Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Synthesis and properties study of Q1D semiconductor nanostructures
(USC Thesis Other)
Synthesis and properties study of Q1D semiconductor nanostructures
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
SYNTHESIS AND PROPERTIES STUDY OF Q1D
SEMICONDUCTOR NANOSTRUCTURES
by
Liubing Huang
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
(MATERIALS SCIENCE)
August 2015
Copyright 2015 Liubing Huang
i
Acknowledgement
First of all, I would like to express deepest gratitude to my academic advisor, Dr. Jia G. Lu,
for her guidance, motivation, support and caring throughout my PhD study at USC. I would have
never been able to finish the research and thesis without her guidance and funding. She always
encourage me to think in-depth and independently to find out the fundamental physics underneath
the experimental data. Besides my advisor, I would like to thank the rest of my thesis committee:
Dr. Robert W. Hellwarth and Dr. Steven R. Nutt. Their comments and suggestions value a lot
while I wrote the thesis. I also want to thank Dr. Robert W. Hellwarth for his financial support for
these years. I also would like to thank Dr. Georg Mö thrath, for his guidance on project management
skills, discussions and support. I would like to thank Dr. Hans Bozler, for his support with
equipment set-up and repair. I would like to thank all of the group members for helpful discussion
on my research and collaboration. Especially, I would like to thank Dr. Paichun Chang, who guided
me into the nano research world and trained me for all the instruments in the lab. He is not only a
good guide but also a role model for me. Finally, I would like to thank my family and friends for
their support and encouragement.
ii
Table of Content
Acknowledgement ........................................................................................................................... i
List of Figures ................................................................................................................................ iv
Abstract .......................................................................................................................................... ix
Chapter 1: Background ................................................................................................................... 1
1.1 Significance of Q1D nanomaterials .......................................................................... 1
1.2 Synthesis of Q1D nanostructures .............................................................................. 2
1.3 Properties of InN, CdTe, Sb2Te3 nanowires ............................................................. 6
References of Chapter 1....................................................................................................... 9
Chapter 2: Quantum Transport in Indium Nitride Nanowires ...................................................... 15
2.1 Introduction ............................................................................................................. 15
2.2 Experiments and discussion .................................................................................... 15
2.3 Conclusion .............................................................................................................. 26
References of Chapter 2..................................................................................................... 26
Chapter 3: Structural and Optical Verification of Residual Strain Effect in Single Crystalline
CdTe Nanowires ........................................................................................................................... 30
3.1 Introduction ............................................................................................................. 30
3.2 Experiments ............................................................................................................ 31
3.3 Results and discussions ........................................................................................... 32
3.4 Conclusion .............................................................................................................. 44
References of Chapter 3..................................................................................................... 44
Chapter 4: Nature of AX Centers in Antimony Doped Cadmium Telluride Nanobelts ............... 47
4.1 Introduction ............................................................................................................. 47
4.2 Experiments ............................................................................................................ 49
4.3 Results and discussion ............................................................................................ 50
4.4 Conclusion .............................................................................................................. 65
References of Chapter 4..................................................................................................... 66
iii
Chapter 5: CVD Growth and Characterization of Sb2Te3 Nanowires .......................................... 70
5.1 Synthesis ................................................................................................................. 70
5.2 Characterizations..................................................................................................... 70
References of Chapter 5..................................................................................................... 73
Chapter 6: Applications and Future Research Outlook ................................................................ 74
6.1 CdTe nanowires applications and future works ...................................................... 74
6.2 Sb2Te3 nanowires future works ............................................................................... 80
References of Chapter 6..................................................................................................... 81
iv
List of Figures
Figure 2-1 (a) SEM image shows large quantity of InN nanowires grown on Si substrate.
Inset: EDS spectrum of the as-grown InN nanowires exhibits 1:1 stoichiometry. (b)
XRD pattern indexed to wurtzite InN. (c) TEM image of a single InN nanowire shows
high aspect ratio structure. (d) HRTEM image shows single crystalline wurtzite
structure with inter-plane distance ~0.307 nm, indicating the stacking direction along
[100]. Inset: corresponding FFT pattern of a hexagonal system. ........................................ 17
Figure 2-2 (a) Temperature dependent resistance measurement of a single InN nanowire
contacted by four probes (Inset shows the SEM image of the device). Scale bar is 5
µ m. (b) ln versus T
-1/4
at the temperature range of 17 - 80 K, showing a linear fitting
to 3D Mott VRH model. ...................................................................................................... 18
Figure 2-3 Magneto-resistance (MR) measurement with the magnetic field applied (a)
perpendicular and (b) parallel to the nanowire long axis at different temperatures. (c)
Linear fitting of
)] 0 , ( / ) , ( ln[ T H T
versus
2 / 1
H
under parallel magnetic field. ............ 22
Figure 3-1 The SEM image of as-grown CdTe nanowires; polygonal nanoparticles are found
on the tips of the nanowires. Insert: high resolution SEM image of the Au catalyst
shows a polygonal shape, verifying the VSS growth mechanism. ...................................... 33
Figure 3-2 (a) Low magnification TEM image of a CdTe nanowire showing a nanoparticle
on the tip of the nanowire. (b) HRTEM image shows the growth direction along <111>.
Insert: SAED pattern of the HRTEM image. (c) EDS spectrum of CdTe nanowire. (d)
EDS spectrum of the Au nanoparticle on the tip CdTe nanowires...................................... 34
Figure 3-3 XRD spectrum of CdTe nanowires. Miller indices of the zinc blende CdTe peak
are indicated. The peaks labelled by triangle originate from wurtzites CdTe, and the
peak labeled with diamond shape originates from Au catalytic particle. ............................ 35
v
Figure 3-4 (a) Photoluminescence spectrum of CdTe nanowires measured at 5.0 K. A
pronounce blue-shift in the bound exciton peak (peak #1) was revealed. (b) The band
structure for CdTe shows transitions processes: donor bound excitons (DX), Y band,
free electron to acceptor transitions (e, A), and donor-acceptor-pair transition DAP
originated from shallow acceptors and deep level Au acceptors. ....................................... 39
Figure 3-5 Temperature dependent photoluminescence spectra of CdTe nanowires ranging
from 5.0 K to 79.7 K. Insert: Temperature dependence of integrated peak intensity for
all five Gaussian fitted peaks (see Fig. 4 for peak assignment). ......................................... 40
Figure 3-6 Power dependent PL spectra with excitation laser power from 0.056 W/cm
2
to
38.9 W/cm
2
. ......................................................................................................................... 43
Figure 3-7 Peak position as a function of excitation power density for all five peaks. All the
PL spectra are Gaussian fitted and the corresponding peak positions are shown. When
PL is measured using excitation power density varying from 0.014 W/cm
2
to 9.73
W/cm
2
, no detectable red shift in PL peak position has been observed, indicating lack
of heating. And the blue shift is due to band filling effect as power density increases. ..... 43
Figure 4-1 (a) SEM image of as-grown CdTe:Sb nanobelts. The corresponding EDS
spectrum in the inset reveals a nearly stoichiometric atomic ratio of Cd and Te, while
the overall atomic incorporation ratio of Sb is found to be approximately 1.8 %. (b)
SEM image and EDS mapping of Cd, Te and Sb for one single CdTe:Sb nanobelt
indicating a homogenous distribution of Cd, Te and Sb along the whole channel. (c)
XRD spectra of undoped (black) and Sb doped (red) CdTe nanobelt ensembles. Spectra
are offset vertically for clarity. The diffraction peaks are indexed to zinc blende
structure of CdTe (JCPDS card No. 65-8879). (d) HRTEM image of one CdTe:Sb
nanobelt and the corresponding selective area diffraction (SAED) pattern is shown in
the inset demonstrating nanobelt growth along the [200] crystal axis. ............................... 51
Figure 4-2 Drain-source current Ids versus drain-source voltage Vds curves of a single
CdTe:Sb nanobelt field effect transistor (FET) for two-probe and four-probe
vi
measurements. The linear Ids - Vds curves measured in two-probe (black) and four-
probe (red) configuration overlap demonstrating good Ohmic contacts between the
nanobelt and the Cu/Au electrodes. The contact resistance is smaller than 1% of the
nanowire resistance and thus negligible. ............................................................................. 53
Figure 4-3 (a) The drain-source current Ids as function of drain-source voltage Vds for various
gate voltages Vg ranging from -20 to 20 V for a CdTe:Sb nanobelt FET. Inset shows
the Ids-Vg curve measured at constant Vds = 5 V at room temperature for the same device.
Evaluating the slope of the linear region gives an estimate for the hole mobility of 0.7
cm
2
V
-1
s
-1
and a hole concentration of 5.6× 10
16
cm
-3
. (b) Arrhenius plot of the
conductivity σ versus inverse temperature for one single CdTe:Sb nanobelt FET shows
unambiguously two linear regions, corresponding to two acceptors level that are
determined to be 56 ± 2 meV and 115 ± 4 meV above the valence band. .......................... 54
Figure 4-4 The PL spectrum of the Sb doped CdTe nanobelt ensemble measured at T = 23
K shows the same features as a single CdTe:Sb nanobelt measured at T = 21 K with
slightly different intensities. Therefore, it is valid to investigate the PL spectra of the
nanobelt ensemble at T = 4 K to match luminescence lines to certain radiative
transitions and conduct conclusions for optical recombinations in single nanobelts. ......... 58
Figure 4-5 Power-dependent nanobelt ensemble PL spectra of the emission around 1.45 eV
and their phonon replica at 4 K (spectra are offset vertically for clarity). A shift of the
peak positions to higher energies with increasing excitation power indicate a DAP type
transition which are attributed to A centers (EA3 ~ 130 meV) caused by defect
complexes of Cd vacancies and donors. .............................................................................. 58
Figure 4-6 (a) PL spectrum of the CdTe:Sb nanobelt ensemble, taken at 3.7 K with a
excitation density of 90 mW/m
2
at 633 nm. Several near band edge transitions are
labeled: the free exciton FX, donor bound excitons (D,X), acceptor bound excitons
(A,X) free electron to acceptor (e,A) transition, donor acceptor pair recombination
(DAP) and their phonon replica. (b) Schematic band diagram depicts the impurity
energy levels within the bandgap and recombination processes. One donor level and 3
vii
acceptor levels are involved, and the binding energies are determined to be ED = 9
meV, EA1 = 61 meV, EA2 = 109 meV and EA3 = 130 meV. ................................................. 59
Figure 4-7 Temperature-dependent PL spectra of CdTe:Sb nanobelt ensemble at excitation
of 633 nm and 65 mW/cm² . Spectra are offset vertically for clarity. Temperature
dependent measurements show a shift of the near band gap transitions to lower
energies with increasing temperature which is attributed to the Varshni shift.................... 60
Figure 4-8 (a) Persistent photoconductivity (PPC) measured at T = 100 K at constant voltage
V = 0.1 V with an illumination wavelength of 400 nm. (b) Dark current as a function
of temperature measured by first cooling down in dark (black squares), after
illumination for 10 minutes at 85 K, then warming up in dark (red circles). The
measurement verifies the existence of AX centers. (c) Configuration coordinate
diagram sketches acceptors at two different states as a function of the atomic
configuration. AX
+
represents the positively charged AX centers with large lattice
relaxation, while a
0
is the neutral acceptor state with negligible lattice distortion. The
microscopy structures of AX
+
and a
0
are drawn in the inset. (d) PPC decay curves
measured at various temperatures (100 K, 300 K, 390 K and 400 K). The data were
fitted by
PPC
I (t) exp( (t/ ) )
(as shown in solid curves). The inset shows the
obtained decay time constant τ versus inverse temperature 1000/T. At low temperatures
T < 300 K, τ reveals only weak temperature dependence. At high temperatures T > 300
K, τ shows an Arrhenius-like behavior with an activation energy of Ecap = 0.46 eV. ........ 63
Figure 5-1 (a) A SEM image of the as grown Sb2Te3 nanowires. Inset: the corresponding
EDS spectrum. (b) XRD spectrum of Sb2Te3 nanowires. (c) HRTEM image of a single
Sb2Te3 nanowire. Inset: the corresponding SAED. ............................................................. 72
Figure 5-2 Temperature dependent resistance measurement on single Sb2Te3 nanowires.
Inset: a SEM image of a single Sb2Te3 nanowires with Ti/Au contacts. ............................. 72
viii
Figure 6-1 (a) shows the configuration of a typical CdTe nanowiere photodetector with a
channel length of ~1.2 μm and a thickness of ~220 nm. (b) Photoresponse of CdTe
nanowires in air, under 450nm light illumination (intensity=200μW/cm
2
). ....................... 75
Figure 6-2. (a) I-V characteristics of CdTe nanowire under illumination of light with
different wavelength at bias=10V and wavelength λ=450nm. (b) Excess carrier density
vs. light intensity. (c) Gain vs. light intensity. (d) Responsivity vs. light intensity. ........... 77
Figure 6-3. (a) I-V characteristics of CdTe nanowire under illumination of light with
different wavelength. (b) PL and absorption spectrum of CdTe nanoribbons. (Reprint
from Nanoscale, 2012, 4, 2914). ......................................................................................... 78
Figure 6-4. Schematics of fabrication of CdTe/CdS core-shell structure via AAO template. ..... 79
Figure 6-5. (a) SEM image of a tilted AAO template with Sn catalyst deposited in the bottom.
(b) CdTe nanowires in AAO template................................................................................. 79
ix
Abstract
Nanoscience and nanotechnology studies materials at the nanoscale, on the order of 10
-9
meter.
There are many new physical phenomena and material properties when the dimension of materials
reduces to nanoscale. Therefore, it is of great importance for both physical exploration and
industrial application. Quasi-one-dimensional (Q1D) semiconductor nanostructures is specifically
interesting because they have great potential for electronic and optoelectronic devices. However,
to fabricate high quality Q1D materials and demonstrate precise control over materials properties
has not yet been mature for many important material systems. Also, lack of in-depth physical
understanding on thes nanostructures also holds back the development of nanodevices. In this
thesis, fabrication of different Q1D nanostructures (InN nanowires, CdTe nanowires, CdTe:Sb
nanobelts and Sb2Te3 nanowires) are demonstrated, and their structural, optical and electrical
properties are investigated systematically, which is important for future devices applications. The
thesis is organized as follow.
In Chapter 1, the background on Q1D nanostructures is introduced. The underlining
motivation and the key technologies for fabricating Q1D nanostructures is discussed. Also, the
basic material properties of the systems we studied is summarized.
Chapter 2 focus on the quantum transport in InN nanowires. InN nanowires grown CVD
method is characterized by SEM, TEM, XRD. Transport properties of single InN nanowires are
studied by varying temperature and magnetic field, revealing a variable range hopping transport
mechanism in low temperature region.
x
Chapter 3 discussed the strain effect in unintentionally doped CdTe nanowires by in-depth
analyzing the XRD and PL data. The experiment results also suggest that the optical properties are
sensitive to strain since strain alters the band structure.
In Chapter 4, we focus on the nature of AX centers in antimony doped cadmium telluride
nanobelts. Temperature-dependent conductivity, PL measurements and photoconductivity reveal
that AX center in CdTe:Sb nanobelts. The existence of AX center is proven for the first time
experimentally in the CdTe nanobelts system.
Chapter 5 focuses on growth and characterization (structural and electrical) of Sb2Te3
nanowires.
Finally, Chapter 6 discussed the device application and future research directions for CdTe
nanowires and Sb2Te3 nanowires.
1
Chapter 1: Background
1.1 Significance of Q1D nanomaterials
Nanoscience and nanotechnology are multidisciplinary fields of research conducted at the
nanoscale, on the order of 10
-9
meter. They not only offer great challenges in many of the
fundamental aspects, but also drive technological innovations that exert significant impact, in
electronics, photonics, spintronics, health care, to as far as space exploration. These
interdisciplinary fields are where tiny elementary entities are fabricated and investigated, such as
nanowires, nanotubes, nanoparticles, proteins, DNA, etc. The ability to synthesize these nanoscale
modular units and then to assemble them into larger structures with unique properties in charge
conduction, optical guiding, and magnetic ordering, is changing the way materials and devices are
produced.
Among various nanomaterials, structures with quasi-one-dimensionality (Q1D) have been a
research focal point ever since the discovery of carbon nanotube in 1991.
1
Due to the lack of
efficient methods to fabricate carbon nanotubes with specific electrical characteristics, tremendous
effort has been devoted to the study of semiconducting nanowires with high crystalline structures
and controllable physical properties. In the current trend of device miniaturization, these Q1D
nanostructures play an important role as the building blocks for devices that could overcome the
fundamental limits of microtechnology. Most material properties deviate from bulk behaviors
when the size of one dimension is reduced below a few hundred nanometers; and in nanowires this
change is much more pronounced because two dimensions are dramatically reduced. To a certain
extent, nanowires open a new system of materials since the properties get altered as a result of the
2
dimensional confinements. To explore this new material system, a fundamental understanding is
crucial. Only with a deeper comprehension of the basic characteristics of nanostructures, one can
better tailor their properties for designing and developing innovative devices. In essence, the
unique physical properties originate from the Q1D structures with highly anisotropic geometry
and enlarged surface areas, furthermore, when the size approaches fundamental length scales, such
as Bohr radius or localization length, quantum effects will reveal. Q1D structure possesses large
polarization anisotropies in optical absorption and emission mainly due to the geometric
anisotropy, leading to excellent polarization sensitive devices.
2
Moreover, Q1D materials system
has a much larger surface-to-volume ratio (inverse proportional to the radius) than its bulk
counterpart, thus dramatically enhancing functionalities that base on surface sensitivity.
3
On the
other side, in the quantum confinement region, electronic band structure is strongly size-dependent,
enabling a precise knob for fine-tuning the electrical, optical, and chemical properties.
4
The device
prototypes based on semiconductor nanowires have already demonstrated groundbreaking
advances, as they have been utilized as chemical sensors,
3, 5
field effect transistors,
3, 6, 7
photodetectors,
8-10
nanolasers,
11, 12
solar cells,
13
etc. Among semiconductor Q1D materials, InN,
CdTe and Sb2Te3 nanostructures have emerged to the research forefront because of its exceptional
properties and broad device applications. They are outstanding materials for developing
nanoelectronic, optoelelctronic, and electromechanical devices.
