University of Southern California Dissertations and Theses

Graphene and carbon nanotubes: synthesis, characterization and applications for beyond silicon electronics

(USC Thesis Other)

Graphene and carbon nanotubes: synthesis, characterization and applications for beyond silicon electronics

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

GRAPHENE AND CARBON NANOTUBES: SYNTHESIS, CHARACTERIZATION
AND APPLICATIONS FOR BEYOND SILICON ELECTRONICS

by

Lewis Mortimer Gomez De Arco

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(CHEMISTRY)

December 2010

Copyright 2010 Lewis Mortimer Gomez De Arco
ii

Epigraph

“Mas valiosa es la sabiduria que las piedras preciosas”
“Wisdom is far more valuable than gemstones”

iii

Dedication

A mi madre: Tu amor, altruismo y ensenanzas han tenido recompensa.

iv

Acknowledgements

First, I would like to express my gratitude to my advisor, Professor Chongwu Zhou,
for his guidance and constant motivation to pursue empowering ideas that have the
invariable aim of producing significant advances in the field of nanoscience and
nanotechnology. His excellent insights were instrumental through my Ph.D. studies to
expand my horizons and to carefully conciliate the scientist and the engineer we all have
within. I gratefully appreciate the support and encouragement I have received during the
past years. Without them, it would have been impossible to carry out this work.
I also want to thank my dissertation committee: Professor Mark Thompson and
Professor Martin Gundersen for their agreement to serve on my Ph.D. guidance
committee and taking the time and effort to evaluate my work. It is an honor to have you
in my committee.
In addition, I want to thank my collaborators, colleagues and friends: Dr. Thomas
Vernier, Dr. Kosmas Galatsis, Dr. Bo Lei, Dr. Koungmin Ryu, Dr. Fumiaki Ishikawa, Dr.
Po-Chiang Chen, Dr. Rajay Kumar, Dr. Meng-Tse Chen, Dr. Matthew Allen, Dr. Vincent
Tung, Jenny Linn, Yuyuang Zhang, Yi Zhang, Alexander Badmaev, Akshay Kumar,
Chuan Wang, Zhen Li, Jia Liu, Yuchi Che, Hsiaokang Chang, Jialu Zhang, Xue Lin,
Anuj Madaria, Haitian Chen, Yue Fu, Jing Xu, Jing Qiu, Prof. Mark Thompson, Prof.
Thomas Chen, Prof. Martin Gundersen, Prof. Steve Cronin, Prof. Steve Bradforth and all
other people who offered their precious help.
Finally, a special word for my family: You mean everything to me, thank you for
your unconditional love, support and patience. To my mother Irene, because your love
v

and sacrifice has paid off overwhelmingly. With great effort and dedication, you and Tita,
my granny, taught me to obey God, your guidance and encouragement to do what is good
and just has made me what I am today, thank you. I love you more than what words can
explain. To my sister Lia, my niece Vianeth and my nephew David, thank you because
from the distance, at all times, you let me know I was missed, loved and valued. To my
dear wife Liliana, my love, my inspiration, thanks for walking along my side giving me
support, strength and calmness. You and my dear son Daniel, whose smile brightens the
darkest ambient and fulfills my soul, have helped me carry on during difficult times and
have made much more gratifying the good ones. I love you.
To all my friends, who have helped me in one way or another to build my character
through unforgettable moments full of daily life teachings, thank you for your support
and for being there for me. For all the omitted ones that should be mentioned here, thank
you, you know who you are.
vi

Table of Contents

Epigraph .............................................................................................................................. ii
Dedication .......................................................................................................................... iii
Acknowledgements ............................................................................................................ iv
List of Tables ..................................................................................................................... ix
List of Figures ......................................................................................................................x
Abstract ..............................................................................................................................xx

Chapter 1 Introduction .....................................................................................................1
1.1 Introduction to carbon nanostructures .......................................................................4
1.1.1 Graphene.............................................................................................................4
1.1.2 Electronic structure of two-dimensional carbon: Graphene ...............................5
1.1.3 Electronic structure of one-dimensional carbon: Carbon Nanotubes ...............11
1.2 Graphene and carbon nanotubes for beyond silicon electronics .............................17
1.3 Thesis outline ..........................................................................................................26
Chapter 1 References .....................................................................................................27

Chapter 2 Synthesis, Transfer and Devices of Single- and Few-Layer Graphene by
Chemical Vapor Deposition ......................................................................... 31
2.1 Introduction .............................................................................................................31
2.2 Experiments .............................................................................................................33
2.2.1 Single layer and few-layer graphene synthesis ................................................33
2.3 Results .....................................................................................................................35
2.3.1 Graphene synthesis ...........................................................................................35
2.3.2 Device fabrication and electrical measurements ..............................................37
2.3.3 Graphene transparent conductive film..............................................................39
2.4 Conclusions .............................................................................................................41
Chapter 2 References .....................................................................................................42

Chapter 3 Synthesis of Single and Bilayer Graphene by Chemical Vapor Deposition
of Carbon on Single Crystal Nickel ..............................................................44
3.1 Introduction .............................................................................................................44
3.2 Experiments .............................................................................................................46
3.2.1 Graphene synthesis ...........................................................................................46
3.3 Results .....................................................................................................................47
3.3.1 X-ray diffraction of single crystal nickel vs. polycrystalline nickel ................47
3.3.2 Grain boundaries and multilayer graphene formation ......................................48
3.3.3 Graphene characterization ................................................................................53
3.4 Conclusions .............................................................................................................57
Chapter 3 References .....................................................................................................59

vii

Chapter 4 Large-Area Graphene Films by Chemical Vapor Deposition for Highly
Flexible Organic Photovoltaics ......................................................................61
4.1 Introduction .............................................................................................................61
4.2 Experiments .............................................................................................................64
4.2.1 CVD graphene synthesis ..................................................................................64
4.2.2 CVD graphene transfer .....................................................................................64
4.2.3 Organic photovoltaic cell fabrication with CVD graphene and ITO ...............65
4.3 Results ......................................................................................................................66
4.3.1 X-ray diffraction of polycrystalline Ni .............................................................66
4.3.2 TEM, SAED and Raman of CVD graphene .....................................................67
4.3.3 Atomic force microscopy of CVD graphene, ITO and SWNT films ...............68
4.3.4 Characterization of CVD graphene film as transparent electrode for
organic photovoltaics ......................................................................................71
4.3.5 Flexible transparent electrodes: Graphene vs. ITO ..........................................73
4.3.6 Graphene OPV cell fabrication on flexible substrate and device
performance .....................................................................................................75
4.3.7 Device performance model: The Lambert-W function ....................................80
4.3.8 Performance of Highly-Flexible Graphene OPV Cell under Bending .............83
4.3.9 CVD graphene organic photovoltaic cells on rigid substrates .........................87
4.4 Conclusions .............................................................................................................89
Chapter 4 References .....................................................................................................90

Chapter 5 Metal to Semiconductor Ratio of Aligned Carbon Nanotubes
on a-Sapphire .................................................................................................93
5.1 Introduction .............................................................................................................93
5.2 Experiments .............................................................................................................95
5.3 Results ......................................................................................................................96
5.4 Conclusions ...........................................................................................................105
Chapter 5 References ...................................................................................................106

Chapter 6 Scalable Light-Induced Metal to Semiconductor Conversion of Carbon
Nanotubes ....................................................................................................108
6.1 Introduction ...........................................................................................................108
6.2 Experiments ...........................................................................................................109
6.2.1 Wafer scale synthesis of aligned carbon nanotubes .......................................109
6.2.2 Wafer scale CNTFET fabrication ...................................................................110
6.3 Results ...................................................................................................................112
6.4 Conclusions ...........................................................................................................128
Chapter 6 References ...................................................................................................129

Chapter 7 Conclusions and Future Work .....................................................................133
7.1 Summary and conclusions.....................................................................................133
7.2 Future work ...........................................................................................................134
7.2.1 Graphene nanoribbons with smooth edges .....................................................135
7.2.2 Strained graphene ...........................................................................................137
viii

7.2.3 Synthesis of carbon nanotubes with specific diameter and chirality ..............139
Chapter 7 References ...................................................................................................140

Bibliography ....................................................................................................................142

ix

List of Tables

Table 1.1 Classification of nanomaterials by dimensionality and
composition.
.
2
Table 1.2 Important parameters of carbon materials of different
dimensionalities
.
5
Table 2.1 Parameters correlating graphene and Ni (111).. 32
Table 4.1 Performance details of OPV cells built on PET. The structure of
the devices is given by [CVD-
graphene/PEDOT/CuPc/C60/BCP/Al] and
[ITO/PEDOT/CuPc/C60/BCP/Al] for CVD graphene and ITO
OPVs, respectively.

80
Table 4.2 Table 4.2, Comparison of performance details of OPV cells built
on glass substrates. The structure of the devices is given by
[CVD GRAPHENE/PEDOT/CuPc/C60/BCP/Al] and
[ITO/CuPc/C60/BCP/Al] for CVD GRAPHENE and ITO OPVs,
respectively.
89

x

List of Figures

Figure 1.1 Graphene is a 2D building material for carbon materials of
all other dimensionalities. It can be wrapped up into 0D
buckyballs, rolled into 1D nanotubes or stacked into 3D
graphite.

6
Figure 1.2 Unit cell (a) and Brillouin zone (b) of two-dimensional carbon
are shown as the dotted rhombus and the shaded hexagon,
respectively. and , (i=1,2) are unit vectors and reciprocal
lattice vectors, respectively. Energy dispersion relations are
obtained along the perimeter of the dotted triangle connecting
the high symmetry points Γ, Κ, and Μ.

8
Figure 1.3 Energy dispersion relations of two-dimensional graphene shown
throughout the Brillouin zone and the inset shows the energy
dispersion relations along the high symmetry axes along the
perimeter of the triangle shown in Figure 1.2b.

9
Figure 1.4 Conical low-energy spectrum E(k), indicating changes in the
position of the Fermi energy E
F
with changing gate voltage V
g
.
Positive (negative) V
g
induce electrons (holes) in concentrations
n = V
g
where the coefficient 7.2 10
10
cm
-2
V
-1
for field-
effect devices with a 300 nm SiO
2
layer used as a dielectric.

11
Figure 1.5 The unrolled nanotube lattice. A nanotube can be constructed by
connecting sites O and A, and B and C. OA and OB define the
chiral and translational vectors, respectively. The rectangle
AOCB defines the unit cell for the nanotube. The vector R
denotes a symmetry vector..

12
Figure 1.6 Schematic showing the atomic arrangement of zigzag (12,0),
armchair (7,7)and chiral (9,5) nanotubes.

13
Figure 1.7 Representation of the Brillouin zone for graphene and carbon
nanotube lattice. The condition for the obtention of metallic
energy bands is met if the ratio of the length of the vector to
that of K
1
is an integer.

Otherwise, the nanotube band structure
would have a bandgap and exhibit semiconducting behavior.

15
xi

Figure 1.8 1-D Energy bands (a and c) and 1-D electronic density of
states (b and d) for a semiconducting (blue) and metallic
(red) carbon nanotube. The semiconducting nanotube with
(n,m) = (9,5) clearly shows a bandgap in the band
structure and absence of energy states in the bandgap.
However, the metallic nanotube (7,7) exhibits a
continuous band structure with no bandgap and a finite
density of states at the Fermi energy

16
Figure 1.9 (a) SEM image of a typical suspended six-probe graphene
device. AFM image of the suspended device before (b)
and after (c) the measurements and graphene removal
oxygen plasma etch. (d) Device schematic, side view,
suspended single-layer graphene is colored pink. (e)
Measured four-probe resistivity as a function of gate
voltage before (blue) and after (red) current annealing.
Data from traditional high-mobility device on the
substrate is plotted with gray dotted line. (f) Mobility μ =
as a function of carrier density n for the same devices

19
Figure 1.10 (a) Schematic diagram of a SWNT aligned on the a-plane
sapphire substrate. The red and blue spheres represent oxygen
and aluminum atoms, respectively. The purple plane shows the
a-plane orientation. (b) AFM image of as-grown SWNTs
oriented in parallel on the a-plane sapphire substrate.

22
Figure 1.11 Separation of carbon nanotubes by density gradient
ultracentrifugation. (a) Colored bands in the centrifuge tube
correspond to separated metallic and semiconducting fractions.
Inset shows a schematic representation of surfactants wrapping a
carbon nanotube. (b) UV-Vis-NIR absorption and (c) Raman
spectra of the separated metallic (black) and semiconducting
(red) fractions.

23
Figure 1.12 Electrical breakdown of CNTFETs. (a) I
ds
-V
g
curves for a
typical transistor after consecutive electrical breakdown. Inset
shows a diagram of a back-gated transistor built on aligned
carbon nanotubes. (b) On current vs On/Off current for multiple
CNTFET devices before and after electrical breakdown.

25

xii

Figure 2.1 Schematic representation of the atomic arrangement of the
hexagonal lattice of the (111) face of nickel (blue spheres) and
how carbon atoms (gray spheres) would arrange on the Ni (111)
surface to form graphene. .

33
Figure 2.2 AFM image of e-beam evaporated nickel film atop a Si/SiO
2

substrate before (a) and after (b) high temperature annealing. (c)
X-ray diffraction spectra of the nickel film after thermal
annealing. (d) Schematic diagram depicting apparatus and
chemical vapour deposition process employed. (e) Wafer scale
growth of graphene on Ni substrate. (f) AFM image of nickel
substrate after CVD synthesis of graphene.
.
34
Figure 2.3 Figure 2.3 (a) Raman spectrum obtained on as-synthesized
graphene films on Si/SiO
2
/Ni substrates. D, G and G’ Raman
bands for graphene are labeled on each spectrum. (b) Raman
spectrum obtained on a single layer graphene. Zoomed images
of the G’ band on the right shows the purely single Lorentzian fit
of this peak (red trace), which is characteristic of single-layer
graphene. (c) Raman spectrum of bilayer or few-layer graphene.
A nearly symmetric splitting and broadening of the G’ band of
this spectrum (shown on the right) is properly fit by a set of four
Lorentzian curves, that can be assigned to bilayer or few-layer
graphene. No G’ band was found with a bulk graphite-like
lineshape on the substrates sampled.
.
36
Figure 2.4 (a) Schematic of a graphene film transferred to a Si/SiO
2

substrate via nickel etching. (b) Photograph of a 4 inch wafer
after synthesis of transferred FLG on polycrystalline nickel;
upper right shows an AFM image of the few-layer graphene
films on Si/SiO
2
; lower right panel shows an optical micrograph
of deposited few-layer graphene on Si/SiO
2
; lower left panel
shows a typical micro Raman spectrum of the transferred films.
.
37
Figure 2.5 (a) 4-inch wafer with back-gated few-layer graphene devices;
insets show SEM and AFM images of a typical device and
device channel, respectively. (b) I
DS
-V
DS
measurements for
different gate voltages, V
G
= 2.5 V, 1.5 V and -1.5 V for the
black, red and blue curves, respectively and (c) I
DS
-V
G
curve of
one of the FET devices for V
DS
= 0.01 V.

39
xiii

Figure 2.6 (a) Photograph of a 1cm
2
FLG film transferred onto a glass
substrate (inside the red box). (b) Transmittance of the FLG film
shown in (a).

40
Figure 3.1 (a-b) XRD spectra and AFM images of Ni (111) substrate and
polycrystalline Ni substrate, respectively. The left inset in 1a is a
schematic diagram of Ni (111) hexagonal lattice arrangement
with graphene carbon atoms on surface. The blue atoms are Ni
and the grey ones are carbon. The color scale bar corresponds to
AFM images in a and b inset.

47
Figure 3.2
(a) XRD spectra collected from polycrystalline Ni with fast
(black), medium (red), and slow (blue) annealing rates and XRD
spectrum collected from Ni (111) (green) after thermal annealing
(XRD spectrum is identical for Ni (111) using different
annealing rates). (b). Zoomed-in XRD spectra of peaks at 2θ =
52.16° (assigned as Ni (200)). (c-j). Optical images taken after
graphene CVD growth from polycrystalline Ni with fast (c, d),
medium (e, f) and slow (g, h) annealing rates. (i, j). Optical
images taken after graphene CVD growth from Ni (111).

49
Figure 3.3 (a-b). Schematic diagrams of graphene growth mechanism on Ni
(111) (a) and polycrystalline Ni surface (b). (c). Optical image of
graphene/ Ni (111) surface after the CVD process. The inset is a
three dimensional schematic diagram of a single graphene layer
on Ni (111) surface. (d). Optical image of graphene/
polycrystalline Ni surface after the CVD process. The inset is a
three dimensional schematic diagram of graphene layers on
polycrystalline Ni surface. Multiple layers formed from the grain
boundaries.

52
Figure 3.4 (a-b). Ten typical Raman spectra of graphene grown on Ni (111)
and polycrystalline Ni, respectively. (c). The G’-to-G peak
intensity ratio (I
G’
/I
G
) v.s. the Full Width at Half Maximum
(FWHM) of G’ bands of graphene on both Ni (111) and
polycrystalline Ni.

53
Figure 3.5 (a). Maps of I
G’
/I
G
of 780 spectra collected on a 60*50 μm
2
area
on the Ni (111) surface and (b) 750 spectra collected on a 60*50
μm
2
area on the polycrystalline Ni surface. Corresponding
optical images to Ni (111) Raman map and polycrystalline Ni
Raman map (c and d).

55
xiv

Figure 3.6 Optical image of graphene transferred form Ni (111) (a) and
polycrystalline Ni film (b) to SiO
2
/Si substrate. Corresponding
Raman spectra taken from graphene transferred from Ni (111)
(c) and polycrystalline Ni (d).

57
Figure 4.1 Schematic of the CVD graphene transfer process onto
transparent substrates

65
Figure 4.2 Schematic representation of the energy level alignment (right)
and construction of the heterojunction organic solar cell
fabricated with graphene as anodic electrode: CVD graphene /
PEDOT / CuPc / C
60
/ BCP / Al.
.
66
Figure 4.3 (a) AFM image of a 300 nm Ni film deposited on Si/SiO2
substrate after high temperature annealing. (b) Typical X-ray
diffraction spectrum of annealed Ni film.
.
67
Figure 4.4 (a) Low magnification TEM image of CVD graphene films.
Inset shows a selected area electron diffraction (SAED) pattern
of the few-layer graphene film. (b) Raman spectrum of CVD
graphene film.

68
Figure 4.5 (a) AFM image of a transferred CVD graphene film onto glass
substrate. (b) Cross section measurement of the height of the
CVD graphene. Typical thickness exhibited by the transferred
films is found within the range 1-3 nm. (c) AFM images of the
surface of CVD graphene, ITO and SWNT films on glass. The
scale bar in z-direction is 50 nm for all images.

69
Figure 4.6 (a) Transmission spectra for CVD graphene, ITO and SWNT
films on glass.Photographs showing highly transparent graphene
films transferred onto glass and PET are shown in (c) and (d),
respectively.

71
Figure 4.7 (a) Transmission spectra of CVD graphene with different
sheet resistance (R
Sheet
). (b) Comparison of R
Sheet
vs. light
transmittance at 550 nm for CVD graphene and reduced
GO films reported in the literature

72
xv

Figure 4.8 (a) Photograph illustrating high flexibility of CVD
graphene transferred on a PET flexible substrate. (b) and
(c) AFM images of the surface of CVD graphene and ITO
films on PET, respectively. (d) and (f) Conductance of the
CVD graphene and ITO films on PET substrates under
bending conditions, respectively. The devices used to
monitor the conductance had channel width (W) = 1 mm,
and length (L) = 1 mm. (e) Optical images of CVD
graphene (upper) and ITO (lower) films on PET before
and after being bent at the angles specified in (b) and (c).
Arrows show the direction of the bending

74
Figure 4.9 Logarithmic (up) and linear (down) current density and
power density vs voltage characteristics of CVD graphene
(a) and ITO (b) OPV cells on PET under dark (red traces)
and 100 mW/cm
2
AM1.5G spectral illumination (blue
traces). The output power density of the cells is plotted on
(a) and (b) as open circle traces. The structure of the
devices is given by [CVD graphene / PEDOT / CuPc / C
60

/ BCP / Al] and [ITO / CuPc / C
60
/ BCP / Al] for CVD
graphene and ITO OPVs, respectively
.
78
Figure 4.10 Comparison of the modeled (solid lines) current density
and power density curves of the graphene and ITO devices
obtained from the Shockley equation against the
experimentally (dots) obtained values.
.
82
Figure 4.11 Current density vs. voltage characteristics of CVD
graphene (a) or ITO (b) photovoltaic cells under 100
mW/cm
2
AM1.5G spectral illumination for different
bending angles. Insets show photographs of the
experimental set up employed in the experiments

84
Figure 4.12 (a) Fill factor dependence of the bending angle for CVD
graphene and ITO devices shown in figure 4.12 (b) SEM
images showing the surface structure of CVD graphene
(up) and ITO (down) photovoltaic cells after being
subjected to the bending angles described in figure 4.11.
.

86
xvi

Figure 4.13 Logarithmic (a) and linear (b) current density vs voltage
characteristics of CVD GRAPHENE and ITO photovoltaic cells
on glass under 100 mW/cm
2
AM1.5 spectral illumination.
Structure of the devices is given by [CVD GRAPHENE /
PEDOT / CuPc / C
60
/ BCP / Al] and [ITO / CuPc / C
60
/ BCP /
Al] for CVD GRAPHENE and ITO OPVs, respectively.

88
Figure 5.1 (a) real time TEM imaging of the early stages of nanotube
growth [1]. (b) Schematic representation showing the three
stages of nanotube growth. I: Carbon saturation, II: detachment
of sp
2
graphene and III: reshaping of catalyst and nanotube
elongation.

93
Figure 5.2 (a) AFM image of a-sapphire substrate showing c-axis and the
atomic step direction. (b) AFM image of CVD-grown SWNTs
on a-sapphire. (c) Top (upper) and side (lower) views of the a-
sapphire surface atomic structure. (d) G band Raman intensity
dependence on the polarization angle. (e) cos
2
(θ) fit to the plot of
the normalized G band intensity v.s. the polarization angle
(scattered dots).

97
Figure 5.3 (a) Optical and SEM images of a device, showing aligned
nanotubes between patterned source and drain electrodes. Scale
bar is 2 μm. RBM (b) and G band (c) of several typical
nanotubes scanned with lasers of 785 nm, 633 nm and 532 nm in
wavelength from top to bottom, respectively.

99
Figure 5.4 FESEM images of aligned nanotubes synthesized on unannealed
(a) and annealed (b) a-sapphire. Deviation angles from the
direction [1
1
00] (θ
D
) in nanotube segments aligned in that
direction were typically 10 times lower in nanotubes grown on
annealed sapphire, evidencing higher straightness than
nanotubes grown on unannealed sapphire.

101
Figure 5.5 Representative G’ band spectra for SWNTs grown on
unannealed (upper) and annealed (lower) a-sapphire. Insets: G’
FWHM distribution measured by Lorentzian fitting of Raman
peaks.

102
Figure 5.6 Intensity profile of the Raman RBM (a) and G band (b) with
respect to the scan position of aligned CNTs grown on
unannealed a-plane sapphire. RBM peaks have been circled for
clarity. (c) and (d) show Raman intensity profiles of aligned cnts
on annealed sapphire for RBM and G-band regions, respectively.
104
xvii

Figure 6.1 Wafer-scale synthesis of aligned nanotubes on quartz and a-
sapphire wafers. Photographs of the wafers after nanotube
growth are shown on the left. SEM images on the right show
large arrays of highly aligned nanotubes.

110
Figure 6.2 (a) Schematic showing the aligned-nanotube transfer process. (b)
Photograph of a Si/SiO
2
wafer with transferred nanotubes. SEM
image shows that, after being transferred, nanotubes maintain a
good degree of alignment on the receiving substrate. Au
electrodes deposition followed by etching of nanotubes outside
the device channel area completes the fabrication of CNTFETs.

111
Figure 6.3 (a) Schematic diagram showing large arrays of field-effect
transistors comprising of horizontally aligned carbon nanotubes
between source and drain electrodes. (b) Photograph of a Si/SiO
2

wafer with fabricated aligned nanotube transistors. The SEM
image shows a typical CNTFET in the arrays. (c) Schematic
diagram illustrating the scalable light irradiation process. (d)
Current vs. gate voltage (I
DS
-V
G
) characteristics of a CNTFET
device, obtained with V
DS
=0.5 V before (black trace) and after
(red trace) light irradiation. The I
On
/I
Off
ratio increased from ~64
to ~10
5
in the nanotube transistor due to the light irradiation.-
mediated oxidation of nanotube sidewalls leads to metal-to-
semiconductor transition.

113
Figure 6.4 AFM image of single-nanotube CNTFET that shows increase of
I
On/
I
Off
after 5 hours of light exposure and converting into a
semiconducting device. Zoomed AFM image shows no visible
damage or cutting on the nanotube structure. (Right) I
On/
I
Off

evolution of the CNTFET shown on the left upon timed light
irradiation

114
Figure 6.5 Light-mediated oxidation of nanotube sidewalls leads to metal-
to-semiconductor transition. (a) Raman D band (left) and G band
(right) of a metallic nanotube in a single-nanotube device before
and after light irradiation. A 5-fold decrease in the I
G
/I
D
ratio
shows that increased defect density on the nanotube sidewalls
was achieved due to light irradiation. (b) I
DS
-V
G
characteristics
of the single-nanotube device shown in the inset of (a), before
and after light irradiation. The I
On
/I
Off
improved from 1.6 to 10.5,
indicating metal to semiconductor transition.

116
xviii

Figure 6.6 (a) Comparison between the G/D ratios of nanotubes before and
after one hour irradiation with the full spectrum, ultraviolet, (250
nm - 400 nm), visible (380 nm – 700 nm) and near infrared (750
nm – 2000 nm). (b) Percentage of nanotubes exhibiting Raman
RBM signal after light irradiation using the same irradiation
conditions as part (c). The inset shows the decrease of RBM
intensity was more significant for the small-diameter nanotube
than for larger nanotubes.

117
Figure 6.7 (a) Schematic showing light-induced oxidation of the nanotube
sidewalls and possible chemical groups introduced on the
nanotube sidewalls upon sp
2
–sp
3
rehybridization by light-
induced oxidation. (b) Comparison of typical I
DS
-V
G

characteristics of two CNTFETs before and after irradiation in
air and in vacuum. The device irradiated in air became fully
depletable, showing an improvement in I
On
/I
Off
from 1.7 to 100,
while the device irradiated in vacuum exhibited virtually no
change in its transistor characteristics.

118
Figure 6.8 FTIR spectra of carbon nanotubes before (black) and after (red)
light exposure at open environmental conditions .

120
Figure 6.9 Influence of the irradiation time and nanotube diameter on the
metal-to-semiconductor conversion observed in CNTFETs. (a, b
and c) I
On
and I
Off
of single- and few-nanotube CNTFETs
showing different evolutions under timed light irradiation (V
ds
=
100 mV). (d) Histogram of CNTFETs that lost electrical
conduction or survived after six-hour light irradiation plotted
versus the nanotube diameter. Clear diameter dependence was
observed. (e) Percentage of CNTFETs that survived (red) and
showed depletable behavior (black) for different light irradiation
durations. The best yield was found after 4-hour exposure, when
the percentage of depletable devices increased from 32% to 88%
while keeping a survival ratio near 81%.
.
122
Figure 6.10

Stacked histograms showing the number of nanotubes exhibiting
RBM vs. the RBM frequency, before and after light irradiation,
as measured with three excitation lines (532 nm, 633 nm, and
785 nm). Frequency regions characteristic for metal or
semiconductor nanotubes are highlighted based on Kataura’s
plot. Comparison of the histograms obtained before and after
irradiation for each laser shows a predominant light-induced
oxidation of small-diameter nanotubes (large RBM frequency).

125

xix

Figure 6.11

Percentage of metallic nanotubes in as-grown samples before
(gray columns) and after (red columns) light exposure using
xenon (upper panel) and halogen (lower panel) lamps.
Nanotubes were grouped into two categories based on their
diameter: small-diameter (0.7 - 1.3 nm) and large-diameter (1.4 -
2.0 nm) nanotubes. A substantial decrease in the percentage of
small-diameter metallic nanotubes found after light irradiation,
for both light sources employed, indicates their preferential
oxidation over semiconducting small-diameter nanotubes.
Contrarily, the percentage of large-diameter metallic nanotubes
was largely unaffected by light, indicating the preferential
oxidation (metal over semiconductor) is more effective for
small-diameter nanotubes.

126
Figure 7.1

Figure 7.2

Anisotropic etching of graphene along crystalline orientations of
the graphene lattice. Etching is attributed to Cu clusters that are
present in the CVD chamber as impurities left behind due to
graphene synthesis on Cu substrates in the same CVD chamber.
Image on the right corresponds to a zoomed region from the
image on the left. Yellow arrows show the anisotropic etching of
graphene by particles during thermal annealing.

Strained graphene. (a) Schematic representation of the effect of
uniaxial tensile stress on a graphene supercell. The dashed
(solid) lattices indicate the unstrained (strained) graphene.
Calculated band structure of unstrained (b) and 1% tensile
strained (c) graphene [18]. A band gap is clearly seen on the
band structure of strained graphewas largely unaffected by light,
indicating the preferential oxidation (metal over semiconductor)
is more effective for small-diameter nanotubes. (d) Transfer
characteristics under back gate bias for unstrained and strained
graphene microribbon FETs. (e) Schematic representation of
uniaxially strained graphene by wet etching of underlying SiO
2

etching. Inset shows a typical strain observed on graphene FET
channel upon wet trench etching.

136

138
xx

Abstract

Graphene and carbon nanotubes have outstanding electrical and thermal
conductivity. These characteristics make them exciting materials with high potential to
replace silicon and surpass its performance in the next generation of semiconductors
devices, such devices ought to be considerably smaller and faster than the ones used in
present technology. Despite of the excellent electrical and thermal conduction properties
of graphene and carbon nanotubes, the advance of nanoelectronics based on them has
been hampered due to fundamental limitations of the current synthesis and integration
technologies of these carbon nanomaterials. Therefore, there is a strong need to do
research at fundamental and applicative levels to help find the roadmap that these
materials need to follow, in order to become a real alternative for silicon in future
technologies.
This dissertation present our approach to overcome some of the most critical
problems that hinder the implementation of graphene and carbon nanotubes as important
components in real-life macro and nanoelectronic devices. Towards this end, we
systematically studied synthesis methods for scalable, high quality graphene and
evaluated our large-scale synthesized graphene as transparent electrodes in functional
energy conversion devices. In addition, we explored scalable methods to obtain carbon
nanotube field-effect transistors with only semiconductor nanotube channels and studied
the substrate influence on the structure and metal to semiconductor ratio of aligned
nanotubes. Although we have successfully tackled some of the most important challenges
xxi

of the above-mentioned one- and two-dimensional carbon nanostructures, more remains
to be done to integrate them as functional components in electronic devices to reach the
goal of transferring them from the laboratory to the manufacturing industry, and
ultimately to the society.
In chapter 1, a general introduction to carbon nanomaterials is presented, followed
by a more focused discussion on the structure and properties of graphene and carbon
nanotubes. Chapter 2, presents the development of a chemical vapor deposition method
for scalable graphene synthesis and the evaluation of its electrical properties as the active
channel in field effect transistor and as a transparent conductor. Chapter 3 presents further
work on graphene synthesis on single crystal nickel and the influence of the substrate
atomic arrangement on the synthesized graphene. Chapter 4 presents the implementation
of the highly scalable graphene synthesized by CVD as the transparent electrode in
flexible organic photovoltaic cells. Chapter 5 evaluates the influence of
substrate/nanotube interactions during align nanotube growth on the Raman signature of
the resulting aligned nanotubes, nanotube structure and metal to semiconductor ratio.
Chapter 6 presents our findings on a scalable method that can be used at wafer scale to
achieve metal to semiconductor conversion of carbon nanotubes by light irradiation and
its application to achieve semiconducting CNTFETs. Finally, in chapter 7, future research
directions in related areas of science and technology are proposed.
1

Chapter 1 Introduction

Faster and smaller computers, smarter medicaments, ultrasensitive sensors, the
dream of a new generation of products that are cleaner, lighter, stronger and more
efficient; those are aspects that represent the aspirations of a great part of human kind that
increasingly strives for better technologies. Interestingly, the concept of nanotechnology
is at the center of this discussion. Nanotechnology has become instrumental on finding
pathways to reach the processes and products that are on demand today and will become
even more essential for the society of the future.
Nanotechnology can be defined as the understanding and manipulation of matter
with at least one dimension of the order of 1 to 100 nanometers, where unique
phenomena enable novel applications. For example, whereas elemental carbon is a poor
conductor of electricity and not particularly strong, the two-dimensional carbon is a
semimetal that presents a high charge carrier mobility, obeying the laws of relativistic
quantum mechanics rather than regular quantum mechanics; furthermore, one-
dimensional carbon has mechanical strength 100 times higher than steel, presenting either
metallic or semiconducting nature depending on their chiral arrangement.
Two principal factors cause the properties of nanomaterials to differ significantly
from bulk materials: increased relative surface area, which can change or enhance
chemical reactivity [1]; and quantum effects that can affect the material optical, magnetic
and electrical properties [2]. It is precisely the collection of new and surprising properties
of nanomaterials what has motivated the scientific and engineering community to
dedicate a tremendous amount of effort to improve the understanding of their physical
2

and chemical properties, as well as to look for controllable synthesis and accurate
characterization techniques. The unique properties of the various types of nanomaterials
makes them highly desirable for applications in commercial, medical, military, and
environmental sectors.
Table 1.1, Classification of nanomaterials by dimensionality and composition.

