Investigation of Color Phosphors for Laser-Driven White Lighting. A thesis presented to. the faculty of. In partial fulfillment

Transcription

1 Investigation of Color Phosphors for Laser-Driven White Lighting A thesis presented to the faculty of the Russ College of Engineering and Technology of Ohio University In partial fulfillment of the requirements for the degree Master of Science Sara S. Al-Waisawy December Sara S. Al-Waisawy. All Rights Reserved.

2 2 This thesis titled Investigation of Color Phosphors for Laser-Driven White Lighting by SARA S. AL-WAISAWY has been approved for the School of Electrical Engineering and Computer Science and the Russ College of Engineering and Technology by Wojciech M. Jadwisienczak Associate Professor of Electrical Engineering and Computer Science Dennis Irwin Dean, Russ College of Engineering and Technology

3 3 ABSTRACT AL-WAISAWY, SARA S., M.S., December 2014, Electrical Engineering Investigation of Color Phosphors for Laser-Driven White Lighting Director ofthesis: Wojciech M. Jadwisienczak Solid-state white lighting devices (SSWLDs) commonly use III-nitride near-uv or blue light emitting diodes (LEDs), combined with one or more phosphors, to generate white light. These devices already offer many advantages over traditional incandescent and fluorescent light sources, including long lifetimes, environmentally friendly designs without the need for mercury, and enormous energy savings. Despite unquestionable recent commercial success and the future potential for further development, current SSWLDs suffer from the droop effect limiting the overall efficacy and a thermallyinduced shift in the peak emission wavelength of the phosphor. Thus, the overall efficiency of these devices can still be improved. One such example is to control the operating temperature of the device. When operating an LED, the temperature inevitably increases, yet the phosphor particles exhibit a loss in efficiency as the temperature of the device increases. In addition, LEDs suffer from efficiency loss and color instability with increased operating current, making high-power devices not achievable using LEDs as the excitation source. Recently, a new concept for developing SSWLD, based on laser diode (LD) substituting for LED as a pump source for exciting colour-stable phosphors, was proposed. In contrast to LEDs, laser diodes do not exhibit efficiency loss; many exhibit increased efficiency as current increases, and maintain color stability. Thus, there is a need in the art for improved solid-state white lighting devices that rely on laser

4 4 diodes. In this project we have characterized individual Eu(WO 4 ) 2 (red phosphor), BaMg 2 Al 16 O 27 :Eu,Mn (green phosphor) and (Sr,Mn) 2 SiO 4 :Eu (blue phosphor) and trichromatic white light phosphors. Characteristics of light produced with each phosphor, variations with incident light power and phosphor temperature, as well as effects from phosphor ageing, are described. Results of comparison between pumping with coherent and incoherent light at the same wavelength are also described.

5 5 ACKNOWLEDGMENTS I thank God for giving me the courage and determination, as well as guidance in conducting this research study, despite all difficulties. I also extend my heartfelt gratitude to my supervisor, Dr. Wojciech M. Jadwisienczak, for encouraging me to work hard and conduct such important and interesting research. I am lucky to work under his supervision since He has opened my eyes on very amazing areas of optoelectronics science and technology. He has increased my motivation about this rapidly developing field. Further, I introduce my sincere thanks to my sponsor HCED. I would like to express my gratitude to Dr. Faiz Rahman for all the help he has offered toward my research. His advices are worthwhile for my research training as well as my future professional career plans. Moreover, I wish to extend my thanks and appreciation to my thesis committee members, Dr. Wojciech M. Jadwisienczak, Dr. Savas Kaya, Dr. Faiz Rahman, and Dr. Eric Stinaff for reviewing my thesis document and providing suggestions. Additionally, I would like to thank Phosphor Technology Limited, England, U.K., for providing us with the RGB phosphor powders for this project. I express sincere thanks for Jason Wright who did the scanning of electron micrographs of studied RGB phosphor particles. I also appreciate the assistance from Mohammed Bsatee in laboratory experimental work, and I am glad to see that my research can be extended by him for magneto-optical studies of RGB phosphors. Finally, I thank my parents for their unconditional support with my studies, and for giving me a chance to improve myself through my studying abroad.

6 6 TABLE OF CONTENTS Page Abstract... 3 Acknowledgments... 5 List of Tables... 8 List of Figures... 9 Chapter 1: Introduction Successful Phosphors Using Transition Metals and Rare Earth Elements Novel Technologies Using Phosphors Challenges for Novel Phosphors Used in Solid State Lighting (SSL) Chapter 2: Color Science Understanding the Chromatics Photopic Vision Scotopic Vision CIE 1931 Chromaticity System Correlated Color Temperature (CCT) Color Rendering Index (CRI) Luminous Efficacy of Radiation (LER) Human Physiological Aspect of Light Chapter 3: Phosphor Excitation Chapter 4: Experimental Setups Samples Information White Phosphor Preparation Photoluminescence Spectra Measurement CIE Chromaticity Coordinates vs. Different Excitation Power Density Measurement CIE Chromaticity Coordinates vs. Different Temperature Measurement Characterization of Phosphors Using Coherent and Non-Coherent Excitations... 45

7 7 Chapter 5: Results and Discussion Chapter 6: Conclusions Chapter 7: Future Work References Appendix A: List of Conferences Presentations and Journal Papers Appendix B: Derivation of the Color X, Y, Z System from Color R, G, B System Appendix C: Software Procedures for Measuring Phosphor Emission Spectra Appendix D: Color Measurement Procedures... 92

8 8 LIST OF TABLES Page Table 5-1: RGB phosphors particle size distributions as measured from an ultrasonic dispersion by a Coulter counter method

9 9 LIST OF FIGURES Page Figure 2-1: Spectral sensitivity curves of cones in human eyes with photopic vision [46] Figure 2-2: Spectral luminous efficiency curve for high level vision (photopic). This figure was reproduced from Ref. [40] Figure 2-3: Spectral luminous efficiency curve for low level vision (scotopic). This figure was reproduced from Ref. [40] Figure 2-4: Color matching function for CIE 1931 RGB with a 2 standard observer. This graph was reproduced from Ref. [48] Figure 2-5: CIE 1931 xʹ, yʹ, zʹ color matching function based on 2 observers. This graph was reproduced from Ref. [40] Figure 2-6: CIE 1931 (x, y) chromaticity diagram. This graph was reproduced from Refs. [40],[48] Figure 2-7: Color temperature of blackbody radiations [53] Figure 2-8: Comparison of three different correlated color temperature light-emitting diodes [55] Figure 2-9: Comparison of same object illuminated by light sources with low and high CRI [58] Figure 3-1: Schematic representation of resonant energy transfer process between D (donor) and A (acceptor). This graph was reproduced from Ref. [70] Figure 3-2: Electronic exchange of resonant energy transfer between D and A with electronic wave-functions of D and A overlap. This graph was reproduced from Ref. [70] Figure 4-1: Schematic diagram of power measurement setup using photo-detector Figure 4-2: Experimental setup diagram used for phosphors emission characterization. 42 Figure 4-3: Experimental setup diagram used for characterization of phosphors response induced by coherent and non-coherent excitations

10 Figure 5-1: Scanning electron micrographs of the red (R) (a), green (G) (b) and blue (B) (c) phosphor particles Figure 5-2: Room temperature photoluminescence excitation (PLE) spectra of (a) KEu(WO 4 ) 2 red phosphor monitored at 615 nm, (b) BaMg 2 Al 16 O 27 :Eu,Mn green phosphor monitored at 515 nm, and (c) (Sr,Mg) 2 SiO 4 :Eu blue phosphor monitored at 460 nm. Photoluminescence (PL) spectra were measured for each phosphor, with excitation photons energy corresponding to the maximum of each individual PLE spectrum. All spectra were normalized to the same intensity. The dashed line indicates the position of LD excitation wavelengths at 404 nm conducted in the color characteristics measurements. Please notice the different horizontal scale in segment (a) Figure 5-3: Room temperature photoluminescence spectra and CIE chromaticity coordinates of the R phosphor. The 404 nm laser excitation at a power level of 30.3 mw/mm 2 was used for these measurements. The black curve on CIE chromaticity chart shown in inserts is the Planckian locus, showing the chromaticity coordinates of an ideal black body at different temperatures Figure 5-4: Room temperature photoluminescence spectra and CIE chromaticity coordinates of the G phosphor. The 404 nm laser excitation at a power level of 30.3 mw/mm 2 was used for these measurements. The black curve on CIE chromaticity chart shown in inserts is the Planckian locus, showing the chromaticity coordinates of an ideal black body at different temperatures Figure 5-5: Room temperature photoluminescence spectra and CIE chromaticity coordinates of the B phosphor. The 404 nm laser excitation at a power level of 30.3 mw/mm 2 was used for these measurements. The black curve on CIE chromaticity chart shown in inserts is the Planckian locus, showing the chromaticity coordinates of an ideal black body at different temperatures Figure 5-6: Room temperature photoluminescence spectra and CIE chromaticity coordinates of the trichromatic white phosphor. The 404 nm laser excitation at a power level of 30.3 mw/mm 2 was used to excite the phosphor. The black curve on CIE chromaticity chart shown in inserts is the Planckian locus, showing the chromaticity coordinates of an ideal black body at different temperatures Figure 5-7: CIE chromaticity diagram showing position of CIE (x, y) coordinates for RGB phosphors obtained for temperature and excitation power density experiments. Dashed lines link the RGB points defining the scope of accepted CIE (x, y) coordinate variation for obtaining white light Figure 5-8: (a) Variation of the CIE (x, y) coordinates for the R(T) phosphor as the function of temperature. (b) Change of the KEu(WO 4 ) 2 phosphor dominant peak intensity and peak potion as the function of temperature

11 Figure 5-9: (a) Variation of the CIE (x, y) coordinates for the G(T) phosphor as the function of temperature. (b) Change of BaMg 2 Al 16 O 27 :Eu,Mn phosphor dominant band intensity and peak potion as the function of temperature Figure 5-10: (a) Variation of the CIE (x, y) coordinates for the B(T) phosphor as the function of temperature. (b) Change of (Sr,Mg) 2 SiO 4 :Eu phosphor dominant band intensity and peak potion as the function of temperature Figure 5-11: Variation of CIE chromaticity coordinates for silica-bound RGB and trichromatic white phosphors with 404 nm laser pumping as the function of phosphor temperature Figure 5-12: (a) Variation of the CIE (x, y) coordinates for R(P) phosphor as the function of excitation power density. (b) Change of KEu(WO 4 ) 2 phosphor dominant peak intensity and peak potion as the function of excitation power density Figure 5-13: (a) Variation of the CIE (x, y) coordinates for the G(P) phosphor as the function of excitation power density. (b) Change of BaMg 2 Al 16 O 27 :Eu,Mn phosphor dominant peak intensity and peak potion as the function of excitation power density Figure 5-14: (a) Variation of the CIE (x, y) coordinates for the B(P) phosphor as the function of excitation power density. (b) Change of (Sr,Mg) 2 SiO 4 :Eu phosphor dominant peak intensity and peak potion as the function of excitation power density Figure 5-15: Variation of CIE chromaticity coordinates for silica-bound RGB and trichromatic white phosphors with 404 nm laser pumping as the function of excitation power density Figure 5-16: Phosphor luminescence intensity variation with pump power density for (a) red (R), (b) green (G), (c) blue (B) and (d) trichromatic white phosphors. Measurement results, for each case, are shown for ordinary one-sided illumination (filled circles) and for two-sided illumination (filled squares), using a 404 nm wavelength-selective mirror (notch filter) Figure 5-17: Trichromatic white phosphor emission intensity change when excited at 404 nm as the function of incident light power with the laser speckle reducer switched off (coherent pumping) and switched on (incoherent pumping) Figure 5-18: RGB phosphors ageing effect after 400 hours of heating at 120 C. The intensity emission spectra were measured periodically

12 12 CHAPTER 1: INTRODUCTION The demand for substitutional light sources in daily life arises from a lack of unlimited natural energy sources and environmental problems [1]. Currently, the world's major sources of energy come from petroleum, charcoal, and natural gas. These fossil fuels are unrenewable, and they will run out in the future. Also, extracting energy from these sources has been causing an increase in the amount of undesirable elements (carbon dioxide, nitrogen oxides, sulfur dioxide, etc.) in nature. Therefore, finding clean renewable energy sources and reducing our energy consumption have been significant research areas in recent years [2], [3]. A promising way for saving energy needed for general lighting purposes is by replacing conventional incandescent and fluorescent light sources with more efficient light sources, because approximately 25% of all energy consumed globally goes for light generating. Moreover, the traditional incandescent light bulbs have very low efficiency (about 5%) for producing visible light, such that most input electric energy gets converted to infrared energy (heat). Despite the fluorescent lamp s relatively high efficiency (about 40%) [4], it is still considered as wasting energy due to a Stokes shift in the phosphor-covered discharge gas tube [3]. By producing a bright blue LED based on (Ga,In)N in the 1990s, the idea of creating a white light source by using light emitting diodes (LEDs) became practicably viable [4], [5]. Solid-state white light emitting diodes (WLEDs) are a prominent new generation of optoelectronic devices having the prospect to replace the conventional light sources due to the many attractive features of WLEDs such as: low consumption of electrical power, long lifetime,

