CHARGE TRANSPORT IN QUANTUM
DOT – LIGHT EMITTING DEVICES
A DISSERTATION
SUBMITTED TO THE FACULTY OF THE GRADUATE SCHOOL
OF THE UNIVERSITY OF MINNESOTA
BY
BRIJESH KUMAR
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
STEPHEN A. CAMPBELL, ADVISOR
P. PAUL RUDEN, CO-ADVISOR
AUGUST 2013
© Brijesh Kumar, 2013
i
Acknowledgements
First of all, I would like to thank my advisors, Prof. Campbell and Prof.
Ruden for their constant support and encouragement over the last six years.
They both provided me with ideas if I was stuck or to help secure the facilities
required to conduct my research. It was a great pleasure working with both
my advisors. I cannot thank both of them enough for being my support
through both rough and smooth phases of my Ph.D.
I would like to thank all my colleagues, beginning with Sang Ho Song. I
had so many insightful discussions with him on my research and various other
topics of life. I would also like to thank Gagan Agarwal, Maryam Jalali,
Min-Woo Jang, Rick Liptak, Mohammad Yunus and Isaiah Steinke for their
help over the years. I would like to thank all the NFC staff for their prompt
help whenever requested. Teaching was an important part of my Ph.D. ex-
perience. So, I would like to thank Prof. William Robbins for assigning me
classes of my liking. I would like to thank all my teachers of different courses
for answering my endless questions.
My friends and roommates, Srijan Aggarwal, Ricky Jain and Ankur
Khare, were my support throughout my Ph.D. I am grateful to all of them
ii
(especially, Srijan) for tolerating me through all these years and making me
feel home even if I was 8,000 miles away from home. I am especially indebted
to my teacher for teaching me what is important in life.
I would like to thank my family for their support over the years. It was
hardest for my mother to allow me to go so far away from home. I cannot
thank her enough for her sacrifices for me over the years. Waking up at 4.30
am everyday so that we could get fresh lunch for school every day for 18
continuous years, even when she was sick, is just a small example of her sac-
rifices. I would like to thank my father for his love and support. He sacrificed
his own desires for getting us better education. My sisters, Kavita and
Rashmi, have been supportive of all the decisions I have taken in my life. I
thank them for being there for our parents, when I was away from home. I
would like to thank my late grandfather for teaching me so much about living
a simple life and God.
Finally, I would like to thank God for everything I have, including my
intelligence, however little it may be…
iii
Dedication
To
Krishna,
my Teachers
& my Parents
iv
Abstract
Inorganic quantum dots have excellent optoelectronic properties. But, due
in part to a lack of a suitable medium for dispersion, they have not been
extensively used in optoelectronic devices. With the advent of organic
semiconductors, the integration of quantum dots into optoelectronic devices
has become possible. Such devices are termed as hybrid organic/inorganic
quantum dot light emitting devices. In hybrid organic/inorganic quantum dot
light emitting devices, the mechanisms of charge and/or energy transfer into
the quantum dots include Forster energy transfer and direct charge injection.
Forster energy transfer involves formation of excitons on organic
semiconductors, followed by an energy transfer onto the inorganic quantum
dots, where the exciton recombines resulting in emission of a photon. Direct
charge injection is the mechanism in which the electrons and holes are directly
injected into the quantum dots and they recombine on the quantum dots to
result in a photon. Which mechanism is operating in a device has been a
subject of contention. In this work, by using various device configurations, we
show that both these mechanisms can operate independently to maximize the
quantum dot light emission in such devices.
We also propose a model for inorganic QD-LEDs, which explores the most
important parameters that control their electrical characteristics. The
v
device is divided into a hole transport layer, several quantum dot layers, and
an electron transport layer. Conduction and recombination in the central
quantum dot region is described by a system of coupled rate equations, and
the drift-diffusion approximation is used for the hole and electron transport
layers. For NiO/Si-QDs/ZnO devices with suitable design parameter the
current and light output are primarily controlled by the quantum dot layers,
specifically, their radiative and non-radiative recombination coefficients.
Radiative recombination limits the device current only at sufficiently large
bias. This model can be extended to apply to hybrid organic/inorganic
QD-LEDs.
vi
Table of Contents
Acknowledgements ................................................................... i
Dedication .............................................................................. iii
Abstract ................................................................................. iv
Table of Contents ................................................................... vi
List of Tables ......................................................................... ix
List of Figures ......................................................................... x
Chapter 1 Introduction ..................................................... 1
1.1. Introduction to Light Emitting Devices ...................................... 1
1.2. Background of Quantum Dots ..................................................... 3
1.3. Synthesis of Quantum Dots ......................................................... 5
1.3.1 Liquid phase synthesis ................................................. 5
1.3.2 Vapor phase synthesis .................................................. 7
1.4. Luminescence from Quantum Dots ........................................... 12
1.5. Structure of Dissertation ........................................................... 16
Chapter 2 Review of QD-LEDs ....................................... 18
vii
2.1. Device Structure ........................................................................ 18
2.2. Organic Transport Layers .......................................................... 20
2.3. Inorganic Transport Layers ....................................................... 25
2.4. Latest Developments ................................................................. 28
Chapter 3 Charge Transport in Hybrid QD-LEDs ......... 40
3.1. Introduction ............................................................................... 40
3.2. Experimental Method ................................................................ 42
3.3. Results and Discussion .............................................................. 47
Chapter 4 Modeling of QD-LEDs ................................... 56
4.1. Introduction ............................................................................... 56
4.2. Model Description ..................................................................... 58
4.2.1 A. Carrier injection from transport layers into
quantum dot layers ............................................................................ 59
4.2.2 Transport among the quantum dot layers ................. 60
4.2.3 Recombination in the quantum dots ......................... 62
4.2.4 Coupled rate equations .............................................. 63
4.2.5 Transport in ETL and HTL ...................................... 64
4.2.6 Carrier injection from the contacts ............................ 65
4.3. Results and Discussion .............................................................. 66
4.3.1 Device Parameters ..................................................... 66
viii
4.3.2 Simulation Results ..................................................... 68
Chapter 5 Conclusion and Scope for Future work .......... 79
5.1. Conclusion ................................................................................. 79
5.2. Scope for Future Work .............................................................. 80
Bibliography .......................................................................... 82
ix
List of Tables
Table 4.1 The parameters of NiO and ZnO used for our model device. 67
Table 4.2 The parameters used to describe transport in QD layers. ...... 68
x
List of Figures
Figure 1.1 Density of states of bulk semiconductor, quantum well,
quantum wire and quantum dot. .................................................................... 3
Figure 1.2 Nucleation and growth phases for quantum dots in hot
injection method. Adapted from [6]. ............................................................... 7
Figure 1.3 Emission of light from bulk semiconductors upon relaxation of
electrons from the conduction band to valence bands. Due to the continuity of
bands, the wavelength of emitted light is different depending on the energy
lost by electrons. ........................................................................................... 12
Figure 1.4 Discrete density of states for quantum dots. Electrons can be
excited to lowest level in the conduction “band” and thus the only
recombination takes place between highest level in the valence “band.” The
interaction between other states is highly unlikely. ...................................... 13
Figure 1.5 The spectrum of quantum dots with different size
distributions. As the variance in size increases, the spectrum becomes broader.
...................................................................................................................... 14
Figure 1.6 Silicon quantum dots of different size emitting various colors of
the spectrum. Courtesy: Dr. Rick Liptak [26]. .............................................. 15
Figure 1.7 Dependence of bandgap on size of Si quantum dots. Adapted
from [27]. ....................................................................................................... 17
xi
Figure 2.1. Schematic of a generic QD-LED. ......................................... 29
Figure 2.2 Band diagram of a generic QD-LED. .................................... 30
Figure 2.3 QD-LED with only ETL (structure A) or only HTL (structure
B). In both cases PPV acts as the transport layer. From [28]. ..................... 31
Figure 2.4 PL (solid lines) and EL (dashed lines) from two different sized
CdSe/CdS core-shell QDs (a) 2.5 nm core/ 0.5 nm shell (b) 3.4 nm core/ 0.7
nm shell [32]. ................................................................................................. 32
Figure 2.5 Energy band diagram of device used by Coe et al. containing
both electron and hole transport layers [33]. ................................................ 33
Figure 2.6 EL spectrum of (a) device with ETL and HTL (b) device with
additional hole blocking layer [33]. ................................................................ 34
Figure 2.7 Schematic of contact printing of QDs [34]. ........................... 35
Figure 2.8 Patterned QD-LED created by Kim et al. (c) Device schematic
(d) Photograph of light emission [34]. ........................................................... 36
Figure 2.9 (a) Photographs of operating QD-LEDs of different colors. (b)
EL (solid lines) and PL (dashed lines) corresponding to these devices. (c)
Photograph of luminescence of different colored QDs. (d) EQE (e) Power
efficiency and (f) J-V characteristics of these devices [35]. ........................... 37
Figure 2.10 (a) Band diagram of device with NiO inorganic HTL by
Caruge et al. (b) J-V characteristics of the device with high resistivity
(squares) and low resistivity (circles) NiO [38]. ............................................ 38
xii
Figure 2.11 (a) Device structure with inorganic HTL and ETL by Caruge
et al. (b) Band diagram of the device [42]. ................................................... 39
Figure 3.1 (a) UV/Visible Absorbance and (b) photoluminescence (PL)
spectra of CdSe quantum dots. ..................................................................... 43
Figure 3.2 Device structures for (a) Device 1, (b) Device 2 and (c) Device
3. ................................................................................................................... 44
Figure 3.3 (a) Height and (b) phase image of spin coated mixture of QDs
and TPD. ...................................................................................................... 46
Figure 3.4 Current density - Voltage (J-V) characteristics of device 1. It is
also representative of the J-V characteristics for devices 2 and 3. ................ 47
Figure 3.5 EL intensity curves at different current densities for device 1.
The peak at 460 nm corresponds to EL from TPD and the peak at 610 nm
corresponds to EL from CdSe/ZnS quantum dots. Inset shows the QD to TPD
emission ratio, which is almost constant at 0.2............................................. 48
Figure 3.6 EL intensity curves at different current densities for device 2.
Inset shows the QD to TPD emission ratio, which is almost constant at 1.0.
...................................................................................................................... 49
Figure 3.7 EL intensity curves at different current densities for device 3.
...................................................................................................................... 50
Figure 3.8 Band diagram for device 2. The band diagrams for other two
devices can be derived from this, either by removing PEDT:PSS (device 1) or
by removing TPD (device 3). ........................................................................ 51
xiii
Figure 3.9 Comparison of EL intensities from devices 1, 2 & 3 with the
sum of EL intensities of device 1 & 3 at 100 mA/cm2 showing that, EL from
QDs for device 2 and sum of device 1 and 3 overlap. Inset shows that the same
is true at 10 and 50 mA/cm2 as well. ............................................................ 53
Figure 4.1 Schematic device structure for a Quantum Dot Light Emitting
Diode. ............................................................................................................ 72
Figure 4.2 Flat band diagram for a sample device which consists of Si
quantum dots, ZnO as the electron transport layer, NiO as the hole transport
layer and ohmic contacts to both anode and cathode. The horizontal lines at
ends signify the metal Fermi energies for anode and cathode. Energies are in
eV with respect to the vacuum level. ............................................................ 73
Figure 4.3 Discretization scheme for QD-LED focusing on the QD layers.
...................................................................................................................... 74
Figure 4.4 Carrier densities and electrostatic potential in a device with 5
QD layers (represented by dots on the plot) at equilibrium and with an
applied voltage of 0.5 V. ............................................................................... 75
Figure 4.5 Total device current density (log and linear) plots with
increasing forward bias. The recombination current density is also plotted.
...................................................................................................................... 76
Figure 4.6 Current density vs applied voltage for different assumed values
of the carrier capture coefficients (τ). ........................................................... 77
xiv
Figure 4.7 (a) Current density vs applied voltage and (b) ratio of
recombination current to total current (Jrec/J) for various radiative
recombination coefficients (γ). ...................................................................... 78
1.1 Introduction to Light Emitting Devices INTRODUCTION
1
Chapter 1
Introduction
1.1. Introduction to Light Emitting Devices
Incandescent light bulbs still enjoy a 90% share of the household lighting
market. One of the major reasons they occupy such a large market share
despite the inefficiency of incandescent lighting, is that the light spectrum
from such sources is very close to sunlight. Hence, the color rendering in in-
candescent lighting is extremely good [1]. Lighting represents 15% of total
energy consumption in the U.S. So, if the efficiency of lighting is improved, it
will greatly reduce the total energy consumption. Compact fluorescent lamps
(CFLs) have much better efficiency but they suffer from problems of poor
color rendering and color temperature. CFLs also have a significant envi-
ronmental impact because they use mercury, which is extremely toxic to both
nature and humans. Inorganic light emitting diodes (ILEDs) usually emit
light with a reasonably sharp spectrum, so they are not as useful in lighting
applications. A broader spectrum of light emission is required for lighting
applications. Moreover, they are difficult to fabricate and tuning the bandgap
of ILEDs to get a desired color output requires great deal of engineering. Even
1.1 Introduction to Light Emitting Devices INTRODUCTION
2
upon having manufactured ILEDs of three primary colors, getting white light
out of ILED based lamps is extremely difficult because of the different effi-
ciencies of the three ILEDs, particularly as the individual sources age. This
results in poor color rendering.
In the last two and a half decades, there has been great interest in or-
ganic light emitting diodes (OLEDs). They are easier to fabricate than ILEDs
and low-cost mass-production techniques like printing can be used to deposit
organic semiconductors. Not only do they suffer from the same problems of
color rendering as ILEDs, they also emit with a wider spectrum than ILEDs,
so control of exact color is even more difficult. Incorporation of quantum dots
as the light emitters in organic diodes can result in structures called quantum
dot–light emitting devices (QD-LEDs). They have much sharper spectra and
use different sized quantum dots of a single material to create the exact colors
required. Even white light can be produced using this method by combining
quantum dots of different sizes in the correct proportions. Some of the unde-
sirable properties of organic semiconductors creep into these devices, but the
beauty of using quantum dots as light emitters is that we can use inorganic
semiconductors instead of organic semiconductors to get rid of the problems
that affect organic semiconductors. Depending on the application, one or the
other may be more suitable.
