In this paper we present thermal analysis of three kinds of ceramic package designs for high power LEDs and thermal characterization of high power LED array system. The analysis was made by transient thermal measurement and thermal simulation using the finite volume method (FVM). For the package design, thermal behaviors, as are described in thermal resistance, of the three packaging designs were compared and evaluated as functions of bulk thermal resistance, spreading resistance, and surface roughness. The deviation between the simulated results and measured data were attributed to the different surface roughness in the interfaces between the packaging components. For the system design, the emphasis is placed upon the investigation of junction temperature rise of LED array for a limited range of boundary conditions which include design effect of heat pipe, convection condition, and ambient temperature. It was found out that the measured junction temperatures and thermal resistance of LED array are increased with the input power and ambient temperature and decreased with the air velocity. An analytical thermal model analogous with an equivalent parallel circuit system was proposed and was verified by comparison with experimental data.

A typical Fresnel lens is not suitable to a reading light system with multiple light sources since it is designed in such a way that each groove is at a slightly different angle from the next but with the same focal length. Therefore, a Fresnel lens with suitably designed groove angles is needed for this kind of light system. In this paper, a more efficient three-layered Hierarchical Genetic Algorithm (3LHGA) is proposed to find an optimal set of groove angles for a designed Fresnel lens to optimize both the illuminace and uniformity for a reading light system with multiple light sources. The groove angles of a designed Fresnel lens are directly derived from a Fresnel lens database by two layers of control genes in the proposed 3LHGA. The proposed 3LHGA not only makes it possible to evolve a lot of groove angles as parametric genes but also further improve the performance of a Fresnel lens and increase the speed of evolution. From the simulation results we can demonstrate that the designed Fresnel lens indeed offers improvement of light-guiding performance for a multiple-LED reading light system.

The two dimensional Finite-Difference-Time-Domain (FDTD) algorithm is used to study the optical behavior of nano-composite encapsulants. As the size of the nano-particles in an encapsulant decreases, the scattering from particles also decreases and the nano-mixture eventually becomes an optically uniform medium. Calculations of FDTD reveal the size limit of nano-particles when the transition from scatterers to an optically uniform medium occurs. As the size of the nano-particles is reduced to 0.02 λ, scattering substantially disappears and the transmission efficiency improves two-fold compared to that without nano-particles. The numerical results show that the use of a nano-composite encapsulant can improve the extraction efficiency of high-brightness light-emitting-diodes (LEDs). In addition, we simulated the roughened surface of a high-index resin layer using FDTD. The transmission efficiency of roughened surface increases 37% compared to that of the flat surface. Therefore, the combination of high-index nano-composites and a roughened surface can increase the extraction efficiency of the LEDs.

We showed a detailed thermal simulation of an Epi-down flip-chip packaged LED. Simulation results show that chip attachment defects can cause significant thermal gradients across the active layer of chip, leading to premature failures.

Thermal transient measurements of high power GaN-based LEDs with multi-chip designs are presented and discussed. Once transient cooling curve was obtained, the structure function theory was applied to determine the thermal resistance of packages. The measured total thermal resistances from junction to ambient were 19.87 K/W, 10.78 K/W, 6.77 K/W for the one-chip, two-chip and four-chip package, respectively. The contribution of each component to the total thermal resistance of the package can be calculated from the cumulative structure function and differential structure function. The total thermal resistance of multi-chip package is found to decrease with the number of chips due to parallel heat dissipation. Very useful thermal design rule for high power multi-chip LED package is analogized from the experiments. It was found that the effect of the number of chips in a package on the thermal resistance depends on the ratio of partial thermal resistance of chip and that of slug. Thermal resistance for full color multi chip LEDs, where each chip is driven independently, was measured as well and the implication was discussed.

This work presents an alternative approach for preparing photometric standard LEDs, which is based on a novel functional seasoning method. The main idea of our seasoning method is simultaneously monitoring the light output and the junction voltage to obtain quantitative information on the temperature dependence and the aging effect of the LED emission. We suggested a general model describing the seasoning process by taking junction temperature variation and aging effect into account and implemented a fully automated seasoning facility, which is capable of seasoning 12 LEDs at the same time. By independent measurements of the temperature dependence, we confirmed the discrepancy of the theoretical model to be less than 0.5 % and evaluate the uncertainty contribution of the functional seasoning to be less than 0.5 % for all the seasoned samples. To demonstrate assigning the reference value to a standard LED, the CIE averaged LED intensity (ALI) of the seasoned LEDs was measured with a spectroradiometer-based instrument and the measurement uncertainty was analyzed. The expanded uncertainty of the standard LED prepared by the new approach amounts to be 4 % ~ 5 % (k=2) depending on color without correction of spectral stray light in the spectroradiometer.

LEDs are used in many applications, and new applications are found every day. To address this market, LEDs often come in vastly different varieties, shapes, sizes, packages and modules. Measurement of these LEDs is required so that they can be compared and selected within a global market. This paper presents the different types of optical quantity that can be measured for these LEDs together with guidelines for measurement. In particular, the protocols for measuring Averaged LED Intensity, Partial LED Flux, luminance and illuminance are presented. Some of these quantities are new, and the reader may be unfamiliar with them. Definitions are provided where appropriate. Many LED measurements have associated standard measurement conditions, which apply to LEDs and not other sources. Other measurements depend critically on setup conditions but lack standardization and hence details of methods used must accompany results. Where standard conditions exist these are detailed and where they do not advice is provided on the best methodologies. Work in establishing standard conditions is on-going, especially in Commission Internationale de l'Eclairage (CIE) technical committees, and information on this work is provided.

Mg-doped Al_xGa_1-xN is grown by metal-organic chemical vapor deposition to investigate the Ohmic characteristics of Pt and Ni/Au. The Al solid composition measured by x-ray varies from 0.04 to 0.19, while the atomic concentration of Mg confirmed by secondary ion mass spectroscopy spans from 3x10¹⁹ to 1x10²⁰ cm^-3. The Ohmic characteristics are measured by current-voltage by varying the Mg activation temperature, Ohmic metal annealing temperature, and annealing time. The specific contact resistance is 3.5 and 7.5x10^-5 Ω cm².with Pt and Ni/Au in p-Al_0.085Ga_0.915N and p-Al_0.14Ga_0.86N measured by circular transmission line model, respectively. These are the lowest ever reported in p-AlGaN.

According to our rich experiences on manufacturing of high reliability GaN based Light Emitting Diodes, we try to expatiate the relations between processes and device reliability on five aspects in this paper. In addition, we lay out some of our solutions on the five aspects and show our stable device.

We present first results on the limits of GaN growth on large diameter sapphire and the challenges that have to be solved for a successful growth of high power LEDs on silicon substrates. Up to 5.4 μm thick crack-free GaN on Si(111) LED structures were grown by metalorganic chemical vapor phase epitaxy. The FWHM of the GaN (0002) ω scan in x-ray diffraction amounts to 380 arcsec. On Si substrates, we achieve low curvatures with radii > 50 m, which is important for a successful processing of the samples on large diameter substrates. Additionally, a low curvature during InGaN multi-quantum-well growth is achieved and enables the growth of homogenous InGaN layers. The main difficulty for GaN-on-Si is light extraction, which leads to an approximately three- to four-fold reduction in direct comparison with GaN LEDs on sapphire.

This paper reports a new model of current versus voltage for light emitting devices with a quantum well. These devices include light emitting diodes (LEDs) and laser diodes (LDs). In LED or LD operation, nearly all electrons and holes are injected into the quantum well and recombine there. As a result, the forward current consists of mainly the recombination current. This is in contrast to the popular Sah-Noyce-Shockley (SNS) diode equation where the dominating current is the diffusion current. The SNS model assumes that recombination in the depletion region is negligible under typical forward bias condition. This is opposite to what actually happens in LEDs or LDs with quantum wells. However, SNS equation has been used directly to describe LEDs and LDs with quantum wells for several decades probably because of its simplicity. Another probably reason is that there is no other I-V model available. In this study, I-V curves calculated using the new model agree well with results measured on GaN/sapphire LEDs with InGaN quantum wells. In the calculation, junction temperature T_j rather than case temperature T_c is used to achieve better agreement.

A simple optical system for generating large-size hexagonal lattice used as the mask for photonic crystal LED fabrication is proposed. The key element in the system is a holographic diffractive optical element consisting of three phase gratings made with holographic means. Under the illumination of a single plane wave, three plane waves can be generated and interfere to form a uniform hexagonal pattern. Theoretical analysis demonstrates that equal intensity of the three plane waves can be obtained if groove depth of the gratings is about 400 nm, which can be achieved by controlling the development time in grating processing. Mask on LED with lattice area up to 15 cm² has been obtained with the method.

Fluorescent polymer-zinc oxide hybrid (PZOH) materials were successfully prepared using simple chemical bath deposition method at low temperature. The formation of ZnO nanoparticles and their deposition on the surface of fluorescent polymer was confirmed by SEM and TEM images, which were further clarified by performing selected area diffraction patterns in TEM image that confirms the incorporation of crystalline ZnO material. The optical emission efficiency showed significant enhancement by employing different ratio of ZnO composition with respect to fluorescent polymer. Finally, bright and efficient white light emitting diodes have been fabricated using these PZOH, as a luminescence converter (LUCO) with the help of commercially available Nitride based blue LED, as a primary pumping source. The output light efficiency of PZOHs fabricated White LED showed drastic improvement when compared to pure fluorescent polymer, due to higher PL quantum efficiency (86 %) of PZOH then the pure fluorescent polymer (54 %).

Low-defect AlGaN films were grown by metal organic chemical deposition (MOCVD) over patterned sapphire substrates (PSS) with V-grooves fabricated by wet chemical etching based on a mixed solution of H₂SO₄ and H₃PO₄. Three high-temperature (HT) growth steps were performed on the wet-etched PSS. In the 1st HT-growth step, GaN layers with triangular cross-sections were grown on sapphire mesas of the surface of the wet-etched PSS. In the 2nd HT-growth step, the GaN layers were more grown in the lateral direction and coalescent at the bottom, but corrugated at the surface. Finally, in the 3rd HT-growth step, AlGaN layer was grown on the corrugated surface of GaN and coalesced completely into AlGaN layer with flat surface. By employing the corrugated surface of GaN grown on the wet-etched PSS as the initial surface for the growth of AlGaN layers, the tensile stress of AlGaN was remarkably reduced. Additionally, in order to completely eliminate the cracks of the AlGaN layer, a low-temperature (LT) AlN layer was inserted between the AlGaN layer and the GaN layer with the corrugated surface. By inserting the LT-AlN interlayer, no crack did generate in the AlGaN film, and the density of threading dislocations in the film was remarkably decreased after the lateral epitaxy using the wet-etched PSS with V-grooves.

The effect of chip size on the thermal-optical properties of GaN-based light emitting diodes was investigated. An increase in chip size was associated with a decrease of total series resistance due to enhancement of current spreading area. Consequently, the junction temperature linearly increased with an increase of the driving current, and the increase rate was slower for lager chip size. Moreover, we found out that the driving current and chip size affect the dominant emission wavelength shift that was understood to be a competition between blue shift behaviors of piezoelectricity-induced quantum confined stark effect and red shift behavior of self-heating effect. Thus, the operating current for color stabilization was increased with increasing chip size such as 80mA, 140mA and 160mA for 350×350μm², 600×600μm² and 1000×1000μm² chip sizes, respectively. Herein, the operating current for color stabilization was determined at the driving current, where blue shift of piezoelectricity-induced quantum confined stark effect became in balance with the temperature induced band gap shrinkage resulted from self-heating effect.

This work reports on the moisture inducing delamination in light-emitting diode (LED) packages and its effects on thermal characteristics. The LED samples were subjected to moisture preconditioning followed by heat block testing. Transient thermal measurements were performed to investigate the thermal behavior of the delaminated LEDs. Increase of thermal resistance with the degree of delamination was observed from the transient measurement. The thermo-mechanical calculated from coupled-field FEA simulation agree well with the micrographical evidence. The calculated hygro-mechanical stress increased with the preconditioning time. It was found that the thermo-mechanical stress plays more important role than the hygro-mechanical stress for the development of delamination in the LED packages. Moisture preconditioning for 3 hrs and 6 hrs under 85°C/85RH conditions was found to make little contribution to the delamination between the chip and lead frame.

The influence of charge injection on efficiency and operating voltage of organic light emitting device was investigated. Using UV-treated and un-treated ITO substrates, organic light emitting diodes were fabricated as we vary the thickness of LiF layer as the electron injection layer. The operating voltages and efficiencies were measured. When ITO surface is UV treated, operating voltage was decreased for all ranges of LiF thickness. By inserting a LiF thin layer with thickness of 5~10 Å lower operating voltages and higher power efficiencies were achieved. For a thicker LiF layer, power efficiency was decreased rapidly for UV treated ITO because of insulating properties of LiF. On the other hand, for untreated ITO - higher hole injection barrier, it was kept relatively high value up to the range of 20Å of due to the charge balance.