In general, nanomaterials can be classified in two different ways (see table 1.1).
The first form of classification groups them according to their dimensional confinement.
They can be regarded as zero-dimensional, one-dimensional and two-dimensional
nanomaterials. Each of them has diverse applications, for example, a two dimensional
nanomaterial such as a multi layered quantum well is promising for next generation solar
cells with high conversion efficiency [3, 4]. Carbon nanotubes, one type of one-
dimensional quantum wires, are expected to play an important role in next generation
electronics, since it offers one of the highest mobility among electronic materials [5-9].
On the other hand, the zero-dimensional nanomaterial quantum dot, which is confined in
Nanomaterials
Dimension
0-D 1-D 2-D
• Nanocrystals
- Metal
- Semiconductor (Quantum dots)
- Oxides
• Buckminsterfullerene
• Hollow spheres
• Nanowires
• Nanorods
• Carbon nanotubes
• Nanofibers
• Supported thin films
• Graphene
• Quantum wells
Composition
Dendrimers Composites Metal-based
Nanomaterials
Carbon
Nanostructures
• Low molecular weight
- Dendrimers
- Dendrons
• High molecular weight
- Dendronized polymers
- Hyperbranched polymers
- Polymer brushes
•Matrix type
- Ceramic composites
- Polymer composites
- Metal composites
• Metal colloids
• Metal oxides
• Quantum dots
• Fullerenes
• Carbon Nanotubes
• Graphene
• Carbon nanohorns
3

three dimensions, has found important use as an alternative to the conventional organic
dye molecules for in-vivo imaging and diagnosis [10, 11].
The second form of classification groups nanomaterials as dendrimers,
composites, metal-based and carbon-based nanomaterials. Dendrimers are synthetic
macromolecules of nanometer dimensions built from branched units that usually extend
from a common core. As a consequence of their unique structural topology and chemical
versatility, dendrimers have found applications as highly efficient catalysts [12] and drug-
delivery applications [13]. Composites are combinations of nanoparticles with other
nanoparticles or with larger, bulk-type materials. New interface surface interactions
enhance the material’s mechanical and thermal properties, including quantum effects [14].
Metal-based nanomaterials include metallic and semiconductor nanoparticles confined in
at least one dimension. Quantum effects associated with dimensional confinement
determine their properties [15], finding applications in fields as spectroscopy, microscopy
and plasmonics, among others [16-19]. Carbon-based nanomaterials are composed
mostly of carbon, most commonly taking the form of a hollow spheres, ellipsoids, tubes
or sheets. One-atom-thick spherical and ellipsoidal carbon nanomaterials are referred to
as fullerenes, while carbon nanostructures in the form of cylinders and sheets are called
nanotubes and graphene, respectively. These structures have many fascinating and
widespread potential applications, including improved films and coatings, stronger and
lighter materials, and applications in electronics.
In this thesis, we will circumscribe to carbon related nanomaterials of one and two
dimensions, i.e. carbon nanotubes and graphene. Charge carrier mobility in these
nanostructures is by far the highest of any known material, which makes them ideal
4

candidates for future electronics applications, especially at the nanoscale. On the other
hand, because of the continuing demand for miniaturization, cost reduction by device
scaling and technology optimization, bulk silicon technology faces the difficult challenge
to scale down to molecular size scales, and still work effectively. Therefore, research to
discover and integrate new functional materials that bring excellent performance to
semiconductor devices at the nanoscale is in high demand.

1.1 Introduction to carbon nanostructures
Carbon is an interesting and important element not only because it forms millions of
organic compounds with other elements, but also because of its capability to form a
variety of allotropes. Diamond and graphite are the two well known allotropes of carbon
from ancient times. Fullerenes, the third form of carbon were discovered in 1985 [20] and
carbon nanotubes in 1991[21] .Though it was realized in 1991 that carbon nanotubes
were formed by rolling a 2D graphene sheet, the isolation of graphene was quite elusive,
resisting any attempt on its experimental work until 2004 [22]. Thus, only three-
dimensional (diamond and graphite), one-dimensional (nanotubes) and zero-dimensional
(fullerenes) allotropes of carbon were known.

1.1.1 Graphene
Despite its lack of isolation, graphene is the best theoretically studied allotrope of
carbon. One of the greatest concerns on graphene research from 1934 was the idea that a
strictly 2D crystal could not freely exist. Some studies had concluded that these crystals
were thermodynamically unstable and could not exist [23]. It was established that, in the
5

standard harmonic approximation, thermal fluctuations will destroy long-range order,
resulting in melting of a 2D lattice. It was also presumed that during synthesis, any
graphene nucleation sites will have large perimeter-to-surface ratios, thus favoring the
formation of other carbon allotropes.
Dimension 0-D 1-D 2-D 3-D
Isomer Fullerene Nanotube Graphene Diamond
Hybridization
sp
2
sp
2
sp
2
sp
3
Density 1.72 1.2-2.0 2.26 3.515
Bond length
1.40 (C=C)
1.46 (C-C)
1.44 (C=C) 1.42 (C=C) 1.54 (C-C)
Electronic
properties
Semiconductor
E
g
=1.9 eV
Metal/Semiconductor
Eg= ~0.3 – 1.1 eV
Zero-gap
semiconductor
Insulator

Table 1.2, Important parameters of carbon materials of different dimensionalities

Graphene is the building block for carbon materials of all other dimensionalities
and therefore the mother of all graphitic materials [24]. It is interesting that carbon with
sp
2
hybridization is able to form the two-dimensional graphene, the planar local structure
in the closed polyhedral of the fullerene family and the cylinder-shaped carbon nanotubes,
all with different physical properties (see table 1.2). Thus, keeping the sp
2
hybridization,
the 2D carbon can be wrapped up into 0D fullerenes, rolled into 1D nanotubes, or stacked
into 3D graphite, as can be seen in figure 1.1.

1.1.2 Electronic structure of two-dimensional carbon: Graphene
Graphene sheets are one-atom thick, 2D layers of sp
2
-bonded carbon. The 3D
material graphite is a layer-structured material and is composed of several layers of
6

graphene that are held together by Van der Waals forces, with an interlayer spacing of
3.35 Å. Carbon has four electrons in its valence level with a configuration of 2s
2
2p
2
.

Figure 1.1 Graphene is a 2D building material for carbon materials of all other dimensionalities.
It can be wrapped up into 0D buckyballs, rolled into 1D nanotubes or stacked into 3D graphite.

The energy of the 2s and 2p sublevels is so close that their orbitals hybridize into
sp, sp
2
and sp
3
to give triple, double or single bonds, respectively. The hexagonal
structure of graphene poses an alternate double bond arrangement that makes it perfectly
conjugated in sp
2
hybridization. In this case its p
x
and p
y
orbitals contain one electron
each, and the remaining p
z
has only one electron. This p
z
orbital overlaps with the p
z

7

orbital of a neighbor carbon atom to form a π-bond, while the remaining orbitals form σ-
bonds with other neighboring carbons.
In figure 1.2a and b the unit cell the Brillouin zone of two-dimensional graphite is
shown as a dotted rhombus and shaded hexagon, respectively, where and are unit
vectors in real space, and and are reciprocal lattice vectors. In the x, y coordinates
shown in the figure 1.2, the real space unit vectors and of the hexagonal lattice are
expressed as:
(1.1)
Where a = is the lattice constant of graphene.
Correspondingly the unit vectors and of the reciprocal lattice are given by:
(1.2)
This corresponds to a lattice constant of in reciprocal space. The direction of the unit
vectors and of the reciprocal hexagonal lattice is rotated by 90 degrees from the
unit vectors and of the hexagonal lattice in real space. The shaded area in figure
1.2b shows the first Brillouin zone of graphene, where the three high symmetry points Γ,
Κ and Μ can be defined as the center, the corner, and the center of the edge of the
hexagonal Brillouin zone, respectively. Thus, the energy dispersion relations of graphene
can be calculated from the triangle Γ Κ Μ shown by the dotted lines in Figure 1.2b.
The number of sigma (σ) bonds in carbon atoms with sp
n
hybridization is given
by (n + 1). Therefore in graphene, each carbon atom has three σ bonds that extend in-
plane connecting to each one of its nearest neighbors (Figure 1.2a). The remaining 2p
z

orbital, which is perpendicular to the graphene plane, makes π covalent bonds. π-
8

electrons in graphene are delocalized and are largely responsible for its conduction
properties, while π-orbitals are the most important for determining the solid state
properties of graphene.

Figure 1.2 Unit cell (a) and Brillouin zone (b) of two-dimensional carbon are shown as the
dotted rhombus and the shaded hexagon, respectively. and , (i=1,2) are unit vectors and
reciprocal lattice vectors, respectively. Energy dispersion relations are obtained along the
perimeter of the dotted triangle connecting the high symmetry points Γ, Κ, and Μ.

Bloch functions for two inequivalent carbon atoms A and B, contained in the
hexagonal lattice unit cell; have to be constructed to calculate the band structure of
graphene. In this way, the diagonal matrix elements of the Hamiltonian constructed on
the system will integrate over a single atom, for which H
AA
= H
BB
= E
2p
gives the orbital
energy of the 2p energetic state. However, for the off-diagonal matrix element H
AB
, the
nearest-neighbor B atoms relative to an A atom have to be considered.
(1.3)
Where t is the transfer integral and f(k) is a function that includes the B atoms phase
factors. f(k) is given in Cartesian coordinates by:
(1.4)
A B
y
x
a
1
a
2
k
y
k
x
b
1
b
2
Γ
Κ
Μ
B
A
B
1.42 Å
120̊
9

The solution of the secular equation to determine the eigenvalues E( ) under the tight
binding approximation is given by:
(1.5)
The upper half of the energy dispersion curves describes the π* anti-bonding energy band
and the lower half is the π bonding energy band. Interestingly, the π* anti-bonding and π
bonding bands are degenerate at the Κ points through which the Fermi energy passes.

Figure 1.3 Energy dispersion relations of two-dimensional graphene shown throughout the
Brillouin zone and the inset shows the energy dispersion relations along the high symmetry axes
along the perimeter of the triangle shown in Figure 1.2b.

Since there are two π electrons per unit cell, these electrons fully occupy the
lower π bonding energy band. Thus, a detailed calculation of the density of states shows
that the density of states at the Fermi energy is zero i.e. graphene is a zero-gap
semiconductor. The existence of a zero gap at the K points comes from the symmetry
requirement that the two carbon atoms A and B in the hexagonal lattice are equivalent to
each other. For instance, if the A and B atoms were different either in location or in
k
x
K
y
E
10

electronic configuration, the energy for the E
2p
states would be different for A and B (H
AA

H
BB
) and therefore, the calculated energy dispersion would show an energy gap
between the π and π* states at the K points. Equation 1.5 is commonly used as a simple
approximation for the electronic structure of graphene. It is clear that when the overlap
integral approaches zero, the π and π* states become symmetrical around E = E
2p
and in
the case where E
2p
is taken as zero-point energy, this point is regarded as the Dirac-point.
Figure 1.3 depicts the energy dispersion of graphene around the Dirac point. The
first significant feature of this result is that, since the energy band is exactly symmetric
about the point E = E
2p
= 0
,
and this condition is met only at Dirac point. It follows that
for exactly half filling of the band the DOS at the Fermi level is exactly zero and undoped
graphene is a perfect semimetal.
At zero doping, the lower half of the band (see figure 1.3) is filled exactly up to
the Dirac Point. If by applying a suitable “gate” voltage to the graphene relative to the
substrate we induce a nonzero charge, this is equivalent to injecting a number of electrons
in the upper half or holes in the lower half of the Dirac cones. This effect is known as the
field effect (Figure 1.4). Graphene is found to exhibit a pronounced ambipolar electric
field effect such that charge carriers can be tuned continuously between electrons to holes.
Single layer graphene atop a thermally grown SiO
2
layer on a highly doped Si substrate
may serve as a prototype of a field effect transistor. SiO
2
also serves as an insulating
layer, so a back-gate voltage can be applied to vary carrier concentration.
The early graphene FET devices demonstrated by Novoselov [22] exhibited
dopant concentrations as high as 10
13
cm
–2
and achieved a mobility that could exceed
10,000 cm
2
/Vs. This field effect was found to be practically temperature independent.
11

This translates into ballistic transport on submicron scales. The room-temperature
mobility is limited by impurities or corrugations of the graphene surface, which means
that it can still be improved significantly up to the order of 10
5
cm
2
/Vs as has been
demonstrated by Du [25] and Bolotin [26].
Figure 1.4 Conical low-energy spectrum E(k), indicating changes in the position of the Fermi
energy E
F
with changing gate voltage V
g
. Positive (negative) V
g
induce electrons (holes) in
concentrations n = V
g
where the coefficient 7.2 10
10
cm
-2
V
-1
for field-effect devices with a
300 nm SiO
2
layer used as a dielectric.

1.1.3 Electronic structure of one-dimensional carbon: Carbon Nanotubes
A single-wall carbon nanotube (SWCNT) is defined as a rolled-up sheet of
graphene in which the edges of the sheet are joined together to form a seamless, hollow
cylinder with a diameter of 0.7 - 10 nm, although most of the observed single-wall
nanotubes have diameters of less than 2 nm. If we neglect the two close ends of the
carbon nanotubes, and in virtue of their large aspect ratio (i.e., length/diameter can be ≥
10
3
-10
6
), these nanotubes can be considered as one-dimensional nanostructures.
0
-V

ρ (Ω)
Gate voltage (V)
+V
E
F
E
F
E
F
12

The band structure of graphene is the basis for understanding the electronic
behavior of carbon nanotubes. Every lattice point in the two-dimensional graphene can be
described by multiples (n,m) of two unit vectors a
1
and a
2
, as shown in figure 1.5. The
multiples are also used to described the chiral vector. In figure 1.5, a carbon nanotube can
be formed if we cut the graphene lattice along the dotted lines of OB and AC that are
perpendicular to the edges of C
h
and roll the structure by joining the points B with C and
O with A. The chiral vector C
h
( in figure 1.5), whose length define the nanotube
circumference, will specify the structure of the carbon nanotube.

Figure 1.5 The unrolled nanotube lattice. A nanotube can be constructed by connecting sites O
and A, and B and C. OA and OB define the chiral and translational vectors, respectively. The
rectangle AOCB defines the unit cell for the nanotube. The vector R denotes a symmetry vector.

The vectors OA and OB define the chiral vector C
h
and the translational vector T
that extends along the length of the nanotube, in the k
z
direction, respectively. C
h
can be
y
x
O
B
C
A
k
z
C
h
a
1
a
2
k
┴
θ θ θ θ
n
m
R
T
13

expressed by the real space unit vectors a
1
and a
2
(see figure 1.5) of the graphene
hexagonal lattice:
(n, m are integers, 0 ≤ ≤ n). (1.6)
The specific chiral vector shown in the figure is (4,2), and the unit cell of this nanotube is
formed by the rectangle bounded by OACB.

Figure 1.6 Schematic showing the atomic arrangement of zigzag (12,0), armchair (7,7)and chiral
(9,5) nanotubes.

Different types of CNTs have different values of n and m. As shown in figure 1.6,
an armchair nanotube corresponds to the case when n = m, that is C
h
= (n,n), which gives
Armchair (7,7)
Zigzag (12,0)
Chiral (9,5)
14

a chiral angle of 30 degrees, and a zigzag nanotube corresponds to the case when m = 0,
or C
h
= (n,0), for a chiral angle of 0 degrees. All other chiral vectors correspond to chiral
nanotubes, with chiral angles between 0 and 30 degrees. The indexes n and m are very
important, containing all the information that determines the properties of the CNTs. For
instance, the diameter of the nanotube can be obtained by L/π, in which L is the
circumferential length of the nanotube:
(1.7)
Figure 1.5 shows two important wavevectors: k
z
, which goes in the direction of
the translation vector T parallel to the nanotube axis, and k
┴,
which goes along the
circumferential direction of the nanotube. By using boundary conditions to ensure the
periodicity on the band structure along the chiral vector C
h
, the wavevector k
┴
associated
with the C
h
direction becomes quantized, while the wavevector k
z
associated with the
translational vector T remains continuous for a nanotube of infinite length. Thus the
energy bands consist of a set of one-dimensional energy dispersion relations which are
cross sections of those for two-dimensional graphene (Figure 1.7).
If we translate these wavevectors to reciprocal lattice vectors and denote them as
K
1
(for the circumferential direction) and K
2
(for the nanotube axis direction), we can
define the energy dispersion relations for CNTs as:
, (1.8)
Where T is the magnitude of the translational vector T, k is a 1D wave vector along the
nanotube axis, and N denotes the number of hexagons within the nanotube unit cell. The
N pairs of energy dispersion curves given by equation 1.8 correspond to the cross section
15

of the two-dimensional energy dispersion surface shown in figure 1.7, where cuts are
made on the lines . If for a particular nanotube (n,m) the cutting passes
through a high symmetry K point of the Brillouin zone (Figure 1.3), where the π and π*
energy bands are degenerate by symmetry, the one-dimensional energy bands have a zero
energy gap and the nanotube will be metallic.

Figure 1.7 Representation of the Brillouin zone for graphene and carbon nanotube lattice. The
condition for the obtention of metallic energy bands is met if the ratio of the length of the vector
to that of K
1
is an integer.

Otherwise, the nanotube band structure would have a bandgap and
exhibit semiconducting behavior.

Contrarily, if the cutting line does not pass through the K point, then the energy
bands will have a finite energy gap between the
valence and conduction band, and the nanotube is expected to exhibit semiconducting
behavior. To obtain a metallic nanotube, the ratio of the length of the vector to that of
K
1
in figure 1.7 is an integer. is given by , the condition for metallic
nanotubes is that (2n+m) or equivalently (n-m) is a multiple of 3. From this, it follows
that approximately one third of the CNTs are metallic and the other two thirds are
semiconducting.
E
k
E
GAP
E
k
Semiconducting Metallic
K’
K
M
K’
M
K
2
K
1
Γ
W
k
y
k
x
Y
16

The density of states (DOS) of carbon nanotubes can be easily calculated
following a tight binding calculation approach from the energy dispersion relations. It is
given by:
(1.9)
Figure 1.8 shows the band structure and the density of states of a semiconducting
(9,5) and a metallic (7,7) nanotube. The density of states near the Fermi level E
F
is
constant, with a finite value for the metallic nanotubes while itis reduced to zero for
semiconducting nanotubes.

Figure 1.8 1-D Energy bands (a and c) and 1-D electronic density of states (b and d) for a
semiconducting (blue) and metallic (red) carbon nanotube. The semiconducting nanotube with
(n,m) = (9,5) clearly shows a bandgap in the band structure and absence of energy states in the
bandgap. However, the metallic nanotube (7,7) exhibits a continuous band structure with no
bandgap and a finite density of states at the Fermi energy.

As can be seen from the (7,7) nanotube, the density of states is small near the
Fermi energy, typically nearly four orders of magnitude smaller than that of metals,
0 0.2 0.4 0.6 0.8
-0.5
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
k
z
/[pi/T] normalized
E(k
x
,k
y
)/γ
0
[eV]
0 0.2 0.4 0.6
-3
-2
-1
0
1
2
3
Density of states (E) [eV cm]-1
Energy (eV)
(9,5) (7,7)
0 0.2 0.4 0.6 0.8
-3
-2
-1
0
1
2
3
k
z
/(pi/T) normalized values between [0,1]
E(k
x
,k
y
)/γ
0
[eV]
0 5 10 15
-3
-2
-1
0
1
2
3
Density of states (E) [eV cm]-1
Energy (eV)
0 2 4
2
4
6
8
10
12
14
16
18
i
E
ii
(eV)
(d) (a) (b) (c)
17

which in principle would make them poor conductors. However, the one-dimensional
nature of the nanotubes leads to an unusually high mean free path of the charge carriers
of the order of 1-10 μm, in contrast to that of metals, which is around 40 μm for copper.
This phenomenon can be associated with the perfect chemical structure and the
one-dimensional nature of the carbon nanotube. In bulk materials, charge carriers can be
scattered in every possible direction and there are states available for these scattering
events, contrarily in nanotubes the only states available for charge conduction are
quantized and in the direction along the nanotube axis for which the phase space for
scattering is greatly restricted, and at low bias the only energetically allowed state to be
scattered to be scattered to is backwards in the opposite direction and not favored [27].

1.2 Graphene and carbon nanotubes for beyond silicon electronics
Graphene and carbon nanotubes have outstanding electrical and thermal
conductivity. These characteristics make them exciting materials with high potential to
replace silicon and surpass its performance in the next generation of semiconductors
devices, such devices ought to be considerably smaller and faster than the ones used in
present technology. Despite of the excellent electrical and thermal conduction properties
of graphene and carbon nanotubes, the advance of nanoelectronics based on them has
been hampered due to fundamental limitations of the current synthesis and integration
technologies of these carbon nanomaterials. Therefore, there is a strong need to do
research at fundamental and applicative levels to help find the roadmap that these
materials need to follow, in order to become a real alternative for silicon in future
technologies.
18

In the case of graphene, electrons behave like massless relativistic particles,
which govern most of its electronic properties. One of the most important consequences
of such unusual dispersion relation is the observation of half-integer Quantum Hall Effect
and the absence of localization, which can be very important for graphene-based field
effect transistors [24]. Mechanical exfoliation of highly ordered pyrolitic graphite
(HOPG) or high purity graphite flakes can lead to obtain graphene crystals with very few
defects, which in turn exhibit high mobility of the charge carriers. Figure 1.9 shows
scanning electron microscopy (SEM) and atomic force microscopy (AFM) of the
graphene-based device reported in the literature as having the highest electron mobility to
date [26]. The graphene film was obtained by mechanical exfoliation of graphite on
Si/SiO
2
substrate in which the oxide layer underneath the graphene was etched in order to
obtain a free-standing graphene flake connecting the metal electrodes.
Electrical measurements of resistivity vs. gate voltage show the intrinsic
ambipolar behavior of graphene. It was also established that the transfer characteristics of
the device is greatly improved after undergoing a high-current annealing process to
remove contaminants from the graphene surface. The mobility μ for this device reaches
the outstanding value of 230,000 cm
2
/Vs measured at the highest carrier density n =
2x10
11
cm
-2
. Such high μ would in principle favor high frequency performance.
19

Figure 1.9 (a) SEM image of a typical suspended six-probe graphene device. AFM image of the
suspended device before (b) and after (c) the measurements and graphene removal oxygen plasma
etch. (d) Device schematic, side view, suspended single-layer graphene is colored pink. (e)
Measured four-probe resistivity as a function of gate voltage before (blue) and after (red) current
annealing. Data from traditional high-mobility device on the substrate is plotted with gray dotted
line. (f) Mobility μ = as a function of carrier density n for the same devices.

Early graphene devices pursuing high frequency demonstrated encouraging
characteristics, exhibiting a cutoff frequency f
T
of 26 GHz, which is the frequency at
which the current gain becomes unity and signifies the highest frequency at which signals
are propagated [28]. Only recently, P. Avouris and collaborators reported the fabrication
of graphene FETs on SiC substrates with cutoff frequency of 100 GHz for a device of
gate length of 240 nm and using a source-drain voltage of 2.5 V [29]. This f
T
exceeds
those previously reported for graphene FETs [28, 30, 31] as well as those for Si metal-
oxide semiconductor FETs for the same gate length (~40 GHz at 240 nm) [29].
Another intrinsic property of graphene is its transparency. A single sheet of
graphene absorbs only 2.3 % of the incident light. Such combination of high conductivity
and low light absorption makes this material an ideal candidate as a transparent
conductive film. It is very tempting to use the unique properties of graphene for
technology applications even beyond graphene FET applications. Composite materials,
20

photodetectors, support for biological samples in TEM, mode-lockers for ultrafast lasers
and many more areas would gain strongly from using graphene for non-FET purposes.
The problem, however, is the mass-production of graphene. The technique of
choice for the great majority of researchers is the mechanical exfoliation of graphene
flakes from graphite [22] and that method is able to produce only research-size graphene
samples. Consequently, the implementation of graphene as a viable technology for
beyond-silicon electronics has been hampered due to the difficulty of producing high-
quality single or few-layers specimens over large areas. The scientific community is
employing a lot of effort in the development of technologies for mass production of
graphene; and it is obvious that the development of a synthesis technology that allows
truly mass production of high quality graphene will constitute a gigantic step forward that
would give us realistic hope that soon we will enjoy products based on this exciting two-
dimensional material.
Numerous methods has been proposed to obtain single-layer or few-layer
graphene (FLG) at large scale [32-35], however, the methods proposed so far either not
scalable, produce thick graphite, or highly defective graphene layers, or the cost of
graphene production is so high that it becomes prohibitive for mass production.
Therefore, approaches that can provide high-quality single- and few-layer graphene over
large areas and the evaluation of their electronic properties are still desired to meet
realistic applications.
In the case of carbon nanotubes, one of their most promising applications in
nanoelectronics is as the semiconducting channel in field-effect transistors (CNTFETs).
CNTFETs have acquired great importance to replace silicon due mainly to two reasons:
21

one is their potential to switch on and off much faster than current silicon technologies
[36-38] and the other is the foreseen limits in the downscaling of silicon transistors [39,
40]. Electronic devices based on nanotube technology have evolved rapidly since the
early single-nanotube devices demonstrated in 1998. However, in spite of the significant
progress made toward integrated nanotube circuits [36, 41-43], the assembly and
integration of nanotube electronics still faces significant challenges. These challenges can
be regarded as the technology components that need to be controlled so that nanotube
devices are produced at a reasonable cost and with repeatable performance. For
nanotubes to become a real alternative to replace silicon in an electronic device there
should be a complete control of their orientation, position, density, diameter and
ultimately their chirality.
Several researchers, including our group, have devoted a great effort to try to
solve these problems. Important advance has been made to control the nanotube
orientation and positioning, which brings the advantage of registration-free fabrication.
This has been achieved by allowing carbon nanotubes to grow, by chemical vapor
deposition, along step edges or preferential crystalline orientations on quartz and sapphire
substrates [44-46], which yields parallel arrays of nanotubes on the substrates (figure
1.10). Furthermore, the successful control of CNTs positioning on the device channel was
achieved by a synergic combination of fabrication techniques and CVD synthesis of
nanotubes [47, 48].
22

Figure 1.10 (a) Schematic diagram of a SWNT aligned on the a-plane sapphire substrate. The red
and blue spheres represent oxygen and aluminum atoms, respectively. The purple plane shows the
a-plane orientation. (b) AFM image of as-grown SWNTs oriented in parallel on the a-plane
sapphire substrate.

Another critical aspect to control the device current carrying density is the density
of the nanotubes in the channel and it can also be controlled by accurately controlling
catalyst density and synthesis conditions. But regardless of the advance on the above-
mentioned components, there is not a reliable method that allows to accurately
controlling the diameter and chirality of the synthesized nanotubes to date. The presence
of approximately one third of metallic nanotubes in nanotube samples represents a
serious drawback toward the fabrication of nanotube-based high performance devices.
Therefore, the obtention of semiconducting or metallic nanotube only samples remains as
a milestone to achieve.
Different approaches have been followed to obtain CNTFETs containing only
semiconducting nanotubes in the channels either by selective synthesis [49-51], post-
synthesis separation methods [52-55], or post-synthesis methods to selectively etch
metallic nanotubes. A significant advance towards this end has been the implementation
of a solution-based density gradient ultracentrifugation (DGU) method, in which CNTs
b a
23

are added to suitable surfactant solutions and centrifuged at ultra-high speeds in a density
gradient media [56]. As a result, carbon nanotubes of different density settle at different
heights of the centrifuge tube according to their buoyancy, which is directly correlated to
their diameter. Even more surprising is the fact that it is possible to obtain fractions with
high enrichment of nanotubes according to their electrical nature, i.e. metallic or
semiconducting nanotubes, by using a convenient mixture of surfactants that have
different affinities for semiconducting and metallic nanotubes (see figure 1.11).
Following this approach, CNEFETs with semiconducting channels have been
demonstrated [57]. However, disadvantages such as a low separation yield, higher
resistivity due to numerous defects introduced during nanotube dispersion as well as the
lack of position and orientation control of the nanotubes in the channel limits the
applications of this kind of separated nanotubes in nanoelectronics.

Figure 1.11 Separation of carbon nanotubes by density gradient ultracentrifugation. (a) Colored
bands in the centrifuge tube correspond to separated metallic and semiconducting fractions. Inset
shows a schematic representation of surfactants wrapping a carbon nanotube. (b) UV-Vis-NIR
absorption and (c) Raman spectra of the separated metallic (black) and semiconducting (red)
fractions.
Recently, significant advance has been made using sub-monolayer films of
randomly grown nanotube networks for flexible devices and circuits. This approach
400 500 600 700 800 900 1000
0.0
0.2
0.4
0.6
0.8
1.0
Absorbance (Normalized)
Wavelength (nm)
M
SC
M
11
S
22
S
33
1300 1400 1500 1600 1700 1800

Raman shift (cm
-1
)
Semiconducting
Metallic
a b c
24

etches stripe regions that go along the charge transport direction in the CNTFET channel
in order to interrupt the heterogeneous percolative transport, between source and drain
electrodes, that would otherwise be dominated by the metallic nanotube pathway. This
approach, though effective for devices with channel length (Lc) ~10-10
2
μm, cannot be
easily scaled to sub-micron regimes; therefore, it is not very useful for nanoelectronics.

Figure 1.12 Electrical breakdown of CNTFETs. (a) I
ds
-V
g
curves for a typical transistor after
consecutive electrical breakdown. Inset shows a diagram of a back-gated transistor built on
aligned carbon nanotubes. (b) On current vs On/Off current for multiple CNTFET devices before
and after electrical breakdown.

The most effective method, suitable for nanoelectronics, to obtain CNTFETs with
semiconducting channels so far is by the use of electrical breakdown of metallic
nanotubes in the channel [58]. This method consists of applying a sufficiently high
positive gate voltage to the nanotube channel to “turn off” the semiconducting nanotubes
while, simultaneously, a high voltage is applied between the source and drain electrodes.
As a result, the high current passing through the metallic nanotubes in the channel
produces electrical stress and heat that burns such metallic nanotubes in air (figure 1.12).
10
-6
10
-5
10
-4
I
ds
(A )
-10 -5 0 5 10
V
g
(V)
Before Breakdown
1st Breakdown
2nd Breakdown
3rd Breakdown
10
-8
10
-7
10
-6
10
-5
10
-4
I
o n
(A )
10
1
10
2
10
3
10
4
I
on
/I
off
+ Before electrical breakdown
+ After electrical breakdown
a b
25

Figure 1.12 shows a typical CNTFET with channel width W = 100 μm and length
L = 0.75 μm that underwent electrical breakdown using a gate voltage V
g
= 15 V and
source-drain voltage V
ds
from 0 to -35 V. As a result, the device I
on
/I
off
increased from 2
to nearly 1000 after several cycles of the breakdown process, but the CNTFET I
on

decreases nearly one order of magnitude, therefore a compromise should be reached
between the conductivity of the devices and their I
on
/I
off
. The decrease of the CNTFET
conductance due to electrical breakdown, although acceptable, can degrade the
performance of nanoelectronics devices. Furthermore, what is more critical is that due to
the inherent polyvalent nature of nanotubes in the channel (different diameter and
chirality) the electrical breakdown process needs to be carried out device by device,
which represents a serious limiting hurdle for scalability.
In this thesis, we present our approach to overcome some of the most critical
problems that hinder the implementation of graphene and carbon nanotubes as important
components in real-life macro and nanoelectronic devices. Towards this end, we
systematically studied synthesis methods for scalable, high quality graphene and
evaluated the substrate influence on the structure of aligned nanotubes. In addition, post-
synthesis methods for metal to semiconductor conversion of CNTFETs and the
evaluation of large-scale graphene in functional energy conversion devices were explored
as well. Although we have successfully tackled some of the most important challenges of
the above-mentioned one- and two-dimensional carbon nanostructures, more remains to
be done to integrate them as functional components in electronic devices to reach the goal
of transferring them from the laboratory to the manufacturing industry, and ultimately to
the society.
26

1.3 Thesis outline
In chapter 1, a general introduction to carbon nanomaterials is presented, followed
by a more focused discussion on the structure and properties of graphene and carbon
nanotubes. Chapter 2, presents the development of a chemical vapor deposition method
for scalable graphene synthesis and the evaluation of its electrical properties as the active
channel in field effect transistor and as a transparent conductor. Chapter 3 presents further
work on graphene synthesis on single crystal nickel and the influence of the substrate
atomic arrangement on the synthesized graphene. Chapter 4 presents the implementation
of the highly scalable graphene synthesized by CVD as the transparent electrode in
flexible organic photovoltaic cells. Chapter 5 evaluates the influence of
substrate/nanotube interactions during align nanotube growth on the Raman signature of
the resulting aligned nanotubes, nanotube structure and metal to semiconductor ratio.
Chapter 6 presents our findings on a scalable method that can be used at wafer scale to
achieve metal to semiconductor conversion of carbon nanotubes by light irradiation and
its application to achieve semiconducting CNTFETs. Finally, in chapter 7, future research
directions in pertinent areas of science and technology are proposed.
27

Chapter 1 References
1. Arenz, M., et al., The effect of the particle size on the kinetics of CO electrooxidation on
high surface area Pt catalysts. Journal of the American Chemical Society, 2005. 127(18):
p. 6819-6829.

2. Yu, H., et al., Two- versus three-dimensional quantum confinement in indium phosphide
wires and dots. Nature Materials, 2003. 2(8): p. 517-520.

3. Nelson, J., et al., Effect of quantum well location on single quantum well p-i-n photodiode
dark currents. Journal of Applied Physics, 1999. 86(10): p. 5898-5905.

