This PDF file contains the front matter associated with SPIE Proceedings Volume 6669, including the Title Page, Copyright information, Table of Contents, Introduction (if any), and the Conference Committee listing.

Vertical GaN based Light Emitting Diodes on metal alloy substrate (VLEDMS) were realized and characterized for solid state lighting application. An efficiency of more than 100 lumens/watt from a white LED was achieved. And, an efficiency of more than 80 lumens/watt from a high efficiency and high power green LED was achieved also. The dissipate heat more effectively than conventional and flip-chip LEDs, thanks to the higher thermal conductivity of a copper alloy substrate. This increases their maximum operating current and output power and makes them more suitable for solid-state lighting applications. In addition, these VLEDMS exhibit many advantages over those on sapphire under extreme operation conditions for general lighting application.

One of the methods for creating white light with light-emitting diodes (LEDs) is mixing radiations from several different colored LEDs. Mixed-color LEDs are expected to have greater luminous efficacy because they do not undergo down-conversion losses like phosphor-converted white LEDs. However, in reality mixed-color LED systems require extra optical elements to reduce spatial color variation and create uniform white light. Optical diffusing techniques commonly used for these purposes cause light loss, because some portion of the light is scattered back toward the LEDs where it is absorbed and lost. In 2004, a technique known as scattered photon extraction (SPE) was used to extract backscattered light from the phosphor layer of phosphor-converted white LEDs to increase overall light output. In this study, it was hypothesized that by using similar SPE optics with optical diffusers, the spatial color uniformity of mixed-color white LED systems could be improved without sacrificing the overall luminous efficiency. With this new approach, both a microsphere-doped diffuser and SPE optics were utilized. The experiments showed that the proposed setup did increase the spatial color uniformity of the mixed-color LED system with more than 79% overall optical efficiency. This study also demonstrated the effects of microsphere size, concentration, and diffuser thickness on spatial color uniformity and optical efficiency.

Many luminaires have an array of LEDs that illuminate a lenslet-array diffuser in order to create the appearance of a single, extended source with a smooth luminance distribution. Designing such a system is challenging because luminance calculations for a lenslet array generally involve tracing millions of rays per LED, which is computationally intensive and time-consuming. This paper presents a technique for calculating an on-axis luminance distribution by tracing only one ray per LED per lenslet. A multiple-LED system is simulated with this method, and with Monte Carlo ray-tracing software for comparison. Accuracy improves, and computation time decreases by at least five orders of magnitude with this technique, which has applications in LED-based signage, displays, and general illumination.

A direct or bottom LED backlight is a key concept in large area LCD displays because it does not use a light guide, is flat, and is easy to assemble. In this paper, a method of luminance management for a bottom LED backlight is proposed and demonstrated. We analytically calculate both the power consumption and brightness uniformity in function of: screen brightness, screen size, backlight thickness, transmittance of the LCD panel, reflective cavity efficiency, gain and cone angle of enhancement films, LED array configuration, and the average luminous flux and radiation pattern of a single LED. Moreover, a 42-inch LCD television with this backlight design approach is made and demonstrated. The bottom backlight incorporates an array of RGGB 4-in-1 multi-chip LEDs within a highly reflective box behind a diffuser and a dual brightness enhancement film. We predict with an accuracy of 94% the brightness uniformity and with 96% the luminance level.

We will discuss the performance, progress and trend of High Power Light Emitting Diodes (HP-LEDs), suitable for high luminance applications like micro-display projection, car headlamps, spot lamps, theatre lamps, etc. Key drivers for the high luminance applications are LED parameters such as internal quantum efficiency, extraction efficiency, drive current, operating temperature and optical coupling efficiency, which are important for most applications as they also enable higher lumen/$ ratios. Historical progress, prospects for improving these parameters and potential optical luminance enhancement methods to meet the demands for the various illumination applications are presented.

We adapt the tenets of Hering's opponent color theory to the processing of data obtained from a tristimulus colorimeter to independently determine the intensity and possible peak wavelength shift of a narrowband LED. This information may then be used for example in an optical feedback loop to maintain constant intensity and chromaticity for a light source consisting of two LEDs with different peak wavelengths. This approach is particularly useful for LED backlighting of LCD display panels using red, green, and blue LEDs, wherein a tristimulus colorimeter can be used to maintain primary chromaticities to within broadcast standard limits in real time.

