Efficiency Improvement of Nitride-Based Solid State Light Emitting Materials -- CRADA Final Report | |
Kisielowski, Christian ; Weber, Eicke | |
Lawrence Berkeley National Laboratory | |
关键词: Light Sources; Electron Microscopy; Ternary Alloy Systems; Indium; Recombination; | |
DOI : 10.2172/1000045 RP-ID : LBNL-4038E RP-ID : DE-AC02-05CH11231 RP-ID : 1000045 |
|
美国|英语 | |
来源: UNT Digital Library | |
【 摘 要 】
The development of In{sub x}Ga{sub 1-x} N/GaN thin film growth by Molecular Beam Epitaxy has opened a new route towards energy efficient solid-state lighting. Blue and green LED's became available that can be used to match the whole color spectrum of visible light with the potential to match the eye response curve. Moreover, the efficiency of such devices largely exceeds that of incandescent light sources (tungsten filaments) and even competes favorably with lighting by fluorescent lamps. It is, however, also seen in Figure 1 that it is essential to improve on the luminous performance of green LED's in order to mimic the eye response curve. This lack of sufficiently efficient green LED's relates to particularities of the In{sub x}Ga{sub 1-x}N materials system. This ternary alloy system is polar and large strain is generated during a lattice mismatched thin film growth because of the significantly different lattice parameters between GaN and InN and common substrates such as sapphire. Moreover, it is challenging to incorporate indium into GaN at typical growth temperatures because a miscibility gap exists that can be modified by strain effects. As a result a large parameter space needs exploration to optimize the growth of In{sub x}Ga{sub 1-x}N and to date it is unclear what the detailed physical processes are that affect device efficiencies. In particular, an inhomogeneous distribution indium in GaN modifies the device performance in an unpredictable manner. As a result technology is pushed forward on a trial and error basis in particular in Asian countries such as Japan and Korea, which dominate the market and it is desirable to strengthen the competitiveness of the US industry. This CRADA was initiated to help Lumileds Lighting/USA boosting the performance of their green LED's. The tasks address the distribution of the indium atoms in the active area of their blue and green LED's and its relation to internal and external quantum efficiencies. Procedures to measure the indium distribution with near atomic resolution were developed and applied to test samples and devices that were provided by Lumilids. Further, the optical performance of the device materials was probed by photoluminescence, electroluminescence and time resolved optical measurements. Overall, the programs objective is to provide a physical basis for the development of a simulation program that helps making predictions to improve the growth processes such that the device efficiency can be increased to about 20%. Our study addresses all proposed aspects successfully. Carrier localization, lifetime and recombination as well as the strain-induced generation of electric fields were characterized and modeled. Band gap parameters and their relation to the indium distribution were characterized and modeled. Electron microscopy was developed as a unique tool to measure the formation of indium clusters on a nanometer length scale and it was demonstrated that strain induced atom column displacements can reliably be determined in any materials system with a precision that approaches 2 pm. The relation between the local indium composition x and the strain induced lattice constant c(x) in fully strained In{sub x}Ga{sub 1-x}N quantum wells was found to be: c(x) = 0.5185 + {alpha}x with {alpha} = 0.111 nm. It was concluded that the local indium concentration in the final product can be modulated by growth procedures in a predictable manner to favorably affect external quantum efficiencies that approached target values and that internal quantum efficiencies exceeded them.
【 预 览 】
Files | Size | Format | View |
---|---|---|---|
1000045.pdf | 75KB | download |