4. Mazzer, M., et al., Progress in quantum well solar cells. Thin Solid Films, 2006. 511: p.
76-83.

5. Durkop, T., et al., Extraordinary mobility in semiconducting carbon nanotubes. Nano
Letters, 2004. 4(1): p. 35-39.

6. Kang, S.J., et al., High-performance electronics using dense, perfectly aligned arrays of
single-walled carbon nanotubes. Nature Nanotechnology, 2007. 2(4): p. 230-236.

7. Cao, Q., et al., Medium-scale carbon nanotube thin-film integrated circuits on flexible
plastic substrates. Nature, 2008. 454(7203): p. 495-U4.

8. Ishikawa, F.N., et al., Transparent Electronics Based on Transfer Printed Aligned
Carbon Nanotubes on Rigid and Flexible Substrates. Acs Nano, 2009. 3(1): p. 73-79.

9. Ryu, K., et al., CMOS-Analogous Wafer-Scale Nanotube-on-Insulator Approach for
Submicrometer Devices and Integrated Circuits Using Aligned Nanotubes. Nano Letters,
2009. 9(1): p. 189-197.

10. Michalet, X., et al., Quantum dots for live cells, in vivo imaging, and diagnostics.
Science, 2005. 307(5709): p. 538-544.

11. Ballou, B., et al., Noninvasive imaging of quantum dots in mice. Bioconjugate Chemistry,
2004. 15(1): p. 79-86.

12. Astruc, D. and F. Chardac, Dendritic catalysts and dendrimers in catalysis. Chemical
Reviews, 2001. 101(9): p. 2991-3023.

13. Patri, A.K., I.J. Majoros, and J.R. Baker, Dendritic polymer macromolecular carriers for
drug delivery. Current Opinion in Chemical Biology, 2002. 6(4): p. 466-471.

14. Sukpirom, N. and M.M. Lerner, Rapid syntheses of nanocomposites with layered
tetratitanate using ultrasound. Materials Science and Engineering a-Structural Materials
Properties Microstructure and Processing, 2003. 354(1-2): p. 180-187.

15. Leobandung, E., et al., Observation of Quantum Effects and Coulomb-Blockade in Silicon
Quantum-Dot Transistors at Temperatures over 100 K. Applied Physics Letters, 1995.
67(7): p. 938-940.
28

16. Mock, J.J., et al., Shape effects in plasmon resonance of individual colloidal silver
nanoparticles. Journal of Chemical Physics, 2002. 116(15): p. 6755-6759.

17. Rycenga, M., et al., Understanding the SERS Effects of Single Silver Nanoparticles and
Their Dimers, One at a Time. Journal of Physical Chemistry Letters. 1(4): p. 696-703.

18. Billot, L., et al., Error signal artifact in apertureless scanning near-field optical
microscopy. Applied Physics Letters, 2006. 89(2).

19. Gomez, L., et al., Apertureless scanning near-field optical microscopy: a comparison
between homodyne and heterodyne approaches. Journal of the Optical Society of
America B-Optical Physics, 2006. 23(5): p. 823-833.

20. Kroto, H.W., et al., C60: Buckminsterfullerene. Nature, 1985. 318(6042): p. 162-163.

21. Iijima, S., Helical microtubules of graphitic carbon. Nature, 1991. 354(6348): p. 56-58.

22. Novoselov, K.S., et al., Electric Field Effect in Atomically Thin Carbon Films. Science,
2004. 306: p. 666-669.

23. Mermin, N.D., Crystalline Order in Two Dimensions. Physical Review, 1968. 176(1): p.
250.

24. Geim, A.K. and K.S. Novoselov, The rise of graphene. Nat Mater, 2007. 6(3): p. 183-
191.

25. Du, X., et al., Approaching ballistic transport in suspended graphene. Nature
Nanotechnology, 2008. 3(8): p. 491-495.

26. Bolotin, K.I., et al., Ultrahigh electron mobility in suspended graphene. Solid State
Communications, 2008. 146(9-10): p. 351-355.

27. Kreupl, F., Carbon Nanotubes in Microelectronic Applications. Carbon Nanotube
Devices. 2008: Wiley-VCH Verlag GmbH & Co. KGaA. 1-41.

28. Lin, Y.-M., et al., Operation of Graphene Transistors at Gigahertz Frequencies. Nano
Letters, 2008. 9(1): p. 422-426.

29. Lin, Y.M., et al., 100-GHz Transistors from Wafer-Scale Epitaxial Graphene. Science.
327(5966): p. 662-.

30. Meric, I., et al., RF performance of top-gated, zero-bandgap graphene field-effect
transistors. Ieee International Electron Devices Meeting 2008, Technical Digest, 2008: p.
513-516.

31. Moon, J.S., et al., Epitaxial-Graphene RF Field-Effect Transistors on Si-Face 6H-SiC
Substrates. Electron Device Letters, IEEE, 2009. 30(6): p. 650-652.

32. Forbeaux, I., J.-M. Themlin, and J.-M. Debever, Heteroepitaxial graphite on 6H-
SiC(0001): Interface formation through conduction-band electronic structure. Physical
Review B, 1998. 58(24): p. 16396-16406.
29

33. Viculis, L.M., J.J. Mack, and R.B. Kaner, A Chemical Route to Carbon Nanoscrolls.
Science, 2003. 299: p. 1361.

34. Gilje, S., et al., A Chemical Route to Graphene for Device Applications. Nano Letters,
2007. 7(11): p. 3394-3398.

35. Wu, Z., et al., Transparent, Conductive Carbon Nanotube Films. Science, 2004. 305: p.
1273-1276.

36. Bachtold, A., et al., Logic Circuits with Carbon Nanotube Transistors. Science, 2001.
294 p. 1317-1320.

37. Javey, A., et al., Ballistic carbon nanotube field-effect transistors. Nature, 2003. 424: p.
654-657.

38. Dulrkop, T., et al., Extraordinary Mobility in Semiconducting Carbon Nanotubes. Nano
Letters, 2004. 4(1): p. 35-39.

39. Avouris, P., et al., Carbon nanotubes: nanomechanics, manipulation, and electronic
devices. Applied Surface Science, 1999. 141: p. 201-209.

40. Bohr, M.T., Nanotechnology Goals and Challenges for Electronic Applications. IEEE
Transactions on Nanotechnology 2001. 1(1): p. 56-62.

41. Javey, A., et al., Carbon Nanotube Transistor Arrays for Multistage Complementary
Logic and Ring Oscillators. Nano Letters, 2002. 2(9): p. 929-932.

42. Chen, Z., et al., An Integrated Logic Circuit Assembled on a Single Carbon Nanotube.
Science 2006. 311: p. 1735.

43. Cao, Q., et al., Medium-scale carbon nanotube thin-film integrated circuits on flexible
plastic substrates. Nature 2008. 454: p. 495–500.

44. Ismach, A., et al., Atomic-Step-Templated Formation of Single Wall Carbon Nanotube
Patterns. Angewandte Chemie International Edition, 2004. 43(45): p. 6140-6143.

45. Han, S., X. Liu, and C. Zhou, Template-Free Directional Growth of Single-Walled
Carbon Nanotubes on a- and r-Plane Sapphire. Journal of the American Chemical
Society, 2005. 127(15): p. 5294-5295.

46. Liu, X., et al., Diameter Dependence of Aligned Growth of Carbon Nanotubes on a-Plane
Sapphire Substrates. The Journal of Physical Chemistry C, 2008. 112(41): p. 15929-
15933.

47. Lei, B., et al., Raman Characterization and Polarity Tuning of Aligned Single-Walled
Carbon Nanotubes on Quartz. Japanese Journal of Applied Physics. 49(2).

48. Kocabas, C., M. Shim, and J.A. Rogers, Spatially Selective Guided Growth of High-
Coverage Arrays and Random Networks of Single-Walled Carbon Nanotubes and Their
Integration into Electronic Devices. Journal of the American Chemical Society, 2006.
128(14): p. 4540-4541.
30

49. Bachilo, S.M., et al., Narrow ( n,m)-Distribution of Single-Walled Carbon Nanotubes
Grown Using a Solid Supported Catalyst. Journal of the american chemical society,
2003. 125: p. 11186-11187.

50. Li, Y., et al., Preferential Growth of Semiconducting Single-Walled Carbon Nanotubes
by a Plasma Enhanced CVD Method. Nano Letters, 2004. 4(2): p. 317-321.

51. Joselevich, E. and C.M. Lieber, Vectorial Growth of Metallic and Semiconducting Single-
Wall Carbon Nanotubes. Nano Letters, 2002. 2: p. 1137-1141.

52. Chen, Z., et al., Bulk Separative Enrichment in Metallic or Semiconducting Single-Walled
Carbon Nanotubes. Nano Letters, 2003. 3(9): p. 1245-1249.

53. Arnold, M.S., et al., Sorting carbon nanotubes by electronic structure using density
differentiation. Nature Nanotechnology, 2006. 1: p. 6.

54. Chen, F., et al., Toward the Extraction of Single Species of Single-Walled Carbon
Nanotubes Using Fluorene-Based Polymers. Nano Letters, 2007. 7(10): p. 3013-3017.

55. Nish, A., et al., Highly selective dispersion of single-walled carbon nanotubes using
aromatic polymers. Nature Nanotechnology, 2007. 2: p. 640-646.

56. Arnold, M.S., et al., Sorting carbon nanotubes by electronic structure using density
differentiation. Nat Nano, 2006. 1(1): p. 60-65.

57. Wang, C., et al., Wafer-Scale Fabrication of Separated Carbon Nanotube Thin-Film
Transistors for Display Applications. Nano Letters, 2009. 9(12): p. 4285-4291.

58. Collins, P.G., M.S. Arnold, and P. Avouris, Engineering Carbon Nanotubes and
Nanotube Circuits Using Electrical Breakdown. Science, 2001. 292(5517): p. 706-709.

31

Chapter 2 Synthesis, Transfer and Devices of Single- and Few-Layer Graphene
by Chemical Vapor deposition

2.1 Introduction.
Graphene is probably the best theoretically studied allotropic form of carbon. It
consists of a two-dimensional hexagonal arrangement of carbon atoms, with a quasi-
linear dispersion relation, for which the carrier effective mass is very low [1]. As a
consequence, it has a predicted mobility at room temperatures in the order of 10
6
cm
2
/Vs
and an experimentally measured mobility of 15,000 cm
2
/Vs. The high mobility of this
material opens the possibility of ballistic transport at submicron scales [2, 3].
Despite the advances in graphene research, and the numerous foreseen important
applications, implementation of graphene has been hampered due to the difficulty of
producing single or few-layers specimens over large areas. Three main methods have
been used to obtain single-layer or few-layer graphene (FLG): i) Epitaxial growth of
graphene obtained on 6H oriented SiC by vacuum annealing at 1400°C [4], with the
drawback of being limited by the cost and size of SiC substrates; ii)

Micromechanical
exfoliation of small mesas of highly oriented pyrolytic graphite (HOPG)

[3], which
cannot be scaled to wafer-size dimensions,

and iii) chemically-assisted exfoliation of
intercalated graphite compounds [5-7], which typically leads to graphene with large
amount of defects. An alternative way is the chemical vapour deposition (CVD) of
camphor on nickel [8], which led to growth of graphene of about twenty layers.
Camphor, however, is a solid precursor that needs to be sublimated, and its vapour
pressure cannot be precisely controlled. Recently, segregation of graphene was reported
32

on Ni surfaces; however, several layers were obtained instead of single-layer graphene,
and the electronic properties of the synthesized material were not evaluated [9].
Therefore, approaches that can provide high-quality single- and few-layer graphene over
large areas and the evaluation of their electronic properties are still desired to meet
realistic applications.

Table 2.1, Parameters correlating graphene and Ni (111).

In this chapter we report the implementation of a simple, scalable and cost-
efficient method to prepare single and few-layer graphene films by CVD on nickel films
by using a low molar-mass hydrocarbon, such as methane, as carbon feedstock. Ni films
provide an excellent geometrical fit of the ordered graphene/graphite phase of carbon to
the crystalline metal surface, as well as convenient interactions that favors bond
formation between carbon atoms at specific conditions [10].
It is assumed that the carbon atoms dissolve into the Ni crystalline surface, and at
certain temperatures, they arrange epitaxially on the Ni (111) surface to form graphene
(figure 2.1). Synthesized graphene films on Ni were recovered on Si/SiO
2
substrates for
device fabrication. In addition, we have achieved transferring the as-synthesized films to
different target substrates such as Si/SiO
2
and glass, which can enable wafer-scale
silicon-compatible fabrication of hybrid silicon / graphene electronics and transparent
conductive film applications.

Ni(111) intersticious
distance (Å)
C-C bond length
(Å)
Ni lattice constant
(Å)
C lattice constant
(Å)
Lattice constant
mismatch (%)
1.412
1.420 2.517 2.4610 2.2
33

Figure 2.1 Schematic representation of the atomic arrangement of the hexagonal lattice of the
(111) face of nickel (blue spheres) and how carbon atoms (gray spheres) would arrange on the Ni
(111) surface to form graphene.

2.2 Experiments

2.2.1 Single-layer and few-layer graphene synthesis
Si/SiO
2
wafers of 4 inch in diameter were used as substrates to deposit 100 nm
thick films of elemental Ni (Figure 2.2a) by electron beam (e-beam) evaporation of an
elemental Ni target with purity 99.999%. Evaporated films were annealed at 300 or 800
°C in a 10:1 Ar:H
2
mixture to induce the formation of crystalline nickel on the substrate
surface. Heating and cooling rates of 0.15 °C min
-1
allowed the formation of
polycrystalline Ni domains throughout the substrate. X-ray diffraction measurements
revealed that the (111) face is the most abundant crystalline orientation on the
polycrystalline surface, accompanied by the (200) face in much less proportion (figure
2.2b-c).

34

Wafers with polycrystalline nickel films were subsequently used in graphene
synthesis experiments. Schematic shown in Figure 2.2d depicts the system set up
employed in the CVD deposition of graphene layers. CVD synthesis of graphene was
carried out at ambient pressure by systematically varying parameters such as temperature,
gas composition, gas flow rate and deposition time.

Figure 2.2 AFM image of e-beam evaporated nickel film atop a Si/SiO
2
substrate before (a) and
after (b) high temperature annealing. (c) X-ray diffraction spectra of the nickel film after thermal
annealing. (d) Schematic diagram depicting apparatus and chemical vapour deposition process
employed. (e) Wafer scale growth of graphene on Ni substrate. (f) AFM image of nickel substrate
after CVD synthesis of graphene.

Below we present results obtained by heating the substrates under a flow of 600
standard cubic centimeters (sccm) of H
2
up to 800 °C. After the target temperature was
reached, methane gas at a flow rate of 100 sccm was added to the hydrogen flow over the
substrate, which was lying horizontally inside the tube. The deposition process was
conducted for 8 minutes over complete wafers (figure 2.2e). We found that using diluted
2 μ μ μ μm
2 μ μ μ μm
2 0 3 0 4 0 5 0 6 0 7 0 8 0

2 θ /d e g re e s Intensity (arb. units)
N i (1 1 1 )
N i (2 0 0 ) S i (1 0 0 )
a
b
c

Gas in
Gas out
Substrate
1 μm 1 μm
(a)
(b) (c)
f e
d
35

methane was key for the growth of single and few-layer graphene (less than 5 layers),
while using concentrated methane led to the growth of multilayer graphene that
resembled bulk graphite. Figure 2.2f shows an AFM image of the Ni surface after
graphene synthesis. Our graphene growth method could be extended to other carbon
feedstocks such as ethylene, acetylene, ethanol, and isopropanol, and other catalytic films
such as Fe and Co.

2.3 Results

2.3.1 Graphene synthesis
We performed micro Raman analysis throughout the wafers after the chemical
vapour deposition process to confirm the formation of graphene layers on the Ni surface
and to obtain information about the quality and the number of layers deposited. Figure
2.3a shows Raman spectra taken at different locations on the synthesized films over
Si/SiO
2
/Ni substrates by using an excitation wavelength of 532 nm, with a power density
of 2.0 mW cm
-2
. Strong peaks near 1580 cm
-1
and 2690 cm
-1
were found. Analysis of the
frequencies and lineshapes of these peaks allows their assignment as the G and G’ bands
of graphene layers, respectively [11]. The peak located at 1345 cm
-1
corresponds to the D
band of graphitic carbon species, which is associated with the amount of defects in the
crystalline structure of the graphene layers. The low cross section of the D band confirms
that synthesized films are largely free of structural defects.
Interlayer interactions affect the Raman fingerprints for single-layer, bilayer, and
few-layer graphene, allowing unambiguous identification of graphene layers [11, 12].
36

Figure 2.3b shows the Raman spectrum of single-layer graphene in the synthesized films.
Single Lorentzian fit of the G’ band is characteristic of monolayer graphene. On the other
hand, a subtle splitting, up-shift of nearly 15 wavenumbers and broadening observed in
the G’ band that can be fit with four Lorentzian peaks, as shown in Figure 2.3c, which
constitute the spectroscopic signature of bilayer graphene [11, 12]. The domain size for
the single-layer, bilayer, and few-layer graphene is typically around 1-2 um, which is
likely due to the grain size of the polycrystalline nickel film. Extensive Raman
characterization over as-synthesized samples consistently showed the presence of
graphene with less than five graphene layers [13]. No signature of multi-layer or bulk
graphite was found in the films deposited.
Figure 2.3 (a) Raman spectrum obtained on as-synthesized graphene films on Si/SiO
2
/Ni
substrates. D, G and G’ Raman bands for graphene are labeled on each spectrum. (b) Raman
spectrum obtained on a single layer graphene. Zoomed images of the G’ band on the right shows
the purely single Lorentzian fit of this peak (red trace), which is characteristic of single-layer
graphene. (c) Raman spectrum of bilayer or few-layer graphene. A nearly symmetric splitting and
broadening of the G’ band of this spectrum (shown on the right) is properly fit by a set of four
Lorentzian curves, that can be assigned to bilayer or few-layer graphene. No G’ band was found
with a bulk graphite-like lineshape on the substrates sampled.
2500 2600 2700 2800 2900
R am an shift (cm
-1
)
E xp
F it
2500 2600 2700 2800 2900
R am an shift (cm
-1
)
Exp
Fit
1200 1800 2400 3000
Raman Shift (cm
-1
)
1 μm 1 μm
a
b
c
G’
G
D
37

2.3.2 Device fabrication and electrical measurements

Figure 2.4 (a) Schematic of a graphene film transferred to a Si/SiO
2
substrate via nickel etching.
(b) Photograph of a 4 inch wafer after synthesis of transferred FLG on polycrystalline nickel;
upper right shows an AFM image of the few-layer graphene films on Si/SiO
2
; lower right panel
shows an optical micrograph of deposited few-layer graphene on Si/SiO
2
; lower left panel shows
a typical micro Raman spectrum of the transferred films.

Two methods were used to transfer the as-synthesized graphene film to target
substrates. The first approach consisted of immersing the graphene-on-nickel sample into
a nickel etchant solution (figure 2.4a). This process removed nickel and left graphene
films deposited on the underlying Si/SiO
2
substrate. Figure 2.4b shows a photograph of a
Ni etching
Ni
Si/SiO
2
Si/SiO
2
Graphene
5 μm 5 μm
1200 1800 2400 3000

Raman Shift (cm
-1
)
a
b
38

Si/SiO
2
/Ni/Graphene wafer right after CVD synthesis. Micro Raman spectrum taken on
the films after transfer is shown in lower left part of figure 2.4b, clearly showing very low
D band intensity, which confirms that graphene is largely free of defects after transfer.
Upper and lower right panels of figure 2.4b shows AFM and white-light microscopy
images, of the graphene films on Si/SiO
2
substrate after etching the Ni film, respectively.
FLG were composed of micron-scale domains of single-, bi- and few-layers of graphene
with a maximum thickness of 5 layers, as confirmed by micro Raman spectroscopy.
Graphene transfer from the original Ni substrate to a Si/SiO
2
substrate allowed the
fabrication of back-gated FETs at large scale (Figure 2.5a). Micro-Raman measurements
performed on the device channel were consistent with a maximum of five graphene
layers comprising the films (data not shown). Four-probe measurements performed on the
FLG films revealed a sheet resistance of ~68 kΩ/sq.
I
DS
-V
DS
characteristics depicted in figure 2.5b shows that the drain current
increases with the increase of negative gate voltage, indicating a weak p-type behavior in
the films. Figure 2.5c shows the transfer characteristics for a device with channel width
of 20 μm and channel length of 4 μm. Most devices were highly conductive and exhibited
a weak modulation of the drain current by the gate bias, which is consistent with a 2D
semimetal [3].
Compared to nanotubes [14, 15], graphene FETs typically exhibit low current
on/off ratios, which can be improved significantly by patterning graphene into
nanoribbons [16]. Single graphene layer is a zero-gap semiconductor, but interlayer
interactions bring in a semimetal behavior in FLG. Therefore, the transfer characteristics
39

observed in Figure 2.5c can be attributed to a screened gating effect due to irregularities
of the film and the presence of more than two graphene layers in the films.

Figure 2.5 (a) 4-inch wafer with back-gated few-layer graphene devices; insets show SEM and
AFM images of a typical device and device channel, respectively. (b) I
DS
-V
DS
measurements for
different gate voltages, V
G
= 2.5 V, 1.5 V and -1.5 V for the black, red and blue curves,
respectively and (c) I
DS
-V
G
curve of one of the FET devices for V
DS
= 0.01 V.

2.3.3 Graphene transparent conductive film
For our second transfer approach, we used poly (methyl metacrylate) (PMMA)
which was spin-coated on top of the synthesized graphene films lying on Si/SiO
2
/Ni
substrates. The wafer was dipped into a nickel etchant solution (Transene Company Inc.)
at 90°C for 2 hours to etch away the nickel film and render a free-standing PMMA film
with the synthesized graphene adhered to it. PMMA/graphene film was transferred to
Drain
Source
2 μm 0.5 μm
-40 -20 0 20 40
3
3.2
3.4
3.6
3.8
I
DS
(μA)
V
G
(V)
0 -1 -2 -3 -4 -5
0
-1
-2
-3
I
DS
(mA)
V
DS
(V)
a
b
c
40

other substrates (Si/SiO
2
, glass, etc.), and then acetone was used to dissolve the PMMA
residues and leave clean graphene films on the target substrate surface.
Figure 2.6a shows a photograph of a ~1cm
2
FLG film transferred on glass
exhibiting high transparency to naked eyes. PMMA-assisted transfer is currently limited
to small areas due to the difficulty in handling PMMA films; however, we are certain that
wafer-scale transfer by similar approaches using PMMA or other polymers will be
developed. Figure 2.6b shows that the transmittance spectrum of the transferred FLG film
in the visible wavelength range is ~80%, which is consistent with the visible-light
transmittance reported for 2-3 graphene layer films[17]. Due to the simultaneous good
electrical conductivity and high transparency of the synthesized graphene films, they are
likely to find application as transparent conductors.

Figure 2.6 (a) Photograph of a 1cm
2
FLG film transferred onto a glass substrate (inside the red
box). (b) Transmittance of the FLG film shown in (a).

400 500 600 700 800
0
20
40
60
80
100
Transmittance (%)
Wavelength (nm)
a b
41

2.4 Conclusions
In summary, this chapter demonstrates a simple, scalable and effective method to
synthesize monolayer and few-layer graphene films by using methane-based CVD on
nickel films, transfer of the synthesized films to different target substrates and their
evaluation as transparent conducting films. Graphene produced over Si/SiO
2
wafers can
be very useful for device fabrication, and our approach may serve as the foundation for
the growth of single-domain graphene over macro-scale areas such as complete wafers.
Growth of graphene on single-crystalline nickel is currently under way with the goal of
significantly increasing graphene grain size. This approach constitutes a significant
advance towards the production of thin films of graphene at industrial scales and has
important implications for future graphene-related electronic devices.

42

Chapter 2 References
1. Zhang, Y., et al., Experimental observation of the quantum Hall effect and Berry’s phase
in graphene. Nature, 2005. 438: p. 201-204.

2. Hwang, E.H., S. Adam, and S. Das Sarma, Transport in chemically doped graphene in
the presence of adsorbed molecules. Physical Review B, 2007. 76(19): p. 195421.

3. Novoselov, K.S., et al., Electric Field Effect in Atomically Thin Carbon Films. Science,
2004. 306: p. 666-669.

4. Forbeaux, I., J.-M. Themlin, and J.-M. Debever, Heteroepitaxial graphite on 6H-
SiC(0001): Interface formation through conduction-band electronic structure. Physical
Review B, 1998. 58(24): p. 16396-16406.

5. Viculis, L.M., J.J. Mack, and R.B. Kaner, A Chemical Route to Carbon Nanoscrolls.
Science, 2003. 299: p. 1361.

6. Gilje, S., et al., A Chemical Route to Graphene for Device Applications. Nano Letters,
2007. 7(11): p. 3394-3398.

7. Wu, J., et al., Organic solar cells with solution-processed graphene transparent
electrodes. Applied Physics Letters 2008. 92: p. 263302.1-263302.3.

8. Somani, P.R., S.P. Somani, and M. Umeno, Planer nano-graphenes from camphor by
CVD. Chemical Physics Letters, 2006. 430: p. 56-59

9. Yu, Q., et al., Graphene segregated on Ni surfaces and transferred to insulators. Applied
Physics Letters, 2008. 93, : p. 113103.1-113103.3.

10. Eizenberg, M. and J.M. Blakely, Carbon Monolayer phase condensation on Ni(111)
Surface Science, 1979. 82: p. 228-236.

11. Ferrari, A.C., et al., Raman Spectrum of Graphene and Graphene Layers. Physical
Review Letters, 2006. 97(18): p. 187401.

12. Gupta, A., et al., Raman Scattering from High-Frequency Phonons in Supported n-
Graphene Layer Films. Nano Letters, 2006. 6(12): p. 2667-2673.

13. Cancado, L.G., et al., Geometrical approach for the study of G ' band in the Raman
spectrum of monolayer graphene, bilayer graphene, and bulk graphite. Physical Review
B, 2008. 77(24).

14. Kang, S.J., et al., High-performance electronics using dense, perfectly aligned arrays of
single-walled carbon nanotubes. Nat Nano, 2007. 2(4): p. 230-236.

15. Wang, C., et al., Device study, chemical doping, and logic circuits based on transferred
aligned single-walled carbon nanotubes. Applied Physics Letters, 2008. 93(3).

16. Han, M.Y., et al., Energy Band-Gap Engineering of Graphene Nanoribbons. Physical
Review Letters, 2007. 98(20): p. 206805.
43

17. Li, X., et al., Highly conducting graphene sheets and Langmuir-Blodgett films. Nat Nano,
2008. 3(9): p. 538-542.

44

Chapter 3 Synthesis of Single and Bilayer Graphene by Chemical Vapor
Deposition of Carbon on Single Crystal Nickel

3.1 Introduction
Tremendous efforts have been made to explore both the physical properties [1-4]
and applications of graphene, such as the opening of bandgap of graphene, [5-7]
graphene as transparent conductive electrodes [8, 9], graphene nanomechanical
resonators [10] and so on. Micromechanical cleavage of graphite has allowed the study of
fundamental properties of graphene due to the high quality, scarce presence of structural
defects, and low levels of unintentional doping of the exfoliated graphene [11, 12].
However, this approach is not scalable and the development of new methods to obtain
graphene by scalable methods has surged as an active research area [13-17].
Recently, chemical vapor deposition (CVD) has raised its popularity in the
synthesis of graphene as a scalable and cost effective approach [18-25]. Polycrystalline
Ni has been shown to be a good substrate for graphene synthesis by CVD, but the
percentage of monolayer or bilayer graphene is limited by the grain size of crystalline Ni
obtained after thermal annealing of Ni thin film [20, 22]. It was pointed out that CVD
graphene on Ni can be divided into two categories: multilayer graphene (≥ 3 layers) and
monolayer/bilayer graphene, as micro-Raman can distinguish multilayer from
monolayer/bilayer, but cannot distinguish monolayer from bilayer[26]. We, among other
groups, have reported the synthesis of wafer-scale few-layer graphene by CVD on the
surface of polycrystalline Ni [22]. Our results suggest that during the synthesis carbon
atoms tend to segregate on nucleation sites on the Ni surface to form multiple-layer
45

graphene grains. The formation of such multilayer domains is believed to be correlated to
different factors including the abundance of defects and grain boundaries on the
polycrystalline Ni substrate. It is therefore particularly interesting to investigate the
formation of graphene on single crystal Ni due to the absence of interface boundaries,
and Ni (111) is especially interesting due to the excellent lattice match between
graphene/graphite and Ni (111) face, where the hexagonal lattice constant is 2.497 Å for
Ni (111) and 2.46 Å for graphite [27].
In this chapter, we study the influence of the concentration of Ni interface
boundaries on the formation of such multilayer graphene domains. Our main finding is
that the synthesis of graphene on the (111) face of single crystal Ni by CVD favors the
formation of highly uniform monolayer/bilayer graphene on the Ni surface, and
simultaneously hinders the formation of multilayer graphene domains. Our results are
understood on the basis of the diffusion-segregation model for carbon precipitation on Ni
surface [28], where the uniform and grain-boundary-free surface of Ni (111) single
crystal provides a smooth surface for uniform graphene formation. In contrast, the rough
surface of polycrystalline Ni with abundant grain boundaries facilitates the formation of
multilayer graphene. Micro-Raman surface mapping reveals that the area percentages of
monolayer/bilayer graphene are 91.4% for the Ni (111) substrate and 72.8% for the
polycrystalline Ni substrate under comparable CVD conditions.

46

3.2 Experiments

3.2.1 Graphene synthesis
To prepare samples for CVD growth, single crystal Ni was obtained from Crystal
Base Co., Ltd, while polycrystalline Ni was prepared by depositing 500 nm of Ni metal
(99.999 % purity) onto a SiO
2
/Si wafer by e-beam evaporation. The growth of graphene
on Ni (111) and polycrystalline Ni was conducted in the same CVD furnace with a 1 inch
quartz tube. Different flow rates of CH
4
were used for Ni (111) and polycrystalline Ni
film in order to favor the formation of thin graphene films on both substrates. The Ni
(111) (or polycrystalline Ni film) substrate was first loaded into the furnace, then 600
sccm H
2
was flew into the furnace for 15 minutes while the temperature increased from
the room temperature to 900 °C. 80 sccm CH
4
(25 sccm for Ni film) was introduced to
the furnace under 900°C for 10 minutes. The furnace was then cooled done to room
temperature under a cooling rate of 16°C/min.
In order to correlate graphene formation with degree of crystallinity of the Ni
substrate, we have employed three different annealing rates on polycrystalline Ni. In the
fast annealing condition, the temperature was increased from room temperature to 900°C
in 5 minutes, followed by a subsequent annealing at 900 °C for 5 minutes. In the medium
annealing condition, the temperature was increased from room temperature to 900°C in
15 minutes, followed by 20 minutes annealing. In the slowest annealing condition, the
temperature was increased from room temperature to 500°C in 8 minutes and further
increased to 900°C in 15 minutes. Samples were then annealed at this temperature for 20
minutes. 600 sccm H
2
was introduced to the system during all three annealing processes.
47

3.3 Results

3.3.1 X-ray diffraction of single crystal nickel vs. polycrystalline nickel
Shown in Figure 3.1 are the X-Ray Diffraction (XRD) spectra taken after the
thermal annealing of both substrates. There is only a single peak corresponding to Ni
(111) in the XRD spectrum taken from the Ni (111) substrate (Figure 3.1.a) as expected.
The XRD spectrum of polycrystalline Ni, however, shows a strong Ni (111) peak and a
very weak Ni (200) after the annealing process, which indicates the presence of
predominantly Ni (111) grains with a smaller population of Ni (100) grains. The Atomic
Force Microscopy (AFM) images shown as insets in Figure 3.1 further indicate the
difference between those two substrates in terms of the surface roughness: The surface of
Ni (111) is considerably smoother (average roughness = 5.3 nm) than the polycrystalline
Ni (average roughness = 36.3 nm) after thermal annealing.
Figure 3.1 (a-b) XRD spectra and AFM images of Ni (111) substrate and polycrystalline Ni
substrate, respectively. The left inset in 1a is a schematic diagram of Ni (111) hexagonal lattice
arrangement with graphene carbon atoms on surface. The blue atoms are Ni and the grey ones are
carbon. The color scale bar corresponds to AFM images in a and b inset.

40 50 60

2θ θ θ θ ( ( ( (degrees)
Intensity (a.u.)
Ni (111)
Ni (200)
40 50 60
2θ θ θ θ ( ( ( (degrees)
Ni (111)

Intensity (a.u.)
a
b
48

In addition, the polycrystalline Ni surface acquired a multigrain-like appearance after
thermal annealing. The left inset schematic diagram in Figure 3.1.a displays the atomic
arrangement of Ni (111) and graphene, where the blue spheres represent Ni atoms and the
grey spheres represent carbon atoms.