This research intends to explore with a uniform color space based on the CIE 1931 x-y chromatic coordinate system. The goal is to improve the non-uniformity of the CIE 1931 x-y chromaticity diagram such as to approach the human color sensation as possible; however, its simple methodology still can be kept. In spite of the existence of various kinds of the uniform color coordinate systems built up early (CIE u'-v', CIE Lab, CIE LUV, etc.), the establishment of a genuine uniform color space is actually still an important work both for the basic research in color science and the practical applications of colorimetry, especially for recent growing request in illumination engineering and in display technology. In this study, the MacAdam ellipses and the Munsell color chips are utilized for the comparison with the human color sensation. One specific linear transformation matrix is found for the CIE 1931 color matching functions (see manuscript) to become the novel uniform ones. With the aid of the optimization method, the transformation matrix can be easily discovered and makes the 25 MacAdam ellipses are similar to each other in the novel uniform color space. On the other hand, the perfectiveness of the equal-hue curves and the equal-chroma contours from the Mnusell color chips evaluates for the best optimization conditions among several different definitions for the similarity of all the MacAdam ellipses. Finally, the color difference between any two colors can be simply measured by the Euclidean distance in the novel uniform color space and is still fitted to the human color sensation.

Light sources which produce a high color rendering index (CRI) have many applications in the lighting industry today. High color rendering accents the rich color which abounds in nature, interior design, theatrical costumes and props, clothing and fabric, jewelry, and machine vision applications. Multi-wavelength LED sources can pump phosphors at multiple stokes shift emission regimes and when combined with selected direct emission sources can allow for greater flexibility in the production of warm-white and cool white light of specialty interest. Unique solutions to R8 and R14 CRI >95 at 2850K, 4750K, 5250K, and 6750K presented.

The optical and structural properties of InGaN/GaN multi-quantum wells (MQWs) grown on sapphire by metalorganic chemical vapor deposition (MOCVD) have been investigated by optical measurements of temperature-dependent photoluminescence (PL), photoluminescence excitation (PLE) and time-resolved photoluminescence (TRPL), and structural analysis methods of high-resolution X-ray diffraction (HRXRD) and high-resolution transmission electron microscopy (HRTEM). Two typical samples are studied, both consisting of five periods of InGaN wells with 1.6 nm thickness with different indium compositions of 21 and 24%, respectively. The thickness of the GaN barrier was 7.9 nm. According to the PL and PLE measurement results, large values of activation energy and Stokes' shift are obtained. This indicates that higher In composition results in the increase of composition fluctuation in the InGaN MQW region, showing the stronger carrier localization effect. The lifetime at the low-energy side of the InGaN peaks is longer for higher indium composition, as expected from the larger Stokes shift.

Key points for high efficiency LED are reviewed as internal and external structure. Then, the effect of silicon doping in the selected barrier on the electroluminescence of InGaN/GaN MQW-LED is explained with specially designed dual wavelength LEDs. The result verifies that the hole carrier transport is easily blocked by the silicon doped barrier, and the dominant electron and hole recombination occurs at the wells between p-GaN and the silicon doped barrier.

The properties of several solid state lighting phosphor compounds have been enhanced through the application of advanced material processing and particle encapsulation techniques. Also higher light extraction efficiencies from the pump LED were achieved by index-matching the refractive index of the phosphor to that of the InGaN chip. New solid-state- reaction protocols have produced a significant increase in the intrinsic efficiency of the orthosilicate yellow phosphor compared to the traditional YAG:Ce phosphor. In addition, the stability of phosphor materials in high humidity environments was increased significantly using metal oxide coatings applied by vapor deposition.

A single-chip white light LED is commonly modeled by considering the phosphor coating as a homogeneous Lambertian light source. However, this approach leads to an incorrect optical simulation of phosphor-coated multi-chip LEDs due to the presence of a previously unreported spatial distribution of emission spots across the phosphor layer. We introduce "weighting" factors based on position-dependent light strength across the phosphor surface in order to improve the model accuracy. Following the modeling algorithm in the mid-field region, we have built up a precise and practical optical model by using Monte Carlo ray tracing and weighting factors. We measure the LED radiation distribution at several representative distances to test the model performance. In all cases, the accuracy is higher than 99.5% in normalized cross correlation between the simulated pattern and experimental measurement.