13 13 high luminous efficacy, and contribution to carbon dioxide emission reduction [1], [6], [7]. Fundamentally, the simplest way to generate white light in solid state lighting is using multiple monochromatic LEDs. These LEDs are semiconductor devices. They basically work on electroluminescence phenomena. When the semiconductor material (pn junction) is forward-biased, charge carriers (electrons and holes) are injected into the p- n junction. Then, they recombine with each other, so that their energy is released as a photon (light) [5], [8], [9]. The wavelength (color) of the photon is dependent on the band gap of the semiconductor. Therefore, in principle, the LED can be a nearly monochromatic light source [4]. Although the first LED was fabricated in the 1960s, these LEDs did not have high brightness suitable for generating white light until the 1990s, when S. Nakamura of Nichia Corp. used (Ga,In)N material to produce a bright blue LED [4], [5]. Use of III-nitrides semiconductor materials opened the door to produce very bright LEDs, and created an opportunity for a new generation of white light sources [5], [8]. However, producing red and green III-nitride LEDs as bright as blue III-nitride LEDs remains a challenge [5]. The problems of producing red and green wavelengths using III-nitrides are related to difficulty in fabricating InGaN ternary alloys with sufficient In content. By increasing the indium content, the InGaN become inhomogeneous and resulting emission intensity is reduced due to the drooping effect. Therefore, red and green LEDs are less bright than blue LEDs [4]. This low efficiency restricts manufacturing of high efficiency WLEDs using multiple color LEDs based on III-nitrides [4], [5]. Furthermore, it is known that existing multiple color LEDs in the

14 14 same device require complex electronic circuits to control the power supply for each individual LED [5]. Additionally, there are two other strategies for producing white light in solid state lighting. One method is by using a yellow emission phosphor that is excited by blue LED, and the second one is by using mixed RGB phosphors that are excited by near UV- LED [4], [5], [9], [10]. However, the white light which is created from the blue LED with yellow phosphor has a poor color rendering index, and it is typically perceived as cool light because of the lack of a red spectral component [4], [6], [10]. Therefore, using a near UV-LED chip coated with RGB phosphors is the most successful technique these days for fabricating white light with a broad emission spectrum and a high color rendering index [1], [4]. Even though the latter method generates good quality white light, it still has not reached the expectations for solid-state lighting to generate artificial white light that is as close as possible to natural sun light [4]. Additionally, producing high quality warm WLEDs for general lighting is restricted by the low efficiency of available red phosphors when excited by blue or near UV wavelengths [1], [4]. Also, most commercially available and successful phosphors like [Y 2 O 3 :Eu, BaMgAl 10 O 17 :Eu 2+, (Ce,Tb)MgAl 11 O 19 ] [11], [12] are used extensively with fluorescent tubes and compact fluorescent lamps (CFL) [13]. However, some of these phosphors currently are employed with LED excitation light sources (e.g. BaMgAl 10 O 17 :Eu 2+ ) [6]. For these reasons, much research has been focused on this interesting scientific area of discovering novel phosphors that can be used with blue and near UV LED excitation sources [4], [14].

15 15 It is known that LEDs presently used in solid-state lighting suffer from a droop effect reducing luminous efficacy with an increase of drive current. There is also excess heat generation in modern III-nitride LEDs [7], [15]. Typically, heat may cause changes in the peak and width of spectral emission from LED chips. Therefore, the color emission characteristics from the phosphor layer are influenced by the altered excitation spectrum and by the excess heat of the device [16]. Currently, there is a proposal to use laser diodes (LDs) instead of LEDs for the generation of white light in solid state lighting technology. The idea is capitalizing on the fact that the luminous efficacy of LD increases linearly with the drive current and that the phosphor layer is placed far from the LD. Thus, this latter use results in reduced temperature increase during the LED operation [16], [17]. This thesis focuses on the characterizations of selected commercial phosphors under continuous wave violet LD (404 nm) excitation for the purpose of generating white light. Specifically, in this project the CIE 1931 (CIE means Commission International de l Eclairage, International Organization of Color Science) [18] parameters of monochromatic color light phosphors were studied at different laser excitation power densities, various excitation light spectroscopic characteristics, and different phosphor sample temperatures. This document is organized as follow: (1) in chapter one, we review the most successful phosphors with transition metals and rare earth elements, novel technologies using phosphors, and challenges for novel phosphors used in SSL; (2) in chapter two, we focus on color science, the CIE 1931 chromaticity system, correlated color temperature

16 16 (CCT), color rendering index (CRI), luminous efficacy of radiation (LER), and human physiological aspects of light and color mixing; (3) chapter three describes general theory of the phosphor excitation process, (4) chapter four describes samples and experimental setups; (5) in chapter five, we discuss the experimental results; (6) chapter six contains the conclusion of this thesis; and (7) chapter seven contains proposals for future work. 1.1 Successful Phosphors Using Transition Metals and Rare Earth Elements For more than a century, scientific research about phosphors has been in progress, but the importance of these materials has increased significantly with growing modern technology. Therefore, many applications require searching for novel phosphors with desirable qualities [19]. However, luminescent materials (here phosphors) are generally inorganic white powders composed from a host lattice [20] which is typically a wide band gap material (more than 3 ev) [21] [23] doped with impurities called sensitizers and activators [11], [20], [21], [24]. Phosphors emit photons after being excited by certain types of energy (photon, electron, cathode ray, etc.) [21], [25]. The sensitizers and activators form luminescence centers with certain concentrations in the host lattice [20], [24]. If an activator luminescence center is not capable of absorbing the excitation light efficiently, then selected sensitizers (impurities or a host lattice) are selected to stimulate energy transfer to activators generating luminescence [12], [20], [26]. Moreover, the phosphors based on rare earth (RE) and transition metal (TM) ions are widely utilized for many phosphor applications, especially for solid state lighting. These ions typically offer high absorption efficiency of near UV light (λ = nm) and blue light (λ =

17 nm). Upon photon absorption, an electronic transition occurs in RE 3+ ion 4f shells or between 5d-4f shells of TM. [3]. Since the 5s5p shells shield the 4f shell in the RE ion, the transition between f-f shells is forbidden and less influenced by the environment giving emission spectra characteristic sharp lines. When an RE ion is in a (2+) ionic state, a broad band emission spectrum in the RE 2+ ion occurs from the d-f transition because this transition is allowed and affected strongly by surrounding environment [11], [21], [26], [27]. Likewise, the d-d transition of TM ions generates wide spectrums, and typically such transition is affected strongly by the host lattice [3], [11]. Examples of successful commercial phosphors used in SSL are as follows: YAG:Ce 3+ (garnet phosphor group) emitting yellow light at peak (~540 nm) when excited by blue light (~460 nm); the YAG:Ce 3+ emitted spectrum is broad band due to the transition between 4f-5d in the Ce 3+ ion [3], [5]. Another commercial yellow emission phosphor is LaSr 2 AlO 5 :Ce 3+ (oxide phosphor). When it is excited by blue light at 450 nm, a broad yellow emission band is observed with peak maximum at 556 nm [3], [5]. The nitride phosphors doped with Eu 2+ (M 2 Si 5 N 8 where M = Ca, Sr, or Ba) absorb in a wide spectral excitation range from 300 nm to 500 nm and with a luminescence peak at 395 nm. The oxynitride phosphors (MSi 2 O 2 N 2 where M = Ca, Sr, Ba) doped with Eu 2+ ion have a broad excitation spectrum, ranging from 370 to 460 nm. The MSi 2 O 2 N 2 doped with Ca and Ba emits yellow peak at 560 nm and blue peak at 500 nm, respectively. The MSi 2 O 2 N 2 doped Sr releases green-yellow emissions from 530 to 570 nm; however, it depends on the concentration of Eu 2+ ion and the ratio of O/N [5]. Because of the commercial success of these phosphors, initially, there was an attempt to use them for

18 18 SSL technology. However, fundamental differences in excitation mechanisms involved with these phosphors do not meet all specifications required by modern SSL devices. Thus, the need for new phosphors suitable for SSL has rapidly increased. There also exists a necessity for developing and studying phosphors suitable for SSL using LDs to pump the phosphors. 1.2 Novel Technologies Using Phosphors The use of phosphors has increased more and more with the development of modern light generating applications. Solid state lighting (SSL) presently creates white light from yellow or RGB phosphors pumped by blue or near UV InGaN LED. Since the laser pumping may solve the problems related to LED pumping of yellow phosphor, recently, there has been an attempt to generate high efficiency stable white light using the LD-phosphor approach [16]. Another interesting idea for using phosphor is to combine it with solar cells to improve the efficiency and performance of photovoltaics. Most of the incident light on photovoltaics is not absorbed because of the photons mismatching the semiconductor s band gaps. Therefore, the energy of incident photons is dissipated as heat and reduces the efficiency of photovoltaics. By using down conversion or up conversion phosphors, the wavelengths (energy) of photons can be shifted to match the energy band gaps of the semiconductor [2]. Moreover, a high energy photon can be converted into two energy photons which fit the band gap by using down conversion phosphors with increased quantum efficiency [2], [28]. Furthermore, new phosphors have been applied in many other devices, and the efficiency of these instruments depends on

19 19 the properties of the phosphor. For example, novel phosphors are used in plasma display panels (PDPs), cathode ray tubes (CRTs) [29], improved fluorescent lamps [30], X-ray radiography, digital imaging [31], traffic signs, printers, clocks, emergency signs [32], computed tomography (CT), and gamma radiation detectors [33]. Also, thermo-graphic phosphors can be utilized to measure the temperature of remote and internal objects such as combustion of internal engines of gas turbines by using laser induced phosphorescence (LIP) measurements [34]. 1.3 Challenges for Novel Phosphors Used in Solid State Lighting (SSL) Each application requires a kind of phosphor that has features different from the one that is used for other applications. For instance, solid state lighting (SSL) needs phosphors with high absorption of excitation energy, high quantum efficiency, good luminous efficacy, and high thermal and chemical stabilization [23]. Even though the white light produced from LEDs with a phosphor layer has a very high luminous efficacy, this luminous efficacy is decreased with the increase of the drive current of the LEDs [35]. LEDs suffer from droop efficiency at high input current [7], [16], [35], [36]. This heat causes the emission from LEDs to change, therefore the CIE parameters of phosphor layer and white light change [36], [16]. The efficiency and luminous efficacy of light are decreased with increasing temperature [5], [16]. Another challenge of using phosphor in SSL is a Stokes shift loss which makes red (R) phosphor have low efficiency (high Stokes shift) compared with blue (B) and green (G) phosphors [1], [37]. Stokes shift is the energy (wavelength) difference between the exciting energy (wavelength) to

20 the emitting energy (wavelength) [5], [17]. Furthermore, there are many other parameters affecting the brightness and quality of white light generated from phosphors. For 20 example, homogeneity of a monochromatic phosphor layer or of mixed phosphor powders is important, since adjusting the RGB ratio in a mixture can produce white light with desired low CCT [1]. Also, the thickness of a phosphor layer controls the amount of light from the excitation source passing through the phosphor, so that the quality of white light depends on the layer s thickness as well[1], [38].

21 21 CHAPTER 2: COLOR SCIENCE 2.1 Understanding the Chromatics It is interesting why some objects are white or black or red or another color, and why the color of an object is seen in daylight while the color of the same object cannot be viewed during the night. Color is a visual stimulus that is detected by the eyes, and the brain analyses the light that reaches our eyes from the objects in the environment [39]. Scientific awareness about colors and the mixing of different colors to produce new color came from the experiments of Isaac Newton in 1666, when he analyzed sunlight into its essential color components [18], [40]. He discovered that sunlight (white light) consists of a spectrum of colors ranging from red, orange, yellow and green to blue and violet [40]. After that, no further developments took place in color science until 1801, when Thomas Young mentioned that the receptors in the retina are stimulated by three primary colors: red, yellow, and blue. Furthermore, scientists found out that all the various colors are derived from mixtures of three primary colors: red-green-blue (RGB) [41], [42]. Color vision depends on three main factors. These are the light sources, the objects which are illuminated by the light sources, and observers. When light from a source with energy (E = hv) falls on an object, then depending on the physical properties of its surface (reflection, transmission, absorption, etc.), the object reflects those wavelengths that represent its color in order to excite the visual system of an observer [43]. Furthermore, the paramount element in seeing the color is the observer [44]. In the

22 following two subsections, we will explain how human eyes analyze the color in photopic and scotopic vision Photopic Vision Human eyes contain three kinds of cones and one kind of rod receptor in the retina [45]. The β cones have (S) responses to short wavelengths (blue); the γ or (M) cones respond to medium wavelengths (green); and ρ or (L) cones respond to long wavelengths (red) [45]. Figure 2-1 represents the sensitivity curves of these three cones S, M, and L in the retina of our eyes. Figure 2-1: Spectral sensitivity curves of cones in human eyes with photopic vision [46]. These cones are sensitive to light at normal levels of illumination such as daylight and indoor artificial light. When the luminance of the stimulus is several (cd/m 2 ) or more,

23 23 photopic vision is achieved through these cone receptors. The highest sensitivity is in the green part; therefore, we see the green color as brightest [40]. Figure 2-2 shows the luminous efficiency of our eyes at a normal photopic illuminated light source. The peak of the curve is at 555 nm (green) where the human eyes are very sensitive, and the sensitivity decreases towards shorter and longer wavelengths. 1 spectral luminous efficiency at photopic vision 0.8 relative sensitivity wavelength (nm) Figure 2-2: Spectral luminous efficiency curve for high level vision (photopic). This figure was reproduced from Ref. [40]. The sensitivity of human eyes is different with high intensity illumination than with low intensity illumination. They have their sensitivity peak at shorter wavelengths than 555 nm in scotopic condition. There is more detail about scotopic vision in the section below.

