Springer Theses

Recognizing Outstanding Ph.D. Research

Mengxi Liu

Controlled Synthesis
and Scanning
Tunneling Microscopy
Study of Graphene
and Graphene-Based
Heterostructures
Springer Theses

Recognizing Outstanding Ph.D. Research

Aims and Scope

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.

Theses are accepted into the series by invited nomination only

and must fulfill all of the following criteria
• They must be written in good English.
• The topic should fall within the confines of Chemistry, Physics, Earth Sciences,
Engineering and related interdisciplinary fields such as Materials, Nanoscience,
Chemical Engineering, Complex Systems and Biophysics.
• The work reported in the thesis must represent a significant scientific advance.
• If the thesis includes previously published material, permission to reproduce this
must be gained from the respective copyright holder.
• They must have been examined and passed during the 12 months prior to
nomination.
• Each thesis should include a foreword by the supervisor outlining the signifi-
cance of its content.
• The theses should have a clearly defined structure including an introduction
accessible to scientists not expert in that particular field.

More information about this series at http://www.springer.com/series/8790

Mengxi Liu

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

and Prof. Yanfeng Zhang

Center for Nanochemistry
National Center for Nanoscience College of Chemistry and Molecular
and Technology Engineering, Peking University
Chinese Academy of Sciences Beijing
Beijing People’s Republic of China
People’s Republic of China

ISSN 2190-5053 ISSN 2190-5061 (electronic)

Springer Theses
ISBN 978-981-10-5180-7 ISBN 978-981-10-5181-4 (eBook)
https://doi.org/10.1007/978-981-10-5181-4
Library of Congress Control Number: 2017952510

© Springer Nature Singapore Pte Ltd. 2018

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made. The publisher remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore
Supervisors’ Foreword

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.

Beijing, People’s Republic of China Prof. Zhongfan Liu

June 2015 Prof. Yanfeng Zhang

v
Abstract

Graphene, a novel two-dimensional atomic crystal, has attracted serious attention

because of its unique physical properties, such as high carrier mobility, high
Young’s modulus, high thermal conductivity, and low optical absorption. Due to
these excellent properties, graphene has been considered as a potential material for
applications on electronics, optics, and composite materials. However, it is negative
for graphene to show high performance in logic circuits because graphene is a
semimetal material with zero bandgap at Fermi level. Therefore, tuning the energy
band structure of graphene near Fermi level is highly favorable.
On one hand, the energy band structure of graphene is highly dependent on its
stacked order. Bernal-stacked bilayer graphene shows zero bandgap between
valence and conduction bands with a quadratic dispersion. Upon an electric field
applied perpendicular to the bilayer graphene would cause a bandgap opening. For
non-Bernal stacked bilayer graphene, the saddle points in the energy band structure
lead to a divergence in the density of states, as known as a van Hove singularity,
which would give rise to new phase of matter, such as superconductivity, mag-
netism, or charge density wave.
On the other hand, diverse approaches were developed for opening the bandgap
of graphene. The existing methods include gate-induced method, substrate-induced
method, etching graphene layer into graphene nanoribbons or nanomeshes, pat-
terned hydrogen adsorption and chemical doping. However, most approaches lead
to the decreasing of carrier mobility of graphene, which goes against the application
on devices. Theoretical calculations predicted that the heterostructure of graphene
and hexagonal boron nitride (h-BN) would be the most effective way to open the
bandgap of graphene and maintain the high carrier mobility spontaneously.
On the basis of the energy band engineering of graphene, this thesis mainly
encompasses two parts: The one is the synthesis of non-Bernal-stacked graphene
and the new electronic states near the Fermi level; the other is the designed fab-
rication of hexagonal boron nitride–graphene heterostructure and the electronic
structure on the interface.

vii
viii Abstract

1. Controlled Growth of Graphene on Rh Substrates and the Generation of van

Hove Singularities

(1) Controlled Growth of Graphene on Rh Substrates

Large-scale uniform graphene was synthesized on high carbon solubility sub-
strates of Rh foils using an ambient pressure chemical vapor deposition method.
It was interestingly to be found that, by increasing the cooling rate after the
growth process, the thickness of graphene could be tuned from multilayer to
bilayer and even to monolayer. Besides, there were plenty of wrinkles formed
on graphene grown on Rh substrates, and the wrinkles usually encircled into
nearly hexagonal or quadrilateral shapes. The wrinkles were proposed to be a
consequence of compressive stress during the cooling process because of the
different thermal expansion coefficients of graphene and Rh substrates. Based
on these results, a segregated growth mechanism was put forward for graphene
growth on Rh foils and it was emphasized that carbon atoms mostly segregated
from Rh grain boundaries, with a little portion of Rh steps and terraces.

(2) Scanning Tunneling Microscopy Study of Graphene on Rh Substrates

With the aid of scanning tunneling microscopy, the bilayer graphene on Rh
substrates was found that preferred to stack deviated from the Bernal stacking
geometry, resulting in the formation of moiré patterns. The moiré periods
decreased with the increasing twisted angles between the coupled graphene
bilayer. Interestingly, the robust van Hove singularities were observed on
various graphene moiré patterns near the Fermi level by scanning tunneling
spectroscopy. More importantly, the energy difference between the two van
Hove singularities (△EVHS) increased with the increasing twisted angles,
indicating that the van Hove singularities on twisted bilayer graphene were
angle-dependent on Rh substrates. Besides, the strain caused by graphene
wrinkles strongly affected the local energy band structures of the twisted bilayer
graphene, resulting in the decreasing △EVHS with the lattice deformations
increasing. Moreover, along a graphene wrinkle, the states condensed into
well-defined pseudo-Landau levels, which mimic the quantization of massive
Dirac fermions in a magnetic field of about 100 T. These results suggest that
twisted bilayer graphene could induce the electronic states near Fermi level, and
strained bilayer graphene could be an ideal platform to realize the
high-temperature zero-field quantum valley Hall effect.

2. Controlled Growth of h-BN-graphene Heterostructures and the Electronic States

on the Interface

(1) Controlled Growth of Quasi-freestanding h-BN-graphene Heterostructures

Monolayer graphene and h-BN were synthesized separately on weakly coupled
Ir(111) substrate using an ultrahigh vacuum chemical vapor deposition method.
With the aid of scanning tunneling microscopy, it was found that the layer
corrugation of h-BN was much greater than that of graphene on Ir(111)
Abstract ix

