University of Southern California Dissertations and Theses

Light Emission from Carbon Nanotubes and Two-Dimensional Materials

(USC Thesis Other)

Light Emission from Carbon Nanotubes and Two-Dimensional Materials

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content LIGHT EMISSION FROM CARBON NANOTUBES AND
TWO-DIMENSIONAL MATERIALS

by

Bo Wang

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
(PHYSICS)

May 2020

Copyright 2020 Bo Wang
ii

Dedication

This thesis is dedicated to my beloved parents for
their constant support.

献给我亲爱的父母以感谢他们长久以来的支持

iii

Acknowledgements
PhD research is a pretty tough, long but happy process for me. In the past six years, I
encountered a lot of issues regarding with both my daily routine and experiments and it is
extremely happy to overcome these difficulties with the help from professors, friends and lab-
mates. Here, I would like to offer my sincerest thanks to those people who helped me during my
PhD life.
First, I would like to thank my academic advisor Steve for his constant support and
invaluable guidance. Thank him for his understanding and tolerance of my mistakes. It is a great
experience to work and study under his supervision. I could not make these achievements without
his help.
I would also like to thank Dr. Stephan Hass, Dr. Aiichiro Nakano, Dr. Wei Wu and Dr. Han
Wang for being my dissertation committee and qualifying exam committee members.
My gratitude also goes to my girlfriend Sisi Yang, who gave me constant support and
encouraged me whenever I faced some trouble. Thank her for taking care of me when I was sick
or upset.
I am very fortunate to have fellows and friends like Yu Wang, Zhi Cai and Boxiang Song
who shared their experience and being good audience. Thank them to have so much fun in the
spare time.
It is a great experience to overlap and work with senior Ph.D. students, Dr. Rohan Dhall and
Dr. Zhen Li, Dr. Lang Shen, Dr. Bingya Hou, Dr. Jihan Chen and Dr. Haotian Shi, in Cronin
research group. Their great work inspired me a lot.
iv

My gratitude also extends to all the fellows in Cronin research lab and other research labs:
Dr. Nirakar Poudel, Mr. Bofan Zhao, Ms. Indu A A, Ms Ruoxi Li, Ms Yichen Gong, Mr. Boxin
Zhang, Mr. Deming Meng, Mr Qingzhou Liu, Mr Fanqi Wu, Mr. Yihang Liu, et al.
Finally, I would like to thank my beloved parents for their support through my entire life.
Their help and love are always the force to maintain my research momentum.

v

TABLE OF CONTENTS
Dedication ....................................................................................................................................... ii
Acknowledgements ........................................................................................................................ iii
List of Tables ................................................................................................................................. vii
List of Figures .............................................................................................................................. viii
Abstract ........................................................................................................................................ xiv

Chapter 1 Introduction .................................................................................................................... 1
1.1 Carbon Nanotubes ................................................................................................................. 1
1.2 Carbon Nanotube Field Effect Transistors ............................................................................ 3
1.3 Photoluminescence from Carbon Nanotubes and Two-Dimensional Materials.................... 8
1.4 Electroluminescence from Carbon Nanotube Field Effect Transistors ............................... 12
Chapter 2 Avalanche Photoemission in Suspended Carbon Nanotubes: Light Without Heat ...... 18
2.1 Abstract ................................................................................................................................ 18
2.2 Introduction ......................................................................................................................... 19
2.3 Experimental Details ........................................................................................................... 20
2.4 Results and Discussion ........................................................................................................ 22
2.5 Conclusion ........................................................................................................................... 37
Chapter 3 Ultra-Low Power Light Emission via Avalanche and Sub-Avalanche Breakdown in
Suspended Carbon Nanotubes ...................................................................................................... 39
3.1 Abstract ................................................................................................................................ 39
3.2 Introduction ......................................................................................................................... 40
3.3 Experimental Details ........................................................................................................... 41
3.4 Results and Discussion ........................................................................................................ 44
3.5 Conclusion ........................................................................................................................... 54
Chapter 4 Broadband Electroluminescence from Reverse Breakdown in Individual Suspended
Carbon Nanotube pn-junctions ..................................................................................................... 55
4.1 Abstract ................................................................................................................................ 55
4.2 Introduction ......................................................................................................................... 56
4.3 Experimental Details ........................................................................................................... 58
4.4 Results and Discussion ........................................................................................................ 60
4.5 Conclusion ........................................................................................................................... 66
vi

Chapter 5 Auger Suppression of Nanotube Incandescence in Individual Suspended CNT pn-
junctions ........................................................................................................................................ 67
5.1 Abstract ................................................................................................................................ 67
5.2 Introduction ......................................................................................................................... 68
5.3 Experimental Details ........................................................................................................... 69
5.4 Results and Discussion ........................................................................................................ 71
5.5 Conclusion ........................................................................................................................... 82
Chapter 6 Defect-Induced Photoluminescence Enhancement and Corresponding Transport
Degradation in Individual Suspended Carbon Nanotubes ............................................................ 83
6.1 Abstract ................................................................................................................................ 83
6.2 Introduction ......................................................................................................................... 84
6.3 Experimental Details ........................................................................................................... 84
6.4 Results and Discussion ........................................................................................................ 87
6.5 Conclusion ........................................................................................................................... 95
Chapter 7 Radiation-Induced Direct Band Gap Transition in Few-layer MoS2 ........................... 97
7.1 Abstract ................................................................................................................................ 97
7.2 Introduction ......................................................................................................................... 98
7.3 Experimental Details ......................................................................................................... 100
7.4 Results and Discussion ...................................................................................................... 103
7.5 Conclusion ..........................................................................................................................112
Chapter 8 Formation of Brightly Luminescent MoS2 Nano-Islands from Multi-Layer Flakes via
Plasma Treatment and Laser Exposure ........................................................................................115
8.1 Abstract ...............................................................................................................................115
8.2 Introduction ........................................................................................................................116
8.3 Experimental Details ..........................................................................................................117
8.4 Results and Discussion .......................................................................................................119
8.5 Conclusion ......................................................................................................................... 128
Chapter 9 Future Work ................................................................................................................ 130
9.1 Single Photon Emission via LED-Like Light Emission from Dual-Gate Carbon Nanotube
Field Effect Transistors ........................................................................................................... 130
Bibliography ............................................................................................................................... 132

vii

List of Tables
Table 1.1 Table of previous papers reporting electroluminescence from single carbon nanotube
devices................................................................................................................................... 17

Table 2.1 Raman peak positions and corresponding temperatures for different gate voltage, bias
voltage, and current. The temperatures established from the G band downshifts are based on
a coefficient of 0.0226 cm
-1
/K, measured in our previous work.
31, 41
................................... 29
viii

List of Figures
Figure 1.1 (a) Lattice diagram of a single-walled carbon nanotube and (b) SEM image of
suspended carbon nanotubes. .................................................................................................. 1
Figure 1.2 A typical Raman spectrum of a carbon nanotube. Adapted from Graupner et al., J.
Raman Spectrosc. (2007)
6
. ....................................................................................................... 3
Figure 1.3 Schematic diagram of a single-gate carbon nanotube field effect transistor. ............... 4
Figure 1.4 Semi-log plot of the source-drain current plotted as a function of gate voltage at a
constant bias voltage of 0.2V taken from a p-channel carbon nanotube field effect transistor.
.................................................................................................................................................. 5
Figure 1.5 Source-drain current plotted as a function of bias voltage at (a) Vg = 10V and (b) Vg =
-10V . ......................................................................................................................................... 6
Figure 1.6 Schematic diagram of a dual-gate carbon nanotube field effect transistor. .................. 6
Figure 1.7 Current plotted as a function of bias voltage under different gate conditions taken from
a dual-gate carbon nanotube FET. ........................................................................................... 7
Figure 1.8 Schematic for the excitation-decay process of photoluminescence. ............................ 8
Figure 1.9 Schematic diagram of the photoluminescence imaging setup for carbon nanotubes. .. 9
Figure 1.10 (a) IR image of a bright luminescent suspended carbon nanotube on pillar structures
and (b) spectrum of the photoluminescence from a suspended carbon nanotube. ................. 10
Figure 1.11 (a) Calculated band structures and (b) photoluminescence spectra of MoS2 with
different thicknesses from bulk to monolayer
11, 12
. .................................................................11
Figure 1.12 Schematic energy band diagram of charge carrier movement and decay processes in
a forward biased pn-junction. ................................................................................................ 13
Figure 1.13 (a) Electroluminescence image and (b) EL spectra from a partially-suspended carbon
nanotube field effect transistor
15
. ........................................................................................... 14
Figure 1.14 (a) Optical image, (b) EL spectra in the visible light wavelength range and (c)
ELspectra in near infrared wavelength range of thermal emission taken from a carbon
nanotube field effect transistor. .............................................................................................. 15
Figure 1.15 (a) Schematic diagram and (b) a scanning electron microscope (SEM) image of a dual-
gate carbon nanotube field effect transistor
18
. ....................................................................... 15
Figure 1.16 (a) An electroluminescence spectrum and (b) a photoluminescence spectrum taken
from the same dual-gate carbon nanotube field effect transistor
18
. ....................................... 16

Figure 2.1 (a) Scanning electron microscope (SEM) image and (c) schematic diagram of a
suspended carbon nanotube field effect transistor. (b) IR illuminated and (d)
electroluminescence images of the device under a bias of 5.1V with the current of 4nA at gate
voltage=10V . .......................................................................................................................... 22
Figure 2.2 (a) Current-gate voltage (I-Vg) characteristics measured from a different suspended
CNT FET. The potentials of the source and drain electrodes are indicated in the plot in blue
and red, respectively. (b) Log-linear plots of the electric current (left axis) and
electroluminescence intensity plotted as a function of applied bias voltage at a gate potential
ix

of Vg=+10V . (d) Zoom in of the data in (b). (c) Schematic energy band diagram illustrating the
emission of photons by ballistic hot carriers in an avalanche emission processes. ............... 24
Figure 2.3 (a) Simulated conduction and valence band profiles for a suspended carbon nanotube
FET with gate voltage Vg=+10V and bias voltage Vbias=+4V . The dashed line indicates the
Fermi level. (b) Electric field intensity profile along the nanotube under the same conditions.
................................................................................................................................................ 26
Figure 2.4 (a) Current-gate voltage (I-Vg) characteristics measured from a suspended CNT FET.
The potentials of the source and drain electrodes are indicated in the plot in blue and red,
respectively. (b) Log-linear plots of the electric current (left axis) and electroluminescence
intensity plotted as a function of applied bias voltage at a gate potential of Vg=-10V . ......... 28
Figure 2.5 (a,b) Electroluminescence images and (c,d) G band Raman spectra of a suspended CNT
FET under various applied gate and bias voltages. ................................................................ 30
Figure 2.6 (a) Current-gate voltage (I-Vg) characteristics measured from a suspended carbon
nanotube field effect transistor. The potentials of the source and drain electrodes are indicated
in the plot in blue and red, respectively. (b) Log-linear plots of the electric current (left axis)
and electroluminescence intensity plotted as a function of applied bias voltage at a gate
potential of Vg=+6V ................................................................................................................ 31
Figure 2.7 (a) Current-gate voltage (I-Vg) characteristics measured from the same suspended
carbon nanotube field effect transistor shown in Figure 2.2, now with a gate voltage of
Vg=+10V . The potentials of the source and drain electrodes are indicated in the plot in blue
and red, respectively. (b) Log-linear plots of the electric current (left axis) and
electroluminescence intensity plotted as a function of applied bias voltage at a gate potential
of Vg=+10V . (c) Linear plot of the current-voltage characteristics shown in (b). .................. 33
Figure 2.8 (a) Current-gate voltage (I-Vg) characteristics measured from another suspended carbon
nanotube field effect transistor. The potentials of the source and drain electrodes are indicated
in the plot in blue and red, respectively. (b) Log-linear plots of the electric current (left axis)
and electroluminescence intensity plotted as a function of applied bias voltage at a gate
potential of Vg=+7.5V . (c) Zoomed in plot of the data shown in (b). .................................... 35
Figure 2.9 (a) Photoluminescence and (b,c) electroluminescence spectra of suspended carbon
nanotube devices taken under (a) no applied voltage and (b,c) Vgate=-10V and Vbias=+3V . .. 36
Figure 2.10 Spatial profiles of the electroluminescence intensity observed along the length of a
suspended carbon nanotube device taken under (a) avalanche conditions (i.e., Vgate=+10V and
Vbias=+4V) and (b) thermal emission conditions (i.e., Vgate=-10V and Vbias=+1.6V). The plots
show data taken by reversing the roles of the left/right electrodes as the source/drain and
drain/source. ........................................................................................................................... 37

Figure 3.1 (a) Schematic diagram, (b) scanning electron microscope (SEM) image and (c) optical
microscope images of a suspended carbon nanotube field effect transistor. ......................... 43
Figure 3.2 (a) Infrared and (b) Electroluminescence images of a CNTFET device. ................... 44
Figure 3.3 Log plots of (a) the electric current and (b) EL intensity plotted as a function of applied
bias voltage at a gate potential of Vg = +7V . The value of the maximum differential
x

conductance is labeled in the plot (a). .................................................................................... 45
Figure 3.4 (a) Minimum bias voltage at which light emission is observed plotted as a function of
gate voltage. Three regions are labeled on the plot: avalanche region, thermal emission and no
detectible light emission. (b) 3D plot of the relative EL efficiency plotted as a function of
applied gate and bias voltages. (c) Plot of the maximum differential conductance at different
values of gate voltage. This data set was measured at the same time as the data in Figure 3.2
from the same CNTFET device. ............................................................................................ 47
Figure 3.5 (a) Minimum bias voltage at which light emission is observed plotted as a function of
a narrower range of gate voltage. Three regions are labeled on the plot: avalanche region,
thermal emission and no detectible light emission. (b) 3D plot of the relative EL efficiency
plotted as a function of applied gate and bias voltages. (c) Log plot of the electric current
plotted as a function of applied bias voltage at a gate potential of Vg = +7V. ....................... 48
Figure 3.6 (a) Minimum bias voltage at which light emission is observed plotted as a function of
gate voltage. Three regions are labeled on the plot: avalanche region, thermal emission and no
detectible light emission. (b) Color contour map plot of the EL efficiency plotted as a function
of gate and bias voltages measured from the same device of Figure 3.2 in the manuscript. . 50
Figure 3.7 Plot of the EL efficiency plotted as a function of bias voltage at gate voltage of 7 V
from the same device of Figure 3.2 in the manuscript. .......................................................... 51
Figure 3.8 Raman G band shift at (a) bias voltage = 3.8 V with current = 3.2 μA at gate voltage =
10 V , showing a temperature difference of around 1200K from room temperature, and (b) bias
voltage = 3.75 V with current = 2 nA, showing almost no temperature difference under an
applied gate voltage of 10 V . .................................................................................................. 52
Figure 3.9 Plot of the electrical current plotted as a function of bias voltage at gate voltage of -10
V from a single suspended CNTFET device showing negative differential conductance. .... 53

Figure 4.1 (a) Schematic diagram, (b) scanning electron microscope (SEM) image, and (c) optical
microscope images of a dual-gate, partially suspended carbon nanotube field effect transistor.
................................................................................................................................................ 59
Figure 4.2 (a) Current-gate voltage (I-Vg) characteristics measured from a suspended dual-gate
CNT FET device obtained by shorting the two gates (i.e., Vg1=Vg2) and applying a constant
bias voltage of 0.2 V . (b) Current plotted as a function of bias voltage by gating the device in
the pn (Vg1 = -10 V and Vg2 = 10 V) and np (Vg1 = 10 V and Vg2 = -10 V) configurations. (c)
Current and EL intensity plotted as a function of bias voltage at Vg1 = -Vg2 = 10V . ............. 61
Figure 4.3 Calculated conduction and valence band profiles of a dual-gate CNT FET device at (a)
Vg1 = -Vg2 = -15V and (b) Vg1 = -Vg2 = 15V under a bias voltage of 3V . ............................. 62
Figure 4.4 (a) Spectrum of thermal emission taken at Vb = 3.5 V and Ib=2.3µA, and (b) spectrum
of sub-avalanche light emission taken at Vb = -3.6 V and Ib=4nA from a suspended dual-gate
CNT FET device. Spectra collected from another device within the visible wavelength range
of (c) thermal emission taken at Vb = 3.7 V and Ib=2µA, and (d) sub-avalanche light emission
taken at Vb = -3.8 V and Ib=4nA. ........................................................................................... 64
Figure 4.5 PL spectra taken from two reprensentative suspended dual-gate CNT FETs. ............ 65
xi

Figure 5.1 (a) Schematic diagram, (b) colorized scanning electron microscope (SEM) image and
(c) optical microscope images of a dual-gate suspended carbon nanotube field effect transistor.
................................................................................................................................................ 71
Figure 5.2 (a,b) IR image with illumination and incandescence image (without illumination) from
a suspended dual-gate CNT FET device taken with an InGaAs camera. (c) Current and
incandescence intensity plotted as a function of gate voltage 1 with Vg1 = -Vg2 and Vbias = 3V .
................................................................................................................................................ 73
Figure 5.3 (a) Current and incandescence intensity plotted as a function of gate voltage 1 with Vg1
= -Vg2 and Vbias = 3V . (b) Current-normalized photon counts plotted as a function of gate
voltage 1 with Vg1 = -Vg2 and Vbias = 3V . (c) Incandescence spectrum taken at the red-circled
point indicated in (a) and photoluminescence spectrum taken from the same dual-gate CNT
FET device. ............................................................................................................................ 75
Figure 5. 4 (a) Calculated conduction and valence band profile and (b) calculated carrier density
profile of a dual-gate CNT FET device at Vg1 = -Vg2 = 15V under a bias voltage of 3V . (c)
Width of the intrinsic region plotted as a function of applied gate voltage with Vg1 = -Vg2 at
Vb = 3V . .................................................................................................................................. 78
Figure 5.5 Calculated conduction and valence band profiles of a dual gate CNT FET device at (a)
Vg1 = -Vg2 = -15V , (b) Vg1 = -Vg2 = 0V and (c) Vg1 = -Vg2 = 15V under a bias voltage of 3V .
The formation of pn-juctions are labeled between dashed lines. ........................................... 79
Figure 5.6 Calculated carrier density profile (a) Vg1 = -Vg2 = -15V , (b) Vg1 = -Vg2 = 0V and (c) Vg1
= -Vg2 = 15V (d) Vg1 = -Vg2 = -5V and (e) Vg1 = -Vg2 = 5V under a bias voltage of 3V . The
intrinsic regions are labeled between dashed lines. ............................................................... 80
Figure 5.7 (a) Raman G band spectra taken from one dual-gate CNT FET device under different
gate conditions (Vg1 = -Vg2) at a constant bias voltage of 3V . (b) The Raman G band frequency
and current plotted as a function of Vg1 (Vg1 = -Vg2). ............................................................ 81
Figure 5.8 Current plotted as a function of bias voltage under different gate conditions taken from
the same device shown in Figure 5.2. .................................................................................... 82

Figure 6.1 (a) Optical, (b) photoluminescence, and (c) SEM images of carbon nanotubes
suspended across quartz pillars. (d) Schematic diagram of the photoluminescence imaging
setup. ...................................................................................................................................... 86
Figure 6.2 (a) Representative photoluminescence image and (b) Raman spectra of four different
................................................................................................................................................ 87
Figure 6.3 (a) Representative photoluminescence image and (b) Raman spectra of five different
................................................................................................................................................ 88
Figure 6.4 (a) Representative PL image and (b) Raman spectra of 4 different suspended carbon
nanotubes that exhibit D/G ratios > 0.25. .............................................................................. 89
Figure 6.5 Photoluminescence intensity plotted as a function of the D/G band Raman intensity 90
Figure 6.6 (a-c) Schematic diagrams and (d) optical microscope image of the flip-chip transfer91
Figure 6.7 (a) SEM image, (b) current-voltage characteristics, and (c) Raman spectrum of a CNT
xii

................................................................................................................................................ 93
Figure 6.8 (a) Luminescence image and (b) spatial profile of the near-IR electroluminescence
(λ>1100nm) intensity of an individual suspended carbon nanotube. .................................... 94
Figure 6.9 The electric current plotted as a function of applied bias voltage from a blank chip
containing the metal electrodes used in the flip-chip transfer technique, showing a resistance
of around 2GΩ. ...................................................................................................................... 95

Figure 7.1 (a) Optical microscope and (b) SEM image of an MoS2 flake deposited on a SiN
membrane. Two holes in this membrane provide a suspended region of the flake to measure,
which behave quite differently from the substrate-supported region. ................................. 102
Figure 7.2 PL spectra of (a) suspended and (b) on-substrate bilayer MoS2 taken before and after
proton irradiation, and after annealing. ................................................................................ 105
Figure 7.3 PL spectra of (a) suspended and (b) on-substrate trilayer MoS2 taken before and after
proton irradiation, and after annealing. ................................................................................ 107
Figure 7.4 PL spectra of (a) suspended and (b) on-substrate MoS2 with four layers taken before
and after proton irradiation, and after annealing. ................................................................. 108
Figure 7.5 PL spectra of (a) suspended and (b) on-substrate monolayer MoS2 taken before and
after proton irradiation, and after annealing. ........................................................................110
Figure 7.6 PL spectra of monolayer and bilayer MoS2 taken before and after proton irradiation
with intermedium fluence (6×10
13
protons/cm
2
) and heavy fluence (6×10
14
protons/cm
2
) . 111
Figure 7.7 PL spectra of monolayer and bilayer MoS2 taken after proton irradiation with heavy
fluence (6×10
14
protons/cm
2
) and ultra-heavy fluence (6×10
15
protons/cm
2
). .....................112
Figure 7.8 PL intensity ratio of suspended/substrate-supported MoS2 plotted as a function of layer
number...................................................................................................................................113
Figure 7.9 Curve fits of the PL spectra of bilayer MoS2 taken before and after proton irradiation,
and after annealing. This is the same data plotted in Figure 7.2 of the manuscript. The peak
positions and intensities are indicated in each plot. ..............................................................114

Figure 8.1 Transmission electron microscope (TEM) images of MoS2: (a) oxygen plasma treated
only and (b) oxygen plasma treated and laser exposed. ....................................................... 120
Figure 8.2 (a) PL spectra for a bilayer MoS2 flake before and after plasma treatment and laser
exposure. (b) Normalized PL spectra for the same bilayer MoS2 flake taken before treatment,
after oxygen plasma treatment only and after oxygen plasma treatment and laser exposure.
.............................................................................................................................................. 121
Figure 8.3 XPS spectra for (a) untreated, (b) oxygen plasma treated only, and (c) oxygen plasma
treated and laser exposed bi-layer MoS2 flakes. .................................................................. 123
Figure 8.4 Time-resolved photoluminescence intensity for (a) untreated, (b) oxygen plasma treated
only, and (c) oxygen plasma treated and laser exposed MoS2 flakes. .................................. 125
Figure 8.5 (a) Optical microscope and (b) AFM image of an MoS2 flake deposited on a silicon
nitride substrate. (c) A line scan of the surface topography along the dashed line indicated in
(b). ........................................................................................................................................ 127
xiii

Figure 8.6 Optical images and PL spectra taken from another suspended MoS2 flake of (a) before
treatment and (b) after oxygen plasma treatment and thermal annealing in Argon. ............ 128

Figure 9.1 (a) Photoluminescence spectra from the sp
3
defect state and (b) photon antibunching
property of a carbon nanotube at room temperature
42
. ........................................................ 131

xiv

Abstract
This dissertation work presents research projects related to electroluminescence and
photoluminescence from low dimensional material devices, including carbon nanotube field effect
transistors and two-dimensional material. These research projects can be helpful to understand the
fundamental properties and potential optical, electrical and optoelectronic applications of these
material and nanoscale devices. For future research, this dissertation can provide good reference
and guidance in both theoretical and method aspects.
Chapter 1 provides some background information that can be helpful to understand this
dissertation work. It gives a brief overview of carbon nanotube including its atomic structure and
optical & electrical properties, which can help understand its optical and optoelectronic
applications. It is then followed by an introduction to carbon nanotube field effect transistors,
including its electronic characteristics. Brief introductions to photoluminescence from low
dimensional materials and electroluminescence from carbon nanotube field effect transistors are
included in order to help understanding this dissertation work.
Chapters 2-4 present several electroluminescence researches of suspended carbon nanotube
field effect transistors. Chapter 2 reports avalanche light emission through reverse breakdown
(impact ionization) process. The electrical-driven light emission occurs when the CNT FET is
gated to its “off” state and remains unheated due to extremely low electric power density. Chapter
3 presents a follow-up research work focused on electrical power and relative light emission
efficiency during avalanche and sub-avalanche light emission process and it exhibits a relative
light emission efficiency with three orders of magnitude higher than all the other previous reports.
xv

Chapter 4 reports spectra studies of avalanche and sub-avalanche light emission process. The
spectra of avalanche light emission exhibits a featureless flat spectrum covering a large wavelength
range from 600-1600nm.
Chapter 5 presents a project related to thermal emission from dual-gate CNT FETs. The
thermal emission (incandescence) intensity can be tuned by the electrostatic doping method. The
charge carriers induced by gate voltages can suppress incandescence intensity via a non-radiative
process – Auger recombination. To the best of my knowledge this is the only system where the
incandescence (i.e., thermal emission) intensity can be tuned by electrostatic doping while the
electrical power is held constant. CNT is quite unique as a one-dimensional material.
Chapter 6 presents a project of photoluminescence from suspended carbon nanotubes. It
reveals a detailed comparison between photoluminescence (PL) intensity and Raman spectroscopy
quantified defects in carbon nanotubes.
Chapter 7 and 8 present projects related to photoluminescence from the most famous
transition metal dichalcogenides MoS2. Chapter 7 reports direct band gap transition induced by
proton radiation. It provides a reliable method to achieve permanent direct band gap transition
from multi-layer MoS2 flakes and can be potentially useful for making optoelectronic devices such
as LEDs and solar cells. Chapter 8 reports a two-step method to create “triangular” MoS2 nano-
islands via oxygen plasma followed by laser exposure, which can be potentially used as quantum
dots and single photon emitters.
Chapter 9 presents future work related to potential single photon emission via LED like light
emission from dual-gate CNT FETs.
1

Chapter 1 Introduction

1.1 Carbon Nanotubes

Carbon nanotubes have been extensively explored due to their unique optical, electrical and
optoelectronic properties as a one-dimensional material, such as high thermal conductivity (~6600
W/m∙K), high electron mobility (~100,000 cm
2
/V∙s), and large Young’s modulus (~1 TPa), since
first discovered by Sumio Ijima in 1991
1-4
. The formation process of a carbon nanotube can be
described as rolling a graphene sheet into a cylindrical tube. Depending on the carbon layer number,
carbon nanotubes can be divided into two different types, multi-walled carbon nanotube and
single-walled carbon nanotube. Figure 1.1 shows a lattice diagram of a single-walled carbon
nanotube and a scanning electron microscope (SEM) image of a suspended carbon nanotube.

