FABRICATION OF GRAPHITE AND GRAPHENE-BASED
MATERIALS FOR FIELD EMISSION APPLICATIONS
KOH TING TING ANGEL
(B. Appl. Sci. (Hons.) National University of Singapore)
A THESIS SUBMITTED
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
DEPARTMENT OF MATERIALS SCIENCE &
ENGINEERING
NATIONAL UNIVERSITY OFSINGAPORE
2012
Declaration
i
DECLARATION
I hereby declare that this thesis is my original work and it has been written by me in its
entirety. I have duly acknowledged all the sources of information which have been used
in the thesis.
This thesis has also not been submitted for any degree in any university previously.
Koh Ting Ting Angel
28th
August 2012
Acknowledgements
ii
ACKNOWLEDGEMENTS
Embarking on the PhD journey has indeed been the most challenging task I have
taken in the 26 years of my life. Through this journey, I have gotten to know a handful of
wonderful people that have made an impact in my graduate life. It has been a privilege and an
honor to work with these brilliant and helpful people and they will always remain dear to me.
First, I would like to express my deepest gratitude to my supervisor Dr. Daniel Chua
who is my first teacher of research when I first started research work five years ago. He has
not only provided me with invaluable guidance and advice necessary to see me through this
program but at the same time gave me the freedom to pursue independent work. I sincerely
thank him for his patience, understanding and consolation during the bad times when nothing
seems to go right. I am also grateful to him for the numerous overseas opportunities that he
has given me, from which I have gained many priceless experiences from. I would also like
to thank him for putting up with my complaints and for tolerating the many lame jokes that I
uncontrollably crack all the time during our serious discussions. Through him, I have learnt
many lessons about research and life.
I am also grateful to Prof. Li Yi, Prof. Ding Jun, Dr. Chiu Cheng-Hsin and Dr. Stefan
Adams from whom I have acquired fundamental knowledge that I used as a basis for my
research ideas.
I would like to express my sincerest thanks to Mr. Tan Choon Wah from Physics
workshop. I am grateful for his help and support all these years in realizing my amateurish
technical design drawings into physical objects and for the many valuable life lessons that he
has enlightened me with. Without his help to fix broken equipment, my experiments would
Acknowledgements
iii
not have been possible. I would also like to thank Mr. Chen Gin Seng from the Department
of Physics for his help and guidance. In addition, I would like to thank Mr. Wong How
Kwong from the Department of Physics for his guidance and for providing new furniture
when the old one breaks for my lab in Science faculty. Despite the tiny size of my lab, he has
not forgotten about us and welcomed me as part of the Physics family all these years.
I would also like to extend my appreciation to the lab officers of DMSE especially,
Ms. Agnes Lim, Mr. Henche Kuan, Mr. Chan Yew Weng, Ms. Serene Chooi, Ms. He Jian
and Mr. Chen Qun their help and teachings all these years. Special thanks go out to Dr.
Zhang Jixuan for her patience and helpfulness.
My graduate student life would be less fulfilling if it was without the ups and downs I
shared with the following people. I am deeply grateful to my research group senior, Yuan
Mei. I would like to thank her for providing me with exceedingly useful discussions where
we bounced ideas off each other and for sharing my joy and horror moments upon the
revelation of a new discovery or some mistakes. I would also like to thank Jovan Hsieh, my
first mentor in research who has undoubtedly played an important part in my embarkation
towards a graduate research career. She has my thanks for giving me her moral support and
for being my listening ear since day one. I will not forget her kindness when she gave her
experiment slot to me when I was pressed for time. Special thanks go to Su Ru, Wang
Hongyu, Yu Jun, Tang Zhe, Le Quang Tri and Dr. Chen Ting, who have help me in one way
or another during my arduous PhD course of study.
Last but not least, I would like to express my heartfelt gratitude to my family and
especially my mum, for their unconditioned love and support in my research endeavor.
August 2012 Angel Koh
Table of Contents
iv
Table of Contents
DECLARATION i
ACKNOWLEDGEMENTS ii
TABLE OF CONTENTS iv
SUMMARY viii
LIST OF TABLES x
LIST OF FIGURES xi
LIST OF ABBREVATIONS xv
LIST OF PUBLICATIONS xviii
CHAPTER 1 INTRODUCTION 1
1.1 Graphite and monolayer graphite 1
1.2 Properties of graphene 3
1.2.1 Electrical properties 3
1.2.2 Mechanical properties 5
1.2.3 Optical properties 6
1.2.4 Thermal properties 7
1.3 Methods of fabricating graphene 7
1.4 Field emission theory 11
1.4.1 Model and principles 11
1.4.2 Field emission of carbon materials 13
1.5 Motivation and objectives 18
References 20
Table of Contents
v
CHAPTER 2 EXPERIMENTAL TECHNIQUES 25
2.1 Thin film deposition techniques 25
2.1.1 Magnetron sputtering (MS) 25
2.1.2 Pulsed laser deposition (PLD) 27
2.2 Graphene/graphene films Fabrication Methods 30
2.2.1 Modified Hummers method 30
2.2.2 Electrophoretic deposition (EPD) 31
2.3 Materials characterization 32
2.3.1 Scanning electron microscope (SEM) 32
2.3.2 Transmission electron microscope (TEM) 33
2.3.3 Raman spectroscopy 34
2.3.4 X-ray photoelectron spectroscopy (XPS) 36
2.3.5 Atomic force microscopy (AFM) 37
2.3.6 Field emission testing (FE) 39
References 41
CHAPTER 3 FABRICATION OF NANOGRAPHITE IN AMORPHOUS CARBON
(a-C) MATRIX FIELD EMITTERS 44
3.1 Introduction 44
3.2 Experimental method 47
3.3 Microstructural characterization of nanographite clusters in a-C matrix 49
3.3.1 Effect of temperature on growth of nanoclusters 49
3.3.2 Discussion 57
3.4 Field emission study of nanographite clusters in a-C matrix 59
3.4.1 Field emission properties 59
3.4.2 Emission mechanism of nanographite clusters in a-C matrix 65
3.5 Summary 69
References 70
Table of Contents
vi
CHAPTER 4 PULSED LASER FABRICATION OF FEW-LAYER GRAPHENE
USING METAL SUBSTRATES 73
4.1 Introduction 73
4.2 Few-layer graphene from Nickel substrates 80
4.2.1 Experimental method 80
4.2.2 Cooling rate dependence of graphene growth 81
4.2.3 Laser energy effects on graphene growth 85
4.3 Few-layer graphene segregation using other metal substrates 87
4.3.1 Experimental method 87
4.3.2 Segregation of graphene on different metal substrates 87
4.3.3 Effect of cooling rate when using different metals 90
4.3.4 Discussion 94
4.4 Few-layer graphene segregation using different pulsed laser wavelengths 98
4.4.1 Experimental method 98
4.4.2 Segregation with Nd:YAG and KrF pulsed laser 98
4.4.3 Discussion 102
4.5 Fabrication of few-layer graphene field emitters using PLD 108
4.5.1 Experimental method 108
4.5.2 Microstructural characterization of cones 108
4.5.3 Field emission study of graphene coated metal cones 112
4.5.4 Discussion 113
4.6 Summary 117
References 120
CHAPTER 5 GRAPHENE-BASED FIELD EMITTERS BY
ELECTROPHORETIC DEPOSITION (EPD) 125
5.1 Introduction 125
5.2 Experimental method 129
5.3 Large area high density graphene for field emission 130
5.3.1 Microstructural characterization 130
Table of Contents
vii
5.3.2 Field emission properties 134
5.3.3 Discussion 135
5.4 Graphene/carbon nanotube hybrid materials for field emission 139
5.4.1 Microstructural characterization 139
5.4.2 Field emission properties 142
5.4.3 Discussion 143
5.5 Summary 146
References 147
CHAPTER 6 CONCLUSIONS AND FUTURE WORK 150
6.1 Conclusions 150
6.2 Future work 154
Summary
viii
Summary
Graphene has attracted a tremendous amount of research interest in the recent
decade due to its one-of-a-kind properties. With its excellent electrical, mechanical and
optical properties, graphene has the potential to outperform current available materials in
a wide range of applications. However, the fabrication and understanding of the growth
mechanism of graphene need to be established before realizing this material for practical
applications. While techniques such as chemical vapor deposition (CVD) and mechanical
exfoliation have been extensively investigated, studies on solid-state transformation of
carbon into graphene through physical vapor deposition (PVD) techniques are limited.
The aims of this dissertation were to understand the growth mechanism of
graphene segregation using pulsed laser deposition (PLD) with a solid carbon source and
to explore alternative methods to fabricate high sp2 and graphene-based materials for
field emission applications. To achieve this objective, investigations on growth and field
emission properties of nanographite in an amorphous carbon (a-C) matrix was first
conducted. Its field enhancement originated from the nanographite clusters within the
films that formed internal conducting channels in the largely insulating a-C matrix and
turn-on voltages decreased with an increase in nanographite density within the deposition
temperature range of 100–700 oC.
Subsequently, segregation of graphene was studied with PLD and the effects of
various parameters such as cooling rate, substrate, laser energy and wavelengths on the
growth of graphene were investigated. Briefly, low laser energy was favored for graphene
formation and different metals required different cooling profiles for graphene
Summary
ix
segregation. Using the knowledge acquired from these experiments, graphene-on-metal
nanocones field emitters with different densities were fabricated. It was observed that the
graphene-on-metal nanocones gave enhanced field emission properties as compared to
the bare metal nanocones due to the reduction of effective field emission tunneling
barrier, which was a result of graphene-metal charge transfer interactions. Controlling the
metal nanocones density was also an important factor, as electron screening from
neighboring cones should be minimized.
Lastly, graphene-based field emitters were fabricated with electrophoretic
deposition (EPD) and this presented a low cost, environmentally safe and highly
upscalable method to produce large area graphene field emitters. Graphene-only emitters
were deposited with different EPD durations and when the graphene density was too high,
screening effect resulted, which increased the turn-on electric field. To improve the
reliability of the graphene-only emitters, hybrid carbon nanotubes (CNT) with graphene
emitters were also fabricated. The CNT was envisioned to hold down the graphene flakes,
like a safety belt, at high voltages to prevent an early short circuit at relatively low
voltages. In contrast to pure graphene emitters, these hybrid emitters performed better
when the EPD duration was increased, as more graphene flakes were present when the
time was increased. These graphene flakes can help to improve the electrical conduction
of the material, as well as provide additional emission sites on the surface.
Hence, through the information gained from these experiments, graphene growth
segregation from solid carbon sources using PLD can be better understood and graphene
field emitter materials can be optimized during fabrication to achieve better performances.
List of Tables
x
List of Tables
Table 3.1 D and G peaks positions, amplitude, relative intensity ratio and estimated
sp2 content of the films deposited at different temperatures ......................51
Table 3.2 Comparison of field enhancement factors of various carbon-based emitters
....................................................................................................................66
Table 4.1 Carbon ion energies for step 1 and 2 during laser ablation of carbon target
for KrF and Nd:YAG systems .................................................................107
Table 4.2 Comparison of work function of free-standing graphene and clean metal
surfaces with the respective graphene on Ni or Co metal systems from ref.
60..............................................................................................................115
Table 4.3 Comparison of turn-on voltages of the graphene on metal cones samples
with the nanographite in amorphous carbon samples in Chapter 3 .........117
List of Figures
xi
List of Figures
Fig. 1.1 A graphene sheet is the building block of all other graphitic structures
ranging from fullerenes (1D), carbon nanotubes (2D) to graphite (3D) ......2
Fig. 1.2 Single layer and bi-layer graphene visible on 300 nm SiO2 under an
optical microscope but hardly visible on a 200 nm SiO2 substrate ..............3
Fig. 1.3 GO produced by Hummers method is yellow-brown in color while
reduced graphene is black in color .............................................................10
Fig. 1.4 Difference in emission barrier between field emission and thermionic
emission .....................................................................................................11
Fig. 1.5 Vertically oriented graphene flakes fabricated by spin coating onto
conductive silicon substrates in ref. 16 ......................................................16
Fig. 2.1 Schematic of RF magnetron sputtering showing the in-chamber
configurations ............................................................................................26
Fig. 2.2 Schematic of PLD used for nanographite in amorphous carbon deposition
and graphene segregation ...........................................................................28
Fig. 2.3 Plume seen during the ablation of target....................................................28
Fig. 2.4 Schematic of EPD setup with planar electrodes ........................................31
Fig. 2.5 Three possible outcomes during the interaction of photon and matter in
Raman spectroscopy ..................................................................................35
Fig. 2.6 Schematic of AFM .....................................................................................38
Fig. 2.7 Parallel plate electron emission tester setup ..............................................39
Fig. 2.8 Probe tip electron emission tester setup .....................................................40
Fig. 3.1 Raman spectrum for films deposited at room temperature, 100, 400 and
700 C ........................................................................................................50
Fig. 3.2 Raman ID/IG ratio and sp2 content changes of samples deposited at
temperature from 100-700 oC ....................................................................52
Fig. 3.3 Raman D and G peak positions of samples deposited at temperature from
100-700 oC .................................................................................................52
List of Figures
xii
Fig. 3.4 XPS spectra for films deposited at room temperature, 100, 400 and 700
C ...............................................................................................................53
Fig. 3.5 TEM micrograph for films deposited at (a) 100 C (b) 400 C (c) 700 C.
Inset shows the low magnification area view and SAED of crystalline
cluster with lattice spacing 0.33 nm corresponding to carbon ...................54
Fig. 3.6 Cluster density vs. temperature of 100, 400 and 700 oC samples ..............55
Fig. 3.7 AFM surface morphologies of (a) 100 C (b) 400 C (c) 700 C
samples........ ...............................................................................................56
Fig. 3.8 I-V characteristics of films deposited at 100, 400 and 700 oC ...................60
Fig. 3.9 Fowler-Nordheim plots at 50 µm for films deposited at 100, 400 and 700 oC................................................................................................................62
Fig. 3.10 Voltage-distance (V-d) curves for nanographite in a-C films deposited at
100, 400 and 700 oC ...................................................................................62
Fig. 3.11 Turn-on field versus anode-cathode distances of films deposited at 100,
400 and 700 oC. Inset (a) shows typical line shapes with 2 different height
approximations determined from simulations in Ref. 37. Inset (b) shows
line details at small distances .....................................................................64
Fig. 4.1 Raman spectra of samples cooled at different rates. The laser energy per
pulse was 50 mJ .........................................................................................81
Fig. 4.2 Cross-section TEM showing (a) the graphene layers above the Ni. (b)
Ni/Nickel Silicide/Si layered structure. Inset (c) is an overview optical
image showing rather uniform coverage ....................................................84
Fig. 4.3 Cross-section SEM showing delamination at Ni film-Si substrate interface
for medium and fast cooling samples ........................................................84
Fig. 4.4 Raman spectra of samples in which different pulsed laser energies were
used. Cooling rate was kept constant .........................................................86
Fig. 4.5 Raman spectra of carbon deposited on various metal substrates. Few-layer
graphene was formed only on Ni metal substrates when cooled with 1
oC/min to 550
oC followed by 20
oC/min...................................................89
Fig. 4.6 Cross-section TEM showing (a) the graphene layers above Ni (b) a more
disordered form of graphite formed when cooling conditions were too
fast............................................................................................................. .89
Fig. 4.7 Comparison of Raman spectra between Cu and Ni samples in which
different cooling rates 50 (top) and 100 oC/min (bottom) were used ........90
List of Figures
xiii
Fig. 4.8 Comparison of FLG grown on Co (top) and Ni (bottom). Full coverage
was obtained on Co substrate, while areas with graphene on Ni are
indicated by the black arrows (inset) .........................................................91
Fig. 4.9 Cross-section TEM image of graphene on top of Co showing 2-3
layers.......................................................................................................... 92
Fig. 4.10 Raman spectra of (a) as grown few-layer graphene on Co (top). (b)
Transferred few-layer graphene on Si, which has a significantly reduced D
peak indicating relatively few defects in the graphene (bottom) ...............92
Fig. 4.11 Raman spectra of areas covered and uncovered during carbon deposition
and segregation ..........................................................................................93
Fig. 4.12 Illustration of carbon atom concentration throughout the Co film as a
result of slow and fast cooling profiles. Carbon atoms are diffused deeper
into the Co film and harder to precipitate when cooling is slow. During
fast cooling, solubility is exceeded in the near-surface regions and
precipitation results ....................................................................................97
Fig. 4.13 Raman spectra of graphene on Ni segregated from carbon deposited by
Nd:YAG laser at 750 oC ..........................................................................100
Fig. 4.14 Raman spectra of graphene on Ni segregated from carbon deposited by
KrF laser at 750 oC ...................................................................................100
Fig. 4.15 Raman spectra of graphene on Ni segregated from carbon deposited by
Nd:YAG laser at 5 different spots. Optical images (right) show the spot
that was scanned ......................................................................................101
Fig. 4.16 Raman spectra of graphene on Ni segregated from carbon deposited by
Nd:YAG laser when cooling rate was adjusted to a faster rate of 20 oC/min
..................................................................................................................102
Fig. 4.17 SEM images of original as-fabricated Ni cones (a) Type A, (b) Type B and
Co cones (c) Type A, (d) Type B .............................................................110
Fig. 4.18 SEM images of heat-only (a) Ni (b) Co metal nanocones .......................110
Fig. 4.19 SEM images of (a) Ni (b) Co metal nanocones after graphene
segregation............................................................................................... 111
Fig. 4.20 Raman spectra for Ni and Co metal cones ...............................................111
Fig. 4.21 Field emission I-V plots for graphene on cone samples ..........................112
Fig. 4.22 Field emission F-N plots for graphene on cone samples .........................113
List of Figures
xiv
Fig. 4.23 Schematic of work function shift when graphene is chemisorbed onto
metal ( G-on-M) and the position of EF in intrinsic versus doped
graphene........................................................................................... ........115
Fig. 5.1 SEM images of EPD deposited graphene field emitters with deposition
time of (a, b) 5, (c, d) 10 and (e, f) 15 min. Images on the left column were
taken at 1000X while those on the right were at 20,000X
magnification...........................................................................................132
Fig. 5.2 Cross-section SEM image of EPD deposited graphene showing near
vertical orientation of microflakes ...........................................................133
Fig. 5.3 (a) TEM images showing two as-prepared graphene flakes, one folded and
one flat. SAED (inset) confirmed that the graphene flakes were of high
crystalline quality. (b) Lattice fringes of scrolled or folded areas at high
magnification ...........................................................................................133
Fig. 5.4 Field emission J-E curves of the 5, 10, 15 min samples ..........................134
Fig. 5.5 F-N plots for 5 and 10 min samples. β can be calculated from the slope of
the straight line fit ....................................................................................135
Fig. 5.6 Schematic diagram showing the possible scenarios when deposition time
is (a) short or (b) extended. When flake density is high, a flatter overall
surface morphology gave poorer field enhancing properties due to
screening effect. .......................................................................................138
Fig. 5.7 SEM images of 2.5, 5 and 10 min graphene/CNT hybrid samples by EPD.
Left column shows magnification at 1000X while right column shows
20,000X ....................................................................................................140
Fig. 5.8 (a) Cross-section SEM images showing morphology with multiple
protrusions and a vertically aligned CNT-strapped graphene sheet (white
arrow) (b) Hybrid graphene/CNT film by filtration at 50% graphene in ref.
20 shows a flatter overall surface morphology ........................................141
Fig. 5.9 TEM images showing (a) high crystalline quality graphene (b) CNT used
in the experiments ....................................................................................141
Fig. 5.10 I-V plots of graphene/CNT hybrid emitters .............................................142
Fig. 5.11 F-N plots of graphene/CNT hybrid emitters ............................................143
List of Abbreviations
xv
List of Abbreviations
0D Zero dimensional
1D One dimensional
2D Two dimensional
3D Three dimensional
a-C Amorphous carbon
AFM Atomic force microscopy
CCD Charge-couple device
CNT Carbon nanotubes
CVD Chemical vapor deposition
DC Direct current
dia. Diameter
DLC Diamond-like carbon
EPD Electrophoretic deposition
FE Field emission
FEG Field emission gun
List of Abbreviations
xvi
FET Field effect transistors
FLG Few-layer graphene
F-N Fowler-Nordheim
FWHM Full width half maximum
GO Graphene oxide
HR High resolution
ITO Indium-tin oxide
KE Kinetic energy
LCD Liquid crystal displays
min Minutes
NEA Negative electron affinity
OLED Organic light emitting devices
PLD Pulsed laser deposition
PVD Physical vapor deposition
RF Radio frequency
RMS Root mean square
rpm rounds per minute
List of Abbreviations
xvii
sec Seconds
SEM Scanning electron microscope
SLG Single layer graphene
TEM Transmission electron microscopy
WKB Wentzel–Kramers–Brillouin
XPS X-ray photoelectron spectroscopy
List of Publications
xviii
LIST OF PUBLICATIONS
1. A. T. T. Koh, Y. M. Foong, Likun Pan, Zhuo Sun and D. H. C. Chua, “Effective
large-area free-standing graphene field emitters by electrophoretic deposition”
Applied Physics Letters 101, 183107 (2012).
2. A. T. T. Koh, Y. M. Foong, and D. H. C. Chua, “Comparison of the mechanism
of low defect few- layer graphene fabricated on different metals by pulsed laser
deposition” Diamond and Related Materials 25, 98-102 (2012).
3. A. T. T. Koh, Y. M. Foong, B. Z. Phang, Daniel H. C. Chua, and M. Tanemura.
“Structural Studies of Aluminium Nitride Embedded in Amorphous Carbon”
Journal of Nanoscience and Nanotechnology 12, 6526-6530 (2012).
4. A. T. T. Koh, Y. M. Foong, J. Yu, D. H. C. Chua, A. T. S. Wee, Y. Kudo, K.
Okano, “Understanding Tube-like Electron Emission from Nanographite
Clustered Films” Journal of Applied Physics 110, 034903 (2011).
5. A. T. T. Koh, Y. M. Foong, and D. H. C. Chua, “Cooling rate and energy
dependence of pulsed laser fabricated graphene on nickel at low temperature”
Applied Physics Letters 97, 114102 (2010).
6. A. T. T. Koh, J. Hsieh and D.H.C. Chua, “Structural characterization of Dual
Metal containing Diamond-like carbon Nanocomposite films by pulsed laser
deposition” Diamond and related Materials 19, 637 – 642 (2010).
7. A. T. T. Koh, J. Hsieh, Daniel H.C. Chua “Electron emission studies of CNTs
grown on Ti and Ni containing amorphous carbon nanocomposite films” Applied
Surface Science 256,178–182 (2009).
List of Publications
xix
8. A. T. T. Koh, Y. M. Foong, K. Mastubara, Z. Yusop, M. Tanemura and D. H. C.
Chua, “Graphene-coated metal nanocones for field emission”
9. A. T. T. Koh, Likun Pan, Zhuo Sun and D. H. C. Chua, “Hybrid
graphene/carbon nanotubes field emitters by electrophoretic deposition”
10. A. T. T. Koh, Y. M. Foong, K. Mastubara, Z. Yusop, M. Tanemura and D. H. C.
Chua, “A study on Nd:YAG and KrF pulsed laser effects on solid state
segregation of few-layer graphene”
11. Y. M. Foong, A. T. T. Koh, J. Hsieh, S. R. Lim, and D. H. C. Chua. “Materials
properties of ZnO/Diamond-like carbon (DLC) nanocomposite fabricated with
different source of targets” Diamond and Related Materials 25, 103-110 (2012).
12. J. Yu, P. Anetab, Angel T. T. Koh, Daniel H.C. Chua, J. Wei. “Enhanced
electron emission from tetrahedral amorphous carbon capped carbon nanotube
core-shelled structure” Diamond and Related Materials. 21, 37-41 (2012).
13. Jie Sun, Niclas Lindvall, Matthew T. Cole, Koh T. T. Angel, Teng Wang, Ken B.
K. Teo, Daniel H. C. Chua, Johan Liu and August Yurgens. “Low Partial
Pressure Chemical Vapor Deposition of Graphene on Copper”. IEEE
Transactions on Nanotechnology 11, 255-260 (2012).
14. Y. M. Foong, A. T. T. Koh, H. Y. Ng and D. H. C. Chua. “Mechanism behind the
surface evolution and microstructure changes of laser fabricated nanostructured
carbon composite” Journal of Applied Physics 110, 054904 (2011).
List of Publications
xx
15. Y. M. Foong, A. T. T. Koh, S. R. Lim, H. Y. Ng and D. H. C. Chua. “Properties
of laser fabricated nanostructured Cu/diamond-like carbon composite”. Journal
of Material Research 26, 2761-2771 (2011).
16. Y. M. Foong, A. T. T. Koh, and D. H. C. Chua. “Experimental and theoretical
study on the energy-dependent surface evolution and microstructure changes in
copper nanostructured composites” Journal of Physics D: Applied Physics 44,
385401 (2011).
17. Y. M. Foong, A. T. T. Koh, L. Niu, and D. H. C. Chua. “Mechanism behind the
formation of self-assembled nano-sized clusters in Diamond-like carbon
nanocomposite”. Journal of Nanoscience and Nanotechnology 11, 10511 (2011)
18. J. Yu, Y. M. Foong, A. T. T. Koh, and D. H. C. Chua. “Field emission
characteristics of tetrahedral amorphous carbon coated carbon nanotubes with
and without hydrogenation treatment”. Journal of Physical Chemistry C 115,
11336 (2011).
19. I. Saito, W. Miyazaki, M. Onishi, Y. Kudo, T. Masuzawa, T. Yamada, A. Koh, D.
Chua, K. Soga, M. Overend, M. Aono, G. A. J. Amaratunga, and K. Okano
Applied Physics Letters 98, 152102 (2011).
20. T. Masuzawa, Y. Sato, Y. Kudo, I. Saito, T. Yamada, A. T. T. Koh, D. H. C.
Chua, T. Yoshino, W. J. Chun, S. Yamasaki, and K. Okano. “Correlation between
low threshold emission and C--N bond in nitrogen-doped diamond films” Journal
of Vacuum Science and Technology. B. 29, 02B119 (2011).
21. Y. M. Foong, A. T. T. Koh, J. Hsieh, and D. H. C. Chua. “A comparative study
on as-deposited and in situ oxidized ZnO/diamond-like carbon (DLC)
List of Publications
xxi
nanocomposite by pulsed laser deposition technique” Journal of Material
Research 25, 899 (2010).
22. J. Hsieh, A. T. T. Koh, D. H. C. Chua. “Surface studies of carbon-based ZnO
nanocomposite films” International Journal of Nanotechnology 6, 661 (2009)
Conference presentations:
1. Poster presentation “Segregation of Few-layer Graphene with Low Defects by
Pulsed Laser Deposition” 2012 Materials Research Society Spring Meeting, San
Francisco, California, USA, 9th
-13th
April 2012.
2. Poster presentation “Experimental and Theoretical Study on the Energy
Dependent Surface Evolution and Microstructure Changes in Copper
Nanostructured Composite” 2012 Materials Research Society Spring Meeting,
San Francisco, California, USA, 9th
-13th
April 2012.