(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

3.2 STM Study of Bilayer Graphene . . . . . . . . . . . . . . . . ......... 43

3.2.1 The Stacking Order of Graphene Synthesized
on Rh Substrates . . . . . . . . . . . . . . . . . . . . . . ......... 43
3.2.2 The Linking Type of Bilayer Graphene on Rh
Substrates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.3 Angle-Dependent VHSs in Twisted Bilayer Graphene . . . . . . . . . 47
3.3.1 VHSs in Twisted Bilayer Graphene . . . . . . . . . . . . . . . . . 47
3.3.2 Angle-Dependent VHSs . . . . . . . . . . . . . . . . . . . . . . . . . . 49
3.4 STM/STS Study of Graphene Wrinkles . . . . . . . . . . . . . . . . . . . . 50
3.4.1 Graphene Wrinkles on Rh Substrates . . . . . . . . . . . . . . . . 50
3.4.2 Strain-Induced LLs on Graphene Wirnkle . . . . . . . . . . . . . 52
3.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
4 Controlled Synthesis of in-Plane h-BN-G Heterostructures . . . . . . . . 55
4.1 Introduction of in-Plane h-BN-G Heterostructures . . . . . . . . . . . . . 55
4.1.1 Intriguing Properties of h-BN-G Heterostructures . . . . . . . . 56
4.1.2 Synthesis of in-Plane h-BN-G Heterostructures . . . . . . . . . 57
4.2 Experiment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
4.2.1 UHV-STM System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
4.2.2 The Choice of Substrate . . . . . . . . . . . . . . . . . . . . . . . . . . 61
4.3 Controlled Growth of Graphene on Ir(111) . . . . . . . . . . . . . . . . . 64
4.3.1 Growth Process of Graphene on Ir(111) . . . . . . . . . . . . . . 65
4.3.2 The Controllable Growth of Graphene on Ir(111) . . . . . . . 65
4.3.3 Moiré Patterns of Graphene on Ir(111) . . . . . . . . . . . . . . . 67
4.4 Controlled Growth of h-BN on Ir(111) . . . . . . . . . . . . . . . . . . . . 69
4.4.1 Growth Process of h-BN on Ir(111) . . . . . . . . . . . . . . . . . 69
4.4.2 Controllable Growth of h-BN on Ir(111) . . . . . . . . . . . . . . 70
4.4.3 Moiré Patterns of h-BN on Ir(111) . . . . . . . . . . . . . . . . . . 71
4.5 Controlled Growth of h-BN-G Heterostructures on Ir(111) . . . . . . 73
4.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
5 Atomic and Electronic Structures of h-BN-G Interfaces . . . . . . . ... 77
5.1 Atomic Structure of h-BN-G Interface . . . . . . . . . . . . . . . . . . ... 77
5.2 The Formation Mechanism of the Zigzag-Type Boundaries . . . ... 81
5.2.1 Edge Types of Submonolayer Graphene Grown
on Metal Substrates . . . . . . . . . . . . . . . . . . . . . . . . . . ... 81
5.2.2 The Formation Process of Zigzag-Type Linking
Boundary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... 84
5.3 Electronic Structure of h-BN-G Interface . . . . . . . . . . . . . . . . ... 85
5.3.1 Research of Electronic Properties of h-BN-G Interface . ... 85
5.3.2 Electronic Structure of Monolayer Graphene and h-BN
on Ir(111) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... 86
Contents xvii

5.4 Discussion and Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Vita . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Chapter 1
Introduction

Graphene, a two-dimensional material consisted of sp2 hybridization carbon atoms,

has fascinated much attention owing to its extraordinary electronic, optical, mag-
netic, thermal, and mechanical properties, such as high carrier mobility
(*105 cm2 V−1 s−1), high Young’s modulus (*1.0 TPa), high thermal conduc-
tivity (*5000 W m−1 K−1) and optical transmittance (*97.7%) [1–7]. Therefore,
graphene has broad application prospects and has been applied to a variety of areas,
such as transparent conductive film, supercapacitors, high-frequency transistors,
logic transistors and proton exchange membrane [8–15].
The band structure of graphene depends on the number of layers and stacking
orders: the valence band and conduction band of the monolayer graphene meet at
the two vertices of the first Brillouin zone at the K and K′ points, and the energy and
momentum of electrons at the Fermi level exhibit a linear dispersion relationship,
which renders the quasiparticles in graphene formally identical to the massless
Dirac Fermions [1–3, 5]; AB-stacked bilayer graphene shows parabolic band
structure at the Fermi level [16–20]; Non-AB-stacking bilayer graphene shows van
hove singularities (VHSs) near Fermi level, which will induce a new phase of the
material, such as superconductivity, magnetism [21–23].
Since the high carrier mobility at room temperature, graphene is expected to be
applied in the next generation of logic devices. However, graphene is a
semi-metallic material with zero-bandgap. A great deal of effort is made to modify
the band structure of graphene, such as chemical doping, dual-gated bilayer gra-
phene, fabricating graphene nanoribbons (GNRs), graphene nanomesh (GNMs) and
heterstructures of graphene and hexagonal boron nitride [12, 14, 16, 17, 20, 24–33].
This chapter is divided into three sections: the first section will introduce the
basic properties of graphene and layer-dependent energy band structure; the second
section mainly introduces the current synthesis methods of graphene; the third
section briefly introduces the methods for bandgap modulation of graphene.

© Springer Nature Singapore Pte Ltd. 2018 1

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_1
2 1 Introduction

1.1 The Properties of Graphene

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].

1.1.1 The Structure of Graphene

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

1.1.2 The Band Structure of Graphene

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

(3) Non-AB-stacked bilayer graphene

For the 2D materials, a saddle point in the electronic band structure leads to a
divergence in the density of state (DOS), also known as a Van Hove singularity
(VHS). When the Fermi energy (EF) is close to the VHSs, the weak interactions can
1.1 The Properties of Graphene 5

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].

1.2 Synthesis of Graphene

In recent years, various techniques have been established for synthesizing graphene,
such as, mechanical exfoliation, chemical exfoliation, chemical vapor deposition
(CVD) and so on.

1.2.1 The Synthesis Methods of Graphene

(1) Mechanical exfoliation

The exfoliation method for preparing graphene includes mechanical exfoliation
and chemical exfoliation. The mechanical exfoliation method is known as the
so-called “scotch tape method” [1]. The number of graphene layers can be con-
trolled to a limited degree via the number of repeated peeling steps and then the
graphene layers can be transferred to the targeted surfaces for the further study. This
method is ideally suited for the investigation of the fundamental properties of
graphene, because it can generate monolayer graphene sheets of high quality easily
and low-cost. However, the disadvantages including poor reproducibility, low-yield
and the labour intensive processes required result in that it is difficult to scale this
process to mass production and thus lead to be used predominantly only for fun-
damental studies.
The chemical exfoliation method is using molecules or atoms to intercalate into
the graphite layer, thus weakening the interlayer van der Waals interactions and
resulting in the larger distance between graphite layers. Further ultrasound process
generates graphene solution [47]. Compared with the mechanical exfoliation
method, the graphene prepared by chemical exfoliation is low-cost, high-yield and
can be applied in coating, printing, transparent conductive film and energy storage.
The disadvantage of this method is that prepared graphene is small-size, uneven
thickness and poor conductivity.
(2) Reduced graphene oxide
Graphene oxide (GO) is obtained by treating graphite with strong oxidizers. The
bulk GO disperses in basic solutions to yield monolayer sheets, known as graphene
oxide analogy. The films consist of millions of randomly stacked flakes, leaving
nano-sized capillaries between them. Closing these nanocapillaries using chemical
1.2 Synthesis of Graphene 7

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

1.2.2 The Synthesis Methods of Bilayer Graphene

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.

1.3 Bandgap Opening of Graphene

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

1.3.1 Physical Method

(1) External electric field induced bandgap opening

Graphene’s unique electronic band structure of Dirac cone has led to fascinating
phenomena. With one more graphene layer added, the inversion symmetric
AB-stacked bilayer graphene also has a zero-bandgap structure in its pristine form,
featuring two nearly parallel conduction bands above two nearly parallel valence
bands near the Fermi level [3]. But it is predicted that a bandgap can be induced by
breaking the inversion symmetric of AB-stacked bilayer graphene. In 2009, Zhang
et al. [20] firstly demonstrated a widely field-tunable bandgap in bilayer graphene
with infrared absorption spectroscopy, as shown in Fig. 1.5. A largest bandgap of
250 meV was observed in their experimental spectral range, which emphasizes the
intrinsic potential of bilayer graphene for nanoelectronics. Since the tunable
bandgap can reach the infrared range, bilayer graphene may enable novel
nanophotonic devices for infrared light generation, amplification and detection.
However, this method is only applicable to AB-stacked bilayer graphene, and
depends on the presence of electric fields. It is not universal for opening the
bandgap of graphene.