Figure 1.1 (a) Lattice diagram of a single-walled carbon nanotube and (b) SEM image of suspended carbon
nanotubes.

(a) (b)
Suspended
CNTs
2

In this thesis, all the CNT-based devices use single-walled carbon nanotubes fabricated via
chemical vapor deposition (CVD). The typical diameter of carbon nanotubes used in these projects
is around 1nm-2nm, which can be calculated by its radial breathing mode (RBM) frequency using
Raman spectroscopy
5
, which can be written as
𝜔 𝑅𝐵𝑀 ~
248
𝑑 cm
−1
∙nm,
where d is the diameter of the CNT. For single-walled carbon nanotubes, the RBM frequency is
typically 150-300 cm
-1
.
There are 3 more basic Raman peaks usually used to characterize carbon nanotubes, which
are the D-band mode, G-band mode and G’-band mode. Figure 1.2 shows a typical Raman
spectrum of a carbon nanotube
6
. The Raman D-band mode is corresponding to the defects within
the carbon nanotube, which will be further discussed in Chapter 6. The Raman G-band mode
presents tangential vibrations of carbon atoms along axial and circumferential directions, which
exhibits two peaks around 1590cm
-1
. One important application of the Raman G-band frequency
is to monitor the temperature change under different conditions, which will be further discussed in
Chapters 2-5.

3

Figure 1.2 A typical Raman spectrum of a carbon nanotube. Adapted from Graupner et al., J. Raman Spectrosc.
(2007)
6
.

1.2 Carbon Nanotube Field Effect Transistors

Carbon nanotube field effect transistors were first demonstrated by Tans et al. 1998
7
. Figure
1.3 shows a typical diagram of a single-gate suspended carbon nanotube field effect transistor,
indicating the cross section of the device. As illustrated in Figure 1.3, a gate electrode is patterned
at the bottom of the trench while a suspended carbon nanotube lies across the trench connecting
source and drain electrodes.
Raman Shift (cm
-1
)
4

Figure 1.3 Schematic diagram of a single-gate carbon nanotube field effect transistor.

In this thesis, all the carbon nanotube field effect transistors involved are p-channel FETs.
Figure 1.4 shows a typical semi-log plot of the source-drain current plotted as a function of gate
voltage at a constant bias voltage of 0.2V taken from a p-channel carbon nanotube field effect
transistor, showing an on/off ratio around 10
6
.
Source
Drain
Gate
CNT
Pt

p-type
5

Figure 1.4 Semi-log plot of the source-drain current plotted as a function of gate voltage at a constant bias
voltage of 0.2V taken from a p-channel carbon nanotube field effect transistor.

When the device is gated with negative voltages, the FET is in its “on” state and the
resistance of the device is around 100kΩ. The device becomes very resistive (>GΩ) when it is
gated with large positive voltages. Figure 1.5 shows typical IV curves for both positive and
negative gating conditions. It should be noted that, when it is gated to “on” state, the IV curve
shows negative differential conductance (NDC) after a threshold bias voltage, which will be further
discussed in Chapter 2.
-10 -8 -6 -4 -2 0 2 4 6 8 10
0.01
0.1
1
10
100
1000
Current (nA)
Gate Voltage (V)
Bias=0.2V
On Off
6

Figure 1.5 Source-drain current plotted as a function of bias voltage at (a) V g = 10V and (b) V g = -10V .

Recently, carbon nanotube field effect transistors using advanced gating method are studies
widely by many research groups. Figure 1.6 shows a diagram of the dual-gate carbon nanotube
field effect transistor used in this thesis. Two gate electrodes that are close to each other are
patterned in the bottom of trench.

Figure 1.6 Schematic diagram of a dual-gate carbon nanotube field effect transistor.
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
0
500
1000
1500
2000
2500
3000
3500
Current (nA)
Bias Voltage (V)
V
g
=10V
0.0 0.5 1.0 1.5 2.0
0
500
1000
1500
2000
2500
3000
3500
Current (nA)
Bias Voltage (V)
Vg=-10V
(a) (b)
V
g
=-10V
CNT
Pt
Si
Pt
S
D
G1
G2
7

By gating the two gate electrodes with equal and opposite voltages, we are able to create a
pn-junction within the carbon nanotube. Figure 1.7 shows the current plotted as a function of bias
voltage obtained by setting the two gate voltages to equal and opposite values of Vg1 = -Vg2 = 10
V , which results in the formation of a pn-junction within the CNT. By reversing the gate voltages,
we can achieve reversible rectifying behavior, indicating that the rectifying behavior is not simply
due to one of the Schottky contacts associated with the Pt/CNT junction.

Figure 1.7 Current plotted as a function of bias voltage under different gate conditions taken from a dual-gate
carbon nanotube FET.

-2 -1 0 1 2
-1500
-1000
-500
0
500
1000
Current (nA)
Bias Voltage (V)
PN -10V 10V
NP 10V -10V
V
g
=0
8

1.3 Photoluminescence from Carbon Nanotubes and Two-Dimensional Materials

Photoluminescence from Carbon Nanotubes:
Photoluminescence have been extensively explored as it can be helpful understanding the
optical properties of materials and can be potentially applied to quantum information technology.
The general process of photoluminescence is illustrated in Figure 1.8. A photon with energy larger
than the band gap energy can be absorbed by an electron sitting in the valence band and the electron
will be excited in the conduction band while leaving a hole in the valence band. Then the excited
electron can easily decay from the conduction band and recombine with the hole in the valence
band. During the decay process, another photon with energy equal to the band gap energy will be
emitted.

Figure 1.8 Schematic for the excitation-decay process of photoluminescence.

Photoluminescence (PL) from carbon nanotubes, which is first reported by O’Connell et al.
in 2002
8
, is an effective method to characterize carbon nanotubes and it undergoes a different
process from that discussed above. The electrons and holes in a carbon nanotube will form bound
electron-hole pairs (i.e., excitons). An absorbed photon with enough energy will excite an exciton
E
C
E
V
Photon
e
-
Hole
E
C
E
V
Photon
e
-
Hole
Excitation Decay
9

at ground state into a higher energy state. The excited exciton will then decay to ground state
rapidly within 20-200ps and emit a photon with the energy equal to the transition energy
9
. Figure
1.9 shows a diagram of a typical PL imaging/spectroscopy system at room temperature. In this
system, a 785 nm wavelength laser source is used to irradiate the sample. A 1100nm long pass
filter is used to eliminate any Rayleigh or Raman scattered light. The PL signal is then collected
with an InGaAs camera/spectrometer, which is sensitive to light emission in the near infrared
wavelength range (i.e., 1100-1600nm).

Figure 1.9 Schematic diagram of the photoluminescence imaging setup for carbon nanotubes.

Because of low-lying dark excitons, it is intrinsically inefficient and highly dependent on
their chiralities for photoluminescence from carbon nanotubes
10
. Due to substrate quenching,
Dichroic Mirror (785nm)
CNT
Long-pass Filter(1100nm)
Objective Lens
785nm
Laser
InGaAs
Camera/Spectrometer
10

suspended carbon nanotube samples are usually used in experiment in order to achieve bright
photoluminescence. Figure 1.10 shows an IR image of a “bright” nanotube suspended across
quartz pillars with an E 11 emission at 1560nm.

Figure 1.10 (a) IR image of a bright luminescent suspended carbon nanotube on pillar structures and (b)
spectrum of the photoluminescence from a suspended carbon nanotube.

Photoluminescence from Transition Metal Dichalcogenides:
Among two dimensional materials, photoluminescence from MoS2 is the most widely
studied because of its robustness and reliability. Monolayer MoS2 is a direct band gap material
with a band gap around 1.85eV , while multilayer MoS2 shows indirect band gap emission around
1.3eV
11
. The calculated band structures and photoluminescence spectra of MoS2 with different
thicknesses from bulk to monolayer are shown in Figure 1.11
12
.
10µ m
suspended
CNT
1350 1400 1450 1500 1550 1600 1650
0
500
1000
1500
2000
2500
3000
3500
PL Intensity (Counts)
Wavelength (nm)
(a) (b)
11

Figure 1.11 (a) Calculated band structures and (b) photoluminescence spectra of MoS 2 with different thicknesses
from bulk to monolayer
11, 12
.
(a)
Bulk
4-layer Bilayer Monolayer
(b)
12

1.4 Electroluminescence from Carbon Nanotube Field Effect Transistors

Electroluminescence has been widely explored in order to understand the optoelectronic
properties of semiconductor materials and can be potentially applied to LEDs and quantum
information technologies. Considering a forward biased pn-junction, the process of
electroluminescence is illustrated in Figure 1.12. When a pn-junction is forward biased, electrons
in the n-type side and holes in the p-type side can move to the opposite side easily due to the
relatively small energy barrier, which is reduced by the applied bias voltage. Then photons can be
generated by decay of electrons on both p-type and n-type sides.
13

Figure 1.12 Schematic energy band diagram of charge carrier movement and decay processes in a forward biased
pn-junction.

Electroluminescence (EL) from on-substrate and partially-suspended single carbon
nanotube field effect transistors was first reported by the IBM group
13-15
. A light emission
efficiency around 10
-4
photons/injected electron-hole pair was reported by them. In their study, a
critical threshold applied electrical power of several microwatts was required in order to reach
light emission detection limit. Figure 1.13 shows the electroluminescence image and the spectra
of the EL taken from a partially suspended carbon nanotube field effect transistor. It should be
E
c
E
v
E
f
E
f
eV
Hole
e
-
E
c
E
v
E
f
E
f
eV
Hole
e
-
Hole
Photon
e
-
Photon
Decay
Charge Carriers
Movement
14

noted that the electroluminescence comes from the interface between the suspended and on-
substrate region which has the largest electric field due to band bending.

Figure 1.13 (a) Electroluminescence image and (b) EL spectra from a partially-suspended carbon nanotube field
effect transistor
15
.

According to recent studies, thermal emission caused by heating is most likely the main
mechanism of light emission occurring in this previous work because at these applied powers
substantial Joule heating occurs, which can be monitored by the G-band shift in Raman spectra
16
.
In 2007, Liu et al. from our group reported the spectra of thermal emission from suspended single
carbon nanotube field effect transistors
17
. Figure 1.14 shows an optical image and spectra of
thermal emission taken from a carbon nanotube field effect transistor. The spectra of thermal
emission from carbon nanotube field effect transistors in visible light range can be well fitted via
a Planck model as blackbody radiation, while in the near infrared wavelength range there is a
significant deviation from blackbody radiation due to the contribution of exciton recombination at
E11 transition.
(a)
(b)
15

Figure 1.14 (a) Optical image, (b) EL spectra in the visible light wavelength range and (c) ELspectra in near
infrared wavelength range of thermal emission taken from a carbon nanotube field effect transistor.

Most recently, LED-like light emission via dual-gate carbon nanotube field effect transistors
was reported by Kato’s group in 2017
18
. Figure 1.15 shows the configuration and a scanning
electron microscope (SEM) image of the dual-gate CNT field effect transistor used in their study.
Multiple suspended CNTs are connecting between both sides of the source and drain electrodes,
and two local gates are behind the electrodes for electrostatic doping. A pn-junction can be formed
via controlling two gate electrodes individually.

Figure 1.15 (a) Schematic diagram and (b) a scanning electron microscope (SEM) image of a dual-gate carbon
nanotube field effect transistor
18
.

Figure 1.16 shows both photoluminescence and electroluminescence spectra taken from the
same dual-gate carbon nanotube field effect transistor. Interestingly, the EL spectrum from the
(a) (b)
(c)
1um
16

forward-biased split gate device is exactly the same as its PL spectrum, which indicates that the
EL is attributed to E 11 exciton recombination without heating the carbon nanotube up. The result
observed by Kato’s group is important because they achieved pure light emission through E 11
transition without any thermal emission component.

Figure 1.16 (a) An electroluminescence spectrum and (b) a photoluminescence spectrum taken from the same
dual-gate carbon nanotube field effect transistor
18
.

17

Author /Group Device Power Year
IBM (Freitag) FET 70µW 2004
IBM (Chen) Suspended FET 35µW 2005
Tersoff (Misewich) FET 80µW 2003
Martel (Adam) NNFETs and CNFETs 80µW 2008
Rogers (Zaumseil) FETs array 2.6mW 2009
Avouris (Mueller) FET 0.6µW 2010
Krupke (Pfeiffer) SWCNT 27µW 2011
Cronin (Zuwei) FET 9µW 2011
Kato (Higashide) Dual gate FET N/A 2017
Cronin (Bo) FET 20nW 2017
Table 1.1 Table of previous papers reporting electroluminescence from single carbon nanotube devices.
18

Chapter 2 Avalanche Photoemission in Suspended Carbon Nanotubes: Light
Without Heat
This chapter is similar to Wang et al., published in ACS photonics

2.1 Abstract

We observe bright electroluminescence from suspended carbon nanotube (CNT) field effect
transistors (FETs) under extremely low applied electrical powers (~nW). Here, light emission
occurs under positive applied gate voltages, with the FET in its “off” state. This enables us to apply
high bias voltages (4V) without heating the CNT. Under these conditions, we observe light
emission at currents as small as 1nA, and corresponding electrical powers of 4nW, which is three
orders of magnitude lower than previous studies. Thermal emission is ruled out by monitoring the
G band Raman frequency, which shows no evidence of heating under these small electrical currents.
The mechanism of light emission is understood on the basis of steep band bending that occurs in
the conduction and valence band profiles at the contacts, which produces a peak electric field of
500kV/cm, enabling the acceleration of carriers beyond the threshold of exciton emission. The
exciton-generated electrons and holes are then accelerated in this field and emit excitons in an
avalanche process. This is evidenced by an extremely sharp increase in the current with bias
voltage (45mV/dec). We also observe light emission at negative applied gate voltages when the
FET is in its “on” state at comparable electrical powers to those reported previously (~5µW).
However, substantial Joule heating (T>1000K) is also observed under these conditions, and it is
difficult to separate the mechanisms of thermal emission from hot carrier photoemission in this
regime.
19

2.2 Introduction

The one-dimensional nature of carbon nanotubes gives rise to several interesting optical and
optoelectronic phenomena including excitons, trions, and single photon emission at room
temperature.
19-22
Electroluminescence (EL) was first observed from substrate-supported individual
carbon nanotube (CNT) pn-junctions in 2003 by the IBM group, who reported an
electroluminescence efficiency of ~10
-4
photons per injected electron–hole pair.
13, 14
In addition to
pn-junction devices, there have been several papers reporting hot carrier electroluminescence from
unipolar CNT devices under high bias voltages (up to 60V).
15, 23-26
These hot carriers, in turn,
create excitons through impact ionization under unipolar conditions. These excitons recombine
radiatively producing bright electroluminescence locally at the CNT-contact region. This
excitation mechanism is estimated to be ~1000 times more efficient than recombination of
independently injected electrons and holes, and it results from weak electron-phonon scattering
and strong electron-hole binding caused by the one-dimensional confinement.
From photoluminescence studies, it is known that the luminescence efficiency of suspended
CNTs is several orders of magnitude higher than substrate-supported CNTs.
27-29
In the work of
Lefebrve et al., PL emission was only observed from the suspended segments of carbon
nanotubes.
28
Based on their signal-to-noise ratio, they estimate the PL enhancement factor of
suspended CNTs to be at least 100X. The IBM group also studied electroluminescence from
partially-suspended CNTs under high bias. Here, they used the high local fields (E 50kV/cm)
produced at the junction between the suspended and supported parts of a single carbon nanotube
to accelerate carriers beyond the threshold for photon emission.
15
In this previous work, the
20

threshold of detectible light emission occurred under applied electrical powers of >10µW. In this
regime, however, it is hard to rule out the effects of thermal emission caused by Joule heating.
Previously, our group studied electrical heating of suspended carbon nanotubes under high bias
conditions using Raman spectroscopy and thermal emission spectroscopy, and found that that the
onset of heating typically occurred around electrical powers of 1µW.
30-33

2.3 Experimental Details

In the work presented here, we demonstrate light emission at extremely low powers (~4nW)
from individual CNTs suspended between two metal contacts. Raman spectroscopy is used to
measure the temperature of the CNTs under high bias voltages in order to rule out thermal emission
caused by Joule heating of the nanotubes. By gating the FET into the “off” state, we are able to
apply high electric fields (>500kV/cm) without heating the device. In addition, the ballistic nature
of the electron (and hole) transport in these suspended CNTs provides an ideal system for studying
hot carrier phenomena. In order to verify the proposed mechanism of light emission, we calculate
the band and electric field profiles along the length of the nanotube using a self-consistent Poisson
solver.
The CNT devices are fabricated by first etching a 500nm deep, 1µm wide trench in a
Si/SiO2/Si3N4 wafer. Platinum source and drain electrodes are then deposited perpendicular to the
trench, and a Pt gate electrode is deposited in the bottom of the trench, as shown in Figure 2.1.
34-
36
Ferric nitrate (Fe(NO3)3) catalyst dispersed in de-ionized water is deposited in lithographically-
21

defined windows patterned on the drain and source electrodes, as can be seen in Figure 2.1a. CNTs
are then grown by chemical vapor deposition (CVD) using a mixture of argon gas bubbled through
ethanol and hydrogen gas at 825°C, yielding suspended single wall carbon nanotubes in a field
effect transistor (FET) geometry, as depicted in Figures 2.1c.
37, 38
The nanotube growth is the final
processing step in the sample fabrication, which ensures that these nanotubes are not contaminated
by any chemical residues from lithographic processes. Also, since these nanotubes are suspended,
there are no effects introduced by the underlying substrate. In our measurements, the drain is kept
at 0V (i.e., ground) and the gate and bias voltages are applied with respect to the drain electrode.
Near infrared images were collected with a thermoelectrically-cooled InGaAs 2D array (Xenics,
Inc.) sensitive over the wavelength range from 1000-1600nm. Figures 2.1b and 2.1d show IR
illuminated and electroluminescence images of this device, respectively. Here, bright
electroluminescence can be observed under an applied gate voltage of Vg=+10, a bias voltage of
Vbias=5.1V , and a current of I=4nA.
22

Figure 2.1 (a) Scanning electron microscope (SEM) image and (c) schematic diagram of a suspended carbon
nanotube field effect transistor. (b) IR illuminated and (d) electroluminescence images of the device under a bias
of 5.1V with the current of 4nA at gate voltage=10V .

2.4 Results and Discussion

Figure 2.2a shows the low bias I-Vg characteristics of a suspended CNT FET device with a
threshold voltage around Vg=+1.5V . Figure 2.2b shows the electric current (linear) and
electroluminescence (EL) intensity (log) plotted as a function of bias voltage with a gate voltage
of Vg=+10V . Here, we see the onset of EL at a bias voltage of Vbias=4V and a current of Ibias<1nA,
which is below the noise level (1nA) in the bias voltage range. This corresponds to an applied
electrical power of P < 4nW, which is three orders of magnitude below all previous reports of
electroluminescence from individual carbon nanotubes.
13, 14
Here, the light emission increases
exponentially with bias voltage. The intensity of light emission increases exponentially with bias
voltage over this range. Here, the electrons and holes accelerating in the large electric field gain a
Pt
SiN
x
SiO
2
p-type Si
I-V for 2-OH5
0.0E+00
1.0E-06
2.0E-06
3.0E-06
4.0E-06
5.0E-06
0 500 1000 1500
Vb (mV)
Id (A)
Vg = -5V
Vg = 2V
2 mm
Trench
Catalyst
Catalyst
Nanotube
(a) (b)
10µ m
(d)
10µ m
CNT Electro-
luminescence
source
drain
gate
(c)
23

kinetic energy that is proportional to the distance they travel and the electric field E, i.e., K.E. =
eEx. Once the carriers have enough energy to emit an exciton (Eex 0.6eV), they scatter, creating
an exciton which later radiates as a photon or creates an electron/hole pair that contributes to the
current. In the avalanche process, many electron/hole pairs are generated for each initial electron,
which greatly increases the conductance of the nanotubes causing the rapid increase in current
shown in Figure 2.2b. Here, the large resistance of the CNT at Vg=+10V (R~2GΩ) enables us to
apply relatively high bias voltages without inducing any measurable heating in the CNT. Above
Vbias=7.5V , the current begins to increase sharply with bias voltage due to the avalanche breakdown
mechanism described above.
24

In order to obtain a more detailed understanding of the avalanche breakdown mechanism in
these suspended CNT FET devices, we performed electrostatic simulations of these devices using
the Sentaurus software package, which solves Poisson’s equation iteratively to provide the self-
consistent charge density profiles along the device. The calculated conduction and valence band
profiles are plotted in Figure 2.3a along the length of the CNT device, and exhibit extremely sharp
3.5 4.0 4.5 5.0
0
2
4
6
8
10
Current
Intensity
Bias Voltage (V)
Current (nA)
20
30
40
50
60
EL Intensity (au)
drain source
4 5 6 7 8
0
20
40
60
80
100
Bias Voltage (V)
Current (nA)
20
30
40
50
60
Current
Intensity
EL Intensity (au)
(c)
-10 -8 -6 -4 -2 0 2 4 6 8 10
0.01
0.1
1
10
100
1000
Current (nA)
Gate Voltage (V)
Bias=0.2V
(d)
-
+
-
+
-
+
-
-
+
-
+
-
+
-
-
-
(b) (a)
Figure 2.2 (a) Current-gate voltage (I-V g) characteristics measured from a different suspended CNT FET. The
potentials of the source and drain electrodes are indicated in the plot in blue and red, respectively. (b) Log-linear
plots of the electric current (left axis) and electroluminescence intensity plotted as a function of applied bias
voltage at a gate potential of V g=+10V . (d) Zoom in of the data in (b). (c) Schematic energy band diagram
illustrating the emission of photons by ballistic hot carriers in an avalanche emission processes.
25

band bending at the source contact. The corresponding electric field profile along the CNT device
is plotted in Figure 2.3b, exhibiting a maximum field strength of 500kV/cm. To put this value of
500kV/cm in some theoretical context, we can divide the exciton binding energy 360–400 meV by
the size of the exciton (approximately 5 nm)
39
to achieve a critical field of Eo = 720–800 kV/cm
required for the complete ionization of the excitons, which is on the order of the built-in fields that
are being applied experimentally. Here, essentially all of the voltage drop (4V) occurs with 50-
100nm of the source contact, and provides enough kinetic energy for each charge carrier to emit
several excitons in the avelanche process.
26

Light emission is also observed under negative applied gate voltages, with the CNT FET in
its “on” state. Figure 2.4 shows the electric current and IR emission intensity of a device under a
gate voltage of Vg=-10V . At bias voltages above Vbias=1.2V , the current exhibits negative
differential conductance (NDC), as reported previously.
30, 40
This NDC is caused by substantial
heating in the CNT due to the onset of optical phonon emission.
30, 32, 40
Here, the onset of IR
-4
-3
-2
-1
0
Energy (eV)
3.0 2.0 1.0 0.0
Position (µ m)
Source
Drain
500x10
3
400
300
200
100
0
E-Field (V/cm)
3.0 2.0 1.0 0.0
Position (µ m)
(b)
(a)
500x10
3
400
300
200
100
0
E-Field (V/cm)
3.0 2.0 1.0 0.0
Position (µ m)
Figure 2.3 (a) Simulated conduction and valence band profiles for a suspended carbon nanotube FET with gate
voltage V g=+10V and bias voltage V bias=+4V . The dashed line indicates the Fermi level. (b) Electric field intensity
profile along the nanotube under the same conditions.
27

emission (likely due to thermal emission) occurs at Vbias=1.2V , I=3.5µA, and P=4.2µW. We believe
that the main mechanism of light emission here is thermal emission.
33
In fact, Figure 2.6 shows
that, under comparable gate and bias conditions, we observe substantial downshifts of the G band
Raman mode by as much as 19cm
-1
, which corresponds to temperatures of more than 1000K.
Interestingly, the photon emission scales exponentially with bias voltage, as observed with the
avalanche emission mechanism.

28

Figure 2.4 (a) Current-gate voltage (I-V g) characteristics measured from a suspended CNT FET. The potentials
of the source and drain electrodes are indicated in the plot in blue and red, respectively. (b) Log-linear plots of
the electric current (left axis) and electroluminescence intensity plotted as a function of applied bias voltage at a
gate potential of V g=-10V .

Figure 2.5 shows the electroluminescence images and G band Raman spectra of a CNT FET
under positive and negative applied gate voltages and various bias voltages. When the FET is gated
in its “on” state (Vg=-10V), we observe light emission at a bias voltage of 2V and current of 2.7µA
(Figure 2.6a). The G band Raman mode under Vg=-10V shows temperature-induced downshifts
source
-10 -8 -6 -4 -2 0 2 4 6 8 10
0.01
0.1
1
10
100
1000
Current (nA)
Gate Voltage (V)
Bias=0.2V
drain
(b)
(a)
0.0 0.4 0.8 1.2 1.6 2.0
0
500
1000
1500
2000
2500
3000
3500
Current
Intensity
Bias Voltage (V)
Current (nA)
100
1000
EL Intensity (au)
29

are listed in Table 2.1 and exhibit temperatures above 1000K.

Gate Voltage (V) Bias Voltage (V) Current (nA) G band Frequency (cm
-1
) Temperature (K)

0.0 0 1591.1 300
1.0 1400 1589.0 393
1.4 2000 1584.4 596
2.0 2700 1570.1 1229

0.0 0 1591.0 300
4.5 2 1590.6 318
Table 2.1 Raman peak positions and corresponding temperatures for different gate voltage, bias voltage, and
current. The temperatures established from the G band downshifts are based on a coefficient of 0.0226 cm
-1
/K,
measured in our previous work.
31, 41

Figure 2.5b shows the electroluminescence image of the same device gated in its “off” state
(Vg=+10V) under an applied bias voltage of Vb=4.5V and current of 2nA. Here, the G band Raman
mode shows a shift of less than 1cm
-1
, corresponding to a temperature increase of less than 20K.
Therefore, under these conditions, we can safely say that the mechanism of electroluminescence
is not thermal emission, and instead is based on the avalanche emission process described in the
main manuscript. It should be noted that the threshold for EL is observed at a power of 5nW, which
is 100-1000 times lower than the threshold for electrical heating.
30-33

-10V
10V
30

Figure 2.5 (a,b) Electroluminescence images and (c,d) G band Raman spectra of a suspended CNT FET under
various applied gate and bias voltages.

