraces are also the way for
carbon segregation.
The bilayer graphene synthesized on Rh are non-AB stacked, which is confirmed
by the observation of moiré patterns by STM. The moiré period D and the twisted
angle h follow the law: D ¼ 0:246=2 sin ðh=2Þ nm: On the graphene moiré patterns,
we observed VHSs near the Fermi level by STS. The energy gap DEVHS increases
with the increase of the twist angles, and in accordance with the equation DEVHS ¼
h vF DK: In addition, a clear Landau level was observed at the apex of the graphene
wrinkles.
It is predicted that the in-plane heterostructures of graphene and h-BN can
effectively open the bandgap while maintaining high carrier mobility. We designed
the two-step growth method in UHV system and prepared the h-BN@G
heterostructures and G-h-BN heterostructures on Ir(111) substrates. Although
graphene and h-BN show different physical height and in-plane corrugation on the
Ir(111) substrate, they can be seamlessly spliced to form in-plane heterostructures.
Moreover, we find that >99% of the linking boundaries are zigzag type, which is
© Springer Nature Singapore Pte Ltd. 2018 91
M. Liu, Controlled Synthesis and Scanning Tunneling Microscopy Study
of Graphene and Graphene-Based Heterostructures,
Springer Theses, https://doi.org/10.1007/978-981-10-5181-4_6
92 6 Summary
consistent with the results of the DFT calculations and shows higher selectivity than
that on Rh(111) substrate (78%). The formation of zigzag-type boundary depends
on thermodynamic stability and two-step growth methods. The dI/dV measure-
ments on the linking interface show that graphene and h-BN tend to exhibit their
intrinsic electronic structures. The result means that h-BN does not directly change
the electronic structure of the graphene, but exists as an insulating state.
Vita