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

(2) Substrate induced bandgap opening

In 2007, Gweon et al. [24] reported that epitaxial graphene on SiC has a bandgap
due to interaction with the substrate. The monolayer, bilayer and trilayer graphene
has bandgap of 0.26, 0.14 and 0.066 eV, respectively. When the number of gra-
phene layers exceeds five layers, the bandgap and Dirac points of graphene are
similar to those of HOPG, that is, more than five layers of graphene exhibit the
nature of HOPG (Fig. 1.6).
This method is simpler than the external electric field method. However, this
method is limited to the use of SiC substrate to growth epitaxial graphene.
Moreover, epitaxial graphene on SiC is electron doped and the Fermi level lies
above the gap. Thus to make graphene a viable semiconductor either it has to be
hole doped or the Fermi level must be moved in between the gap by applying gate
voltage.
(3) Quantum confinement
The mechanism of opening a bandgap in graphene is to destroy the lattice
symmetry of graphene, thus cutting the graphene layer into a specific shape is an
effective method for the modulation of graphene band structure. This method
mainly includes cutting graphene layer into graphene nanoribbons (GNRs) and
graphene nanomeshes (GNMs).
So far, the preparation of GNRs has developed a number of different methods.
One of the first and foremost techniques is etching, including e-beam lithography,
plasma etching and chemical etching. The first work of preparing GNRs by e-beam
lithography was reported by Kim et al. [14] in 2007 and the corresponding elec-
tronic measurements of GNRs were carried out. Finite bandgap was observed in
these ribbons which was inversely proportional to the ribbon width and in good
agreement with theoretical results. However, the limitation of GNRs prepared by
e-beam lithography is that they were not able to reach a width below 20 nm. In
2010, Dai et al. [67] used a mix gas of O2, H2, and NH3 to etch GNRs from the
edges in high temperature CVD system, achieving the thinnest GNRs with the
width less than 5 nm and the on/off ratio of GNRs-FET can reach 104.
CNTs can be considered to be rolled graphene nanoribbons, thus opening up the
CNTs is a practicable way to produce GNRs. The width of GNRs depends on the
diameter of the precursor CNTs. In 2009, Dai et al. [68] achieved this breakthrough.
They used Ar plasma to etch the multiwalled CNTs (MWCNTs), thus the GNRs
with different layers can be obtained. The width of produced GNRs is 10–20 nm
and the on/off ratio of 7 and 6 nm-width GNRs-FET exceeded 10 and 100,
respectively. Another approach is making a point defect on CNTs by oxidation and
then unrolled the CNTs from the point defect by ultrasound method. This process
can be likened to cutting the CNT by scissors [69]. The achieved GNRs usually
show high quality and high carrier mobility (1500 cm2 V−1 s−1).
GNM is a derivative nanostructure of graphene with a tunable bandgap. GNM
comprises graphene with a high-density array of nanoscale holes. The band
structure of GNM is determined by both of edge type and neck width between the
 1.3 Bandgap Opening of Graphene

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.

1.3.2 Chemical Method

(1) Chemical modification

Chemical modification is the most well-known method to modify the band
structure of graphene. The chemical modification of graphene essentially breaks the
symmetry of the AB lattice to open the bandgap. In the current report, when the
hydrogenation of graphene increased from 3 to 27%, the bandgap of graphene
increased from 0.1 to 0.8 eV. The continuous hydrogenation of graphene resulted in
the formation of the insulator, graphane [27].
The chemical modification can change the bandgap, as well as the surface
properties of graphene, therefore can be applicated in the composite materials,
bio-energy and other areas [72]. However, chemical modification has a great
influence on the electronic structure of graphene. For example, graphane has
completely different structure and properties from intrinsic graphene. In addition,
the dopant barrier has a great influence on the carrier mobility of the modified
graphene, which make it difficult to be applied to the electronic devices.
(2) Hybrids
Hexagonal boron nitride (h-BN) is a structural analog of graphene, but shows
insulating property with a band gap of 5.9 eV. In 2007, Brink et al. [73] predicted
theoretically that the graphene/h-BN heterostructures has a band gap of 53 meV,
meanwhile the Dirac fermions of graphene have a certain effective mass. However,
the bandgap opening based on this method is relatively small and has a dependency
on the substrate, which is somewhat difficult to apply to the electronic devices.
1.3 Bandgap Opening of Graphene 13

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].

1.4 Research Strategy

In this thesis, a combination of CVD synthesis and STM/STS measurement is used

to study the graphene and graphene-based heterostructures.
1. Controlled synthesis of graphene on Rh substrates and the induced VHSs
Segregation growth method is an effective method for the preparation of
non-AB-stacked graphene. I chose Rh with high carbon solubility as the substrate
and atmospheric CVD as the growth system for the preparation of graphene. By
controlling the growth pressure, carrier gas and the cooling rate after growth, it can
be effectively achieved to synthesize non-AB-stacked graphene. I used STM/STS to
study the atomic structure and electronic structures of non-AB-stacked graphene.
This part of the work will be introduced in Chaps. 2 and 3.
2. Controlled synthesis of h-BN-G heterostructures and the interfacial state
I prepared the in-plane h-BN-G heterostructure by using two-step growth
method with UHV-CVD system on Ir(111) substrate. The atomic structure and edge
state of the h-BN-G heterostructure are characterized by STM and STS. Moreover,
the relationship between the growth morphology and the interface type of the
heterostructures is studied. Moreover, the mechanism of bandgap opening in
h-BN-G heterostructure is proposed. This part of the work will be introduced in
Chaps. 4 and 5.

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

29. 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
30. Ding Y, Wang Y, Ni J (2009) Electronic properties of graphene nanoribbons embedded in
boron nitride sheets. Appl Phys Lett 95:123105
31. 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
32. 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
33. da Rocha Martins J, Chacham H (2010) Disorder and segregation in B–C–N graphene-type
layers and nanotubes: tuning the band gap. ACS Nano 5:385
34. Howard JB, McKinnon JT, Makarovsky Y, Lafleur AL, Johnson ME (1991) Fullerenes C60
and C70 in flames. Nature 352:139
35. Bachtold A, Hadley P, Nakanishi T, Dekker C (2001) Logic circuits with carbon nanotube
transistors. Science 294:1317
36. Huang PY, Ruiz-Vargas CS, van der Zande AM, Whitney WS, Levendorf MP, Kevek JW,
Garg S, Alden JS, Hustedt CJ, Zhu Y (2011) Grains and grain boundaries in single-layer
graphene atomic patchwork quilts. Nature 469:389
37. Abergel DSL, Fal’Ko VI (2007) Optical and magneto-optical far-infrared properties of bilayer
graphene. Phys Rev B 75:1418
38. Van Hove L (1953) The occurrence of singularities in the elastic frequency distribution of a
crystal. Phys Rev 89:1189
39. Dewar MJS (1987) A new mechanism for superconductivity. Angew Chem Int Ed Engl
26:1273
40. Gonzalez J (2008) Kohn-Luttinger superconductivity in graphene. Phys Rev B 78:205431
41. Fleck M, Ole AM, Hedin L (1997) Magnetic phases near the Van Hove singularity in S- and
D-B and Hubbard models Phys Rev B 56:3159
42. Li G, Luican A, Andrei EY (2009) Scanning tunneling spectroscopy of graphene on graphite.
Phys Rev Lett 102:176804
43. Miller DL, Kubista KD, Rutter GM, Ruan M, De Heer WA, First PN, Stroscio JA (2009)
Observing the quantization of zero mass carriers in graphene. Science 324:924
44. Britnell L, Ribeiro RM, Eckmann A, Jalil R, Belle BD, Mishchenko A, Kim YJ,
Gorbachev RV, Georgiou T, Morozov SV (2013) Strong light-matter interactions in
heterostructures of atomically thin films. Science 340:1311
45. Rechtsman MC, Zeuner JM, Plotnik Y, Lumer Y, Podolsky D, Dreisow F, Nolte S, Segev M,
Szameit A (2013) Photonic floquet topological insulators. Nature 496:196
46. Valla T, Fedorov AV, Johnson PD, Glans PA, McGuinness C, Smith KE, Andrei EY,
Berger H (2004) Quasiparticle spectra, charge-density waves, superconductivity, and
electron-phonon coupling in 2H-NbSe2. Phys Rev Lett 92:086401
47. Hernandez Y, Nicolosi V, Lotya M, Blighe FM, Sun Z, De S, McGovern TI, Holland B,
Byrne M, Gun’Ko YK, Boland JJ, Niraj P, Duesberg G, Krishnamurthy S, Goodhue R,
Hutchison J, Scardaci V, Ferrari AC, Coleman JN (2008) High-yield production of graphene
by liquid-phase exfoliation of graphite Nat Nanotechnol 3:563
48. Compton OC, Nguyen SBT (2010) Graphene oxide, highly reduced graphene oxide, and
graphene: versatile building blocks for carbon-based materials. Small 6:711
49. 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
50. Losurdo M, Giangregorio MM, Capezzuto P, Bruno G (2011) Graphene CVD growth on
copper and nickel: role of hydrogen in kinetics and structure. Phys Chem Chem Phys
13:20836
51. Li X, Cai W, Colombo L, Ruoff RS (2009) Evolution of graphene growth on Ni and Cu by
carbon isotope labeling. Nano Lett 9:4268
References 17