3.3.2 Grain boundaries and multilayer graphene formation
The process of graphene growth on Ni can be divided into two parts: the first is
carbon segregation from bulk Ni to Ni surface in an intermediate temperature range (~
1065-1180 K), and the second is carbon precipitation which happens when the system
temperature decreases (<1065 K). Carbon segregation and precipitation tend to happen at
the grain boundaries [28]. This can be related to the fact that the impurities in transition
metals tend to segregate at grain boundaries, which can be rationalized by considering
that disorder and vacancies at such locations can readily act as active sites for the
interaction and accumulation of impurities during cooling [28]. In the specific case of
carbon dissolved in nickel, this means that grain boundaries can be good nucleation sites
for carbon segregation and hence, for multilayer graphene formation [26]. Therefore, we
believe that grain boundaries play an important role in both the carbon segregation and
precipitation processes during graphene synthesis.
To fully elucidate the effect of grain boundaries, we have performed a series of
experiments to prepare polycrystalline and single crystal Ni samples. Three different
annealing rates (175 °C/min, 58 °C/min, 27 °C/min) were chosen from fast to slow for
the annealing process of polycrystalline Ni, while the medium annealing rate (58 °C/min)
49

Figure 3.2 (a) XRD spectra collected from polycrystalline Ni with fast (black), medium (red), and
slow (blue) annealing rates and XRD spectrum collected from Ni (111) (green) after thermal
annealing (XRD spectrum is identical for Ni (111) using different annealing rates). (b). Zoomed-
in XRD spectra of peaks at 2θ = 52.16° (assigned as Ni (200)). (c-j). Optical images taken after
graphene CVD growth from polycrystalline Ni with fast (c, d), medium (e, f) and slow (g, h)
annealing rates. (i, j). Optical images taken after graphene CVD growth from Ni (111).

was applied to anneal the Ni (111). XRD spectra were collected after the thermal
annealing process as shown in Figure 3.2.a The first three spectra from top to bottom
correspond to fast, medium and slow rate annealing on polycrystalline Ni, and the last
one is collected from Ni (111) after annealing. All the spectra were normalized for further
analysis. While all four samples show a strong peak corresponding to Ni (111), the
polycrystalline Ni samples display an additional Ni (200) peak at 2θ = 52.16°, with
intensity being highest for fast annealing and lowest for slow annealing. In contrast, the
35 40 45 50 55 60
2θ θ θ θ (degrees)
Intensity (a.u.)

fast
medium
slow
single crystal
50 51 52 53 54 55
2θ θ θ θ (degrees)
Intensity (a.u.)

(a)
(b)
(c)
(d)
(e)
(f)
(h)
(j)
(g)
(i)
50

single crystalline Ni XRD spectrum shows no peak for Ni (200) (Figure 3.2.b).
Therefore, we can conclude that slower thermal annealing favors the formation of
crystalline Ni (111) grains with less grain boundaries for polycrystalline Ni samples.
Graphene was synthesized by CVD at 900 °C at atmosphere pressure, and a
cooling rate of 16 °C/min was used down to 500 °C. The details of the synthesis
procedures are similar to the previously reported work [22] and can also be found in the
experimental section above. Figure 3.2.c-j depicts the optical images of Ni substrates
after graphene synthesis. Figure 3.2.c and d, 3.2.e and f, and 3.2.g and h correspond to
polycrystalline Ni samples obtained with fast, medium and slow thermal annealing rates,
respectively. The dark regions are confirmed to consist of multilayer graphene (≥3
layers), while the light regions are confirmed to be monolayer/bilayer graphene using
micro-Raman spectroscopy. According to Figure 3.2.c-h, the percentage of multilayer
graphene formation increases as the polycrystallinity of the Ni substrate increases. These
results suggest that the formation of multilayer graphene can be attributed to the increase
of carbon segregation localized at polycrystalline grain boundaries, while the formation
of monolayer/bilayer graphene is mainly obtained on the flat central areas of large
crystalline grains.
Furthermore, images obtained on Ni (111) substrate after graphene synthesis are
shown in Figure 3.2.i and j. Analysis of the graphene growth on single crystal Ni (111)
(Figures 3.2.i and j) reveals the scarce formation of multilayer graphene grains. The
observation can be understood by considering the absence of inter-plane grain boundaries
on the surface of Ni (111) and therefore a shortage of nucleation sites for multilayer
51

graphene formation. Thus, mostly monolayer/bilayer graphene is uniformly formed on
the surface of Ni (111) single crystal.
Based on the discussion above, a graphene growth mechanism is proposed in
Figure 3.3 Figure 3.3.a and b give schematic diagrams of possible mechanism of the
formation of graphene during the carbon segregation and precipitation on Ni (111) and
polycrystalline Ni surfaces, respectively. Due to the high solubility of carbon in Ni,
carbon first diffuses into bulk Ni, and then segregates and precipitates onto Ni surface. In
the carbon/Ni (111) system, the surface of Ni (111) is very smooth with almost no grain
boundaries, allows uniform segregation of carbon onto the Ni (111) surface, and thus
tends to form single layer graphene. In contrast, in the carbon/ polycrystalline Ni system,
the Ni surface is heavily populated by the grain boundaries, especially inter-plane grain
boundaries, which allow the accumulation of carbon at these sites during the segregation
phase and lead to the formation of multilayer graphene.
Optical images were taken from both Ni (111) and polycrystalline Ni substrates.
Figure 3.3.c shows that the surface of Ni (111) has a relatively uniform color with only a
few dark dots. In contrast, the surface of polycrystalline Ni has many dark grains, as
shown in Figure 3.3.d. Both the dark dots and dark grains have been confirmed to be
multilayer graphene by micro-Raman, while the rest of the surface is confirmed to be
monolayer/bilayer graphene. This experimental result well matches our proposed “grain
boundary mechanism” for graphene formation. The inset schematic diagram in Figure
3.3.c shows the formation of graphene in an ideal case: a well-ordered graphene sheet on
Ni (111) surface without any grain boundaries.
52

Figure 3.3 (a-b). Schematic diagrams of graphene growth mechanism on Ni (111) (a) and
polycrystalline Ni surface (b). (c). Optical image of graphene/ Ni (111) surface after the CVD
process. The inset is a three dimensional schematic diagram of a single graphene layer on Ni
(111) surface. (d). Optical image of graphene/ polycrystalline Ni surface after the CVD process.
The inset is a three dimensional schematic diagram of graphene layers on polycrystalline Ni
surface. Multiple layers formed from the grain boundaries.

The formation of multilayer graphene on polycrystalline Ni is illustrated in the
inset of Figure 3.3.d, red and green represents two graphene layers which segregate and
precipitate from grain boundaries. More layers will continue segregating and
precipitating from the boundaries depends on the concentration of carbon in bulk Ni. The
misalignment between two polycrystalline grains may provide abundant nucleation sites
for carbon atoms to segregate from the boundaries. Therefore, multilayer graphene tends
to form at the boundaries, while monolayer graphene tends to form on Ni (111) surface.

(b)
(c)
(d)
(a)
SLG SLG SLG
53

3.3.3 Graphene characterization
The formation of graphene layers on Ni surface was confirmed by micro-Raman
spectroscopy after the CVD process. The information of the defects of graphene, as well
as the number of graphene layers can also be derived from Raman spectra. Figure 3.4.a
and b show ten typical spectra collected from different locations on the synthesized
graphene films on Ni (111) and polycrystalline Ni, respectively. The low intensity of D
band (~ 1350 cm
-1
) confirms that the graphene formed on both Ni (111) and
polycrystalline Ni surfaces are of low defects. Peaks located at ~ 1590 cm
-1
and ~ 2700
cm
-1
are assigned as G and G’ bands of the graphene layers, respectively. All ten spectra
collected from the graphene on Ni (111) in Figure 3.4.a show single Lorentzian lineshape
and narrow linewidth (25 – 55 cm
-1
). Furthermore, all ten spectra exhibit G’ to G peak
intensity ratios (I
G’
/I
G
) larger than one, which are considered fingerprints for the
formation of monolayer/bilayer graphene, as previously reported [20].
Figure 3.4 (a-b). Ten typical Raman spectra of graphene grown on Ni (111) and polycrystalline
Ni, respectively. (c). The G’-to-G peak intensity ratio (I
G’
/I
G
) v.s. the Full Width at Half
Maximum (FWHM) of G’ bands of graphene on both Ni (111) and polycrystalline Ni.

In contrast, the ten typical Raman spectra collected from the graphene on
polycrystalline Ni in Figure 3.4.b can be divided into two groups. The first group of
50 75 100
0
1
2
3
Single crystal
Polycrystalline
I
G'
/ I
G
G' FWHM
1200 1600 2000 2400 2800
Raman Shift (cm
-1
)

1200 1600 2000 2400 2800

Raman Shift ( cm
-1
)
a b c
54

spectra are similar to spectra in Figure 3.4.a, which have single Lorentzian profile for G’
band, narrow linewiths for G’ peaks, and I
G’
/I
G
> 1. On the other hand, the second group
of spectra have noticeable up-shift of ~ 15 cm
-1
in the G’ band, as well as broadening of
G’ band linewidth (~50 cm
-1
) that can be fit with four or more Lorentzian peaks. It is
known that the G’ peak position exhibits up-shift with the increase of number of layer of
graphene, and the fitting of G’ peak with two or more Lorentzian peaks is a signature of
multilayer graphene [29, 30]. Moreover, I
G’
/I
G
for the second group of Raman spectra is
smaller than one, which is characteristics of multilayer graphene [29, 31]. Altogether, all
the Raman spectra collected from the graphene film on single crystal Ni show the
formation of monolayer/bilayer graphene, while only about 50% of the spectra collected
from the graphene film grown on polycrystalline Ni correspond to monolayer/bilayer
graphene.
Figure 3.4.c shows a plot of I
G’
/I
G
values versus Full Width at Half Maximum
(FWHM) of G’ bands of micro-Raman spectra taken at random locations on the graphene
films grown on single crystal and polycrystalline Ni. It is clearly shown in Figure 3.4.c
that all spectra of graphene on single crystal Ni surface have higher IG’/IG values (all
higher than one) and narrower FWHM, while only half of the spectra of graphene on
polycrystalline Ni have high I
G’
/I
G
values and narrow FWHM. Therefore, the result from
Figure 3.4.c confirms that the graphene on single crystal Ni is dominantly
monolayer/bilayer, while the percentage of multilayer graphene is much higher on
polycrystalline Ni.

55

Figure 3.5 (a). Maps of I
G’
/I
G
of 780 spectra collected on a 60*50 μm
2
area on the Ni (111)
surface and (b) 750 spectra collected on a 60*50 μm
2
area on the polycrystalline Ni surface.
Corresponding optical images to Ni (111) Raman map and polycrystalline Ni Raman map (c and
d).

In order to have a better idea of how the graphene grows in large area on both Ni
(111) and polycrystalline Ni, about 800 Raman spectra were collected over 3000 μm
2

area with 2 μm spacing between each data point on both surfaces. The I
G’
/I
G
values were
then extracted from the spectra. Figure 3.5.a and b show the I
G’
/I
G
contour maps of
graphene on Ni (111) and polycrystalline Ni, respectively. 91.4% of Raman spectra
collected from the graphene on Ni (111) surface has I
G’
/I
G
higher than one, which is a
hallmark of monolayer/bilayer graphene. In contrast, the percentage of monolayer/bilayer
graphene is only 72.8% from the graphene grown on polycrystalline Ni. Moreover, the
narrow spacing of 2 μm between data points for the Raman measurements enables the
0 10 20 30 40 50
50
40
30
20
10
0

Y (μm)
X (μm)
0
0.25
0.50
0.75
1.0
91.4%
(a)
0 10 20 30 40 50
40
30
20
10
0

Y (μm)
X (μm)
0
0.25
0.50
0.75
1.0
72.8%
(b)
(c) (d)
56

confirmation of continuous graphene deposition on both substrates. Figure 3.5.c and d are
the corresponding optical images of the two Raman maps, which is in accordance with
the Raman maps: dark regions in the optical images correspond to multilayer graphene in
Raman maps, while the rest is confirmed to be monolayer/bilayer graphene.
Optical images and Raman spectra were obtained on the transferred graphene film
from Ni (111) and polycrystalline Ni film. Transfer of graphene from Ni (111) and
polycrystalline Ni was done using the same method reported above [22]. Comparing
Figure 3.6.a and b, the graphene transferred from Ni (111) is more uniform than graphene
transferred from polycrystalline Ni. It is confirmed by Raman spectroscopy that the
graphene transferred from Ni (111) is monlayer/bilayer as shown in Figure 3.6.c. The G’
band can be fitted by a single Lorentzian peak and has a linewidth of ~30 cm
-1
. I
G’
/I
G
is
also larger than one. Raman spectra collected from different locations on the substrate are
shown in Figure 3.6.d. Raman spectra colored in red, black, and blue were taken from
regions pointed by arrows in red, black, and blue respectively. The shape of G’ peak in
red corresponds to the Raman signature of graphite. The G’ peak in black has a
symmetric shape with a linewidth of ~60 cm
-1
, indicating the formation of few-layer
graphene (<5 layers). The G’ peak in blue has a linewidth of ~30 cm
-1
and can be fitted
by a single Lorentzian peak indicating monolayer/bilayer graphene.

57

Figure 3.6 Optical image of graphene transferred form Ni (111) (a) and polycrystalline Ni
film (b) to SiO
2
/Si substrate. Corresponding Raman spectra taken from graphene
transferred from Ni (111) (c) and polycrystalline Ni (d).

3.4 Conclusions
In summary, we found that preferential formation of monolayer/bilayer graphene
on the single crystal surface is attributed to its atomically smooth surface and the absence
of grain boundaries. In contrast, CVD graphene formed on polycrystalline Ni leads to
higher percentage of multilayer graphene (≥3 layers), which is attributed to the presence
1200 1600 2000 2400 2800

Raman Shift (cm
-1
)
1200 1600 2000 2400 2800

Raman Shift (cm
-1
)
a b
e c
58

of grain boundaries in Ni that can serve as nucleation sites for multilayer growth. Micro-
Raman surface mapping reveals that the area percentages of monolayer/bilayer graphene
are 91.4% for the Ni (111) substrate and 72.8% for the polycrystalline Ni substrate under
comparable CVD conditions.

59

Chapter 3 References
1. Geim, A.K. and K.S. Novoselov, The rise of graphene. Nat Mater, 2007. 6(3): p. 183-
191.

2. Novoselov, K.S., et al., Electric Field Effect in Atomically Thin Carbon Films. Science,
2004. 306: p. 666-669.

3. Bolotin, K.I., et al., Ultrahigh electron mobility in suspended graphene. Solid State
Communications, 2008. 146(9-10): p. 351-355.

4. Novoselov, K.S., et al., Two-dimensional gas of massless Dirac fermions in graphene.
Nature, 2005. 438(7065): p. 197-200.

5. Wang, X. and H. Dai, Etching and narrowing of graphene from the edges. Nat Chem.
2(8): p. 661-665.

6. Bai, J., et al., Graphene nanomesh. Nat Nano. 5(3): p. 190-194.

7. Kim, M., et al., Fabrication and Characterization of Large-Area, Semiconducting
Nanoperforated Graphene Materials. Nano Letters. 10(4): p. 1125-1131.

8. Gomez De Arco, L., et al., Continuous, Highly Flexible, and Transparent Graphene
Films by Chemical Vapor Deposition for Organic Photovoltaics. ACS Nano. 4(5): p.
2865-2873.

9. Bae, S., et al., Roll-to-roll production of 30-inch graphene films for transparent
electrodes. Nat Nano. 5(8): p. 574-578.

10. Chen, C., et al., Performance of monolayer graphene nanomechanical resonators with
electrical readout. Nat Nano, 2009. 4(12): p. 861-867.

11. Novoselov, K.S., et al., Two-dimensional atomic crystals. Proceedings of the National
Academy of Sciences of the United States of America, 2005. 102(30): p. 10451-10453.

12. Meyer, J.C., et al., The structure of suspended graphene sheets. Nature, 2007. 446(7131):
p. 60-63.

13. Stankovich, S., et al., Synthesis of graphene-based nanosheets via chemical reduction of
exfoliated graphite oxide. Carbon, 2007. 45(7): p. 1558-1565.

14. Forbeaux, I., J.-M. Themlin, and J.-M. Debever, Heteroepitaxial graphite on 6H-
SiC(0001): Interface formation through conduction-band electronic structure. Physical
Review B, 1998. 58(24): p. 16396-16406.

15. Rollings, E., et al., Synthesis and characterization of atomically thin graphite films on a
silicon carbide substrate. Journal of Physics and Chemistry of Solids. 67(9-10): p. 2172-
2177.

16. Hass, J., et al., Highly ordered graphene for two dimensional electronics. Applied
Physics Letters, 2006. 89(14): p. 143106-3.
60

17. Sutter, P.W., J.-I. Flege, and E.A. Sutter, Epitaxial graphene on ruthenium. Nat Mater,
2008. 7(5): p. 406-411.

18. Karu, A.E. and M. Beer, Pyrolytic Formation of Highly Crystalline Graphite Films.
Journal of Applied Physics, 1966. 37(5): p. 2179-2181.

19. Yu, Q., et al., Graphene segregated on Ni surfaces and transferred to insulators. Applied
Physics Letters, 2008. 93, : p. 113103.1-113103.3.

20. Reina, A., et al., Large Area, Few-Layer Graphene Films on Arbitrary Substrates by
Chemical Vapor Deposition. Nano Letters, 2009. 9(1): p. 30-35.

21. Kim, K.S., et al., Large-scale pattern growth of graphene films for stretchable
transparent electrodes. Nature, 2009. 457: p. 706-710.

22. Gomez, L., et al., Synthesis, Transfer and Devices of Single- and Few-Layer Graphene by
Chemical Vapor Deposition. IEEE Transactions on Nanotechnology, 2009. 8(2): p. 135-
138.

23. Li, X., et al., Large-Area Synthesis of High-Quality and Uniform Graphene Films on
Copper Foils. Science, 2009. 324(5932): p. 1312-1314.

24. Li, X., et al., Evolution of Graphene Growth on Ni and Cu by Carbon Isotope Labeling.
Nano Letters, 2009. 9(12): p. 4268-4272.

25. Levendorf, M.P., et al., Transfer-Free Batch Fabrication of Single Layer Graphene
Transistors. Nano Letters, 2009. 9(12): p. 4479-4483.

26. Reina, A., et al., Growth of large-area single- and Bi-layer graphene by controlled
carbon precipitation on polycrystalline Ni surfaces. Nano Research, 2009. 2(6): p. 509-
516.

27. Eizenberg, M. and J.M. Blakely, Carbon Monolayer phase condensation on Ni(111)
Surface Science, 1979. 82: p. 228-236.

28. Shelton, J.C., H.R. Patil, and J.M. Blakely, Equilibrium segregation of carbon to a nickel
(111) surface: A surface phase transition. Surface Science, 1974. 43(2): p. 493-520.

29. Ferrari, A.C., et al., Raman Spectrum of Graphene and Graphene Layers. Physical
Review Letters, 2006. 97(18): p. 187401.

30. Gupta, A., et al., Raman Scattering from High-Frequency Phonons in Supported n-
Graphene Layer Films. Nano Letters, 2006. 6(12): p. 2667-2673.

31. Cancado, L.G., et al., Geometrical approach for the study of G ' band in the Raman
spectrum of monolayer graphene, bilayer graphene, and bulk graphite. Physical Review
B, 2008. 77(24).

61

Chapter 4 Large-Area Graphene Films by Chemical Vapor Deposition for
Highly Flexible Organic Photovoltaics.

4.1 Introduction
Solar energy harvesting using organic photovoltaic (OPV) cells has been
proposed as a means to achieve low-cost energy due to their ease of manufacture, light
weight and compatibility with flexible substrates [1]. A critical aspect of this type of
optoelectronic device is the transparent conductive electrode through which light couples
into the device. Conventional OPVs typically use transparent indium tin oxide (ITO) or
fluorine doped tin oxide (FTO) as such electrodes [1, 2]. However, the scarcity of indium
reserves, intensive processing requirements, and highly brittle nature of metal oxides [3-
5] impose serious limitations on the use of these materials for applications where cost,
physical conformation, and mechanical flexibility are important.
Carbon nanotubes [6-8] and nanowires [9, 10] have been used as alternative
materials for electrodes in OPVs, but the roughness of such films were comparable to or
larger than a typical device thickness, which may lead to significant shunt losses. In
contrast, graphene is a one-atom thick, two-dimensional crystalline arrangement of
carbon atoms with a quasi-linear dispersion relation, and predicted mobility on the order
of 10
6
cm
2
/V·s for a charge carrier concentration n
i
~ 10
12
cm
-2
[11].
Graphene monolayer has a transparency of 97-98 percent [12] and the sheet
resistance of undoped graphene is of the order of ~6kΩ [13-16]. Graphene films are
suitable for applications as transparent conductive electrodes where low sheet resistance
and high optical transparency are essential. Conventional methods to obtain graphene thin
62

films such as epitaxial growth [17], micromechanical exfoliation of graphite [18] and
exfoliation of chemically oxidized graphite [19, 20] are either expensive, unscalable or
yield graphene with limited conductivity due to a high defect density.
Recently, graphene films obtained from reduced graphene oxide (GO) have been
explored for applications as transparent electrodes in solar cells [21-23]. However, the
devices obtained exhibited rather moderate performance, leakage current under dark
conditions, and moderate power conversion efficiency of < 0.4 %. The moderate
performance of these devices may be attributed to several factors, including: i) reduction
of oxygen functionalities on the graphene oxide flakes does not completely restore the π
conjugation in the films, and ii) the vacuum filtration or spin coating methods used to
prepare reduced graphene oxide films lead to stacked graphene flakes and thus significant
flake-to-flake resistance.
Eda et. al. [21] reported doped reduced graphene oxide films with a sheet
resistance of 40 kΩ/sq, a transparency of 64%, and a solar cell conversion efficiency of
0.1%, while Wu et. al. [23] reported reduced graphene oxide films of 5–10
3
kΩ/sq, >80%
transparency, and a conversion efficiency of 0.4%. Efforts to improve percolation on the
graphene electrode include the use of reduced GO combined with carbon nanotube films,
but this approach requires extra processing steps [24]. As a result, continuous, highly
flexible, and transparent graphene films are still highly desirable for photovoltaic
applications.
Chemical vapor deposition (CVD) has surged as an important method to obtain
high quality graphene films [25-28]. In particular, films with sheet resistance of 280 Ω/sq
(80% transparent) and 770 Ω/sq (90% transparent) have been reported for graphene
63

synthesized on Ni films, while sheet resistance of 350 Ω/sq (90% transparent) has been
reported for CVD graphene on Cu films, which represents a good advance in the use of
graphene as transparent conductive films. Another advantage of CVD is its scalability,
we have reported wafer-scale synthesis and transfer of single- and few-layer graphene for
device fabrication [25].
In this chapter we present our work on the implementation of large area and
highly smooth few-layer graphene films, synthesized by CVD, as the anode material in
flexible and rigid OPV cells. The multilayer structure of the solar cell is shown in Figure
4.1. Such films exhibit sheet resistance and transparency controlled in the range of 230
Ω/sq at 72% transparency, and 8.3 kΩ/sq at 91% transparency. The use of CVD graphene
is attractive because other graphene films, which are formed by stacked micron-size
flakes, suffer from flake-to-flake contact resistance and high roughness. In contrast, grain
boundaries of CVD graphene films have the advantage of being formed in situ during
synthesis; such a process is expected to minimize contact resistance between neighboring
graphene domains and may result in smoother films with better conducting properties.
Solar cells made with CVD graphene exhibited performance that compares to ITO
devices and surpasses that of ITO devices under bending conditions, exhibiting power
conversion efficiencies of 1.18% and being operational under bending conditions up to
138°.

64

4.2 Experiments

4.2.1 CVD graphene synthesis
Elemental Ni was thermally evaporated onto pre-cleaned Si/SiO
2
substrates up to
a thickness of ~1000 Å. Subsequently, Ni/Si/SiO
2
substrates were taken into a sealed
high-temperature furnace and heated to 900 °C under a hydrogen flow rate of 600 sccm.
Graphene synthesis was obtained at 900 °C by flowing methane at a flow rate of 100
sccm for 8 minutes.

4.2.2 CVD graphene transfer
After synthesis, CVD Graphene was transferred to transparent target substrates
such as glass and polyethylene terephtalate (PET) by depositing a thin layer (300 nm) of
poly-methylmethacrylate (PMMA) on top of the as-synthesized graphene on Si/SiO
2
/Ni
substrates by spin coating. The nickel layer underneath the graphene was subsequently
dissolved by dipping the substrates in aqueous nitric acid solution (14% w/w) for
approximately 1 hour under constant heating at 90 °C or until the Ni film was etched to
levels below the limit of detection of an Omicron XPS/UPS system (< 0.1 - 1 atom %).
The free-standing PMMA/graphene film was then transferred by direct graphene
contact onto transparent substrates such as glass and polyethylene terephthalate (PET)
sheets. Then, the PMMA layer was finally dissolved with acetone for ~ 10 minutes,
leaving the graphene film on the target substrate surface. Figure 4.1 shows a schematic
representation of the graphene transfer process to transparent substrates, either rigid or
flexible.
65

Figure 4.1 Schematic of the CVD graphene transfer process onto transparent substrates.

4.2.3 Organic photovoltaic cell fabrication with CVD graphene and ITO
Graphene electrodes were fabricated by transferring as-grown CVD graphene
films onto pre-cleaned PET substrates. PET substrates coated with ITO were obtained
from Southwall Technologies Inc. Both substrates were solvent cleaned and passivated
by spin coating a thin layer of poly(3,4-ethylenedioxythiophene)-poly(styrenesulfonate)
(PEDOT:PSS). Organic thin films and the aluminum cathode were consecutively
deposited by thermal evaporation to form a multilayered configuration: CVD graphene or
ITO / PEDOT:PSS / Copper phthalocyanine (CuPc) / Fullerene (C
60
) / Bathocuproine
(BCP) / Aluminum (Al). Aluminum cathodes were deposited through a shadow mask
with circular openings of 0.75 mm
2
. Figure 4.2 shows a representation of the OPV cell
and the band structure of the stacked materials.
PMMA coating PMMA coating
Graphene film Graphene film
Ni Ni
Si/SiO Si/SiO
2 2
Ni etching Ni etching
PMMA / Graphene PMMA / Graphene
Substrate Substrate
Graphene Graphene
Transfer Transfer
C C
3 3
H H
6 6
O O
66

Figure 4.2 Schematic representation of the energy level alignment (right) and construction of the
heterojunction organic solar cell fabricated with graphene as anodic electrode: CVD graphene /
PEDOT / CuPc / C
60
/ BCP / Al.

4.3 Results

4.3.1 X-ray diffraction of polycrystalline Ni
X-ray diffraction spectra were collected on the annealed Ni substrates over which
graphene films are synthesized. To this end, elemental Ni was thermally evaporated onto
precleaned Si/SiO
2
substrates up to a thickness of ~1000 Å. Subsequently, Ni/Si/SiO
2

substrates were taken into a sealed high-temperature furnace and heated to 900 °C under
a hydrogen flow rate of 600 sccm.
A Rigaku x-ray diffractometer equipped with a 12 kW rotating anode x-ray
generator was employed to investigate the distribution of crystalline planes on the
annealed polycrystalline Ni. Figure 4.3a shows an AFM image of a typical annealed Ni
surface. Irregular and faceted-shape surface are consistent with polycrystalline formation.
Graphene
CuPc
C
60
Al 4.5eV
BCP
4.3eV
PEDOT
5.2eV
2.2eV
5.2eV
3.3eV
4.0eV
6.2eV
6.4eV
3.0eV
Graphene
CuPc
C
60
Al 4.5eV
BCP
4.3eV
PEDOT
5.2eV
2.2eV
5.2eV
3.3eV
4.0eV
6.2eV
6.4eV
3.0eV
Multilayer band structure
Graphene OPV structure
67

X-ray diffraction spectrum shown in Figure 4.3b reveals the presence of (111) and (200)
planes. Furthermore, it is clear that the surface of annealed Ni film is comprised
predominantly by the (111) plane.

Figure 4.3 (a) AFM image of a 300 nm Ni film deposited on Si/SiO2 substrate after high
temperature annealing. (b) Typical X-ray diffraction spectrum of annealed Ni film.

4.3.2 TEM, SAED and Raman of CVD graphene
Graphene films were synthesized by chemical vapor deposition on a thermally
annealed polycrystalline nickel surface comprised mostly of the (111) plane. The
graphene films were grown as detailed in a previous work [25]. This synthesis yields a
continuous film comprised of monolayer and few-layer graphene with low defect density,
as indicated by TEM imaging and diffraction and micro Raman spectroscopy
measurements performed on the transferred films. Figure 4.4a shows a low magnification
TEM image of the deposited CVD graphene films, while the selected area diffraction
pattern along the z- direction shows well ordered typical graphite lattice structure.
40 50 60 70 80

2θ (degrees)
Intensity (a.u.)
Ni (111)
Ni(200)
Si(100)
a

b

68

Further spectroscopic evidence is obtained by micro-Raman. Figure 4.4b shows the
Raman spectrum corresponding to the sample analyzed in Figure S2a. The observed G-
band centered at 1581 cm
-1
is characteristic of the C-C stretching in the sp
2
structure of
graphene.

Figure 4.4 (a) Low magnification TEM image of CVD graphene films. Inset shows a selected
area electron diffraction (SAED) pattern of the few-layer graphene film. (b) Raman spectrum of
CVD graphene film.

Another band of interest, the G’-band that appears at 2697 cm
-1
, presents a fairly
symmetric lineshape that suggests the films are comprised of few layers of graphene (1-5
layers). The very low D-band intensity at 1354 cm
-1
with respect to the intensity of the G-
band, indicates a low defect density in the transferred CVD graphene.

4.3.3 Atomic force microscopy of CVD graphene, ITO and SWNT films
In general, CVD graphene obtained under the experimental conditions described
here contains 1-5 graphene layers. To further test this point, we perform height profile
measurements on the transferred films. Cross section analysis of AFM images taken on

1500 2000 2500 3000

Raman shift (cm
-1
)
b
a
69

transferred CVD graphene on glass substrates allows a quantitative estimate of the film
thickness. Figure 4.5a shows an AFM image of an opening in the graphene film with
clear edges. Figure 4.5b shows the height profile along the straight line depicted in Figure
4.5a. A height step of ~1nm can be clearly observed between the substrate surface and the
graphene film. Larger vertical distances can be found between the substrate surface and
the CVD graphene edge step as well. However, due to irregularities on Ni surface, and as
transferred graphene films may suffer from bending, folding and mechanical stress, they
may not be lying fully extended and flattened on the receiving substrate. Thus, the lowest
vertical distance within the profile edge steps can be regarded as a good estimate of the
film thickness.

Figure 4.5 (a) AFM image of a transferred CVD graphene film onto glass substrate. (b) Cross
section measurement of the height of the CVD graphene. Typical thickness exhibited by the
transferred films is found within the range 1-3 nm. (c) AFM images of the surface of CVD
graphene, ITO and SWNT films on glass. The scale bar in z-direction is 50 nm for all images.

1.0 nm
a
b
c
Graphene roughness: 0.9 nm ITO roughness: 0.7 nm SWNT roughness: 8.4 nm
70

In this case, a thickness of ~1 nm indicates the CVD graphene can be as thin as single or
bilayer graphene. In addition, domains showing a large height profile may correspond to
mechanical distortions, such in-plane film compression or film folding, or to multilayer
graphene stacks with more than 5 layers.
The thin film nature of OPV devices requires control of layer thickness and
morphology to reduce the possibility of leakage current and shorts.[22, 23] Therefore,
thickness and surface smoothness of the transparent electrode in OPVs are important for
good device performance. As a point of reference, we compared the thickness and
roughness given by CVD graphene films against single-walled carbon nanotubes
(SWNT) and ITO films, which are materials that have been amply reported in the
literature as transparent electrodes [29].
The thickness of CVD graphene obtained at the above-mentioned synthesis
conditions is on the order of 1-3 nm. In Figure 4.5c, AFM images of the CVD graphene,
a commercial 150-nm-thick ITO film (R
Sheet
= 20 Ω/sq) and a 30-nm-thick SWNT film
obtained by the filtration/PDMS transfer method (R
Sheet
= 1.1 kΩ/sq) are compared. The
r.m.s. surface roughness measured for CVD graphene, ITO and SWNT films was 0.9, 0.7
and 8.4 nm for graphene, ITO and SWNT, respectively; such values were consistent with
roughness previously reported for graphene [22], ITO [30] and SWNT [31] films.
Roughness values obtained indicate that CVD graphene was nearly 10 times
smoother and thinner than SWNT films, with surface roughness comparable to that of
ITO. In addition, CVD graphene and SWNT films passivated with poly(3,4-
ethylenedioxythiophene)-poly(styrenesulfonate) (PEDOT:PSS) showed r.m.s surface
roughness of 0.8 and 5.1, respectively.
71

4.3.4 Characterization of CVD graphene film as transparent electrode for organic
photovoltaics
High transparency is also necessary for the use of CVD graphene as a substitute
for transparent metal oxide electrodes in OPVs. Figure 4.6 shows photographs displaying
see-through areas (dotted lines) of 2 and 1.3 cm
2
of graphene films after being transferred
to glass and polyethylene terephthalate (PET), respectively. Thorough inspection of the
graphene films using scanning electron microscopy also confirmed the formation of
continuous films without any visible cracks.

Figure 4.6 (a) Transmission spectra for CVD graphene, ITO and SWNT films on
glass.Photographs showing highly transparent graphene films transferred onto glass and PET are
shown in (c) and (d), respectively.

Optical transmittance of the transferred graphene films in the visible and near
infrared range was measured using a Varian50 spectrophotometer in the wavelength
range of 400 – 1100 nm. Figure 1e depicts the wavelength dependence of the optical
transparency of the CVD graphene, ITO and SWNT films displayed in Figure 1f. For
ITO films, the transmittance peaks at 535 nm, while the transmittance increases
monotonically with the increase in wavelength of the incident light from T = 86% (at 400
b c
400 500 600 700 800 900 1000 1100
0
20
40
60
80
100
Transmittance (%)
Wavelength (nm)
CVD Graphene
ITO
SWNT
a
72

nm) to T = 95% (at 1100 nm) and from T = 70% (at 400 nm) to T = 82% (at 1100) nm, in
graphene and SWNT films respectively.
Light transmission in graphene is dictated by absorption, due to the sp
2

conjugated system. As a consequence of this, as CVD graphene films become thinner,
transparency is expected to increase. In principle, the sheet resistance of a graphene film
comprised of several graphene layers should decrease for each layer added [32]; therefore
it is expected that the thicker the film, the larger the number of layers, the smaller the
sheet resistance, but simultaneously, the lower the transparency.
We were able to tune the transparency and sheet resistance of graphene films by
varying the synthesis conditions. Figure 1g shows that highly transparent CVD graphene
films can be obtained at the expense of higher resistance. Sheet resistance as low as 230
Ω/sq (with T=72%) and optical transparency as high as 91% (with R
Sheet
= 8.3 kΩ/sq) can
be achieved, and therefore a compromise between these parameters must be met for
specific applications.