1.2 Synthesis of Q1D nanostructures
A variety of methods have been developed to synthesize Q1D nanostructures. According to
the synthesis environments, they can be mainly divided into two categories: solution phase growth
and vapor phase growth. Solution phase methods provide low temperature, cost-effective venues
3
for Q1D nanostructure growth. Various solution phase growth techniques have been successfully
implemented. They include electrochemical deposition,
14-19
chemical bath deposition,
20
sol-gel
method,
21
solvothermal,
22-31
sonothermal,
32
hydrothermal growth,
33, 34
etc. Vapor phase growth is
a well-adopted method to fabricate nanowires, formed by condensation of vaporized sources.
Various vapor phase growth techniques have been developed to fabricate cadmium chalcogenides
nanowires, including chemical vapor deposition (CVD),
35-44
metal-organic chemical vapor
deposition (MOCVD),
45, 46
physical vapor deposition (PVD),
47
pulse laser deposition (PLD),
48
molecular beam epitaxial (MBE),
49
etc. In general, the vapor phase growth of nanowires is
governed by two different growth mechanisms: vapor-liquid-solid (VLS) and vapor-solid (VS). In
our lab, AAO membrane assisted electrochemical growth and catalytic CVD vapor phase growth
are most commonly used. Therefore, these two fabrication mechanisms are reviewed here.
Anodic aluminum oxide (AAO) membrane that consists of hexagonally ordered nanopores
has been the most commonly used template for fabricating aligned semiconductor nanowire arrays
because of its chemical stability, insulating characteristics and good mechanical strength. More
importantly, the morphologies of AAO (pore size, pore-to-pore distance and channel length) can
be readily adjusted via anodization parameters, providing a straightforward route to fabricate
nanowires of various sizes, including those within the quantum confinement region ( < 10 nm in
diameter).
14
Nanowires have been fabricated by filling the pores of AAO by different deposition
techniques such as electrochemical deposition,
14, 18-21, 50-55
sol-gel method,
21
chemical bath
deposition,
20
etc. Among them, electrochemical deposition is the most important and widely used
method. However, AAO templates have intrinsic insulating barrier layers at the bottom, which
limit direct DC deposition. This issue can be either overcome by firstly thinning the barrier layer
4
to ~10 nm by gradually reducing anodization voltage and then performing AC deposition;
14, 18
or
by removing the barrier layer via chemical etching and then evaporating electrode for DC
deposition.
15-17
Both techniques have advantages and disadvantages. In DC deposition, removing
the barrier requires more complicated processes, but it provides a more stable and uniform
deposition. In both techniques, deposition rate depends highly on pore size, deposition voltage and
current, concentration and temperature of electrolytes. Therefore, selection of electrolytes and
control of the deposition parameters are critical to ensure a successful fabrication.
The vapor–liquid–solid (VLS) mechanism was first proposed in 1964 by Wagner and Ellis as
an explanation for the Si whisker growth,
56
which was then already widely applied for various
whisker material growth. Metal nanoparticles, typically Au, are used in the VLS processes, which
act as catalysts that facilitate uni-directional material growth. The VLS growth consists of several
processes: A metal catalyst first absorbs source vapor and forms a eutectic alloy in the form of a
liquid droplet; The liquid droplet has a much larger sticking coefficient and thus serves as a
preferred site for deposition from the source vapor, Upon super-saturation, nucleation starts and
drives the source to precipitate at the liquid alloy-solid interface and triggers the axial growth. The
growth involves subsequently vapor, liquid and solid phases, thus is referred to as the VLS growth
mechanism. It is essential to have eutectic alloy droplets in such VLS processes, therefore,
equilibrium phase diagrams are essential for selecting optimal catalysts and growth conditions.
Morphologies of nanowires grown by VLS depend upon sizes and physical properties of the liquid
alloy. The resulted nanowires exhibit high crystalline qualities because of epitaxial growth. On the
other hand, VLS growth can also occur without additional metal nanoparticles, via the so-called
self-catalyst VLS growth. For example, CdS nanowires have been fabricated in a self-catalytic
method, where S deficit CdS1-x liquid droplets serve as catalysts.
57
5
Furthermore, AAO template directed VLS growth also attracted a great research attention
since it is advantageous to fabricate vertical aligned nanowire arrays with high material quality
and controllable geometries.
58
Metal nanoparticles are first deposited into the AAO channels by
electrochemical deposition, and they subsequently catalyze nanowire growth along the cylindrical
channels in a vapor transport process. The VLS grown nanowire array usually possesses better
crystallinity than the ones grown by electrochemical deposition in AAO, offers a platform for
making high performance devices. However, to avoid deformation of AAO templates, the growth
temperature is restricted to ~600 ° C, which sets limitation for materials that can be fabricated by
this method.
The vapor-solid (VS) growth is another important method, in which solid nanowires are
formed by direct condensation from vapor sources. In the above mentioned VLS growth,
unavoidable doping from foreign catalysts may affect the electrical and optical properties. For
example, Au is found to induced a deep impurity level in CdTe nanowires.
59
In contrast, VS growth
does not involve metal catalysts, thus it is preferable when metal impurities in materials cause a
major concern. CdS nanowires
47, 60-62
, CdSe nanowires
63
and CdTe nanowires
64-66
have been
successfully synthesized by VS growth. However, the fundamental mechanism is not yet well
established. Some experimental work has suggested that screw dislocations are responsible for the
uni-directional growth.
67, 68
And some have proposed that after the initial nucleation of vapor
source, the vapor molecules preferably deposit on the tip due to the Gibbs-Thomson effect, driving
the axial growth.
47
Another possible mechanism suggested is that nanowire growth becomes
energetically favorable when the precursor vapor pressure is so low that it is insufficient to
facilitate film growth.
62
6
1.3 Properties of InN, CdTe, Sb2Te3 nanowires
InN nanowires
Indium nitride (InN) is a direct bandgap semiconductor material and its bandgap has been
established to be ~0.7 eV. When alloyed with gallium nitride, the ternary system InGaN has a
direct bandgap span from the ultraviolet (3.4 eV) to the infrared (0.69 eV). Therefore, it has great
potential in photoelectronic device applications such as LED and Photovoltaics. Indium nitride has
high conductivity and electron mobility (3200 cm
2
/(Vs) at 300 K), therefore, it has attracted great
interest for high speed electronic components. InN nanostructures such as nanowires are
particularly tantalizing due to the potential of integrating the abovementioned properties with
various quantum effects. Therefore, high quality InN nanowires are synthesized and their transport
properties are investigated in Chapter 2.
CdTe nanowires
Cadmium telluride (CdTe) has attracted enormous research interest because of its exceptional
optical and electrical properties for developing optoelectronic devices. In particular, it holds
promising potential for low-cost photovoltaics, due to its high absorption coefficient and a direct
band gap of 1.49 eV at room temperature, which is close to the theoretically calculated optimal
value for solar cells under AM 1.5 irradiation. Vapor grown CdTe nanowires are single crystalline
zinc blend structure with lattice constant a0 = 0.648. In addition, wurzite CdTe nanowires with
kinks are also fabricated. The growth directions of the segments are respectively along [111] and
[100] direction.
69
Such configuration may lead to novel device design and applications. Moreover,
Luo et al. have prepared both zinc blend and wurzite structure CdTe nanowire arrays by sputtering
method.
65
It is reported that pure zinc blende phase nanowire arrays that stack along the [111]
7
direction are obtained at a high growth temperature with a low deposition rate. Conversely,
wurtzite phase CdTe nanowire arrays with growth in the [0002] direction are obtained at a lower
growth temperature with a higher deposition rate. A Gibbs free energy nucleation model is applied
to explain the formation of these different crystal phases under the employed growth conditions.
Furthermore, very narrow CdTe nanowires fabricated by SLS mechanism are commonly found to
be admixtures of zinc blend and wurtzite phases, similar to the CdS nanowires obtained via SLS
method.
70
Electrical and optical properties of CdTe nanowires have been investigated but more scientific
effort are still needed to get deeper understanding. Undoped CdTe nanowires exhibit high
resistivity
69
with low hole concentration presence due to intrinsic defects such as Cd vacancies
(VCd) and defect complexes involving VCd. However, the impurity levels are not well-defined and
the strain effect in the undoped CdTe nanowires are rarely studied, therefore, we carried out a
comprehensive study on the optical and structural properties of undoped CdTe nanowires, as
discussed in Chapter 3. In order to enhance the electronic property and device performance,
efficient doping is critical, and the preserving difficulty lies in achieving low resistivity as well as
high mobility. So it is important to understand the fundamental mechanisms of impurity doping.
Among the work on CdTe nanostructures doped by Zn
71-73
, Sb
9, 74, 75
, or Mn
76
, Sb has been proven
to be an effective p-type dopant. Controlled p-type doping has been achieved by co-evaporation
of Sb and CdTe.
75
The conductivity increases by two order of magnitude after Sb doping. Hole
concentration is measured to be around 10
16
- 10
17
cm
-3
even with high concentration of Sb dopants.
The physical origin of self-compensation in CdTe:Sb nanostructure is still a puzzle for the
community. Therefore, we investigated the nature of Sb dopants in CdTe nanowires, and have
8
found surprisingly that in addition to shallow acceptor state, Sb induces deep level AX centers,
resulting in significant compensations, as discussed in Chapter 4.
Sb2Te3 nanowires
Topological insulators (TI) are electronic materials that have a bulk band gap like an ordinary
insulator but have protected conducting states on their edge or surface. The surface Dirac fermion
states are protected by time reversal symmetry and their spin are lock at a right angle due to strong
spin-orbital coupling
77
. Such unique properties make it interesting to investigate for both
fundamental physics and device applications. Exploration and control of the surface states
important because of its potential application on improved spintronic devices and potentially
useful for quantum computing. Single crystalline nanostructure (nanowires, nanoribbon etc) offers
an attractive system to study the exotic surface states because of (1) the large surface-to-volume
ratio and (2) bandgap engineering due to size confinement effect.
Sb2Te3 is one of the 3D topological insulator materials yet discovered, which has bulk
bandgap of 0.28eV and simple surface states consisting of a single Dirac cone existing in the band
gap
78
. Therefore, Sb2Te3 nanowire is an ideal system to reveal the topological phenomena. Sb2Te3
nanowires have been synthesized by electrochemical deposition
79
, VLS/VS growth mechanisms
80
.
However, studying their topological insulator properties begins only in the recent years, therefore,
to understand their transport properties still require great research efforts. To probe the topological
insulting states by transport measurement is challenging because of the residual bulk conduction
related to disorder or unintentional doping. It is widely suggested that counter doping would be
helpful to suppress the bulk conduction
81
, and thus provide a clean system to study the topological
9
insulator surface states. To investigate the intrinsic properties of Sb2Te3 nanowires, it is important
to growth high quality single crystalline nanowires and characterize the physical properties
comprehensively.
References of Chapter 1
1. Iijima, S. Nature 1991, 354, (6348), 56-58.
2. Robert, R.; Daniel, P.; Arian, K.; Robert, B.; Sebastian, G.; Ulf, P.; Carsten, R. Journal of
Physics D: Applied Physics 2014, 47, (39), 394012.
3. Fan, Z.; Wang, D.; Chang, P.-C.; Tseng, W.-Y.; Lu, J. G. Appl Phys Lett 2004, 85, (24),
5923-5925.
4. Blö mers, C.; Lu, J. G.; Huang, L.; Witte, C.; Grü tzmacher, D.; Lü th, H.; Schä pers, T. Nano
Lett 2012, 12, (6), 2768-2772.
5. Fan, Z.; Lu, J. G. Appl Phys Lett 2005, 86, (12), 123510.
6. Chang, P. C.; Fan, Z.; Chien, C. J.; Stichtenoth, D.; Ronning, C.; Lu, J. G. Appl Phys Lett
2006, 89, (13), 133113.
7. Huang, L.; Li, D.; Chang, P.; Chu, S.; Bozler, H.; Beloborodov, I. S.; Lu, J. G. Phys Rev B
2011, 83, (24), 245310.
8. Xie, X.; Kwok, S.-Y.; Lu, Z.; Liu, Y.; Cao, Y.; Luo, L.; Zapien, J. A.; Bello, I.; Lee, C.-S.;
Lee, S.-T.; Zhang, W. Nanoscale 2012, 4, (9), 2914-2919.
9. Xie, C.; Luo, L. B.; Zeng, L. H.; Zhu, L.; Chen, J. J.; Nie, B.; Hu, J. G.; Li, Q.; Wu, C. Y.;
Wang, L.; Jie, J. S. Crystengcomm 2012, 14, (21), 7222-7228.
10. Yang, G.; Kim, B.-J.; Kim, D.; Kim, J. Opt Express 2014, 22, (16), 18843-18848.
11. Geburt, S.; Thielmann, A.; Roder, R.; Borschel, C.; McDonnell, A.; Kozlik, M.; Kuhnel,
J.; Sunter, K. A.; Capasso, F.; Ronning, C. Nanotechnology 2012, 23, (36), 365204.
12. Rö der, R.; Wille, M.; Geburt, S.; Rensberg, J.; Zhang, M.; Lu, J. G.; Capasso, F.;
Buschlinger, R.; Peschel, U.; Ronning, C. Nano Lett 2013, 13, (8), 3602-3606.
10
13. Fan, Z. Y.; Razavi, H.; Do, J. W.; Moriwaki, A.; Ergen, O.; Chueh, Y. L.; Leu, P. W.; Ho,
J. C.; Takahashi, T.; Reichertz, L. A.; Neale, S.; Yu, K.; Wu, M.; Ager, J. W.; Javey, A.
Nature Materials 2009, 8, (8), 648-653.
14. Routkevitch, D.; Bigioni, T.; Moskovits, M.; Xu, J. M. The Journal of Physical Chemistry
1996, 100, (33), 14037-14047.
15. Xu, D. S.; Xu, Y. J.; Chen, D. P.; Guo, G. L.; Gui, L. L.; Tang, Y. Q. Adv Mater 2000, 12,
(7), 520-522.
16. Mondal, S. P.; Das, K.; Dhar, A.; Ray, S. K. Nanotechnology 2007, 18, (9), 095606.
17. Zhao, Y.; Yang, X. C.; Huang, W. H.; Zou, X.; Lu, Z. G. J Mater Sci 2010, 45, (7), 1803-
1808.
18. Yang, Y.; Chen, H. L.; Mei, Y. F.; Chen, J. B.; Wu, X. L.; Bao, X. M. Solid State Commun
2002, 123, (6-7), 279-282.
19. Klein, J. D.; Herrick, R. D.; Palmer, D.; Sailor, M. J.; Brumlik, C. J.; Martin, C. R. Chem
Mater 1993, 5, (7), 902-904.
20. Zhang, H.; Ma, X. Y.; Xu, J.; Niu, J. J.; Sha, J.; Yang, D. R. J Cryst Growth 2002, 246, (1-
2), 108-112.
21. Cao, H. Q.; Xu, Y.; Hong, J. M.; Liu, H. B.; Yin, G.; Li, B. L.; Tie, C. Y.; Xu, Z. Adv Mater
2001, 13, (18), 1393-1394.
22. Shi, H. Q.; Zhou, X. D.; Fu, X.; Wang, D. B.; Hu, Z. S. Mater Lett 2006, 60, (15), 1793-
1795.
23. Datta, A.; Chavan, P. G.; Sheini, F. J.; More, M. A.; Joag, D. S.; Patra, A. Cryst Growth
Des 2009, 9, (9), 4157-4162.
24. Fan, L. B.; Feng, T. H.; Wang, P.; Feng, Z. B.; Zhang, C. L. Aust J Chem 2009, 62, (5),
448-454.
25. Yan, S. C.; Sun, L. T.; Qu, P.; Huang, N. P.; Song, Y. C.; Xiao, Z. D. J Solid State Chem
2009, 182, (10), 2941-2945.
26. Mandi, M. A.; Hassan, J. J.; Ng, S. S.; Hassan, Z. J Cryst Growth 2012, 359, 43-48.
11
27. Yan, S. C.; Xiao, Z. D.; Shi, Y.; Hu, D.; Xu, X.; Wu, J. S.; Zhou, M. M. Nanosci Nanotech
Let 2013, 5, (2), 213-221.
28. Zhao, J. G.; Hua, Z. H.; Yao, Y. Superlattice Microst 2013, 61, 146-151.
29. Hadia, N. M. A.; Garcia-Granda, S.; Garcia, J. R. J Nanosci Nanotechno 2014, 14, (7),
5449-5454.
30. Kazeminezhad, I.; Hekmat, N.; Kiasat, A. Fiber Polym 2014, 15, (4), 672-679.
31. Mahdi, M. A.; Hassan, J. J.; Kasim, S. J.; Ng, S. S.; Hassan, Z. B Mater Sci 2014, 37, (2),
337-345.
32. Jiang, L. P.; Xu, S.; Miao, J. J.; Wang, H.; Zhu, J. J. J Nanosci Nanotechno 2006, 6, (8),
2584-2587.
33. Fan, H. M.; Ni, Z. H.; Feng, Y. P.; Fan, X. F.; Kuo, J. L.; Shen, Z. X.; Zou, B. S. Appl Phys
Lett 2007, 91, (17), 171911.
34. Murali, G.; Reddy, D. A.; Giribabu, G.; Vijayalakshmi, R. P.; Venugopal, R. J Alloy
Compd 2013, 581, 849-855.
35. Wang, Y. W.; Meng, G. W.; Zhang, L. D.; Liang, C. H.; Zhang, J. Chem Mater 2002, 14,
(4), 1773-1777.
36. Barrelet, C. J.; Wu, Y.; Bell, D. C.; Lieber, C. M. J Am Chem Soc 2003, 125, (38), 11498-
11499.