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

disadvantages of graphene growth on Rh substrate were summarized by comparison

with the growth of graphene grown on other noble metal substrates (Pt, Pd) with
high carbon solubility.

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.

2.1.1 Process Engineering for 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.

2.1.2 Factors for Graphene Segregation

The advantage of segregation growth is that it is easily to obtain the non-AB

stacked graphene, while the disadvantage is that it is difficult to control the
homogeneity of graphene. Thus how to improve the homogeneity of graphene film
2.1 Experiment 21

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:

Quantity of segregated carbon / MC  MT  n

where △C refers to the difference of carbon solubility in the metal substrate at

different temperature, △T refers to the temperature difference, and n refers to the
thickness of the substrate. It can be seen from the equation, with the reduced carbon
solubility and the thickness of the metal substrate, the amount of segregated carbon
will be reduced. For example, the carbon solubility of Fe, Co and Ni at 1000 °C are
7, 1.6, 0.9 at.%, respectively. Under the same growth conditions, monolayer
22 2 Controllable Synthesis of Graphene on Rh

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

The existing research on segregation growth of graphene is mainly focused on three

kinds of substrates: (1) Cheap metals with high carbon solubility, such as Ni [8–10],
Fe [11], and Co [12, 13]. This kind of metals can be used to prepare large-scale
graphene with high coverage. However, due to the high carbon solubility, it is
difficult to control the uniformity and thickness. (2) Noble metals, such as Ru(0001)
[14–17] and Ir(111) [18]. Such metals can be used to prepare monolayer or bilayer
singlecrystalline graphene. The preparation of graphene on noble metals is usually
carried out in UHV chamber. Moreover, the singlecrystalline noble metals are
expensive and the as-grown graphene is difficult to be transferred onto other
24 2 Controllable Synthesis of Graphene on Rh

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

2.1.4 Experimental System

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

Fig. 2.5 The experimental system for graphene growth

26 2 Controllable Synthesis of Graphene on Rh

Fig. 2.6 The growth process

of graphene in APCVD
system

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.

2.2 Controllable Growth of Graphene on Rh Substrates

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

Fig. 2.8 Graphene grown on

Rh foils under LPCVD
system. a, b SEM images of
as-grown graphene on Rh
foils. c OM image of
graphene after transferred
onto SiO2/Si substrate

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

Fig. 2.10 The identification of graphene thickness. a OM image of graphene on SiO2/Si.

b Raman spectrum of graphene. c, d Large-scale TEM image and the magnified image of the
folded edge. e, f The SAED of graphene shown in (c)

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)

2.3 The Mechanism of Graphene Growth on Rh

Substrates

Based on the high carbon solubility of Rh substrates and the characteristics of

graphene growth under different cooling rates, I proposed a mechanism for gra-
phene growth on Rh substrates under APCVD conditions. As shown in Fig. 2.15, at
1000 °C, CH4 are decomposed on Rh surfaces for its perfect ability to catalyze
hydrocarbon decomposition, and contemporary, carbon atoms dissolve into bulk Rh
because of its high carbon solubility. In the cooling process of CVD growth, the
solubility of carbon becomes lower with decreasing the sample temperature, and
therefore excess carbon atoms would segregate from bulk to surface forming gra-
phene layers. In the relative long cooling process, carbon atoms have enough time
to segregate to metal surfaces to form multilayer graphene. In contrast, under fast
cooling, the balance between dissolution and segregation of carbon is destroyed,
and only a small amount of carbon atoms can be segregated from bulk to surface,
leading to the formation of mainly monolayer graphene.
In addition to the Rh substrate, I also studied the growth of graphene on Pt and
Pd substrates. The carbon solubility of Pt and Pd are higher than that of Rh. But the
two metal substrates show absolutely different characteristics on graphene growth.
Pt has a carbon solubility of 0.7 at.% at 1000 °C, slightly lower than Ni. However,
monolayer graphene is obtained on Pt foils in the APCVD system, whether the
rapid cooling or slow cooling process is used. I speculate that the reason may be
due to the slow velocity or the high energy required for carbon atoms migrating in
the Pt bulk, which prevents the excess carbon atoms from segregating to the Pt
surface. Pd has an ultrahigh carbon solubility of 3.5 at.% at 1000 °C. We can only
obtain multilayer graphene or graphite using LPCVD system on Pt foils, whether
2.3 The Mechanism of Graphene Growth on Rh Substrates 33

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

Based on the growth characteristics of graphene on Rh substrate, I proposed the

segregation growth mechanism of graphene. By comparison with the graphene
growth on other noble metals of Pt and Pd, Rh substrate is the most suitable
substrate for the segregation growth of bilayer graphene.

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.

3.1 Introduction to STM Study of Graphene

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.

3.1.1 The Working Principle of STM

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

(2) 3D piezoelectric scanner

The piezoelectric ceramics change dimensions in response to an applied voltage
and conversely, then develop an electrical potential in response to mechanical
pressure. Piezoelectric scanners are designed to move in x, y, and z by expanding in
some directions and contracting in others. Electrodes are attached to the outside of
the tube, segmenting it electrically into vertical quarters, for +x, +y, −x, and −y
travel. Another electrode is attached to the center of the tube to provide motion in
the z direction.
(3) Coarse sample-to-tip control and computer
A voltage bias is applied and the tip is brought close to the sample by coarse
sample-to-tip control, which is turned off when the tip and sample are sufficiently
close. At close range, fine control of the tip in all three dimensions is typically
piezoelectric, maintaining tip-sample separation (0.4–0.7 nm) range. In this situa-
tion, the voltage bias will cause electrons to tunnel between the tip and sample,
creating a current that can be measured. Once tunneling is established, the tip’s bias
and position with respect to the sample can be varied and data are obtained from the
changes in current.
(4) Vibration isolation system
Due to the extreme sensitivity of tunnel current to height, proper vibration
insulation or an extremely rigid STM body is imperative for obtaining usable results.
In the first STM by Binnig and Rohrer, magnetic levitation was used to keep the
STM free from vibrations. Now, mechanical spring or gas spring systems are often
used.
There are two typical working modes of STM: constant-height mode and
constant-current mode, as shown in Fig. 3.2. In constant-current mode, feedback
electronics adjust the height by a voltage to the piezoelectric height controller. This
leads to a height variation and thus the image comes from the tip topography across
the sample and gives a constant charge density surface. In constant-height mode,
the voltage and height are both held constant while the current changes to keep the
voltage unchanged. It leads to an image made of current changes over the surface,
which can be related to charge density. The benefit of constant-height mode is that

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

it is faster than the constant-current mode, because the piezoelectric movements

require more time to register the height change in constant-current mode.
As a new type of microscopy, the advantage of STM over other measurements
are listed as follows:
(1) The ultrahigh resolution of STM is considered to be 0.1 nm lateral resolution
and 0.01 nm (10 pm) depth resolution.
(2) STM is versatile can be used in ultrahigh vacuum, air, water and other liquids
and gasses. It can also be operated in temperatures as low as zero Kelvin up to a
few hundred Kelvin.
(3) STM can give researchers a 3D profile of a surface, which allows researchers to
examine a multitude of characteristics, including roughness, surface defects and
determining things about the molecules such as size and conformation.
(4) Atoms and molecules can be manipulated by tip. It is attractive for several
reasons. Firstly the STM has an atomically precise positioning system which
allows very accurate atomic scale manipulation. Furthermore, after the surface
is modified by the tip, it is a simple which can be imaged with the same tip,
without changing the instrument.