24 Scotopic Vision The rods receptors are sensitive to light at low levels of illumination. For example, starlight, moonlight, etc. These rods are used to form scotopic vision of objects [40]. Figure 2-3 represents the luminous efficiency of our eyes for scotopic vision. Under this scotopic vision condition, the sensitivity curve peaks at 510 nm, and it drops far from 510 nm until it becomes zero for wavelengths shorter than 400 nm and longer than 620 nm. 1 spectral luminous efficiency at scotopic vision 0.8 relative sensitivity wavelength (nm) Figure 2-3: Spectral luminous efficiency curve for low level vision (scotopic). This figure was reproduced from Ref. [40]. In the past, it was published that all the cones are located in the foveola in the retina within a 2 arc, and that the rods are located outside of this area. Since 1960, it has been believed that the field of view is larger than a 2 arc. It is within a 10 arc in the foveola [40], [47]. Then, the receptors connect to nerve fibers which join together to

25 25 compose the optic nerve which communicates the eye to the brain [40]. Since light is the only way to stimulate the vision system, there are many concepts which define the qualifications of a light source required to use it for indoor application. The system which examines the characteristics of artificial light source is designed by a CIE in CIE 1931 Chromaticity System Finding measurement instruments equivalent to the human eyes for seeing and analyzing color was the motivation of the CIE to recommend a mathematical method to match the color of all wavelengths in the visible spectrum [18]. Based on the experiments conducted by Wright and Guild in the 1920s to find the equations for matching colors, in 1931, the CIE discovered the first color matching function corresponding to a 2 observer [42]. Figure 2-4 indicates the CIE 1931 RGB color matching function. It is explained at the bottom of Fig. 2-4.

26 r' g' b' 0.25 tristimulus values wavelengths (nm) Figure 2-4: Color matching function for CIE 1931 RGB with a 2 standard observer. This graph was reproduced from Ref. [48]. In Fig. 2-4, the rʹ(λ), gʹ(λ), bʹ(λ) curves represent the amount of red, green, and blue colors required to match the color of each wavelength in the visible spectrum. The area under the curve for these rʹ(λ), gʹ(λ), bʹ(λ) curves is equal to each other, which represents equal power white light generated by mixing these RGB amounts. The negative parts in the curves represent the need for adding one of the matching stimuli to a test color. These negative parts arise because of the overlap between β and ρ cones, so that the γ cones are never stimulated alone. For example, the test color of light at 470 nm can be matched by mixing some amount of B and G only by adding an amount of R to the test color. In order to avoid the negative parts, the CIE converted later the rʹ (λ), gʹ (λ), bʹ (λ) color matching function into xʹ (λ), yʹ (λ) and zʹ (λ) color matching function [40]. This color matching function is shown in Fig. 2-5, which was reproduced from Ref. [40].

27 x' y' z' wavelengths (nm) Figure 2-5: CIE 1931 xʹ, yʹ, zʹ color matching function based on 2 observers. This graph was reproduced from Ref. [40]. In Fig. 2-5, all the curves become positive to eliminate any error as seen in Fig. 2-4 because it is not meaningful to plot a color curve in the negative region. Even though Fig. 2-5 matches any color within visible wavelengths, it does not show the direction of a color produced from different focus mixing of tristimulus colors (RGB). For example, the equal power white light (equal amount mixing of RGB) is not seen on Fig In a color system, it is important to plot the direction of color. Therefore, the CIE derived a two-dimensional chromaticity diagram from the color matching function in Fig. 2-5 [41]. Appendix B contains the equations and procedures to derive the CIE 1931 (x, y) chromaticity diagram from the first color matching function for the CIE 1931 RGB. the chromaticity diagram is shown in Fig. 2-6 below.

28 28 Figure 2-6: CIE 1931 (x, y) chromaticity diagram. This graph was reproduced from Refs. [40],[48]. Figure 2-6 represents the color map for whole colors in the visible range, and the ideal white light point ( equal energy ) is located close to the middle of the CIE 1931 (x, y) chromaticity diagram with coordinates (x=1/3, y=1/3). 2.3 Correlated Color Temperature (CCT) CCT is the temperature of an artificial light source which has chromaticity coordinates close to the Planckian locus [49], [50], [48]. The Planckian locus is the line connecting the chromaticity coordinates of the light emitted from a blackbody radiator corresponding to its temperature. This locus is located close to the center of the chromaticity diagram [48], [51]. It starts from low temperature (less than 2000 K) as a

29 deep red, and then becomes orange, yellow, white, blue (more than K) with increasing temperatures of the blackbody [51], [52], [50]. This can be seen in Fig Figure 2-7: Color temperature of blackbody radiations [53]. However, a source that produces light by heating up like an incandescent lamp has the same temperature as a black-body radiator, and the color coordinates of the lamp s light is then on the Planckian locus [50]. The light emitted in other ways and very close to the Planckian locus has CCT, and there is also a neglected CCT of the light that is far from the Planckian locus, such as green light [51]. Light with a low color temperature is considered a warm color (reddish), and light with a high color temperature is considered a cool color (bluish) [54]. Figure 2-8 shows the white color of light-emitting diodes with different color temperatures. A warm white LED (2000 K~ 4000 K) has more red component in color mixing of white light. A natural white LED (4000 K~ 5000 K) has almost equal RGB color in the white light, and the white point is located in the center of the white region in the chromaticity diagram. A cool white LED (5500 K ~ K) has more blue color in components of white light generation.

30 30 Figure 2-8: Comparison of three different correlated color temperature light-emitting diodes [55]. In addition to correlated color temperature (CCT), there is another factor to characterize a light source called the color rendering index (CRI). 2.4 Color Rendering Index (CRI) In the lighting industry, CRI is one of the most important parameters used to evaluate the quality of the light source. The CRI is defined as how the color of objects appears to human eyes when these objects are illuminated by a test light source compared with the color of the same objects illuminated by a reference light source (black body radiator with CRI =100%) [40], [56]. Moreover, the CIE has introduced a standard definition of color rendering index, which is the effect of an illuminant on the color appearance of an object by conscious or subconscious comparison with their color appearance under a reference illuminant. [40]. However, the light source should have a color temperature equal or close to the color temperature of the reference light source

31 31 [48], [57], so that the CRI exists only for a light source with chromaticity coordinates on or near the Planckian locus (only for white or near white light). The CRI of a light source does not depend on its CCT. The CRI is measured by considering the differences in appearance of color objects under standard and artificial light sources. Therefore, these differences can be the same for two light sources having different CIE chromaticity coordinates (different CCT) [40]. The CIE determines the CRI of a generating light source using eight colored samples, and illuminates them with a CIE standard light source and then with a light source for which to determine the CRI. According to the Eq. 2.1, the special color rendering index for each sample (R i ) is calculated using the formula R i = d i (2.1) where di is the difference of the color sample under both illuminations. The average of R i gives the general color rendering index (R a ) or (CRI) as seen in Eq. 2.2 R a = [ d 1 + d 2 + d 3 + d 4 + d 5 + d 6 + d 7 + d 8 ] (2.2) 8 Further, the CIE uses additional six colored samples to get a more accurate determinate color rendering index of the light source [40]. As in any mathematical process, inserting more points gives better results. It can be seen in Fig. 2-9, how the object is seen by human eyes when it is illuminated by LED light sources with differing CRIs.

32 32 Figure 2-9: Comparison of same object illuminated by light sources with low and high CRI [58]. The Figure 2-9 indicates how the CRI of the light source is important for reflecting the real color of objects. Additionally, the factor that defines how well a light source generates visible light rather than other wavelengths is called the luminous efficacy of radiation (LER). 2.5 Luminous Efficacy of Radiation (LER) The luminous efficacy of radiation (LER) is equally important as the color rendering index (CRI) for assessing the quality of a light source. The LER is the ocular efficiency of the visible light source [59], [60]. However, the LER or (luminous efficacy) LE is defined as the luminous flux (output light) divided by radiant flux (radiant power) [61], [62]. The unit of LER is lumen per watt [lm/w] in an SI unit system, where 100% luminous efficiency is equal to 683 lm/w (maximum LE) [62]. Furthermore, the LER

33 33 depends on the sensitivity of the human eyes. The maximum LER is 683 lm/w at 555 nm monochromatic wavelength (green light) [59], [60] which is the peak photopic sensitivity of human eyes to visible radiation. The sensitivity decreases away from the 555 nm wavelength as seen in Fig. (2-2), so that the LER diminishes as well [60]. The wavelengths in the invisible regions do not have LER because there is no luminous flux in these areas [62]. Moreover, the LER of an ideal white light source is lm/w [60]. The theoretical LER of white LEDs is lm/w, whereas commercial LEDs typically have LER equal to 150 lm/w at 20 ma forward-biased current. The LER of LEDs is higher than the LER of incandescent lamps (13 lm/w), and fluorescent lamp (90 lm/w). There are still many improvements which can be made to increase the LER of white LEDs in order to reach the maximum possible limit [63]. 2.6 Human Physiological Aspect of Light Human behavior is affected strongly by the nature of illumination of a light source. The percentage of red and blue component in white light influences our visual perception. The electromagnetic wavelengths from the sun are the source of the life on our planet. Humans use sunlight directly as visible light to brighten the environment, and indirectly by depending on plants for our food [64]. However, the circadian rhythms of humans are influenced by light, such as during rest and active time, sleep and awakening. We understand that the circadian rhythms are biological oscillations with periods close to, but not exactly, 24 hrs. Sunlight is one of the signals which entrains circadian cadences to 24 hrs a day [65]. The action of the circadian rhythm is different with

34 34 different wavelengths. Blue light has the most effect on circadian and nervous system stimulation. Attentiveness of a person increases with high color temperature light (which includes more blue light), but long time exposure to blue light causes reduced sleep hours in the night [66]. Furthermore, we choose different types of lights for different places [67]. People prefer to use a low color temperature white light source (warm light) at home to enhance the repose, for example, using candles in the living room or bedroom. Since high color temperature white light (cool light) reinforces performance, this cool light is used in offices and other working places [68]. The information and parameters (CIE 1931 (x, y) chromaticity coordinates, CCT, CRI, and LER) mentioned in this chapter are used to test light sources for specific application. Therefore, the phosphors applied in solid state lighting should have good qualities in order to generate white light with appropriate characteristics for use in general light sources.

35 35 CHAPTER 3: PHOSPHOR EXCITATION It is important to understand the excitation process of phosphors as luminescent materials. Generally, the excitation and energy transfer mechanism in a phosphor involves different interactions between a sensitizer (donor) and an activator (acceptor), depending on their features and the distance between them [3]. These interactions can be magnetic, electric multi-pole type interactions as well as charge exchange interactions [69]. The combined Forster-Dexter theory is the most common approach providing the main explanation for most of the interactions involving different bodies between which energy migration occurs. Theodore Forster developed the theory of interaction of the electric-electric dipole type in 1940, and then D. L. Dexter, primarily, generalized it to include higher order multi-dipole electric interactions and exchange interactions [69]. Since then, the resonant excitation of typical phosphors is clarified based on this theory. Figure 3-1 shows the resonant interaction between two bodies. Here, D indicates a sensitizer (donor) and A is an activator (acceptor). The simplest type of interaction between D A can be dipole-dipole (electric dipole), and the probability of transferring energy per second between them is expressed by Eq. (3.1) as P dd (R) = 3c4 ħ 4 σ A 4πn 4 τ D R 6 f D(E)F A (E) E 4 de (3.1)

36 36 where σ A represents the absorption cross-section of A, n is the refractive index of the crystal, τ D is radiative life time of D, R represents the distance between D and A, and ƒ D (E) and F A (E) are the shapes of D emission and A absorption spectrum, respectively. e e D R Figure 3-1: Schematic representation of resonant energy transfer process between D (donor) and A (acceptor). This graph was reproduced from Ref. [70]. A Sometimes the electronic wave-functions overlap when D and A are close to each other. In this case, the resonant energy transfer is an electronic exchange between these bodies. This is seen in Fig The probability of transferring energy in this interaction is given by P ex (R) = ( 2π ħ ) K2 exp ( 2R L ) f D(E)F A (E) de (3.2) In Eq 3.2, the K 2 is a constant and L represents Bohr radius (radii average of the excited state of D and the ground state of A).

37 37 D R A Figure 3-2: Electronic exchange of resonant energy transfer between D and A with electronic wave-functions of D and A overlap. This graph was reproduced from Ref. [70]. When energies of transition between D and A are not equal (this is the condition of resonant transfer), the energy transfer is assisted with a phonon which represents the energy difference. However, energy transition also can take place between the same types of ions like (D D). This transfer is called excitation migration or energy migration. It influences the time delay for phosphor emission [70]. The sensitizers (donors of energy) can be a host lattice or an ion [26], [69]. If the sensitizer is the host lattice, then the energy transfer is called host sensitization [69]. Sometimes, the sensitizer in the system is an impurity which is added to increase the efficiency of light emission. For example, the Ce 3+ ion sensitizes the Mn 2+ ion in sulfide phosphors [70]. Moreover, the sensitizers transfer exciting energy to the luminescent

38 center (activator). Then the electrons of the activator are excited to the excited state. These electrons release their energy as photons when they return to the ground state [3]. 38

39 39 CHAPTER 4: EXPERIMENTAL SETUPS 4.1 Samples Information In this section, the work on three different color phosphors RGB and the white light produced by mixing RGB phosphors is presented. The RGB phosphors were obtained as powders: KEu(WO 4 ) 2 (red phosphor, R), BaMg 2 Al 16 O 27 :Eu,Mn (green phosphor, G), and (Sr,Mg) 2 SiO 4 :Eu (blue phosphor, B). Phosphor specimens for experiments were prepared by mixing small amounts of phosphor powders with a potassium silicate solution (KASIL 1624) and coating the slurry on small glass substrates. Then, specimens were left to dry in order to be used in experiments. The samples were made of each individual, R, G, and B, phosphor and studied under different experimental conditions White Phosphor Preparation White emission phosphor powder mixture was produced by combining different amounts of three monochromatic RGB phosphor powders. Since the emission spectra of individual RGB phosphors are not identical, several attempts of mixing different amounts of RGB phosphors was performed. The CIE coordinates of the starting powder mixture of white phosphor were initially measured, and the composition was readjusted and remeasured several times until a white light with the desired CCT close to 5037 K was achieved (see Fig. 5-6). The developed trichromatic white phosphor that emitted the

40 desired color spectrum was used to prepare experiment specimens as described in Section Photoluminescence Spectra Measurement In this section, the procedures and instruments required for spectroscopic emission experiments will be described. Figure 4.1 shows the general experiment flow diagram and describes the experimental components were used. Figure 4-2 shows the experimental setup for spectral and color characterizations of fresh and aged samples. The experimental steps configurations are as follows: 1- At the beginning, the instruments were adjusted as in Fig In this step, the laser beam needed to hit the sample (S) inside the integrating sphere. In order to do that, one side of the integrating sphere was opened, and the sample was positioned in the middle of an integrating sphere. 2- The integrating sphere with a sample was aligned with a laser and with optical components for sample excitation. 3- The optical power of the laser beam was measured using an optical power meter and an optical sensor placed at the exact position where the sample was placed. The laser beam diameter and excitation laser power density were controlled (changed) by changing the position of the focusing lens, placed outside of the integrating sphere, or by inserting a diaphragm and a set of optical density filters into the laser beam path.

41 41 Newport 818-uv Laser diode InGaN/GaN Dual filter wheel Diaphragm Photo detector New focus 5254 NF Power meter Newport power/energy meter 1825-C Figure 4-1: Schematic diagram of power measurement setup using photo-detector. 4- After selecting the laser optical excitation power and laser beam diameter needed for sample excitation, the photodetector was removed and replaced with the sample within integrating sphere (see Fig. 4-2). 5- The final special adjustments of the integrating sphere and sample were made prior to conducting measurements. 6- A long pass filter (LPF, GG 435) was placed inside the integrating sphere in front of the output port, blocking the laser light from being detected by a spectrometer. Generated luminescence was transmitted via optical fiber to an Ocean Optics spectrometer (USB 2000). 7- The spectrometer was connected to a PC computer and the optical signal was analyzed using software. The Ocean View software from Ocean Optics, Inc. was used to collect

42 data and analyze the experimental data. Appendix C contains executed software procedures for measuring the emission spectra. 42 Figure 4-2: Experimental setup diagram used for phosphors emission characterization. The individual emission spectra of RGB phosphors and white phosphor were measured using the same procedure.

43 CIE Chromaticity Coordinates vs. Different Excitation Power Density Measurement This section describes the process and experimental setup configuration for measuring color coordinates of RGB and white phosphors at different laser excitation power intensities, using the following steps: 1- The 1 & 2 steps described in Section 4.2 were repeated. 2- The optical power of the laser beam was measured using an optical power meter and an optical sensor placed at the exact position where the sample was placed. The laser beam diameter and excitation laser power density were controlled (changed) by changing the position of the focusing lens placed outside of the integrating sphere or by inserting a diaphragm and a set of different optical density filters into the laser beam path. 3- The 4 through 6 steps described in Section 4.2 were repeated. 4- The spectrometer was connected to a computer and the software was run. Then the Color Wizard option was chosen from the Ocean View software (see Appendix D). 5- The laser excitation power density was changed by employing a different combination of power density filters mounted on a dual 1 optical filter wheel (New Focus 5254 NF) without inducing any changes to the laser beam/sample arrangement. Systematic measurement of the CIE 1931 chromaticity coordinates of RGB and white phosphors were performed using the steps described in this section at different excitation power densities achievable from a used LD up to ~31mW/cm 2.

44 CIE Chromaticity Coordinates vs. Different Temperature Measurement The experimental setup and procedures for measuring CIE 1931 chromaticity coordinates of RGB and white phosphors as a function of temperature were performed using the following steps: 1- Initially, a Polyimide tape heater (TS), 1W, 25 resistance, biased with adjusted dc power supply (5V), was mounted on a thermally insulated stage and was used to heat up phosphor specimens inside the integrating sphere. Thermocouple Type-K was used to monitor the phosphor specimen s surface temperature at the laser excitation spot. 2- Steps 1 through 6 described in Section 4.2 were repeated. The CIE coordinates measurement procedure using software was performed (see Appendix D). 3- The phosphor specimen temperature from 23 C to 160 C, with accuracy of ±1 C, was achieved by controlling supply current from a dc power supply. 4- The CIE coordinates measurement was systematically studied for RGB and white phosphors for each temperature without any measurement setup adjustment. It is worth noticing that the phosphor specimen temperature was primarily affected by the external heater but not by the laser beam excitation, even at the maximum laser power density excitation achievable (~31mW/mm 2 ). We believe that since the laser beam heat generated at the sample surface is small, it can be at this point excluded from further considerations.

45 Characterization of Phosphors Using Coherent and Non-Coherent Excitations The experiment focused on comparison of the effects from optical excitation of phosphors using coherent and non-coherent light sources in a single laser beam optical path or in laser beam double optical paths. It was performed in the following steps: 1- The experimental setup was adjusted as shown in Fig The specimen was replaced with a photo-detector as similarly described in Section 4.3. The variation of coherent and non-coherent optical power density was achieved by rotating the 1 optical filter wheel (New Focus 5254 NF) without inducing any changes to the optical beam/sample arrangement. 3- Next, the photodetector was replaced with a specimen. A long pass filter (GG 435*) (LPF) was inserted in front of the optical fiber s holder and the optical fiber (OF) was connected to the spectrometer (USB 2000). 4- A luminescence collecting lens (f/# =3, f=150 mm) was inserted between the specimen and the long pass filter. 5- The data was collected using software (see Appendix C). 6- In the case of a single laser beam optical path experiment, the excitation laser light that was not absorbed by the phosphor was freely exiting the specimen from the specimen s back. In the case of a double laser beam optical paths experiment, a notch filter (IF) was inserted behind the specimen and step 5 was repeated. 7- The experiment with incoherence excitation light source utilizes steps 1-5 described in this section. The Optotune LSR S-VIS laser speckle reducer was inserted after the dual filter wheel (New Focus 5254 NF). Since the light diffuses widely, a focusing lens

46 was used between the speckle reducer and the phosphor sample in order to focus the stimulating light on the sample. 46 Figure 4-3: Experimental setup diagram used for characterization of phosphors response induced by coherent and non-coherent excitations.

47 47 CHAPTER 5: RESULTS AND DISCUSSION To solve well known problems that current white LEDs suffer from, one can fabricate white LEDs using a near ultraviolet LED chip combined with RGB tricolor phosphor. In this way, the resultant white LED will have improved color uniformity, high color rendering index, and generally excellent light quality. However, the existing difference in individual color phosphor host degradation eventually will produce color aberrations. Thus, it is extremely important to fabricate and characterize a single-host phosphor with various emission bands. Unfortunately, such a phosphor has not been developed yet. Thus, the typical phosphor mixtures used currently for white LED fabrication rely on engineering individual colors, where the energy transfer occurs between the host/sensitizer and activator or coactivator via a multipolar interaction and energy transfer. The investigations reported here were based on studies of three rare-earth phosphors that can be combined to produce balanced white light. Specifically, we used the following three phosphors: (1) Potassium europium tungstate phosphor (KEu(WO 4 ) 2, type: CPK63/N-U1) emission color: red (R); (2) Europium and manganese-doped barium magnesium aluminate phosphor (BaMg 2 Al 16 O 27 :Eu,Mn type: KEMK63M/F-U1) emission color: green (G); and (3) Europium-doped strontium magnesium silicate phosphor ((Sr,Mg) 2 SiO 4 :Eu, type: HEBK63/N-D1) emission color: blue (B). All three phosphors were obtained as powders from Phosphor Technology Limited, England, U.K. Fig. 5-1 shows scanning electron micrographs of these RGB phosphor particles. These phosphors have distinctive particle size distributions that have been evaluated with a

48 48 Coulter counter and are listed here in Table 5-1. Particles of the R and B phosphors have very well-defined crystal habits. The G phosphor has comparatively smaller platelet-like particles of somewhat random shapes. The spectral characterization, including excitation and emission spectra for all the monochromatic RGB phosphors (without binder), were measured under the same experimental conditions at room temperature. a b c 1 m Figure 5-1: Scanning electron micrographs of the red (R) (a), green (G) (b) and blue (B) (c) phosphor particles. The R and B phosphors are typical of a large class of near-uv pumped phosphors containing europium as the active luminescent ion [71], [72], whereas the green phosphor belongs to a class of Eu/Mn co-doped phosphors in which Eu 2+ ions act as sensitizers, transferring energy non-radiatively to Mn 2+ activator ions that produce the main luminescence [73], [74]. The investigated RGB phosphors show efficient luminescence when excited at selected wavelengths between 250 nm to 450 nm (B), 465 nm (G) and 550 nm (R), respectively, at 300 K. The typical excitation spectra of studied RGB phosphors are depicted in Fig. 5-2.

49 49 Figure 5-2: Room temperature photoluminescence excitation (PLE) spectra of (a) KEu(WO 4 ) 2 red phosphor monitored at 615 nm, (b) BaMg 2 Al 16 O 27 :Eu,Mn green phosphor monitored at 515 nm, and (c) (Sr,Mg) 2 SiO 4 :Eu blue phosphor monitored at 460 nm. Photoluminescence (PL) spectra were measured for each phosphor, with excitation photons energy corresponding to the maximum of each individual PLE spectrum. All spectra were normalized to the same intensity. The dashed line indicates the position of LD excitation wavelengths at 404 nm conducted in the color characteristics measurements. Please notice the different horizontal scale in segment (a). The excitation spectrum of R phosphor (see Fig. 5-2(a)) monitored at 615 nm can be divided into two parts. One is a charge transfer band from 230 to 320 nm centered at

50 nm. This band is attributed to the O 2 W 6+ charge transfer within the WO 2 4 groups. The other part is composed of a series of narrow bands from 320 to 550 nm, which correspond to the characteristic f f transitions of the Eu 3+ ion. When monitoring the red emission of Eu 3+ (615 nm, 5 D 0-7 F 2 ), the photoluminescence excitation (PLE) spectrum of the R phosphor exhibits peaks at 323, 364, 385, 395, 416, 466, 480, 570 nm corresponding to the intra-4f transitions of the Eu 3+ ion from the ground level 7 F 0 to the 5 H 3, 5 D 4, 5 L 7, 5 L 6, 5 D 3, and 5 D 2 excited levels, respectively. The R phosphor emitting primarily in 610nm to 630 nm spectral regions can be effectively excited due to the intraconfigurational f f transitions of the Eu 3+ ion by photons with energies from near UV. It is seen that the excitation wavelength at 465 nm is the most efficient one; however, the 404 nm excitation wavelength is equally efficient. The representative PL spectrum, when excited at 404 nm (see Fig.2(a)), shows emission lines in the range of , , and nm due to intra-4f-shell transitions in the Eu 3+ ion. The transitions are split into components depending on the host matrix crystal field. These emission lines of the Eu 3+ ion cover the orange and red spectral region with less intense 5 D 0 7 F 1 magneticdipole transition emissions central at 595 nm and dominant 5 D 0 7 F 2 electric-dipole transition lines at 614nm and 618nm, respectively. Two less intense emission peaks for the 5 D 0 7 F 4 transition are also observed at 696 and 705 nm along with a minor unresolved emission peak at 650 nm corresponding to the 5 D 0 7 F 3 transition. The observed 5 D 0 7 F 2 and 5 D 0 7 F 1 transition intensity ratio confirms that the Eu 3+ ion occupies predominantly a site without inversion symmetry in the KEu(WO 4 ) 2 phosphor [75]. The presence of 5 D 0 7 F 1 emissions of the Eu 3+ ion indicates that in the KEu(WO 4 ) 2

51 51 host crystal there also exist Eu 3+ ions occupying a site with an inversion center. However, the intensity of electric dipole transition of the Eu 3+ ions is almost 10 times higher than that of the magnetic dipole transition, intimating that the Eu 3+ ions predominantly occupy the low symmetry sites. The barium magnesium aluminate phosphor BaMg 2 Al 16 O 27 studied in this project is doped with europium and manganese. In this phosphor, the Mn 2+ ions act as an efficient activator emitting in the green spectral region. In general, the photoluminescence process of single doped Mn 2+ in a phosphor host is characterized by the transition of 3d5s electrons in the Mn 2+ ion. This typically restricts the application of Mn 2+ in phosphors to be used within narrow spectral range limits. To overcome this limitation, some sensitizers are used to enhance the emission intensity of the Mn 2+ ion. It is known that emissions of Eu 2+ ions arise from the allowed transition between the 5d and 4f levels, and thus emission efficiency of the Eu 2+ ion is high in many hosts. Furthermore, co-doping of the Eu 2+ and Mn 2+ pair has recently been adopted as a way of improving photoluminescence properties of BaMg 2 Al 16 O 27 doped with Mn 2+ only. Figure 5-2(b) shows the excitation spectrum of the G phosphor excited with photons of wavelength between 250 nm to 470 nm and monitored at 515 nm. The PLE spectrum consists of a broad band spanning from 250 nm to 430 nm and overlapped with a sharp peak at 385 nm. A low intensity broad band peaking at 450 nm is also seen. The dominant broad PLE band when monitored at 515 nm (emission from the Mn 2+ ion due to the 4 T 1 6 A 1 transition) is composed of spectral features resulting from the direct excitation of Eu 2+ and Mn 2+ ions and energy transfer between them. The broad band peaking at 350 nm and

52 high energy wing from 250 nm to 350 nm are due to 4f 7 4f 6 5d 1 transitions of the Eu 2+ ion that replaced Ba 2+ ion in the host crystal. The blue emission band at 450 nm is 52 attributed to the 4f 6 5d 1 4f 7 transition of Eu 2+ ions. The sharp peak at 385 nm is due to 6 A 1 4 T 2 transitions of Mn 2+ ions which replaced Al 3+ ion in the host crystal. When the Mn 2+ ion replaces the Al 3+ ion, then a negative charge is developed in the lattice because of the nonequivalent replacement of the ions involved. To maintain the electrical neutrality of the phosphor, it is expected that positive charge compensation is induced by structural defects and grain boundaries in the absence of extrinsic impurities. The observed divalent manganese emission critically depends on the size of the crystallographic cation site that Mn 2+ is likely to occupy (here Mn Al ) and on the coordination number of the Mn 2+ ion in the BaMg 2 Al 16 O 27 :Eu,Mn host matrix. In the present case, the Mn 2+ ion occupies a tetrahedrally coordinated site [76]. It is known that the mechanism of energy transfer in phosphors basically requires a spectral overlap between the sensitizer (donor) emission and the activator (acceptor) excitation. In the case of BaMg 2 Al 16 O 27 :Eu,Mn, the Eu 2+ ion is a sensitizer and the Mn 2+ ion is an activator. It was shown that there exists efficient non-radiative energy transfer between the Eu 2+ and Mn 2+ ions in BaMg 2 Al 16 O 27 and other barium aluminate hosts [77], [78]. The presence of the Eu 2+ ion emission band in the PL spectrum at 450 nm, when excited at 404 nm (see Fig. 5-2(b)), and the large intensity ratio between the green band of Mn 2+ to the blue band of Eu 2+, indicates that the aforementioned sensitization process is very effective. However, not all energy migrating between Eu 2+ donors reaches Mn 2+ acceptors, resulting in green emission at 515 nm. This is most probably because of the excess of Eu 2+ ions sensitizing individual Mn 2+ ions

53 53 and/or Mn 2+ -Mn 2+ ion pairs, as well as the possibility of energy cross-relaxation processes occurring among Eu 2+ -Eu 2+ ions. Figure 5-2(c) shows the excitation spectrum for a B phosphor while monitoring at 460 nm. It can be seen that the excitation spectrum is composed of a few broad bands in the region spanning from 250 nm to 450 nm with a well-defined maximum at 410 nm and a shoulder at ~350 nm on the high energy wing. Chen et al. reported that there are two major crystallographic sites in monoclinic Sr 2 SiO 4, namely, ten-coordinated Sr(I) and nine-coordinated Sr(II) [79]. Thus, it is expected that Eu 2+ ions in monoclinic Sr 2 SiO 4 :Eu 2+ will replace primarily both the Sr(I) and Sr(II) sites due to the similar radius of Sr 2+ (0.113nm) and Eu 2+ (0.095nm) ions, rather than being incorporated into the Mg 2 (Si 2 O 4 ) sub-lattice due to the big difference in the radius of Mg 2+ (0.065nm) and Eu 2+. The PLE spectrum can be assigned to the absorptions from the Eu 2+ ions occupying Sr(I) (minority) and Sr(II) (majority) sites. The short wavelength band-shoulder at ~350 nm corresponds to the Eu 2+ ions occupying Sr(I) sites, whereas the band at 410 nm corresponds to the Eu 2+ ions occupying Sr(II) sites [79]. The observed luminescence spectrum, when excited at 410 nm, shows emission predominantly from Sr(I) site with emission from Sr(II) site strongly suppressed. According to B. J. Chen and colleagues [79], there exists effective energy migration between the Eu 2+ ions occupying Sr(I) and the Eu 2+ ions occupying Sr(II) sites. Relying on the arguments of intensity ratio between 350 nm (Sr(I)) and 410 nm (Sr(II)), peaks and the absence of emission from Eu 2+ ions in the Sr(II) site, one can assume that the emission from Eu 2+ in the Sr(I) site dominates. Furthermore, it can be speculated that Eu is effectively excited when occupying both

54 54 Sr(I) and Sr(II) sites; however, observed emission at 460nm originates from Eu 2+ ions occupying the Sr(I) site with strong energy transfer from Eu 2+ occupying Sr(II) site to Sr(I) site, and weak back energy transfer between these two centers. It is worthwhile to note that typical sharp line emissions corresponding to the transitions of Eu 3+ ion are observed in the PL spectrum of the B phosphor (see Fig. 5-2(c)). It should be mentioned here that PLE and PL spectra shown in Fig. 5-2 were measured explicitly in this project. These spectra were provided together with RGB phosphors by Phosphor Technology Ltd., United Kingdom. To confirm the photoluminescence of fabricated RGB phosphor specimens (see Experiments & Samples, Chapter 4), PL measurements of RGB phosphors were conducted by exciting them with a 404 nm laser diode using the experimental setup shown in Fig Fig. 3 shows the PL spectrum of R phosphor excited with 404 nm photons at room temperature. It is seen that the spectrum is dominated by 5 D 0 7 F 2 electric-dipole transition lines at 614nm and 618nm. The PL spectrum has identical spectral features as shown in Fig. 5-2(a). The inset shows CIE chromaticity coordinates of the R phosphor measured at 300 K. The measured CIE parameters are x= and y=0.327 and differ by x=0.032 and y=0.003 from the CIE red color standard shown in Fig.5-7. Also shown in the inset to Fig. 5-3 is the position of the chromaticity coordinate for the R phosphor under the same measurement conditions as that used for taking the PL spectrum.

55 55 Figure 5-3: Room temperature photoluminescence spectra and CIE chromaticity coordinates of the R phosphor. The 404 nm laser excitation at a power level of 30.3 mw/mm 2 was used for these measurements. The black curve on CIE chromaticity chart shown in inserts is the Planckian locus, showing the chromaticity coordinates of an ideal black body at different temperatures. Figure 5-4 shows the PL spectrum of the G phosphor excited with 404 nm photons at room temperature. It is seen that the spectrum is dominated by emission from the Mn 2+ ion due to the 4 T 1 6 A 1 transition. The PL spectrum has identical spectral features as shown in Fig. 5-2(b). The inset shows CIE chromaticity coordinates of the G phosphor measured at 300 K. The measured CIE parameters are x= and y=0.706 and differ by x=0.078 and y=0.004 from the CIE red color standard shown in Fig Also shown in the inset to Fig. 5-4 is the position of the chromaticity coordinate for the G phosphor under the same measurement conditions as that used for taking the PL spectrum.

56 56 Figure 5-4: Room temperature photoluminescence spectra and CIE chromaticity coordinates of the G phosphor. The 404 nm laser excitation at a power level of 30.3 mw/mm 2 was used for these measurements. The black curve on CIE chromaticity chart shown in inserts is the Planckian locus, showing the chromaticity coordinates of an ideal black body at different temperatures. Figure 5-5 shows the PL spectrum of the B phosphor excited with 404 nm photons at room temperature. It is seen that the spectrum is dominated by emission at 460nm originating from Eu 2+ ions occupying the Sr(I) site. The PL spectrum has identical spectral features as shown in Fig. 5-2(c). The inset shows CIE chromaticity coordinates of the B phosphor measured at 300 K. The measured CIE parameters are x= and y=0.078 and differ by x=0.015 and y=0.018 from the CIE blue color standard shown in Fig Also shown in the inset to Fig. 5-5 is the position of the chromaticity coordinate for B phosphor under the same measurement conditions as that used for taking the PL spectrum.

57 57 Figure 5-5: Room temperature photoluminescence spectra and CIE chromaticity coordinates of the B phosphor. The 404 nm laser excitation at a power level of 30.3 mw/mm 2 was used for these measurements. The black curve on CIE chromaticity chart shown in inserts is the Planckian locus, showing the chromaticity coordinates of an ideal black body at different temperatures. Figure 5-6 shows the room temperature PL spectrum of the trichromatic white phosphor excited with 404 nm photons and having the measured correlated color temperature of 5037 K. It is seen that the white phosphor s luminescence qualitatively reflects the individual spectral features of RGB phosphors shown in Figs. 5-3 to 5-5. The RGB phosphors characteristic emission band intensities were optimized to generate white light with the CIE coordinates shown in the inset of Fig. 5-6.

58 58 Figure 5-6: Room temperature photoluminescence spectra and CIE chromaticity coordinates of the trichromatic white phosphor. The 404 nm laser excitation at a power level of 30.3 mw/mm 2 was used to excite the phosphor. The black curve on CIE chromaticity chart shown in inserts is the Planckian locus, showing the chromaticity coordinates of an ideal black body at different temperatures. The measured CIE (x, y) parameters are x= and y=0.357 and differ by x= and y=0.024 from the CIE white light standard (equal power point). The insert shows the position of the CIE (x, y) chromaticity coordinate for the trichromatic white phosphor under the same measurement conditions as that used for taking relevant PL spectrum. Figure 5-7 is the CIE chromaticity diagram showing individual RGB phosphor (x, y) coordinates. The insert shows the position of (x, y) coordinates determined experimentally using two various excitation conditions. The R(T), G(T) and B(T) represents (x, y) coordinates obtained for RGB phosphors from temperature dependent CIE experiments (see Fig.s 5-8, 5-9, 5-10), whereas the R(P), G(P) and B(P) represents

59 59 (x, y) coordinates obtained for RGB phosphors from excitation power density dependent CIE experiments (see Fig.s 5-11, 5-12, 5-13), discussed in the following text. The insert shows also the values of standard CIE (x, y) parameters for comparison. Figure 5-7: CIE chromaticity diagram showing position of CIE (x, y) coordinates for RGB phosphors obtained for temperature and excitation power density experiments. Dashed lines link the RGB points defining the scope of accepted CIE (x, y) coordinate variation for obtaining white light. On room temperature excitation with a laser source at 404 nm, the RGB phosphors produced emission spectra that have been depicted in Fig. 5-3 to 5-5. The wavelength peaks for the RGB phosphors were located at 616 (R), 515 (G) and 460 nm (B), respectively. The full-width-at-half-maximums (FWHM) of the primary emission peaks were 18 nm, 38 nm and 62 nm for the R, G and B phosphors, respectively. It should be noted that there was no measurable increase in phosphor temperature just from exposure to a laser pump light. The RGB phosphors generate their characteristic colors

60 60 from respective atomic transitions in europium (Eu 3+ (red), Eu 2+ (blue)) ions and manganese (Mn 2+ (green) ion. As discussed previously, in the case of the G phosphor, europium ions act as sensitizers transferring energy gained from absorbed pump photons efficiently to luminescent manganese ions [80], [81]. The Mn 2+ ions are themselves hard to pump directly because their d d transitions are forbidden on both spin and parity grounds for electrical dipole transitions. Presence of Eu 2+ ions in the same host allows first their pumping through f d transitions followed by efficient radiationless transfer of energy to Mn 2+ ions, causing them to fluoresce [82]. While the green emission in this phosphor originates from Mn 2+ ions, some emission also comes from the minority Sr(II) site or possibly Eu 2+ ions (low intensity broad band between 425 and 500 nm in Fig. 5-4). In the following, we present experimental data showing CIE (x, y) coordinates change as a function of temperature and excitation power density. Figure 5-8 shows the evolution of CIE (x, y) coordinates of the R(T) phosphor as a function of temperature changing from 25 C to 153 C when excited with an LD at 404 nm generating excitation power density of 30 mw/mm 2. It is seen in Fig. 5-8(a) that the CIE (x, y) coordinate values change linearly from ( , ) to ( , ) with increasing temperature. Figure 5-8(b) shows the dominant R phosphor emission line peak intensity and peak position changes when temperature increases. It is seen that the R phosphor peak (at 616 nm) intensity increases when temperature changed from 25 C to ~90 C and then decreased reaching initial intensity at 153 C. We can speculate that the observed intensity change by ±16.2%, as compared to the initial

61 intensity, is due to varying energy migration schemes in the KEu(WO 4 ) 2 phosphor. The dominant R phosphor peak position does not change significantly up to 153 C. 61 Figure 5-8: (a) Variation of the CIE (x, y) coordinates for the R(T) phosphor as the function of temperature. (b) Change of the KEu(WO 4 ) 2 phosphor dominant peak intensity and peak potion as the function of temperature. Figure 5-9 shows the evolution of CIE (x, y) coordinates of the G(T) phosphor as a function of temperature changing from 25 C to 152 C when excited with LD at 404

62 62 nm generating excitation power density of 30 mw/mm 2. It is seen in Fig. 9(a) that the CIE (x, y) coordinate values change nonlinearly from (x=0.1296, y=0.7059) to (x=0.1284, y=0.6807) with increasing temperature. Figure 5-9(b) shows the dominant G phosphor emission band peak intensity and position changes when temperature increases. It is seen that the G phosphor band (at 515 nm) intensity decreases monotonically with increased temperature. The dominant G phosphor band peak position does not change significantly up to 153 C. Figure 5-9: (a) Variation of the CIE (x, y) coordinates for the G(T) phosphor as the function of temperature. (b) Change of BaMg 2 Al 16 O 27 :Eu,Mn phosphor dominant band intensity and peak potion as the function of temperature.

63 63 Figure 5-10 shows the evolution of CIE (x, y) coordinates of the B(T) phosphor as a function of temperature changing from 25 C to 147 C when excited with LD at 404 nm generating excitation power density of 30 mw/mm 2. It is seen in Fig. 5-10(a) that the CIE (x, y) coordinate values change from ( , ) to ( , ) with increasing temperature. Initially, the CIE (x, y) coordinates change only due to the x parameter change and when temperature increased above 50 C both the CIE (x, y) coordinates change linearly. Figure 5-10(b) shows the dominant B phosphor emission band peak intensity and position changes when temperature increases. It is seen that the B phosphor band (at 460 nm) intensity decreases monotonically with increased temperature. The dominant B phosphor band peak position does not change significantly up to 147 C.

64 64 Figure 5-10: (a) Variation of the CIE (x, y) coordinates for the B(T) phosphor as the function of temperature. (b) Change of (Sr,Mg) 2 SiO 4 :Eu phosphor dominant band intensity and peak potion as the function of temperature. The CIE chromaticity coordinates of phosphor emissions change with temperature. This leads to the well-known problem with white LEDs changing their color point with changes in operating current and thus LED temperature. This phenomenon has been well-investigated with LED light excitation of phosphors [83]. In studying CIE chromaticity as the function of temperature, we looked at changes in the chromaticity of RGB phosphors under coherent laser excitation (404 nm) at different phosphor temperatures. It should be noted that in these studies the temperature increase was caused by an external heater and was not due to heating from the laser diode or from the

65 65 absorption of pump radiation. Figure 5-11 shows the variation of CIE chromaticity coordinates, due to phosphor temperature variations, for the RGB and white phosphors. The range of parameters causing the variation (temperature in C), corresponding to the temperature variations ranges discussed in Fig.5-8, 5-9 and 5-10, is shown alongside each group of chromaticity points. Figure 5-11: Variation of CIE chromaticity coordinates for silica-bound RGB and trichromatic white phosphors with 404 nm laser pumping as the function of phosphor temperature. Figure 5-12 shows the evolution of CIE (x, y) coordinates of the R(P) phosphor as a function of excitation power density from 1.4 mw/mm 2 to 30 mw/mm 2 when excited with LD at 404 nm at 300 K. It is seen in Fig. 5-12(a) that the CIE (x, y) coordinates values changes linearly from ( , ) to (x-=0.6727, y=0.3270) with increasing excitation power density. Figure 5-12(b) shows the dominant R phosphor

66 66 emission line peak intensity and peak position changes when excitation power density increases. It is seen that the R phosphor peak (at 616 nm) intensity increases linearly with excitation power density up to 30 mw/mm 2. The dominant R phosphor peak position does not change significantly up to maximum power density used in experiment. Figure 5-12: (a) Variation of the CIE (x, y) coordinates for R(P) phosphor as the function of excitation power density. (b) Change of KEu(WO 4 ) 2 phosphor dominant peak intensity and peak potion as the function of excitation power density. Figure 5-13 shows the evolution of CIE (x, y) coordinates of the G(P) phosphor as a function of excitation power density from 1.4 mw/mm 2 to 30 mw/mm 2 when excited with LD at 404 nm at 300 K. It is seen in Fig. 5-13(a) that the CIE (x, y) coordinate values change non-monotonically from ( , ) to (x=0.1325, y=0.7059)

67 67 with increasing excitation power density. It is interesting to notice here that the trend of CIE (x, y) coordinates change reverses when the excitation power density exceeded 26 mw/mm 2. Figure 5-13(b) shows the dominant G phosphor emission band peak intensity and band position changes when excitation power density increases. The G phosphor dominant band peak intensity increased and saturated for excitation power exceeding 30 mw/mm 2. Above this value the G phosphor emission intensity decreased. The dominant G phosphor band peak position changed non-systematically ±1.5 nm when excitation maximum power density increased. Figure 5-13: (a) Variation of the CIE (x, y) coordinates for the G(P) phosphor as the function of excitation power density. (b) Change of BaMg 2 Al 16 O 27 :Eu,Mn phosphor dominant peak intensity and peak potion as the function of excitation power density.

68 68 Figure 5-14 shows the evolution of CIE (x, y) coordinates of the B(P) phosphor as a function of excitation power density from 1.5 mw/mm 2 to 30 mw/mm 2 when excited with LD at 404 nm at 300 K. It is seen in Fig. 5-14(a) that the CIE (x, y) coordinate values change linearly from ( , ) to (x=0.1345, y=0.0777) with increasing excitation power density. Slight deviation for the observed linearity has been noticed for CIE (x, y) coordinates for the power of 30 mw/mm 2. Figure 5-14(b) shows the dominant B phosphor emission band peak intensity and band position changes when excitation power density increases. The B phosphor dominant band peak intensity increased linearly with excitation power density. The dominant B phosphor band peak position changed non-systematically ±1.5 nm when excitation maximum power density increased.

69 69 Figure 5-14: (a) Variation of the CIE (x, y) coordinates for the B(P) phosphor as the function of excitation power density. (b) Change of (Sr,Mg) 2 SiO 4 :Eu phosphor dominant peak intensity and peak potion as the function of excitation power density. Figure 5-15 shows the variation of CIE chromaticity coordinates, due to excitation power density, for the RGB and white phosphors. The range of parameters causing the variation (power density in mw/mm 2 ), corresponding to the excitation power intensity variations ranges discussed in Fig. 5-12, 5-13 and 5-14, is shown alongside each group of chromaticity points.

70 70 Figure 5-15: Variation of CIE chromaticity coordinates for silica-bound RGB and trichromatic white phosphors with 404 nm laser pumping as the function of excitation power density. The R and B phosphors showed the least variation whereas the G phosphor showed more variation. The trichromatic white phosphor showed even larger variation in chromaticity as the temperature (see Fig. 5-11) or pump power (see Fig. 5-15) was varied. If used in real illumination devices, the variation is expected to be much smaller than what these results show because the phosphor temperature and incident laser pump power are not likely to vary much at all. In this sense, these results are an illustration of worst case scenario results that will only be observed in real lighting devices if they malfunction for some reason, causing the phosphor temperature or incident pump power to increase catastrophically.

71 71 We also studied double-sided excitation of RGB and white phosphors by exposing a phosphor film to pump radiation from both sides of the film. This was achieved by placing a phosphor-coated glass substrate in front of a notch filter (used as a narrowband filter) that allowed selective reflection of the incident laser light back on to the phosphor. The notch filter, designed for blocking a very narrow wavelength band around 404 nm, acted as a highly reflective mirror for 404 nm radiation and illuminated the phosphor from the other side as well. In this way, the phosphor film in silica matrix binder was exposed to 404 nm laser light from both sides, while the down-converted light was able to escape freely from both sides of the phosphor film. For this purpose, use was made of a 404 nm blocking 12.5 mm diameter notch filter with optical density (OD) 6 light rejection capability. Double-sided pump illumination ensures a more thorough phosphor excitation, as compared to only one-sided illumination. This happens in spite of the fact that phosphor crystals are largely transparent. Illumination from the other side ensures that phosphor crystals with the right orientation preferentially absorb photons and thus get pumped. Results for all phosphors for this experiment are shown here in Fig

72 72 Figure 5-16: Phosphor luminescence intensity variation with pump power density for (a) red (R), (b) green (G), (c) blue (B) and (d) trichromatic white phosphors. Measurement results, for each case, are shown for ordinary one-sided illumination (filled circles) and for two-sided illumination (filled squares), using a 404 nm wavelength-selective mirror (notch filter). We found that most pump photons are absorbed in the first pass through the phosphor. On the other side of the phosphor sample, photons from phosphor luminescence were allowed to go through the notch filter and then be absorbed by a black absorber. Residual pump photons that survived the first pass through the phosphor were, however, reflected by the notch filter and were thus again incident on the phosphor from the back side. This resulted in a small increase in luminescence for R and B phosphors;

73 73 however, the increase in luminescence was much more pronounced in the case of the G phosphor. The small increase in the case of the R phosphor (Fig. 5-16(a)) was due to the fact that the absorption spectrum of that phosphor is not well-matched to pumping at 404 nm wavelength (see Fig. 5-2(a)). This causes relatively few pump photons to be absorbed, allowing many photons to get through the phosphor film sample unabsorbed. After reflecting from the narrowband notch filter mirror, the photons are then incident on the phosphor film from the other side, but again due to absorption mismatch relatively few photons are absorbed by the R phosphor, causing only a small increase in luminescence. In the case of the B phosphor (Fig. 5-2 (c)), it is the high absorption of the blue phosphor for pump photons that results in a similar small effect in phosphor luminescence (see Fig. 5-16(c)). In this case, many pump photons are absorbed in the first pass through the blue phosphor film, and thus there are fewer photons left to reflect off the narrowband mirror. A small number of photons pumping the film from the other side then cause only a small increase in phosphor luminescence. The larger increase in luminescence in the case of the G phosphor is likely a result of this phosphor having smaller and more planar particles (see Table 5-1) that possess a higher cross-section for double-sided pumping. Figure 16 also shows that none of the phosphors is close to luminescence saturation at the highest optical power densities used in our experiments. Thus these phosphors are suitable for use in high fluency pumping configurations in laser-based light sources. In another set of experiments, we investigated the difference between pumping phosphors with coherent and incoherent light. While this can be done by pumping, in

74 74 turn, with a laser diode and an LED emitting light at the same wavelength, the wider wavelength spread of LED radiation can make the comparison somewhat difficult. We chose to use a dynamic light diffuser, commonly used for de-speckling laser beams, as a means of reducing the spatial coherence of our pump beam. Measurements of phosphor luminescence as a function of incident light power with the laser speckle reducer switched off (coherent pumping) and switched on (incoherent pumping) show almost no difference as is shown in Fig This observation strengthens the conclusion that lasing speckle pattern does not cause a decrease in phosphor pumping efficiency. Figure 5-17: Trichromatic white phosphor emission intensity change when excited at 404 nm as the function of incident light power with the laser speckle reducer switched off (coherent pumping) and switched on (incoherent pumping). We also studied the ageing characteristics of our RGB phosphors and various RGB phosphor mixtures. Accelerated life tests were done by periodically measuring the intensity of the main emission peak for each phosphor as samples were kept heated at

75 C in ambient air. The results are shown in Fig where the gradual decrease in the phosphor s luminescence efficiency over an extended period of time can be clearly seen. Phosphors designed for pumping with LEDs emitting around 460 nm show a similar behavior (not shown here). However, in the case of remote laser excitation, the lifetime can be much longer as phosphors operate at temperatures only a little above room temperature. Typical temperatures for laser excited phosphors are in the range of 30 C to 35 C (assuming the experimental arrangement itself is kept at around 23 C). At such temperatures, we estimate that the phosphors studied in this work will still emit 80% of their initial intensity after 400 hrs. Based on the our investigation about the characterizations of these individual and mixing white phosphors under violet laser diode excitation light source become possible to manufacture white light source for general lighting.