3. Poster presentation “Field Emission Mechanism of Nitrogen-doped Diamond with
Different C-N Concentration” 2011 Materials Research Society Spring Meeting,
San Francisco, California, USA, 25th
-29th
April 2011.
4. Oral presentation “Electron Emission from Nanographite clusters embedded in a-
C films” 2010 8th International Vacuum Electron Sources Conference (IVESC),
Nanjing, China, 14th
–16th
October 2010.
5. Poster presentation ” XPS and TEM analysis of Multi-elemental Nanocomposites
in Diamond-like amorphous carbon films” Ecoss 2008:The 25th European
Conference on Surface Science, Liverpool, England, 27th
July–1st August 2008.
List of Publications
xxii
6. Poster presentation “Ti & Ni in Diamond Like Carbon as Growth Substrates for
Carbon Nanotubes”. The 2008 Asian Conference on Nanoscience and
Nanotechnology (AsiaNano 2008), Biopolis, Singapore 3rd–7th November 2008.
Chapter 1 Introduction
1
Chapter 1
Introduction
Over the last few years, extensive research efforts have been devoted to graphene
due to its unique and attractive material properties. In this chapter, the discovery of
graphene and its properties are reviewed in section 1.1 and 1.2, respectively. Fabrication
methods of graphene will be explained in section 1.3. The principles and equations used
in field emission (FE) studies, as well as the application of graphene as a potential field
emitter will be reviewed in Section 1.4. Lastly, the motivations and objectives of this
project are detailed in section 1.5.
1.1 Graphite and monolayer graphite
Graphite, an allotrope of carbon, has been long known since 1789. Being the
thermodynamically most stable form of carbon at standard conditions, it was believed
that a single graphite layer cannot exist freely. It was till the year 2004 when Novoselov
et al.1 demonstrated the isolation of single layer graphite planes through mechanical
exfoliation of a highly ordered pyrolytic graphite target (often referred to as the scotch-
tape technique) that refuted this belief. Graphene is the name given to the monolayer
graphite sheets, which are the building blocks of all graphitic materials.
The physical structure of graphene resembles that of a soft membrane and a sheet
of graphene is not flat but rippled.2 These ripples are caused by the pre-existing strains in
Chapter 1 Introduction
2
Fig. 1.1 A graphene sheet is the building block of all other graphitic structures ranging
from fullerenes (0D), carbon nanotubes (1D) to graphite (3D).2
graphene and can cause charge distribution inhomogeneities, which are not favorable
towards electronic quality. It can be rolled up to form 0D fullerenes, 1D carbon
nanotubes (CNT) or stacked to form 3D graphite (Fig. 1.1).3 In fact, when a pencil is
pressed on paper during writing, there could be many single layer graphene sheets in the
numerous graphite stacks produced on paper. However, due to the lack of inspection
tools, graphene cannot be easily differentiated from graphite, until the method of viewing
graphene flakes placed on SiO2 substrate with a specific thickness (300 nm) that makes
these elusive flakes visible (Fig. 1.2) was discovered.4 Since then, there have been
increased interests in these single and few-layer planar aromatic carbon films due to the
special properties that it exhibit.
Chapter 1 Introduction
3
Fig. 1.2 Single layer and bi-layer graphene visible on 300 nm SiO2 under an optical
microscope but hardly visible on a 200 nm SiO2 substrate.3
To clarify the nomenclature, graphene, when strictly defined, refers to a single
layer (SLG), bilayer graphene refers to 2 layers while few-layer graphene (FLG) has 3–
10 layers. Above 10 layers, it is generally referred to as graphite. While there are
similarities and differences between the properties of graphene and FLG, FLG grown on
4H-SiC has been shown to behave like SLG due to a different (from A-B stacking)
stacking structure, giving rise to rotational faults in every graphene layer, thus decoupling
the individual sheets and resulting in isolated graphene layers.5 In the next section, the
properties of graphene will be detailed.
1.2 Properties of graphene
1.2.1 Electronic properties
Graphene is a zero bandgap semiconductor that exhibits unique electronic
properties such as massless Dirac fermions (zero rest mass and effective speed of light),6
ambipolar electric field effect7 and anomalous quantum hall effect at room temperature,
which is a characteristic behavior of Dirac fermions.8, 9
The origin of the massless
Chapter 1 Introduction
4
fermions is due to the linear energy relationship between electron energy, E, and
wavevector, k, in the form of
22
yxf kkE (1.1)
near the vertices of the hexagonal Brillouin zone of graphene.6, 10
The six corners of the
2D Brillouin zones are therefore known at Dirac points. Ambipolar electric field effect
occurs when the charge carriers can be changed continuously between electron and holes
in concentrations as high as 1013
cm-2
by changing the externally applied voltage.1, 7
The
charge carriers of freely suspended graphene have an intrinsic mobility limit of 2 × 105
cm2/V.s at low temperatures, which is the highest known of any semiconductor. However,
the mobility is limited to 4 × 104 cm
2/V.s for graphene on SiO2 substrate, due to extrinsic
scattering by surface phonons, showing that substrate interactions does affect the
properties of graphene.11
As a result of the high mobility, graphene charge carriers are
able to exhibit ballistic transport on a submicron scale, in which the electrons transport in
a medium with negligible resistivity and have mean free path much larger than the
dimensions of the box that holds the medium.12
However, the wavy or rippled surface of
a graphene sheet results in breaking of the translational invariance and hence limits
ballistic transport. Nevertheless, these electronic properties make graphene a highly
suitable material for electronic applications and it has been shown to be a promising
material for transparent electrodes,13
ultracapacitors,14
field emission15, 16
and transistor
applications.17
As a result of not possessing a band gap, pristine SLG cannot be used as field
effect transistors (FET) as it cannot be switched off. However, this problem has been
Chapter 1 Introduction
5
overcame with different methods that can be divided into those that disrupt and break the
hexagonal structure of graphene to induce a band gap and those that preserve the 6 fold
symmetry of the structure.18
For the former, this includes slicing graphene into
nanoribbons,19
chemical doping20
and creating graphene nanomesh21
structures with
lithography. The use of bilayer graphene for applications that require a band gap is also
possible without having to destroy the hexagonal rings of carbon. For instance, Xia et
al.22
has demonstrated a band gap in bilayer graphene and an on/off current ratio of 100
(compared to 4 for SLG) at room temperature when a biasing condition was applied for
their dual gate graphene FET design. Zhang et al.23
have also demonstrated bilayer
graphene FET with a tunable band gap up to 0.25 eV, using infrared micro-spectroscopy,
while Quhe et al.18
made use of graphene-substrate interaction between single graphene
and hexagonal boron nitride to induce a band gap. These methods preserve the high
carrier mobility of graphene which is desirable in electronic applications.
1.2.2 Mechanical properties
The Young’s modulus of defect-free graphene using atomic force microscopy
(AFM) nanoindentation on graphene suspended over holes was measured to be 1.0 TPa at
a thickness of 0.335 nm, which makes graphene one of the strongest materials ever.24
As
graphene is a two dimension (2D) material, the measured values were normalized with
the area instead of volume and the Young’s modulus is equivalent to the in-plane
stiffness (E2D
). The same value of 1 TPa was obtained for single wall CNT, which is
essentially a rolled up graphene sheet. Given its excellent mechanical properties, it can
value add the use of graphene in many applications that may not be primarily mechanical
Chapter 1 Introduction
6
in nature. The use of graphene in composites has also been reported to greatly improve
tensile strength by 150 % and increase the Young’s modulus by 10 times at a low
graphene loading of 1.8 vol.%25
and has outperformed multiwall CNT as a reinforcement
additive.26
Possessing excellent mechanical properties as an electronic material can
increase the robustness of the fabricated device by allowing it to withstand harsh
conditions with lower chances of breakage.
1.2.3 Optical properties
A single layer or few-layer graphene is extremely thin and thus optically
transparent, with a transparency of 97.7 % in the visible light range. The transmittance
value of graphene decreases linearly with an increase in the number of layers.27
The inter-
band optical transitions in SLG and bilayer graphene can be varied dramatically by
electrical gating. This is because the low density of states near the Dirac point of
graphene causes the Fermi level, EF, to change significantly when the carrier density
changes.28
The combination of high transparency, high conductivity and ultrafast
photoresponse makes graphene extremely suitable for optoelectronic applications such as
sensors, communication devices and transparent electrodes found in liquid crystal
displays (LCD), touchscreens and for organic light emitting devices (OLED), replacing
the expensive indium-tin oxide (ITO) which is hard to recycle and natural indium sources
are depleting.13
Chapter 1 Introduction
7
1.2.4 Thermal properties
The thermal conductivity of graphene, κ, is dominated by phonon transport,
specifically diffusive conduction at high temperatures and ballistic conduction at
sufficiently low temperatures.27, 29
The measured value of κ for a suspended mechanically
exfoliated graphene is ~5000 Wm−1
K−1
, which is higher than that of bulk graphite at
room temperature (2000 Wm− 1
K−1
) and twice that of diamond (2500 Wm− 1
K−1
). This
value was measured by observing the red-shift of the Raman G peak which depends
linearly on the sample’s temperature. Through the slope of the excitation power versus G
peak frequency, κ was estimated.30
The high thermal conductivity of graphene is useful
and beneficial when efficient heat dissipation is required in electronic applications.
It has also been shown that by making use of the negative thermal expansion
coefficient and orientation of the ripples on graphene surface, wavelength and amplitude
of the ripples in graphene can be controlled by thermal manipulation. These
morphological changes in the ripples with temperature may be the key to explaining
some of the unusual properties of graphene as well as opening a new area for strain based
devices.31
1.3 Methods of fabrication
Graphene is seen as the material of the next generation given its excellent
properties, but the bigger issue that needs to be addressed is in the area of graphene
fabrication. Conventional techniques to fabricate graphene are expensive while some are
Chapter 1 Introduction
8
tedious. The most popular and easiest method is the mechanical exfoliation method
which involves rubbing a highly pyrolytic graphite target against a piece of sticky
cellophane tape and depositing the graphite flakes on SiO2 of carefully chosen thickness
(typically 300 nm). By looking under the microscope, single and FLG flakes can be
distinguished from the color observed (Fig. 1.2). Flakes produced by this method can be
up to 10 m in size but there is a large variation in the thickness of the flakes produced.
While any one can do this with the correct materials and without specialized equipment,
this method is tedious, time consuming and above all not scalable, thus making it
unsuitable for industries.32
The second technique used in fabricating graphene is by growing graphene
epitaxially on SiC. This method involves high process temperatures (>1000 oC and up to
1600 oC) and also ultra-high vacuum conditions of 10
-9 Torr. While this method is not
complicated and graphene grown on SiC can be patterned with standard nanolithography
techniques,33
the large lattice mismatch between epitaxial graphene and the underlying
SiC substrate makes the quality incomparable to films obtained with the mechanical
exfoliation method.34
It is also difficult to transfer the epitaxial graphene on SiC onto
other substrates due to the high chemical stability of SiC. Substrate bonding may also
affect the electronic properties of graphene grown on SiC as mentioned earlier.11
The third growth method of graphene is carried out on metal substrates. The
recent interests in free-standing graphene have led to the boom in research on graphene
adsorbed on metal surfaces. After graphene is absorbed on the metal, the metal support
layer can be easily etched away using an acid solution. The remaining graphene layers
can either be free-standing and dredged up from the solution, or transferred onto another
Chapter 1 Introduction
9
substrate of choice. Graphene on metal substrates can be prepared by both solid and
gaseous carbon sources. On one hand, segregation of dissolved carbon on the substrate
surface through high temperature annealing and cooling of a carbon-containing metal can
be carried out either by solid or gaseous sources. If the carbon source is gaseous, for e.g.
methane or ethylene, decomposition of the gas first occurs on the metal surface to release
carbon. On the other hand, chemical vapor deposition (CVD) is one technique that uses
only gaseous carbon sources. When copper (Cu) catalyst is used with CVD, the growth of
graphene is reported to stop once the metal layer is covered with a SLG, but continues to
give FLG if nickel (Ni) catalyst is used instead. This is due to the difference in their
growth mechanisms, which is surface adsorption and segregation respectively for Cu and
Ni.35
Thus, single layer growth can be achieved by controlling the temperature or catalyst,
which give CVD an edge over mechanical exfoliation that results in a wide range of
thicknesses.36
Additionally, CVD is a highly scalable method that can be used to fabricate
large area graphene (~cm2).
37, 38 Metals such as ruthenium,
39 platinum,
40 copper and
nickel37, 41
have been used in the synthesis of graphene. In the case of nickel, to prevent
the growth of thick layers of graphite, thin layers of Ni in the order of a few hundred
nanometers deposited on SiO2/Si substrates can be used.37
Chemical exfoliation is also frequently used to obtain a large amount of free-
standing graphene. This technique first began with the Hummers method, which was
first demonstrated by W. S. Hummers in 1958 to prepare graphite oxide.42
The graphite
was first oxidized to form graphite oxide using potassium permanganate (KMnO4) and
hydrosulfuric acid (H2SO4), and subsequently exfoliated to form graphene oxide (GO).
The actual chemicals used today may differ from the original Hummers method and thus
Chapter 1 Introduction
10
it is called modified Hummers method. GO, which is yellow-brown as shown in Fig. 1.3,
can be reduced to graphene through treatment with hydrazine or more recently,
microwave assisted reduction.43
The end product can be easily transferred onto other
substrates by filtration, spray coating and other solution processing methods.44
Chemical
exfoliation provides a low cost and high yield method to prepare graphene and is suitable
for large scale production.
Fig. 1.3 GO produced by Hummers method is yellow-brown in color while reduced graphene is
black in color.45
Chapter 1 Introduction
11
1.4 Field emission theory
1.4.1 Model and principles
Field emission is the extraction of electrons from a metal or semiconductor under
a strong electric field into vacuum by tunneling through the surface triangular potential
barrier.46
When an electric field is applied, the surface potential is deformed such that the
shape of the surface potential barrier changes from square (no electric field) to
triangular.47
This is in contrast to thermionic emission where heat is applied and electrons
gain sufficient energy to surmount the barrier instead of tunneling through the barrier.
This difference in barriers between these two types of emission is depicted in Fig. 1.4.
Fig. 1.4 Difference in emission barrier between field emission and thermionic emission.
Chapter 1 Introduction
12
Fowler and Nordheim proposed the first model for field emission in 1928 for
metal surfaces at low temperatures.48
Nonetheless, this model is widely used in field
emission studies of graphitic materials till today. In this model, the emission current
density and macroscopic applied field can be described by the Fowler-Nordheim (F-N)
relation according to the following expression,49
E
BEAJ
23
2
exp)(
(1.2)
Where J is the macroscopic current density, E is the applied field, β is the field
enhancement factor, is the work function and constants A=1.54×10-6
AeV.V-2
,
B=6.83×107 V.eV
-3/2cm
-1.
By taking natural logarithm, equation 1.2 can be rewritten into the following
linear equation 1.3,
E
BA
E
J
23
2
2lnln
(1.3)
Field enhancement factor (β), which is used as a qualitative measure of field
emission property of the material, can be calculated from the slope (∆) of the straight
line plot of ln (J/E2) versus 1/E for a sample of known value using,
23
B (1.4)
Likewise, if β is known, can be calculated from the slope of the F-N plot instead.
Chapter 1 Introduction
13
The local electric field, Elocal, experienced on a microscopic scale at the emitting
site is thus related to the macroscopic applied field, Eappl, through β by the following
expression.
.appllocal EE (1.5)
1.4.2 Field emission of carbon materials
Field emission from carbon-based materials has been studied intensively for
decades and some of the carbon materials used as field emission cold cathodes are
diamond, doped or undoped DLC, and the well-known CNT owing to its high aspect ratio
and increased geometrical enhancement factor.
In field emission applications, diamond has attracted a lot of attention as cold
cathode emitters due to advantageous properties like high thermal conductivity and
negative electron affinity (NEA). NEA means that the minimum energy of electrons in
the vacuum is below the minimum energy of electrons in the conduction band.50
Phosphorous-doped diamond and nitrogen-doped diamond has been extensively studied
for emission along with different surface terminations such as oxygen and hydrogen.51
It
has been reported that field emission of heavily N-doped diamond is not limited by its
resistance and thus follows a metal-insulator-vacuum mechanism, whereby the electrons
are injected from the back contact into the conduction band of diamond and subsequently
flew into vacuum without a potential barrier due to NEA.52
It should be noted that NEA
Chapter 1 Introduction
14
is not a prerequisite to low applied fields as there are materials without NEA that still
exhibit good emission properties e.g., diamond-like carbon (DLC) and CNT.
Generally, low macroscopic threshold fields (<1–40 V/m) observed in carbon
materials are attributed to either geometrical field enhancement factors (external field
enhancement) or field enhancement due to internal conductive elements (internal field
enhancement). For CNTs, they have a high aspect ratio and are fully sp2 bonded; hence
electron transport is not a limiting factor within the tubes. Instead, the density of the
CNTs affects the emission threshold due to shielding effects. In addition, though CNTs
show remarkable performance as field emitters, their thermal stability and reliability are
often inferior, i.e. easily degraded under high applied voltages.
In the case of DLC, which possesses extremely smooth surfaces, the field
enhancement factors are largely contributed by internal field enhancement. The electric
field lines will concentrate at surface conductive elements, which are the sp2 bonded
regions in this case, and lead to enhancement of the applied field. Moreover, owing to the
presence of sp2 in combination with sp
3 bonding, DLC films have a certain edge over
fully sp3 bonded diamond since the sp
2 can provide electron conduction paths within the
films. Furthermore, sp2 clusters have different dielectric constants from sp
3 carbon,
giving rise to enhancement due to dielectric inhomogenity.53
The ability to grow large
area DLC at room temperature through economical methods is an added bonus to its
excellent properties, thus making it a potential candidate for cold cathode emitters.54
DLC
is known to emit electrons with threshold fields in the order of 10 Vµm-1
.55
There have
been various attempts to lower the emission of DLC films by doping metals such as Ag,56
Chapter 1 Introduction
15
Ti57
and Ni.58
In such cases, the increase in sp2 percentage and improved conductivity in
the films by metal incorporation plays a large part in the reduction of threshold field.
Graphene is essentially an unrolled single walled CNT and possesses excellent
conductivity. However, a single graphene sheet may not have good field enhancement
due to its rather flat profile. This is further aggravated by the fact that most graphene
fabrication methods including mechanical exfoliation, decomposition of SiC, segregation
from metals and CVD result in graphene that has a planar morphology parallel to the
substrate. To fully exploit the capabilities of graphene as a field emitter, the graphene
sheets should be perpendicular to the substrate such that the planes of highest
conductivity are along the direction of emission and electrons can thereby emit from the
sharp graphene edges. There are several studies that strive to achieve this in current
literature. Some early works include Eda et al.16
who fabricated graphene cathodes by
spin coating chemically exfoliated graphene onto Si substrate to gain a certain degree of
vertical alignment. The orientation of the sheets was related to the spin coating speeds
and best results were obtained from films spin coated at 600 rpm. Even though the turn-
on field (Eon) was 4 V/m, the maximum current density obtained was 1 mA/cm2, which
was below the requirement for actual high current applications. The authors have recently
carried out the same method of spin coating graphene sheets onto Si microtips and
lowered the Eon to 2.3 V/m.59
Another research group, Malesevic et al.60
managed to
fabricate vertically aligned FLG on titanium and silicon substrates by microwave plasma
enhanced CVD (MW-PECVD). The quality of FLG was optimized by controlling the H2
and CH2 precursor gas ratios. Using this method, the Eon was lowered to 1 V/m and
current density was increased to 14 mA/cm2. Although this method has some advantages
Chapter 1 Introduction
16
Fig. 1.5 Vertically oriented graphene flakes fabricated by spin coating onto conductive silicon
substrates in ref. 16.
such as the direct fabrication of graphene without extra steps, there is no simple way to
control the FLG density, which gives rise to screening effect (similar to the case of a
dense forest of vertically aligned CNT).
Chemically exfoliated graphene has also been studied as field emitters by using
screen printing techniques61
, electrophoretic deposition62
and coating on conductive
tape62
with turn-ons field ranging from 1–5.2 V/μm. Graphene, first exfoliated by
modified Hummers method and reduced by hydrazine, were compared against those
reduced by microwave assistance and it was reported that microwave-reduced graphene
(Eon = 0.39 V/μm) performed better than hydrazine-reduced graphene (Eon = 0.94 V/μm).
The authors attributed it to the larger size of graphene flakes produced by microwave
reduction leading to more protrusions when stacked, thus improving field enhancement.45
Another example of achieving protrusions was recently demonstrated by Pandey et al.
who generated morphological defects (such as edges, discontinuity and ripples) due to the
poor transfer and adhesion of CVD-grown graphene onto pristine Si (100) substrates.63
Chapter 1 Introduction
17
However, this method does not give reliable control over the density of defects and it can
be difficult to reproduce the results. Growth of graphene on nanotube or wire structures
have also been another area explored lately. For instance, Deng et al. reported the direct
growth of FLG on the tips of CNT arrays using a RF H plasma sputtering system.64
The
fabricated structure had sharp graphene edges extending outwards on the tips of a well
aligned CNT array and gave slightly better field emission results compared to a pure
CNT array. Other methods to create graphene-based cathodes with a field enhancing
morphology include the draping of graphene oxide on Ni nanotips and the Eon was
reported 5 × 105 V/m at 1 μA, which is a relatively high field and this is probably due to
the much lower conductivity of graphene oxide.65
Furthermore, Yang et al. reported the
large area transfer of CVD fabricated monolayer graphene onto a well aligned array ZnO
nanotips to produce 100–200 nm graphene-covered protrusions.66
These emitters were
capable of very high current density (~ 500 µA/cm2), clearly demonstrating the potential
of graphene for FE applications. Due to the unique properties of graphene, it is not
surprising that it has received considerable attention for field emission; however the
current graphene field emission technology is still at its infancy and will require more
time before reaching actual applications.
Chapter 1 Introduction
18
1.5 Motivations and objectives
From the above literature, it can be seen that there are still gaps in the area of
graphene fabrication. Since the first report on the formation of monolayer graphene by
Novoselov et al., extensive efforts have been devoted to the fabrication of graphene using
CVD, mechanical and chemical exfoliated methods. However, limited information is
available for physical vapor deposition (PVD) fabrication techniques, such as pulsed laser
deposition (PLD). The use of PLD could reduce carbon dissolution time through the
generation of charged energetic species that penetrate into the substrate. Despite
numerous studies on the growth mechanism of CVD deposited graphene, information on
the mechanism of solid-state transformation, segregation and growth mechanism of
graphene using PLD technique is lacking. Furthermore, even though carbon-based
composites such as metal-doped DLC and metal-coated CNT have been explored as new
generation field emitters, high sp2 carbon-based composites of graphene/CNT have not
been investigated.
As such, my research work was devoted to the following objectives:
1) Fabricate high sp2 carbon/graphene/few-layer graphene using physical deposition
methods i.e. PLD and understand the growth mechanism of graphene using PLD
segregation methods.
2) Study the field emission properties of the fabricated high sp2 and graphene-based
emitter materials.
3) Fabricate and explore high sp2 hybrid materials such as graphene/CNT for field
emission applications.
Chapter 1 Introduction
19
By achieving the objectives in this thesis, this work presented alternative methods
to the fabrication of graphene and has shed some light to understanding the growth
mechanism of graphene using PLD method. Through the fabrication and study of the
emission properties of graphene and graphene/CNT hybrid materials, this work will be
helpful towards contributing to the future development of low cost field emitters for
actual industrial applications.
Chapter 1 Introduction
20
References
1. K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, Y. Zhang, S. V. Dubonos, I.
V. Grigorieva and A. A. Firsov, Science 306, 666 (2004).
2. A. Fasolino, J. H. Los and M. I. Katsnelson, Nature Mater. 6, 858-861 (2007).
3. A. K. Geim and K. S. Novoselov, Nature Mater. 6, 183-191 (2007).
4. P. Blake, E. W. Hill, A. H. C. Neto, K. S. Novoselov, D. Jiang, R. Yang, T. J. Booth
and A. K. Geim, Appl. Phy. Lett. 91, 063124 (2007).
5. J. Hass, F. Varchon, J. E. Millán-Otoya, M. Sprinkle, N. Sharma, W. A. de Heer, C.
Berger, P. N. First, L. Magaud and E. H. Conrad, Phys. Rev. Lett. 100, 125504
(2008).
6. L. S. Panchakarla, K. S. Subrahmanyam, S. K. Saha, A. Govindaraj, H. R.
Krishnamurthy, U. V. Waghmare and C. N. R. Rao, Adv. Mater. 21, 4726 (2009).
7. K. S. Novoselov, S. V. Morozov, T. M. G. Mohinddin, L. A. Ponomarenko, D. C.
Elias, R. Yang, I. I. Barbolina, P. Blake, T. J. Booth, D. Jiang, J. Giesbers, E. W. Hill
and A. K. Geim, phys. status solidi (b) 244, 4106-4111 (2007).
8. K. S. Novoselov, Z. Jiang, Y. Zhang, S. V. Morozov, H. L. Stormer, U. Zeitler, J. C.
Maan, G. S. Boebinger, P. Kim and A. K. Geim, Science 315, 1379 (2007).
9. A. H. Castro Neto, F. Guinea, N. M. R. Peres, K. S. Novoselov and A. K. Geim, Rev.
Mod. Phys. 81, 109-162 (2009).
10. M. I. Katsnelson, Materials Today 10, 20-27 (2007).
11. J. H. Chen, C. Jang, S. Xiao, M. Ishigami and M. S. Fuhrer, Nature Nanotech. 3, 206-
209 (2008).
12. K. S. Novoselov, E. McCann, S. V. Morozov, V. I. Falko, M. I. Katsnelson, U.
Zeitler, D. Jiang, F. Schedin and A. K. Geim, Nature Phys. 2, 177 (2006).
Chapter 1 Introduction
21
13. J. Wu, M. Agrawal, H. c. A. Becerril, Z. Bao, Z. Liu, Y. Chen and P. Peumans, ACS
Nano 4, 43-48 (2009).
14. M. D. Stoller, S. Park, Y. Zhu, J. An and R. S. Ruoff, Nano Lett. 8, 3498-3502 (2008).
15. M. Araidai, Y. Nakamura and K. Watanabe, Phys. Rev. B 70, 245410 (2004).
16. G. Eda, H. Emrah Unalan, N. Rupesinghe, G. A. J. Amaratunga and M. Chhowalla,
Appl. Phys. Lett. 93, 233502 (2008).
17. M. Freitag, Nature Nanotech. 3, 455-457 (2008).
18. R. Quhe, J. Zheng, G. Luo, Q. Liu, R. Qin, J. Zhou, D. Yu, S. Nagase, W.-N. Mei, Z.
Gao and J. Lu, NPG Asia Mater. 4, e6 (2012).
19. X. Wang, Y. Ouyang, X. Li, H. Wang, J. Guo and H. Dai, Phys. Rev. Lett. 100,
206803 (2008).
20. F. Withers, M. Dubois and A. K. Savchenko, Phys. Rev. B 82, 073403 (2010).
21. J. Bai, X. Zhong, S. Jiang, Y. Huang and X. Duan, Nature Nanotech. 5, 190-194
(2010).