3.1.2 The Working Principle of STS

Scanning tunnelling spectroscopy (STS) is an extension of STM, which is used to

provide the local density of states (LDOS) of samples as a function of energy.
The STS curve is obtained by placing a STM tip above a particular place on the
sample. With the height of the tip fixed, the electron tunneling current is then
measured as a function of electron energy by varying the voltage between the tip
and the sample, which is often referred to as an I–V curve. Using the modified
Bardeen transfer Hamiltonian method which treats tunneling as a perturbation, the
tunneling current (I) is found to be:
ZeV
I/ qs ðEF  eV þ EÞ qt ðEF þ EÞ dE ð3:2Þ
0

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

By using modulation techniques, I–V and dI/dV–V curves can be acquired

simultaneously. A small, high-frequency modulation voltage is superimposed on
the D.C. tip-sample bias. The A.C. component of the tunneling current is recorded
using a lock-in amplifier, and the component in-phase with the tip-sample bias
modulation gives dI/dV directly. Equation (3.3) implies that:

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.

3.1.3 Applications of STM/STS Technique in Graphene

Research

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

(2) The superlattice of graphene

There are two main mechanisms for the formation graphene superlattice, the
twist between graphene layers and the lattice mismatch between graphene and the
underneath substrate.
Andrei et al. [6] reported that moiré patterns formed by twisted bilayer graphene
using STM. The period of graphene moiré pattern (D) and the twisted angle (h) is in
accordance with the following equation:

d
D¼  ð3:5Þ
2 sin h2

where d is the lattice constant of graphene (0.246 nm).

In 2012, LeRoy et al. [7] reported the study of moiré patterns formed by gra-
phene on h-BN substrates using STM. Since the lattice mismatch (d) of *1.7%
between graphene and h-BN, the moiré period (k) is found to be:

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.

3.2 STM Study of Bilayer 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.

3.2.1 The Stacking Order of Graphene Synthesized on Rh

Substrates

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

Fig. 3.4 Gradually magnified STM images of graphene synthesized on Rh substrates

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.

3.2.2 The Linking Type of Bilayer Graphene on Rh

Substrates

The graphene synthesized on the Rh substrate follows the mechanism of segrega-

tion growth, thus it is difficult to obtain large-scale singlecrystalline graphene
domains. In addition to the above-mentioned uniform bilayer graphene moiré
structures, there are a variety of linking types of bilayer graphene on Rh foils.
In the process of graphene growth on the Rh substrate, the domain size of the
bilayer graphene can be controlled by modulating the CH4/H2 ratio and the growth
time. Therefore, there are a large number of coexisting regions of monolayer and
bilayer graphene on Rh substrates before the bilayer graphene reaches 100%.
Figure 3.6 lists two typical linking boundaries. Figure 3.6a–c is the step-by-step
magnified STM images of a monolayer-bilayer-coexistence region. The interface
between the monolayer and the bilayer is connected by the highlighted grain
boundary, and both sides near the boundary show the electronic interference frin-
ges. It is indicated that the first-layer graphene is made up of two domains with
different orientations by the high-LDOS boundary. Figure 3.6d, e show another
type of linking interface between monolayer and bilayer graphene, as determined by
the same way with the former. Differently, a mild and straight linking boundary is
observed on this region, instead of a high-LDOS boundary. It is indicated that the
first layer is a seamless graphene layer with singlecrystalline feature. The bilayer
region is formed by the first-layer graphene and the underneath submonolayer
graphene flake. The electron scattering and interference of the underlying graphene
edge cannot be observed due to the presence of the upper layer of graphene.
The linking boundaries are also observed in twisted bilayer graphene domains.
Figure 3.7 lists two bilayer graphene linking interfaces with different kind of
linking type. If the first graphene layer is consist of two domains with different
lattice orientation, then the two domains will form different moiré patterns with the
underneath graphene layer. For example, the upper left corner of Fig. 3.7a shows a
moiré pattern with large period (D > 20 nm), while the lower right corner shows a
moiré pattern with period of *5 nm. The two regions are linked by a linking
boundary with width of *5 nm. Differently, in Fig. 3.7b, the periods of the gra-
phene moiré patterns in the upper and lower regions have little difference, which
46 3 STM Study of Twisted Bilayer Graphene

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.

3.3 Angle-Dependent VHSs in Twisted Bilayer Graphene

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].

3.3.1 VHSs in Twisted Bilayer Graphene

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

3.3.2 Angle-Dependent VHSs

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)

Additionally, the abrupt reduction of △EVHS at h * 1.3° observed in our

experiment is very interesting. The observed △EVHS * 0.11 eV is much less than
the expected value ħvF△K * 0.28 eV. There are two possible origins for this
observation. The first one is that some unknown effects enhance the interlayer
coupling strength The second one is that there are flat bands in twisted the graphene
bilayer with a twisted angle around 1.3° and the reduction of Fermi velocity around
1.3° leads to the small △EVHS. It requires further theoretical analysis and experi-
ments to confirm the deductions.

3.4 STM/STS Study of Graphene Wrinkles

The growth temperature of graphene on the Rh substrate in APCVD system is

1000 °C. During the cooling process from 1000 °C to room temperature after
growth, the thermal expansion coefficient of the graphene changed from
−1.25  10−6/K to 1.25  10−6/K, while that of Rh substrate changed from
7.86  10−6/K to 13.65  10−6/K in the same temperature range. This thermal
expansion mismatch leads to the formation of graphene winkle to release the
interface strain.

3.4.1 Graphene Wrinkles on Rh Substrates

The thickness of graphene synthesized on Rh substrates can be tuned from

monolayer to multilayer by controlling the cooling rate after the growth process.
3.4 STM/STS Study of Graphene Wrinkles 51

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

3.4.2 Strain-Induced LLs on Graphene Wirnkle

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

structures. The pseudo-magnetic field of the strained wrinkle estimated according to

the spectrum is about 100 T. The wrinkle is one of the simplest models to study the
strain-induced pseudo-magnetic flux of a corrugated graphene sheet.

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 heterostructure of graphene and h-BN is predicted to show many excellent

physical properties, such as, bandgap opening, ultra-high carrier mobility, antifer-
romagnetic and half-semimetallic characteristics. In the first section of this chapter,
I will give a brief review of the novel properties and the reported synthesis methods
of h-BN-G heterostructures. The process of preparation of h-BN-G in-plane
heterostructures is maturing, but some important basic scientific problems are still
not solved. For example, the atomic structures and electronic properties on the
interface between graphene and h-BN. The second section of this chapter introduces
the UHV two-step growth method and the weak influence substrate Ir(111) single
crystal with little electron doping effect on graphene.
In the third and fourth section, the method of UHV-CVD is used to synthesize
the graphene and h-BN on Ir(111) surface, respectively. The growth behavior and
superlattice structure of the two analogues on Ir(111) are systematically studied by
STM. In the fifth section, the preparation of h-BN @ G and G @ h-BN
heterostructures are realized on Ir(111), and the growth behavior, domain size
control and other issues of heterostructures are discussed.