With the advent of a new generation of high brightness LEDs especially in the blue spectral range, white light technology based on phosphor conversion gains maturity for a successful penetration of, e.g., the general lighting market within the next years. Major challenges ahead are originating from the specific demands of the markets on small emission areas, highest possible intensities, long-term color stability, and spatial homogeneity of color coordinates. The LED industry more or less relies on the conversion phosphor classes of YAG, Sulfates, and Silicates, embedded in silicone matrices. A number of conformal coating technologies are in use. The optimization of the coating geometries with the help of software simulation offers a high potential for increased angular color homogeneity and high package densities, especially when secondary optics is in use. We report on recent progress in simulating parameters for improved white LEDs manufactured by coating technologies.

We have studied structure-property relations of Eu(II) doped nitridosilicates M₂Si₅N₈ and MSi₇N₁₀ (M = Sr, Ba). For both systems that are described as being efficient LED phosphors, we show that a detailed examination of the local activator environment in combination with net positive charge calculations for Eu with the EHTB-MO method allows a qualitative prediction of the luminescence properties of nitridosilicate LED phosphors. The non-linear shift of the amber to red emission of solid solutions Ba_2-xSr_xSi₅N₈:Eu is explained by a non-statistical distribution of Eu over the available lattice sites. The Stokes shift differences between the two available Eu sites are significantly larger for Ba₂Si₅N₈:Eu than for Sr₂Si₅N₈:Eu. In contradiction to literature data, BaSi₇N₁₀:Eu shows emission in the cyan spectral region and a large Stokes shift, in accordance with the host lattice geometry and the electronic structure calculations that were carried out. SiAlON formation as an additional design tool to tune nitridosilicate phosphor emission properties is demonstrated for Sr₂Si_5-xAl_xO_xN_8-x:Eu. (Al,O) incorporation leads to anisotropic changes of lattice constants and a red shift and broadening of the Eu emission.

We have recently demonstrated white-light devices based on single size, ultra-small cadmium selenide (CdSe) nanocrystals that possess broadband emission (420-710 nm). When encapsulated in a polymer and coated on commercial UV LEDs, these nanocrystals act as frequency downconverters to produce high quality white light. These devices emit very pure white light with excellent CIE coordinates (0.35, 0.31) and a very high color rendering index of 93 when excited with a 365 nm LED. The relatively low luminous efficiency of preliminary devices, up to 1 lm/W, is a result of unoptimized device geometry, low light extraction efficiency, low UV LED efficiency, and low nanocrystal quantum yield (~10%). In order to address the issue of efficiency, we have pursued two studies. First, we investigate the use of photonic crystal mirrors, chirped mirrors, and thin film filters to increase the extraction efficiency and decrease UV light leakage from the device. Simulation results of chirped mirrors consisting of alternating layers of silicon dioxide (SiO₂) and hafnium dioxide (HfO₂) have demonstrated high reflectance throughout the visible region, thus increasing extraction efficiency by reflecting emitted white light out of the device that would normally be lost. Second, we examine the tradeoff of color quality and efficiency that results from using blue LEDs for excitation: unabsorbed blue light will imbalance the white-light spectrum, but the conversion of blue to white light yields a lower Stokes loss than UV excitation light.

Solid State Lighting is becoming increasingly more advanced, both in terms of lumen output as well as energy efficiency. At the same time, packages emitting enough lumens for lighting applications are decreasing in size. This smaller packaging enables several new applications. In this paper we will discuss one of these new applications: low cost, large, flexible and very thin light emitting surfaces. Our approach consists of using very thin transparent lightguides. Due to their limited thickness, these lightguides are quite flexible. Tiny low power, side-emitting LEDs are used to couple light into these lightguides. A carefully calculated outcoupling structure ensures light is coupled out uniformly. Although this general principle is known, some aspects are new to our approach. The flexibility of our thin lightguides can be very useful for numerous lighting applications; a radius of curvature of just a few centimeters is easily obtained, while still maintaining good outcoupling and uniformity. Furthermore, we show that for several geometries, a perfect homogeneous brightness can only be obtained using a precise pattern and density distribution of outcoupling structures.

On airport runways, blue light fixtures denote taxiways between the runway and the airport terminal. Blue optics transmit mostly short-wavelength radiation, which makes traditional incandescent lamps a poor choice of light source; the resulting fixture efficiency could be less than one percent. LEDs are replacing incandescent lamps in this application. But unlike incandescent sources, LEDs do not radiate enough heat to melt ice and snow from the fixture optics. To meet Federal Aviation Administration (FAA) regulations for weatherability, some LED-based fixtures incorporate electric heaters that, when switched on, nearly negate the energy-savings benefit of converting to LED sources. In this study, we explored methods for conduction and convection of LED junction heat to taxiway fixture optics for the purpose of minimizing snow and ice buildup. A more efficient LED-based system compared to incandescent that would require no additional heaters was demonstrated.

A newly developed thin-GaN LED structure has shown great advantages over traditional LED schemes in the lighting efficiency performance. Yet, in the fabrication process of thin-GaN LED chip, several processes still remain to be optimized. In this paper, the process issues of the promising thin-GaN LED chip will be discussed.

ZnO has been considered as a substrate for epitaxial growth of III-Nitrides due to its close lattice and stacking order match. This paper will cover growth of In_xGa_1-xN epitaxial layer on lattice-matched ZnO substrates by metal-organic chemical vapor deposition (MOCVD). InGaN of various indium compositions from different growth temperatures were well controlled in the InGaN films on ZnO substrates. High-resolution X-ray diffraction (HRXRD) confirmed the epitaxial growth of InGaN film on ZnO. The optical and structural characterization of InGaN epilayer on ZnO substrates was measured by room temperature photoluminescence, temperature-dependent photoluminescence, and field-emission secondary electron microscope. In addition, a transition layer of Al₂O₃ on ZnO substrates have been employed for InGaN growth to help prevent Zn and O diffusion into the epilayers as well as assist nitride epilayer growth. HRXRD results show a single crystal InGaN film has been successfully grown on annealed Al₂O₃ coated ZnO substrates.

This paper reports a study on visual evaluation of colors under LED lighting operated by an energy-saving control strategy. Digitally controlled LED systems can produce various qualities of light, adjustable to users' requirements. In this context, a novel control concept inspired this research: strategic control of Red, Yellow, Green & Blue LEDs forming white light can further increase energy efficiency. The resulting (more efficient) light, however, would have decreased "color rendering" (ability of accurately reproduce colors). The notable point is that while reducing color rendering, color temperature and light levels can stay constant and hence the appearance of the modified light could stay the same, and only the colors of illuminated objects would change. But how spaces would be perceived under such light with changing color rendering is a key question. This research investigated the hypothesis that a significant range of color distortions would be unnoticeable under such dynamically controlled illumination, especially outside of users' main field of view. If successful, such control technique could be implemented for unoccupied spaces that would not tolerated dimming, and minimize peak hours energy waste, potentially enabling significant power reductions. Three incremental series of experiments were performed based on subjective assessment of colors under changing color rendering. Tests were carried out for central and peripheral vision, using laboratory booths (phase 1) and full scale architectural mockups (phase 2). Results confirmed the fundamental hypothesis, showing that the majority of subjects did not detect the color changes in their periphery while the same color changes were noticeable with direct observation.

The package and system level temperature distributions of a high power (>1W) light emitting diode (LED) array has been investigated using numerical heat flow models. For this analysis, a thermal resistor network model was combined with a 3D finite element submodel of an LED structure to predict system and die level temperatures. The impact of LED array density, LED power density, and active versus passive cooling methods on device operation were calculated. In order to help understand the role of various thermal resistances in cooling such compact arrays, the thermal resistance network was analyzed in order to estimate the contributions from materials as well as active and passive cooling schemes. An analysis of thermal stresses and residual stresses in the die are also calculated based on power dissipation and convection heat transfer coefficients. Results show that the thermal stress in the GaN layer are compressive which can impact the band gap and performance of the LEDs.