22. F. Xia, D. B. Farmer, Y.-m. Lin and P. Avouris, Nano Lett. 10, 715-718 (2010).
23. Y. Zhang, T.-T. Tang, C. Girit, Z. Hao, M. C. Martin, A. Zettl, M. F. Crommie, Y. R.
Shen and F. Wang, Nature 459, 820-823 (2009).
24. C. Lee, X. Wei, J. W. Kysar and J. Hone, Science 321, 385-388 (2008).
25. X. Zhao, Q. Zhang, D. Chen and P. Lu, Macromolecules 43, 2357-2363 (2010).
26. M. A. Rafiee, J. Rafiee, Z. Wang, H. Song, Z.-Z. Yu and N. Koratkar, ACS Nano 3,
3884-3890 (2009).
27. Y. Zhu, S. Murali, M. D. Stoller, A. Velamakanni, R. D. Piner and R. S. Ruoff,
Carbon 48, 2118-2122 (2010).
28. F. Wang, Y. Zhang, C. Tian, C. Girit, A. Zettl, M. Crommie and Y. R. Shen, Science
320, 206-209 (2008).
Chapter 1 Introduction
22
29. C. Yu, L. Shi, Z. Yao, D. Li and A. Majumdar, Nano Lett. 5, 1842-1846 (2005).
30. A. A. Balandin, S. Ghosh, W. Bao, I. Calizo, D. Teweldebrhan, F. Miao and C. N.
Lau, Nano Lett. 8, 902-907 (2008).
31. W. Bao, F. Miao, Z. Chen, H. Zhang, W. Jang, C. Dames and C. N. Lau, Nature
Nanotech. 4, 562-566 (2009).
32. A. K. Geim, Science 324, 1530-1534 (2009).
33. C. Berger, Z. Song, X. Li, X. Wu, N. Brown, C. Naud, D. Mayou, T. Li, J. Hass, A.
N. Marchenkov, E. H. Conrad, P. N. First and W. A. de Heer, Science 312, 1191-
1196 (2006).
34. H. Huang, W. Chen, S. Chen and A. T. S. Wee, ACS Nano 2, 2513-2518 (2008).
35. X. Li, W. Cai, L. Colombo and R. S. Ruoff, Nano Lett. 9, 4268-4272 (2009).
36. J. Wintterlin and M. L. Bocquet, Surf. Sci. 603, 1841-1852 (2009).
37. K. S. Kim, Y. Zhao, H. Jang, S. Y. Lee, J. M. Kim, K. S. Kim, J.-H. Ahn, P. Kim, J.-
Y. Choi and B. H. Hong, Nature 457, 706-710 (2009).
38. X. Li, W. Cai, J. An, S. Kim, J. Nah, D. Yang, R. Piner, A. Velamakanni, I. Jung, E.
Tutuc, S. K. Banerjee, L. Colombo and R. S. Ruoff, Science 324, 1312-1314 (2009).
39. P. W. Sutter, J.-I. Flege and E. A. Sutter, Nature Mater. 7, 406-411 (2008).
40. H. Ueta, M. Saida, C. Nakai, Y. Yamada, M. Sasaki and S. Yamamoto, Surf. Sci. 560,
183-190 (2004).
41. A. Reina, X. Jia, J. Ho, D. Nezich, H. Son, V. Bulovic, M. S. Dresselhaus and J.
Kong, Nano Lett. 9, 30-35 (2008).
42. W. S. Hummers and R. E. Offeman, J. Am. Chem. Soc. 80, 1339-1339 (1958).
43. W. Chen, L. Yan and P. R. Bangal, Carbon 48, 1146-1152 (2010).
44. S. Pei, J. Zhao, J. Du, W. Ren and H.-M. Cheng, Carbon 48, 4466-4474 (2010).
Chapter 1 Introduction
23
45. K. Wang, T. Feng, M. Qian, H. Ding, Y. Chen and Z. Sun, Appl. Surf. Sci. 257, 5808
(2011).
46. R. G. Forbes, Solid-State Electron. 45, 779-808 (2001).
47. J.-M. Bonard, M. Croci, I. Arfaoui, O. Noury, D. Sarangi and A. Châtelain, Diam.
Relat. Mater. 11, 763-768 (2002).
48. R. H. Fowler and L. Nordheim, Proc. R. Soc. A 119, 173-181 (1928).
49. C. X. Xu, X. W. Sun and B. J. Chen, Appl. Phy. Lett. 84, 1540-1542 (2004).
50. M. W. Geis, N. N. Efremow, K. E. Krohn, J. C. Twichell, T. M. Lyszczarz, R. Kalish,
J. A. Greer and M. D. Tabat, Nature 393, 431-435 (1998).
51. T. Yamada, H. Kato, S. Shikata, C. Nebel, H. Yamaguchi, Y. Kudo and K. Okano, J.
Vac. Sci. Technol. B Microelectro.n Nanometer. Struct. Process. Meas. Phenom. 24,
967 (2006).
52. K. Okano, T. Yamada, A. Sawabe, S. Koizumi, J. Itoh and G. A. J. Amaratunga,
Appl. Phy. Lett. 79, 275 (2001).
53. J. D. Carey, R. D. Forrest, R. U. A. Khan and S. R. P. Silva, Appl. Phy. Lett. 77,
2006 (2000).
54. J. Robertson, Carbon 37, 759 (1999).
55. J. Robertson, Mater. Sci. Eng. R 37, 129 (2002).
56. S. Ahmed, M. Moon and K. Lee, Appl. Phy. Lett. 92, 193502 (2008).
57. X. Z. Ding, Y. J. Li, Z. Sun, B. K. Tay, S. P. Lau, G. Y. Chen, W. Y. Cheung and S.
P. Wong, J. Appl. Phys. 88, 3842 (2000).
58. A. T. T. Koh, J. Hsieh and D. H. C. Chua, Diam. Relat. Mater. 19, 637 (2010).
59. E. Stratakis, G. Eda, H. Yamaguchi, E. Kymakis, C. Fotakis and M. Chhowalla,
Nanoscale 4, 3069-3074 (2012).
Chapter 1 Introduction
24
60. A. Malesevic, R. Kemps, A. Vanhulsel, M. P. Chowdhury, A. Volodin and C. Van
Haesendonck, J. Appl. Phys. 104, 084301 (2008).
61. Q. Min, F. Tao, D. Hui, L. Lifeng, L. Haibo, C. Yiwei and S. Zhuo, Nanotechnol. 20,
425702 (2009).
62. Z.-S. Wu, S. Pei, W. Ren, D. Tang, L. Gao, B. Liu, F. Li, C. Liu and H.-M. Cheng,
Adv. Mater. 21, 1756-1760 (2009).
63. S. Pandey, P. Rai, S. Patole, F. Gunes, G.-D. Kwon, J.-B. Yoo, P. Nikolaev and S.
Arepalli, Appl. Phy. Lett. 100, 043104 (2012).
64. J.-h. Deng, R.-t. Zheng, Y.-m. Yang, Y. Zhao and G.-a. Cheng, Carbon 50, 4732
(2012).
65. D. Ye, S. Moussa, J. D. Ferguson, A. A. Baski and M. S. El-Shall, Nano Lett. 12,
1265-1268 (2012).
66. Z. Yang, Q. Zhao, Y. Ou, W. Wang, H. Li and D. Yu, Appl. Phys. Lett. 101, 173107
(2012).
Chapter 2 Experimental techniques
25
Chapter 2
Experimental techniques
In this chapter, the principles behind each experimental technique used will be
described. The techniques used for thin film deposition will be detailed in section 2.1,
while section 2.2 describes the experimental methods used to fabricate nanographite,
graphene and graphene emitters. Lastly, the material characterization techniques will be
detailed in section 2.3.
2.1 Thin film deposition techniques
2.1.1 Magnetron sputtering
Magnetron sputtering is a powerful and flexible technique that can be used to
make coatings of solid metal, alloys and other compounds. It can be used with a direct
current (DC) or radio frequency (RF) source. DC sputtering is not suitable for insulators
as it causes positive charge buildup on the target, which in turn repels the positive argon
(Ar) ions and stops the sputtering process.1 On the other hand, RF allows the sputtering
of insulators and operation at lower Ar pressures, leading to less collisions and better
line-of-sight deposition.2 In this technique, a target, which is the cathode, is bombarded
by energetic ions. These ions, typically Ar ions, are produced in a glow discharge plasma
between the cathode and anode in the vacuum chamber in the range of 10-6
Torr. Upon
ion bombardment, some of the target atoms are dislodged or sputtered and subsequently
condensed onto the substrate. Secondary electrons produced as part of the bombardment
Chapter 2 Experimental techniques
26
process also plays a part in the continuation of the plasma. The difference between
conventional sputtering and magnetron sputtering is that in order to increase the
sputtering rate, a magnetic field is applied. The magnetic field, which is applied
perpendicularly to the electric field, constraints the motion of the electrons causing them
to move in a helical manner near to the target surface region. By limiting the motion of
the electrons near the target, higher sputtering rates can be achieved due to an increased
probability of electron-atom collision that results in higher ionization ratios and a dense
plasma is produced near the target surface. Subsequently, the increase in ionization rates
results in higher deposition rates.3 A schematic of magnetron sputtering is shown in Fig.
2.1.4 In this project, magnetron sputtering was used to deposit nickel (Ni), cobalt (Co),
iron (Fe) and copper (Cu) thin films onto highly doped (n++
) silicon substrates. The metal
thin films were then used as substrates for further experiments.
Fig. 2.1 Schematic of RF magnetron sputtering showing the in-chamber configurations.4
Chapter 2 Experimental techniques
27
2.1.2 Pulsed laser deposition (PLD)
Pulsed laser deposition is a physical vapor deposition technique, which has been
effectively used to fabricate thin films material such as DLC, metal multilayers, ceramics,
superconductive and nanostructured materials.5-9
The advantages of using PLD for thin
film deposition include the ability to grow high quality films, to ablate any material and
to obtain a stoichiometric transfer of target material onto the substrate, which is
especially useful in the case of composite materials.10-12
PLD systems come in different
laser wavelengths depending on the source of laser. Some common wavelengths used are
ArF (193 nm), KrF (248 nm) and ND:YAG (1064, 532, 355 and 266 nm) lasers from the
first to fourth harmonic.13-15 ArF and KrF lasers are also known as excimer lasers that
radiate energy in the ultra-violet range and they make use of a gaseous mixture as laser
source.
The PLD system used in the experiments in this work consisted of a KrF (λ=
248 nm) Lambda Physik laser, an optics system to guide the laser and a homebrewed
deposition chamber as shown in Fig. 2.2. The system consists of a box where the pulsed
laser is produced and a vacuum chamber where the substrate and target are placed. The
substrate and the target are positioned 5 cm apart and facing each other in the chamber,
which is evacuated to a pressure of 10-6
Tor by a turbomolecular pump and a roughing
pump. Short pulses of laser are guided by various focusing optic lens before it enters the
vacuum chamber and strikes the target. The target then absorbs the focused laser and
when the energy density used is sufficiently high, ablation of the target occurs, thereby
expelling a highly energetic and forward directed flux of material, known as the plume,
Chapter 2 Experimental techniques
28
Fig. 2.2 Schematic of PLD used for nanographite in amorphous carbon deposition and graphene
segregation.
Fig. 2.3 Plume seen during the ablation of target.
Chapter 2 Experimental techniques
29
orthogonally from the target surface. The plume, which condenses on the substrate
surface placed directly in front of the target, consists of ions, atoms, excited states,
electrons and macroparticles or fragments of expelled material.6 Figure 2.3 shows the
bluish green plume seen during deposition and also the positions of the laser, substrate
and target in the chamber.
The deposited film properties are affected by several parameters of the PLD
deposition. Ion energy, which is determined by laser fluence, is one of the critical factors
known to affect the type of carbon film formed, whether diamond-like or graphitic
films.16
PLD has the widest range of acceptable energies without requiring any external
biasing compared to various other techniques.9 Furthermore, in addition to the laser
fluence (J/cm2), the repetition rate of the laser pulses, substrate temperature, target to
substrate distance and deposition environment (such as the use of reactive gases) can be
varied, thus making PLD a versatile technique. The typical pulse duration of such lasers
is in the nanosecond range. However in the recent decade, the use of femtosecond pulse
(10-15
sec) duration lasers, which have a 800 nm wavelength have been explored.17-19
The
extremely fast pulse duration increases the energy of the ejected particles up to a few keV
and highly wear resistant DLC materials can be more easily obtained. A detailed review
on PLD and its mechanisms related to film growth can be found in the review by
Willmott and Hubber in reference 20.20
Chapter 2 Experimental techniques
30
2.2 Graphene/graphene films fabrication methods
2.2.1 Modified Hummers method
Generally, modified Hummers method first involves the oxidation of graphite into
graphite oxide through the intercalation and expansion of the graphite lattice. Graphite
oxide is then exfoliated into graphene oxide (GO) through ultrasonication or stirring in
water and GO is reduced to form graphene. The reduction of GO is an important step as
GO is an insulator and has different properties from graphene. This method uses natural
graphite powder, potassium permanganate (KMnO4), sodium nitrate (NaNO3),
hydrosulfuric acid (H2SO4), hydrogen peroxide (H2O2) aqueous solution and ultrapure
water. Graphite powder and NaNO3 was added to H2SO4 with the jar in an ice bath.
While stirring, KMnO4 was slowly added to the mixture to prevent the temperature from
exceeding 20 oC. Upon completion of this step, the ice bath was removed and the
temperature was maintained at 35 oC for 30 min. Water was then stirred in, causing
effervescence and raising the temperature to 98 oC. The brown colored suspension was
kept at this temperature for 15 min and more water was added. H2O2 was then added to
reduce the residual KMnO4. After the reduction process, the suspension consisting of
graphite oxide is a bright yellow color. This graphite oxide mixture was then filtered and
washed with water and remaining metal ions were treated and removed. Exfoliation of
graphite oxide into GO can be done by ultra-sonication. GO was reduced to graphene and
this was achieved by microwave assisted reduction of its aqueous suspension for the
graphene flakes used in this work.21
Chapter 2 Experimental techniques
31
2.2.2 Electrophoretic deposition (EPD)
Electrophoretic deposition is a widely used technique in the industry for the
fabrication of phosphors for high resolution displays as well as coatings.22, 23
Its
advantages include high deposition rate, low cost, ability to deposit on substrates with
complex shapes, ease of up-scaling due to its simplicity and the flexibility to deposit
many materials. It is also capable of producing homogeneous films with a high packing
density.24
In this technique, two electrodes are placed in a colloidal suspension of the
material. In our setup, the cathode is the substrate and the anode is a graphite electrode.
During the preparation of the suspension, a charger material, such as Mg(NO3)2,
La(NO3)3, MgCl2, or Al(NO3)3, is added to the solution. The metal ions are preferentially
Fig. 2.4 Schematic of EPD setup with planar electrodes.
Chapter 2 Experimental techniques
32
adsorbed onto the powder particles, charging it in the process. EPD is then achieved by
two steps. First, when an electric field is applied, charged particles in the suspension
move towards the electrode. Second, the particles are accumulated at the respective
electrode to form a coherent deposit.24
A schematic of the EPD setup is shown in Fig. 2.4.
2.3 Materials characterization
2.3.1 Scanning electron microscope (SEM)
Scanning electron microscope is a useful and flexible imaging technique for
surface and subsurface analysis of solid materials. The probe in SEM is an electron
source from a tungsten hairpin (thermionic source), LaB6 crystal or field emission gun
(FEG).25
To obtain surface images, Philips XL-30 and Zeiss Supra SEM, both with FEG
sources, were employed. SEM allows the viewing of specimens too small to be examined
by a light microscope. Since the wavelength of electrons is much smaller than that of
visible light, it has a higher resolving power. Over the years, the resolution for SEM has
improved from 50 nm to 1–5 nm today and from a simple tool it has now become an
equipment widely used in many areas of such as biology, chemistry and metallurgy.26
In its primary mode, the secondary electrons are detected in SEM to give us
images of the sample.27
The images appear three dimensional due to the small angular
aperture of the electron probe giving the large depth of field that the SEM possesses .As a
result, rough surfaces such as pollen grains stay in focus throughout the whole sample
image.
Chapter 2 Experimental techniques
33
No lenses are involved in SEM image magnification. Instead, the electrons raster
across the surface of the specimen and at the same time, a spot of the cathode ray tube
(controlled by current from the detector) scans across a screen in a similar rectangular
fashion except that the area is far larger. SEM imaging is always carried out in a vacuum
environment as the presence of gas molecules may cause the instability of the electron
beam from the source. Also, samples are required to be conductive to avoid charging,
which may cause a distorted image to form. Thus, samples that are non-conductive have
to be sputtered with a layer of gold first. In addition, a SEM equipped with energy
dispersive spectroscopy capabilities can also give us the elemental composition of the
sample by analyzing the characteristic x-rays given off by the sample.27
2.3.2 Transmission electron microscope (TEM)
Transmission electron microscopy is an imaging technique whereby a beam of
electrons, focused by magnetic lens, passes through the thickness of the sample,
interacting with the specimen as this happens and forms a two dimensional projected
image of a three dimensional specimen.28
Some basic instrumentation in TEM are an
electron gun to illuminate the sample, magnetic lens to demagnify and control the size of
the beam hitting the sample, fluorescent viewing screen, a charge-coupled device (CCD)
camera. A detailed setup of TEM can be found in ref. 27.27
The point to point resolution
of TEM is reported to be 0.1 nm or better. Such a resolution allows viewing of the
arrangement of individual atoms, thus allowing the periodic lineup of atoms in crystalline
structures to be seen.
Chapter 2 Experimental techniques
34
TEM viewing is also carried out in a vacuum environment. Also, another
requirement is that the sample must be thin enough to be transparent to electrons. For
nanoparticles, specially designed TEM copper (Cu) grids are used to support the film
which is deposited on top of the grids and such samples can be viewed directly without
polishing or grinding.27
2.3.3 Raman spectroscopy
Raman spectroscopy is a vibrational spectroscopy technique that can obtain
information about molecular vibrational motion and acquire fingerprints to help identify
and quantify samples.29
It is a flexible technique that can be used on solids, liquids,
powders and films. It can also be used on mixtures and composites as well and the
resulting spectrum is just a superposition of the individual components. It does not
require a vacuum environment to operate in and the spectra can be acquired relatively
quickly. No sample preparation is needed and the spectra are characteristic of each type
of material. Raman spectroscopy is especially useful in the study of carbon materials as it
is highly sensitive to symmetric covalent bonds that have little or no dipole moment
which is the case of carbon-carbon bonds.30
Chapter 2 Experimental techniques
35
Fig. 2.5 Three possible outcomes during the interaction of photon and matter in Raman
spectroscopy.
In this technique, a monochromatic light or laser, in the visible, infrared or
ultraviolet range is focused by a microscope onto the sample. The laser interacts with the
sample and causes excitation of the molecules to a virtual energy state. Depending on the
nature of the interaction of the photon with the sample, three outcomes, as depicted in Fig.
2.5, can occur when the molecule relaxes and the scattered photon is collected.31
Firstly,
elastic scattering or Rayleigh scattering is said to have occurred if the scattered photon is
of the same frequency as the incident photon, and this happens to the majority of photons.
If the molecules relax into a different vibrational state, the energies of the scattered
photons are shifted either up or down and this is also known as the Raman Effect. The
difference in the energy between the incident photon and scattered photon provides
information on the microstructure of the sample. In the case where the final vibrational
state of the molecule is higher, the scattered photon is of a lower frequency in order to
maintain the energy balance of the system and this is called a Stokes shift. Conversely, if
the final vibrational state of the molecule is lower and the scattered photon is of a higher
frequency, this is known as an Anti-Stokes shift. A Renishaw Raman spectrometer 2000
Chapter 2 Experimental techniques
36
with 514.51 nm green argon laser was used in this project. The laser beam was focused
onto the sample surface using an optical microscope with a magnification of 50 times at
spot size of 1 μm. The Raman spectra were acquired in the range of 1000–3000 cm-1
.
2.3.4 X-ray photoelectron spectroscopy (XPS)
X-ray photoelectron spectroscopy is a surface analysis technique that provides
chemical state and compositional information from an outer surface thickness of 1–10 nm.
It is capable of detecting all elements except helium and hydrogen.32, 33
Comparing
various surface analysis instruments like Auger Electron Spectroscopy (AES) and
Secondary Ion Mass Spectroscopy (SIMS), it is the easiest to obtain chemical state
information from XPS data. Thus, XPS is the preferred choice when information in this
area is required.32
For our purpose, a monochromatic XPS (Al Kα (1486.6 eV)) operated
at 275 W with a pass energy of 10 eV (narrow scan) was used.
Instrumentation in XPS consists of several components. They include a
preparation chamber for sample cleaning and experiments, analytical chamber where the
photon source is present, a hemispherical electron analyzer to measure the kinetic energy
(KE) of the photoelectrons and a detector at the end. X-ray photons bombard the surface
atoms of the sample. With sufficient energy, the photons can excite the core electrons of
the atoms and cause the ejection of a photoelectron from an atom. The KE of the
photoelectron ejected is then measured and converted into binding energy. This binding
energy is characteristic of the element and its oxidation state. Binding energy, EB, can be
calculated from equation 2.1,34
Chapter 2 Experimental techniques
37
EB = hυ – Ek – (2.1)
where, EB is the binding energy of the photoelectron, h is Planck’s constant, υ is the
frequency of the x-ray source, Ek is the kinetic energy of the photoelectron and is the
work function of the spectrometer.
From the equation, it can be seen that the ejected photoelectron must reach the
detector with an undisrupted path or else analytical information will be lost. As such, an
ultra-high vacuum environment is required in XPS, to fulfill this condition as well as to
prevent surface contamination. The sample surface can also be cleaned by argon sputter
to remove any oxidized layer or contaminants before the information is acquired.
2.3.5 Atomic force microscopy (AFM)
Atomic force microscopy is a type of scanning probe microscope that has superior
resolving power. It consists of a scanning tip rastering across the surface to obtain a
morphology image of atomic resolution. A Veeco, Nanoscope 3, Multimode AFM was
used to characterize the films produced.
Unlike SEM and scanning tunneling microscopy, AFM works for both conductive
and non-conductive samples.35
AFM can also operate in an ambient environment. The
probe in AFM is also known as a cantilever, which is essentially a probe with an
extremely sharp tip scanning across the surface to produce a three dimensional surface
Chapter 2 Experimental techniques
38
Fig. 2.6 Schematic of AFM.36
map. When the tip is brought close to the sample surface, the cantilever may be deflected
according to surface changes. The deflection of the laser off the back of the cantilever is
monitored by the photodiode as seen in the schematic diagram of AFM shown in Fig. 2.6.
The diameter of the tip is less than 40 nm and usually made from silicon or silicon
nitride.36
AFM can be operated in a few modes including contact and tapping mode. For
our thin films sample, tapping mode is used. In tapping mode, the tip only contacts the
sample intermittently, for a much shorter time as compared to contact mode. The
cantilever in tapping mode oscillates at its resonant frequency and minimizes the chances
of damage especially to soft samples.36
Chapter 2 Experimental techniques
39
2.3.6 Field emission testing (FE)
1) Parallel Plate Setup
Electron emission, which is a quantum tunneling effect, was tested with a self-
made electron emission tester. The setup used was in the form of a parallel plate
configuration with the cathode being the sample and the anode was an ITO-coated glass
slide. The cathode and anode are separated by a 100 µm polymer spacer layer with a cut
hole in the center of the film to allow electron emission only from that area. During
testing, the voltage was incrementally increased until significant current was produced.
Turn-on field with this setup is defined as the voltage needed to produce a current density
of 10 µA/cm2. The readings were taken with a LabView 8 software Keithley 2410
voltage source and current measurement unit. The measurements were taken under
vacuum conditions at a pressure of 1×10-6
Torr maintained by a turbomolecular pump. A
schematic cross-sectional diagram of the setup is shown below in Fig. 2.7.
Fig. 2.7 Parallel plate electron emission tester setup.
e-
Chapter 2 Experimental techniques
40
2) Probe tip setup
The difference between the probe tip and parallel plate setup is that in the former,
the anode is a tip (tungsten in our case) and the cathode-anode distance can be varied
from 0–100 µm. The self-made setup was specially fitted with a micrometer screw
gauges outside the vacuum chamber that allowed adjustments in the X-Y-Z axis without
breaking the vacuum. Turn-on voltage with this setup is defined as the voltage needed to
produce a current of 1 nA. The self-made probe tip emission setup used in our
experiments is shown in Fig. 2.8.
Fig. 2.8 Probe tip electron emission tester setup.
Chapter 2 Experimental techniques
41
References
1. P. D. Davidse, Vacuum 17, 139-145 (1967).
2. J. L. Vossen, J.Vac. Sci. Technol. 8, S12-S30 (1971).
3. P. J. Kelly and R. D. Arnell, Vacuum 56, 159-172 (2000).
4. D. Song, Appl. Surf.Sci. 254, 4171-4178 (2008).
5. N. Benchikh, F. Garrelie, C. Donnet, B. Bouchet-Fabre, K. Wolski, F. Rogemond, A.
S. Loir and J. L. Subtil, Thin Solid Films 482, 287 (2005).
6. F. Bonaccorso, C. Bongiorno, B. Fazio, P. G. Gucciardi, O. M. Maragò, A. Morone
and C. Spinella, Appl. Surf.Sci. 254, 1260-1263 (2007).
7. R. Janmohamed, J. J. Steele, C. Scurtescu and Y. Y. Tsui., Appl. Surf.Sci. 253, 7964
(2007).
8. A. H. Jayatissa, F. Sato, N. Saito, Y. Hirano and K. Takizawa, Carbon 38, 1145
(2000).
9. A. A. Voevodin and M. S. Donley, Surf. Coat. Techol. 82, 199 (1996).
10. P. Murray and D. Peeler, J. Electron. Mater. 23, 855-859 (1994).
11. H.-U. Krebs, M. Weisheit, J. Faupel, E. Süske, T. Scharf, C. Fuhse, M. Störmer, K.
Sturm, M. Seibt, H. Kijewski, D. Nelke, E. Panchenko and M. Buback, edited by B.
Kramer (Springer Berlin / Heidelberg, 2003), Vol. 43, pp. 101-107.
12. H. M. Christen and G. Eres, J. Phys. Condens. Matter 20, 264005 (2008).
13. F. Xiong, Y. Y. Wang, V. Leppert and R. P. H. Chang, J. Mater. Res. 8, 2265-2272
(1993).
14. J. J. Gaumet, A. Wakisaka, Y. Shimizu and Y. Tamori, J. Chem. Soc., Faraday Trans.
89, 1667-1670 (1993).
15. A. T. T. Koh, J. Hsieh and D. H. C. Chua, Diam. Relat. Mater. 19, 637-642 (2010).
Chapter 2 Experimental techniques
42
16. G. Reisse, B. Keiper, S. Weissmantel and U. Falke, Appl. Surf.Sci. 127–129, 500-506
(1998).
17. F. Garrelie, A. S. Loir, C. Donnet, F. Rogemond, R. Le Harzic, M. Belin, E.
Audouard and P. Laporte, Surf. Coat. Techol. 163–164, 306-312 (2003).
18. N. Benchikh, F. Garrelie, C. Donnet, B. Bouchet-Fabre, K. Wolski, F. Rogemond, A.
S. Loir and J. L. Subtil, Thin Solid Films 482, 287-292 (2005).
19. F. Qian, V. Craciun, R. K. Singh, S. D. Dutta and P. P. Pronko, J. Appl. Phys. 86,
2281-2290 (1999).
20. P. R. Willmott and J. R. Huber, Rev. Mod. Phys. 72, 315 (2000).
21. W. Chen, L. Yan and P. R. Bangal, Carbon 48, 1146-1152 (2010).
22. K. Y. Sasaki and J. B. Talbot, Adv. Mater. 11, 91-105 (1999).
23. B. J. C. Thomas, A. R. Boccaccini and M. S. P. Shaffer, J. Am. Ceram. Soc. 88, 980-
982 (2005).
24. A. R. Boccaccini, J. Cho, J. A. Roether, B. J. C. Thomas, E. Jane Minay and M. S. P.
Shaffer, Carbon 44, 3149-3160 (2006).
25. P. J. Goodhew, F. J. Humphreys and R. Beanland, Electron Microscopy and Analysis.
(Taylor & Francis, UK, 2001).
26. A. McPherson, A. J. Malkin, Y. G. Kuznetsov and M. Plomp, Acta Crystallogr. Sect.
D 57, 1053-1060 (2001).
27. J. Goldstein, Scanning Electron Microscopy and X-Ray Microanalysis. (Kluwer
Academic/Plenum Publishers, New York, 2003).
28. X. F. Zhang and Z. Zhang, Progress in Transmission Electron Microscopy. (Springer,
New York, 2001).
Chapter 2 Experimental techniques
43
29. J. R. Ferraro, K. Nakamoto and C. W. Brown, Introductory Raman Spectroscopy.
(Academic Press, Amsterdam, 2003).
30. Y. Wang, D. C. Alsmeyer and R. L. McCreery, Chem. Mater. 2, 557-563 (1990).
31. E. Smith and G. Dent, Modern Raman Spectroscopy: A Practical Approach. (John
Wiley & Sons, Chichester, 2005).
32. P. E. J. Flewitt and R. K. Wild, Physical Methods for Materials Characterisation.
(Institute of Physics Pub., Bristol, 2003).
33. A. J. Garratt-Reed and D. C. Bell, Energy-dispersive X-ray Analysis in the Electron
Microscope. (BIOS, Oxford, 2003).
34. H. S. Nalwa, Advances in Surface Science. (Academic Press, San Diego, 2001).
35. D. G. Seiler, Characterization and metrology for ULSI technology: 2003
International Conference on Characterization and Metrology for ULSI Technology,
Austin, Texas, 24-28 March 2003. (American Institute of Physics, USA, 2003).
36. Uppsala University, Department of Pharmacy,
(http://www.farmfak.uu.se/farm/farmfyskem-web/instrumentation/afm.shtml, 2012)
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
44
Chapter 3
Fabrication of nano-graphite in amorphous
carbon (a-C) matrix field emitters
In this chapter, carbon was deposited on Si substrates at various temperatures
using the PLD technique. The effect of temperature on the microstructure, surface and
field emission properties of the resulted films were studied. By relating the field emission
characteristics and the microstructure of the films, the emission mechanism of
nanographite in a-C film was investigated.
3.1 Introduction
Carbon materials for field emission applications have been a subject of great
interest for decades. The various polymorphic forms of carbon such as carbon nanotubes
(CNT), diamond and amorphous carbon (a-C) and diamond-like carbon (DLC) have been
demonstrated as potential field emitters.1-3
On one hand, while it is rather clear that the good emission properties of CNT
stems out from its high aspect ratio morphology and good conductivity from orientated
sp2 bonded graphene sheets, electron emission from planar amorphous carbon films on
the other hand, tends to be a topic of much debate. CNT has been favored for its very low
turn-on fields, in fact the lowest reported for the carbon family, but it suffers from
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
45
stability issues and tends to be burnt out easily.4 As the focus shifts towards flat emitters
which are mechanically stable, easily deposited on a variety of substrates and possess a
low work function (4–5 eV), planar amorphous carbon films, whether doped, undoped,
hydrogenated (a-C:H) or unhydrogenated, has been given a fair share of attention ever
since it demonstrated potential to be a cold cathode material during the 1990s.
Over the years, various models have been proposed for electron emission from
amorphous carbon. Several popular and widely accepted models include the Amaratunga
and Silva model5 on nitrogen doped a-C:H films, which has been used as a reference for
other amorphous carbon films. Nitrogen doped a-C:H was a popular choice of doped
films due to nitrogen being a shallow donor, thus allowing for the easy formation of a
space-charged layer that leads to sharper band bending. In this model, the n++
silicon
substrate is the true cathode, while the nitrogen doped a-C:H film is regarded as a space
charged layer through which electrons coming from the substrate is accelerated. These
electrons, termed “hot electrons”, subsequently acquire energies near to the vacuum level
and are thus able to emit at lower turn-on voltages. While models that involve these space
charged controlled mechanism may very well explain nitrogen doped a-C:H films, they
may not be apt for undoped nanostructured a-C or DLC films which are non-
hydrogenated. It was also briefly suggested that DLC films with the widest bandgap and
highest sp3 would be best for field emission.
6 This was later disproved in an in-depth
study which led to the consensus that the sp2 phase in a-C films plays an important role in
the field emission process.5 The conductive sp
2 elements facilitate conduction through the
film by forming electron paths or channels to deliver electrons to the surface to be
emitted. Sp2 bonded carbon atoms in planar films can come in the form of short olefinic
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
46
chains as in DLC (low sp2) to nanoclustered graphite in a-C matrix (intermediate sp
2
content) and finally fully sp2 film made of graphite sheets in the case of graphene or
nanographite. In the case of flat films in which the field enhancement factor () is unity,
it was documented that a field in the order of 500 V/µm is required for emission,7
However, it only takes less than a tenth of that value for a-C films.
Since DLC is known for its smoothness in the range of root mean square (RMS)
0.1 nm, there have been doubts about its field enhancing ability. In several studies of
DLC, β has been given the value of 1 which translates to no enhancement and lowered
emission turn-on fields are a result of lowered work function or electron affinity values.8,9
In addition, there have been various reports on improved emission with doped DLC films
with silver10
, nickel11, 12
or titanium11
which formed conductive metal nanoclusters. The
clusters in turn formed conductive paths within the largely insulating diamond-like
matrix and led to relatively lower emission turn-on voltages. However, these electron
channels have not been directly proven or seen.
In this chapter, it was experimentally showed that the electron channels present in
a-C films gave it a CNT-like or tube-like emission behavior when the emission testing
distance was varied. Changes in emission characteristics with deposition temperature
were also studied. It is worth noting that no conditioning was done on the films as there
have been doubts about the nature of emission from a-C films whether it is intrinsic or
caused by surface and/or internal damage due to a conditioning process which involves
the ramping of voltages in cycles from low to high and vice versa prior to testing. 7, 13
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
47
3.2 Experimental method
The films were deposited using a carbon target (99.9% purity, 325 mesh) with a
KrF (λ = 248 nm) Lambda Physik excimer pulsed laser system with a laser fluence of 10
J/cm2, pulse duration of 25 ns and frequency of 20 Hz. The target was rotated with a
speed of 60 rpm and fixed at a distance of 5 cm from the highly n doped (n++
) Si (100)
substrate. The sp2 content was controlled by varying the temperature from 100 to 700
oC
in increasing steps of 100 oC using a heater stage in vacuum. The base pressure was kept
below 5 × 10-6
Torr during deposition.
Raman spectra were recorded at room temperature with a Renishaw system using
an excitation source of 514.5 nm lines from Ar+ laser. The spectra were acquired in the
range of 1000–2000 cm-1
and accumulated for 10 times. The laser power used was 25
mW to avoid sample damage. Surface morphology and surface roughness were obtained
with a Digital Instruments, Nanoscope III, Multimode AFM operated in tapping mode.
AFM images acquired were flattened using the software provided by the manufacturer
and not modified by any filtering. The RMS roughness was evaluated over an area of 1 ×
1 m. XPS was carried out with a Kratos axis ultra DLD system monochromatic XPS (Al
Kα at 1486.6 eV source). The XPS spectrometer was operated at 225 W to obtain the
spectra of the samples. By deconvoluting the carbon peaks, chemical bonding and sp2
content of the films were determined. The spectra were obtained in a concentric
hemispherical analyzer with a pass energy (Ep) of 20 eV in narrow scan mode. A JEOL
JEM 3010F HRTEM (LaB6) operated at 300 kV was used for TEM imaging to observe
the nanoclusters directly.
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
48
Field emission current versus anode voltage (I–V) was measured with a probe tip
setup. The anode was an electrochemically etched tungsten needle with a tip diameter of
200 µm, while the cathode was the sample. Testing was conducted in at a base pressure
of ~4 × 10-6
Pa. The zero-distance point was determined by first moving the anode with 7
V of applied voltage in contact with the cathode and subsequently moving the anode
away from the cathode to a point where the current diminishes to zero. On the same spot
of the sample, the distance between the anode tip and substrate was adjusted using a
micrometer stage with an accuracy of 0.5 µm. The anode-substrate distances were varied
from 1 to 100 µm without breaking vacuum. The turn-on voltage (Von), at each distance
was the voltage needed to produce a current of 1 nA and the current limit was set at 1 µA
to prevent damage to the anode.
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
49
3.3 Microstructural characterization of nano-graphite
clusters in a-C matrix
3.3.1 Effect of temperature on microstructure
The changes in microstructures and the degree of sp2 clustering in films with
increasing deposition temperature were examined using Raman Spectroscopy. As shown
in Fig. 3.1, a broad peak centered on 1560 cm-1
, which is the typical spectrum for
amorphous carbon, was observed for the room temperature and 100 oC samples, while
two distinct peaks were observed for 400 and 700 oC samples.
14, 15 These peaks can be
deconvoluted into two Gaussian peaks known as the disorder and graphitic peak, i.e. the
D and G peak, which corresponded to the breathing mode of the six fold aromatic rings
and the in-plane C–C stretching mode of sp2 hybridized carbon atoms, respectively.
Unlike the D peak which only occurs due to the hexagonal rings, the G peak occurs for
all sp2 sites inclusive of sp
2 chains and rings.
14 The full width half maximum (FWHM),
the intensities of D (ID) and G (IG) peaks, and the ID/IG ratio of all the samples are
summarized in Table 3.1. Different studies have shown that the ratio of D to G peak
intensity, i.e. the ID/IG ratio, can be used to estimate the sp3 bonding in a-C, in which a
higher ID/IG ratio corresponds to a lower fraction of sp3 bonding present in the films.
16, 17
It was observed that samples deposited at room temperature and those at lower
temperatures (100–300 oC) had lower ID/IG ratio and possessed a high percentage of sp
3
bonding, akin to diamond type bonding. The sp2 component here mainly exists in the
form of olefinic chains intertwined with a sp3 carbon network. As the temperature
increased, the ID/IG ratio of the samples increased and the separation between D and G
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
50
peaks increased. The changes in ID/IG ratio and D and G peak positions in relation with
temperature were illustrated in Fig. 3.2 and 3.3, respectively. The D peak became more
obvious and shifted towards a lower 1350 cm-1
and G peak shifted towards a higher 1600
cm-1
, which are peaks observed in a commercial graphite spectrum.12
Above 400 oC, a
high degree of aromatic clustering was observed from the distinctly separated D and G
peaks, which indicated that certain elements in the films were becoming graphite-like.
Fig. 3.1 Raman spectrum for films deposited at room temperature, 100, 400 and 700 C.
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
51
Table 3.1 D and G peaks positions, amplitude, relative intensity ratio and estimated sp2 content of
the films deposited at different temperatures.
Sample
D peak
position
(cm-1
)
D peak
amplitude
(ID)
G peak
position
(cm-1
)
G peak
amplitude
(IG)
ID/IG sp
2 content
(%)
0 1432 71 1574 180 0.39 32
100 1424 552 1574 861 0.64 49
200 1422 905 1575 956 0.95 69
300 1413 569 1578 558 1.02 74
400 1385 398 1576 369 1.08 78
500 1382 373 1580 340 1.10 79
600 1383 591 1583 526 1.12 81
700 1376 473 1585 411 1.15 83
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
52
Fig. 3.2 Raman ID/IG ratio and sp2 content changes of samples deposited at temperature from
100–700 oC.
Fig. 3.3 Raman D and G peak positions of samples deposited at temperature from 100–700 oC.
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
53
XPS was employed on all the samples in order to confirm the estimated sp2
percentages. The XPS spectra for selected samples are as shown in Fig. 3.4 and were
deconvoluted into 5 Gaussian peaks. The binding energy at 284.5, 285.2, 286.5 and
288.0 eV are corresponding to sp2, sp
3, C–contaminated and C=O, respectively.
18 By
calculating the area under the peak sp2/(sp
2+sp
3), the sp
2 content were determined and are
summarized in Table 3.1. From the ratios of the peak contributions in XPS and Raman,
the sp2 percentage calculations ranged from 32 to 83 % from room temperature to
deposition at 700 oC. Thus, this confirmed the effects of heating in graphitizing the sp
3
bonding of DLC.
Fig. 3.4 XPS spectra for films deposited at room temperature, 100, 400 and 700 C.
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
54
Fig. 3.5 TEM micrograph for films deposited at (a) 100 C (b) 400 C (c) 700 C. Inset
shows the low magnification area view and SAED of crystalline cluster with lattice spacing 0.33
nm corresponding to carbon.
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
55
The clusters in the samples can be directly observed in TEM images shown in Fig.
3.5. The TEM image of 100 ºC sample in Fig. 3.5(a) showed no discrete nanocrystalline
clusters. On the other hand, clusters were observed in Fig. 3.5(b) and 3.5(c), which
corresponded to the 400 and 700 ºC samples respectively. A higher cluster density was
also observed for the 700 ºC sample in lower magnification images as seen in figure
insert 3.5c(ii). These results related well with the Raman results, where increased
clustering was seen from the separated D and G peaks. A plot of the cluster density
versus temperature is shown in Fig. 3.6, where the density at 100, 400 and 700 oC is 0,
1.27 x 1011
, and 3.65 x 1011
clusters/cm2, respectively. In addition, selected area electron
diffraction (SAED) (inset 3.5c(i)) showed that the clusters are crystalline with a lattice
spacing of 0.333 nm, thus confirming that the clusters are indeed made up of graphite and
thus fully sp2 bonded. This is in great contrast to a largely amorphous carbon film
Fig. 3.6 Cluster density vs. temperature of 100, 400 and 700 C samples.
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
56
Fig. 3.7 AFM surface morphologies of (a) 100 C (b) 400 C (c) 700 C samples.
deposited with minimal external heating where only short olefinic sp2 chains are present
in such films as seen in Fig. 3.5(a) for the 100 oC sample. These short chains in the
carbon network cannot be directly observed from the images and thus appear to be
featureless with no lattice fringes (lines) when viewed under TEM as the ordering of the
atoms are random with no particular order.
AFM topological images in Fig. 3.7 show the surface morphology of the films.
There was a slight increase in RMS roughness when the temperature was increased with
the surface roughness values 0.25 and 0.35 and 0.39 nm for the 100, 400 and 700 ºC
samples, respectively.
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
57
3.3.2 Discussion
Experimental results showed that sp2 bonding increased with increasing
deposition temperature. Raman spectroscopy revealed that the ID/IG ratio also increased
with increasing deposition temperature and ID/IG ratio has been reported to be
proportional to the square of correlation length (La) or size of nanocluster.19
As such, the
increase in ID/IG ratio with increasing deposition temperatures corresponded to an
increase in La in diamond-like films and increase in cluster size in the case of
nanographite films, as evidenced by TEM.
This formation of clusters is a temperature driven phenomenon.19, 20
At room
temperature deposition, energetic carbon atoms that penetrate into the growing film on
the substrate possess poor mobility, and carbon adatom mobility is also reduced due to a
low substrate temperature. As a result, high sp3 films are formed by a subsurface growth
process otherwise known as subplantation.21
However, with an increase in temperature,
the conversion of randomly distributed sp2 and sp
3 a-C into ordered clusters and aromatic
rings occurs and is facilitated by the increase in thermal energy.22
With higher deposition
temperatures, the subplanted carbon atoms have improved mobility and can diffuse
within the growing film, hence aromatic cluster aggregation is enhanced.23
The formation
of sp2 bonding can be enhanced by 1) heating during deposition or 2) annealing after
deposition and growth is completed. The microstructure differences between the two do
not differ greatly, but the transition temperature from a highly sp3 to sp
2 film is
significantly lower for heating during deposition. This transition temperature from
amorphous DLC to nanographite clustered a-C films for post-deposition annealing occurs
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
58
near 800 oC
24 whereas in our case, the transition occurred below 400
oC. This lowered
transition temperature can be explained through the activation of infrequent events
between pulses of impinging carbon ions onto the substrate, during which a period of 1 ps
of elevated temperature or thermal spike occurred, as shown by Molecular Dynamics
simulation results.25
As such, the heating during deposition is a more effective method to
achieve nanostructuring through heat. It was also observed that through sufficient heating,
substantial clustering of sp2 occurs and a preferential direction may be adopted by the sp
2
phase, resulting in a parallel or perpendicular orientation with respect to the substrate.26
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
59
3.4 Field emission study of nanographite clusters in a-
C matrix
3.4.1 Field emission properties
In order to investigate the field emission properties of the films, I-V characteristic
curves were obtained at various distances. The I-V characteristics of the films deposited
at 100, 400 and 700 oC is shown in Fig. 3.8. The spread of the curves at various distances
reflected the ease of emission with the 700 oC sample being the easiest emitter and 100
oC being the most difficult. To illustrate the differences, Von at 50 µm for the 3 samples
were 1760, 1420 and 1110 V, respectively for the samples in increasing temperatures.
Fowler-Nordheim (F-N) plots were also obtained using data from a single anode-cathode
distance of 50 µm in Fig. 3.9, which showed a linear trend between ln (I/V2) and (1/V),
indicating electron emission was of a metallic nature. By taking ratio of the slopes of the
plots according to equation 1.4, we can obtain a relation as follows,
(3.1)
In the temperature range tested, the change in work function of the films tested
was 0.4 eV.27
As such, when taking ratio, the term φ1/φ2 can be approximated to 1. The
enhancement factors of such flat film emitters can be divided into two parts; the internal
enhancement (βin) and external geometrical enhancement factor (βex) and the total β is
composed of the product of the two.26
Factors that affect the external enhancement factor
1
22/3
2
1
2
1 )(
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
60
Fig. 3.8 I-V characteristics of films deposited at 100, 400 and 700 oC.
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
61
are geometrical in nature and are related to the physical dimensions of the emitter feature
such as a tube or needle. In this case, external enhancement is usually taken to be unity.
Internal field enhancement, on the other hand, is typically used for planar emitters whose
surfaces usually appear featureless. The combination of microstructure and composition
of such materials thus determines the internal field enhancement factor.26
In this work,
the surface morphology in AFM images showed a slight roughness increment when
deposition was carried out at higher temperatures. However, such small changes in
surface topology are insufficient to qualify as geometrical enhancement sites for efficient
field emission. The best example of enhancement due to physical or geometrical
structuring is CNT. They have a very high aspect ratio28
ranging from 20 to ~10000 in
extremely high aspect ratio29
CNT. In contrast, the aspect ratio of the protrusions found in
our samples was less than 0.1, hence geometrical enhancement was not considered for
such flat films. Thus, βex was taken to be unity and equation 3.1 becomes,
(3.2)
This meant that the ratio of the slopes corresponded to a ratio of internal enhancement
factors (βin) such that β100: β400: β700 = 1: 1.48: 1.90. This ratio is expected not to
change with different inter-electrode distances as it represents the difference between the
samples. However, the values of the individual β of the samples at different distances will
change.
Using the data obtained from varied anode-cathode distances, the Von
versus distance (Von-D) curves with an inverted y-axis was plotted to illustrate the
voltage drop in vacuum as in Fig. 3.10. In obtaining data for these plots, anode-cathode
1
2
2
1
i n
i n
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
62
Fig. 3.9 Fowler-Nordheim plots at 50 µm for films deposited at 100, 400 and 700 oC.
Fig. 3.10 Voltage-distance (V-D) curves for nanographite in a-C films deposited at 100, 400 and
700 oC.
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
63
distances were varied in vacuum from 1 to 100 µm on the same spot of the sample. By
inverting the axis, the graph is the potential profile of electrons in vacuum. A detailed
illustration of the graph conversion can be found in ref. 30.30
From this, the electric field
in vacuum during emission was obtained by determining the slopes (Λ) of the Von-D
curves and the majority of the voltage drop was seen in vacuum. The electric fields in
vacuum for samples deposited at 100, 400 and 700 oC were 33.1, 23.9 and 16.4 V/µm
respectively. It can also been seen that for the 700 oC sample, the bending of the vacuum
level was the least required for emission. Thus, at any tested distance the barrier for
emission was lowest for the 700 oC sample.
Subsequently, Wentzel–Kramers–Brillouin (WKB) approximation, which has been
used to describe the tunneling probability of a triangular barrier formed when the vacuum
level potential bends under a strong electric field, was used to obtain a relationship
between β and electric field in vacuum (Λ).31
In this approximation, the equation for
transmission coefficient is,
D= (3.3)
where is the barrier height, m is the electron effective mass, and Λ is the electric field
in vacuum, which was determined from the slopes of the Von-D curves. From the WKB
equation, the barrier height ratio was related to the electric field in the form β 1/ β 2 = Λ 2/
Λ1, identical to the relationship derived from the F-N equation. Calculating the barrier
height ratios then gave β100: β400: β700 = 1: 1.46: 2.02, agreeing well with the barrier
height ratios obtained independently from the F-N relationship.
2
3
3
28exp
he
m
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
64
From the typical I-V curves, the turn-on field, defined as the voltage divided by the
anode-cathode separation, was plotted as a function of the separation for samples
deposited at 100, 400 and 700 oC. The applied turn-on fields (Eappl) increased
exponentially with decreasing anode-cathode distance as seen from Fig. 3.11. In opposite,
with the increase in anode-cathode distance, Eappl for the samples decreased and settled to
a saturated value after the distance of about 10 µm. Such turn-on fields versus distance
line shapes are typically observed for CNT and other tip emitters, thus hinting a possible
tube-like emission mechanism from the nanostructured graphite films.
Fig. 3.11 Turn-on field versus anode-cathode distances of films deposited at 100, 400 and 700 oC.
Inset (a) shows typical line shapes with 2 different height approximations determined from
simulations in Ref. 37. Inset (b) shows line details at small distances.
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
65
3.4.2 Emission mechanism of nanographite clusters in a-C
matrix
The tube-like emission from atomic smooth DLC can be revealed by considering
the parameters and emission mechanism in materials. In field emission studies, emission
properties are commonly summarized to a parameter that serves as a measure of field
enhancing ability of the sample, otherwise known as the field enhancement factor, β.
However, comparison of β values should be made with caution as β depends strongly on
experimental setup and even more so on anode-cathode distances.32
As such, it is more
meaningful to present the series of β values as a ratio across our samples since anode-
cathode distance was varied. In addition, a comparison of β for samples in this work
against various carbon-based field emitters can be found in Table 3.2.
The process of field emission occurs in 3 steps; first, electrons are injected from
the back contact into the film. Second, the electrons will have to travel within the film
and lastly be transported to the surface where the electrons emit into vacuum in this last
step. To improve field emission properties, either of the steps should be enhanced, or
ideally all steps should be enhanced, either through use of different substrates, dopants or
surface treatments. For instance, using a semiconductor as substrate can allow only
electrons with sufficient energy to tunnel through the substrate-film heterojunction and
give a higher energy electron population transversing the film.33
Additionally, the
limiting step differs for various films such as for polymeric films like nitrogen doped a-
C:H films, electron injection is the crucial factor while for diamond-like carbon, the
barrier mainly lies at the film front. Sp2 bonded graphite is highly conductive while sp
3
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
66
Table 3.2 Comparison of field enhancement factors of various carbon-based emitters.
Sample Type of Anode Inter-electrode
distance
Field
enhancement
factor, β
Ref.
Nanographite in a-C
(100 oC)
200 μm etched W
needle 50 μm 160 This work
Nanographite in a-C
(400 oC)
200 μm etched W
needle 50 μm 236 This work
Nanographite in a-C
(700 oC)
200 μm etched W
needle 50 μm 303 This work
Oriented short sp2
planes dispersed in
amorphous matrix
2 mm dia. stainless
steel ball 20-40 μm 215-680 Shpilman et al.
26
N-containing DLC 4 mm dia. stainless
steel ball - 171 Gröning et al.
34
Ag Doped DLC ITO Parallel Plate 200 μm 2081–5876 Ahmed et al.10
Multi-wall CNT film 3 mm dia. stainless
steel cylinder 125 μm 1400 Bonard et al.
32
Single multi-wall
CNT
3 mm dia.stainless
steel cylinder 1 mm 30, 000 Bonard et al.
32
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
67
bonded diamond has negative electron affinity. Both of which are properties that can
facilitate field emission.
Conduction can occurs through 2 ways; i) directly through overlapping graphite
clusters or ii) by a hopping mechanism that occurs both between the nanographite
clusters and sp2 short chains in the a-C matrix. An increased amount of nanographite
clusters can increase the conductivity of the films as well as improve electron hopping,
which is the main conduction mechanism in a-C films.35
As such, the nanographite
clusters can form conductive channels through the matrix, transporting electrons to the
emission sites in vacuum. These conductive channels are estimated to be about 100–600
nm but physical features of such length have not been observed directly in literature. On
the surface, field lines will concentrate at the more conductive elements which are the
surface clusters. Dielectric differences between the sp2 and sp
3 elements would also mean
different local fields experienced, such that the emission can be seen to occur from freely
suspended conducting channels in vacuum (as the channels are encased in an insulating
matrix), mimicking the behavior of CNT. This emission enhancement mechanism for
diamond-like carbon and its related composites was first proposed by Gröning.36
This CNT or tube-like emission is further ascertained as we observed an
exponential increase in turn-on field when the anode-cathode distance was reduced below
10 um, while a saturated value was obtained for distances above 10 um as shown in Fig.
3.11. This corresponded to a variation of β as the anode-cathode distance changes in
order to preserve a constant local electric field (Eloc) at all distances. This is contrary to
previous beliefs that β does not change with distance and emission data for various
materials were compared across inconsistent distances.32
This decreasing Eappl trend with
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
68
distance has been observed only for CNT37
and iron tip emitters38
and the expected thin
film behavior line shape (dashed) is illustrated in inset 3.11(a). This proves that
nanoclustered graphite films may not be as “flat” in terms of field emission as we think it
might be. As such, a new understanding for field emission from graphitic planar films is
needed.
The shape of the plot in Fig. 3.11 can be explained by the empirical equation below,
VhD
Elocal
)(
1 (3.4)
where E is the field, V is the applied voltage, D is the distance between anode and
cathode and h is the height of the feature or conduction channel. A typical way of
defining distance is to start the measurement from the bottom of the substrate as
illustrated in inset 3.11(a) while it is in fact more appropriate to start the measurement
from the tip of the emitter feature. The curve obtained can be divided into 2 regions,
namely the near-anode region and far anode region. In the near anode region, the term
1/(D-h) dominates the equation and small changes in D can greatly affect the shape of the
plot, giving a sharp slope for small distances. When D is very much larger than h in the
far anode regions, the distance related term approximates to 1/D regardless of the changes
in h. This approximation agrees well with simulation results from CNT.37
It was also observed that for distances below 10 µm (Inset 3.11(b)), turn-on fields
for the 700 oC sample was larger than the 100
oC sample while the trend was reversed at
distances greater than 10 µm. This can be explained by considering the feature height and
dominating factors in the 2 regions. First, given that h700>h100, this would imply that 1/(D-
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
69
h700) >1/(D-h100) for the near anode region and this would result in a higher Eon for 700 oC
sample. Likewise, at the far anode region, feature height is no longer a dominating factor
and distance terms for both cases are 1/D. Thus, in this region, β dominates the equation
instead and given that the β700 is twice that of β100 as determined earlier, Eappl-700 is
smaller than Eappl-100 in accordance to equation 1.5.
3.5 Summary
In summary, in-situ heating during deposition led to the formation of
nanographite clusters within an a-C matrix. Sp2 content and aromatic ring formation
within the films increased with temperature. With the formation of nanostructures within
the film and sp2/sp
3 inhomogenities in the microstructure, field emission was enhanced.
Internal enhancement factor ratios were both determined with F-N relationship and WBK
approximation showing that the enhancement was doubled with increasing temperatures.
Through analysis of the turn-on field at various anode-cathode distances of field emission,
a CNT-like emission mechanism was observed for the nanographite clusters embedded a-
C films.
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
70
References
1. O. M. Küttel, O. Gröning, C. Emmenegger, L. Nilsson, E. Maillard, L. Diederich
and L. Schlapbach, Carbon 37, 745-752 (1999).
2. A. N. Obraztsov, A. P. Volkov and I. Pavlovsky, Diam. Relat. Mater. 9, 1190-
1195 (2000).
3. Y. Saito and S. Uemura, Carbon 38, 169-182 (2000).
4. S. Berber and D. Tománek, Phys. Rev. B 77, 073407 (2008).
5. G. A. J. Amaratunga and S. R. P. Silva, Appl. Phys. Lett. 68, 2529 (1996).
6. J. Robertson and M. Rutter, Diam. Relat. Mater. 7, 620-625 (1998).
7. J. Carey, R. Forrest and S. Silva, Appl. Phys. Lett. 78, 2339 (2001).
8. S. Kundoo, P. Saha and K. K. Chattopadhyay, Mater. Lett. 58, 3920 (2004).
9. Y. Umehara, S. Murai, Y. Koide and M. Murakami, Diam. Relat. Mater. 11, 1429
(2002).
10. S. F. Ahmed, M.-W. Moon and K.-R. Lee, Appl. Phys. Lett. 92, 193502 (2008).
11. A. T. T. Koh, J. Hsieh and D. H. C. Chua, Diam. Relat. Mater. 19, 637-642.
12. R. Narula, R. Panknin and S. Reich, Phys. Rev. B 82, 045418 (2010).
13. W. T. Diamond, J. Vac. Sci. & Technol. A 16, 707-719 (1998).
14. A. C. Ferrari, Diam. Relat. Mater. 11, 1053-1061 (2002).
15. F. Atchison, T. Brys, M. Daum, P. Fierlinger, A. Foelske, M. Gupta, R. Henneck,
S. Heule, M. Kasprzak, K. Kirch, R. Kötz, M. Kuzniak, T. Lippert, C. F. Meyer, F.
Nolting, A. Pichlmaier, D. Schneider, B. Schultrich, P. Siemroth and U.
Straumann, Diam. Relat. Mater. 16, 334-341 (2007).
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
71
16. S. Xu, B. K. Tay, H. S. Tan, L. Zhong, Y. Q. Tu, S. R. P. Silva and W. I. Milne, J.
Appl. Phys. 79, 7234 (1996).
17. C. Casiraghi, A. C. Ferrari and J. Robertson, Phys. Rev. B 72, 085401 (2005).
18. K. Wu, W. Chen, J. Hwang, H. Wei, C. Kou, C. Y. Lee, Y. L. Liu and H. Huang,
Appl. Phys. A. Mater. 95, 707-712 (2009).
19. A. Ilie, A. Ferrari, T. Yagi, S. Rodil, J. Robertson, E. Barborini and P. Milani, J.
Appl. Phys. 90, 2024 (2001).
20. E. Cappelli, S. Orlando, V. Morandi, M. Servidori and C. Scilletta, J. Phys.: Conf.
Ser. 59, 616 (2007).
21. Y. Lifshitz, Diam. Relat. Mater. 5, 388 (1996).
22. J. Robertson, Mater. Sci. Eng. R 37, 129 (2002).
23. B. K. Tay, X. Shi, H. S. Tan and D. H. C. Chua, Surf. Interface Anal. 28, 231
(1999).
24. S. Anders, J. Diaz, J. W. Ager, R. Y. Lo and D. B. Bogy, Appl. Phys. Lett. 71,
3367 (1997).
25. N. A. Marks, M. F. Cover and C. Kocer, Appl. Phys. Lett. 89, 131924 (2006).
26. Z. Shpilman, B. Philosoph, R. Kalish, S. Michaelson and A. Hoffman, Appl. Phys.
Lett. 89, 252114 (2006).
27. A. Ilie, A. C. Ferrari, T. Yagi and J. Robertson, Appl. Phys. Lett. 76, 2627 (2000).
28. M. Chhowalla, K. B. K. Teo, C. Ducati, N. L. Rupesinghe, G. A. J. Amaratunga,
A. C. Ferrari, D. Roy, J. Robertson and W. I. Milne, J. Appl. Phys. 90, 5308-5317
(2001).
Chapter 3 Fabrication of Nano-graphite in a-C matrix field emitters
72
29. D. N. Hutchison, N. B. Morrill, Q. Aten, B. W. Turner, B. D. Jensen, L. L.
Howell, R. R. Vanfleet and R. C. Davis, J. Microelectromech. S. 19, 75-82 (2010).
30. Y. Kudo, Y. Sato, T. Masuzawa, T. Yamada, I. Saito, T. Yoshino, W. J. Chun, S.
Yamasaki and K. Okano, J.Vac. Sci. Technol. B 28, 506-510 (2010).
31. Y. Kudo, T. Yamada, H. Yamaguchi, T. Masuzawa, I. Saito, S. Shikata, C. E.
Nebel and K. Okano, Jpn. J. Appl. Phys. 47, 8921 (2008).
32. J. Bonard, M. Croci, I. Arfaoui, O. Noury, D. Sarangi and A. Chātelain, Diam.
Relat. Mater. 11, 763-768 (2002).
33. S. Silva and J. Carey, Diam. Relat. Mater. 12, 151-158 (2003).
34. O. Groning, O. M. Kuttel, P. Groning and L. Schlapbach, App.Phys. Lett. 71,
2253 (1997).
35. J. D. Carey and S. R. P. Silva, Phys. Rev. B 70, 235417 (2004).
36. O. Gröning, O. M. Küttel, P. Gröning and L. Schlapbach, Appl. Surf. Sci. 111,
135-139 (1997).
37. R. C. Smith and S. R. P. Silva, J. Appl. Phys. 106, 014314 (2009).
38. R. Smith, R. Forrest, J. Carey, W. Hsu and S. Silva, Appl. Phys. Lett. 87, 013111
(2005).
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
73
Chapter 4
Pulsed laser fabrication of few-layer
graphene using metal substrates
In this chapter, pulsed laser deposition (PLD) was used to fabricate few-layer
graphene (FLG). The effects of cooling rate and laser energy on graphene growth on Ni
metal substrates were studied in section 4.2. In section 4.3, the feasibility of Ni, Co, Fe
and Cu metal as substrates for graphene growth with PLD was explored. The use of
different laser wavelength PLD systems was studied in relation to graphene growth in
section 4.4. Lastly, FLG field emitters were fabricated and studied in section 4.5.
4.1 Introduction
In recent years, increased attention has been directed towards exploring
alternative methods to fabricate graphene, in particular, graphene through the deposition
of carbon on metals. Metals like ruthenium1, platinum
2 and nickel
3, 4 have been used in
the synthesis of graphene. Using metals to catalyst the growth of graphene or growing
adsorb graphene on metal surfaces is the third most used method of graphene synthesis
after mechanical exfoliation and epitaxial growth on SiC. High temperature requirements
(1000–1600 oC), ultra-high vacuum condition (10
-9 Torr), as well as the difficulties in
removing graphene from chemically stable SiC often complicate the fabrication process
and escalate costs.5 For graphene on metal substrate, free-standing graphene can be
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
74
obtained by etching with acid and transferring the graphene onto another substrate of
choice and can thus be a more effective method of fabrication. The range of graphene
thicknesses obtained from graphene grown on metals is also smaller as compared to
methods such as mechanical exfoliation.5 Hence, growing graphene on metals presents
several advantages including reduced temperatures, less stringent vacuum conditions and
ease of transferring graphene onto other arbitrary substrates by dissolving the metal
layer.6
Currently there are two main techniques that make use of metal substrates for
graphene growth; they are (i) chemical vapor deposition (CVD) and (ii) segregation.
CVD graphene has been proven to be a successful technique for growing large area
graphene (~cm2) by using Ni
3, 4 and Cu
7 as the main catalysts and hydrocarbon gas
sources. For instance, Reina et al.4 reported the fabrication of 1–12 graphene layers over
large area on Ni substrates using atmospheric pressure CVD. Li et al.7 also fabricated
large area graphene on Cu using CVD with methane and hydrogen gas source. In addition,
it was observed that the thickness of graphene layers did not increase with the thickness
of Cu as the growth process was self-limiting.8 Although CVD graphene is very
successful on Cu metals, the self-limiting formation of monolayer graphene on Cu may
be a deterrent for certain electronic applications. This is because monolayer graphene
lacks a bandgap, which is required for on/off switching in devices.9 Conversely, FLG is a
2D semimetal, which has been used to fabricate devices such as field effect transistors
(FET). 10
Studies have also suggested that FLG may be better for graphene FET as it
gives better manufacturing reliability.11
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
75
For segregation, the carbon source can be in the form of a hydrocarbon gas or in
the form of a solid-state carbon target, with the former being an unlimited source of
carbon and the latter being a limited source. Generally, this method makes use of the
carbon solubility of metal where carbon is absorbed by the metal at a higher temperature.
When the temperature is reduced, the carbon solubility is exceeded and precipitation of
the excess carbon occurs at the surface of the metal. However, it should be noted that
carbon uptake by the metals occurs at temperatures above 600 oC, thus this is the
minimum temperature at which such methods can succeed.12
Single or few-layer growth
can be achieved by controlling the cooling profile during segregation. Previously, Yu et
al.13
have fabricated graphene on Ni by segregation using a CH4: H2: Ar gas source at
1000 ºC.
Methods that make use of gaseous carbon sources whether in CVD or segregation
techniques, require the Ni-catalyzed decomposition of the hydrocarbon gas source and
subsequent diffusion into the metal, thus needing a longer time (>10 times) for the carbon
dissolution process. Here, the uptake of carbon by the metal can only be increased by an
increase in temperature. In this chapter, FLG was fabricated using pulsed laser ablation of
solid carbon targets at a relatively lower temperature of 750 oC, whereby the carbon
source was limited to that supplied within 1.5 minutes of target ablation (~7 nm of
carbon). PLD has been a popular method for fabricating a-C materials such as DLC and
its related nanostructured composites.14
This is because PLD provides carbon species that
possess energy up to a few hundred eV15
and particularly the energy of carbon species for
short wavelength lasers fall within the favored energy range for carbon sub-
implantation,16
a mechanism required for DLC formation.
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
76
From the point of view of segregation, the charged energetic carbon species
generated in the PLD plume has an added advantage of reduced carbon uptake duration.
The time taken for the absorption of carbon atoms into the metal layer is reduced for PLD
deposited carbon as solid solution formation of carbon in metal is facilitated by the
charged carbon ions that are able to penetrate the surface i.e. implantation of carbon ions.
Neutral carbon atoms on the other hand are unable to penetrate the surface and will be
deposited on the surface and absorbed by diffusion thereafter. In the case of solid-state
carbon segregation, the metal substrate was solely used as a medium to absorb incoming
carbon, instead of a catalyst to break down hydrocarbon gases.
Graphene was first fabricated using PLD in 2010 by Zhang et al17
with deposition
temperatures ranging from 1000 to 1300 oC and subsequently a study on FLG at reduced
temperature using PLD was carried out by Koh et al.18
While extensive research has been
carried out with CVD or segregation with gaseous carbon sources, extremely limited
information and studies are available on the parameters that affect segregation using
energetic carbon sources (e.g. from PLD) and solid-state metal-carbon diffusion methods.
Factors such as temperature, cooling rate, amount of carbon and laser energy (when using
a pulsed laser system) has to be controlled well in accordance with the type of method
used to achieve the desired results.
In section 4.2, changes in the quality and thickness of graphene formed on Ni
substrates using different cooling rates as well as different energies of the pulsed laser
beam were studied. Subsequently, in section 4.3, the viability of various metals to form
FLG from a solid carbon source provided by PLD was investigated. The metals chosen as
substrates were Ni, Cu, Co and Fe and the lattice constants of graphene, Ni, Cu, Co and
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
77
Fe are 0.357, 0.352, 0.361, 0.251 and 0.287 nm respectively, with Ni and Cu having the
smallest lattice mismatches. Ni has been largely successful in fabricating graphene in
several techniques ranging from common methods such as CVD and segregation, to less
popular methods such as graphene from metal melts.19
It was also interesting to find out
if the self-limiting behavior of Cu can be extended to other forms of carbon sources
besides gaseous sources used in CVD. Co and Fe are common catalysts for CNT
fabrication and Co has the highest solubility for carbon among the metals tested.
Although epitaxial growth of graphene on single crystal Pt,2 Ir and Ru
1 is well known, it
was neither included nor the focus of this study as the use of such substrates would
increase fabrication costs tremendously.
In section 4.4, the effects of using different laser sources were investigated. The
two lasers compared were a shorter wavelength excimer KrF laser (248 nm) and a longer
wavelength Nd:YAG laser (266 nm). Different laser wavelengths are known to produce
carbon films with different amounts of sp2 and sp
3 bonding and thus a different
microstructure entirely. Using experimental results and theoretical calculations, the
microstructure differences of segregated graphene using these two different lasers, as
well as the reasons behind the observed phenomenon were studied.
It has been established in earlier chapters that graphene possesses the ideal
combination of electronic, mechanical and optical properties as robust materials for
electronic applications such as field emission or even transparent emitters. Materials used
as cold cathode emitters are required to exhibit field enhancing effects whether internal or
external and the absence of such features can lead to extremely high turn-on fields (Eon)
or even no possible electron emission at all. The high aspect ratio (area to thickness ratio)
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
78
of graphene could potentially give dramatic field enhancing abilities when combined with
its good electron transport within the graphene sheet. However, many fabrication
methods produce graphene sheets that are morphologically flat or planar as graphene
growth proceeds laterally unlike in the case of CNT where 100 % vertical alignment can
be easily obtainable. Planar surfaces with no field enhancement can require turn-on
electric fields of 500–1000 V/µm20, 21
and such high electric fields are undesirable as it
may cause vacuum breakdown or electrical discharge.22
Thus, numerous research efforts
have been put into achieving vertically aligned graphene through mainly three methods; 1)
control the CVD growth process to get vertical petal-like FLG nanowalls,23-25
2) transfer
planar graphene sheets grown on metal substrates and process them thereafter to obtain
the desired alignment on a different substrate24, 26
and 3) use chemical processes to obtain
large quantities of free-standing exfoliated graphene from graphite and deposit graphene
films with solution processing methods such as filtration,27 spin coating28
or screen
printing.29
However, graphene nanowalls (GNW) by CVD in method 1 leads to a high
density of closely packed GNW that can produce a certain degree of electron screening
effect, give a higher Eon, depending on the control of density during fabrication. Method
2 which uses polymer support layers to transfer planar graphene sheets grown on metals
onto other substrates, can lead to extra defects generated during the transfer and also
incomplete transfer leads to non-uniform emitter surfaces.7 The third method will be
covered in Chapter 5 and thus not discussed in detail in this chapter.
Besides aligning graphene, a pre-shaped substrate, in the form of sharp tip
structures can also be used to help graphene acquire the required protrusions for field
emission. For instance, Ye et al.30
has reported on the fabrication of an emitter made
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
79
from graphene oxide sheets draped onto nickel nanotip arrays such that the flat GO sheets
acquired sharp protrusions throughout and thereby effectively reducing the Eon. However,
the conductivity of GO is significantly lesser as compared to graphene and it is possible
that the authors used GO instead of graphene to demonstrate a large area emitter (35 mm2)
as it is relatively easier to obtain large pieces of free standing GO.
In order to preserve the conductivity of as-grown graphene, it is best to directly
align graphene during growth and thus avoid transfer methods altogether. This can be
achieved by growing graphene onto tip or cone-shaped substrates. However, due to the
high temperatures used for some methods (>1000 oC) to fabricate graphene, metal cones
will undergo significant diffusion and reflow during the heating process and lose their
shapes as a result. The method of decomposition of SiC to form graphene is also not
suitable as it is very difficult to fabricate SiC cones due to its extremely high hardness.31
With the recent knowledge and understanding gained in the area of solid-state
transformation of carbon to graphene using a PLD deposited carbon source at
comparatively reduced temperatures, fabrication of FLG covered Ni and Co metal
nanocones structures that can be used for field emission applications were demonstrated
in section 4.5. The FLG was grown directly on metal cones and this eliminated problems
encountered during the transfer of graphene using support layers that in the process may
result in additional graphene breakage, low surface coverage fraction and incomplete
transfer that can affect the electron transport properties of graphene. Field emission
studies were also carried out on the fabricated FLG on the metal nanocones and their
field enhancing mechanisms were studied.
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
80
4.2 Few-layer graphene from Nickel substrates
4.2.1 Experimental method
Ni films of about 600 nm thickness were sputtered onto highly doped (n++
) Si
substrates. Subsequently, the Ni thin film substrates were transferred to the KrF (λ=
248nm) Lambda Physik excimer PLD system, which was evacuated to 5 x 10-6
Torr. A
carbon target (99.9% purity) was ablated with a laser energy of 50 mJ, a pulse duration
of 25 ns and a frequency of 10 Hz. The laser spot size was 1 mm2 and the target was
rotated at 6 rpm with the laser ablating a circular outline of 2 cm in radius. Substrate
temperature was kept at 750 oC during ablation. Subsequently, the Ni substrates were
cooled with three different controlled cooling processes 1, 50 and 100 oC/min. In
comparing the effects of laser energy, the cooling rate used was that of 1 oC/min and
energies 50, 100, 200 mJ were used for the comparison.
A Reinshaw Raman spectroscopy with 514.2 nm green laser beam was used to
examine the microstructure of the fabricated carbon. The green laser at 50 mW was
focused onto the sample with an optical microscope at 50X magnification and the spot
size was 1 µm. Spectra were obtained in the range of 1000–3000 cm-1
. A JEOL JEM
2010F high resolution TEM (with FEG source) operated at 200 kV was used to obtain
bright field cross-section images of the sample so as to view the graphene layers on these
samples.
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
81
4.2.2 Cooling rate dependence of graphene growth
Visible Raman spectroscopy (514.5 nm) is a suitable technique to probe the
structure of sp2 rich carbon as visible excitation is highly sensitive towards sp
2 bonding
and its photons resonate with the π states.32
Figure 4.1 shows the Raman spectra of the
samples cooled at different rates. The spectra of graphene consist of 3 peaks at 1350 cm-1
,
1590 cm-1
and 2700 cm-1
. The D (disorder) peak at 1350 cm-1
depends on the breathing
mode of the six fold aromatic rings, while it is independent of the thickness of graphene
and is activated purely by disorder. The D peak is thus absent in perfectly crystalline
graphite. The G (graphitic) peak situated around 1590 cm-1
depends on the in-plane
stretching motion and occurs for all sp2 sites inclusive of sp
2 chains and rings. The peak
at 2700 cm-1
(at a wavenumber exactly double that of D peak) is the second order D peak,
Fig. 4.1 Raman spectra of samples cooled at different rates. The laser energy per pulse was 50 mJ.
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
82
which is generally referred to as the 2D peak, is present even in the absence of defect.
There is a significant difference in the shape of the 2D peak between graphite and
graphene. In graphite, the 2D peak is not symmetrical and has a shoulder peak at a
slightly lower wavenumber. However, for graphene, the 2D peak is symmetrical. The
non-symmetrical peak with a shoulder peak attached at lower wavenumber can be clearly
distinguished when the graphene layers increase to 5 and beyond.32
As such, looking at
the peaks in the Raman spectra, it can be deduced that there were less than 5 layers in the
FLG samples. The intensity of the Raman peaks also suggested that the graphene layers
were not single layers and a higher 2D to G peak ratio would mean fewer layers were
present. Monolayer graphene has I(2D)/I(G) ratio of about 2 and the intensity of G peak
is known to increase almost linearly as the number of layers increases.33
By this
approximation, the I(2D)/I(G) ratio for the 1 oC/min and 50
oC/min cooling rates were
about 0.30 and 0.25 respectively which translated to 3–4 layers, clearly showing features
of FLG.
The 3–4 layers of graphene of the 1 oC/min sample can also be directly observed
in the TEM image (Fig. 4.2(a)), in which the carbon lattice fringes on the surface of the
Ni films had a lattice spacing of 0.33 nm, confirming the presence of carbon. Figure 4.2(b)
shows the layered structure of the samples and the Ni layer was seen to react with Si to
form nickel silicide at the interface, reducing the Ni layer to ~ 400–480 nm. In using a
cooling rate of 50 oC/min, the D peak was observed to reduce such that the I(D)/I(G) ratio
was 0.75 compared to a 1.15 ratio seen in the 1 oC/min cooling rate sample. This
indicated that a lower amount of disorder or defects were present in the sample cooled
with 50 oC/min. The samples cooled at 50 and 100
oC/min were also largely delaminated
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
83
at the Ni film-Si substrate interface when cooling was completed (Fig. 4.3), while
samples cooled at 1 oC/min remained attached to the underlying Si substrate. The lattice
constants of the materials used are as follow; Ni = 0.352 nm, C = 0.356 nm and Si =
0.543 nm. Referring to the Raman spectra in Fig 4.1, stress and strain can affect the G
peak position.33-35
As a result of the compressive stress experienced by the layered
structure, a slightly blue shifted G peak is seen in the Raman spectra in Fig 4.1 for the
sample cooled at 1 oC/min initially while for the samples cooled at 50 and 100
oC/min,
stress was released through delamination.
The differences in the Raman spectrum can be explained in terms of the
deposition process by pulsed laser and cooling process thereafter. Firstly, when the laser
ablates the target, carbon atoms with a certain amount of energy are deposited onto the Ni
substrate. These carbon atoms can either be 1) deposited onto the surface of Ni before
diffusing into the Ni thin film due to the heat present or 2) be implanted slightly into
near-surface regions of the Ni film. As such, the majority of the carbon will be
concentrated at near-surface regions, while the rest of nickel remained largely carbon free.
When cooled at 50 oC/min, the solubility of carbon in Ni dropped rapidly and the amount
of carbon atoms that exceeded the solubility limit are precipitated onto the Ni surface. As
the temperature decreased, carbon atoms continue to precipitate out and heat was
maintained such that the carbon atoms had sufficient mobility to rearrange themselves to
form a more ordered crystalline structure. This less defective ordering was translated to a
lower I(D)/I (G) ratio in the Raman spectra.
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
84
Fig. 4.2 Cross-section TEM showing (a) the graphene layers above the Ni. (b) Ni/Nickel
Silicide/Si layered structure. Inset (c) is an overview optical image showing rather
uniform coverage.
Fig. 4.3 Cross-section SEM showing delamination at Ni film-Si substrate interface for
medium and fast cooling samples.
Si substrate
Ni film (with graphene)
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
85
For samples that were cooled at 1 oC/min, the temperature (750
oC) was almost
maintained right after laser ablation was stopped and close to equilibrium conditions were
experienced. As such, large amounts of carbon atoms remained in the Ni film and had
sufficient time to diffuse from near-surface regions deeper into the bulk of the Ni film
where a larger amount of carbon can be dissolved in Ni without reaching its solubility
limit. As such, a minimal amount of carbon segregated on the surface as the temperature
was decreased slowly. Owing to the competition between segregation and diffusion
deeper into the bulk Ni, the larger amount of un-precipitated carbon that remained in the
Ni film after slow cooling (1 oC/min) resulted in a higher D peak (Fig. 4.1) as compared
to the sample with medium cooling (50 oC/min), which had less carbon remaining in the
Ni film after segregation. No 2D peak was detected for sample cooled at 100 oC/min as
the large amount of carbon that was precipitated onto the surface lacked the mobility to
form crystalline carbon within a much shorter cooling time and instead a more disordered
form of carbon was formed.
4.2.3 Laser energy effects on graphene growth
It is well understood that the laser energy affects the penetration depth of atoms
onto the substrate and this affects the carbon precipitation process.36, 37
The Raman
spectra shown in Fig. 4.4 are samples fabricated with laser energies, 50, 100 and 200 mJ.
It can be seen that using 50 and 100 mJ yielded similar results, while using 200 mJ
produced no 2D peaks. In the pulsed laser process, a higher laser energy would give rise
to more energetic carbon atoms and subsequently lead to a deeper implantation of carbon
atoms into the Ni film bulk which has not exceeded the solubility limit. While 100 mJ
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
86
and 50 mJ ablations would result in a shallower implantation in regions nearer to the
surface, making the precipitation process easier for the amount of carbon that was
supplied.
Fig. 4.4 Raman spectra of samples in which different pulsed laser energies were used. Cooling
rate was kept constant.
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
87
4.3 Few-layer graphene segregation using other metal
substrates
4.3.1 Experimental method
Metal films of approximately 500–600 nm thickness were sputter-deposited onto
highly doped (n++
) Si substrates. The thin metal film substrates were subsequently
transferred to the PLD chamber. The laser energy used in this section was 50 mJ. The
other laser parameters and chamber conditions used remained the same as those detailed
in section 4.2.1. The metal substrates were then cooled under controlled cooling at an
initial rate of 1 oC/min until the temperature was 550
oC, followed by a faster cooling rate
of 20 oC/min to room temperature.
4.3.2 Segregation of graphene on different metal substrates
Figure 4.5 shows the Raman spectra of the various metal substrates that were used
during deposition. As observed, there were two peaks present at ~1350 and ~1590 cm-1
across all four metals. In the case of Ni, there was an additional peak at ~2700 cm-1
.
These peaks at 1350, 1590 and 2700 cm-1
corresponded to the disorder (D), graphitic (G)
and second-order disorder (2D) peak respectively.33
The G peak, which is a non-resonant
scattering process (does not involve real electronic states), occurs due to the in-plane
vibrations of the sp2 carbon whether in chain or ring form. This peak is caused by the
phonons near the Brillouin zone centre (otherwise known as Γ point) whose momenta are
small. The D and 2D peaks are due to double resonant scattering processes, which
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
88
involve two continuous electronic transitions between two real states with different
energies and momenta, and are caused by phonons near the K point of the Brillouin
zone.38
Its shape and intensity can be used to determine the presence of graphene or
graphite and the thickness of graphene respectively. Based on the Raman spectra, it
appeared that only Ni was successful in forming FLG while carbon on the other metal
substrates was more disordered even though there was some evidence of clustering.
The shape of the 2D peak was symmetrical and the absence of the shoulder peak
at lower wavenumber suggested that FLG was present.39
Based on the peak shape in our
Raman spectra, it can be seen that there were less than 5 layers in the FLG on Ni samples.
For a more detailed estimation from the I(2D)/I(G) ratio of ~0.30, the graphene on Ni
corresponded to 3–4 layers.7
TEM images of the graphene on Ni when the cooling conditions are (a) suitable and
(b) too fast are shown in Fig. 4.6(a) and (b) respectively. In Fig. 4.6(a), few-layer
graphene which was imaged as lattice fringes was seen on top of Ni and it agreed well
with the Raman results of 3–4 layers. Fringes on the surface of the Ni films had a lattice
spacing of 0.33 nm, thereby confirming the presence of carbon. On the other hand, Fig.
4.6(b) shows the carbon lattice fringes that appeared to be more disordered and were
oriented in various directions.
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
89
Fig. 4.5 Raman spectra of carbon deposited on various metal substrates. Few-layer
graphene was formed only on Ni metal substrates when cooled with 1 oC/min to 550
oC followed
by 20 oC/min.
Fig. 4.6 Cross-section TEM showing (a) the graphene layers above Ni (b) a more disordered form
of graphite formed when cooling conditions were too fast.
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
90
4.3.3 Effect of cooling rate when using different metals
Figure 4.7 shows the Raman spectra of Ni and Cu substrate samples that were
subjected to different cooling profiles after the deposition of carbon. For the 50 oC/min
cooling rate it was seen that only Ni was observed to be successful in forming graphene
as shown by the 2D peak it exhibited, whereas the absence of the 2D peak indicated that
there was no graphene on Cu. For the 100 oC/min cooling rate, both Ni and Cu failed to
yield any positive results for graphene. This showed that cooling rate is critical in the
formation of graphene or polycrystalline graphite.
Fig. 4.7 Comparison of Raman spectra between Cu and Ni samples in which different
cooling rates 50 (top) and 100 oC/min (bottom) were used.
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
91
Fig. 4.8 Comparison of FLG grown on Co (top) and Ni (bottom). Full coverage was obtained on
Co substrate, while areas with graphene on Ni are indicated by the black arrows (inset).
Few-layer graphene was also formed on Co when a faster cooling profile
(20 oC/min) was adjusted to cater to Co. The Raman spectrum of FLG on Co is shown in
Fig. 4.8 where D, G and 2D peaks were present. This was also confirmed by TEM images
in Fig 4.9 which showed 2–3 layers of FLG. Under the optical microscope, it was
observed that the FLG grown on Co gave a higher surface coverage as compared to the
FLG on Ni sample. The as-grown FLG on Co was also transferred to Si substrate by
dissolving the Co metal in dilute HCl and it was observed that the D peak in the Raman
spectrum decreased significantly (Fig. 4.10). This showed that a large part of the disorder
contributing to the D peak was attributed to the amorphous carbon that was embedded
deep within the metal substrate during the pulsed laser ablation and that the segregated
FLG was relatively free of defects by itself.
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
92
Fig. 4.9 Cross-section TEM image of graphene on top of Co showing 2-3 layers.
Fig. 4.10 Raman spectra of (a) as grown few-layer graphene on Co (top). (b) Transferred few-
layer graphene on Si, which has a significantly reduced D peak indicating relatively few defects
in the graphene (bottom).
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
93
This phenomenon was also observed on areas where the Ni and Co metals were
covered during carbon deposition using PLD. With the use of stainless steel clips, certain
areas of the Ni and Co metals were blocked from the forward directed carbon plume. It
was observed that areas with carbon directly deposited onto, gave Raman spectra that
were a superposition of disordered carbon and graphene peaks, while areas that were
covered by the clips gave graphene with significantly reduced defects as seen from the
Raman spectra of the respective areas in Fig. 4.11. Graphene was able to segregate on the
covered area by diffusion of the carbon atoms nearby. This confirmed that most of the D
peak counts belonged to the underlying carbon dissolved in metal.
Fig. 4.11 Raman spectra of areas covered and uncovered during carbon deposition and
segregation.
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
94
4.3.4 Discussion
Graphene on metal substrates can be fabricated by two methods. The first method
is by segregating dissolved carbon in metal onto the surface under controlled temperature
annealing after carbon deposition (e.g. with PLD) and the second method is by
decomposing a precursor gas such as ethylene or methane on the metal surface, as in the
case of CVD. The former is a physical process that involves properties such as the
solubility of carbon in metal while the latter is a chemical reaction that makes use of the
metal layer as a catalyst to break down the gaseous carbon source. Another difference
between a CVD process and a PLD segregation process lies in the carbon source. In CVD,
the carbon source is in the form of a gaseous gas mixture, which way surpasses the
amount of carbon needed to form FLG. Conversely, PLD uses a solid carbon target,
which limits the carbon source during segregation to that supplied during target ablation.
Lastly, the carbon supplied by PLD is an energetic source of carbon ions, which can
penetrate deeper into the metal layer rather than stay on the metal film surface.
Across the four metals tested, graphene was formed only on top of Ni but not on
Cu, Co, Fe. To better understand this, the solubilities of carbon in these metals should be
examined. The carbon solubilities in Ni, Cu, Co and Fe at about 700 ºC are 1, 0.05, 1.9
and 0.3 at.% respectively.40
The relatively high solubility of carbon in Ni at 700 ºC
supported the formation of graphene on Ni. On the other hand, given the extremely low
carbon solubility in Cu, the precipitation process could not produce enough carbon to
give crystalline graphene as insufficient carbon was absorbed into Cu. Instead, the as-
deposited carbon remained largely on top of Cu and as heating proceeded and the initially
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
95
amorphous carbon on top of Cu began to exhibit signs of clustering as observed by
separate D and G peaks. This is in stark contrast to copper’s success when using CVD,
which gave self-limiting growth of single layer graphene. This is due to the non-chemical
nature of the segregation method, which does not rely on the catalytic ability of metals
and depends solely on its carbon solubility. As such, a metal with sufficient carbon
solubility would be desired over its catalytic ability when PLD deposition followed by
segregation is employed.
It is however interesting to note that there was no positive result for graphene on
Co in spite of its high carbon solubility when the same cooling recipe (1 oC/min followed
by 20 oC/min) for Ni was used. To understand this phenomenon, the rates of diffusion of
carbon into Co and Ni are considered. The rate of diffusion is characterized by a constant
known as the diffusion coefficient, D, which can be expressed as:
)exp(
RT
QAD (4.1)
where A is cm2s
-1, Q is in kJmol
-1, R= 8.31441 Jmol
-1K
-1.
The A and Q values for Ni are 0.12 and 137.3 respectively while the A and Q
values for Co are 0.31 and 153.7, respectively.41
As the activation energy, Q, is in the
exponent factor, it dominates the equation, making diffusion of carbon in Co slower than
that of Ni. The diffusion coefficients of carbon at 1000 ºC in Ni and Co are calculated to
be 2.79 × 10-11
and 1.53 × 10-11
m2s
-1, respectively.
42 This implies that Co required a
different cooling recipe from Ni due to differences in their physico-chemical properties
and this can be explained through the formation process. During the cooling process,
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
96
there is a competition between two processes; a) carbon atoms diffusing in towards the
metal substrate closer to the Si side that contains low concentration of carbon and b) the
precipitation process that results as the carbon solubility limit in metal is exceeded.
In the case of Co, which has a higher carbon solubility and lower diffusion
coefficient, when the cooling profile is slower, carbon atoms can make use of the longer
time to diffuse further into the metal film. As a result, it is more difficult for near surface
carbon atoms to exceed the solubility limit when a higher temperature is constantly held
for an extended period of time since carbon with low D can have more time to
redistribute. Conversely, a faster cooling rate (at 20 oC/min from 750
oC to room
temperature), would maintain a carbon saturated near-surface region to allow for
adequate precipitation, which led to the formation of FLG (Fig. 4.12). On the other hand,
for Ni, which has almost half the carbon solubility compared to Co, can easily exceed the
solubility limit when the temperature gradually decreases. From the distinct D, G and 2D
peaks observed in the Raman spectra for Co, it was suggested that the fabrication FLG on
Co was crystalline or nanocrystalline. As such, the Tuinstra-Koenig (T-K) relation43
can
be used to provide a size estimate (La) of the graphitic domains where La= C
()[I(D)/I(G)]-1
and C(514.2nm) is ~ 4.4nm. The La value for FLG on Co was calculated
to be 19.5 nm. However, it should be noted that the T-K relation has a tendency to
underestimate the domain size due to the dominant effect of small crystallites and the
equation assumes a uniform domain size throughout the sample.44
Therefore, larger
domains should be present in the fabricated FLG.
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
97
Fig. 4.12 Illustration of carbon atom concentration throughout the Co film as a result of slow and
fast cooling profiles. Carbon atoms are diffused deeper into the Co film and harder to precipitate
when cooling is slow. During fast cooling, solubility is exceeded in the near-surface regions and
precipitation results.
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
98
4.4 Few-layer graphene segregation using different
pulsed laser wavelengths
4.4.1 Experimental method
Nickel films of approximately 500–600 nm thickness were sputter-deposited onto
highly doped (n++
) Si substrates. The thin metal film substrates were subsequently
transferred to the pulsed laser systems which were evacuated to 5 x 10-4
Pa prior to
deposition. The pulsed laser systems used were 1) krypton fluoride (KrF, λ= 248nm)
Lambda Physik excimer PLD system with a pulsed duration of 25 ns and 2) neodymium-
doped yttrium aluminum garnet (Nd:YAG, λ= 266nm) with a pulse duration of 10–18 ns.
The frequency and laser energy was 10 Hz and 50 mJ for both lasers. A carbon target
(99.9% purity) was ablated with the respective lasers. Substrate temperature was kept at
750 oC during ablation. Deposition was carried out to yield an estimated thickness of 7
nm of carbon film if it had been deposited on silicon. The metal substrates were then
cooled under controlled cooling at an initial rate of 4 oC/min until the temperature was
550 oC, followed by a faster cooling rate of 20
oC/min to room temperature, unless stated
otherwise.
4.4.2 Segregation with Nd:YAG and KrF pulsed laser
The Raman spectra of graphene segregated using carbon from the Nd: YAG laser
seen in Fig. 4.13 showed 3 distinctive peaks at ~1347 (D peak) and ~1582 cm-1
(G peak)
and ~2687 cm-1
(2D peak) which are peaks characteristic of graphene layers. The peak at
2900 cm-1
is a combination peak of G and D peak. Similarly, the three characteristic
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
99
peaks were also observed in the Raman spectra (Fig. 4.14) obtained from the graphene
segregated from carbon from the KrF laser under similar conditions. The D, G and 2D
peak for KrF fabricated sample was at 1338, 1577, 2698 cm-1
respectively. The D peak,
indicates the presence of defects or numerous graphene edges. As the shape and position
of the 2D peak depends on the electronic properties of the carbon structure, the
symmetrically shaped 2D peak (with no shoulder peak) suggested that the peak belonged
to graphene layers and not bulk graphite.
The position of the 2D peak was used to determine the thickness of the graphene
layers. The peak wavenumber is also known to increase with the number of layers.45
As
the Raman spectra of KrF showed a 2D peak at 2698 cm-1
as compared to 2689 cm-1
in
the Nd:YAG sample, the Nd:YAG sample can thus be deduced to consist of relatively
thinner graphene layers.46
The G peak, which is also known as the graphitic peak has
intensity that increases with the number of graphene layers while peak position
wavenumber decreases with thickness.45
The position of G peak in the KrF sample was
1576 cm-1
as compared to the 1582 cm-1
peak in Nd:YAG sample, which pointed to a
thinner graphene stack in the Nd:YAG fabricated graphene once again. Furthermore,
through the relative peak intensities of the 2D and G peak, the thickness of the samples
can be further verified. As a reference, the I(2D)/I(G) ratio of a single layer graphene
sample is 2.33, 47
From Fig.4.13, I(2D)/I(G) ratio is 1.57, which translated to ~2 layers of
graphene for the Nd:YAG fabricated sample. In addition, the I(2D)/I(G) ratio is 0.30 in
Fig. 4.14, which gave ~4 layers of graphene for the KrF fabricated sample. Raman
spectra in Fig 4.15 obtained from various spots on the sample deposited with Nd:YAG
laser showed that the graphene thickness at these spots were rather uniform within a
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
100
Fig. 4.13 Raman spectra of graphene on Ni segregated from carbon deposited by Nd:YAG laser
at 750 oC.
Fig. 4.14 Raman spectra of graphene on Ni segregated from carbon deposited by KrF laser at
750 oC.
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
101
range from 2–3 layers. However, it should be noted that not 100% coverage was observed
through optical microscope images as seen in Fig 4.15.
Additionally, segregation was also carried out using Nd:YAG laser at a different
cooling rate of 20 oC/min and it was observed that few-layer graphene was also formed at
this cooling rate as shown in Fig. 4.16. However, the resultant graphene was thicker as
seen from the lower I(2D)/I(G) ratio of 0.78. The faster cooling rate (20 oC/min) allowed
for more graphene layers to precipitate during cooling.
Fig. 4.15 Raman spectra of graphene on Ni segregated from carbon deposited by Nd:YAG laser
at 5 different spots. Optical images (right) show the spot that was scanned.
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
102
Fig. 4.16 Raman spectra of graphene on Ni segregated from carbon deposited by Nd:YAG laser
when cooling rate was adjusted to a faster rate of 20 oC/min.
4.4.3 Discussion
It is known that the type of laser (i.e. wavelength) and laser fluence (J/cm2) will
affect the characteristics of the ablated plume components such as energies of the ions
and type of ions and neutrals present within the plume.15
These will in turn affect the
microstructure and resultant type of film obtained.48
As such, it is imperative to
understand the type of lasers used in this experiment. On one hand, Nd:YAG is a solid-
state laser that makes use of a crystal as the laser source. The wavelength of such lasers
naturally occurs at 1064 nm. However, through the use of Q-switching, output powers
can be increased and pulsed duration reduced to the nanosecond range. The laser
frequency can thus be doubled, tripled or quadrupled to give 532, 355 and 266 nm
wavelengths respectively, which are higher harmonics of laser radiation.49
On the other
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
103
hand, KrF is a gas based laser that makes use of the reaction between krypton and
fluorine gases to emit a laser at wavelength 248 nm, which is in the near ultraviolet range.
The process of plasma formation from a target when a pulsed laser ablates the
target can be divided into three main steps. In the first step, the incident laser radiation is
absorbed and interacts with the surface atoms of the target. This electronic absorption is
followed by extremely rapid heating of the target surface and leads to the formation of a
plume that consists of positive carbon ions, electrons and neutral particles. When the
laser fluence used is high enough for the surface temperature to reach beyond the critical
temperature instantly (as is the case of this study), thermal evaporation by phase
explosion occurs. The density of excited atoms increases with the laser fluence to a point
where the target surface is no longer a bound solid but should be regarded as a dense gas
in which strong repulsive forces of the like-charged ions are present. This repulsive force
in combination with the evaporation of material leads to the expulsion of energetic
particles.50
A shorter wavelength laser, thus a higher photon energy, will have a lower
optical penetration depth and result in a smaller ablated area and hence give particles with
higher energy. For carbon target which has a high melting point and thermal conductivity,
droplet emission is reduced.15
In the second step, the incident laser beam interacts with the expelled target
material leading to the formation and expansion of an isothermal plume. As the expulsion
of target material through the initial interaction between incident laser and target occurs
in the range of picoseconds and the pulse duration of the laser is in the range of
nanoseconds, the plume will be irradiated by the incident laser beam unless a pulse
duration smaller than the picosecond range is used.51
Hence, this laser interaction with
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
104
the expelled plume is a significant process in the ablation process. At this stage,
absorption of the laser by the plume leads to the attenuation of the laser intensity incident
on the target and also leads to additional excitation, fragmentation and ionization of
plume species.
Lastly, in the third step, the plasma expands adiabatically in an anisotropic three
dimensional manner with an overall forward direction that is characteristic of pulsed laser
ablation.50, 52
The first two steps occur during the laser beam irradiation period (i.e. pulse
duration), while the third step occurs after the laser pulse irradiation ceases. The
microstructure of the resultant film depends on temperature of the substrate and energies
of the plume species which is dependent on the wavelength, pulse duration, intensity and
energy of the laser source.
In laser plume studies of Nd:YAG ablated carbon, it was observed using time-of-
flight mass spectroscopy data that the size of carbon clusters decreases with increasing
laser power for 1024 and 532 nm. But in the 266 nm laser ablated plume, carbon ion
clusters were below C7+ even when the laser power density was at the lowest. This
showed that photoelectronic excitations played an important role in shorter wavelength
ablation as large carbon clusters (Cn+ where n≥7) are dissociated photoelectronically by
absorbing the 266 nm radiation. The dominant carbon ion species in the carbon plume
ablated by 266 nm wavelength is C+
at a fluence of ≥ 5 Jcm-2
together with some C3+ and
small amounts of C5+.49, 53
In the laser plume studies of KrF laser, it was observed that C+,
C2+ and C3
+ ion species were present.
48, 53, 54
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
105
It has been reported that the ablation rate and energetics of the plume are affected
by the absorption of laser fluence by the target, which is dependent on the optical
properties of laser-target interactions, rather than the thermal properties of the target
material. As such, the energy absorbed by atoms from the incident laser pulse was
calculated for each laser system by dividing the laser fluence by the number of neutrals,
nn. As mentioned earlier, the laser pulse will interact with the expelled plume as well,
since the pulse duration is greater than the plume expansion duration. The absorption of
the laser by the plume subsequently occurs by an inverse bremsstrahlung process and its
absorption, αp can be estimated by,
kT
h
T
nZ i
p
exp11069.3
32/1
23
8 (4.2)
where Z, ni, T, h, k and υ are the average ion charge, ion density and temperature of the
plasma, Planck constant, Boltzmann constant and frequency of the laser light,
respectively.52, 55
The ion density, ni, in cm-3
can be approximated by using Saha
equation55
, which can be expressed as,
2/1
2/315 exp104.2
kT
UnTn i
ni (4.3)
where nn is the density of neutrals in cm-3
and Ui is the first ionization potential of carbon
atom which is 11.26 eV.56
The temperature was taken to be 3600 K, which is the
sublimation temperature of carbon. As αp is directly proportional to ni2, absorption mainly
occurs very near to the target surface where ni is high.
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
106
The laser energy absorbed by electronic absorption, Q, can be estimated by the
product of αp and fluence of the laser I, in J/cm-3
(equation 4.4). From these calculations,
the energy absorbed per carbon target atom due to the incident laser (first step of
ablation), E1, and the additional energy absorbed due to laser-plume interactions (second
step of ablation), E2, can be estimated by,57
IQ p (4.4)
From the estimations of energies absorbed by a carbon target atom/ion by the
absorption and interaction with different laser wavelengths in Table 4.1, it can be seen
that the energy of the carbon species in the Nd:YAG laser were ~1.26 eV lesser as
compared to the KrF ablated carbon species. Despite the differences in wavelength and
energies of depositing carbon species, both types of laser were viable for graphene
segregation. The lower energy of carbon species from the Nd:YAG laser could also give
a higher sp2 near surface carbon concentration that favored the formation of more
graphitic rather than amorphous carbon bonds. Thus, it can be seen that the D and G
peaks of the Nd:YAG graphene was distinctly separated while the KrF graphene spectra
gave a superposition of graphene and amorphous carbon peaks with a more conjoint D
and G peaks. In addition, it is expected that similar results will be obtained with the
Nd:YAG laser as with the KrF system with Ni, Co, Cu and Fe substrates. That is, Ni and
Co will be suitable but Cu and Fe will not be due to their poor carbon solubilities.
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
107
Table 4.1 Carbon ion energies for step 1 and 2 during laser ablation of carbon target for KrF and
Nd:YAG systems.
Laser Type E1
(eV/ion)
E2
(eV/ion)
KrF 56.34 2.72 × 10-5
Nd:YAG 55.08 3.36 × 10-5
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
108
4.5 Fabrication of few-layer graphene field emitters
using PLD
4.5.1 Experimental method
Ni and Co metal nanocones were fabricated from sputter-deposited metal films
using ion irradiation technique by bombardment with Ar+ ions using a Kaufman type ion
gun (ION TECH. INC. Ltd., model 3-1500-100FC). The incidence angle was 45 oC to the
surface, and irradiation was performed at room temperature for 7 min. The diameter and
the energy of the ion beam employed were 6 cm and 1000 eV, respectively. The basal
and working pressures on the ion beam chamber were 10−5
and 5 × 10−2
Pa, respectively.
Two types of Ni and Co metal nanocones were prepared for each metal, one with distinct
taller cones with some small cones in-between and one with a high density of uniform
shorter cones. From here forth, the two types of cones will be referred to as Type A and
Type B cones respectively. The prepared metal nanocones were then transferred to the
KrF excimer PLD chamber and graphene segregation was carried out at cooling rates
suitable for each metal according to the results obtained earlier and with the same laser
parameters as detailed in section 4.2 and 4.3. The metal nanocones were examined before
and after the graphene segregation using a Zeiss Supra 40 FEG SEM.
4.5.2 Microstructural characterization of cones
Figure 4.17 shows the surface morphology of the as-fabricated Type A (Fig.
4.17(a, c)) and Type B (Fig. 4.17(b, d)) Ni and Co metal cones. It can be seen that for
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
109
both Ni and Co metals, high density cones arrays resulted from the ion irradiation and
Type A consisted of tall cones separated by smaller cones while Type B consisted of
small cone tips with uniform height. For Ni nanocones, the heights of the cones range
from 400–700 nm for Type A and 200–300 nm for Type B. For Co, the sizes of the
nanocones range from 400–840 nm for Type A and 200–300 nm for Type B. The apex
radius of the cones were ~40 nm for the cones in Type A and slightly smaller than that
for the small cones in Type B.
The effects of heating on the metal nanocones were studied by carrying out the
heating and cooling process in the chamber without the deposition of carbon. The
corresponding SEM images are shown in Fig. 4.18. It can be seen that the heating process
did not destroy the conical structures entirely and the conical shapes were largely
maintained. After graphene was segregated using PLD-deposited carbon, the morphology
of the surface was examined again. Figure 4.19 shows the respective SEM images of both
Type A and B Ni and Co cones after carbon deposition and graphene segregation.
Comparing the images to the heat-only Type A cones, it was observed that the energetic
species did a small amount of additional damage to the cone structures in addition to the
heat effect. Thus, this made it more critical for the laser energy to be kept to a minimum
to not cause significant sputtering damage on the cones. The Raman spectra of Ni and Co
cones after segregation are shown in Fig. 4.20. The three characteristic peaks, D, G and
2D peaks confirmed the presence of FLG on the metal cones.
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
110
Fig. 4.17 SEM images of original as-fabricated Ni cones (a) Type A, (b) Type B and Co cones (c)
Type A, (d) Type B.
Fig. 4.18 SEM images of heat-only (a) Ni (b) Co metal nanocones.
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
111
Fig. 4.19 SEM images of (a) Ni (b) Co metal nanocones after graphene segregation.
Fig. 4.20 Raman spectra for Ni and Co metal cones.
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
112
4.5.3 Field emission study of graphene coated metal cones
The field emission I-V plots and F-N plots are shown in Fig. 4.21 and 4.22 for the
various cone samples. The turn-on voltage (Von) for Type A cones is 70 V and 105 V for
Co and Ni respectively. Also, the Von for Type B cones is 470 V and 580 V for Co and Ni
respectively. From the slopes of the F-N plots for the samples, β was estimated to be 275,
65 for Co and Ni Type A cones, 8 and 4 for Co and Ni Type B cones respectively, when
the work function of the material was assumed to be 5 eV.
Fig. 4.21 Field emission I-V plots for graphene on cone samples.
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
113
Fig. 4.22 Field emission F-N plots for graphene on cone samples.
4.5.4 Discussion
Spindt tips made of metals such as molybdenum (Mo) were the first cold cathode
material that was used for field emission displays. Electrons are extracted from these
sharp tips under a high electric field and used to bombard phosphor screens. However,
they were susceptible to poisoning by certain elements (e.g. sulfur) that are given off the
sulfur-based phosphor screen and ion erosion by ion bombardment.58, 59
This causes
blunting of the tips and field emission performance is compromised. As a result, Spindt
tips are plagued with stability and reliability issues that led to the rise of carbon-based
field emitters. Given that graphene has excellent mechanical properties and chemically
stability, growing FLG directly onto the metal cones can help to provide a layer of
mechanical and chemical protection for the metal cones.
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
114
The as-prepared Ni and Co cones did not reach the electron emission turn-on
current within the maximum limits of the voltage. But after graphene was segregated onto
the cones, field emission at a considerably lower voltage took place. This was attributed
to a change of the tunneling barrier at the interface between the cone tip surface and
vacuum. From first principles studies of graphene on Ni and Co substrates by
Khomyakov et al.,60
it has been shown that the chemisorption of graphene onto these
metals resulted in an electron transfer process from metal to graphene which n-doped
graphene. The work function of free-standing graphene and clean metal surfaces as
compared to the adsorbed graphene on metal systems are shown in Table 4.2. It can be
seen that the work function of pure metal cones was significantly reduced by 1.66–
1.81 eV after the adsorption of graphene. As the emission barrier height is affected by
work function of a material, a reduced emission barrier height is resulted.61
The shift of
EF has been illustrated in Fig. 4.23 where a schematic of the band structure of the metal
cone with and without a graphene coating is shown. Similar effects have been cited in the
cases of gold (Au)-graphene and aluminum (Al)-graphene. A higher reduction potential
(relative to standard hydrogen electrode) of Au at 1.5 V compared to 0.14 V of graphene
resulted in electron transfer from graphene to Au, while a negative reduction potential of
Al at -1.66 V compared to graphene led to electron transfer from Al to graphene.62
The
reduction potentials of Ni, Co and graphene are -0.25, -0.28 and 0.14 V respectively.62, 63
Hence, charge transfer from Ni and Co metals to graphene occurred, raising the Fermi
level, EF, of graphene and led to the reduction of work function (barrier height) at the
surface of the coated metal tips.
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
115
Table 4.2 Comparison of work function of free-standing graphene and clean metal surfaces with
the respective graphene on Ni or Co metal systems from ref. 60.
Work Function (eV) Graphene Ni Co
Free-standing 4.48 5.47 5.44
Graphene on metal system 3.66 3.78
Difference 1.81 1.66
Fig. 4.23 Schematic of work function shift when graphene is chemisorbed onto metal ( G-on-M)
and the position of EF in intrinsic versus doped graphene.
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
116
From the SEM images, the Type B Ni cones lost most of their definitive cone tip
outlines as compared to the Type B Co cones, where the small nanocones could still be
identified after graphene segregation. This was attributed to the slower cooling profile
used for the Ni cones (as mentioned in section 4.3), thus allowing a larger degree of
diffusion and reflow of the Ni conical structures to take place. Despite that, field emission
still occurred as the surface of the Ni remained rough, consisting of nano protrusions as a
reduced β factor. This proved that the change in tunneling barrier at the emitter surface by
applying a graphene coating onto the metal cones, instead of the physical aspect ratio, is
the dominant factor in reducing the turn-on voltages for the samples.
Type A samples consisted of nanocones that were of a range of heights, with a
mix of some very tall cones in the midst of some shorter cones, while Type B samples
consisted of high density nanocones that were of a relatively uniform height. This gave
the former an advantage of an improved field enhancement as a result of the spaced-out
tall cones. On the other hand, the numerous cones of the same height resulted in a large
degree of electron screening effect as evidenced by the large drop in β and higher field
emission turn-on voltage for these samples. In addition, when comparing the electrical
conductivities of Co and Ni, the significantly higher conductivity of Co at 17.9×106 Sm
-1,
as compared to 14.6×106 Sm
-1 for Ni also helped to provide more efficient electron
transport from the substrate to the surface, allowing for a lower turn on voltage.64
In
addition, a comparison of the turn-on voltages for the graphene on metal nanocones
samples with the nanographite in amorphous carbon samples in Chapter 3 is shown in
Table 4.3.
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
117
Table 4.3 Comparison of turn-on voltages of the graphene on metal cones samples with the
nanographite in amorphous carbon samples in Chapter 3.
Sample Von
(V)
Co Type A 70
Ni Type A 105
Co Type B 470
Ni Type B 580
Nano-graphite in a-C 420-500
4.6 Summary
In section 4.2, FLG was fabricated on Ni substrates using the PLD technique. The
effects of cooling rate and laser energy of the graphene product were studied with Raman
spectroscopy. It was observed that the cooling rate was critical in the production of
graphene, graphite (>10 layers) or a slightly more disordered carbon. On the other hand,
laser energy was less critical as long as the energy fell below 100 mJ. With a precise
control of temperature and amount of source carbon, the formation of graphene layers can
be controlled. The results have shown that it is possible to produce graphene using solid
sources and a physical deposition technique at relatively lower temperatures compared to
graphene epitaxial growth using SiC, as long as the temperature control and laser energy
are optimized.
Few-layer graphene was fabricated on Ni and Co metal substrates using PLD as
the required carbon source in section 4.3. Although graphene growth on Cu has been very
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
118
successful with a CVD process involving chemical reactions with precursors, the failure
to produce graphene when using a solid carbon source with Cu was because the catalytic
behavior of the metal was not utilized for such a setup. Instead, the metal’s solubility was
critical to the successful fabrication of graphene by segregation. When using a cooling
rate of 1 oC/min followed by 20
oC/min, FLG was only formed on Ni and not Co. With an
increase in cooling rate, largely defect-free FLG with 100 % substrate coverage was also
formed on Co. As such, besides carbon solubility of the substrate, the cooling recipe used
for different metals was also extremely important owing to the competing processes
during segregation.
In section 4.4, FLG was fabricated by both Nd:YAG and KrF pulsed lasers at
750 oC. The thickness of the resultant graphene layers was estimated from the relative
positions of characteristic Raman peaks of graphene as well the the relative I(2D)/I(G)
ratios. The corresponding ratio of graphene in the Nd:YAG sample was 1.57 which
translates to ~2 layers of graphene, while the ratio for the KrF sample was 0.30, which
gave ~4 layers of graphene. Through theoretical calculations, the carbon species in the
Nd:YAG laser were of a lower energy than the KrF laser. Even though the two laser
sources had different wavelengths and carbon ion energies, both laser systems were
capable of producing graphene when the same cooling profile was used.
Lastly, in section 4.5, cones of various density and heights from Ni and Co metals
were fabricated and the segregation technique studied in sections 4.2 and 4.3 was
subsequently applied to grow graphene directly onto the metal cones. The fabricated
graphene on metal cones were studied for their field emission properties. It was observed
that the mix of tall cones separated by small cones in-between, otherwise known as
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
119
Type A cones, provided a better field emission enhancement as compared to the array of
highly dense cones with uniform heights (i.e. Type B cones). By applying a chemisorbed
coating of graphene onto metals such as Ni and Co, field emission properties were
enhanced greatly by the reduction of tunneling barrier of the metal cones and lower turn-
on voltages were required for electron emission as a result.
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
120
References
1. P. W. Sutter, J.-I. Flege and E. A. Sutter, Nature Mater. 7, 406-411 (2008).
2. H. Ueta, M. Saida, C. Nakai, Y. Yamada, M. Sasaki and S. Yamamoto, Surf. Sci. 560,
183-190 (2004).
3. K. S. Kim, Y. Zhao, H. Jang, S. Y. Lee, J. M. Kim, K. S. Kim, J.-H. Ahn, P. Kim, J.-
Y. Choi and B. H. Hong, Nature 457, 706-710 (2009).
4. A. Reina, X. Jia, J. Ho, D. Nezich, H. Son, V. Bulovic, M. S. Dresselhaus and J.
Kong, Nano Lett. 9, 30-35 (2008).
5. J. Wintterlin and M.-L. Bocquet, Surf. Sci. 603, 1841-1852 (2009).
6. C. He, W. Wang, S. Deng, N. Xu, Z. Li, G. Chen and J. Peng, J. Phys. Chem. A 113,
7048-7053 (2009).
7. X. Li, W. Cai, J. An, S. Kim, J. Nah and D. Yang, Science 324, 1312 (2009).
8. C. Mattevi, H. Kim and M. Chhowalla, J. Mater. Chem. 21, 3324 (2011).
9. J. B. Oostinga, H. B. Heersche, X. Liu, A. F. Morpurgo and L. M. K. Vandersypen,
Nature Mater. 7, 151-157 (2008).
10. K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, Y. Zhang, S. V. Dubonos, I.
V. Grigorieva and A. A. Firsov, Science 306, 666-669 (2004).
11. Z. Jin, Y. Su, J. Chen, X. Liu and D. Wu, Appl. Phys. Lett. 95, 233110 (2009).
12. J. A. odr guez-Manzo, C. Pham-Huu and F. Banhart, ACS Nano 5, 1529-1534
(2011).
13. Q. Yu, J. Lian, S. Siriponglert, H. Li, Y. Chen and S. Pei, Appl. Phys. Lett. 93,
113103 (2008).
14. A. T. T. Koh, J. Hsieh and D. H. C. Chua, Diam. Relat. Mater. 19, 637-642 (2010).
15. A. A. Voevodin and M. S. Donley, Surf. Coat. Technol. 82, 199-213 (1996).
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
121
16. J. Robertson, Mater. Sci. Eng. R 37, 129-281 (2002).
17. H. Zhang and P. X. Feng, Carbon 48, 359-364 (2010).
18. A. T. T. Koh, Y. M. Foong and D. H. C. Chua, Appl. Phy. Lett. 97, 114102 (2010).
19. S. Amini, J. Garay, G. Liu, A. A. Balandin and R. Abbaschian, J. Appl. Phys. 108,
094321 (2010).
20. J. Carey, R. Forrest and S. Silva, Appl. Phys. Lett. 78, 2339 (2001).
21. S. Bellucci, A. Malesevic, F. Micciulla, I. Sacco, R. Kemps, A. Vanhulsel and C. Van
Haesendonck, Nanosci. Nanotechnol. Lett. 3, 907-912 (2011).
22. R. V. Latham, High Voltage Vacuum Insulation: Basic Concepts and Technological
Practice. (Academic Press, London, 1995).
23. A. Malesevic, R. Kemps, A. Vanhulsel, M. P. Chowdhury, A. Volodin and C. Van
Haesendonck, J. Appl. Phys. 104, 084301 (2008).
24. P. Hojati-Talemi and G. P. Simon, Carbon 49, 2875-2877 (2011).
25. H.-C. Hsu, C.-H. Wang, S. K. Nataraj, H.-C. Huang, H.-Y. Du, S.-T. Chang, L.-C.
Chen and K.-H. Chen, Diam. Relat. Mater. 25, 176-179 (2012).
26. S. Pandey, P. Rai, S. Patole, F. Gunes, G.-D. Kwon, J.-B. Yoo, P. Nikolaev and S.
Arepalli, Appl. Phy. Lett. 100, 043104 (2012).
27. K. Wang, T. Feng, M. Qian, H. Ding, Y. Chen and Z. Sun, Appl. Surf. Sci. 257,
5808-5812 (2011).
28. E. Stratakis, G. Eda, H. Yamaguchi, E. Kymakis, C. Fotakis and M. Chhowalla,
Nanoscale 4, 3069-3074 (2012).
29. Q. Min, F. Tao, D. Hui, L. Lifeng, L. Haibo, C. Yiwei and S. Zhuo, Nanotechnology
20, 425702 (2009).
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
122
30. D. Ye, S. Moussa, J. D. Ferguson, A. A. Baski and M. S. El-Shall, Nano Lett. 12,
1265-1268 (2012).
31. I. Yonenaga, Phys. B Condens. Matter 308–310, 1150-1152 (2001).
32. A. C. Ferrari, Solid State Commun. 143, 47 (2007).
33. D. Graf, F. Molitor, K. Ensslin, C. Stampfer, A. Jungen, C. Hierold and L. Wirtz,
Nano Lett. 7, 238-242 (2007).
34. X. Duan, H. Son, B. Gao, J. Zhang, T. Wu, G. G. Samsonidze, M. S. Dresselhaus, Z.
Liu and J. Kong, Nano Lett. 7, 2116-2121 (2007).
35. D. S. Lee, C. Riedl, B. Krauss, K. von Klitzing, U. Starke and J. H. Smet, Nano Lett.
8, 4320-4325 (2008).
36. C. Uzan-Saguy, C. Cytermann, R. Brener, V. Richter, M. Shaanan and R. Kalish,
Appl. Phy. Lett. 67, 1194 (1995).
37. J. W. Ager, S. Anders, A. Anders, B. Wei, X. Y. Yao, I. G. Brown, C. S. Bhatia and
K. Komvopoulos, Diam. Relat. Mater. 8, 451-456 (1999).
38. A. Obraztsov, E. Obraztsova, A. Zolotukhin and A. Tyurnina, J. Exp. Theor. Phys.
106, 569-574 (2008).
39. A. C. Ferrari, Solid State Commun. 143, 47 (2007).
40. W. F. Gale and T. C. Totemeier, (Butterworth-Heinemann, Great Britain, 1992), pp.
148- 154.
41. W. F. Gale and T. C. Totemeier, (Elsevier Butterworth-Heinemann, The Netherlands,
2004), pp. 19- 25.
42. A. Moisala, A. Nasibulin and E. Kauppinen, J. Phys.: Cond. Matt. 15, S3011-S3035
(2003).
43. F. Tuinstra and J. Koenig, J. Chem. Phys. 53, 1126 (1970).
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
123
44. A. C. Ferrari and J. Robertson, Phys. Rev. B 61, 14095 (2000).
45. Z. Ni, Y. Wang, T. Yu and Z. Shen, Nano Res. 1, 273-291 (2008).
46. S. J. Chae, F. Güneş, K. K. Kim, E. S. Kim, G. H. Han, S. M. Kim, H.-J. Shin, S.-M.
Yoon, J.-Y. Choi, M. H. Park, C. W. Yang, D. Pribat and Y. H. Lee, Adv. Mater. 21,
2328-2333 (2009).
47. S. He, K.-K. Liu, S. Su, J. Yan, X. Mao, D. Wang, Y. He, L.-J. Li, S. Song and C.
Fan, Anal. Chem. 84, 4622-4627 (2012).
48. K. Yamamoto, Y. Koga, S. Fujiwara and F. Kokai, Jpn. J. Appl. Phys. 36, L1333
(1997).
49. J. J. Gaumet, A. Wakisaka, Y. Shimizu and Y. Tamori, J. Chem. Soc., Faraday Trans.
89, 1667-1670 (1993).
50. K. J. Koivusaari, J. evoska and S. . vuori, J. Appl. hys. 85, 2915 (1999).
51. M. N. R. Ashfold, F. Claeyssens, G. M. Fuge and S. J. Henley, Chem. Soc. Rev. 33,
23-31 (2004).
52. R. K. Singh and J. Narayan, Phys. Rev. B 41, 8843-8859 (1990).
53. K. Yamamoto, Y. Koga, S. Fujiwara, F. Kokai and R. B. Heimann, Appl. Phys. A
Mater. Sci. Process. 66, 115-117 (1998).
54. P. Murray and D. Peeler, J. Electron. Mater. 23, 855-859 (1994).
55. P. R. Willmott and J. R. Huber, Rev. Mod. Phys. 72, 315 (2000).
56. D. R. Lide, CRC Handbook of Chemistry and Physics: A Ready-reference Book of
Chemical and Physical Data. (CRC Press, Boca Raton, Fla, 2008).
57. A. Peled, Photo-excited Processes, Diagnostics, and Applications: Fundamentals and
Advanced Topics. (Kluwer Academic, Norwell, MA, 2003).
58. S. R. P. Silva, Properties of Amorphous Carbon. (INSPEC, London, 2003).
Chapter 4 Pulsed laser fabrication of few-layer graphene using metal substrates
124
59. J. Robertson, Thin Solid Films 296, 61-65 (1997).
60. P. A. Khomyakov, G. Giovannetti, P. C. Rusu, G. Brocks, J. van den Brink and P. J.
Kelly, Phys. Rev. B 79, 195425 (2009).
61. K. L. Jensen, J. Lebowitz, Y. Y. Lau and J. Luginsland, J. Appl. Phys. 111, 054917-
054919 (2012).
62. H. J. Jeong, H. D. Jeong, H. Y. Kim, S. H. Kim, J. S. Kim, S. Y. Jeong, J. T. Han and
G.-W. Lee, Small 8, 272-280 (2012).
63. G. Eda, H. Emrah Unalan, N. Rupesinghe, G. A. J. Amaratunga and M. Chhowalla,
Appl. Phy. Lett. 93, 233502 (2008).
64. J. Blitz, Electrical and Magnetic Methods of Non-Destructive Testing. (Chapman &
Hall, London, 1997).
Chapter 5 Graphene-based field emitters by EPD
125
Chapter 5
Graphene-based field emitters by
electrophoretic deposition
In this chapter, chemically exfoliated graphene was used to fabricate graphene-
based field emitters using electrophoretic deposition (EPD). Free-standing graphene
emitters were studied in section 5.3, while a hybrid of CNT and graphene emitters were
fabricated and investigated in section 5.4.
5.1 Introduction
While there are currently a number of methods for graphene fabrication, only
certain methods are suitable for the mass production of graphene layers. For instance,
while CVD on metals is widely used to form single to FLG, a transfer technique is
required to release the grown graphene from the underlying metal substrate to obtain
free-standing graphene.1-3
Thus, this may be time-consuming for applications where a
large amount of individual graphene flakes are required. This may especially be so in the
case of field emission, where the emitter is generally preferred to be of a high density and
a high aspect ratio.4 Both of these requirements can be fulfilled by chemically exfoliated
graphene. This is because the small size of resultant flakes can provide numerous
emission sites and the high aspect ratio of a graphene flake (lateral size to thickness ratio)
facilitates field enhancement. In chemical exfoliation, graphene flakes are prepared by
the oxidation of graphite to graphite oxide, which is then exfoliated into graphene oxide
Chapter 5 Graphene-based field emitters by EPD
126
(GO). Reduction of GO was typically achieved with hydrazine, which is highly toxic and
unstable.5 However, in recent years, a safer microwave assisted reduction method that
involves heating GO organic or aqueous suspension in a microwave has been developed.6
Thus, with the recent advances in chemical exfoliation of graphene, a high yield of
graphene can be obtained with safer methods. As a result, low-cost and environmentally
green field emitters can be achieved.7, 8
While field emission for other carbon materials such as carbon nanotubes,
diamond-like carbon and its composites has been intensively studied, the area of
graphene field emission applications is studied to a lesser extent. Some early works
include the fabrication of graphene composites by spin coating to lift the horizontal
graphene from the substrate by Goki et al.9 and the CVD preparation of vertically aligned
graphene nanowalls by Malesevic et al..10
Wu et al.11
had also prepared graphene field
emitters with EPD and compared it to an emitter made by directly coating graphene
powder onto conductive tape, thus proving the effectiveness of EPD for field emission
applications. In section 5.3, graphene field emitters were fabricated using EPD, which is
a low cost, high deposition rate and easily up-scaled method to deposit the graphene
emitters. The effects of density and deposition time, on field emission properties of the
samples were then studied.
In section 5.4, emitters made of CNT and graphene materials were fabricated and
investigated for their FE properties. Since the discovery of CNT in 1991, it has attracted
immense interest in various applications ranging from field emission,12
energy storage,13
composite reinforcement,14
and sensors.15
Even with the discovery of its carbon
Chapter 5 Graphene-based field emitters by EPD
127
counterpart, isolated graphene sheets, which is cited to be able to outperform CNT in
various applications, it may sometimes be beneficial to combine both the materials to
produce a synergistic effect, instead of replacing either one or the other entirely in the
fabrication of materials for applications.
Methods to combine these two materials such as layer-by-layer assembly,16
spin
coating17
and filtration18
have been studied by various groups of researchers. For example,
Yu et al. have proposed a layer-by-layer electrostatic self-assembly of CNT/graphene
hybrid films by immersing a substrate into aqueous solutions of complementarily
functionalized materials for supercapacitor applications.16
Furthermore, the synergistic
behavior of the hybrid material films has also been demonstrated in current literature. For
instance, Cai et al. fabricated conductive hybrid GO/CNT films on glass substrates, in
which CNT imparted conductivity to the otherwise insulator film and GO acted as a
carrier, which imparted the ability of layer-by-layer assembly and compatibility with the
glass substrate due to the hydrophilic oxygen groups on GO.19
In addition, CNT films have also been proposed for transparent electrode
applications as voids in the CNT network increases the transparency of the film. However,
a large amount of voids also reduces the overall conductance, giving rise to a tradeoff for
transparency and conductivity in such materials. Even though pristine graphene possess
excellent electronic and mechanical properties, the difficulties in obtaining high quality
graphene can lead to structural defects that compromise conductivity and mechanical
strength. As such, this leads to much lower actual values for the various graphene
properties as compared to theoretical predictions. Hence, this limits the graphene
replacement of CNT for applications. Instead, desirable results can be obtained if both
Chapter 5 Graphene-based field emitters by EPD
128
materials work complementarily as shown by Li et al.17
who fabricated free-standing
graphene/CNT hybrid films through the transfer of CNT films onto CVD fabricated
graphene on Cu and subsequent etching of Cu. The CNT voids were found to be filled
with nanosized graphene, resulting in a highly transparent film with improved
conductivity as compared to CNT or graphene alone.18
Hybrid materials fabricated using solution processing methods such as vacuum
filtration, have also been systematically studied to directly compare the electrical and
mechanical properties of pure CNT films with graphene/CNT hybrid samples. Khan et al.
reported that hybrid graphene/CNT films raised the stiffness of the CNT-only films from
2 to 4.8 GPa at 20 wt. % and the conductivity of the films increased until a 95 wt. %
loading of graphene.20
However, it can be seen from their SEM images that when such
filtration methods were used to prepare hybrid films, the surface morphology was flat and
may not be suitable for FE applications. The films by filtration can be easily transferred
to any substrate but this would also mean that the electrical connectivity with the
substrate may not be ideal for FE.
In section 5.3, free-standing graphene emitters with very low turn-on field were
fabricated for FE application. However, it was observed that when the voltage was raised
to a relatively high value, the graphene microflakes would be elevated towards the ITO
glass slide. The burnt-out voltage value decreased with an increase in EPD deposition
time. Thus in section 5.4, a hybrid graphene/CNT emitter where the CNT was envisioned
to act like a safety belt mechanism to strap down the graphene microflakes in order to
improve the reliability at higher voltages was proposed. The hybrid films were fabricated
by EPD and their field emission properties in relation with deposition time were studied.
Chapter 5 Graphene-based field emitters by EPD
129
5.2 Experimental method
The graphite powder, potassium permanganate (KMnO4) and potassium sulfate
(K2SO4) used in this experiment were purchased from Sinopharm Chemical Reagent Co.
Ltd. Other solutions used in the preparation process are 98 % hydrosulfuric acid (H2SO4),
30 % hydrogen peroxide (H2O2) aqueous solution and ultrapure water. Modified
Hummers method21
was used to oxidize the graphite power into graphite oxide. The
graphite oxide was exfoliated into hydrophilic GO by ultra-sonication. The prepared GO
was then added to ultrapure water. 0.2 ml of ammonia was then added to this suspension,
which adjusted the overall pH value of the dispersion to 10. The exfoliation and
reducibility of the GO has been reported to be greatly improved with the presence of
ammonia.22
This suspension was then placed in a microwave system (Microwave
Synthesis For Scale Up CEM Inc.) at a temperature of 150 oC for 30 min, with the RF
power adjusted to 100 W. The suspension was then cooled to room temperature and
centrifuged at 3000 rpm/min to get a homogenously dispersed graphene suspension.
After drying, 80 mg of graphene powder was added to 200 ml of ethanol to give a
0.4 mg/ml graphene suspension. 10 mg of Al(NO3)3 was added as charger material and
the suspension was ultrasonicated for 1 hr. The deposition durations used were 5, 10 and
15 min. For the hybrid emitters, graphene/CNT mixture solutions were prepared by
adding 40 mg of graphene and 40 mg of CNT into 200 ml of ethanol to give a 0.4 mg/ml
hybrid suspension with 50-50 wt.% composition. The deposition timings used were 2.5, 5
and 10 min.
Chapter 5 Graphene-based field emitters by EPD
130
A piece of conductive carbon paper was used as the positive electrode (anode),
while the substrate, highly doped Si (n++
) was used as the negative electrode (cathode).
The electrodes were then submerged into the graphene suspension under an applied
voltage of 20 V. The as-prepared size of the graphene field emitters for demonstration
was 2 × 1 cm each. The graphene layers on these samples were viewed with a JEOL JEM
2010F high resolution TEM (200KV FEG) operated at 200 kV. A Philips XL-30 FE-
SEM was used to obtain sub-surface images of the samples. For larger area FE testing, a
parallel plate configuration was used. The graphene cathode was separated from the
indium tin oxide (ITO) coated glass slide anode by a 100 µm thick polymer spacer. The
measurements were taken under high vacuum conditions (1×10-6
Torr), with a LabView 8
software and Keithley 2410 high voltage source and current measurement unit. The turn-
on voltage was defined as the voltage needed to produce a current density of 10 µA/cm2.
5.3 Large area high density graphene for field emission
5.3.1 Microstructural characterization
Figure 5.1 shows the low magnification SEM images of the graphene emitters
after 5, 10 and 15 minutes of deposition. It was observed that a uniform spread of
graphene flakes was obtained throughout the substrate. At higher magnifications, it can
be seen that the flakes were randomly stacked and oriented, with some flakes protruding
higher above others and some almost perpendicular to the substrate. This can be directly
observed with a cross-section SEM images in Fig. 5.2. A configuration like this is
Chapter 5 Graphene-based field emitters by EPD
131
beneficial towards field emission as it improves the field enhancement of the surface.
Each flake remained thin after the EPD deposition process and the distribution of the
flakes was largely homogenous and no particular aggregation in specific areas was
observed. With the increased deposition time, the density of graphene flakes increased
such that the flakes begin to stack onto one another without contact with the substrate,
while the orientation still remained random. Figure 5.3(a) shows the TEM image of two
as-prepared graphene flakes (before EPD deposition), one of which is folded into half
while the other is lying almost entirely flat. The flakes consisted of areas, which were
scrolled and crumpled. This is typical for graphene sheets due to their extreme thinness
and thermodynamic nature. Lattice fringes can also be observed at higher magnifications
of the scrolled or folded regions in Fig. 5.3(b), while unfolded regions were generally
featureless. The largest dimension of the flake is about 2 μm. The almost featureless and
highly transparent areas on the flakes were monolayer graphene areas. Selected area
electron diffraction (SAED) of the flake gave a six fold symmetry diffraction dots
showing the high crystalline quality of the individual flake.
Chapter 5 Graphene-based field emitters by EPD
132
Fig. 5.1 SEM images of EPD deposited graphene field emitters with deposition time of (a, b) 5, (c,
d) 10 and (e, f) 15 min. Images on the left column were taken at 1000X, while those on the right
were at 20,000X magnification.
Chapter 5 Graphene-based field emitters by EPD
133
Fig. 5.2 Cross-section SEM image of EPD deposited graphene showing near vertical orientation
of microflakes.
Fig. 5.3 (a) TEM images showing two as-prepared graphene flakes, one folded and one flat.
SAED (inset) confirmed that the graphene flakes were of high crystalline quality. (b) Lattice
fringes of scrolled or folded areas at high magnification.
Chapter 5 Graphene-based field emitters by EPD
134
5.3.2 Field emission properties
The field emission characteristics of the samples deposited with 5, 10 and 15 min
are shown in Fig. 5.4. The turn-on field was 0.68 and 0.88 V/μm for the 5 and 10 min
samples respectively, while the sample deposited with 15 min burnt-out (short circuited)
before reaching the required current value.
From the Fowler-Nordheim (F-N) plots seen in Fig. 5.5, it can be seen that
electron emission did occur by tunneling as the data fitted well into a straight line when
ln (J/E2) versus 1/E. By taking the work function of graphene to be 5 eV (the same as
graphite),10
the field enhancement factors, β, were calculated from the slopes of the F-N
plots using equation 1.4 to be 39,500 and 17,300, for the 5 min and 10 min samples
respectively.
Fig. 5.4 Field emission J-E curves of the 5, 10, 15 min samples.
Chapter 5 Graphene-based field emitters by EPD
135
Fig. 5.5 F-N plots for 5 and 10 min samples. β can be calculated from the slope of the straight
line fit.
5.3.3 Discussion
Many low dimensional carbon structures, such as CNT, carbon and metal
nanoclusters embedded in DLC have been studied for field emission applications. The
turn-on fields for CNT fabricated by EPD ranged from 0.83–3.4 V/um.23-26
As such, from
the above results, it can be seen that the turn-on fields (0.68 and 0.88 V/um) for EPD
graphene emitters were comparable, if not lower as compared to the turn-on fields for
CNT deposited by EPD technique. Graphene grown by CVD on Cu substrates have also
been studied previously, and field enhancement was achieved by patterning the planar
graphene surface with photolithography and wet etching to obtain a field emission array
with each element being 30 × 30 μm2 and 20 μm apart.
27 They obtained a turn-on field of
7.2 V/μm at a current of 100 nA/cm2. This may be attributed to the relatively large
Chapter 5 Graphene-based field emitters by EPD
136
individual emitting element as compared to the dimension of a graphene or a CNT,
leading to poorer field enhancement properties.
It was observed that the 15 min sample burnt out at higher voltages before
reaching the required turn-on current value for emission. This was attributed to the
deposition time for being excessively long, leading to poor adhesion of the top layer of
graphene flakes. This can be better understood by looking at the reactions occurring at
the electrodes when a voltage is applied. Since the voltage applied is high enough (>5 V),
electrolysis of water occurs at the cathode as described in the following equation 5.1. It
should be noted that if the current of the electrochemical cell is too high, excessive H2
will be produced and bubbling may be observed at the substrate surface. This is undesired
as the bubbles will affect the adhesion of depositing species onto the substrate.
2 H2O+2e- H2 (g) + 2OH
- (5.1)
The Al(NO3)3 charger also dissociates in water to give the following ions as
shown in equation 5.2 and Al3+
is preferentially adsorbed onto the graphene flakes.
Al(NO3)3 Al(NO3)+
+ 2NO3-
(5.2)
Using nitrates as the charger material can also provide additional OH- ions by the
following reduction reactions of the nitrate ion.
NO3- + H2O +2e
- NO2
- + 2OH
- (5.3)
NO3- + 6H2O + 8e
- NH3 + 9OH
- (5.4)
The Al3+
adsorbed onto the graphene flakes, then reacts with at the cathode where
there is a high concentration of OH- ions to form metal hydroxide bonds.
Al(NO3)+
+ 3OH- Al(OH)3 +NO3
- (5.5)
Chapter 5 Graphene-based field emitters by EPD
137
The adhesion of the depositing species to the substrate is thus achieved by the
formation of metal hydroxides from metal charger ions (adsorbed on graphene flakes) at
the cathode surface, which contributes to hydrogen bonding between graphene flakes and
substrate.28-32
However, once the substrate surface is covered uniformly with graphene
flakes, the subsequent incoming flakes will begin to pile on top of the existing flakes and
have no contact with the substrate. These piled up flakes have a weaker bonding and can
lead to a higher contact resistance between the top and bottom graphene flakes, which in
turn hinders electron emission from the surface.33
Flakes that were piled onto each other
were more easily attracted toward the anode (ITO glass slide) and cause a short circuit
before a satisfactory field emission current was achieved. On the other hand, the higher
turn-on field for the 10 min sample, as compared to the 5 min sample, was attributed to
the more pronounced screening effect due to the much denser graphene flakes
configuration present.
The low turn-on field for EPD graphene emitters was mainly attributed to (i) the
high conductivity of graphene, (ii) the high density of emitting elements and (iii) random
orientation of the individual flakes, leading to protrusions and increasing the local field
enhancement effect. Some flakes were oriented normally to the substrate and electron
emission can occur from the graphene flake edges. Additionally, when a field is applied,
the free-standing flakes could align to the direction of the electric field while pivoting on
their bond with the substrate, thus effectively enhancing field emission. Emission from
graphene edges has been well agreed by researchers to be the key contributing emitting
element.34-36
However, when the density of the emitting elements (graphene flakes in this
case) became too high, electron screening occurred and this reduced the overall
Chapter 5 Graphene-based field emitters by EPD
138
effectiveness of the field emitter as the entire surface appeared electronically flat with
poor local field enhancements. Moreover, the reliability of EPD fabricated emitters was
also reduced when the stacking of graphene flakes occurred with an increased deposition
duration, which results in the emitter being easily burnt out. On the other hand, if the
deposition time is too short, the graphene flakes do not cover the surface of the substrate
uniformly. The possible scenarios are depicted in Fig. 5.6 Deposition time should be
controlled to maximize the field enhancement effects of the overall emitter surface. As
such, with the correct optimizations during fabrication, graphene-based field emitters
could potentially overtake the performance of CNT as emitters.
Fig. 5.6 Schematic diagram showing the possible scenarios when deposition time is (a) short or (b)
extended. When flake density is high, a flatter overall surface morphology gave poorer field
enhancing properties due to screening effect.
Chapter 5 Graphene-based field emitters by EPD
139
5.4 Graphene/carbon nanotube hybrid materials for field
emission
5.4.1 Microstructural characterization
From the SEM images in Fig. 5.7, it can be seen that the microstructure of the
hybrid films consisted of the typical tangled spaghetti-like CNT network with some areas
filled with graphene microflakes. At higher magnifications, the graphene microflakes
were observed to be either on the surface or embedded within the network of CNT. This
morphology was consistent throughout the entire sample, as well as through all the
deposition timings tested. The surface morphology of the samples appeared to be rougher
and had more graphene microflakes protrusions when the deposition time was increased
from 2.5 to 5 and 10 min. In contrast to the 5 and 10 min samples, at the shortest duration
of 2.5 min, larger CNT-only areas between protrusions were observed from the low
magnification SEM images. Even though the surface graphene sheets were “strapped”
down by the CNT strands, some sheets maintained a vertical position as seen in the high
magnification SEM image of the 2.5 min sample in Fig. 5.7(b) and the cross section SEM
image in Fig. 5.8(a). When the surface morphologies of hybrid samples deposited by
EPD and vacuum filtration method were compared using a SEM image in Fig 5.8(b),20
it
was observed that vacuum filtration method gave a much flatter morphology as compared
to EPD.
Chapter 5 Graphene-based field emitters by EPD
140
The TEM images of pre-deposited graphene flakes and CNT use in preparing the
EPD solution are shown in Fig. 5.9. The inset shows the high crystalline quality of the
graphene flakes and the CNT were multiwalled with a diameter of 20–40 nm each.
Fig. 5.7 SEM images of 2.5, 5 and 10 min graphene/CNT hybrid samples by EPD. Left column
shows magnification at 1000X while right column shows 20,000X.
Chapter 5 Graphene-based field emitters by EPD
141
Fig. 5.8 (a) Cross-section SEM images showing morphology with multiple protrusions and a
vertically aligned CNT-strapped graphene sheet (white arrow) (b) Hybrid graphene/CNT film by
filtration at 50% graphene in ref. 20 shows a flatter overall surface morphology.20
Reprinted from
Carbon 48, 2825(2010) with permission from Elsevier.
Fig. 5.9 TEM images showing (a) high crystalline quality graphene (b) CNT used in the
experiments.
Chapter 5 Graphene-based field emitters by EPD
142
5.4.2 Field emission properties
The field emission I-V curves of the hybrid samples in comparison to a CNT-only
sample were presented in Fig. 5.10. The turn-on fields of the 2.5, 5 and 10 min samples
were 2.84, 2.50, 1.30 V/µm respectively, as compared to 3.25 V/µm for a 5 min CNT-
only sample. The turn-on field was generally improved for the hybrid films as compared
to pure CNT films. Even though the turn-on field was still higher than pure graphene
films, better reliability and less burnt-out situations were observed for the hybrid samples
at higher voltages.
The respective F-N plots for the hybrid samples are shown in Fig. 5.11. Using a
work function of 5 eV for carbon, the β obtained from equation 1.4 was 1800, 3400,
19700 for the 2.5, 5 and 10 min hybrid samples respectively.
Fig. 5.10 I-V plots of graphene/CNT hybrid emitters.
Chapter 5 Graphene-based field emitters by EPD
143
Fig. 5.11 F-N plots of graphene/CNT hybrid emitters.
5.4.3 Discussion
It has been established that by combining both CNT and graphene, hybrid
materials that gave improved properties can be produced. The method of fabrication can
also affect the performance of the hybrid materials in certain applications. For instance,
samples deposited with vacuum filtration methods (Fig. 5.8) gave smoother surface
morphologies as compared to EPD and may not be ideal for FE applications. In contrast,
EPD gave hybrid samples with many surface protrusions that can aid the local field
enhancement. In these experiments, it was observed that there was more CNT than
graphene on the hybrid samples, despite the fact that the CNT and graphene hybrid
suspension was prepared to contain a 50-50 wt. % of each type of material. This became
less pronounced with increasing deposition times as the amount of graphene deposited
Chapter 5 Graphene-based field emitters by EPD
144
was observed to increase when the deposition duration was increased to 10 and 15 min.
The duration needed to fully cover the substrate macroscopically was also reduced for
hybrid suspensions as compared to pure graphene suspensions. This may be due to the
different drag forces experienced by the CNT and graphene flakes in the solution, when
moving towards the cathode. CNT have a more streamlined body (20–40 nm) as
compared to graphene microflakes, which can be of sizes up to 10 μm. Thus, the
graphene flakes faced a larger drag force when moving through the solution. Larger
flakes can also take a longer time to arrive. As such, larger flakes, in addition to small
flakes were deposited in the 10 min sample as compared to the 2.5 and 5 min samples.
However, the delay in arrival times of the two species was not significant enough to
cause phase separation. The microstructure of the hybrid was therefore well mixed
throughout when the deposition duration was set beyond the minimum time needed for
the graphene flakes to arrive on the substrate. It is worthy to note that this issue would not
be present in methods such as spin coating and vacuum filtration, which does not involve
the electrostatic attraction between species in the solution and substrate. As such, the 50-
50 wt. % of CNT and graphene would be maintained in these methods.
To achieve improved FE characteristics, it is ideal that the material has a good
conductivity within, as well as possessing areas of local field enhancement. These can be
in the form of (i) physical surface protrusions or (ii) areas where field lines are
concentrated at the surface of the emitter, due to different conductivities of phases in a
composite. An example of the former would be the samples studied in this chapter, while
the latter is the material fabricated and discussed in Chapter 3. Graphene and CNT each
played a slightly different role in this hybrid material even though both of them
Chapter 5 Graphene-based field emitters by EPD
145
ultimately contributed to field emission at the surface. Within the film, the microstructure
can be viewed as graphene embedded within the CNT matrix. However, at the surface
regions where FE occurs, the CNT strands coated the graphene flakes instead. As such,
graphene had at least two roles; 1) to act as connection particles between the CNT strands
and improve the conductivity within the porous CNT network when they were embedded
inside, 2) to act as additional local field enhancing spots when they were located at near
surface regions. On the other hand, the role of CNT was to strap the graphene flakes
down at high voltages such that they would not detach from the substrate and elevate
towards the ITO glass slide, thereby improving the reliability of the emitters.
From the results, it was observed that with an increase in deposition time, the
amount of graphene flakes on the surface and the thickness of the film increased. As the
conductivity of porous CNT films depend on the contact resistance of inter-tube junctions,
having more graphene flakes embedded would create extra connecting conduction paths
arising from sheet-tube junctions and thus increased film conductivity.20, 37
Also, having a
larger amount of graphene sheets on the surface, when the deposition time was increased,
served to provide more emission sites at the surface. This in turn led to lower emission
turn-on fields. However, the turn-on fields of the hybrid films were higher than pure
graphene films. This could be due to the inability of the strapped-down graphene sheets
to align to the direction of the applied field as in the case of the free-standing graphene-
only emitters. The obtained results from section 5.3 and 5.4 also suggested that with a
decrease in CNT fraction and more graphene flakes at the surface, turn-on field could be
further lowered for the hybrid samples.
Chapter 5 Graphene-based field emitters by EPD
146
5.5 Summary
In summary, uniform 2 × 1 cm2 graphene field emitters were fabricated by EPD in
section 5.3. The turn-on field was 0.68 and 0.88 V/um for the 5 min and 10 min samples
respectively. The higher turn-on field of the 10 min sample was attributed to the very
high density of graphene flakes, which led to electron screening and lowered the overall
field enhancement of the emitter. The random orientation of the graphene flake led to
local field enhancement due to surface protrusions, as well as numerous emission sites
which include those from the edges of graphene flakes oriented normal to the substrate.
Graphene field emitters fabricated by EPD, in combination with chemical exfoliation and
microwave assisted reduction, presented a low-cost, high yield and safe method for
graphene field emitters that can easily up-scaled.
Graphene/CNT hybrid films were prepared using EPD in section 5.4. The
fabricated films consisted of well-dispersed graphene sheets within a spaghetti CNT
network and possessed a surface morphology that was favorable for FE applications. The
results revealed that with an increase in deposition duration, the turn-on field decreased
for the hybrid materials. The turn-on fields of the 2.5, 5 and 10 min samples were 2.84,
2.50, 1.30 V/µm respectively. Lower turn-on fields for the hybrid samples as compared to
pure CNT samples were attributed to the increased conductivity within the samples by
intersheet-tube contacts and additional field enhancing graphene emission spots on the
surface. These hybrid materials offer a range of possibilities for new materials for
nanotechnology.
Chapter 5 Graphene-based field emitters by EPD
147
References
1. L. G. De Arco, Z. Yi, A. Kumar and Z. Chongwu, IEEE Trans. Nanotechnol. 8, 135-
138 (2009).
2. K. S. Kim, Y. Zhao, H. Jang, S. Y. Lee, J. M. Kim, K. S. Kim, J.-H. Ahn, P. Kim, J.-
Y. Choi and B. H. Hong, Nature 457, 706-710 (2009).
3. A. Reina, X. Jia, J. Ho, D. Nezich, H. Son, V. Bulovic, M. S. Dresselhaus and J.
Kong, Nano Lett. 9, 30-35 (2008).
4. J. M. Bonard, J. P. Salvetat, T. Stöckli, L. Forró and A. Châtelain, Appl. Phys. A
Mater. Sci. Process. 69, 245-254 (1999).
5. S. Pei and H.-M. Cheng, Carbon 50, 3210-3228 (2012).
6. Y. Zhu, S. Murali, W. Cai, X. Li, J. W. Suk, J. R. Potts and R. S. Ruoff, Adv. Mater.
22, 3906-3924 (2010).
7. Y. Zhu, S. Murali, M. D. Stoller, A. Velamakanni, R. D. Piner and R. S. Ruoff,
Carbon 48, 2118-2122 (2011).
8. G. M. Morales, P. Schifani, G. Ellis, C. Ballesteros, M. Gerardo, C. Barbero and H. J.
Salavagione, Carbon 49, 2809-2816 (2011).
9. G. Eda, H. E. Unalan, N. Rupesinghe, G. A. J. Amaratunga and M. Chhowalla, Appl.
Phys. Lett. 93, 233502 (2008).
10. A. Malesevic, R. Kemps, A. Vanhulsel, M. P. Chowdhury, A. Volodin and C. Van
Haesendonck, J. Appl. Phys. 104, 084301 (2008).
11. Z.-S. Wu, S. Pei, W. Ren, D. Tang, L. Gao, B. Liu, F. Li, C. Liu and H.-M. Cheng,
Adv. Mater. 21, 1756-1760 (2009).
12. Z. W. Pan, F. C. K. Au, H. L. Lai, W. Y. Zhou, L. F. Sun, Z. Q. Liu, D. S. Tang, C. S.
Lee, S. T. Lee and S. S. Xie, J. Phys. Chem. C 105, 1519-1522 (2001).
Chapter 5 Graphene-based field emitters by EPD
148
13. Z. Chen, Y. Qin, D. Weng, Q. Xiao, Y. Peng, X. Wang, H. Li, F. Wei and Y. Lu, Adv.
Funct. Mater. 19, 3420-3426 (2009).
14. W. Ma, L. Liu, Z. Zhang, R. Yang, G. Liu, T. Zhang, X. An, X. Yi, Y. Ren, Z. Niu, J.
Li, H. Dong, W. Zhou, P. M. Ajayan and S. Xie, Nano Lett. 9, 2855-2861 (2009).
15. M. Pumera, Chem. Rec. 9, 211-223 (2009).
16. D. Yu and L. Dai, J. Phys. Chem. C 1, 467-470 (2009).
17. V. C. Tung, L.-M. Chen, M. J. Allen, J. K. Wassei, K. Nelson, R. B. Kaner and Y.
Yang, Nano Lett. 9, 1949-1955 (2009).
18. C. Li, Z. Li, H. Zhu, K. Wang, J. Wei, X. Li, P. Sun, H. Zhang and D. Wu, J. Phys.
Chem. C 114, 14008-14012 (2010).
19. D. Cai, M. Song and C. Xu, Adv. Mater. 20, 1706-1709 (2008).
20. U. Khan, I. O’Connor, Y. K. Gun’ko and J. N. Coleman, Carbon 48, 2825-2830
(2010).
21. W. S. Hummers and R. E. Offeman, J. Am. Chem. Soc. 80, 1339-1339 (1958).
22. I. Janowska, K. Chizari, O. Ersen, S. Zafeiratos, D. Soubane, V. Costa, V. Speisser, C.
Boeglin, M. Houllé, D. Bégin, D. Plee, M.-J. Ledoux and C. Pham-Huu, Nano Res. 3,
126-137 (2010).
23. S. M. Jung, J. Hahn, H. Y. Jung and J. S. Suh, Nano Lett. 6, 1569-1573 (2006).
24. L. Wang, Y. Chen, T. Chen, W. Que and Z. Sun, Mater. Lett. 61, 1265-1269 (2007).
25. W. Y. Sung, S. M. Lee, W. J. Kim, J. G. Ok, H. Y. Lee and Y. H. Kim, Diam. Relat.
Mater. 17, 1003-1007 (2008).
26. Y. Chen, H. Jiang, D. Li, H. Song, Z. Li, X. Sun, G. Miao and H. Zhao, Nanoscale
Res. Lett. 6, 1-8 (2011).
27. C.-K. Huang, Y. Ou, Y. Bie, Q. Zhao and D. Yu, Appl. Phys. Lett. 98, 263104 (2011).
Chapter 5 Graphene-based field emitters by EPD
149
28. B. J. C. Thomas, A. R. Boccaccini and M. S. P. Shaffer, J. Am. Chem. Soc. 88, 980-
982 (2005).
29. O. O. Van der Biest and L. J. Vandeperre, Annu. Rev. Mater. Sci. 29, 327-352.
30. J. B. Talbot, E. Sluzky and S. K. Kurinec, J. Mater. Sci. 39, 771-778 (2004).
31. J. Cho, S. Schaab, J. Roether and A. Boccaccini, J. Nanopart. Res. 10, 99-105 (2008).
32. R. Hashaikeh and J. Szpunar, J. Coating. Tech. Res. 8, 161-169 (2011).
33. L. Gomez De Arco, Y. Zhang, C. W. Schlenker, K. Ryu, M. E. Thompson and C.
Zhou, ACS Nano 4, 2865-2873 (2010).
34. C. Wu, F. Li, Y. Zhang and T. Guo, Carbon 50, 3622-3626 (2012).
35. H. Yamaguchi, K. Murakami, G. Eda, T. Fujita, P. Guan, W. Wang, C. Gong, J.
Boisse, S. Miller, M. Acik, K. Cho, Y. J. Chabal, M. Chen, F. Wakaya, M. Takai and
M. Chhowalla, ACS Nano 5, 4945-4952 (2011).
36. D. Ye, S. Moussa, J. D. Ferguson, A. A. Baski and M. S. El-Shall, Nano Lett. 12,
1265-1268 (2012).
37. P. E. Lyons, S. De, F. Blighe, V. Nicolosi, L. F. C. Pereira, M. S. Ferreira and J. N.
Coleman, J. Appl. Phys. 104, 044302-044308 (2008).
Chapter 6 Conclusion and future work
150
Chapter 6
Conclusions and future work
6.1 Conclusions
The primary objective of this work was to fabricate sp2 and graphene-based
materials for field emission applications. The first step taken towards this aim was to
perform carbon deposition on Si substrates at 100, 400 and 700 oC with a PLD system to
grow nanographite in a-C matrix. It was shown that the sp2 content in the samples
increased with the increase in deposition temperature. Additionally, the sp2 content of the
film was mainly agglomerated into nanoclusters as seen in TEM observations. From
AFM surface morphology images, the surface of the samples were highly smooth and
external field enhancement factor, βex, was taken to be 1. The 700 oC sample had the
lowest turn-on voltage among the samples and this was attributed to the improved
internal enhancement by the high density of nanographite clusters present. The electrical
conduction within the films was improved through conduction channels formed either by
overlapping nanographite clusters or by electron hopping between the clusters. Hence,
the presence of such conduction channels gave the flat films a CNT-like emission
mechanism. The relative field emission barrier heights were also estimated for the 3
samples with β100: β400: β700 = 1: 1.46: 2.02.
By extending the method of carbon deposition with heating during pulsed laser
ablation of carbon to a transition metal substrate (e.g. Ni) in place of Si substrate, few-
layer graphene was fabricated from solid state transformation of PLD deposited carbon.
Chapter 6 Conclusion and future work
151
The slow cooling (1 oC/min) and medium cooling rates (50
oC/min) gave few-layer
graphene, while the fast cooling rate gave disordered carbon due to insufficient time for
the carbon atoms to arrange into ordered crystalline structures. The graphene fabricated
were 3–4 layers thick by estimation from the I(2D)/I(G) ratio of Raman peaks. The FLG
could also be directly observed under TEM. In addition, lower pulsed laser energies of 50
or 100 mJ gave FLG formation, while no FLG was detected for the pulsed laser energy of
200 mJ. This was due to a shallower implantation of carbon atoms into the metal
substrate at lower energies, which in turn facilitated the precipitation process.
Graphene segregation using PLD was subsequently studied with the use of
various metal substrates such as Ni, Cu, Co and Fe. By using the 1 oC/min cooling profile,
graphene was only formed on Ni but absent in the other three metals. On one hand, the
absence of FLG on Cu was due to its extremely low carbon solubility. On the other hand,
for a high carbon solubility metal like Co, a suitable cooling recipe is needed for FLG
formation. By increasing the cooling rate to suit Co, FLG was also formed on Co metal
with a high coverage. The fabricated graphene on Co was transferred onto a Si substrate
by etching the Co metal substrate with dilute acid. After the transfer, it was observed that
the D peak signal dropped significantly. This proved that a large part of the D peak signal
was contributed by the amorphous carbon that was still embedded within the metal
substrate after the completion of the segregation process. This was further confirmed by
the reduced D peak signal of Raman spectra taken from areas exposed and unexposed to
the carbon plume during deposition.
In addition, graphene segregation using PLD was also studied with the use of
different wavelength lasers such as Nd:YAG (λ=266 nm) laser as compared to the KrF
Chapter 6 Conclusion and future work
152
(λ=248 nm). Even though the energy of the carbon species in the Nd:YAG system was >1
eV smaller than those in the KrF system, FLG was formed on Ni substrates with the same
cooling profile for both lasers. On a side note, Nd:YAG might have a slight advantage
over KrF with its lower energy carbon species, which favors the formation of graphitic
sp2 carbon rather than amorphous carbon.
Using the information gained from PLD segregation of graphene on metals, Ni
and Co metal nanocones with varied densities and height distribution were fabricated for
field emission purposes. The field emission properties of the metal nanocones were
greatly enhanced with the graphene layer coating. This was attributed to the charge
transfer interactions between the metal nanocones and graphene, which resulted in a
reduced work function for the combined material. This in turn led to a reduced electron
tunneling barrier height that manifested as a lower turn-on voltage for the graphene-on-
metal nanocones. For both metals, a lower turn-on voltage was obtained from the taller
cones separated with shorter cones in-between them (Type A), while a higher turn-on
voltage was acquired from the high density uniform height nanocones (Type B), due to
more serious electron screening effects experienced in the latter configuration.
Large area graphene emitters were also fabricated using EPD. The fabricated
emitters consisted of randomly oriented graphene flakes with some flakes being vertically
oriented with respect to the substrate. This gave a favorable configuration for field
emission in contrast to methods such as vacuum filtration that generally gave a flatter
surface morphology. With the increase in deposition duration, the density of graphene
flakes on the substrate increased. However, the field emission turn-on voltage also
Chapter 6 Conclusion and future work
153
increased with deposition time. The sample that was deposited for 5 min gave the lowest
turn-on voltage followed by the sample that was deposited for 10 min. No considerable
emission was observed for the sample deposited at 15 min before a short circuit occurred.
The poorer emission characteristics of the 10 and 15 min samples were due to the
reduction in surface geometrical enhancement caused by overcrowding of graphene
flakes on the substrate. In addition, as the stacked graphene flakes had no contact with the
substrate, poor contact resistance between the flakes can also lead to higher overall
resistance for the thicker samples. Furthermore, when the deposition duration was
excessively long in the case of 15 min, the samples were easily short circuited at higher
voltages as the piled-up flakes elevated towards the ITO glass slide used as the anode.
To improve the reliability of the graphene EPD emitters, hybrid graphene/CNT
emitters were prepared by EPD. The CNT strands were envisioned to resemble a “safety
belt” mechanism that would hold down the graphene flakes at high voltages. Despite the
50-50 wt.% of CNT and graphene used in the prepared EPD source solution, the
fabricated emitters were predominately CNT with some graphene flakes embedded
within the CNT network. The hybrid materials showed improved field emission
properties as compared to pure CNT emitters fabricated with the same method. This was
because the graphene flakes embedded within the film provided additional conduction
paths for electron transport to the surface in the otherwise porous CNT film. The
presence of graphene flakes also increased the number of emitting spots on the surface of
the hybrid film.
Chapter 6 Conclusion and future work
154
From the studies carried out in this dissertation, various effective sp2 and
graphene-based field emitters can be fabricated from physical methods such as PLD,
microstructure enhancements by using sharp cones shapes and solution processing
methods such as EPD. With careful control of cooling profile, parameters affecting
emitter density and combination of materials, the performance of graphene-based field
emitters can be enhanced.
6.2 Future work
In this work, the fabrication of graphene through solid state transformation of
carbon provided by PLD at 750 oC was studied. It has also been shown that using this
technique, FLG can be grown directly onto nano-patterned substrates for various
applications. However, some loss in the structural outline of the metal cones, calls to
attention that the deposition temperature has to be further decreased. As such, it may be
beneficial to explore longer laser wavelengths lasers such as 532 or 1064 nm in future
that can allow for an easier sp2 formation as well as shallower carbon ion implantation
depth. The carbon ions in pulsed laser plumes have sufficient energy to implant carbon
atoms into the metal substrates and this can aid the absorption of sufficient carbon at a
lower temperature. However, if implantation occurs too deeply into the metal substrate,
precipitation can be difficult. Sufficient carbon has to be absorbed by the near-surface
metal regions to exceed the C solubility for segregation to occur and a shallower
implantation depth is preferred since saturation can occur at areas nearer to the surface
and cooling time can be reduced as well.
Chapter 6 Conclusion and future work
155
In the other section for large area emitter fabrication with EPD, hybrid
graphene/CNT emitter materials were studied. However, the 1:1 CNT to graphene ratio in
the source solution was not directly transferred onto the substrates during fabrication.
Hence, more control over the EPD process of hybrid materials is desired and the growth
mechanism of hybrid materials using EPD should be investigated. Furthermore, it would
be interesting to investigate the field emission properties of a hybrid emitter, whereby a
large amount of graphene flakes are strapped down by a minimal amount of CNT strands.
Once again, a better understanding of the EPD process is required to achieve such a
calculated configuration.
In closing, the development of graphene-based materials for field emission is in
the stage of infancy. As the field emission properties of graphene are highly affected by
its configuration, fabrication techniques play a critical role in realizing its full potential in
this area. There remain yet many unexplored methods and challenges towards graphene
field emitter fabrication before it can be employed in actual applications. Nonetheless,
with a suitable technique and control of the correct parameters involved, graphene can
prove to triumph other materials of its class and play a pivotal role in the advancement of
field emission devices.