Figure 4.7 (a) Transmission spectra of CVD graphene with different sheet resistance (R
Sheet
). (b)
Comparison of R
Sheet
vs. light transmittance at 550 nm for CVD graphene and reduced GO films
reported in the literature.

70 75 80 85 90 95 100
10
-1
10
1
10
3
10
5
10
7
R
sheet
(kΩ/sq)
Transmittance (%)
CVD Graphene
Eda et al.
Blake et al.
Li et al.
Wu et al.
400 600 800 1000
0
20
40
60
80
100
Transmittance (%)
Wavelength (nm)
8.30 kΩ/sq
3.40 kΩ/sq
0.70 kΩ/sq
0.23 kΩ/sq
a
b
73

Further characterization of the CVD graphene films is shown in Figure 4.7, where
we compare sheet resistance and transparency of CVD graphene against reduced GO
films reported in the literature. Analysis of Figure 1h shows that graphene synthesized by
CVD exhibits better transparency/R
Sheet
ratio than the reduced graphene oxide films
reported so far [14, 23, 33, 34]. The transparency/R
Sheet
ratio of CVD graphene can be
further improved to yield films with 700 Ω/sq and ~ 90% transparency [27].

4.3.5 Flexible transparent electrodes: Graphene vs. ITO
To investigate the flexibility of the CVD graphene electrodes and its influence on
the performance of flexible OPV cells, we transferred CVD graphene films onto PET
substrates (Figure 4.7a) and compared the electrical conductivity of graphene and ITO
films under bending conditions. Figure 4.7b and 4.7c show AFM images of CVD
graphene (R
Sheet
= 500 Ω/sq and T = 75 %) and ITO (R
Sheet
= 25 Ω/sq, T = 86 %), on PET.
The r.m.s. surface roughness of ITO on PET was 1.1 nm, nearly 60% higher than
on glass. 100-nm-thick aluminum metal contacts were thermally deposited through a
shadow mask onto the above mentioned films. Two-probe electrical measurements were
performed on both films by direct contact of tungsten micro-probes to the aluminum
electrodes, soldering the probe tips to the aluminum pads to assure good electrical contact
on each measurement. Performing this process for each bending angle allowed us to
monitor the change in conductance of the film with the bending angle (inset figure 4.7d).
The conductance of the graphene/PET film remained virtually unperturbed by
bending (Figure 4.7d) even after several complete bending cycles and decreased by only
7.9% after 100 bending cycles.
74

Figure 4.8 (a) Photograph illustrating high flexibility of CVD graphene transferred on a PET
flexible substrate. (b) and (c) AFM images of the surface of CVD graphene and ITO films on
PET, respectively. (d) and (f) Conductance of the CVD graphene and ITO films on PET
substrates under bending conditions, respectively. The devices used to monitor the conductance
had channel width (W) = 1 mm, and length (L) = 1 mm. (e) Optical images of CVD graphene
(upper) and ITO (lower) films on PET before and after being bent at the angles specified in (b)
and (c). Arrows show the direction of the bending.

In contrast, Figure 4.7f shows three clearly defined regions that describe the
typical behavior of ITO conductance under bending conditions. For bending angles from
0° to ~130° a steady decrease in the conductance of the ITO film by three orders of
magnitude with increased bending angle was observed. Interestingly, immediately after a
critical angle (128°) conductance suddenly fell by six orders of magnitude. Finally, after
the critical angle was reached, the conductance of the film continued to decrease even

f

d

e

20 μm

20

20 μm

20 μm

b
c

ITO

Graphene

Graphene

0.5 μm

ITO

0.5 μm

0 40 80 120 160
10
-8
10
-6
10
-4
10
-2
10
0
10
2
Conductance (mS)
Bending angle (2θ)

Bending Graphene/PET
Recovery Graphene/PET
θ θ
0 40 80 120 160
10
-8
10
-6
10
-4
10
-2
10
0
10
2
Conductance (mS)

Bending angle (2θ)
Bending ΙΤΟ/ΠΕΤ
Recovery ΙΤΟ/ΠΕΤ
b
20 μm 20 μm
20 μm 20 μm
c
a
75

when bending angle decreases; an open circuit (σ ≤ 10
-12
S) was obtained after only one
bending cycle.
The fact that the conductivity of the ITO film did not recover after bending the
ITO film back to lower radius of curvature can be associated with the development of
multiple discontinuity scattering sites on the brittle ITO film that were generated by
tensile strain under bending and may further develop under compressive stress while
decreasing the bending angle. Optical microscopy images were collected on the ITO/PET
and Graphene/PET films. Figure 4.7e shows optical micrographs of graphene and ITO
films before and after the first bending cycle (0°150°0°). As can be seen, under the
microscope resolution, very pronounced cracks were developed on the ITO film, while
the graphene film remained intact. These results demonstrate the advantage of CVD
graphene in terms of mechanical flexibility over ITO films, which opens new avenues for
robust, flexible, and lightweight transparent CVD graphene electrodes in OPVs.

4.3.6 Graphene OPV cell fabrication on flexible substrate and device performance
Although CVD graphene films clearly outperform ITO as transparent conductive
electrodes on flexible PET substrates under bending conditions, it is important to
implement this material into working OPVs in order to evaluate its performance. Thus,
we fabricated OPV cells on PET substrates using graphene and ITO as transparent
electrodes, under identical experimental conditions. Graphene electrodes were fabricated
by transferring as-grown CVD graphene films onto pre-cleaned, 100 μm thick, PET
substrates. PET substrates coated with ITO were obtained from Southwall Technologies
Inc. Both substrates were solvent cleaned and passivated by spin coating a thin layer (10
76

nm) of poly(3,4-ethylenedioxythiophene)-poly(styrenesulfonate) with R
Sheet
= 1 kΩ/sq
(PEDOT:PSS).
Use of the PEDOT:PSS coating as the electron blocking layer decreased the
conductivity of the PEDOT:PSS/CVD Graphene film to 2.1 kΩ/sq, while for the
PEDOT/ITO film it remained ~1 kΩ/sq. PEDOT:PSS was expected to help mitigate the
brittle nature of the ITO electrode to enhance its performance under bending conditions,
and interestingly, PEDOT:PSS passivation of ITO was also found to improve the
rectification behavior of the devices. Finally, the planarizing effect afforded by the
PEDOT:PSS treatment is desirable to compensate for possible folding or wrinkles that
may accompany the CVD graphene film transfer process or irregular wetting between the
electrode and the cell active layers, which would yield device shorting or shunt losses.
The substrates were taken into high vacuum conditions where the organic thin
films and the aluminum cathode were consecutively deposited by thermal evaporation.
The multilayered configuration employed (Figure 1a) is given as: CVD graphene [<5 nm]
or ITO / PEDOT:PSS / Copper phthalocyanine (CuPc) [40 nm] / Fullerene (C
60
) [40 nm]
/ Bathocuproine (BCP) [10 nm] / Aluminum (Al). Aluminum cathodes were deposited
through a shadow mask with circular openings of 0.75 mm
2
. Optical excitation of the
CuPc (C
60
) leads to the donation of an electron (hole) to C
60
(CuPc) and the
photogenerated charge carriers are swept to the external contacts producing a measurable
light-generated current.
Current density vs. voltage or J(V) characteristics were measured in air at room
temperature in the dark and under spectral mismatch corrected 100 mW/cm
2
white light
illumination from an AM 1.5G filtered 300 W Xenon arc lamp (Newport Co.). Routine
77

spectral mismatch correction for ASTM G173-03 was performed using a filtered silicon
photodiode calibrated by the National Renewable Energy Laboratory (NREL) to reduce
measurement errors. Chopped monochromatic light (250 Hz, 10 nm FWHM) and lock-in
detection was used to perform all spectral responsivity and spectral mismatch correction
measurements.
We compared the J(V) characteristics of a typical photovoltaic cell obtained with
CVD graphene (R
sheet
: 3.5 kΩ/sq, T: 89%) against a typical cell obtained with an ITO
anode (R
Sheet
: 25 Ω/sq, T: 96%), that were fabricated under identical experimental
conditions. Figure 4.9a and 4.9b show semi-log (up) and linear (down) J(V) plots
obtained from CVD graphene and ITO OPV cells, respectively. Red and blue traces
correspond to the current density measured in the dark and under illumination,
respectively. The output power density of the cells (P), which is given by P = J·V, is
shown in Figures 4.9a and 4.9b as open circle traces for which the maximum point on the
curve corresponds to the maximum output power density (P
max
) of the device.
For an incident power density, P
inc
= 100 mW/cm
2
, the power conversion
efficiency (η = P
max
/P
inc
) and other performance parameters are summarized in Table 4.1.
It is clearly observed from the semi-log plots in Figures 4.9a and 4.9b that both devices
have nearly identical open circuit voltage (V
oc
) (for J=0) of 0.48 V under illumination
conditions, which suggests similar recombination behavior in both cells. Furthermore,
from Figures 4.9a, it can be seen that unlike OPVs reported for reduced GO anodes [23,
24], we did not observe large leakage current densities from any of the CVD graphene
OPV cells.
78

Figure 4.9 Logarithmic (up) and linear (down) current density and power density vs voltage
characteristics of CVD graphene (a) and ITO (b) OPV cells on PET under dark (red traces) and
100 mW/cm
2
AM1.5G spectral illumination (blue traces). The output power density of the cells is
plotted on (a) and (b) as open circle traces. The structure of the devices is given by [CVD
graphene / PEDOT / CuPc / C
60
/ BCP / Al] and [ITO / CuPc / C
60
/ BCP / Al] for CVD graphene
and ITO OPVs, respectively

-0.8 -0.4 0.0 0.4 0.8
-8
-4
0
4
8
10
-8
10
-6
10
-4
10
-2
-1
0
1
(A/cm
2
) (mA/cm
2
)
Voltage (V)

Current density
Dark (mA/cm
2
)
Light (mA/cm
2
)
ITO
Power (mW/cm
2
)
Power Density (mW/cm
2
)
-0.8 -0.4 0.0 0.4 0.8
-8
-4
0
4
8
10
-8
10
-6
10
-4
10
-2
-1
0
1
(A/cm
2
) (mA/cm
2
)
Voltage (V)

Current density
Dark (mA/cm
2
)
Light (mA/cm
2
)
Graphene
Power (mW/cm
2
)
Power Density (mW/cm
2
)
a
b
79

The J(V) characteristics of the CVD graphene cell under illumination showed a
short-circuit photocurrent density (J
sc
) (for V=0) of 4.73 mA/cm
2
, an open-circuit voltage
(V
oc
) of 0.48 V and a maximum power (P
max
) of 1.18 mW/cm
2
, to yield a fill factor (FF)
of 0.52 and overall power conversion efficiency (η) of 1.18%. The control device, using
an ITO anode on PET, gave J
sc
of 4.69 mA/cm
2
, V
oc
of 0.48 V and P
max
of 1.27 mW/cm
2
,
for a FF of 0.57 and an efficiency of 1.27%. Analysis of figures 4.9a and 4.9b reveals
that despite the lower transparency and higher R
Sheet
of the CVD graphene electrode,
CVD graphene solar cell exhibits an output power density nearly 93% of that shown by
the ITO device. We also observed that CVD graphene OPV cells were more sensitive to
the anode conductivity, and hence, to its capacity to pull holes from the active layers than
to its transparency.
The fact that the two cells gave very similar device performance is encouraging,
especially considering that the ITO substrate gave ~100-fold lower R
Sheet
and higher
transparency than the CVD graphene film, which would favor the performance of the
ITO device [35]. This may be rationalized by considering that, as demonstrated above,
the sheet resistance increases to similar values on both electrodes after being coated with
PEDOT:PSS. In this case, charge injection from the active layers of the OPV cells may
be limited by the PEDOT:PSS layer, thus yielding similar performance on both cells. We
fabricated OPV cells on PET/PEDOT:PSS substrates without graphene or ITO and all of
them produced open circuit characteristics. Although PEDOT:PSS was used on both,
graphene and ITO OPV cells, the performance of the cells was measured by puncturing
the PEDOT:PSS layer to contact the underlying electrode material, which confirms that
80

CVD graphene and ITO anodes, instead of PEDOT:PSS are the ultimate electrodes in the
hole extraction process of the devices.

Anode
J
sc
(mA/cm
2
)
V
oc
(V) FF η η η η
CVD graphene 4.73 0.48 0.52 1.18
ITO 4.69 0.48 0.57 1.27

Table 4.1. Performance details of OPV cells built on PET. The structure of the devices is given by
[CVD-graphene/PEDOT/CuPc/C60/BCP/Al] and [ITO/PEDOT/CuPc/C60/BCP/Al] for CVD
graphene and ITO OPVs, respectively.

4.3.7 Device performance model: The Lambert-W function
To estimate the impact of resistive losses on device performance the J(V)
dependence under illumination was modeled according to a modified form of the
Shockley equation, which is commonly applied to describe the current density (J) vs.
voltage (V) characteristics of organic solar cells, given by:

ph
p
s
t
s
s
J
R
JR V
nV
JR V
J J −
−
+






−







 −
= 1 exp (4.1)

where R
s
, R
p
, J
s
, J
ph
, n, and V
t
are the lumped series resistance, lumped parallel
resistance, reverse-bias saturation current-density, photocurrent-density, diode ideality
factor, and thermal voltage respectively for a single diode circuit model.
81

The transcendental nature of equation 4.1 limits the optimization of the model
parameters. As a practical matter, equation 4.1 was resolved by expressing it in terms of
the Lambert-W function [36]. Rearranging equation 4.1 gives:

J +
(J
ph
+ J
s
)R
p
−V
R
p
+ R
s








exp
R
s
nV
t
J −
V
R
p
+ R
s
















=
J
s
R
p
R
p
+ R
s
exp
R
p
nV
t
V
R
p
+ R
s








(4.2)

After multiply both sides of equation 4.2 by R
s
/nV
t
exp(α/β), where α = (J
ph
+ J
s
)R
s
R
p
and
β = nV
t
(R
s
+R
p
) we define the quantities

w ≡
R
s
nV
t
J +
(J
ph
+ J
s
)R
p
−V
R
p
+ R
s








(4.3)

and

x ≡
R
s
nV
t
R
p
R
p
+ R
s
J
s
exp
R
p
nV
t
V + (J
ph
+ J
s
)R
s
R
p
+ R
s





 . (4.4)

Since w=W
0
(x) and W(x)e
W(x)
=x(V), substituting and solving for J gives

) (
) (
) (
) (
exp
) (
0
p s
s ph p
p s
s ph s
t
p
p s t
p s s
s
t
R R
V J J R
R R
J J R V
nV
R
R R nV
R R J
W
R
nV
J
+
− +
−


















+
+ +
+
= (4.5)
82

Where W
0
represents Lambert’s function of the form W(x)e
W(x)
=x(V) [36-39], which
expresses the measured current-density dependence on applied voltage in terms of the
model parameters for a single diode equivalent circuit model. The ProductLog[z]
implementation of the Lambert-W function in Mathematica 5.1 was used to perform
nonlinear regression according to equation 4.5 to obtain the model parameters for CVD
graphene and ITO based OPV devices on PET substrates.

Figure 4.10 Comparison of the modeled (solid lines) current density and power density curves of
the graphene and ITO devices obtained from the Shockley equation against the experimentally
(dots) obtained values.

In Figure 4.10 the modeled J(V) and output power density obtained according to
equation 4.5 are plotted as solid lines for the CVD graphene and ITO cells, depicted in
Figures 4.9a and b, respectively. The modeled data are compared against the
experimentally measured values and plotted as open symbols in Figure 4.10. The strong

0.0 0.1 0.2 0.3 0.4 0.5
-5
-4
-3
-2
-1
0
0.0
0.5
1.0
1.5
ITO
Graphene
Voltage (V)

Current density (mA/cm
2
)
Power density (mW/cm
2
)
83

similarity between the graphene and ITO device demonstrates that the CVD graphene
based devices may be described by the generalized Shockley equation in the same way
that their ITO based counterparts are commonly discussed.
Modeling the data in this way allows us to estimate to what extent series resistive
losses, parallel conductance, and recombination processes may impact device
performance. The model ideality factors, parallel resistances and saturation current-
densities were all comparable for the ITO and CVD graphene devices under illumination,
having values of n = 2.4 and 2.6, R
p
=1.47 kΩcm
2
and 1.62 kΩcm
2
, and J
s
=2.0 μA/cm
2

and 3.1 μA/cm
2
,
respectively, suggesting that the recombination and leakage processes
are similar for both devices.
The model series resistance calculated from equation 4.5 for the CVD graphene
device is 12.6 Ωcm
2
, which is less than 5 times that of the ITO device with R
s
= 2.6
Ωcm
2
, while the model photocurrent density (J
ph
) for the CVD graphene device (4.75
mA/cm
2
) is higher than J
ph
for the ITO device (4.66 mA/cm
2
). This indicates that the
power output of the graphene based device is primarily limited by charge transport losses
rather than optical transmittance losses. This constitutes a very promising result for CVD
graphene transparent electrodes, which perform comparably to ITO, despite carrying a
relatively higher sheet resistance.

4.3.8 Performance of highly-flexible graphene OPV cell under bending
Given the good performance of OPVs with graphene electrodes, the question
remains if such devices will perform well under strain-stress conditions. Current-voltage
84

characteristics under bending of CVD graphene and ITO solar cells are shown in Figures
4.11a and b, respectively.

Figure 4.11 Current density vs. voltage characteristics of CVD graphene (a) or ITO (b)
photovoltaic cells under 100 mW/cm
2
AM1.5G spectral illumination for different bending angles.
Insets show photographs of the experimental set up employed in the experiments.

a
-0.4 0.0 0.4 0.8
-4
-2
0
2
4

Current density (mA/cm
2
)
Voltage (V)

Dark
2θ = θ = θ = θ = 0
2θ = θ = θ = θ = 83
2θ = θ = θ = θ = 138
CVD-G
Bending angle(2Θ )=138° ° ° °
Θ
Light
Bending angle(2Θ )=138° ° ° °
Θ
Bending angle(2Θ )=138° ° ° °
Θ
Light
-0.4 0.0 0.4 0.8
-4
-2
0
2
4

Current density (mA/cm
2
)
Voltage (V)

Dark
2θ = θ = θ = θ = 0
2θ = θ = θ = θ = 83
2θ = θ = θ = θ = 138
CVD-G
Bending angle(2Θ )=138° ° ° °
Θ
Light
Bending angle(2Θ )=138° ° ° °
Θ
Bending angle(2Θ )=138° ° ° °
Θ
Light

-0.4 0.0 0.4 0.8
-4
0
4

Dark
2θ = θ = θ = θ = 0
2θ = θ = θ = θ = 36
2θ = θ = θ = θ = 60
Voltage (V)
Current density (mA/cm
2
)

ITO
Light
Bending angle(2Θ )=36° ° ° °
Θ
Light
Bending angle(2Θ )=36° ° ° °
Θ
Bending angle(2Θ )=36° ° ° °
Θ
-0.4 0.0 0.4 0.8
-4
0
4

Dark
2θ = θ = θ = θ = 0
2θ = θ = θ = θ = 36
2θ = θ = θ = θ = 60
Voltage (V)
Current density (mA/cm
2
)

ITO
Light
Bending angle(2Θ )=36° ° ° °
Θ
Light
Bending angle(2Θ )=36° ° ° °
Θ
Bending angle(2Θ )=36° ° ° °
Θ
b
85

We observed that the performance of both devices was slightly degraded upon
bending. For instance, solar cells using CVD graphene electrodes withstood bending
angles (curvature radii, surface strain) up to 138° (4.1 mm, 2.4%) while exhibiting good
solar cell performance. In sharp contrast, ITO cells only withstood bending to 36° (15.9
mm, 0.8%) while showing poor performance, and failed completely to become an open
circuit after being bent to 60° (9.5 mm, 1%).
It is important to note that, with increased bending angle, the current density
dropped for CVD graphene and ITO devices, while their open circuit voltage remained
virtually unchanged. In some cases this effect can be associated with decreased
illumination of the devices during bending. However, as both cells are subjected to
similar bending conditions, the marked difference exhibited in the conversion efficiency
between them cannot be attributed to irregular illumination induced by bending, but may
be related to the presence of micro cracks on the ITO device.
To further investigate this, we plotted the fill factor vs. the bending angle of the
OPV cells with CVD graphene and ITO electrodes (Figure 4.12a). The fill factor
(FF=P
max
/J
sc
V
oc
) depends strongly on the output power of the cell, and is directly related
to the cell conversion efficiency (η) by
η = FF
J
sc
V
oc
P
inc
×100 . (4.6)
Gradual degradation of the initial fill factor, and hence, the conversion efficiency was
observed on the CVD graphene cell as the bending angle increased; in contrast, the fill
factor of the ITO device rapidly decayed to zero when bent at around 60°.
Furthermore, we performed SEM measurements to investigate changes in film
morphology that may have been introduced by bending of the devices. Figure 4.12b
86

shows the appearance of micro-cracks throughout the ITO device, while no signs of
micro-cracks or fissures were observed on the graphene device. Development of micro-
cracks generated by mechanical stress in ITO, even at small bending angles, can
substantially increase the film resistance, which has a key impact in reducing the fill
factor.

Figure 4.12 (a) Fill factor dependence of the bending angle for CVD graphene and ITO devices
shown in figure 4.12 (b) SEM images showing the surface structure of CVD graphene (up) and
ITO (down) photovoltaic cells after being subjected to the bending angles described in figure 4.12.

This agrees well with the observed decrease in output current density and power
conversion efficiency of the solar cells without observing appreciable change in the V
oc
.
CVD graphene, being of organic nature and more flexible, surpasses the performance of
ITO, which may easily crack under slight bending albeit PEDOT:PSS passivation.
Therefore, the brittle nature of ITO plays a major role in the resulting poor performance
of ITO-flexible organic solar cells, while the CVD graphene thin films exhibited good
performance as flexible transparent electrodes.

0 20 40 60 80 100 120 140
0.0
0.1
0.2
0.3
0.4
0.5
0.6
Fill Factor
Bending Angle (2θ)

Graphene
ITO
a b
Graphene/PET
ITO/PET
100 μm 50 μm
50 μm 100 μm
87

4.3.9 CVD graphene organic photovoltaic cells on rigid substrates
In order to explore the performance of graphene OPVs on rigid substrates, we
fabricated solar cells on CVD graphene films transferred on glass (R
Sheet
: 1.2 kΩ/sq, T:
82% at 550 nm) and glass substrates coated with ITO (Thickness: 150 nm, R
Sheet
: 20
Ω/sq, T: 84% at 550 nm). J(V) characteristics of the fabricated devices are plotted in
semi-log (figure 4.13a) and linear (Fig. 4.13b) scale for both devices.
Again, CVD graphene solar cells on rigid transparent substrate showed distinct
diode behavior with little leakage current at reverse bias in the dark, while exhibiting
high dark current under forward bias. Under illumination, the ITO device gives J
sc
of
5.41 mA/cm
2
, V
oc
of 0.47 V, FF of 0.54 and power conversion efficiency (η) of 1.39%.
On the other hand, the CVD graphene device exhibited J
sc
of 3.45 mA/cm
2
, V
oc
of 0.47 V,
FF of 0.47 and (η) of 0.75%.
Comparison of these devices shows that even though conversion efficiency of the
CVD graphene device is lower, the overall performance of the CVD graphene
photovoltaic cell is competitive, with FF comparable to that of the control ITO device.
Higher transparency of the ITO film may lead to a higher exciton generation rate, which
in turn is reflected in higher J
sc
values. However, smoothness and thickness of the
graphene film may favor charge injection and transport. The disparate power conversion
efficiency observed between the two cells can be attributed to the higher sheet resistance
and lower transparency of the graphene electrode in the G-OPV.

88

Figure 4.13 Logarithmic (a) and linear (b) current density vs voltage characteristics of CVD
GRAPHENE and ITO photovoltaic cells on glass under 100 mW/cm
2
AM1.5 spectral
illumination. Structure of the devices is given by [CVD GRAPHENE / PEDOT / CuPc / C
60
/
BCP / Al] and [ITO / CuPc / C
60
/ BCP / Al] for CVD GRAPHENE and ITO OPVs, respectively.

Table 4.2 summarizes representative solar cell performance parameters measured for the
CVD graphene and ITO cells as well as reduced GO devices reported in the literature.
-0.4 -0.2 0.0 0.2 0.4 0.6
-6
-4
-2
0
2
4
6
Voltage (V)

Current density (mA/cm
2
)
ITO Dark
ITO Light
Graphene Dark
Graphene Light

b
a
-0.4 -0.2 0.0 0.2 0.4 0.6 0.8
10
-8
10
-6
10
-4
10
-2
10
0
Current density (A/cm
2
)
Voltage (V)
Graphene Dark
Graphene Light
ITO Dark
ITO Light
89

Clearly, CVD graphene in all cases, on rigid or flexible substrates, compares favorably
against reduced GO as transparent anode in OPV cells.

Anode
J
sc
(mA/cm
2
)
V
oc
(V) FF η η η η (%)
CVD graphene 3.45 0.47 0.47 0.75
ITO 5.41 0.47 0.54 1.39
Red. GO (Wu et al.) 2.10 0.48 0.34 0.40
Red. GO (Wang et al.) 1.00 0.70 0.36 0.26

Table 4.2, Comparison of performance details of OPV cells built on glass substrates. The
structure of the devices is given by [CVD GRAPHENE/PEDOT/CuPc/C60/BCP/Al] and
[ITO/CuPc/C60/BCP/Al] for CVD GRAPHENE and ITO OPVs, respectively.

4.4 Conclusions
In this chapter, we demonstrate a feasible, scalable and effective method to
employ CVD graphene as highly transparent, continuous and flexible electrodes for
OPVs. This approach constitutes a significant advance towards the production of
transparent conductive electrodes in solar cells. CVD graphene meets the most important
criteria of abundance, low cost, conductivity, stability, electrode/organic film
compatibility and flexibility that are necessary to replace ITO in organic photovoltaics,
which may have important implications for future organic optoelectronic devices.

90

Chapter 4 References
1. Peumans, P., A. Yakimov, and S.R. Forrest, Small molecular weight organic thin-
film photodetectors and solar cells. Journal of Applied Physics, 2003. 93(7): p.
3693-3723.

2. Andersson, A., et al., Fluorine Tin Oxide as an Alternative to Indium Tin Oxide in
Polymer LEDs. Advanced Materials, 1998. 10(11): p. 859-863.

3. Jansseune, T., Indium price soars as demand for displays continues to grow
Compound Semiconductor, 2005. 11: p. 34-35.

4. Scott, J.C., et al., Degradation and failure of MEH-PPV light-emitting diodes.
Journal of Applied Physics, 1996. 79: p. 2745-2751.

5. Boehme, M. and C. Charton, Properties of ITO on PET film in dependence on the
coating conditions and thermal processing. Surface & Coatings Technology
2005. 200 p. 932- 935.

6. Wu, Z., et al., Transparent, Conductive Carbon Nanotube Films. Science, 2004.
305: p. 1273-1276.

7. Barnes, T.M., et al., Single-wall carbon nanotube networks as a transparent back
contact in CdTe solar cells. Applied Physics Letters, 2007. 90: p. 243503.1-
243503.3.

8. Rowell, M.W., et al., Organic solar cells with carbon nanotube network
electrodes. Applied Physics Letters, 2006. 88: p. 233506.1-233506.3

9. Kang, M.-G. and L.J. Guo, Nanoimprinted Semitransparent Metal Electrodes and
Their Application in Organic Light-Emitting Diodes. Advanced Materials, 2007.
19: p. 1391-1396.

10. Lee, J.-Y., et al., Solution-Processed Metal Nanowire Mesh Transparent
Electrodes. Nano Letters, 2008. 8: p. 689-692.

11. Zhang, Y., et al., Experimental observation of the quantum Hall effect and
Berry’s phase in graphene. Nature, 2005. 438: p. 201-204.

12. Nair, R.R., et al., Fine Structure Constant Defines Visual Transparency of
Graphene. Science, 2008. 320: p. 1308.

13. Geim, A.K. and K.S. Novoselov, The rise of graphene. Nat Mater, 2007. 6(3): p.
183-191.

91

14. Blake, P., et al., Graphene-Based Liquid Crystal Device. Nano Letters, 2008.
8(6): p. 1704-1708.

15. Tan, Y.W., et al., Measurement of Scattering Rate and Minimum Conductivity in
Graphene. Physical Review Letters, 2007. 99(24): p. 246803.1-246803.4

16. Echtermeyer, T.J., et al., Graphene field-effect devices. The European Physical
Journal - Special Topics, 2007. 148: p. 19-26.

17. Forbeaux, I., J.-M. Themlin, and J.-M. Debever, Heteroepitaxial graphite on 6H-
SiC(0001): Interface formation through conduction-band electronic structure.
Physical Review B, 1998. 58(24): p. 16396-16406.

18. Novoselov, K.S., et al., Electric Field Effect in Atomically Thin Carbon Films.
Science, 2004. 306: p. 666-669.

19. Viculis, L.M., J.J. Mack, and R.B. Kaner, A Chemical Route to Carbon
Nanoscrolls. Science, 2003. 299: p. 1361.

20. Gilje, S., et al., A Chemical Route to Graphene for Device Applications. Nano
Letters, 2007. 7(11): p. 3394-3398.

21. Eda, G., et al., Transparent and conducting electrodes for organic electronics
from reduced graphene oxide. Applied Physics Letters 2008. 92: p. 233305.1-
233305.3.

22. Wang, X., L. Zhi, and a.K. Mllen, Transparent, Conductive Graphene Electrodes
for Dye-Sensitized Solar Cells. Nano Letters, 2008. 8(1): p. 323-327.

23. Wu, J., et al., Organic solar cells with solution-processed graphene transparent
electrodes. Applied Physics Letters 2008. 92: p. 263302.1-263302.3.

24. Tung, V.C., et al., Low-Temperature Solution Processing of Graphene#Carbon
Nanotube Hybrid Materials for High-Performance Transparent Conductors.
Nano Letters, 2009. 9: p. 1949–1955.

25. Gomez, L., et al., Synthesis, Transfer and Devices of Single- and Few-Layer
Graphene by Chemical Vapor Deposition. IEEE Transactions on
Nanotechnology, 2009. 8(2): p. 135-138.

26. Yu, Q., et al., Graphene segregated on Ni surfaces and transferred to insulators.
Applied Physics Letters, 2008. 93, : p. 113103.1-113103.3.

27. Reina, A., et al., Large Area, Few-Layer Graphene Films on Arbitrary Substrates
by Chemical Vapor Deposition. Nano Letters, 2008. 9 (1): p. 30-35.

92

28. Kim, K.S., et al., Large-scale pattern growth of graphene films for stretchable
transparent electrodes. Nature, 2009. 457: p. 706-710.

29. Sangeeth, C.S.S., J. Manu, and M. Reghu, Charge transport in transparent
conductors: A comparison. Journal of Applied Physics, 2009. 105(6): p.
063713.1-063713.6.

30. Kim, K.-B., et al., Relationship between Surface Roughness of Indium Tin Oxide
and Leakage Current of Organic Light-Emitting Diode. Japanese Journal of
Applied Physics, 2003. 42: p. L438-L440.

31. Zhang, D., et al., Transparent, Conductive, and Flexible Carbon Nanotube Films
and Their Application in Organic Light-Emitting Diodes. Nano Letters, 2006.
6(9): p. 1880-1886.

32. Li, X., et al., Transfer of Large-Area Graphene Films for High-Performance
Transparent Conductive Electrodes. Nano Letters, 2009. 9(12): p. 4359-4363.

33. Eda, G., G. Fanchini, and M. Chhowalla, Large-area ultrathin films of reduced
graphene oxide as a transparent and flexible electronic material. Nature
Nanotechnology, 2008. 3: p. 270-274.

34. Li, X., et al., Highly conducting graphene sheets and Langmuir-Blodgett films.
Nat Nano, 2008. 3(9): p. 538-542.

35. Xuan, W., et al., Transparent Carbon Films as Electrodes in Organic Solar
Cells13. Angewandte Chemie International Edition, 2008. 47(16): p. 2990-2992.

36. Hayes, B., Why W? American Scientist, 2005. 93: p. 104-108.

37. Banwell, T.C. and A. Jayakumar, Exact analytical solution for current flow
through diode with series resistance. Electronics Letters 2000. 36(4): p. 291-292.

38. Jain, A. and A. Kapoor, Exact analytical solutions of the parameters of real solar
cells using Lambert W-function. Solar Energy Materials and Solar Cells, 2004.
81(2): p. 269-277.

39. Ortiz-Conde, A., F.J. García Sánchez, and J. Muci, Exact analytical solutions of
the forward non-ideal diode equation with series and shunt parasitic resistances.
Solid-State Electronics, 2000. 44(10): p. 1861-1864.

93

Chapter 5 Metal to Semiconductor Ratio of Aligned Carbon Nanotubes on
a-Sapphire

5.1 Introduction
The unique electronic, mechanical, and optical properties of carbon nanotubes
relate closely to their atomic structure. Single-walled carbon nanotubes are fully defined
by their diameter and chiral angle, which determines whether the tube is ordered or
defective, a metal or semiconductor and its band gap. With the advent of aligned
nanotubes by direct synthesis on crystalline insulator substrates, it is important to
understand the alignment mechanism and to evaluate the influence of aligned growth on
the ultimate structure and properties of the nanotubes obtained.