4.1 Introduction of in-Plane h-BN-G Heterostructures

H-BN, a structural analogue of graphene, is composed of alternating B and N atoms

in a honeycomb lattice, possessing a lattice constant of *0.25 nm and a band gap
of 5.9 eV. Researches on in-plane and vertically-stacked heterostructures of gra-
phene and hexagonal boron nitride (h-BN) have attracted intense attentions for
energy band engineering and device performance optimization of graphene
(Table 4.1).

© Springer Nature Singapore Pte Ltd. 2018 55

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_4
56 4 Controlled Synthesis of in-Plane h-BN-G Heterostructures

Table 4.1 Comparison of Graphene h-BN

graphene and h-BN
Lattice constant 0.246 nm 0.250 nm
Interlayer distance 0.335 nm 0.330 nm
Conductivity Semimetal Insulator
Bandgap 0 5.9 eV

The potential for the formation of in-plane heterostructures between graphene

and h-BN is that two analogues show similar lattice structure: honeycomb lattice
with similar lattice constant; layered structure with similar interlayer distance.

4.1.1 Intriguing Properties of 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

(4) Half-semimetallicity at the interface of C-BN heterostructures

Half-semimetallicity has been predicted to occur in zigzag-shaped GNRs
embedded by h-BN. It is reported that the magnetism of the edge states in GNRs
and polarity effects in h-BNNRs team up to give a spin asymmetric screening that
induces an electronic reconstruction and half-semimetallicity at the interface. The
half-semimetallicity is also related to the width of GNRs. These results prove that
the unconventional physical effects which are similar to those observed at insulating
oxide interfaces can also exist in lower dimensions, which opens alternative routes
for tuning electronic properties at interfaces [4].
Consequently, the in-plane h-BN-G heterostructures can not only open the band
gap of graphene, but also exhibit many excellent physical properties such as
ultra-high carrier mobility, antiferromagnetic and half-semimetallic properties.

4.1.2 Synthesis of in-Plane h-BN-G Heterostructures

The CVD synthesis of graphitic boron-carbon-nitrogen materials was pioneered by

Kaner et al. in 1987 [5], using boron trichloride, acetylene, and ammonia precursors
at reaction temperature of 400–700 °C. Recently, the rise of graphene and other
two-dimensional materials has rekindled the research interest in the synthesis of
atomically thin hybrids or heterostructures of graphene and h-BN. Various
approaches have been developed for the synthesis of h-BN-G heterostructures.
(1) One-batch CVD growth of h-BN-G films
In 2010, Ajayan et al. [6] firstly reported the one-batch growth of h-BNC hybrids
on Cu foils, wherein methane and ammonia borane (H3NBH3) were simultaneously
introduced to serve as precursors for graphene and h-BN, respectively. As shown in
Fig. 4.1, a hybrid film containing mixed graphene and h-BN domains with rela-
tively small domain sizes was synthesized, which shows high quality and high
crystallization as convinced by OM, AFM and TEM. The hybrid composition was
tunable by controlling the precursor ratio. The hybrid films can be transferred onto
SiO2/Si for further characterization. It is noted that the composition of the h-BNC
film had a significant influence on its electronic property, such as the electrical
conductivity and bandgap. A bandgap opening of *18 meV was detected in an
h-BNC film (56 at.% C), which is considered to be attributed to the quantum
confinement effect or spin polarization at specific C-BN boundaries. However, for
such highly mixed h-BN and graphene domains, it is difficult to distinguish the two
analogues. In this regard, it is imperative to achieve a strict control of the hybrid
film thickness, and the location of the two analogues, as well as to recognize the
composite domains and their interfaces directly.
58 4 Controlled Synthesis of in-Plane h-BN-G Heterostructures

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

(2) Patterned regrowth of in-plane h-BN-G heterostructures

Through delicate control of the synthesis protocols, h-BN-G heterostructures
with defined h-BN and graphene domains were successively synthesized by pat-
terned regrowth process, which was sequentially reported by Park et al. [7] and
Ajayan et al. [8].
This approach was implemented by synthesizing monolayer graphene on Cu
foils, patterning with photolithography (PL) or reactive ion etching (RIE), and then
growing h-BN on the bare substrate regions to achieve a monolayer in-plane
heterostructure, as illustrated in Fig. 4.2. The as-grown h-BN-G film could be
transferred onto various substrates for further characterizations and device appli-
cations. This synthetic approach allows variable domain shapes and domain sizes
for the two analogues in the h-BN-G heterostructures. Such versatile patterning
techniques promote the integration of two-dimensional heterostructures into useful,
engineered device architectures, such as a split closed-loop resonator. As for the
heterostructure synthesis of patterned regrowth approach, the photolithography step
usually induced unwanted structural and chemical disorders on the edge of gra-
phene, thus leading to disordered interfaces between graphene and h-BN.
Accordingly, the intrinsic interface properties would be blurred by the topological
defects and chemical impurities. In this regard, continuous growth of h-BN and
graphene without the involvement of any photolithography process is desired for
generating a clean heterostructure interface.
4.1 Introduction of in-Plane h-BN-G Heterostructures 59

Fig. 4.2 a Schematic diagram of a patterned regrowth synthesis approach of h-BN-G

heterostructures. b OM image of h-BN-G heterostructures after transferred onto SiO2/Si substrates.
Reproduced with permission of [7] Copyright 2012, Rights Managed by Nature Publishing
Group. c OM image of h-BN-G heterostructures on Cu foils with various patterns. d Patterned
arrays of h-BN-G heterostructures after transferred onto SiO2/Si substrates. e Radio frequency
(RF) measurements of an h-BN-G-based split closed-loop resonator (Inset: optical image of the
device). Reproduced with permission of [8] Copyright © 2013, Rights Managed by Nature
Publishing Group

(3) Lateral heteroepitaxial growth of h-BN-G

In 2014, Gu et al. [9] introduced a hydrogen etching step and a designed lateral
epitaxy growth pathway to synthesize in-plane h-BN-G heterostructures with
ordered linking boundaries. Monolayer graphene was first grown on Cu foils, and
then etched by H2 to form hexagonal holes down to the Cu foil substrate. Then,
h-BN was patched onto the bare Cu regions, leading to lateral h-BN-G
heterostructures with sharp boundaries. By using this lateral epitaxy method, the
edge types of the heterojunctions were controlled precisely, considering that the
etched holes of graphene usually evolve into hexagonal shapes with zigzag edges.
Furthermore, low-energy electron microscopy (LEEM) observations demonstrate
that the lattice coherence of graphene and epitaxial h-BN could be perfectly
maintained, forgoing the diverse lattice orientations of the supporting substrates.
Consequently, both top-down and bottom-up approaches have been well
developed for the CVD syntheses of h-BN-G heterostructures. All these synthetic
efforts not only open up a rich field of further studies on the interfacial properties of
in-plane h-BN-G heterostructures, but also provide great opportunities for exploring
their applications in graphene-based electronics.
The successful CVD synthesis of h-BN-G heterostructures has inspired blos-
soming research interests in the 1D heterojunctions, due to their abundant fasci-
nating properties as predicted theoretically. As mentioned above, Cu foils were
commonly used as substrates for the CVD growth of lateral h-BN-G films, because
of their relatively low cost, good availability and compatibility with batch pro-
duction. However, it is difficult to investigate the structural details of the lateral
60 4 Controlled Synthesis of in-Plane h-BN-G Heterostructures

h-BN-G heterostructures, mainly because of the highly corrugated and multi-facet

textured features of metal foils. Consequently, ultra-flat single crystal substrates are
highly desired for constructing ideal in-plane h-BN-G, as well as for investigating
their atomic-scale features.

