Polarization engineering and approaches for high-performance III-nitride light emitters.

摘要

Light emitting diodes (LEDs) have been increasingly integrated into mainstream lighting. In all applications requiring single-colored light, LEDs have outperformed filtered incandescent lamps. However, there are two major challenges. First is the issue of cost. High-performance nitride-based white LEDs cost roughly two orders of magnitude more expensive than incandescent lamps. The second challenge is color rendering---quantified by Color Rendering Index (CRI). Today's nitride white light LEDs still rely on the mixing of blue light from blue InGaN LEDs and yellow phosphor, and the CRI is relatively low. The best white LEDs to date have a CRI of 70--80, in comparison to traditional lamps, which generally have a CRI close to 100, and able to represent the true color of an object. An ideal way to improve the CRI is by mixing the luminescence of primary color LEDs. However, in order to make this approach viable, all the LEDs have to be based on a single materials platform. AlInGaN is the only materials system to date with the potential to fulfill this, since the bandgap of this nitride compound (with varying amount of Al, In, and Ga) can be varied from UV to IR range. There is still a lot of room for improvement in the efficiencies of nitride blue and green LEDs, while nitride-based active region emitting in the red wavelength (lambda ∼ 650-nm) regime is not realizable yet.;In this dissertation, methods to increase internal quantum efficiency by polarization field engineering have been proposed. Two novel structures based on (1) staggered InGaN QW and (2) type-II InGaN-GaNAs QW have been investigated. Staggered InGaN QWs have shown improvement in the photoluminescence, cathodoluminescence, and LED output power, which agree well with numerical model prediction. All materials and devices in this work have been designed, grown and fabricated in-house. For the LED fabrication, a method based on selective area epitaxy---which bypasses dry-etching---has been utilized. In the second approach, theoretical analysis of type-II InGaN-GaNAs has also predicted improvement in the spontaneous recombination rate and optical gain in the visible wavelength regime. Theoretical investigation of efficiency droop, which is important for high-power LEDs, has also been performed based on current injection efficiency analysis in InGaN QW.