A LED material evaluation technology called EpiEL was introduced to measure electroluminescence (EL) directly on LED epiwafers. Based on EpiEL technology, a virtual LED device fabrication & characterization system has been developed which can be used to measure LED device parameters on epiwafers without any costly and time-consuming device fabrication. As through a finished device, the system reveals not only the electro-luminescence (EL) but also the electrical properties of the material. The developed EpiEL mapping system can rapidly obtain EL spectrum, LIV, output characteristics, wavelength & FWHM shift curves, as well as wafer-level uniformity about slope quantum efficiency, emission intensity, peak/dominant wavelength (WLP/WLD), FWHM, driving voltage/current, etc... EL measurement is usually performed on finished device (such as LED) since it needs a device structure to inject current. EpiEL mapping system overcomes this limitation by instantly forming a well-defined LED device inside the material. With such capability, EpiEL technology provides a unique electroluminescence solution for optoelectronic (especially emerging solid-state lighting) industry which brings better capability and efficiencies: instant response for material development and device-level quality control right after material growth.

In this paper, we discuss the variations of thermal resistance coefficient from junction to board (Rθ_jb) for high-power light-emitting diodes (LEDs) as a result of changes in power dissipation and ambient temperature. Three-watt white and blue LED packages from the same manufacturer were tested for Rθ_jb at different input power levels and ambient temperatures. Experimental results show that Rθ_jb increases with increased input power for both LED packages, which can be attributed to current crowding mostly and some to the conductivity changes of GaN and TIM materials caused by heat rise. With increasing ambient temperature, Rθ_jb increased but not as much as what was observed with drive current increase.

Superheterodyne techniques were originally developed for radio transmission and reception nearly a century ago. In this paper we explore the adaptation of this technology to the problem of simultaneously monitoring the intensities of multiple LED channels with a single photosensor. The use of superheterodyne techniques obviates the need for multiple photosensors filters and tristimulus color filters to monitor the relative intensities of red, green, and blue LEDs. In addition, they alleviate the problems of electrical and optical noise, as well as the influence of ambient illumination on the photosensors. They can also be used to advantage with phosphor-coated white light LEDs in solid state lighting systems. Taking a broader view, the use of such techniques demonstrates the value of looking outside the realm of conventional LED power and control technologies when designing solid state lighting systems.

In this paper we report the analysis of thermal stability of High Brightness Light Emitting Diode subjected to thermal and bias ageing. The degradation mechanisms of several families of commercial available devices were investigated. In the first part of the work we estimated thermal resistance and thermal behaviour under dc bias condition. After this thermal characterisation two different ageing tests were carried out on devices: thermal aging at high temperature levels without biasing the devices and accelerated dc stress at nominal current value (400mA). At each step a complete electrical and optical characterisation of aged devices was performed, in order to find a correlation between different aging and a better understanding of degradation mechanism. This characterisation included I-V measurements, optical power vs current characteristics and spectral analysis. During thermal stress we observed the increase of forward voltage at nominal current and the degradation of optical power with nearly exponential kinetics. We found that lifetimes were well correlated with stress temperature: therefore it was possible to find an activation energy of degradation mechanism of about 1.5eV. Moreover, modifications of spectral properties during electrical and thermal stress were found. Thus, a package level analysis was carried out in order to clarify the role of modification in optical properties of reflector cup and the efficiency of phosphors. Finally, evaluation of differential structure functions indicated that stress induces also the worsening of the properties of the chip-to-package thermal path: this phenomenon has been attributed to the partial detachment and degradation of the ohmic contacts.

We investigate improvement in performance attainable by etching Photonic Crystals and Photonic Quasicrystals (PQC) into the top emitting surface of LEDs (PQC-LEDs). We describe the physical mechanisms of extraction enhancement through ordered surface patterning and investigate benefits in terms of total extraction enhancement, beam directionality, and far field beam quality. We show that a great variety of far field beam profiles tailored for specific applications such as projection TV light engines and direct flat panel display illumination can be obtained simply by adjusting geometric design parameters. Our results show that PQC-LEDs can provide around 50% improvement in extraction enhancement and coupling efficiency for applications requiring non Lambertian beam shapes when etched into standard epitaxy wafers in comparison to "state of the art" surface roughened GaN LED devices. We also show that by tailoring the epi-structure to complement the properties of the PQC, extraction as well as directional light coupling enhancements can be improved by a factor of 9 amounting to a total coupling enhancement of around 78%.

Nanofibers made from non-absorbing polymers such as poly(methyl methacrylate) are solid structures that have one dimension (diameter) in the 10-1,000 nanometer (nm) range, while the other dimension (length) can be quite long. These nanofibers can be formed in either an oriented or random packing structure, and the surface morphology of the fiber can range from smooth to nanoporous. Quantum dots (QD) or other luminescent nanoparticles (diameter 1-10 nm) can be added to the nanofiber to create the photoluminescent nanofiber (PLN). Because PLNs are nanocomposites of fluorescent nanoparticles and polymer nanofibers, the optical properties of the nanocomposite, including absorption, emission, and light scattering, can be tailored for application-specific requirements. Nanofibers may have several applications in solid-state lighting, including serving as a light diffuser, providing optical filtering of low photopic sensitive wavelengths (i.e., blue) to increase conversion to higher luminosity wavelengths, and providing a convenient vehicle for handling and blending QDs to achieve a high color-rendering index.

The popularity of multi-primary color LED light engines presents unique challenges to optical design. Multiple design requirements must be satisfied including the achievement of uniform color projection in the field for archetainment, accent, and show lighting, as well as high illuminance at distances greater than 5 meters. A novel collimation optic is presented which includes multiple combinations of color enhancement and collimation features.

We introduce an alternative to pulse width modulation (PWM) for LED intensity control called "Pseudorandom Pulse Code Modulation." Rather than using a single pulse with a variable duty factor to control the time-averaged LED drive current, we generate a pseudorandom pulse code to perform the same function. The advantages of this technique include: 1) the ability to control multiple channels without the need for dedicated PWM hardware circuitry; 2) drive current averaging; and 3) reduction of acoustic noise due to magnetostriction.

A tool to predict the behavior of LED-based luminaires is critical to their design. In the absence of such a tool, the design process becomes quite laborious and highly dependant on expensive experimental work. Unfortunately, thermal effects can make the system level behavior very difficult to predict: a change in temperature causes a change in spectral characteristics, which in turn causes an adjustment to the balance of the LEDs, affecting the heat load, and thereby once again changing the spectral characteristics. In order to accurately predict how a SSL luminaire will behave, it is necessary to model it at the system level. An accurate model must consider heat loading/dissipation, the response of electrical components to temperature, the effect of temperature on spectral characteristics (including intensity, spectral bandwidth, and peak wavelength), and then recursively recalculate the heat load. We have developed just such a model for a luminaire employing optical feedback and thermal feedforward. The model makes use of measured data for the components, and computes its system-level behavior. The model also computes the change in behavior due to aging, based on the junction temperature. The model has been verified by experiment, and found to agree to within ten percent. The aging predictions have not yet been verified.

The photometric modeling of LEDs as generalized Lambertian sources (GL-Sources) is discussed. Non-Lambertian LED sources, with axial symmetry, have important real-world applications in general lighting. In particular, so-called generalized Lambertian sources, following a cosine to the nth power distribution (n≥1), can be used to describe the luminous output profiles from solid-state lighting devices like LEDs. For such sources, the knowledge of total power (in Lumens [Lms]), the knowledge of the output angular characteristics, as well as source area, is sufficient information to determine all other critical photometric quantities such as: maximum radiant intensity (in Candelas [Cd = Lm/Sr]) and maximum luminance (in nits [nts = Cd/m²]), as well as illuminance (in lux [lx = Lm/m²]). In this paper, we analyze this approach to modeling LEDs in terms of its applicability to real sources.

With the invention of Ultra-Bright LED, solid state lighting has come to something which is much more efficient and energy saving when compared to conventional incandescent or fluorescent lighting. With the use of proper driver electronics now a days it is possible to install solid state lighting systems with the cost same as that of any other lighting technology. This paper is a part of the research project we are doing in our lab, which deals with using ultra bright LEDs of different colors for lighting applications. The driver electronics are made in such a way that, the color and brightness of the lights will change according to context. For instance, if one of the users is reading a story or listening to music in a Personal Computer or in a hand held device such as a PDA, the lighting systems and the HVAC (Heating Ventilation Air-conditioning) systems will change dramatically according to the content of the story or the music. The vulnerability of solid-state lighting helps to accomplish such an effect. Such a type of system will help the reader to feel the story mentally and physically as well. We developed complete driver electronics for the system using multiple microcomputers and a full software suite which uses complex algorithms to decode the context from text or music and synchronize it to lighting and HVAC information. The paper also presents some case-study statistics which shows the advantage of using the system to teach kindergarten children, deaf and dumb children and for language learning classes.

We present an electrical model for InGaN based high power light-emitting diodes (LEDs) with self-heating effect. The model consists of an ideal diode connected with a series resistor. Because of the self-heating effect, the temperature of the p-n junction (T_j) is much higher than case temperature (T_C), the maximum T_j in our experiment is 148K high above case temperature. For the T_j rises caused by self-heating effect, some temperature-dependent parameters, such as the band gap (E_g) of InGaN which influence the reverse saturation current, the hole mobility (μ_p) which influence the series resistance of p-type GaN, can be written as the function of forward current (I) and voltage (U). The voltage drops on p-n junction and equivalent series resistance are extracted from the measured I-V curve, and then the values of equivalent series resistance at different forward power are obtained. By the self-heating effect, the values of equivalent series resistance show a very strong forward power-dependent characteristics, the maximum equivalent series resistance in this paper is about 1.6Ωat high forward power, nearly twice of the minimum one. It is worthwhile to point out that the equivalent series resistance changed by self-heating effect and its influence on electronically property of GaN based LEDs can not be neglected even at the low resistance level about 1Ωin our experiment.

We describe the mode solutions for the Helmholtz Equation using the operator formalism. The study is extended to the structural solution for the focused non-linear Schrödinger equation (NLSE). With this treatment, we obtain for the NLSE a reduced partial differential equation, whose characteristic solution has an eikonal structure which allows us a geometrical analysis. Focusing region in non-linear media is described by means of an envelope region of eikonal trajectories establishing similar behaviors with caustic structures. In particular, if the boundary condition consists of a slit shape curve, the focusing profile corresponds with the evolute of the curve. In general, the profile satisfies a non-linear partial differential equation whose structure remains non-variable under changes of variables which may represents scaling or rotations. This feature permits us to extend the analysis to other kind of focusing regions, such as focusing vortex.

The morphology and optical properties of Si samples implanted by low-energy Fe⁺ ions with different fluencies (1×10¹⁵ - 1.8×10¹⁷ cm^-2) and further subjected to pulsed ion-beam treatment (PIBT) have been studied by atomic force microscopy and optical reflectance spectroscopy. It was proved that the iron disilicide ( β-FeSi₂) crystallites have been formed on the surface of Si substrate as a result of ion implantation and PIBT. The method of ultrahigh vacuum and low-temperature (Τ = 850°C) cleaning of Fe⁺-implanted Si samples has been used for the first time. It was found that it is possible to form smooth epitaxial Si films with reconstructed surface and thickness up to 1.7 μm by molecular beam epitaxy (MBE) on the surface of Si samples implanted at a fluence of up to 1×10¹⁶ CM^-2. Further increasing implantation fluence results into disruption of epitaxial Si growth and strong increase of surface relief roughness due to 3D silicon growth mechanism. Preservation of β-FeSi₂ precipitates inside Si matrix after the formation of a cap epitaxial Si layer has been confirmed by optical spectroscopy data. Low temperature photoluminescence measurements in the range of 1400-1700 nm showed that light emission of the Si/ β-FeSi₂/Si heterostructures formed is due to contributions from β- FeSi₂ precipitates and dislocations.

The material properties of zincblende Al_xGa_1-xN alloys from first-principles calculations are investigated. It is found that the bulk moduli of zincblende Al_xGa_1-xN increase with an increase of aluminum composition x, but the pressure derivative of bulk modulus decreases with an increase of aluminum composition x. The bulk modulus of 199.82 ± 0.64 GPa for GaN increases to that of 213.03 ± 1.09 GPa for AlN, and the deviation parameter of bulk modulus for zincblende Al_xGa_1-xN is 11.18 ± 3.61 GPa. The pressure derivative of bulk modulus of 4.190 ± 0.020 for GaN decreases to that of 3.589 ± 0.030 for AlN. The refractive indices in the wavelength range of 300-600 nm are also investigated. The one-Sellmeier-term expression is used for curve-fitting analysis of the refractive index as a function of the wavelength in the transparent region, and the results should provide for the design and growth of vertical cavity structures and distributed Bragg reflectors on GaAs substrates. Finally, the direct bowing parameter of 0.446 ± 0.126 eV and indirect bowing parameter of -0.360 ± 0.242 eV are obtained. There is a direct-indirect crossover near x = 0.654 for which the band gap energy is 4.573 eV.

The effect of spontaneous and piezoelectric polarizations on optical characteristics of blue InGaN light-emitting diodes (LEDs) is investigated in this work with the APSYS simulation program. Specifically, the band diagram, carrier distribution, radiative recombination current, and light-current performance curves are discussed for various polarization situations. According to the simulation results, the amount of polarization charges is dominated by piezoelectric polarization. Therefore, the band diagram and overlap between electrons and holes in quantum wells are improved effectively and the light-current performance is raised apparently as the piezoelectric polarization is removed and only the spontaneous polarization is taken into account. The possible reason is that the influence of piezoelectric polarization due to lattice constant mismatch on optical properties is severe than the spontaneous polarization due to asymmetry of the wurtzite structure along the c-axis. Moreover, the simulation results suggest that the blue InGaN LED structures with spontaneous polarization may provide higher output power but are more sensitive to temperature, in a range from 300 K to 350 K, when compared to the LED structures with piezoelectric polarization.

Optical characteristics of deep ultraviolet light-emitting diodes with the consideration of spontaneous and piezoelectric polarizations are studied in this article with the APSYS (Advanced Physical Model of Semiconductor Devices) simulation program. The amounts of surface charges caused by different polarizations are calculated and compared. Moreover, the band diagram, carrier distribution, radiative recombination current, and light-current performance curves of the InAlGaN UV LED structures with different polarizations are also discussed and investigated. According to the simulated results, we find that the influence of the spontaneous polarization is more apparent than the piezoelectric polarization on band properties, carrier distribution, radiative recombination and output power in deep UV spectral region. In other words, for nitride materials in deep UV region, the polarization resulted from lattice-mismatch is smaller than that caused by asymmetry of the structure along the c-axis. This conclusion is quite different from the situation of blue InGaN light-emitting diodes. For blue LEDs, the piezoelectric polarization is the dominant polarization mechanism because the lattice mismatch between compound layers is a severe problem for these long-wavelength LED devices.

Light emitting diodes (LEDs) based on indium gallium nitride (InGaN) have been used widely in lighting applications and have shown great potential with high wall-plug efficiency and low manufacturing cost. However, one of the roadblocks in achieving higher light-extraction efficiency is total internal reflection, which prevents a majority of the photons from being extracted from the LEDs. The objective of this paper is to investigate measures for improving LED chip efficiency. Here, a high efficiency LED structure has been theoretically demonstrated using an infiltrated opal layer. A layer of indium-tin oxide (ITO) infiltrated opals was deposited on the surface of the LED chips as the light extraction layer in order to enhance the light extraction efficiency. Opals of similar radius periodically arranged on the structure will work as a rough surface. By enhancing the light extraction angle, the spontaneous emission from the active region can be emitted vertically. At the same time, opals with different radius and different refractive indexes are also studied. ITO, as the infiltration, is used to provide good lateral current spreading. The plane wave expansion (PWE), finite difference time domain (FDTD), and rigorous coupled wave analysis (RCWA) methods are used to calculate the energy band structures and the extraction efficiency of the infiltrated opals layer. The scanning method and the Monte Carlo method are employed to optimize the structural parameters of the infiltrated opal layer. The infiltrated opals layer on the LED shows a high light extraction efficiency of over 60%.