76 76 Figure 5-18: RGB phosphors ageing effect after 400 hours of heating at 120 C. The intensity emission spectra were measured periodically. Table 5-1: RGB phosphors particle size distributions as measured from an ultrasonic dispersion by a Coulter counter method. Volume [%] [ m] (Red) [ m] (Green) [ m] (Blue)

77 77 CHAPTER 6: CONCLUSIONS This thesis focused on studying the characteristics of light emission from KEu(WO 4 ) 2, BaMg 2 Al 16 O 27 :Eu,Mn and (Sr,Mg) 2 SiO 4 :Eu rare-earth phosphors pumped by blue-violet LD. Changes in chromaticity point, as a result of temperature and incident optical power density, were studied and found to be quite small. Variation of phosphor luminescence intensity with temperature and pump power density was studied for RGB and trichromatic white phosphors. No phosphor saturation effects were observed up to the highest laser pump power employed (96 mw). This was borne out by both one-sided and two-sided pumping of phosphors embedded in a silicate binder matrix. We carried out phosphor pumping experiments with both coherent and incoherent illumination at the same wavelength of 404 nm and found no significant difference between these two cases. This strengthens our conviction that laser speckle does not pose any problem for energy conversion in laser-based phosphor pumping schemes. Finally, our phosphor ageing studies showed that laser pumping does not cause any appreciable degradation over time due to thermal or optical irradiation effects.

78 78 CHAPTER 7: FUTURE WORK One of the possible future research directions resulting from the project is study of the RGB monochromatic color and white trichromatic phosphors with a magnetic field stimulus. During the course of the project, we have observed preliminarily that luminescence of all studied phosphors changes with an applied magnetic field. In general, red R, G and B phosphors changed luminescence intensity when retaining all major spectral features with increasing magnetic field strength from 0 to 1.5 Tesla at 300 K. Thus, the investigations and further experiments focus on exploring properties of these materials when applying magnetic fields may bring new insights to better understanding of energy migration, luminescence intensity quenching and long term stability of materials. Another research opportunity arises from investigation of the quantum efficiency (QE) for selected RGB and white trichromatic phosphors. Quantum efficiency represents one of the primary criteria for selecting phosphors in many applications, including SSL, flat panel display, fluorescent lamps, etc. In general, QE of luminescence materials under photo-excitation source means the ratio of the number of photons emitted to the photons absorbed [84]. The best way to measure QE for phosphor powder is using an integrating sphere [84], [85] to avoid the errors in measuring the absorbed photons by the phosphor sample, since the data of standard reflectance material is typically calculated from experiments using the integrating sphere approach. However, the sample absorbed photons can be calculated based on reflectivity of the sample compared with standard reflected materials (BaSO 4, MgO, etc.). In order to measure accurate QE of phosphors,

79 79 the system has to be calibrated using fluorescence standard material having known QE when excited by a known light source being an integral part of the experimental system [84]. In was demonstrated that an integrating sphere can be successfully used to measure QE for liquid and powder phosphors with near-uv and blue light excitation sources [85]. The quinine sulfate dehydrate was used as a fluorescence standard to test and calibrate the system. Therefore, we believe that the QE measurements of RGB and white phosphors can be continued using, as the base, the experimental setup developed in this project in the future.

80 80 REFERENCES [1] C. C. Lin and R.-S. Liu, Advances in Phosphors for Light-emitting Diodes, J. Phys. Chem. Lett., vol. 2, no. 11, pp , Jun [2] H. Lian, Z. Hou, M. Shang, D. Geng, Y. Zhang, and J. Lin, Rare earth ions doped phosphors for improving efficiencies of solar cells, Energy, vol. 57, pp , Aug [3] S. Ye, F. Xiao, Y. X. Pan, Y. Y. Ma, and Q. Y. Zhang, Phosphors in phosphorconverted white light-emitting diodes: Recent advances in materials, techniques and properties, Mater. Sci. Eng. R Reports, vol. 71, no. 1, pp. 1 34, Dec [4] C. J. Humphreys, Solid-State Lighting, vol. 33, no. April, pp , [5] N. C. George, K. a. Denault, and R. Seshadri, Phosphors for Solid-State White Lighting, Annu. Rev. Mater. Res., vol. 43, no. 1, pp , Jul [6] L. Zhou, J. Wei, J. Wu, F. Gong, L. Yi, and J. Huang, Potential red-emitting phosphor for white LED solid-state lighting, J. Alloys Compd., vol. 476, no. 1 2, pp , May [7] B.-M. Song, B. Han, and J.-H. Lee, Optimum design domain of LED-based solid state lighting considering cost, energy consumption and reliability, Microelectron. Reliab., vol. 53, no. 3, pp , Mar [8] M. S. Shur and A. R. T. Uras, Solid-State Lighting : Toward Superior, vol. 93, no. 10, pp , [9] Y. Jeff, solid-state lighting (lamps, chips, and materials for tomorrow), no. June, [10] X. Wang, J. Wang, J. Shi, Q. Su, and M. Gong, Intense red-emitting phosphors for LED solid-state lighting, Mater. Res. Bull., vol. 42, no. 9, pp , Sep [11] C.. Ronda, T. Jüstel, and H. Nikol, Rare earth phosphors: fundamentals and applications, J. Alloys Compd., vol , pp , Jul [12] R. Ha, F. Republic, and O. Chemie, Luminescent Materials, vol. 21, pp , 2012.

81 [13] R. Haitz and J. Y. Tsao, Solid-state lighting: The case 10 years after and future prospects, Phys. Status Solidi, vol. 208, no. 1, pp , Jan [14] J. Kundu, Y. Ghosh, A. M. Dennis, H. Htoon, and J. a Hollingsworth, Giant nanocrystal quantum dots: stable down-conversion phosphors that exploit a large stokes shift and efficient shell-to-core energy relaxation., Nano Lett., vol. 12, no. 6, pp , Jun [15] D. S. Meyaard, G.-B. Lin, J. Cho, E. Fred Schubert, H. Shim, S.-H. Han, M.-H. Kim, C. Sone, and Y. Sun Kim, Identifying the cause of the efficiency droop in GaInN light-emitting diodes by correlating the onset of high injection with the onset of the efficiency droop, Appl. Phys. Lett., vol. 102, no. 25, p , [16] K. a. Denault, M. Cantore, S. Nakamura, S. P. DenBaars, and R. Seshadri, Efficient and stable laser-driven white lighting, AIP Adv., vol. 3, no. 7, p , [17] H.-Y. Ryu and D.-H. Kim, High-brightness Phosphor-conversion White Light Source Using InGaN Blue Laser Diode, J. Opt. Soc. Korea, vol. 14, no. 4, pp , Dec [18] M. R. Luo, Applying colour science in colour design, Opt. Laser Technol., vol. 38, no. 4 6, pp , Jun [19] B. Moine and G. Bizarri, Why the quest of new rare earth doped phosphors deserves to go on, Opt. Mater. (Amst)., vol. 28, no. 1 2, pp , Jan [20] C. Ronda, Ed., luminescence from theory to application. Wilet-vch, 2008, p [21] B. G. Yacobi and D. B. Holt, Cathodoluminescence Microscopy of Inorganic Solids. new york and london: Springer US, 1990, pp [22] A. Lakshmanan, S.-B. Kim, B. G. Kum, H. M. Jang, and B. K. Kang, Rare earth doped CaSO 4 luminescence phosphors for applications in novel displays - new recipes, Phys. Status Solidi, vol. 203, no. 3, pp , Feb [23] J. McKittrick and L. E. Shea-Rohwer, Review: Down Conversion Materials for Solid-State Lighting, J. Am. Ceram. Soc., vol. 97, no. 5, pp , May [24] S.O.Kasap, Optoelectronics and Photonics, 2nd ed. New Jersey: Courier, 2013, pp

82 [25] Ben G Streetman and Sanjay Banerjee, solid state electronic devices, 5th ed. New Jersey: Tom Robbins, 2000, pp [26] S.-E. Tsou, Cathodoluminescence and Kinetics of gallium nitride doped with thulium, ohio, [27] J.-C. G. Bünzli, Benefiting from the unique properties of lanthanide ions., Acc. Chem. Res., vol. 39, no. 1, pp , Jan [28] B. S. Richards, Luminescent layers for enhanced silicon solar cell performance: Down-conversion, Sol. Energy Mater. Sol. Cells, vol. 90, no. 9, pp , May [29] P. Psuja, D. Hreniak, and W. Strek, Rare-Earth Doped Nanocrystalline Phosphors for Field Emission Displays, J. Nanomater., vol. 2007, pp. 1 7, [30] P. a. Nagpure and S. K. Omanwar, Combustion synthesis of borate phosphors for use in plasma display panels and mercury-free fluorescent lamps, Int. J. Self- Propagating High-Temperature Synth., vol. 22, no. 1, pp , Mar [31] D. Cavouras, I. Kandarakis, T. Maris, G. S. Panayiotakis, and C. D. Nomicos, Assessment of the gain transfer function of phosphors for application in medical imaging radiation detectors., Eur. J. Radiol., vol. 35, no. 1, pp , Jul [32] T. Aitasalo, P. Dereń, J. Hölsä, H. Jungner, J.-C. Krupa, M. Lastusaari, J. Legendziewicz, J. Niittykoski, and W. Stręk, Persistent luminescence phenomena in materials doped with rare earth ions, J. Solid State Chem., vol. 171, no. 1 2, pp , Feb [33] G. Blasse, Luminescence of rare earth ions at the end of the century, vol. 192, no. 1993, pp , [34] G. Särner, M. Richter, and M. Aldén, Investigations of blue emitting phosphors for thermometry, Meas. Sci. Technol., vol. 19, no. 12, p , Dec [35] a Neumann, J. J. Wierer, W. Davis, Y. Ohno, S. R. J. Brueck, and J. Y. Tsao, Four-color laser white illuminant demonstrating high color-rendering quality., Opt. Express, vol. 19 Suppl 4, no. May, pp. A982 90, Jul [36] Y. Xu, L. Chen, Y. Li, G. Song, Y. Wang, W. Zhuang, and Z. Long, Phosphorconversion white light using InGaN ultraviolet laser diode, Appl. Phys. Lett., vol. 92, no. 2, p ,

83 83 [37] S. Chhajed, Y. Xi, Y.-L. Li, T. Gessmann, and E. F. Schubert, Influence of junction temperature on chromaticity and color-rendering properties of trichromatic white-light sources based on light-emitting diodes, J. Appl. Phys., vol. 97, no. 5, p , [38] R. Hu, X. Luo, H. Feng, and S. Liu, Effect of phosphor settling on the optical performance of phosphor-converted white light-emitting diode, J. Lumin., vol. 132, no. 5, pp , May [39] E. E. C. Craig F. Bohren, Fundamentals of atmospheric radiation: an introduction with 400 problems, vol. 43, no. 11. Wily-VCH, [40] R. W. G. Hunt and M. R. Pointer, Measuring Colour. Chichester, UK: John Wiley & Sons, Ltd, [41] G. W. Meyer, Tutorial on color science, Vis. Comput., vol. 2, no. 5, pp , Sep [42] K. McLaren, Industrial applications of colour science, Phys. Technol., vol. 19, no. 2, pp , Mar [43] M. V. Orna, Chemical origins of color, J. Chem. Educ., vol. 55, no. 8, p. 478, Aug [44] A. Byrne and D. R. Hilbert, Color realism and color science., Behav. Brain Sci., vol. 26, no. 1, pp. 3 21; discussion 22 63, [45] B. R. Conway, S. Chatterjee, G. D. Field, G. D. Horwitz, E. N. Johnson, K. Koida, and K. Mancuso, Advances in color science: from retina to behavior., J. Neurosci., vol. 30, no. 45, pp , Nov [46] File:Cones SMJ2 E.svg - Wikimedia Commons. [Online]. Available: [Accessed: 24-Jan- 2014]. [47] C. I. E. S. Observers, Application Note. [48] W. S. S. Günter Wyszecki, Color Science: Concepts and Methods, Quantitative Data and Formulae, 2 Ed. Wiley-Interscience: John Wiley & Sons, [49] I. Processing, The Concept of Correlated Colour Temperature Revisited kos Borbe rpa ly, A mson, vol. 26, no. 6, pp , 2001.