Figure 5.1 (a) real time TEM imaging of the early stages of nanotube growth [1]. (b) Schematic
representation showing the three stages of nanotube growth. I: Carbon saturation, II: detachment
of sp
2
graphene and III: reshaping of catalyst and nanotube elongation.

I II III
a
b
94

Figure 5.1 shows the early stages of nanotube growth, where there is a strong
interaction between carbon atoms, catalyst particle and substrate surface. On the other
hand, nanotube misalignment represents a serious drawback toward large scale
fabrication of nanotube-based high performance devices [2, 3]
Different methods have been developed to produce aligned single-walled carbon
nanotubes (SWNTs) arrays, among which the epitaxial approach has emerged as a
scalable process to produce massively aligned nanotubes on insulating substrates such as
sapphire and quartz [2, 4-8]. Epitaxial growth alignment can be understood by
considering the presence of strong binding energies localized along low potential energy
directions, defect sites and atomic steps on the substrate surface [9]. In particular,
alignment of SWNTs grown by chemical vapor deposition (CVD) on a-plane sapphire
has been explained as a result of the interaction along the pseudo-one-dimensional array
of aluminum atoms on the substrate surface [10]. There is, however, controversy, as
similar aligned growth was observed on quartz substrates and explained as guided growth
following the substrate step edges [6].
Although it is well known that intermolecular interactions play an important role
in defining the growth direction of aligned nanotubes, little effort has been employed to
investigate the influence of alignment on the nanotube structure and properties. In
addition, there has been a lack of information about the metal-to-semiconductor ratio and
the effect of the nanotube-substrate interaction on Raman bands for such aligned
nanotubes.
In this chapter, we present our work on combined micro-fabrication techniques
and multi-wavelength resonance micro Raman spectroscopy (532 nm, 633 nm, and 785
95

nm) to characterize as-grown aligned nanotubes on a-plane sapphire and shed light on the
above-mentioned scientific issues: the alignment mechanism (lattice or step-edge guided
growth), the metal-to-semiconductor ratio, and substrate surface influence in alignment
and straightness of as-grown nanotubes on a-plane sapphire.

5.2 Experiments
Aligned SWNTs used in this study were synthesized on a-plane sapphire by a
CVD method. Ferritin protein (Alpha Aesar, Inc.) was used as the iron source for catalyst
particles. Growth of aligned nanotubes was fulfilled by flowing 2000 sccm of CH
4
, 10
sccm of C
2
H
4
and 600 sccm of H
2
at 900 °C. Simultaneous control of nanotube
orientation and position was achieved by patterning of catalyst at desired sites on the
sapphire substrates with the use of photolithographic techniques. Photolithography
allowed opening windows on the sapphire substrates were catalyst solutions were
allowed to dry. Subsequent photoresist removal leaves the patterned catalyst islands that
allow control over nanotube density and positioning, instrumental factors for accurate
nanotube characterization.
After synthesis, atomic force microscopy (AFM) and field-emission scanning
electron microscopy (FESEM) combined with micro Raman spectroscopy were used to
provide structural and spectroscopic information of the samples at individual nanotube
level. Altogether eight samples from four rounds of CVD growth were used in this study,
and statistical analysis was derived from usually over 150 nanotubes from multiple
samples and confirmed to be consistent from sample to sample.

96

5.3 Results
Atomic force microscopy images were taken on a-Sapphire substrates before and
after carbon nanotube growth by chemical vapor deposition. Figure 5.2a shows an AFM
image of a clean a-plane sapphire substrate where atomic steps can be observed along the
[
1
101] lattice direction. AFM image in Figure 5.2b shows that as-grown nanotubes did
not align along the substrate atomic steps direction but rather followed the [1
1
00]
direction on the a-face. In addition, it can be observed that figure 5.2b shows that a few
nanotubes had short sections exhibiting alignment along the atomic steps direction
[
1
101], this secondary alignment direction suggests the presence of competitive
alignment directions or competitive mechanisms of alignment of nanotubes on a-
sapphire.
Competition between lattice-directed and atomic-step-templated alignment
mechanisms has been reported for nanotubes grown on substrates with different miscut
angles [10]. However, in our case the substrates lacked of intentional miscut and a lattice
oriented alignment mechanism was strikingly favored over the atomic step one.
Furthermore, Figure 5.2c depicts the top and side views of the a-sapphire atomic layout
as well as a schematic view showing the nanotube principal alignment direction. Low van
der Waals energy grooves of about 4.34 Å wide and 0.91 Å deep present between oxygen
atoms on the upmost layer of the sapphire substrate along the [1
1
00] direction, provide a
high binding energy path that, added to the lack of miscut on the substrate, could explain
the favored lattice-directed alignment.

97

Figure 5.2 (a) AFM image of a-sapphire substrate showing c-axis and the atomic step
direction. (b) AFM image of CVD-grown SWNTs on a-sapphire. (c) Top (upper) and side (lower)
views of the a-sapphire surface atomic structure. (d) G band Raman intensity dependence on the
polarization angle. (e) cos
2
(θ) fit to the plot of the normalized G band intensity v.s. the
polarization angle (scattered dots).

Polarized micro Raman was used to evaluate the SWNTs alignment along [
1
101]
direction [11]. Figure 1d shows plots of G band intensity as a function of the angle of
polarization (θ). It is observed that the Raman signal is strongly suppressed when the
excitation laser is polarized perpendicular to the nanotube axis, following the antenna
effect. As the probability of the absorption and emission processes varies linearly with
the intensity of the incident field, which in turn varies as cos
2
(θ), it is expected that the
Raman G band signal show a close fit to this polarization dependence [6, 11].

This
dependence is clearly shown in Figure 1e and remained throughout the samples. The
(e)
(c)
(b) (a)
(d)
1540 1560 1580 1600 1620 1640
Intensity (arb. units)
Ram an shift (cm
-1
)
0°
20°
40°
60°
80°
(d)
(e)
(c)
[11 ¯ 00]
[11 ¯ 00]
(a)
[1 ¯ 101]
[0001]
0.2 µm 2 µm
(b)
[11 ¯ 00]
[1 ¯ 101]
[0001]
0.0 0.2 0.4 0.6 0.8
0.0
0.3
0.5
0.8
1.0
---- cos
2
(θ θ θ θ)

Intensity (Normalized)
Rotation angle (radians)
98

results obtained reveal a high degree of unidirectional alignment from the collectivity of
nanotubes along [
1
101] of the a-plane, therefore confirming lattice-guided alignment as
the predominant alignment mechanism on a-plane sapphire without intentional miscut.
Now that the alignment mechanism has been elucidated, we further used micro
Raman to determine the metal-to-semiconductor ratio for the aligned nanotubes. The
distribution of the electronic nature of isolated carbon nanotubes was investigated by
analyzing the tangential vibration modes (G
-
and G
+
bands) and the RBM frequencies of
carbon nanotubes grown from patterned catalyst particles between metallic electrodes.
Figure 5.3a shows optical and SEM images of nanotubes aligned between source and
drain electrodes. Raman spectra were acquired on the device channels with a spatial
resolution of ~0.5 μm. G
-
band for semiconducting nanotubes exhibited a typical
Lorentzian lineshape, while metallic nanotubes showed a broadened Breit-Wigner-Fano
(BWF) G
-
lineshape due to the presence of free electrons in the conduction band [12, 13].

FESEM images and micro Raman provided enough resolution to correlate
individual nanotubes with their Raman spectra and determine the number of nanotubes
exhibiting semiconducting or metallic characteristics in the samples. Figures 5.3b and
5.3c show typical micro Raman spectra in the G and RBM band frequency regions for
individual aligned nanotubes, respectively. We correlated nanotube diameters with their
RBM frequency by: d
t
= A/(ω
RBM
– B), where d
t
is the nanotube diameter, ω
RBM
is the

RBM frequency, A = 248 , and B = 0 [12].

99

Figure 5.3 (a) Optical and SEM images of a device, showing aligned nanotubes between
patterned source and drain electrodes. Scale bar is 2 μm. RBM (b) and G band (c) of several
typical nanotubes scanned with lasers of 785 nm, 633 nm and 532 nm in wavelength from top to
bottom, respectively.

The process of a typical assignment is described for the spectra at the bottom of
Figures 5.3b and 5.3c. For this nanotube, the G band lineshape reveals a semiconducting
nature and its RBM frequency (174.65 cm
-1
) obtained with a 2.33 eV (532 nm) laser
corresponds to a resonant semiconducting nanotube with d
t
= 1.42 nm. This assignment
was confirmed by calculating the bandgap energies of a nanotube with this diameter.
1400 1500 1600 1700
785 nm
633 nm
Raman shift (cm
-1
)
532 nm
a
100 150 200 250 300 350
532 nm
633 nm

Raman shift (cm
-1
)
785 nm
SC M SC
SC M SC
M SC M
c
b
100

Tight binding calculations and tunable Raman spectroscopy show that only
semiconducting nanotubes for which E
33
~ 2.30 eV will be resonantly excited by this
laser [13]. Applying this procedure to more than 150 nanotubes and using lasers with
energies 1.58 eV (785 nm), 1.98 eV (633 nm) and 2.33 eV (532 nm) we determined the
averaged percentage of metallic nanotubes, for two different CVD grown samples, to be
about 27.9±0.6%; corresponding to a ratio between aligned metallic and semiconducting
nanotubes of 1:2.6, which is moderately lower than the theoretically predicted 1:2 ratio
[14].
Now that the metallic to semiconducting ratio of aligned nanotubes on a-sapphire
has been determined, below we used micro Raman spectroscopy to probe nanotube-
substrate interactions present in aligned growth. Raman G’ band intensity and lineshape
of carbon nanotubes can be related to strain and external perturbation [15]. In order to
probe the effect of nanotube-substrate interactions on Raman G’ band, aligned nanotubes
were grown on as-received (i.e., unannealed) and annealed a-plane sapphire substrates.
The annealing condition was 900 ˚C in air for 13 hours.
Figures 5.4a and b show FESEM images of aligned nanotubes synthesized on
unannealed and annealed a-sapphire, respectively. Comparison of Figures 5.4a and b
reveals a higher degree of straightness on nanotubes grown on annealed a-sapphire, but at
the same time, alignment of small fractions of nanotubes in directions other than [1
1
00]
becomes evident. That is because the annealing process can induce reconstruction on a-
sapphire surface [16], leading to improved surface atomic ordering along [1
1
00] and thus
straighter nanotubes. At the same time, annealing can also induce more pronounced step
101

edges, therefore yielding nanotube segments along other directions. We note that even for
annealed samples, the predominant alignment direction is still [1
1
00].

Figure 5.4 FESEM images of aligned nanotubes synthesized on unannealed (a) and annealed (b)
a-sapphire. Deviation angles from the direction [1
1
00] (θ
D
) in nanotube segments aligned in that
direction were typically 10 times lower in nanotubes grown on annealed sapphire, evidencing
higher straightness than nanotubes grown on unannealed sapphire.

On the other hand, the unannealed a-sapphire surface consists of irregular
corrugations and scratches that allow nanotubes to hop between vicinal and disordered
pseudo-unidimensional surface potential grooves, which may result in lowered
straightness in the nanotubes as shown in Figure 5.4a.

1 μm
(a)
3 μm
3 μm
(b)
102

Figure 5.5 Representative G’ band spectra for SWNTs grown on unannealed (upper) and
annealed (lower) a-sapphire. Insets: G’ FWHM distribution measured by Lorentzian fitting of
Raman peaks.

To further confirm the importance nanotube-substrate interactions in the
straightness of aligned nanotubes, we analyzed the G’ band linewidth distribution of
aligned nanotubes grown on unannealed and annealed a-sapphire. Upper and lower
panels of Figure 5.4c show representative G’ band spectra of SWNTs grown on
2500 2600 2700 2800
Raman shift (cm
-1
)
39.3 cm
-1
20 30 40 50
G' linewidth (cm
-1
)
2500 2600 2700 2800
Raman shift (cm
-1
)
28.5 cm
-1
20 30 40 50
G' linewidth (cm
-1
)
103

unannealed and annealed a-sapphire respectively. The average full width at half
maximum (FWHM) of the G’ band increased from 30.3 ± 5.0 cm
-1
to 35.1 ± 5.9 cm
-1
, a
total of around 4.8 cm
-1
for nanotubes aligned on annealed a-sapphire. Line broadening of
this band can occur as a result of a perturbation exerted on the nanotubes due to
nanotube-substrate van der Waals interactions. Thus, an improved surface atomic
ordering favors straightness on nanotubes due to a stronger substrate-SWNT interaction,
along the alignment direction.
The effect of the strength of nanotube-substrate interactions in Raman low
frequency modes is shown in Figure 5.6. Figures 5.6a and b show the intensity profile of
the resonant Raman RBM and G bands with respect to the scan position of isolated
aligned nanotubes grown on as-received a-plane sapphire. While 15 nanotubes exhibited
strong G band in Figure 4a, only 9 of them exhibited clearly distinguishable RBM peaks.
On the other hand, spectral maps of nanotubes grown on annealed a-Sapphire are
shown in Figures 5.6c and d. Analysis of Figure 5.6c revealed that only 4 out of 14
nanotubes in the scanned area had distinguishable RBM bands. We have consistently
observed that the percentage of nanotubes showing distinguishable RBM bands is lower
for nanotubes grown on annealed sapphire than for unannealed sapphire, and these
surface effects were found to be more evident on small diameter nanotubes.
RBM of nanotubes is a totally symmetric vibration A
1
in which all the carbon
atoms undergo an equal radial displacement [17]. This mode has been predicted by theory
to be strong, but it is prone to decrease, shift and disappear from the low frequency region
under forces such as hydrostatic pressure and intermolecular Van der Waals interactions
that can induce subtle geometrical deformations [18-20].

104

Figure 5.6 Intensity profile of the resonant Raman RBM (a) and G band (b) with respect
to the scan position of aligned nanotubes grown on unannealed a-plane sapphire. RBM peaks
have been circled for clarity. (c) and (d) show the Raman intensity profile of aligned nanotubes
on annealed sapphire for RBM and G-band regions, respectively.

Results displayed in Figure 5.6 suggest that annealed sapphire exerted stronger
interaction with the aligned nanotubes than unannealed sapphire, therefore leading to a
faster damping of the RBM vibration via mode symmetry breaking, which in turn yields
lower intensity or disappearance of the RBM bands. This conclusion is consistent with
the observation of G’ band broadening for aligned nanotubes on annealed sapphire shown
1400 1500 1600 1700 1800
0
5
10
15
20
25
30
35
40
Data set
Raman shift (cm
-1
)
(c)
100 150 200 250 300
0
5
10
15
20
25
30
35
40
Data set
Raman shift (cm
-1
)
(b)
(a)
100 150 200 250 300
0
5
10
15
20
25
30
35
40
Data set
Raman shift (cm
-1
)
(d)
1400 1500 1600 1700 1800
0
5
10
15
20
25
30
35
40
Data set
Raman shift (cm
-1
)
105

in Figure 5.5. By being RBM the only well resolved symmetric mode among the main
carbon nanotube bands, it also explains why other modes remain visible.

5.4 Conclusions
In summary, we synthesized aligned carbon nanotube arrays on sapphire and
showed that micro-Raman spectroscopy can be used to probe the nanotube-substrate
interactions involved in the alignment mechanism of nanotubes on substrates surfaces.
Combination of SEM, AFM and specific nanotube arrays fabrication techniques allowed
a simple and reliable way to characterize individual, aligned nanotubes in terms of their
metallic to semiconducting ratio. The anisotropic constraint exerted by the a-sapphire
substrate surface along the carbon nanotubes circumference may be responsible for RBM
damping of aligned nanotubes, as well as their G’ band line broadening. Results obtained
in this work agree with the growth mechanistic concept of surface potential interactions
in the self alignment of nanotubes on a-plane sapphire.

106

Chapter 5 References

1. Hofmann, S., et al., In situ Observations of Catalyst Dynamics during Surface-
Bound Carbon Nanotube Nucleation. Nano Letters, 2007. 7(3): p. 602-608.

2. Han, S., X. Liu, and C. Zhou, Template-Free Directional Growth of Single-
Walled Carbon Nanotubes on a- and r-Plane Sapphire. Journal of the american
chemical society, 2005. 127: p. 5296-5296.

3. Ismach, A. and E. Joselevich, Orthogonal Self-Assembly of Carbon Nanotube
Crossbar Architectures by Simultaneous Graphoepitaxy and Field-Directed
Growth. Nano Letters, 2006. 6 (8): p. 1706-1710

4. Ryu, K., et al., Synthesis of Aligned Single-Walled Nanotubes Using Catalysts
Defined by Nanosphere Lithography. Journal of the American Chemical Society,
2007. 129(33): p. 10104-10105.

5. Ismach, A., D. Kantorovich, and E. Joselevich, Carbon Nanotube Graphoepitaxy:
Highly Oriented Growth by Faceted Nanosteps. Journal of the american chemical
society, 2005. 127: p. 11554-11555.

6. Kocabas, C., et al., Guided Growth of Large-Scale, Horizontally Aligned Arrays
of Single-Walled Carbon Nanotubes and Their Use in Thin-Film Transistors.
small, 2005. 1(11): p. 1110 – 1116.

7. Ago, H., et al., Synthesis of horizontally-aligned single-walled carbon nanotubes
with controllable density on sapphire surface and polarized Raman spectroscopy.
Chemical Physics Letters 2006. 421: p. 399-403.

8. Ago, H., et al., Aligned growth of isolated single-walled carbon nanotubes
programmed by atomic arrangement of substrate surface. Chemical Physics
Letters 2005. 408: p. 433-438.

9. Somorjai, G.A., Modern Surface Science and Surface Technologies: An
Introduction. Chemical Reviews 1996. 96: p. 1223-1235.

10. Ago, H., et al., Competition and cooperation between lattice-oriented growth and
step-templated growth of aligned carbon nanotubes on sapphire. Applied Physics
Letters 2007. 90: p. 123112

11. Wang, Y., et al., Receiving and transmitting light-like radio waves: Antenna effect
in arrays of aligned carbon nanotubes. Applied Physics Letters, 2004. 85(13): p.
2607-2609.

107

12. Jorio, A., et al., Characterizing carbon nanotube samples with resonance Raman
scattering. New Journal of Physics, 2003. 5 p. 139.1-139.17.

13. Dresselhaus, M.S., et al., Raman spectroscopy on isolated single wall carbon
nanotubes. Carbon, 2002. 40: p. 2043-2061.

14. Araujo, P.T., et al., Third and Fourth Optical Transitions in Semiconducting
Carbon Nanotubes. Physical Review Letters, 2007. 98: p. 067401-1 - 067401-4.

15. Wood, J.R., et al., Carbon nanotubes: From molecular to macroscopic sensors.
Physical Review B, 2000. 62(11): p. 7571-7575.

16. Saw, K.G., Surface reconstruction of α-(0001) sapphire: An AFM, XPS, AES, and
EELS investigation and EELS investigation. Journal of Materials Science, 2004.
39 p. 2911 – 2914.

17. Rao, A.M., et al., Diameter-Selective Raman Scattering from Vibrational Modes
in Carbon Nanotubes Science, 1997. 275: p. 187-191.

18. Venkateswaran, U.D., et al., Probing the single-wall carbon nanotube bundle:
Raman scattering under high pressure. Physical Review B 15 APRIL 1999-II
VOLUME , NUMBER 1999. 59(16): p. 10928-10934.

19. Peters, M.J., et al., Structural phase transition in carbon nanotube bundles under
pressure. Physical Review B 2000. 61( 9): p. 5939-5954.

20. Yang, X., et al., Single-walled carbon nanotube bundle under hydrostatic
pressure studied by first-principles calculations. Physical Review B , , 2006. 73:
p. 235403.1-235403.6.

108

Chapter 6 Scalable Light-Induced Metal to Semiconductor Conversion of
Carbon Nanotubes

6.1 Introduction
The outstanding properties of single-walled carbon nanotubes (SWNTs) have
earned them numerous applications in different technological areas [1-3] . Carbon
nanotube field-effect transistors (CNTFETs) have acquired great importance due to their
potential to switch on and off much faster than current silicon technologies [4-6] and the
foreseen limits in the downscaling of silicon transistors [7, 8]. In spite of significant
progress made toward integrated nanotube circuits [4, 9-11], the assembly and integration
of nanotube electronics still faces significant challenges due to the coexistence of metallic
and semiconducting nanotubes in as-synthesized samples.
Different approaches have been followed to obtain CNTFETs containing only
semiconducting nanotubes in the channels either by selective synthesis [12-14], post-
synthesis separation methods [15-18], or post-synthesis methods to selectively etch
metallic nanotubes [19, 20]. Pioneering works that include the use of monochromatic
light irradiation [21] and broadband light irradiation [22] to selectively etch metallic
nanotubes have been reported. An alternative approach is to induce a metal-to-
semiconductor transition in carbon nanotubes. Electron beam irradiation and hydrogen
plasma have yielded metal-to-semiconductor conversion of SWNTs [23-25], but the
limited size of the electron beam and instability of the plasma represent limiting hurdles
for scalability.

109

In this chapter we present our work on the use of light irradiation to induce the
metal-to-semiconductor conversion of carbon nanotubes for transistors based on aligned
nanotubes and individual nanotube devices. This conversion process is easy to implement
and scalable to complete wafers.

6.2 Experiments

6.2.1 Wafer scale synthesis of aligned carbon nanotubes
Random orientation of as-grown carbon nanotubes on silicon substrates constitute
one of the main obstacles for CMOS integration; therefore aligned SWNTs were
synthesized on quartz and sapphire wafers and transferred onto silicon wafers. A thin
layer of either elemental iron or Fe (III) was deposited on sapphire and quartz substrates
prior to nanotube growth. After this, wafers were loaded inside a horizontal quartz tube of
6 inch in diameter in a heating coil furnace.
Carbon nanotube growth was performed under a gaseous mixture of H
2
, CH
4
, and
C
2
H
4
at 900 °C for 30 minutes. As shown in figure 6.1, large arrays of horizontally
aligned nanotubes with densities between 1 and 10 SWNTs/μm can be obtained at full
wafer scales. Figure 6.1 shows photographs of quartz and sapphire wafers after nanotube
growth. SEM images show the high degree of nanotube alignment that can be obtained
over full wafer scale areas on sapphire and quartz substrates.
Given the scalability of this process, we followed a 4-inch diameter wafer scale
approach to synthesize nanotube samples as well as wafer-scale transfer to Si/SiO
2

substrates for device fabrication.
110

Figure 6.1 Wafer-scale synthesis of aligned nanotubes on quartz and a-sapphire wafers.
Photographs of the wafers after nanotube growth are shown on the left. SEM images on the right
show large arrays of highly aligned nanotubes.

6.2.2 Wafer scale CNTFET fabrication
Wafer-scale transfer of aligned nanotubes to Si/SiO
2
wafers brings the potential to
achieve high-density two-dimensional arrays of nanotubes on silicon by repeated transfer
on the same substrate, which is needed to significantly surpass the performance of current
silicon-based CMOS technologies. A 100 nm film of Au was deposited on top of the as-
grown nanotubes over the entire wafers, after which a thermally activated adhesive

a-sapphire
quartz
111

polymer (Revalpha tape from Nitto Denko) was placed on the Au film. Peeling-off the
tape resulted in picking up the nanotube/Au film. The thermal tape/Au film/aligned
SWNT film was placed onto the target Si/SiO
2
substrate and the whole structure was
heated to detach the thermal tape.

Figure 6.2 (a) Schematic diagram showing the aligned-nanotube transfer process. (b) Photograph
of a Si/SiO
2
wafer with transferred nanotubes. SEM image shows that, after being transferred,
nanotubes maintain a good degree of alignment on the receiving substrate. Au electrodes
deposition, followed by etching of nanotubes outside the device channel area complete the
fabrication of CNTFETs.

Au
Peel-off
Au + SWNTs
Aligned
SWNTs
Transfer
onto
Si/SiO
2
Au
deposition
Full wafer
processing
a
b
112

Gold etchant was employed to dissolve the Au film leaving behind the aligned
SWNT arrays on the target substrate. By using this approach, the transfer efficiency
obtained was nearly 100%. After nanotube transfer, photolithography was used to
fabricate back-gated carbon nanotube field-effect transistors (figure 6.2).

6.3 Results
Aligned nanotubes offer significant potential for nanotube assembly and
integration [26, 27], as nanotube devices can be easily fabricated at wafer scale, as shown
in figures 6.3a and b shows the photograph of an array of devices based on aligned
nanotubes transferred to a 4 inch Si/SiO
2
wafer. The SEM image of a typical device is
shown in figure 6.3b. Figure 6.3c illustrates the light irradiation process, where a
collimated white light beam from either a xenon or halogen lamp is used to irradiate the
fabricated wafer with nanotube devices for durations from 30 min to several hours.
Figure 6.3 shows the drain current (I
DS
) vs. gate voltage (V
G
) for a typical device
with 6 aligned nanotubes before and after light irradiation with an accumulated energy of
30 kJ/cm
2
. A remarkable increase in the channel current on/off ratio (I
On
/I
Off
) was
observed from 50 before irradiation to 1.2 x 10
5
after irradiation. This process is highly
scalable as compared to the traditional electrical breakdown approach
19
, which has to be
carried out device by device. Inspection of the nanotubes after irradiation revealed no
visible cut in the nanotubes (figure 6.4), suggesting the increase in I
On
/I
Off
is due to metal-
to-semiconductor conversion of nanotubes in the channel.

113

Figure 6.3 (a) Schematic diagram showing large arrays of field-effect transistors comprising of
horizontally aligned carbon nanotubes between source and drain electrodes. (b) Photograph of a
Si/SiO
2
wafer with fabricated aligned nanotube transistors. The SEM image shows a typical
CNTFET in the arrays. (c) Schematic diagram illustrating the scalable light irradiation process.
(d) Current vs. gate voltage (I
DS
-V
G
) characteristics of a CNTFET device, obtained with V
DS
=0.5
V before (black trace) and after (red trace) light irradiation. The I
On
/I
Off
ratio increased from ~64
to ~10
5
in the nanotube transistor due to the light irradiation.-mediated oxidation of nanotube
sidewalls leads to metal-to-semiconductor transition.

We synthesized aligned nanotubes on 3 inch quartz and sapphire wafers after
synthesis, source/drain Ti/Au electrodes were patterned. Figure 6.4 shows an AFM image
from a metallic CNT-FET, comprising of only one nanotube within the channel, after
being exposed for 5 hours to light. The nanotube (1.60 nm in diameter) was found to be
free from obvious structural damage after exposure. In general, irradiated nanotubes that

c
b
a
d
Source
Drain
5μm
-20 -10 0 10 20
1E-11
1E-10
1E-9
1E-8
1E-7
1E-6
1E-5
I
DS
(A)
V
G
(V)
Before
After
114

showed metal-to-semiconductor conversion after being irradiated up to 5 hours were
found to be continuous even after being annealed at temperatures in hydrogen
atmosphere.

Figure 6.4 AFM image of single-nanotube CNTFET that shows increase of I
On/
I
Off
after 5 hours of
light exposure and converting into a semiconducting device. Zoomed AFM image shows no
visible damage or cutting on the nanotube structure. (Right) I
On/
I
Off
evolution of the CNTFET
shown on the left upon timed light irradiation.

The AFM image here reveals that the metal tube did not show any cutting, and the
observed increase in I
On
/I
Off
(figure 6.4) is therefore attributed to metal-to-semiconductor
d = 1.60 nm
300 nm
-10 -5 0 5 10
1E-9
1E-8
1E-7
1E-6
1E-5
Before
3 Hours
5 Hours
V
G
(V)
I
DS
(A)
0 1 2 3 4 5
1E-9
1E-8
1E-7
1E-6
1E-5
I
DS
(A)
Time (h)
I
On
I
Off
115

conversion. Right part of figure 6.4 shows the evolution of the transfer characteristics of
the CNT-FET shown on the left. A clear drop in I
On
and increase in gate bias dependence
of the drain current are observed as evidence of metal-to-semiconductor conversion
associated with a higher defect density on the photo-oxidized nanotube.
To systematically investigate the effect of light irradiation, we carried out
electrical measurements and micro Raman characterization for nearly 200 devices with
one to five nanotubes in the channel. Fig. 6.5a and b show the D-band, G-band and I
DS
-
V
G
curves for a device with a single metallic nanotube before and after light irradiation.
Analysis of the Raman spectra of this nanotube before and after irradiation reveals an
increase of the Raman band intensity at 1345 cm
-1
(D band) and a decrease of the G band
intensity (1590.4 cm
-1
), accompanied by an upshift of 5.5 cm
-1
for the G band The ratio
between the G and D Raman band intensities (I
G
/I
D
) is regarded as an assessment of the
sp
2
/sp
3
ratio in carbon nanotubes, and thus the five-fold decrease in I
G
/I
D
after irradiation
(figure 6.5a) is attributed to an increase in the defect density due to an increase in the sp
3

nature of irradiated nanotubes [24].
It is known that rehybridization defects due to conversion of sp
2
to sp
3
sites lead
to π-electrons localization that can readily open or increase the bandgap of nanotubes
[28], and result in the conversion of metallic to semiconductor nanotubes. The I
DS
-V
G

curves of figure 6.5b shows that the metallic single-nanotube FET exhibited stronger gate
bias dependence after light irradiation, thus indicating that an increase in the sp
3
nature of
metallic nanotubes leads to metal to semiconductor conversion [4, 28-31].
To elucidate which part of the light spectrum plays the major role in the sp
2
/sp
3

conversion of carbon nanotubes, we irradiated devices with ultraviolet (250 nm - 400
116

nm), visible (380 nm – 700 nm) and near-infrared (750 nm – 2000 nm) radiation for one
hour
1300 1400
Raman shift (cm
-1
)
Before
After
D band
1500 1550 1600 1650 1700

Raman shift (cm
-1
)
Before
After
5.5 cm
-1
G band
-10 -5 0 5 10
1E-8
1E-7
1E-6
I
DS
(A)
V
G
(V)
Before
After
a b

Figure 6.5 Light-mediated oxidation of nanotube sidewalls leads to metal-to-semiconductor
transition. (a) Raman D band (left) and G band (right) of a metallic nanotube in a single-nanotube
device before and after light irradiation. A 5-fold decrease in the I
G
/I
D
ratio shows that increased
defect density on the nanotube sidewalls was achieved due to light irradiation. (b) I
DS
-V
G

characteristics of the single-nanotube device shown in the inset of (a), before and after light
irradiation. The I
On
/I
Off
improved from 1.6 to 10.5, indicating metal to semiconductor transition.

by using different band-pass filters between the light source and the devices. Fig. 6.6a
displays the Raman I
G
/I
D
before (gray bars) and after (red bars) light exposure using the
full spectrum, ultraviolet, visible and near infrared irradiation. It is observed that both full
spectrum and ultraviolet irradiation led to a decreased I
G
/I
D
similar in magnitude, whereas
the effect of visible and near infrared irradiation is minor and within the statistical margin
of error.
Further confirmation about the role of ultraviolet light is carried out by examining
the Raman radial breathing mode (RBM), which is observed to be highly sensitive to
light irradiation. Prolonged irradiation led to a decrease in intensity and eventual
disappearance of RBM for many nanotubes (figure 6.6b inset), which can be attributed to
the presence of rehybridization defects that perturb the symmetry of this vibration mode
[32].
117

Figure 6.6b shows the percentage of nanotubes that remained exhibiting RBM peaks after
light irradiation. Again, while the full spectrum and ultraviolet irradiation delivered
similar effect (40% of examined nanotubes showed disappearance of RBM bands after
irradiation), the visible and near infrared irradiation delivered much less significant effect.
Careful examination of figure 6.6b inset also reveals that the nanotube with
diameter d = 1.42 nm (174 cm
-1
) displayed more significant decrease in RBM intensity
than the nanotube with d = 1.74 nm (144 cm
-1
), indicating that small-diameter nanotubes
are more reactive under ultraviolet irradiation than large-diameter nanotubes [33].
Binding energies for carbon-carbon bonds with sp
2
hybridization (6.37 eV) lie well above
the energy associated with photons in the UV region employed here (3.3~5.0 eV), for
which the defect density increase observed upon irradiation cannot be regarded as a
consequence of direct photolysis.
Full UV Vis NIR
0.0
0.2
0.4
0.6
0.8
1.0
1.2
G/D ratio (Normalized)
Before
After
F u ll U V V is N IR
0
5 0
1 0 0
1 5 0
Percentage %
R B M afte r
irrad ia tio n
100 150 200 250
B efore
A fter
R am an sh ift (cm
-1
)
a b

Figure 6.6 (a) Comparison between the G/D ratios of nanotubes before and after one hour
irradiation with the full spectrum, ultraviolet, (250 nm - 400 nm), visible (380 nm – 700 nm) and
near infrared (750 nm – 2000 nm). (b) Percentage of nanotubes exhibiting Raman RBM signal
after light irradiation using the same irradiation conditions as part (c). The inset shows the
decrease of RBM intensity was more significant for the small-diameter nanotube than for larger
nanotubes.

118

We attribute the observed metal-to-semiconductor conversion to UV-assisted
photo-oxidation of nanotubes in oxygen-containing environment. UV-irradiated oxygen
forms oxygen radicals and ozone, which are strong gas-phase oxidants [34, 35]. Figure
6.7a shows a schematic of carbon nanotube oxidation due to light irradiation in air. The
radical driven reactions can be triggered by UV radiation and readily contribute to the
surface functionalization of nanotubes with oxygen-containing groups.
Oxidation of nanotube sidewall is also consistent with the upshift observed on the
Raman G band (figure 6.5a). Such frequency displacement can be related to interactions
between electron-withdrawing oxygen functionalities and the ̊ -electron system on the
nanotube sidewall [36, 37].

