University of Southern California Dissertations and Theses

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

(USC Thesis Other)

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

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

OPTICAL, MECHANICAL, AND ELECTRICAL PROPERTIES OF NANO-
STRUCTURED MATERIALS

by
Chia-Chi Chang

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)

December 2012

Copyright 2012 Chia-Chi Chang

ii

Dedication

This thesis is dedicated to my parents, sisters, wife, and son
for their constant support.

iii

Acknowledgements

First, I would like to thank my research advisor, Prof. Stephen B. Cronin, who
provides me instructive guidance and solid training during my Ph. D study. He
demonstrates how to solve complicated problems and interpret the results. Under his
supervision, I learn how to design and conduct experiments independently. I very much
enjoy working with him in these four years.
I also want to thank all the members in Cronin research group, Adam Bushmaker,
I-Kai Hsu, Wei-Hsuan Hung, Zuwei Liu, Wenbo Hou, Mehmet Aykol, Jesse Theiss,
Chun-Chung Chen, Prathamesh Pavaskar, Moh Amer, Rohan Dhall, Shunwen Chang,
Zhen Li, Jing Qiu, and Shermin Arab. They share the experimental experience with me to
facilitate my research progress.
I am thankful to Prof. Daniel Dapkus and Prof. Chongwu Zhou. With the
collaboration and discussion, I have the chance to explore many interesting research
projects in the Energy Frontier Research Center. In addition, I want to thanks Chun-Yung
Chi, Maoquing Yao, and Haitian Chen for providing me the semiconducting nanowires
and nanosheets.
I am grateful to Prof. Stephan Haas, my academic advisor. He gives me very
useful advices about life in U.S. and choosing research groups during my first two years.
iv

Thanks to Prof. Aiichiro Nakano for serving my committee of both qualifying exam and
final defense. He provides thoughtful feedback and comments to my work.
Finally, I want to thank my family, my parents, sisters, wife, and son. Without
their supports and encouragements, I could not have the chance to study in U.S. and have
this wonderful journey in my life.

v

Table of Contents

Dedication ........................................................................................................................... ii
Acknowledgements ............................................................................................................ iii
List of Tables .................................................................................................................... vii
List of Figures .................................................................................................................. viii
Abstract ............................................................................................................................. xv
Chapter 1: Background ....................................................................................................... 1
1.1 Physical, Electronic, and Phonon Structures of SWCNTs ....................................... 1
1.2 Phonon Confinement in Silicon Nanocrystals ....................................................... 14
1.3 Surface States in Semiconductors .......................................................................... 19
Chapter 2: Mechanical Breaking Strain of Individual SWCNTs ...................................... 26
2.1 Abstract .................................................................................................................. 26
2.2 Introduction to Mechanical Properties of SWCNTs .............................................. 27
2.3 Suspended SWCNTs Growth and Experimental Setup ......................................... 28
2.4 Experimental Results and Discussion .................................................................... 33
2.5 Conclusion ............................................................................................................. 43
Chapter 3: Strain-Induced Raman D Band Observed in SWCNTs .................................. 44
3.1 Abstract .................................................................................................................. 44
3.2 Introduction to Raman D mode .............................................................................. 44
3.3 Sample Preparation and Experimental Setup ......................................................... 47
3.4 Experimental Results and Discussion .................................................................... 48
3.5 Conclusion ............................................................................................................. 61
Chapter 4: Tailoring the Crystal Structure of Individual Silicon Nanowires by
Polarized Laser Annealing .................................................................................... 62
4.1 Abstract .................................................................................................................. 62
4.2 Thermal Annealing for Semiconducting Materials ................................................ 62
4.3 Sample Preparation and Experimental Setup ......................................................... 64
4.4 Experimental Results and Discussion .................................................................... 66
4.5 Conclusion ............................................................................................................. 76
Chapter 5: Electrical and Optical Characterization of Surface Passivation in GaAs
Nanowires ............................................................................................................. 78
vi

5.1 Abstract .................................................................................................................. 78
5.2 Electron Beam Induced Current (EBIC) Measurement ......................................... 78
5.3 Time-Resolved Photoluminescence (PL) Measurement ........................................ 82
5.4 Experimental Results and Discussion .................................................................... 84
5.5 Conclusion ........................................................................................................... 101
Chapter 6: Electrical and Optical Characterization of Twin-Free GaAs Nanosheets ..... 102
6.1 Abstract ................................................................................................................ 102
6.2 Material Contrast in Scicanning Electron Mroscopy ........................................... 102
6.3 Experimental Results and Discussion .................................................................. 106
6.4 Conclusion ........................................................................................................... 116
Chapter 7: Conclusion..................................................................................................... 118
Bibliography ................................................................................................................... 121

vii

List of Tables

Table 2-1. Summary of Raman data taken on nanotubes under strain. ............................ 41

viii

List of Figures

Figure 1-1. A sheet of graphene rolled to form different types single walled
carbon nanotubes[24]. ............................................................................................. 2
Figure 1-2. (a) Honeycomb lattice of graphene. (b) Corresponding Brillouin
zone[16]. ................................................................................................................. 5
Figure 1-3. Electronic band structure of a graphene lattice[16]. ........................................ 6
Figure 1-4. The allowed cutting lines (wave vectors), energy dispersion, and
density of states for (6, 6) metallic and (7, 0) semiconducting
nanotubes[65]. ......................................................................................................... 9
Figure 1-5. Kataura plot. The transition energies as a function of nanotube
diameter................................................................................................................. 10
Figure 1-6. (a) Phonon dispersion of graphene, including longitudinal optical
(LO) phonon, in-plane transverse optical (iTO) phonon, out-of-plane
transverse optical (oTO) phonon, and corresponding acoustic phonons,
LA, iTA, oTA. (b) Phonon DOS of graphene. (c) Phonon dispersion of a
(10, 10) CNT (d) Phonon DOS of a (10, 10) CNT[41]......................................... 11
Figure 1-7. Radial breathing mode (a) and tangential vibrational modes (b) of
carbon nanotubes. (c) Typical Raman spectrum of CNTs and
corresponding vibrational modes. ......................................................................... 13
Figure 1-8. Diameter dependence for the G
-
and G
+
band frequencies for different
isolated semiconducting and metallic nanotubes. Filled and open symbols
represent for semiconducting and metallic nanotubes, respectively[1]. ............... 14
Figure 1-9. Raman spectra and fitting curves of silicon NWs with crystal size of
15 nm (a) and 10 nm (b)[92]. ................................................................................ 16
Figure 1-10. Calculated Raman spectra with different crystal sizes as indicated by
modified phonon confinement model[48]. ........................................................... 17
Figure 1-11. Raman shifts as a function of diameters of nanospheres and
nanowires calculated by bond polarizability model[188]. .................................... 18
ix

Figure 1-12. Raman shifts (a) and linewidth (b) as a function of the size of
nanocrystals according to the modified model (Equation 1.20), RWL
model, and BP model, and experimental data from literature[48]. ....................... 19
Figure 1-13. Energy distribution curves of freshly cleaved and oxygen
contaminated silicon surfaces[167]. ...................................................................... 21
Figure 1-14. Silicon energy band diagram with band bending compared with the
surface electron energy distribution[167]. ............................................................ 22
Figure 1-15. Energy band structures of GaAs surfaces based on energy loss
spectra: solid line represents bulk density of states for (111) As-saturated
surface; dot-dashed line represents (111) Ga-saturated surface; dashed line
represents (100) Ga-saturated surface. Shaded area indicates the filled
states[167]. ............................................................................................................ 23
Figure 1-16. Electronic density mapping of GaAs surface grown by MBE,
obtained by scanning tunneling microscopy. The dots represent As antisite
defects in the top three layers[8]. .......................................................................... 24
Figure 1-17. Summary of band energies of reported surface states in different
types of GaAs surfaces, inclucing dangling bond-induced[102],
intrinsic[46, 152], antisite defect-induced[23, 174], oxygen-induced
surface states[152]. ............................................................................................... 25
Figure 2-1. (a) Our carbon nanotubes CVD growth system. (b) Growth diagram.
(c) SEM image of ultra-long aligned carbon nanotubes. ...................................... 30
Figure 2-2. Fabrication process of breakable H-shaped chip for CNTs growth and
strain experiment. .................................................................................................. 32
Figure 2-3. Photograph images of our experimental setup and SEM images of
suspended SWCNT: (a) breakable H-chip, (b) broken H-chip with
suspended SWCNTs mounted on the translation stage, and (c) SEM image
of SWCNTs spanning a trench. ............................................................................ 32
Figure 2-4. (a) G band Raman frequency of an individual, suspended carbon
nanotube under applied strain. (b) Raman spectra before, during, and after
applying 12% strain. (c) Raman spectra (vertical axis) mapped
chronologically as the sample is strained and unstrained. (d) FWHM
plotted as a function of G band downshift, | Δω
G
|. ................................................ 35
Figure 2-5. (a) G band Raman frequency of another suspended carbon nanotube
under applied strain. (b) Raman spectra before, during, and after applying
6.2% strain. (c) Raman spectra (vertical axis) mapped chronologically as
x

the sample is strained and unstrained. (d) FWHM plotted as a function of
G band downshift, | Δω
G
|. ...................................................................................... 37
Figure 2-6. (a) Selected Raman spectra of the nanotube in Figure 2-4 at various
degrees of strain show the G band softening and intensity diminishing. (b)
Selected Raman spectra of the nanotube in Figure 2-5 at various degrees
of strain show the G band softening and splitting. ................................................ 39
Figure 3-1. Raman double resonance for linear dispersive bands with Fermi
velocities V
1
and V
2
[154]. ..................................................................................... 45
Figure 3-2. SEM images of ultra-long CNTs before (a) and after (b) PDMS
transfer. ................................................................................................................. 48
Figure 3-3. (a) Experimental setup of on-substrate sample. (b) AFM image of
nanotubes on the elastic PDMS substrate. ............................................................ 48
Figure 3-4. (a) Raman spectra of suspended carbon nanotubes at strains of 0, 2,
3.5, and 5%. (b)-(d) D, G
-
, and G
+
band frequencies plotted as a function
of applied strain. In run1, strain is decreased from 4.5% to 0.5%. In run2,
strain is increased from 0% to 5%. ....................................................................... 50
Figure 3-5. (a) D band Raman frequency plotted as a function of G
-
band
frequency. (b)-(c) Raman intensity of G and D band plotted as a function
of the G
-
band shift ( 
G-
). (d)-(e) Intensity ratio of D to G band plotted as
a function of applied strain and the square of the G2 band shift,
respectively. .......................................................................................................... 51
Figure 3-6. (a) Defect-induced intra-valley double resonance process in the
electronic band structure of graphene[165]. (b) Strain-free graphene
hexagonal (in red) and strained graphene orthorhombic (in green)
Brillouin zones. The orthorhombic BZ is generated by folding the
hexagonal BZ along the green lines. .................................................................... 53
Figure 3-7. (a) Raman spectra of on-substrate carbon nanotubes at strains of 0, 3,
and 6%. (b)-(d) D, G2, and G1

band frequencies plotted as a function of
applied strain. Run1 starts from 0% strain and is increased to 3% strain.
Then, the strain is released to 0% and increased to 6% for run2. Run 3
starts from 3%, and is increased up to 9%. ........................................................... 55
Figure 3-8. (a) D band Raman frequency plotted as a function of G2 band
frequency. (b)-(c) Raman intensity of G2 and D band plotted as a function
of the G2 band shift (
G2
). (d)-(e) Intensity ratio of D to G2 band plotted
as a function of applied strain and the square of the G2 band shift,
respectively. .......................................................................................................... 56
xi

Figure 3-9. Raman spectra of the on-substrate carbon nanotubes shown in Figures
5 and 6 under various degrees of strain. The * symbol indicates peaks
originating from the underlying PDMS substrate. ................................................ 59
Figure 3-10. D band linewidth plotted as a function of strain for (a) suspended and
(b) on-substrate carbon nanotubes. Raw spectra of (c) suspended and (d)
on-substrate CNTs at 5% strain. ........................................................................... 60
Figure 4-1. (a) SEM image of as-grown Si NWs. (b) TEM images of individual
NWs. (c) Optical image of a silicon nanowire with focused laser spot. (d)
AFM image of the NW with a diameter of 78 nm. (e) Raman spectra of the
NW before and after annealing at the same spot. ................................................. 65
Figure 4-2. (a) Raman spectra before and after laser annealing at different powers
at the same laser spot. These spectra have been artificially offset for
clarity. (b) Crystalline fraction X
c
, (c) Raman linewidth, and (d) Raman
shift relative to single crystal bulk and estimated mean crystalline size
after laser annealing. The inset in (c) shows the crystalline fraction plotted
as a function of laser irradiation time at two constant laser powers. .................... 67
Figure 4-3. (a) Raman spectra before and after focused laser annealing. (b)
Integrated intensity of crystalline peak ( ) as a function of position,
plotted together with the intensity of amorphous peak ( . (c) Raman
intensity mapping along the NW. ......................................................................... 69
Figure 4-4. (a) TEM image of a locally annealed Si NW in which the annealed
area is 1 µm indicated by the box. (b) Schematic diagram of the NW
boundary between annealed and unannealed regions. Corresponding
HRTEM images in 4c and 4d are indicated by the dashed boxes. (c)
HRTEM image of the unannealed area near the core of the NW. (d)
HRTEM image of the annealed segment near the NW surface. (e) Stokes
and anti-Stokes Raman spectra taken during the laser annealing process.
(f) Crystalline peak intensity, (g) Raman shift and linewidth plotted as a
function of position along the NW........................................................................ 72
Figure 4-5. (a) Experimental setup of polarization-dependent laser heating on
crystalline NWs. (b) AS/S intensity ratio and corresponding temperature
as a function of polarization angle. (c) Raman intensity, linewidth, and
Raman shift plotted as a function of polarization angle with respect to the
NW axis. ............................................................................................................... 73
Figure 4-6. (a) Schematic diagram of polarization dependent laser-induced
preferential annealing. Blue areas represent crystalline silicon. (b) Spatial
Raman intensity mapping of a laser annealed NW at various polarization
angles. (c) Crystalline fraction X
c
before and after annealing and calibrated
xii

annealing temperature plotted as a function of polarization angle and
position. ................................................................................................................. 76
Figure 5-1. Band diagrams of a metal, a semiconductor, and a metal-
semiconductor interface. ....................................................................................... 79
Figure 5-2. (a) SEM image of as-grown silicon nanowire and schematics of
nanowire before and after ion implantation. (b) Correlated SEM and EBIC
images [63]............................................................................................................ 80
Figure 5-3. (a, b) False-color SEM and EBIC images. (c) EBIC profiles along
(solid line) and perpendicular to (dashed line) the nanowire axis. (d)
Minority carrier (hole) diffusion length as a function of nanowire
diameter[2]. ........................................................................................................... 81
Figure 5-4. (a) JOEL 6610 SEM in Center of Electron Microscope and
Microanalysis at USC. (b) Inside of SEM chamber with homemade chip
carrier and wring. (c) Side-view of wire-bonded sample. (d) Schematic of
measurement setup. ............................................................................................... 82
Figure 5-5. (a) Schematic diagram of single photon counting technique. (b)
Exponential decay of photoluminescence lifetime histogram. [PicoQuant,
2009] ..................................................................................................................... 83
Figure 5-6. Schematic diagram of time-resolved PL setup at Aerospace research
lab. ......................................................................................................................... 83
Figure 5-7. The calculated free carrier density plotted as a function of dopant
impurity concentration for GaAs nanowires with various diameters. .................. 86
Figure 5-8. (a) SEM image of a tapered Al
x
Ga
1-x
As-passivated GaAs nanowire.
(b) Continuous-wave PL spectra taken at the tip and base of the nanowire
in (a). (c, d) Spatially mapped PL and Raman data along the nanowire axis
plotted as a function of position. ........................................................................... 90
Figure 5-9. Diameter of GaAs nanowires as a function of position. Insets are SEM
images of passivated (top) and bare (bottom) nanowires. AFM height
profiles at different position for passivated (b) and bare (c) nanowires. .............. 91
Figure 5-10. (a) TEM image of Al
x
Ga
1-x
As-passivated GaAs nanowire slice in
longitudinal direction prepared by focused ion beam (FIB) technique. The
sample thickness is around 70nm. HRTEM images near the tip (b) and the
base (c) of the nanowire. We performed this experiment on a shorter
sample than those in Figure 5-8 due to the limitation of our FIB technique.
xiii

We still observed that the sample was bent after FIB slicing, which results
in invisibility of stacking faults at the tip of the nanowire. .................................. 92
Figure 5-11. (a) EDX and (b) Raman spectra of Al
x
Ga
1-x
As-passivated GaAs
nanowires. ............................................................................................................. 93
Figure 5-12. (a) SEM image of a Al
x
Ga
1-x
As-passivated GaAs nanowire. (b) the
corresponding CL mapping................................................................................... 95
Figure 5-13. (a) SEM image of an Al
x
Ga
1-x
As-passivated GaAs nanowire device,
and (b) corresponding EBIC image. (c) Schematic diagram illustrating the
electron beam induced current measurement. (d) EBIC profiles along the
nanowire axis for passivated and bare nanowires. ................................................ 97
Figure 5-14. GaAs nanowires grown on (a) a GaAs (111) surface and (b) a silicon
(111) surface. Both scale bars are 20 nm. The lines perpendicular to the
nanowire axis are twin sacking faults. .................................................................. 98
Figure 5-15. Time-resolved photoluminescence spectra of passivated and
unpassivated nanowires with different impurity concentrations. ......................... 99
Figure 6-1. 20 kV BSE image of 17-period Al
0.5
Ga
0.5
As/GaAs superlattice
obtained from Hitachi S-4500 FE-SEM[125]. .................................................... 104
Figure 6-2. Schematic diagrams of surface states induced band bending effects in
n and p type semiconductors. .............................................................................. 104
Figure 6-3. SE SEM images of GaAs-based heterostructure containing several p-
and n- layers. The p and n dopants were Be and Si, respectively, at
concentrations of 2 10
18
cm
-3
[125]. ................................................................... 105
Figure 6-4. 1 keV SE image of carbon nanotubes on SiO
2
/Si substrate and
schematic illustration of the origin of contrast difference between metallic
and semiconducting nanotubes[93]. .................................................................... 105
Figure 6-5. SE SEM image of GaAs nanosheets grown on GaAs substrate with a
nitride layer as growth mask. Nanosheets in the status of just-terminated
(a), beginning of overgrowth (b), and further overgrowth (c). ........................... 109
Figure 6-6. (a) PL spectra of GaAs nanosheet and commercial wafers at 77 K. The
carrier concentration of 1.2 10
18
cm
-3
is estimated. (b) Hall voltage as a
function of applied current in the nanosheet. Carrier concentration is
9.2610
17
cm
-3
. .................................................................................................... 111
xiv

Figure 6-7. (a) SE SEM image of GaAs nanosheets contacted with metal
electrodes. Dashed lines indicate the boundary between initial- and over-
growth. (b) Rectifying I-V curve measured across the boundary. Inset
shows linear IV when electrodes are both in initial growth region. ................... 112
Figure 6-8. (a-b) Correlated SEM and EBIC images of a GaAs nanosheet. (c)
EBIC profiles across the inclined boundary of three different sheets. (d)
EBIC profiles across the boundary (solid squares) and metal-nanosheet
Schokkty contact. ................................................................................................ 115
Figure 6-9. (a) Optical microscope image of GaAs nanosheets on a silicon
substrate. The colored dots represent the locations of laser spot. (b) Raman
spectra of the nanosheet taken at the different locations. ................................... 116

xv

Abstract

This thesis includes three main topics covering a broad range of nanoscience:
mechanical and optical properties of single walled carbon nanotubes (SWCNTs), optical
manipulation and characterization of the crystal structure of silicon nanowires, and a
systematic study of minority carrier dynamics in surface-passivated GaAs nanostructures,
including nanowires and nanosheets. The physical properties of these materials and
related phenomena are briefly reviewed in Chapter 1, including physical, electronic, and
phonon structures of SWCNTs, light absorption and phonon confinement in nanocrystals,
and surface states in semiconductors.
In Chapter 2, we study the ultimate breaking strain of SWCNTs with in-situ
Raman spectroscopy. Chemical vapor deposition (CVD) with a high flow rate is used to
synthesize ultra-long suspended SWCNTs across an extendible slit. The G band Raman
frequency downshifts linearly with strain and its downshifting rate spans a wide range
from -6.2 to -23.6 cm
-1
/%. This observation corroborates the theoretical prediction that
the downshifting rate is strongly chiral dependent. A threshold downshift 
G
>75 cm
-1
is
observed for the G band lineshape broadening. In this experiment, we achieve strains up
to 13.7 0.3% without slippage, breakage, and defect formation.
In Chapter 3, we observe emergence of the D band Raman mode in SWCNTs
under axial strain. The D to G mode Raman intensity ratio (I
D
/I
G
) is seen to increase with
strain quadratically by more than a factor of 100. However, the ratio returns to its original
xvi

pre-strain value, indicating that there is no permanent defect formation. Strain-induced
symmetry lowering is proposed to explain the appearance of the D band without a real
defect in the lattice. Another possible scenario is a strain-induced reversible transition
from sp
2
to sp
3
bond configuration in a twisted nanotube bundle.
In Chapter 4, we develop a method of locally tailoring crystal structure of
individual crystalline-amorphous core-shell silicon nanowires with a polarized laser spot.
We are able to control the crystallinity of the silicon nanowires from 0 to 0.93 by
controlling the incident laser power. Raman spectroscopy is used to determine the
annealing temperature while annealing the nanowires and to characterize the crystal
structure before and after annealing. High resolution transmission electron microscopy
(HRTEM) is used to confirm the laser-induced crystallization. Due to the one-
dimensional nature of nanowires, a strong polarization dependent heating/annealing is
observed. The most efficient annealing occurs when the laser polarization is aligned
along the axis of the nanowires.
In Chapter 5, we focus on the minority carrier dynamics in Al
x
Ga
1-x
As-passivated
GaAs nanowires. With passivation, the minority carrier diffusion length (L
diff
) increases
from 30 nm to 180 nm, as measured by electron beam induced current (EBIC) mapping,
and the photoluminescence (PL) lifetime increases from sub-60 ps to 1.3 ns. A 48-fold
enhancement in the continuous-wave PL intensity is observed on the same individual
nanowire with and without the Al
x
Ga
1-x
As passivation layer, indicating a significant
reduction in surface recombination. These results indicate that the minority carrier
xvii

lifetime is not limited by twin stacking faults in passivated nanowires. From the PL
lifetime and minority carrier diffusion length, we estimate the surface recombination
velocity (SRV) to range from 1.7 10
3
to 1.1 10
4
cm s
-1
, and the minority carrier mobility
 is estimated to lie in the range from 10.3 to 67.5 cm
2
V
-1
s
-1
for the passivated nanowires.
In Chapter 6, we explore the interesting phenomena rising from the over-growth
of GaAs nanosheets. Due to the shape anisotropy, we are able to grow stacking fault free
nanosheets in the initial growth region. The twin-free crystal structure is confirmed by
HRTEM images. However, clear boundaries appear at the interface between initial
growth and over-growth regions in both EBIC and SEM contrast images. Possible
scenarios, such as sudden changes in the crystal structure, doping, and surface quality, are
discussed.

1

Chapter 1: Background

As to the broad range of topics covered in this thesis, including mechanical and
optical properties of single walled carbon nanotubes (SWCNTs), optical manipulation
and characterization of the crystal structure of silicon nanowires, and minority carrier
dynamics in surface-passivated GaAs nonostructures, the physical properties of these
materials and related phenomena are briefly reviewed in this chapter. In Section 1.1, we
review the physical, electronic, and phonon structures of SWCNTs. In Section 1.2, we
review the concept of phonon confinement effect in nanocrystals. In Section 1.3, we
review the surface states and surface depletion in semiconductors.

1.1 Physical, Electronic, and Phonon Structures of SWCNTs
Carbon is the sixth element in the periodic table with six electrons which occupy
1s
2
, 2s
2
,and 2p
2
atomic orbitals. Two electrons are strongly bound to the atomic core and
occupy the 1s
2
orbital. The other four electrons are weakly bound and occupy valence
2s
2
2p
2
orbitals. The energy difference between the four valence electrons is small. Their
wavefunctions can mix and form hybridized orbitals, which enhances the binding energy
of the carbon atom with its neighboring atoms. There are three possible hybridizations in
carbon: sp, sp
2
, and sp
3
, whereas a single 2s electron mixes with one, two, or three 2p
electrons. Various hybridizations lead to certain structural arrangements of carbon atoms,
2

in which sp binding forms chain structures, sp
2
binding forms planar structures (e.g.
graphite) and sp
3
binding forms tetrahedral structures (e.g. diamond). Different binding
states give very different physical properties. For instance, diamond is one of the hardest
materials, optically transparent and a poor conductor. However, graphite is flexible, black,
and a good conductor. Gaphene is a single layer of graphite in which the sp
2
bonds are as
strong as sp
3
bonds. This gives graphene extraordinary mechanical properties.
Carbon nanotubes have the same sp
2
binding as graphene and they can be formed
by rolling up a single layer of graphene into a seamless cylindrical tube as shown in
Figure 1-1. In a graphene, carbon atoms arrange in a hexagonal lattice with C-C bond
length of a
cc
=0.142 nm. There are two atoms per unit cell with unit vectors
a
1

√ ,
, a
2

√ ,
(1.1)

Figure 1-1. A sheet of graphene rolled to form different types single walled carbon
nanotubes[24].
3

where √ 3a
cc
=0.246 nm is the lattice constant. A carbon nanotube is a rolled-up
graphene sheet along a certain direction. The vector going around the circumference of
the nanotube is called the chiral vector, C=na
1
+ma
2
. (n, m) is the chiral indices. The
circumference of the nanotube is given by the length of the chiral vector:
C 
√ , (1.2)
and the diameter of the nanotube is
d=C/ = √ / . (1.3)
The chiral angle is the angle between C and a
1
, which is

/ √ . (1.4)
The chiral angle is between 0
o
and 30
o
. Tubes with  = 0
o
(n, 0) are called zigzag tubes;
tubes with  = 30
o
(n, n) are called armchair tubes. Both zigzag and armchair nanotubes
are achiral nanotubes. The general nanotubes are called chiral nanotubes. The smallest
graphene lattice vector T perpendicular to C is called translation vector and it is
determined from the chiral indices
T=
a
1
+

a
2
= -t
1
a
1
+t
2
a
2
(1.5)
where d
R
is the greatest common divisor of (2m+n, 2n+m). The length of a unit cell is
T =
. (1.6)
4

The number of carbon atoms in a unit cell is
N
c
=

. (1.7)
Zone-folding technique is used to derive the electronic band structure of carbon
nanotubes from that of graphene. As we discussed earlier, a carbon atom has four valence
electrons. In sp
2
hybridized configuration of graphene lattice, three electrons of each
carbon atom form three covalent bonds (  bonds) with its nearest neighbors. Therefore,
only the fourth electron in the 2p
z
orbital contributes to electronic properties of graphene.
In a unit cell, there are two electrons from carbon atoms A and B, as shown in Figure 1-
2(a). The reciprocal lattice of graphene is also in hexagonal honeycomb structure as
shown in Figure 1-2(b). The lattice vectors are
b
1
=

√ ,

, b
2
=

√ ,

. (1.8)
The tight-binding Hamiltonian for electrons in graphene considering that electrons
can hop to both nearest- and second nearest-neighbor atoms is described by
∑
, ,.′ ∑ , ,
, ,. , , , ,
(1.9)
where , , annihilates (creates) an electron with spin on site R
i
of sublattice A (an
equivalent definition is used for sublattice B), t( 2.8 eV is the nearest-neighbor hopping
energy), and t' is the next nearest-neighbor hopping energy. The energy bands obtained
from this Hamiltonian is
k 3 k
k , (1.10)
5

k 2 cos √ 3 4cos √cos √ (1.11)

Figure 1-2. (a) Honeycomb lattice of graphene. (b) Corresponding Brillouin zone[16].

The value of t' is not well known. Based on the ab initio calculations, it's between
0.02t and 0.2t. For a finite value of t', the electron-hole symmetry is broken. Figure 1-3
depicts the full band structure. The zoom-in part shows a linear dispersion near a Dirac
point K (or K'). The linear dispersion can be obtained by substituting k by K+q in
Equation (1.10) to read q | q |, where q is the momentum measured relatively
to Dirac point K and (1 10
m/s) is the Fermi velocity.

(b)
(a)
6

Figure 1-3. Electronic band structure of a graphene lattice[16].

In nanotubes, the real (C and T) and reciprocal lattice vectors satisfy the
conditions:
C K

= 2, T K

= 0, C K
//
= 0, T K
//
= 2 (1.12)
where K

is the reciprocal lattice vectors corresponds to the chiral vector C and K
//
corresponds to the translation vector T.
These relations yield
K

=
(-t
2
b
1
+t
1
b
2
) and K
//
=
(mb
1
-nb
2
) (1.13)
where N is the number of hexagons per unit cell, N=N
c
/2.
For the band structure of nanotubes, we have to consider the boundary condition
around the circumference. The wavelength of the electron wavefunction in C direction
must satisfy q= C , where q is a integer. Therefore, the wave vector must satisfy
7

K

=


=

 
=

, q= 1, … ,
(1.14)
Only discrete K

wave vectors are allowed. The one-dimensional Brillouin zone
of nanotubes forms cutting lines with a length of

 
in the reciprocal lattice of graphene
with a period of

, where d is the diameter of nanotubes. We can obtain the energy
dispersion of these N sub-bands for nanotubes:

//

//

, ( 0, … 1 ,  
 
) (1.15)
Figures 1-4(a, b) show the cutting lines for the (6, 6) metallic and (7, 0)
semiconducting nanotubes. Each of the cutting lines gives rise to a different energy sub-
band, as shown in Figures 1-4(c, d), and results in a one-dimensional van Hove
singularity (VHS), as shown in Figures 1-4(e, f). The electronic properties are thus
determined. For metallic nanotubes, one of the cutting lines crosses the degenerate K
point in the graphene Brillouin zone, in which there is no band gap. However, no cutting
line crosses the K point for semiconducting nanotubes thus a finite band gap exists.
Figures 1-4(e, f) show the electronic density of states (DOS) related to the band structures
plotted in Figures 1-4(c, d). The DOS approach infinite at a band edge, which results in
many van Hove singularities. In the single electron picture, the optical resonances occur
when the energy of incoming light matches the electron transition between valence sub-
bands and conduction sub-bands. The transition energy is labeled by E
ij
, where ij refers to
the ith conduction sub-band and jth valence sub-band. i=j is required due to the selection
rule for light polarization parallel with nanotubes.

is used to denote transition
8

energies for semiconducting nanotubes and
is used for metallic nanotubes. Figure 1-5
is the well known Kataura plot, which shows the transition energies of different sub-
bands from nanotubes with different diameters.
The method of constructing one-dimensional electronic band structures by cutting
the two-dimensional band structure is called "zone-folding technique". The method does
not consider the curvature effect of nanotubes, which makes it invalid when the diameter
of nanotubes is smaller than 0.5 nm[7, 130]. The electron hybridization of in-plane sp
2

orbitals (  bonds) with out-of-plane sp
3
orbitals (  bonds) becomes non-negligible in
small nanotubes.

9

Figure 1-4. The allowed cutting lines (wave vectors), energy dispersion, and density of
states for (6, 6) metallic and (7, 0) semiconducting nanotubes[65].
(d)
(e) (f)
(c)
10

Figure 1-5. Kataura plot. The transition energies as a function of nanotube diameter.

A force constant model is used to calculate the phonon dispersion of
graphene[110]. There are six phonon branches because there are two carbon atoms in the
unit cell. Three of them are acoustic phonon modes. The other three are optical phonon
modes: one out-of-plane mode and two in-plane modes (longitudinal and transverse
modes). The longitudinal (LO) mode represents that the vibration of carbon atoms is
parallel with the propagation direction; the transverse (TO) mode represents that the
vibration is perpendicular to the propagation direction. The phonon dispersion of carbon
nanotubes can be derived from graphene by using a similar zone-folding technique.
11

Figure 1-6 shows the phonon dispersion and density of states for graphene and a (10, 10)
nanotube. There are many phonon sub-bands in the nanotube, as shown in Figure 1-6(c).
Raman spectroscopy is the most common tool to study these phonons. However, only a
small portion of them are Raman active, satisfying the selection rules, and can be
observed in the Raman spectrum.

Figure 1-6. (a) Phonon dispersion of graphene, including longitudinal optical (LO)
phonon, in-plane transverse optical (iTO) phonon, out-of-plane transverse optical (oTO)
phonon, and corresponding acoustic phonons, LA, iTA, oTA. (b) Phonon DOS of
graphene. (c) Phonon dispersion of a (10, 10) CNT (d) Phonon DOS of a (10, 10)
CNT[41].

Figure 1-7(c) shows the typical Raman spectrum of carbon nanotubes. The radial
breathing mode (RBM), ranging from 100 to 350 cm
-1
, is widely used to determine the
diameter (d) of nanotubes by the following relation[76]
(a)
(b) (c) (d)
12

(1.16)
The D mode at 1350 cm
-1
is not a Raman-active mode in defect-free carbon
nanotubes. It is known as a defect-induced double-resonant Raman mode, is usually an
indicator of crystal quality of carbon nanotubes. The D band frequency increases with
laser energy by about 50 cm
-1
/eV[127, 146, 166, 172], and the intensity increases with the
density of defects[70, 142]. The frequency of the first order Raman modes, such as G
modes, is usually independent of the excitation energy because they are vibration modes
at the  point (k=0). The explanation of laser-energy dependence of the D band
frequency, called dispersive behavior, is that the defect allows large phonon wave vector
q to contribute to the spectra, and the double resonance process selectively enhances
specific large-q modes. The frequency shift is due to the slope of phonon dispersion[154].
The 2D mode is the second order of the D mode, so its frequency is twice as D band.
The high-energy vibration modes, G
-
and G
+
, are optical phonon modes
corresponding to the vibrating direction of carbon atoms perpendicular and parallel to the
tube axis, respectively. The G
-
band has significant difference in the lineshape between
metallic and semiconducting nanotubes due to the electron-phonon coupling effect. In
metallic nanotubes, a broad, asymmetrical Breit-Wigner-Fano (BWF) lineshape of G
-
band can be observed due to the coupling of discrete phonon with conduction
electrons[11, 72]. However, a sharp G
-
peak is normally seen as shown in Figure 1-7(c).
In addition to the electronic property, the frequency of G
-
mode also depends on the
diameter of nanotubes, as shown in Figure 1-8. The G
-
band frequency decreases with
13

decreasing diameter due to the increasing curvature of nanotubes[74, 76]. Furthermore,
the vibration modes we mentioned here are sensitive to strain. The strain-induced phonon
softening and other effects will be presented and discussed in the next two chapters.

Figure 1-7. Radial breathing mode (a) and tangential vibrational modes (b) of carbon
nanotubes. (c) Typical Raman spectrum of CNTs and corresponding vibrational modes.
(a)
(b)
(c)
RBM
G
+

G
-

14

Figure 1-8. Diameter dependence for the G
-
and G
+
band frequencies for different
isolated semiconducting and metallic nanotubes. Filled and open symbols represent for
semiconducting and metallic nanotubes, respectively[1].

1.2 Phonon Confinement in Silicon Nanocrystals
In the study of nanoscale materials, Raman spectroscopy is one of the most
powerful tools due to its advantages of simple and fast measurement, low-power laser
excitation, and non-destruction to the samples. In silicon bulk crystal, the Raman
spectrum exhibits a peak at 520.5 cm
-1
due to the optical phonon dispersion relation with
a linewidth ( ) of ~3 cm
-1
. In silicon nanocrystals, quantum confinement effects on
optical phonons have been observed in the silicon nanocrystals and the one-phonon
confinement models[12, 131, 150] have been used to explain the size dependence of
phonon redshift and lineshape broadening. The so-called RWL model was originally
15

proposed by Richter, Wang, and Ley in 1981. In a bulk crystal, the first-order Raman
scattering process obeys the momentum conservation law (q=0, q is the phonon wave
vector). However, the rule q=0 is violated in nanocrystals due to the strong size
confinement. The theoretical first-order Raman spectrum can be obtained from the
following equation.
,

(1.17)
where is the phonon dispersion relation of bulk silicon;  is the natural Raman
linewidth; 0, is the Fourier coefficient of the phonon wave function with
confinement. The phonon dispersion relation can be expressed by the analytical form:
522

.
, with q in the range from 0 to 2 /a, where lattice constant
a=0.543 nm[122]. Li et al., reported that the best confinement function is a Guassian[92]:
0,
(1.18)
where L is the size of crystals. The fitting results for the different samples are shown in
Figure 1-9. The curve on the top represents the Raman spectrum of silicon NWs with
smaller crystal size (L=10 nm) and the bottom curve represents the one with larger crystal
size (L=15 nm). The Raman peak shifts from 504 to 511 cm
-1
when crystal size increases
from 10 to 15 nm.

16

Figure 1-9. Raman spectra and fitting curves of silicon NWs with crystal size of 15 nm (a)
and 10 nm (b)[92].

However, silicon NWs are crystals in column shape unless there are abundant of
structure defects in the NWs to form crystal grains as shown in Li's report[92]. The
Fourier coefficient can be modified as follows[171]:
|0, ,
|

/

/
1
√

(1.19)
where L
1
and L
2
are, respectively, the diameter and length of the NWs. In this case, the
confinement effect only occurs in the radial direction of the NWs.
17

More recently, a modified one-phonon confinement model is developed to
calculate Raman frequency, intensity, and linewidth of silicon nanocrystals
simultaneously in the report by Faraci et al[48]. In a linear vibrating string of length D,
the possible wave vectors of confined phonons can be assumed as
, with wave
functions requiring zero vibrating amplitude at the border. A confinement function F
c
(r,D)
in real space is used to modify the Fourier coefficient as follows:
,e xp ∙ (1.20)
where , ∑

, for r  D/2, or 0 otherwise.
, 2,4,6…

( is equal to the maximum integer smaller than 2D/a). In this model, the function
, represents the probability distribution as a function of r. In Figure 1-10, the
Raman peak and linewidth of calculated spectra show strong size dependence.

Figure 1-10. Calculated Raman spectra with different crystal sizes as indicated by
modified phonon confinement model[48].

18

The other model, bond polarizability (BP) model, was proposed by Zi et al. in
1996[188]. They use a partial density approach to calculate the force constant in
crystalline silicon and develop a simple relation between Raman downshift and crystal
size:
  = (L)- 
0
= -B(a/L)

, where (L) is the Raman frequency of nanocrystals with size L;

0
is the Raman frequency of bulk silicon (optical phonon frequency at  point). The
parameters B and  are used to describe the vibrational confinement due to the finite size
in a nanocrystal. This model can apply in both spheres and columns. For spheres, the
values of B and  are 47.4 and 1.44. For columns, the values of B and  are 20.9 and
1.08[188]. Figure 1-11 shows the Raman shifts as a function of the diameter for
nanospheres and nanowires.

Figure 1-11. Raman shifts as a function of diameters of nanospheres and nanowires
calculated by bond polarizability model[188].
nanospheres
nanowires
19

Several models were proposed to have better fit with the experimental data as
shown in Figure 1-12. In the region below 3 nm, the current theories are not enough to
have a precise prediction of Raman shifts. This may be due to the experimental errors or
other fundamental effects, like surface effects. However, a fairly good agreement at large
sizes between all the models and the experimental data has been reached.

Figure 1-12. Raman shifts (a) and linewidth (b) as a function of the size of nanocrystals
according to the modified model (Equation 1.20), RWL model, and BP model, and
experimental data from literature[48].

1.3 Surface States in Semiconductors
Surface states are unavoidable in semiconductors, which are electronic states at
the surface of materials. The surface states rise from the sudden termination of material
with a surface. Therefore, the band structure of surface states is very different from that
(a) (b)
20

of the bulk Bloch states. The surface states often lie energetically in the forbidden bulk
band gap. There are so-called intrinsic and extrinsic surface states. The intrinsic surface
states are defined as the states originated from clean and well ordered surfaces which can
be obtained in ultra-high vacuum[38, 46]. The origin of extrinsic surface states includes
dangling bonds[102, 167], chemisorption of the impurities[152], antisite defects in
compound semiconductors[23, 174], etc. The density of surface states could be as high as
810
14
/cm
2
, i.e., approximately one per surface atom[167], which leads to surface
depletion because of the Fermi level pinning at the surface. The surface states also act as
recombination sites for free carriers, leading to undesired electrical properties, such as
short minority carrier diffusion length and lifetime, hysteresis in current-voltage curves,
and transconductance dispersion. In order to solve these problems, we have to understand
the nature of the surface states.
Energy distribution curves of silicon surface were obtained by the AC method
using a hemispherical potential analyzer[167], as shown in Figure 1-13. A copper emitter
is used as a Fermi energy reference, which is zero in x-axis. There are four prominent
peaks in the spectra. Peaks C and D are from bulk-state direct transition (k-conserving),
which will vary with photon energy. Peaks A and B are identified as the transition
between the energy bands of surface states and Fermi energy, which are independent of
the excitation photon energy. Figure 1-13 shows the energy distribution curves of freshly
cleaved and contaminated silicon surfaces. As a result, oxygen contamination effectively
terminates dangling bonds, which neutralizes a negative surface charge. Release of band
bending is evidenced by the shift of contaminated curve to negative direction by 0.6 eV.
21

Because of the large density of surface states, the Fermi level is pinned at surface
state energy. Consequently, a space charge layer forms at the surface, which is called
surface depletion. Figure 1-14 shows the band bending diagram for a n-typed silicon
surface with depletion width of 2.8 nm. The depletion width is inversely proportional to
the majority carrier concentration. However, the silicon-silicon oxide interface can be
easily formed with a very low density of surface states. Thus, the Fermi level is not
pinned and can be controlled by a metal gate on top of the oxide. This is the reason that
silicon dominates modern electronics.

Figure 1-13. Energy distribution curves of freshly cleaved and oxygen contaminated
silicon surfaces[167].

22

Figure 1-14. Silicon energy band diagram with band bending compared with the surface
electron energy distribution[167].

As to GaAs, notorious for its surface states, there is no such easy way to solve the
surface problem. Compared to silicon, GaAs is a III-V compound semiconductor, in
which its surfaces are more sensitive to contaminations and defects. Great efforts have
been made to classify the type of surface states in GaAs. Eastman and Freeouf used
energy-resolved photoemission yield spectroscopy[46] to observe an unoccupied energy
band of surface states with a peak about 0.9 eV above the valence band maximum of
doped GaAs (110). These states are identified as intrinsic surface states. Ludeke and
Esaki studied the surface states of polar surfaces (100) and (111) using electron energy
loss spectroscopy[102]. Empty and filled surface states, attributed to dangling Ga and As
bonds, are observed near the conduction band and valence band edges, respectively, as
shown in Figure 1-15. These surface-related energy bands are very different from that
23

found in the nonpolar (110) surface which contains approximately equal amount of
unsatisfied Ga and As bonds. Nevertheless, the surface states of Ga-saturated (111)
surface is different from that of Ga-saturated (100) surface.

Figure 1-15. Energy band structures of GaAs surfaces based on energy loss spectra: solid
line represents bulk density of states for (111) As-saturated surface; dot-dashed line
represents (111) Ga-saturated surface; dashed line represents (100) Ga-saturated surface.
Shaded area indicates the filled states[167].

Other than the intrinsic and dangling bond surface states, Szuber observed
oxygen-induced and defect-induced surface states near the valence band on thermally
24

cleaned GaAs (100) surface in ultrahigh vacuum[152]; Weber and his coworkers
observed two surface states located at 0.75 eV below the conduction band and at 0.5 eV
above the valence band[174]. They attributed these two surface states to the As
Ga
antisite
defects (As atom sits on Ga sites). Scanning tunneling microscopy is used to resolve the
As antisite defects formed on the GaAs surface grown by molecular beam epitaxy (MBE),
as shown in Figure 1-16. For this sample, the concentration of antisite defects is up to
510
19
cm
-3
. Figure 1-17 summarizes the band energies of different types of surface
states in different GaAs surfaces.

Figure 1-16. Electronic density mapping of GaAs surface grown by MBE, obtained by
scanning tunneling microscopy. The dots represent As antisite defects in the top three
layers[8].
25

Figure 1-17. Summary of band energies of reported surface states in different types of
GaAs surfaces, inclucing dangling bond-induced[102], intrinsic[46, 152], antisite defect-
induced[23, 174], oxygen-induced surface states[152].
26

Chapter 2: Mechanical Breaking Strain of Individual
SWCNTs

2.1 Abstract
We apply immense strain to ultra-long, suspended, single walled carbon
nanotubes while monitoring their Raman spectra. We can achieve strains up to 13.7±0.3
% without slippage, breakage, or defect formation based on the observation of reversible
change in Raman spectra. This is more than twice that of previous observations. The rate
of G band downshift with strain is found to span a wide range from -6.2 to -23.6 cm
-1
/%
strain. Under these immense strains, the G band is observed to downshift by up to 157
cm
-1
(from 1592 to 1435 cm
-1
). Interestingly, under these significant lattice distortions,
we observe no detectable D band Raman intensity. Also, we do not observe any
broadening of the G band linewidth until a threshold downshift of 
G
> 75 cm
-1
is
achieved at high strains, beyond which the FWHM of the G band increases sharply and
reversibly. Based on a theoretical nonlinear stress-strain response, we estimate the
maximum applied stress of the nanotubes in this study to be 99 GPa with a strength-to-
weight ratio of almost 74,000 kN·m/kg, which 30 times that of Kevlar and 117 times that
of steel.

27

2.2 Introduction to Mechanical Properties of SWCNTs
As we discussed previously, a carbon nanotube is a rolled-up graphene sheet with
slight curvature. However, graphene was demonstrated to be the strongest material
measured so far[89]. Its breaking strength is 130 GPa at 25% strain. Therefore, carbon
nanotubes were predicted to possess high Young's modulus and breaking strain due to the
low defect density. Carbon nanotubes present a unique mechanical system that is harder
than diamond in the axial direction (Y=0.97 TPa) yet extremely Young's modulus and
flexible in the transverse direction[58, 177]. Unlike diamond, however, carbon nanotubes
are not brittle and can be elongated considerably before breaking[84]. Nanotubes’
tremendous strength-to-weight ratio has been explored for a number of applications
including electromechanical resonators with attoNewton (10
-18
N) and zeptogram (10
-17

g) sensitivity[22, 133], high strength, light weight composites[99, 185], and even space
elevators[128, 178]. However, the exceptional mechanical properties of nanotubes have
been difficult to realize experimentally and utilize for practical applications, mainly
because the strength of the nanotube interface has been the limiting factor.
For more than ten years, theoretical calculations of carbon nanotubes under stain
have predicted the formation of 5-7 defects for strains above 6%[116]. Several
experimental groups corroborated these predictions, observing breaking strains of 5.3%
and 5.8%[168, 182]. However, a few years later, revised theoretical calculations reported
that nanotubes should be stable against defect formation beyond strains of 15%[187].
They calculated an activation barrier for 5-7 defect formation of 2 eV at 15% strain,
making nanotubes very stable against defect formation at room temperature. The authors
28

stated that the “ultimate strength limit of carbon nanotubes has yet to be reached
experimentally”[187]. Further detailed theoretical studies have predicted the breaking
strain of CNTs to lie in the range from 14.5 to 22%, depending on the chiral angle[45,
186]. It is likely that these early experimental measurements merely reflected the strength
of the nanotube-substrate contact and, hence, represent only a lower limit of the true
breaking strain of carbon nanotubes.
Several research groups have performed Raman measurements of nanotubes
under strain. In a majority of this previous work[28, 61, 99, 100, 134, 183-185], bulk
quantities of nanotubes dispersed in polymer composites were measured simultaneously,
giving an ensemble average of many nanotubes. These measurements indicated that only
a small fraction of the strain applied to the composite was transferred to the nanotubes
within the composite. In later measurements, individual nanotube bundles were measured
under strain[29, 30, 82, 85]. These studies showed that only 10% of the applied strain
was transferred to the individual nanotubes within the bundle and that the primary effect
of the strain was to debundle the nanotubes[85]. As a result of this reduced transfer of
strain, previous experimental studies have not been able to provide sufficient strains to
achieve mechanical breakdown in nanotubes.

2.3 Suspended SWCNTs Growth and Experimental Setup
In this experiment, ultra-long suspended single walled carbon nanotubes are
grown by chemical vapor deposition (CVD). Figure 2-1(a) shows our CVD system
29

including a Thermo Scientific Lindberg Blue M quartz tube furnace, a MKS 647C multi-
gas flow controller, and several mass flow controllers. The important parameters for
carbon nanotube growth are the type of catalyst, carbon source, temperature, and flow
rate. Catalysts are usually transitional metals, such as Fe, Ni, and Co, from metal salts or
evaporated metal particles. Different carbon sources (e.g. carbon monoxide, methane,
ethylene, ethanol etc.) have different decomposition temperatures. Here, we use ferric
nitrate Fe(NO
3
)
3
catalyst and flow H
2
:CH
4
:C
2
H
4
at rates of 1500:1500:50 sccm at 900
o
C
for 25 minutes, which is the procedure reported by Brintlinger[10, 138] as shown in
Figure 2-1(b). Ferric nitrate is dissolved in isopropyl alcohol (IPA) with the concentration
of 0.5 mg/ml. We grow nanotubes on Si
x
N
y
coated silicon substrate. A shallow trench is
created by photolithography followed by CF
4
reactive ion etch (RIE) and KOH wet etch.
A substrate with double-sided coating of Si
x
N
y
is necessary for this experiment. The
trench serves as a trap for catalyst while a drop is depositing on one side of the trench.
After loading the substrate at the center of the CVD system with the catalyst side at the
upstream of gas flow, 1000 sccm argon (Ar) is applied to purge the one inch quartz tube
for 6 minutes. The furnace is then heated from room temperature to 900
o
C with 400 sccm
of Ar. After the temperature is stabilized at 900
o
C, methane (CH
4
) of 1500 sccm,
ethylene (C
2
H
4
) of 50 sccm, and hydrogen (H
2
) of 1500 sccm are introduced into the
furnace to start the nanotube growth for 25 minutes. 1000 sccm argon is applied to
terminate the growth and keeps flowing till temperature below 100
o
C. Centimeter-long
and well aligned single walled carbon nanotubes can be obtained by using this high flow
rate method, as shown in Figure 2-1(c).
30

Figure 2-1. (a) Our carbon nanotubes CVD growth system. (b) Growth diagram. (c) SEM
image of ultra-long aligned carbon nanotubes.
(c)
(b)
(a)
SWCNTs
catalysts in a trench
31

In order to apply large strain along the nanotubes, an H-shaped pattern is etched
through a double-sided Si
x
N
y
/Si/Si
x
N
y
wafer, as shown in Figure 2-2(h). This technique is
similar to the one developed by Huang et al[69]. However, we introduce the design of
partially etched bridges to reduce the failure rate during the cutting to separate the chips.
Also, we use translation stage to apply larger strain, instead of limited applied strain
induced by thermal expansion of steel base plate. A breakable segment of the chip,
consisting of partially etched Si, enables the two sides of the slit to be separated easily
after they have been mounted on a translation stage equipped with half micron step
resolution and up to 5 mm maximum displacement, as shown in Figure 2-3(b). That gives
up to 0.25 % strain resolution in long suspended nanotubes (195 µm).
Figure 2-2 shows the details of the fabrication process of the breakable chip.
Photolithography is applied to pattern trenches on Si
x
N
y
double-sided coated wafer. 8
minutes CF
4
RIE transfers the pattern into 400 nm thick Si
x
N
y
layer which serves as
KOH etching mask at 80
o
C till etching through the Si substrate. The remaining Si
x
N
y

membrane is removed before deposition of catalysts. Note that the catalysts are deposited
few millimeters away from the slit in order to have longer contact length between carbon
nanotubes and underlying substrate.
A Renishaw inVia micro-Raman spectrometer is used to take Raman spectra of
these individual nanotubes under various degrees of applied strain with a 532 nm Spectra
Physics solid-state laser at a power of 2 mW. Low laser power is used here to avoid the
heat-induced Raman frequency softening.
32

Figure 2-2. Fabrication process of breakable H-shaped chip for CNTs growth and strain
experiment.

Figure 2-3. Photograph images of our experimental setup and SEM images of suspended
SWCNT: (a) breakable H-chip, (b) broken H-chip with suspended SWCNTs mounted on
the translation stage, and (c) SEM image of SWCNTs spanning a trench.
33

2.4 Experimental Results and Discussion
According to SEM images in Figure 2-3(c), the density of nanotubes is around 1
per 10 µm separation, although they sometimes form small bundles. The small laser spot
size (0.4 µm) and sharp Raman signals indicate that only one nanotube contributes to the
data in each measurement. The long nanotube-substrate contact length is essential to
achieving large strains in nanotubes. Son et al. determined the frictional force between a
nanotube and its underlying Si/SiO
2
substrate to be 10 pN/nm. This relatively small
frictional force per unit length, integrated over these long lengths, yields large net
strains[145]. For every 1 m of nanotube-substrate contact length, up to 10 nN can be
exerted without slippage. Assuming a diameter of 1 nm, this corresponds to
approximately 10 GPa per 1 m of contact length. Using this technique, we are able to
achieve sufficient strains in carbon nanotubes, far exceeding those of previous reports,
which were limited by the effects of nanotube-substrate slippage and nanotube
debundling.
Spatial mapping of these Raman spectra allows us to find the angle between the
nanotube and the direction of strain, , and determine a more accurate value of strain by
the relation =ΔLcos /L, where L is the initial length of suspended SWCNTs and ΔL is
the length change due to strain. Also we find that suspended SWCNTs grown by the high
flow rate method lie straight across the trench with very little slack. We quantify the
amount of slack by monitoring the G band Raman shift. Typically, during the first few
displacements of the translation stage, we observe no downshift of the G band mode, as
34

the slack is taken up. Once a downshift is observed, we define this as 0% strain. From
this point on, Raman spectra are used to ensure that only reversible changes occur in the
nanotube. In these samples, the nanotubes lie on Si
x
N
y
-coated silicon substrates, which
produce a strong background signal that obscures the relatively small nanotube signals.
By checking that the 0% strain spectrum remains the same before and after applying
strain, we carefully rule out any possible slippage of the nanotubes on the underlying
substrate, as observed previously by Kumar, et al[85]. In each strain cycle, the strain was
incrementally increased while checking that the slack spectra remained unchanged. The
last data point taken before an irreversible change is observed is defined as the maximum
applied strain ( 
max
). It should be noted, however, that 
max
is a lower limit of the actual
breaking strength of the nanotube, since the irreversible changes observed here
correspond to slippage events rather than breakage of the nanotubes, as evidenced by the
presence of a Raman signal, as described below.
35

Figure 2-4. (a) G band Raman frequency of an individual, suspended carbon nanotube
under applied strain. (b) Raman spectra before, during, and after applying 12% strain. (c)
Raman spectra (vertical axis) mapped chronologically as the sample is strained and
unstrained. (d) FWHM plotted as a function of G band downshift, | ΔωG|.

Figure 2-4(a) plots the G band frequency as a function of strain, showing a strain-
induced downshift rate of -6.2 cm
-1
/% strain. This downshift is remarkably linear with
strain and is understood on the basis of weakening of the carbon-carbon bonds, which
lowers their vibrational frequency. Several data sets depicting stretching and relaxing of
the nanotube show consistent and reversible data in which the G band frequency reverts
to its original pre-strain position after the strain is relaxed. This indicates that the
nanotube was able to endure these enormous strains without breaking, forming 5-7
0 10 203040 50 60 7080 90
4
8
12
16
20
24
28

FWHM (cm
-1
)
| 
G
|
stretching
relaxing
stretching
relaxing
stretching
(d)
(a)
02 4 6 8 10 12 14
1480
1500
1520
1540
1560

G band Frequency (cm
-1
)
Strain (%)
Stretching
Relaxing
Stretching
Relaxing
Stretching
-6.2cm
-1
/% strain
1400 1450 1500 1550 1600 1650

Intensity (a.u.)
Raman Shift (cm
-1
)
before strain
strain 12%
after strain
(b)
(c)
1480
1500
1520
1540
1560
1580
1600
14 0 12 5 0

Raman Shift (cm
-1
)
Strain (%)
0
36

defects, or slipping on the underlying Si
x
N
y
substrate, which would result in irreversible
shifts in the Raman frequency. Figure 2-4(b) shows the G band Raman spectra of the
nanotube before, during, and after applying 12% strain. Before applying strain, the
Raman spectrum shows a single, sharp G band peak, signifying that this is a
semiconducting nanotube[40]. No D band was observed for this nanotube, even at large
applied strains, indicating that there are very few defects in this nanotube[39]. At 12%
strain, the spectrum shows two peaks at 1490 cm
-1
and 1580 cm
-1
. The 1490 cm
-1
peak is
attributed to the G
-
Raman mode, while the 1580 cm
-1
peak is attributed to amorphous
carbon, since it does not respond to strain at all. Remarkably, after such a large distortion
of the lattice, the G band reverts back to its original pre-strain lineshape and frequency.
Figure 2-4(c) shows the Raman intensity plotted as a function of Raman shift (cm
-1
) and
strain (%). Here, the continuous modulation of the G band peak can be seen clearly,
varying all the way down to 1487 cm
-1
. Our last data point was taken at a strain of
13.7%. Beyond this strain, the G band shifted back to 1567 cm
-1
, indicating that the
nanotube slipped off the substrate but was not broken. Figure 2-4(d) shows the G band
linewidth under the applied strains. Here, we have plotted the FWHM as a function of the
G band downshift. We see essentially no change in the linewidth until a downshift of 75
cm
-1
, which corresponds to a strain of 9.5%. Beyond this threshold, we observe a sharp
increase in the FWHM of this Raman mode. Somewhat surprisingly, the FWHM reverts
back to its original pre-strain linewidth after relaxing the applied strain, indicating that no
permanent defects or damage have been induced in the nanotube.
37

Figure 2-5. (a) G band Raman frequency of another suspended carbon nanotube under
applied strain. (b) Raman spectra before, during, and after applying 6.2% strain. (c)
Raman spectra (vertical axis) mapped chronologically as the sample is strained and
unstrained. (d) FWHM plotted as a function of G band downshift, | ΔωG|.

Figure 2-5 shows the data from another suspended, ultra-long CNT across a 170
m wide slit. While this sample could only endure strains of 6.2%, we observe a much
higher rate of G band downshift with strain (-23.6 cm
-1
/% strain), resulting in a maximum
G band downshift of 157 cm
-1
(from 1592 to 1437 cm
-1
). Figure 2-5(a) shows the G band
frequency plotted as a function of strain. Again, several data sets depicting stretching and
relaxing of the nanotube show reversible (elastic) changes in the G band, indicating no
occurrence of defect formation or slippage. The raw Raman spectra taken before, during,
0123456
1440
1470
1500
1530
1560
1590

G band Frequency (cm
-1
)
Strain (%)
Stretching
Relaxing
Stretching
Relaxing
Stretching
-23.6cm
-1
/% strain
(a)
0 2040 6080 100120140160
3
6
9
12
15
18
21
FWHM (cm
-1
)
| 
G
|
stretching
relaxing
stretching
relaxing
stretching

(d)
(b)
1400 1450 1500 1550 1600 1650

Intensity (a.u.)
Raman shift (cm
-1
)
before strain
strain 6.2%
after strain
(c)
1400
1440
1480
1520
1560
1600
6 6.2 4 0 0 0

Raman Shift (cm
-1
)
Strain (%)
38

and after 6.2% strain are shown in Figure 2-5(b). Here, two peaks can be seen in the
spectra, corresponding to G
+
and G
-
. Figure 2-5(c) shows the Raman spectra mapped as a
function of strain. The G
+
and G
-
bands here downshift at different rates of -10.8 and -
23.6 cm
-1
/% strain, respectively. These differing rates of downshift with strain were
predicted theoretically[175, 181] and also observed experimentally[51], as described
below. The G band linewidth, plotted in Figure 2-5(d), shows no change until a downshift
of 80 cm
-1
(4% strain) is achieved, above which the FWHM increases sharply, but
reversibly. Upon relaxation of this strain both the frequency and linewidth of the G band
revert back to their pre-strain conditions. No D band was observed in either of the
nanotubes shown in Figures 2-4 and 2-5 during the course of these measurements.
The deviation of the data points taken at 0% strain in Figure 2-4(a) is due to the
poor mechanical precision of translation stage, rather than irreversible changes in the
nanotube. In Figure 2-5(a), the different datasets deviate significantly from each other
due to the large downshift rate of -23.6 cm
-1
/% strain as compared to that in Figure 2-
4(a), which is only -6.2 cm
-1
/% strain. Therefore, in Figure 2-5(a), a relatively small
uncertainty in strain results in a large variation in the G band frequency.
Based on our observations, the Raman intensity also changes with applied strain.
Figure 2-4(c) shows the Raman intensity drop by a factor of 19 at 13.7% strain, due to the
strain-induced change in the electronic transition energy of the nanotube[15, 69, 104,
105, 112, 129, 147, 160, 179]. This shifts the resonance window of the nanotube away
from the laser energy. Since the strain-induced shift depends strongly on chiral angle, a
39

13.7% strain can shift the electronic transition energy by anywhere between 0 and 760
meV[69]. Therefore, the 19-fold decrease in Raman intensity observed for the nanotube
in Figure 2-4 is reasonable. The nanotube in Figure 2-5 shows a weaker strain
dependence of the Raman intensity on strain, implying that the chiral angle of this
nanotube is closer to 30
o
[69, 147]. Selective Raman spectra of both nanotubes at different
strains in Figures 2-4 and 2-5 are shown in Figure 2-6.

Figure 2-6. (a) Selected Raman spectra of the nanotube in Figure 2-4 at various degrees
of strain show the G band softening and intensity diminishing. (b) Selected Raman
spectra of the nanotube in Figure 2-5 at various degrees of strain show the G band
softening and splitting.

Table 2-1 summarizes the results observed in six suspended nanotube samples
strained in this way. The Table lists the diameter of the nanotube/bundle as determined
by atomic force microscopy (AFM) (d
B
), the initial length of the unstrained suspended
segment of the nanotube (L), the maximum observed downshift of the Raman frequency
( 
max
), the rate of downshift with strain (d /d ), the maximum strain observed ( 
max
),
1500 1600
2.2%
13.7%
11.4%
9.5%
7.8%
6%
4%
1.2%

Intensity (a.u.)
Raman Shift (cm
-1
)
0%
1400 1500 1600
6.2%
5.6%
5.1%
4.4%
3%
1.7%

Intensity (a.u.)
Raman Shift (cm
-1
)
0%
(a) (b)
40

and the estimated tensile stress corresponding to the maximum observed strain. Based on
previous experimental[89] and theoretical[94] reports on the nonlinear stress-strain
behavior of carbon based materials, a quadratic function was used to represent the
nonlinear stress-strain behavior, in which  = A ε + B ε
2
, where  is the stress in units of
GPa, ε is the dimensionless applied strain, A is the Young’s modulus in units of GPa, and
B is the third-order elastic modulus in units of GPa. Liu et al[94]. have used ab initio
methods to calculate coefficients of A = 1047.3 GPa and B = -2375.2 GPa. The maximum
applied stresses of our nanotubes were obtained using these coefficients and our
maximum applied strain values in this equation. Again, our samples exhibited partial
slippage of the nanotube between 3±0.5 and 13.7±0.3 % strain. That is, we observe a
relaxation of strain, but the nanotube was still suspended. The large variation of observed
maximum strain results from the different contact length and contact quality from sample
to sample.

41

sample d
B
(nm) L ( μm) Δω
max
(cm
-1
) dω/d  (cm
-1
/%) 
max
(%)  Tensile stress (GPa)
#1_suspend - 62
68
(1583-1515)
-17 ±  38.1
#2_suspend - 98
32
(1591-1559)
-10.6 3±0.5 29.3
#3_suspend 1.3 42
51
(1582-1531)
-6.5 8.3±1.2 70.6
#4_suspend 3.3 59
47
(1575-1528)
-6.9 6.8±0.8 60.2
#5_suspend - 174
157
(1592-1435)
-10.8/-23.6 6.2±0.3 55.8
#6_suspend 3.4 195
86
(1573-1487)
-6.2 13.7±0.3 98.9

Table 2-1. Summary of Raman data taken on nanotubes under strain.

In Table 2-1, the rate at which the G band downshifts with strain (d /d ) spans a
wide range from -6.2 to -23.6 cm
-1
/% strain for different nanotubes measured in this
study. Using molecular dynamics simulations, Yang et al. have predicted a strong
chirality dependence of the rate of G band downshift with strain, ranging from -6.25 to -
27.5 cm
-1
/% strain[181]. Similar results were also observed by Wu et al. using ab initio
method[175] and by Gao et al. experimentally[51]. This range is in excellent agreement
with our experimental results. In particular, they found strain-induced downshifts of
  

  /
G
-8 cm
-1
/% strain and   

  /
G
-24 cm
-1
/% strain for armchair nanotubes
(n,n), and   

  /
G
-25 cm
-1
/% strain and   

  /
G
-10 cm
-1
/% strain for zigzag
42

nanotubes (n,0)[181]. However, they only performed calculations on these highly
symmetric armchair (n,n) and zigzag (n,0) nanotubes, which have relatively few atoms
per unit cell compared to achiral nanotubes. Nevertheless, we can estimate the
approximate chirality from the relative downshifts of the G
+
and G
-
band Raman modes.
For the nanotube in Figure 2-5,         
 
/ /
G G
, indicating that its chirality is close
to that of an armchair nanotube ( ≈30
o
). Since the downshift in the G band frequency
gives a measure of the weakening of the C-C bond, it is reasonable to expect breakage to
occur at similar G band downshifts rather than similar values of strain. The linewidth
data in Figures 2-4(d) and 2-5(d) further support this notion, showing thresholds for
linewidth broadening that occur at similar G band downshifts, rather than strain. Since we
do not see any evidence of defect formation in these nanotubes, it is not clear from our
data whether these maximum applied strains ( 
max
) were limited by slippage or indicate
the true threshold strain for mechanical breakdown. While sample #6 exhibited a
maximum strain of 13.7±0.3%, we believe that the lower values of maximum strain
observed in the other nanotube samples are due to slippage between the nanotubes and
their underlying substrates, since a Raman signal can still be observed from the nanotubes
after the maximum strain is reached. We, therefore, ascribe the 
max
values given in Table
2-1 as lower limits for the ultimate strength of carbon nanotubes. 4 out of the 6 nanotubes
in this study have relatively low initial G band frequencies (1573-1583 cm
-1
). We have
ruled out the possibility of pre-strain induced during the nanotube growth process by
making the nanotubes slack, and verifying that no further upshift occurs. In a previous
publication, we reported laser heating of suspended carbon nanotubes[64], which
43

required laser powers higher than 3mW focused to a 0.4 µm spot size in order to observe
G band downshifts due to laser heating of the nanotube. We are, therefore, confident that
laser heating can be ignored in the work presented here.

2.5 Conclusion
In conclusion, by growing ultra-long carbon nanotubes with long substrate-
nanotube contacts, we have been able to apply immense strains to individual, suspended
carbon nanotubes. Monitoring the Raman spectra while reversibly straining and relaxing
the nanotube enables us to identify and quantify the effects of slippage, which have
limited the strain achieved in previous measurements. No evidence of defect formation
was observed under strains up to 13.7±0.3%. This is twice the value of strain achieved in
previous measurements, establishing a new lower limit for the ultimate strength of carbon
nanotubes. The rates of G band downshift with strain span a large range from -6.2 to -
23.6 cm
-1
/% strain, which is consistent with theoretical predictions and previous
observation. Linewidth broadening is observed with strains beyond a threshold downshift
of the G band of 
G
> 75 cm
-1
. Beyond this the FWHM of the G band increases sharply
and reversibly.

44

Chapter 3: Strain-Induced Raman D Band Observed in
SWCNTs

3.1 Abstract
We report the emergence of the D band Raman mode in single walled carbon
nanotubes under large axial strain. The D to G mode Raman intensity ratio (I
D
/I
G
) is
observed to increase with strain quadratically by more than a factor of 100-fold. Up to
5% strain, all changes in the Raman spectra are reversible. The emergence of the D band,
instead, arises from the reversible and elastic symmetry-lowering of the sp
2
bonds
structure. Beyond 5%, we observe irreversible changes in the Raman spectra due to
slippage of the nanotube from the underlying substrate, however, the D band intensity
resumes its original pre-strain intensity, indicating that no permanent defects are formed.

3.2 Introduction to Raman D mode
The D mode has been observed in graphite and has been showed that it is induced
by disorder. The D mode intensity increases with the number of defects. Interestingly, the
D mode frequency shows dispersive behavior, in which its frequency increases with
increasing excitation energy by about 50 cm
-1
/eV[166]. Furthermore, there is a difference
of 7 cm
-1
between Stoke and anti-Stoke frequencies with the same excitation
energy[153]. Almost 20 years later, double resonance Raman scattering was proposed to
explain this dispersive behavior of D mode[154]. Figure 3-1 shows the intra-valley
45

double resonance process starting from light absorption of incident light to excite an
electron from valence band to conduction band at k point (from i to a). For a given
excitation energy and a monotonically increasing phonon dispersion relation ћ 
ph
(q),
there is at most one q and ћ 
ph
leading to the double resonance. The electron is thus
scattered to a state on the other band (from a to b). This is the phonon for which double
resonance occurs. Then defects are needed to scatter the electron elastically back to the k
point (from b to c). Finally, the electron recombines with the hole at the k point (from c
to i).

Figure 3-1. Raman double resonance for linear dispersive bands with Fermi velocities V
1

and V
2
[154].

In carbon nanotubes, a Raman mode related to the graphite D mode is also
observed. The defects in nanotubes can be 5-7 defects, kinks, impurities, or the finite
length of tubes. However, the electronic band structure of nanotubes strongly depends on
46

the diameter and chirality, which could change the double-resonance condition
significantly. Maultzsch et al., found only the nanotubes with (n-m)/3 integer contribute
to the D mode because of the particular electronic structure of nanotubes[109].
Unlike the more commonly studied G band and radial breathing mode, the D band
Raman mode involves phonons with finite momentum[108]. Because photons carry little
momentum, the D band is observed only when the momentum conservation requirement
of the optical Raman process is broken by defects and disorder. Hence, the relative
intensity of the D band gives a measure of the amount of disorder in the nanotubes. As
such, the relative intensity of the so-called defect-induced D band has been used as an
indication of the quality of carbon nanotubes for many years[41, 42, 70]. More recently,
the D to G mode Raman intensity ratio (I
D
/I
G
) in graphene has been used to determine the
defect density ( ), by the relation: (cm
-2
)=(1.8 0.5) 
-4
(I
D
/I
G
) 10
22
, where  is the
excitation wavelength in nanometers[13, 101]. However, no such correlation exists for
CNTs. Furthermore, we expect different types of defects to affect the D band intensity
differently. In a previous study, an atomic force microscope (AFM) tip was used to
induce strains ranging from 0.06% to 1.65% in CNTs clamped at both ends by metal
electrodes[31]. However, no general trend in the I
D
/I
G
Raman intensity ratio was
observed. Several other prior studies have investigated the Raman spectra of carbon
nanotubes under strain, none of which have reported a consistent change in the D band
intensity due to strain[32, 50, 83, 98].

47

3.3 Sample Preparation and Experimental Setup
In this chapter, we investigate the D band Raman intensity using two
experimental approaches to apply large strains to carbon nanotubes in a continuous,
reversible fashion. In the first approach, single walled CNTs are grown suspended across
a gap separating two adjacent silicon substrates. Strain is then induced by increasing the
separation between the two substrates using a micromechanical translation stage, as
introduced in chapter 2[20]. In the second approach, CNTs are grown on an oxidized
silicon substrate and then transferred to an elastic polydimethyl siloxane (PDMS)
substrate. Nearly 100% transfer of nanotubes can be achieved by first treating the PDMS
substrate with an oxygen plasma, as shown in Figure 3-2. The two ends of the PDMS
substrate are then mounted on a translation stage, in order to induce strain, as shown in
Figure 3-3[86]. The ultra-long nature of these nanotubes enables the application of high
strains because the total Van der Waals force that can be applied depends on the length of
the CNT-substrate contact (10 pN/ m)[145]. Nanotubes grown in this fashion are aligned
along the direction of the gas flow. Figure 3-3(b) shows the atomic force microscope
(AFM) image of a PDMS sample. Bundles are typically seen in both suspended and on-
substrate samples, ranging from 2-5 nm in diameter. A 532 nm laser (100 W) was
focused through a 100X high numerical aperture objective lens (NA=0.9) and used to
collect Raman spectra from the nanotubes under strain for both samples.
48

Figure 3-2. SEM images of ultra-long CNTs before (a) and after (b) PDMS transfer.

Figure 3-3. (a) Experimental setup of on-substrate sample. (b) AFM image of nanotubes
on the elastic PDMS substrate.

3.4 Experimental Results and Discussion
Figure 3-4(a) shows the Raman spectra of an ultra-long suspended CNT bundle
with a single resonant nanotube spanning an 82 m wide gap. At 0% strain there are two
(b) (a)
49

prominent peaks corresponding to G
+
and G
-
bands and an almost undetectable D band.
The strain dependence of the D, G
-
, and G
+
band frequencies are plotted in Figures 3-
4(b), 4(c), and 4(d), which exhibit downshifts of 23, 40, and 43 cm
-1
under 5% strain,
respectively. All peaks in these spectra downshift linearly and reversibly with applied
strain up to 5%, indicating that no slippage occurs between the CNT bundle and the
underlying substrate. The broad G
-
band peak at 1577 cm
-1
exhibits a Breit-Wigner-Fano
(BWF) lineshape, typical of metallic nanotubes, which becomes more narrow under
strain due to the opening of a band gap. Strain-induced band gap opening can be created
in SWCNTs under axial strain at a rate of 0 to 30 meV/% strain, depending on the chiral
angle[66]. These changes in the linewidth of the G
-
band are also reversible with strain,
as shown in the inset of Figure 3-4(c), in agreement with previous observations in
metallic SWCNTs[86]. The raw spectra in Figure 3-4(a) show that the D band not only
downshifts but also grows in intensity as the strain increases.

50

Figure 3-4. (a) Raman spectra of suspended carbon nanotubes at strains of 0, 2, 3.5, and
5%. (b)-(d) D, G
-
, and G
+
band frequencies plotted as a function of applied strain. In
run1, strain is decreased from 4.5% to 0.5%. In run2, strain is increased from 0% to 5%.

(a)
1300 1400 1500 1600 1700
G
+
Offset Intensity (a.u.)
Raman Shift (cm
-1
)
0%
2%
3.5%
5% D band
G
-
0 123 45
1340
1350
1360
D Band (cm
-1
)
Strain (%)
run1
run2
-4.7 cm
-1
/ % strain
012 345
1540
1550
1560
1570
1580
-8.0 cm
-1
/ % strain

G
-
Band (cm
-1
)
Strain (%)
run1
run2
0 12345
16
24
32
40

G
-
Linewidth (cm
-1
)
run1
run2
0 123 45
1570
1580
1590
1600
1610
1620
G
+
Band (cm
-1
)
Strain (%)
run1
run2
-8.6 cm
-1
/ % strain
(b)
(c) (d)
51

Figure 3-5. (a) D band Raman frequency plotted as a function of G
-
band frequency. (b)-
(c) Raman intensity of G and D band plotted as a function of the G
-
band shift (
G-
). (d)-
(e) Intensity ratio of D to G band plotted as a function of applied strain and the square of
the G
-
band shift, respectively.

(a)
1540 1550 1560 1570 1580
1340
1345
1350
1355
1360
1365
0%

D Band (cm
-1
)
G
-
Band (cm
-1
)
run1
run2
5%
012 345
0.00
0.01
0.02
0.03
0.04

Intensity Ratio I
D
/I
G
Strain (%)
run1
run2
0 400 800 1200 1600
0.00
0.01
0.02
0.03
0.04

Intensity Ratio I
D
/I
G
run1
run2
( 
G-
)
2
(cm
-2
)
010 20 30 40
10000
20000
30000
40000
50000

G Band Intensity

G-
(cm
-1
)
run1
run2
0 1020 3040
200
300
400
500
600
700
D Band Intensity

G-
(cm
-1
)
run1
run2
(b)
(d)
(c)
(e)
52

Figure 3-5(a) shows the D band frequency plotted as a function of the G
-
band
frequency. The co-linearity of this data indicates that the G
-
and D bands originate from
the same nanotube. The same co-linearity is observed between the G
+
and D bands (not
shown here). Figure 3-5(b) shows the D band intensity normalized by the G band
intensity plotted as a function of strain. This Raman intensity ratio (I
D
/I
G
) exhibits a
quadratic dependence on the applied strain, and changes reversibly from 0.004 to 0.043.
Since the G band downshift is expected to have a linear dependence on the C-C bond
length, we plot I
D
/I
G
as a function of the square of the G
-
band shift (
G-
2
), and obtain
the linear fit shown in Figure 3-5(c). The quadratic dependence of I
D
/I
G
on strain is likely
due to two effects. First, there is a strain-induced reduction of the G band Raman
intensity as the nanotube inter-band transition (
) is shifted off resonance from the laser
energy inversely proportional to 
G-
, as shown in Figure 3-5(d). Second, there is an
increase of the D band intensity proportional to 
G-
due to the strain-induced lowering
of the symmetry, as shown in Figure 3-5(e). The G band Raman intensity depends on the
optical transition energy (E
ii
) and on the laser excitation energy according to the
resonance Raman process[32, 75]. With a fixed laser energy, the resonance Raman
formula can be approximated to first order by a linear or an inverse linear relationship,
depending on whether E
ii
is moving toward or away from resonance. Strain-induced
variations of the optical transition energies in carbon nanotubes have been studied
theoretically and experimentally, showing linear dependences on the value of uniaxial
strain[14, 66]. In the particular region we observed here, the G band intensity is inversely
proportional to the applied strain.
53

Figure 3-6. (a) Defect-induced intra-valley double resonance process in the electronic
band structure of graphene[165]. (b) Strain-free graphene hexagonal (in red) and strained
graphene orthorhombic (in green) Brillouin zones. The orthorhombic BZ is generated by
folding the hexagonal BZ along the green lines.