4.2 Experiment

4.2.1 UHV-STM System

As mentioned in Chap. 3, UHV-STM is a powerful instrument for characterizing

the apparent morphology of nanomaterials, and the resolution in x and y directions
is 0.1 nm, while in z direction is 0.01 nm. In addition, the LT-STS can characterize
the electronic state of the nanomaterials.
Compared with the traditional CVD-growth system, the UHV in situ CVD
growth has three advantages: firstly, the metal substrate in the UHV system does
not face the problem of being oxidized or contaminated, which ensures the
cleanliness of the substrate; secondly, only carbon source and h-BN source need to
be led into the UHV system and carrier gas (N2) and reducing gas (H2) are
undesired, which avoid the etching of graphene or h-BN during the growth process;
thirdly, the gases being led into the UHV chamber are much less than that in the
traditional CVD system, thus the growth behavior of the nanomaterials at the
nanometer scale would be more controllable. Therefore, the in situ growth of
two-dimensional materials in UHV system combined with STM/STS characteri-
zation is a very favorable model system for studying the growth characteristics,
surface atomic structure and physical and chemical properties of nano materials.
A schematic and picture of a well-designed UHV-CVD growth system are
shown in Fig. 4.3 which can be used to grow graphene and h-BN, as well as the
h-BN-G heterostructures. Before the sample preparation, the tubes connected with
UHV chamber need be vacuumized and baked for more than 2 h. The metal

Fig. 4.3 UHV-CVD growth system: sketch map and picture

4.2 Experiment 61

substrates need be cleaned by repetitive argon ion sputtering at room-temperature [p

(Ar) = 5
10−6 mbar, I (ion current) = 14.0 lA] following with annealing at 800–
1000 K. The growth of graphene or h-BN follows similar procedures: exposing the
substrate to ethylene (C2H4) or ammonia borane (BH3NH3) at 800–1000 K for
several minutes; cutting the precursors and holding the substrates at 800–1000 K
for more than half hour. Ethylene and ammonia boron are led into the UHV
chamber by different leak valves to prevent cross-contamination of the two
precursors.

4.2.2 The Choice of Substrate

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 "

Intriguingly, at a low borazine dosage, a brick-and–mortar pattern was acquired

after the substitution reaction. Herein, the bright ribbons are regarded as segregated
GNRs and the enclosed areas are h-BCxN alloys. This unique structure is possibly
mediated by the periodic buckling in the graphene moiré superstructures. Carbon
atoms on the moiré valley are partially bonded to the substrate with strong p–d
hybridization at the interface, potentially resulting in significant electron localiza-
tion and high reactivity towards hetero-atomic exchange, followed with the evo-
lution of h-BCxN domains. And the residual carbon atoms are preferentially
segregated along the moiré humps to evolve GNRs. Moreover, by further increasing
the borazine dosage at 900 K, complete phase segregated h-BN-G heterostructures
62 4 Controlled Synthesis of in-Plane h-BN-G Heterostructures

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.

4.3 Controlled Growth of Graphene on Ir(111)

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

4.3.1 Growth Process of Graphene on Ir(111)

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.

4.3.2 The Controllable Growth of Graphene on Ir(111)

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

Fig. 4.7 Flow chart of graphene growth on Ir(111)

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.

4.3.3 Moiré Patterns of Graphene on Ir(111)

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

4.4 Controlled Growth of h-BN on Ir(111)

4.4.1 Growth Process of h-BN on Ir(111)

It is very important to choose precursors for h-BN preparation in UHV system.

Most molecules containing B and N elements are toxic and dangerous. Previous
works on h-BN preparation usually use borazine as precursor, which is a colorless,
but explosive liquid at room temperature. Borazine is easily decomposed to produce
flammable borane and toxic nitrogen oxides when exposed in air. In this work, we
use solid ammonia boron (NH3BH3) as precursors to prepare h-BN (Fig. 4.12).

Fig. 4.12 The thermal

decomposition product of
ammonia boron molecules
70 4 Controlled Synthesis of in-Plane h-BN-G Heterostructures

Fig. 4.13 Flow chart of

h-BN growth on Ir(111)

The ammonia boron molecules decompose at 400 K to produce hydrogen,

borazine and boron-nitride polymers. The decomposition gas products can be
introduced into the UHV chamber for preparing h-BN. The growth of h-BN also
follows the UHV-CVD process: first, the substrate temperature is raised to 900 K,
meanwhile the decomposition products of ammonia boron are introduced into the
UHV chamber for the growth of h-BN on Ir(111); secondly, the precursor mole-
cules are cleaved on the metal surface and formed h-BN; after the growth process,
the sample is held at growth temperature for half an hour. The key factors con-
trolling the growth of h-BN in UHV system are similar to that of graphene and are
not described here (Fig. 4.13).

4.4.2 Controllable Growth of h-BN on Ir(111)

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

Fig. 4.14 H-BN on Ir(111)

substrate a, b Large-scale
STM image of submonolayer
h-BN on Ir(111) and the
corresponding magnified
STM image. c STM image of
monolayer h-BN on Ir(111)
and the corresponding LEED
measurement

4.4.3 Moiré Patterns of h-BN on Ir(111)

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

4.5 Controlled Growth of h-BN-G Heterostructures on Ir

(111)

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

heterostructure (Fig. 4.19c, d) is obtained after second patching growth of gra-

phene, and the triangular domain shape of h-BN remains unchanged.
Maintaining the same precursor supply and growth time, the domain size of the
submonolayer graphene can be effectively controlled by growth temperature. The
graphene domain size can increase from 25 to 500 nm following the growth tem-
perature increasing from 900 to 1100 K, meanwhile, the domains still maintains the
hexagonal shapes. Similarly, in the growth processes of h-BN @ G heterostruc-
tures, the domain size of h-BN increases from 30 to 500 nm following the
increasing of growth temperature from 900 to 1100 K. The h-BN domains maintain
triangular shapes.
The growth temperature supplies two factors for controlling the domain size of
graphene or h-BN: the one is that the elevated temperature increases the migration
rate of cracked precursors, simultaneously accelerates the growth rate of graphene
or h-BN, and increases the temperature by reducing the graphite; the other is that
the elevated temperature reduces the nucleation density of the graphene or h-BN on
the metal surface, providing a sufficient growth environment for the expansion of
individual domains. If the growth temperature and the precursor supply are both
increased simultaneously, graphene will grow rapidly at high-density nucleation
sites and will cover the whole surface of metal substrates, thus the submonolayer
graphene cannot be obtained.
The ratio of in-plane corrugation for graphene and h-BN on Ir(111) is 1:7, and
the height of graphene is 0.8 Å higher than that of h-BN. However, they can
overcome the difference of corrugation and height to form in-plane heterostructures.
In h-BN-G heterostructures, graphene and h-BN maintain their apparent height and
in-plane corrugations in their own domains. At the linking boundaries, graphene is
pulled down and h-BN is pulled up to form monolayer seamless heterostructures.

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

In order to investigate the atomic and electronic structure on the interface of

intrinsic h-BN-G heterostructures, we choose Ir(111), which has a weak interaction
with graphene, as the substrate for sample synthesis. Monolayer graphene and h-BN
can be prepared on the Ir(111) separately by UHV-CVD process. STM character-
ization and DFT calculations reveal that graphene and h-BN show different moiré
periods, physical height and in-plane corrugations. By the modified UHV-CVD
two-step growth method, h-BN @ G heterostructures and G-h-BN heterostructures
can be synthesized by on Ir(111).