Figure 2.6 shows the low bias I-Vg characteristics of a suspended CNT FET device with a
threshold voltage around Vg=+1V . Figure 2.6b shows a log-linear plot of the electric current and
electroluminescence (EL) intensity plotted as a function of bias voltage over a relatively small
range between 3.5 and 3.8V . Here, we see the onset of EL at 1.5nA, which is 2-3 orders of
magnitude lower than any previous report. The current, which is plotted on a log scale, increases
faster than exponentially, increasing by 3 orders of magnitude over a very small bias voltage range
from 3.5V to 3.8V (100mV/decade), as plotted in Figure S1b. The intensity of light emission
1540 1560 1580 1600 1620
Raman Intensity
Raman Shift (cm
-1
)
0V at room temperature
Bias=1V/Current=1.4uA
Bias=1.4V/Current=2uA
Bias=2V/Current=2.7uA
1550 1560 1570 1580 1590 1600 1610 1620
Raman Intensity
Raman Shift (cm
-1
)
0V at room temperature
Bias=4.5V/Current=2nA
“On” State
Gate Voltage=-10V
“Off” State
Gate Voltage=+10V
(a) (b)
(c) (d)
31

increases exponentially with bias voltage over this range. While Figure S1 shows
electroluminescence at extremely low currents and an electric current that increases 1000-fold at
high fields, the electroluminescence intensity only increases by a factor of 3.5X, which is not
consistent with the avalanche process. Here, it is possible that Auger recombination or some other
mechanism may be limiting the efficiency of light emission for this particular device.

Figure 2.6 (a) Current-gate voltage (I-V g) characteristics measured from a suspended carbon nanotube field
effect transistor. The potentials of the source and drain electrodes are indicated in the plot in blue and red,
respectively. (b) Log-linear plots of the electric current (left axis) and electroluminescence intensity plotted as a
function of applied bias voltage at a gate potential of V g=+6V .
3.50 3.55 3.60 3.65 3.70 3.75 3.80
1
10
100
1000
Current
Intensity
Bias Voltage (V)
Current (nA)
10
15
20
25
30
35
EL Intensity (au)
(b)
-10 -8 -6 -4 -2 0 2 4 6 8 10
0.01
0.1
1
10
100
1000
Current (nA)
Gate Voltage (V)
Bias=0.2V
drain source
(a)
32

Figure 2.7 shows data taken from the same CNT FET device measured in Figure 2.6, but
under a gate voltage of Vg=+10V and bias voltages ranging from 2.0 to 3.4V . For applied bias
voltages below 2.6V , the current increases exponentially with voltage, and no photon emission is
observed in this range. At Vbias=2.6V , however, we observe an abrupt jump in the current, with a
slope of 45mV/decade, and the onset of photoemission, which are both characteristics of the
avalanche process. The current then nearly saturates at a value around 6µA, over the bias voltage
range from 2.6 to 3.4V . Figure 2.7c shows a zoomed-in plot of the data in Figure S2b, showing
that the current actually increases slightly with bias over this voltage range. In this region, the CNT
is highly conductive, due to the large density of free carriers created by the avalanche process, and
the contact resistance becomes the current limiting factor. The light intensity increases
exponentially with bias voltage in this range, despite the nearly constant current conditions. Here,
the threshold for light emission occurs at an applied bias voltage of Vbias=2.7V and I=6µA, which
corresponds to a power of 16µW. Based on our previous Raman studies of suspended CNT FETs
under high bias, and those shown in the Figure 5 of the main text, we expect substantial heating to
occur at these high biases.
30-33
It is, therefore, difficult to separate the effects of thermal emission
and avalanche emission in this range of applied bias and gate potentials, and it is likely that both
processes are contributing to the light emission in the range.
33

Figure 2.7 (a) Current-gate voltage (I-V g) characteristics measured from the same suspended carbon nanotube
field effect transistor shown in Figure 2.2, now with a gate voltage of V g=+10V . The potentials of the source and
drain electrodes are indicated in the plot in blue and red, respectively. (b) Log-linear plots of the electric current
(left axis) and electroluminescence intensity plotted as a function of applied bias voltage at a gate potential of
V g=+10V . (c) Linear plot of the current-voltage characteristics shown in (b).

-10 -8 -6 -4 -2 0 2 4 6 8 10
0.01
0.1
1
10
100
1000
10000
Current (nA)
Gate Voltage (V)
Bias=0.2V
drain source
(b)
(c)
(a)
2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4
1
10
100
1000
10000
Current
Intensity
Bias Voltage (V)
Current (nA)
100
EL Intensity (au)
2.6 2.8 3.0 3.2 3.4
5800
6000
6200
6400
Current
Intensity
Bias Voltage (V)
Current (nA)
100
EL Intensity (au)
34

Figure 2.8 shows data taken from another CNT FET device measured under a gate voltage
of Vg=+7.5V . Here, we see the onset of electroluminescence at a bias voltage of 4.1V and less than
1nA of current. This electroluminescence increases exponentially with increasing bias voltage. We
also observe an abrupt increase in the electric current in this bias voltage range. Both of these
behaviors are characteristics of the avalanche process.
35

Figure 2.8 (a) Current-gate voltage (I-V g) characteristics measured from another suspended carbon nanotube
field effect transistor. The potentials of the source and drain electrodes are indicated in the plot in blue and red,
respectively. (b) Log-linear plots of the electric current (left axis) and electroluminescence intensity plotted as a
function of applied bias voltage at a gate potential of V g=+7.5V . (c) Zoomed in plot of the data shown in (b).

Figure 2.9a shows the photoluminescence spectrum of a typical suspended nanotube device
taken with a 785nm wavelength excitation laser. This spectrum shows a typical PL emission
-10 -8 -6 -4 -2 0 2 4 6 8 10
0.01
0.1
1
10
100
1000
10000
Current (nA)
Gate Voltage (V)
Bias=0.2V
3 4 5
1
10
100
Current
Intensity
Bias Voltage (V)
Current (nA)
100
EL Intensity (au)
4.1 4.2 4.3 4.4 4.5
1
2
3
4
5
Current
Intensity
Bias Voltage (V)
Current (nA)
100
EL Intensity (au)
Drain Source
(b)
(c)
(a)
36

centered around 1500nm with a linewidth of 50nm, consistent with previous literature.
28,27
An EL
spectrum taken under Joule heating conditions (i.e., Vgate=-10V and Vbias=+3V) is plotted in Figure
2.9b and shows broad thermal emission extending over the range from 900 to 1500nm with a
FWHM of 400nm, consistent with our previous reports.
33
The blackbody tail of this thermal
emission can even be seen in the visible wavelength range ( <900nm), as plotted in Figure 2.9c.

Figure 2.9 (a) Photoluminescence and (b,c) electroluminescence spectra of suspended carbon nanotube devices
taken under (a) no applied voltage and (b,c) V gate=-10V and V bias=+3V .

In order to provide further evidence that the light emission originates from the source contact,
electroluminescence images were collected from a device under high field/avalanche conditions
(i.e., Vgate=+10V and Vbias=+4V). By reversing the left/right electrodes from source/drain to
drain/source, we plot the spatial profile of the electroluminescence intensity, as shown in Figure
2.10. Here, we observe a clear shift in the peak position, indicating that the emission is originating
near the source electrode under both configurations. As a comparison, electroluminescence
intensity profiles were also taken under thermal emission conditions (i.e., Vgate=-10V and
Vbias=+2V), which show no shift in the peak position.
1400 1450 1500 1550 1600
0
2000
4000
6000
8000
10000
PL Intensity (Counts)
Wavelength (nm)
900 1000 1100 1200 1300 1400 1500
0
2000
4000
6000

Thermal Emission Intensity
Wavelength (nm)
600 650 700 750 800 850 900
0
200
400
600
800
1000
Thermal Emission Intensity
Wavelength (nm)
(a) (b)
(c)
37

Figure 2.10 Spatial profiles of the electroluminescence intensity observed along the length of a suspended
carbon nanotube device taken under (a) avalanche conditions (i.e., V gate=+10V and V bias=+4V) and (b) thermal
emission conditions (i.e., V gate=-10V and V bias=+1.6V). The plots show data taken by reversing the roles of the
left/right electrodes as the source/drain and drain/source.

2.5 Conclusion

In conclusion, bright electroluminescence is observed from suspended carbon nanotube
(CNT) field effect transistors (FETs) under high applied bias voltages due to an avalanche
photoemission process. Here, we apply positive applied gate voltages in order to turn the FET
“off”, which enables us to apply high bias voltages (4V) without substantially heating the CNT.
Light emission is observed at currents as small as 1nA and electrical powers of 4nW, which is three
orders of magnitude lower than previous studies. The G band Raman mode shows no evidence of
heating under these high bias conditions, thus ruling out the mechanism of thermal emission.
Instead, the light emission is understood based on the steep band bending of the conduction and
valence bands at the source contact, which enables the charge carriers to be accelerated beyond
the threshold of exciton emission. Light emission is also observed at negative applied gate voltages,
with the FET in its “on” state, however, this emission coincides with substantial Joule heating and
2 3 4 5 6 7
0
500
1000
1500
2000
2500
3000
DS
SD
EL Intensity (a.u.)
Position ( mm)
1 2 3 4 5 6
0
100
200
300
400
500
600
700
DS
SD
Fit of SD
Fit of DS
EL Intensity (a.u.)
Position ( mm)
Avalanche Emission Thermal Emission
(a) (b)
38

is attributed to thermal emission of photons.
39

Chapter 3 Ultra-Low Power Light Emission via Avalanche and Sub-
Avalanche Breakdown in Suspended Carbon Nanotubes
This chapter is similar to Wang et al., published in ACS photonics

3.1 Abstract

We explore near infrared light emission in suspended carbon nanotube field effect
transistors (FETs) over a wide range of gate and bias voltages. An abrupt increase in both the
electric current (90 µA/V) and electroluminescence (EL) intensity is observed at high bias voltages
(~3.5 V), when gated in the “off’ state (i.e., Vgate > 0 V). For bias voltages below the threshold for
avalanche breakdown, we observe light emission due to the creation of exitons by impact
ionization under these high electric fields. Here, we find that there is a relatively small region over
which low power (~nW) light emission is observed. By plotting the relative luminescence
efficiency (i.e., light intensity/electrical power) as a function of the gate and bias voltages, we
observe a very sharp feature corresponding to avalanche emission at which the
electroluminescence efficiency is 2-3 orders of magnitude higher than that under all other
conditions of gate and bias voltage. A steep increase in the current with bias voltage (i.e., large
dI/dVb) is observed at the same gate and bias conditions of the highly efficient electroluminescence
and signifies the onset of the avalanche process. We believe that these results demonstrate
additional mechanistic evidence for achieving highly efficient light emission in carbon nanotubes
via avalanche breakdown.

40

3.2 Introduction

Recent research of light emission from nanoscale systems, such as carbon nanotubes (CNTs)
and III-V semiconducting quantum dots, shows several interesting optical and optoelectronic
phenomena including exciton and trion formation and room-temperature single photon emission
(SPE) at telecommunications wavelengths (1300-1550 nm)
19-22, 42
. High purity single photon
sources (SPSs) are a key component needed for quantum information and computing technologies.
Electroluminescence (EL) from substrate-supported single CNT pn-junction devices was first
demonstrated by the IBM group in 2003
13, 14
. They reported a light emission efficiency around 10
-
4
photons per injected electron-hole pair. In addition to these pn-junction devices, several CNT
FET-based devices exhibiting electroluminescence have also been reported from several groups
30-
33
. In these previous single nanotube studies, an empirical threshold of 1 µW of applied electrical
power was required in order to detect light emission from these CNT FET devices. At these applied
powers, however, substantial Joule heating occurs, which can be monitored by the G-band shift in
their Raman spectra
16
. As a result, thermal emission caused by heating is likely the main
mechanism of light emission occurring in these previous works. In order to develop electrically-
driven single photon emitters based on carbon nanotubes for quantum communication, it is
absolutely essential that highly efficient electroluminescent carbon nanotube devices be
demonstrated.
Photoluminescence from carbon nanotubes was first observed in 2002 by O’Connell et al.
8
,
which was followed by many further optical and optoelectronic studies of CNTs. Lefebrve et al.
41

reported bright photoluminescence from suspended regions of carbon nanotubes, which indicated
that suspended CNTs had a much higher luminescence efficiency than substrate-supported CNTs
28
.
In 2005, Chen et al. observed bright electroluminescence from partly suspended carbon nanotubes,
which they attributed to the high fields produced at the region between the suspended and non-
suspended parts of CNT that provide enough electrical energy for exciton creation and subsequent
light emission
15
. Khasminskaya et al. observed electrically-driven single photon emission from a
CNT-based photonic circuit under cryogenic conditions in 2016
43
. In 2017, He et al. showed that
by introducing aryl sp
3
defects in single-walled CNTs through a solution-based doping method,
they can achieve single photon emission with a low probability of multiphoton emission at room
temperature
42
. Electroluminescence with various of approaches are also reported by many other
groups
44-49
. Most recently, Wang et al. reported avalanche emission in CNT FETs under extremely
high electric fields (0.5 MV/cm)
16
. Here, the drive currents were three orders of magnitude lower,
while the EL efficiencies were three orders of magnitude higher, than previous reports in the
literature and Raman spectroscopy was used to explicitly rule out thermal emission
8, 32, 50, 11, 23
.
While several demonstrative measurements were shown, this previous report did not study this
effect over a wide range of gate voltage conditions, nor did it present a consistent picture of
electron transport and EL emission.

3.3 Experimental Details

In the work presented here, we have systematically explored light emission in these
42

suspended CNT FET devices over a wide range of experimental device parameters (i.e., gate
voltage, bias voltage, bias current, and light intensity), in order to further establish and understand
the origin of this avalanche emission mechanism. Here, we are able to tune through various
mechanisms of light emission from thermal emission to avalanche electroluminescence. By
comparing the bias voltage dependence of the electric current and EL intensity, we observe a region
under which sub-avalanche light emission is observed. There appears to be a “sweet spot” in gate
voltage at which the avalanche emission process occurs. Here, the current-voltage characteristics
of each device were used to verify that they consist of individual nanotubes rather than multiple
nanotubes. Based on several years of measurements involving these devices, we have found that
individual, suspended nanotube devices exhibit negative differential conductance above
approximately 1.2 V with a maximum current of 3-5 μA
51-53
. A plot of the current vs. bias voltage
showing NDC can be found at the end of this chapter as Figure 3.9.
Figure 3.1(a) shows a schematic diagram of the CNT FET device. In the device fabrication
process, a 1 µm wide trench is etched in a Si/SiO2/Si3N4 wafer approximately 500 nm deep. Thirty
pairs of platinum source and drain electrodes are then patterned on the surface of the wafer along
with one common gate electrode on the bottom of the trench using photolithography. A scanning
electron microscope (SEM) image of a CNT suspended over the trench is shown in Figure 3.1(b).
Figure 3.1(c) shows optical microscope images of a typical chip
34-36
, indicating the approximate
size and location of the focused laser spot. Lithographically-defined windows are introduced into
a photoresist layer on the source and drain electrodes for the controlled deposition of ferric nitrate
(Fe(NO3)3) catalyst. To avoid chemical contamination by the lithographic processes, the last step
43

of the sample fabrication is CNT growth by chemical vapor deposition (CVD) at 825 ℃ using
hydrogen and argon gas, which is passed through pure ethanol
38
. A semiconductor parameter
analyzer is used for the electrical characterization of these devices. During these measurements,
the drain electrode is grounded and all the values of gate and bias voltage are applied with respect
to the drain electrode. Electroluminescence images are collected with a thermoelectrically-cooled
InGaAs camera (Xenics, Inc) with an effective wavelength range from 1000-1600 nm. This IR
camera is capable of outputting the light intensity for selected regions or individual pixels in the
EL image.

Figure 3.1 (a) Schematic diagram, (b) scanning electron microscope (SEM) image and (c) optical microscope
images of a suspended carbon nanotube field effect transistor.

Source Drain
Gate
CNT
Pt
S

S O

p-type
trench
Pt
Electrode
(a)
trench
Suspended
Carbon
Nanotube
Pt Electrodes
(b)
1µ m
0.5µ m
10µ m
(c)
30 pairs of
electrodes
gate
electrode
Catalyst
Catalyst
Pt Electrodes
44

3.4 Results and Discussion

Figure 3.2 (a) Infrared and (b) Electroluminescence images of a CNTFET device.

Figure 3.3 shows semi-log plots of the electric current and the electroluminescence intensity
plotted as a function of applied bias voltage at a gate potential of Vg = +7 V. In Figure 3.3(a), there
is a very steep jump in the current with a very large differential conductance of 90 µA/V , which
indicates the onset of the avalanche breakdown of Vbias = 3.75 V . In Figure 3.3(b), we observe the
onset of electroluminescence at Vbias = 3.4 V , which is 0.35 V below the threshold for avalanche
breakdown, as observed in the current-bias voltage data. Here, in the “sub-avalanche” regime
(between Vbias =3.4 V and 3.8 V), ballistic electrons accelerating in the high electric fields gain
enough kinetic energy for an exciton decay (~0.8 eV), resulting in light emission. The electric field
required to separate these excitons into free electron-hole pairs is substantially high (E ≈ Exciton
binding energy/exciton size ≈ 0.8 MV/cm). This threshold for exciton separation and subsequent
carrier multiplication can be seen by the abrupt increase in current at Vbias = 3.75 V . It should be
noted that, the differential conductance associated with the avalanche process is so steep that the
device reaches the thermal emission regime extremely quickly, as labeled in this plot. This occurs
(a)
(b)
10µ m
10µ m
Electroluminescence
45

when the current reaches the µA range, and thus the voltage range over which avalanche emission
is the dominant mechanism of light emission is actually quite small compared with thermal
emission, as discussed in the following paragraph.

2.5 3.0 3.5 4.0
10
100
1,000
10,000
100,000
EL Intensity (au)
Bias Voltage (V)
2.5 3.0 3.5 4.0
1
10
100
1000
Current (nA)
Bias Voltage (V)

Avalanche
Emission
Sub-
Avalanche
Emission
(a)
(b)
Sub-
Avalanche
Emission
Onset of
Emission
Avalanche
Emission
Thermal
Emission
Figure 3.3 Log plots of (a) the electric current and (b) EL intensity plotted as a function of applied bias voltage
at a gate potential of V g = +7V . The value of the maximum differential conductance is labeled in the plot (a).
46

Figure 3.4(a) shows the minimum bias voltage needed to reach the light detection limit of
our experimental setup plotted as a function of the gate voltage for the same CNT FET device
shown in Figure 3.3. We also measured the current and EL intensity as a function of bias voltage
at different gate voltages. We measured the current and EL intensity as a function of bias voltage
at different gate voltages. By monitoring the light emission as a function of current, we are able to
find the bias voltage at which avalanche and sub-avalanche light emission occurs at each value of
gate voltage and then we connected these points to get the triangular region. Here, we see that
there is a relatively small “triangular” region over which low power avalanche and sub-avalanche
emission is observed, with a current of less than 5 nA and an electrical power less than 20 nW,
which are both well below the threshold current and power for thermal emission
8, 32, 50, 11, 23
.
Outside this “triangular” region and above the solid curve, which is shaded in this plot, the
electrical power is always larger than several µW and the CNT exhibits thermal emission under
these conditions. Figure 3.4(b) shows a 3D plot of the relative electroluminescence efficiency (i.e.,
light intensity/electrical power) plotted as a function of the gate and bias voltages, in which there
is a sharp feature corresponding to avalanche emission. Figure 3.4(c) shows a plot of the maximum
differential conductance (MaxDC) under different values of the applied gate voltage, which is
caused by the carrier multiplication in the avalanche process. Here, the peak in the MaxDC occurs
at the same gate voltage as the avalanche emission and highest EL efficiency. These data sets show
that there appears to be a “sweet” spot around Vg = +7 V , which has the highest light emission
onset voltage, the highest electroluminescence efficiency, and the largest maximum differential
conductance.
47

Figure 3.5 shows similar data from another CNT FET device. Figure 3.5(a) shows the
minimum bias voltage needed to reach the light detection limit plotted as a function of gate voltage
over a relative narrow range (5-10 V). This data also shows a “triangular” avalanche light emission
region and a shaded thermal emission region. Beyond the avalanche light emission region, the
current increases abruptly, as shown in Figure 3.5(c), and thermal heating becomes the dominant
mechanism of emission. Figure 3.5(b) shows a 3D plot of the relative electroluminescence
efficiency plotted as a function of the gate voltage and bias voltage, in which another sharp peak
Figure 3.4 (a) Minimum bias voltage at which light emission is observed plotted as a function of gate voltage.
Three regions are labeled on the plot: avalanche region, thermal emission and no detectible light emission. (b)
3D plot of the relative EL efficiency plotted as a function of applied gate and bias voltages. (c) Plot of the
maximum differential conductance at different values of gate voltage. This data set was measured at the same
time as the data in Figure 3.2 from the same CNTFET device.

48

appears corresponding to avalanche emission. Figure 3.5(c) shows a log plot of the electric current
vs. applied bias voltage at a gate potential of Vg = +7 V for this FET device, exhibiting a steep
current jump with a large differential conductance of 30 µA/V .

Figure 3.5 (a) Minimum bias voltage at which light emission is observed plotted as a function of a narrower
range of gate voltage. Three regions are labeled on the plot: avalanche region, thermal emission and no detectible
light emission. (b) 3D plot of the relative EL efficiency plotted as a function of applied gate and bias voltages.
(c) Log plot of the electric current plotted as a function of applied bias voltage at a gate potential of V g = +7V .

These results provide new mechanistic insight regarding light emission in carbon nanotubes
that extend beyond our previous work
16
. Here, we have identified a regime of sub-avalanche light
emission that we were previously unable to resolve. In addition, we have provided a detailed map
of the light emission as a function of the bias and gate voltages, which exhibits a sharp singularity
5 6 7 8 9 10
2.8
3.0
3.2
3.4
3.6
3.8
4.0
4.2
Minimum Bias Voltage (V)
Gate Voltage (V)
Avalanche
Region
Thermal
Emission
Below Light
Detection Limit
2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.2
1
10
100
1000
10000
Current (nA)
Bias Voltage (V)
(a)
(b)
(c)

Avalanche
Region
49

in the electroluminescence efficiency as a function of these variables. This singularity exemplifies
how stringent the conditions required are to observe avalanche and sub-avalanche emission (drops
off by 2-3 orders of magnitude over just 0.1 V)
50

Figure 3.6 (a) Minimum bias voltage at which light emission is observed plotted as a function of gate voltage.
Three regions are labeled on the plot: avalanche region, thermal emission and no detectible light emission. (b)
Color contour map plot of the EL efficiency plotted as a function of gate and bias voltages measured from the
same device of Figure 3.2 in the manuscript.
-2 0 2 4 6 8 10
1.5
2.0
2.5
3.0
3.5
4.0
Bias Voltage (V)
Gate Voltage (V)
0.1000
0.1887
0.3561
0.6719
1.268
2.393
4.515
8.519
16.08
30.34
57.24
108.0
203.8
280.0
-2 0 2 4 6 8 10
1.5
2.0
2.5
3.0
3.5
4.0
Minimum Bias Voltage (V)
Gate Voltage (V)
Avalanche
Region
Thermal
Emission
Below Light
Detection Limit
(a)
(b)
51

Figure 3.7 Plot of the EL efficiency plotted as a function of bias voltage at gate voltage of 7 V from the same
device of Figure 3.2 in the manuscript.
2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0
0
100
200
300
400
Relative Efficiency (counts/nW)
Bias Voltage (V)
Avalanche
region
52

Figure 3.8 Raman G band shift at (a) bias voltage = 3.8 V with current = 3.2 μA at gate voltage = 10 V , showing
a temperature difference of around 1200K from room temperature, and (b) bias voltage = 3.75 V with current =
2 nA, showing almost no temperature difference under an applied gate voltage of 10 V .
1400 1500 1600 1700
Raman Intensity
Raman Shift (cm-1)
0V at room temperature
Bias=3.8V/current=3.2uA
(a)
1400 1500 1600 1700
Raman Intensity
Raman Shift (cm
-1
)
0V at room temperature
Bias=3.75V/Current=2nA
(b)
53

Figure 3.9 Plot of the electrical current plotted as a function of bias voltage at gate voltage of -10 V from a single
suspended CNTFET device showing negative differential conductance.

54

3.5 Conclusion
In conclusion, in mapping out the electrical bias and gate conditions for light emission in
suspended carbon nanotubes, a narrow region occurs corresponding to avalanche breakdown
which appears with extremely low current (<5 nA) and electrical powers (<20 nW). Here, we find
a sweet spot with the highest electroluminescence efficiency, the highest light emission onset
voltage, and the largest maximum differential conductance occurring under the same gate and bias
voltage conditions, which provides solid evidence for the avalanche emission mechanism. We also
observe efficient light emission below the threshold bias voltage of avalanche emission (i.e., sub-
avalanche light emission) by impact ionization. Above the threshold for avalanche breakdown, the
electroluminescence efficiency quickly decreases by 2-3 orders of magnitude, and thermal
emission becomes the dominant mechanism of light emission.