In carbon nanotubes, the D band is not actually a Raman active phonon mode.
However, this phonon mode can be seen in nanotubes with a large amount of disorder
and defects, which break the symmetry of the lattice and relax the selection rules. Large
amounts of uniaxial strain also break the symmetry of the lattice, and can result in
relaxation of the selection rules. The intensity of the so-called defect-induced D band has
been used to indicate the quality of carbon nanotubes for many years[41, 42, 70].
However, what we observe here are reversible changes in the D band intensity. This
indicates that we are not actually creating permanent defects, but that the strain (5%)
induces a lattice distortion large enough to distort the sp
2
symmetry, making the D band
observable. The diagram in Figure 3-6(a) shows the dominant intra-valley defect-induced
double resonance process in graphene[165]. Defects are needed to bring back the electron
(a) (b)
54

from B to A. Figure 3-6(b) shows the phonon wavevector connects the points A and B in
a hexagonal Brillouin zone (BZ) of a non-strained hexagonal lattice.
Let us now consider a graphene layer, with a uniaxial strain applied along either
the zig-zag or armchair direction. The origin of the D band is a double resonance
intervalley process that, in hexagonal lattices, requires a finite momentum, which is
represented by the blue arrow connecting the A and B points in Figure 3-6(b). This
distortion will decrease the lattice symmetry from hexagonal to orthorhombic, the
orthorhombic unit cell being twice that of the hexagonal one. The orthorhombic BZ zone
can be generated by folding the hexagonal BZ, as shown in Figure 3-6(b). When the
hexagonal BZ is folded into the orthorhombic BZ under strain, however, the points A and
B are folded onto the same point in the orthorhombic zone of a strained hexagonal lattice.
In this case, no defect is needed to activate the D band phonon. Symmetry lowering from
a hexagonal to an orthorhombic lattices can be achieved under large axial strain. This
argument is strictly valid only for a graphene layer, since carbon nanotubes are a one-
dimensional system, with symmetries that depends on their chiral angle. However, since
many nanotubes properties can be obtained from the graphene electronic dispersion, by
considering the cutting lines in the extended Brillouin zone, we expect that the strain-
induced symmetry lowering will also activate the D band for carbon nanotubes. There
have been several reports of Raman spectroscopy of graphene under strain, none of which
report any change in the D band Raman intensity under strain[49, 67, 68, 114, 119].
Therefore, it is likely that our observation of the D band mode under strain may not arise
55

from the distortion of the Brillouin zone, but rather to the locally induced deformations,
as described below.

Figure 3-7. (a) Raman spectra of on-substrate carbon nanotubes at strains of 0, 3, and 6%.
(b)-(d) D, G2, and G1

band frequencies plotted as a function of applied strain. Run1 starts
from 0% strain and is increased to 3% strain. Then, the strain is released to 0% and
increased to 6% for run2. Run 3 starts from 3%, and is increased up to 9%.
1300 1400 1500 1600 1700
*
D
*
PDMS
G2
G1
Offset Intensity (a.u.)
Raman Shift (cm
-1
)
0%
3%
6%
024 6 8
1280
1300
1320
1340
run 1
run 2
run 3
-6.6cm
-1
/ % strain

Strain (%)
D Band (cm
-1
)
02468
1500
1520
1540
1560
1580
G2 Band (cm
-1
)
Strain (%)
run1
run2
run3
-9.1 cm
-1
/ % strain
02468
1550
1560
1570
1580
1590
1600
G1 Band (cm
-1
)

run1
run2
run3
2.3%
Strain (%)
(b)
(c)
(d)
(a)
56

Figure 3-8. (a) D band Raman frequency plotted as a function of G2 band frequency. (b)-
(c) Raman intensity of G2 and D band plotted as a function of the G2 band shift ( 
G2
).
(d)-(e) Intensity ratio of D to G2 band plotted as a function of applied strain and the
square of the G2 band shift, respectively.

(a)
1540 1550 1560 1570 1580 1590
1300
1310
1320
1330
1340
1350
6%

run1
run2
run3
D Band (cm
-1
)
G2 Band (cm
-1
)
0%
012 3
0.0
0.3
0.6
0.9
1.2
run1
run2
run3
Intensity Ratio I
D
/I
G2
Strain (%)
0 200 400 600 800 1000
0.0
0.3
0.6
0.9
1.2

G 


cm
-2

Intensity Ratio I
D
/I
G2
run1
run2
run3

010 20 30
0
15000
30000
45000
60000

G2
(cm
-1
)
run1
run2
run3
G2 Band Intensity

010 20 30
2000
4000
6000
8000


G2
(cm
-1
)
run1
run2
run3
D Band Intensity
(b)
(c) (d)
(e)
57

Another possibility is that the nanotubes are twisted in a bundle, and as strain is
applied, local deformations (i.e., pinching) are created along the length of the resonant
nanotube that subside after releasing the tension. Such local deformations have been
created using an AFM tip, resulting in reversible changes in the conductance by up to two
orders of magnitude. According to molecular dynamic simulations, the significant change
in conductance was attributed to a reversible transition from sp
2
to sp
3
bond
configurations in the local bending region[95, 159].
Figure 3-7(a) shows the Raman spectra of an ultra-long SWCNTs bundle strained
on a PDMS substrate. At zero strain, there are two sharp G bands (G1 and G2), indicating
that the bundle consists of two Raman resonant nanotubes (NT1 and NT2). Again, a
small D band can be seen in this bundle at 0% strain. As the strain is increased, the D
band downshifts in frequency and grows in intensity. Figures 3-7(b) and 3-7(c) show
reversible changes in the Raman frequency up to 8% strain, for the D and G2 bands,
respectively. However, irreversible changes of the G1 band frequency were observed
after 5% strain. Beyond 5%, the G1 band upshifts from 1552 to 1571 cm
-1
, due to
slippage between NT1 and the underlying substrate, as shown in Figure 3-7(d).
Hysteresis plays an important role in these measurements, by indicating the occurrence of
slippage. There is no hysteresis in the Raman downshift of the D and G2 bands, as shown
in Figures 3-7(b) and 3-7(c). The hysteresis in the G1 band frequency, however, indicates
slippage at 5% strain. AFM has been used to reveal the buckling of strain-relaxed
nanotubes[111]. The Raman frequency of D band remains downshifted after 5% strain, as
with the G2 band, indicating that the D band origins from NT2 only. Due to the strong
58

dependence on chirality, strain-induced G band downshifts span a wide range from -6.2
to -23.6 cm
-1
/% strain[20, 51, 175, 180]. Before slippage, the G band downshifted at a
rate of approximately -8 to -9 cm
-1
/% strain, which lies within the range of previously
reported values.
The D band frequency can be plotted as a linear function of the G2 band
frequency, as shown in Figure 3-8(a), indicating that these two modes originate from the
same SWCNT. The intensity ratio I
D
/I
G2
is plotted as a function of strain in Figure 3-8(b)
and varies from 0.01 to 1.34 with a quadratic dependence similar to that shown in Figure
3-5(b). The G band downshift provides a more reliable measure of the C-C bond
elongation, and yields a more clear plot of the intensity ratio as a function of the square of
the G2 band downshift ( 
G2
2
), as shown in Figure 3-8(c). The quadratic dependence
shown here is consistent with the data shown in Figure 3-5(c). Again, we observe that the
G2 band Raman intensity decreases inversely with 
G2
and the D band intensity
increases proportional to 
G2
, as shown in Figures 3-8(d) and 3-8(e). All samples
measured in this work show a reversible increase in the D band linewidth with strain, as
plotted in Figures 3-10(a) and 3-10(b). Interestingly, the linewidths observed in the
suspended sample (Figure 3-4) are around 6 cm
-1
. This is considerably narrower than
previous reports of D band linewidths, which are typically in the range of 10-35 cm
-1
[73].
Recently, narrow peaks (8 cm
-1
) have been observed in the Raman spectra of bilayer
graphene, and attributed to symmetry breaking by Moire patterns of twisted layers of
graphene, which activates a double resonance process[132].
59

Figure 3-9. Raman spectra of the on-substrate carbon nanotubes shown in Figures 5 and 6
under various degrees of strain. The * symbol indicates peaks originating from the
underlying PDMS substrate.
1300 1400 1500 1600
0
1000
2000
3000
4000
Raman Shift (cm
-1
)
*
*
D
G2
G1
Raman Intensity (a.u.)
2.5% strain
1300 1400 1500 1600
0
1000
2000
3000
4000
4% strain
*
D
*
G2
G1
Raman Intensity (a.u.)
Raman Shift (cm
-1
)
1300 1400 1500 1600
0
1000
2000
3000
4000
5.5% strain
Raman Shift (cm
-1
)
*
*
D
G2
G1
Raman Intensity (a.u.)
1300 1400 1500 1600
0
1000
2000
3000
4000
8% strain
Raman Shift (cm
-1
)
*
*
D
G2
G1
Raman Intensity (a.u.)
(a) (b)
(d) (c)
60

Figure 3-10. D band linewidth plotted as a function of strain for (a) suspended and (b) on-
substrate carbon nanotubes. Raw spectra of (c) suspended and (d) on-substrate CNTs at
5% strain.

Figure 3-9 shows the raw Raman spectra of the nanotube shown in Figures 3-7
and 3-8 taken at various degrees of strain. At strains of 2.5% and 4%, the G2 band
intensity decreases significantly with strain. The G1 and G2 bands overlap at strains
between 4 and 5%, making it difficult to obtain reliable intensity data in this range of
strain. As we can see in the spectrum of Figure 3-9(c) taken at 5.5% strain, the G1 and
G2 bands become well separated after the slippage of NT1 at 5% strain. However, for
(a)
012 345
5.2
5.6
6.0
6.4
Suspended CNT

D Band Linewidth (cm
-1
)
Strain (%)
run1
run2
02 468
24
32
40
48
56

D Band Linewidth (cm
-1
)
Strain (%)
run1
run2
run3
On-substrate CNT
1300 1320 1340 1360 1380 1400
0
20
40
60
80

Raman Intensity (a.u.)
Raman Shifts (cm
-1
)
data
fit
5% strain
1280 1300 1320 1340 1360 1380
0
20
40
60
80
100
5% strain
data
fit
Raman Intensity (a.u.)
Raman Shifts (cm
-1
)
(b)
(c) (d)
61

strains higher than 7%, we begin to observe an additional peak between 1500 and 1600
cm
-1
, as indicated by the arrow in the Figure 3-9(d), which may be due to other nanotubes
within the bundle shifting onto resonance with the laser energy or to other non-Raman
active modes becoming observable due to the lowering of the lattice symmetry under
stain[3]. The D band downshifts from 1348 to 1282 cm
-1
before it merges with the PDMS
peak at 1260 cm
-1
. After relaxing the strain, there was no detectable increase in I
D
/I
G2
of
this bundle from its original intensity ratio, as shown in Figure 3-8(b). We exclude the
possibility of breaking C-C bonds at high strains and reconstruction of bonds when strain
is released since this process would require a large amount of energy and the experiments
were carried at room temperature[148]. The strain-induced D band requires a detailed
theoretical calculation to provide a deeper understanding.

3.5 Conclusion
In conclusion, we observe an increase in the D to G mode Raman intensity ratio
(I
D
/I
G
) in carbon nanotubes under the application of uniaxial strain. A 100-fold increase
in the I
D
/I
G
ratio is observed at strains of 5%. However, all changes in the Raman spectra
are reversible with strain, indicating that no permanent defects are formed in the lattice
under the applied strains. Instead, the change in I
D
/I
G
ratio arises from a lowering of the
symmetry of the lattice under these large strains.

62

Chapter 4: Tailoring the Crystal Structure of Individual
Silicon Nanowires by Polarized Laser Annealing

4.1 Abstract
We study the effect of polarized laser annealing on the crystalline structure of
individual crystalline-amorphous core-shell silicon nanowires (NWs) using Raman
spectroscopy. The crystalline fraction of the annealed spot increases dramatically from 0
to 0.93 with increasing incident laser power. We observe Raman lineshape narrowing and
frequency hardening upon laser annealing due to the growth of the crystalline core, which
is confirmed by high resolution transmission electron microscopy (HRTEM). The anti-
Stokes:Stokes Raman intensity ratio is used to determine the local heating temperature
caused by the intense focused laser, which exhibits a strong polarization dependence in Si
NWs. The most efficient annealing occurs when the laser polarization is aligned along the
axis of the NWs, which results in an amorphous-crystalline interface less than 0.5 µm in
length. This paper demonstrates a new approach to control the crystal structure of NWs
on the sub-micron length scale.

4.2 Thermal Annealing for Semiconducting Materials
Annealing is the heat treatment commonly used to improve the crystal quality and
the electrical conductivity of semiconductors. For instance, the crystal defects (such as
63

stacking faults and dislocations) can be removed after appropriated thermal annealing[26,
91]. For silicon, polycrystalline films can be obtained from amorphous silicon
synthesized by low-pressure chemical vapor deposition (LPCVD) via furnace thermal
annealing in the range of 700-1000
o
C[78, 81]. After selected dopant impurities have been
implanted, thermal annealing is needed to redistribute the impurities to the active sites.
However, the over diffusion of dopant atoms induced by conventional furnace annealing
negates an importance advantage of ion-implantation which possesses precise control of
doping profile. Therefore, rapid thermal annealing (RTA) and pulse laser annealing are
developed to have better control of dopant diffusion. A continuous (CW) laser annealing
is thought to have the similar crystal regrowth mechanism with conventional furnace
annealing, where the crystallization takes place in the solid phase. Pulse laser, instead,
produces a melted region due to the high power density within extremely short duration.
The laser-induced temperature depends both on the laser power and irradiation time, and
on the optical absorption of the material and its thermal conductance. In this chapter, we
would like to study the CW laser annealing in individual silicon nanowires which provide
special absorption properties due to the one-dimensional nature.
Compare to bulk materials, nanotubes and nanowires possess high aspect ratio
geometry. Therefore, they behave like antennas, where the light absorption and emission
are strongly enhanced when the light polarization is parallel with the long axis of
nanotubes or nanowires. The antenna effect is pronounced in nanotubes[4, 107] and small
diameter nanowires[157, 170].
64

4.3 Sample Preparation and Experimental Setup
In this paper, we use Raman spectroscopy and high resolution transmission
electron microscopy (HRTEM) to study the crystallization process of core-shell Si NWs
under laser irradiation in an argon environment. A CW 532 nm wavelength laser is
focused through a 100X objective lens (0.9 NA) and used as a local heat source to modify
the crystallinity of core-shell Si NWs. In our Raman spectrometer, a holographic notch
filter is used, which enables us to take both Stokes and anti-Stokes Raman spectra. A
half-wavelength polarizer is placed before the notch filter to rotate the laser polarization.
In-situ Raman spectra are taken during the annealing process with the same incident
laser, which give information about the heating temperature based on the anti-
Stokes:Stokes Raman intensity ratio[136] and the temperature-induced Raman lineshape
broadening and frequency downshifting[21, 135, 149]. Low laser power (0.5 mW)
Raman spectra with an integration time of 60 seconds are taken before and after
annealing to reveal changes in the crystallinity due to the local annealing process. A
strong polarization dependence of this local heating is observed in the core-shell NWs.
By using focused laser annealing, we are able to control the crystallinity of core-shell
NWs by varying the laser power and/or laser polarization, and, thus, create amorphous-
single crystalline junctions in individual NWs.
Core-shell Si NWs were synthesized by the thermal CVD method. A 2 nm film of
gold was first evaporated on Si (100) substrates and annealed at 530
o
C for 30 minutes to
produce isolated gold nanoparticles, which serve as catalysts for the NW growth. The
growth was then carried out at 530
o
C for 30 minutes while flowing 20 sccm hydrogen
65

(H
2
) and 111 sccm silane (SiH
4
). The resulting nanowires have a diameter distribution
extending from 40 to 100 nm based on TEM and AFM characterizations, as shown in
Figure 4-1. As-grown nanowires were sonicated in isopropyl alcohol and then transferred
onto substrates (0.25 mm thick glass or 100 nm thick silicon nitride membranes). The
Raman spectrometer was calibrated with a single crystal (100) silicon substrate routinely
before each experiment to ensure that the transverse optical (TO) phonon mode of silicon
exhibits a Raman peak with a frequency of 520.5 cm
-1
and a FWHM of 4.5 cm
-1
.

Figure 4-1. (a) SEM image of as-grown Si NWs. (b) TEM images of individual NWs. (c)
Optical image of a silicon nanowire with focused laser spot. (d) AFM image of the NW
with a diameter of 78 nm. (e) Raman spectra of the NW before and after annealing at the
same spot.
(a)
(c) (d)
(e)
(b)
400 500 600

Intensity (a.u.)
Raman Shift (cm
-1
)
as-grown
annealed
66

4.4 Experimental Results and Discussion
Figures 4-1(a) and 4-1(b) show electron microscope images of Si NWs grown in
this study. Metal strips with grid markers were deposited on top of one of the NWs, as
shown in Figure 4-1(c). Atomic force microscope (AFM) was used to determine the
diameter of this NW to be 78 nm, as shown in Figure 4-1(d). Figure 4-1(e) shows the
Raman spectra of an individual Si NW before and after laser annealing at 50 mW, while
the laser polarization is aligned along the NW axis. In the following, we align the laser
polarization along the axis of the NWs, unless stated otherwise. The laser annealing
process is carried out in an argon environment at 1 atmosphere to prevent oxidation
during annealing. Figure 4-2(a) shows the evolution of the Raman spectra taken after 30-
second irradiations with different laser powers at the same focused laser spot. Before
annealing, a broad peak centered at 480 cm
-1
, the Raman signature of amorphous silicon,
is observed in this NW. As the incident laser power is increased, this peak can be seen to
upshift and become more narrow due to the reduction in phonon confinement effects[48,
57, 92, 113, 126, 171]. The raw spectra in Figure 4-2(a) were fitted with two Lorentzian
peaks, as shown in the inset of Figure 4-2(b). The broad peak is attributed to amorphous
Si, and the sharp peak is attributed to crystalline Si. The crystalline fraction can be
obtained from the relative peak intensities by following the equation:

⁄ );
where and are the integrated Raman intensities of the crystalline and amorphous
components, respectively. is the scattering ratio of crystalline and amorphous silicon,
which ranges from 0.8 to 1[6, 97, 143, 176]. We use 1 here. In Figure 4-2(b), the
crystallinity increases with increasing annealing power during the crystallization process.
67

Also, the lineshape and frequency of the crystalline component becomes more narrow
and upshifts as the annealing power is increased.

Figure 4-2. (a) Raman spectra before and after laser annealing at different powers at the
same laser spot. These spectra have been artificially offset for clarity. (b) Crystalline
fraction X
c
, (c) Raman linewidth, and (d) Raman shift relative to single crystal bulk and
estimated mean crystalline size after laser annealing. The inset in (c) shows the
crystalline fraction plotted as a function of laser irradiation time at two constant laser
powers.

(a)
0 1020 30 405060 70
0.0
0.3
0.6
0.9

Crystalline Fraction X
C
Annealing Power (mW)
400 60 0 8 00

R a m an S h ift (cm
-1
)
afte r 55 m W
afte r 4 5 m W
afte r 35 m W
afte r 2 5 m W
before la se r
annealin g
Intensity (a.u.)
afte r 6 5 m W
300 400 500 600

Intensity (a.u.)
Raman Shift (cm
-1
)
a-Si
c-Si
fitting result
raw data
20 30 40 50 60 70
8
12
16
20
24

Linewidth (cm
-1
)
Annealing Power (mW)

20 30 40 50 60 70
0
2
4
6
8

Annealing Power (mW)
  (cm
-1
)
8
16
24
32
40
Crystalline size (nm)
0 100 200 300 400 500
0.0
0.2
0.4
0.6
0.8

X
c
10mW
50mW
Irradiation Time (sec)
(b)
(c) (d)
68

These observations indicate that the crystalline core of the NW grows, while the
amorphous shell becomes thinner due to the crystallization process. Note that the
crystalline Raman peak becomes visible after laser annealing at 20 mW but is
downshifted from the single crystal value to 513 cm
-1
and significantly broadened to 20
cm
-1
. We attribute this to the effect of phonon quantum confinement because of the small
crystalline core. As the crystalline core increases due to successive annealing, the
crystalline peak blue shifts to 520 cm
-1
, while the linewidth narrows to 7 cm
-1
, which is
closer to the values of single crystal bulk Si. The linewidth and Raman shift ∆ of the
crystalline peak are plotted as a function of annealing power in Figures 4-2(c) and 4-2(d),
where ∆ is the peak downshift from 520.5 cm
-1
. The crystalline size D can be calculated
from the bond polarization model: ∆ = B (a/D)

, where B=2.0 cm
-1
, =1.44, and a is the
silicon lattice constant (a=0.543 nm)[122]. Various models[48, 57, 92, 113, 126, 171]
have been developed to obtain the mean crystal grain size from these Raman features.
However, they are all in agreement when the crystal size is above 3 nm[48].
The mechanism of CW laser-induced crystallization is different from that of a
pulsed laser. The crystallization takes place in the solid phase under CW laser irradiation.
The annealing temperature depends on the laser power and the heat loss to the substrate.
The crystalline core acts as a seed for crystallization and the growth of crystalline core is
expected to depend on the annealing time at a given temperature. However, the
crystalline fraction of the annealed NWs does not show any time dependence, as shown
in the inset of Figure 4-2(c). This is because the irradiated nanowire reaches thermal
69

equilibrium on a time scale of microseconds, and the resulting changes in crystalinity
take place far below our minimum time resolution.

Figure 4-3. (a) Raman spectra before and after focused laser annealing. (b) Integrated
intensity of crystalline peak ( ) as a function of position, plotted together with the
intensity of amorphous peak ( . (c) Raman intensity mapping along the NW.

Figure 4-3(a) shows the Raman spectra of a core-shell Si NW before and after
annealing with 50 mW. A Raman spatial map scan is used to resolve the crystallinity
changes along the NW after the single spot laser annealing. The integrated crystalline
300 400 500 600

Intensity (a.u.)
Raman Shift (cm
-1
)
before
after
-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6

Crystalline
Amorphous
Position ( m)
Raman Intensity (a.u.)

-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6
300
400
500

Raman Shift (cm
-1
)
Position ( m)
680 nm
(a) (b)
(c)
70

peak intensity plotted as a function of position gives a full width at half maximum of
680 nm by fitting with a Lorentzian function. This is slightly larger than the nominal laser
spot size (500 nm) obtained with the 100X objective lens. We believe that the slightly
wider profile of the crystalline peak intensity is due to the finite thermal gradient
produced while annealing, which results in different levels of crystallization that have
occurred near the annealing spot. The integrated amorphous peak intensity follows the
opposite trend of the crystalline peak, as expected in the crystallization process. Further
spatial mapping of the Raman intensity is plotted in Figure 4-3(c), which shows the
modulations of both a-Si and c-Si peaks along the length of the NW.
Nanowires were also deposited on 100 nm thick silicon nitride (Si
x
N
y
)
membranes (SPI, inc), enabling HRTEM imaging to be taken after laser annealing.
Figure 4-4(a) shows a TEM image of a locally annealed NW. Although the same laser
conditions (50 mW with polarization along NW axis) were used, the annealed region is
twice as large as that of the NW deposited on the glass substrate. This is because the heat
conduction through the Si
x
N
y
membrane is far less than that of the glass substrate, which
also increases the amorphous-crystalline interface to 0.5 µm in length. The drawing in
Figure 4-4(b) illustrates the cross section at the boundary between the unannealed and
annealed regions. The unannealed region remains as a core-shell structure, as seen in
Figure 4-4(c). Here, the crystalline structure can only been seen near the core of the NW.
However, in the annealed region (Figure 4-4(d)), the crystalline core fills nearly the entire
NW cross section, with a thin silicon oxide layer. The same uniform crystalline structure
as in Figure 4-4(d) is observed within 1 µm of the annealed region, indicated by a gray
71

box in Figure 4-4(a). The axial modulation in crystal structure is also reflected in the
spatial Raman mapping (intensity, Raman shift, and linewidth), as shown in Figures 4-
4(f) and 4-4(g).

72

Figure 4-4. (a) TEM image of a locally annealed Si NW in which the annealed area is 1
µm indicated by the box. (b) Schematic diagram of the NW boundary between annealed
and unannealed regions. Corresponding HRTEM images in 4c and 4d are indicated by the
dashed boxes. (c) HRTEM image of the unannealed area near the core of the NW. (d)
HRTEM image of the annealed segment near the NW surface. (e) Stokes and anti-Stokes
Raman spectra taken during the laser annealing process. (f) Crystalline peak intensity, (g)
Raman shift and linewidth plotted as a function of position along the NW.
(a)
10 nm
Oxidation layer
NW surface
10 nm
(b)
(d)
(e)
(f) (g)
-750 -500 500 750
0.0
5.0x10
3
1.0x10
4
1.5x10
4

Intensity (counts)
Raman Shift (cm
-1
)
S
AS
T
AS/S
=1103 K
shift =496.6 cm
-1
linewidth=18 cm
-1
(c)
-2 -1 0 1 2

Raman Intensity (a.u.)
Position ( m)
-2 -1 012
508
510
512
514
516
518
520

Position ( m)
Raman Shift (cm
-1
)
5
10
15
20
25
Linewidth (cm
-1
)
73

The uniformity of Raman shift and linewidth within the annealed region agrees
with the TEM inspection and further corroborates the high crystal quality. The Raman
Stokes (S) and anti-Stokes (AS) spectra shown in Figure 4-4(e) were taken during the
laser annealing process. These Raman modes downshift and broaden to 496.6 cm
-1
and
18 cm
-1
due to the high temperature reached[21, 135, 149]. The annealing temperature is
estimated to be 1103 K, based on the AS/S intensity ratio, I
AS
/I
S
=Ce
-
ħ

0
/k
B
T
, where C is a
calibrated coefficient from room temperature data, ħ 
0
=64.6 meV is the phonon energy,
and k
B
is Boltzmann's constant.

Figure 4-5. (a) Experimental setup of polarization-dependent laser heating on crystalline
NWs. (b) AS/S intensity ratio and corresponding temperature as a function of
polarization angle. (c) Raman intensity, linewidth, and Raman shift plotted as a function
of polarization angle with respect to the NW axis.

(a)
-120 -60 0 60 120 180
0.00
0.05
0.10
0.15

AS/S intensity ratio
Temperature
300
350
400
450
500
550

Temperature (K)
Polarization Angle
AS/S Intensity Ratio
(b)
-120 -60 0 60 120 180
517
518
519
520
521
5
6
7
8
9

Shift (cm
-1
)
Polarization Angle

Width (cm
-1
)

Raman Intensity
(c)
74

We also observe a strong polarization-dependence of the laser-induced heating.
The Raman intensity follows the expression , where is the angle
between laser polarization and NW axis as shown in Figure 4-5(a). The temperature
profile obtained from the Raman AS/S intensity ratio over a wide range of polarization
angles, in Figure 4-5(b), shows a strong preferential heating effect when the laser
polarization is aligned at =0
o
and 180
o
. There is a maximum temperature difference of
200 K under 40 mW irradiation. In Figure 4-5(c), the Raman linewidth oscillates between
5 to 8 cm
-1
and the Raman shift varies from 520 to 517 cm
-1
can also be understood by the
preferential heating. The polarization dependence of the Raman intensity has been
reported in crystalline silicon films[80, 97]. The intensity of Raman scattering depends on
the orientation of the incident laser polarization relative to the crystal axes following by
the simplified expression  2  , where  is the angle between laser
polarization and crystal axes. However, we do not observe other oscillation patterns of
Raman intensity rather than the one in Figure 4-5(c), which implies that the former effect
dominates in the NW system. The polarization dependence of laser heating is attributed
to the different degrees of absorption of the polarized laser. A recent finite-difference
time-domain (FDTD) study on light absorption in a 50 nm Si NW shows that the
absorption per unit volume of transverse magnetic (TM) light is almost 20-fold larger
than that of transverse electric (TE) light at a wavelength of 532 nm[71].
In Figure 4-6, we demonstrate the effect of polarization-dependent annealing on a
12 µm-long core-shell NW with a fixed laser power of 50 mW at various laser
polarization angles specified on the top of Figure 4-6(a). Each annealing spot is separated
75

by 1 µm along the wire axis to avoid proximity heating from the Gaussian laser beam.
Prior to annealing, spatial Raman mapping was performed and correlated with the
polarization angles to reveal the uniformity of crystallinity over the length of the NW.
After preferential annealing, the same spatial Raman mapping was performed again to
determine the crystallinity modulation along the NW, as plotted in Figure 4-6(c). In
Figure 4-6(b), visible a-Si peaks are observed at integer position numbers, which are
unannealed spots. The intensity modulation of a-Si peaks is merely due to the
polarization angle effect described in the previous paragraph. At the annealed spots (half
integer position numbers), the a-Si peaks either diminish or become very weak. For these
c-Si peaks, there are two factors responsible for the intensity modulation. Besides the
polarization angle effect, the crystalline fractions after annealing are different, as depicted
in Figure 4-6(a). The drawing illustrates the crystal structure modulations in both axial
and radial directions. Figure 4-6(c) also shows the corresponding annealing temperatures
as a function of polarization angle. The improvement of crystallinity shows strong
polarization dependence due to preferential heating.
76

Figure 4-6. (a) Schematic diagram of polarization dependent laser-induced preferential
annealing. Blue areas represent crystalline silicon. (b) Spatial Raman intensity mapping
of a laser annealed NW at various polarization angles. (c) Crystalline fraction X
c
before
and after annealing and calibrated annealing temperature plotted as a function of
polarization angle and position.

4.5 Conclusion
In conclusion, Raman spectroscopy was used to study the effect of laser annealing
on crystalline-amorphous core-shell Si NWs grown by the CVD method. The crystalline
fraction was found to increase gradually with increasing laser power based on the local
Raman spectra. The Raman shift and linewidth show upshifting and sharpening,
respectively, due to the reduction of phonon confinement effects, as the crystalline core
01 234 567 89
450
475
500
525
100 60 70 0 30 10 50 20
Polarization Angle

Raman Shift (cm
-1
)
Position ( m)
40
012345 6 789
780
840
900
960
1020
1080

after
Position ( m)
Annealing Temperature (K)
before
0.1
0.6
0.8
1.0
Crystalline Fraction X
c
10 20 30 40 50 60 70 100 0
Polarization Angle
(a)
(b)
(c)
0
o
10
o
20
o
30
o
40
o
50
o
60
o
70
o
100
o
77

of the NW increases. Strong preferential heating due to the laser polarization was utilized
to control the crystallinity of individual NWs. Using the focused laser annealing, we are
able to create an amorphous-crystalline interface less than 0.5 µm in length.

78

Chapter 5: Electrical and Optical Characterization of Surface
Passivation in GaAs Nanowires

5.1 Abstract
In this chapter, we report a systematic study of carrier dynamics in Al
x
Ga
1-x
As-
passivated GaAs nanowires. With passivation, the minority carrier diffusion length (L
diff
)
increases from 30 nm to 180 nm, as measured by electron beam induced current (EBIC)
mapping, and the photoluminescence (PL) lifetime increases from sub-60 ps to 1.3 ns. A
48-fold enhancement in the continuous-wave PL intensity is observed on the same
individual nanowire with and without the Al
x
Ga
1-x
As passivation layer, indicating a
significant reduction in surface recombination. These results indicate that, in passivated
nanowires, the minority carrier lifetime is not limited by twin stacking faults. From the
PL lifetime and minority carrier diffusion length, we estimate the surface recombination
velocity (SRV) to range from 1.7 10
3
to 1.1 10
4
cm s
-1
, and the minority carrier mobility
 is estimated to lie in the range from 10.3 to 67.5 cm
2
V
-1
s
-1
for the passivated nanowires.

5.2 Electron Beam Induced Current (EBIC) Measurement
Electron beam induced charge collection was first observed by Everhart during
his graduate study at Cambridge University in 1958. Later on, this technique has been
widely applied in semiconductor analysis, including identifying buried junctions and
79

defects, or carrier lifetime. Charge carriers in semiconductors are created by the collision
of an energetic primary electron with valence electrons. The energy lost of primary
electron produces an electron in the conduction band and a hole in the valence band. The
charge carriers travel at velocities v, given by

, where k is Boltzmann
constant and T is temperature. Their motion can be described as diffusive with a diffusion
coefficient, D
e
or D
h
, for each carrier type. For GaAs, D
e
= 220 cm
2
/sec and D
h
= 10
cm
2
/sec[88]. During the diffusion, recombination of excess electrons and holes occurs
when one or the other carrier is trapped at a specific point in the crystal. An internal
electric field is needed to segregate the electrons and holes before recombination, such as
p-n junction or Schottky junction. Figure 5-1 shows the band bending of a metal-
semiconductor interface due to the Fermi level alignment. The build-in electric field
within the depletion region (W
0
) can separate the electron-hole pairs created by energetic
electron beam. Similar electric field also forms in p-n junction.

Figure 5-1. Band diagrams of a metal, a semiconductor, and a metal-semiconductor
interface.
metal semiconductor
vacuum
E
F E
V
E
F
E
C
E
V
E
F
E
C
80

Hoffmann and his coworkers used EBIC measurement to resolve the p-n junction
in silicon nanowires prepared by ion implantation technique[63]. Instead of conventional
lithography to make electrical contacts to the nanowires, a conductive PtIr tip attached to
a nanomanipulator is used to directly contact the nanowires in the SEM, as shown in
Figure 5-2.

Figure 5-2. (a) SEM image of as-grown silicon nanowire and schematics of nanowire
before and after ion implantation. (b) Correlated SEM and EBIC images [63].

EBIC measurement is also used to study the minority carrier diffusion length
(L
diff
) of Au-catalyzed silicon nanowires. Au impurities in silicon nanowires have been
thought to introduce mid-gap traps in bulk silicon that act as recombination sites for
electrons and holes. However, the L
diff
shows a strong dependence on nanowire diameter,
as shown in Figure 5-3. The L
diff
increases from 25 nm to 80 nm as nanowire diameter
increases from 30 nm to 100 nm, and is independent with the doping level. This implies
that nanowire surface controls the L
diff
rather than Au impurities[2].
(b) (a)
81

Figure 5-3. (a, b) False-color SEM and EBIC images. (c) EBIC profiles along (solid line)
and perpendicular to (dashed line) the nanowire axis. (d) Minority carrier (hole) diffusion
length as a function of nanowire diameter[2].

We modified the conventional SEM (JOEL 6610) to measure EBIC signal by
using an Ithaco current preamplifier that connects to the source and drain of a device
through the vacuum electrical feedthrough, as shown in Figure 5-4. This device consists a
Schottky interface that can separate electron beam induced electron-hole pairs to form
current. The output signal of preamplifier is sent back to video channel of the SEM and
the correlated SEM and EBIC images can be obtained simultaneously.
(b)
(c)
(d)
(a)
82

Figure 5-4. (a) JOEL 6610 SEM in Center of Electron Microscope and Microanalysis at
USC. (b) Inside of SEM chamber with homemade chip carrier and wring. (c) Side-view
of wire-bonded sample. (d) Schematic of measurement setup.

5.3 Time-Resolved Photoluminescence (PL) Measurement
The idea of time-resolved photoluminescence is recording the time dependent
intensity profile of the emitted light upon the excitation by a laser pulse. By analyzing the
intensity decay, the lifetime of radiative recombination can be obtained. However, there
are several practical problems in this measurement. For instance, there might be only a
(a) (b)
(c) (d)
83

few photons emitted from a single excitation, or the time interval between emitted
photons is very short. In fact, the interval could be less than hundred picoseconds. Single
photon counting has been used to solve these problems. Instead of only one excitation, a
repetitive excitation is applied; meanwhile, photomultiplier tube (PMT) is used to collect
single photons with precisely timed registration, as illustrated in Figure 5-5(a).

Figure 5-5. (a) Schematic diagram of single photon counting technique. (b) Exponential
decay of photoluminescence lifetime histogram. [PicoQuant, 2009]

Figure 5-6. Schematic diagram of time-resolved PL setup at Aerospace research lab.
(a) (b)
sample
Substrate
Lens
Lens
Longpass
filter
Tsunami fs laser
Trigger
84

5.4 Experimental Results and Discussion
GaAs is one of the most widely used semiconductors, second to silicon, with
applications in fast electronics, infrared LEDs[162], and high-efficiency solar cells[53].
However, GaAs suffers from pronounced effects associated with its surface states, which
has prevented GaAs metal-oxide-semiconductor field-effect transistors (MOSFET) from
becoming a viable technology. This problem is also reflected in GaAs' extremely high
surface recombination velocity (10
6
cm/s), which is three orders of magnitude higher than
most other III-V semiconductors[96, 118]. The surface depletion effect in GaAs is
exacerbated in nanostructures with high surface-to-volume ratios. For example, semi-
insulating electrical behavior was observed in highly doped GaAs nanowires without
surface treatment[79]. Ammonium polysulfide (NH
4
)
2
S
x
has been used to passivate the
surfaces of III-V semiconductors with covalently bonded sulfur atoms[139, 151].
However, sulfur-passivation only provides short-term surface stability. Passivating GaAs
nanowires with a wide band gap semiconductor such as Al
x
Ga
1-x
As provides long-term
surface stability[37, 121, 123, 124, 161].
Here, we present a systematic study of Al
x
Ga
1-x
As-passivated GaAs nanowires
using spatial mapping of Raman and photoluminescence (PL) spectroscopy, time-
resolved photoluminescence (TRPL) spectroscopy, and electron beam induced current
(EBIC) mapping. These measurements directly probe the minority carrier diffusion
length and lifetime in these nanowires. The surface recombination velocity and carrier
mobility are also calculated based on the experimental results.
85

As mentioned above, the problems associated with surface states and surface
depletion are more severe in GaAs nanowires because of their high surface-volume
ratios. In moderately doped nanowires, the depletion region will consist of a cylindrical
ring with a conducting channel in the middle of the nanowire. If the doping is too low,
however, the entire nanowire cross section will be depleted, and therefore, insulating.
Figure 5-7 shows the free carrier density plotted as a function of the dopant impurity
concentration for several nanowire diameters calculated by solving the Poisson equation
with Fermi-Dirac statistics, assuming mid-gap pinning of surface Fermi level. For a 100
nm diameter nanowire, doping concentrations below 10
17
cm
-3
, yield completely depleted
nanowires. Because of these surface states, dopant impurity concentrations above
approximately 7 10
17
cm
-3
are needed in order to generate a significant amount of free
carriers in this material. There are two ways to mitigate this problem. First, the nanowires
can be heavily doped, N
I
≥ 10
18
cm
-3
, or the GaAs surface can be passivated with a wider
bandgap semiconductor, such as Al
x
Ga
1-x
As or GaP.
86

Figure 5-7. The calculated free carrier density plotted as a function of dopant impurity
concentration for GaAs nanowires with various diameters.

The synthesis of GaAs nanowires has been initiated more than ten years ago with
several different methods[43, 44, 90, 120, 140, 141, 163]. In this study, GaAs nanowires
are synthesized by metal organic chemical vapor deposition (MOCVD) with selective
area growth (SAG)[25, 161]. Trimethylgallium (TMGa), trimethylaluminum (TMAl),
and arsine are used as precursors for Ga, Al, and As deposition. High density arrays of
GaAs nanowires are grown along the (111) direction on silicon substrates. A thermally
grown silicon oxide layer is used as a mask for the SAG growth. A Raith electron beam
lithography (EBL) system is used to pattern a 1 mm 1 mm array of holes with 600 nm
pitch. A short (20-30 second) buffered HF etch is performed to expose the crystalline
silicon surface before loading the sample into the MOCVD reactor. The sample is first
10
15
10
16
10
17
10
18
10
19
10
5
10
7
10
9
10
11
10
13
10
15
10
17
10
19
N
D
d=100nm
d=200nm
d=300nm
d=400nm
d=500nm
Free Carrier Density (cm
-3
)
Impurity Concentration (cm
-3
)
87

annealed in hydrogen at 920
o
C for 5 minutes to remove the native oxide. Arsine flows
while the temperature is cooled from 850
o
C to 440
o
C. Nucleations of GaAs are grown at
440
o
C for 8 minutes with partial pressures of 3.74 10
-7
atm and 4.78x10
-5
atm for TMGa
and arsine, respectively. After nucleation, the temperature is increased to 790
o
C for
nanowire growth with the same partial pressures of TMGa and arsine. The growth rate is
approximately 8.33 Å/s. Using these conditions, nanowires are grown with very uniform
cross-section along the length of each nanowire up to 10 m long. After GaAs nanowire
growth, an Al
x
Ga
1-x
As layer is deposited as a passivation layer with partial pressures of
3.7410
-7
atm, 4.78 10
-5
atm, and 2.18 10
-7
atm for TMGa, arsine, and TMAl,
respectively. An additional thin layer of GaAs (partial pressures of 1.41 10
-5
atm for
TMG and 4.78x10
-4
atm for arsine) is grown to prevent oxidation of the Al
x
Ga
1-x
As shell.
For individual nanowire measurements, the nanowires were transferred onto a SiO
2
/Si
substrate with lithographically-defined grid markers, which enables us to record the
location of individual nanowires for Raman spectroscopy and photoluminescence
mapping[18]. In order to measure the minority carrier diffusion length, the GaAs
nanowire was contacted by two metal electrodes using EBL. Before evaporation of the
metal contacts, a short oxygen plasma was performed to remove the residual resist and
oxidize the outer shell (Al
x
Ga
1-x
As and GaAs). The oxidized shell was then removed
using a mixture of HCl:H
2
O (1:1). Ohmic contacts can be easily obtained between
GeAu/Ni/Au metal contacts and bulk GaAs after annealing. GaAs nanowires, however,
are completely dissolved in the electrodes at these high annealing temperatures, and the
88

lower annealing temperatures required result in non-linear, high resistance Schottky
contacts.
Micro-Raman spectroscopy and micro-PL mapping are performed using a
translation stage with 100 nm step resolution and a high numerical aperture objective lens
(100X). A silicon CCD detector is used to detect photoluminescence in the range from
500 to 900 nm and Raman shift from 150 to 3200 cm
-1
. Low power (<0.5 mW) excitation
by a continuous 532 nm laser is used to avoid optical heating of the nanowires. EBIC
measurements are carried out in a JEOL JSM-6610 scanning electron microscope (SEM)
equipped with a voltage source (Keithley 2400) and a low noise current preamplifier
(Ithaco 1201). In time-resolved PL measurements, sample excitation was carried out with
a pump pulse (Tsunami fs laser) of center wavelength 800 nm and pulse energy 32 pJ.
The PL signal was detected by a streak camera (Hamamatsu C5680) with an extended
response NIR streak tube. The photon collection was centered at 860 nm with a
bandwidth of 75 nm. The data integration time was 6000 seconds, and the minimum
system response is 60 ps.
Figure 5-8(a) shows an SEM image of a 7 m long Al
x
Ga
1-x
As-passivated,
undoped GaAs nanowire. The diameter of the nanowire is tapered from the tip to the base
of the nanowire due to the inhomogeneity of the Al
x
Ga
1-x
As passivation layer. The
tapered structure of the passivated nanowires was studied by atomic force microscopy
(AFM), as shown in Figure 5-9 which shows that the Al
x
Ga
1-x
As coating varies from 0 to
11 nm in thickness. This thickness variation was also verified by high resolution
89

transmission electron microscopy (HRTEM), as shown in Figure 5-10. The Raman
intensity ratio of the AlAs peaks (337 and 371 cm
-1
) to that of GaAs (266.5 cm
-1
) gives a
relative measure of the Al
x
Ga
1-x
As thickness. This ratio changes from 0.7 at the tip of the
nanowire to almost zero (no passivation) at the base, as shown in Figure 5-8(c) (right
axis). Based on these Raman spectra, there is no change in the composition of Al
x
Ga
1-x
As
along the length of the nanowire. The aluminum composition x can be obtained from the
vibrational frequency of the AlAs-like LO Raman mode via the following relation:

364.7 46.7 9 .4
[144, 173]. Raman spectrum shown in Figure 5-
11(b), we found a composition of 14.5% aluminum, which is close to the value of 12%
characterized by X-ray diffraction measurements presented in a previous publication
using similar growth conditions[120]. The EDX spectrum of passivated nanowires, as
shown in Figure 5-11(a), provides the lower limit of aluminum composition (7.5%) in
Al
x
Ga
1-x
As shell, since the GaAs core also contributes to the EDX signal. The PL spectra
also show a strong position dependence due to the varying degree of passivation along
the length of the nanowire. Figure 5-8(b) shows the photoluminescence spectra taken at
both ends of the nanowire, which show a 48-fold enhancement in the PL intensity with
passivation. Surface states in the less-passivated regions form non-radiative
recombination sites causing most of the photo-excited carriers to recombine non-
radiatively at the nanowire surface, thereby lowering the PL intensity. Based on our
previous HRTEM studies[103], we found that the density of stacking faults are uniform
throughout the length of the bare nanowires. Also, since EDX measurements on bare
GaAs nanowires show an equal composition of Ga and As throughout the length of these
90

nanowrires, we believe that the large change in PL intensity is mainly due to the Al
x
Ga
1-
x
As passivation layer.

Figure 5-8. (a) SEM image of a tapered Al
x
Ga
1-x
As-passivated GaAs nanowire. (b)
Continuous-wave PL spectra taken at the tip and base of the nanowire in (a). (c, d)
Spatially mapped PL and Raman data along the nanowire axis plotted as a function of
position.
012 34567
843
846
849
852

Position (um)
PL Peak (nm)
56
58
60
62
64
66
PL Linewidth (nm)
750 800 850 900 950
0
20000
40000
60000
80000

PL Intensity (counts)
Wavelength (nm)
tip
base
X48
012 34567
0
10
20
30

base
PL Intensity
Raman Intensity Ratio
Normalized PL Intensity (a.u)
Position (um)
tip
0.0
0.2
0.4
0.6
0.8

AlGaAs/GaAs
Raman Intensity Ratio
(a) (b)
(d)
(c)
91

Figure 5-9. Diameter of GaAs nanowires as a function of position. Insets are SEM images
of passivated (top) and bare (bottom) nanowires. AFM height profiles at different
position for passivated (b) and bare (c) nanowires.
02 46 8 10
0
40
80
120
160
passivated NW
tip
base

AFM Height (nm)
X ( m)
0123 456
0
40
80
120
160
bare NW
tip base
AFM Height (nm)
X ( m)
(c)
(b)
(a)
92

Figure 5-10. (a) TEM image of Al
x
Ga
1-x
As-passivated GaAs nanowire slice in
longitudinal direction prepared by focused ion beam (FIB) technique. The sample
thickness is around 70nm. HRTEM images near the tip (b) and the base (c) of the
nanowire. We performed this experiment on a shorter sample than those in Figure 5-8
due to the limitation of our FIB technique. We still observed that the sample was bent
after FIB slicing, which results in invisibility of stacking faults at the tip of the nanowire.
(a)
(c)
(b)
93

Figure 5-11. (a) EDX and (b) Raman spectra of Al
x
Ga
1-x
As-passivated GaAs nanowires.

This PL enhancement indicates that these surface states have been successfully
passivated by the Al
x
Ga
1-x
As layer. In addition to enhancement in the PL intensity, we
also observe a blueshift (from 852 to 842 nm) and broadening (from 56 to 65 nm) in the
PL emission, as shown in Figure 5-8(d). Strain-induced optical band gap
0.3 0.6 0.9 1.2 1.5 1.8 2.1
0
100
200
300
AlK
GaL
AsL

Intensity (count)
Energy (keV)
OK
Si
Wt %
O K 0.66
AsL 36.07
GaL 56.12
Al K 7.46
250 300 350 400 450
AlAs TO
* Si substrate
AlAs LO
GaAs LO

Intensity (a.u.)
Raman Shift (cm
-1
)
GaAs TO
AlAs-like LO
(a)
(b)
94

modulation[115] is excluded here in accordance with the spatially mapped Raman data,
which do not show any shift in the Raman modes of the GaAs. The GaAs Raman peak
remains almost constant at 266.5 cm
-1
throughout the length of the nanowire. Thus, we
attribute the broadening and blueshift to filling of the conduction band with free
carriers[35]. Without Al
x
Ga
1-x
As passivation, most of the free carriers are depleted in the
GaAs core, resulting in band edge-to-band edge PL emission. After the surface states are
passivated, free carriers begin filling up the conduction band, causing a blueshift in the
PL emission[35]. Since the amount of blueshift depends on the number of free carriers,
the inhomogeneity in surface passivation leads to broadening of the PL peak. A similar
effect was observed by Titova et al. in InP nanowires, which showed a broadening and
blueshift in PL emission under illumination due to the presence of a high-density
electron-hole plasma[158]. In addition to PL mapping, we also apply
cathodoluminescence (CL) to examine the passivated nanowires, as shown in Figure 5-
12. Instead of using 532 nm focused laser (spot size ~ 0.5 m), focused electron beam (1
keV) is used to excite the electron-hole pairs in the nanowires with much smaller
interaction volume (less than 100 nm), which gives better spatial resolution. The CL
result make a good agreement with PL data. Brighter at one end and dimmer at the other.
95

Figure 5-12. (a) SEM image of a Al
x
Ga
1-x
As-passivated GaAs nanowire. (b) the
corresponding CL mapping.

Figure 5-13(a) shows an SEM image of a GaAs nanowire with a metal (Cr/Au)
electrode patterned on top using EBL. The electrode forms a Schottky contact at the
nanowire surface, and the current-voltage characteristics show a typical non-linear
behavior. Figure 5-13(c) shows a schematic diagram of the EBIC measurement
technique, where the focused electron beam creates electron-hole pairs in a p-type
nanowire. Electron-hole pairs generated within one minority carrier diffusion length of
the metal-semiconductor Schottky junction will, on average, result in a measurable
current. Minority carriers generated far away from the Schottky junction will recombine
and, therefore, not contribute to the measured (EBIC) current. By spatially mapping the
EBIC current, we can determine the minority carrier diffusion length directly. Figure 5-
13(b) shows the EBIC map corresponding to the SEM image in Figure 5-13(a).
(b)
(a)
96

Here, an acceleration voltage of 5 kV was used to create electron-hole pairs in this
measurement. The EBIC signal is strongest near the Schottky interface and gradually
diminishes away from the contact. Figure 5-13(d) shows the EBIC intensity profile
plotted along the nanowire axis. The minority carrier diffusion length, L
diff
, is extracted
by fitting with an exponentially decaying function. For passivated nanowires, we find that
L
diff
= 180 nm, while for unpassivated nanowires L
diff
= 30 nm. This value of 30 nm
corresponds to the minimum resolution of our EBIC system. Despite this six-fold
increase in L
diff
with passivation, 180 nm is a relatively short diffusion length, roughly on
the scale of the diameter (90 nm). Dense twin stacking faults have been observed in these
nanowires by high resolution transmission electron microscopy, as shown in Figure 5-14,
which are expected to affect the electron transport. However, the observed L
diff
= 180 nm
is more than two orders of magnitude longer than the average separation between twin
stacking faults, indicating that they do not form strong centers for electron-hole
recombination.
97

Figure 5-13. (a) SEM image of an Al
x
Ga
1-x
As-passivated GaAs nanowire device, and (b)
corresponding EBIC image. (c) Schematic diagram illustrating the electron beam induced
current measurement. (d) EBIC profiles along the nanowire axis for passivated and bare
nanowires.
0 200 400 600
0.0
0.4
0.8

L
diff
=30nm
e
-x/L
diff
passivated NW
unpassivated NW
Normalized EBIC
Position x (nm)
L
diff
=180nm
(a) (b)
(c)
(d)
98

Figure 5-14. GaAs nanowires grown on (a) a GaAs (111) surface and (b) a silicon (111)
surface. Both scale bars are 20 nm. The lines perpendicular to the nanowire axis are twin
sacking faults.

We also perform time-resolved photoluminescence spectroscopy in order to
determine the carrier lifetimes in passivated and unpassivated GaAs nanowires. For these
measurements, shorter nanowires (1.5 m) were grown to mitigate the effects associated
with non-uniform passivation. This data is shown in Figure 5-15. For unpassivated
nanowires, the PL lifetime is shorter than the instrument response time of our setup (60
ps). However, passivated nanowires exhibit a significantly longer lifetime due to the
passivation of surface states. In addition to the instrument component, two time constants
are needed to fit the TRPL curve. The average time constants are 0.2 ns for the short
lifetime and 1.3 ns for the long lifetime. Discrepancies between the lifetimes measured in
different regions of the nanowire array are due to the inhomogeneous crystal quality and
passivation, as has been seen in single nanowire TRPL measurements[124]. In
passivated, highly doped nanowires, both the photoluminescence intensity and lifetime
significantly decrease due to the increase of impurity scattering, resulting in average time
(a) (b)
99

constants of 0.1 ns and 0.9 ns for the short and long lifetimes, respectively. This dataset
indicates that, for the doped, passivated nanowires, the minority carrier lifetime is not
limited by surface recombination, since the doped nanowire exhibits a shorter PL lifetime
than the undoped nanowire.

Figure 5-15. Time-resolved photoluminescence spectra of passivated and unpassivated
nanowires with different impurity concentrations.

Based on the measured carrier lifetimes determined from TRPL, the surface
recombination velocity S can be estimated by the following equation[34, 36]
1
1
4

0.0 0.5 1.0 1.5 2.0

unpassivated
n-doped
undoped
PL Intensity (counts)
Time (ns)

TR-PL
AlGaAs-passivated
100

where  is the effective carrier lifetime, 
b
is the carrier lifetime in bulk material, and d is
the diameter of the GaAs core (90 nm). Using a bulk carrier lifetime of 
b
= 1.3 s, we
calculate the surface recombination velocity S to be 1.7 10
3
cm s
-1
and 1.1 10
4
cm s
-1
for
the long and short decay constants[118]. For bulk GaAs with appropriate passivation,
SRV as low as 500 have been reported[117]. The relatively high SRV values observed in
the nanowire samples indicate that either the surface states have only been partially
passivated or that stacking faults and bulk impurities are limiting the SRV[118]. This is
confirmed by the highly doped nanowires, which show a significantly reduced PL
lifetime compared to the corresponding undoped sample. The minority carrier lifetime
and surface recombination velocity (SRV) of twin-free Al
x
Ga
1-x
As passivated GaAs
nanowires grown by the vapor-liquid-solid (VLS) growth method were reported by three
other groups ranging from 1 to 2.5 ns and around 3 10
3
cm/sec, respectively[9, 37, 77,
124]. While dense twin stacking faults are normally seen in our nanowires, using
catalyst-free selective area MOCVD growth, we did not observe a significant difference
in the lifetime and SRV compared to the numbers reported in twin-free nanowires. This
indicates that these twin stacking faults are not the main factor limiting minority carrier
dynamics. We can also correlate L
diff
with  to estimate the minority carrier mobility  by
the following equation

√
where the diffusion coefficient is given by D= kT/e and both L
diff
and  are measured at
room temperature. The short and long decay constants give higher and lower limits of the
101

minority carrier mobility  to be 67.5 and 10.3 cm
2
V
-1
s
-1
, respectively, for the passivated
nanowires. These mobilities are far lower than the values in bulk GaAs, which have been
both experimentally and theoretically reported in the range from 1000 to 7500 cm
2
V
-1
s
-1
depending on the acceptor concentration[62, 169].

5.5 Conclusion
In conclusion, a systematic study of Al
x
Ga
1-x
As-passivated GaAs nanowires
shows the significance of surface passivation on free carrier dynamics in GaAs
nanowires. Weak PL emission and short PL lifetimes are observed in bare (unpassivated)
GaAs nanowires. We observe a 48-fold enhancement in the PL intensity and a six-fold
increase in the minority carrier diffusion length with surface passivation. The surface
recombination velocity is calculated to lie in the range from 1.7 10
3
to 1.1 10
4
cm s
-1
.
For the passivated nanowires, we estimate the minority carrier mobility to lie in the range
from 10.3 to 67.5 cm
2
V
-1
s
-1
, based on the measured lifetimes and diffusion length.
Furthermore, the relatively long minority carrier diffusion lengths indicate that twin
stacking faults do not limit the minority carrier lifetimes in passivated nanowires.

102

Chapter 6: Electrical and Optical Characterization of Twin-
Free GaAs Nanosheets

6.1 Abstract
Electrical properties of twin-free MOCVD-grown GaAs nanosheets are
investigated by secondary electron scanning electron microscopy (SE-SEM) and electron
beam induced current (EBIC) imaging. An abrupt boundary between an initial growth
region and an overgrowth region is observed in the nanosheets. The electron
concentration (N
D
) of the initial growth region is around 1 10
18
cm
-3
, as determined by
both Hall effect measurements and low temperature photoluminescence (PL)
spectroscopy. A built-in potential of 1.4 V is obtained from the rectifying current-voltage
curve across the boundary, indicating that the overgrowth region is p-doped with a hole
concentration (N
A
) of approximately 1 10
18
cm
-3
. The SEM contrast is attributed to the
doping reversion at the boundary, which exhibits a symmetric EBIC profile. Based on the
estimated carrier concentrations, the space charge region is less than 20 nm wide, which
is significantly shorter than the minority carrier diffusion length of 177 nm.

6.2 Material Contrast in Scicanning Electron Mroscopy
Scanning electron microscopy (SEM) has been widely used in material
characterizations but not limited to the morphology examination. It has been used to
103

image the composition and doping in semiconductor superlattices[47, 125, 137, 164]. The
atomic number affects the number of high-energy back scattering electrons (BSE) due to
the large angle Rutherford scattering, which also contributes to the emission of low-
energy secondary electrons (SE). The atomic number contrast has been used to resolve
the Al
0.5
Ga
0.5
As/GaAs superlattice, as shown in Figure 6-1. Furthermore, the SE emission
yield strongly depends on the doping level and type due to the surface Fermi level
pinning effect, as illustrated in Figure 2. In p-type materials, lower energy is need to
excite an electron from valence band to vacuum compared to n-type materials.
Furthermore, the build-in field near the surface of p-type semiconductors tends to
accelerate the secondary electrons out of the surface and the opposite field in n-type
semiconductors prohibits the electrons escaping from the surface. Under the same
electron irradiation, the SE emission in p-type materials is stronger than n-type
materials[125], therefore, p-type materials have brighter contrast than n-type materials.
Figure 6-3 shows SE-SEM images of GaAs-based heterostructure containing several p-
and n- layers. A wide range of acceleration voltages (1-20 keV) is used in this experiment,
in which 1 keV gives highest signal-to-noise ratio.
SE-SEM imaging has also been used to identify the electrical properties of carbon
nanotubes, either metallic or semiconducting, due to the surface potential difference[93].
In their study, carbon nanotubes are placed on the insulating substrate with electrical
contacts, as shown in Figure 6-4. The insulating substrate becomes positively charged
with 1 keV electron beam irradiation while the nanotubes remain relatively neutral
because the electrodes outside of the irradiated region serves as electron reservoirs. The
104

potential difference between the insulating surface and nanotubes results in voltage
contrast in SEM imaging. The metallic nanotubes with higher conductivity show uniform
contrast through the length of the nanotubes. However, gradual change in contrast of
semiconducting nanotubes is observed due to the lower conductivity.

Figure 6-1. 20 kV BSE image of 17-period Al
0.5
Ga
0.5
As/GaAs superlattice obtained from
Hitachi S-4500 FE-SEM[125].

Figure 6-2. Schematic diagrams of surface states induced band bending effects in n and p
type semiconductors.

n-type p-type
105

Figure 6-3. SE SEM images of GaAs-based heterostructure containing several p- and n-
layers. The p and n dopants were Be and Si, respectively, at concentrations of 2 10
18
cm
-
3
[125].

Figure 6-4. 1 keV SE image of carbon nanotubes on SiO
2
/Si substrate and schematic
illustration of the origin of contrast difference between metallic and semiconducting
nanotubes[93].

106

6.3 Experimental Results and Discussion
Semiconducting nanowires have drawn a lot of attention due to their potential
applications in solar cells[5, 27, 52, 59, 106], batteries[17, 33], and light emitting
diodes[56, 87]. The abilities of control and characterization the nanowires become
crucial. Various growth methods have been demonstrated and discussed. Either core-shell
or axial p-n junction nanowires can be synthesized by controlling dopant flow rates and
growth temperature[27, 55, 59, 60, 106, 155]. Breuer et al. claimed that the Au impurities
in GaAs nanowires grown by Au-catalyzed vapor-solid-liquid (VLS) method act as
recombination centers for free carries which could degrade the device performance[9].
Selective area growth metal organic chemical vapor deposition (SAG MOCVD) has been
used to eliminate this concern[19, 103]. However, the formation of twin stacking faults is
inevitable in nanowire structure. Very recently, Chi et al. have firstly demonstrated that
twin-free GaAs nanosheets can be grown by SAG MOCVD method along 112
on
GaAs substrates.
Here, we present a systematic characterization of the twin-free GaAs nanosheets.
SE SEM is used to observe the structure morphology and sharp boundary between initial-
and over-growth regions due to the doping reversion. The carrier type and concentration
of the initial-growth region are determined by low temperature photoluminescence (PL)
and Hall measurement. Based on the observation of rectifying current-voltage (I-V) curve
across the boundary and uniform signal of electron beam induced current (EBIC) along
the boundary, we attribute the boundary to a naturally formed p-n junction.
107

As mentioned above, the GaAs nanosheets are synthesized by SAG MOCVD.
Trimethylgallium (TMGa), arsine, and disilane are used as precursors for Ga, As, and Si
deposition. GaAs nanosheets are grown vertically along the (111) direction on GaAs
substrates. A plasma-enhanced CVD grown nitride layer is used as a mask for the SAG
growth. A Raith electron beam lithography (EBL) system is used to pattern 100 nm wide
and 5 m long slits. CF
4
reactive ion etching transfers the patterns from the resist to the
silicon nitride layer. The growth temperature of nanosheets is 790
o
C with the partial
pressures of 2.39 10
-6
atm, 3.74 10
-7
atm, and 7.14 10
-12
atm for arsine, TMGa, and
disilane. Note that the V/III ratio (6.38) used here is significantly lower than our previous
nanowire growth (127) condition[19]. Using these conditions, nanosheets are grown with
very uniform cross-section along (111) direction but form two self-terminated inclines, as
shown in Figure 6-5(a). We call the triangle with two bottom angles of 19
o
and 35
o
initial
growth. As the growth condition keeps going, additional growth starts from the apex of
the initial growth region, in which the growth mechanism is similar to nanowire growth
due to the small growth area. The twin formation becomes possible at this point. This
additional growth creates steps which facilitate the crystal growth on the inclines. Thus,
the growth rate and dopant incorporation may be different from the initial growth. We
call it overgrowth region as indicated by arrows in Figures 6-5(b) and 6-5(c).
Interestingly, multiple triangles are formed with further overgrowth and twin lines are
observed between these triangles. It could be either single twin or multiple twins. Odd
number of twins makes the triangle flip 180
o
with respect to the one at the bottom, as
108

shown in Figures 6-5(b) and 6-5(c). Actually, this phenomenon may shine some light on
the mechanism of twin formation in nanowires and more studies are needed.
The SEM contrast in Figure 6-5 can be discussed in two aspects. First, the whole
nanosheet is brighter than the substrate, GaAs substrate covered by an insulating layer.
This is because that the insulating layer is positively charged due to the electron beam
irradiation with low energy (less than 3 keV)[54]. The secondary electrons emitted from
surface are more than the incoming primary electrons. Note that the nanosheets are
connected to the underlying GaAs substrate and the substrate acts as an electron reservoir
for the nanosheets. Thus, the nanosheets remain relatively neutral compared to the
irradiated insulating layer. The surface potential is normally around few volts, which
hinders emission of secondary electrons out of insulating layer. This is called voltage
contrast SEM imaging. Second, there is a contrast difference within the overgrown
nanosheets, as shown in Figure 6-5(b) and 6-5(c). The initial growth region appears to be
darker than the overgrowth region above the terminated inclines. This implies that the
doping changes abruptly across the terminated inclines and form these boundaries.
Constant partial pressure of disilane is used intentionally to have n-type GaAs throughout
the whole growth, which makes the initial growth n-doped. However, a trace amount of
residual carbon impurities in TMGa may be responsible for the doping reversion across
the boundary due to the different growth mechanism.
109

Figure 6-5. SE SEM image of GaAs nanosheets grown on GaAs substrate with a nitride
layer as growth mask. Nanosheets in the status of just-terminated (a), beginning of
overgrowth (b), and further overgrowth (c).

(a)
(b)
(c)
110

We use low temperature PL and Hall effect measurement to determined the
carrier type and concentration of initial grown region. First, liquid nitrogen is used to cool
the sample down to 77 K in order to avoid thermal smearing effect and resolve the
effective optical band gap of the doped GaAs nanosheets. First, unintentional doped and
n-doped GaAs wafers with known carrier concentration are measured to calibrate our
system. The sharp PL spectrum in Figure 6-6(a) represents the unintentional doped GaAs
wafer. Free exciton transition contributes to the sharp peak and the longer wavelength
emission is attributed to band acceptor transitions involving carbon impurities. The PL
spectrum of a n-doped GaAs wafer with carrier concentration of around 1 10
18
cm
-3

blueshifts by 37 meV. More pronounced band acceptor transitions is seen in this sample.
Surprisingly, the PL linewidth of n-type nanosheets we grown is narrower than the
commercial wafer, indicating better crystal quality. Also, it blueshifts by 43 meV
corresponding to the carrier concentration of 1.2 10
18
cm
-3
based on the reported
model[35]. In addition, Hall effect measurements are used to determined the carrier
concentration of the nanosheets grown in the same batch. We measure Hall voltage (V
H
)
as a function of applied current (I) through the nanosheet under different magnitude of
magnetic fields (B), as shown in Figure 6-6(b). We estimated the concentration (n=N
D
) of
9.310
17
cm
-3
by following the equation: V
H
=-IB/ned, where d=150 nm, the thickness of
the nanosheet.
111

Figure 6-6. (a) PL spectra of GaAs nanosheet and commercial wafers at 77 K. The carrier
concentration of 1.2 10
18
cm
-3
is estimated. (b) Hall voltage as a function of applied
current in the nanosheet. Carrier concentration is 9.26 10
17
cm
-3
.
(a)
0 5 10 15 20
0.0
0.4
0.8
1.2

Hall Voltage (mV)
Current ( A)
0.90T
0.79T
0.68T
0.41T
0.18T
0.08T
(b)
760 800 840 880
n-NS
undoped wafer
n-wafer
PL Intensity (a.u.)
Wavelength (nm)
112

Figure 6-7. (a) SE SEM image of GaAs nanosheets contacted with metal electrodes.
Dashed lines indicate the boundary between initial- and over-growth. (b) Rectifying I-V
curve measured across the boundary. Inset shows linear IV when electrodes are both in
initial growth region.

-2 -1 0 1 2
0
1
2
3
4
I-V_1-2

I (nA)
V (V)
V
bi
=1.4 V
-0.08 -0.04 0.00 0.04 0.08
-1.0
-0.5
0.0
0.5
1.0

I (mA)
V (V)
I-V_2-3
(a)
(b)
113

The dimension of overgrowth region is much smaller than our focused laser spot
(1.25 m with 40X long working distance objective), which inhibits the precise PL
measurement. Multiple contacts on different regions of the overgrown nanosheets is used
to probe local I-V characteristics, as shown in Figure 6-7. In order to obtain ohmic
contacts, diluted hydrochloric acid (HCl:H
2
O/1:1) is used to etch the oxide layer for 20
seconds before metal deposition. For n-type GaAs, GeAu/Ni/Au with thickness of
100/30/100 nm are deposited, followed by rapid thermal annealing in forming gases at
375
o
C for 20 seconds. In the SEM image of Figure 6-7(a), a clear boundary appears
between the initial growth and overgrowth regions, indicated by the dashed lines. A
rectifying I-V curve with a built-in potential (V
bi
) is obtained while measuring across the
boundary. We can thus calculate the hole concentration (N
A
) by following the equation:

ln
, where k is Boltzmann constant and n
i
=1.8 10
6
cm
-3
, intrinsic carrier
concentration of GaAs. The N
A
is around 1 10
18
cm
-3
by using N
A
= 110
18
cm
-3
obtained
from PL and Hall measurements. Note that this is a rough estimation of the hole
concentration.
EBIC measurements are used to verify these p-n junctions. We expect the built-in
potential at the junction could effectively separate electron-hole pairs induced by electron
beam, thus, generating strong EBIC signals at these boundaries. Two metal contacts are
placed at two ends of the nanosheet, as shown in Figure 6-8(a). We do observe uniform
EBIC signals along these boundaries. The dark and bright contrasts represent the EBIC
current directions. We plot the EBIC profiles as a function of position across the
114

boundary for three different nanosheets, as shown in Figure 6-8(c). The profile consists
of space charge area (depletion width) and minority carrier diffusion length. The
depletion width is estimated to be less than 20 nm based on the carrier concentrations,
which is about the resolution limit of our EBIC measurement. Therefore, the EBIC
profile mainly represents the minority carrier diffusion length of the nanosheet. It can
also be obtained from the metal-semiconductor Schottky contact, as plotted in Figure 6-
8(d). The diffusion length, L
diff
= 177 nm, is extracted by fitting the profile with an
exponentially decaying function.

115

Figure 6-8. (a-b) Correlated SEM and EBIC images of a GaAs nanosheet. (c) EBIC
profiles across the inclined boundary of three different sheets. (d) EBIC profiles across
the boundary (solid squares) and metal-nanosheet Schokkty contact.

Here, we present a p-n junction formed naturally during the MOCVD growth.
This could be due to the low V/III ratio growth condition used in this study. The carbon
background doping somehow dominates in the overgrowth regions, thus makes them p-
type doped. The other possibility is the surface states induced doping because GaAs is
nortorious for its surface states, including oxygen defects[152], Ga/As dangling
bonds[38, 46, 102], and As
Ga
antisite defects[23, 174]. These surface states have been
(c)
(a)
(d)
-0.6 -0.3 0.0 0.3 0.6
0.0
0.4
0.8
p-n
junction
EBIC Signal
Position ( m)
NS_1
NS_2
NS_3
-0.6 -0.3 0.0 0.3 0.6
-0.4
0.0
0.4
0.8
NS_3_p-n
Schottky
EBIC Signal
Position ( m)
L
diff
= 177 nm
covered by metal
(b)
116

reported to have surface density of states lying at different energy positions in the band
gap, which could cause different degrees of Fermi level pinning. We observe an
additional Raman peak with frequency between TO (267 cm
-1
) and LO (292 cm
-1
) modes,
as shown in Figure 6-9. This peak is attributed to the surface-related phonon mode[156].

Figure 6-9. (a) Optical microscope image of GaAs nanosheets on a silicon substrate. The
colored dots represent the locations of laser spot. (b) Raman spectra of the nanosheet
taken at the different locations.

6.4 Conclusion
In summary, we observe a sharp boundary in MOCVD grown GaAs nanosheets in
SE-SEM images. Low temperature PL and Hall measurement are used to determine the
carrier type and concentration of the initial grown nanosheets. Based on the observation
of rectifying I-V behavior across boundary, we estimate both electron and hole
(b)
(a)
117

concentration to be around 1 10
18
cm
-3
. Finally, the EBIC mapping confirms a p-n
junction along the terminated boundary.

118

Chapter 7: Conclusion
We have applied a high flow rate CVD method to synthesize ultra-long SWCNTs.
A stretchable chip has been designed to study the ultimate breaking strain of carbon
nanotubes. The ultra-long nature of the nanotubes provides sufficient Van der Waals
force between nanotubes and substrate, which enables us to apply large strains on the
SWCNTs and set a new lower limit of breaking strain. In-situ Raman spectroscopy is
used to monitor the strain-induced G band downshifts. The G band Raman frequency
downshifts linearly with strain, and its downshifting rate spans a wide range from -6.2 to
-23.6 cm
-1
/%. This observation corroborates the theoretical prediction that the
downshifting rate is strongly chiral dependent. The highest strain we achieve in this
experiment is 13.7 0.3%. The reversible strain-induced change in the G band Raman
frequency downshifts indicates that no slippage, breakage, and defect formation occurs
during strain cycles. We also observe a threshold downshift 
G
> 75 cm
-1
for the G band
lineshape broadening which may be due to the strain-induced band gap opening of
nanotubes. We have demonstrated that more than 10% strain can be achieved by our
experimental setup, opening a new avenue to explore the electronic, vibrational, and
optical properties of SWCNTs under immerse strain (>10%).
Furthermore, we observe emergence of the D band Raman mode in SWCNTs
under axial strain. The D to G mode Raman intensity ratio (I
D
/I
G
) is observed to increase
with strain quadratically by more than a factor of 100. However, the ratio resumes its
original pre-strain value, indicating that there is no permanent defect formation. This
119

surprising result opposes the long standing notion that the D band is attributed to defects
and disorder in carbon nanotubes. We attribute the appearance of the D band without a
real defect in the lattice to the strain-induced symmetry lowering. Another possible
scenario is a strain-induced reversible transition from sp
2
to sp
3
bond configuration in a
twisted nanotube bundle, which has been observed electronically.
In the semiconductor nanowires project, we propose and demonstrate a new
method to locally tailor crystal structure of individual crystalline-amorphous core-shell
silicon nanowires with a polarized laser spot. By controlling the incident laser power, the
crystallinity of the silicon nanowires can be controlled from 0 to 0.93. This concept is
similar to conventional thermal annealing in which a laser serves as the heat source.
However, a strong polarization dependent heating/annealing is observed due to the one-
dimensional nature of the nanowires. The most efficient annealing occurs when the laser
polarization is aligned along the axis of a nanowire. The annealing temperature is
simultaneously monitored by Raman spectroscopy during annealing. Before and after
annealing, the crystal size of silicon nanowires can be estimated by Raman bond
polarization model. High resolution transmission electron microscopy (HRTEM) is also
used to confirm the laser-induced crystallization.
In order to build up a solid fundamental understanding of the nanomaterials we
used in nanostructure tandem solar cells, we modify a conventional scanning electron
microscope (SEM) to an electron beam induced current (EBIC) microscope. This allows
us to study the minority carrier diffusion length in Al
x
Ga
1-x
As-passivated GaAs
120

nanowires. With passivation, the minority carrier diffusion length (L
diff
) increases from 30
nm to 180 nm, as measured by EBIC mapping, and the photoluminescence (PL) lifetime
increases from sub-60 ps to 1.3 ns. This is due to a significant reduction in surface
recombination, which also results in a 48-fold enhancement in the continuous-wave PL
intensity on the same individual nanowire with and without the Al
x
Ga
1-x
As passivation
layer. These results indicate that the minority carrier lifetime is not limited by twin
stacking faults in passivated nanowires. We estimate the surface recombination velocity
(SRV) to range from 1.7 10
3
to 1.1 10
4
cm s
-1
, and the minority carrier mobility  to lie
in the range from 10.3 to 67.5 cm
2
V
-1
s
-1
for the passivated nanowires based on the PL
lifetime and minority carrier diffusion length.
We successfully grow twin-free GaAs nanosheets and observe a naturally formed
p-n junction between initial growth region and overgrowth region by using secondary
electron SEM (SE-SEM) and EBIC imaging. The twin-free crystal structure is confirmed
by HRTEM images. Surface states-induced Fermi level pinning effect is used to interpret
the clear boundaries observed between initial growth and over-growth regions in SE-
SEM images. According to the SE-SEM contrast, doping contrast, the initial growth is n-
doped and overgrowth is p-doped. Strong and uniform EBIC signals are observed along
these boundaries due to the built-in field of the p-n junctions. A rectifying current-voltage
characteristic across the boundaries further confirms the existence of p-n junctions. The
surface states-induced effective doping and doping reversion during the overgrowth are
the possible origins of the formation of the p-n junctions.
121

Bibliography
[1] A. Jorio, M. A. P., A. G. Souza Filho, R. Saito, G. Dresselhaus and M. S.
Dresselhaus, "Characterizing carbon nanotube samples with resonance Raman
scattering," New Journal of Physics, vol. 5, p. 139, 2003.
[2] Allen, J. E., E. R. Hemesath, D. E. Perea, J. L. Lensch-Falk, LiZ.Y, F. Yin, M. H.
Gass, P. Wang, A. L. Bleloch, R. E. Palmer, and L. J. Lauhon, "High-resolution
detection of Au catalyst atoms in Si nanowires," Nat Nano, vol. 3, pp. 168-173,
2008.
[3] Araujo, P. T., N. M. Barbosa Neto, H. Chacham, S. S. Carara, J. S. Soares, A. D.
Souza, L. G. Cançado, A. B. de Oliveira, R. J. C. Batista, E. Joselevich, M. S.
Dresselhaus, and A. Jorio, "In Situ Atomic Force Microscopy Tip-Induced
Deformations and Raman Spectroscopy Characterization of Single-Wall Carbon
Nanotubes," Nano Letters, vol. 12, pp. 4110-4116, 2012/08/08 2012.
[4] Arco, L. G.-D., B. Lei, S. Cronin, and C. Zhou, "Resonant micro-Raman
spectroscopy of aligned single-walled carbon nanotubes on a-plane sapphire,"
Applied Physics Letters, vol. 93, pp. 123112-3, 2008.
[5] Baxter, J. B. and E. S. Aydil, "Nanowire-based dye-sensitized solar cells,"
Applied Physics Letters, vol. 86, pp. 053114-3, 2005.
[6] Beeman, D., R. Tsu, and M. F. Thorpe, "Structural information from the Raman
spectrum of amorphous silicon," Physical Review B, vol. 32, p. 874, 1985.
[7] Blase, X., L. X. Benedict, E. L. Shirley, and S. G. Louie, "Hybridization effects
and metallicity in small radius carbon nanotubes," Physical Review Letters, vol.
72, pp. 1878-1881, 1994.
[8] Bourgoin, J. C., H. Hammadi, M. Stellmacher, J. Nagle, B. Grandidier, D.
Stievenard, J. P. Nys, C. Delerue, and M. Lannoo, "As antisite incorporation in
epitaxial growth of GaAs," Physica B: Condensed Matter, vol. 273–274, pp. 725-
728, 1999.
[9] Breuer, S., C. Pfüller, T. Flissikowski, O. Brandt, H. T. Grahn, L. Geelhaar, and H.
Riechert, "Suitability of Au- and Self-Assisted GaAs Nanowires for
Optoelectronic Applications," Nano Letters, vol. 11, pp. 1276-1279, 2011/03/09
2011.
[10] Brintlinger, T. H., "Carbon Nanotube Electronics: Growth, Imaging and
Electronic Properties," Phd., Maryland, Ph.D Thesis, 2005.
[11] Brown, S. D. M., A. Jorio, P. Corio, M. S. Dresselhaus, G. Dresselhaus, R. Saito,
and K. Kneipp, "Origin of the Breit-Wigner-Fano lineshape of the tangential G-
band feature of metallic carbon nanotubes," Physical Review B, vol. 63, p. 155414,
2001.
[12] Campbell, I. H. and P. M. Fauchet, "The effects of microcrystal size and shape on
the one phonon Raman spectra of crystalline semiconductors," Solid State
Communications, vol. 58, pp. 739-741, 1986.
[13] Cançado, L. G., A. Jorio, E. H. M. Ferreira, F. Stavale, C. A. Achete, R. B. Capaz,
M. V. O. Moutinho, A. Lombardo, T. S. Kulmala, and A. C. Ferrari, "Quantifying
122

Defects in Graphene via Raman Spectroscopy at Different Excitation Energies,"
Nano Letters, vol. 11, pp. 3190-3196, 2011/08/10 2011.
[14] Cao, J., Q. Wang, and H. Dai, "Electromechanical Properties of Metallic,
Quasimetallic, and Semiconducting Carbon Nanotubes under Stretching,"
Physical Review Letters, vol. 90, p. 157601, 2003.
[15] Cao, J., Q. Wang, and H. Dai, "Electromechanical Properties of Metallic,
Quasimetallic, and Semiconducting Carbon Nanotubes under Stretching,"
Physical Review Letters, vol. 90, p. 157601, 2003.
[16] Castro Neto, A. H., F. Guinea, N. M. R. Peres, K. S. Novoselov, and A. K. Geim,
"The electronic properties of graphene," Reviews of Modern Physics, vol. 81, p.
109, 2009.
[17] Chan, C. K., H. Peng, G. Liu, K. McIlwrath, X. F. Zhang, R. A. Huggins, and Y.
Cui, "High-performance lithium battery anodes using silicon nanowires," Nat
Nano, vol. 3, pp. 31-35, 2008.
[18] Chang, C.-C., H. Chen, C.-C. Chen, W.-H. Hung, I.-K. Hsu, J. Theiss, C. Zhou,
and S. B. Cronin, "Tailoring the crystal structure of individual silicon nanowires
by polarized laser annealing," Nanotechnology, vol. 22, p. 305709, 2011.
[19] Chang, C.-C., C.-Y. Chi, M. Yao, N. Huang, C.-C. Chen, J. Theiss, A. W.
Bushmaker, S. LaLumondiere, T.-W. Yeh, M. L. Povinelli, C. Zhou, P. D.
Dapkus, and S. B. Cronin, "Electrical and Optical Characterization of Surface
Passivation in GaAs Nanowires," Nano Letters, vol. 12, pp. 4484-4489,
2012/09/12 2012.
[20] Chang, C.-C., I. K. Hsu, M. Aykol, W.-H. Hung, C.-C. Chen, and S. B. Cronin,
"A New Lower Limit for the Ultimate Breaking Strain of Carbon Nanotubes,"
ACS Nano, vol. 4, pp. 5095-5100, 2010/09/28 2010.
[21] Chen, Y., B. Peng, and B. Wang, "Raman Spectra and Temperature-Dependent
Raman Scattering of Silicon Nanowires," The Journal of Physical Chemistry C,
vol. 111, pp. 5855-5858, 2007.
[22] Chiu, H. Y., P. Hung, H. W. C. Postma, and M. Bockrath, "Atomic-Scale Mass
Sensing Using Carbon Nanotube Resonators," Nano Letters, vol. 8, pp. 4342-
4346, Dec 2008.
[23] Choi, K. J. and J.-L. Lee, "Determination of energy levels of surface states in
GaAs metal--semiconductor field-effect transistor using deep-level transient
spectroscopy," Applied Physics Letters, vol. 74, pp. 1108-1110, 1999.
[24] Choudhary, V. and A. Gupta, Polymer/Carbon Nanotube Nanocomposites, 2011.
[25] Chu, H.-J., T.-W. Yeh, L. Stewart, and P. D. Dapkus, "Wurtzite InP nanowire
arrays grown by selective area MOCVD," physica status solidi (c), vol. 7, pp.
2494-2497, 2010.
[26] Colli, A., A. Fasoli, C. Ronning, S. Pisana, S. Piscanec, and A. C. Ferrari, "Ion
Beam Doping of Silicon Nanowires," Nano Letters, vol. 8, pp. 2188-2193,
2008/08/01 2008.
[27] Colombo, C., M. Heibeta, M. Gratzel, and A. Fontcuberta i Morral, "Gallium
arsenide p-i-n radial structures for photovoltaic applications," Applied Physics
Letters, vol. 94, pp. 173108-3, 2009.
123

[28] Cooper, C. A., R. J. Young, and M. Halsall, "Investigation into the Deformation
of Carbon Nanotubes and Their Composites through the Use of Raman
Spectroscopy," Compos. A, Appl. Sci. Manuf., vol. 32A, pp. 401-11, 2001.
[29] Cronin, S. B., A. K. Swan, M. S. Unlu, B. B. Goldberg, M. S. Dresselhaus, and M.
Tinkham, "Measuring the Uniaxial Strain of Individual Single-Wall Carbon
Nanotubes: Resonance Raman Spectra of Atomic-Force-Microscope Modified
Single-Wall Nanotubes," Physical Review Letters, vol. 93, pp. 167401-1, 2004.
[30] Cronin, S. B., A. K. Swan, M. S. Unlu, B. B. Goldberg, M. S. Dresselhaus, and M.
Tinkham, "Resonant Raman Spectroscopy of Individual Metallic and
Semiconducting Single-Wall Carbon Nanotubes under Uniaxial Strain," Physical
Review B (Condensed Matter and Materials Physics), vol. 72, pp. 35425-1, 2005.
[31] Cronin, S. B., A. K. Swan, M. S. Ünlü, B. B. Goldberg, M. S. Dresselhaus, and M.
Tinkham, "Measuring the Uniaxial Strain of Individual Single-Wall Carbon
Nanotubes: Resonance Raman Spectra of Atomic-Force-Microscope Modified
Single-Wall Nanotubes," Physical Review Letters, vol. 93, p. 167401, 2004.
[32] Cronin, S. B., A. K. Swan, M. S. Ünlü, B. B. Goldberg, M. S. Dresselhaus, and M.
Tinkham, "Resonant Raman spectroscopy of individual metallic and
semiconducting single-wall carbon nanotubes under uniaxial strain," Physical
Review B, vol. 72, p. 035425, 2005.
[33] Cui, L.-F., R. Ruffo, C. K. Chan, H. Peng, and Y. Cui, "Crystalline-Amorphous
Core −Shell Silicon Nanowires for High Capacity and High Current Battery
Electrodes," Nano Letters, vol. 9, pp. 491-495, 2009/01/14 2008.
[34] Dan, Y., K. Seo, K. Takei, J. H. Meza, A. Javey, and K. B. Crozier, "Dramatic
Reduction of Surface Recombination by in Situ Surface Passivation of Silicon
Nanowires," Nano Letters, vol. 11, pp. 2527-2532, 2011/06/08 2011.
[35] De-Sheng, J., Y. Makita, K. Ploog, and H. J. Queisser, "Electrical properties and
photoluminescence of Te-doped GaAs grown by molecular beam epitaxy,"
Journal of Applied Physics, vol. 53, pp. 999-1006, 1982.
[36] Demichel, O., V. Calvo, A. Besson, P. Noé, B. Salem, N. Pauc, F. Oehler, P.
Gentile, and N. Magnea, "Surface Recombination Velocity Measurements of
Efficiently Passivated Gold-Catalyzed Silicon Nanowires by a New Optical
Method," Nano Letters, vol. 10, pp. 2323-2329, 2010/07/14 2010.
[37] Demichel, O., M. Heiss, J. Bleuse, H. Mariette, and A. Fontcuberta i Morral,
"Impact of surfaces on the optical properties of GaAs nanowires," Applied
Physics Letters, vol. 97, pp. 201907-3, 2010.
[38] Dose, V., H. J. Gossmann, and D. Straub, "Investigation of Intrinsic Unoccupied
Surface States at GaAs(110) by Isochromat Spectroscopy," Physical Review
Letters, vol. 47, pp. 608-610, 1981.
[39] Dresselhaus, M. S., G. Dresselhaus, A. Jorio, A. G. Souza, M. A. Pimenta, and R.
Saito, "Single nanotube Raman spectroscopy," Accounts of Chemical Research,
vol. 35, pp. 1070-1078, 2002.
[40] Dresselhaus, M. S., G. Dresselhaus, A. Jorio, A. G. Souza, and R. Saito, "Raman
spectroscopy on isolated single wall carbon nanotubes," Carbon, vol. 40, pp.
2043-2061, 2002.
124

[41] Dresselhaus, M. S., G. Dresselhaus, R. Saito, and A. Jorio, "Raman spectroscopy
of carbon nanotubes," Physics Reports, vol. 409, pp. 47-99, 2005.
[42] Dresselhaus, M. S., A. Jorio, M. Hofmann, G. Dresselhaus, and R. Saito,
"Perspectives on Carbon Nanotubes and Graphene Raman Spectroscopy," Nano
Letters, vol. 10, pp. 751-758, 2010/03/10 2010.
[43] Duan, X. and C. M. Lieber, "General Synthesis of Compound Semiconductor
Nanowires," Advanced Materials, vol. 12, pp. 298-302, 2000.
[44] Duan, X., J. Wang, and C. M. Lieber, "Synthesis and optical properties of gallium
arsenide nanowires," Applied Physics Letters, vol. 76, pp. 1116-1118, 2000.
[45] Dumitrica, T., M. Hua, and B. I. Yakobson, "Symmetry-, time-, and temperature-
dependent strength of carbon nanotubes," Proceedings of the National Academy
of Sciences, vol. 103, pp. 6105-6109, April 18, 2006 2006.
[46] Eastman, D. E. and J. L. Freeouf, "Photoemission Partial Yield Measurements of
Unoccupied Intrinsic Surface States for Ge(111) and GaAs(110)," Physical
Review Letters, vol. 33, pp. 1601-1605, 1974.
[47] Elliott, S. L., R. F. Broom, and C. J. Humphreys, "Dopant profiling with the
scanning electron microscope---A study of Si," Journal of Applied Physics, vol.
91, pp. 9116-9122, 2002.
[48] Faraci, G., S. Gibilisco, P. Russo, A. R. Pennisi, and S. La Rosa, "Modified
Raman confinement model for Si nanocrystals," Physical Review B, vol. 73, p.
033307, 2006.
[49] Frank, O., M. Mohr, J. Maultzsch, C. Thomsen, I. Riaz, R. Jalil, K. S. Novoselov,
G. Tsoukleri, J. Parthenios, K. Papagelis, L. Kavan, and C. Galiotis, "Raman 2D-
Band Splitting in Graphene: Theory and Experiment," ACS Nano, vol. 5, pp.
2231-2239, 2011/03/22 2011.
[50] Frogley, M. D., Q. Zhao, and H. D. Wagner, "Polarized resonance Raman
spectroscopy of single-wall carbon nanotubes within a polymer under strain,"
Physical Review B, vol. 65, p. 113413, 2002.
[51] Gao, B., L. Jiang, X. Ling, J. Zhang, and Z. Liu, "Chirality-Dependent Raman
Frequency Variation of Single-Walled Carbon Nanotubes under Uniaxial Strain,"
The Journal of Physical Chemistry C, vol. 112, pp. 20123-20125, 2008.
[52] Garnett, E. and P. Yang, "Light Trapping in Silicon Nanowire Solar Cells," Nano
Letters, vol. 10, pp. 1082-1087, 2010/03/10 2010.
[53] Geisz, J. F., S. Kurtz, M. W. Wanlass, J. S. Ward, A. Duda, D. J. Friedman, J. M.
Olson, W. E. McMahon, T. E. Moriarty, and J. T. Kiehl, "High-efficiency
GaInP/GaAs/InGaAs triple-junction solar cells grown inverted with a
metamorphic bottom junction," Applied Physics Letters, vol. 91, pp. 023502-3,
2007.
[54] Glavatskikh, I. A., V. S. Kortov, and H. J. Fitting, "Self-consistent electrical
charging of insulating layers and metal-insulator-semiconductor structures,"
Journal of Applied Physics, vol. 89, pp. 440-448, 2001.
[55] Gudiksen, M. S., L. J. Lauhon, J. Wang, D. C. Smith, and C. M. Lieber, "Growth
of nanowire superlattice structures for nanoscale photonics and electronics,"
Nature, vol. 415, pp. 617-620, 2002.
125

[56] Guo, W., A. Banerjee, P. Bhattacharya, and B. S. Ooi, "InGaN/GaN disk-in-
nanowire white light emitting diodes on (001) silicon," Applied Physics Letters,
vol. 98, pp. 193102-3, 2011.
[57] Gupta, R., Q. Xiong, C. K. Adu, U. J. Kim, and P. C. Eklund, "Laser-Induced
Fano Resonance Scattering in Silicon Nanowires," Nano Letters, vol. 3, pp. 627-
631, 2003.
[58] Gupta, S., K. Dharamvir, and V. K. Jindal, "Elastic Moduli of Single-Walled
Carbon Nanotubes and Their Ropes," Physical Review B (Condensed Matter and
Materials Physics), vol. 72, pp. 165428-16, 2005.
[59] Gutsche, C., A. Lysov, D. Braam, I. Regolin, G. Keller, Z.-A. Li, M. Geller, M.
Spasova, W. Prost, and F.-J. Tegude, "n-GaAs/InGaP/p-GaAs Core-Multishell
Nanowire Diodes for Efficient Light-to-Current Conversion," Advanced
Functional Materials, vol. 22, pp. 929-936, 2012.
[60] Gutsche, C., R. Niepelt, M. Gnauck, A. Lysov, W. Prost, C. Ronning, and F.-J.
Tegude, "Direct Determination of Minority Carrier Diffusion Lengths at Axial
GaAs Nanowire p–n Junctions," Nano Letters, vol. 12, pp. 1453-1458,
2012/03/14 2012.
[61] Hadjiev, V. G., M. N. Iliev, S. Arepalli, P. Nikolaev, and B. S. Files, "Raman
Scattering Test of Single-Wall Carbon Nanotube Composites," Applied Physics
Letters, vol. 78, p. 3193, 2001.
[62] Harmon, E. S., M. L. Lovejoy, M. R. Melloch, M. S. Lundstrom, T. J. de Lyon,
and J. M. Woodall, "Experimental observation of a minority electron mobility
enhancement in degenerately doped p-type GaAs," Applied Physics Letters, vol.
63, pp. 536-538, 1993.
[63] Hoffmann, S., J. Bauer, C. Ronning, T. Stelzner, J. Michler, C. Ballif, V. Sivakov,
and S. H. Christiansen, "Axial p-n Junctions Realized in Silicon Nanowires by Ion
Implantation," Nano Letters, vol. 9, pp. 1341-1344, 2009/04/08 2009.
[64] Hsu, I. K., M. T. Pettes, A. Bushmaker, M. Aykol, L. Shi, and S. B. Cronin,
"Optical Absorption and Thermal Transport of Individual Suspended Carbon
Nanotube Bundles," Nano Letters, vol. 9, pp. 590-594, 2009.
[65] Huang, M., "Studies of Mechanically Deformed Single Wall Carbon Nanotubes
and Graphene by Optical Spectroscopy," Ph. D Dissertation, 2009.
[66] Huang, M., Y. Wu, B. Chandra, H. Yan, Y. Shan, T. F. Heinz, and J. Hone,
"Direct Measurement of Strain-Induced Changes in the Band Structure of Carbon
Nanotubes," Physical Review Letters, vol. 100, p. 136803, 2008.
[67] Huang, M., H. Yan, C. Chen, D. Song, T. F. Heinz, and J. Hone, "Phonon
softening and crystallographic orientation of strained graphene studied by Raman
spectroscopy," Proceedings of the National Academy of Sciences, vol. 106, pp.
7304-7308, May 5, 2009 2009.
[68] Huang, M., H. Yan, T. F. Heinz, and J. Hone, "Probing Strain-Induced Electronic
Structure Change in Graphene by Raman Spectroscopy," Nano Letters, vol. 10, pp.
4074-4079, 2010/10/13 2010.
126

[69] Huang, M. Y., Y. Wu, B. Chandra, H. Yan, Y. Shan, T. F. Heinz, and J. Hone,
"Direct measurement of strain-induced changes in the band structure of carbon
nanotubes," Physical Review Letters, vol. 100, p. 136803, Apr 2008.
[70] Hulman, M., V. Skakalova, S. Roth, and H. Kuzmany, "Raman spectroscopy of
single-wall carbon nanotubes and graphite irradiated by gamma rays," Journal of
Applied Physics, vol. 98, pp. 024311-5, 2005.
[71] Hyun, J. and L. Lauhon. (2011). Energy and Spatially Resolved Measurements of
Plasmonically Enhanced Photocurrent in a Single Si Nanowire with Au
Nanoparticles.
[72] Jiang, C., K. Kempa, J. Zhao, U. Schlecht, U. Kolb, T. Basché, M. Burghard, and
A. Mews, "Strong enhancement of the Breit-Wigner-Fano Raman line in carbon
nanotube bundles caused by plasmon band formation," Physical Review B, vol. 66,
p. 161404, 2002.
[73] Jorio, A., C. Fantini, M. S. S. Dantas, M. A. Pimenta, A. G. Souza Filho, G. G.
Samsonidze, V. W. Brar, G. Dresselhaus, M. S. Dresselhaus, A. K. Swan, M. S.
Ünlü, B. B. Goldberg, and R. Saito, "Linewidth of the Raman features of
individual single-wall carbon nanotubes," Physical Review B, vol. 66, p. 115411,
2002.
[74] Jorio, A., M. A. Pimenta, A. G. Souza Filho, R. Saito, G. Dresselhaus, and M. S.
Dresselhaus, "Characterizing carbon nanotube samples with resonance Raman
scattering," New Journal of Physics, vol. 5, p. 139, 2003.
[75] Jorio, A., A. G. Souza Filho, G. Dresselhaus, M. S. Dresselhaus, R. Saito, J. H.
Hafner, C. M. Lieber, F. M. Matinaga, M. S. S. Dantas, and M. A. Pimenta, "Joint
density of electronic states for one isolated single-wall carbon nanotube studied
by resonant Raman scattering," Physical Review B, vol. 63, p. 245416, 2001.
[76] Jorio, A., A. G. Souza Filho, G. Dresselhaus, M. S. Dresselhaus, A. K. Swan, M.
S. Ünlü, B. B. Goldberg, M. A. Pimenta, J. H. Hafner, C. M. Lieber, and R. Saito,
"G-band resonant Raman study of 62 isolated single-wall carbon nanotubes,"
Physical Review B, vol. 65, p. 155412, 2002.
[77] Joyce, H. J., Q. Gao, H. H. Tan, C. Jagadish, Y. Kim, X. Zhang, Y. Guo, and J.
Zou, "Twin-Free Uniform Epitaxial GaAs Nanowires Grown by a Two-
Temperature Process," Nano Letters, vol. 7, pp. 921-926, 2007/04/01 2007.
[78] Kakkad, R., J. Smith, W. S. Lau, S. J. Fonash, and R. Kerns, "Crystallized Si
films by low-temperature rapid thermal annealing of amorphous silicon," Journal
of Applied Physics, vol. 65, pp. 2069-2072, 1989.
[79] Katzenmeyer, A. M., F. o. Léonard, A. A. Talin, P.-S. Wong, and D. L. Huffaker,
"Poole −Frenkel Effect and Phonon-Assisted Tunneling in GaAs Nanowires,"
Nano Letters, vol. 10, pp. 4935-4938, 2010/12/08 2010.
[80] Kolb, G., T. Salbert, and G. Abstreiter, "Raman-microanalysis of strain and
crystal orientation in laser-crystallized silicon," Fresenius' Journal of Analytical
Chemistry, vol. 341, pp. 166-170, 1991.
[81] Korin, E., R. Reif, and B. Mikic, "Crystallization of amorphous silicon films
using a multistep thermal annealing process," Thin Solid Films, vol. 167, pp. 101-
106, 1988.
127

[82] Kumar, R., M. Aykol, and S. B. Cronin, "Effect of Nanotube-Nanotube Coupling
on the Radial Breathing Mode of Carbon Nanotubes," Physical Review B
(Condensed Matter and Materials Physics), vol. 78, pp. 165428-4, 2008.
[83] Kumar, R., M. Aykol, K. Ryu, C. Zhou, and S. B. Cronin, "Top-down
lithographic method for inducing strain in carbon nanotubes," Journal of Applied
Physics, vol. 106, pp. 014306-5, 2009.
[84] Kumar, R. and S. B. Cronin, "Optical properties of carbon nanotubes under axial
strain," Journal of Nanoscience and Nanotechnology, vol. 8, pp. 122-130, Jan
2008.
[85] Kumar, R. and S. B. Cronin, "Raman Scattering of Carbon Nanotube Bundles
under Axial Strain and Strain-Induced Debundling," Physical Review B
(Condensed Matter), vol. 75, pp. 155421-1, 2007.
[86] Kumar, R. and S. B. Cronin, "Raman scattering of carbon nanotube bundles under
axial strain and strain-induced debundling," Physical Review B, vol. 75, p. 155421,
2007.
[87] Lai, E., W. Kim, and P. Yang, "Vertical nanowire array-based light emitting
diodes," Nano Research, vol. 1, pp. 123-128, 2008.
[88] Leamy, H. J., "Charge collection scanning electron microscopy," Journal of
Applied Physics, vol. 53, pp. R51-R80, 1982.
[89] Lee, C., X. Wei, J. W. Kysar, and J. Hone, "Measurement of the Elastic Properties
and Intrinsic Strength of Monolayer Graphene," Science, vol. 321, pp. 385-388,
July 18, 2008 2008.
[90] Lee, H. G., H. C. Jeon, T. W. Kang, and T. W. Kim, "Gallium arsenide crystalline
nanorods grown by molecular-beam epitaxy," Applied Physics Letters, vol. 78, pp.
3319-3321, 2001.
[91] Lee, J. W., H. Shichijo, H. L. Tsai, and R. J. Matyi, "Defect reduction by thermal
annealing of GaAs layers grown by molecular beam epitaxy on Si substrates,"
Applied Physics Letters, vol. 50, pp. 31-33, 1987.
[92] Li, B., D. Yu, and S.-L. Zhang, "Raman spectral study of silicon nanowires,"
Physical Review B, vol. 59, p. 1645, 1999.
[93] Li, J., Y. He, Y. Han, K. Liu, J. Wang, Q. Li, S. Fan, and K. Jiang, "Direct
Identification of Metallic and Semiconducting Single-Walled Carbon Nanotubes
in Scanning Electron Microscopy," Nano Letters, vol. 12, pp. 4095-4101,
2012/08/08 2012.
[94] Liu, F., P. Ming, and J. Li, "Ab Initio Calculation of Ideal Strength and Phonon
Instability of Graphene under Tension," Physical Review B (Condensed Matter
and Materials Physics), vol. 76, p. 064120, 2007.
[95] Liu, L., C. S. Jayanthi, M. Tang, S. Y. Wu, T. W. Tombler, C. Zhou, L.
Alexseyev, J. Kong, and H. Dai, "Controllable Reversibility of an sp^{2} to sp^{3}
Transition of a Single Wall Nanotube under the Manipulation of an AFM Tip: A
Nanoscale Electromechanical Switch?," Physical Review Letters, vol. 84, pp.
4950-4953, 2000.
[96] Lloyd-Hughes, J., S. K. E. Merchant, L. Fu, H. H. Tan, C. Jagadish, E. Castro-
Camus, and M. B. Johnston, "Influence of surface passivation on ultrafast carrier
128

dynamics and terahertz radiation generation in GaAs," Applied Physics Letters,
vol. 89, pp. 232102-3, 2006.
[97] Lu, Z. Q., T. Quinn, and H. S. Reehal, "Polarization-dependent Raman spectra of
thin crystalline silicon films," Journal of Applied Physics, vol. 97, p. 033512,
2005.
[98] Lucas, M. and R. J. Young, "Effect of uniaxial strain deformation upon the
Raman radial breathing modes of single-wall carbon nanotubes in composites,"
Physical Review B, vol. 69, p. 085405, 2004.
[99] Lucas, M. and R. J. Young, "Effect of Uniaxial Strain Deformation upon the
Raman Radial Breathing Modes of Single-Wall Carbon Nanotubes in
Composites," Physical Review B (Condensed Matter and Materials Physics), vol.
69, pp. 85405-9, 2004.
[100] Lucas, M. and R. J. Young, "Raman Spectroscopic Study of the Effect of Strain
on the Radial Breathing Modes of Carbon Nanotubes in Epoxy/SWNT
Composites," Composites Science and Technology, vol. 64, pp. 2297-2302, 2004.
[101] Lucchese, M. M., F. Stavale, E. H. M. Ferreira, C. Vilani, M. V. O. Moutinho, R.
B. Capaz, C. A. Achete, and A. Jorio, "Quantifying ion-induced defects and
Raman relaxation length in graphene," Carbon, vol. 48, pp. 1592-1597, 2010.
[102] Ludeke, R. and L. Esaki, "Electron Energy-Loss Spectroscopy of GaAs and Ge
Surfaces," Physical Review Letters, vol. 33, pp. 653-656, 1974.
[103] Madaria, A. R., M. Yao, C. Chi, N. Huang, C. Lin, R. Li, M. L. Povinelli, P. D.
Dapkus, and C. Zhou, "Toward Optimized Light Utilization in Nanowire Arrays
Using Scalable Nanosphere Lithography and Selected Area Growth," Nano
Letters, vol. 12, pp. 2839-2845, 2012/06/13 2012.
[104] Maiti, A., A. Svizhenko, and M. P. Anantram, "Electronic Transport through
Carbon Nanotubes: Effects of Structural Deformation and Tube Chirality,"
Physical Review Letters, vol. 88, p. 126805, 2002.
[105] Maki, H., T. Sato, and K. Ishibashi, "Direct Observation of the Deformation and
the Band Gap Change from an Individual Single-Walled Carbon Nanotube under
Uniaxial Strain," Nano Letters, vol. 7, pp. 890-895, 2007.
[106] Mariani, G., P.-S. Wong, A. M. Katzenmeyer, F. Léonard, J. Shapiro, and D. L.
Huffaker, "Patterned Radial GaAs Nanopillar Solar Cells," Nano Letters, vol. 11,
pp. 2490-2494, 2011/06/08 2011.
[107] Marinopoulos, A. G., L. Reining, A. Rubio, and N. Vast, "Optical and Loss
Spectra of Carbon Nanotubes: Depolarization Effects and Intertube Interactions,"
Physical Review Letters, vol. 91, p. 046402, 2003.
[108] Matthews, M. J., M. A. Pimenta, G. Dresselhaus, M. S. Dresselhaus, and M. Endo,
"Origin of dispersive effects of the Raman D band in carbon materials," Physical
Review B, vol. 59, pp. R6585-R6588, 1999.
[109] Maultzsch, J., S. Reich, and C. Thomsen, "Chirality-selective Raman scattering of
the D mode in carbon nanotubes," Physical Review B, vol. 64, p. 121407, 2001.
[110] Maultzsch, J., S. Reich, C. Thomsen, H. Requardt, and P. Ordejón, "Phonon
Dispersion in Graphite," Physical Review Letters, vol. 92, p. 075501, 2004.
129

[111] Maune, H. and M. Bockrath, "Elastomeric carbon nanotube circuits for local
strain sensing," Applied Physics Letters, vol. 89, pp. 173131-3, 2006.
[112] Minot, E. D., Y. Yaish, V. Sazonova, J.-Y. Park, M. Brink, and P. L. McEuen,
"Tuning Carbon Nanotube Band Gaps with Strain," Physical Review Letters, vol.
90, p. 156401, 2003.
[113] Miska, P., M. Dossot, T. D. Nguyen, M. Grun, H. Rinnert, M. Vergnat, and B.
Humbert, "Embedded Silicon Nanocrystals Studied by Photoluminescence and
Raman Spectroscopies: Exciton and Phonon Confinement Effects," The Journal
of Physical Chemistry C, vol. 114, pp. 17344-17349, 2010.
[114] Mohiuddin, T. M. G., A. Lombardo, R. R. Nair, A. Bonetti, G. Savini, R. Jalil, N.
Bonini, D. M. Basko, C. Galiotis, N. Marzari, K. S. Novoselov, A. K. Geim, and
A. C. Ferrari, "Uniaxial strain in graphene by Raman spectroscopy: G peak
splitting, Grüneisen parameters, and sample orientation," Physical Review B, vol.
79, p. 205433, 2009.
[115] Montazeri, M., M. Fickenscher, L. M. Smith, H. E. Jackson, J. Yarrison-Rice, J.
H. Kang, Q. Gao, H. H. Tan, C. Jagadish, Y. Guo, J. Zou, M.-E. Pistol, and C. E.
Pryor, "Direct Measure of Strain and Electronic Structure in GaAs/GaP
Core −Shell Nanowires," Nano Letters, vol. 10, pp. 880-886, 2010/03/10 2010.
[116] Nardelli, M. B., B. I. Yakobson, and J. Bernholc, "Mechanism of strain release in
carbon nanotubes," Physical Review B-Condensed Matter, vol. 57, pp. R4277-
R4280, Feb 1998.
[117] Nelson, R. J. and R. G. Sobers, "Interfacial recombination velocity in
GaAlAs/GaAs heterostructures," Applied Physics Letters, vol. 32, pp. 761-763,
1978.
[118] Nelson, R. J. and R. G. Sobers, "Minority-carrier lifetimes and internal quantum
efficiency of surface-free GaAs," Journal of Applied Physics, vol. 49, pp. 6103-
6108, 1978.
[119] Ni, Z. H., T. Yu, Y. H. Lu, Y. Y. Wang, Y. P. Feng, and Z. X. Shen, "Uniaxial
Strain on Graphene: Raman Spectroscopy Study and Band-Gap Opening," ACS
Nano, vol. 2, pp. 2301-2305, 2008/11/25 2008.
[120] Noborisaka, J., J. Motohisa, S. Hara, and T. Fukui, "Fabrication and
characterization of freestanding GaAs/AlGaAs core-shell nanowires and AlGaAs
nanotubes by using selective-area metalorganic vapor phase epitaxy," Applied
Physics Letters, vol. 87, pp. 093109-3, 2005.
[121] Ouattara, L., A. Mikkelsen, N. Sköld, J. Eriksson, T. Knaapen, E. Ćavar, W.
Seifert, L. Samuelson, and E. Lundgren, "GaAs/AlGaAs Nanowire
Heterostructures Studied by Scanning Tunneling Microscopy," Nano Letters, vol.
7, pp. 2859-2864, 2007/09/01 2007.
[122] Paillard, V., P. Puech, M. A. Laguna, R. Carles, B. Kohn, and F. Huisken,
"Improved one-phonon confinement model for an accurate size determination of
silicon nanocrystals," Journal of Applied Physics, vol. 86, pp. 1921-1924, 1999.
[123] Parkinson, P., H. J. Joyce, Q. Gao, H. H. Tan, X. Zhang, J. Zou, C. Jagadish, L. M.
Herz, and M. B. Johnston, "Carrier Lifetime and Mobility Enhancement in Nearly
130

Defect-Free Core −Shell Nanowires Measured Using Time-Resolved Terahertz
Spectroscopy," Nano Letters, vol. 9, pp. 3349-3353, 2009/09/09 2009.
[124] Perera, S., M. A. Fickenscher, H. E. Jackson, L. M. Smith, J. M. Yarrison-Rice, H.
J. Joyce, Q. Gao, H. H. Tan, C. Jagadish, X. Zhang, and J. Zou, "Nearly intrinsic
exciton lifetimes in single twin-free GaAs/AlGaAs core-shell nanowire
heterostructures," Applied Physics Letters, vol. 93, pp. 053110-3, 2008.
[125] Perovic, D. D., M. R. Castell, A. Howie, C. Lavoie, T. Tiedje, and J. S. W. Cole,
"Field-emission SEM imaging of compositional and doping layer semiconductor
superlattices," Ultramicroscopy, vol. 58, pp. 104-113, 1995.
[126] Piscanec, S., M. Cantoro, A. C. Ferrari, J. A. Zapien, Y. Lifshitz, S. T. Lee, S.
Hofmann, and J. Robertson, "Raman spectroscopy of silicon nanowires," Physical
Review B, vol. 68, p. 241312, 2003.
[127] Pócsik, I., M. Hundhausen, M. Koós, and L. Ley, "Origin of the D peak in the
Raman spectrum of microcrystalline graphite," Journal of Non-Crystalline Solids,
vol. 227–230, Part 2, pp. 1083-1086, 1998.
[128] Pugno, N. M., "On the Strength of the Carbon Nanotube-Based Space Elevator
Cable: from Nanomechanics to Megamechanics," Journal of Physics: Condensed
Matter, vol. 18, pp. S1971-S1990, 2006.
[129] Randal, J. G., W. Qian, C. Jien, W. Dunwei, and D. Hongjie, "Piezoresistance of
Carbon nanotubes on Deformable Thin-Film Membranes," Applied Physics
Letters, vol. 86, p. 093104, 2005.
[130] Reich, S., J. Maultzsch, C. Thomsen, and P. Ordejón, "Tight-binding description
of graphene," Physical Review B, vol. 66, p. 035412, 2002.
[131] Richter, H., Z. P. Wang, and L. Ley, "The one phonon Raman spectrum in
microcrystalline silicon," Solid State Communications, vol. 39, pp. 625-629, 1981.
[132] Righi, A., S. D. Costa, H. Chacham, C. Fantini, P. Venezuela, C. Magnuson, L.
Colombo, W. S. Bacsa, R. S. Ruoff, and M. A. Pimenta, "Graphene Moiré
patterns observed by umklapp double-resonance Raman scattering," Physical
Review B, vol. 84, p. 241409, 2011.
[133] Sazonova, V., Y. Yalsh, H. Ustunel, D. Roundy, T. A. Arias, and P. L. McEuen,
"A Tunable Carbon Nanotube Electromechanical Oscillator," Nature, vol. 431, pp.
284-7, 2004.
[134] Schadler, L. S., S. C. Giannaris, and P. M. Ajayan, "Load Transfer in Carbon
Nanotube Epoxy Composites," Applied Physics Letters, vol. 73, pp. 3842-4, 1998.
[135] Scheel, H., S. Khachadorian, M. Cantoro, A. Colli, A. C. Ferrari, and C. Thomsen,
"Silicon nanowire optical Raman line shapes at cryogenic and elevated
temperatures," physica status solidi (b), vol. 245, pp. 2090-2093, 2008.
[136] Scheel, H., S. Reich, A. C. Ferrari, M. Cantoro, A. Colli, and C. Thomsen,
"Raman scattering on silicon nanowires: The thermal conductivity of the
environment determines the optical phonon frequency," Applied Physics Letters,
vol. 88, p. 233114, 2006.
[137] Schonjahn, C., R. F. Broom, C. J. Humphreys, A. Howie, and S. A. M. Mentink,
"Optimizing and quantifying dopant mapping using a scanning electron
131

microscope with a through-the-lens detector," Applied Physics Letters, vol. 83, pp.
293-295, 2003.
[138] Sfeir, M. Y., F. Wang, L. Huang, C.-C. Chuang, J. Hone, S. P. O'Brien, T. F.
Heinz, and L. E. Brus, "Probing Electronic Transitions in Individual Carbon
Nanotubes by Rayleigh Scattering," Science, vol. 306, pp. 1540-1543, November
26, 2004 2004.
[139] Sheldon, M. T., C. N. Eisler, and H. A. Atwater, "GaAs Passivation with
Trioctylphosphine Sulfide for Enhanced Solar Cell Efficiency and Durability,"
Advanced Energy Materials, pp. n/a-n/a, 2012.
[140] Shi, W. S., Y. F. Zheng, N. Wang, C. S. Lee, and S. T. Lee, "A General Synthetic
Route to III–V Compound Semiconductor Nanowires," Advanced Materials, vol.
13, pp. 591-594, 2001.
[141] Shi, W. S., Y. F. Zheng, N. Wang, C. S. Lee, and S. T. Lee, "Oxide-assisted
growth and optical characterization of gallium-arsenide nanowires," Applied
Physics Letters, vol. 78, pp. 3304-3306, 2001.
[142] Skákalová, V., Y. S. Woo, Z. Osváth, L. P. Biró, and S. Roth, "Electron transport
in Ar+-irradiated single wall carbon nanotubes," physica status solidi (b), vol. 243,
pp. 3346-3350, 2006.
[143] Smit, C., R. A. C. M. M. v. Swaaij, H. Donker, A. M. H. N. Petit, W. M. M.
Kessels, and M. C. M. v. d. Sanden, "Determining the material structure of
microcrystalline silicon from Raman spectra," Journal of Applied Physics, vol. 94,
pp. 3582-3588, 2003.
[144] Solomon, G. S., D. Kirillov, H. C. Chui, and J. J. S. Harris, "Determination of
AlAs mole fraction in Al[sub x]Ga[sub 1 - x]As using Raman spectroscopy and x-
ray diffraction," Stanford, California (USA), 1994, pp. 1078-1081.
[145] Son, H., G. G. Samsonidze, J. Kong, Y. Zhang, X. Duan, J. Zhang, Z. Liu, and M.
S. Dresselhaus, "Strain and Friction Induced by Van der Waals Interaction in
Individual Single Walled Carbon Nanotubes," Applied Physics Letters, vol. 90, pp.
253113-3, 2007.
[146] Sood, A. K., R. Gupta, and S. A. Asher, "Origin of the unusual dependence of
Raman D band on excitation wavelength in graphite-like materials," Journal of
Applied Physics, vol. 90, pp. 4494-4497, 2001.
[147] Souza Filho, A. G., N. Kobayashi, J. Jiang, A. Grüneis, R. Saito, S. B. Cronin, J.
Mendes Filho, G. G. Samsonidze, G. Dresselhaus, and M. S. Dresselhaus, "Strain-
Induced Interference Effects on the Resonance Raman Cross Section of Carbon
Nanotubes," Physical Review Letters, vol. 95, p. 217403, 2005.
[148] Stone, A. J. and D. J. Wales, "Theoretical studies of icosahedral C60 and some
related species," Chemical Physics Letters, vol. 128, pp. 501-503, 1986.
[149] Su, Z., J. Sha, G. Pan, J. Liu, D. Yang, C. Dickinson, and W. Zhou,
"Temperature-Dependent Raman Scattering of Silicon Nanowires," The Journal
of Physical Chemistry B, vol. 110, pp. 1229-1234, 2006.
[150] Sui, Z., P. P. Leong, I. P. Herman, G. S. Higashi, and H. Temkin, "Raman
analysis of light-emitting porous silicon," Applied Physics Letters, vol. 60, pp.
2086-2088, 1992.
132

[151] Suyatin, D. B., C. Thelander, M. T. Björk, I. Maximov, and L. Samuelson,
"Sulfur passivation for ohmic contact formation to InAs nanowires "
Nanotechnology, vol. 18, p. 105307, 2007.
[152] Szuber, J., "Electronic properties and origin of surface states on GaAs(100)
surfaces thermally cleaned in ultrahigh vacuum," Thin Solid Films, vol. 105, pp.
33-37, 1983.
[153] Tan, P., Y. Deng, and Q. Zhao, "Temperature-dependent Raman spectra and
anomalous Raman phenomenon of highly oriented pyrolytic graphite," Physical
Review B, vol. 58, pp. 5435-5439, 1998.
[154] Thomsen, C. and S. Reich, "Double Resonant Raman Scattering in Graphite,"
Physical Review Letters, vol. 85, pp. 5214-5217, 2000.
[155] Tian, B., X. Zheng, T. J. Kempa, Y. Fang, N. Yu, G. Yu, J. Huang, and C. M.
Lieber, "Coaxial silicon nanowires as solar cells and nanoelectronic power
sources," Nature, vol. 449, pp. 885-889, 2007.
[156] Tiginyanu, I. M., G. Irmer, J. Monecke, A. Vogt, and H. L. Hartnagel, "Porosity-
induced modification of the phonon spectrum of n-GaAs," Semiconductor Science
and Technology, vol. 12, p. 491, 1997.
[157] Titova, L. V., T. B. Hoang, H. E. Jackson, L. M. Smith, J. M. Yarrison-Rice, Y.
Kim, H. J. Joyce, H. H. Tan, and C. Jagadish, "Temperature dependence of
photoluminescence from single core-shell GaAs--AlGaAs nanowires," Applied
Physics Letters, vol. 89, pp. 173126-3, 2006.
[158] Titova, L. V., T. B. Hoang, J. M. Yarrison-Rice, H. E. Jackson, Y. Kim, H. J.
Joyce, Q. Gao, H. H. Tan, C. Jagadish, X. Zhang, J. Zou, and L. M. Smith,
"Dynamics of Strongly Degenerate Electron −Hole Plasmas and Excitons in Single
InP Nanowires," Nano Letters, vol. 7, pp. 3383-3387, 2007/11/01 2007.
[159] Tombler, T. W., C. Zhou, L. Alexseyev, J. Kong, H. Dai, L. Liu, C. S. Jayanthi,
M. Tang, and S.-Y. Wu, "Reversible electromechanical characteristics of carbon
nanotubes under local-probe manipulation," Nature, vol. 405, pp. 769-772, 2000.
[160] Tombler, T. W., C. Zhou, L. Alexseyev, J. Kong, H. Dai, L. Liu, C. S. Jayanthi,
M. Tang, and S.-Y. Wu, "Reversible Electromechanical Characteristics of Carbon
Nanotubes under Local-Probe Manipulation," Nature, vol. 405, pp. 769-772, 2000.
[161] Tomioka, K., Y. Kobayashi, J. Motohisa, S. Hara, and T. Fukui, "Selective-area
growth of vertically aligned GaAs and GaAs/AlGaAs core–shell nanowires on
Si(111) substrate," Nanotechnology, vol. 20, p. 145302, 2009.
[162] Tomioka, K., J. Motohisa, S. Hara, K. Hiruma, and T. Fukui, "GaAs/AlGaAs
Core Multishell Nanowire-Based Light-Emitting Diodes on Si," Nano Letters, vol.
10, pp. 1639-1644, 2010/05/12 2010.
[163] Trentler, T. J., K. M. Hickman, S. C. Goel, A. M. Viano, P. C. Gibbons, and W. E.
Buhro, "Solution-Liquid-Solid Growth of Crystalline III-V Semiconductors: An
Analogy to Vapor-Liquid-Solid Growth," Science, vol. 270, pp. 1791-1794,
December 15, 1995 1995.
[164] Turan, R., D. D. Perovic, and D. C. Houghton, "Mapping electrically active
dopant profiles by field-emission scanning electron microscopy," Applied Physics
Letters, vol. 69, pp. 1593-1595, 1996.
133

[165] Venezuela, P., M. Lazzeri, and F. Mauri, "Theory of double-resonant Raman
spectra in graphene: Intensity and line shape of defect-induced and two-phonon
bands," Physical Review B, vol. 84, p. 035433, 2011.
[166] Vidano, R. P., D. B. Fischbach, L. J. Willis, and T. M. Loehr, "Observation of
Raman band shifting with excitation wavelength for carbons and graphites," Solid
State Communications, vol. 39, pp. 341-344, 1981.
[167] Wagner, L. F. and W. E. Spicer, "Observation of a Band of Silicon Surface States
Containing One Electron Per Surface Atom," Physical Review Letters, vol. 28, pp.
1381-1384, 1972.
[168] Walters, D. A., L. M. Ericson, M. J. Casavant, J. Liu, D. T. Colbert, K. A. Smith,
and R. E. Smalley, "Elastic strain of freely suspended single-wall carbon
nanotube ropes," Applied Physics Letters, vol. 74, pp. 3803-3805, Jun 1999.
[169] Walukiewicz, W., J. Lagowski, L. Jastrzebski, and H. C. Gatos, "Minority-carrier
mobility in p-type GaAs," Journal of Applied Physics, vol. 50, pp. 5040-5042,
1979.
[170] Wang, J., M. S. Gudiksen, X. Duan, Y. Cui, and C. M. Lieber, "Highly Polarized
Photoluminescence and Photodetection from Single Indium Phosphide
Nanowires," Science, vol. 293, pp. 1455-1457, August 24, 2001 2001.
[171] Wang, R.-p., G.-w. Zhou, Y.-l. Liu, S.-h. Pan, H.-z. Zhang, D.-p. Yu, and Z.
Zhang, "Raman spectral study of silicon nanowires: High-order scattering and
phonon confinement effects," Physical Review B, vol. 61, p. 16827, 2000.
[172] Wang, Y., D. C. Alsmeyer, and R. L. McCreery, "Raman spectroscopy of carbon
materials: structural basis of observed spectra," Chemistry of Materials, vol. 2, pp.
557-563, 1990/09/01 1990.
[173] Wasilewski, Z. R., M. M. Dion, D. J. Lockwood, P. Poole, R. W. Streater, and A.
J. SpringThorpe, "Composition of AlGaAs," Journal of Applied Physics, vol. 81,
pp. 1683-1694, 1997.
[174] Weber, E. R., H. Ennen, U. Kaufmann, J. Windscheif, J. Schneider, and T.
Wosinski, "Identification of As[sub Ga] antisites in plastically deformed GaAs,"
Journal of Applied Physics, vol. 53, pp. 6140-6143, 1982.
[175] Wu, G., J. Zhou, and J. Dong, "Raman Modes of the Deformed Single-Wall
Carbon Nanotubes," Physical Review B, vol. 72, p. 115411, 2005.
[176] Wu, X. L., G. G. Siu, S. Tong, X. N. Liu, F. Yan, S. S. Jiang, X. K. Zhang, and D.
Feng, "Raman scattering of alternating nanocrystalline silicon/amorphous silicon
multilayers," Applied Physics Letters, vol. 69, pp. 523-525, 1996.
[177] Wu, Y., M. Y. Huang, F. Wang, X. M. H. Huang, S. Rosenblatt, L. M. Huang, H.
G. Yan, S. P. O'Brien, J. Hone, and T. F. Heinz, "Determination of the Young's
Modulus of Structurally Defined Carbon Nanotubes," Nano Letters, vol. 8, pp.
4158-4161, Dec 2008.
[178] Yakobson, B. I. and R. E. Smalley, "Fullerene Nanotubes: C1,000,000 and
Beyond," American Scientist, vol. 85, p. 324, 1997.
[179] Yang, L. and J. Han, "Electronic Structure of Deformed Carbon Nanotubes,"
Physical Review Letters, vol. 85, p. 154, 2000.
134

[180] Yang, W., R.-Z. Wang, and H. Yan, "Strain-induced Raman-mode shift in single-
wall carbon nanotubes: Calculation of force constants from molecular-dynamics
simulations," Physical Review B, vol. 77, p. 195440, 2008.
[181] Yang, W., R. Z. Wang, and H. Yan, "Strain-induced Raman-mode shift in single-
wall carbon nanotubes: Calculation of force constants from molecular-dynamics
simulations," Physical Review B, vol. 77, p. 195440, May 2008.
[182] 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, vol. 84, pp. 5552-5555, Jun 2000.
[183] Zhao, Q. and H. Daniel Wagner, Phil. Trans. Royal Soc. London Series A-Math.
Phys. Engg. Sci., vol. 306, p. 2407, 2004.
[184] Zhao, Q. and H. Daniel Wagner, "Two-Dimensional Strain Mapping in Model
Fiber-Polymer Composites Using Nanotube Raman Sensing," Composites Part A:
Applied Science and Manufacturing, vol. 34, pp. 1219-1225, 2003.
[185] Zhao, Q., M. D. Frogley, and H. D. Wagner, "Direction-Sensitive Strain-Mapping
with Carbon Nanotube Sensors," Composites Science and Technology, vol. 62, pp.
147-150, 2002.
[186] Zhao, Q., M. B. Nardelli, and J. Bernholc, "Ultimate Strength of Carbon
Nanotubes: A Theoretical Study," Physical Review B, vol. 65, p. 144105, 2002.
[187] Zhao, Q. Z., M. B. Nardelli, and J. Bernholc, "Ultimate strength of carbon
nanotubes: A theoretical study," Physical Review B, vol. 65, p. 144105, Apr 2002.
[188] Zi, J., H. Buscher, C. Falter, W. Ludwig, K. Zhang, and X. Xie, "Raman shifts in
Si nanocrystals," Applied Physics Letters, vol. 69, pp. 200-202, 1996.

Abstract (if available)

Abstract This thesis includes three main topics covering a broad range of nanoscience: mechanical and optical properties of single walled carbon nanotubes (SWCNTs), optical manipulation and characterization of the crystal structure of silicon nanowires, and a systematic study of minority carrier dynamics in surface-passivated GaAs nanostructures, including nanowires and nanosheets. The physical properties of these materials and related phenomena are briefly reviewed in Chapter 1, including physical, electronic, and phonon structures of SWCNTs, light absorption and phonon confinement in nanocrystals, and surface states in semiconductors. ❧ In Chapter 2, we study the ultimate breaking strain of SWCNTs with in-situ Raman spectroscopy. Chemical vapor deposition (CVD) with a high flow rate is used to synthesize ultra-long suspended SWCNTs across an extendible slit. The G band Raman frequency downshifts linearly with strain and its downshifting rate spans a wide range from -6.2 to -23.6 cm⁻¹/%. This observation corroborates the theoretical prediction that the downshifting rate is strongly chiral dependent. A threshold downshift of ΔɷG75 cm-1 is observed for the G band lineshape broadening. In this experiment, we achieve strains up to 13.7±0.3% without slippage, breakage, and defect formation. ❧ In Chapter 3, we observe emergence of the D band Raman mode in SWCNTs under axial strain. The D to G mode Raman intensity ratio (ID/IG) is seen to increase with strain quadratically by more than a factor of 100. However, the ratio returns to its original pre-strain value, indicating that there is no permanent defect formation. Strain-induced symmetry lowering is proposed to explain the appearance of the D band without a real defect in the lattice. Another possible scenario is a strain-induced reversible transition from sp² to sp³ bond configuration in a twisted nanotube bundle. ❧ In Chapter 4, we develop a method of locally tailoring crystal structure of individual crystalline-amorphous core-shell silicon nanowires with a polarized laser spot. We are able to control the crystallinity of the silicon nanowires from 0 to 0.93 by controlling the incident laser power. Raman spectroscopy is used to determine the annealing temperature while annealing the nanowires and to characterize the crystal structure before and after annealing. High resolution transmission electron microscopy (HRTEM) is used to confirm the laser-induced crystallization. Due to the one-dimensional nature of nanowires, a strong polarization dependent heating/annealing is observed. The most efficient annealing occurs when the laser polarization is aligned along the axis of the nanowires. ❧ In Chapter 5, we focus on the minority carrier dynamics in AlxGa1-xAs-passivated GaAs nanowires. With passivation, the minority carrier diffusion length (Ldiff) increases from 30 nm to 180 nm, as measured by electron beam induced current (EBIC) mapping, and the photoluminescence (PL) lifetime increases from sub-60 ps to 1.3 ns. A 48-fold enhancement in the continuous-wave PL intensity is observed on the same individual nanowire with and without the AlᵪGa₁₋ᵪAs passivation layer, indicating a significant reduction in surface recombination. These results indicate that the minority carrier lifetime is not limited by twin stacking faults in passivated nanowires. From the PL lifetime and minority carrier diffusion length, we estimate the surface recombination velocity (SRV) to range from 1.7x10³ to 1.1x10⁴ cms⁻¹, and the minority carrier mobility μ is estimated to lie in the range from 10.3 to 67.5 cm²V⁻¹s⁻¹ for the passivated nanowires. ❧ In Chapter 6, we explore the interesting phenomena rising from the over-growth of GaAs nanosheets. Due to the shape anisotropy, we are able to grow stacking fault free nanosheets in the initial growth region. The twin-free crystal structure is confirmed by HRTEM images. However, clear boundaries appear at the interface between initial growth and over-growth regions in both EBIC and SEM contrast images. Possible scenarios, such as sudden changes in the crystal structure, doping, and surface quality, are discussed.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

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

PDF

GaAs nanowire optoelectronic and carbon nanotube electronic device applications

PDF

Growth, characterization of gallium arsenide based nanowires and application in photovoltaic cells

PDF

Photodetector: devices for optical data communication

PDF

Light Emission from Carbon Nanotubes and Two-Dimensional Materials

PDF

Light management in nanostructures: nanowire solar cells and optical epitaxial growth

PDF

Plasmonic enhancement of catalysis and solar energy conversion

PDF

Controlled synthesis, characterization and applications of carbon nanotubes

PDF

Design and characterization of metal and semiconducting nanostructures and nanodevices

PDF

Optoelectronic properties and device physics of individual suspended carbon nanotubes

PDF

Optoelectronic, thermoelectric, and photocatalytic properties of low dimensional materials

PDF

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

PDF

Large-scale molecular dynamics simulations of nano-structured materials

PDF

Hot electron driven photocatalysis and plasmonic sensing using metallic gratings

PDF

Modeling graphene: magnetic, transport and optical properties

PDF

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

PDF

Artificial photosynthesis on titanium oxide passivated III-V semiconductors

PDF

Fabrication and application of plasmonic nanostructures: a story of nano-fingers

PDF

Asymmetric response and the reorientation of interfacial water with respect to applied electric fields

PDF

Vibrational spectroscopy and molecular orientation at interfaces: correlating the structure of a material's surface to its properties

Asset Metadata

Creator Chang, Chia-Chi (author)

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

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Physics

Publication Date 11/22/2012

Defense Date 10/22/2012

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

Tag carbon nanotubes,defects,GaAs nanowires,OAI-PMH Harvest,Raman spectroscopy,silicon nanowires,solar cells,strain,surface passivation

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Cronin, Stephen B. (committee chair), Dapkus, Paul Daniel (committee member), Haas, Stephan W. (committee member), Nakano, Aiichiro (committee member), Zhou, Chongwu (committee member)

Creator Email ccchang05@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-118197

Unique identifier UC11291519

Identifier usctheses-c3-118197 (legacy record id)

Legacy Identifier etd-ChangChiaC-1337.pdf

Dmrecord 118197

Document Type Dissertation

Rights Chang, Chia-Chi

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