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

Notably, the physical characteristics of h-BN-G heterostructures are related to the

linking type in the interface. Thus, it becomes an important subject that how to form
particular-type boundaries during growth process and how to identify the atomic
linking type. This chapter mainly introduces the study of interface type and inter-
facial electron state of h-BN-G heterostructures.
The first section mainly introduces the study of the different interface types of
h-BN-G in-plane heterostructures by using high-resolution STM. The experimental
results show that the proportion of zigzag-type boundaries is above 99%. In the
second section, I analyzed the reason for the formation of almost 100% zigzag-type
boundary in h-BN-G heterostructures based on both of experimental and theoretical
results. In the third section, I first briefly introduced the previous research of the
electronic structure of h-BN-G interface. Then I studied the electronic states of
monolayer graphene, h-BN and h-BN-G heterostructures on Ir(111) substrate
separately by means of STS. Finally, I discussed the mechanism for h-BN-G
heterostructure having tunable bandgap and high carrier mobility.

5.1 Atomic Structure of h-BN-G Interface

Controlling and identifying the atomic type of linking boundary is critical to

h-BN-G heterostructures. I have investigated 100 regions (100  100 nm) for the
atomic types of the h-BN-G heterostructures on Ir(111) by high resolution STM.
The tested regions are selected from both of G@h-BN and h-BN @ G
heterostructures grown under different temperature. The graphene and h-BN
domains in the 100 scanning regions show different shapes. The following will
describe the interface type in detail.
Figure 5.1a, c are the typical G@h-BN and h-BN@G heterostructures, respec-
tively. Graphene and h-BN domains display different moiré patterns. Magnified STM
images of the area within the blue boxes show the atomic structure of linking
© Springer Nature Singapore Pte Ltd. 2018 77
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_5
78 5 Atomic and Electronic Structures of h-BN-G Interfaces

Fig. 5.1 Atomic structure of h-BN-G interfaces on Ir(111) a G@h-BN heterostructures.

b h-BN@G heterostructures. c, d Corresponding magnified STM images of (a) and (b),
respectively

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

Fig. 5.5 Armchair-type boundary of h-BN-G heterostructure a Large-scale STM image of a

linking boundary between graphene and h-BN. b Atomic-resolved STM image of the area marked
by blue dotted box showing armchair-type boundary with overlaid atomic model

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%)

5.2 The Formation Mechanism of the Zigzag-Type

Boundaries

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.

5.2.1 Edge Types of Submonolayer Graphene Grown

on Metal Substrates

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:

RS19 [ RSL [ RAC [ RZZ

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

graphene domain. I identified above 50 graphene domains and 50 h-BN domains on

the Ir(111) substrate, which are all terminated by zigzag-type edges.
In conclusion, the theoretical and experimental results confirm that the sub-
monolayer graphene and h-BN on Ir(111) substrate tends to form zigzag-type
84 5 Atomic and Electronic Structures of h-BN-G Interfaces

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.

5.2.2 The Formation Process of Zigzag-Type Linking

Boundary

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

5.3 Electronic Structure of h-BN-G Interface

5.3.1 Research of Electronic Properties of h-BN-G Interface

According to the growth system of h-BN-G heterostructures introduced in Chap. 4:

for the h-BNC hybrid materials grown by one-batch method on Cu foils, the
domains of graphene and h-BN cannot be recognized; for the h-BN-G
heterostructure grown by patterned regrowth method, the linking boundary is too
rough and the atomic continuity cannot be determined; for the lateral h-BN-G
heterostructure grown by heteroepitaxial method, it is difficult to identify the two
analogues using STM since the similar lattice constants of graphene and h-BN.
Li et al. [6] proposed that they can distinguish graphene and h-BN by high bias
(1 V, 200 pA) dI/dV mapping on the interface of h-BN-G heterostructures. They
did a dI/dV line mapping perpendicular to the boundary at the interface. The
spectrum of the Dirac point located at −300 mV is confirmed to be recorded on
graphene domain, while the spectrum with the metal state is considered to be
recorded on the h-BN domain. The metal state is contributed of the interfacial
interaction between monolayer h-BN and the underlayer Cu substrate. From the
above results, it can be seen that the electronic properties of the h-BN-G
heterostructure on the Cu foil is greatly affected by the substrate.
In the h-BN-G heterostructure prepared on Ir(111) by UHV-CVD method,
graphene and h-BN domains show different moiré patterns that can be identified by
STM images, and the Ir(111) substrate has a weak doping effect on the
heterostructures. Thus, the h-BN-G heterostructure on Ir(111) will be an excellent
model for studying the electronic states of h-BN-G interfaces.
86 5 Atomic and Electronic Structures of h-BN-G Interfaces

5.3.2 Electronic Structure of Monolayer Graphene

and h-BN on Ir(111)

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

electronically freestanding graphene. In contrast, on h-BN domain the dI/dV curves

at locations D, E, and F all exhibit a flat background close to zero from −0.8 to
0.8 eV, suggesting the insulating nature of h-BN. Interestingly, when the STM tip
scan approaches to the boundary from graphene side, the V-shaped dI/dV curve
becomes broadened prominently (curves B and C). This remarkable decrease of
graphene conductance would be attributed to the edge effect of graphene island [7].

5.4 Discussion and Conclusion

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

graphene nanoribbons. It is proved that the graphene in h-BN-G heterostructures

exhibits intrinsic electronic structure and edge state, which is not affected by
adjacent h-BN.
Therefore, in the h-BN-G heterostructure, h-BN is atomically bonded with
graphene and acts as an insulator: on the one hand, it breaks the symmetry of
graphene lattice, resulting in the bandgap opening of graphene; on the other hand,
the metallic boundary is formed between graphene and h-BN, which can transmit
the electrons from defect-induced intervalley scattering, greatly reducing the elec-
tron scattering at graphene domain edge and ensuring high carrier mobility.
In conclusion, in this Chapter I investigated the interface type and interfacial
electron state of h-BN-G heterostructures on Ir(111) substrates, and explore the
mechanism of interface formation and the deep meaning of bandgap opening. First,
I observed the atomic structure on h-BN-G interface on the Ir(111) by using STM.
The statistical results confirm that more than 99% of the boundary type is zigzag
boundary. The boundary type is not affected by the growth order, domain size and
domain shape. It is consistent with the results on Rh(111) substrate and has a higher
selectivity than h-BN-G heterostructures on Rh(111) substrate (78%).
Since the weak interaction between Ir(111) and the epitaxial layer, the growth of
graphene and h-BN on the substrate is only affected by the intrinsic lattice sym-
metry and formation energy of different edge, thus they all prefer to be terminated
by zigzag edge. The second step patching growth on the existing domains does not
change the domain shape. Therefore, the zigzag-type edges can be preserved and
further evolving into zigzag-type boundaries. Both the theoretical and experimental
results demonstrate that it is feasible to control the boundary type from the growth
point of view.
Finally, the electronic states of graphene, h-BN and the in-plane heterostructures
on weakly-coupled substrate Ir(111) were studied by using STS. It is found that
graphene and h-BN prefer to exhibit their intrinsic electronic structures near the
interface. The related literature also suggests that the intrinsic edge state of gra-
phene was observed at the boundary of the heterostructure, that is, the graphene in
the h-BN-G heterostructure exhibits an intrinsic electronic structure.

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

May 25, 1989 Born—Weifang, China.

2006–2010 B.S. in Applied Chemistry,
South China University of Technology, China.
2010–2015 Ph.D. in Physical Chemistry,
Peking University, China.
2015–present Assistant Professor,
National Center for Nanoscience and Technology, China.

© Springer Nature Singapore Pte Ltd. 2018 93

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