1.2 Background of Quantum Dots
The structure and functioning of QD
detail in Chapter 2. This chapter will delve deeper into the properties of
quantum dots that make them extremely desirable as light emitters.
1.2. Background of Quantum Dots
Figure 1.1 Density of states of bulk semiconductor, quantum well, quantum wire and quantum dot.
Figure 1.1 shows the density of states comparison of bulk materials and
various quantum structures. The density of states of bulk materials is co
tinuous, whereas it starts becoming quantized as we confine the directions in
which the electrons can move freely. Once all
the energy states become completely discrete as shown in
an atom. The existence of continuous bands in bulk semiconductors can be
considered a splitting of degenerate levels due to all of the atoms in a cry
talline lattice. As we reduc
the bands are no longer continuous and they have states like an atom.
Background of Quantum Dots INTRODUCTION
3
The structure and functioning of QD-LEDs will be described in more
detail in Chapter 2. This chapter will delve deeper into the properties of
tum dots that make them extremely desirable as light emitters.
Background of Quantum Dots
Density of states of bulk semiconductor, quantum well, quantum wire and quantum dot.
shows the density of states comparison of bulk materials and
various quantum structures. The density of states of bulk materials is co
s, whereas it starts becoming quantized as we confine the directions in
ns can move freely. Once all three dimensions are confined,
the energy states become completely discrete as shown in Figure 1.1
an atom. The existence of continuous bands in bulk semiconductors can be
considered a splitting of degenerate levels due to all of the atoms in a cry
talline lattice. As we reduce the number of atoms interacting within a lattice,
the bands are no longer continuous and they have states like an atom.
INTRODUCTION
LEDs will be described in more
detail in Chapter 2. This chapter will delve deeper into the properties of
tum dots that make them extremely desirable as light emitters.
Density of states of bulk semiconductor, quantum well,
shows the density of states comparison of bulk materials and
various quantum structures. The density of states of bulk materials is con-
s, whereas it starts becoming quantized as we confine the directions in
three dimensions are confined,
1, just like
an atom. The existence of continuous bands in bulk semiconductors can be
considered a splitting of degenerate levels due to all of the atoms in a crys-
e the number of atoms interacting within a lattice,
the bands are no longer continuous and they have states like an atom.
1.2 Background of Quantum Dots INTRODUCTION
4
The exciton Bohr radius (aB) is the size at which the transition from
bulk to quantum structure takes place. This is the typical size of a bound
electron hole pair. It is defined from Bohr’s model of a hydrogen atom. Bohr’s
radius for most semiconductors is several nanometers. Thus, the particles of
semiconductors with diameters equal to or less than the Bohr radius act as
quantum dots.
Along with the quantization of energy levels upon reducing the particle
size to the nanometer range, the value of the bandgap also changes. This can
be attributed to an increase in energy level splitting, causing the conduction
and valence band to move away from each other. The bandgap changes can be
estimated using the quantum mechanical “particle in a box” theory [2]. For
the weak confinement regime (a>aB), the energy of exciton in a spherical
quantum dot is given by
Enml
=Eg−
Ry*
n2 + ħ2χ
2
2MR2 (1.1)
where R is the radius of the quantum dot, χml
are the roots of spherical
Bessel function (m – number of the root, l – order of the function), Ry* is the
exciton Rydberg energy (Ry*= e2
2εaB
), and M=m*e+m
*h. The exciton energy
in the strong confinement regime (a≤aB) is given by
1.3 Synthesis of Quantum Dots INTRODUCTION
5
Enl
=Eg+
ħ2
2µR2χ2nl (1.2)
As can be seen from Equation 1.2, the bandgap of quantum dots is always
greater than that of bulk semiconductor and it increases as we reduce the
radius of the quantum dots.
1.3. Synthesis of Quantum Dots
There are numerous ways by which quantum dots can be synthesized.
They can be divided into two major categories. The following two subsections
describe each of those categories.
1.3.1 Liquid phase synthesis
Liquid phase synthesis is the most common technique for making quan-
tum dots. It can be further subdivided into two primary categories:
• Aerosol based synthesis: In this method, liquid precursors are
aerosolized in presence of gaseous precursors to produce the desired
quantum dots. The size of quantum dots is controlled by the con-
centration of the precursors and the size of aerosol droplets. This
method was used by both Ahonen et al. and by Kim et al. to create
TiO2 [3] and copper quantum dots [4], respectively.
• Solution based synthesis: In this method, solution based pre-
cursors are added together to allow for a chemical reaction. The
1.3 Synthesis of Quantum Dots INTRODUCTION
6
concentration and timing of adding the reagents is such that the
particles stop growing after a certain size. This is by far the most
commonly used technique.
Of these two the latter provides much better control of the reaction and so
the size of the quantum dots. Thus, we will only discuss solution based
methods in further detail.
There are various solution based methods to synthesize quantum dots,
but the hot injection method gives the best quality and highly monodispered
quantum dots. For most applications, monodispersity of quantum dots is de-
sired as many of the properties of quantum dots are size dependent, especially
the bandgap, which determines the luminescence wavelength. If the quantum
dots are not monodispersed, the luminescence spectra will be broad.
Murray, et al. created highly monodispersed CdE (E= sulfur, selenium,
tellurium) particles using the hot injection method [5]. They accomplish it by
rapid injection of a precursor solution into a hot solution of surfactants.
Figure 1.2 shows the different phases of a hot injection reaction. When pre-
cursors are injected into the reaction chamber, the concentration goes over the
nucleation threshold and quantum dots begin to nucleate. As time progresses,
the concentration of precursor reduces because it is being consumed in the
nucleation reaction. Once the concentration of the precursors becomes smaller
than the nucleation concentration, the nuclei begin to grow to form larger
quantum dots. If the time to reach the saturation concentration is much
1.3 Synthesis of Quantum Dots
higher than the time for nucleation, all the particles will have approximately
the same size, usually with a 10% or less standard deviation.
Figure 1.2 Nucleation and growth phases for quantum dots in hot injection method. Adapted from
1.3.2 Vapor phase synthesis
In vapor-phase synthesis of quantum dots, conditions are created such
that the vapor phase mixture is thermodynamically unstable compared to the
precursors. A supersaturated vapor is an example of such a condition. Then it
is condensed in a controlled mann
Synthesis of Quantum Dots INTRODUCTION
7
higher than the time for nucleation, all the particles will have approximately
the same size, usually with a 10% or less standard deviation.
Nucleation and growth phases for quantum dots in hot injection method. Adapted from [6].
Vapor phase synthesis
phase synthesis of quantum dots, conditions are created such
that the vapor phase mixture is thermodynamically unstable compared to the
precursors. A supersaturated vapor is an example of such a condition. Then it
is condensed in a controlled manner to create quantum dots [7].
INTRODUCTION
higher than the time for nucleation, all the particles will have approximately
Nucleation and growth phases for quantum dots in hot
phase synthesis of quantum dots, conditions are created such
that the vapor phase mixture is thermodynamically unstable compared to the
precursors. A supersaturated vapor is an example of such a condition. Then it
1.3 Synthesis of Quantum Dots INTRODUCTION
8
1.3.2.1 Using solid precursors
The basic philosophy for this method of synthesis of quantum dots is to
induce nucleation by supersaturation of hot gases by vaporization of solids
and then rapidly cooling the gas.
Condensation in inert atmosphere:
In this method, a solid is heated and the vapor mixed with a background
gas and then, a cooled gas is added to this vapor-gas mixture to reduce the
temperature, which results in creation of quantum dots. This method is
primarily used for synthesis of metal quantum dots because most of the metals
are relatively easy to vaporize with a reasonable vapor pressure. Wegner et al.
prepared bismuth quantum dots using this method [8]. Ohno fabricated
composite quantum dots (Al/Pb, Si/In, Ge/In and Al/Pb) by creating
quantum dots of one material using inert gas condensation in one chamber,
allowed them to flow into second chamber, where the second material was
deposited onto the quantum dots to create two material composite quantum
dots [9].
Pulsed laser ablation:
Instead of using heat to vaporize a material, pulsed laser can be used to
vaporize materials that are not easily vaporized, or, are in a confined space.
This method is primarily used to create Si quantum dots as demonstrated by
1.3 Synthesis of Quantum Dots INTRODUCTION
9
Nakata et al. [10] and Marine et al. [11]. Hydrogenated Si quantum dots were
also prepared using pulsed laser ablation by Makimura et al. [12].
Spark discharge generation:
This method is again used for creating quantum dots of metals. In the
presence of an inert gas, a high enough voltage so as to create a spark is ap-
plied between the two electrodes (made of the metal to be vaporized). The
spark vaporizes the metal and quantum dots are formed upon condensation of
the vapors. Weber et al. used spark discharge to create Ni quantum dots [13].
Ion sputtering:
This method involves using a target of the quantum dot material, usually
a metal, during ion sputtering using inert gas. Fabrication of quantum dots of
12 different materials using ion sputtering was demonstrated by Urban et al.
[14].
1.3.2.2 Using liquid or vapor precursors
In this case, the supersaturation of gas-vapor mixture is achieved by a
chemical reaction, rather than a physical vaporization.
Chemical vapor synthesis:
This process is very similar to chemical vapor deposition (CVD), but, in
this case a hot wall reactor is used, so that the nucleation takes place in the
1.3 Synthesis of Quantum Dots INTRODUCTION
10
vapor phase rather than the wall of the reactor or the substrate. The substrate
is kept at a lower temperature, so it acts as a collection location rather than a
nucleation location as used by Singh et al. while fabricating tin nitride
quantum dots [15]. This method can be used to create many different kinds of
quantum dots, with various gas and precursor combinations possible.
Laser pyrolysis:
Another method of heating the precursors to help in reaction and nucle-
ation is to use laser energy. This method consumes a lot less energy compared
to chemical vapor synthesis because only a small portion of gas is heated in-
stead of a whole chamber. Quantum dots of various elemental and compound
materials can be fabricated using this method as shown by: Borsella et al. to
create MoS2 quantum dots [16], Kamlag et al. to create SiC quantum dots [17]
and Ledoux et al. to create Si quantum dots [18,19].
Thermal plasma synthesis:
An alternate method for providing the required energy is to inject the
precursors into a thermal plasma. The thermal plasma decomposes the pre-
cursors and the radicals react to form the desired quantum dots. This method
was used by Heberlein et al. to create hard coatings using quantum dots [20].
1.3 Synthesis of Quantum Dots INTRODUCTION
11
Flame synthesis:
Here the precursors are combusted and the required energy is provided by
the combustion reaction. Most of the carbon quantum dots in the atmosphere
are created by the combustion of hydrocarbons in this manner. This method
was used to create ferric oxide quantum dots by Janzen et al. [21], and titania
quantum dots by Wegner et al. [22].
Low temperature reactive synthesis:
Some materials can be reacted without external application of heat. As
the energy from the reaction and injection method should be sufficient to
continue the reaction, this method is limited a few reactions. ZnSe quantum
dots were fabricated using this method by Sarigiannis et al. [23].
Non-thermal plasma synthesis:
Other methods of vapor phase synthesis suffer from agglomeration of
quantum dots, because there are no capping agents as in liquid phase syn-
thesis methods. But, non-thermal plasma synthesis can solve this problem
[24]. A non-thermal plasma is created when electrons are at a much higher
temperature than the ions, which results in quantum dots being negatively
charged. Due to this unipolar charging of quantum dots, the agglomeration of
quantum dots is highly reduced. This method was used by Gorla et al. to
create Si and Ge quantum dots [25].
1.4 Luminescence from Quantum Dots
Figure 1.3 Emission of light from bulk semiconductors upon relaxation of electrons from the conduction band to valence bands. Due to the continuity of bands, the wavelength o
different de
1.4. Luminescence from Quantum Dots
Bulk semiconductors have a conduction and a valence band due to the
interaction of atoms in a lattice. The conduction and valence bands are se
arated by a region of zero d
in the conduction and valence bands are so close to each other that they are
considered continuous. Electrons can exist in the conduction band and the
valence band, but, they cannot stay in the bandgap of t
When energy is supplied to the electrons, they can move from the valence
band to the conduction band. When the electrons are relaxed from the co
duction band to the valence band, they emit light as shown in
Luminescence from Quantum Dots INTRODUCTION
12
Emission of light from bulk semiconductors upon relaxation of electrons from the conduction band to valence bands. Due to the continuity of bands, the wavelength of emitted light is
different depending on the energy lost by electrons.
Luminescence from Quantum Dots
Bulk semiconductors have a conduction and a valence band due to the
interaction of atoms in a lattice. The conduction and valence bands are se
arated by a region of zero density of states known as the bandgap. The states
in the conduction and valence bands are so close to each other that they are
considered continuous. Electrons can exist in the conduction band and the
valence band, but, they cannot stay in the bandgap of the semiconductor.
When energy is supplied to the electrons, they can move from the valence
band to the conduction band. When the electrons are relaxed from the co
duction band to the valence band, they emit light as shown in Figure
INTRODUCTION
Emission of light from bulk semiconductors upon relaxation of electrons from the conduction band to valence bands.
f emitted light is
Bulk semiconductors have a conduction and a valence band due to the
interaction of atoms in a lattice. The conduction and valence bands are sep-
ensity of states known as the bandgap. The states
in the conduction and valence bands are so close to each other that they are
considered continuous. Electrons can exist in the conduction band and the
he semiconductor.
When energy is supplied to the electrons, they can move from the valence
band to the conduction band. When the electrons are relaxed from the con-
Figure 1.3. The
1.4 Luminescence from Quantum Dots
wavelength of the emitted light is given by the energy lost by the electron in
this transition. The most likely scenario for this relaxation to take place is
from the bottom of the c
there are non empty states around the conduction band edge, so some ele
trons from those states also relax to emit light at energies
bandgap as shown in Figure
for spectrum of luminescence bulk semiconductor light emitting diodes
(LEDs) is nonzero because of the “continuous” valence and conduction bands.
Figure 1.4 Discrete density of states for quantum dots. Electrons can be excited to lowest level in the conduction “band” and thus the only
recombination takes place between highest level in the valence “band.” The interaction between other states is highly unlikely.
Luminescence from Quantum Dots INTRODUCTION
13
wavelength of the emitted light is given by the energy lost by the electron in
this transition. The most likely scenario for this relaxation to take place is
from the bottom of the conduction band to the top of the valence band. But
there are non empty states around the conduction band edge, so some ele
trons from those states also relax to emit light at energies higher than
Figure 1.3. So, the full width at half maximum (FWHM)
for spectrum of luminescence bulk semiconductor light emitting diodes
because of the “continuous” valence and conduction bands.
Discrete density of states for quantum dots. Electrons can be excited to lowest level in the conduction “band” and thus the only
recombination takes place between highest level in the valence d.” The interaction between other states is highly unlikely.
INTRODUCTION
wavelength of the emitted light is given by the energy lost by the electron in
this transition. The most likely scenario for this relaxation to take place is
onduction band to the top of the valence band. But
there are non empty states around the conduction band edge, so some elec-
higher than the
. So, the full width at half maximum (FWHM)
for spectrum of luminescence bulk semiconductor light emitting diodes
because of the “continuous” valence and conduction bands.
Discrete density of states for quantum dots. Electrons can be excited to lowest level in the conduction “band” and thus the only
recombination takes place between highest level in the valence d.” The interaction between other states is highly unlikely.
1.4 Luminescence from Quantum Dots
Figure 1.5 The spectrum of quantum dots with different size distributions. As the variance in size increases, the spectrum becomes
But in quantum dots, due to the electrons being confined in a small
space (on the order of wavelength of electron wavefunction), the energy levels
in the conduction and valence bands move far apart and can no longer be
considered continuous as shown in
an electron in an energy level higher than the “conduction band” edge is
negligible. Hence, the only source of luminescence from QD
of electrons from the bottom of conduction “band” to top of the valence
“band.” If the same material is used, the bandgap is only dependent on the
size of the quantum dots. As
bandgap increases as shown in
section 1.2. So, the source of a spread in luminescence spectrum from qua
tum dots is the distribution in size of the quantum. The tighter the distrib
Luminescence from Quantum Dots INTRODUCTION
14
The spectrum of quantum dots with different size distributions. As the variance in size increases, the spectrum becomes
broader.
in quantum dots, due to the electrons being confined in a small
space (on the order of wavelength of electron wavefunction), the energy levels
in the conduction and valence bands move far apart and can no longer be
considered continuous as shown in Figure 1.4. Now the probability of finding
an electron in an energy level higher than the “conduction band” edge is
negligible. Hence, the only source of luminescence from QDs is the relaxation
of electrons from the bottom of conduction “band” to top of the valence
“band.” If the same material is used, the bandgap is only dependent on the
size of the quantum dots. As the size of the quantum dots decreases, the
s as shown in Figure 1.7 and described in more details in
, the source of a spread in luminescence spectrum from qua
tum dots is the distribution in size of the quantum. The tighter the distrib
INTRODUCTION
The spectrum of quantum dots with different size distributions. As the variance in size increases, the spectrum becomes
in quantum dots, due to the electrons being confined in a small
space (on the order of wavelength of electron wavefunction), the energy levels
in the conduction and valence bands move far apart and can no longer be
. Now the probability of finding
an electron in an energy level higher than the “conduction band” edge is
s is the relaxation
of electrons from the bottom of conduction “band” to top of the valence
“band.” If the same material is used, the bandgap is only dependent on the
creases, the
and described in more details in
, the source of a spread in luminescence spectrum from quan-
tum dots is the distribution in size of the quantum. The tighter the distribu-
1.4 Luminescence from Quantum Dots
tion of quantum dot sizes, the sharper is the spectrum of quantum dot l
minescence as shown in
Figure 1.6 Silicon quantum dots of different size emitting various colors of the spectrum. Courtesy: Dr. Rick Liptak
The ability to control the size of quantum dots also gives quantum dots
the ability to emit lights of different colors due to different bandgaps. An
example of different colors emitted by the quantum dots of same material
produced by the same process is shown in
quired to produce different colors is to change the growth conditions. This
property makes quantum dots extremely versatile for use in various applic
tions requiring different colors of light. Electronic displays are one such a
plication, where color accuracy and sharp spectrum are extremely important.
Quantum dot based devices can be used to creat
cause research and development cost can be highly reduced as one single
process can produce all the primary colors of the spectrum.
Luminescence from Quantum Dots INTRODUCTION
15
tion of quantum dot sizes, the sharper is the spectrum of quantum dot l
minescence as shown in Figure 1.5.
Silicon quantum dots of different size emitting various colors of the spectrum. Courtesy: Dr. Rick Liptak [26].
The ability to control the size of quantum dots also gives quantum dots
the ability to emit lights of different colors due to different bandgaps. An
example of different colors emitted by the quantum dots of same material
produced by the same process is shown in Figure 1.6. The only change r
quired to produce different colors is to change the growth conditions. This
ntum dots extremely versatile for use in various applic
tions requiring different colors of light. Electronic displays are one such a
plication, where color accuracy and sharp spectrum are extremely important.
Quantum dot based devices can be used to create such devices cheaply b
cause research and development cost can be highly reduced as one single
process can produce all the primary colors of the spectrum.
INTRODUCTION
tion of quantum dot sizes, the sharper is the spectrum of quantum dot lu-
Silicon quantum dots of different size emitting various
The ability to control the size of quantum dots also gives quantum dots
the ability to emit lights of different colors due to different bandgaps. An
example of different colors emitted by the quantum dots of same materials
. The only change re-
quired to produce different colors is to change the growth conditions. This
ntum dots extremely versatile for use in various applica-
tions requiring different colors of light. Electronic displays are one such ap-
plication, where color accuracy and sharp spectrum are extremely important.
e such devices cheaply be-
cause research and development cost can be highly reduced as one single
1.5 Structure of Dissertation INTRODUCTION
16
1.5. Structure of Dissertation
• Chapter 2 introduces quantum dot - light emitting devices
(QD-LEDs). It also details the device structure and functioning of
QD-LEDs.
• Chapter 3 details the fabrication and testing methods of QD-LEDs.
It also covers the results and analysis of charge transport in or-
ganic-inorganic hybrid QD-LEDs.
• Chapter 4 covers the model description of QD-LEDs and simulation
results for the model.
• Chapter 5 concludes the dissertation with most important results
and future work.
1.5 Structure of Dissertation
Figure 1.7 Dependence of
Structure of Dissertation INTRODUCTION
17
Dependence of bandgap on size of Si quantum dotAdapted from [27].
INTRODUCTION
quantum dots.
2.1 Device Structure REVIEW OF QD-LEDS
18
Chapter 2
Review of QD-LEDs
2.1. Device Structure
A generic device structure of quantum dot - light emitting device
(QD-LED) is shown in Figure 2.1. Any QD-LED basically consists of an
electron transport layer (ETL), which facilitates the transport of electrons
from the cathode and blocks the transport of holes. The hole transport layer
facilitates the transport of holes from the anode to the quantum dot layer and
blocks the transport of electrons. The quantum dot layer acts the recombi-
nation regime, where the transported electrons and holes recombine to create
photons. A band diagram of a QD-LED is shown in Figure 2.2. It shows
the basis on which the materials for ETL and HTL are chosen. The ETL is
chosen such that its conduction band aligns with the conduction band of the
QDs which makes it easy for the electrons to be transferred from ETL to QD
layer. The valence band of ETL is much lower than the valence band of the
QDs, so that the holes require a large amount of energy to be transferred from
QDs into the ETL. Hence, the electron transport layer also acts as a hole
2.1 Device Structure REVIEW OF QD-LEDS
19
blocking layer. Similarly, the hole transport layer is chosen such that the
electrons are transported easily from the HTL to the QD layer and electrons
are blocked from being transported from the QD layer to HTL.
When a power supply is connected to the device, with the positive ter-
minal connected to the anode and the negative terminal connected to the
cathode, electrons are injected into the ETL and holes are injected into the
HTL. Electrons and holes are finally transported into the QD layer, where
they recombine to emit light. The electrons and holes are blocked by HTL
and ETL respectively, from overshooting out of the QD layer (which reduces
the efficiency of light emission). As explained in Chapter 1, the wavelength
of light emission can be changed by changing the size of QDs, everything else
staying the same, as long as the band alignment remains reasonable.
QD-LEDs can be divided into two major types:
• Organic transport layers.
• Inorganic transport layers.
Each of these will be explored in detail in the following sections.
2.2 Organic Transport Layers REVIEW OF QD-LEDS
20
2.2. Organic Transport Layers
Colvin et al. were the first to demonstrate a working QD-LED based on
organic transport layers [28]. p-paraphenylene vinylene (PPV) was used as
the transport layer and the QDs were made of CdSe. Indium tin oxide (ITO)
acted as the transparent anode and Mg was used as the opaque and reflective
cathode. ITO coated glass was used as the substrate for device fabrication.
CdSe QDs were dissolved in toluene and spin coated 5 times, to create ordered
sheets of CdSe QDs. 100 nm PPV layer was spin coated on top to reduce
electric fields in the device at the voltages required for device operation.
They used two device configurations (shown in Figure 2.3):
• structure A: PPV acted as electron transport layer
• structure B: PPV acted as hole transport layer
In structure A, once the electrons are injected from Mg cathode into the
PPV layer, they get transported into the QD layer, but there is no barrier for
electrons to stop them from overshooting into the ITO anode. Similarly,
once holes are injected from ITO anode into QD layer, they can easily over-
shoot into the PPV layer. Thus, the light emission efficiency of this device is
very poor. In structure B, both electrons and holes experience a barrier at
the QD/PPV interface, thus the probability of overshoot of carriers is highly
reduced. As a result structure B has higher light emission compared to
structure A.
2.2 Organic Transport Layers REVIEW OF QD-LEDS
21
There was a lot of advancement in core/shell quantum dot growth be-
tween 1994 and 1997 [29-31]. Core/shell quantum dots show increased
photoluminescence (PL) efficiency, primarily due to resistance to oxidation of
the core. As discussed in Chapter 1, the light emission wavelength of QDs is
dependent on the size of the quantum dot. Upon oxidation the emission
wavelength decreases due to reduction in effective size of the QDs. QD-LEDs
should be able to give consistent performance at the desired wavelength for
long periods of time, so, maintaining a constant size is essential. For com-
pound semiconductors, however, there is a second problem. Oxidation is al-
most always preferential to one of the elements leaving a high concentration of
defects at the oxides/semiconductor interface. These centers often provide a
high concentration of states in the gap, acting as effective recombination
centers. Thus, resistance to oxidation one of the most important qualities of
QDs required for QD-LEDs.
In 1997, Schlamp et al. improved the efficiency and lifetime of the device
used by Colvin et al. (described above) by using core/shell CdSe quantum
dots [32]. The core was composed of CdSe and the shell was made of CdS.
The CdS shell helps avoid oxidation of CdSe core, which results in increased
luminescence efficiency of devices based on core/shell quantum dots. CdS also
separates the carriers, which are confined to the core, from defects that are
highly likely to occur at the shell/air or shell/matrix interface. They used two
different QD sizes to demonstrate that different colored LEDs could be cre-
2.2 Organic Transport Layers REVIEW OF QD-LEDS
22
ated using the same structure and materials, just by changing the QD size.
The PL and EL (electroluminescence) spectra for the two different sized QDs
are shown in Figure 2.4. An increase of a factor of 20 was observed in
quantum efficiency and device lifetime increased by a factor of 100 when the
device structure used was the same as structure B in Figure 2.3.
In 2002, Coe et al. improved upon the above designs, to include both
electron and hole transport layers, which could confine the electrons and holes
in the quantum dots, increasing the light emission efficiency [33]. The en-
ergy level diagram of the device is shown in Figure 2.5. They used a novel
technique of phase separation to create a monolayer of quantum dots. So,
instead of using polymers as the hole transport layer, small molecule organic
material, N,N’-diphenyl-N,N’-bis(3-methylphenyl)-(1,1’- biphenyl)-
4,4’-diamine (TPD) was used, which would not trap the quantum dots in the
organic semiconductor-QD solution. A TPD-QD solution was spin coated on
ITO, which results in phase separation of QDs from TPD. If the right ratio
of QDs and TPD are used, it results in a monolayer of QDs on top of TPD
HTL. Coe et al. were able to achieve 22% - 100% coverage (a complete
monolayer of QDs) using this method, by varying the ratio of QDs and TPD
in the TPD-QD solution. After growing QD and HTL layers, the ETL which
again consisted of small molecular organic semiconductor,
tris-(8-hydroxyquinoline)aluminium (Alq3), was deposited using thermal
2.2 Organic Transport Layers REVIEW OF QD-LEDS
23
evaporation. A cathode consisting of 10:1 Mg:Ag followed by Ag cap was
deposited using thermal evaporation.
Figure 2.6(a) shows the electroluminescence (EL) spectrum of the device.
As shown in Figure 2.5, the barrier for the holes to be overshot into the Alq3
ETL is very small, which results in a significant recombination of electrons
and holes in Alq3 layer. Thus, only 62% of the light output is a result of
emission from QDs. But, upon introducing an additional hole blocking
layer, 3-(4-biphenylyl)-4-phenyl-5-t-butylphenyl-1,2,4-triazole (TAZ), between
QDs and Alq3 increases the light output from quantum dots to 85% as shown
in Figure 2.6(b). By using this device structure, Coe et al. improved the
efficiency 25 fold compared to the device structure used by Schlamp et al.
[32,33]. So, it is extremely important to control the location of recombina-
tion in QD-LEDs to maximize the light output efficiency.
In the phase-separation method as described by Coe et al., the materials
that can be used are limited by the ability of QDs to form a solution with the
HTL and to phase-separate during spin-coating. This limits our flexibility in
choosing materials. We may not be able to choose a material which has
better band alignment, but not chosen because QDs cannot phase-separate
during spin-coating, or QDs and HTL are not soluble in the same solvent.
Kim et al. presented a solution to this problem by using contact printing of
QDs [34]. A schematic of this contact printing method is shown in Figure
2.7. The major steps of contact printing are:
2.2 Organic Transport Layers REVIEW OF QD-LEDS
24
1) A patterned silicon master is created by using photolithography
and plasma etching. Poly(dimethylsiloxane) (PDMS) is molded
using the silicon master mold.
2) The PDMS stamp is coated with parylene-C using chemical vapor
deposition (CVD). Parylene-C is more resilient to chemicals and
moisture than PDMS, so, it increases the longevity of PDMS
stamp.
3) Then, the PDMS stamp is spin-coated with QDs dissolved in a
solvent and allowed to dry.
4) Finally, the PDMS stamp is brought in contact with the substrate
which results in the transfer of QDs from the stamp to the sub-
strate. The QDs in the trenches in the PDMS stamp are not
transferred because they do not come in contact with the sub-
strate.
A side, but important, benefit of using the contact printing method is that
QDs can be transferred onto the substrate in a pattern, which can be used to
create different colored pixels on the same substrate if aligned properly. An
example of a patterned matrix of different colored QD-LEDs is shown in
Figure 2.8. Figure 2.8(c) shows a schematic of the patterned device with
perpendicular red and green patterns and Figure 2.8(d) shows a photograph of
a working patterned QD-LED. The light blue color is emitted by TPD,
whereas red and green color is emitted by QDs.
2.3 Inorganic Transport Layers REVIEW OF QD-LEDS
25
Anikeeva et al. extended the contact printing technique to vary the
QD-LED color over the entire visible spectrum, by changing the QD size and
composition [35]. Figure 2.9 shows various photographs of QD-LEDs of
different colors. Their PL, external quantum efficiency (EQE), power
efficiency and current density – voltage characteristics are also shown in
Figure 2.9.
2.3. Inorganic Transport Layers
QD-LEDs with organic transport layers suffer from all the problems that
ail organic light emitting diodes (OLED) such as, limited lifetime, degradation
on exposure to air and moisture, and instability of metal contacts [36,37].
Thus, Caruge et al. were the first to explore the possibility of using wide band
gap inorganic semiconductors as transport layers for QD-LEDs [38]. In their
work, instead of replacing both electron and hole transport layers, they just
replaced the hole transport layer by NiO. The electron transport layer was
kept organic (Alq3). NiO was chosen as HTL because it had previously been
used as hole transport/injecting layer in optoelectronic devices [39-41].
The role of transport layers is to avoid quenching of QD
electroluminescence due to proximity to the contacts, due to the infinite
recombination velocity at the contacts. The NiO transport layers should not
have too many charge carriers so as to cause substantial quenching of QD EL.
At the same time, the carrier density in NiO should not be so low that it
2.3 Inorganic Transport Layers REVIEW OF QD-LEDS
26
causes a dramatic increase in the overall resistance of the device. Sheet
resistance is inversely proportional to carrier density so it can be used a good
estimator of carrier density assuming that the mobility is nearly constant.
Caruge et al. conducted PL experiments by depositing QDs on NiO films of
varying resistivity [38]. It was found that for NiO films with high resistivity
( cm⋅Ω≈1ρ ), the PL of QDs remains high, whereas for NiO with low
resistivity ( cm⋅Ω×≈ −4105ρ ), it was quenched.
NiO resistivity can be varied by changing the oxygen to argon ratio
during sputtering. The controllability of resistivity is useful not only to
avoid quenching of EL of QDs, but, also gives us control on charge balance in
the QDs. If the hole concentration in QDs is higher than electron
concentration, the resistivity of NiO can be increased to reduce the hole
concentration in QDs to achieve better charge balance. If the hole
concentration in QDs is lower than that of electrons, the resistivity can be
decreased to achieve better charge balance and consequently, better efficiency.
The device is fabricated by sputtering NiO on ITO coated glass, followed
by spin coating of CdSe QDs, followed by thermal evaporation of Alq3 and Mg
contact. The band diagram of this device is shown in Figure 2.10(a). Two
different devices were fabricated by Caruge et al. with different NiO
resistivities; the current-voltage characteristics of which are plotted in Figure
2.10(b). At low voltages, the current density is limited by the resistivity of
NiO layer, hence the current density for low resistivity NiO device is much
2.3 Inorganic Transport Layers REVIEW OF QD-LEDS
27
higher than the current density for the device with high NiO resistivity. But
at high voltages, the current is limited by transport through the quantum
dots, hence, the two J-V curves overlap each other at high voltages.
In 2008, Caruge et al. made a device with both transport layers being
organic [42]. NiO was used as HTL and ZnO:SnO2 was used as ETL. The
device structure and corresponding band diagram is shown in Figure 2.11.
Because, the ETL is deposited by sputtering, there is a lot of flexibility in
choosing the material for ETL. Thus, a ZnO:SnO2 alloy was used as the
ETL, because it acts as a good transporter of electrons with high electron
mobility as well as the conduction band aligns well with the Fermi level of the
cathode and the conduction band of quantum dots. At the same time, the
valence band is deep enough that it blocks the overshoot of holes from the
QDs into the ETL. They observed that inorganic QD-LEDs were more ro-
bust to atmospheric conditions as compared to QD-LEDs with organic
transport layers. They retested after storing unpackaged devices in air for
four days and saw no degradation in performance, which is impossible to see
for unpackaged QD-LEDs based on organic transport layers. They were also
able to drive higher currents through the device compared to organic
QD-LEDs.
2.4 Latest Developments REVIEW OF QD-LEDS
28
2.4. Latest Developments
QD Vision (http://www.qdvision.com/) is a startup that exclusively
manufactures QD-LEDs. Sony is set to use QD technology in their latest
line of televisions, Triluminos [43]. Conventional LCD TVs use a blue LED
coated with phosphors to convert the blue light to white which acts as the
backlight. In case of QD based LCD TVs, the blue LED will be uncoated
and will be placed in a thin glass tube containing QDs. The QDs will be
such that they will absorb blue light and emit pure red and green light. The
resultant white light is more intense at these wavelengths (which are the
primary colors of LCD TVs) compared to conventional phosphor coated
LEDs. Thus, the color is brighter and the TV is more efficient compared to a
conventional LCD TV.
2.4 Latest Developments REVIEW OF QD-LEDS
29
Figure 2.1. Schematic of a generic QD-LED.
2.4 Latest Developments REVIEW OF QD-LEDS
30
Figure 2.2 Band diagram of a generic QD-LED.
2.4 Latest Developments REVIEW OF QD-LEDS
31
Figure 2.3 QD-LED with only ETL (structure A) or only HTL (structure B). In both cases PPV acts as the transport layer. From [28].
2.4 Latest Developments REVIEW OF QD-LEDS
32
Figure 2.4 PL (solid lines) and EL (dashed lines) from two different sized CdSe/CdS core-shell QDs (a) 2.5 nm core/ 0.5 nm shell (b) 3.4 nm core/ 0.7 nm shell [32].
2.4 Latest Developments REVIEW OF QD-LEDS
33
Figure 2.5 Energy band diagram of device used by Coe et al. containing both electron and hole transport layers [33].
2.4 Latest Developments REVIEW OF QD-LEDS
34
Figure 2.6 EL spectrum of (a) device with ETL and HTL (b) device with additional hole blocking layer [33].
2.4 Latest Developments REVIEW OF QD-LEDS
35
Figure 2.7 Schematic of contact printing of QDs [34].
2.4 Latest Developments REVIEW OF QD-LEDS
36
Figure 2.8 Patterned QD-LED created by Kim et al. (c) Device schematic (d) Photograph of light emission [34].
2.4 Latest Developments REVIEW OF QD-LEDS
37
Figure 2.9 (a) Photographs of operating QD-LEDs of different colors. (b) EL (solid lines) and PL (dashed lines) corresponding to these devices. (c) Photograph of luminescence of different colored QDs. (d) EQE (e) Power efficiency and (f) J-V characteristics of these devices [35].
2.4 Latest Developments REVIEW OF QD-LEDS
38
Figure 2.10 (a) Band diagram of device with NiO inorganic HTL by Caruge et al. (b) J-V characteristics of the device with high resistivity
(squares) and low resistivity (circles) NiO [38].
2.4 Latest Developments REVIEW OF QD-LEDS
39
Figure 2.11 (a) Device structure with inorganic HTL and ETL by Caruge et al. (b) Band diagram of the device [42].
3.1 Introduction CHARGE TRANSPORT IN HYBRID QD-LEDS
40
Chapter 3
Charge Transport in Hybrid
QD-LEDs
This chapter has been reprinted with the permission from Kumar,
Brijesh, et al. "Comparing direct charge injection and Forster energy transfer
into quantum dots in hybrid organic/inorganic quantum dot light emitting
devices." Journal of Applied Physics 112.3 (2012): 034501-034501, Copyright
2012, AIP Publishing LLC. [44]
3.1. Introduction
The advent of organic electronics has allowed the integration of inorganic
materials which have good optoelectronic properties. The ability of materials
such as quantum dots (QDs) to be deposited from a solution allows one to
incorporate them into low-cost optoelectronic devices. Over the last few years
several groups have demonstrated such devices.[32,45,46]
There are various factors that influence the emission from Quantum Dot
Light Emitting Devices (QD LEDs). One of the most important of those is the
mechanism for energy and/or charge transfer to the recombination sites
3.1 Introduction CHARGE TRANSPORT IN HYBRID QD-LEDS
41
(QDs). This transfer generally takes place either by direct charge injection of
electrons and holes or by Forster energy transfer of excitons from organic
molecules onto the quantum dots.[46-48] Anikeeva et al. suggest that the
light output from QDs can be increased by confining carriers near the QDs to
promote both mechanisms, but asserts that Forster energy transfer is the
primary mechanism.[48] Chin et al. suggest that light output is maximized by
using an electron transport layer that enhances direct charge injection and
reduces Forster energy transfer.[47] In this work, we investigate the charge
transfer mechanisms into CdSe quantum dots. In agreement with Chin, we
show that carrier confinement increases QD electroluminescence (EL) effi-
ciency. We also find that, for the devices studied here, the optical outputs of
the two mechanisms are independent. That is, the total optical output is well
described by the sum of the two mechanisms, independent of current injection
density. This may be a property of devices in which the Forster mechanism
does not change the population of carriers. In such devices both mechanisms
can be combined to increase the amount of light output from hybrid QD LED.
3.2 Experimental Method CHARGE TRANSPORT IN HYBRID QD-LEDS
42
3.2. Experimental Method
Cadmium selenide quantum dots with an outer shell of zinc sulfide were
synthesized using the SILAR method modified from Reiss et al. [49] The
successive ion layer adsorption and reaction (SILAR) technique deposits a
monolayer at a time of the shell on to the core nanocrystals, and gives ex-
cellent thickness control.[50] This technique has the advantage of avoiding the
nucleation of shell particles, since only one precursor is introduced at a time
and the process can be done in the same reaction vessel as the growth.
To prepare the quantum dots, a mixture of cadmium oxide (CdO, 0.411
g), dodecylphosphonic acid (DPA, 1.6 g) and hexadecylamine (18 g) in
trioctylphosphine oxide or TOPO (8.4 g) was heated to 270°C under a ni-
trogen atmosphere. A solution of trioctylphosphine:selenium (TOP:Se, 4
mL) in TOP (10 mL) was rapidly injected and the particles were grown for 15
minutes at 250°C. The solution was then cooled to 220°C for SILAR injec-
tions of 0.1 M zinc stearate and sulfur solutions. Each monolayer was grown
for 15 minutes per precursor, for a total of two ZnS monolayers. After
completion, the particles were precipitated and washed with methanol, cen-
trifuged, and redispersed in chloroform. Figure 3.1 (a) and (b) show the
UV/Vis and photoluminescence (PL) spectra of the CdSe nanoparticles. The
UV/Vis spectrum showed a first absorption peak at 626 nm, corresponding to
5.6 nm nanocrystals. The second and third absorption peaks of CdSe are also
3.2 Experimental Method CHARGE TRANSPORT IN HYBRID QD-LEDS
43
Figure 3.1 (a) UV/Visible Absorbance and (b) photoluminescence (PL) spectra of CdSe quantum dots.
visible at 550 and 510 nm respectively. The PL spectrum showed emission at
635 nm.
The substrates used for the fabrication of the devices, were glass slides
precoated with indium tin oxide (ITO), obtained from Delta Technologies
Limited. ITO is a transparent conducting oxide commonly used as an elec-
trode in light emitting devices. In our device, it acts as an anode, injecting
holes into the device. Before use the substrates were cut to squares of about
half an inch, then cleaned by sequentially ultrasonicating in water, acetone,
methanol and iso-propanol with a N2 blow dry step after each solvent. After
3.2 Experimental Method CHARGE TRANSPORT IN HYBRID QD-LEDS
44
Figure 3.2 Device structures for (a) Device 1, (b) Device 2 and (c) Device 3.
the cleaning, the ITO coated glass slides were exposed to UV-ozone for 10
minutes to remove any residual carbon contamination from the solvents.
Meanwhile, a solution of N,N’-Bis(3-methylphenyl)-N,N’-diphenyl-
benzidine (TPD) (obtained from Sigma-Aldrich Co.) was prepared in chlo-
roform. The suspension of CdSe/ZnS quantum dots in chloroform was added
to this solution such that the final concentration of dots was 50 mg/ml and
that of TPD was 10 mg/ml. A 50 mg/ml solution of quantum dots in chlo-
roform with no added TPD was also prepared. PEDT:PSS
(Poly(3,4-ethylenedioxythiophene) poly(styrenesulfonate)) which is available
in solution form (with water as a solvent) from H.C. Starck GmbH under the
commercial name CLEVIOS™ P VP CH 8000, acts as a hole transport layer.
It was filtered just before deposition using a syringe filter (pore size: 0.2 µm)
to remove any particles formed during transport and storage.
Three sets of devices were prepared, one without the PEDT:PSS layer
(device 1, Figure 3.2(a)) and two with the PEDT:PSS layer. The two devices
with the PEDT:PSS layer had different emission layers, one with a matrix of
3.2 Experimental Method CHARGE TRANSPORT IN HYBRID QD-LEDS
45
TPD and quantum dots (device 2, Figure 3.2(b)) and one with just quantum
dots (device 3, Figure 3.2(c)). The procedure for three sets of devices was the
same except for the deposition of the PEDT:PSS layer. To avoid degradation
due to exposure to air and moisture, all of the processing was done inside a
nitrogen-filled glove box. A 30 nm thick PEDT:PSS layer was deposited by
spin coating at 3000 rpm. It was then baked on a hot plate at 120°C for about
1 hour to remove any residual solvent. This bake also hardens the film so that
the next layer to be deposited does not affect the morphology of the film. An
advantage of using PEDT:PSS layer as the hole transport layer is that it is
soluble in water, but not chloroform. The deposition of the next layer from a
chloroform solution does not affect the PEDT:PSS film.
For device 2, a 40 nm thick film of TPD matrix embedded with quantum
dots (emission matrix) was deposited from the solution by spin coating on top
of the PEDT:PSS film. For device 3, a film of quantum dots was deposited by
spin coating a suspension of quantum dots. In both cases this was followed by
baking on hot plate at 120°C for about 45 minutes to remove residual solvents.
This was followed by the deposition of a 40 nm thick bathocuproine (BCP)
film by thermal evaporation of BCP powder (Sigma-Aldrich Co.) BCP acts as
the electron transport layer. A stencil mask with one mm diameter holes was
placed on top of the substrate followed by the deposition of 400 nm thick Al
cathodes by thermal evaporation of Al pellets (Sigma-Aldrich Co.) The
thickness of the spin coated films was measured using Variable Angle Spec-
3.2 Experimental Method
Figure 3.3 (a) Height and (b) phase image of spin coated mixture of QDs and TPD.
troscopic Ellipsometer (VASE, J.A. Woollam Co.). The thermally evaporated
film thickness was measured in real time using a quartz cry
power to the deposition boats was controlled such that the deposition rate was
constant to get high quality films. The devices were taken out of the glove box
just before testing to minimize exposure to air and moisture.
Testing of the devices was done using an Agilent 4156A parameter an
lyzer attached to two probes: a hard probe (tungsten) and a
gold wire). The hard probe is used to contact the anode and so pierces the
deposited films, making contact with the ITO. The soft probe is chosen to
make contact with the cathode so that it does not pierce the thin Al layer. As
a constant electric current is supplied to the device using the parameter a
alyzer, the light spectrum is measured using USB2000 spectrometer (Ocean
Experimental Method CHARGE TRANSPORT IN HYBRID QD
46
(a) Height and (b) phase image of spin coated mixture of
troscopic Ellipsometer (VASE, J.A. Woollam Co.). The thermally evaporated
film thickness was measured in real time using a quartz crystal monitor. The
power to the deposition boats was controlled such that the deposition rate was
constant to get high quality films. The devices were taken out of the glove box
just before testing to minimize exposure to air and moisture.
Testing of the devices was done using an Agilent 4156A parameter an
lyzer attached to two probes: a hard probe (tungsten) and a soft probe (thin
gold wire). The hard probe is used to contact the anode and so pierces the
deposited films, making contact with the ITO. The soft probe is chosen to
make contact with the cathode so that it does not pierce the thin Al layer. As
electric current is supplied to the device using the parameter a
alyzer, the light spectrum is measured using USB2000 spectrometer (Ocean
HYBRID QD-LEDS
(a) Height and (b) phase image of spin coated mixture of
troscopic Ellipsometer (VASE, J.A. Woollam Co.). The thermally evaporated
stal monitor. The
power to the deposition boats was controlled such that the deposition rate was
constant to get high quality films. The devices were taken out of the glove box
Testing of the devices was done using an Agilent 4156A parameter ana-
soft probe (thin
gold wire). The hard probe is used to contact the anode and so pierces the
deposited films, making contact with the ITO. The soft probe is chosen to
make contact with the cathode so that it does not pierce the thin Al layer. As
electric current is supplied to the device using the parameter an-
alyzer, the light spectrum is measured using USB2000 spectrometer (Ocean
3.3 Results and Discussion CHARGE TRANSPORT IN HYBRID QD-LEDS
47
Figure 3.4 Current density - Voltage (J-V) characteristics of device 1. It is also representative of the J-V characteristics for devices 2 and 3.
Optics Inc.). The spectrometer is attached using a fiber optic cable, with a
core diameter of 600 µm, to the back of the device, so as to collect the light
from the transparent side of the device. Care was taken in placement so as to
cover the end of the optical fiber completely by the device that was being
tested.
3.3. Results and Discussion
Figure 3.3 shows the height and phase atomic force microcopy (AFM)
images of the spin coated mixture of TPD and QDs. Both height and phase
images show a uniform pattern. In a similar process, Coe [46] observed ex-
3.3 Results and Discussion CHARGE TRANSPORT IN HYBRID QD-LEDS
48
Figure 3.5 EL intensity curves at different current densities for device 1. The peak at 460 nm corresponds to EL from TPD and the peak at 610 nm corresponds to EL from CdSe/ZnS quantum dots. Inset shows the QD to TPD emission ratio, which is almost constant at 0.2
tremely smooth (rms roughness 0.57 nm by atomic force microscopy) regions
and extremely rough (no value given, but obviously very rough) regions and
ascribe the behavior to phase segregation. Our films appear to be uniform,
with no observable rough regions. The rms roughness of 300 nm by 300 nm
areas varied from 1.6 to 2.1 nm. We take this value to be consistent with in-
dividual nanoparticles rather than agglomerates. Although the reason for this
difference is uncertain, we note that our suspensions were far more concen-
trated than those of Coe, 50% by volume rather than 10%. Thus, our analysis
3.3 Results and Discussion CHARGE TRANSPORT IN HYBRID QD-LEDS
49
Figure 3.6 EL intensity curves at different current densities for device 2. Inset shows the QD to TPD emission ratio, which is almost constant at 1.0.
assumes a uniform mixture of QDs and TPD in the QD+TPD layers shown in
Figure 3.2.
Figure 3.4 shows the J-V characteristics of device 1 which is representa-
tive of J-V characteristics of other devices as well. The semilog plot in Figure
3.4 shows the diode like characteristics of our device. As we increase the ap-
plied voltage, the series resistance of the device plays a more significant role
and thus the slope of the curve constantly decreases with increasing voltage.
3.3 Results and Discussion CHARGE TRANSPORT IN HYBRID QD-LEDS
50
Figure 3.7 EL intensity curves at different current densities for device 3.
Figure 3.5 shows the EL spectrum of the device without the PEDT:PSS
layer (device 1) for the three different current densities (10 mA/cm2, 50
mA/cm2 and 100 mA/cm2). There are two peaks in the EL spectrum. The
peak at 460 nm corresponds to EL from TPD. [51] The peak at 610 nm cor-
responds to EL from CdSe/ZnS quantum dots. When compared to the PL
spectrum, the peak is shifted by about 25 nm. This small shift may be due to
different spectrometers being used for measuring solution based PL and solid
state EL, or it could be due to some minor oxidation of the QDs when testing
3.3 Results and Discussion CHARGE TRANSPORT IN HYBRID QD-LEDS
51
Figure 3.8 Band diagram for device 2. The band diagrams for other two devices can be derived from this, either by removing PEDT:PSS (device 1) or by removing TPD (device 3).
them in devices. The EL intensity from TPD is much stronger than that of
quantum dots. The inset shows that the ratio of QD emission to TPD
emission is almost constant (~0.28). Figure 3.6 shows the EL spectrum of
device with the PEDT:PSS layer (device 2) for three different current densi-
ties. Here again two peaks are observed, but this time the intensity of EL from
quantum dots is comparable to that of TPD. The inset showing the QD/TPD
emission ratio again shows that the emission ratio (~1.0) is not dependent on
3.3 Results and Discussion CHARGE TRANSPORT IN HYBRID QD-LEDS
52
current density. Finally, Figure 3.7 shows the EL spectra for three current
densities for the device with just the quantum dot layer and no TPD (device
3). As expected, there is only one EL emission peak observed at 610 nm,
which corresponds to emission from the quantum dots. The intensity of
emission increases with increasing current density as one would expect.
Figure 3.8 shows the band diagram for device 2. The band diagrams of the
other two devices can be deduced from Figure 3.8. The carrier transport
mechanism can be most easily understood for device 3. Electrons are injected
by the Al cathode and transported via the BCP layer to the CdSe quantum
dots. Holes are injected by the ITO anode and transported via the PEDT:PSS
layer to the quantum dots. Electrons and holes recombine on the quantum
dots resulting in the emission spectrum observed in Figure 3.7. There is no
possibility of Forster energy transfer from PEDT:PSS layer into the CdSe
quantum dots because the bandgap of the quantum dots is higher than that of
PEDT:PSS. Thus, the only mechanism of charge transfer into the quantum
dots in device 3 is direct charge injection. Figure 3.9 shows a comparison of
EL from devices 1, 2 & 3 and the sum of EL from devices 1 & 3 at 100
mA/cm2. In devices 1 and 2, there is a possibility of both direct charge in-
jection and Forster energy transfer from TPD into the quantum dots. Looking
at TPD emission first, we find that the addition of the PEDT:PSS hole
transport layer appears to decrease the emission from TPD. This is may be
due to competition between the emission mechanisms. Indeed, one can clearly
3.3 Results and Discussion CHARGE TRANSPORT IN HYBRID QD-LEDS
53
Figure 3.9 Comparison of EL intensities from devices 1, 2 & 3 with the sum of EL intensities of device 1 & 3 at 100 mA/cm2 showing that, EL from QDs for device 2 and sum of device 1 and 3 overlap. Inset shows that the same is true at 10 and 50 mA/cm2 as well.
see that the QD emission for device 2, where both carrier types are confined
near the QDs, is much stronger than it is for devices without the confining
layers.
An unexpected result however, is that the EL emission from the QDs for
device 2, which has both TPD and the confining layers, is exactly the same as
3.3 Results and Discussion CHARGE TRANSPORT IN HYBRID QD-LEDS
54
the EL of the device with only TPD plus the EL of the device with only the
confining layer. This is not a numerical fluke. It is true for all values of current
density that were measured (see inset). Thus, it can be concluded that the two
charge transfer mechanisms in device 1 and 3 act in parallel to give the light
output in device 2. Thus, the EL from quantum dots in device 1 is completely
due to Forster energy transfer from the TPD layer. The amount of light
output by the QDs in device 1 is limited by the efficiency of Forster energy
transfer. This depends on various factors including the materials used and the
device configuration as demonstrated by Jing et al. [52] Some of the excitons
in TPD are transferred onto QDs by the Forster process, some radiatively
recombine, and the remaining non-radiatively recombine. In device 2, the
amount of light output by QDs is increased as we are no longer limited by the
Forster energy transfer efficiency. Here, direct charge injection provides car-
riers that can radiatively recombine. Hence, instead of competing, these two
mechanisms, direct charge injection and Forster energy transfer, act inde-
pendently to maximize the light output from the QDs. [47,48]
The above conclusions are also supported by the band diagram drawn in
Figure 3.8. For device 1, the barrier for hole transfer from ITO to CdSe is 1.8
eV whereas the barrier for hole transfer from ITO to TPD is 0.7 eV. Thus,
most of the holes are first transferred into TPD and then, upon formation of
excitons, are transferred into the quantum dots via the Forster mechanism. In
device 2, there is an intermediate PEDT:PSS layer about 30 nm thick, so the
3.3 Results and Discussion CHARGE TRANSPORT IN HYBRID QD-LEDS
55
only place for the holes to go is into this layer even though the barrier is 1.7
eV. Once holes are in this layer, they can either go into the CdSe quantum
dots by overcoming a small (0.1 eV) barrier or into TPD by losing energy. The
holes that travel directly into the quantum dots recombine there and account
for the direct charge injection component of the EL intensity, whereas, the
holes traveling into the TPD layer account for the Forster energy transfer
component of the EL intensity.
4.1 Introduction MODELING OF QD-LEDS
56
Chapter 4
Modeling of QD-LEDs
This chapter has been reprinted with the permission from Kumar,
Brijesh, et al. "Modeling Charge Transport in Quantum Dot Light Emitting
Devices with NiO and ZnO transport layers and Si Quantum Dots" Journal of
Applied Physics (2013): Copyright 2013, AIP Publishing LLC.
4.1. Introduction
Colloidal quantum dots (QDs) are well suited for display technologies
because of their tunable band gap and hence wavelength tunable luminescence
properties. [32,33,44,53] Optimization and other development efforts
potentially can be minimized for different colored light emitting diodes
(LEDs) because essentially the same structure may be used to create LEDs of
various colors. If appropriately chosen, electron and hole transport layers
and their contacts need not be varied. The only structural feature that
determines the emission wavelength is the size of the QDs. The procedure to
prepare the QDs may stay the same (although the conditions need to be
varied to create different sizes), the procedure for deposition of the QDs onto
the device could stay the same, the electrode deposition could stay the same,
4.1 Introduction MODELING OF QD-LEDS
57
and the transport layer deposition could stay the same. Hence, all the design
and optimization required for each of these individual device components can
be useful for LEDs of different colors.
Quantum dot light emitting devices (QD-LEDs) were first presented in
the form of hybrid organic/inorganic light emitting devices by Colvin et al.
[28] and were later improved upon by Coe-Sullivan et al.[54] For the hybrid
devices, the transport layers were fabricated from organic semiconductors and
the QDs from inorganic semiconductors. However, organic transport layers
suffer from sensitivity to air exposure and stability problems. Hence,
inorganic semiconductor transport layers, NiO and ZnO, were developed and
used in a device by Caruge et al. [42] This work deals with modeling of
QD-LEDs with ZnO and NiO transport layers and Si QDs. It can readily
be extended to other inorganic transport layers and QDs. The model can also
be used for hybrid QD-LEDs by modifying the transport model to allow for
the field and density dependence of the carrier mobility that usually need to
be considered for organic semiconductors.
QDs have been a subject of intense study ever since their discovery by
Ekimov et al. in 1981.[55] Most of the studies concerning transport in
quantum dots are related to microscopic single electron transport at
extremely low temperatures.[56-59] Relatively few studies deal with charge
transport at room temperature, which is the range of primary interest for
applications like light emitting or memory devices.[60,61] In this work, a
4.2 Model Description MODELING OF QD-LEDS
58
complete model for modeling charge transport in QD-LEDs is proposed. It
incorporates charge transport in the bulk, followed by charge transfer into the
quantum dot layers and finally, the transfer amongst the various QD-layers
and recombination in the QDs.
4.2. Model Description
The schematic device structure for a QD-LED is show in Figure 4.1.
Figure 4.2 shows the corresponding band diagram. The device essentially
consists of a number of quantum dot layers sandwiched between two thin
transport layers, an electron transport layer (ETL) and a hole transport layer
(HTL). The HTL is contacted by an indium tin oxide (ITO), transparent
conducting layer, which acts as a hole injector, whereas, the ETL is connected
to an Aluminium cathode, which acts as an electron injector. The device
functions as follows. Electrons are injected from the cathode and holes from
the anode and they travel through their respective transport layers to the
QDs, where they recombine. If the recombination is radiative, photons are
emitted. The photon wavelength depends on the QDs and is completely
independent of the transport layers.
The device model can be divided into the following parts:
A. Carrier injection from transport layers into the quantum dot layers
B. Transport among the quantum dot layers
C. Recombination in the quantum dots
4.2 Model Description MODELING OF QD-LEDS
59
D. Coupled rate equations
E. Transport in the ETL and HTL
F. Carrier injection from the contacts
Each of the different components of the model are described in the
following subsections.
4.2.1 A. Carrier injection from transport layers into quantum dot
layers
Focusing on the QD layer adjacent to the HTL, the quantum dots closest
to the HTL are considered to be traps for holes. In equilibrium, using
detailed balance, the rate of capture of holes by these QDs is the same as the
rate of emission,
(4.1a)
where, NT is the density of trap states in the quantum dot layer (equivalent to
the density of quantum dots since each QD can only accommodate one hole),
p(0) is the hole density in the quantum dot layer adjacent to the HTL, p(0-) is
the hole density in the HTL very close to the quantum dot layer, ẽ is the
emission rate coefficient for the trap states, and, c is the capture rate
coefficient for the trap states. Whenever a subscript (0) is used it stands for
equilibrium
The ratio of emission and capture coefficients is,
)0()0(
)0(~
~
00
0
0
0−−
=ppN
p
e
c
T (4.1b)
)0()0(~)0(~00000 pNpcpe T −= −
4.2 Model Description MODELING OF QD-LEDS
60
Under non-equilibrium conditions, the rate of change of carrier
concentration is given by the difference of capture and emission, which can be
described as:
)0(~)0()0(~ pepNpct
pT −−=
∂∂ −
(4.1c)
If the device is not too far out of equilibrium, the capture and emission
coefficients can be taken to be the same as for equilibrium. Furthermore, if
the carrier concentration is non-degenerate, eqn. 4.1c can be simplified to,
−=∂∂
−
−
)0()0()0(
)0(10
0
ppp
p
t
p
plτ (4.1d)
where τpl (= 0~/1 e ) is capture/emission time constant, a parameter which
depends on the materials used. Equations for electrons at the interface
between the ETL and the adjacent QD layer are entirely analogous.
4.2.2 Transport among the quantum dot layers
The quantum dots can be considered as semiconductor particles with an
insulating layer surrounding them.[62] The insulating layer may consist of
silicon dioxide[63] or organic ligands,[64] depending on the method used to
prepare the quantum dots. Figure 4.3 shows the model potential energy
diagram for the quantum dots. The transport from one QD to another is
viewed as a direct tunneling process, where the QDs act as potential wells and
the insulating layers as tunnel barriers.
4.2 Model Description MODELING OF QD-LEDS
61
Using a one-dimensional WKB approximation, the tunneling probability
from 1→2 can be written as,
( ) insdT 22exp12 κ−= , (4.2)
where kappa is the inverse characteristic length for tunneling, and, dins is the
thickness of the insulating layer around the QDs.
The electron is ‘oscillating’ in the well with ‘frequency,’ Sith dv 2/=ν ,
where, vth is the thermal velocity of electrons, dSi is the diameter of the QD,
and it has probability T12 of making a transition to the neighboring QD layer.
The total density of electrons per second tunneling from 1→ 2, assuming an
unoccupied layer 2 is,
12121 TnN ν=→ , (4.3a)
where n1 is the number of electrons in layer 1.
The ‘oscillation frequency’ is the same for both the layers because the
particle size and temperature is the same in layers 1 and 2. Thus, the total
number of electrons per second tunneling from 2→1, assuming an unoccupied
layer 1, can be written as,
21212 TnN ν=→ , (4.3b)
Consequently, the net flow of electrons from 1→2 is,
)( 2211121221 nTnTNN −=− →→ ν (4.3c)
For elastic tunneling T12 = T21, and again by detailed balance we may write
−=− →→ 2
2
11121221 n
n
nnwNN
o
o
(4.3d)
4.2 Model Description MODELING OF QD-LEDS
62
where, )2(2exp
21212 oxSi
th dd
vTw κν −==
and n1o and n2o are equilibrium
concentrations of electrons in layers 1 and 2, respectively. Similar equations
can be written for transport of holes in the QDs.
4.2.3 Recombination in the quantum dots
There are two types of recombination in quantum dots, radiative and
non-radiative. Radiative recombination can be either monomolecular or
bimolecular depending upon the doping of quantum dots. If the quantum
dots are doped, then the recombination rate is dependent only on the minority
carrier density, hence it is monomolecular. But if the quantum dots are
undoped, or if both non-equilirium carrier densities are comparable to the
equilibrium carrier density, then the recombination rate depends on both
kinds of carriers, i.e., it is bimolecular. In the rest of this work, we assume
that the quantum dots are undoped and hence the recombination is
bimolecular, as also shown by Kočka et al.[65]
Bimolecular recombination rate can be defined as follows:
Ur = γ(np – ni2), (4.4a)
where γ is recombination rate coefficient, n and p are electron and hole
concentrations, respectively, and ni is intrinsic carrier concentration.
Non-radiative recombination can be modeled as Shockley-Read-Hall
(SRH) recombination,[66,67] .
4.2 Model Description MODELING OF QD-LEDS
63
( ) ( )11
2
nnnp
npnU
pn
inr +++
−=ττ (4.4b)
Assuming, the electron and hole recombination lifetimes to be equal, τp =τn
and p,n >> n1 , the total non-radiative recombination rate can be written as,
( )np
npnU
nr
inr +
−=τ
2
(4.4c)
Both, radiative and non-radiative recombination take place in parallel in the
quantum dots, hence, the total recombination rate can be written as,
( ) ( )np
npnnpnUUU
nr
iinrr +
−+−=+=τ
γ2
2
(4.4d)
4.2.4 Coupled rate equations
The discretization scheme for the quantum dot layers is shown in Figure
4.2. From what has been described in the above subsections, the transport
equations for holes can be written in terms of coupled rate equations:
))1()1((
)1()1()1()1()2(
)2(
)1()1()1()1(
)0(
)0(1)1( 212
110
0120
01,0 np
nnpnpnp
p
ppwpp
p
p
dt
dp
nr
ii
p
p +−−−−
−−
−=τ
γτ
(4.5a)
))2()2((
)2()2()2()2()3(
)3(
)2()2()2(
)2(
)1()1(
)2( 222
220
023
0
012 np
nnpnpnp
p
ppwp
p
ppw
dt
dp
nr
ii
pp
+−−−−
−−
−=τ
γ
(4.5b)
…..
4.2 Model Description MODELING OF QD-LEDS
64
))()((
)()()()(
)1()1(
)()()(
)(
)1()1(
)(
22
0
01,
0
0,1
NnNp
nNnNpnNpNn
NpNp
NpNpwNp
Np
NpNpw
dt
Ndp
nr
iNiNN
pNN
pNN
+−−−−
++
−−
−−−= +−
τγ
(4.5c)
…..
))1()1((
)1()1()()(
)()1(
)1()(
1)(
)(
)1()1(
)(
212
001,0
0,1
np
nnpnLpLn
LpLp
LpLpLp
Lp
LpLpw
dt
Ldp
nr
iiLL
LpL
pLL
+−−−−
++−−
−−−=
+−
τγ
τ
(4.5d)
The transport equations for electrons can be written in analogous fashion. It
is important to note that each of the above equations contain a product of
electron and hole densities, thus, we need to employ a solution based on
non-linear system of equations.
4.2.5 Transport in ETL and HTL
Bulk drift and diffusion is used to describe the transport in the two
transport layers. It is represented by the following set of equations:
UGx
J
qt
p p −+∂
∂−=
∂∂ 1
(4.6a)
UGx
J
qt
n n −+∂∂=
∂∂ 1
(4.6b)
( )npNNq
x ad −+−−=∂∂
εψ
2
2
(4.6c)
4.2 Model Description MODELING OF QD-LEDS
65
where,
xqp
x
pqDJ ppp ∂
∂−∂∂−= ψµ
(4.6d)
xqn
x
nqDJ nnn ∂
∂−∂∂= ψµ
(4.6e)
Generation is assumed to be negligible and thus neglected from our
calculations (G=0). Recombination is assumed to be of Shockley-Read-Hall
form,[66,67]
( ) ( )ipin
i
nnnp
npnU
+++−=
ττ
2
(4.7)
4.2.6 Carrier injection from the contacts
The contacts at both the anode and cathode are assumed to be ohmic.
Thus, the carrier concentrations inside the transport layers, very close to the
contacts (positions 0 & w), are equal to the equilibrium carrier
concentrations.
( ) ( ) ( ) ( ) 22 )()( ;000 wnwnwpnnp ii == (4.8a)
There is no space charge at electrodes,
( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) 0 ;00000 =−+−=−+− wnwpwNwNnpNN adad (4.8b)
The electrostatic potential at the electrodes is thus given by:
( ) ( ) ( ) ( )
∆+∆−
=
−=
2)0(ln ;
)0(
0ln0 VC
ii
EE
n
wn
q
kTw
n
p
q
kTV ψψ
(4.8c)
4.3 Results and Discussion MODELING OF QD-LEDS
66
The last term in Eqn. 4.8c accounts for the discontinuities in the
conduction (∆EC) and valence (∆EV) band, which are estimated from
Anderson's rule[68] as described in more detail by Smith.[69] The
conduction band and valence band edges of quantum dots do not affect the
potential boundary conditions because the electric field in the device is
controlled by the difference in the Fermi levels of the transport layers, just like
a p-i-n diode where the electric field in the i-region is controlled by the
difference in the Fermi levels of p and n-regions.
4.3. Results and Discussion
4.3.1 Device Parameters
To illustrate the above described model, the example device structure
shown in Figure 4.4 is used. A 47 nm thick layer of NiO, which is naturally a
p-type, is used as the HTL, while a 50 nm thick ZnO layer, a naturally n-type
material, provides the ETL. The emission layer consists of 5 layers of silicon
QDs, each 5 nm in diameter including a 0.5 nm thick SiO2 shell acting as the
passivation layer. Thus the total device thickness is 122 nm. The
parameters for the transport layers are presented in Table 4.1. The majority
carrier mobilities are taken from the literature, but, the minority carrier
mobilities are not easily available in the literature, so, the minority carrier
mobilities are assumed to the same as majority carrier mobilities. As
demonstrated in the section below, the minority carrier mobility does not
4.3 Results and Discussion MODELING OF QD-LEDS
67
matter because the current is dominated by majority carriers. Using the
method described by Smith[42,69] and the electron affinity and band gap
values given by Caruge et al.,[42] the discontinuities in the bands are
calculated to be ∆EC = 2.5 eV and ∆EV = 2.2 eV. The carrier lifetimes were
not available for NiO, so they were assumed to be the same as for ZnO as
reported by Liu et al.[70]
Table 4.1 The parameters of NiO and ZnO used for our model device.
NiO ZnO
Hole mobility, µp[71] 0.2 cm2/V-s 8.6 cm2/V-s
Electron Mobility, µn[71,72] 0.2 cm2/V-s 8.6 cm2/V-s
Doping concentration p-type, NA=1017 cm-3 n-type ND=1017 cm-3
Bandgap, Eg[42] 3.7 eV 3.4 eV
Intrinsic carrier concentration, ni 6.106 × 10-13 cm-3 2.0 × 10-10 cm-3
Electron affinity, χ[42] 1.8 eV 4.3 eV
Carrier Lifetime, τp=τn [70] 1.2 µs 1.2 µs
Table 4.2 shows the important parameters relating to the transport in the
QD layers. The intrinsic carrier concentration, ni, is based on the band gap
provided by Ligman et al.[73] The QDs are assumed to be undoped. The
tunneling coefficient is calculated using WKB approximation assuming a 0.5
nm thick oxide shell on the Si QDs. The capture lifetime of an electron/hole
by a QD is not easily available in literature. So, the capture lifetimes are
varied from 1 ms to 1 ns and the effect of such a change is studied on the
device. This is discussed in further detail in the following section. The
4.3 Results and Discussion MODELING OF QD-LEDS
68
non-radiative SRH recombination lifetime is assumed to be ~1 µs as observed
by Sobolev et al. [74]
Table 4.2 The parameters used to describe transport in QD layers.
Intrinsic carrier concentration, ni 1.42 × 104 cm-3
Tunneling coefficient, w 1.25 × 1010 s-1
Capture lifetime, τ 1 × 10-3 s – 1 × 10-9 s
Bimolecular Recombination Coefficient, γ[65] 4 × 10-10 cm3 s-1
Non-radiative recombination lifetime, τnr [74] 1 × 10-6 s
Band gap, Eg[73] 1.8 eV
Electron affinity, χ[73] 3.5 eV
4.3.2 Simulation Results
Figure 4.4 shows the electron and hole carrier densities and electrostatic
potential at equilibrium and for an applied potential of 0.5 V. The built-in
potential for the device is 0.98 V as can be seen from Figure 4.4. Upon
applying the 0.5 V forward bias, the electrostatic potential across the device
goes down to 0.48 V as expected. At equilibrium, the carrier densities
decrease as we move away from the contacts and deeper into the device.
Because of the high band gap of the ETL and HTL, only the majority carriers
play any significant role in the overall current of the device As can be seen
from the Figure 4.4, minority carrier densities in both ETL and HTL are more
than 20 orders of magnitude lower than the majority carrier densities. In the
QD layers, both electron and hole carrier densities are at comparable levels
because of the much smaller band gap. When a 0.5 V forward bias is
applied, the majority carrier densities in the transport layers increase, both
4.3 Results and Discussion MODELING OF QD-LEDS
69
electron and hole densities in the QD layers increase, and as a result, a current
of 9.47 µA/cm2 flows through the device.
The total current density is plotted in Figure 4.5 as semilog and linear
plots. At low voltages, the current is limited by diffusion and thus, the current
density increases exponentially with increasing forward bias. At high
voltages, the current is limited by the series resistance of the device and hence
the current increases linearly with forward bias. A semilog plot of the
radiative recombination current is also shown in Figure 4.5 to compare with
the total current. At low voltages, the radiative recombination current is much
smaller than the total current, because the SRH recombination dominates
over the bimolecular radiative recombination. When a higher forward bias is
applied, the carrier densities in the QD increase, thus, the total bimolecular
radiative recombination increases (Eqn. 4.4a). The total SRH recombination
(Eqn. 4.4b) in QDs also increases, but it increases at a lower rate than
bimolecular radiative recombination because p and n terms appear both in the
numerator and denominator, whereas in bimolecular radiative recombination,
they appear only in the numerator. Because of the different rates of increase
in the two recombination mechanisms, at around 1 V forward bias, SRH
recombination in QDs becomes negligible compared to bimolecular radiative
recombination
As described in section III.A, it is difficult to estimate the capture
lifetime, τ, of carriers from the bulk to the quantum dots, so, the capture
4.3 Results and Discussion MODELING OF QD-LEDS
70
lifetime is varied from 1 ms to 1 ns, to assess the effect it has on the device
characteristics. Figure 4.6 shows the current density vs applied voltage plot
for the device with varying carrier capture lifetimes, keeping all others
parameters constant. The plots at τ = 1 ns and τ = 1 µs completely overlap
each other, and the current density only reduces slightly if τ = 1 ms. Hence,
it can be concluded that over a reasonably large range of values the carrier
capture lifetime has negligible effect on the current-voltage characteristics of
the device.
We have assumed a radiative recombination coefficient of γ = 4 × 10-10
cm3 s-1. Depending on the fabrication conditions, this coefficient may vary..[75]
Thus, we next explore the effect of changing the value of γ. Keeping all the
other parameters constant, the effect of the changing γ from 4 × 10-12 cm3 s-1
to 4 × 10-8 cm3 s-1 on current density and ratio of recombination current to
total current is studied and is plotted in Figure 4.7 (a) and (b) respectively.
At high forward bias, the total current density increases upon increasing the
recombination coefficient, which again indicates that the total current is
limited by the radiative recombination current at high forward bias. At low
voltages, the total current is limited by the non-radiative recombination
current, so the total current density stays nearly constant at voltages below
0.5 V for γ between 4 × 10-12 cm3 s-1 and 4 × 10-10 cm3 s-1. Both these
observations are consistent with the our conclusions from Figure 4.5 that at
low voltages, the current is primarily due to non-radiative recombination and
4.3 Results and Discussion MODELING OF QD-LEDS
71
at high voltages, the current is primarily a result of radiative recombination.
As observed in Figure 4.7(b), the ratio of recombination current to total
current increases upon increasing the applied voltage. With increasing γ, the
transition point of no recombination current to a positive value of
recombination current (the turn on point of the LED) shifts to the left. For
γ between 4 × 10-12 cm3 s-1 and 4 × 10-10 cm3 s-1, the turn-on point is beyond or
around 0.5 V forward bias. But, for γ =4 × 10-8 cm3 s-1, the turn-on point is at
almost 0 V. Thus, the current density plot corresponding to for γ =4 × 10-8
cm3 s-1 is higher than the other values of γ even at smaller voltages, as the
recombination current is not negligible. Thus, our model can be used to
explore the turn on-voltage of QD-LEDs.
4.3 Results and Discussion MODELING OF QD-LEDS
72
Figure 4.1 Schematic device structure for a Quantum Dot Light Emitting Diode.
4.3 Results and Discussion MODELING OF QD-LEDS
73
Figure 4.2 Flat band diagram for a sample device which consists of Si quantum dots, ZnO as the electron transport layer, NiO as the hole transport layer and ohmic contacts to both anode and cathode. The horizontal lines at ends signify the metal Fermi energies for anode and cathode. Energies are in eV with respect to the vacuum level.
1.8
NiO
3.5
Si QDs
4.3
ZnO
Anode
Cathode
4.3 Results and Discussion
Figure 4.3 Discretization scheme for QDlayers.
Results and Discussion MODELING OF QD
74
Discretization scheme for QD-LED focusing on the QD
MODELING OF QD-LEDS
LED focusing on the QD
4.3 Results and Discussion MODELING OF QD-LEDS
75
0 20 40 60 80 100 120 140100
102
104
106
108
1010
1012
1014
1016
1018
Car
rier
Den
sity
(cm
-3)
x(nm)
p (0.5 V) n (0.5 V) p (0.0 V) n (0.0 V)
-1.8
-1.6
-1.4
-1.2
-1.0
-0.8
-0.6
-0.4
-0.2
V (0.5 V) V (0.0 V)
Pot
entia
l (V
)
Figure 4.4 Carrier densities and electrostatic potential in a device with 5 QD layers (represented by dots on the plot) at equilibrium and with an applied voltage of 0.5 V.
4.3 Results and Discussion MODELING OF QD-LEDS
76
0 1 2 3 4 5 6 7
10-10
10-8
10-6
10-4
10-2
100
102
C
urre
nt D
ensi
ty (
A/c
m2 )
J(log) J
rec(log)
Forward Bias (V)
0
200
400
600
800
1000
1200
J(linear)
Cur
rent
Den
sity
(A
/cm
2 )
Figure 4.5 Total device current density (log and linear) plots with increasing forward bias. The recombination current density is also plotted.
4.3 Results and Discussion MODELING OF QD-LEDS
77
0.0 0.5 1.0 1.5 2.010-8
10-7
10-6
10-5
10-4
10-3
10-2
10-1
100
101
102
103
Applied voltage (V)
Cur
rent
den
sity
(A
/cm
2 )
τ=10-3 s τ=10-6 s τ=10-9 s
Figure 4.6 Current density vs applied voltage for different assumed values of the carrier capture coefficients (τ).
4.3 Results and Discussion MODELING OF QD-LEDS
78
0.0 0.5 1.0 1.5 2.0
10-7
10-6
10-5
10-4
10-3
10-2
10-1
100
101
102
103
Cur
rent
Den
sity
(A
/cm
2 )
Voltage applied (V)
γ=4x10-12 cm3s-1
γ=4x10-11 cm3s-1
γ=4x10-10 cm3s-1
γ=4x10-8 cm3s-1
0.0 0.5 1.0 1.5 2.0
0.0
0.2
0.4
0.6
0.8
1.0
J rec/
J
Voltage applied (V)
γ=4x10-12 cm3s-1
γ=4x10-11 cm3s-1
γ=4x10-10 cm3s-1
γ=4x10-8 cm3s-1
Figure 4.7 (a) Current density vs applied voltage and (b) ratio of recombination current to total current (Jrec/J) for various radiative recombination coefficients (γ).
5.1 Conclusion CONCLUSION AND SCOPE FOR FUTURE WORK
79
Chapter 5
Conclusion and Scope for
Future work
5.1. Conclusion
Two charge transfer mechanisms in hybrid organic/inorganic QD-LEDs,
direct charge injection and Forster energy transfer were investigated. With the
help of AFM, it was shown that in our devices phase separation does not take
place. By using different device configurations, it was shown that direct
charge injection and Forster energy transfer act independently to produce the
light output from QDs in a hybrid QD LED where the QDs do not get phase
separated from the organic matrix. So, based on our findings and that of
Anikeeva et al. [48] we can conclude that depending on the distribution of the
QDs in the organic matrix, it is possible to have one or both of direct charge
transfer and Forster energy transfer mechanisms play a significant role in the
light output of QD LEDs.
5.2 Scope for Future Work CONCLUSION AND SCOPE FOR FUTURE WORK
80
We also proposed a model for inorganic QD-LEDs. Our model is able to
incorporate the most important parameters that affect the current density
and light output of such devices. The device current is mostly limited by the
QD layers, where the current is primarily due to recombination of two kinds,
radiative and non-radiative. At low voltages, radiative recombination
dominates the device operation, whereas as the voltage increases, radiative
recombination dominates. As the voltage is increased even further, the
current is limited by the series resistance of the device and the current voltage
plot becomes essentially linear. There is practically no effect of capture
lifetime from the transport layers to the QD layers because the device
operation is limited by the QD layers, not how the carriers got into the QDs.
Our model is also able to explore the turn-on voltage of QD-LEDs based on
the ratio of radiative recombination current to the total device current.
5.2. Scope for Future Work
There is great potential for future work on modeling of QD-LEDs.
QD-LEDs with various device configurations have been fabricated, but their
functioning is not fully understood. The model proposed by us can be
extended to model organic transport layer based QD-LEDs, by using field
dependent mobility to model the transport layers.
The model for charge transport in quantum dots is a first-order model,
where quantum dots are simplified to multiple layers of uniform sheets of
5.2 Scope for Future Work CONCLUSION AND SCOPE FOR FUTURE WORK
81
quantum dots. Thus, only one dimensional charge transport is considered by
us. A more complex model can be developed where charge transport from
individual quantum dots is considered in all three dimensions.
We only considered charge transport in QD-LEDs but the light output
was not modelled. In a future work, the model can be extended to include the
light output from QD-LED along with the other complex phenomenon such as
reflection of light from the electrodes, the loss of light from the sides of the
QD-LED, absorption of the light by the transport layers and glass substrate,
etc.
BIBLIOGRAPHY
82
Bibliography
[1] PO Anikeeva, "Physical properties and design of light-emitting devices based on organic materials and nanoparticles.", Physical properties and design of light-emitting devices based on organic materials and nanoparticles. (2009).
[2] SV Gaponenko, Optical properties of semiconductor nanocrystals, Cam-bridge university press 1998.
[3] P Ahonen, J Joutsensaari, O Richard, U Tapper, DP Brown, JK Jokiniemi, et al. Mobility size development and the crystallization path dur-ing aerosol decomposition synthesis of TiO2 particles, J.Aerosol Sci. 32 (2001) 615-630.
[4] JH Kim, TA Germer, GW Mulholland, SH Ehrman. Size-monodisperse metal nanoparticles via hydrogen-free spray pyrolysis, Adv Mater. 14 (2002) 518-521.
[5] C Murray, D Norris, M Bawendi. Synthesis and characterization of nearly monodisperse CdE (E= sulfur, selenium, tellurium) semiconductor nanocrystallites, J.Am.Chem.Soc. 115 (1993) 8706-8715.
[6] M Kunu, F Mikulec, NNIN Nanotechnology Open Textbook. In Colloidal Semiconductor Nanoparticle Formation [Online], 2007.
[7] MT Swihart. Vapor-phase synthesis of nanoparticles, Current Opinion in Colloid and Interface Science. 8 (2003) 127-133.
BIBLIOGRAPHY
83
[8] K Wegner, B Walker, S Tsantilis, SE Pratsinis. Design of metal nanopar-ticle synthesis by vapor flow condensation, Chemical Engineering Science. 57 (2002) 1753-1762.
[9] T Ohno. Morphology of composite nanoparticles of immiscible binary systems prepared by gas-evaporation technique and subsequent vapor con-densation, Journal of Nanoparticle Research. 4 (2002) 255-260.
[10] Y Nakata, J Muramoto, T Okada, M Maeda. Particle dynamics during nanoparticle synthesis by laser ablation in a background gas, J.Appl.Phys. 91 (2002) 1640-1643.
[11] W Marine, L Patrone, B Luk'Yanchuk, M Sentis. Strategy of nanocluster and nanostructure synthesis by conventional pulsed laser ablation, Appl.Surf.Sci. 154 (2000) 345-352.
[12] T Makimura, T Mizuta, K Murakami. Laser ablation synthesis of hy-drogenated silicon nanoparticles with green photoluminescence in the gas phase, Jpn J Appl Phys Part 2 Letter. 41 (2002) L144-L146.
[13] AP Weber, M Seipenbusch, G Kasper. Application of aerosol techniques to study the catalytic formation of methane on gasborne nickel nanoparticles, Journal of Physical Chemistry A. 105 (2001) 8958-8963.
[14] FK Urban III, A Hosseini-Tehrani, P Griffiths, A Khabari, Y- Kim, I Petrov. Nanophase films deposited from a high-rate, nanoparticle beam, Journal of Vacuum Science and Technology B: Microelectronics and Na-nometer Structures. 20 (2002) 995-999.
[15] SV Nand, K Ankur, K Brijesh, MB Raj. Synthesis of single phase cubic tin nitride nanoparticles by atmospheric pressure-halide vapor phase epitaxy, Solid State Sciences. 10 (2008) 569-572.
BIBLIOGRAPHY
84
[16] E Borsella, S Botti, MC Cesile, S Martelli, A Nesterenko, PG Zappelli. MoS2 nanoparticles produced by laser induced synthesis from gaseous pre-cursors, J.Mater.Sci.Lett. 20 (2001) 187-191.
[17] Y Kamlag, A Goossens, I Colbeck, J Schoonman. Laser CVD of cubic SiC nanocrystals, Appl.Surf.Sci. 184 (2001) 118-122.
[18] G Ledoux, J Gong, F Huisken, O Guillois, C Reynaud. Photolumines-cence of size-separated silicon nanocrystals: Confirmation of quantum con-finement, Appl.Phys.Lett. 80 (2002) 4834-4836.
[19] G Ledoux, D Amans, J Gong, F Huisken, F Cichos, J Martin. Nanostructured films composed of silicon nanocrystals, Materials Science and Engineering C. 19 (2002) 215-218.
[20] J Heberlein, O Postel, S Girshick, P McMurry, W Gerberich, D Iordanoglou, et al. Thermal plasma deposition of nanophase hard coatings, Surface and Coatings Technology. 142-144 (2001) 265-271.
[21] C Janzen, P Roth. Formation and characteristics of Fe2O3 nano-particles in doped low pressure H2/O2/Ar flames, Combust.Flame. 125 (2001) 1150-1161.
[22] K Wegner, WJ Stark, SE Pratsinis. Flame-nozzle synthesis of nanopar-ticles with closely controlled size, morphology and crystallinity, Mater Lett. 55 (2002) 318-321.
[23] D Sarigiannis, JD Peck, G Kioseoglou, A Petrou, TJ Mountziaris. Characterization of vapor-phase-grown ZnSe nanoparticles, Appl.Phys.Lett. 80 (2002) 4024-4026.
[24] U Kortshagen. Nonthermal plasma synthesis of semiconductor nanocrystals, J.Phys.D. 42 (2009) 113001.
BIBLIOGRAPHY
85
[25] CR Gorla, S Liang, GS Tompa, WE Mayo, Y Lu, Silicon and germanium nanoparticle formation in an inductively coupled plasma reactor, J. Vac. Sci. Technol. A. 15 (1997) 860-864.
[26] X Pi, R Liptak, JD Nowak, N Wells, C Carter, S Campbell, et al. Air-stable full-visible-spectrum emission from silicon nanocrystals synthesized by an all-gas-phase plasma approach, Nanotechnology. 19 (2008) 245603.
[27] M Wolkin, J Jorne, P Fauchet, G Allan, C Delerue. Electronic states and luminescence in porous silicon quantum dots: the role of oxygen, Phys.Rev.Lett. 82 (1999) 197-200.
[28] VL Colvin, MC Schlamp, AP Alivisatos. Light-emitting diodes made from cadmium selenide nanocrystals and a semiconducting polymer, Nature. 370 (1994) 354-357.
[29] A Mews, A Eychmueller, M Giersig, D Schooss, H Weller. Preparation, characterization, and photophysics of the quantum dot quantum well system cadmium sulfide/mercury sulfide/cadmium sulfide, J.Phys.Chem. 98 (1994) 934-941.
[30] MA Hines, P Guyot-Sionnest. Synthesis and Characterization of Strongly Luminescing ZnS-Capped CdSe Nanocrystals, J.Phys.Chem. 100 (1996) 468-471.
[31] M Danek, KF Jensen, CB Murray, MG Bawendi. Synthesis of Lumines-cent Thin-Film CdSe/ZnSe Quantum Dot Composites Using CdSe Quantum Dots Passivated with an Overlayer of ZnSe, Chem.Mater. 8 (1996) 173-180.
[32] M Schlamp, X Peng, AP Alivisatos. Improved efficiencies in light emit-ting diodes made with CdSe(CdS) core//shell type nanocrystals and a semi-conducting polymer, J.Appl.Phys. 82 (1997) 5837-5842.
BIBLIOGRAPHY
86
[33] S Coe, WK Woo, VB Moungi Bawendi. Electroluminescence from single monolayers of nanocrystals in molecular organic devices, Nature. 420 (2002) 800-803.
[34] L Kim, PO Anikeeva, S Coe-Sullivan, JS Steckel, MG Bawendi, V BulovicÌ . Contact Printing of Quantum Dot Light-Emitting Devices, Nano Lett. 8 (2008) 4513-4517.
[35] PO Anikeeva, JE Halpert, MG Bawendi, V Bulovic. Quantum dot light-emitting devices with electroluminescence tunable over the entire visible spectrum, Nano letters. 9 (2009) 2532-2536.
[36] AP Ghosh, LJ Gerenser, CM Jarman, JE Fornalik. Thin-film encapsula-tion of organic light-emitting devices, Appl.Phys.Lett. 86 (2005) 223503.
[37] J McElvain, H Antoniadis, MR Hueschen, JN Miller, DM Roitman, JR Sheats, et al. Formation and growth of black spots in organic light-emitting diodes, J.Appl.Phys. 80 (1996) 6002-6007.
[38] JM Caruge, JE Halpert, V Bulovic, MG Bawendi. NiO as an inorganic hole-transporting layer in quantum-dot light-emitting devices, Nano letters. 6 (2006) 2991-2994.
[39] L Davis. Properties of transparent conducting oxides deposited at room temperature, Thin Solid Films. 236 (1993) 1-5.
[40] K Suga, N Koshizaki, K Yasumoto, E Smela. Gas-sensing characteristics of ZnO-NiO junction structures with intervening ultrathin SiO2 layer, Sensors Actuators B: Chem. 14 (1993) 598-599.
[41] WY Lee, D Mauri, C Hwang. High-current-density ITO[sub x]/NiO[sub x] thin-film diodes, Appl.Phys.Lett. 72 (1998) 1584-1586.
BIBLIOGRAPHY
87
[42] J Caruge, J Halpert, VB V Wood, M Bawendi. Colloidal quantum-dot light-emitting diodes with metal-oxide charge transport layers, Nature pho-tonics. 2 (2008) 247-250.
[43] K Bourzac. Quantum dots go on display. Nature. 493 (2013) 283-283.
[44] B Kumar, R Hue, WL Gladfelter, SA Campbell. Comparing direct charge injection and Forster energy transfer into quantum dots in hybrid organ-ic/inorganic quantum dot light emitting devices, J.Appl.Phys. 112 (2012) 034501-034501-5.
[45] J Zhao. Efficient CdSe//CdS quantum dot light-emitting diodes using a thermally polymerized hole transport layer, Nano Lett. 6 (2006) 463-467.
[46] S Coe-Sullivan, WK Woo, JS Steckel, MG Bawendi, V Bulovic. Tuning the performance of hybrid organic//inorganic quantum dot light-emitting devices, Organic Electron. 4 (2003) 123-130.
[47] PTK Chin, RAM Hikmet, RAJ Janssen. Energy transfer in hybrid quantum dot light-emitting diodes, J.Appl.Phys. 104 (2008) 013108-013108-6.
[48] P Anikeeva, C Madigan, J Halpert, M Bawendi, V Bulović. Electronic and excitonic processes in light-emitting devices based on organic materials and colloidal quantum dots, Physical Review B. 78 (2008) 085434.
[49] P Reiss, J Bleuse, A Pron. Highly luminescent CdSe/ZnSe core/shell nanocrystals of low size dispersion, Nano letters. 2 (2002) 781-784.
[50] JJ Li, YA Wang, W Guo, JC Keay, TD Mishima, MB Johnson, et al. Large-scale synthesis of nearly monodisperse CdSe/CdS core/shell nanocrystals using air-stable reagents via successive ion layer adsorption and reaction, J.Am.Chem.Soc. 125 (2003) 12567-12575.
BIBLIOGRAPHY
88
[51] J Kido, M Kohda, K Okuyama, K Nagai. Organic electroluminescent devices based on molecularly doped polymers, Appl.Phys.Lett. 61 (1992) 761-763.
[52] P Jing, X Yuan, W Ji, M Ikezawa, X Liu, L Zhang, et al. Efficient energy transfer from hole transporting materials to CdSe-core CdS/ZnCdS/ZnS-multishell quantum dots in type II aligned blend films, Appl.Phys.Lett. 99 (2011) 093106.
[53] AH Mueller. Multicolor light-emitting diodes based on semiconductor nanocrystals encapsulated in GaN charge injection layers, Nano Lett. 5 (2005) 1039-1044.
[54] S Coe-Sullivan, JS Steckel, WK Woo, MG Bawendi, V Bulovic. Large-area ordered quantum-dot monolayers via phase separation during spin-casting, Adv.Funct.Mater. 15 (2005) 1117-1124.
[55] A Ekimov, A Onushchenko. Quantum size effect in three-dimensional microscopic semiconductor crystals, Jetp Lett. 34 (1981) 363-366.
[56] D Weinmann, W Häusler, B Kramer. Spin blockades in linear and non-linear transport through quantum dots, Phys.Rev.Lett. 74 (1995) 984-987.
[57] A Alivisatos. Semiconductor clusters nanocrystals, and quantum dots, Science. 271 (1996).
[58] S Gustavsson, R Leturcq, B Simovič, R Schleser, T Ihn, P Studerus, et al. Counting statistics of single electron transport in a quantum dot, Phys.Rev.Lett. 96 (2006) 76605.
[59] SM Cronenwett, TH Oosterkamp, LP Kouwenhoven. A tunable Kondo effect in quantum dots, Science. 281 (1998) 540-544.
BIBLIOGRAPHY
89
[60] J Carreras, C Bonafos, J Montserrat, C Domínguez, J Arbiol, B Garrido. Auger quenching-based modulation of electroluminescence from ion-implanted silicon nanocrystals, Nanotechnology. 19 (2008) 205201.
[61] X Pi, O Zalloum, A Knights, P Mascher, P Simpson. Electrical conduc-tion of silicon oxide containing silicon quantum dots, Journal of Physics: Condensed Matter. 18 (2006) 9943.
[62] U Kortshagen, L Mangolini, A Bapat. Plasma synthesis of semiconductor nanocrystals for nanoelectronics and luminescence applications, Journal of Nanoparticle Research. 9 (2007) 39-52.
[63] X Li, Y He, SS Talukdar, MT Swihart. Process for preparing macroscopic quantities of brightly photoluminescent silicon nanoparticles with emission spanning the visible spectrum, Langmuir. 19 (2003) 8490-8496.
[64] JD Holmes, KJ Ziegler, RC Doty, LE Pell, KP Johnston, BA Korgel. Highly luminescent silicon nanocrystals with discrete optical transitions, J.Am.Chem.Soc. 123 (2001) 3743-3748.
[65] J Kočka, I Pelant, A Fejfar. Light emitting silicon, recent progress, J.Non Cryst.Solids. 198 (1996) 857-862.
[66] W Shockley, W Read Jr. Statistics of the recombinations of holes and electrons, Physical Review. 87 (1952) 835.
[67] RN Hall. Electron-Hole Recombination in Germanium, Phys.Rev. 87 (1952) 387-387.
[68] RL Anderson. Experiments on Ge-GaAs heterojunctions, Solid-State Electronics. 5 (1962) 341-351.
[69] DL Smith. Recombination currents in a p–n heterojunction diode, physica status solidi (a). 44 (1977) 381-390.
BIBLIOGRAPHY
90
[70] Y Liu, C Gorla, S Liang, N Emanetoglu, Y Lu, H Shen, et al. Ultraviolet detectors based on epitaxial ZnO films grown by MOCVD, J Electron Mater. 29 (2000) 69-74.
[71] HL Chen, YM Lu, WS Hwang. Characterization of sputtered NiO thin films, Surface and Coatings Technology. 198 (2005) 138-142.
[72] K Ellmer. Resistivity of polycrystalline zinc oxide films: current status and physical limit, J.Phys.D. 34 (2001) 3097.
[73] RK Ligman, L Mangolini, UR Kortshagen, SA Campbell. Electrolumi-nescence from surface oxidized silicon nanoparticles dispersed within a pol-ymer matrix, Appl.Phys.Lett. 90 (2007) 061116-061116-3.
[74] N Sobolev, A Emel’yanov, E Shek, V Vdovin. Extended structural defects and their influence on the electroluminescence in efficient Si light-emitting diodes, Physica B: Condensed Matter. 340 (2003) 1031-1035.
[75] G Juska, M Viliunas, K Arlauskas, J Kocka. Space-charge-limited pho-tocurrent transients: The influence of bimolecular recombination, Phys.Rev.B. 51 (1995) 16668-16676.