84 84 [50] J. Smith, Calculating Color Temperature and Illuminance using the TAOS TCS3414CS Digital Color Sensor, Intell. OPTO Sens., no. 25, pp. 1 7, [51] E. D. Jones, LIGHT EMITTING DIODS ( LED S ) FOR GENERAL ILLUMINATION, [52] I. Morison, Introduction to Astronomy and Cosmology - Ian Morison - Google Books. John Wiley & Sons, Ltd, [53] File Blackbody-colours-vertical. [Online]. Available: [54] Lighting Research Center _ Education _ Learning _ Terminology _ Correlated Color Temperature. [Online]. Available: [55] What is the difference between Warm White and Cool White - Dongguan Jihai Bright LED Electronics Co. [Online]. Available: [56] Color Rendering _ Lighting Design Lab. [Online]. Available: [57] J. Schanda, Ed., Colorimetry Understanding the CIE System. John Wiley & Sons, [58] Technical _ 299 Lighting. [Online]. Available: [59] R. Mirhosseini, M. F. Schubert, S. Chhajed, J. Cho, J. K. Kim, and E. F. Schubert, Improved color rendering and luminous efficacy in phosphor-converted white light-emitting diodes by use of dual-blue emitting active regions., Opt. Express, vol. 17, no. 13, pp , Jun [60] T. W. Murphy, Maximum spectral luminous efficacy of white light, J. Appl. Phys., vol. 111, no. 10, p , [61] V. Lampret, J. Peternelj, and A. Krainer, Luminous flux and luminous efficacy of black-body radiation: an analytical approximation, vol. 73, no. 5, pp , [62] J. Song, What is Luminous efficacy meaning,definition, [Online]. Available:

85 [63] Y. Narukawa, M. Ichikawa, D. Sanga, M. Sano, and T. Mukai, White light emitting diodes with super-high luminous efficacy, J. Phys. D. Appl. Phys., vol. 43, no. 35, p , Sep [64] L. Hawkins, Seasonal affective disorders: the effects of light on human behaviour., Endeavour, vol. 16, no. 3, pp , Sep [65] J. F. Duffy, R. E. Kronauer, and C. a Czeisler, Phase-shifting human circadian rhythms: influence of sleep timing, social contact and light exposure., J. Physiol., vol. 495 ( Pt 1, pp , Aug [66] I. Iskra-Golec, a. Wazna, and L. Smith, Effects of blue-enriched light on the daily course of mood, sleepiness and light perception: A field experiment, Light. Res. Technol., vol. 44, no. 4, pp , Jun [67] P. R. Boyce, Review: The Impact of Light in Buildings on Human Health, Indoor Built Environ., vol. 19, no. 1, pp. 8 20, Mar [68] R. Paschotta, Encyclopedia of Laser Physics and Technology. Wiley-VCH, [69] X. C. Zundu Luo, Yidong Huang, Spectroscopy of Solid-state Laser and Luminescent Materials - Zundu Luo, Yidong Huang, Xueyuan Chen - Google Books. 2007, p [70] Shigeo Shionoya and William M yen, Ed., Phosphor Handbook - Google Books. 1999, pp [71] X. Zhang, J. Zhang, and M. Gong, Synthesis and luminescent properties of UVexcited thermal stable red-emitting phosphor Ba3Lu(PO4)3: Eu3+ for NUV LED, Opt. Mater. (Amst)., vol. 36, no. 4, pp , Feb [72] J. S. Kim, P. E. Jeon, J. C. Choi, H. L. Park, S. I. Mho, and G. C. Kim, Warmwhite-light emitting diode utilizing a single-phase full-color Ba[sub 3]MgSi[sub 2]O[sub 8]:Eu[sup 2+], Mn[sup 2+] phosphor, Appl. Phys. Lett., vol. 84, no. 15, pp , [73] D. W. D. N. Guo, H.You, C. Jia C, R. Ouyang, A Eu(2+) and Mn(2+)- coactivated fluoro-apatite-structure Ca6Y2Na2(PO4)6F2 as a standard whiteemitting phosphor via energy transfer, Dalt. Trans., vol. 43, no. 32, pp , [74] K. Li, D. Geng, M. Shang, Y. Zhang, H. Lian, and J. Lin, Color-Tunable Luminescence and Energy Transfer Properties of Ca 9 Mg ( PO 4 ) 6 F 2 : Eu 2 +, 85

86 86 Mn 2 + Phosphors for UV-LEDs, J. Phys. Chem. C, vol. 118, p , [75] G.Blasse and B. C. Grabmaier, Luminescent materials. Berlin ; springer, [76] N. Suriyamurthy and B. S. Panigrahi, Luminescence of BaAl2O4:Mn2+,Ce3+ phosphor, J. Lumin., vol. 127, no. 2, pp , Dec [77] J. Zhou, Y. Wang, B. Liu, and F. Li, Energy transfer between Eu Mn and photoluminescence properties of Ba[sub 0.75]Al[sub 11]O[sub 17.25] BaMgAl[sub 10]O[sub 17]:Eu[sup 2+],Mn[sup 2+] solid solution, J. Appl. Phys., vol. 108, no. 3, pp , [78] X. Yu, C. J. Summers, and W. Park, Controlling energy transfer processes and engineering luminescence efficiencies with low dimensional doping, J. Appl. Phys., vol. 111, no. 7, pp , [79] B. J. Chen, K. W. Jang, H. S. Lee, M. Jayasimhadri, E. J. Cho, S. S. Yi, and J. H. Jeong, SiO 2 effect on spectral and colorimetric properties of europium doped SrO 2 MgO x SiO 2 (0.8 x 1.6) phosphor for white LEDs, J. Phys. D. Appl. Phys., vol. 42, no. 10, p , May [80] B. M. J. S. and J. G. Verlijsdonk, THE LUMINESCENCE PROPERTIES OF Eu 2+- AND Mn2+-DOPED BARIUM HEXAALUMINATES, Mat. Res. Bull, vol. 21, no. c, pp , [81] J. M. P. J. Verstegen and L. Division, A Survey of a Group of Phosphors, Based on Hexagonal Aluminate and Gallate Host Lattices, J. Electrochem. Soc., vol. 121, no. 12, pp , [82] U.CALDINO G. and J.RUBIO O., Optical spectroscopy of Mn2+ ions in CaCl2 single crystals, Radiat. Eff. Defects Solids, vol. 127, no. 1, pp , [83] C.-C. Chiang, M.-S. Tsai, and M.-H. Hon, Luminescent Properties of Cerium- Activated Garnet Series Phosphor: Structure and Temperature Effects, J. Electrochem. Soc., vol. 155, no. 6, p. B517, [84] L. S. Rohwer and J. E. Martin, Measuring the absolute quantum efficiency of luminescent materials, J. Lumin., vol. 115, no. 3 4, pp , Nov [85] Y. R. Do and J. W. Bae, Application of photoluminescence phosphors to a phosphor-liquid crystal display, J. Appl. Phys., vol. 88, no. 8, p. 4660, 2000.

87 [86] O. Optics, Spectrometer Operating Software Installation Installation and Operation Manual. [Online]. Available: 87

88 APPENDIX A: LIST OF CONFERENCES PRESENTATIONS AND JOURNAL PAPERS W. M. Jadwisienczak, F. Rahman and S. Al-Waisawy, Title: Investigation of Color Phosphors for Laser-driven White Lighting. Paper was presented during the 17 th International Conference on Luminescence and Optical Spectroscopy of Condensed Matter (ICL2014), July, 2014 Wroclaw, Poland. 2. F. Rahman, W. M. Jadwisienczak, S. Al-Waisawy and J. Wright, Title: Laser Diodedriven White Light Source System Design, Phosphors, Light Characteristics and Variability Studies. Paper was presented during the 10 th International Symposium on Semiconductor Light-emitting Devices December 14-19, Kaohsiung, Taiwan. 3. S. Al-Waisawy, W. M. Jadwisienczak, J. Wright, D. Pendrill and F. Rahman, Title: Laser Excitation of Red, Green, Blue and Trichromatic White Rare-earth Phosphors for Solid-state Lighting Applications. Paper was submitted to Journal of Luminescence in November F. Rahman, S. Al-Waisawy, J. Wright, and W. M. Jadwisiencsak, Title: A Laserdriven White Light Source with Trichromatic Phosphor for Solid-state Lighting. Paper was submitted to Applied Optics in November 2014.

89 89 APPENDIX B: DERIVATION OF THE COLOR X, Y, Z SYSTEM FROM COLOR R, G, B SYSTEM The CIE coordinates convert tristimulus R,G,B values to X,Y,Z values to eliminate the color negative values that are seen in the color matching function for CIE 1931RGB. The following equations for conversion are: X = 0.49R G B Y = R G B Z = 0.00R G B (A-1) (A-2) (A-3) These numbers were chosen carefully to make all color values positive. Moreover, the CIE 1931 xʹ, yʹ, zʹ color matching function (see Fig. 2-5) was derived from the color matching function for CIE 1931RGB in (Fig. 2-4) based on the equations shown below: xʹ(λ) = 0.49rʹ(λ) gʹ(λ) bʹ(λ) yʹ(λ) = rʹ(λ) gʹ(λ) bʹ(λ) zʹ(λ) = 0.00rʹ(λ) gʹ(λ) bʹ(λ) (A-4) (A-5) (A-6) Then the CIE 1931 (x, y) chromaticity coordinates can be derived from X, Y, and Z in order to show the relativity of tristimulus values using the equations shown below: x = X/(X + Y + Z) y = Y/(X + Y + Z) z = Z/(X + Y + Z) (A-7) (A-8) (A-9) The sets of above equations indicate that (x + y + z = 1). Therefore, it becomes easy to represent color parameters in two different dimension chromaticity diagrams based on (x, y) and (z = 1 x y). At equal energy white light, the trichromatic colors are equal

90 90 (R = G = B) as well as their corresponding values (X = Y = Z). As a result, the chromaticity coordinates of equal energy white light are represented using the relationship (x = y = z = 1/3).

91 APPENDIX C: SOFTWARE PROCEDURES FOR MEASURING PHOSPHOR EMISSION SPECTRA 91 The following are general procedures for measuring intensity spectra. The specifics of the procedures, typical measurement parameters, and computer hardware/software system requirements can be found in the Ocean View software manual (page 43) titled Quick View Minus Background which is reproduced below: Quick View Minus Background Select Quick View mode from the Quick View software and subtract the stored background spectra from each spectrometer channel before displaying them using Ocean View. To execute this operation perform as following: 1. Invoke the Quick View Minus Background wizard by clicking the Create New Spectroscopy application icon. 2. Select either an active acquisition or start a new one. 3. Click Spectroscopy, then Quick View Minus Background, then next. 4. Set acquisition parameters. 5. Store background spectrum. The Ocean View software manual can be accessed through Ref [86].

92 92 APPENDIX D: COLOR MEASUREMENT PROCEDURES The following describes Ocean View software vr features enabling operator to measure color parameters, (CIE 1931 chromaticity coordinates, and Correlated Color Temperature (CCT), etc.) of the luminescent sample. The Ocean View software has different procedures for measuring color parameters, depending on desired choice, described in Step 3 of the Color Wizard manual (see page 48). In this project, we have measured the color parameters by choosing the New Relative Irradiance Processing option, then letting the software leads us through the process of measuring color parameters based on the relative irradiance spectrum collected during the measurement procedure. The following steps show the method of color parameters measurement for an arbitrary luminescent sample. Color Wizard With Ocean View software active, the operator can calculate and report the Color Rendering Index (CRI) and Correlated Color Temperature (CCT) for investigated samples. The Ocean View software will compute all color parameters, except RGB coordinates, using the following options: Emissive color measurements require a spectrometer that has been calibrated to take irradiance readings (either relative or absolute). The Ocean View software gives the operator options of choosing to measure either emissive or reflective color from either absolute or relative irradiance.

93 93 Reflective color measurements do not require an irradiance calibration, but they do require a reference spectrum. This reference scan is taken with the same optical configuration, except that a perfect reflector is placed at the same location as the samples. After choosing from the above two options, the procedure will be as follow: 1. Invoke the Color Wizard by clicking the Create New Spectroscopy application icon. 2. Select Color option. 3. In the Color Source panel, select the processing method used to calculate color data. By choosing any of the first three options, the operator will be guided by the wizard through the steps associated with that choice before being brought back to complete the color wizard. If there is a suitable existing process that has already been completed, this will be available to select under the last option, namely Existing Processing. Select the desired process and click next. Note that Correlated Color Temperature (CCT) and Color Rendering Index (CRI) will not be available if operator chooses Percent Reflection as the desired processed data source. 4. In the Color Setup panel, the Display Parameters portion allows the operator to choose which color values are displayed upon completion of the wizard. All applicable values will be added to a color table for display. Note that Correlated Color Temperature (CCT) and Color Rendering Index (CRI) will not be available if the operator chooses Percent Reflection as the processed data source.

94 !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! Thesis and Dissertation Services!