Figure 6.7 (a) Schematic showing light-induced oxidation of the nanotube sidewalls and possible
chemical groups introduced on the nanotube sidewalls upon sp
2
–sp
3
rehybridization by light-
induced oxidation. (b) Comparison of typical I
DS
-V
G
characteristics of two CNTFETs before and
after irradiation in air and in vacuum. The device irradiated in air became fully depletable,
showing an improvement in I
On
/I
Off
from 1.7 to 100, while the device irradiated in vacuum
exhibited virtually no change in its transistor characteristics.

-10 -5 0 5 10
1E-8
1E-7
1E-6
I
DS
(A)
V
G
(V)
Device in vacuum before
Device in vacuum after
Device in air before
Device in air after
O
2
+ hν 2 O
-
O
-
+ O
2
O
3
O
-
+ H
2
O 2 OH
-
a b
119

Further confirmation of the role of oxygen in the photo-assisted metal-to-
semiconductor conversion was obtained by comparing the effects of light irradiation on
nanotube transistors in air and in vacuum (3x10
-5
Torr). Typical examples are shown in
figure 6.7b. The device exposed to light irradiation in vacuum (black and red curves)
exhibited little change in the on-state current and the I
On/
I
Off
. In contrast, the device
exposed to irradiation in air (blue and green curves) displayed an increase in I
On
/I
Off
and a
drop in the on-state current. This unambiguously confirms the important role of oxygen
for the light-assisted metal-to-semiconductor conversion of nanotubes.
Qualitative identification of oxygenated groups that functionalized carbon
nanotubes upon light irradiation at ambient conditions was obtained using Fourier
transform infrared spectroscopy (FTIR), and results are shown in figure 6.8. To obtain
FTIR signal of detectable amplitude, we used carbon nanotube samples derived from the
HiPco process instead of horizontally aligned nanotubes on substrates. The HiPco
nanotubes were deposited onto sapphire spectral windows and exposed to the full spectra
of the Xe lamp for 12 h. FTIR spectra were collected with a Jasco FTIR 670 Plus
spectrophotometer and background corrected to subtract the substrate signal.
A prominent broadband in the range (3100-3600 cm
-1
) can be observed to
increase in carbon nanotubes exposed to light irradiation. This broad band can be
attributed to contributions from a variety of hydroxylic (-OH) stretching modes, such as
O-H stretching in coupled carboxylic acid groups and O-H stretching in alcoholic or
phenolic groups. Similar observation was reported before for oxidized nanotube samples
[36]. These results suggest that hydroxyl groups constitute an important form of
oxygenated groups introduced to the nanotubes by light irradiation.
120

Figure 6.8 FTIR spectra of carbon nanotubes before (black) and after (red) light exposure at open
environmental conditions.

To further elucidate the underlying mechanism and to optimize the yield of
depletable transistors by light irradiation, we have carried out detailed experiments on 38
working devices with single or a few nanotubes with varying irradiation time. A “lost”
device is defined as a device with I
On
< 100 pA at V
ds
= 0.1 V. In contrast, devices with
I
On
≥ 100 pA at V
ds
= 0.1 V are categorized as survived devices, among which depletable
devices are defined as having I
On
/I
Off
≥ 10 and non depletable devices are those with
I
On
/I
Off
< 10. AFM was performed to obtain the diameter of each nanotube. Light-induced

2800 3200 3600 4000
0
40
80
Transmittance (%)
Wavenumber (cm
-1
)
Before light irradiation
After light irradiation
-OH Stretch
121

metal-to-semiconductor conversion was found to be stable and irreversible at ambient
conditions, in contrast to the behavior observed by Ajayan [38].
CNTFETs were grouped into three categories according to their transport
characteristics at time = 0, 3hours, and 5 hours under exposure: i) CNTFETs that showed
metallic behavior, but were converted into semiconducting and remained depletable
throughout the irradiation process (figure 6.9a, non-depletable to non-depletable to
depletable [DDND]); ii) Non-depletable CNTFETs that first became depletable, and
then lost electrical conduction upon continued irradiation (figure 6.9b, non-depletable to
depletable to lost [NDDL]); and iii) CNTFETs that were depletable before
irradiation, and then lost electrical conduction after continued light irradiation (figure
6.9c, depletable to depletable to lost [DDLost]).
The CNTFET in figure 6.9a has 2 nanotubes with d = 1.17 nm and 1.20 nm
connecting source and drain electrodes. I
On
/I
Off
for this device changed as 1.53.8203
for 0, 3 and 5 hours of light exposure, respectively. On the other hand, the nanotube in
the single-nanotube device shown in figure 6.9b has a diameter of 0.92 nm and its I
On
/I
Off

changed as 2.4191 for 0 and 3 hours, respectively. Radial tension possessed by small-
diameter carbon nanotubes decrease their stability and increase their reactivity compared
to those with larger diameters [39], which explains why light irradiation for the device in
figure 6.9b would first make it depletable but later too resistive for charge transport.
Furthermore, I
On
/I
Off
for the single nanotube CNTFET in figure 6.9c (d = 1.1 nm)
changed as 2135846 upon irradiation for 0 and 3 hours, respectively. Comparing
figures 6.9a and b confirms that small-diameter nanotubes are more reactive, as electrical
conduction was lost in figure 6.9b, but persisted in figure 6.9a after 5 hour light exposure.
122

Before 3h 3.5h 4h 4.5h
0
25
50
75
100
Percentage %
Time (h)
D e p l e t a b l e
S u r v i v a l

0
5
10

Number of devices
L o s t
0.6 0.9 1.2 1.5 1.8 2.1 2.4
0
5
10

S u r v i v e
D i a m e t e r ( n m )
a
b
c
d
e
0 1 2 3 4 5
1E-9
1E-8
1E-7
1E-6
1E-5 ND D
I
DS
(A)
Time (h)
I
On
I
Off
ND
d = 1.17 nm
1.20 nm
0 1 2 3 4 5
1E-12
1E-11
1E-10
1E-9
1E-8
D Lost
I
DS
(A)
Time (h)
I
On
I
Off
D
d = 1.1 nm
0 1 2 3 4 5
1E-13
1E-11
1E-9
1E-7
1E-5
Lost D
I
DS
(A)
Time (h)
I
On
I
Off
ND
d = 0.92 nm
Figure 6.9 Influence of the irradiation time and nanotube diameter on the metal-to-semiconductor
conversion observed in CNTFETs. (a, b and c) I
On
and I
Off
of single- and few-nanotube CNTFETs
showing different evolutions under timed light irradiation (V
ds
= 100 mV). (d) Histogram of
CNTFETs that lost electrical conduction or survived after six-hour light irradiation plotted versus
the nanotube diameter. Clear diameter dependence was observed. (e) Percentage of CNTFETs
that survived (red) and showed depletable behavior (black) for different light irradiation
durations. The best yield was found after 4-hour exposure, when the percentage of depletable
devices increased from 32% to 88% while keeping a survival ratio near 81%.

The effect of prolonged light exposure on a chip with 38 working CNTFETs is
shown in figure 6.9d. Devices with nanotube diameters ranging from 0.6 nm to ~1.3 nm
123

became nearly open circuits, while those with nanotube diameters larger than 1.4 nm in
general survived light exposure. After nearly 6 hours of irradiation, most (90%) of the
surviving devices (20 CNTFETs) were depletable and exhibited clear semiconducting
behavior.
Similar diameter dependence has also been observed for nanotube devices
exposed to H
2
plasma [24, 25] and CH
4
plasma [22] due to the higher curvature of small-
diameter nanotubes that makes them more reactive than their large-diameter counterparts.
As demonstrated above, prolonged light-assisted oxidation of the nanotube sidewall may
lead to highly resistive devices. Thus, it is important to find the exposure time that best
optimize the trade-off between depletable and surviving devices.
Figure 6.9d shows the change in the percentage of depletable (black) and
surviving (red) CNTFETs from a chip with 31 working devices, as a function of light
exposure time with a power density of 2.2 W/cm
2
. The best yield was obtained after 4
hours of light exposure, which offered ~88% depletable devices and a survival rate of
~81%. This yield is typical for devices with five nanotubes or less in the channel.
However, the effect of the metal to semiconductor conversion in I
On
/I
Off
is generally less
pronounced in devices with larger number of nanotubes, which can be improved by using
CNTFETs with optimum and narrower diameter distribution.
Now that the diameter dependence has been elucidated, we have carried out micro
Raman measurements to ascertain whether light irradiation provides selectivity between
metal and semiconductor tubes of similar diameters (figure 6.10). Covalent
functionalization of SWNT sidewall is accompanied by a decrease in the RBM intensity
below the noise level (disappearance of RBM bands) [40, 41]. For complete
124

characterization, we have carried out micro Raman spectroscopy with three different
excitation lines (532 nm, 633 nm, and 785 nm).
Figure 6.10 shows plots of the number of nanotubes exhibiting RBM v.s. the
RBM frequency before and after irradiation as measured with all three excitation lines.
The frequencies characteristic for metal and semiconductor nanotubes are highlighted
based on Kataura’s plot. Detailed analysis of the data shown in figure 6.10, for all lasers
employed, reveals a diameter-dependent decrease of nanotubes showing RBM, which
correlates to the increased sp
3
character of nanotubes upon light irradiation. By
comparing the histograms in figure 6.10a before and after light irradiation, one can
clearly see that predominantly small-diameter nanotubes (with large RBM frequency)
underwent disappearance of RBM bands, which is consistent with the diameter
dependence shown in Fig. 6.9.
Interestingly, most metallic nanotubes with M
11
bandgaps in resonance with
visible laser wavelengths 532 nm (2.32 eV) and 633 nm (1.96 eV) have diameters lower
than ~1.3 nm, which means they are more likely to be oxidized by light irradiation. NIR
laser energy with wavelength of 785 nm (1.58 eV) is, in contrast, in good resonance with
S
11
and S
22
bandgaps of semiconducting nanotubes with diameters lower than 1.3-1.4 nm,
and metallic nanotubes with larger diameters. Results shown in figure 6.10 further
confirm, for all lasers employed, a well marked diameter-dependent oxidation of
nanotubes.
125

Figure 6.10 Stacked histograms showing the number of nanotubes exhibiting RBM vs. the RBM
frequency, before and after light irradiation, as measured with three excitation lines (532 nm, 633
nm, and 785 nm). Frequency regions characteristic for metal or semiconductor nanotubes are
highlighted based on Kataura’s plot. Comparison of the histograms obtained before and after
irradiation for each laser shows a predominant light-induced oxidation of small-diameter
nanotubes (large RBM frequency).

M M
M M M M
100 125 150 175 200 225 250 275 300 325
0
10
RBM Raman shift (cm
-1
)
0
10
785 nm
100 125 150 175 200 225 250 275 300 325
0
30
0
30
633 nm
100 125 150 175 200 225 250 275 300 325
0
20

0
20
532 nm
M M
SC SC
SC SC SC SC
SC SC
SC SC
M M M M
M M M M M M M M M M M M
100 125 150 175 200 225 250 275 300 325
0
10
RBM Raman shift (cm
-1
)
0
10
785 nm
100 125 150 175 200 225 250 275 300 325
0
30
0
30
633 nm
100 125 150 175 200 225 250 275 300 325
0
20

0
20
532 nm
M M M M
SC SC
SC SC SC SC
SC SC
SC SC
126

Figure 6.11 Percentage of metallic nanotubes in as-grown samples before (gray columns) and
after (red columns) light exposure using xenon (upper panel) and halogen (lower panel) lamps.
Nanotubes were grouped into two categories based on their diameter: small-diameter (0.7 - 1.3
nm) and large-diameter (1.4 - 2.0 nm) nanotubes. A substantial decrease in the percentage of
small-diameter metallic nanotubes found after light irradiation, for both light sources employed,
indicates their preferential oxidation over semiconducting small-diameter nanotubes. Contrarily,
the percentage of large-diameter metallic nanotubes was largely unaffected by light, indicating
the preferential oxidation (metal over semiconductor) is more effective for small-diameter
nanotubes.

0.7-1.3 1.3-2.0
0
10
20
30
40
50
Metallic nanotubes %
Diameter (nm)
Before
After
Xenon
0.7-1.3 1.3-2.0
0
10
20
30
40
50
Metallic nanotubes %
Diameter (nm)
Before
After
Halogen
127

We irradiated as-grown nanotubes with the full spectrum of xenon and halogen
light sources. Figure 6.11 shows the percentage of metallic nanotubes in the samples,
before and after irradiation. In order to better analyze the results, carbon nanotubes were
grouped into two categories based on their diameter (0.7 - 1.3 nm and 1.4 - 2.0 nm).
Interestingly, we observed for both light sources a marked preferential oxidation of
metallic nanotubes with diameters between 0.7 and 1.3 nm over their semiconducting
counterparts, with a decrease in the percentage of metallic nanotubes from 45 to 7% and
35 to 18% for xenon and halogen irradiation, respectively.
The difference observed in the effect of xenon and halogen light sources over the
oxidation of small-diameter nanotubes can be related to the higher intensity of UV
photons of the former (Spectral irradiance of xenon and halogen lamps was provided by
Newport Co.), which will in principle provoke a faster and/or stronger oxidation on the
nanotubes than the halogen source. On the other hand, there was no significant difference
between the oxidation of large-diameter metallic and semiconducting nanotubes in the
diameter range of 1.4 – 2.0 nm.
The results described above may be rationalized by considering that higher radial
tension, added to the presence of free electrons on the conduction band of small-diameter
metallic nanotubes makes them more reactive, upon light-induced oxidation, than
semiconducting nanotubes of similar diameters; for which, appropriate irradiation times
as well as narrow diameter distribution are key to obtain semiconducting CNTFET arrays
by light irradiation.

128

6.4 Conclusions
Light irradiation of nanotubes constitutes a breakthrough scalable process for
nanotube-based electronic devices via a defect-assisted metal-to-semiconductor
conversion stimulated by light-induced oxidation. This process was found to be diameter
dependent and faster in small-diameter metallic nanotubes. I
on
/I
off
improvements obtained
in CNTFETs were typically in the range of 10
2
up to 10
5
and can be easily scaled and
integrated as a customizable technology over larger-diameter wafers. The approach
presented in this work offers clear advantages over conventional processes to eliminate
metallic nanotubes from CNTFETs and constitutes a significant advance towards large
scale fabrication of carbon nanotube based electronic devices.

129

Chapter 6 References

1. Thostenson, E.T., Z. Ren, and T.-W. Chou, Advances in the science and
technology of carbon nanotubes and their composites: a review. Composites
Science and Technology 2001. 61: p. 1899–1912.

2. Avouris, P., Z. Chen, and V. Perebeinos, Carbon-based electronics. Nature
Nanotechnology, 2007. 2 p. 605-615.

3. Mahar, B. and C. Laslau, Development of Carbon Nanotube-Based Sensors—A
Review. IEEE Sensors Journal, 2007. 7(2): p. 19.

4. Bachtold, A., et al., Logic Circuits with Carbon Nanotube Transistors. Science,
2001. 294 p. 1317-1320.

5. Javey, A., et al., Ballistic carbon nanotube field-effect transistors. Nature, 2003.
424(6949): p. 654-657.

6. Dulrkop, T., et al., Extraordinary Mobility in Semiconducting Carbon Nanotubes.
Nano Letters, 2003. 4(1): p. 35-39.

7. Avouris, P., et al., Carbon nanotubes: nanomechanics, manipulation, and
electronic devices. Applied Surface Science, 1999. 141: p. 201-209.

8. Bohr, M.T., Nanotechnology Goals and Challenges for Electronic Applications.
IEEE Transactions on Nanotechnology 2001. 1(1): p. 56-62.

9. Javey, A., et al., Carbon Nanotube Transistor Arrays for Multistage
Complementary Logic and Ring Oscillators. Nano Letters, 2002. 2(9): p. 929-932.

10. Chen, Z., et al., An Integrated Logic Circuit Assembled on a Single Carbon
Nanotube. Science, 2006. 311(5768): p. 1735-.

11. Cao, Q., et al., Medium-scale carbon nanotube thin-film integrated circuits on
flexible plastic substrates. Nature, 2008. 454(7203): p. 495-500.

12. Bachilo, S.M., et al., Narrow ( n,m)-Distribution of Single-Walled Carbon
Nanotubes Grown Using a Solid Supported Catalyst. Journal of the american
chemical society, 2003. 125: p. 11186-11187.

13. Li, Y., et al., Preferential Growth of Semiconducting Single-Walled Carbon
Nanotubes by a Plasma Enhanced CVD Method. Nano Letters, 2004. 4(2): p. 317-
321.

14. Joselevich, E. and C.M. Lieber, Vectorial Growth of Metallic and Semiconducting
Single-Wall Carbon Nanotubes. Nano Letters, 2002. 2: p. 1137-1141.
130

15. Chen, Z., et al., Bulk Separative Enrichment in Metallic or Semiconducting
Single-Walled Carbon Nanotubes. Nano Letters, 2003. 3(9): p. 1245-1249.

16. Arnold, M.S., et al., Sorting carbon nanotubes by electronic structure using
density differentiation. Nature Nanotechnology, 2006. 1: p. 6.

17. Chen, F., et al., Toward the Extraction of Single Species of Single-Walled Carbon
Nanotubes Using Fluorene-Based Polymers. Nano Letters, 2007. 7(10): p. 3013-
3017.

18. Nish, A., et al., Highly selective dispersion of single-walled carbon nanotubes
using aromatic polymers. Nature Nanotechnology, 2007. 2: p. 640-646.

19. Collins, P.G., M.S. Arnold, and P. Avouris, Engineering Carbon Nanotubes and
Nanotube Circuits Using Electrical Breakdown. Science, 2001. 292 p. 706-709.

20. Zhang, G., et al., Selective Etching of Metallic Carbon Nanotubes by Gas-Phase
Reaction. Science, 2006. 314: p. 974-977.

21. Huang, H., et al., Preferential Destruction of Metallic Single-Walled Carbon
Nanotubes by Laser Irradiation. J. Phys. Chem. B 2006. 110: p. 7316-7320.

22. Zhang, Y., et al., Sorting out Semiconducting Single-Walled Carbon Nanotube
Arrays by Preferential Destruction of Metallic Tubes Using Xenon-Lamp
Irradiation. Journal of Physical Chemistry C, 2008. 112: p. 8.

23. Chen, B.-H., et al., Novel Method of Converting Metallic-Type Carbon Nanotubes
to Semiconducting-Type Carbon Nanotube Field-Effect Transistors. Japanese
Journal of Applied Physics, 2006. 45(4B): p. 3680–3685.

24. Zhang, G., et al., Hydrogenation and Hydrocarbonation and Etching of Single-
Walled Carbon Nanotubes. Journal of the American Chemical Society, 2006. 128:
p. 6026-6027.

25. Zheng, G., et al., Transition of Single-Walled Carbon Nanotubes from Metallic to
Semiconducting in Field-Effect Transistors by Hydrogen Plasma Treatment. Nano
Letters, 2007. 7(6): p. 1622-1627.

26. Han, S., X. Liu, and C. Zhou, Template-Free Directional Growth of Single-
Walled Carbon Nanotubes on a- and r-Plane Sapphire. Journal of the American
Chemical Society, 2005. 127: p. 5294-5295.

27. Liu, X., S. Han, and C. Zhou, Novel Nanotube-on-Insulator (NOI) Approach
toward Single-Walled Carbon Nanotube Devices. Nano Letters, 2006. 6(1): p. 34-
39.
131

28. Park, K.A., K. Seo, and Y.H. Leeand, Adsorption of Atomic Hydrogen on Single-
Walled Carbon Nanotubes. Journal of Physical Chemistry B 2005. 109: p. 8967-
8972.

29. Derycke, V., et al., Carbon Nanotube Inter- and Intramolecular Logic Gates.
Nano Letters, 2001. 1(9): p. 453-456.

30. Zhou, W., et al., Structural characterization and diameter-dependent oxidative
stability of single wall carbon nanotubes synthesized by the catalytic
decomposition of CO. Chemical Physics Letters, 2001. 350: p. 6-14.

31. Park, S., D. Srivastava, and K. Cho, Generalized Chemical Reactivity of Curved
Surfaces: Carbon Nanotubes. NanoLetters, 2003. 3: p. 1273-1277.

32. Gomez De Arco, L., et al., Resonant micro-Raman spectroscopy of aligned
single-walled carbon nanotubes on a-plane sapphire. Applied Physics Letters,
2008. 93: p. 123112.1-123112.3.

33. Chan, S.-P., et al., Oxidation of Carbon Nanotubes by Singlet O2. Physical
Review Letters, 2003. 90(8): p. 4.

34. Zellner, R., Global Aspects of Atmospheric Chemistry. Topics in Physical
Chemistry, ed. H. Baumgartel, W. Grunbein, and F. Hensel. 1999, New York:
Springer. 334.

35. Tedetti, M., et al., Hydroxyl radical-induced photochemical formation of
dicarboxylic acids from unsaturated fatty acid (oleic acid) in aqueous solution.
Journal of Photochemistry and Photobiology A, 2007. 188: p. 4.

36. Kim, U.J., et al., Raman and IR Spectroscopy of Chemically Processed Single-
Walled Carbon Nanotubes. Journal of the American Chemical Society, 2005. 127:
p. 15437-15445.

37. Chen, G., et al., Chemically Doped Double-Walled Carbon Nanotubes:
Cylindrical Molecular Capacitors. Physical Review Letters, 2003. 90(25): p.
257403-1 - 257403-4.

38. Vijayaraghavan, A., et al., Metal-Semiconductor Transition in Single-Walled
Carbon Nanotubes Induced by Low-Energy Electron Irradiation. Nano Letters,
2005. 5(8): p. 1575-1579.

39. Seo, K., et al., Chirality- and Diameter-Dependent Reactivity of NO2 on Carbon
Nanotube Walls. Journal of the American Chemical Society, 2005. 127: p. 15724-
15729.

132

40. Wang, C., et al., Electronically Selective Chemical Functionalization of Carbon
Nanotubes: Correlation between Raman Spectral and Electrical Responses.
Journal of the American Chemical Society, 2005. 127: p. 11460-11468.

41. Buffa, F., H. Hu, and D.E. Resasco, Side-Wall Functionalization of Single-Walled
Carbon Nanotubes with 4-Hydroxymethylaniline Followed by Polymerization of -
Caprolactone. Macromolecules, 2005. 38: p. 8258-8263.

133

Chapter 7 Conclusions and Future Work

7.1 Summary and conclusions
This dissertation presented our approach to study and solve some of the
fundamental problems that hinder the realization of one- and two-dimensional carbon
nanostructures, graphene and carbon nanotubes, as viable technologies in next generation
electronic devices. This effort included the development and implementation of a scalable
method to produce high quality graphene at large scale and the development of a scalable
method to convert metal nanotubes into nanotubes with semiconducting behavior in order
to achieve semiconducting CNTFETs at wafer scales.
In chapter 1, we presented a general introduction to carbon nanomaterials
followed by a more focused discussion on the structure and properties of graphene and
carbon nanotubes. Chapter 2 illustrated the development of a simple, scalable, and cost-
efficient method to prepare graphene using methane-based CVD, with which we achieved
high quality graphene at large scale. In addition, we evaluated its electrical properties as
the active channel in field effect transistors and as a transparent conductor. Further work
on the synthesis of graphene on single crystal nickel showed a strong influence of the
substrate atomic arrangement, lattice order and surface smoothness on the thickness of
the synthesized graphene, favoring the formation of single and bilayer graphene over
few-layer films; those results were presented in Chapter 3. An interesting application of
the excellent conducting properties of large scale graphene films is presented in Chapter 4,
where we implemented our highly scalable graphene films as the transparent electrode in
flexible organic photovoltaic cells. CVD graphene solar cells demonstrated outstanding
134

capability to operate under bending conditions, outperforming largely ITO-based cells
which displayed cracks and irreversible failure under bending. Our work indicated the
great potential of CVD graphene films for flexible photovoltaic applications. Chapter 5
described our work on the evaluation of the influence of substrate/nanotube interactions,
present during align nanotube growth, on the Raman signature of the resulting aligned
nanotubes, nanotube structure and metal to semiconductor ratio. Last, in Chapter 6, we
presented our findings on a light-induced scalable method that can be used at wafer scale
to achieve a metal to semiconductor conversion of carbon nanotubes and its application
to achieve arrays of semiconducting CNTFETs.

7.2 Future work
Albeit the progress achieved and reported here, more remains to be done to
integrate graphene and carbon nanotubes as functional components in electronic devices
to meet the goal of taking them from the laboratory to the manufacturing industry. Both
materials have significant potential to replace silicon in nanoelectronics (1-5) but, in the
case of graphene, its semimetal nature with zero bandgap hampers its use for effective
field-effect transistor applications at room temperature [1, 2] (6-7). Therefore, a future
research direction for graphene will consist of searching for ways to open a bandgap on
its band structure, which could lead to high mobility FETs that may reach ballistic charge
transport.

135

7.2.1 Graphene nanoribbons with smooth edges
Quantum confinement effects lead to bandgap opening in graphene structures
with lateral dimensions of less than 10 nm [3-8]. Following this idea, several ways of
opening a gap in graphene via graphene nanoribbons are reported in the literature, from
which we can highlight the synthesis and lithographic patterning of graphene
nanoribbons, nanoribbon synthesis by opening carbon nanotubes and nanopatterning of
graphene with a periodic array of holes. A different approach consists of the covalent
functionalization of a graphene sheet to perturb the graphene sp
2
structure. Nanoribbon
patterning however, often give devices with rough edges and low driving current or
transconductance [9, 10]. Given the marked degradation in conduction, dense arrays of
ordered nanoribbons are needed, if practical devices and circuits were to be made. Such
ordered array of high density GNRs constitute a significant challenge on its own [11].
It is clear that we need to obtain smooth edge states in order to advantage the
properties of graphene nanoribbons, which minimizes charge scattering while providing
sufficient bandgap for efficient transistor operation. None of the abovementioned
approaches can achieve that because plasma etching or strong chemical oxidation of
carbon nanotubes leads to GNR with large defect density at the edges, degrading the
transport properties of graphene. On the other hand, nickel particles have shown the
ability to cut graphene films at high temperatures producing smooth edges and hence,
resulting conduction properties similar to those observed for pristine graphene samples
[12, 13].

136

Figure 7.1 Anisotropic etching of graphene along crystalline orientations of the graphene lattice.
Etching is attributed to Cu clusters that are present in the CVD chamber as impurities left behind
due to graphene synthesis on Cu substrates in the same CVD chamber. Image on the right
corresponds to a zoomed region from the image on the left. Yellow arrows show the anisotropic
etching of graphene by particles during thermal annealing.

Curiously, in our lab, we bumped into the serendipitous discovery of anisotropic
etching along specific crystallographic orientations of graphene films during a high
temperature annealing process of graphene samples transferred to SiO
2
substrates in a
H
2
/Ar atmosphere (figure 7.1). In our case, copper particles, provided presumably from
Cu contamination inside the CVD chamber or remains of the etched Cu film during
graphene transfer, led to anisotropic cutting but also to efficient etching of graphene
along certain lattice orientations. The possibility to pattern cupper particles on graphene,
followed by a high temperature annealing step, in order to obtain GNRs with smooth
edges is worth to be explored. By following this approach, opening of bandgap as well as
good conduction properties in GMRs can be achieved.

137

7.2.2 Strained graphene
Given the inherent loss of transport properties of graphene due to nanoribbon
patterning and covalent functionalization, finding a ways to open a bandgap in graphene
that does not imply creating defects in the graphene lattice is highly desired. An
interesting approach to create a bandgap on graphene would be one where destruction of
the conjugated sp
2
structure of graphene is avoided. In this way, retention of the
outstanding charge transport properties of graphene would remain, while the bandgap
opening guarantees efficient transistor operation.
Theoretical calculations showed that uniaxial strain of graphene produces opening
of a bandgap [14, 15]. This band-gap opening can be attributed to the breaking of
sublattice symmetry of graphene under uniaxial strain [16, 17] with the size of the band
gap increasing almost linearly with the increase of tensile strain and almost reaching 0.3
eV gap for a strain of 1%.
According to the above mentioned work, the band-gap variation with uniaxial
strain suggests that, by applying a uniaxial strain, a highly controllable and tunable band
gap of single-layer graphene can be achieved, i.e. the bandgap energy will increase with
the increase in uniaxial strain. Figure 7.2a-c shows a schematic representation and
theoretical results from the literature of uniaxial strain of the graphene lattice and its
effect on the bandgap [15]. Figure 7.2d and e shows our preliminary results to achieve
uniaxial strain on graphene and its effects on the transfer characteristics of FETs with
graphene as active channel material.

138

Figure 7.2 Strained graphene. (a) Schematic representation of the effect of uniaxial tensile stress
on a graphene supercell. The dashed (solid) lattices indicate the unstrained (strained) graphene.
Calculated band structure of unstrained (b) and 1% tensile strained (c) graphene. A band gap is
clearly seen on the band structure of strained graphene [15]. (d) Transfer characteristics under
back gate bias for unstrained and strained graphene microribbon FETs. (e) Schematic
representation of uniaxially strained graphene by wet etching of underlying SiO
2
etching. Inset
shows a typical strain observed on graphene FET channel upon wet trench etching.

Other possible approaches to achieve uniaxial strain in graphene may include
strain induced by a bent flexible substrate, or graphene may also be deposited on a
piezoelectric material so that the strain can be easily and precisely controlled. In principle,
the band-gap opening by uniaxial strain would be more efficient compared to other
methods, such as electric field tuning on bilayer [18] or molecule adsorption [17]. It is
also easier to be realized than fabrication of graphene nanoribbons [6], and the gap is
more controllable than those in epitaxial graphene [16]. The uniaxial strained graphene
may provide an alternative way to fabricate graphene-based FETs.

-20 -10 0 10 20
1E-7
1E-6
1E-5

Drain Current (A)
Gate Voltage (V)
0.2 V
0.4 V
-20 -10 0 10 20
2E-4
4E-4
6E-4
8E-4
1E-3

Drain Current (A)
Gate Voltage (V)
0.2 V
0.4 V
Unstrained Strained
(d)
(e)
139

7.2.3 Synthesis of carbon nanotubes with specific diameter and chirality
Coexistence of metallic and semiconducting carbon nanotubes in as-grown
samples by chemical vapor deposition continues to be the most important hurdle for
nanotube nanoelectronics, thus imposing important limits to their application in high-
performance applications. Although the metal nanotube conversion process induced by
light irradiation described here is easy to implement, scalable to wafer-size scales and
capable of yielding improvements in the channel-current on/off ratio up to five orders of
magnitude in nanotube-based field-effect transistors, optimization of irradiation
conditions is still needed to yield nearly 90% of depletable nanotube field-effect
transistors. However, as the reactivity of carbon nanotubes is strongly dependent on their
diameter, a lack of control of the diameter distribution of nanotubes precludes this
approach to achieve 100% efficiency with little or no damage to semiconducting
nanotubes. Efforts to controllably synthesize nanotubes with a diameter distribution of
< ± 0.1 nm will greatly improve the efficiency of the light induced metal to
semiconductor conversion to obtain nearly 100% semiconducting CNTFETs.
On the other hand, chirality controlled synthesis through new and innovative
synthesis techniques needs to be investigated. Catalyst-free nanotube synthesis, or
synthesis where the catalyst particle is either bounded on the substrate surface, or
perfectly and regularly distributed over the substrate to avoid unpredictable catalyst-
substrate interactions during the early stages of growth need to be explored as well. This
will certainly constitute the most important milestone yet to achieve in nanotube
technology

140

Chapter 7 References
1. Novoselov, K.S., et al., Two-dimensional atomic crystals. Proceedings of the National
Academy of Sciences of the United States of America, 2005. 102(30): p. 10451-10453.

2. Meric, I., et al., Current saturation in zero-bandgap, top-gated graphene field-effect
transistors. Nat Nano, 2008. 3(11): p. 654-659.

3. Nakada, K., et al., Edge state in graphene ribbons: Nanometer size effect and edge shape
dependence. Physical Review B, 1996. 54(24): p. 17954.

4. Son, Y.-W., M.L. Cohen, and S.G. Louie, Energy Gaps in Graphene Nanoribbons.
Physical Review Letters, 2006. 97(21): p. 216803.

5. Barone, V.n., O. Hod, and G.E. Scuseria, Electronic Structure and Stability of
Semiconducting Graphene Nanoribbons. Nano Letters, 2006. 6(12): p. 2748-2754.

6. Han, M.Y., et al., Energy Band-Gap Engineering of Graphene Nanoribbons. Physical
Review Letters, 2007. 98(20): p. 206805.

7. Chen, Z., et al., Graphene nano-ribbon electronics. Physica E: Low-dimensional Systems
and Nanostructures, 2007. 40(2): p. 228-232.

8. Bai, J., et al., Graphene nanomesh. Nat Nano. 5(3): p. 190-194.

9. Jiao, L., et al., Narrow graphene nanoribbons from carbon nanotubes. Nature, 2009.
458(7240): p. 877-880.

10. Bai, J., X. Duan, and Y. Huang, Rational Fabrication of Graphene Nanoribbons Using a
Nanowire Etch Mask. Nano Letters, 2009. 9(5): p. 2083-2087.

11. Lu, W. and C.M. Lieber, Nanoelectronics from the bottom up. Nat Mater, 2007. 6(11): p.
841-850.

12. Datta, S.S., et al., Crystallographic Etching of Few-Layer Graphene. Nano Letters, 2008.
8(7): p. 1912-1915.

13. Campos, L.C., et al., Anisotropic Etching and Nanoribbon Formation in Single-Layer
Graphene. Nano Letters, 2009. 9(7): p. 2600-2604.