37. Dong, L. F.; Jiao, J.; Coulter, M.; Love, L. Chem Phys Lett 2003, 376, (5-6), 653-658.
38. Li, C.; Liu, Z. T.; Yang, Y. Nanotechnology 2006, 17, (8), 1851-1857.
39. Chavan, P. G.; Kashid, R. V.; Badhade, S. S.; Mulla, I. S.; More, M. A.; Joag, D. S. Vacuum
2014, 101, 38-45.
40. Ma, C.; Wang, Z. L. Adv Mater 2005, 17, (21), 2635-2639.
41. Dai, G. Z.; Zhang, Q. L.; Peng, Z. W.; Zhou, W. C.; Xia, M. X.; Wan, Q.; Pan, A. L.; Zou,
B. S. J Phys D Appl Phys 2008, 41, (13), 135301.
42. Wang, M.; Fei, G. T. Nanoscale Res Lett 2009, 4, (10), 1166-1170.
12
43. Xiao, B. B.; Xu, Y. B. Physica E 2011, 44, (3), 696-699.
44. Yan, Y.; Liao, Z. M.; Bie, Y. Q.; Wu, H. C.; Zhou, Y. B.; Fu, X. W.; Yu, D. P. Appl Phys
Lett 2011, 99, (10), 103103.
45. Shan, C. X.; Liu, Z.; Hark, S. K. Appl Phys Lett 2007, 90, (19), 193123.
46. Shan, C. X.; Liu, Z.; Hark, S. K. Phys Rev B 2006, 74, (15), 153402.
47. Ye, C. H.; Meng, G. W.; Wang, Y. H.; Jiang, Z.; Zhang, L. D. J Phys Chem B 2002, 106,
(40), 10338-10341.
48. Lee, K. Y.; Lim, J. R.; Rho, H.; Choi, Y. J.; Choi, K. J.; Park, J. G. Appl Phys Lett 2007,
91, (20), 201901.
49. Bellet-Amalric, E.; Elouneg-Jamroz, M.; Rueda-Fonseca, P.; Bounouar, S.; Den Hertog,
M.; Bougerol, C.; Andre, R.; Genuist, Y.; Poizat, J. P.; Kheng, K.; Cibert, J.; Tatarenko, S.
J Cryst Growth 2013, 378, 233-237.
50. Simpkins, B. S.; Brintlinger, T.; Stroud, R. M.; Sherrill, S.; Lee, S. B.; Pehrsson, P. E. J
Phys Chem C 2013, 117, (22), 11843-11849.
51. Hu, Z. D.; Yu, Y.; Hu, H. N. Mater Lett 2010, 64, (7), 863-865.
52. Zeng, Y.; Su, Y. K.; Wu, X. M.; Tang, J. N. Rare Metal Mat Eng 2014, 43, (2), 413-417.
53. Xu, D.; Guo, Y.; Yu, D.; Guo, G.; Tang, Y.; Yu, D. P. Journal of Materials Research 2002,
17, (07), 1711-1714.
54. Zhao, A. W.; Meng, G. W.; Zhang, L. D.; Gao, T.; Sun, S. H.; Pang, Y. T. Appl Phys A
2003, 76, (4), 537-539.
55. Hackney, Z.; Mair, L.; Skinner, K.; Washburn, S. Mater Lett 2010, 64, (18), 2016-2018.
56. Wagner, R. S.; Ellis, W. C. Appl Phys Lett 1964, 4, (5), 89-90.
57. Wu, X. C.; Tao, Y. R. J Cryst Growth 2002, 242, (3-4), 309-312.
58. Fan, Z.; Razavi, H.; Do, J.-w.; Moriwaki, A.; Ergen, O.; Chueh, Y.-L.; Leu, P. W.; Ho, J.
C.; Takahashi, T.; Reichertz, L. A.; Neale, S.; Yu, K.; Wu, M.; Ager, J. W.; Javey, A. Nat
Mater 2009, 8, (8), 648-653.
13
59. Huang, L.; Lu, S.; Chang, P.; Banerjee, K.; Hellwarth, R.; Lu, J. Nano Res 2014, 7, (2),
228-235.
60. Zhou, S. M. Physica Status Solidi a-Applications and Materials Science 2006, 203, (6),
R45-R47.
61. Pan, H.; Xing, G. C.; Ni, Z. H.; Ji, W.; Feng, Y. P.; Tang, Z.; Chua, D. H. C.; Lin, J.; Shen,
Z. X. Appl Phys Lett 2007, 91, (19), 193105.
62. Lin, Y. F.; Hsu, Y. J.; Lu, S. Y.; Kung, S. C. Chem Commun 2006, (22), 2391-2393.
63. Fasoli, A.; Colli, A.; Martelli, F.; Pisana, S.; Tan, P. H.; Ferrari, A. C. Nano Res 2011, 4,
(4), 343-359.
64. Wang, X.; Wang, J.; Zhou, M.; Zhu, H.; Wang, H.; Cui, X.; Xiao, X.; Li, Q. The Journal
of Physical Chemistry C 2009, 113, (39), 16951-16953.
65. Luo, B.; Deng, Y.; Wang, Y.; Tan, M.; Cao, L.; Zhu, W. Crystengcomm 2012, 14, (23),
7922-7928.
66. Hou, J.; Yang, X.; Lv, X.; Peng, D.; Huang, M.; Wang, Q. Appl Surf Sci 2011, 257, (17),
7684-7688.
67. Zhang, H. Z.; Kong, Y. C.; Wang, Y. Z.; Du, X.; Bai, Z. G.; Wang, J. J.; Yu, D. P.; Ding,
Y.; Hang, Q. L.; Feng, S. Q. Solid State Commun 1999, 109, (11), 677-682.
68. Bierman, M. J.; Lau, Y. K. A.; Kvit, A. V.; Schmitt, A. L.; Jin, S. Science 2008, 320, (5879),
1060-1063.
69. Ye, Y.; Dai, L.; Sun, T.; You, L. P.; Zhu, R.; Gao, J. Y.; Peng, R. M.; Yu, D. P.; Qin, G.
G. J Appl Phys 2010, 108, (4), 044301.
70. Kuno, M.; Ahmad, O.; Protasenko, V.; Bacinello, D.; Kosel, T. H. Chem Mater 2006, 18,
(24), 5722-5732.
71. Zhou, S. M.; Zhang, X. H.; Meng, X. M.; Wu, S. K.; Lee, S. T. Appl Phys a-Mater 2005,
81, (8), 1647-1650.
72. Zhou, S. M. Phys Low-Dimens Str 2006, 2, 29-33.
73. Gandhi, T.; Raja, K. S.; Misra, M. Electrochim Acta 2006, 51, (26), 5932-5942.
14
74. Chao, X.; Biao, N.; Long, Z.; Long-Hui, Z.; Yong-Qiang, Y.; Xian-He, W.; Qun-Ling, F.;
Lin-Bao, L.; Yu-Cheng, W. Nanotechnology 2013, 24, (35), 355203.
75. Zhu, L.; Jie, J.; Wu, D.; Luo, L.; Wu, C.; Zhu, Z.; Yu, Y.; Wang, L. Journal of
Nanoengineering and Nanomanufacturing 2012, 2, (2), 191-196.
76. Ramasamy, P.; Mamum, S. I.; Jang, J.; Kim, J. Crystengcomm 2013, 15, (11), 2061-2066.
77. Hasan, M. Z.; Kane, C. L. Reviews of Modern Physics 2010, 82, (4), 3045-3067.
78. Zhang, H.; Liu, C.-X.; Qi, X.-L.; Dai, X.; Fang, Z.; Zhang, S.-C. Nat Phys 2009, 5, (6),
438-442.
79. Jin, C.; Zhang, G.; Qian, T.; Li, X.; Yao, Z. The Journal of Physical Chemistry B 2005,
109, (4), 1430-1432.
80. Lee, J. S.; Brittman, S.; Yu, D.; Park, H. Journal of the American Chemical Society 2008,
130, (19), 6252-6258.
81. Takagaki, Y.; Jahn, U.; Ramsteiner, M. Semiconductor Science and Technology 2012, 27,
(8), 085006.
15
Chapter 2: Quantum Transport in Indium Nitride Nanowires
2.1 Introduction
Charge transport properties of low dimensional systems are of profound interest due to the
quantum mechanical phenomena that occur when their sizes reduce to nanometer scales.
1-3
Advance in nanofabrication has opened the pathway to probe the fundamental properties
manifested in such strongly confined systems. Among them, quasi-one-dimensional (quasi-1D)
materials (e.g. nanowires and nanotubes) synthesized via bottom-up technology with highly
crystalline structures and highly anisotropic geometries are excellent candidates for understanding
quantum transport properties and for developing potential nanoelectronic devices.
2, 4-7
Among the
semiconductor nanowires, indium-based III-V semiconductors, such as InN, InAs and InSb, are
known to have high conductivity and extremely high electron mobility.
8-13
Thus, it is of significant
interest to investigate their quantum transport properties in order to evaluate them as the future
building blocks for high-speed electronic devices.
2.2 Experiments and discussion
InN nanowires are synthesized by a low-pressure chemical vapor deposition (CVD) method.
The source InN powder is placed in an alumina boat at the center of a horizontal tube furnace,
where the temperature is elevated to 580 ° C in 10 min under continuous ammonia (NH3) flow at
50 sccm in a controlled pressure of 1 torr. Silicon substrate coated with 10 nm Au catalyst film is
placed close to the source powder. Figure 2-1 (a) shows a scanning electron microscope (SEM)
image of InN nanowires with diameter ranging from 50 to 200 nm. The stoichiometry of the as-
16
grown InN nanowires is investigated by energy dispersive X-ray spectroscopy (EDS) analysis. As
shown in the inset of Figure 2-1 (a), distinct indium (In) and nitrogen (N) peaks confirm the
elemental composition, while the silicon (Si) peak originates from the substrate. The atomic ratio
of In:N is close to 1:1, suggesting good stoichiometry of the InN nanowires. Crystal structure of
the nanowires is examined by X-ray diffraction (XRD) pattern as plotted in Figure 2-1 (b). All
diffraction peaks match well with those of the standard wurtzite structure InN (JCPDS file No. 50-
1239). TEM image of a single InN nanowire demonstrating high aspect ratio structure is shown in
Figure 2-1 (c). High-resolution transmission electron microscopy (HRTEM image in Figure 2-1
(d)) analysis is also performed to further examine the crystal structure. It displays a single
crystalline system with pure close-packed wurtzite structure. The nanowire grows along [100]
direction and the interplanar spacing of the (100) planes is estimated to be 0.307 nm, which
matches that of InN single crystal with a lattice constant a= 0.354 nm.
14
The Fast Fourier
Transform (FFT) result (inset in Figure 2-1 (d)) shows typical diffraction pattern of the hexagonal
system at the [001] projection direction. The InN nanowires suspended in solution are then
dispersed onto a Si/SiO2 substrate for subsequent contact electrode patterning. Photolithography
followed by metallization of Pd/Au is processed to define four-probe contacts with equal spacing
(as depicted in Figure 2-2 (a) inset). Palladium (Pd) is selected as the adhesive contact metal to
form ohmic electrical contacts because its work function (5.12-5.65 eV)
15, 16
matches with the large
electron affinity of InN (5.8 eV).
17, 18
17
Figure 2-1 (a) SEM image shows large quantity of InN nanowires grown on Si substrate. Inset: EDS
spectrum of the as-grown InN nanowires exhibits 1:1 stoichiometry. (b) XRD pattern indexed to wurtzite
InN. (c) TEM image of a single InN nanowire shows high aspect ratio structure. (d) HRTEM image shows
single crystalline wurtzite structure with inter-plane distance ~0.307 nm, indicating the stacking direction
along [100]. Inset: corresponding FFT pattern of a hexagonal system.
18
Figure 2-2 (a) Temperature dependent resistance measurement of a single InN nanowire contacted by four
probes (Inset shows the SEM image of the device). Scale bar is 5 µ m. (b) ln versus T
-1/4
at the temperature
range of 17 - 80 K, showing a linear fitting to 3D Mott VRH model.
19
Figure 2-2 (a) shows the resistance change of an individual InN nanowire (diameter ~117
nm) in the measurement temperature range of 5 - 300k. First, one observes in the temperature
range from 15 to 80 K, the resistance decreases monotonically with temperature rising from 15 to
80 K, namely an insulating characteristic. The conduction at this range of temperature is governed
by variable range hopping mechanism. Next, a distinct semiconductor-to-metal transition is
observed around 80 K. Above this transition temperature, enhanced scatterings due to Coulomb
interactions play a dominant role, giving rise to a metallic-like characteristic. This transition has
been reported in other semiconducting nanowires.
19
There have been previous transport
measurements on InN nanowire showing metallic behavior
20
at low temperature range, indicating
the formation of an accumulation layer. The origin of its existence is not yet clear.
20-26
In contrast,
n-type semiconducting behavior
27
has also been demonstrated. Likewise in our work, a
semiconductor-to-metal transition is observed, suggesting that the Fermi level is situated at the
donor states below the conduction band edge Ec. Under this context, electrons localized close to
the Femi level conduct via hopping at low temperatures from one localized state to the other. The
InN nanowire presented here has a diameter of 117 nm, much larger than the Bohr radius (~4.6
nm)
28, 29
, so that the electron transport reveals bulk 3D conduction.
To understand quantitatively the hopping transport mechanism, we will focus on the hopping
regime for nanowire system with diameter D satisfying the inequality:
B hop
a T r D ) (
(1)
where
) / ( ) (
0 0
T T T r
hop
is the electron hopping length, with being the numerical
constant discussed below in equation (2);
0
is the localization length deep in the insulating regime.
B
a ~
0
, where
B
a is the Bohr radius; and
0
T is the characteristic temperature scale discussed
20
below in equation (3). When this inequality Eq. (1) is satisfied, the wire can be considered as a 3D
system
38
. At low temperatures, the typical hopping distance
0
10 ~
hop
r , so that the hopping
conduction (i.e. Eq. (1)) is easy to satisfy. And the temperature dependent resistivity can be
expressed as
33
] ) / exp[(
0 0
T T (2)
where the power depends on different hopping conductivity mechanisms: 2 / 1 at very
low temperature (Efros-Shklovskii varible range hopping (ES VRH) regime), ) 1 /( 1 d at
relatively higher temperature (Mott VRH regime) where d is the dimensionality; and 1 at even
higher temperature (nearest neighbor hopping (NNH) regime).
To clarify the hopping mechanism, the crossover temperature between ES to Mott VRH is
first considered. The crossover temperature is given by:
) 1 /( ) 1 (
) (
d d
M
ES
M cross
T
T
T T (3)
where TM and TES are the characteristic temperature scales depending on the localization
length and the density of states 0 at the Fermi level,
30, 31, 33
d M
T
0 0
1
is the characteristic
temperature in the Mott VRH model. And
0 0
2
4
e
T
ES
is the characteristic temperature in the
ES VRH model. Taking d=3 gives the crossover temperature K T T T
M ES cross
5 /
2
, which is
below the temperature range in our experiment. Next, to distinguish Mott VRH from the nearest
neighbor hopping, we plot ) ln( versus T
-1
, which is not a straight line, so the NN hopping can be
easily ruled out. Thus Mott VRH model is valid in our InN nanowires. Figure 2-2 (b) plots ) ln(
21
versus T
-1/4
at low temperature range (17 - 80 K), fitting well with a linear expression. This fitting
confirms Mott VRH hopping model, and shows that the electron conduction through the
nanochannel follows bulk 3D hopping transport model. Therefore, the temperature dependent
resistivity can be written as:
] ) / exp[(
4 / 1
0
T T
M
(4)
To further explore the transport mechanisms of the electrons at low temperature, magneto-
resistance (MR) measurements are carried out by applying external magnetic fields ranging from
-5 to 5 T at temperatures from 4.2 to 58 K, which is in the vicinity of semiconductor-to-metal
transition temperature in terms of the phonon energy. Figure 2-3 (a) and (b) show the MR data
with the magnetic field perpendicular and parallel to the nanowire long axis, respectively, both
exhibiting negative MR (i.e. resistances decrease with increasing magnetic field)
32
. The negative
MR effect becomes less pronounced with increasing temperature due to shorter hopping distance,
4
0
/ ) ( T T T r
M hop
. From Figs. 3a and b, we observe that the rate change of MR in field parallel
to the nanowire long axis is higher, i.e. the resistance change is much stronger in parallel field than
in perpendicular field.
22
Figure 2-3 Magneto-resistance (MR) measurement with the magnetic field applied (a) perpendicular and
(b) parallel to the nanowire long axis at different temperatures. (c) Linear fitting of
)] 0 , ( / ) , ( ln[ T H T
versus
2 / 1
H
under parallel magnetic field.
23
The negative magnetoresistance phenomenon measured in InN nanowires in the vicinity of
semiconductor-to-metal transition can be analyzed using the VRH model, which is consistent with
the temperature dependent conduction at zero field described earlier. The magnetic field produces
two effects: i) classical and ii) quantum. The classical effect is squeezing the electron wave
functions
33
. This mechanism leads to the positive magnetoresitance, i.e. the larger the magnetic
field the larger the sample resistance. However, in addition to the classical mechanism, there is a
second mechanism leading to the negative MR observed. This mechanism has a quantum
mechanical origin and is related to the fact that different electron wave functions can interfere with
each other during the electron hopping process. The magnetic field destroys this interference thus
leading to the negative magnetoresistance. The negative magnetoresistance is well known in the
theory of weak localization
34
. However, this theory assumes that all electronic states are
delocalized, thus it is not directly applicable to our samples. On the contrary, as follows from our
resistance measurement at zero magnetic field, all electronic states in the nanowire channels are
localized. This is why we have used VRH mechanism (Eq. (4)) to analyze the data. And in the
VRH regime, the interference between different electron wave functions still exists. A magnetic
field 𝐻 strongly affects the interference if the magnetic flux through a typical closed trajectory 𝑆
is comparable to the quantum flux
e
c
0
. The typical closed trajectory is defined as
3 / 2
~
kn S , where n is the impurity concentration, and
4
1
3
1
~
T
T
n k
M
is the average number of
hops within the phase coherence time
4
1
exp ~
T
T
M
over which the electron makes a transition
24
into a quantum state that is incoherent with respect to the initial state. Thus the negative
magnetoresistance is applicable for magnetic field satisfying the condition: 1
3
1
4
1
n
T
T
c
eH
M
.
Based on the VRH model we discussed earlier, we assume that the inequality Eq. (1) is still
valid in the vicinity of the semiconductor-to-metal transition, although in this regime, this
inequality may be more difficult to satisfy since the localization length can be rather large
B
c
a
n n
/ 1
0
(5)
where
B
a ~
0
is the localization length deep in the insulation regime and 1 0 is the
critical index. n and nc denote the impurity concentration and the critical impurity concentration.
It is known that nc depends on the magnetic field in the following way
35
2 / 1
3 / 2
0
) 0 (
) 0 ( ) (
C
c
c c
n
c
eH
A
n
n H n
(6)
where A is a numerical constant. In the high field regime, 0 A , ) 0 ( ) (
c c
n H n leads to
0 ) 0 ( ) ( H
.
In the Mott VRH regime, the resistance is given by Eq. (4). The temperature
scale TM depends on the localization length as
3
~
d
M
T with d=3 as discussed earlier.
Using the fact that in the vicinity of semiconductor-to-metal transition, the parameter
1
) 0 (
) ( ) 0 (
c
c c
n
H n n
, and expanding Eq. (4) yields an expression:
0
2 / 1
3 / 2
) (
ln
) 0 , (
) , (
ln
T
n
c
eH
B
T
H T
(7)
25
where A B
4
3
is a numerical coefficient
36
. The temperature dependence of
)] 0 , ( / ) , ( ln[ T H T is given by the factor ] / ) ( ln[
0
T , and can be measured independently as
has been shown in the zero field resistivity measurement. And the magnetic field dependent of the
)] 0 , ( / ) , ( ln[ T H T at constant temperature has a linear expression with respect to
) 2 /( 1
H . Fig
3c shows the plot of this magnetoresistivity ratio, with the best fit obtained for 1 .
Finally we explore the MR dependence on the applied field direction, much stronger effect
when the field is in the parallel direction. The origin of this direction asymmetry lies in the fact
that the nanowire conduction channel is strongly anisotropic. So the shift in the threshold, Eq. (6)
depends on the angles between the direction of magnetic field H and the axes of the ellipsoid of
the diffusion coefficient 𝐷 𝑖𝑗
.
37
The magnetic field in this case transforms into
2
1
~
D
D
H H ,
where ) sin cos (
2
//
2
1
D D D D
,
3
2
// 2
D D D , with being the angle between H and
the wire long axe
37
. Then in this anisotropic scenario, a concentration threshold
C
n and its shift are
defined by Eq. (6) with magnetic field H replaced by H
~
, that leads to the displacement of the
threshold depending on the angle . Therefore, in parallel field when 0 ,
D D
1
and as a
result for magnetic field, we obtain
3 / 1
//
~
D
D
H H . In a perpendicular magnetic field when
2 / and
// 1
D D D
, we get
6 / 1
//
~
D
D
H H . Thus one obtains the following result for the
magnetoresistance ratio for the perpendicular and parallel magnetic fields, showing the anisotropic
effect of the applied field direction.
26
4 / 1
//
)] 0 , ( / ) 0 , , ( ln[
)] 0 , ( / ) 2 / , , ( ln[
D
D
T H T
T H T
(8)
2.3 Conclusion
In summary, InN nanowires have been successfully synthesized using catalytic CVD method,
showing single crystalline structure in wurtzite phase as confirmed by XRD and HRTEM. To
investigate the fundamental transport mechanisms of through nanowire, four probe transport
measurements on individual InN nanowire are performed with respect to temperature and magnetic
field. A clear semiconductor-to-metal transition at T ~ 80 K is revealed. Below the transition, the
conduction is dominated by 3D Mott variable range hopping. In addition, a negative magneto-
resistance is observed due to the destruction of the interference of the electron wavefunctions as
magnetic field increases. Asymmetry of the MR effect in parallel and perpendicular fields is
investigated, showing that the MR change is much stronger in wire axis-parallel field due to the
anisotropy in the conduction channel.
References of Chapter 2
1. Park, H.; Park, J.; Lim, A. K. L.; Anderson, E. H.; Alivisatos, A. P.; McEuen, P. L. Nature
2000, 407, (6800), 57-60.
2. Lu, J. G.; Chang, P. C.; Fan, Z. Y. Mater. Sci. Eng. R. 2006, 52, (1-3), 49-91.
3. Luth, H.; Blomers, C.; Richter, T.; Wensorra, J.; Estevez Hernandez, S.; Petersen, G.;
Lepsa, M.; Schapers, T.; Marso, M.; Indlekofer, M.; Calarco, R.; Demarina, N.;
Grutzmacher, D. physica status solidi (c) 7, (2), 386-389.
4. Cui, Y.; Zhong, Z. H.; Wang, D. L.; Wang, W. U.; Lieber, C. M. Nano Letters 2003, 3, (2),
149-152.
5. Thompson, R. S.; Li, D. D.; Witte, C. M.; Lu, J. G. Nano Letters 2009, 9, (12), 3991-3995.
27
6. Tsai, L. T.; Chiu, S. P.; Lu, J. G.; Lin, J. J. Nanotechnology 2010, 21, (14), -.
7. Qian, F.; Li, Y.; Gradecak, S.; Wang, D. L.; Barrelet, C. J.; Lieber, C. M. Nano Letters
2004, 4, (10), 1975-1979.
8. Li, Y.; Xiang, J.; Qian, F.; Gradecak, S.; Wu, Y.; Yan, H.; Yan, H.; Blom, D. A.; Lieber,
C. M. Nano Letters 2006, 6, (7), 1468-1473.
9. Cimpoiasu, E.; Stern, E.; Cheng, G. S.; Munden, R.; Sanders, A.; Reed, M. A. Brazilian
Journal of Physics 2006, 36, (3B), 824-827.
10. Richter, T.; Blomers, C.; Luth, H.; Calarco, R.; Indlekofer, M.; Marso, M.; Schapers, T.
Nano Letters 2008, 8, (9), 2834-2838.
11. Chang, C. Y.; Chi, G. C.; Wang, W. M.; Chen, L. C.; Chen, K. H.; Ren, F.; Pearton, S. J.
Applied Physics Letters 2005, 87, (9), 093112-3.
12. Chang, C. Y.; Chi, G. C.; Wang, W. M.; Chen, L. C.; Chen, K. H.; Ren, F.; Pearton, S. J.
Journal of Electronic Materials 2006, 35, (4), 738-743.
13. King, P. D. C., Veal, T. D. and McConville, C. F. Journal of Physics: Condensed Matter
2009, 21, (17).
14. Hu, M. S.; Hsu, G. M.; Chen, K. H.; Yu, C. J.; Hsu, H. C.; Chen, L. C.; Hwang, J. S.; Hong,
L. S.; Chen, Y. F. Applied Physics Letters 2007, 90, (12), 123109.
15. Michaelson, H. B. Journal of Applied Physics 1977, 48, (11), 4729-4733.
16. Gu, D. F.; Dey, S. K.; Majhi, P. Applied Physics Letters 2006, 89, (8), 082907.
17. Li, S. X.; Yu, K. M.; Wu, J.; Jones, R. E.; Walukiewicz, W.; Ager, J. W.; Shan, W.; Haller,
E. E.; Lu, H.; Schaff, W. J. Physical Review B 2005, 71, (16), 161201(R).
18. Ager, J. W.; Miller, N.; Jones, R. E.; Yu, K. M.; Wu, J.; Schaff, W. J.; Walukiewicz, W.
Physica Status Solidi B-Basic Solid State Physics 2008, 245, (5), 873-877.
19. Chang, P. C.; Lu, J. G. Appl. Phys. Lett. 2008, 92, (21), 212113.
20. Richter, T.; Luth, H.; Schapers, T.; Meijers, R.; Jeganathan, K.; Hernandez, S. E.; Calarco,
R.; Marso, M. Nanotechnology 2009, 20, (40), 405206.
28
21. Werner, F.; Limbach, F.; Carsten, M.; Denker, C.; Malindretos, J.; Rizzi, A. Nano Letters
2009, 9, (4), 1567-1571.
22. Segev, D.; Van de Walle, C. G. Europhysics Letters 2006, 76, (2), 305-311.
23. Lu, H.; Schaff, W. J.; Eastman, L. F.; Stutz, C. E. Applied Physics Letters 2003, 82, (11),
1736-1738.
24. Veal, T. D.; Mahboob, I.; Piper, L. F. J.; McConville, C. F.; Lu, H.; Schaff, W. J. Journal
of Vacuum Science & Technology B 2004, 22, (4), 2175-2178.
25. Rickert, K. A.; Ellis, A. B.; Himpsel, F. J.; Lu, H.; Schaff, W.; Redwing, J. M.; Dwikusuma,
F.; Kuech, T. F. Applied Physics Letters 2003, 82, (19), 3254-3256.
26. Mahboob, I.; Veal, T. D.; McConville, C. F.; Lu, H.; Schaff, W. J. Physical Review Letters
2004, 92, (3), 036804.
27. Lee, S.; Lee, W.; Seo, K.; Kim, J.; Han, S. H.; Kim, B. Nanotechnology 2008, 19, (41),
415202.
28. Lu, H.; Schaff, W. J.; Eastman, L. F.; Wu, J.; Walukiewicz, W.; Look, D. C.; Molnar, R.
J. Mater. Res. Soc. Symp. Proc. 2003, 743, L4.10.1.
29. Chaudhry, A.; Islam, M. S. Journal of Nanoscience and Nanotechnology 2008, 8, (1), 222-
227.
30. Adachi, S., Handbook of Physical Properties of Semiconductors. Kluwer: Dordecht, 2004;
Vol. 3.
31. Lin, Y. F.; Jian, W. B.; Wang, C. P.; Suen, Y. W.; Wu, Z. Y.; Chen, F. R.; Kai, J. J.; Lin,
J. J. Applied Physics Letters 2007, 90, (22), 223117.
32. Note that the measurements in perpendicular and parallel fields are done on two different
samples. They are synthesized in the same run, but show slightly different resistances.
33. ShklovskiI, B. I.; Efros, A. L., Electronic properties of doped semiconductors. Springer:
New York, 1984.
34. Khmel'nitskii, D. E. Physica B+C 1984, 126, (1-3), 235-241.
35. Khmel'nitskii, D. E.; Larkin, A. I. Solid State Communications 1981, 39, (10), 1069-1070.
29
36. Altshuler, B. L.; Aronov, A. G.; Khmelniskii, D. E. JETP letters 1982, 36, (5), 195.
37. Altshuler, B. L.; Aronov, A. G.; Larkin, A. I.; Khmelnitskii, D. E. Sov. Phys. JETP 1981,
54, 411.
38. For thinner nanowires with diameter D < r_hop the nanowire is effectively 1D system
therefore the electron transport is defined by the weakest link leading to activation
resistivity behavior.
30
Chapter 3: Structural and Optical Verification of Residual Strain
Effect in Single Crystalline CdTe Nanowires
3.1 Introduction
Cadmium telluride (CdTe) has attracted significant research interest because of its superior
optical and electrical properties for developing optoelectronic devices. In particular, it holds
promising potential for low-cost photovoltaic due to its high absorption coefficient, absorbing 99%
of the incident light in a layer thickness of only ~1 µ m, as compared to ~10 µ m for Si. In addition,
it has a direct band gap of 1.49 eV at room temperature, which is close to the theoretically-
calculated optimal value for solar cells under AM 1.5 irradiation. Furthermore, the band gap of
CdTe could be readily engineered by alloying with mercury or zinc, thus it can function as an
effective photo-detector with a wide spectral range. Moreover, CdTe nanowire with high aspect
ratio is exceptional for building high performance photoelectric devices since it possesses the
unique properties of nanostructured system.
Various techniques have been developed to synthesize CdTe nanowires including template-
directed electrochemical deposition
1
, catalytic solution growth
2
, self-assembling nanoparticles into
nanowires
3
and hydrothermal method
4,5
. However, most of these methods are solution based, and
the resulted nanowires are typically polycrystalline. Little has been done on synthesis of single
crystalline CdTe nanowires
6,7
. Even with the chemical vapour deposition (CVD) method, which
is known as an efficient way to synthesize single crystalline nanowires
8,9
, the growth mechanisms
for CdTe nanowires are not yet clearly understood. Specifically, the effect of the ambient gas on
the reaction mechanism has not been fully studied.
31
In this article, we first describe Au-catalyzed CVD growth of single crystalline CdTe
nanowires, followed by careful structural and optical characterizations of the nanowires.
Experimental evidences are discussed to elucidate the nanowire growth mechanism via hydro-
assisted vapor-solid-solid (VSS) process. Most significantly, we demonstrate the existence of non-
negligible inhomogeneous residual strain in the as-synthesized CdTe nanowires, and the impact of
such strain on the electronic structure of the nanowires manifested in their photoluminescence
response. This study indicates that residual strain effect has to be incorporated into the
consideration for sensible nanowire device design and performance control. The details of
nanowire growth and characterization results are discussed below.
3.2 Experiments
The catalytic CVD growth of CdTe utilized a quartz tube furnace. And an alumina boat with
source CdTe powder (99.999%, from Alfa Aesar) was placed at the centre of the furnace. A Si
wafer was etched in buffered oxide etch (BOE) solution to remove the native oxide layer followed
by 10 nm Au film coating as the catalyst. The Si substrate was placed at the furnace 10 cm away
from the source powder. The system was pumped for 30 minutes to ~40 Pascal. The furnace was
set to be 465 º C, resulting in the source temperature of 465 º C and the substrate temperature of 350
º C. The CdTe nanowires were grown for 1 hour under continuous 10 sccm H2 gas flow.
The morphology and composition of as-grown CdTe nanowires were characterized by field
emission scanning electron microscopy (FESEM, Hitachi S4800), transmission electron
microscopy (TEM, JEOL 2000), and energy dispersive X-ray fluorescence (EDS, equipped in
JEOL 2000). The structural properties were studied by powder X-ray diffraction (Rigaku Ultima
32
IV diffractometer, CuKa, λ=1.5418 Å, acceleration voltage V=40 kV, emission current I=20 mA)
in θ/2θ mode with a scan speed of 04°/min and Bragg angle 2θ ranging from 20° to 80°. All the
peak parameters (peak position, its full width at half maximum (FWHM) and the corresponding
lattice constant) are indentified by MDI Jade 5.0 software. And the optical properties were studied
by photoluminescence (PL) at a temperature range of 5.0 to 79.7 K under cw green laser (532 nm,
~ 1.1 W/cm2) excitation. The laser spot radius is ~ 200 µ m and the PL spectra are taken from
hundreds of nanowires.
3.3 Results and discussions
Figure 3-1 shows a scanning electron microscope (SEM) image of the as-grown CdTe
nanowires with diameter ranging from 50 to 200 nm and length from 5 to 20 μm. Polygonal
nanoparticles are found on the tips of the nanowires, suggesting a catalytic growth mechanism.
Figure 3-2 (a) is a low magnification TEM image of a typical CdTe nanowire with smooth surface
and uniform thickness along the nanowire axis. A polygonal nanoparticle was found on the tip of
the nanowire, consistent with the SEM observation. Figure 3-2 (b) shows the high resolution
transmission electron microscope (HRTEM) images of a single CdTe nanowire in the zone axis
[
1 1 0
]. The interplanar distance is determined to be 0.37 nm, corresponding to the <111> direction
in zinc blende CdTe. The insert displays a selected area electron diffraction (SAED) pattern of
CdTe nanowires, verifying the single crystalline nature. The pattern is indexed to zinc blende
structure CdTe with growth direction along <111>. Figure 3-2 (c) (d) show the local EDS spectra
of the nanowire and the Au catalytic nanoparticle, respectively, in which the C and Cu peaks
originated from the carbon coated copper TEM grid. The EDS spectrum of the nanowire shows
distinct Cadmium (Cd) and Tellurium (Te) peaks, and the atomic ratio of Cd:Te is calculated to be
33
~1:1, suggesting good stoichiometry. The EDS spectrum of the nanoparticle indicates the
composition primarily Au (~92 at. %) with a small amount of Cd (~8 at. %), and the Te content is
below the detection limit. According to the Au-Cd phase diagram, at the low growth temperature
(350 º C), the Au-Cd alloy is in solid form. Thus, the growth of CdTe nanowires is governed by a
vapor-solid-solid (VSS) growth mechanism. From the high resolution SEM image (insert of
Figure 3-1), Au catalyst shows a polygonal shape after growth, which is a further evidence of VSS
growth.
Figure 3-1 The SEM image of as-grown CdTe nanowires; polygonal nanoparticles are found on the tips of
the nanowires. Insert: high resolution SEM image of the Au catalyst shows a polygonal shape, verifying
the VSS growth mechanism.
34
Figure 3-2 (a) Low magnification TEM image of a CdTe nanowire showing a nanoparticle on the tip of the
nanowire. (b) HRTEM image shows the growth direction along <111>. Insert: SAED pattern of the
HRTEM image. (c) EDS spectrum of CdTe nanowire. (d) EDS spectrum of the Au nanoparticle on the tip
CdTe nanowires.
We should point out that these nanowires grow via a hydrogen-assisted VSS process, where
hydrogen ambient plays an important role. The formation is detailed as the following: first, the
CdTe source reacts with H2 gas, forming Cd and H2Te vapors; subsequently, H2Te decomposes
into H2 and Te2 vapor. The process is known as hydrogen-assisted thermal evaporation
10, 11
. The
reactions are expressed as:
) ( ) ( ) ( ) (
2 2
g Te H g Cd g H s CdTe
) ( 2 / 1 ) ( ) (
2 2 2
g Te g H g Te H
35
The vapors are transported downstream in the furnace to the region where the Au-coated Si
substrate is placed. Cd, Te and H2Te vapors then diffuse on the surface or inside the Au catalyst
clusters, forming CdTe following the reaction:
As excess source atoms present, CdTe crystallizes at the catalyst/semiconductor interface
since it is energetically favourable, resulting in nanowires growth.The hydrogen assistant growth
mechanism is verified by the fact that very few CdTe nanowires are grown in pure Ar gas ambience,
in contrast to the high density CdTe nanowires obtained in H2 ambience. This suggests that the
dissociation of CdTe source is suppressed in Ar ambience. This is as expected, since Cd and Te
form strong covalence bond with a bond energy of ~1.1 eV, the thermal energy (at 465 º C) alone
is not sufficient to break the bond.
Figure 3-3 XRD spectrum of CdTe nanowires. Miller indices of the zinc blende CdTe peak are indicated.
The peaks labelled by triangle originate from wurtzites CdTe, and the peak labeled with diamond shape
originates from Au catalytic particle.
) ( ) ( 2 2 / 1 2
2 2 2
g H s CdTe Te Te H Cd
36
Residual strain commonly exists in nanowire growth on substrate with thermal expansion
coefficient and lattice mismatch
12,13
. Lattice mismatch induced strain field is usually found to
concentrate around the base of the nanowires
12
. In addition, surface relaxation and
reconstruction
14,15
, impurities (such as Au impurity) and defects (such as point defects, stacking
faults, and dislocations) generate lattice distortion, which can all contribute to residual strain in the
nanowires. We focus in this letter the strain effects in the CdTe nanowires through a systematic
investigation of the structural and optical properties of the nanowires. Figure 3-3 shows the XRD
θ/2θ scan pattern of the CdTe nanowire. The diffraction peaks (2θ) at 24.0°, 39.5°, 46.7°, 57.0°,
62.7° , 71.5° and 76.4° are respectively indexed to (111), (220), (311), (400), (331), (422) and (511)
of zinc blende structure of CdTe. The peaks of the unstrained zinc blende structured CdTe are
located at 23.800° , 39.356° , 46.514° , 56.877° , 62.523° , 71.358° , and 76.431° , respectively (JCPDS
card No. 65-8879). All the peaks are broaden and shifted, suggesting the presence of residual
strains in the nanowires. A small fraction of wurtzite CdTe phase (<2%) (JCPDS card No. 19-
0193) can also be identified, originating from the stacking fault in some nanowires. Strain along
the nanowires axis (i.e. <111> direction) is determined from the XRD (111) peak. The (111) peak
has a full-width-half-maximum (FWHM) of 0.418° , which is much broader compared to the peak
width that would be induced by size confinement. They originate in fact from the inhomogeneous
stress distribution in the CdTe nanowires. The lattice constant calculated from the (111) peak is
a=0.6417± 0.0056 nm and the lattice constant for unstrained CdTe is a0=0.6482 nm
17
, so that the
compressive strain along the nanowire axis is determined to be
0 0 //
/ ) ( a a a
2
10 ) 86 . 0 1 (
.
(Note that the subscript // and are used to denote strain components parallel or perpendicular to
the nanowire axis). Similarly, the strain component in the plane perpendicular to the nanowire axis
37
is calculated to be
2
10 ) 0 . 1 4 . 0 (
from the (220) diffraction peak. Since strains alter the band
structure of the nanowires
14
, they affect the optical properties of the nanowires.
To verify the strain effect, low temperature PL spectrum is measured at 5.0 K (Figure 3-4
(a)). It is dominated by a peak around 1.552 eV, with a small shoulder at 1.598 eV (peak 1 as noted
in Figure 3-4 (a)), and two broad bands centred at 1.494 eV (peak 4) and 1.335 eV (peak 5). The
PL spectrum is Gaussian fitted, as shown in Fig. 4a, with fitted peaks delineated in the dashed
lines. Note that the Gaussian fit yields two emission peaks at 1.552 eV and 1.553 eV (labelled as
peak 2 and 3). The origins of the peaks are designated according to the PL peak position of zinc
blende CdTe taken from literature
18-22
. The corresponding band diagram are shown in Fig. 4b,
illustrating different observed transitions. Note that the PL spectrum is expected to be dominantly
from zinc blende CdTe as XRD (Figure 3-3) indicates wurzite content is less than 2%. The peak
at 1.598eV (peak 1) originates from the emission of donor bound excitons
18
. And the weak bound
exciton emission is attributed to the enhanced surface recombination in the nanowires. The peak
centred around 1.55 eV is actually double peak (peak 2 and 3) consisting of a donor acceptor pair
(DAP) transition channel and a free electron to acceptor (e,A) transition channel related to the
same acceptor energy level
19-21
. The corresponding acceptor state is located at ~50 meV above the
valence band edge, which is likely to originate from the defect state due to Cd vacancies (V-Cd)
or defect complexes involving V-Cd
18
. Note that the double peak have been reported to be in the
range of 1.550 eV to 1.559 eV, varying with sample preparation.
19-21
The 1.494 eV peak can be
identified to be the well-known Y luminescence band, originated from the excitons bound to the
structural dislocations. And the peak at 1.335 eV can be attributed to DAP transition involving Au
acceptors (with energy level E Au about 263 meV above the valence band) and unidentified shallow
donor which locates at a few meV below the conduction band.
22
38
The designation of the peak origins are further verified by the temperature dependent PL
measurement, as shown in Figure 3-5. All the PL spectra are Gaussian fitted and the integrated
PL intensity of each peak as a function of temperature is shown in the insert. The integrated PL
intensity of peak 1 decreases as temperature rises from 5.0 K to 20.0 K, and cannot be reliably
distinguished in curve fitting for measurements at or above 34.6 K. This is consistent with the
temperature dependence of bound exciton emission due to the delocalization and dissociation of
bound excitons at increased temperature. In addition, the integrated PL intensity of double peaks
(2 and 3) has distinct temperature-dependent behaviours. The 1.552 eV peak (peak 3) shows a
pronounced decrease as temperature increases from 5.0 K to 20.0 K, in contrast, the integrated PL
intensity of 1.553 eV (peak 2) peak shows only a slight decrease. Different temperature
dependence of the integrated PL intensity are key characteristics to distinguish DAP transitions
and (e,A) type transitions. With increase in temperature, shallow donor ionizes, resulting in a
suppression in DAP transition. Consequently, the integrated PL intensity of DAP transition peak
decreases faster than that of the (e,A) type transition. As temperature further increases to 34.6 K,
more donors are ionized, leading to the diminishing of DAP peak (peak 3) and an increase in the
(e,A) peak (peak 2). Therefore, the 1.552 eV peak (peak 3) originates from the DAP transition,
and the 1.553 eV peak (peak 2) is attributed to (e,A) type transition.
39
Figure 3-4 (a) Photoluminescence spectrum of CdTe nanowires measured at 5.0 K. A pronounce blue-shift
in the bound exciton peak (peak #1) was revealed. (b) The band structure for CdTe shows transitions
processes: donor bound excitons (DX), Y band, free electron to acceptor transitions (e, A), and donor-
acceptor-pair transition DAP originated from shallow acceptors and deep level Au acceptors.
40
The temperature dependence of Y luminescence band has been studied relating to the
recombination kinetics. Y luminescence is caused by the radiative decay of dislocation bound
excitons, thus the recombination processes involves the capture of free excitons to the dislocation
sites and the thermal release of excitons from dislocations.
23
Thus, the temperature dependence of
integrated Y luminescence intensity can be derived by carefully balancing these two processes,
and the activation energy for the thermal quenching process can be determined. Following Ref.
[23], the temperature dependent of Y band (peak 4) is fitted with expression
Where IY is the integrated intensity of the Y band peak, I0 and are temperature independent
parameters. KB is Boltzmann constant and EB is the exciton bounding energy to the structural
dislocations. The term T
2/3
comes from the effective density of exciton states, which is usually
neglected in a simple theoretical model. EB is fitted to be 9.8 meV, consistent with previous
reported value in thin film CdTe samples
24
.
Figure 3-5 Temperature dependent photoluminescence spectra of CdTe nanowires ranging from 5.0 K to
79.7 K. Insert: Temperature dependence of integrated peak intensity for all five Gaussian fitted peaks (see
Fig. 4 for peak assignment).
) / exp( 1
3 / 2
0
T k E T
I
I
B B
Y
41
The thermal quenching of the integrated PL intensity of peak 5 cannot be fitted with only one
activation channel. Rather, it fits well with the experiment data by assuming the presence of two
thermal activation channaels for dissociating the electron-hole pairs. The integrated intensity can
be formulated by the expression:
24, 25
where I0 is the zero temperature integrated intensity of the DAP peak. and are
temperature independent fitting parameters. E1 and E2 correspond to the activation energies for
those two different thermal processes resulting in the non-radiative recombination. The fitting
results yield E1=4.76 meV and E2=20.6 meV. To elucidate the actual physical channels for these
two thermal processes, further theoretical investigation is needed.
Taking a closer look at the PL spectra, it is evident that the fitted PL peaks 1, 2, 3,
corresponding to bound exciton, DAP, and (e,A) transitions, are broadened and shifted compared
to those measured in high quality bulk CdTe crystals
18
. Evidently, the observed blue shift and
broadening of PL peaks cannot be a consequence of confinement effect in the nanowires (estimated
to be less than 0.5 meV for these CdTe nanowires of diameters 50-200 nm, which are far greater
than the CdTe exciton diameter of 15 nm). To rule out the heating effect, power dependent PL
measurement are perform (as shown in Figure 3-6 in Electronic Supplementary Materials (ESM)).
There is no observable red shift in all peaks (as shown in Figure 3-7), indicating lack of heating.
Otherwise with heating and rise of nanowire temperature, a red shift of PL peak is expected with
increase excitation power. Alternatively, the XRD measured strain in the nanowire ensembles
provides a hint for the original of the blue shift and broadening in PL peaks which we attribute to
residue strain in the nanowires. It is worth noting that broadening of PL peak (corresponding to
) / exp( ) / exp( 1
2 2 1 1
0
T k E T k E
I
I
B B
1
2
42
strain distribution) does not necessarily occur on a single nanowire, but could potential arise from
different amounts of strain existing in different nanowires of various sizes in the ensemble.
In order to estimate how much energy shift the strain field causes, we apply the deformation
potential theory
26
. Following the aforementioned XRD results, the magnitude of strain and its
distribution along the nanowire axis and in-plane perpendicular to the axis are
2
//
10 ) 9 . 0 0 . 1 (
and
2
10 ) 0 . 1 4 . 0 (
. As expected, the mean axial stress (-1.0× 10
-2
) and the mean in-plane stress
(0.4× 10
-2
) measured are consistent with the Poisson effect and a Poisson ratio of ~0.41 for CdTe.
The strain induced fractional volume change
2 /
//
V V
can be extracted using a lower-bound
estimate by adding the maximum axial strain to the minimum transverse strain, and vice versa.
This determines that the distribution of the inhomogeneous strain induced fractional volume
change in the samples is greater than the range of -0.66% to -2.94%. As the strain induced energy
shift roughly follows where ac and av are the conduction band and valence band
deformation potentials
26
, the distribution of corresponding to the inhomogeneous strain
distribution in the samples is greater than the range of +3 to +13 meV. In other words, the exciton,
shallow donor and acceptor related transition peaks are broadened by at least 10 meV as a result
of the inhomogeneous distribution of the residual strain. This qualitatively explains the satisfactory
Gaussian fitting of the bound exciton peak of the nanowire samples (peak 1 centred at ~1.598 eV
at 5.0 K) and a peak with FWHM ~16 meV in contrast to the reported bound exciton peak for high
quality bulk crystalline CdTe sample (1.592 eV with ~1 meV FWHM)
18
.
V V a a E
V c g
/ ) (
g
E
43
Figure 3-6 Power dependent PL spectra with excitation laser power from 0.056 W/cm
2
to 38.9 W/cm
2
.
Figure 3-7 Peak position as a function of excitation power density for all five peaks. All the PL spectra are
Gaussian fitted and the corresponding peak positions are shown. When PL is measured using excitation
power density varying from 0.014 W/cm
2
to 9.73 W/cm
2
, no detectable red shift in PL peak position has
been observed, indicating lack of heating. And the blue shift is due to band filling effect as power density
increases.
44
3.4 Conclusion
In summary, zinc blend single crystalline CdTe nanowires have been successfully synthesized
by CVD technique. The nanowire growth is governed by a Au catalytic VSS mechanism. We find
that hydrogen gas plays an important role in the growth of CdTe nanowires, serving not only as a
carrier gas, but also as a reaction intermediate assisting thermal evaporation of CdTe sources. The
flow rate of hydrogen gas is an essential parameter to adjust in order to obtain single crystalline
CdTe nanowires. We show by XRD that our CVD grown CdTe nanowires have residual strains
which alter the electronic band structure. The shift and broadening in the exciton and impurity
energy levels are clearly demonstrated in the temperature dependent photoluminescence results.
These studies elucidate that the strain effects in nanowires need to be carefully considered in the
interpretation of the optical and electrical behaviors of nanomaterials and in the design of future
nanoscale optoelectronic devices.
References of Chapter 3
1. Xu, D.; Chen, D.; Xu, Y.; Shi, X.; Guo, G.; Gui, L.; Tang, Y. Preparation of II-VI group
semiconductor nanowire arrays by dc electrochemical deposition in porous aluminum
oxide templates. Pure Appl. Chem. 2000, 72, 127-135.
2. Wang, F.; Dong, A.; Sun, J.; Tang, R.; Yu, H.; Buhro, W. E. Solution-liquid-solid growth
of semiconductor nanowires. Inorg. Chem. 2006, 45, 7511-7521.
3. Tang, Z.; Kotov, N. A.; Giersig, M. Spontaneous organization of single CdTe nanoparticles
into luminescent nanowires. Science 2002, 297, 237-240.
4. Yang, Q.; Tang, K.; Wang, C.; Qian, Y.; Zhang, S. PVA-Assisted synthesis and
characterization of CdSe and CdTe nanowires. J. Phys. Chem. B 2002, 106, 9227-9230.
5. Yong, S. M.; Muralidharan, P.; Jo, S. H.; Kim, D. K. One-step hydrothermal synthesis of
CdTe nanowires with amorphous carbon sheaths. Mater. Lett. 2010, 64, 1551-1554.
45
6. Hou, J.; Yang, X.; Lv, X.; Peng, D.; Huang, M.; Wang, Q. One-step synthesis of CdTe
branched nanowires and nanorod arrays. Appl. Surf. Sci. 2011, 257, 7684-7688.
7. Ye, Y.; Dai, L.; Sun, T.; You, L. P.; Zhu, R.; Gao, J. Y.; Peng, R. M.; Yu, D. P.; Qin, G.
G. High-quality CdTe nanowires: synthesis, characterization, and application in
photoresponse devices. J. Appl. Phys. 2010, 108, 044301.
8. Hochbaum, A. I.; Fan, R.; He, R.; Yang, P. Controlled growth of Si nanowire arrays for
device integration. Nano Lett. 2005, 5, 457-460.
9. Wang, D.; Dai, H. Low-temperature synthesis of single-crystal germanium nanowires by
chemical vapor deposition. Angew. Chem. 2002, 114, 4977-4980.
10. Utama, M. I. B.; Peng, Z.; Chen, R.; Peng, B.; Xu, X.; Dong, Y.; Wong, L. M.; Wang, S.;
Sun, H.; Xiong, Q. Vertically aligned cadmium chalcogenide nanowire arrays on
muscovite mica: a demonstration of epitaxial growth strategy. Nano Lett. 2011, 11, 3051-
3057.
11. Lovergine, N.; Prete, P.; Cola, A.; Mazzer, M.; Cannoletta, D.; Mancini, A. M. Hydrogen
transport vapor phase epitaxy of CdTe on hybrid substrates for X-ray detector applications.
J. Electron. Mater. 1999, 28, 695-699.
12. Taraci, J. L.; Hÿ tch, M. J.; Clement, T.; Peralta, P .; McCartney, M. R.; Drucker, J. Picraux,
S. T. Strain mapping in nanowires. Nanotechnology 2005, 16, 2365-2371.
13. Seo, H. W.; Bae, S. Y.; Park, J.; Yang, H.; Park, K. S.; Kim, S. Strained gallium nitride
nanowires. J. Chem. Phys. 2002, 116, 9492-9499.
14. Li, S.; Yang, G. W. Universal scaling of semiconductor nanowires bandgap. Appl. Phys.
Lett., 2009, 95, 073106.
15. Sarkar, S.; Pal, S.; Sarkar, P. Electronic structure and band gap engineering of CdTe
semiconductor nanowires. J. Mater. Chem. 2012, 22, 10716-10724.
16. Shi, W. S.; Zheng, Y. F.; Wang, N.; Lee, C. S.; Lee, S. T. Oxide-assisted growth and optical
characterization of gallium-arsenide nanowires. Appl. Phys. Lett. 2001, 78, 3304-3306.
17. Ebina, A.; Takahashi, T. Studies of clean and adatom treated surfaces of II-VI compounds.
J. Cryst. Growth 1982, 59, 51-64.
18. Shin, H.Y.; Sun, C.Y. The exciton and edge emissions in CdTe crystals. Mater. Sci. Eng.
B 1998, 52, 78-83.
46
19. Aguilar-Herná ndez, J.; Cá rdenas-Garcí a, M.; Contreras-Puente, G.; Vidal-Larramendi, J.
Analysis of the 1.55 eV PL band of CdTe polycrystalline films. Mater. Sci. Eng. B 2003,
102, 203-206.
20. Kraft, C.; Metzner, H.; Hä drich, M.; Reislö hner, U.; Schley, P.; Gobsch, G.;. Goldhahn, R.
Comprehensive photoluminescence study of chlorine activated polycrystalline cadmium
telluride layers. J. Appl. Phys. 2010, 108, 124503.
21. Van Gheluwe, J.; Versluys, J.; Poelman, D.; Clauws, P. Photoluminescence study of
polycrystalline CdS/CdTe thin film solar cells. Thin Solid Films 2005, 480-481, 264-268.
22. Molva, E.; Francou, J. M.; Pautrat, J. L.; Saminadayar, K.; Dang, L. S. Electrical and
optical properties of Au in cadmium telluride. J. Appl. Phys. 1984, 56, 2241-2249.
23. Hildebrandt, S.; Uniewski, H.; Schreiber, J.; Leipner, H. S. Localization of Y
Luminescence at Glide Dislocations in Cadmium Telluride. J. Phys. III France 1997, 7,
1505-1514.
24. Halliday, D. P.; Potter, M. D. G.; Mullins, J. T.; Brinkman, A. W. Photoluminescence study
of a bulk vapour grown CdTe crystal. J. Cryst. Growth 2000, 220, 30-38.
25. Bimberg, D.; Sondergeld, M. Thermal dissociation of excitons bounds to neutral acceptors
in high-purity GaAs. Phys. Rev. B 1971, 4, 3451-3455.
26. Van de Walle, C. G. Band lineups and deformation potentials in the model-solid theory.
Phys. Rev. B 1989, 39, 1871-1883.
47
Chapter 4: Nature of AX Centers in Antimony Doped Cadmium
Telluride Nanobelts
4.1 Introduction
The accelerated demand in device miniaturization urgently requires functional nanoscale
elements as the building blocks. The devices based on semiconductor nanostructures have already
shown excellent promises, for they have been utilized as chemical sensors
1, 2
, field effect transistors
(FET)
2, 3
, photodetectors
4-6
, solar cells
7
, etc. Among the II-VI semiconductor materials, CdTe has
attracted significant research interest for decades because of its exceptional properties and broad
device applications. CdTe has a high absorption coefficient and a direct bandgap of 1.49 eV at
room temperature, thus it has been widely used in heterojunction photovoltaics.
8
CdTe
nanostructure has demonstrated as a promising candidate to build core-shell nanostructured solar
cells which are expected to have higher efficiencies than the thin film counterparts.
7
Also, the
bandgap of CdTe can be tuned by alloying with Zn, Hg or Mg, rendering versatile device
applications. For example, Hg1-xCdxTe makes outstanding infrared photodetectors and Cd1-z ZnxTe
makes excellent solid-state X-ray and gamma-ray detectors. In addition, benefitting from the
nanostructure morphology, CdTe nanostructures have been employed for developing scalable
nanoscale electronics, with tunable optical absorption resonance and polarization sensitive
detection.
9-12
In order to enhance the electronic property and device performance, doping has been
commonly utilized, yet the preserving difficulty lies in achieving efficient doping with low
resistivity. Unintentionally doped CdTe nanostructures exhibit slightly p-type conductivity,
48
attributed to intrinsic defects such as Cd vacancies (VCd) and defect complexes involving VCd.
4, 13
Among the work on CdTe nanostructures doped by Zn
14-16
, Sb
5, 17, 18
, or Mn
19
, Sb has been proven
to be an effective p-type dopant. In the recent years, there have also been many reports on CdTe:Sb
nanostructure based functional devices,
5, 17, 18
such as p-CdTe/n-CdS photovoltaics, p-CdTe:Sb/n-
Si photovoltaics and photodetectors,
5
Al/AlOx/CdTe:Sb memory devices
17
and Al/CdTe:Sb nano-
Schottky barrier diodes
18
, however, the nominal doping level is only on the order of 10
16
- 10
17
cm
-
3
even with high Sb amounts.
18
This problem is generally attributed to self-compensation of
acceptor impurities, but microscopic origin of the factors limiting the Sb doping level is not clear
and demands experimental proof. Particularly, it has been theoretically predicted that in Sb doped
CdTe, the efficient doping rate would be limited by the formation of AX centers, i.e. part of the
substitutional acceptors undergo a transition from shallow acceptor states to deep AX centers with
large lattice relaxation in the vicinity of the impurtiy.
20
(Note that the similar notation (A,X) refers
to the emission of acceptor bound excitons in a PL spectrum, and should be careful to distinguish
them.) The existence of AX centers leads to a significant alteration of electronic properties, which
could deteriorate device performance.
21, 22
To elucidate these effects, we have carried out a
comprehensive investigation on Sb doped CdTe nanobelts, with correlating the optical and
electrical properties, via parallel studies in the temperature dependent optical and charge transport
measurements. This work has clarified the impurity energy levels in Sb doped CdTe nanobelt, and
correlated them with microscopic origins, thus demonstrated for the first time the nature of deep
AX centers as theoretically predicted.
20
49
4.2 Experiments
CdTe:Sb nanobelts were synthesized by catalytic chemical vapor deposition (CVD) method
in a quartz tube furnace. An alumina boat with a source mixture of CdTe and Sb powder with
weight ratio of 10:1 was placed at the center of the furnace. A Si chip was used as growth substrate.
It was etched in buffered oxide etch (BOE) solution to remove the native oxide layer, followed by
depositing a layer of 10 nm Au film which serves as catalyst. The substrate was placed at the
furnace 10 cm away (downstream) from the source powder. The system was pumped for 30
minutes to ~ 0.4 mbar. The temperature for the source material was set to 470 º C, resulting in a
temperature of 350 º C at the growth substrate. The CdTe:Sb nanobelt synthesis lasted 1 hour under
continuous 10 sccm H2 gas flow.
The morphology and composition of as-grown CdTe:Sb nanobelts were characterized by
scanning electron microscopy (SEM, JEOL JSM-7001F) and energy dispersive X-ray
spectroscopy (EDS, equipped in the SEM). The structural properties were studied by transmission
electron microscopy (TEM, JEOL JEM-2100F) and powder X-ray diffraction (Rigaku Ultima IV
diffractometer, CuKa, λ = 1.5418 Å, acceleration voltage V = 40 kV, emission current I = 20 mA)
in θ/2θ mode with a scan speed of 4 ° /min. The fabrication of CdTe:Sb nanobeltFETs for electrical
measurements is a multistep procedure: CdTe nanobelts were first removed from the substrate by
sonication in isopropyl alcohol (IPA). The nanobelt suspension in IPA was then drop-casted to
another Si/SiO2 substrate (a highly doped p-type Si chip capped with a 500 nm oxide layer).
Photolithography and E-beam evaporation techniques were then used to define contacts (Cu/Au
20nm/170nm) to the single CdTe:Sb nanobelts. The metal contacts serve as source and drain
electrodes, while the p-type Si substrate as the back gate. Temperature-dependent transport
measurements with a semiconductor analyzer (HP 4156C) were carried out by loading the sample
50
into a cryostat cooled by liquid nitrogen. Furthermore, the temperature-dependent ensemble
photoluminescence (PL) and microphotoluminescence (μ-PL) measurements of single nanobelts
were performed using a HeNe 633 nm cw laser for excitation, while the sample was mounted in a
Janis ST-500 helium flow cryostat with a temperature range from 4 to 70 K. The luminescence
light was collected by spherical lenses, dispersed by a 500 mm monochromator and detected using
a liquid nitrogen-cooled CCD. Moreover, to probe the existence of AX centers, photoconductance
measurements were conducted in the same system as conductivity measurements, using UV light
source of 400 nm wavelength and 0.4 mW/cm
2
power density.
4.3 Results and discussion
A scanning electron microscopy (SEM) image of as-grown CdTe:Sb nanobelts is shown in
Figure 4-1 (a). The CdTe:Sb nanobelts range from 100 nm to 1 µ m in width and 10 - 30 µ m in
length. A catalytic particle is observed at the tip of each nanobelt investigated, suggesting a catalyst
driven growth mechanism.
23
The corresponding energy dispersive X-ray (EDS) spectrum in the
inset of Figure 4-1 (a) reveals a nearly stoichiometric atomic ratio of Cd and Te, while the overall
atomic incorporation ratio of Sb is approximately 1.8 at%. The concentration of Sb in different
nanobelts was estimated to be 1.0 - 2.5 at% based on point-EDS spectra of six nanobelts. The EDS
element mapping of one single nanobelt, as shown in Figure 4-1 (b), indicates a homogenous
distribution and co-localization of Cd, Te and Sb along the whole nanobelt. Figure 4-1 (c) shows
the XRD θ/2θ scan pattern of the Sb doped CdTe nanobelt ensemble in contrast to an undoped
CdTe nanobelt ensemble. The diffraction peaks are indexed to zinc blende structure of CdTe
(JCPDS card No. 65-8879). According to the XRD measurement, the Sb doped CdTe nanobelts
remain pure zinc blende structure without any additional phases. Figure 4-1 (d) shows the high
51
resolution transmission electron microscopy (HRTEM) image of another single CdTe:Sb nanobelt,
verifying the single crystalline nature of the doped zinc blende CdTe nanobelts. The inter-planar
spacing of 0.32 nm is indexed to the (200) plane, demonstrating that the nanobelt grows along the
[200] crystal axis.
Figure 4-1 (a) SEM image of as-grown CdTe:Sb nanobelts. The corresponding EDS spectrum in the inset
reveals a nearly stoichiometric atomic ratio of Cd and Te, while the overall atomic incorporation ratio of
Sb is found to be approximately 1.8 %. (b) SEM image and EDS mapping of Cd, Te and Sb for one single
CdTe:Sb nanobelt indicating a homogenous distribution of Cd, Te and Sb along the whole channel. (c)
XRD spectra of undoped (black) and Sb doped (red) CdTe nanobelt ensembles. Spectra are offset vertically
for clarity. The diffraction peaks are indexed to zinc blende structure of CdTe (JCPDS card No. 65-8879).
(d) HRTEM image of one CdTe:Sb nanobelt and the corresponding selective area diffraction (SAED)
pattern is shown in the inset demonstrating nanobelt growth along the [200] crystal axis.
52
The electrical transport characteristics of single CdTe:Sb nanobelts were investigated in a
FET configuration. The contact resistance is found to be negligible (see Figure 4-2). Figure 4-3
(a) shows the linear -
ds ds
IV curves, where
ds
I is the drain-source current and
ds
V is the drain-
source voltage for various gate voltages ( 20 10 0 10 20 - =-
g
V , , , , V ), demonstrating
unambiguously gating with p-type characteristics of the CdTe:Sb nanobelt. The drain-source
current
ds
I decreases with increasing gate voltage
g
V , as directly shown in the inset of Figure 4-3
(a). From the linear region of the -
ds ds
IV curve, the p-channel transconductance (
m ds g
g dI / dV )
of the CdTe:Sb nanobelt FET can be estimated to be 4.74 × 10
-9
S at room temperature. For
nanobelt geometry, the channel capacitance per unit area is given by
00 r
C / h ,
4
where
0
is the
vacuum permittivity,
r
is the relative permittivity of gate dielectric and h is the thickness of gate
dielectric (h = 500 nm). And from the transconductance
0 m h ds
g W / L C V ,
4
the field effect hole
mobility (
h
) is extracted to be approximately 0.7 cm
2
V
-1
s
-1
for a channel width (W ) and length
( L ) of 273 nm and 1.4 µ m, respectively. Furthermore, using the conductivity of the nanobelt
FET at
g
V = 0 V and the hole mobility
h
calculated above, hole concentration
h
p / q is
estimated to be 5.6 × 10
16
cm
-3
. Even for high Sb incorporation of 1.0 –2.5 at%, the hole
concentration still remains at this level, far below the nominal doping concentration as expected
from EDS measurement. Together with the low mobility in CdTe:Sb nanobelt of 0.7 cm
2
V
-1
s
-1
compensation of acceptor impurities is exhibited.
53
Figure 4-2 Drain-source current I ds versus drain-source voltage V ds curves of a single CdTe:Sb nanobelt
field effect transistor (FET) for two-probe and four-probe measurements. The linear I ds - V ds curves
measured in two-probe (black) and four-probe (red) configuration overlap demonstrating good Ohmic
contacts between the nanobelt and the Cu/Au electrodes. The contact resistance is smaller than 1% of the
nanowire resistance and thus negligible.
54
Figure 4-3 (a) The drain-source current I ds as function of drain-source voltage V ds for various gate voltages
V g ranging from -20 to 20 V for a CdTe:Sb nanobelt FET. Inset shows the I ds-V g curve measured at constant
V ds = 5 V at room temperature for the same device. Evaluating the slope of the linear region gives an
estimate for the hole mobility of 0.7 cm
2
V
-1
s
-1
and a hole concentration of 5.6× 10
16
cm
-3
. (b) Arrhenius plot
of the conductivity σ versus inverse temperature for one single CdTe:Sb nanobelt FET shows
unambiguously two linear regions, corresponding to two acceptors level that are determined to be 56 ± 2
meV and 115 ± 4 meV above the valence band.
55
It has been a long-lasting problem to produce low resistivity p-type CdTe. In fact,
compensation effects are common in II-VI compound semiconductors.
24-26
Compensation in II-VI
semiconductors could arise from amphoteric dopant incorporation, either by dopants substituting
both group II and VI element, or by dopants going into either substitutional or interstitial sites,
which results in mutual electrical compensation.
25, 27
Since Sb is a group V element, the electronic
configuration is very unlikely to occupy a Cd site. Plus, Sb has a similar ionic radius to Cd and Te,
therefore, it is highly unfavorable to get into interstitial sites. Thus amphoteric doping can be
neglected in Sb doped CdTe. Another mechanism proposed is that the doping also leads to the
formation of native defects such as vacancies, antisites or dopant-vacancy pairs that have
compensation effect. Native vacancies and antisites are found not to dominate, but dopant-vacancy
pairs called A-centers are likely to contribute to the compensation.
24
The third compensation
mechanism is predicted to be the formation of deep level defect called AX centers, as historically
it was referred to an unknown defect complex. It involves a large lattice relaxation that lowers the
total energy of the defect.
24-26
A microscopic picture of the AX center proposed is a double-
breaking-bond structure, formed by breaking two anion-cation bonds and forming one anion-anion
bond around the dopant.
20, 28
The large lattice relaxation and rebonding consume free holes and
compensate the substitutional (acceptor) states, which is considered to be significant compensation
factor when the substitutional state is metastable with respect to a configuration with a large lattice
relaxation.
20, 28
It has been computed that forming AX centers is energetically favorable for Sb
doped CdTe, resulting in a strong compensation effect.
20
Yet this deep level AX center has not
been reported experimentally.
To gain further insights into the nature of the compensation in our samples, temperature-
dependent conductivity measurement were performed with temperature range from 85 K to 360 K,
56
as shown in Figure 4-3 (b). The Arrhenius plot of conductivity versus temperature exhibits two
linear regions, verifying that the electrical transport behaviors are governed by thermal excitations
involving two acceptor impurity levels. Each linear region shows an Arrhenius-type behavior with
a conductivity
a
exp( E / kT) , where Ea, k and T are the activation energy, Boltzmann constant
and absolute temperature, respectively. In strongly compensated semiconductors, activation
energies simply represent ionization energies of acceptor levels.
29
Therefore, the result identifies
two acceptor levels at 56 ± 2 meV and 115 ± 4 meV above the valence band. The shallow level at
56 meV is induced by a substitutional Sb atom on a Te site (Sb Te), which is expected to be located
at 51 - 61 meV above the valence band.
30-34
The deeper acceptor level at approximately 115 meV
is also induced by Sb incorporation.
30
As we will soon see that both impurity levels correspond to
the optical emission measurement.
Low-temperature photoluminescence (PL) is the best tool to study energy levels of radiative
defects. Figure 4-6 (a) shows the PL spectrum taken at 3.7 K of a CdTe:Sb nanobelt ensemble in
which the most significant recombination processes are labeled. Note, it is valid to investigate the
PL spectra of the nanobelt ensemble to conduct conclusions for single nanobelts (see Figure 4-4).
The origin of the optical recombination mechanisms were identified according to the literature
31,
32, 35-37
and subsequently the impurity levels are depicted in the schematic diagram Figure 4-6 (b).
The luminescence of free exciton transitions (FX) is energetically found at ~ 1.596 eV in
agreement with the bandgap energy of 1.609 eV reduced by the exciton binding energy (EB
FX
) of
13 meV.
31
Two phonon replica of the FX arise at ~ 1.576 eV and ~ 1.555 eV lowered by the
phonon energy (ELO) of 21 meV.
31
A donor bound exciton (D,X) peak is found at 1.594 eV and is
probably caused by Te antisites (TeCd)
32, 35
, while the 1.590 eV line is related to the recombination
of an acceptor bound exciton (A,X).
31
Furthermore, at lower energies several recombination
57
signatures occur: The (e,A) transition is induced by a free electron recombining with an acceptor
bound hole, while the donor-acceptor-pair (DAP) transition is the emission feature of an electron
bound to a donor recombining with a hole bound to an acceptor. The higher energetic (e,A)1
transition and the corresponding DAP1 transition are located at 1.548 eV and 1.539 eV. Since both
rely on the same acceptor level, the energy difference between these two series corresponds to the
binding energy of an electron to a donor ED = 9 meV. This involved acceptor level exhibits an
ionization energy of E A1 = 60 meV, which coincides very well with the shallower acceptor level
determined by the electrical measurements. The 1.500 eV and 1.492 eV emission lines for CdTe:Sb
materials have not been reported before in the literature. They are attributed to a second pair of
(e,A)2 and DAP2 transitions due to the deeper acceptor level at E A2 = 109 meV above the valance
band edge, which was also observed in our electrical measurement with a comparable activation
energy of 115 meV. (We will address in details this unkown defect level later in the article.) Both
observed acceptor state energies determined by PL spectroscopy are in good agreement with the
activation energies obtained by temperature dependent electrical measurements. Moreover, the
most intense luminescence around 1.450 eV, which is accompanied by its phonon replicas, clearly
shows a blue-shift with increasing excitation intensity (see Figure 4-5). This strongly indicates a
DAP character of this recombination, thus it is labelled as DAP3. The corresponding acceptor
levels EA3 = 130 meV are related to the A center.
38
Temperature dependent PL measurements show
a red shift of the near-band edge emission with increasing temperature due to a shrinkage of the
bandgap (Figure 4-7).
58
Figure 4-4 The PL spectrum of the Sb doped CdTe nanobelt ensemble measured at T = 23 K shows the
same features as a single CdTe:Sb nanobelt measured at T = 21 K with slightly different intensities.
Therefore, it is valid to investigate the PL spectra of the nanobelt ensemble at T = 4 K to match luminescence
lines to certain radiative transitions and conduct conclusions for optical recombinations in single nanobelts.
Figure 4-5 Power-dependent nanobelt ensemble PL spectra of the emission around 1.45 eV and their
phonon replica at 4 K (spectra are offset vertically for clarity). A shift of the peak positions to higher
energies with increasing excitation power indicate a DAP type transition which are attributed to A centers
(E A3 ~ 130 meV) caused by defect complexes of Cd vacancies and donors.
59
Figure 4-6 (a) PL spectrum of the CdTe:Sb nanobelt ensemble, taken at 3.7 K with a excitation density of
90 mW/m
2
at 633 nm. Several near band edge transitions are labeled: the free exciton FX, donor bound
excitons (D,X), acceptor bound excitons (A,X) free electron to acceptor (e,A) transition, donor acceptor
pair recombination (DAP) and their phonon replica. (b) Schematic band diagram depicts the impurity
energy levels within the bandgap and recombination processes. One donor level and 3 acceptor levels are
involved, and the binding energies are determined to be E D = 9 meV, E A1 = 61 meV, E A2 = 109 meV and
E A3 = 130 meV.
60
Figure 4-7 Temperature-dependent PL spectra of CdTe:Sb nanobelt ensemble at excitation of 633 nm and
65 mW/cm² . Spectra are offset vertically for clarity. Temperature dependent measurements show a shift of
the near band gap transitions to lower energies with increasing temperature which is attributed to the
Varshni shift.
By correlating the electrical characterization and optical emission measurements, we can
establish a band diagram of the impurity levels, as shown in Figure 4-6 (b). In brief, the donor
level (ED = 9 meV) is attributed to Te antisites. The deepest acceptor level (EA3 ~ 130 meV) is
related to the A centers, and two acceptor levels at EA1 ~60 meV and EA2 ~110 meV are induced
by the Sb doping, while EA1 is by substitutional Sb atom on a Te site, and EA2 is not yet clarified.
A natural question is thus whether this EA2 is the AX center predicted. The fact that dopants induce
two distinct acceptor levels has been observed in different materials such as phosphorous doped
ZnSe and CdS
39, 40
: It is explained that the levels are related to the same dopant, but with different
atomic configurations: the shallow level is related to substitutional dopants and the deeper one
corresponds to a state with large lattice relaxation (AX centers) in the vicinity of the dopant. The
most important experimental signature for the presence of AX centers is the observation of
61
persistent photoconductivity (PPC) and its temperature dependence. Figure 4-8 (a) depicts a
typical photoconductivity transient response of a CdTe:Sb nanobelt measured at low temperature
(T = 100 K). When irradiated under light for 50 seconds, the current increases gradually and only
by a factor of two. After the light is turned off, the photocurrent exhibits persistent
photoconductivity (PPC). PPC effects have been observed in nanostructures attributed to surface
band bending which induces charge separation.
41-43
However, surface band bending related PPC
is known to be sensitive to ambient gases,
41-43
while in the CdTe:Sb nanobelts, photo response is
not sensitive to ambience atmosphere (see supplementary information Fig. S6). Therefore, the PPC
effect observed in the CdTe:Sb nanobelts gives the first evidence of AX centers. To further unravel
its property, a temperature cycle measurement was performed, as shown in Figure 4-8 (b). The
sample is cooled down in dark to a temperature of 85 K (black squares), followed by 10 minutes
illumination under 400 nm light, and then warmed up in dark to 380 K (red dots). A hysteresis of
the dark current is observed, showing a thermal activation nature of charge trapping, which is in
fact a hallmark of the AX centers.
44, 45
To explain the microscopic mechanisms involving the AX centers, we use the configuration
coordinate diagram,
46
as drawn in Figure 4-8 (c), which depicts hole capture and de-capture
kinetics. The atomic state a , AX represent respectively the shallow Sb dopant and AX centers, and
the charge state is indicated by a superscript, i.e.,
0
a and a
represent the neutral and (-1) charge
state of shallow acceptors. And AX
0
and AX
+
represent the neutral and (+1) charged AX centers.
The ground state of AX
+
was proposed to be double-breaking-bond geometry with triangular C3V
symmetry, as schematically shown in the inset of Figure 4-8 (c).
20, 28
In the configuration
coordinate diagram, two parabolas represent the total energy of the system at different charge states
as a function of the atomic configuration around the dopants. The left parabola depicts the total
62
energy of the neutral charge state, and the ground state for neutralized acceptor state is
0
a . The
right parabola shows the total energy of the +1 charge state with the ground state of AX
+
. Note that
the impurity levels illustrated in the band diagram (Figure 4-6 (b)) are induced by the ionization
of defects. In the process
0
a a h
, the ionization energy is EA1 ~ 60meV. Effectively, it can
be regarded as an acceptor level located ~60 meV above the valence band. Similarly, EA2 and E A3
correspond to the ionization energies of AX centers (
0
A A h
) and A centers
(
0
AX AX h
), respectively.
In equilibrium, the existence of AX
+
in CdTe:Sb is predicted to be energetically favorable
since the process 2 a h AX
is exothermic.
20
The mechanism is that a negatively charged
acceptor a
captures two holes h
and is transformed into a positively charged AX
state, which
is the origin of the strong self-compensation in the system. Under light excitation with energy
larger than the optical ionization energy
exc opt
EE , as shown in Figure 4-8 (c), AX
+
will undergo
the following transitions:
00
free free
AX AX h a h
.
Optical absorption does not involve lattice relaxation, so it always occurs at constant Q, i.e.
only vertical transition is allowed (refer to process 1 labeled in Figure 4-8 (c)). Therefore, by
releasing a free hole, AX
+
will first transit into AX
0
state, which is a neutrally charged state with
the same configuration as AX
+
. However, the ground state of the neutral acceptor is the a
0
state
with tetrahedral Td symmetry, which has negligible lattice relaxation involved, as schematically
shown in the inset of Figure 4-8 (c). Therefore, AX
0
will transit to a
0
state via phonon emission,
as indicated by process 2 in Figure 4-8 (c). In both processes, the hole concentration increases,
hence the conductivity increases. When light is turned off, re-capture of the free holes is prohibited
by an energy barrier Ecap, giving rise to a persistent photocurrent observed.
63
Figure 4-8 (a) Persistent photoconductivity (PPC) measured at T = 100 K at constant voltage V = 0.1 V
with an illumination wavelength of 400 nm. (b) Dark current as a function of temperature measured by first
cooling down in dark (black squares), after illumination for 10 minutes at 85 K, then warming up in dark
(red circles). The measurement verifies the existence of AX centers. (c) Configuration coordinate diagram
sketches acceptors at two different states as a function of the atomic configuration. AX
+
represents the
positively charged AX centers with large lattice relaxation, while a
0
is the neutral acceptor state with
negligible lattice distortion. The microscopy structures of AX
+
and a
0
are drawn in the inset. (d) PPC decay
curves measured at various temperatures (100 K, 300 K, 390 K and 400 K). The data were fitted by
PPC
I (t) exp( (t/ ) )
(as shown in solid curves). The inset shows the obtained decay time constant τ
versus inverse temperature 1000/T. At low temperatures T < 300 K, τ reveals only weak temperature
dependence. At high temperatures T > 300 K, τ shows an Arrhenius-like behavior with an activation energy
of E cap = 0.46 eV.
64
The decay of the photocurrent occurs when holes overcome the energy barrier and recapture
by the AX center. At low temperature, this can be achieved dominantly by hole tunneling through
the capture barrier Ecap in the configuration space.
44
At higher temperature, it is achieved by
thermal excitation of holes over Ecap.
44
This explains the hysteresis curve displayed in Figure 4-8
(b). Illumination at low temperature excites holes from deep acceptor AX
+
into a
0
states, which
persist for temperatures T < 300 K. At higher temperatures, thermal energy is sufficient to
overcome the capture barrier, resulting in hole capture into AX
+
states and subsequently quenching
of the conductivity.
To better illustrate the two decay processes and thus determine the hole capture energy barrier
Ecap, temperature-dependent PPC measurements similar to Figure 4-8 (a) were carried out at
temperatures ranging from 100 K to 410 K. The normalized PPC decay curves are shown in Figure
4-8 (d). The normalized photocurrent
0
ppc d d
I t I t I / I I with the dark current
d
I
and the current level when the light is turned off 0 I is fitted by a stretched exponential function
to obtain a decay time τ:
44, 47, 48
PPC
I (t) exp( (t/ ) )
.
Here τ is the characteristic decay time constant and β (0 < β < 1) is the decay exponent. The
fitting curves are shown as solid curves in Figure 4-8 (d) as well. The fitting values of β vary only
very slightly, and the best fit yields β = 0.38. The temperature dependence of the decay time
constant τ is plotted in the inset of Figure 4-8 (d), revealing unambiguously two distinct
temperature regions. In the low temperature region for T < 300 K, exhibits a weak temperature
dependence, suggesting the tunneling mechanism. While in the higher temperature region above
65
300 K, thermal excitation dominates with an Arrhenius type behavior
cap
exp( E / kT) .
44
Extracted from the slope, the hole capture barrier height is determined to be Ecap= 0.46 ± 0.03 eV.
The existence and the nature of the AX centers are thus validated by our temporal and
temperature dependent photoconductance measurement, and we can conclude that the unknown
EA2 is in essence the AX centers which contribute to significant compensation as observed in the
samples. According to Chadi’s simulation
20
, the reaction 2 a h AX
is exothermic with an
energy change of E = 0.3 eV for Sb doped CdTe. The formation of AX centers compensates the
free holes and limits its concentration to be
hV
n n exp( E/ kT) , where
V
n the effective density of
valence states is. Taking
19
1 8 10
V
n. cm
-3
for CdTe,
29
the theoretical maximum hole
concentration is 4 × 10
16
cm
-3
, well consistent with our experimental observation.
4.4 Conclusion
In conclusion, single crystalline CdTe:Sb nanobelts have been synthesized via catalytic CVD
method. They are measured in a FET configuration, revealing p-type characteristics with low hole
concentrations and mobilities, manifesting significant compensation. Temperature-dependent
conductivity and PL measurements reveal that Sb doping induces two distinct acceptor levels in
CdTe:Sb nanobelts. The shallow level consists of a substitutional Sb atom on a Te site and the
deeper one corresponds to AX center which is unraveled by photoconductance measurements
showing PPC effect and dark current hysteresis. The existence of AX center is proven for the first
time experimentally in the CdTe nanobelts system. This work elucidates the Sb doping effects on
the electronic structure of the CdTe nanobelts, which is still lacking in the field. Only with a solid
66
fundamental understanding of the electronic band structure and doping effect, it is possible to
design and develop sensible devices.
References of Chapter 4
1. Fan, Z. Y.; Lu, J. G. Appl Phys Lett 2005, 86, (12).
2. Fan, Z. Y.; Wang, D. W.; Chang, P. C.; Tseng, W. Y.; Lu, J. G. Appl Phys Lett 2004, 85,
(24), 5923-5925.
3. Chang, P. C.; Fan, Z.; Chien, C. J.; Stichtenoth, D.; Ronning, C.; Lu, J. G. Appl Phys Lett
2006, 89, (13).
4. Xie, X.; Kwok, S. Y.; Lu, Z. Z.; Liu, Y. K.; Cao, Y. L.; Luo, L. B.; Zapien, J. A.; Bello, I.;
Lee, C. S.; Lee, S. T.; Zhang, W. Nanoscale 2012, 4, (9), 2914-2919.
5. Xie, C.; Luo, L. B.; Zeng, L. H.; Zhu, L.; Chen, J. J.; Nie, B.; Hu, J. G.; Li, Q.; Wu, C. Y.;
Wang, L.; Jie, J. S. Crystengcomm 2012, 14, (21), 7222-7228.
6. Ye, Y.; Dai, L.; Sun, T.; You, L. P.; Zhu, R.; Gao, J. Y.; Peng, R. M.; Yu, D. P.; Qin, G.
G. J Appl Phys 2010, 108, (4).
7. Fan, Z. Y.; Razavi, H.; Do, J. W.; Moriwaki, A.; Ergen, O.; Chueh, Y. L.; Leu, P. W.; Ho,
J. C.; Takahashi, T.; Reichertz, L. A.; Neale, S.; Yu, K.; Wu, M.; Ager, J. W.; Javey, A.
Nat Mater 2009, 8, (8), 648-653.
8. Romeo, N.; Bosio, A.; Tedeschi, R.; Romeo, A.; Canevari, V. Sol Energ Mat Sol C 1999,
58, (2), 209-218.
9. Panoiu, N. C.; Osgood, J. R. M. Opt. Lett. 2007, 32, (19), 2825-2827.
10. Kim, S. K.; Day, R. W.; Cahoon, J. F.; Kempa, T. J.; Song, K. D.; Park, H. G.; Lieber, C.
M. Nano Letters 2012, 12, (9), 4971-4976.
11. Wang, J.; Gudiksen, M. S.; Duan, X.; Cui, Y.; Lieber, C. M. Science 2001, 293, (5534),
1455-1457.
12. Rö der, R.; Ploss, D.; Arian, K.; Buschlinger, R.; Geburt, S.; Peschel, U.; Ronning, C.
Journal of Physics D 2014.
67
13. Luo, L.-B.; Huang, X.-L.; Wang, M.-Z.; Xie, C.; Wu, C.-Y.; Hu, J.-G.; Wang, L.; Huang,
J.-A. Small 2014, 10, (13), 2645-2652.
14. Zhou, S. M.; Zhang, X. H.; Meng, X. M.; Wu, S. K.; Lee, S. T. Appl Phys a-Mater 2005,
81, (8), 1647-1650.
15. Zhou, S. M. Phys Low-Dimens Str 2006, 2, 29-33.
16. Gandhi, T.; Raja, K. S.; Misra, M. Electrochim Acta 2006, 51, (26), 5932-5942.
17. Chao, X.; Biao, N.; Long, Z.; Long-Hui, Z.; Yong-Qiang, Y.; Xian-He, W.; Qun-Ling, F.;
Lin-Bao, L.; Yu-Cheng, W. Nanotechnology 2013, 24, (35), 355203.
18. Zhu, L.; Jie, J.; Wu, D.; Luo, L.; Wu, C.; Zhu, Z.; Yu, Y.; Wang, L. Journal of
Nanoengineering and Nanomanufacturing 2012, 2, (2), 191-196.
19. Ramasamy, P.; Mamum, S. I.; Jang, J.; Kim, J. Crystengcomm 2013, 15, (11), 2061-2066.
20. Chadi, D. J. Phys Rev B 1999, 59, (23), 15181-15183.
21. Li, J. Z.; Lin, J. Y.; Jiang, H. X.; Khan, M. A. Appl Phys Lett 1998, 72, (22), 2868-2870.
22. Klein, P. B.; Freitas, J. A.; Binari, S. C.; Wickenden, A. E. Appl Phys Lett 1999, 75, (25),
4016-4018.
23. Huang, M. H.; Wu, Y.; Feick, H.; Tran, N.; Weber, E.; Yang, P. Advanced Materials 2001,
13, (2), 113-116.
24. Desnica, U. V. Prog Cryst Growth Ch 1998, 36, (4), 291-357.
25. Chadi, D. J. Annu Rev Mater Sci 1994, 24, 45-62.
26. Faschinger, W.; Gundel, S.; Nurnberger, J.; Albert, D. Commad 2000 Proceedings 2000,
41-48.
27. Zhang, M.; Wille, M.; Rö der, R.; Heedt, S.; Huang, L.; Zhu, Z.; Geburt, S.; Grü tzmacher,
D.; Schä pers, T.; Ronning, C.; Lu, J. G. Nano Letters 2014, 14, (2), 518-523.
28. Chadi, D. J.; Chang, K. J. Appl Phys Lett 1989, 55, (6), 575-577.
29. Chin, K. K. Sol Energ Mat Sol C 2010, 94, (10), 1627-1629.
68
30. Fochuk, P.; Grill, R.; Nykonyuk, Y.; Krustok, J.; Armani, N.; Zakharuk, Z.; Grossberg, M.;
Panchuk, O. Ieee T Nucl Sci 2007, 54, (4), 763-768.
31. Soltani, M.; Certier, M.; Evrard, R.; Kartheuser, E. J Appl Phys 1995, 78, (9), 5626-5632.
32. Dhese, K. A.; Devine, P.; Ashenford, D. E.; Nicholls, J. E.; Scott, C. G.; Sands, D.; Lunn,
B. J Appl Phys 1994, 76, (9), 5423-5428.
33. Dhese, K. A.; Ashenford, D. E.; Nicholls, J. E.; Devine, P.; Lunn, B.; Scott, C. G.;
Jaroszyński, J. Journal of Crystal Growth 1994, 138, (1–4), 443-447.
34. Kanie, H.; Ogino, K.; Kuwabara, H.; Tatsuoka, H. physica status solidi (b) 2002, 229, (1),
145-148.
35. Berding, M. A. Phys Rev B 1999, 60, (12), 8943-8950.
36. Fiederle, M.; Eiche, C.; Salk, M.; Schwarz, R.; Benz, K. W.; Stadler, W.; Hofmann, D. M.;
Meyer, B. K. J Appl Phys 1998, 84, (12), 6689-6692.
37. Hildebrandt, S.; Uniewski, H.; Schreiber, J.; Leipner, H. S. J Phys Iii 1997, 7, (7), 1505-
1514.
38. Hofmann, D. M.; Omling, P.; Grimmeiss, H. G.; Meyer, B. K.; Benz, K. W.; Sinerius, D.
Phys Rev B 1992, 45, (11), 6247-6250.
39. Kosai, K.; Fitzpatrick, B. J.; Grimmeiss, H. G.; Bhargava, R. N.; Neumark, G. F. Appl Phys
Lett 1979, 35, (2), 194-196.
40. Hou, S. L.; Marley, J. A. Appl Phys Lett 1970, 16, (11), 467-469.
41. Cammi, D.; Ronning, C. Adv Cond Matter Phys 2014.
42. Wang, Y.; Liao, Z. L.; She, G. W.; Mu, L. X.; Chen, D. M.; Shi, W. S. Appl Phys Lett 2011,
98, (20).
43. Viana, E. R.; Gonzalez, J. C.; Ribeiro, G. M.; de Oliveira, A. G. J Phys Chem C 2013, 117,
(15), 7844-7849.
44. Li, J. Z.; Lin, J. Y.; Jiang, H. X.; Salvador, A.; Botchkarev, A.; Morkoc, H. Appl Phys Lett
1996, 69, (10), 1474-1476.
45. Nelson, R. J. Appl Phys Lett 1977, 31, (5), 351-353.
69
46. Lang, D. V.; Logan, R. A. Phys Rev Lett 1977, 39, (10), 635-639.
47. Dissanayake, A. S.; Huang, S. X.; Jiang, H. X.; Lin, J. Y. Phys Rev B 1991, 44, (24), 13343-
13348.
48. Lin, J. Y.; Dissanayake, A.; Brown, G.; Jiang, H. X. Phys Rev B 1990, 42, (9), 5855-5858.
70
Chapter 5: CVD Growth and Characterization of Sb2Te3 Nanowires
5.1 Synthesis
Sb2Te3 nanowires are synthesized by Au-catalyzed chemical vapor deposition (CVD) method,
followed the procedure discribed in reference
1
. A Si/SiO2 substrates were sonicated for 2 min in
acetone and then rinsed with IPA and distilled deionized (DD) water. The substrate was then
treated in a piranha solution (H2SO4:30% H2O2= 3:1) at 95–100 ° C for 30 min to remove organic
residue. The acid-treated substrates were washed with DD water and dried with high-purity
nitrogen. The substrates were immersed in 0.1 wt % aqueous poly-L-lysine solution for 1 min and
rinsed with water, then immersed in Au nanoparticles (60 nm in diameter) solution for 2 min and
rinse with DD water. Sb2Te3 nanowires were then synthesized in a quartz tube furnace
(Lindberg/Blue M). Sb powder (0.6 g) was placed at the center of the heating zone, while Te
powder (0.9 g) was located 13.5 cm upstream from the Sb powder. The substrate was placed in the
downstream of the furnace 10.5 cm away from the Sb source. The furnace was evacuated and
flushed repeatedly with Ar gas to minimize oxygen contamination. Then the furnace was heated
to 430 °C for 6 hours. The Ar flow rate was kept at 80 sccm and the pressure was kept at ~10 torr.
5.2 Characterizations
The morphology and composition of as-grown Sb2Te3 nanowires were characterized by
scanning electron microscopy (SEM, JEOL JSM-7001F) and energy dispersive X-ray
spectroscopy (EDS, equipped in the SEM). The structural properties were studied by transmission
electron microscopy (TEM, JEOL JEM-2100F) and powder X-ray diffraction (Rigaku Ultima IV
diffractometer, CuKa, λ = 1.5418 Å, acceleration voltage V = 40 kV, emission current I = 20 mA)
71
in θ/2θ mode with a scan speed of 4 ° /min. Figure 5-1 (a) shows a typical SEM image of the as
grown Sb2Te3 nanowires and the inset depicts the corresponding EDS spectrum, revealing the
atomic ratio of Sb and Te is ~ 2:3. Figure 5-1 (b) shows the XRD spectrum of Sb2Te3 nanowires,
verifying the Rhombohedral structure of space group m R3 (JCPDS card No. 00-015-0874).
Figure 5-1 (c) shows a HRTEM image of a single Sb2Te3 nanowire and inset is the corresponding
SAED, confirming its single crystalline nature and the growth direction is along [110].
The Sb2Te3 nanowire field effect transistors (FETs) were fabricated in the following
procedures. The as grown Sb2Te3 nanowires were suspended in isopropanol alcohol and drop-
casted onto Si/SiO2 substrate for electrical contact fabrication. Photolithography and E-beam
evaporation techniques were then used to define contacts (Ti/Au 20nm/150nm) to the single
Sb2Te3 nanowires. Ti/Au electrodes forms Ohmic contacts to Sb2Te3 nanowires and the contact
resistance is negligible. The metal contacts serve as source and drain electrodes, while the p-type
Si substrate as the back gate. Temperature dependent transport measurement was conducted, as
shown in Figure 5-2. The Sb2Te3 nanowires shows metallic-like transport property, i.e. resistance
increase monotonically with temperature, due to femi level pining to valence band originated to
native defects in Sb2Te3 nanowires.
72
Figure 5-1 (a) A SEM image of the as grown Sb 2Te 3 nanowires. Inset: the corresponding EDS spectrum.
(b) XRD spectrum of Sb 2Te 3 nanowires. (c) HRTEM image of a single Sb 2Te 3 nanowire. Inset: the
corresponding SAED.
Figure 5-2 Temperature dependent resistance measurement on single Sb 2Te 3 nanowires. Inset: a SEM
image of a single Sb 2Te 3 nanowires with Ti/Au contacts.
73
References of Chapter 5
1. Lee, J. S.; Brittman, S.; Yu, D.; Park, H. Journal of the American Chemical Society 2008,
130, (19), 6252-6258.
74
Chapter 6: Applications and Future Research Outlook
6.1 CdTe nanowires applications and future works
Due to their excellent electrical, optical and piezoelectric properties, CdTe nanowires offer a
wealth of device applications. We have demonstrate CdTe nanowire based photodetectors and
preliminary results on fabricating CdTe nanowire based photovoltaics.
CdTe nanowire photodetector
Nanowires are prominent material for constructing photodectors with high sensitivity owing
to their large surface-to-volume ratio. And due to its small scale, it has promising potential for
integrating into CMOS technology. Despite various device architectures that have been proposed
for nanowire photodectors, the basic principle for photodetection lies in the hotoconductivity of
semiconductor materials. Collecting photogenerated carriers provides an effective approach for
photodetection.
To construct CdTe nanowire photodetectors, a standard FET fabrication technique was
employed. CdTe nanowires were first dispersed in IPA solution and drop-casted onto Si substrate
with 500 thermally grown oxide layer. Then Pd/Au (10nm/100nm) double layer was pattern to
make contacts to the nanowires by E-beam lithography (EBL) and E-beam evaporation. SEM
image of a typical CdTe nanowire with two contacts was shown in Figure 6-1 (a). To evaluate the
performance of photodetector based on a single CdTe nanowire, monochromatic light was guided
onto the CdTe nanowire perpendicularly. Figure 6-1 (b) shows the photoresponce of the device
under the bias of 10 V by switching the light on and off at a time interval of 20 second, under
75
monochromatic light (450 nm, 200 μW/cm
2
). As light was turned on, the current increase from
~0.02 nA to 0.3 nA, and a fast response time (< 1 s) was revealed.
Figure 6-1 (a) shows the configuration of a typical CdTe nanowiere photodetector with a channel length
of ~1.2 μm and a thickness of ~220 nm. (b) Photoresponse of CdTe nanowires in air, under 450nm light
illumination (intensity=200μW/cm
2
).
Figure 6-2 (a) shows the I-V characteristics of the CdTe nanowire photodetector as the light
intensity increase from 50 μW/cm
2
to 340 μW/cm
2
(wavelength=450nm). The excess carrier
density at the bias of 10 V can be determined from the I-V characteristics, and shown in Figure
6-2 (b). It shows a near-linear relationship, revealing the defect density is low in the CdTe
nanowires, in consistent with the HRTEM observations. The performance of a photodetector are
commonly characterized by responsivity (Rλ), defined as the photocurrent generated per unit
power of incident light, and the gain (G), defined as the number of electrons collected by electrodes
76
due to excitation by one photon.
1
Under the light input intensity (450nm, 50 μW/cm
2
to 340
μW/cm
2
), gain is determined to be ~2 ×10
3
to 7×10
3
, and the responsivity is determined to be ~2
×10
3
A/W. The results are much better than the performances of the photodetector based on CdTe
nanoribbons reported in Ref.
2
, due to the single crystalline nature of the CdTe nanowires
fabricated from catalytic CVD method.
The wavelength dependence of the I-V characteristics are plotted in Figure 6-3 (a). A
decrease in photocurrent can be observed as the wavelength increases from violet (450nm) to to
near infared (1000nm), due to the absorption coefficiency changes as the wavelength, as shown in
the absorption spectrum (Figure 6-3 (b)).
We have demonstrated the fabrication and characterization of CdTe nanowire based
photodetectors. However, the devices can be further improved in the following aspects:
(1) Ohmic contacts. I-V characteristics (Figure 6-1. (a)) shows a non-liner and asymmetry
effect, attributing to the Schottky barrier in the metal semiconductor interfaces. To improve the
carrier collection efficiency, Ohmic contacts are needed. And Cu/Au contact, which is
conventionally used in back contact of CdTe/CdS solar cell, should be considered in replacing the
Pd/Au contacts.
(2) Surface effect. Surface states can be traps for photogenerationed carriers and thus lower
the photoresponsivity and gain of the device. Therefore, study the effect of the surface state by
introducing surface passivation layer to the nanowires will be an interesting focus. Also, for the
unpassivated nanowires, it is of great interest to study the effect of the surface absorbates by
comparing the photoresponce in air and in vacuum.
77
(3)Size effect. The size effect of the nanowires is important since it influents the light
absorption, carrier transport processes in photodetector. Therefore, to study how the sizes influence
the photodetector performances will also be an important issue to investigate in the future.
Figure 6-2. (a) I-V characteristics of CdTe nanowire under illumination of light with different wavelength
at bias=10V and wavelength λ=450nm. (b) Excess carrier density vs. light intensity. (c) Gain vs. light
intensity. (d) Responsivity vs. light intensity.
78
Figure 6-3. (a) I-V characteristics of CdTe nanowire under illumination of light with different wavelength.
(b) PL and absorption spectrum of CdTe nanoribbons. (Reprint from Nanoscale, 2012, 4, 2914).
CdTe nanowire based photovoltaics
The implementation of traditional planar structure solar electricity is always limited by its
high cost. In planar structure, light absorption and charge carriers collection are in the same
direction, therefore, in order to obtain high absorption and charge collection efficiency, materials
with high purity but expensive are required. This problem could be solved by introducing a new
architecture design: core-shell nanostructure. Core-shell nanostructure orthogonalize the direction
of light absorption and charge separation, thus can tolerate lower material quality.
2
To demonstrate
a CdTe nanowire based core-shell structure photovoltaic device, a template assisted fabrication
method was proposed. Figure 6-4shows the procedures of fabricating of CdTe/CdS core-shell
structure via anodic aluminum oxide (AAO) template. CdTe nanowire array is first fabricated by
AAO template guided growth. The template will be partially etched away and a CdS thin film is
coated to make the p-n heterojunction. Then transparent conductive oxide and metal contact would
be deposited to complete the solar cell.
79
Figure 6-4. Schematics of fabrication of CdTe/CdS core-shell structure via AAO template.
Figure 6-5. (a) SEM image of a tilted AAO template with Sn catalyst deposited in the bottom. (b) CdTe
nanowires in AAO template.
AAO nanoporous membranes (as shown in Figure 6-5 (a)) are good templates for fabricating
the single crystalline nanowire arrays. Stable growth of CdTe nanowire arrays in AAO templates
has been achieved by catalytic CVD method using Sn as catalyst, as shown in Figure 6-5 (b).
However, the filling rateare, composition, and morphology of the filled CdTe needed to be further
optimized by controlling over the growth conditions (i.e., temperature, pressure, gas flow rates).
80
(1) Control growth of the CdTe nanowire array. The filling rateare, composition, and
morphology of the filled CdTe needed to be further optimized by controlling over the growth
conditions (i.e., temperature, pressure, gas flow rates).
(2) CdS deposition. High quality CdS thin film are usually grown by close space sublimation
(CSS) technique. Therefore, I will use this technique to coat CdTe nanowires with n-CdS to form
core-shell structure solar cell. And the parameters will be controlled to obtain a good interface
between them.
(3)Device testing and optimization. A completed CdTe/CdS core-shell structure solar cell will
be tested and the result will be analyzed to guide the design of devices in the future.
6.2 Sb2Te3 nanowires future works
Magneto-resistance (MR) is another important tool to study the TI surface states. In
topological insulator nanowires, it is reported that a magnetic flux piercing the nanowire resulted
in an Aharonov-Bohm effect caused by the unique surface states
3
. To reveal the topological
properties, magneto-resistance measurement with a magnetic field up to 10 Tesla will be
performed. We will conduct temperature and magnetic field dependent transport measurement to
study the WAL and EEL effects in order to study the transport properties of the surface states.
Also, we will also perform angle-dependent magnetoconductance, which is an important evidence
to distinguish different quant magnetic field angle and temperature dependent measurements are
performed and analyzed to determine whether these effects originate from the surface (2D) or the
bulk (3D). For 2D surface state transport, the magnetotransport will only respond to the
perpendicular component of the magnetic field 𝐵𝑐𝑜𝑠𝜃 .
81
All these temperature and magnetic field dependent measurement can be performed to the
magnetic doped Sb2Te3 nanowires to reveal the surface state under perpetuation of magnetic
doping. Also, we will also study the magnetic transport properties with various doping
concentration, therefore manipulate the quantum transport properties of the topological insulators.
When inducing magnetic dopant into the topological insulator Sb2Te3 nanowires, the intricate
interplay between topological order and ferromagnetism is expected to give rise to a variety of
unconventional spintronic. For example, a weak magnetic perturbation can open up an energy gap
in the surface spectrum of a TI, due to the breaking of time invariant symmetry. Therefore,
magnetic doped Sb2Te3 nanowires will be a good material system to study in the further.
References of Chapter 6
1. Xie, X.; Kwok, S.-Y.; Lu, Z.; Liu, Y.; Cao, Y.; Luo, L.; Zapien, J. A.; Bello, I.; Lee, C.-S.;
Lee, S.-T.; Zhang, W. Nanoscale 2012, 4, (9), 2914-2919.
2. Tian, B.; Zheng, X.; Kempa, T. J.; Fang, Y.; Yu, N.; Yu, G.; Huang, J.; Lieber, C. M.
Nature 2007, 449, (7164), 885-889.
3. Peng, H.; Lai, K.; Kong, D.; Meister, S.; Chen, Y.; Qi, X.-L.; Zhang, S.-C.; Shen, Z.-X.;
Cui, Y. Nat Mater 2010, 9, (3), 225-229.
Abstract (if available)
Abstract
Nanoscience and nanotechnology studies materials at the nanoscale, on the order of 10⁻⁹ meter. There are many new physical phenomena and material properties when the dimension of materials reduces to nanoscale. Therefore, it is of great importance for both physical exploration and industrial application. Quasi-one-dimensional (Q1D) semiconductor nanostructures is specifically interesting because they have great potential for electronic and optoelectronic devices. However, to fabricate high quality Q1D materials and demonstrate precise control over materials properties has not yet been mature for many important material systems. Also, lack of in-depth physical understanding on these nanostructures also holds back the development of nanodevices. In this thesis, fabrication of different Q1D nanostructures (InN nanowires, CdTe nanowires, CdTe:Sb nanobelts and Sb₂Te₃ nanowires) are demonstrated, and their structural, optical and electrical properties are investigated systematically, which is important for future devices applications.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Light management in nanostructures: nanowire solar cells and optical epitaxial growth
PDF
Nanostructure interaction modeling and estimation for scalable nanomanufacturing
PDF
One-dimensional nanostructures for chemical sensing, transparent electronics, and energy conversion and storage devices
PDF
Semiconducting metal oxide nanostructures for scalable sensing applications
PDF
Photodetector: devices for optical data communication
PDF
Zero-dimensional and one-dimensional nanostructured materials for application in photovoltaic cells
PDF
The growth and characterization of III-V semiconductor nanowire arrays by nanoscale selective area metalorganic chemical vapor deposition
PDF
Synthesis and mechanical behavior of highly nanotwinned metals
PDF
Nanostructure electronic band modeling for solar cell and lighting applications
PDF
Development of organic-inorganic optical microcavities for studying polymer thin films
PDF
2D layered materials: fundamental properties and device applications
PDF
Design and characterization of metal and semiconducting nanostructures and nanodevices
PDF
Biosensing and biomimetic electronics
PDF
Development of hybrid optical microcavities for Plasmonic laser and improving biodetection
PDF
Low-dimensional asymmetric crystals: fundamental properties and applications
PDF
Modeling and analysis of nanostructure growth process kinetics and variations for scalable nanomanufacturing
PDF
Nanomaterials for macroelectronics and energy storage device
PDF
Synthesis, characterization, and mechanical properties of nanoporous foams
PDF
Optical and electrical characterization of one-dimensional (1D) and two-dimensional (2D) nanostructures
PDF
Processing and properties of phenylethynyl-terminated PMDA-type asymmetric polyimide and composites
Asset Metadata
Creator
Huang, Liubing
(author)
Core Title
Synthesis and properties study of Q1D semiconductor nanostructures
School
Viterbi School of Engineering
Degree
Doctor of Philosophy
Degree Program
Materials Science
Publication Date
07/31/2015
Defense Date
11/03/2014
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
CdTe nanowires,CdTe:Sb nanobelts,electrical properties,InN nanowires,OAI-PMH Harvest,optical properties,Q1D semiconductor nanostructures,Sb₂Te₃ nanowires,structure characterization
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Lu, Jia G. (
committee chair
), Hellwarth, Robert W. (
committee member
), Nutt, Steven R. (
committee member
)
Creator Email
huangliubing7@gmail.com,liubingh@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c3-622314
Unique identifier
UC11305086
Identifier
etd-HuangLiubi-3778.pdf (filename),usctheses-c3-622314 (legacy record id)
Legacy Identifier
etd-HuangLiubi-3778.pdf
Dmrecord
622314
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Huang, Liubing
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
Tags
CdTe nanowires
CdTe:Sb nanobelts
electrical properties
InN nanowires
optical properties
Q1D semiconductor nanostructures
Sb₂Te₃ nanowires
structure characterization