55

Chapter 4 Broadband Electroluminescence from Reverse Breakdown in
Individual Suspended Carbon Nanotube pn-junctions
This chapter is similar to Wang et al., under reviewed by Nano Research

4.1 Abstract

There are various mechanisms of light emission in carbon nanotubes (CNTs), which give
rise to a wide range of spectral emission characteristics that provide important information
regarding the underlying physical processes that lead to photon emission. Here, we report spectra
obtained from individual suspended CNT dual-gate field effect transistor (FET) devices under
different gate and bias conditions. By applying opposite voltages to the gate electrodes (i.e., Vg1 =
-Vg2), we are able to create a pn-junction within the suspended region of the CNT. Under forward
bias conditions, the spectra exhibit a peak corresponding to E 11 exciton emission via thermal (i.e.,
blackbody) emission occurring at electrical powers around 8µW, which corresponds to a power
density of approximately 0.5 MW/cm
2
. On the other hand, the spectra observed under reverse bias
correspond to impact ionization and avalanche emission, which occurs at electrical powers of
~10nW and exhibits a featureless flat spectrum extending from 1600nm to shorter wavelengths up
to 600nm. Here, the hot electrons generated by the high electric fields (~0.5MV/cm) are able to
produce high energy photons far above the E 11 (ground state) energy. It is somewhat surprising
that these devices do not exhibit light emission by the annihilation of electrons and holes under
forward bias, as in a light emitting diode (LED). Possible reasons for this are discussed, including
Auger recombination.
56

4.2 Introduction

Our ability to control, produce, and enhance light emission from carbon nanotubes (CNTs)
is based largely on photoluminescence measurements in which an intense laser is used to
photoexcite the nanotubes. Over the past ten years, several research groups have reported that
oxygen doping of CNTs using ozonolysis produces localized exciton states, which exhibit long
photoluminescence lifetimes (>1nsec), enhanced photoluminescence intensities (~20X), and
promising g
(2)
-factors up to room temperature.
19-22, 42, 54-59
Kato’s group recently reported single
photon emission at room temperature from air-suspended carbon nanotubes (CNTs).
60
Prior to
these studies, however, single photon emission in carbon nanotubes and other nanoscale materials
(quantum dots, TMDCs) had been relegated to cryogenic temperatures.
21, 61, 62
Hogele et al.
demonstrated that carbon nanotubes emit non-classical light at 4.2K through the observation of
photon antibunching in the photoluminescence of a suspended single carbon nanotube.
21
This was
the first report of quantum correlations of photoemission in a single carbon nanotube (CNT), and
the probability of multiphoton emission was found to be smaller than 3%, indicating that carbon
nanotubes could be used as a source of single photons for applications in quantum cryptography
and quantum information processing. Strauf’s group also observed photon antibunching with g
2
(0)
= 0.15 from cavity-embedded (6,5) CNTs dispersed in a sodium dodecylsulfate (SDS) solution at
9K.
62
It should be noted, however, that in these previous studies light emission was produced by
optical pumping.
While the study of optically-pumped light emission from CNTs has evolved to an
57

unprecedented level of control and sophistication, electrically-driven light emission from
individual carbon nanotubes has lagged way behind. Early reports of electroluminescence (EL)
from individual carbon nanotube devices around the early 2000s included several papers from the
IBM group, who reported an electroluminescence efficiency of ~10
-4
photons per injected
electron–hole pair.
13, 14
Typical electrical powers dissipated in these devices (P=IV) were on the
order of 10s of µW.
14-16, 25, 26, 33, 63
While several mechanisms of light emission have been discussed
in these early EL studies, including electron-hole annihilation and impact ionization, it is likely
that the main mechanism of emission in these early studies was from thermal emission due to
substantial heating. At these applied powers, substantial Joule heating occurs as evidenced by
monitoring the G band shift in their Raman spectra.
16, 30-33
As a result, thermal emission caused by
heating (~1000K) is likely the main mechanism of light emission occurring in these previous works.
In 2018, avalanche photoemission was reported from individual suspended carbon
nanotubes under large applied electric fields, resulting in efficient generation of light without heat
at applied electrical powers of just 4nW.
64
This corresponds to unipolar light emission, thus
circumventing the difficulty associated with injecting electrons into CNTs due to their small
electron affinities and large associated Schottky barriers. Thermal emission by Joule heating was
ruled out by Raman spectroscopy, however, no spectral characterization was done of this avalanche
photoemission process.

58

4.3 Experimental Details

In the work presented here, we have recorded the light emission spectra from suspended
dual-gate CNT FET devices under both forward and reverse bias conditions. Here, we are able to
tune through various mechanisms of light emission from thermal emission to impact ionization
and, eventually, avalanche electroluminescence. In addition to the spectral profiles, we compare
the relative electroluminescence efficiencies of these two basic emission mechanisms.
Figure 4.1(a) shows a diagram of the dual-gate CNT FET device. In the microfabrication
process, a pair of 100nm thick platinum gate electrodes are first deposited on an undoped Si
substrate using photolithography and electron-beam metal deposition. A 600nm thick SiO2 layer
is then deposited using plasma enhanced chemical vapor deposition (PECVD). A trench is etched
through the SiO2 layer approximately 1µm wide and 600 nm deep using reaction ion etching (RIE).
Photolithography is then used to pattern 29 pairs of platinum source and drain electrodes on top of
the SiO2 layer with 100nm in thickness. The gap between the source and drain electrodes is 2 µm,
and the gap between the two gates is 200 nm. Figure 4.1(b) shows a scanning electron microscope
(SEM) image of a suspended CNT across a pair of source and drain electrodes. Figure 4.1(c) shows
optical microscope images of a typical dual-gate CNT FET chip
34-36
. Catalyst windows (5µm ×
5µm) are patterned in a photoresist layer on top of the source and drain electrodes near the trench
enabling us to deposit ferric nitrate Fe(NO3)3/Al2O3-based catalyst. The final step of the sample
fabrication process is chemical vapor deposition (CVD) of CNTs at 825 ℃ by bubbling hydrogen
and argon gas through pure ethanol.
38
For the electrical characterization of these devices, a
semiconductor parameter analyzer (HP, Inc) is used. Electroluminescence (EL) images are
59

collected with a thermoelectrically-cooled InGaAs camera (Xenics, Inc) with a 1100-1600nm
effective wavelength range. EL spectra are collected using a homebuilt spectrometer system with
a liquid nitrogen-cooled InGaAs array (Princeton Instrument, Inc) over the same effective
wavelength range.

Figure 4.1 (a) Schematic diagram, (b) scanning electron microscope (SEM) image, and (c) optical microscope
images of a dual-gate, partially suspended carbon nanotube field effect transistor.

trench
Pt Source
Electrode
(a)
trench
Suspended
Carbon
Nanotube
(b)
5µ m
1µ m
50µ m
(c)
a pair of gate
electrodes
Catalyst Catalyst
CNT
Pt
SiO
2
Si
Pt
S
D
G1
G2
Pt Gate
Electrodes
Pt Source
Electrode
Pt Drain
Electrode
Set of Drain
Electrodes
Set of Source
Electrodes
Trench w/ Gate
Electrodes
Pt Drain
Electrode
60

4.4 Results and Discussion

Figure 4.2(a) shows a plot of current-gate voltage (I-Vg) characteristics measured from a
suspended dual-gate CNT FET device obtained by shorting the two gates together and keeping the
bias voltage at a constant value of 0.2 V . This I-Vg curve indicates that this is an ambipolar device
with a charge neutral point at Vg=1.75V . However, the contact resistance associated with n-type
conduction (Vg=+10V) is approximately 1000X higher than that of p-type conduction (Vg=-10V).
Figure 4.2(b) shows the current plotted as a function of bias voltage obtained by setting the two
gate voltages to equal and opposite values of Vg1 = -Vg2 = 10 V , which results in the formation of
a pn-junction within the CNT. By reversing the gate voltages, we can achieve reversible rectifying
behavior, indicating that the rectifying behavior is not simply due to one of the Schottky contacts
associated with the Pt/CNT junction. Calculated conduction and valence band profiles of a dual-
gate CNT FET device showing the formation of pn junctions at (a) Vg1 = -Vg2 = -15V and (b) Vg1
= -Vg2 = 15V under a bias voltage of 3V are shown in Figure 4.3. Figure 4.2(c) shows the electric
current and the EL intensity plotted together as a function of bias voltage while gating another
device in the np-configuration (i.e., Vg1 = -Vg2 = +10 V). Under forward bias, the current increases
abruptly above Vb>0.5V , however, electroluminescence is not observed until Vb>2V when the
current is above 2µA. This corresponds to a regime in which the electrical power exceeds 4µW
and light emission occurs via thermal emission. Under reverse bias, we observe light emission for
bias voltages below -3V with currents less than 4nA (P=12nW). This corresponds to the sub-
avalanche regime in which light emission occurs via impact ionization.
61

Figure 4.2 (a) Current-gate voltage (I-V g) characteristics measured from a suspended dual-gate CNT FET device
obtained by shorting the two gates (i.e., V g1=V g2) and applying a constant bias voltage of 0.2 V . (b) Current
plotted as a function of bias voltage by gating the device in the pn (V g1 = -10 V and V g2 = 10 V) and np (V g1 =
10 V and V g2 = -10 V) configurations. (c) Current and EL intensity plotted as a function of bias voltage at V g1 =
-V g2 = 10V .

-10 -8 -6 -4 -2 0 2 4 6 8 10
1E-9
1E-8
1E-7
Current (A)
Gate Voltage (V)
-2.0 -1.6 -1.2 -0.8 -0.4 0.0 0.4 0.8 1.2 1.6 2.0
-1.50E-008
-1.00E-008
-5.00E-009
0.00E+000
5.00E-009
1.00E-008
1.50E-008
Current (A)
Bias Voltage (V)
PN -10V 10V
NP 10V -10V
(a)
(b)
-4 -3 -2 -1 0 1 2 3 4
0
500
1000
1500
2000
2500
Current
Light Intensity
Bias Voltage (V)
Current (nA)
0
200
400
600
800
1000
1200
1400
Light Intensity (Counts)
(c)
62

Figure 4.3 Calculated conduction and valence band profiles of a dual-gate CNT FET device at (a) V g1 = -V g2 =
-15V and (b) V g1 = -V g2 = 15V under a bias voltage of 3V .

0.5 1.0 1.5 2.0 2.5
-6
-5
-4
-3
-2
-1
Band Voltage (V)
Position ( mm)
Valence Band
Conduction Band
NP
Junction
V
g1
= -V
g2
= -15V
0.5 1.0 1.5 2.0 2.5
-6
-5
-4
-3
-2
-1
0
Band Voltage (V)
Position ( mm)
Valence Band
Conduction Band
PN
Junction
(a)
(b)
V
g1
= -V
g2
= +15V
Reverse
Bias
Forward
Bias
63

Figure 4.4(a) shows the spectrum of thermal emission taken under forward bias at Vb = 3.5
V and Ib=2.3µA (P=8µW), which exhibits a broad peak corresponding to a thermally-broadened
E 11 exciton transition. Photoluminescence (PL) spectra taken from two representative dual-gate
CNT FETs are plotted in Figure 4.5. Figure 4.4(b) shows the spectrum of sub-avalanche light
emission taken under reverse bias at Vb = -3.6 V and Ib=4nA (P=14.4nW), which exhibits a flat
spectrum extending significantly beyond the ground state E 11 transition. Here, sub-avalanche light
emission occurs at relative light emission efficiency (EL intensity/electric power) that is 150X
higher than thermal emission. It is somewhat surprising that these devices do not exhibit light
emission under lower forward bias currents (<1µA) by the annihilation of electrons and holes, as
in a light emitting diode (LED). One possible reason for this is the high non-radiative
recombination rates in the nanotubes, as evidenced by the relatively high reverse saturation
currents, which are typically above 1nA.
65
Auger recombination, which is a non-radiative process,
is another possible mechanism by which the light emission efficiency is reduced by more than one
order of magnitude at high gate voltages.
66-69
Also, the high n-type contact resistances, which result
in a substantial voltage drop across the contact instead of the pn-junction itself. Figure 4.4(c) and
4.4(d) show visible-light-range spectra of thermal emission and sub-avalanche light emission
collected from another dual gate CNT FET device using a silicon CCD detector. Figure 4.4(c)
shows the spectrum of thermal emission taken in the forward biased region at Vb = 3.7 V and Ib =
2.0µA (P = 7.4µW), which exhibits a peak around 785nm, corresponding to a thermally-broadened
E22 exciton feature. Figure 4.4(d) shows the spectrum of sub-avalanche light emission taken under
reverse bias at Vb = -3.8 V and Ib=4nA (P=15.2nW), which exhibits a featureless flat spectrum.
64

Here, again, sub-avalanche light emission also exhibits a much higher (~200X) relative light
emission efficiency (EL intensity/electric power) than that of thermal emission within the visible
wavelength range.

1200 1300 1400 1500
0
100
200
300
400
EL Intensity (Counts)
Wavelength (nm)
Thermal Emission
Background
1200 1300 1400 1500
0
50
100
150
200
250
EL Intensity (Counts)
Wavelength (nm)
Avalanche Emission
Background
600 650 700 750 800 850 900
0
50
100
150
200
250
300
EL Intensity (Counts)
Wavelength (nm)
Thermal Emission
Background
600 650 700 750 800 850 900
0
30
60
90
120
150
EL Intensity (Counts)
Wavelength (nm)
Avalanche Emission
Background
(a)
(c)
V
b
= 3.7V
I = 2µ A
P = 7.4µ W
V
b
= -3.8V
I = 4nA
P = 15nW
Efficiency = 6Counts/nW
Efficiency = 1184Counts/nW
E
11
(d)
(b)
V
b
=3.5V
I=2.3µ A
P=8µ W
V
b
=-3.6V
I=4nA
P=14nW
E
22
Figure 4.4 (a) Spectrum of thermal emission taken at V b = 3.5 V and I b=2.3µA, and (b) spectrum of sub-avalanche
light emission taken at V b = -3.6 V and I b=4nA from a suspended dual-gate CNT FET device. Spectra collected
from another device within the visible wavelength range of (c) thermal emission taken at V b = 3.7 V and I b=2µA,
and (d) sub-avalanche light emission taken at V b = -3.8 V and I b=4nA.
65

Figure 4.5 PL spectra taken from two reprensentative suspended dual-gate CNT FETs.

1350 1400 1450 1500 1550 1600 1650
0
3000
6000
9000
PL Intensity (Counts)
Wavelength (nm)
1350 1400 1450 1500 1550 1600 1650
0
1000
2000
3000
PL Intensity (Counts)
Wavelength (nm)
(a)
(b)
66

Photon emission from avalanche breakdown in silicon was reported in 1956 by Chynoweth
and McKay. In this work, photon emission up to 3.2eV was observed from a silicon pn-junction
under reverse bias.
70
This is almost three times higher than the band gap energy of silicon, which
exemplifies how electrons accelerating in high electric fields can gain enough kinetic energy to
emit photons many times greater than the band gap (or ground state) energy (i.e., E 11 exciton in
the case of a CNT) of a material. More recently, van Drieënhuizen et al. reported simulations of
above-band gap emission in avalanche-mode silicon pn junctions under high fields (~10
5
V/cm).
71

In carbon nanotubes, however, electrons can accelerate in these high fields without scattering
because of the limited number of scattering states in k-space that conserve momentum. That is,
CNTs provide a unique one-dimensional system in which phonon scattering is suppressed,
enabling hot electrons to emit high energy photons.

4.5 Conclusion

In conclusion, we report spectra obtained from individual suspended CNT pn-junction
devices under forward and reverse bias. Under forward bias, the spectra exhibit a relatively
inefficient thermal emission peak centered around the E 11 exciton at electrical powers of
approximately 8µW. Under reverse bias, however, we observe efficient broadband emission
extending up to wavelengths as short as 600nm corresponding to impact ionization and avalanche
emission, at electrical powers of ~10nW. These devices do not exhibit light emission by the
annihilation of electron and holes under forward bias, as occurs in light emitting diodes (LEDs),
indicating high non-radiative recombination rates possibly due to Auger recombination.
67

Chapter 5 Auger Suppression of Nanotube Incandescence in Individual
Suspended CNT pn-junctions
This chapter is similar to Wang et al., published in ACS Applied Materials & Interfaces

5.1 Abstract

There are various mechanisms of light emission in carbon nanotubes (CNTs), which give
rise to a wide range of spectral characteristics that provide important information. Here, we report
suppression of incandescence via Auger recombination in suspended carbon nanotube pn-junctions
generated from dual-gate CNT field effect transistor (FET) devices. By applying equal and
opposite voltages to the gate electrodes (i.e., Vg1 = -Vg2), we create a pn-junction within the CNT.
Under these gating conditions, we observe a sharp peak in the incandescence intensity around zero
applied gate voltage, where the intrinsic region has the largest spatial extent. Here, the emission
occurs under high electrical power densities around 0.1MW/cm
2
(or 6µW) and arises from thermal
emission at elevated temperatures above 800K (i.e., incandescence). It is somewhat surprising that
this thermal emission intensity is so sensitive to the gating conditions, and we observe a 1000-fold
suppression of light emission between Vg1=0V and Vg1=15V , over a range in which the electrical
power dissipated in the nanotube is roughly constant. This behavior is understood on the basis of
Auger recombination, which suppresses light emission by the excitation of free carriers. Based on
the calculated carrier density and band profiles, the length of the intrinsic region drops by a factor
of 7-25X over the range from |Vg|=0V to 15V . We, therefore, conclude that the light emission
intensity is significantly dependent on the free carrier density profile and the size of the intrinsic
region in these CNT devices.
68

5.2 Introduction

Light emission from nanoscale devices (i.e., III-V semiconducting quantum dots and carbon
nanotubes (CNTs)) exhibits interesting optoelectronic and optical phenomena, including localized
exciton-based light emission, room-temperature single photon emission, and trions
19-22, 54-58, 60, 72
.
The IBM group first reported electrically-driven light emission from on-substrate single CNT pn-
junctions
13, 14, 73
with a relative efficiency of light emission around 10
-4
photon/pair of injected
electrons and holes. After that, electrically-driven light emission was reported from several CNT
field effect transistor (FET) devices by several groups
18, 23-26, 33, 44-46, 48, 74
. In these studies,
detectable light emission was only achieved at relatively high applied electrical powers, above
1µW. Under these cases, significant Joule heating occurs, as evidenced by Raman G band
frequency shifts,
16, 31-33, 75
and it has become clear that, in these previous works, Joule heating
caused thermal emission as the main mechanism of light emission. Additionally, electrically-
driven light emission from silicon-based pn-junctions and FETs were also reported
76-79
. Other than
FET-based devices, electrically-driven light emission were also reported using other approaches,
including asymmetric contacts and arrays of aligned CNTs
47, 80-83
. Under very high electric fields
(0.5MV/cm), light emission via avalanche breakdown was also reported in air-suspended CNT
FETs
16, 84, 85
. Under these conditions, the light emission efficiencies were much higher (three orders
of magnitude) than former reports of electrically-driven light emission in the literature, and the G
band frequency was monitored to rule out thermal emission explicitly
31-33, 75
.

69

5.3 Experimental Details

In the work presented here, we explore the incandescence characteristics of dual-gated CNTs
as the pn-junction is swept from reverse bias to forward bias. Before any LED-like light emission
can be observed (i.e., reaches the light emission detection limit), the incandescence becomes
extremely significant due to high electrical power density and consequent heating. While no LED-
like light emission is observed, clear signatures of avalanche emission and thermal emission are
seen, including an interesting peak around zero gate voltage (i.e., Vg1 = -Vg2 ≈ 0). To the best of
our knowledge this is the only system where the incandescence (i.e., thermal emission) intensity
can be tuned by electrostatic doping while the electrical power (i.e., device temperature) is held
constant. An incandescence spectrum is collected around the peak point in order to corroborate the
broad band thermal emission.
Figure 5.1(a) shows the suspended dual-gate CNT field effect transistor (FET) device used
in this work. Two separate platinum (Pt) gate electrodes shared by 29 pairs of drain and source
electrodes are first deposited on a high resistivity silicon wafer. Then, on top of the silicon, an
approximately 600nm thick SiO2 layer is grown using PECVD. A trench is etched through the SiO2
layer 1000 nm wide and 600 nm deep using dry etching (RIE). Photolithography is then used to
deposit 29 pairs of drain and source Pt electrodes on the surface of the wafer. The gap between the
drain and source Pt electrodes is 2 µm, the suspended length is 1µm, and the gap between gates is
200 nm. Figure 5.1(b) shows a colorized scanning electron microscope (SEM) image of a
suspended CNT across drain and source electrodes. Optical microscope images of a representative
dual-gate CNT FET chip are shown in Figure 5.1(c)
35, 37, 86
. In order to deposit Fe(NO3)3/Al2O3-
70

based catalyst, lithographically-defined windows (5µm X 5µm) are written on the drain and source
Pt electrodes in a photoresist layer. The final step of the device fabrication is CNT growth, using
argon and hydrogen gas bubbled through ethanol with high purity at 825℃, via a homebuilt
chemical vapor deposition (CVD) system.
38
The devices were electrical characterized using a
semiconductor measurement unit system. Here, we ground the drain electrode for all the electric
measurement process while the bias voltage is applied to the source. Incandescence images are
taken with an InGaAs camera (Xenics, Inc), which is thermoelectrically-cooled and provides
sensitivity over the 1000-1600nm wavelength range. The Xenics software is capable of reading
light intensities from a selected area or a pixel within the incandescence images.
Photoluminescence and incandescence spectra are collected using the same homebuilt
spectrometer system with an InGaAs array (Princeton Instruments, Inc), which is sensitive over
the same effective wavelength range as our imaging detector. A 785nm CW laser with a power of
0.2mW is used in the NIR spectrometer system.

71

Figure 5.1 (a) Schematic diagram, (b) colorized scanning electron microscope (SEM) image and (c) optical
microscope images of a dual-gate suspended carbon nanotube field effect transistor.

5.4 Results and Discussion

Figures 5.2(a) and 5.2(b) show a 1100nm-wavelength illuminated image and an
incandescence image of a dual-gate CNT FET using an InGaAs camera, respectively. Here,
incandescence occurs under applied gate voltages of Vg1 = 2V , Vg2 = -2V and a current of I = 2.2µA
at a bias voltage of Vbias = 3.5V . Figure 5.2(c) shows the incandescence intensity (blue curve) and
the electric current (black curve) plotted as a function of gate voltage 1 (i.e., Vg1) with opposite
and equal gate voltages (i.e., Vg1 = -Vg2) and keeping the bias voltage at a constant value of 3.5V .
While the electrical current and power remain almost constant over the range -3V ≤ Vg1 ≤ 10V (P
1µ m
trench
Pt Source
Electrode
(a)
trench
Pt Gate
Electrodes
(b)
5µ m
1µ m
50µ m
(c)
Set of Drain
Electrodes
a pair of gate
electrodes
Catalyst
Catalyst
Pt Source
Electrode
Trench w/ Gate
Electrodes
Pt Drain
Electrode
Pt
SiO
2
Si
Pt
S
D
G1
G2
Set of Source
Electrodes
Pt Drain
Electrode
Suspended
Carbon
Nanotube
600nm
2µ m
72

≈ 6µW), the incandescence intensity exhibits a peak around Vg1 = -Vg2 = 0. A plot of the G band
Raman frequency as a function of gate voltage taken from a typical dual-gate CNT FET is shown
at the end of this chapter as Figure 5.7. Under conditions when the applied current is larger than
1µA, substantial downshifts is shown (>10cm
-1
), which corresponds to temperatures above 800K
based on a coefficient of around 0.02 cm
-1
/K.
31, 87
Here, we observe a 1000-fold suppression in the
incandescence intensity between Vg1=0V and Vg1=15V . Here, the light emission intensity is
asymmetric with respect to positive and negative gate voltages (i.e., forward and reverse bias),
consistent with the simulation results shown below in Figure 5.4.
73

Figure 5.2 (a,b) IR image with illumination and incandescence image (without illumination) from a suspended
dual-gate CNT FET device taken with an InGaAs camera. (c) Current and incandescence intensity plotted as a
function of gate voltage 1 with V g1 = -V g2 and V bias = 3V .

-15 -10 -5 0 5 10 15
0
500
1000
1500
2000
2500
Current
Incandescence
Intensity
Gate Voltage 1 (V) (V
g1
= -V
g2
)
Current (nA)
0
2000
4000
6000
8000
10000
12000
14000
16000
Intensity (Counts)
CNT
Incandescence
(a)
(b)
(c)
10µ m
10µ m
Forward
Bias
Reverse
Bias
Source
Electrodes
Drain
Electrodes
Suspended
CNT
74

Figure 5.3(a) plots the current (black curve) and incandescence intensity (blue curve) as a
function of gate voltage (Vg1) for another device, which exhibits similar behavior to that shown in
Figure 5.2c. Here, the two gate voltages are again equal and opposite (i.e., Vg1 = -Vg2) and the bias
voltage is held at a constant value of +3V . Here, the data is plotted across a narrower gate voltage
range (-10 ≤ Vg1 ≤ +8) than the device shown in Figure 5.2. Figure 5.3(b) plots the current-
normalized photon counts (counts/current) as a function of gate voltage. Here, we should note that,
although the light intensity plot shows a sharp peak around zero gate voltage, the relative efficiency
of sub-avalanche emission at Vg1=-Vg2=-10V (570 counts/nA) is much higher than that at the peak
point around zero gate voltage, which is around 10 counts/nA. This result agrees with our previous
report that the efficiency for sub-avalanche (i.e., impact ionization) emission is substantially higher
than thermal emission
16, 84
. Figure 5.3(c) shows the spectra of photoluminescence and
incandescence taken at Vb = +3V , I = 2µA, and Vg1=0, which exhibits a broad (thermally broadened)
band gap emission corresponding to E 11 exciton. It should be noted that, when Vg1 is negative with
a corresponding current of less than 10 nA, the device becomes reverse biased and sub-avalanche
emission appears due to impact ionization.
75

1200 1300 1400 1500 1600
0
100
200
300
400
500
600
700
800
Intensity (Counts)
Wavelength (nm)
Incandescence
Photoluminescence
-10 -8 -6 -4 -2 0 2 4 6 8
0
500
1000
1500
2000
2500
3000
Current
Incandescence
Intensity
Gate Voltage 1 (V) (V
g1
= -V
g2
)
Current (nA)
0
2000
4000
6000
8000
10000
12000
14000
16000
Intensity (Counts)
(a)
(b)
(c)
Sub-avalanche
emission
E
11
-10 -8 -6 -4 -2 0 2 4 6 8
0
100
200
300
400
500
600
Current-normalized
photon counts (Counts/nA)
Gate Voltage 1 (V) (V
g1
= -V
g2
)
0.05X
cutoff
Figure 5.3 (a) Current and incandescence intensity plotted as a function of gate voltage 1 with V g1 = -V g2 and
V bias = 3V . (b) Current-normalized photon counts plotted as a function of gate voltage 1 with V g1 = -V g2 and V bias
= 3V . (c) Incandescence spectrum taken at the red-circled point indicated in (a) and photoluminescence spectrum
taken from the same dual-gate CNT FET device.

76

We hypothesize that this peak around zero gate voltage (shown in Figures 5.2c and 5.3a)
appears when Auger recombination in the nanotube is minimized, particularly in the suspended
region of the nanotube. At finite gate voltages, the gate-induced carriers are known reduce the
emission efficiency via Auger recombination, which is a non-radiative process
66-69
. This presents
a mechanism by which the light emission efficiency is reduced by more than one order of
magnitude as the intrinsic width becomes smaller. It is interesting to note that, in both Figures 5.2c
and 5.3a, the intensity peak occurs at only zero gate voltage. This stands in contrast to previous
reports of photoluminescence quenching by Yasukochi et al. in which the intensity peak extends
over a range of voltages around zero gating
68
. While we do not fully understand this dissimilar
behavior, there are several differences between this device configuration. Namely, our device is a
dual-gate pn-junction rather than a single-gate device. Also, our intensity profile was obtained
under high bias conditions and, hence, high temperatures ~800K. Lastly, incandescence and
photoluminescent are fundamentally different processes, and can potentially be affected differently
by the Auger recombination mechanism, and there are likely different competing recombination
mechanisms that dominate under these gate and bias conditions.
In order to further understand the underlying Auger suppression of the incandescence
intensity observed in these pn-devices around zero gate voltage, we simulated the valence and
conduction band profiles and carrier density profiles along the length of the CNT under various
gate voltages and a constant bias voltage via a self-consistent Poisson solver in the TCAD
Sentaurus software package
88
. Figure 5.4(a) shows the valence and conduction band profiles along
the device, and Figure 5.4(b) shows a semi-log plot of the carrier density profiles at Vg1 = -Vg2 =
77

15V under a bias voltage of 3V . Here, we find that the size of the width of the intrinsic region drops
by a factor of 7-25X over the range from |Vg| = 0V to 15V , as shown in Figure 5.4c (Vb = 3V).
Here, the voltage dependence of length of the intrinsic region (i.e., minimal Auger recombination)
closely resembles the incandescence intensity profiles of Figures 5.2c and 5.3a. Even the gate
voltage-dependent asymmetry can be seen in the measured (Figure 5.2c) and calculated (Figure
5.4c) results indicating that, under forward bias, the intrinsic region extends further than under
reverse bias. We, therefore, conclude that the suppressed light emission intensity is caused by the
decreased width of the intrinsic region in these CNT devices around zero gate voltage. Additional
valence and conduction band profiles and the corresponding carrier density profiles are included
as Figure 5.5.
78

Figure 5. 4 (a) Calculated conduction and valence band profile and (b) calculated carrier density profile of a
dual-gate CNT FET device at V g1 = -V g2 = 15V under a bias voltage of 3V . (c) Width of the intrinsic region
plotted as a function of applied gate voltage with V g1 = -V g2 at V b = 3V .
0.5 1.0 1.5 2.0 2.5
10
-20
10
-15
10
-10
10
-5
10
0
10
5
10
10
Electron
Hole
Electron+Hole
Carrier Density (m
-1
)
Position ( mm)
-15 -10 -5 0 5 10 15
0.0
0.5
1.0
1.5
2.0
Intrisic Region Length ( mm)
Gate Voltage 1 (V) (V
g1
= -V
g2
)
0.5 1.0 1.5 2.0 2.5
-6
-5
-4
-3
-2
-1
0
Band Voltage (V)
Postion ( mm)
Valence Band
Conduction Band
(a)
(c)
Forward
Bias
Reverse
Bias
V
g1
= 15V
Bias = 3V
P doped
N doped
I
V
b
= +3V
V
g1
= 15V
Bias = 3V
P doped
N doped
I
(b)
0.3µm
79

Figure 5.5 Calculated conduction and valence band profiles of a dual gate CNT FET device at (a) V g1 = -V g2 =
-15V , (b) V g1 = -V g2 = 0V and (c) V g1 = -V g2 = 15V under a bias voltage of 3V . The formation of pn-juctions are
labeled between dashed lines.
0.5 1.0 1.5 2.0 2.5
-6
-5
-4
-3
-2
-1
0
Band Voltage (V)
Postion ( mm)
Valence Band
Conduction Band
0.5 1.0 1.5 2.0 2.5
-6
-5
-4
-3
-2
-1
Band Voltage (V)
Postion ( mm)
Valence Band
Conduction Band
V
g1
= -V
g2
= -15V
0.5 1.0 1.5 2.0 2.5
-6
-5
-4
-3
-2
-1
Band Voltage (V)
Postion ( mm)
Valence Band
Conduction Band
V
g1
= -V
g2
= 0V
V
g1
= -V
g2
= 15V
(a)
(b)
(c)
80

The intrinsic region is defined as the part with minimized induced charge within the CNT.
The length of the intrinsic region is then calculated by measuring the distance between the two
turning points of the region with smaller carrier densities from the carrier density profiles, as shown
in Figure 5.6.

Figure 5.6 Calculated carrier density profile (a) V g1 = -V g2 = -15V , (b) V g1 = -V g2 = 0V and (c) V g1 = -V g2 = 15V
(d) V g1 = -V g2 = -5V and (e) V g1 = -V g2 = 5V under a bias voltage of 3V . The intrinsic regions are labeled between
dashed lines.
0.5 1.0 1.5 2.0 2.5
10
-20
10
-15
10
-10
10
-5
10
0
10
5
10
10
Electron
Hole
Electron+Hole
Carrier Density (m
-1
)
Position ( mm)
V
g1
= -V
g2
= 15V
Forward Bias
0.3µm
0.5 1.0 1.5 2.0 2.5
0
5
10
15
20
25
30
35
40
Electron
Hole
Electron+Hole
Carrier Density (m
-1
)
Position ( mm)
0.5 1.0 1.5 2.0 2.5
0
5x10
8
1x10
9
2x10
9
2x10
9
Carrier Density (m
-1
)
Position ( mm)
Electron
Hole
Electron+Hole
V
g1
= -V
g2
= 0V
V
b
= 3V
V
g1
= -V
g2
= -15V
V
b
= 3V
Reverse Bias
0.08µm
2µm
0.5 1.0 1.5 2.0 2.5
0
1x10
8
2x10
8
3x10
8
4x10
8
5x10
8
6x10
8
Electron
Hole
Electron+Hole
Carrier Density (m
-1
)
Postion ( mm)
3V 3V
3V 3V
V
g1
= -V
g2
= -5V
V
b
= 3V
Reverse Bias
0.18µm
0.5 1.0 1.5 2.0 2.5
10
-15
10
-10
10
-5
10
0
10
5
Electron
Hole
Electron+Hole
Carrier Density (m
-1
)
Postion ( mm)
V
g1
= -V
g2
= 5V
V
b
= 3V
Forward Bias
0.56µm
3V
(c)
(b)
(a)
(d)
(e)
81

Figure 5.7 (a) Raman G band spectra taken from one dual-gate CNT FET device under different gate conditions
(V g1 = -V g2) at a constant bias voltage of 3V . (b) The Raman G band frequency and current plotted as a function
of V g1 (V g1 = -V g2).

Figure 5.8 shows the IV curves taken at different gate conditions. For zero gate voltage, the
CNT remains p-type and it works as a resistor while it works as a diode when it is gated to forward
biased region. Before a threshold bias voltage of around 2V , the conductance remains larger for
zero gate condition.

1300 1400 1500 1600 1700
0
2,000
4,000
6,000
8,000
10,000
12,000
Raman Intensity (Counts)
Raman Shift (cm
-1
)
-10_12nA
-9_3nA
-7_5nA
-5_4nA
-3_10nA
-1V_1.03uA
0V_1.42uA
3V_1.4uA
5V_1.48uA
10V_1.51uA
original
-10 -5 0 5 10
1578
1580
1582
1584
1586
1588
1590
1592
-10 -5 0 5 10
0
200
400
600
800
1000
1200
1400
1600
G-band Frequency
Current
Gate Voltage 1 (V)
G-band Frequency (cm
-1
)
Current (nA)
(a) (b)
82

Figure 5.8 Current plotted as a function of bias voltage under different gate conditions taken from the same
device shown in Figure 5.2.

5.5 Conclusion

We report a suppression behavior in the incandescence of suspended CNT dual-gated pn-
devices. Here, the peak around zero gate voltage arises from the alignment of the intrinsic region
of the pn-junction with the trench. It is somewhat surprising that this thermal emission intensity is
so sensitive to the gating conditions, and we observe a 1000-fold suppression due to Auger
recombination between Vg1=0V and Vg1=15V at nearly constant electrical power. This alignment
suppresses thermal emission from the device by increasing Auger recombination. That is, Auger
recombination (i.e., non-radiative recombination) suppresses light emission under high gating
conditions. From the calculated carrier density and band profiles, a strong correlation is established
between incandescence intensity and the size of the intrinsic region, consistent with the Auger
recombination mechanism.
-2 -1 0 1 2
-1500
-1000
-500
0
500
1000
Current (nA)
Bias Voltage (V)
PN -10V 10V
NP 10V -10V
V
g
=0
83

Chapter 6 Defect-Induced Photoluminescence Enhancement and
Corresponding Transport Degradation in Individual Suspended Carbon
Nanotubes
This chapter is similar to Wang et al., published in Physical Review Applied

6.1 Abstract

The utilization of defects in carbon nanotubes to improve their photoluminescence
efficiency has become a widespread study towards the realization of efficient light emitting devices.
Here, we report a detailed comparison of defects in nanotubes (quantified by Raman spectroscopy)
and photoluminescence (PL) intensity of individual suspended carbon nanotubes (CNTs). We have
also the evaluated the impact of these defects on the electron/hole transport in the nanotubes, which
is crucial for the ultimate realization of optoelectronic devices. We find that brightly luminescent
nanotubes exhibit a pronounced D-band in their Raman spectra, and vice versa, dimly luminescent
nanotubes exhibit almost no D-band. Here, defects are advantageous for light emission by trapping
excitons, which extends their lifetimes. We quantify this behavior by plotting the PL intensity as a
function of the D/G band Raman intensity ratio, which exhibits a Lorentz distribution peaked at
D/G=0.17. For CNTs with a D/G ratio >0.25, the PL intensity decreases, indicating that, above
some critical density, non-radiative recombination at defect sites dominates over the advantages
of exciton trapping. In an attempt to fabricate optoelectronic devices based on these brightly
luminescent CNTs, we transferred these suspended CNTs to platinum electrodes and found that
the brightly photoluminescent nanotubes exhibit nearly infinite resistance due to these defects
84

while those without bright photoluminescence exhibit finite resistance. These findings indicate a
potential limitation in the use of brightly luminescent CNTs for optoelectronic applications.

6.2 Introduction

Enhanced photoluminescence in carbon nanotubes and 2D materials due to defect-localized
exciton states has been reported by several research groups over the past few years.
61, 89-98
This
work typically requires defects to be created through some form of post-growth treatment. For
example, oxygen doping of carbon nanotubes (CNTs) through ozonolysis has been shown to
produce localized exciton states that exhibit enhanced photoluminescence intensities (~20X), long
photoluminescence lifetimes (>1nsec), and single photon emission, even at room temperature.
22,
54-56
These oxygen dopant sites in carbon nanotubes are well understood with a detailed atomic
scale picture based on DFT calculations, and robust methods exist for creating these defect/dopant
sites. UV/ozone treatment of air-suspended CNTs has also been shown to provide up to 5-fold
enhancements in the PL intensity by the creation of such defects. While there have been many
purely optical studies of CNTs, studies of optoelectronic phenomena are relatively few.
15, 23, 26, 43,
99
Also, a vast majority of previous carbon nanotube studies were performed on large ensembles
of nanotubes rather than individual CNTs.

6.3 Experimental Details

In the work presented here, we correlate the defect density in as-grown CNTs with their PL
intensity without requiring the need for any post-growth processing. Previous studies were carried
85

out on ensembles of nanotubes and, thus, it was not possible to obtain a 1-to-1 correlation between
the PL intensity and the D-band Raman mode, which provides a relative measure of the amount of
defects in carbon nanotubes.
100-102
Here, we study air-suspended CNTs rather than surfactant-
suspended CNTs. The electrical resistance is used to further characterize the nature of these
defects and the extent to which they perturb the electron and hole transport in this system.
Samples are fabricated by etching 8µm deep, 2µm × 2µm pillars in a quartz substrate
using a Cl-based reactive ion etch plasma. Optical and electron microscopy images of these pillars
are shown in Figure 6.1. Before etching, a 1nm-thick film of Fe is deposited by electron beam
evaporation to serve as a catalyst for the nanotube growth. Carbon nanotubes are grown by
chemical vapor deposition (CVD) using ethanol as the carbon feedstock at 825
o
C, as reported
previously.
86, 103
Figure 6.1c shows a scanning electron microscope (SEM) image of CNTs
suspended across these pillars grown using this technique. The PL related data is collected on a
home built PL imaging system at room temperature, as shown in Figure 6.1(d). In this system, a
785 nm wavelength laser source (Spectra-Physics, Model 3900) is used to irradiate the sample.
The illumination area is approximately 60μm in diameter. A 1100nm long pass filter is used to
eliminate any Rayleigh or Raman scattered light. The PL signal is then collected with a
thermoelectrically-cooled InGaAs camera (Xenics, Inc.), which is sensitive to IR light emission in
the 1100-1600nm wavelength range. This IR camera is capable of outputing the light intensity for
selected regions or individual pixels in the PL image. During these PL measurements, all the bright
nanotubes of interest are moved to the same position relative to the center of the excitation area,
and the PL intensity is obtained from the same pixel in order to minimize the variation between
86

measurements. An optical microscope image of a 4 × 5 array of quartz pillars is shown in Figure
6.1a. Figure 6.1b shows a PL image taken from the same region of this sample. A bright line can
be seen connecting two adjacent pillars, corresponding to PL emission from a suspended CNT.
Raman spectra from these same individual suspended CNTs are collected using a Renishaw InVia
micro-spectrometer. All Raman spectra are measured at room temperature using a 785nm laser
source with the same incident laser power, objective lens, grating, and integration time. Before
collecting each Raman spectrum, several attempts are made to optimize the position and obtain the
strongest Raman signal.

50 100 150 200 250 300
50
100
150
200
250
0
200
400
600
800
1000
1200
1400
1600
PL Image
(b)
10µ m
suspended
CNT
Optical Image
(a)
suspended
CNT
10µ m
suspended
CNT
2µ m
(c)
785nm Laser
Dichroic
Mirror (785nm)
CNT
Longpass
Filter (1100nm)
InGaAs
Camera
Objective Lens
(d)
Figure 6.1 (a) Optical, (b) photoluminescence, and (c) SEM images of carbon nanotubes suspended across
quartz pillars. (d) Schematic diagram of the photoluminescence imaging setup.

87

6.4 Results and Discussion

Figure 6.2 shows the Raman spectra of 4 different suspended carbon nanotubes that exhibit
dim photoluminescence. Figure 6.2a shows a representative photoluminescence image of one such
dim nanotube. For all 4 nanotubes, the D-band is almost undetectable in these Raman spectra.
The D/G band Raman intensity ratio for these nanotubes spans a range from 0 to 0.041.

1300 1400 1500 1600
Normalized Raman Intensity
Raman Shift (cm
-1
)
D/G~0
10µ m
suspended
CNT
D/G=0.01
D/G=0.038
D/G=0.046
(b)
(a)
Figure 6.2 (a) Representative photoluminescence image and (b) Raman spectra of four different suspended
carbon nanotubes that exhibit dim photoluminescence.

88

Figure 6.3 shows the Raman spectra of 5 brightly photoluminescent CNTs, along with a
representative PL image (Figure 6.3a). All of these Raman spectra exhibit pronounced D-bands,
indicating the presence of a substantial amount of defects in these nanotubes. Here, the D/G band
Raman intensity ratio spans a range from 0.075 to 0.16. Here, we believe that exciton localization
at these defect sites prevents non-recombination that occurs at the ends of the CNT, ultimately
extending their luminescence efficiencies and lifetimes. Interestingly, even though the defect-
enhanced PL is provided by localized excitons (i.e., 0D system), these nanotubes appear as lines
in the PL images, indicating that there are many such defects along the length of each nanotube.

Figure 6.3 (a) Representative photoluminescence image and (b) Raman spectra of five different suspended
carbon nanotubes that exhibit bright photoluminescence.
1300 1400 1500 1600
Normalized Raman Intensity
Raman Shift (cm
-1
)
suspended
CNT
D/G=0.075
10µ m
suspended
CNT
D/G=0.09
D/G=0.1
D/G=0.106
D/G=0.16
(b)
(a)
89

Figure 6.4 shows the Raman spectra of 4 nanotubes that exhibit D/G band Raman intensity
ratios greater than 0.25. These highly defective nanotubes exhibit relatively dim
photoluminescence. Here, we believe that non-radiative recombination at these defect sites
outweighs the advantageous effects of exciton trapping created by these defect sites. It is also
possible that these nanotubes have more substantial types of defects, such as vacancies or 5-7
defects.

Figure 6.4 (a) Representative PL image and (b) Raman spectra of 4 different suspended carbon nanotubes that
exhibit D/G ratios > 0.25.
1300 1400 1500 1600
Normalized Raman Intensity
Raman Shift (cm
-1
)
10µ m
D/G=0.258
D/G=0.3
D/G=0.31
10µ m
suspended
CNT
(b)
(a)
D/G=0.35
90

Figure 6.5 plots the PL intensity of 13 nanotubes as a function of the D/G band Raman
intensity ratio. Here, the data exhibits a Lorentzian distribution peaked at D/G=0.17. The wide
dynamic range of the PL, here spanning almost 2 orders of magnitude, is immediately apparent in
this plot. For CNTs with D/G ratios < 0.15, the PL intensity increases with defect density due to
exciton localization. However, for CNTs with D/G ratios > 0.25, the PL intensity decreases,
indicating that, above some critical density, non-radiative recombination at defect sites dominates
over the advantages of exciton localization.

Figure 6.5 Photoluminescence intensity plotted as a function of the D/G band Raman intensity ratio for 13
different suspended carbon nanotubes. The data here is fit to a Lorentzian function centered around D/G = 0.17.

91

In an attempt to fabricate optoelectronic devices based on these brightly luminescent
nanotubes, we developed a flip-chip transfer process to transfer brightly photoluminescent
nanotubes suspended across quartz pillars to pre-patterned metal electrodes (i.e., Pt) on a separate
chip, as illustrated in Figure 6.6. In this process, the CNT-containing pillars are slowly brought
into contact with the electrode chip using a home built contact aligner. The quartz substrate is then
lifted off slowly using a z-axis micromanipulator, resulting in complete transfer of the desired
nanotube to the metal electrodes, suspended over the trench. Figure 6.6d shows an optical
microscope image of the two chips in contact during the transport process.

Figure 6.6 (a-c) Schematic diagrams and (d) optical microscope image of the flip-chip transfer process.

Quartz Pillar
CNT
Source (Pt)
Gate (Si)
Drain (Pt)
(a) (b)
(c) (d)
10µ m
92

Figure 6.7a shows an SEM image of a suspended CNT that has been successfully transferred
using this technique. Disappointingly, 11 out of 11 bright nanotubes that were successfully
transferred to these electrodes showed an extremely high resistance (R>1GΩ) due to the defects
associated with exciton trapping. It is worth mentioning that the one-dimensional nature of CNTs
make them particularly susceptible to defects, since electrons can only scatter backwards. Figure
6.7b shows the current-gate voltage (I-Vgate) characteristics of a non-bright nanotube that was
transferred to these electrodes. Here, the suspended nanotube was first identified by SEM rather
than PL imaging. The I-Vgate characteristics show field effect transistor behavior that is typical of
a semiconducting nanotube with a weak gate. Here, the effect of the underlying silicon gate is
weak because the electrodes are 8µm tall and, hence, the underlying silicon is relatively far away
from the CNT channel. Nevertheless, at a gate voltage of Vgate=-4V , the suspended nanotube
exhibits a resistance of 200kΩ, which is at least three orders of magnitude lower than that of the
brightly photoluminescent nanotubes that were transferred to the same electrodes. This data
indicates the strong role that these defects play in preventing electron/hole transport, which will
likely limit practical applications in electronically-driven light emission from carbon nanotubes.
93

Figure 6.7 (a) SEM image, (b) current-voltage characteristics, and (c) Raman spectrum of a CNT that was
successfully transferred to metal electrodes.

1µ m
Suspended
CNT
Metal
Electrodes
-4 -2 0 2 4
0
20
40
60
80
100
Current (nA)
Gate Voltage (V)
R = 200k W
1300 1400 1500 1600
0
400
800
1200
1600
Raman Intensity (Counts)
Raman Shift (cm
-1
)
(a)
(b)
(c)
94

The spatial resolution of our PL imaging setup is around 1µm and the resistance of a blank
chip used in the flip-chip transfer technique is around 2GΩ, as shown in Figure 6.8 and Figure 6.9.

Figure 6.8 (a) Luminescence image and (b) spatial profile of the near-IR electroluminescence (λ>1100nm)
intensity of an individual suspended carbon nanotube.

1 2 3 4 5 6
0
100
200
300
400
500
600
700
EL Intensity (arb. units)
Position ( mm)
FWHM~1 μm
(a)
(b)
10 μm
95

Figure 6.9 The electric current plotted as a function of applied bias voltage from a blank chip containing the
metal electrodes used in the flip-chip transfer technique, showing a resistance of around 2GΩ.

6.5 Conclusion

In conclusion, we have compared the Raman spectra and photoluminescence (PL) intensities
of individual suspended carbon nanotubes (CNTs). We find that brightly luminescent nanotubes
exhibit pronounced D-bands in their Raman spectra, whereas dimly luminescent nanotubes exhibit
almost no D-band. The relative defect density is quantified using the D/G band Raman
intensity ratio. By plotting the PL intensity as a function of the D/G band Raman intensity ratio,
we observe a Lorentzian distribution peaked at D/G=0.17. CNTs with a D/G ratio above 0.25 show
a decreased PL intensity, indicating the point beyond which defects cause non-radiative
recombination rather than exciton trapping. When these brightly luminescent nanotubes are
96

transferred to metal electrodes, their resistance is found to be infinite (R>1GΩ) because of the
presence of these defects. However, non-luminescent CNTs exhibit finite resistance. These results
indicate that there may be an inherent limitation in the ultimate realization of optoelectronic
devices with a tradeoff between luminescence efficiency and low resistance carrier transport.

97

Chapter 7 Radiation-Induced Direct Band Gap Transition in Few-layer MoS2
This chapter is similar to Wang et al., published in Applied Physics Letters

7.1 Abstract

We report photoluminescence (PL) spectroscopy of air-suspended and substrate-supported
molybdenum disulfide (MoS2) taken before and after exposure to proton radiation. For 2-, 3-, and
4-layer MoS2, the radiation causes a substantial (>10X) suppression of the indirect band gap
emission, likely due to a radiation-induced decoupling of the layers. For all samples measured
(including monolayer), we see the emergence of a defect-induced shoulder peak around 1.7eV ,
which is redshifted from the main direct band gap emission at 1.85eV . Here, defects induced by
the radiation trap the excitons and cause them to be redshifted from the main direct band emission.
After annealing, the defect-induced sideband disappears but the indirect band emission remains
suppressed indicating a permanent transition to direct band gap material. While suspended 2-, 3-,
and 4-layer MoS2 show no change in the intensity of the direct band emission after radiation
exposure, substrate-supported MoS2 exhibits an approximately 2-fold increase in the direct band
gap emission after irradiation. Suspended monolayer MoS2 shows a 2-3X drop in PL intensity,
however, substrate-supported monolayer MoS2 shows a 2-fold increase in the direct band emission.

98

7.2 Introduction

Radiation effects can severely and negatively impact the performance of microelectronic
and optoelectronic devices operating in high radiation environments, such as near particle
accelerators and in the space radiation environment. Radiation induced displacement damage is
particularly harmful for minority carrier devices such as optoelectronic components, diodes, and
bipolar junction transistors (BJTs), where lattice defects can provide non-radiative decay pathways,
greatly reducing the minority carrier lifetime.
104
This is of particular concern for direct band gap
2D semiconductors, such as MoS2, due to the interest in these materials for optoelectronic
applications. Also, radiation-induced charging of insulators in close proximity to semiconductor
materials can induce unintentional changes in the carrier density due to the field effect. This is of
concern because 2D semiconductors are typically supported on insulating substrates when used as
the channel material in advanced field effect transistors (FETs).
105
From a fundamental scientific
perspective, these low dimensional materials provide a good platform for studying the effects of
radiation due to their high surface-to-volume ratio and the unique physics associated with indirect-
to-direct band gap transitions, excitons and trions.
106, 107

In addition to studying the effects of energetic particle radiation, it is worth noting that
energetic particles are commonly used for surface treatment in the semiconductor industry. For
example, plasma treatments are used to improve material surface wettability and bonding, surface
functionalization, and material etching and doping. In the work presented here, we report the
effects of radiation exposure that improve the optoelectronic properties of multi-layer MoS2 by
inducing an indirect-to-direct band gap transition and, thereby, increasing the PL efficiency of the
99

material.
There have been several studies of the effects of radiation on the electronic properties of 2D
materials. Kim et al. studied the effects of proton radiation on MoS2 FETs, and observed a
substantial decrease in the conductance as well as shifts in the threshold voltage after irradiation.
108

Zhang et al. reported the effects of X-ray irradiation on similar back-gated MoS2 transistors, which
resulted in decreased conductance and substantial shifts in the threshold voltage.
109
Lee and
coworkers studied the effects of 662 keV -radiation on suspended MoS2 micromechanical
resonators, and found that charges induced by the radiation caused a frequency upshift of 2% of
the resonance frequency.
110
The radiation hardness of MoS2 FETs was investigated by Ochedowski
with 1.14 GeV U
28+
ions, and significant changes in the structural and electrical properties were
observed by electrical characterization, atomic force microscopy (AFM), and Raman
spectroscopy.
111
Tsai et al. showed that trilayer MoS2 metal-semiconductor-metal photodetectors
exhibit high responsivity (∼1.04 A/W) and photogain, and are stable under 2MeV proton
illumination radiation indicating that they can be used in harsh environments.
112
Zan et al.
demonstrated that electron beam radiation damage of MoS2 flakes can be mitigated by
encapsulation in graphene.
113
While these previous studies have focused on the electronic
properties of irradiated MoS2, there have been few reports of the optoelectronic properties after
irradiation. Tongay et al. investigated the effects of α-particle radiation on substrate-supported,
monolayer MoS2.
93
In this previous work, they observed defect-activated photoluminescence and
a blue shift in the direct transition PL peak after irradiation, as described by the interplay between
bound and free excitons. However, to date there have been no reports on the effects of energetic
100

particle radiation on PL from multi-layer 2D semiconductors, or suspended monolayer 2D
semiconductors. Several other groups have studied the effect of ion irradiation on 2D materials
both theoretically and experimentally.
114-119
However none of the these previous studies discuss an
indirect-to-direct bandgap transition.

7.3 Experimental Details

In the work presented here, we study the effects of proton radiation on the optoelectronic
properties of MoS2. The PL spectra give sensitive information about the changes in the band
structure (i.e., indirect-to-direct band gap) and defects through the suppression of the indirect
1.4eV peak and defect-induced sidebands. We study the radiation effects on both suspended and
substrate-supported MoS2, for a variety of different thicknesses. Post-radiation annealing is also
performed in order to investigate the nature of the defects. By investigating suspended material
alongside substrate-supported material, we are able to understand the fundamental changes in the
optical properties of the material without substrate perturbation, and also independently understand
the influence of the substrate.
Exfoliated MoS2 flakes are transferred to 20µm wide silicon nitride membranes containing
several holes approximately 2-3µm in size, as shown in Figure 7.1. Here, MoS2 is first exfoliated
onto transparent polydimethylsiloxane (PDMS) substrates. Once a flake of the desired thickness
is located, the flake is then transferred to one of the “hole” regions in the SiN membrane using a
home built contact aligner (i.e., dry transfer method) with a 50X objective lens.
120
PL spectra were
collected from both suspended and non-suspended regions of the MoS2 using a Renishaw InVia
101

micro-spectrometer using 532nm wavelength light before and after proton irradiation. The
thickness of the flakes are determined from the contrast in the optical microscope images and the
energy of the indirect band emission peak in the PL spectra.
121, 122
Scanning electron microscope
(SEM) images were taken after all optical characterization, in order to avoid contamination and
potential quenching of the MoS2 photoluminescence. Figure 7.1 shows optical microscope and
SEM images of one of the flakes measured in this study. No visible differences in the sample can
be seen in the optical or electron microscopy after irradiation.

102

The LEAF (Low Energy Accelerator Facility) contains a 400keV ion accelerator capable of
producing beam currents from 10pA up to 1mA. Its primary use is implanting with protons (H
+
),
(a)
(b)
Suspended
Regions
MoS
2
Flake
SiN Membrane
10 µm
Figure 7.1 (a) Optical microscope and (b) SEM image of an MoS 2 flake deposited on a SiN membrane. Two
holes in this membrane provide a suspended region of the flake to measure, which behave quite differently from
the substrate-supported region.
103

although other heavier ions are also possible. Here, we focus on energetic protons since they are
the most common type of space radiation, and thus allow us to best evaluate the effect caused by
exposure of this material to the space radiation environment. Using a scanned particle beam, the
system delivers fluences as low as 1×10
9
up to 1×10
17
particles/cm
2
. For the MoS2 irradiations, we
irradiate at 100keV with fluences ranging from 2×10
12
to 6×10
14
particles/cm
2
. Here, the radiation
beam aperture (1”) is larger than the sample (3mm). Using a beam current of 1µA at a flux of
~10nA/cm
2
, the sample sees an incoming power of 1mW from the beam during irradiation. The
low power of the particle beam eliminates the need for cooling of the sample during irradiation.

7.4 Results and Discussion

Figure 7.2 shows the PL spectra of a bilayer flake measured before and after irradiation with
a fluence of 6x10
14
protons/cm
2
. Several PL spectra were collected from each sample, which
exhibited a high degree of uniformity. Each of these spectra were fitted with three Lorentzian peaks,
and the details of the curve fits of bilayer MoS2 are given at the end of this chapter as Figure 7.9.
Here, a Lorentzian function provides better fits to the spectra than a Gaussian. Both the suspended
and substrate-supported regions show almost complete suppression of the indirect emission (1.55-
1.6eV) after irradiation and the emergence of a defect-induced sideband peak around 1.7eV . Here,
defects induced by the radiation trap the excitons and cause them to be redshifted from the main
direct band emission at 1.83eV . In the suspended region, the direct band emission at 1.83eV
increases by 1.6X after irradiation, while the substrate-supported direct band emission increases
by a factor of 2.7X. After annealing the samples at 300
o
C for 1 hour in argon, the defect-induced
104

sideband peak disappears. However, the indirect band emission remains suppressed after annealing,
indicating that the material has undergone an irreversible indirect-to-direct band gap transition.
These measurements were repeated on several other bilayer flakes, and the same results were
consistently observed. Here, we believe that the proton irradiation is causing a slight decoupling
of the layers, as we have previously observed using an oxygen plasma treatment.
107
The density
functional theory (DFT) calculations of Lake and coworkers established that an increase in the
interlayer separation of just 1Å is sufficient to induce an indirect-to-direct band gap transition in
this material
107
, which was confirmed experimentally with AFM measurements.
107
While these
DFT calculations provide a qualitative explanation of the mechanism underlying this indirect-to-
direct band gap transition, we do not currently have an atomistic picture of the ion interaction with
the lattice of this material. The suppression of the indirect band gap emission is potentially useful
for making optoelectronic devices such as LEDs and solar cells. Spectra taken after an
intermediate proton fluence of 6×10
13
protons/cm
2
and a heavier fluence of 6×10
15
protons/cm
2

are also shown as Figure 7.6 and Figure 7.7. For the heaviest fluence, the PL intensity decreases,
indicating that the effects of radiation-induced defects eventually outweigh the indirect-to-direct
band gap transition.

105

1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1
0
7,000
14,000
21,000
28,000
35,000
Intensity (Counts)
Energy (eV)
After
Irradiation
After
Annealing
Suspended
After
Irradiation
After
Annealing
On-substrate
1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1
0
5,000
10,000
15,000
20,000
Intensity (Counts)
Energy (eV)
Before
Irradiation
Before
Irradiation
Indirect
Direct
(a)
(b)
Figure 7.2 PL spectra of (a) suspended and (b) on-substrate bilayer MoS 2 taken before and after proton
irradiation, and after annealing.

106

Figures 7.3 and 7.4 show MoS2 samples with 3 or more layers in thickness (after irradiation
and before annealing) which, again, show slightly enhanced direct band gap emission, complete
suppression of the indirect emission around 1.4eV , and a broadened, red-shifted defect induced
sideband around 1.7eV . After annealing, the indirect emission remains suppressed, indicating a
indirect-to-direct band gap transition stable against annealing at 300 °C. Also, the defect-induced
sideband disappears and the direct emission decreases to roughly their pre-radiation intensities
after annealing.
107

1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1
0
5,000
10,000
15,000
20,000
Intensity (Counts)
Energy (eV)
After
Irradiation
After
Annealing
Suspended
1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1
0
4,000
8,000
12,000
16,000
Intensity (Counts)
Energy (eV)
After
Irradiation
After
Annealing
On-substrate
Before
Irradiation
Before
Irradiation
Indirect
Direct
(a)
(b)
Figure 7.3 PL spectra of (a) suspended and (b) on-substrate trilayer MoS 2 taken before and after proton
irradiation, and after annealing.
108

1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1
0
4,000
8,000
12,000
16,000
Intensity (Counts)
Energy (eV)
1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1
0
5,000
10,000
15,000
20,000
Intensity (Counts)
Energy (eV)
After
Irradiation
After
Annealing
Suspended
After
Irradiation
After
Annealing
On-substrate
Before
Irradiation
Before
Irradiation
Indirect
Direct
(a)
(b)
Figure 7.4 PL spectra of (a) suspended and (b) on-substrate MoS 2 with four layers taken before and after proton
irradiation, and after annealing.

109

Figure 7.5 shows the PL spectra of a monolayer MoS2 flake measured before and after
irradiation with a fluence of 6×10
14
protons/cm
2
. This monolayer data serves a good control group,
showing the material’s response to changes in the defect concentration without corresponding
changes in the band structure. In the suspended region, we see a 2-fold drop in the intensity of the
direct band gap emission after irradiation, and the emergence of a defect-induced sideband peak at
1.71eV . As mentioned above, defects induced by the radiation trap the excitons and cause them to
be redshifted from the main direct band emission at 1.85eV . For the suspended region, we
understand this decrease in intensity due to defects, which cause non-radiative recombination and
shorten the lifetime of the photoexcited carriers. Under all conditions, the PL intensity in the
substrate-supported region of the MoS2 is much weaker than the suspended region, which is
consistent with the original reports of Mak et al.
121
Here, we see a 2X increase in the direct band
gap emission at 1.85eV for the substrate-supported region after irradiation. As with the suspended
region, we observe an increase in the defect-induced peak around 1.7eV , as was reported by Tongay
et al.
93
After annealing the samples at 300
o
C for 1 hour in argon, the defect-induced sideband peak
disappears in the suspended region and is reduced by 5X in the substrate supported region. Also,
the direct band emission reverts back to its original pre-irradiated intensity, indicating that we can
fully restore the material to its original state by annealing the defects. These measurements were
repeated on several other monolayer flakes and the same results were consistently observed.
Spectra taken at a fluence of 6x10
13
protons/cm
2
can be compared with the PL results reported by
Tongay et al,
93
who reported PL spectra of MoS2 exposed to 5x10
13
particles/cm
2
for 3.04 MeV -
particle radiation, which have similar stopping power (i.e., deposit a similar energy per path length)
110

compared to the 100 keV protons using in the work reported here.

1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1
0
50,000
100,000
150,000
200,000
250,000
Intensity (Counts)
Energy (eV)
Before
Irradiation
After
Irradiation
After
Annealing
Suspended
After
Irradiation
After
Annealing
On-substrate
1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1
0
5,000
10,000
15,000
20,000
Intensity (Counts)
Energy (eV)
Before
Irradiation
Direct
(a)
(b)
Figure 7.5 PL spectra of (a) suspended and (b) on-substrate monolayer MoS 2 taken before and after proton
irradiation, and after annealing.

111

Figure 7.6 PL spectra of monolayer and bilayer MoS 2 taken before and after proton irradiation with intermedium
fluence (6× 10
13
protons/cm
2
) and heavy fluence (6× 10
14
protons/cm
2
).
1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1
0
2,000
4,000
6,000
8,000
10,000
Before Irradiation
Intermedium Dose
Heavy Dose
Intensity (Counts)
Energy (eV)
(a)
(c)
1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1
0
15,000
30,000
45,000
60,000
75,000
90,000
Intensity (Counts)
Energy (eV)
Before Irradiation
Intermedium Dose
Heavy Dose
Suspended On-substrate
1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1
0
4,000
8,000
12,000
16,000
Intensity (Counts)
Energy (eV)
Before Irradiation
Intermedium Dose
Heavy Dose
1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1
0
2,000
4,000
6,000
8,000
10,000
Intensity (Counts)
Energy (eV)
Before Irradiation
Intermedium Dose
Heavy Dose
Monolayer Bilayer
(b)
(d)
112

Figure 7.7 PL spectra of monolayer and bilayer MoS 2 taken after proton irradiation with heavy fluence (6× 10
14

protons/cm
2
) and ultra-heavy fluence (6× 10
15
protons/cm
2
).

7.5 Conclusion

In summary, we demonstrate enhanced photoluminescence and direct band gap emission in
proton irradiated MoS2 flakes. For all substrate-supported samples and most of the suspended
samples, the PL intensity increased after irradiation and the indirect emission is almost completely
suppressed, due to decoupling of the layers (i.e., direct-to-indirect transition). In all samples
measured, we see the emergence of a defect-induced sideband peak around 1.7eV that can be
annealed out by heating the samples to 300
o
C. Only suspended monolayer MoS2 shows a decrease
in the PL intensity by a factor of 2.8X, likely due to the creation of defects and recombination
centers.
1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1
0
8,000
16,000
24,000
32,000
40,000
Intensity (Counts)
Energy (eV)
Ultra-heavy Dose
Heavy Dose
Suspended
1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1
0
3,000
6,000
9,000
12,000
Intensity (Counts)
Energy (eV)
Ultra-heavy Dose
Heavy Dose
On-substrate
1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1
0
3,000
6,000
9,000
12,000
Intensity (Counts)
Energy (eV)
Ultra-heavy Dose
Heavy Dose
1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1
0
2,000
4,000
6,000
8,000
10,000
Intensity (Counts)
Energy (eV)
Ultra-heavy Dose
Heavy Dose
(a)
(c)
Monolayer Bilayer
(b)
(d)
113

Figure 7.8 PL intensity ratio of suspended/substrate-supported MoS 2 plotted as a function of layer number.

In addition, the spectra of suspended multi-layer MoS2 look very similar to those of
substrate-supported multi-layer MoS2 (See Figure 7.8), reflecting the decreased sensitivity to
substrate interactions as the material approaches the bulk limit. Perhaps the most interesting
finding is that, after annealing, the indirect emission remains completely suppressed, indicating
that this is a robust way to convert this material from an indirect band gap semiconductor to a
direct band gap semiconductor.

1 2 3 4
0
5
10
15
20
25
Ratio of Sus/Sub
Number of Layers
Before Irradiation
After Irradiation
After Annealing
114

Figure 7.9 Curve fits of the PL spectra of bilayer MoS 2 taken before and after proton irradiation, and after
annealing. This is the same data plotted in Figure 7.2 of the manuscript. The peak positions and intensities are
indicated in each plot.
1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1
0
2,000
4,000
6,000
8,000
Intensity (Counts)
Energy (eV)
1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1
0
4,000
8,000
12,000
16,000
Intensity (Counts)
Energy (eV)
1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1
0
2,000
4,000
6,000
8,000
10,000
Intensity (Counts)
Energy (eV)
1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1
0
800
1,600
2,400
3,200
4,000
Intensity (Counts)
Energy (eV)
1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1
0
2,000
4,000
6,000
8,000
10,000
12,000
Intensity (Counts)
Energy (eV)
1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1
0
600
1,200
1,800
2,400
Intensity (Counts)
Energy (eV)
Suspended On-substrate
After Annealed
(a) (b) (c)
Before After Irradiation
(d) (e) (f)
e
nt

115

Chapter 8 Formation of Brightly Luminescent MoS2 Nano-Islands from
Multi-Layer Flakes via Plasma Treatment and Laser Exposure
This chapter is similar to Wang et al., under reviewed by ACS Applied Electronic
Materials

8.1 Abstract

A robust and reliable method for enhancing the photoluminescence (PL) of multilayer MoS2
is demonstrated using an oxygen plasma treatment process followed by laser exposure. Here, the
plasma and laser treatments result in an indirect-to-direct band gap transition. The oxygen plasma
creates a slight decoupling of the layers and converts some of the MoS2 to MoO3. Subsequent laser
irradiation further oxidizes the MoS2 to MoO3, as confirmed via X-ray photoelectron spectroscopy
(XPS), and results in localized regions of brightly luminescent MoS2 monolayer triangular islands
as seen in high resolution transmission electron microscope (HRTEM) images. The PL lifetimes
are found to decrease from 494 psec to 190 psec after plasma and laser treatment, reflecting the
smaller size of the MoS2 grains/regions. Atomic force microscope (AFM) imaging shows a 2nm
increase in thickness of the laser-irradiated regions, which provides further evidence of the MoS2
being converted to MoO3.

116

8.2 Introduction

Since 2010, intense research in the optical and optoelectronic properties of two-dimensional
transition metal dichalcogenides (TMDCs) has led to great advancements in the materials
preparation
105, 108-111, 121, 123-125
, processing
106, 107, 126, 127
and devices
93, 112, 113, 126, 128-130
. Many
interesting physical phenomena are seen in these 2D materials, including lasing from monolayer
MoS2 deposited on photonic crystal cavities
131, 132
and interlayer excitons bound at the interface
between WSe2 and MoSe2
133-137
. While this intriguing class of materials presents new degrees of
freedom in the design and optimization of optoelectronic devices, their extremely high surface-to-
volume ratio makes them particularly sensitive to defects, surface contaminants, and other
extraneous perturbations from the underlying substrate and environment
138-140
. As such, post
processing techniques are needed to mitigate these potential problems and, in general, provide a
more robust material platform for reliable device fabrication.
Several techniques have been developed to treat MoS2 flakes in order to increase their
photoluminescence (PL) efficiency. Amani et al. reported that treatment with an organic super acid
can uniformly enhance the PL and minority carrier lifetime of monolayer MoS2 flakes with “near
unity” quantum efficiency
129
. Mouri et al. reported that chemical-doped monolayer MoS2 flakes
can show tunable and enhanced PL by changing the doping level of the doped MoS 2 flakes
141
. In
2014, Nan et al. reported a high temperature annealing process that significantly increased the PL
efficiency of monolayer MoS2 flakes through the formation of Mo-O bonds
142
. Our group reported
both oxygen plasma-based and proton irradiation-based treatment of multilayer MoS2 flakes that
show stable direct band gap transitions and significantly enhanced PL efficiencies
8,138, 139
. In 2016,
117

we reported that the PL intensity from multilayer MoSe2 flakes can be enhanced through selective
oxidation of one MoS2 layer
143
.

8.3 Experimental Details

In the work presented here, we combine an oxygen plasma treatment and subsequent laser
exposure to create “triangular” MoS2 nano-islands in two steps, which can be potentially used as
quantum dots and single photon emitters. This approach enables patterning of specific regions of
the TMDC material by the selective exposure to light. These plasma and laser-treated materials are
characterized using PL spectroscopy, optical microscopy (i.e., transparency), X-ray photoelectron
spectroscopy (XPS), atomic force microscopy (AFM), and transmission electron microscopy
(TEM). Several control experiments are carried out to understand the role of each step of the
treatment (i.e., O2 plasma versus laser exposure) in the MoS2-to-MoO3 transformation. The effect
of oxygen plasma treatment on MoS2 was heavily explored in the past. However, in the work
presented here, the laser exposure followed by the oxygen plasma will further change the lattice
structure of MoS2 flakes as well as the E11 exciton lifetime, as evidenced by TEM images and
XPS spectra.
In the sample preparation process, crystalline MoS2 is first mechanically exfoliated onto
transparent polydimethylsiloxane (PDMS). Based on the indirect band peak from PL spectra, once
an MoS2 flake of the desired size and thickness is located via optical microscopy, the flake is
transferred using a home built contact aligner (i.e., dry transfer method) with a 50x objective lens
onto a TEM-compatible silicon nitride substrate
120
. The flakes are then treated with a remote
118

oxygen plasma for 3 minutes using an XEI Evactron Soft Clean plasma cleaner, in which the
plasma is generated from room air flowing past a 15 W RF source at 200 mTorr. The flake is placed
10cm downstream from the oxygen plasma source. In this remote configuration, the radical species
lose most of their kinetic energy but still remain chemical reactivity. The flake is then exposed to
laser irradiation in ambient air using a 532 nm CW laser focused through a 100X (0.86NA)
objective lens to an approximately 1µm beam size and intensity of 3.5 mW/µm
2
. PL spectra were
collected from these MoS2 flakes before and after the plasma/laser treatment using a InVia micro-
spectrometer (Renishaw, Inc.) using 532 nm-wavelength laser with a power density of 0.03
mW/µm
2
, which is two orders of magnitude weaker than the laser exposure treatment and, thus,
does not change the material during the PL measurements. X-ray photoelectron spectroscopy
(XPS) was performed using a 10 µm spot size on a ULV AC PHI VersaProbe III (Physical
Electronics, East Chanhassen, MN USA) with a monochromated Al Kα source on the MoS2 flakes
before the treatment, after plasma treatment, and after the plasma and laser treatment. AFM images
of the laser exposed regions in the MoS2 were taken using dimension ICON AFM (Bruker, Inc).
Time-resolved photoluminescence (TRPL) spectra were taken from the MoS 2 flakes before the
treatment, after plasma treatment, and after the plasma and laser treatment, and were carried out
on a laser scanning, confocal optical microscope using a 50 fsec pulsed 405 nm diode laser with a
pump power of < 60 nW and repetition rate of 20 MHz. The time-resolved PL spectra were
detected with single-photon avalanche photodiodes (SPAD, SPCM-AQR-14, PerkinElmer)
connected to a HydraHarp 400 (PicoQuant) time-correlated single-photon counting system. To rule
out laser-induced heating effects, we have included results taken from a MoS2 flake that was
119

thermally annealed in an argon environment at 500 °C for 1 hour immediately following the plasma
treatment at the end of this chapter as Figure 8.6.
8.4 Results and Discussion

Figure 8.1a shows a transmission electron microscopy (TEM) image of a suspended MoS2
bilayer, which is treated by oxygen plasma only. In this TEM image, we are able to clearly see the
atomic lattice structure, which is almost the same as an untreated bilayer MoS 2 flake. Figure 8.1b
shows a transmission electron microscopy (TEM) image of a suspended MoS2 bilayer after plasma
and laser treatment, showing the formation of triangular nano-islands, which is indicated by the
different contrast (i.e., the difference of energy loss of the electron beam) of the TEM images, in
this two-step process (i.e., plasma followed by laser exposure). Here, the nano-islands of
monolayer material are embedded within the bilayer MoS2, indicating the partial conversion of
one MoS2 layer to MoO3, consistent with the XPS data, described below. It is also apparent from
these images that the light-induced oxidation of the MoS2 material follows the axes of the crystal,
forming triangular patterns. These resulting monolayer nano-islands of MoS2 are more than one
order of magnitude more luminescent than the original bilayer material.

120

Figure 8.1 Transmission electron microscope (TEM) images of MoS 2: (a) oxygen plasma treated only and (b)
oxygen plasma treated and laser exposed.

Figure 8.2a shows the PL spectra of bilayer MoS2 (indicated by the position of the indirect
band gap emission) taken before and after plasma and laser exposure. Here, we see a 7.2-fold
increase in the peak PL intensity after the treatment with a 22% linewidth narrowing and complete
suppression of the indirect band gap transition. Figure 8.2b shows the PL spectra taken before
treatment, after a 3 minute oxygen plasma exposure only, and after a 3 minute oxygen plasma and
a 3 minute laser exposure. Here, we see a 40 meV blueshift from 1.83 eV to 1.87 eV and the direct
gap PL linewidth also narrows from 98.5 meV to 65.5 meV while the PL intensity remains almost
unchanged. As a control experiment, two few-layer MoS2 flakes were thermally annealed at 500 °C
in Argon gas at a flow rate at 1 SLM after the plasma treatment step instead of laser exposure. We
find PL intensity to be an order-of-magnitude weaker after thermally annealed flakes, indicating
that the transition to MoO3 with MoS2 islands is a photo-driven rather than thermally-driven
process with oxygen plasma-treated MoS2 flakes.
5nm 5nm
monolayer
monolayer
Bilayer
(a) (b)
121

Figure 8.2 (a) PL spectra for a bilayer MoS 2 flake before and after plasma treatment and laser exposure. (b)
Normalized PL spectra for the same bilayer MoS 2 flake taken before treatment, after oxygen plasma treatment
only and after oxygen plasma treatment and laser exposure.

(a)
(b)
1.4 1.6 1.8 2.0
Normalized
Photoluminescence Intensity
Energy (eV)
Pre-Treatment
Oxygen Plasma Only
Post-Treatment
1.4 1.6 1.8 2.0
Photoluminescence Intensity
Energy (eV)
Pre-Treatment X 7.2
Post-Treatment
122

Figure 8.3 shows XPS spectra of a 10 µm diameter region of multi-layer MoS2 flakes taken
before and after plasma and laser exposures. Figure 8.3a shows the Mo 3d spectrum taken before
treatment, where we observe Mo 3d3/2, Mo 3d5/2 and S 2s states that are consistent with
stoichiometric MoS2
144
. After oxygen plasma treatment, we observe two additional peaks that
correspond to the MoO3 doublet (i.e., MoO3 3d3/2 and MoO3 3d5/2 states), as indicated in Figure
8.2b
144-147
. Based on these spectra, we believe that we have created sub-regions of MoO3 within
the MoS2. Figure 8.3c shows the Mo 3d spectrum taken after oxygen plasma and a 60 second laser
exposure. The MoO3 contribution becomes more prominent, increasing from 31 % of the total Mo
3d5/2 counts in Figure 8.2b to 65% in Figure 8.3c, indicating an additional transformation of MoS2
to MoO3.
123

238 236 234 232 230 228 226 224
XPS Intensity
Binding Energy (eV)
238 236 234 232 230 228 226 224
XPS Intensity
Binding Energy (eV)
Mo 3d
5/2
Mo 3d
3/2
MoO
3
3d
3/2
MoO
3
3d
5/2
Mo 3d
5/2
Mo 3d
3/2
S 2s
S 2s
238 236 234 232 230 228 226 224
XPS Intensity
Binding Energy (eV)
Mo 3d
5/2
Mo 3d
3/2
MoO
3
3d
3/2
MoO
3
3d
5/2
S 2s
(a)
(b)
(c)
Before
Treatment
After Plasma
Treatment
After Plasma
Treatment and
Laser Expose
31% MoO
3
69% MoS
2
65% MoO
3
35% MoS
2
Figure 8.3 XPS spectra for (a) untreated, (b) oxygen plasma treated only, and (c) oxygen plasma treated and
laser exposed bi-layer MoS 2 flakes.
124

Figure 8.4 shows the result of PL lifetime measurements taken from treated and untreated
MoS2 flakes with the same thickness (i.e., number of layers). Here, the PL lifetime decreases after
oxygen plasma treatment and laser exposure. The time-dependent PL intensities were fitted to bi-
exponential decays, as indicated in the Figure. The first decay decreases from 494 ps before
treatment to 190 ps after treatment. Here, the plasma-only treatment shows almost no effect on the
PL lifetime. We believe that the formation of small triangular regions of monolayer MoS 2 is
primarily responsible for the shortened lifetimes due to the increased surface-to-volume ratio (or
more accurately, the perimeter-to-area ratio). That is, non-radiative recombination occurring at the
edges of each MoS2 grain is primarily responsible for the decrease in the PL lifetime.
125

0 2 4 6 8 10 12
100
1000
Intensity (Counts)
t (ns)
Decay
Fitting
0 2 4 6 8 10 12
100
1000
Intensity (Counts)
t (ns)
Decay
Fitting
0 2 4 6 8 10 12
100
1000
Intensity (Counts)
t (ns)
Decay
Fitting
τ
1
=3513ps
τ
2
=494ps
Untreated
Plasma treated
(plasma only)
Plasma treated
and laser exposed
τ
1
=3741ps
τ
2
=491ps
τ
1
=2524ps
τ
2
=190ps
(a)
(b)
(c)
Figure 8.4 Time-resolved photoluminescence intensity for (a) untreated, (b) oxygen plasma treated only, and
(c) oxygen plasma treated and laser exposed MoS 2 flakes.
126

Figure 8.5 shows optical microscope and AFM images of a few-layer MoS2 flake after plasma
treatment and laser exposure, in which the letters “USC” were raster scanned on a bilayer portion.
Figure 8.5(b) shows an AFM image of a localized region in which the ~1 µm focused laser spot of
the flake was exposed for 60 seconds. A line scan of the surface topography is shown in Figure
8.5(c), showing an increase in thickness of 2 nm in the region that had undergone laser exposure.
This increase in thickness is consistent with the formation of MoO3, which has a larger thickness
than MoS2
148, 149
. The “USC” region did not get burned or damaged from the laser exposure and
exhibit a significantly enhanced photoluminescence than the surrounding bilayer regions. This
increase in thickness is somewhat surprising since it occurs in regions that are generally lighter
than the surrounding unexposed region. This increased transparency is consistent with the
transition to MoO3, which has a relatively wide band gap.
127

Figure 8.5 (a) Optical microscope and (b) AFM image of an MoS 2 flake deposited on a silicon nitride substrate.
(c) A line scan of the surface topography along the dashed line indicated in (b).
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
-2
-1
0
1
2
3
Height (nm)
Position ( mm)
5µm
Unexposed
2nm
increase in
thickness
Exposed
Height (nm)
-3nm 3.3nm
(a)
(b)
(c)
Exposed area
Line scan
1µm
128

Figure 8.6 Optical images and PL spectra taken from another suspended MoS 2 flake of (a) before treatment and
(b) after oxygen plasma treatment and thermal annealing in Argon.

8.5 Conclusion

In conclusion, we report a two-step method for creating “triangular” MoS2 nano-islands
from exfoliated multilayer MoS2 flakes using a sequence of plasma and light exposure treatments
as observed in TEM images. This plasma/laser treatment induces a transition from indirect to direct
band gap material, yielding a 7-fold enhancement in the direct band gap emission (at 1.87 eV) and
suppression of the indirect emission (around 1.54 eV). Here, a slight spatial decoupling of the
layers is created by intercalated O-species from the plasma treatment. The laser subsequently
oxidizes exposed regions of the MoS2 to MoO3, as confirmed via X-ray photoelectron
1.4 1.6 1.8 2.0
0
50
100
150
200
PL Intensity (Counts)
Energy (eV)
1.4 1.6 1.8 2.0
0
1000
2000
3000
4000
5000
6000
7000
PL Intensity (Counts)
Energy (eV)
Before Oxygen Plasma After plasma and annealing
(a) (b)
Si Substrate
Si Substrate
Silicon Nitride
Membrane
10µm 10µm
Multilayer MoS
2
Flake
Suspended
Region
Si Substrate
Si Substrate
Silicon Nitride
Membrane
Multilayer MoS
2
Flake
Suspended
Region
129

spectroscopy (XPS), which results in localized areas with brightly luminescent monolayer MoS2
triangular nano-islands embedded in bilayer MoS2. A 40meV blueshift is also observed in the
direct-gap emission with a 3 minute laser exposure with a 33 meV decrease in linewidth.
Furthermore, we observe a decrease in the PL lifetime from 494 psec to 190 psec after the plasma
and laser treatment, suggesting an increased density of monolayer MoS 2 grain boundaries. An
increase in flake thickness is observed in the laser-irradiated regions, providing more evidence that
the MoS2 is partially being converted into MoO3.
130

Chapter 9 Future Work
9.1 Single Photon Emission via LED-Like Light Emission from Dual-Gate Carbon
Nanotube Field Effect Transistors

In 2008, Alexander Hogele et al. reported nonclassical light emission at low temperatures
from carbon nanotubes via antibunching behavior in the photoluminescence of a suspended carbon
nanotube
150
. This was the first report of quantum correlations of light emission from a single
carbon nanotube and single photon emission. Their study indicates that Auger recombination (as
evidenced by a fast luminescence decay component under strong optical excitation) is at least
partially responsible for single-photon generation. The presence of exciton localization at low
temperatures ensures that nanotubes emit photons predominantly one by one. Most recently, He et
al. showed that by introducing aryl sp
3
defects in single-walled CNTs through a solution-based
doping method, they can achieve single photon emission originated from the redshifted defect
states (localized excitons) with a low probability of multiphoton emission at room temperature
42
,
as shown in figure 9.1.
131

Figure 9.1 (a) Photoluminescence spectra from the sp
3
defect state and (b) photon antibunching property of a
carbon nanotube at room temperature
42
.

The next step of our light emission project from dual-gate carbon nanotube field effect
transistor is to map out the electrical conditions for LED-like light emission. Our previous study
shows that under a large constant bias voltage, charge carriers induced by large gate voltages will
significantly quench the light emission efficiency through Auger recombination. As such, the
method we are attempting recently is to lower the applied gate voltages as well as the bias voltage
to satisfy the minimal light detection limit of our image system. Once we was able to observe light
emission under low electrical power (i.e., less than 200nW), we will try to demonstrate we could
achieve LED-like light emission via this approach.
(a)
(b)
g
2
(0)=0.01
132

Bibliography
1. Iijima, S., Helical Microtubules of Graphitic Carbon. Nature, 354, 56-58 (1991).
2. Dürkop, T., S.A. Getty, E. Cobas and M.S. Fuhrer, Extraordinary Mobility in
Semiconducting Carbon Nanotubes. Nano Letters, 4, 35-39 (2004).
3. Berber, S., Y .-K. Kwon and D. Tománek, Unusually High Thermal Conductivity of Carbon
Nanotubes. Physical Review Letters, 84, 4613-4616 (2000).
4. Yu, M.-F., B.S. Files, S. Arepalli and R.S. Ruoff, Tensile Loading of Ropes of Single Wall
Carbon Nanotubes and Their Mechanical Properties. Physical Review Letters, 84, 5552-
5555 (2000).
5. Jorio, A., R. Saito, J.H. Hafner, C.M. Lieber, M. Hunter, T. McClure, G. Dresselhaus and
M.S. Dresselhaus, Structural Mos2 Determination of Isolated Single-Wall Carbon
Nanotubes by Resonant Raman Scattering. Physical Review Letters, 86, 1118-1121 (2001).
6. Graupner, R., Raman Spectroscopy of Covalently Functionalized Single-Wall Carbon
Nanotubes. Journal of Raman Spectroscopy, 38, 673-683 (2007).
7. Tans, S.J., A.R.M. Verschueren and C. Dekker, Room-Temperature Transistor Based on a
Single Carbon Nanotube. Nature, 393, 49-52 (1998).
8. O'Connell, M.J., S.M. Bachilo, C.B. Huffman, V .C. Moore, M.S. Strano, E.H. Haroz, K.L.
Rialon, P.J. Boul, W.H. Noon, C. Kittrell, J.P. Ma, R.H. Hauge, R.B. Weisman, and R.E.
Smalley, Band Gap Fluorescence from Individual Single-Walled Carbon Nanotubes.
Science, 297, 593-596 (2002).
9. Hagen, A., M. Steiner, M.B. Raschke, C. Lienau, T. Hertel, H. Qian, A.J. Meixner and A.
Hartschuh, Exponential Decay Lifetimes of Excitons in Individual Single-Walled Carbon
Nanotubes. Physical Review Letters, 95, 197401 (2005).
10. Bachilo, S.M., M.S. Strano, C. Kittrell, R.H. Hauge, R.E. Smalley and R.B. Weisman,
Structure-Assigned Optical Spectra of Single-Walled Carbon Nanotubes. Science, 298,
2361 (2002).
11. Mak, K.F., C. Lee, J. Hone, J. Shan and T.F. Heinz, Atomically Thin Mos2: A New Direct-
Gap Semiconductor. Physical Review Letters, 105, 136805 (2010).
12. Splendiani, A., L. Sun, Y . Zhang, T. Li, J. Kim, C.-Y. Chim, G. Galli and F. Wang, Emerging
Photoluminescence in Monolayer Mos2. Nano Letters, 10, 1271-1275 (2010).
13. Mueller, T., M. Kinoshita, M. Steiner, V . Perebeinos, A.A. Bol, D.B. Farmer and P. Avouris,
Efficient Narrow-Band Light Emission from a Single Carbon Nanotube P-N Diode. Nature
Nanotechnology, 5, 27-31 (2010).
133

14. Misewich, J.A., R. Martel, P. Avouris, J.C. Tsang, S. Heinze and J. Tersoff, Electrically
Induced Optical Emission from a Carbon Nanotube Fet. Science, 300, 783-786 (2003).
15. Chen, J., V . Perebeinos, M. Freitag, J. Tsang, Q. Fu, J. Liu and P. Avouris, Bright Infrared
Emission from Electrically Induced Excitons in Carbon Nanotubes. Science, 310, 1171-
1174 (2005).
16. Wang, B., F. Rezaeifar, J.H. Chen, S. Yang, R.S. Kapadia and S.B. Cronin, Avalanche
Photoemission in Suspended Carbon Nanotubes: Light without Heat. ACS Photonics, 4,
2706-2710 (2017).
17. Liu, Z., A. Bushmaker, M. Aykol and S.B. Cronin, Thermal Emission Spectra from
Individual Suspended Carbon Nanotubes. ACS Nano, 5, 4634-4640 (2011).
18. Higashide, N., M. Yoshida, T. Uda, A. Ishii and Y .K. Kato, Cold Exciton
Electroluminescence from Air-Suspended Carbon Nanotube Split-Gate Devices. Applied
Physics Letters, 110, 191101 (2017).
19. Matsunaga, R., K. Matsuda and Y . Kanemitsu, Observation of Charged Excitons in Hole-
Doped Carbon Nanotubes Using Photoluminescence and Absorption Spectroscopy.
Physical Review Letters, 106, 037404 (2011).
20. Yuma, B., S. Berciaud, J. Besbas, J. Shaver, S. Santos, S. Ghosh, R.B. Weisman, L. Cognet,
M. Gallart, M. Ziegler, B. Honerlage, B. Lounis, and P. Gilliot, Biexciton, Single Carrier,
and Trion Generation Dynamics in Single-Walled Carbon Nanotubes. Physical Review B,
87, 205412 (2013).
21. Hogele, A., C. Galland, M. Winger and A. Imamoglu, Photon Antibunching in the
Photoluminescence Spectra of a Single Carbon Nanotube. Physical Review Letters, 100,
217401 (2008).
22. Ma, X.D., N.F. Hartmann, J.K.S. Baldwin, S.K. Doorn and H. Htoon, Room-Temperature
Single-Photon Generation from Solitary Dopants of Carbon Nanotubes. Nature
Nanotechnology, 10, 671-675 (2015).
23. Adam, E., C.M. Aguirre, L. Marty, B.C. St-Antoine, F. Meunier, P. Desjardins, D. Menard
and R. Martel, Electroluminescence from Single-Wall Carbon Nanotube Network
Transistors. Nano Letters, 8, 2351-2355 (2008).
24. Zaumseil, J., X.N. Ho, J.R. Guest, G.P. Wiederrecht and J.A. Rogers, Electroluminescence
from Electrolyte-Gated Carbon Nanotube Field-Effect Transistors. Acs Nano, 3, 2225-
2234 (2009).
25. Pfeiffer, M.H.P., N. Sturzl, C.W. Marquardt, M. Engel, S. Dehm, F. Hennrich, M.M.
Kappes, U. Lemmer and R. Krupke, Electroluminescence from Chirality-Sorted (9,7)-
Semiconducting Carbon Nanotube Devices. Optics Express, 19, A1184-A1189 (2011).
134

26. Freitag, M., V . Perebeinos, J. Chen, A. Stein, J.C. Tsang, J.A. Misewich, R. Martel and P.
Avouris, Hot Carrier Electroluminescence from a Single Carbon Nanotube. Nano Letters,
4, 1063-1066 (2004).
27. Lefebvre, J., J.M. Fraser, P. Finnie and Y . Homma, Photoluminescence from an Individual
Single-Walled Carbon Nanotube. Physical Review B, 69 (2004).
28. Lefebvre, J., D.G. Austing, J. Bond and P. Finnie, Photoluminescence Imaging of
Suspended Single-Walled Carbon Nanotubes. Nano Letters, 6, 1603-1608 (2006).
29. Tsyboulski, D.A., J.D.R. Rocha, S.M. Bachilo, L. Cognet and R.B. Weisman, Structure-
Dependent Fluorescence Efficiencies of Individual Single-Walled Cardon Nanotubes.
Nano Letters, 7, 3080-3085 (2007).
30. Bushmaker, A.W., V .V . Deshpande, M.W. Bockrath and S.B. Cronin, Direct Observation
of Mode Selective Electron - Phonon Coupling in Suspended Carbon Nanotubes. Nano
Letters, 7, 3618-3622 (2007).
31. Hsu, I.K., M.T. Pettes, M. Aykol, L. Shi and S.B. Cronin, The Effect of Gas Environment
on Electrical Heating in Suspended Carbon Nanotubes. Journal of Applied Physics, 108,
084307 (2010).
32. Amer, M., A. Bushmaker and S. Cronin, Anomalous Kink Behavior in the Current-Voltage
Characteristics of Suspended Carbon Nanotubes. Nano Research, 5, 172-180 (2012).
33. Liu, Z.W., A. Bushmaker, M. Aykol and S.B. Cronin, Thermal Emission Spectra from
Individual Suspended Carbon Nanotubes. ACS Nano, 5, 4634-4640 (2011).
34. Bushmaker, A.W., V .V . Deshpande, S. Hsieh, M.W. Bockrath and S.B. Cronin, Direct
Observation of Born- Oppenheimer Approximation Breakdown in Carbon Nanotubes.
Nano letters, 9, 607-611 (2009).
35. Bushmaker, A.W., V .V . Deshpande, S. Hsieh, M.W. Bockrath and S.B. Cronin, Large
Modulations in the Intensity of Raman-Scattered Light from Pristine Carbon Nanotubes.
Physical Review Letters, 103, 67401 (2009).
36. Chang, S.W., J. Theiss, J. Hazra, M. Aykol, R. Kapadia and S.B. Cronin, Photocurrent
Spectroscopy of Exciton and Free Particle Optical Transitions in Suspended Carbon
Nanotube Pn-Junctions. Applied Physics Letters, 107 (2015).
37. Bushmaker, A.W., V .V . Deshpande, S. Hsieh, M.W. Bockrath and S.B. Cronin, Direct
Observation of Born-Oppenheimer Approximation Breakdown in Carbon Nanotubes. Nano
Letters, 9, 607-611 (2009).
38. Deshpande, V .V ., B. Chandra, R. Caldwell, D.S. Novikov, J. Hone and M. Bockrath, Mott
Insulating State in Ultraclean Carbon Nanotubes. Science, 323, 106-110 (2009).
39. Spataru, C.D., S. Ismail-Beigi, L.X. Benedict and S.G. Louie, Excitonic Effects and Optical
135

Spectra of Single-Walled Carbon Nanotubes. Physical Review Letters, 92 (2004).
40. Pop, E., D. Mann, J. Cao, Q. Wang, K.E. Goodson and H.J. Dai, Negative Differential
Conductance and Hot Phonons in Suspended Nanotube Molecular Wires. Physical Review
Letters, 95, 155505 (2005).
41. Hsu, I.K., R. Kumar, A. Bushmaker, S.B. Cronin, M.T. Pettes, L. Shi, T. Brintlinger, M.S.
Fuhrer and J. Cumings, Optical Measurement of Thermal Transport in Suspended Carbon
Nanotubes. Applied Physics Letters, 92 (2008).
42. He, X.W., N.F. Hartmann, X.D. Ma, Y . Kim, R. Ihly, J.L. Blackburn, W.L. Gao, J. Kono,
Y . Yomogida, A. Hirano, T. Tanaka, H. Kataura, H. Htoon, and S.K. Doorn, Tunable Room-
Temperature Single-Photon Emission at Telecom Wavelengths from Sp(3) Defects in
Carbon Nanotubes. Nature Photonics, 11, 577 (2017).
43. Khasminskaya, S., F. Pyatkov, K. Slowik, S. Ferrari, O. Kahl, V . Kovalyuk, P. Rath, A.
Vetter, F. Hennrich, M.M. Kappes, G. Gol'tsman, A. Korneev, C. Rockstuhl, R. Krupke,
and W.H.P. Pernice, Fully Integrated Quantum Photonic Circuit with an Electrically
Driven Light Source. Nature Photonics, 10, 727 (2016).
44. Mann, D., Y .K. Kato, A. Kinkhabwala, E. Pop, J. Cao, X.R. Wang, L. Zhang, Q. Wang, J.
Guo and H.J. Dai, Electrically Driven Thermal Light Emission from Individual Single-
Walled Carbon Nanotubes. Nature Nanotechnology, 2, 33-38 (2007).
45. Xie, L.M., H. Farhat, H.B. Son, J. Zhang, M.S. Dresselhaus, J. Kong and Z.F. Liu,
Electroluminescence from Suspended and on-Substrate Metallic Single-Walled Carbon
Nanotubes. Nano Letters, 9, 1747-1751 (2009).
46. Wang, S., Q.S. Zeng, L.J. Yang, Z.Y . Zhang, Z.X. Wang, T.A. Pei, L. Ding, X.L. Liang, M.
Gao, Y . Li and L.M. Peng, High-Performance Carbon Nanotube Light-Emitting Diodes
with Asymmetric Contacts. Nano Letters, 11, 23-29 (2011).
47. Xie, X., A.E. Islam, M.A. Wahab, L.N. Ye, X.N. Ho, M.A. Alam and J.A. Rogers,
Electroluminescence in Aligned Arrays of Single-Wall Carbon Nanotubes with Asymmetric
Contacts. Acs Nano, 6, 7981-7988 (2012).
48. Jakubka, F., S.B. Grimm, Y . Zakharko, F. Gannott and J. Zaumseil, Trion
Electroluminescence from Semiconducting Carbon Nanotubes. ACS Nano, 8, 8477-8486
(2014).
49. Higashide, N., M. Yoshida, T. Uda, A. Ishii and Y .K. Kato, Cold Exciton
Electroluminescence from Air-Suspended Carbon Nanotube Split-Gate Devices. Applied
Physics Letters, 110 (2017).
50. Deshpande, V .V ., S. Hsieh, A.W. Bushmaker, M. Bockrath and S.B. Cronin, Spatially
Resolved Temperature Measurements of Electrically Heated Carbon Nanotubes. Physical
Review Letters, 102 (2009).
136

51. Pop, E., D. Mann, J. Cao, Q. Wang, K.E. Goodson and H.J. Dai, Negative Differential
Conductance and Hot Phonons in Suspended Nanotube Molecular Wires. Physical Review
Letters, 95 (2005).
52. Bushmaker, A.W., V .V . Deshpande, S. Hsieh, M.W. Bockrath and S.B. Cronin, Gate
Voltage Controllable Non-Equilibrium and Non-Ohmic Behavior in Suspended Carbon
Nanotubes. Nano Letters, 9, 2862-2866 (2009).
53. Bushmaker, A.W., M.R. Amer and S.B. Cronin, Electrical Transport and Channel Length
Modulation in Semiconducting Carbon Nanotube Field Effect Transistors. IEEE
Transactions on Nanotechnology, 13, 176-181 (2014).
54. Ghosh, S., S.M. Bachilo, R.A. Simonette, K.M. Beckingham and R.B. Weisman, Oxygen
Doping Modifies near-Infrared Band Gaps in Fluorescent Single-Walled Carbon
Nanotubes. Science, 330, 1656-1659 (2010).
55. Miyauchi, Y ., M. Iwamura, S. Mouri, T. Kawazoe, M. Ohtsu and K. Matsuda, Brightening
of Excitons in Carbon Nanotubes on Dimensionality Modification. Nature Photonics, 7,
715-719 (2013).
56. Ma, X., L. Adamska, H. Yamaguchi, S.E. Yalcin, S. Tretiak, S.K. Doorn and H. Htoon,
Electronic Structure and Chemical Nature of Oxygen Dopant States in Carbon Nanotubes.
ACS Nano, 8, 10782-10789 (2014).
57. Ma, X.D., J.K.S. Baldwin, N.F. Hartmann, S.K. Doorn and H. Htoon, Solid-State Approach
for Fabrication of Photostable, Oxygen-Doped Carbon Nanotubes. Advanced Functional
Materials, 25, 6157-6164 (2015).
58. Ma, X.D., A.R. James, N.F. Hartmann, J.K. Baldwin, J. Dominguez, M.B. Sinclair, T.S.
Luk, O. Wolf, S. Liu, S.K. Doorn, H. Htoon and I. Brener, Solitary Oxygen Dopant
Emission from Carbon Nanotubes Modified by Dielectric Metasurfaces. Acs Nano, 11,
6431-6439 (2017).
59. Ju, S.-Y ., W.P. Kopcha and F. Papadimitrakopoulos, Brightly Fluorescent Single-Walled
Carbon Nanotubes Via an Oxygen-Excluding Surfactant Organization. Science, 323, 1319
(2009).
60. Ishii, A., T. Uda and Y .K. Kato, Room-Temperature Single-Photon Emission from
Micrometer-Long Air-Suspended Carbon Nanotubes. Physical Review Applied, 8, 54039
(2017).
61. Hofmann, M.S., J.T. Gluckert, J. Noe, C. Bourjau, R. Dehmel and A. Hogele, Bright, Long-
Lived and Coherent Excitons in Carbon Nanotube Quantum Dots. Nature Nanotechnology,
8, 502-505 (2013).
62. Walden-Newman, W., I. Sarpkaya and S. Strauf, Quantum Light Signatures and
Nanosecond Spectral Diffusion from Cavity-Embedded Carbon Nanotubes. Nano Letters,
137

12, 1934-1941.
63. Wang, B., S. Yang, L. Shen and S.B. Cronin, Ultra-Low Power Light Emission Via
Avalanche and Sub-Avalanche Breakdown in Suspended Carbon Nanotubes. ACS
Photonics, DOI: 10.1021/acsphotonics.8b00896 (2018).
64. Wang, B., F. Rezaeifar, J. Chen, S. Yang, R. Kapadia and S.B. Cronin, Avalanche
Photoemission in Suspended Carbon Nanotubes: Light without Heat. ACS Photonics, 4,
2706-2710 (2017).
65. Chang, S., K. Bergemann, R. Dhall, J. Zimmerman, S. Forrest and S.B. Cronin, Nonideal
Diode Behavior and Bandgap Renormalization in Carbon Nanotube P-N Junctions. IEEE
Transactions on Nanotechnology, 13, 41-45 (2014).
66. Freitag, M., M. Steiner, A. Naumov, J.P. Small, A.A. Bol, V . Perebeinos and P. Avouris,
Carbon Nanotube Photo- and Electroluminescence in Longitudinal Electric Fields. ACS
Nano, 3, 3744-3748 (2009).
67. Steiner, M., M. Freitag, V . Perebeinos, A. Naumov, J.P. Small, A.A. Bol and P. Avouris,
Gate-Variable Light Absorption and Emission in a Semiconducting Carbon Nanotube.
Nano Letters, 9, 3477-3481 (2009).
68. Yasukochi, S., T. Murai, S. Moritsubo, T. Shimada, S. Chiashi, S. Maruyama and Y .K. Kato,
Gate-Induced Blueshift and Quenching of Photoluminescence in Suspended Single-Walled
Carbon Nanotubes. Physical Review B, 84, 121409 (2011).
69. Yoshida, M., A. Popert and Y .K. Kato, Gate-Voltage Induced Trions in Suspended Carbon
Nanotubes. Physical Review B, 93, 041402 (2016).
70. Chynoweth, A.G. and K.G. McKay, Photon Emission from Avalanche Breakdown in
Silicon. Physical Review, 102, 369-376 (1956).
71. van Drieënhuizen, B.P. and R.F. Wolffenbuttel, Optocoupler Based on the Avalanche Light
Emission in Silicon. Sensors and Actuators A: Physical, 31, 229-240 (1992).
72. He, X.W., N.F. Hartmann, X.D. Ma, Y . Kim, R. Ihly, J.L. Blackburn, W.L. Gao, J. Kono,
Y . Yomogida, A. Hirano, T. Tanaka, H. Kataura, H. Htoon, and S.K. Doorn, Tunable Room-
Temperature Single-Photon Emission at Telecom Wavelengths from Sp(3) Defects in
Carbon Nanotubes. Nature Photonics, 11, 577-582 (2017).
73. Lee, K.G., X.W. Chen, H. Eghlidi, P. Kukura, R. Lettow, A. Renn, V . Sandoghdar and S.
Götzinger, A Planar Dielectric Antenna for Directional Single-Photon Emission and near-
Unity Collection Efficiency. Nature Photonics, 5, 166-169 (2011).
74. Chen, J., V . Perebeinos, M. Freitag, J. Tsang, Q. Fu, J. Liu and P. Avouris, Bright Infrared
Emission from Electrically Induced Excitons in Carbon Canotubes. Science, 310, 1171-
1174 (2005).
138

75. Bushmaker, A.W., V .V . Deshpande, M.W. Bockrath and S.B. Cronin, Direct Observation
of Mode Selective Electron-Phonon Coupling in Suspended Carbon Nanotubes. Nano
Letters, 7, 3618-3622 (2007).
76. Xu, K., L. Huang, Z. Zhang, J. Zhao, Z. Zhang, L. Snyman and J. Swart, Light Emission
from a Poly-Silicon Device with Carrier Injection Engineering. Materials Science and
Engineering: B, 231, 28-31 (2018).
77. Xu, K., Y . Chen, T.A. Okhai and L.W. Snyman, Micro Optical Sensors Based on
Avalanching Silicon Light-Emitting Devices Monolithically Integrated on Chips. Optical
Materials Express, 9, 3985-3997 (2019).
78. Xu, K., Silicon Mos Optoelectronic Micro-Nano Structure Based on Reverse-Biased Pn
Junction. Physica Status Solidi (a), 216, 1800868 (2019).
79. Kuai, S. and A. Meldrum, High-Speed Color-Switching Silicon Leds. Advanced Materials,
20, 3844-3847 (2008).
80. Khasminskaya, S., F. Pyatkov, B.S. Flavel, W.H. Pernice and R. Krupke, Waveguide-
Integrated Light-Emitting Carbon Nanotubes. Advanced Materials, 26, 3465-3472 (2014).
81. Yu, D., S. Wang, L. Ye, W. Li, Z. Zhang, Y . Chen, J. Zhang and L.-M. Peng,
Electroluminescence from Serpentine Carbon Nanotube Based Light-Emitting Diodes on
Quartz. Small, 10, 1050-1056 (2014).
82. Graf, A., C. Murawski, Y . Zakharko, J. Zaumseil and M.C. Gather, Infrared Organic Light-
Emitting Diodes with Carbon Nanotube Emitters. Advanced Materials, 30, 1706711 (2018).
83. Wang, R., H. Wu, R. Chen and Y . Chi, Strong Electrochemiluminescence Emission from
Oxidized Multiwalled Carbon Nanotubes. Small, 0, 1901550 (2019).
84. Wang, B., S. Yang, L. Shen and S.B. Cronin, Ultra-Low Power Light Emission Via
Avalanche and Sub-Avalanche Breakdown in Suspended Carbon Nanotubes. ACS
Photonics, 5, 4432-4436 (2018).
85. Banerjee, S. and S.S. Wong, Synthesis and Characterization of Carbon
Nanotube−Nanocrystal Heterostructures. Nano Letters, 2, 195-200 (2002).
86. Chang, S.W., J. Theiss, J. Hazra, M. Aykol, R. Kapadia and S.B. Cronin, Photocurrent
Spectroscopy of Exciton and Free Particle Optical Transitions in Suspended Carbon
Nanotube Pn-Junctions. Applied Physics Letters, 107, 053107 (2015).
87. Hsu, I.K., R. Kumar, A. Bushmaker, S.B. Cronin, M.T. Pettes, L. Shi, T. Brintlinger, M.S.
Fuhrer and J. Cumings, Optical Measurement of Thermal Transport in Suspended Carbon
Nanotubes. Applied Physics Letters, 92, 063119 (2008).
88. Sentaurus, T. C. A. D. "Version L-2016." Synopsys Inc., Mountain View, Ca, USA. (2016).
139

89. Piao, Y .M., B. Meany, L.R. Powell, N. Valley, H. Kwon, G.C. Schatz and Y .H. Wang,
Brightening of Carbon Nanotube Photoluminescence through the Incorporation of Sp(3)
Defects. Nature Chemistry, 5, 840-845 (2013).
90. Sarpkaya, I., Z.Y . Zhang, W. Walden-Newman, X.S. Wang, J. Hone, C.W. Wong and S.
Strauf, Prolonged Spontaneous Emission and Dephasing of Localized Excitons in Air-
Bridged Carbon Nanotubes. Nature Communications, 4 (2013).
91. Vialla, F., Y . Chassagneux, R. Ferreira, C. Roquelet, C. Diederichs, G. Cassabois, P.
Roussignol, J.S. Lauret and C. V oisin, Unifying the Low-Temperature Photoluminescence
Spectra of Carbon Nanotubes: The Role of Acoustic Phonon Confinement. Physical
Review Letters, 113 (2014).
92. Nan, H.Y ., Z.L. Wang, W.H. Wang, Z. Liang, Y . Lu, Q. Chen, D.W. He, P.H. Tan, F. Miao,
X.R. Wang, J.L. Wang and Z.H. Ni, Strong Photoluminescence Enhancement of Mos2
through Defect Engineering and Oxygen Bonding. Acs Nano, 8, 5738-5745 (2014).
93. Tongay, S., J. Suh, C. Ataca, W. Fan, A. Luce, J.S. Kang, J. Liu, C. Ko, R. Raghunathanan,
J. Zhou, F. Ogletree, J.B. Li, J.C. Grossman, and J.Q. Wu, Defects Activated
Photoluminescence in Two-Dimensional Semiconductors: Interplay between Bound,
Charged, and Free Excitons. Scientific Reports, 3 (2013).
94. Harutyunyan, H., T. Gokus, A.A. Green, M.C. Hersam, M. Allegrini and A. Hartschuh,
Defect-Induced Photoluminescence from Dark Excitonic States in Individual Single-Walled
Carbon Nanotubes. Nano Letters, 9, 2010-2014 (2009).
95. He, Y .M., G. Clark, J.R. Schaibley, Y . He, M.C. Chen, Y .J. Wei, X. Ding, Q. Zhang, W. Yao,
X.D. Xu, C.Y . Lu and J.W. Pan, Single Quantum Emitters in Monolayer Semiconductors.
Nature Nanotechnology, 10, 497-502 (2015).
96. Chakraborty, C., L. Kinnischtzke, K.M. Goodfellow, R. Beams and A.N. Vamivakas,
Voltage-Controlled Quantum Light from an Atomically Thin Semiconductor. Nature
Nanotechnology, 10, 507-U38 (2015).
97. Srivastava, A., M. Sidler, A.V . Allain, D.S. Lembke, A. Kis and A. Imamoglu, Optically
Active Quantum Dots in Monolayer Wse2. Nature Nanotechnology, 10, 491-496 (2015).
98. Koperski, M., K. Nogajewski, A. Arora, V . Cherkez, P. Mallet, J.Y . Veuillen, J. Marcus, P.
Kossacki and M. Potemski, Single Photon Emitters in Exfoliated Wse2 Structures. Nature
Nanotechnology, 10, 503-506 (2015).
99. Marty, L., E. Adam, L. Albert, R. Doyon, D. Menard and R. Martel, Exciton Formation
and Annihilation During 1d Impact Excitation of Carbon Nanotubes. Physical Review
Letters, 96 (2006).
100. Dresselhaus, M.S., A. Jorio, A.G. Souza and R. Saito, Defect Characterization in Graphene
and Carbon Nanotubes Using Raman Spectroscopy. Philosophical Transactions of the
140

Royal Society a-Mathematical Physical and Engineering Sciences, 368, 5355-5377 (2010).
101. Dresselhaus, M.S., G. Dresselhaus, R. Saito and A. Jorio, Raman Spectroscopy of Carbon
Nanotubes. Physics Reports-Review Section of Physics Letters, 409, 47-99 (2005).
102. Chang, C.C., C.C. Chen, W.H. Hung, I.K. Hsu, M.A. Pimenta and S.B. Cronin, Strain-
Induced D Band Observed in Carbon Nanotubes. Nano Research, 5, 854-862 (2012).
103. Amer, M.R., S.W. Chang and S.B. Cronin, Competing Photocurrent Mechanisms in Quasi-
Metallic Carbon Nanotube Pn Devices. Small, 11, 3119-3123 (2015).
104. Knoll, G.F., Radiation Detection and Measurement. 4. ed. 2010, Hoboken, N.J.: Wiley.
xxvi, 830 s.
105. Dhall, R., Z. Li, E. Kosmowska and S.B. Cronin, Charge Neutral Mos2 Field Effect
Transistors through Oxygen Plasma Treatment. Journal of Applied Physics, 120, 195702
(2016).
106. Ross, J.S., S.F. Wu, H.Y . Yu, N.J. Ghimire, A.M. Jones, G. Aivazian, J.Q. Yan, D.G.
Mandrus, D. Xiao, W. Yao and X.D. Xu, Electrical Control of Neutral and Charged
Excitons in a Monolayer Semiconductor. Nature Communications, 4, 1474 (2013).
107. Dhall, R., M.R. Neupane, D. Wickramaratne, M. Mecklenburg, Z. Li, C. Moore, R.K. Lake
and S. Cronin, Direct Band Gap Transition in Many-Layer Mos2 by Plasma Induced Layer
Decoupling Advanced Materials, DOI:10.1002/adma.201405259 (2015).
108. Kim, T.Y ., K. Cho, W. Park, J. Park, Y . Song, S. Hong, W.K. Hong and T. Lee, Irradiation
Effects of High-Energy Proton Beams on Mos2 Field Effect Transistors. Acs Nano, 8, 2774-
2781 (2014).
109. Zhang, C.X., A.K.M. Newaz, B. Wang, E.X. Zhang, G.X. Duan, D.M. Fleetwood, M.L.
Alles, R.D. Schrimpf, K.I. Bolotin and S.T. Pantelides, Electrical Stress and Total Ionizing
Dose Effects on Mos2 Transistors. IEEE Transactions on Nuclear Science, 61, 2862-2867
(2014).
110. Lee, J., M.J. Krupcale and P.X.L. Feng, Effects of Gamma-Ray Radiation on Two-
Dimensional Molybdenum Disulfide (Mos2) Nanomechanical Resonators. Applied Physics
Letters, 108 (2016).
111. Ochedowski, O., K. Marinov, G. Wilbs, G. Keller, N. Scheuschner, D. Severin, M. Bender,
J. Maultzsch, F.J. Tegude and M. Schleberger, Radiation Hardness of Graphene and Mos2
Field Effect Devices against Swift Heavy Ion Irradiation. Journal of Applied Physics, 113
(2013).
112. Tsai, D.S., D.H. Lien, M.L. Tsai, S.H. Su, K.M. Chen, J.J. Ke, Y .C. Yu, L.J. Li and J.H. He,
Trilayered Mos2 Metal-Semiconductor-Metal Photodetectors: Photogain and Radiation
Resistance. Ieee Journal of Selected Topics in Quantum Electronics, 20 (2014).
141

113. Zan, R., Q.M. Ramasse, R. Jalil, T. Georgiou, U. Bangert and K.S. Novoselov, Control of
Radiation Damage in Mos2 by Graphene Encapsulation. Acs Nano, 7, 10167-10174
(2013).
114. Fox, D.S., Y .B. Zhou, P. Maguire, A. O'Neill, C. O'Coileain, R. Gatensby, A.M. Glushenkov,
T. Tao, G.S. Duesberg, I.V . Shvets, M. Abid, M. Abid, H.C. Wu, Y . Chen, J.N. Coleman,
J.F. Donegan, and H.Z. Zhang, Nanopatterning and Electrical Tuning of Mos2 Layers with
a Subnanometer Helium Ion Beam. Nano Letters, 15, 5307-5313 (2015).
115. Ghorbani-Asl, M., S. Kretschmer, D.E. Spearot and A.V . Krasheninnikov, Two-
Dimensional Mos2 under Ion Irradiation: From Controlled Defect Production to
Electronic Structure Engineering. 2d Materials, 4 (2017).
116. Lin, Z., B.R. Carvalho, E. Kahn, R.T. Lv, R. Rao, H. Terrones, M.A. Pimenta and M.
Terrones, Defect Engineering of Two-Dimensional Transition Metal Dichalcogenides. 2d
Materials, 3 (2016).
117. Ma, Q., P.M. Odenthal, J. Mann, D. Le, C.S. Wang, Y .M. Zhu, T.Y . Chen, D.Z. Sun, K.
Yamaguchi, T. Tran, M. Wurch, J.L. McKinley, J. Wyrick, K. Magnone, T.F. Heinz, T.S.
Rahman, R. Kawakami, and L. Bartels, Controlled Argon Beam-Induced Desulfurization
of Monolayer Molybdenum Disulfide. Journal of Physics-Condensed Matter, 25 (2013).
118. Santosh, K.C., R.C. Longo, R. Addou, R.M. Wallace and K. Cho, Impact of Intrinsic
Atomic Defects on the Electronic Structure of Mos2 Monolayers. Nanotechnology, 25
(2014).
119. Scheuschner, N., O. Ochedowski, A.M. Kaulitz, R. Gillen, M. Schleberger and J.
Maultzsch, Photoluminescence of Freestanding Single- and Few-Layer Mos2. Physical
Review B, 89 (2014).
120. Yang, R., X.Q. Zheng, Z.H. Wang, C.J. Miller and P.X.L. Feng, Multilayer Mos2
Transistors Enabled by a Facile Dry-Transfer Technique and Thermal Annealing. Journal
of Vacuum Science & Technology B, 32 (2014).
121. Mak, K.F., C. Lee, J. Hone, J. Shan and T.F. Heinz, Atomically Thin Mos2: A New Direct-
Gap Semiconductor. Physical Review Letters, 105 (2010).
122. Park, J., N. Choudhary, J. Smith, G. Lee, M. Kim and W. Choi, Thickness Modulated Mos2
Grown by Chemical Vapor Deposition for Transparent and Flexible Electronic Devices.
Applied Physics Letters, 106 (2015).
123. Xu, X., G. Das, X. He, M.N. Hedhili, E.D. Fabrizio, X. Zhang and H.N. Alshareef, High-
Performance Monolayer Mos2 Films at the Wafer Scale by Two-Step Growth. Advanced
Functional Materials, 29, 1901070 (2019).
124. Meng, J., G. Wang, X. Li, X. Lu, J. Zhang, H. Yu, W. Chen, L. Du, M. Liao, J. Zhao, P.
Chen, J. Zhu, X. Bai, D. Shi, and G. Zhang, Rolling up a Monolayer Mos2 Sheet. Small,
142

12, 3770-3774 (2016).
125. Li, H., G. Lu, Y . Wang, Z. Yin, C. Cong, Q. He, L. Wang, F. Ding, T. Yu and H. Zhang,
Mechanical Exfoliation and Characterization of Single- and Few-Layer Nanosheets of
Wse2, Tas2, and Tase2. Small, 9, 1974-1981 (2013).
126. Gao, G., J. Yu, X. Yang, Y . Pang, J. Zhao, C. Pan, Q. Sun and Z.L. Wang, Triboiontronic
Transistor of Mos2. Advanced Materials, 31, 1806905 (2019).
127. Rezende, N.P., A.R. Cadore, A.C. Gadelha, C.L. Pereira, V . Ornelas, K. Watanabe, T.
Taniguchi, A.S. Ferlauto, A. Malachias, L.C. Campos and R.G. Lacerda, Probing the
Electronic Properties of Monolayer Mos2 Via Interaction with Molecular Hydrogen.
Advanced Electronic Materials, 5, 1800591 (2019).
128. Stergiou, A. and N. Tagmatarchis, Molecular Functionalization of Two-Dimensional Mos2
Nanosheets. Chemistry – A European Journal, 24, 18246-18257 (2018).
129. Amani, M., D.-H. Lien, D. Kiriya, J. Xiao, A. Azcatl, J. Noh, S.R. Madhvapathy, R. Addou,
S. Kc, M. Dubey, K. Cho, R.M. Wallace, S.-C. Lee, J.-H. He, J.W. Ager, X. Zhang, E.
Yablonovitch, and A. Javey, Near-Unity Photoluminescence Quantum Yield in
Mos≪Sub≫2≪/Sub&Gt. Science, 350, 1065 (2015).
130. Desai, S.B., S.R. Madhvapathy, A.B. Sachid, J.P. Llinas, Q. Wang, G.H. Ahn, G. Pitner,
M.J. Kim, J. Bokor, C. Hu, H.S.P. Wong and A. Javey, Mos2 Transistors with 1-Nanometer
Gate Lengths. Science, 354, 99 (2016).
131. Wu, S., S. Buckley, A.M. Jones, J.S. Ross, N.J. Ghimire, J. Yan, D.G. Mandrus, W. Yao, F.
Hatami, J. Vučković, A. Majumdar and X. Xu, Control of Two-Dimensional Excitonic Light
Emission Via Photonic Crystal. 2D Materials, 1, 011001 (2014).
132. Wu, S., S. Buckley, J. Schaibley, L. Feng, J.Q. Yan, D. G. Mandrus, F. Hatami, W. Yao, J.
Vuckovic, A. Majumdar and X. Xu, Ultra-Low Threshold Monolayer Semiconductor
Nanocavity Lasers. 2015.
133. Rivera, P., J.R. Schaibley, A.M. Jones, J.S. Ross, S. Wu, G. Aivazian, P. Klement, K. Seyler,
G. Clark, N.J. Ghimire, J. Yan, D.G. Mandrus, W. Yao, and X. Xu, Observation of Long-
Lived Interlayer Excitons in Monolayer Mose2–Wse2 Heterostructures. Nature
Communications, 6, 6242 (2015).
134. Nayak, P.K., Y . Horbatenko, S. Ahn, G. Kim, J.-U. Lee, K.Y . Ma, A.R. Jang, H. Lim, D.
Kim, S. Ryu, H. Cheong, N. Park, and H.S. Shin, Probing Evolution of Twist-Angle-
Dependent Interlayer Excitons in Mose2/Wse2 Van Der Waals Heterostructures. ACS
Nano, 11, 4041-4050 (2017).
135. Rigosi, A.F., H.M. Hill, Y . Li, A. Chernikov and T.F. Heinz, Probing Interlayer Interactions
in Transition Metal Dichalcogenide Heterostructures by Optical Spectroscopy: Mos2/Ws2
and Mose2/Wse2. Nano Letters, 15, 5033-5038 (2015).
143

136. Hanbicki, A.T., H.-J. Chuang, M.R. Rosenberger, C.S. Hellberg, S.V . Sivaram, K.M.
McCreary, I.I. Mazin and B.T. Jonker, Double Indirect Interlayer Exciton in a Mose2/Wse2
Van Der Waals Heterostructure. ACS Nano, 12, 4719-4726 (2018).
137. Chen, J., C.S. Bailey, Y . Hong, L. Wang, Z. Cai, L. Shen, B. Hou, Y . Wang, H. Shi, J.
Sambur, W. Ren, E. Pop, and S.B. Cronin, Plasmon-Resonant Enhancement of
Photocatalysis on Monolayer Wse2. ACS Photonics, 6, 787-792 (2019).
138. Wang, B., S. Yang, J. Chen, C. Mann, A. Bushmaker and S.B. Cronin, Radiation-Induced
Direct Bandgap Transition in Few-Layer Mos2. Applied Physics Letters, 111, 131101
(2017).
139. Foran, B., C. Mann, M. Peterson, A. Bushmaker, B. Wang, J. Chen, S. Yang and S.B.
Cronin, Effects of Proton Radiation-Induced Defects on Optoelectronic Properties of Mos2.
IEEE Transactions on Nuclear Science, 66, 413-419 (2019).
140. Sun, L., H. Hu, D. Zhan, J. Yan, L. Liu, J.S. Teguh, E.K.L. Yeow, P.S. Lee and Z. Shen,
Plasma Modified Mos2 Nanoflakes for Surface Enhanced Raman Scattering. Small, 10,
1090-1095 (2014).
141. Mouri, S., Y . Miyauchi and K. Matsuda, Tunable Photoluminescence of Monolayer Mos2
Via Chemical Doping. Nano Letters, 13, 5944-5948 (2013).
142. Nan, H., Z. Wang, W. Wang, Z. Liang, Y . Lu, Q. Chen, D. He, P. Tan, F. Miao, X. Wang, J.
Wang and Z. Ni, Strong Photoluminescence Enhancement of Mos2 through Defect
Engineering and Oxygen Bonding. ACS Nano, 8, 5738-5745 (2014).
143. Li, Z., S. Yang, R. Dhall, E. Kosmowska, H. Shi, I. Chatzakis and S.B. Cronin, Layer
Control of Wse2 Via Selective Surface Layer Oxidation. ACS Nano, 10, 6836-6842 (2016).
144. Ganta, D., S. Sinha and R.T. Haasch, 2-D Material Molybdenum Disulfide Analyzed by Xps.
Surface Science Spectra, 21, 19-27 (2014).
145. Benson, E.E., H. Zhang, S.A. Schuman, S.U. Nanayakkara, N.D. Bronstein, S. Ferrere, J.L.
Blackburn and E.M. Miller, Balancing the Hydrogen Evolution Reaction, Surface
Energetics, and Stability of Metallic Mos2 Nanosheets Via Covalent Functionalization.
Journal of the American Chemical Society, 140, 441-450 (2018).
146. Yang, J., S. Kim, W. Choi, S.H. Park, Y . Jung, M.-H. Cho and H. Kim, Improved Growth
Behavior of Atomic-Layer-Deposited High-K Dielectrics on Multilayer Mos2 by Oxygen
Plasma Pretreatment. ACS Applied Materials & Interfaces, 5, 4739-4744 (2013).
147. Khondaker, S.I. and M.R. Islam, Bandgap Engineering of Mos2 Flakes Via Oxygen Plasma:
A Layer Dependent Study. The Journal of Physical Chemistry C, 120, 13801-13806 (2016).
148. Ma, W., P. Alonso-González, S. Li, A. Nikitin, J. Yuan, J. Martín-Sánchez, J. Taboada
Gutiérrez, I. Amenabar, P. Li, S. Vélez, C. Tollan, Z. Dai, Y . Zhang, S. Sriram, K. Kalantar-
144

zadeh, S.-T. Lee, R. Hillenbrand, and Q. Bao, In-Plane Anisotropic and Ultra-Low-Loss
Polaritons in a Natural Van Der Waals Crystal. Nature, 562 (2018).
149. Huang, P., Y . He, C. Cao and Z.-H. Lu, Impact of Lattice Distortion and Electron Doping
on Α-Moo3 Electronic Structure. Scientific reports, 4, 7131 (2014).
150. Hogele, A., C. Galland, M. Winger and A. Imamoglu, Photon Antibunching in the
Photoluminescence Spectra of a Single Carbon Nanotube. Phys Rev Lett, 100, 217401
(2008).

Abstract (if available)

Abstract This dissertation work presents research projects related to electroluminescence and photoluminescence from low dimensional material devices, including carbon nanotube field effect transistors and two-dimensional material. These research projects can be helpful to understand the fundamental properties and potential optical, electrical and optoelectronic applications of these material and nanoscale devices. For future research, this dissertation can provide good reference and guidance in both theoretical and method aspects. ❧ Chapter 1 provides some background information that can be helpful to understand this dissertation work. It gives a brief overview of carbon nanotube including its atomic structure and optical & electrical properties, which can help understand its optical and optoelectronic applications. It is then followed by an introduction to carbon nanotube field effect transistors, including its electronic characteristics. Brief introductions to photoluminescence from low dimensional materials and electroluminescence from carbon nanotube field effect transistors are included in order to help understanding this dissertation work. ❧ Chapters 2-4 present several electroluminescence researches of suspended carbon nanotube field effect transistors. Chapter 2 reports avalanche light emission through reverse breakdown (impact ionization) process. The electrical-driven light emission occurs when the CNT FET is gated to its “off” state and remains unheated due to extremely low electric power density. Chapter 3 presents a follow-up research work focused on electrical power and relative light emission efficiency during avalanche and sub-avalanche light emission process and it exhibits a relative light emission efficiency with three orders of magnitude higher than all the other previous reports. Chapter 4 reports spectra studies of avalanche and sub-avalanche light emission process. The spectra of avalanche light emission exhibits a featureless flat spectrum covering a large wavelength range from 600-1600nm. ❧ Chapter 5 presents a project related to thermal emission from dual-gate CNT FETs. The thermal emission (incandescence) intensity can be tuned by the electrostatic doping method. The charge carriers induced by gate voltages can suppress incandescence intensity via a non-radiative process—Auger recombination. To the best of my knowledge this is the only system where the incandescence (i.e., thermal emission) intensity can be tuned by electrostatic doping while the electrical power is held constant. CNT is quite unique as a one-dimensional material. ❧ Chapter 6 presents a project of photoluminescence from suspended carbon nanotubes. It reveals a detailed comparison between photoluminescence (PL) intensity and Raman spectroscopy quantified defects in carbon nanotubes. ❧ Chapter 7 and 8 present projects related to photoluminescence from the most famous transition metal dichalcogenides MoS₂. Chapter 7 reports direct band gap transition induced by proton radiation. It provides a reliable method to achieve permanent direct band gap transition from multi-layer MoS₂ flakes and can be potentially useful for making optoelectronic devices such as LEDs and solar cells. Chapter 8 reports a two-step method to create “triangular” MoS₂ nano-islands via oxygen plasma followed by laser exposure, which can be potentially used as quantum dots and single photon emitters. ❧ Chapter 9 presents future work related to potential single photon emission via LED like light emission from dual-gate CNT FETs.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Optoelectronic, thermoelectric, and photocatalytic properties of low dimensional materials

PDF

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

PDF

Carbon material-based nanoelectronics

PDF

Optoelectronic properties and device physics of individual suspended carbon nanotubes

PDF

Printed and flexible carbon nanotube macroelectronics

PDF

GaAs nanowire optoelectronic and carbon nanotube electronic device applications

PDF

Transient pulsed plasma for pollution remediation and energy conversion applications

PDF

Carbon nanotube nanoelectronics and macroelectronics

PDF

Nanoelectronics based on gallium oxide and carbon nanotubes

PDF

Optical, mechanical, and electrical properties of nano-structured materials

PDF

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

PDF

Electronic and optoelectronic devices based on quasi-metallic carbon nanotubes

PDF

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

PDF

Nanomaterials for macroelectronics and energy storage device

PDF

Analysis of carbon nanotubes using nanoelectromechanical oscillators

PDF

Controlled synthesis, characterization and applications of carbon nanotubes

PDF

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

PDF

Plasma emission across metal nanoparticle surfaces and semiconductor -liquid interface characterizations

PDF

Hydrogen storage in carbon and silicon carbide nanotubes

PDF

Hot carriers in bare metals and photocatalytically active defect sites in dielectric/metal structures

Asset Metadata

Creator Wang, Bo (author)

Core Title Light Emission from Carbon Nanotubes and Two-Dimensional Materials

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Physics

Publication Date 04/25/2020

Defense Date 03/10/2020

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

Tag carbon nanotubes,defects,electroluminescence,OAI-PMH Harvest,photoluminescence,two-dimensional materials

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Cronin, Stephen Burke (committee chair), Haas, Stephan (committee member), Nakano, Aiichiro (committee member), Wang, Han (committee member), Wu, Wei (committee member)

Creator Email wang510@usc.edu,wangb19880621@163.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-290272

Unique identifier UC11665909

Identifier etd-WangBo-8342.pdf (filename),usctheses-c89-290272 (legacy record id)

Legacy Identifier etd-WangBo-8342.pdf

Dmrecord 290272

Document Type Dissertation

Rights Wang, Bo

Type texts

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

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA