Design of a high-power white light source with colloidal
quantum dots and non-rare-earth phosphors
by
Kristopher T. Bicanic
A thesis submitted in conformity with the requirements for the degree of
Master of Applied Science
Edward S. Rogers Sr. Department of Electrical and Computer Engineering,
University of Toronto
Toronto, Ontario
Canada
Copyright © 2017 by Kristopher T. Bicanic
Kristopher Bicanic University of Toronto Page ii
Abstract
Design of a high-power white light source with colloidal quantum dots and non-rare-earth phosphors
Kristopher T. Bicanic
Master of Applied Science
Graduate Department of Electrical and Computer Engineering
University of Toronto
2017
This thesis describes the design process of a high-power white light source, using novel
phosphor and colloidal quantum dot materials. To incorporate multiple light emitters, we
generalized and extended a down-converting layer model. We employed a phosphor
mixture comprising of YAG:Ce and K2TiF6:Mn4+ powders to illustrate the effectiveness of
the model. By incorporating experimental photophysical results from the phosphors and
colloidal quantum dots, we modeled our system and chose the design suitable for high-
power applications. We report a reduction in the correlated color temperature by ~600K
for phosphor and quantum dot systems, enabling the creation of a warm white light
emission at power densities up to 5 kW/cm2. Furthermore, at this high-power, their
emission achieves the digital cinema initiative (DCI) requirements with a luminescence
efficacy improvement up to 32% over the stand-alone ceramic YAG:Ce phosphor.
Kristopher Bicanic University of Toronto Page iii
Content
Abstract .......................................................................................................................................................................................... ii
Acronyms ....................................................................................................................................................................................... v
List of Tables .............................................................................................................................................................................. vi
List of Figures ........................................................................................................................................................................... vii
Chapter 1: INTRODUCTION
1.1. Solid state light sources .......................................................................................... 1
1.2. Phosphors and colloidal quantum dots for display technology ....................... 4
1.3. Device requirements .............................................................................................. 6
1.4. Organization of thesis ............................................................................................ 8
Chapter 2: BACKGROUND
2.1. Introduction ......................................................................................................... 10
2.2. Light source fundamentals ................................................................................. 10
2.2.1. Figure of merit ........................................................................................ 14
2.3. Red phosphor materials and properties ........................................................... 15
2.4. Multilayer phosphors .......................................................................................... 17
2.5. Conclusions .......................................................................................................... 18
Chapter 3: METHODOLOGY
3.1. Introduction ......................................................................................................... 20
3.2. Photoluminescence quantum efficiency ........................................................... 20
3.3. Power conversion efficiency and photoluminescence .................................... 21
3.4. Photoluminescence lifetime ............................................................................... 23
Chapter 4: MULTI-LAYERED PHOSPHOR DESIGN AND MODELLING
CONSIDERATIONS
4.1. Introduction ......................................................................................................... 25
4.2. Design considerations ......................................................................................... 25
4.3. Optical characterization of phosphor material ................................................ 27
4.3.1. Photoluminescence lifetime ................................................................... 27
4.3.2. Power conversion efficiency ................................................................... 29
4.4. Multilayered phosphor modeling ...................................................................... 30
4.5. Conclusions .......................................................................................................... 33
Kristopher Bicanic University of Toronto Page iv
Chapter 5: FABRICATION OF MULTILAYER STRUCTURE AND PHOSPHORS
5.1. Introduction ......................................................................................................... 34
5.2. Manganese based phosphors ............................................................................. 35
5.2.1. Synthesis ................................................................................................. 35
5.2.2. Physical and optical characterization ................................................... 36
5.3. Colloidal quantum dots in matrix ..................................................................... 39
5.3.1. Synthesis ................................................................................................. 40
5.3.2. Optical characterization ........................................................................ 41
5.4. Layer fabrication .................................................................................................. 42
5.4.1. Solution-processed phosphor ................................................................. 42
5.4.2. Solution-processed quantum dots ......................................................... 43
5.4.3. Solid state film processing ...................................................................... 45
5.5. Conclusions .......................................................................................................... 45
Chapter 6: CHARACTERIZATION AND PERFORMANCE OF PHOSPHOR
DEVICES
6.1. Introduction ......................................................................................................... 47
6.2. Device performance ............................................................................................ 48
6.2.1. Correlated color temperature and luminous efficacy ........................... 50
6.3. Device application in display technology ......................................................... 53
6.3.1. Overview ................................................................................................ 53
6.3.2. Design consideration and color space .................................................... 54
6.4. Conclusions .......................................................................................................... 55
Chapter 7: CONCLUSIONS AND FUTURE WORK
7.1. Summary .............................................................................................................. 57
7.2. Original contributions ........................................................................................ 58
7.3. Future work .......................................................................................................... 58
APPENDIX A
BIBLIOGRAPHY
Kristopher Bicanic University of Toronto Page v
Acronyms
CCT Correlated color temperature
CdSe Cadmium selenide
CdS Cadmium sulfide
CQD Colloidal quantum dot
Cs Caesium
DCI Digital cinema initiative
DMM Digital micro-mirror
Eu Europium
FWHM Full-width at half maximum
Mn4+ Magnesium ion
PCE Power conversion efficiency
PL Photoluminescence
PLQE Photoluminescence quantum efficiency
YAG:Ce Y3Al5O12:Ce3+ phosphor
Kristopher Bicanic University of Toronto Page vi
List of Tables
Table 1 | Model results for K2TiF6:Mn4+ and YAG:Ce phosphor structures. Excitation power
converted to green and red emission shown. Desired color corrected power ratio of 55%
green to 45% red emission. ........................................................................................................ 33
Kristopher Bicanic University of Toronto Page vii
List of Figures
Figure 1.1 | Design structures for solid state LED-based light sources[4] ............................. 2
Figure 1.2 | State of the art YAG:Ce phosphor emission profiles and methods of color
correction. a) CIE color coordinates of YAG:Ce compared to standard black body radiation
sources (black line). b) and c) Spectrally balanced emissions by filtering and spectral re-
engineering with a secondary down-converter, respectively. ................................................. 3
Figure 2.1 | (a) CIE 1931 XYZ standard observer color matching functions 𝑥(𝜆), 𝑦(𝜆), and
𝑧(𝜆)[19]. (b) CIE 1931 color space chromaticity plot and Planckian locus (shown in black)
[19]. ............................................................................................................................................... 12
Figure 2.2 | Display on chromaticity plot of the digital cinema initiative gamut range
requirements and locations of RGB sources and DCI white point[20]. .............................. 14
Figure 3.1 | High-power laser diode system with up to 5 kW/cm2 excitation. .................... 23
Figure 4.1 | a) Measured photoluminescence lifetime characterization of K2TiF6:Mn4+
phosphors with varying power densities. b) photoluminescence lifetime with varying
temperature environment. ......................................................................................................... 28
Figure 4.2 | a) Measured power conversion efficiency of K2TiF6:Mn4+ phosphor at varying
temperature. b) Measured emission spectra of K2TiF6:Mn4+ at varying temperatures with a
power density of 5kW/cm2. ....................................................................................................... 29
Kristopher Bicanic University of Toronto Page viii
Figure 4.3 | Phosphor 1D model setup for test structures used to calculate red and yellow
emission. ...................................................................................................................................... 32
Figure 5.1 | Red phosphor physical characterization a) XRD data for K2TiF6 and
K2TiF6:Mn4+ phosphor. b) SEM image of K2TiF6:Mn4+ phosphor powder. c) Elemental line
scanning from a typical Mn4+ doped K2TiF6 microparticle and SEM of analyzed cross
section. .......................................................................................................................................... 37
Figure 5.2 | SEM images of cross-section of K2TiF6:Mn4+ microparticles for K, Ti, F, and
Mn elemental mappings in the same selected areas. .............................................................. 38
Figure 5.3 | EDS spectrum of the microparticle displays the presence of K, Ti, F, and Mn
elements. ...................................................................................................................................... 38
Figure 5.4 | Absorption and PL emission spectra of K2TiF6:Mn4+ phosphors. Inset shows
the photographs of the K2TiF6:Mn4+ sample under UV lamp illumination ........................ 39
Figure 5.5 | Absorption and PL emission of CQDs in a silica matrix. ................................. 42
Figure 5.6 | a) Power conversion efficiency measurement showing quantum dot
degradation. b) Operation lifetime of quantum dots in optically transparent adhesive with
YAG:Ce capping layer. ............................................................................................................... 44
Figure 6.1 | Laser setup for excitation of phosphor materials and incorporation of
adjustable blue diode source for white light source. ............................................................... 48
Figure 6.2 | CQD 4-hour stability test at 5 kW/cm2. .............................................................. 50
Figure 6.3 | Spectra of the YAG:Ce compared to the QD and K2TiF6:Mn4+ structures. ..... 52
Kristopher Bicanic University of Toronto Page ix
Figure 6.4 | CIE color coordinates of the samples at high flux excitation (left) and
corresponding CCT of samples at all power densities, with Duv<0.02 (right). .................. 52
Figure 6.5 | Corresponding luminous efficacy of best Mn4+ and CQD samples at all power
densities. ....................................................................................................................................... 53
Figure 6.6 | Wavelength separation of K2TiF6:Mn4+ phosphor and YAG:Ce separated into
RGB sources................................................................................................................................. 55
Kristopher Bicanic University of Toronto Page 1 of 64
Introduction
1.1. Solid state light sources
Since the late 2000s, traditional incandescent and fluorescent light sources have been
increasingly supplanted by solid-state white light technology, which offers superior power
efficiencies, low energy consumption, and long operation lifetime[1, 2]. These benefits have
led to applications in a wide range of areas, from low-power spot illuminators to high-
power area illuminators [3]. For example, high-power illuminators can be used as light
sources for projection. For decades Xenon light bulbs have been used in this industry.
However, these sources fail to provide the brightness and color depth needed. The
unacceptably high energy cost also results in significant heat buildup in the system, which
requires further energy to cool sufficiently.
The requirements for white light emitting diodes (LEDs) are red, green, and blue
components. Ideally these will allow full and true rendering of an object's color and a full
gamut range for displays. In a first approach to combining color, one combines efficient
red, green, and blue LEDs into a single package; however, this approach results in distorted
Kristopher Bicanic University of Toronto Page 2 of 64
color and incurs high-cost requirements to fabricate three LEDs[2, 4]. Alternatively, the
more widely practiced method incorporates down-converting phosphor materials, which
converts the output of a single source to lower-energy light. This is ultimately more cost-
efficient than fabricating a multi-LED system. This can be accomplished through a source
that photoexcites three down-converters; alternatively a blue light source can be used,
where a portion of its light is down-converted to red and green[4].
Figure 1.1 | Design structures for solid state LED-based light sources[4]
For applications involving high-power densities (e.g. projection), the choice of
phosphor is critical, as unfavorable photophysical properties can lead to precipitous drops
in efficiency with increased power. For high-power systems of interest, many factors can
Kristopher Bicanic University of Toronto Page 3 of 64
cause a degradation in the phosphor material s emission. Once such high performing
phosphor, yellow-based Y3Al5O12:Ce3+ (YAG:Ce), can operate efficiently at high radiances
due to its high thermal conductivity and short PL lifetime. However, the
spectrum prevents it from replicating a lower correlated color temperature (CCT) [1, 5-
11]. The phosphor spectrum can be corrected using two approaches: the first method filters
out a portion of the green emission as waste (Figure 1.2b), whereas the preferred method
introduces a secondary down-converter to increase the red emission, minimizing waste
energy (Figure 1.2c) [12-15] .
Figure 1.2 | State of the art YAG:Ce phosphor emission profiles and
methods of color correction. a) CIE color coordinates of YAG:Ce compared
to standard black body radiation sources (black line). b) and c) Spectrally
balanced emissions by filtering and spectral re-engineering with a
secondary down-converter, respectively.
Kristopher Bicanic University of Toronto Page 4 of 64
1.2. Phosphors and colloidal quantum dots for display
technology
Today, down-converting light-emitting materials are extensively used for television
and portable electronic backlights [16]. The main down-converting materials used are
atomic emitting phosphors, organic molecules, and colloidal quantum dots. For the
creation of a low-cost light source, solution-processed methods are desired due to the
simplicity of production and large area fabrication. The current industry standard for
down-converting lighting is made from transition and rare earth metal activated
phosphors. This is in part due to their robustness and well-established emission properties.
In this thesis, I focus instead on the use of non-rare earth metal phosphors and colloidal
quantum dot materials. I investigate them in light of their potential low-cost synthesis and
the advantageous properties including tunable emission, high color purity for a large
gamut, and high photoluminescence quantum efficiency (PLQE).
Phosphors are materials which exhibit luminescence when excited with an optical
source. Within this work, atomic emitting phosphors were used for down-conversion
sources. These phosphor materials typically consist of two parts: a host material and a
dopant species to act as an activation source. The host material forms the majority of the
phosphor and provides a matrix into which the dopant is implanted. The host material
possesses a level of influence which impacts the emission of the activation atom based on
Kristopher Bicanic University of Toronto Page 5 of 64
electronegativity. The atomic centers of a phosphor are the dopant atoms embedded in the
host lattice. The dopant atoms dictate the emission properties of the phosphor material.
These atomic centers are the locations where electronic transitions occur, and these allow
for the down-conversion of incident radiation. Rare earth metals, as well as transition
metals, are typical materials which act as atomic emitters. The optical properties of these
materials are determined by the identity of the emitting center, and the tunability is limited
to variations in the host matrix composition and the concentration of the dopant used. For
red emitting centers, only a limited number of phosphors are known, each with their own
limitations, such as broad emission peaks and phosphorescent emission leading to power
saturation.
Colloidal quantum dots, in contrast with rare earth metal phosphors, are significantly
larger. They are an ensemble of many atoms, which form particles several nanometers in
diameter. These particles are of particular interest due to the unique and highly engineered
optical properties they possess. Such properties of interest are the fully tunable visible
spectrum emission, high color purity, and high photoluminescence quantum efficiency
(PLQE)[17]. These properties of the quantum dots are associated with their unique physics
resulting from their size and shape. The small size of quantum dots provides a level of
confinement for the electron and the hole present in the material. The electron and hole in
a nanocrystal are confined to a small region of space below the Bohr radius. This provides
tunable control over the excitation energy in the material that, by extension, creates a
Kristopher Bicanic University of Toronto Page 6 of 64
tunable bandgap that can vary by more than 1 eV [18]. Currently, the applications in which
colloidal quantum dots have been deployed have been limited to low intensity systems, due
to instabilities and non-radiative effects. In part this is due to higher radiances that create
multi-excitons in the quantum dots, increasing the rate of Auger recombination a non-
radiative process that reduces efficiency. This process further plagues high-intensity
systems, due to the resulting generated heat. The interaction in this nano-system is more
pronounced due to the confinement on the excitons[18]. Therefore, by engineering these
materials for optimal performance at high-power densities, one can design a device
structure to take advantage of the desirable properties and provide superior performance
at these higher power levels for applications such as projection.
1.3. Device requirements
In this thesis, I investigate the design and fabrication of high-power light sources. My
goal is to provide superior color quality and efficiency. The design of a warmer white light
source necessitates a thermally conductive red emitting material that meets the high
performance requirements. The red emitter needs to satisfy the following design
requirement first: it must be stable under high illuminant fluxes, which can reach up to 5
kW/cm2. This requirement ensures that the power regime in which the emitter operates
can provide the luminous flux required for projection while maintaining high
performance. The second requirement is device stability at elevated temperatures. The
Kristopher Bicanic University of Toronto Page 7 of 64
thermal requirement ties in with the high flux requirement; if the solution-processed
material is not thermally conductive, it will result in early failure. This in turn ties into the
thickness, as being thin ensures that sufficient cooling can be achieved throughout
the entire device. A limit of 250 µm was implemented in the present work in light of the
limitations of the cooling system available. Additionally, the creation of a single layer of
film is ideal to ensure that a minimal number of thermally resistive interfaces are
maintained.
With the goal to provide a white light source for projection, the red emitter needs to
provide a significant benefit over the current state-of-the-art methods. This is achieved by
providing higher color quality that meets current standards, such as the digital cinema
initiative (DCI). Incorporating red emitting sources can accomplish this, by providing
sufficient red to create an equivalent power relative to the green emission power. This
results in the full utilization of power. Taking into account the gamut of the light source, a
separation of the emission is required to provide a separate and distinguishable red and
green light source. This separation of red and green emission is needed to allow for
sufficient isolation of high quality red and green sources.
In summary, for the design of this high-power light source, the following requirements
need to be met for a superior alternative:
i. High-power density of sustained 5 kW/cm2.
Kristopher Bicanic University of Toronto Page 8 of 64
ii. Conduct heat effectively away from the emitter.
iii. Maintain a thin device profile below 250 µm.
iv. Correct color balance between red and green emission for high-quality emission
sources which satisfy DCI requirements.
v. Separate the emissions of green and red light.
1.4. Organization of thesis
This thesis is separated into seven chapters to describe the process of designing,
modelling, fabricating, and characterizing red emitting materials for the creation of a high-
power white light source. Within Chapter 2, a literature review is presented with a
description of current state-of-the-art red emitting phosphors with their benefits and
drawbacks. Additionally, a background into characterizing the device performance will
also be discussed and the display standards will be explained. A clear understanding of the
standards used will elucidate the reasoning behind certain requirements described in the
previous section. Chapter 3 details the experimental methods for the photophysical
measurements of the phosphor materials investigated. These characterization methods are
essential in characterizing the raw material as well as characterizing the final device for the
implementation as a white light source. The experimental methods described pertain to the
conversion efficacy of pump lights, raw power extracted from a simulated projector setup,
and emission collection. Within Chapter 4, the photo-physical measurements described in
Kristopher Bicanic University of Toronto Page 9 of 64
Chapter 3 are used to determine the fundamental properties of the chosen red manganese
phosphor and quantum dot films. Using these properties various phosphor structures were
modelled in a 1-dimensional calculation to obtain the highest performing structure.
Chapter 5 utilizes the device models described in Chapter 4 and investigates the various
methods of film fabrication capable of generating a low cost, thin, high performance film.
In addition, selected materials are optically and physically characterized. Finally, Chapter
6 provides details into the final device characterization, post fabrication, and the potential
of the device to be a candidate for solid-state lighting, as well as digital projection. Using
the figures of merit described in Chapter 2, the designed devices will be compared to state-
of-the-art ceramic phosphor materials.
Kristopher Bicanic University of Toronto Page 10 of 64
Background
2.1. Introduction
This chapter contains a review of the background information necessary to understand
the concepts described in this thesis. The first half of this chapter discusses the parameters
required for high-quality digital projection, with metrics used to quantify them and their
physical meaning. The second section will discuss the current state-of-the-art red sources
that are being considered for use in a white light source. In this section, the benefits and
drawbacks will be discussed. Finally, the last section of this chapter will detail current
literature on the modeling of multi-phosphor systems.
2.2. Light source fundamentals
The operating principle of phosphor materials, for light sources, is one that utilizes light
down-conversion. This process takes a photon of a given energy and re-emits it at a lower
energy relative to the material s bandgap and electronic states. In the design of a system,
Kristopher Bicanic University of Toronto Page 11 of 64
the color produced is important and can be quantified in various ways. The most common
standard method for quantifying color relates an emission spectrum to a position in a color
space, such as the CIE 1931 XYZ color space. The CIE color space is a representation of
matching a color perceivable by the human eye to an emission spectrum. This is done using
three color matching functions �̅�(𝜆), �̅�(𝜆), 𝑧̅(𝜆) (Figure 2.1a). The color matching
functions correspond with a mathematical interpretation to represent stimuli to the human
eye [19]. This system was designed initially in 1931 to give physical meaning to the color of
plot of 1931 color space is shown below (Figure 2.1b). A few notable features are present in
this depiction. The first is that any (x) and (y) location can be reached and depicts a given
color on the plot. A pure monochromatic source would produce a color on the curved edge
of the plot, which is known as the locus of spectral colors. The colors within this curved
surface require a mixture of multiple wavelengths to accurately reproduce the closer the
color is to the center the less saturated it is. For this reason,
This idea of white can vary based on the criteria chosen.
Kristopher Bicanic University of Toronto Page 12 of 64
Figure 2.1 | (a) CIE 1931 XYZ standard observer color matching functions
�̅�(𝜆), �̅�(𝜆), and 𝑧̅(𝜆)[19]. (b) CIE 1931 color space chromaticity plot and
Planckian locus (shown in black) [19].
The black curve in the CIE 1931 color space chromaticity plot (Figure 2.1b) is known
as a Planckian locus and is a representation of the colors that a blackbody radiator will emit.
These colors are identified as a white light of a given temperature, relative to the blackbody
this is the concept of a color temperature. As the Planckian locus
extends into the redder regions, the color temperature becomes colder as the blackbody
radiation is at a lower temperature, compared to a blue flame. Such temperatures can be
the following: 1000 K a yellow flame, 2740 K a 40 W gas light bulb, 6000 K is midday light
during a clear day [19].
The color space representation of color offers a simple model to visualize the gamut
provided with multiple sources. The representation of color is gaining importance,
specifically due to display technologies, where only a small number of sources are made
Kristopher Bicanic University of Toronto Page 13 of 64
and they are required to reproduce a large range of colors. Based on the media, various
requirements of a gamut range need to be achieved. For the given application of our high-
power emitter, displays and projection are the main goals. For these display technologies,
the Digital Cinemas Initiative (DCI) requirements need to be achieved (Figure 2.2) [20].
This corresponds to a red, green, and blue source, as well as a balance in their power to
eliminate waste power for the DCI white point.
The current industry standard, briefly described in Chapter 1, is YAG:Ce, which has a
broad emission spectrum from 500 nm to 700 nm. The broadness of this emission spectrum
is advantageous for creating a white light source. However, a heavier weight is on the green
emission if used for display applications, this would create a plethora of powerful green
emission when compared to red. This results in a significant portion of light, upward of
40%, that is converted into waste energy to meet the DCI requirements. As a result, we need
to design a system with an improved red emitter when compared to the YAG:Ce system.
Kristopher Bicanic University of Toronto Page 14 of 64
Figure 2.2 | Display on chromaticity plot of the digital cinema initiative
gamut range requirements and locations of RGB sources and DCI white
point[20].
2.2.1. Figure of merit
Correlated color temperature
The Planckian locus depicts white light emission from a blackbody radiator, which is
typically used as a reference source to compare white light. Colors emitted from a source
close to these blackbody radiators do not indicate the temperature of the source. Rather,
isothermal lines can be drawn perpendicular to the locus, to give a range of colors which
are representative of this color temperature. The sources close to the Planckian locus can
mimic the color, which is known as a correlated color temperature (CCT). A specific color
Kristopher Bicanic University of Toronto Page 15 of 64
temperature can be desired, which would give a method of tuning the CCT an advantage
over materials limited to a higher CCTs.
Luminous efficacy
flux 𝜙𝑒 to a luminous flux 𝜙𝑝. This can have two meanings for the radiant emission and
the efficiency of the source. This dictates how efficient the source is at turning an incident
power (energy of the photons) into a visually perceived emission (luminous flux). This is
partly affected by the non-uniform response of the human eye to color. The luminous
efficacy of radiant emission is the theoretical maximum performance of a source based on
the emission; in other words, it is the luminous flux of the spectrum relative to the energy
of the photons emitted. Luminous efficacy of the source, converting a power to light, takes
into account the efficiency loss when down-converting and is given by:
𝐿𝐸 =𝜙𝑝
𝜙𝑒∗ 𝐸𝑄𝐸 (1)
This considers the lumens emitted by the source for a given input power.
2.3. Red phosphor materials and properties
For the advancement of both solid-state lighting and display technologies, a major
focus lies on designing high performing red emitting light sources. These red light sources
are desired to create warmer white light solutions, as well as to improve the color quality of
display technologies. Hence, an abundance of research has gone into investigating the
Kristopher Bicanic University of Toronto Page 16 of 64
potential use of red light emitters. Examples of red light emitters are Eu2+ and Mn4+
activated phosphors.
Numerous host materials have been examined for Eu2+ activated phosphors; some
examples that have undergone intensive studies are sulfides, nitrides, and oxynitrides [13-
15]. These materials all have some similarities in their synthesis conditions, which are
typically quite harsh, e.g. Eu2+- -SiAlON is typically prepared under 1900 °C, 10
atm, N2 pressure [12] and for M2Si5N8:Eu2+ temperature ranges from 1400-2000°C are used
[13]. In addition to these demanding synthesis conditions, the raw materials are expensive
and lack stability when exposed to humidity [13].
Furthermore, the Eu2+ based phosphors possess a less than ideal absorption spectrum,
which overlaps
which greatly reduces efficiency [13-15, 21]
emission spectra is their wide emission range. Typically, the full width at half maximum
(FWHM) of the emission peaks ranges from 50 nm to ~175nm[13-15].
On the other hand, potential exists for using a transition manganese dopant atom
(Mn4+) to create a red emitting phosphor. Unlike the Eu2+ doped phosphors, Mn4+ doped
red phosphors have the advantage that the synthesis process does not require expensive,
high-temperature equipment, allowing for easier preparation [22, 23]. For the host
4+-
Kristopher Bicanic University of Toronto Page 17 of 64
and the emission wavelength [24-28]. Previous literature states that to attain the largest
blue shift possible, fluoride compounds should be adopted, making them an ideal host for
Mn4+ to produce a highly luminous emission. With the large blue shift, the Mn4+ exhibits
the most intense excitation band located at ∼460 nm and sharp red emission line peaks at
∼630 nm [29]. Unfortunately, Mn4+-doped phosphors always suffer from a compromise
between excellent emission properties and long lifetime, which is believed to induce
saturation problems when performed at high-power densities.
Recently, Zhu et al. successfully demonstrated K2TiF6:Mn4+ as a secondary down-
converter in conjunction with YAG:Ce; however, high flux illumination, necessary for
high-power applications such as projection, was not reported [22]. While K2TiF6Mn4+ has
many advantageous properties [22-28], it also possesses a long (~5 ms) excited state lifetime
[29]. This long excited state lifetime results in saturation at high-power, which can be
deleterious for efficiency if not properly compensated.
2.4. Multilayer phosphors
Phosphor down-converters have been extensively researched for their use in lighting
applications. In turn, the combination of multiple phosphors has been investigated,
including their performance based on the phosphor structure. The studies that have been
previously carried out present interesting insight into how packaging phosphors affect the
luminous efficacy. Literature has demonstrated that a separate multi-layered phosphor
Kristopher Bicanic University of Toronto Page 18 of 64
structure provides the highest performance, using a calcium sulfide red phosphor and an
inorganic silicate yellow phosphor [30]. These results indicate that a heavy dependence
exists on the light extraction for their LED packaging. This highlights that grading the index
from high to low luminous flux up to 18% [30].
In addition to the light extraction enhancement, from the grading of indexes, other
various properties are shown to affect the optimal arrangement of phosphors. Such
properties include the phosphor materials used, due to their individual excitation and
emission spectra. This leads to a number of considerations in the design, including
reabsorption, device architecture, quantum efficiencies, and densities of the phosphor
materials [31].
The creation of high-quality white light sources can be accomplished with a multi-
phosphor system by designing an appropriate device architecture for high-power operation
[30]. However, a strong dependence is on the phosphor materials used, device architecture,
quantum efficiencies, and densities of the phosphor materials that need to be considered
in the design [31]. To realize a high-power source, both scattering and saturation also need
to be incorporated into the design protocol.
2.5. Conclusions
In summary, within this chapter a background on color science and light color was
provided and common red phosphors were investigated. Each red phosphor mentioned
Kristopher Bicanic University of Toronto Page 19 of 64
had some complication for the inclusion in a phosphor mixture at high-power density. The
Mn4+ activated phosphor was a compelling candidate due to its favorable emission
spectrum with 10 nm FWHM, at a luminous wavelength (630 nm). This is in contrast to
Eu-activated phosphors with FWHMs ranging from 50 to 200 nm and a significant portion
of the emission >650 nm. Therefore, as a red emitter, Mn4+ would be a promising candidate
if the saturation effect can be overcome. In addition to this, previous literature showed
multi-phosphor systems have a plethora of parameters required to accurately describe the
emission characteristics. For application in high-power systems, additional parameters
need to be accounted for based on the materials selected, e.g. such as saturation and thermal
effects.
Kristopher Bicanic University of Toronto Page 20 of 64
Methodology
3.1. Introduction
Within this chapter, the experimental techniques used to characterize the phosphor
devices will be discussed in detail. This section discusses the optical characterization that
was performed for analysis of the phosphor materials. Using this information, a model will
be created for designing a high-power emission source. In addition to this, performance
metrics of the device will be established.
3.2. Photoluminescence quantum efficiency
For the photoluminescence quantum efficiency (PLQE), an experimental method
adapted from literature was used to allow for higher accuracy measurements of films. The
equation for PLQE is given by:
𝜂𝑃𝐿𝑄𝑌 =𝑛𝑢𝑚𝑏𝑒𝑟 𝑝ℎ𝑜𝑡𝑜𝑛𝑠 𝑒𝑚𝑖𝑡𝑡𝑒𝑑
𝑛𝑢𝑚𝑏𝑒𝑟 𝑝ℎ𝑜𝑡𝑜𝑛𝑠 𝑎𝑏𝑠𝑜𝑟𝑏𝑒𝑑(2)
Kristopher Bicanic University of Toronto Page 21 of 64
For the measurement of films a difficulty in measuring the emission is due to
anisotropy, when compared to solutions [32]. To correct this, three measurements were
taken: the first was a blank of the laser itself, the second was the sample placed in an
integrating sphere and directly illuminated, and third was the sample placed in an
integrating sphere and not being directly illuminated [32]. These measurements are used
to extract the absorbed light with the absorption of scattered light within the film. This is
represented by the equations: 𝐿2 = 𝐿1(1 − 𝜇), 𝐿3 = 𝐿1(1 − 𝐴)(1 − 𝜇) [32] with L being
the integrated power of the laser collected and 𝜇 the absorption of scattered light. This gives
the equation for the absorption:
𝐴 = 1 −𝐿3
𝐿2 (3)
Using this information and the power of the emission, P, one can calculate the PLQE
of the films with the following equation[32]:
𝜂 =𝑃𝑐 − (1 − 𝐴)𝑃𝑏
𝐿𝑎𝐴 (4)
3.3. Power conversion efficiency and photoluminescence
The power conversion efficiency (PCE) measurement was performed to determine the
conversion efficiency of an excitation pump source to the emission of the sample; therefore,
the efficiency is given by the equation:
Kristopher Bicanic University of Toronto Page 22 of 64
𝑃𝐶𝐸 =𝑃𝑒𝑚𝑖𝑡𝑡𝑒𝑑
𝑃𝑒𝑥𝑐𝑖𝑡𝑎𝑡𝑖𝑜𝑛=
𝐸𝑄𝐸 ∗ ℎ𝜈𝑒𝑚
ℎ𝜈𝑒𝑥
(5)
For the excitation source a high-power laser diode system was employed; the setup is
shown in Figure 3.1. A total of 16 diodes, with each capable of providing 2 W of power, in
two laser banks were combined in parallel. The emission of these diodes was collimated in
a lens system, combined, and focused to a 1 mm2 area. This pump area was confirmed
through a CCD and a 20-80 knife edge measurement. To circumvent critical thermal issues,
the pump was pulsed with a repetition rate of 60 Hz for pulse durations of 300 µs. This
setup provided a simulated environment of a phosphor projector source on a rotating ring.
This setup provided in excess of 5 kW/cm2 peak power density. Using a dichroic mirror,
the emission was separated from the excitation source and sent through a lens system; the
emission from the sample was then collected and focused onto a 30W UP55N thermal
power meter with an aperture of 55 mm. For the efficiency measurement, high temperature
environments were also investigated. To simulate these thermal effects, the sample was
mounted onto an aluminum block with a resistive heating element. This heater was
controlled with thermal couples behind the sample and a temperature controller for active
feedback.
For the purpose of obtaining spectral data from the device, the setup was manipulated
to allow for the collection of the emission spectrum. At the collection in front of the thermal
power meter, a white diffused sheet was used to reflect the emission. To ensure the
Kristopher Bicanic University of Toronto Page 23 of 64
spectrometer was not saturated a neutral density filter was added. A 105 µm fiber was used
to collect the reflection from the white sheet and pass it to an Ocean Optics USB2000
spectrometer. This emission was used to characterize the high-power emission.
Figure 3.1 | High-power laser diode system with up to 5 kW/cm2 excitation.
3.4. Photoluminescence lifetime
Due to the exceedingly long lifetime of the manganese activated phosphors, the laser
diode system described above was adapted for photoluminescence lifetime. The collection
of the lens system was instead focused onto a high-speed photodetector. This
photodetector was a silicon based DET10A photodiode, with an aperture of 0.8 mm2 and a
rise time of 1 ns. This detector provided sufficient resolution when investigating long,
Dichroic Mirror
Sample Stage
High Power thermal
power meter
Fiber coupler for
Spectrometer
16 diodes, 𝜆 = 442 nm
and 2 W max power
each
Kristopher Bicanic University of Toronto Page 24 of 64
millisecond regime lifetimes. The use of the 300 µs pulse at 60 Hz was sufficiently small
enough to enable the excitation to fully relax between pulses.
For better resolution, a Horiba Fluorolog TCSPC system with an iHR 320
monochromator and a PPO·900 detector was used to characterize our quantum dot
samples. However, the emission lifetime of these dots was extremely fast when compared
to the phosphors; for that reason, it was not a concern for allowing saturation effects from
occurring. In this case, other non-radiative effects that are power dependent were of
concern, e.g. Auger recombination
Kristopher Bicanic University of Toronto Page 25 of 64
Multi-layered phosphor design and
modelling considerations
4.1. Introduction
Within this section, the red emitter s performance was evaluated with elevated
power and temperature conditions. These performance metrics were then investigated in
a one-dimensional model to determine the optimal device structure to obtain the highest
performance, while meeting the requirements for the device set out in Chapter 1.
4.2. Design considerations
As previously mentioned in Chapter 1, the white light source that is being created needs
to provide a stable and pure red emission. As a result, the phosphor down-converter should
be capable of high-power illumination at a peak power density of 5 kW/cm2, have a red
color purity which reaches the digital cinema initiative (DCI), and maintain a thin device
profile for integration into a projector system. With these requirements, the red phosphor
Kristopher Bicanic University of Toronto Page 26 of 64
material will be integrated with the industry standard (YAG:Ce) to obtain a high-quality
white light source for display technologies. Given these requirements of a pure red source
to incorporate into the YAG:Ce, a narrow red emission spectrum is desired. Two materials
that achieve this property are manganese activated phosphors, with a narrow emission (~10
nm FWHM), and CdSe quantum dots (~20 nm FWHM). Unfortunately, each of these
materials has a drawback when illuminated under high excitation, namely a drastic increase
in non-radiative processes. To circumvent this complication, a one-dimensional phosphor
model was designed to determine a device structure that limits the non-radiative processes.
To achieve this an investigation into the phosphor material was performed under harsh
temperature and illumination conditions.
In addition to obtaining the correct materials to create this light source, a design
limitation in the thickness of the active material was set to be 250 µm. This limitation was
implemented for two reasons: to ensure the cooling across the film was maintained at the
high excitation levels and to meet the limitation of the collection optics for use in
projectors. As a result of this limitation the density of the phosphor materials was required
to be high in the film to allow for sufficient absorption. However, this results in higher
levels of scattering which leads to increased absorption and poor collection in lower levels
of the film. The increased scattering would need to be incorporated into a model for
determining a device architecture of these materials.
Kristopher Bicanic University of Toronto Page 27 of 64
4.3. Optical characterization of phosphor material
Within this section, an investigation into the effects of power and temperature on the
The materials underwent high temperature and high
flux conditions to understand their effects on the emission. The emission properties of
photoluminescence lifetime were investigated for the manganese-based phosphor to
observe trends, which may correspond to an increase in non-radiative processes.
Sequentially, both the quantum dot
conversion efficiency was performed at elevated powers.
4.3.1. Photoluminescence lifetime
The design process considers the extent of both scattering and saturation, allowing for
the design of a high-power illumination source. The relative scattering coefficients were
obtained by measuring the transmission of the dense phosphor layers for a given thickness.
To determine the cause and effect of saturation, the photoluminescence (PL) lifetime and
efficiency of the red phosphor were investigated as a function of power. From this
information, and incorporating the scattering effects of the dense phosphor layers, a 1D
model was built to investigate varying phosphor architectures. Once the best architecture
is obtained, a study can be carried out to determine the appropriate ratio of red and green
phosphors to achieve the desired white temperature.
Kristopher Bicanic University of Toronto Page 28 of 64
The PL lifetime of K2TiF6:Mn4+ was observed to be 4.8 s; this can cause saturation as a
result of non-radiative decays. Here we investigated the power and temperature dependent
lifetime of K2TiF6:Mn4+ using 442 nm excitation source. When we adjust the peak power
density from 0.4 kW/cm2 to 5 kW/cm2, the PL lifetime decreases by 31%, from 4.8 ms to
3.3 ms (Figure 4.1 left). At 5 kW/cm2, it further decreases from 3.3 ms to 2.7 ms upon a
temperature increase from room temperature to 100 C (Figure 4.1 right). The decrease in
the PL lifetime for both high-power density and high temperature indicate an enhancement
of non-radiative processes, which are detrimental to high-power density operation at
elevated temperature.
Figure 4.1 | a) Measured photoluminescence lifetime characterization of
K2TiF6:Mn4+ phosphors with varying power densities. b)
photoluminescence lifetime with varying temperature environment.
Kristopher Bicanic University of Toronto Page 29 of 64
4.3.2. Power conversion efficiency
To determine the effect of saturation on performance, the power conversion efficiency
(PCE) was measured with varying power and temperature (Figure 4.2a). Each temperature
condition shows a decrease in PCE with increasing power, which is attributed to non-
radiative decay. However, no temperature-dependent loss of efficiency occurs until both
the highest power (5 kW/cm2 , when
mixed with a yellow phosphor, the temperature induced saturation effect is expected to be
negligible, as the power will be distributed between the red and yellow phosphors.
Furthermore, the sample shows a minimal loss in its emission at 5 kW/cm2 until the highest
temperatures are reached (Figure 4.2 right).
Figure 4.2 | a) Measured power conversion efficiency of K2TiF6:Mn4+
phosphor at varying temperature. b) Measured emission spectra of
K2TiF6:Mn4+ at varying temperatures with a power density of 5kW/cm2.
Kristopher Bicanic University of Toronto Page 30 of 64
4.4. Multilayered phosphor modeling
To obtain white light emission, we designed a system that incorporated a pulsed blue
excitation source used to excite down-converting phosphors YAG:Ce and K2TiF6:Mn4+
(Figure 4.3). The pulsed laser is used here to minimize saturation and thermal issues. Both
phosphors were mixed into a silicone adhesive matrix (13% wt.) to obtain a dense and
smooth film, with a thickness of less than 250 µm. The emission from the phosphors was
efficiently transmitted through a dichroic mirror. For digital projection, this emission can
be combined with an adjustable blue diode; for our analysis, we calculated the power
required from a 467 nm source to produce the desired white color temperature (Figure
4.3).
We modeled various test structures, with the goal of operating as a solid-state lighting
solution, as well as a source for high-power displays and projectors. Three different types
of designs were adopted: layering the YAG:Ce phosphor on top and the K2TiF6:Mn4+
phosphor as a base, the converse, and a homogeneous blend between the two materials
(Figure 4.3). We based our calculations on results provided by Kang [33] and extended it
to incorporate the alternate architectures, as well as scattering and saturation. The model
assumes a 1D transmission within the phosphor, using the following equations:
𝑑𝑃𝑒𝑥
𝑑𝑧= −𝛼𝑒𝑥𝑃𝑒𝑥(𝑧) (6)
Kristopher Bicanic University of Toronto Page 31 of 64
𝑑𝑃𝑒𝑚
𝑑𝑧= −𝛼𝑒𝑚𝑃𝑒𝑚 +
1
2𝛼𝑒𝑥𝛽𝑃𝑒𝑥(𝑧) (7)
𝑑𝑃𝑒𝑚−
𝑑𝑧= 𝛼𝑒𝑚𝑃𝑒𝑚 −
1
2𝛼𝑒𝑥𝛽𝑃𝑒𝑥(𝑧), (8)
These three equations describe the intensity of light propagating within the phosphor
materials. Equation (5) describes the power of the excitation source, 𝑃𝑒𝑥 . In this
experiment, we used a 442 nm laser diode, focused on intensities of up to 5 kW/cm2. This
equation accounts for absorption by the red and green emitting phosphors. The absorption
of both is incorporated into the coefficient 𝛼𝑒𝑥 per unit length of the device, and thus 𝛼𝑒𝑥
depends on the phosphor composition in this layer. Equation (6) and (7) describe the
emission of the red and yellow phosphors in the forward (6) and reverse (7) directions,
respectively. Within these equations, 𝛼𝑒𝑚 is the energy loss of the phosphor emission, and
𝛽 is the conversion efficiency of the excitation source to phosphor emission
shows a power dependence based on the excitation power, which decreases as the excitation
photons are partially absorbed throughout the phosphor. on power,
𝑃𝑒𝑚± , was calculated for both red and yellow emission separately in the layered structures
with 𝑃𝑒𝑚+ phosphor emission travelling forward into the device and 𝑃𝑒𝑚
− emission travelling
backward to the surface of the device. 𝑃𝑒𝑚− at z=0 is the collected emission, as the system
architecture operates in reflection mode. As a mirror is located at the back of the device,
𝑃𝑒𝑚− (𝑡) = 𝑃𝑒𝑚
+ (𝑡) at this position. The scattering of the excitation from the rear phosphor
Kristopher Bicanic University of Toronto Page 32 of 64
was calculated based on the transmission properties of the top layer; this scattering is more
pronounced in our system due to the higher loading densities, when compared to lower
power phosphor systems. The absorption saturation was incorporated into the K2TiF6:Mn4+
phosphor by using a linear fit to the PCE measured in Figure 3c, and is incorporated into
the 𝛽 term for the K2TiF6:Mn4+ phosphor only.
Figure 4.3 | Phosphor 1D model setup for test structures used to calculate
red and yellow emission.
The results of the model (Table 1) show that when a red phosphor is used as a base, the
scattering induced from the YAG:Ce phosphor coating significantly reduces the red
collection. Alternatively, when the YAG:Ce is the base, the yellow collection suffers due to
the scattering in the K2TiF6:Mn4+ layer, and saturation of K2TiF6:Mn4+ limits the red
emission. The homogeneous mixture, however, distributes the absorption and scattering
losses throughout the entire device structure. This allows for the best balance between red
and green power efficiency, which is a result of reduced red saturation.
Kristopher Bicanic University of Toronto Page 33 of 64
Table 1 | Model results for K2TiF6:Mn4+ and YAG:Ce phosphor structures.
Excitation power converted to green and red emission shown. Desired color
corrected power ratio of 55% green to 45% red emission.
Mixture Red Coating/
Yellow Base
Yellow Coating/
Red Base
Green Emission 25% 18% 44%
Red Emission 22% 39% 16%
Total Converted
Power Output 47% 57% 60%
Color Corrected
Power Output 44% 31% 38%
4.5. Conclusions
In this chapter, an investigation into the emission characteristics of the phosphor
materials was performed. These emission properties were obtained under high temperature
and intensity conditions. This information presented insight into the material s
performance as a red source in a white light emitter. This information was also
implemented into a model to calculate the performance when incorporated into a device.
In this model, given the device structure, a mixture of the red phosphor materials and
YAG:Ce would provide the optimal color-corrected emission. When compared to the
layered structures, a lower emission intensity is observed. However, to obtain a balance of
red and green emission, the mixture system requires no adjustment, whereas the layered
structures will need green attenuation to emission levels below that of the mixture system.
Kristopher Bicanic University of Toronto Page 34 of 64
Fabrication of multilayer structure
and phosphors
5.1. Introduction
Chapter 5 delves into the fabrication and synthesis processes that were performed for
the high-power light sources used. For the first section, the synthesis procedure for the
manganese activated phosphor is established. The physical and optical characterization are
provided. Since the materials are created in microparticle sizes, and the device thickness is
a few hundreds of nanometers, the particle size is critical. A similar procedure was
described for the implementation of colloidal quantum dots into a silica matrix, and optical
characterization was carried out. Finally, the processing methods to incorporate the red
emitters with YAG:Ce powder was described. An initial goal of pure solution-processing is
desired; however, a method of implementing the CQDs in a highly conductive matrix was
described.
Kristopher Bicanic University of Toronto Page 35 of 64
5.2. Manganese based phosphors
In Chapter 2 various rare and non-rare earth metal phosphors were investigated. Each
type of phosphor material has benefits and drawbacks with applications at high-power.
Short photoluminescence lifetimes are desired to have lower impact when operating at high
flux; however, many rare earth metals in literature with short PL lifetimes do not have the
desired emission spectrum to create a high quality red source for displays. Such examples
are Eu and Cs based phosphors with emission bandwidths of 80 nm and a large portion of
the wavelengths around 650 nm, a regime insensitive to the human eye, as shown by �̅�(𝜆)
in Figure 2.1a. Manganese based phosphors, on the other hand, depicted very narrow
emission peaks with a bandwidth of 50 nm [22]. When implementing the manganese
activated phosphor into a fluoride based matrix, the emission peak is near 630 nm, an
optimal wavelength for the sensitivity to the human eye. The downside with implementing
a manganese based phosphor was the long PL lifetime, which at the power regime required
typically enters roll off and a significantly decreased efficiency.
5.2.1. Synthesis
The synthesis of the manganese-based phosphor was conducted in two steps: the first
was the preparation of the manganese into a matrix compatible with the host, and the
second step involved the doping process of the manganese atoms into the host matrix. The
Kristopher Bicanic University of Toronto Page 36 of 64
synthesis was adapted from previous literature in the doping of manganese phosphor into
a K2TiF6 matrix to obtain high quantum efficiencies [22].
For the synthesis of the K2MnF6 powders, a mixture of 9.0 g KHF2 and 0.45 g KMnO4
was dissolved in 30 mL 48% HF to form a purple solution. A 0.1 mL of 30% H2O2 solution
was added dropwise to the solution . The solution gradually produced a yellow
precipitate at the bottom. The obtained yellow powder was washed with acetone several
times and then dried at room temperature.
Then, the K2MnF6 powder was used to dope into the K2TiF6 matrix. To perform this,
36 mg of K2MnF6 was dissolved in 2 mL 48% HF solution, and then commercial K2TiF6
powder (0.5 g, 0.75g, 0.9 g, 1g, 2.5g, 5g) was added to the HF solution separately. The liquid-
solid mixtures were stirred at room temperature for about 20 minutes. The white powder
turned to light yellow after the liquid-solid exchange reaction, indicating that the Mn4+
doped successfully into the K2TiF6 matrix. The obtained light yellow powder was washed
with 5% HF twice and then several times with acetone. Finally, it was dried at room
temperature.
5.2.2. Physical and optical characterization
The red emitter K2TiF6:Mn4+ was synthesized as previously reported [22, 34]. Pure
K2TiF6 and Mn4+ doped powders were characterized by XRD measurements (Figure 5.1a):
diffraction peaks of the two samples can be indexed into tetragonal-phase K2TiF6 (JCPDS
Kristopher Bicanic University of Toronto Page 37 of 64
No. 00-008-0488). The particle size of the phosphors from SEM images (Figure 5.1b) ranges
from 30-500 microns and shows a wide size distribution. Elemental line scanning for a
typical Mn4+ doped K2TiF6 microparticle highlights a similar distribution between
manganese and titanium along the cross section (Figure 5.1). This is in agreement with the
mapping measurements from scanning transmission electron microscopy (STEM, Figure
5.2). The molar ratio of manganese to titanium was estimated to be 10% based on the
calibrated signal intensities of energy dispersive spectroscopy (Figure 5.3). These results
demonstrate Mn4+ ion doping in the K2TiF6 matrix with uniform distribution throughout
the microparticles.
Figure 5.1 | Red phosphor physical characterization a) XRD data for K2TiF6
and K2TiF6:Mn4+ phosphor. b) SEM image of K2TiF6:Mn4+ phosphor
powder. c) Elemental line scanning from a typical Mn4+ doped K2TiF6
microparticle and SEM of analyzed cross section.
Kristopher Bicanic University of Toronto Page 38 of 64
Figure 5.2 | SEM images of cross-section of K2TiF6:Mn4+ microparticles for
K, Ti, F, and Mn elemental mappings in the same selected areas.
Figure 5.3 | EDS spectrum of the microparticle displays the presence of K,
Ti, F, and Mn elements.
The intense absorption band of the K2TiF6:Mn4+ microparticles (Figure 5.4), positioned
at 468 nm, overlaps well with the emission of conventional blue diodes. The overlap
between the absorption of K2TiF6:Mn4+ and emission of YAG:Ce is negligible, thereby
avoiding the re-absorption problem that usually occurs between (oxy)nitride red
phosphors and yellow (or green) phosphors [21]. The red emission of K2TiF6:Mn4+ (Figure
5.4) comprises three sharp peaks with the main peak positioned at 631.5 nm (FWHM: 10
Kristopher Bicanic University of Toronto Page 39 of 64
nm), with PLQEs up to 95%, similar to literature values [28, 29]. Their narrowband spectra
and the high PLQEs enable high efficiency and excellent color quality for white LEDs [14].
Figure 5.4 | Absorption and PL emission spectra of K2TiF6:Mn4+ phosphors.
Inset shows the photographs of the K2TiF6:Mn4+ sample under UV lamp
illumination
5.3. Colloidal quantum dots in matrix
The secondary phosphor material that was used as a red emitter was cadmium selenide
(CdSe) core cadmium sulfide (CdS) shell quantum dots. These colloidal quantum dots were
prepared through solution-processing means, maintaining a low cost. For high-power
applications, a thin shell was used to reduce the absorption cross section and thereby
maintain high efficiency at elevated powers. This reduced absorption cross section is
important for high-power applications to prevent excessive absorption within the QDs,
leading to an increase in non-radiative effects. For quantum dots, the process that causes
Kristopher Bicanic University of Toronto Page 40 of 64
the non-radiative effects is Auger recombination. Unfortunately, the thin shell is normally
undesired since a precipitous drop in the PLQE results when in film. This is due to energy
transfer between quantum dots. To correct this issue, the dots used were engineered into a
silica matrix. This matrix provides passivation of the quantum dots from the environment,
and a means by which to separate the dots to ensure the high efficiency is maintained. Here
we will describe the synthesis procedure to incorporate colloidal quantum dots into a silica
matrix.
5.3.1. Synthesis
For the synthesis of the colloidal quantum dots, an alternate method was implemented
to improve their optical properties at high-power. This process provided a thin passivating
shell to the quantum dots, in order to reduce the absorption across section, thereby
reducing the potential of multi-excitons from generating within a single dot. The synthesis
of the colloidal quantum dots is described in Appendix A. The quantum dots are then
implemented into a passivating silica matrix to prevent quenching by energy transfer. The
matrix process is outlined as follows:
For the incorporation of CQDs into the silica matrix, a solution of 100 µL QDs (Aexciton
~ 2.5) was used. By using acetone, centrifugation, and decanting, the QD samples were
precipitated and extracted. The precipitate was re-dispersed in a mixture of 300 µL pyridine
and 50 µL 6-mercapto-1-hexanol. The solution was stirred for 48 hours until transparent.
Kristopher Bicanic University of Toronto Page 41 of 64
Following this, hexane was used to precipitate the quantum dot from the pyridine/6-
mercapto-1-hexanol solution. After extracting the quantum dots, 200 µL ethanol was used
to re-disperse into solution. With the ethanol solution, 50 µL 3-MPS, 200 µL TEOS, and
100 µL propylamine were added. Following a minute of stirring, 20-30 µL of water was
incorporated and the solution was centrifuged for 10 s. This solution was then sealed in a
vial and dried under room temperature conditions for 4~5 days.
5.3.2. Optical characterization
After incorporation into a matrix, the optical characteristics of the colloidal quantum
dots were analyzed. The emission band can be adjusted based on the growth during the
synthesis of the CdSe cores. For that reason, a tolerance for the emission band must be
established. For the creation of pure red emission, wavelengths spanning 620 630 nm
were desired to provide a highly luminous red from the entire emission spectrum, as
emission at 630 nm is twice as luminous as emission > 640 nm as shown in the color
matching function �̅�. The PL emission and absorption of one such dot batch is given in
Figure 5.5. The measured emission is shown to be around 621 nm and possesses a FWHM
of 26 nm. This narrow emission in a luminous region of the spectrum shows strong
luminous flux. In addition, the absorption band overlaps well with the 442 nm excitation
source used. The dots in film also exhibit a PLQE of 67% in film, which was determined to
be sufficiently high for our application where PLQE>60% was required.
Kristopher Bicanic University of Toronto Page 42 of 64
Figure 5.5 | Absorption and PL emission of CQDs in a silica matrix.
5.4. Layer fabrication
With the phosphor and quantum dot material characterized, the model set, and a device
architecture designed, the process for fabricating these devices was established. The main
goal of fabrication was two-fold. The first was to ensure that a low cost was maintained by
not relying on epitaxial growth or high-temperature procedures. The second goal was the
demonstration of a material which could be scaled up easily. These goals were easily
achieved by solution-processeing methods and as such were the first investigated.
5.4.1. Solution-processed phosphor
For the red phosphor, a solution-processed film preparation method was used. The
phosphor was combined with the commercial YAG:Ce which has a mean particle size of
Kristopher Bicanic University of Toronto Page 43 of 64
~10 µm, a peak emission at 550 nm, and a width from 500 nm to 780 nm. This process used
an optically transparent silicone with a refractive index of 1.7. This silicone was shown to
have good index matching with the phosphor material, allowing for better extraction. This
silicone was mixed in low concentrations with the phosphor (13% wt.) to obtain a thick
heavily loaded phosphor paste. The paste was loaded uniformly into a stencil of a desired
shape, size, and thickness. The solution was placed on a 1-inch aluminum mirror, with a
dielectric reflective coating. Using a razor blade, the film was smoothed to give a 250 µm
or 125 µm thickness. This was used to create both mixed and layered films with the desired
thickness. The film was then heated to 70°C for 2 hours to fully cure the silicone.
5.4.2. Solution-processed quantum dots
For the solution-processing of the quantum dots in the film, the method previously
described was also implemented; however, the films produced were not fully cured and so
an alternative method was required. The complications with curing were attributed to
excess chemicals in the quan Thus, screen printing methods
were used with a UV curable adhesive.
The quantum dots, in a silica matrix, were crushed to make a fine powder and mixed
in Norland optical adhesive 81. At this stage the YAG:Ce was added in varying ratios. Using
a silkscreen for processing, and Kapton tape as a mask, the quantum dot and YAG:Ce
solution was loaded and printed onto the high reflectivity aluminum mirror. Placing the
Kristopher Bicanic University of Toronto Page 44 of 64
sample under a UV lamp for 5 minutes was sufficient for curing, and films (~250 µm thick)
were created.
Unfortunately, the quantum dots in the UV curable adhesive showed signs of
degradation when tested at elevated powers. For quantum dots alone a large drop (~6%
PCE) was lost when increasing power up to 3 kW/cm2. After this power level, the quantum
dots were irreversibly damaged due to thermal buildup (Figure 5.6 left). In addition to this,
when protecting the quantum dots with a top capping layer of YAG:Ce phosphor, a drop
in red emission was seen over a time span of 40 minutes (Figure 5.6 right). This indicated
poor thermal transport resulting in very short operation lifetime.
Figure 5.6 | a) Power conversion efficiency measurement showing quantum
dot degradation. b) Operation lifetime of quantum dots in optically
transparent adhesive with YAG:Ce capping layer.
Kristopher Bicanic University of Toronto Page 45 of 64
5.4.3. Solid state film processing
Due to the short comings of quantum dots in the screen printing method, an
alternative matrix was investigated to improve the thermal transport and achieve higher
power. This method utilized a pellet press and sodium chloride. This process was initiated
similarly to the processing quantum dots for screen printing. This involved grinding the
dots in matrix using a mortar and pestle with sodium chloride (anhydrous). Using a
hydraulic press, the sodium chloride, quantum dot, and phosphor mixture was pressed into
a 13 mm diameter pellet using 8 tonnes of pressure. This created a highly scattering film
that was 200 µm thick. By applying a high refractive index optically transparent silicone,
the voids within the pressed pellet were filled with the use of a vacuum. This resulted in
significantly lower scattering and higher extraction of red emission. Using this method a
top coating layer was added of our YAG:Ce phosphor to reduce the fraction of light
incident on the quantum dots and allow long term stable operation at 5 kW/cm2.
5.5. Conclusions
In this chapter an in-depth description into the device fabrication process was
presented. The K2TiF6:Mn4+ phosphor and CQDs in silica matrix were synthesized were
shown to have ideal emission characteristics for our application. Such characteristics
include an intense absorption at the excitation wavelength, narrow PL emission
wavelength, and the correct wavelength for a luminous red source and high PLQE.
Kristopher Bicanic University of Toronto Page 46 of 64
Following this the processes to form the device structures was also described; for the
manganese activated phosphors a solution-processed film was created in silicone. For the
CQDs, an additional fabrication process was required to account for their intrinsic low
thermal conductivity. As such, a process of incorporating the CQDs into a thermally
conductive NaCl matrix was demonstrated, and a significant increase in device stability
was realized.
Kristopher Bicanic University of Toronto Page 47 of 64
Characterization and performance of
phosphor devices
6.1. Introduction
In this chapter, the K2TiF6:Mn4+ phosphor and colloidal quantum dot device structures
will be characterized for their performance in an assembled device structure. The test setup
of this device will be in the high-power laser system described in Chapter 3, resulting in a
sample which operates in back reflection mode, with powers reaching 5 kW/cm2. Their
potential for application in solid state lighting will be investigated in terms of color quality
and efficiency at elevated powers. Following this the device structure will be analyzed for
its quality as a light source for digital projectors.
Kristopher Bicanic University of Toronto Page 48 of 64
6.2. Device performance
As already mentioned, the phosphor samples were fabricated onto a highly reflective
aluminum mirror with a dielectric coating to provide a reflection of 99%; this allows for
the phosphor to efficiently operate in reflection mode (Figure 6.1). This structure allows
for the sample structure to be a densely packed phosphor film, to provide the broadband
emission spectrum of the YAG:Ce.
Figure 6.1 | Laser setup for excitation of phosphor materials and
incorporation of adjustable blue diode source for white light source.
Given the test setup and based on the calculations performed in Chapter 4, we chose to
design the high-power devices on the red and yellow phosphor mixture. For the quantum
dot samples, however, a YAG:Ce capping ceramic was used to reduce incident power and
improve the operation lifetime. This structure allows us to equally distribute the scattering
of both phosphors along the length of the device without placing the red emitter under
intense illumination, avoiding saturation problems. The highest performing phosphor
devices were composed of ~60% K2TiF6:Mn4+ by weight. For quantum dot samples placed
in the NaCl matrix encapsulated with silicone, they operated optimally at 15% QD
Kristopher Bicanic University of Toronto Page 49 of 64
concentration, with a 40% YAG:Ce phosphor by weight (remaining weight was NaCl filler).
The coating for this device was loaded at the highest possible ratio of YAG:Ce to NaCl
(75%-25%). Potential for improvement are present by utilizing a pure YAG:Ce capping
layer pressed and sintered into a ceramic with no NaCl filler.
For the K2TiF6:Mn4+ samples, no degradation was observed when testing the phosphor
alone at high-power. However, when solution-processed techniques were used to test the
CQDs a large degradation was observed at moderate powers. As a result, a stability test was
performed on these devices to ensure color quality and efficacy are maintained. The results
can be shown in Figure 6.2 below. These results indicate that within the 4-hour period
tested, there was no degradation in red emission. In fact, there was a slight increase in red
emission which can be attributed to the filling of trap states.
Kristopher Bicanic University of Toronto Page 50 of 64
Figure 6.2 | CQD 4-hour stability test at 5 kW/cm2.
6.2.1. Correlated color temperature and luminous efficacy
The spectra of these samples (Figure 6.3) were used to calculate the color coordinates
of the phosphor mixtures with different red K2TiF6:Mn4+ concentrations at 5kW/cm2
(Figure 6.4), in addition to the color temperatures (Figure 6.4). Increasing the
concentration of red phosphor raises the saturation point, thereby increasing the overall
red emitted power and by extension the balanced external quantum efficiency. At
K2TiF6:Mn4+ concentrations of 60%, the yellow phosphor still absorbs sufficient light to
maintain its emission intensity at high-power densities. This allows for high external
efficiency, while also allowing the red emission to be increased, thereby correcting the
spectrum (Figure 6.3).
Kristopher Bicanic University of Toronto Page 51 of 64
Contrary to the red phosphor, the CQD samples prepared showed a concentration
dependence, where increasing the percentage past 15% shows diminishing returns. In fact,
a decrease in emissions results when higher concentrations are reached. This decrease in
emission intensity can be attributed to reabsorption due to the high concentration of
CQDs. At this 15% concentration a sufficient level of absorption is present to adequately
alter the emission profile. This is due to the higher absorption cross section of the CQDs
relative to the Mn4+ activated phosphor.
The CCT was determined for each of the samples using the spectra (Figure 6.3). The
CCT obtained indicated negligible fluctuation in the CCT at varying power for the
K2TiF6:Mn4+ phosphor and CQD mixtures, 4369 K and 4248 K, respectively (Figure 6.4
warmest CCT with Duv<0.02 from Planckian locus). This implies that the effect of any
saturation with both K2TiF6:Mn4+ and CQD is insignificant in this mixed system, since the
effective pump power was decreased by the absorption of the yellow phosphor. Due to the
invariance in color emission, the device structure is able to alter successfully the YAG:Ce
white emission closer to that of a warmer black body radiator.
Kristopher Bicanic University of Toronto Page 52 of 64
Figure 6.3 | Spectra of the YAG:Ce compared to the QD and K2TiF6:Mn4+
structures.
Figure 6.4 | CIE color coordinates of the samples at high flux excitation (left)
and corresponding CCT of samples at all power densities, with Duv<0.02
(right).
Furthermore, the devices exhibit high luminous efficacies (Figure 6.5). These efficacies
were calculated by combining the power conversion efficiency with the calculated
luminous flux, obtained from the spectra. When compared to the ceramic YAG:Ce, the
samples show a significantly lower efficacy; however, when increasing the power to 5
Kristopher Bicanic University of Toronto Page 53 of 64
kW/cm2 the samples show a slower decrease in efficacy making the samples relatively
comparable.
Figure 6.5 | Corresponding luminous efficacy of best Mn4+ and CQD
samples at all power densities.
6.3. Device application in display technology
6.3.1. Overview
To operate the phosphor mixture as a source for displays, the spectrum must be split
into green and red sources that meet the Digital Cinema Initiatives (DCI) color space
As described in Chapter 3, the gamut range achieved by the separated
source dictates the range of colors that can be achieved by a given light source. With the
creation of an optical filter to split the emission of the device s source, the white light source
can generate a red and green source. Furthermore, in this section a comparison of the
Kristopher Bicanic University of Toronto Page 54 of 64
generated light source to the industry standard YAG:Ce ceramic phosphor will be made to
determine the luminous efficacy of the emission.
6.3.2. Design consideration and color space
To reach the DCI color space requirements, the white light emission must be separated
into red and green components. To do this a 590 nm long and short pass filter can be
introduced to separate the spectrum into green and red components, while removing a
portion of the yellow emission.
reflection profile was implemented artificially with the spectrum. This was then used to
calculate the red and green sources for the Mn4+ phosphor and CQD samples (Figure 6.6).
The onset of the short and long pass filters was adjusted to accurately obtain the DCI red
point (x: 0.680, y: 0.320) and green point (x: 0.265, y: 0.690). In a digital projector, pixels
are controlled by a digital micro-mirror (DMM) and can control the quantity of red and
green source used. This can be used to adjust the individual source s level to balance the
power and obtain the correct DCI white point, which is a slight modification of the D65
white-point (daylight). This allowed for the full gamut range required in digital projection.
For the devices fabricated, a balanced emission between red and green allowed for the
rendering of a DCI white point with no waste energy.
Kristopher Bicanic University of Toronto Page 55 of 64
Figure 6.6 | Wavelength separation of K2TiF6:Mn4+ phosphor and YAG:Ce
separated into RGB sources
Using these separated spectra an effective luminous flux can be calculated for operating
a projector at the DCI standard. Unlike the devices created, when separating the emission
for the standalone YAG:Ce system, a significant reduction (~49%) of the green emission is
necessary to obtain the correct emission profile for the DCI white point. Similarly, when
compared to the mixture system, we achieved ~21% higher luminous efficacy in the Mn4+
activated phosphor when rendering white light, and ~32% higher luminous efficacy for the
CQD samples.
6.4. Conclusions
In this Chapter, the quantum dot and Mn4+ activated phosphor were successfully
implemented into a device structure with YAG:Ce phosphor. These devices displayed a
large decrease in their CCT of ~650 K, while maintaining a high luminescence efficacy at
high-powers. These devices also showed superior effective luminescence efficacy for
Kristopher Bicanic University of Toronto Page 56 of 64
applications in display technologies when meeting the DCI requirements. Improvements
of 21% and 32% in the effective luminescence efficacies were shown.
Kristopher Bicanic University of Toronto Page 57 of 64
Conclusions and future work
7.1. Summary
We established a design protocol for analyzing device architectures, with the goal of
producing a highly efficient white light source. The effectiveness of this design protocol
was demonstrated through the combination of an efficient YAG:Ce phosphor with a red
emitting K2TiF6:Mn4+ phosphor. The structure was shown to have superior performance
for adjusting the temperature of white light at high fluxes. This enables the device, one
which conventionally would suffer from saturation and scattering of the red emitter, to
exhibit improved color emission as a source for higher power applications. This mixed
phosphor system, when compared to the stand alone YAG:Ce at 5 kW/cm2, provides a 21%-
32% luminescence improvement when used in cinematography; meanwhile, the
illumination is capable of reducing the correlated color temperature to produce a warmer
white light source.
Kristopher Bicanic University of Toronto Page 58 of 64
7.2. Original contributions
1. A design protocol was created for the incorporation of red light emitting
materials into highly efficient broadband source. A one dimensional model was
created to account for the high-power saturation effects. Using known
properties of the material, various device architectures were investigated. This
allowed for the successful design and implementation of a high-power (5
kW/cm2) red emitting source using both a transition metal phosphor and
colloidal quantum dots [35].
2. We report the first solid-state phosphor and CQD system that creates warm
white light emission at powers up to 5 kW/cm2. Furthermore, at this high-
power
requirements with a luminescence efficacy improvement of 32% over the stand-
alone ceramic YAG:Ce phosphor.
7.3. Future work
Potential future directions for this research are as follows:
1. Within this thesis a measurement of the phosphor and colloidal quantum dot
performance was investigated. For future application, it would be important for
the device to display long term stability (shelf life) and a long operation lifetime.
It would be important to demonstrate the full lifetime of the material, longer
Kristopher Bicanic University of Toronto Page 59 of 64
than what has presently been demonstrated. During the synthesis of the
material a shelf time of a month was demonstrated. This can potentially be
improved through encapsulation methods and further refinement of the CQD
in silica matrix process.
2. In addition to the long term stability of the CQDs, an improvement in the
design of the dots should be investigated, to minimize deleterious effects that
occur at high-power, i.e. Auger. Furthermore, manipulating the dots
synthetically offers the potential to stave off the onset of Auger, or the level to
which it affects the system. This would drastically improve the performance of
the CQD device, giving it a bright future in large area display technologies.
Kristopher Bicanic University of Toronto Page 60 of 64
Appendix A
Synthesis of CdSe core QDs: CdSe QDs were synthesized according to a modified
method reported previously [40]. Typically, CdO (240 mg), tri-octylphosphine oxide
(TOPO, 24 g) and octadecylphosphonic acid (ODPA, 1.12 g) were mixed in a Schlenk
flask (100 mL). The mixture was heated to 150 ˚C for 0.5 h under vacuum with continuous
stirring. Then the system was refilled with N2 and heated at 320 ˚C for 2 h. Subsequently,
tri-octylphosphine (TOP, 4 mL) was injected into the Schlenk flask, and the resulting
mixture was further heated to 380˚C. Upon reaching 380˚C, 2 mL of TOP solution
containing selenium (60 mg mL-1) was injected into the system. CdSe QDs with the first
excitonic peak at 586 nm were formed after about 2 min growth. Finally, the reaction was
terminated by cooling and adding acetone. The resultant CdSe QDs were redispersed in
hexane for shell growth.
CdS and ZnS shell growth on CdSe cores: The shell growth was carried out as described
previously [22]. CdSe core QDs were quantified by measuring absorbance (A1st exciton peak)
at exciton peak (586 nm) in a cuvette with path length of 1 mm. A hexane solution of CdSe
core QDs (8.8 mL, A1st exciton peak = 1) was mixed with oleylamine (OLA, 12 mL) and
octadecene (ODE, 12 mL) in a 250 mL Schlenk flask. The mixture was heated to 100 oC
under vacuum to remove hexane. The Cd-oleate and octanethiol were dissolved seperately
in ODE to obtain desired concentations, and then were injected simultaneously and
continuously into the system at a rate of 12 mL h-1 whilst ramping the system temperature
from 100 oC to 310 oC. Different amounts of Cd-oleate (1, 3, 6, and 7.5 mL) and octanethiol
(106, 320, 640, 800 L) were used for growth of different shell thickness.
Kristopher Bicanic University of Toronto Page 61 of 64
To further grow ZnS shell on CdSe/CdS QDs, the above solution was cooled down to
290 oC. 1.5 mL of as-prepared Zn-OLA diluted in 10.5 mL of ODE was mixed with 0.03
g of sulfur dissolved in 2 mL of OLA. The mixture was slowly and continuously injected
into the system for 1 h at 290 oC. After injection, the solution was annealed at 290 oC for
10 min, followed by an injection of 4 mL of oleic acid (OA) and further annealing for 10
min. The reaction was terminated by cooling and adding acetone at 80 oC. The resultant
CdSe/CdS/ZnS QDs were purified by 3 cycles of centrifugation at 6000 rpm, redispersion
in hexane, and precipitation by acetone. The final core-shell-shell CdSe/CdS/ZnS QDs
were redispersed in toluene (A1st exciton peak = 2.5).
Kristopher Bicanic University of Toronto Page 62 of 64
Bibliography
[1] N. C. George, K. A. Denault, and R. Seshadri, "Phosphors for Solid-State White Lighting," Annual Review of Materials Research, vol. 43, pp. 481-501, 2013.
[2] M. S. Shur and A. Žukauskas, "Solid-state lighting: Toward superior illumination," Proceedings of the IEEE, vol. 93, pp. 1691-1703, 2005.
[3] J. Y. Tsao, M. E. Coltrin, M. H. Crawford, and J. A. Simmons, "Solid-State Lighting: An Integrated Human Factors, Technology, and Economic Perspective," Proceedings of the IEEE, vol. 98, pp. 1162-1179, 2010.
[4] E. F. Schubert and J. K. Kim, "Solid-state light sources getting smart," Science, vol. 308, pp. 1274-1278, May 27 2005.
[5] P. Schlotter, R. Schmidt, and J. Schneider, "Luminescence conversion of blue light emitting diodes," Applied Physics A: Materials Science & Processing, vol. 64, pp. 417-418, 1997.
[6] R.-J. Xie, N. Hirosaki, and T. Takeda, "Wide color gamut backlight for liquid crystal displays using three-band phosphor-converted white light-emitting diodes," Applied physics express, vol. 2, p. 022401, 2009.
[7] L. Protesescu, S. Yakunin, M. I. Bodnarchuk, F. Krieg, R. Caputo, C. H. Hendon, et al., "Nanocrystals of cesium lead halide perovskites (CsPbX3, X= Cl, Br, and I): novel optoelectronic materials showing bright emission with wide color gamut," Nano letters, vol. 15, pp. 3692-3696, 2015.
[8] J. Meyer and F. Tappe, "Photoluminescent Materials for Solid‐State
Lighting: State of the Art and Future Challenges," Advanced Optical Materials, vol. 3, pp. 424-430, 2015.
[9] M. Cantore, N. Pfaff, R. M. Farrell, J. S. Speck, S. Nakamura, and S. P. DenBaars, "High luminous flux from single crystal phosphor-converted laser-based white lighting system," Optics Express, vol. 24, pp. A215-A221, 2016/01/25 2016.
[10] C. Hu, Y. Shi, X. Feng, and Y. Pan, "YAG:Ce/(Gd,Y)AG:Ce dual-layered composite structure ceramic phosphors designed for bright white light-emitting diodes with various CCT," Optics Express, vol. 23, pp. 18243-18255, 2015/07/13 2015.
[11] H. K. Park, J. R. Oh, and Y. R. Do, "2D SiNx photonic crystal coated Y3Al5O12:Ce3+ ceramic plate phosphor for high-power white light-emitting diodes," Optics Express, vol. 19, pp. 25593-25601, 2011/12/05 2011.
[12] K. Sakuma, K. Omichi, N. Kimura, M. Ohashi, D. Tanaka, N. Hirosaki, et al., "Warm-white light-emitting diode with yellowish orange SiAlON ceramic phosphor," Optics Letters, vol. 29, pp. 2001-2003, 2004/09/01 2004.
[13] H. Nersisyan, H. I. Won, and C. W. Won, "Highly effective synthesis and photoluminescence of Sr2Si5N8:Eu2+ red-emitting phosphor for LEDs," Chemical Communications, vol. 47, pp. 11897-11899, 2011.
Kristopher Bicanic University of Toronto Page 63 of 64
[14] P. Pust, V. Weiler, C. Hecht, A. Tücks, A. S. Wochnik, A.-K. Henß, et al., "Narrow-band red-emitting Sr[LiAl3N4]:Eu2+ as a next-generation LED-phosphor material," Nat Mater, vol. 13, pp. 891-896, 09//print 2014.
[15] S. E. Brinkley, N. Pfaff, K. A. Denault, Z. Zhang, H. T. Hintzen, R. Seshadri, et al., "Robust thermal performance of Sr2Si5N8:Eu2+: An efficient red emitting phosphor for light emitting diode based white lighting," Applied Physics Letters, vol. 99, p. 241106, 2011.
[16] S. Coe-Sullivan, W. Liu, P. Allen, and J. S. Steckel, "Quantum dots for LED downconversion in display applications," ECS Journal of Solid State Science and Technology, vol. 2, pp. R3026-R3030, 2013.
[17] Y. Yang, Y. Zheng, W. Cao, A. Titov, J. Hyvonen, R. MandersJesse, et al., "High-efficiency light-emitting devices based on quantum dots with tailored nanostructures," Nat Photon, vol. 9, pp. 259-266, 04//print 2015.
[18] V. I. Klimov, "Spectral and Dynamical Properties of Multiexcitons in Semiconductor Nanocrystals," Annual Review of Physical Chemistry, vol. 58, pp. 635-673, 2007/05/01 2007.
[19] N. Ohta and A. R. Robertson, Colorimetry: Fundamentals and Applications, 2006.
[20] "RP 431-2:2011 - SMPTE Recommended Practice - D-Cinema Quality — Reference Projector and Environment," SMPTE RP 431-2:2011, pp. 1-14, 2011.
[21] A. P. Piquette, M. E. Hannah, and K. C. Mishra, "An investigation of self-absorption and corresponding spectral shift in phosphors," ECS Transactions, vol. 41, pp. 1-9, 2012.
[22] H. Zhu, C. C. Lin, W. Luo, S. Shu, Z. Liu, Y. Liu, et al., "Highly efficient non-rare-earth red emitting phosphor for warm white light-emitting diodes," Nat Commun, vol. 5, 07/08/online 2014.
[23] H.-D. Nguyen, C. C. Lin, M.-H. Fang, and R.-S. Liu, "Synthesis of Na2SiF6:Mn4+ red phosphors for white LED applications by co-precipitation," Journal of Materials Chemistry C, vol. 2, pp. 10268-10272, 2014.
[24] B. Wang, H. Lin, J. Xu, H. Chen, and Y. Wang, "CaMg2Al16O27:Mn4+-based Red Phosphor: A Potential Color Converter for High-Powered Warm W-LED," ACS Applied Materials & Interfaces, vol. 6, pp. 22905-22913, 2014/12/24 2014.
[25] D. Sekiguchi, J.-i. Nara, and S. Adachi, "Photoluminescence and Raman scattering spectroscopies of BaSiF6:Mn4+ red phosphor," Journal of Applied Physics, vol. 113, p. 183516, 2013.
[26] M. Peng, X. Yin, P. A. Tanner, C. Liang, P. Li, Q. Zhang, et al., "Orderly-Layered Tetravalent Manganese-Doped Strontium Aluminate Sr4Al14O25:Mn4+: An Efficient Red Phosphor for Warm White Light Emitting Diodes," Journal of the American Ceramic Society, vol. 96, pp. 2870-2876, 2013.
[27] J. E. Murphy, F. Garcia-Santamaria, A. A. Setlur, and S. Sista, "62.4: PFS, K2SiF6:Mn4+: the Red-line Emitting LED Phosphor behind GE's TriGain Technology™ Platform," SID Symposium Digest of Technical Papers, vol. 46, pp. 927-930, 2015.
Kristopher Bicanic University of Toronto Page 64 of 64
[28] H. F. S. Smet, J. J. Jonas, I. D. J. M. Lisa, E. Koen Van den, P. Dirk, and F. Philippe, "Luminescent Behavior of the K2SiF6:Mn4+ Red Phosphor at High Fluxes and at the Microscopic Level," 2016-01-01 2016.
[29] C. C. Lin, A. Meijerink, and R.-S. Liu, "Critical Red Components for Next-Generation White LEDs," The Journal of Physical Chemistry Letters, vol. 7, pp. 495-503, 2016/02/04 2016.
[30] Z. Yiting and N. Nadarajah, "Investigation of Remote-Phosphor White Light-Emitting Diodes with Multi-Phosphor Layers," Japanese Journal of Applied Physics, vol. 49, p. 100203, 2010.
[31] J. P. You, N. T. Tran, and F. G. Shi, "Light extraction enhanced white light-emitting diodes with multi-layered phosphor configuration," Optics Express, vol. 18, pp. 5055-5060, 2010/03/01 2010.
[32] J. C. de Mello, H. F. Wittmann, and R. H. Friend, "An improved experimental determination of external photoluminescence quantum efficiency," Advanced materials, vol. 9, pp. 230-232, 1997.
[33] D. Y. Kang, E. Wu, and D. M. Wang, "Modeling white light-emitting diodes with phosphor layers," Applied Physics Letters, vol. 89, 2006.
[34] H. Bode, H. Jenssen, and F. Bandte, "Über eine neue Darstellung des Kalium-hexafluoromanganats(IV)," Angewandte Chemie, vol. 65, pp. 304-304, 1953.
[35] K. T. Bicanic, X. Li, R. P. Sabatini, N. Hossain, C. Wang, F. Fan, et al., "Design of phosphor white light systems for high power applications," Submitted, 2016.