14. Pereira, V.M., A.H. Castro Neto, and N.M.R. Peres, Tight-binding approach to uniaxial
strain in graphene. Physical Review B, 2009. 80(4): p. 045401.

15. Ni, Z.H., et al., Uniaxial Strain on Graphene: Raman Spectroscopy Study and Band-Gap
Opening. ACS Nano, 2008. 2(11): p. 2301-2305.

16. Zhou, S.Y., et al., Substrate-induced bandgap opening in epitaxial graphene. Nat Mater,
2007. 6(10): p. 770-775.

17. Ribeiro, R.M., et al., Inducing energy gaps in monolayer and bilayer graphene: Local
density approximation calculations. Physical Review B, 2008. 78(7): p. 075442.
141

18. Castro, E.V., et al., Biased Bilayer Graphene: Semiconductor with a Gap Tunable by the
Electric Field Effect. Physical Review Letters, 2007. 99(21): p. 216802.

142

Bibliography

Ago, H., K. Imamoto, et al. (2007). "Competition and cooperation between lattice-
oriented growth and step-templated growth of aligned carbon nanotubes on
sapphire." Applied Physics Letters 90: 123112

Ago, H., K. Nakamura, et al. (2005). "Aligned growth of isolated single-walled carbon
nanotubes programmed by atomic arrangement of substrate surface." Chemical
Physics Letters 408: 433-438.

Ago, H., N. Uehara, et al. (2006). "Synthesis of horizontally-aligned single-walled carbon
nanotubes with controllable density on sapphire surface and polarized Raman
spectroscopy." Chemical Physics Letters 421: 399-403.

Andersson, A., N. Johansson, et al. (1998). "Fluorine Tin Oxide as an Alternative to
Indium Tin Oxide in Polymer LEDs." Advanced Materials 10(11): 859-863.

Araujo, P. T., S. K. Doorn, et al. (2007). "Third and Fourth Optical Transitions in
Semiconducting Carbon Nanotubes." Physical Review Letters 98: 067401-1 -
067401-4.

Arenz, M., K. J. J. Mayrhofer, et al. (2005). "The effect of the particle size on the kinetics
of CO electrooxidation on high surface area Pt catalysts." Journal of the American
Chemical Society 127(18): 6819-6829.

Arnold, M. S., A. A. Green, et al. (2006). "Sorting carbon nanotubes by electronic
structure using density differentiation." Nature Nanotechnology 1: 6.

Arnold, M. S., A. A. Green, et al. (2006). "Sorting carbon nanotubes by electronic
structure using density differentiation." Nat Nano 1(1): 60-65.

Astruc, D. and F. Chardac (2001). "Dendritic catalysts and dendrimers in catalysis"
Chemical Reviews 101(9): 2991-3023.

Avouris, P., Z. Chen, et al. (2007). "Carbon-based electronics." Nature Nanotechnology 2
605-615.

Avouris, P., T. Hertel, et al. (1999). "Carbon nanotubes: nanomechanics, manipulation,
and electronic devices." Applied Surface Science 141: 201-209.

Bachilo, S. M., L. Balzano, et al. (2003). "Narrow ( n,m)-Distribution of Single-Walled
Carbon Nanotubes Grown Using a Solid Supported Catalyst." Journal of the
american chemical society 125: 11186-11187.

143

Bachtold, A., P. Hadley, et al. (2001). "Logic Circuits with Carbon Nanotube
Transistors." Science 294 1317-1320.

Bae, S., H. Kim, et al. "Roll-to-roll production of 30-inch graphene films for transparent
electrodes." Nat Nano 5(8): 574-578.

Bai, J., X. Duan, et al. (2009). "Rational Fabrication of Graphene Nanoribbons Using a
Nanowire Etch Mask." Nano Letters 9(5): 2083-2087.

Bai, J., X. Zhong, et al. "Graphene nanomesh." Nat Nano 5(3): 190-194.

Ballou, B., B. C. Lagerholm, et al. (2004). "Noninvasive imaging of quantum dots in
mice." Bioconjugate Chemistry 15(1): 79-86.

Banwell, T. C. and A. Jayakumar (2000). "Exact analytical solution for current flow
through diode with series resistance." Electronics Letters 36(4): 291-292.

Barnes, T. M., J. Z. X. Wu, et al. (2007). "Single-wall carbon nanotube networks as a
transparent back contact in CdTe solar cells." Applied Physics Letters 90:
243503.1-243503.3.

Barone, V. n., O. Hod, et al. (2006). "Electronic Structure and Stability of
Semiconducting Graphene Nanoribbons." Nano Letters 6(12): 2748-2754.

Billot, L., M. L. de la Chapelle, et al. (2006). "Error signal artifact in apertureless
scanning near-field optical microscopy." Applied Physics Letters 89(2).

Blake, P., P. D. Brimicombe, et al. (2008). "Graphene-Based Liquid Crystal Device"
Nano Letters 8(6): 1704-1708.

Boehme, M. and C. Charton (2005). "Properties of ITO on PET film in dependence on
the coating conditions and thermal processing." Surface & Coatings Technology
200 932- 935.

Bohr, M. T. (2001). "Nanotechnology Goals and Challenges for Electronic Applications."
IEEE Transactions on Nanotechnology 1(1): 56-62.

Bolotin, K. I., K. J. Sikes, et al. (2008). "Ultrahigh electron mobility in suspended
graphene." Solid State Communications 146(9-10): 351-355.

Buffa, F., H. Hu, et al. (2005). "Side-Wall Functionalization of Single-Walled Carbon
Nanotubes with 4-Hydroxymethylaniline Followed by Polymerization of -
Caprolactone." Macromolecules 38: 8258-8263.

Campos, L. C., V. R. Manfrinato, et al. (2009). "Anisotropic Etching and Nanoribbon
Formation in Single-Layer Graphene." Nano Letters 9(7): 2600-2604.
144

Cancado, L. G., A. Reina, et al. (2008). "Geometrical approach for the study of G ' band
in the Raman spectrum of monolayer graphene, bilayer graphene, and bulk
graphite." Physical Review B 77(24).

Cao, Q., H.-s. Kim, et al. (2008). "Medium-scale carbon nanotube thin-film integrated
circuits on flexible plastic substrates." Nature 454(7203): 495-500.

Cao, Q., H.-s. Kim, et al. (2008). "Medium-scale carbon nanotube thin-film integrated
circuits on flexible plastic substrates." Nature 454: 495–500.

Cao, Q., H. S. Kim, et al. (2008). "Medium-scale carbon nanotube thin-film integrated
circuits on flexible plastic substrates." Nature 454(7203): 495-U4.

Castro, E. V., K. S. Novoselov, et al. (2007). "Biased Bilayer Graphene: Semiconductor
with a Gap Tunable by the Electric Field Effect." Physical Review Letters 99(21):
216802.

Chan, S.-P., G. Chen, et al. (2003). "Oxidation of Carbon Nanotubes by Singlet O2."
Physical Review Letters 90(8): 4.

Chen, B.-H., J.-H. Wei, et al. (2006). "Novel Method of Converting Metallic-Type
Carbon Nanotubes to Semiconducting-Type Carbon Nanotube Field-Effect
Transistors." Japanese Journal of Applied Physics 45(4B): 3680–3685.

Chen, C., S. Rosenblatt, et al. (2009). "Performance of monolayer graphene
nanomechanical resonators with electrical readout." Nat Nano 4(12): 861-867.

Chen, F., B. Wang, et al. (2007). "Toward the Extraction of Single Species of Single-
Walled Carbon Nanotubes Using Fluorene-Based Polymers." Nano Letters 7(10):
3013-3017.

Chen, G., S. Bandow, et al. (2003). "Chemically Doped Double-Walled Carbon
Nanotubes: Cylindrical Molecular Capacitors." Physical Review Letters 90(25):
257403-1 - 257403-4.

Chen, Z., J. Appenzeller, et al. (2006). "An Integrated Logic Circuit Assembled on a
Single Carbon Nanotube." Science 311(5768): 1735-.

Chen, Z., J. Appenzeller, et al. (2006). "An Integrated Logic Circuit Assembled on a
Single Carbon Nanotube." Science 311: 1735.

Chen, Z., X. Du, et al. (2003). "Bulk Separative Enrichment in Metallic or
Semiconducting Single-Walled Carbon Nanotubes." Nano Letters 3(9): 1245-
1249.

145

Chen, Z., Y.-M. Lin, et al. (2007). "Graphene nano-ribbon electronics." Physica E: Low-
dimensional Systems and Nanostructures 40(2): 228-232.

Collins, P. G., M. S. Arnold, et al. (2001). "Engineering Carbon Nanotubes and Nanotube
Circuits Using Electrical Breakdown." Science 292(5517): 706-709.

Collins, P. G., M. S. Arnold, et al. (2001). "Engineering Carbon Nanotubes and Nanotube
Circuits Using Electrical Breakdown." Science 292 706-709.

Datta, S. S., D. R. Strachan, et al. (2008). "Crystallographic Etching of Few-Layer
Graphene." Nano Letters 8(7): 1912-1915.

Derycke, V., R. Martel, et al. (2001). "Carbon Nanotube Inter- and Intramolecular Logic
Gates." Nano Letters 1(9): 453-456.

Dresselhaus, M. S., G. Dresselhaus, et al. (2002). "Raman spectroscopy on isolated single
wall carbon nanotubes." Carbon 40: 2043-2061.

Du, X., I. Skachko, et al. (2008). "Approaching ballistic transport in suspended
graphene." Nature Nanotechnology 3(8): 491-495.

Dulrkop, T., S. A. Getty, et al. (2003). "Extraordinary Mobility in Semiconducting
Carbon Nanotubes." Nano Letters 4(1): 35-39.

Dulrkop, T., S. A. Getty, et al. (2004). "Extraordinary Mobility in Semiconducting
Carbon Nanotubes." Nano Letters 4(1): 35-39.

Durkop, T., S. A. Getty, et al. (2004). "Extraordinary mobility in semiconducting carbon
nanotubes." Nano Letters 4(1): 35-39.

Echtermeyer, T. J., M. C. Lemme, et al. (2007). "Graphene field-effect devices." The
European Physical Journal - Special Topics 148: 19-26.

Eda, G., G. Fanchini, et al. (2008). "Large-area ultrathin films of reduced graphene oxide
as a transparent and flexible electronic material." Nature Nanotechnology 3: 270-
274.

Eda, G., Y.-Y. Lin, et al. (2008). "Transparent and conducting electrodes for organic
electronics from reduced graphene oxide." Applied Physics Letters 92: 233305.1-
233305.3.

Eizenberg, M. and J. M. Blakely (1979). "Carbon Monolayer phase condensation on
Ni(111) " Surface Science 82: 228-236.

Ferrari, A. C., J. C. Meyer, et al. (2006). "Raman Spectrum of Graphene and Graphene
Layers." Physical Review Letters 97(18): 187401.
146

Forbeaux, I., J.-M. Themlin, et al. (1998). "Heteroepitaxial graphite on 6H-SiC(0001):
Interface formation through conduction-band electronic structure." Physical
Review B 58(24): 16396-16406.

Geim, A. K. and K. S. Novoselov (2007). "The rise of graphene." Nat Mater 6(3): 183-
191.

Gilje, S., S. Han, et al. (2007). "A Chemical Route to Graphene for Device Applications."
Nano Letters 7(11): 3394-3398.

Gomez De Arco, L., B. Lei, et al. (2008). "Resonant micro-Raman spectroscopy of
aligned single-walled carbon nanotubes on a-plane sapphire." Applied Physics
Letters 93: 123112.1-123112.3.

Gomez De Arco, L., Y. Zhang, et al. "Continuous, Highly Flexible, and Transparent
Graphene Films by Chemical Vapor Deposition for Organic Photovoltaics." ACS
Nano 4(5): 2865-2873.

Gomez, L., R. Bachelot, et al. (2006). "Apertureless scanning near-field optical
microscopy: a comparison between homodyne and heterodyne approaches."
Journal of the Optical Society of America B-Optical Physics 23(5): 823-833.

Gomez, L., Y. Zhang, et al. (2009). "Synthesis, Transfer and Devices of Single- and Few-
Layer Graphene by Chemical Vapor Deposition." IEEE Transactions on
Nanotechnology 8(2): 135-138.

Gupta, A., G. Chen, et al. (2006). "Raman Scattering from High-Frequency Phonons in
Supported n-Graphene Layer Films." Nano Letters 6(12): 2667-2673.

Han, M. Y., Ouml, et al. (2007). "Energy Band-Gap Engineering of Graphene
Nanoribbons." Physical Review Letters 98(20): 206805.

Han, S., X. Liu, et al. (2005). "Template-Free Directional Growth of Single-Walled
Carbon Nanotubes on a- and r-Plane Sapphire." Journal of the American
Chemical Society 127(15): 5294-5295.

Han, S., X. Liu, et al. (2005). "Template-Free Directional Growth of Single-Walled
Carbon Nanotubes on a- and r-Plane Sapphire." Journal of the American
Chemical Society 127: 5294-5295.

Han, S., X. Liu, et al. (2005). "Template-Free Directional Growth of Single-Walled
Carbon Nanotubes on a- and r-Plane Sapphire." Journal of the american chemical
society 127: 5296-5296.
Hass, J., R. Feng, et al. (2006). "Highly ordered graphene for two dimensional
electronics." Applied Physics Letters 89(14): 143106-3.
147

Hayes, B. (2005). "Why W?" American Scientist 93: 104-108.

Hofmann, S., R. Sharma, et al. (2007). "In situ Observations of Catalyst Dynamics during
Surface-Bound Carbon Nanotube Nucleation." Nano Letters 7(3): 602-608.

Huang, H., R. Maruyama, et al. (2006). "Preferential Destruction of Metallic Single-
Walled Carbon Nanotubes by Laser Irradiation." J. Phys. Chem. B 110: 7316-
7320.

Hwang, E. H., S. Adam, et al. (2007). "Transport in chemically doped graphene in the
presence of adsorbed molecules." Physical Review B 76(19): 195421.

Iijima, S. (1991). "Helical microtubules of graphitic carbon." Nature 354(6348): 56-58.

Ishikawa, F. N., H. K. Chang, et al. (2009). "Transparent Electronics Based on Transfer
Printed Aligned Carbon Nanotubes on Rigid and Flexible Substrates." Acs Nano
3(1): 73-79.

Ismach, A. and E. Joselevich (2006). "Orthogonal Self-Assembly of Carbon Nanotube
Crossbar Architectures by Simultaneous Graphoepitaxy and Field-Directed
Growth." Nano Letters 6 ( 8): 1706-1710

Ismach, A., D. Kantorovich, et al. (2005). "Carbon Nanotube Graphoepitaxy: Highly
Oriented Growth by Faceted Nanosteps." Journal of the american chemical
society 127: 11554-11555.

Ismach, A., L. Segev, et al. (2004). "Atomic-Step-Templated Formation of Single Wall
Carbon Nanotube Patterns." Angewandte Chemie International Edition 43(45):
6140-6143.

Jain, A. and A. Kapoor (2004). "Exact analytical solutions of the parameters of real solar
cells using Lambert W-function." Solar Energy Materials and Solar Cells 81(2):
269-277.

Jansseune, T. (2005). "Indium price soars as demand for displays continues to grow "
Compound Semiconductor 11: 34-35.

Javey, A., J. Guo, et al. (2003). "Ballistic carbon nanotube field-effect transistors."
Nature 424(6949): 654-657.

Javey, A., J. Guo, et al. (2003). "Ballistic carbon nanotube field-effect transistors."
Nature 424: 654-657.

Javey, A., Q. Wang, et al. (2002). "Carbon Nanotube Transistor Arrays for Multistage
Complementary Logic and Ring Oscillators." Nano Letters 2(9): 929-932.

148

Jiao, L., L. Zhang, et al. (2009). "Narrow graphene nanoribbons from carbon nanotubes."
Nature 458(7240): 877-880.

Jorio, A., M. A. Pimenta, et al. (2003). "Characterizing carbon nanotube samples with
resonance Raman scattering." New Journal of Physics 5 139.1-139.17.

Joselevich, E. and C. M. Lieber (2002). "Vectorial Growth of Metallic and
Semiconducting Single-Wall Carbon Nanotubes." Nano Letters 2: 1137-1141.

Kang, M.-G. and L. J. Guo (2007). "Nanoimprinted Semitransparent Metal Electrodes
and Their Application in Organic Light-Emitting Diodes." Advanced Materials
19: 1391-1396.

Kang, S. J., C. Kocabas, et al. (2007). "High-performance electronics using dense,
perfectly aligned arrays of single-walled carbon nanotubes." Nat Nano 2(4): 230-
236.

Kang, S. J., C. Kocabas, et al. (2007). "High-performance electronics using dense,
perfectly aligned arrays of single-walled carbon nanotubes." Nature
Nanotechnology 2(4): 230-236.

Karu, A. E. and M. Beer (1966). "Pyrolytic Formation of Highly Crystalline Graphite
Films." Journal of Applied Physics 37(5): 2179-2181.

Kim, K.-B., Y.-H. Tak, et al. (2003). "Relationship between Surface Roughness of
Indium Tin Oxide and Leakage Current of Organic Light-Emitting Diode."
Japanese Journal of Applied Physics 42: L438-L440.

Kim, K. S., Y. Zhao, et al. (2009). "Large-scale pattern growth of graphene films for
stretchable transparent electrodes." Nature 457: 706-710.

Kim, M., N. S. Safron, et al. "Fabrication and Characterization of Large-Area,
Semiconducting Nanoperforated Graphene Materials." Nano Letters 10(4): 1125-
1131.

Kim, U. J., C. A. Furtado, et al. (2005). "Raman and IR Spectroscopy of Chemically
Processed Single-Walled Carbon Nanotubes." Journal of the American Chemical
Society 127: 15437-15445.

Kocabas, C., S.-H. Hur, et al. (2005). "Guided Growth of Large-Scale, Horizontally
Aligned Arrays of Single-Walled Carbon Nanotubes and Their Use in Thin-Film
Transistors." small 1(11): 1110 – 1116.

149

Kocabas, C., M. Shim, et al. (2006). "Spatially Selective Guided Growth of High-
Coverage Arrays and Random Networks of Single-Walled Carbon Nanotubes and
Their Integration into Electronic Devices." Journal of the American Chemical
Society 128(14): 4540-4541.

Kreupl, F. (2008). Carbon Nanotubes in Microelectronic Applications, Wiley-VCH
Verlag GmbH & Co. KGaA.

Kroto, H. W., J. R. Heath, et al. (1985). "C60: Buckminsterfullerene." Nature 318(6042):
162-163.

Lee, J.-Y., S. T. Connor, et al. (2008). "Solution-Processed Metal Nanowire Mesh
Transparent Electrodes." Nano Letters 8: 689-692.

Lei, B., K. Ryu, et al. "Raman Characterization and Polarity Tuning of Aligned Single-
Walled Carbon Nanotubes on Quartz." Japanese Journal of Applied Physics
49(2).

Leobandung, E., L. J. Guo, et al. (1995). "Observation of Quantum Effects and Coulomb-
Blockade in Silicon Quantum-Dot Transistors at Temperatures over 100 K."
Applied Physics Letters 67(7): 938-940.

Levendorf, M. P., C. S. Ruiz-Vargas, et al. (2009). "Transfer-Free Batch Fabrication of
Single Layer Graphene Transistors." Nano Letters 9(12): 4479-4483.

Li, X., W. Cai, et al. (2009). "Large-Area Synthesis of High-Quality and Uniform
Graphene Films on Copper Foils." Science 324(5932): 1312-1314.

Li, X., W. Cai, et al. (2009). "Evolution of Graphene Growth on Ni and Cu by Carbon
Isotope Labeling." Nano Letters 9(12): 4268-4272.

Li, X., G. Zhang, et al. (2008). "Highly conducting graphene sheets and Langmuir-
Blodgett films." Nat Nano 3(9): 538-542.

Li, X., Y. Zhu, et al. (2009). "Transfer of Large-Area Graphene Films for High-
Performance Transparent Conductive Electrodes." Nano Letters 9(12): 4359-
4363.

Li, Y., D. Mann, et al. (2004). "Preferential Growth of Semiconducting Single-Walled
Carbon Nanotubes by a Plasma Enhanced CVD Method." Nano Letters 4(2): 317-
321.

Lin, Y.-M., K. A. Jenkins, et al. (2008). "Operation of Graphene Transistors at Gigahertz
Frequencies." Nano Letters 9(1): 422-426.

150

Lin, Y. M., C. Dimitrakopoulos, et al. "100-GHz Transistors from Wafer-Scale Epitaxial
Graphene." Science 327(5966): 662-.

Liu, X., S. Han, et al. (2006). "Novel Nanotube-on-Insulator (NOI) Approach toward
Single-Walled Carbon Nanotube Devices." Nano Letters 6(1): 34-39.

Liu, X., K. Ryu, et al. (2008). "Diameter Dependence of Aligned Growth of Carbon
Nanotubes on a-Plane Sapphire Substrates." The Journal of Physical Chemistry C
112(41): 15929-15933.

Lu, W. and C. M. Lieber (2007). "Nanoelectronics from the bottom up." Nat Mater 6(11):
841-850.

Mahar, B. and C. Laslau (2007). "Development of Carbon Nanotube-Based Sensors—A
Review." IEEE Sensors Journal 7(2): 19.

Mazzer, M., K. W. J. Barnham, et al. (2006). "Progress in quantum well solar cells." Thin
Solid Films 511: 76-83.

Meric, I., N. Baklitskaya, et al. (2008). "RF performance of top-gated, zero-bandgap
graphene field-effect transistors." Ieee International Electron Devices Meeting
2008, Technical Digest: 513-516.

Meric, I., M. Y. Han, et al. (2008). "Current saturation in zero-bandgap, top-gated
graphene field-effect transistors." Nat Nano 3(11): 654-659.

Mermin, N. D. (1968). "Crystalline Order in Two Dimensions." Physical Review 176(1):
250.

Meyer, J. C., A. K. Geim, et al. (2007). "The structure of suspended graphene sheets."
Nature 446(7131): 60-63.

Michalet, X., F. F. Pinaud, et al. (2005). "Quantum dots for live cells, in vivo imaging,
and diagnostics." Science 307(5709): 538-544.

Mock, J. J., M. Barbic, et al. (2002). "Shape effects in plasmon resonance of individual
colloidal silver nanoparticles." Journal of Chemical Physics 116(15): 6755-6759.

Moon, J. S., D. Curtis, et al. (2009). "Epitaxial-Graphene RF Field-Effect Transistors on
Si-Face 6H-SiC Substrates." Electron Device Letters, IEEE 30(6): 650-652.

Nair, R. R., P. Blake, et al. (2008). "Fine Structure Constant Defines Visual Transparency
of Graphene." Science 320: 1308.

Nakada, K., M. Fujita, et al. (1996). "Edge state in graphene ribbons: Nanometer size
effect and edge shape dependence." Physical Review B 54(24): 17954.
151

Nelson, J., I. Ballard, et al. (1999). "Effect of quantum well location on single quantum
well p-i-n photodiode dark currents." Journal of Applied Physics 86(10): 5898-
5905.

Ni, Z. H., T. Yu, et al. (2008). "Uniaxial Strain on Graphene: Raman Spectroscopy Study
and Band-Gap Opening." ACS Nano 2(11): 2301-2305.

Nish, A., J.-Y. Hwang, et al. (2007). "Highly selective dispersion of single-walled carbon
nanotubes using aromatic polymers." Nature Nanotechnology 2: 640-646.

Novoselov, K. S., A. K. Geim, et al. (2005). "Two-dimensional gas of massless Dirac
fermions in graphene." Nature 438(7065): 197-200.

Novoselov, K. S., A. K. Geim, et al. (2004). "Electric Field Effect in Atomically Thin
Carbon Films." Science 306: 666-669.

Novoselov, K. S., D. Jiang, et al. (2005). "Two-dimensional atomic crystals."
Proceedings of the National Academy of Sciences of the United States of America
102(30): 10451-10453.

Ortiz-Conde, A., F. J. García Sánchez, et al. (2000). "Exact analytical solutions of the
forward non-ideal diode equation with series and shunt parasitic resistances."
Solid-State Electronics 44(10): 1861-1864.

Park, K. A., K. Seo, et al. (2005). "Adsorption of Atomic Hydrogen on Single-Walled
Carbon Nanotubes." Journal of Physical Chemistry B 109: 8967-8972.

Park, S., D. Srivastava, et al. (2003). "Generalized Chemical Reactivity of Curved
Surfaces: Carbon Nanotubes." NanoLetters 3: 1273-1277.

Patri, A. K., I. J. Majoros, et al. (2002). "Dendritic polymer macromolecular carriers for
drug delivery." Current Opinion in Chemical Biology 6(4): 466-471.

Pereira, V. M., A. H. Castro Neto, et al. (2009). "Tight-binding approach to uniaxial
strain in graphene." Physical Review B 80(4): 045401.

Peters, M. J., L. E. McNeil, et al. (2000). "Structural phase transition in carbon nanotube
bundles under pressure." PHYSICAL REVIEW B 61( 9): 5939-5954.

Peumans, P., A. Yakimov, et al. (2003). "Small molecular weight organic thin-film
photodetectors and solar cells." Journal of Applied Physics 93(7): 3693-3723.

Rao, A. M., E. Richter, et al. (1997). "Diameter-Selective Raman Scattering from
Vibrational Modes in Carbon Nanotubes " Science 275: 187-191.
152

Reina, A., X. Jia, et al. (2008). "Large Area, Few-Layer Graphene Films on Arbitrary
Substrates by Chemical Vapor Deposition." Nano Letters 9 (1): 30-35.
Reina, A., X. T. Jia, et al. (2009). "Large Area, Few-Layer Graphene Films on Arbitrary
Substrates by Chemical Vapor Deposition." Nano Letters 9(1): 30-35.

Reina, A., S. Thiele, et al. (2009). "Growth of large-area single- and Bi-layer graphene by
controlled carbon precipitation on polycrystalline Ni surfaces." Nano Research
2(6): 509-516.

Ribeiro, R. M., N. M. R. Peres, et al. (2008). "Inducing energy gaps in monolayer and
bilayer graphene: Local density approximation calculations." Physical Review B
78(7): 075442.

Rollings, E., G. H. Gweon, et al. "Synthesis and characterization of atomically thin
graphite films on a silicon carbide substrate." Journal of Physics and Chemistry of
Solids 67(9-10): 2172-2177.

Rowell, M. W., M. A. Topinka, et al. (2006). "Organic solar cells with carbon nanotube
network electrodes." Applied Physics Letters 88: 233506.1-233506.3

Rycenga, M., P. H. C. Camargo, et al. "Understanding the SERS Effects of Single Silver
Nanoparticles and Their Dimers, One at a Time." Journal of Physical Chemistry
Letters 1(4): 696-703.

Ryu, K., A. Badmaev, et al. (2007). "Synthesis of Aligned Single-Walled Nanotubes
Using Catalysts Defined by Nanosphere Lithography." Journal of the American
Chemical Society 129(33): 10104-10105.

Ryu, K., A. Badmaev, et al. (2009). "CMOS-Analogous Wafer-Scale Nanotube-on-
Insulator Approach for Submicrometer Devices and Integrated Circuits Using
Aligned Nanotubes." Nano Letters 9(1): 189-197.

Sangeeth, C. S. S., J. Manu, et al. (2009). "Charge transport in transparent conductors: A
comparison." Journal of Applied Physics 105(6): 063713.1-063713.6.

Saw, K. G. (2004). "Surface reconstruction of α-(0001) sapphire: An AFM, XPS, AES,
and EELS investigation

and EELS investigation." JOURNAL OF MATERIALS SCIENCE 39 2911 – 2914.
Scott, J. C., J. H. Kaufman, et al. (1996). "Degradation and failure of MEH-PPV light-
emitting diodes." Journal of Applied Physics 79: 2745-2751.

Seo, K., K. A. Park, et al. (2005). "Chirality- and Diameter-Dependent Reactivity of NO2
on Carbon Nanotube Walls." Journal of the American Chemical Society 127:
15724-15729.

153

Shelton, J. C., H. R. Patil, et al. (1974). "Equilibrium segregation of carbon to a nickel
(111) surface: A surface phase transition." Surface Science 43(2): 493-520.

Somani, P. R., S. P. Somani, et al. (2006). "Planer nano-graphenes from camphor by
CVD." Chemical Physics Letters 430: 56-59

Somorjai, G. A. (1996). "Modern Surface Science and Surface Technologies: An
Introduction." Chemical Reviews 96: 1223-1235.

Son, Y.-W., M. L. Cohen, et al. (2006). "Energy Gaps in Graphene Nanoribbons."
Physical Review Letters 97(21): 216803.

Stankovich, S., D. A. Dikin, et al. (2007). "Synthesis of graphene-based nanosheets via
chemical reduction of exfoliated graphite oxide." Carbon 45(7): 1558-1565.

Sukpirom, N. and M. M. Lerner (2003). "Rapid syntheses of nanocomposites with
layered tetratitanate using ultrasound." Materials Science and Engineering a-
Structural Materials Properties Microstructure and Processing 354(1-2): 180-187.

Sutter, P. W., J.-I. Flege, et al. (2008). "Epitaxial graphene on ruthenium." Nat Mater
7(5): 406-411.

Tan, Y. W., Y. Zhang, et al. (2007). "Measurement of Scattering Rate and Minimum
Conductivity in Graphene." Physical Review Letters 99(24): 246803.1-246803.4

Tedetti, M., K. Kawamura, et al. (2007). "Hydroxyl radical-induced photochemical
formation of dicarboxylic acids from unsaturated fatty acid (oleic acid) in aqueous
solution." Journal of Photochemistry and Photobiology A 188: 4.

Thostenson, E. T., Z. Ren, et al. (2001). "Advances in the science and technology of
carbon nanotubes and their composites: a review." Composites Science and
Technology 61: 1899–1912.

Tung, V. C., L.-M. Chen, et al. (2009). "Low-Temperature Solution Processing of
Graphene#Carbon Nanotube Hybrid Materials for High-Performance Transparent
Conductors." Nano Letters 9: 1949–1955.

Venkateswaran, U. D., A. M. Rao, et al. (1999). "Probing the single-wall carbon
nanotube bundle: Raman scattering under high pressure." PHYSICAL REVIEW
B 15 APRIL 1999-II VOLUME , NUMBER 59(16): 10928-10934.

Viculis, L. M., J. J. Mack, et al. (2003). "A Chemical Route to Carbon Nanoscrolls."
Science 299: 1361.

Abstract (if available)

Abstract Graphene and carbon nanotubes have outstanding electrical and thermal conductivity. These characteristics make them exciting materials with high potential to replace silicon and surpass its performance in the next generation of semiconductors devices, such devices ought to be considerably smaller and faster than the ones used in present technology. Despite of the excellent electrical and thermal conduction properties of graphene and carbon nanotubes, the advance of nanoelectronics based on them has been hampered due to fundamental limitations of the current synthesis and integration technologies of these carbon nanomaterials. Therefore, there is a strong need to do research at fundamental and applicative levels to help find the roadmap that these materials need to follow, in order to become a real alternative for silicon in future technologies.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Electronic and optoelectronic devices based on quasi-metallic carbon nanotubes

PDF

Carbon nanotube macroelectronics

PDF

Raman spectroscopy of carbon nanotubes under axial strain and surface-enhanced Raman spectroscopy of individual carbon nanotubes

PDF

Raman spectroscopy and electrical transport in suspended carbon nanotube field effect transistors under applied bias and gate voltages

PDF

Carbon nanotube nanoelectronics and macroelectronics

PDF

Carbon material-based nanoelectronics

PDF

Synthesis, assembly, and applications of single-walled carbon nanotube

PDF

GaAs nanowire optoelectronic and carbon nanotube electronic device applications

PDF

Controlled synthesis, characterization and applications of carbon nanotubes

PDF

Printed and flexible carbon nanotube macroelectronics

PDF

In-situ characterization of nanoscale opto-electronic devices through optical spectroscopy and electron microscopy

PDF

Chemical vapor deposition of graphene: synthesis, characterization, and applications

PDF

Printed electronics based on carbon nanotubes and two-dimensional transition metal dichalcogenides

PDF

Single-wall carbon nanotubes separation and their device study

PDF

A study of junction effect transistors and their roles in carbon nanotube field emission cathodes in compact pulsed power applications

PDF

Nanomaterials for macroelectronics and energy storage device

PDF

Synthesis, characterization, and device application of two-dimensional materials beyond graphene

PDF

Chemical vapor deposition of graphene and two-dimensional materials: synthesis, characterization, and applications

PDF

One-dimensional nanomaterials for electronic and sensing applications

Asset Metadata

Creator Gomez De Arco, Lewis Mortimer (author)

Core Title Graphene and carbon nanotubes: synthesis, characterization and applications for beyond silicon electronics

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Chemistry

Publication Date 09/13/2010

Defense Date 08/31/2010

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag carbon nanotube field effect transistors,carbon nanotube synthesis,carbon nanotubes,electronic properties of carbon nanotubes,electronic properties of graphene,graphene,graphene field efect transistors,graphene synthesis,graphene transparent conductive films,OAI-PMH Harvest,raman of carbon nanotubes,raman of graphene,resonant raman spectroscopy,semiconducting carbon nanotubes

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Zhou, Chongwu (committee chair), Gundersen, Martin A. (committee member), Thompson, Mark (committee member)

Creator Email gomezdea@usc.edu,lmortimerg@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m3433

Unique identifier UC1200309

Identifier etd-Arco-4103 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-390295 (legacy record id),usctheses-m3433 (legacy record id)

Legacy Identifier etd-Arco-4103.pdf

Dmrecord 390295

Document Type Dissertation

Rights Gomez De Arco, Lewis Mortimer

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu