Strain Engineering of InGaN/GaN Nanopillars for Optoelectronic Applications
Add to Google Calendar
Indium gallium nitride (InGaN)/gallium nitride (GaN) material system is critical for optoelectronic applications in LEDs and lasers because it has direct band gap and large oscillator strength, and the emission wavelength covers the entire visible spectrum. Due to the large lattice mismatch between InGaN and GaN, a large built-in strain exists in the InGaN layer. It is known that nanostructures have large surface-to-volume ratio and could help relax strain via free surfaces. In this work, I will present top-down InGaN/GaN nanostructures as an effective way to manipulate the strain and how we harness the strain effect to create more functionalities of optoelectronics devices. First, we demonstrate that the emission colors from top-down nanopillars could be tuned from red to blue by designing proper nanopillar diameters and strain profile. The wavelength shift is well-described by strain-involved modeling. We also demonstrate electrical nanoLED devices based on the nanopillars. It provides a simple solution to monolithic integration of multiple color pixels on a single chip. Second, we discuss the benefit of strain engineering for quantum light source applications. We focus on the intrinsic control of single photons' polarization states via asymmetric strain. Experimental data is provided to show that pre-defined polarization states are achieved by engineering quantum dot geometry and strain. Single photon emissions with orthogonal polarization states and high degree of linear polarization are recorded. It suggests the potential of top-down InGaN quantum dots for quantum information applications. Finally, the nonideal factors in quantum dots, including random alloy fluctuation and well-width fluctuation, are discussed. These effects impose a fundamental limit to quantum dot inhomogeneity, especially for ternary alloys. A methodology to model random alloy distribution and random well-width fluctuation is developed. The modeling results provide insight to the interplay between strain relaxation and those nonideal effects. They are also compared to experimental data and show very good agreement.