Author: Mayada Mohammed Tahir Taher

Publisher:

Published: 2019

Total Pages: 270

ISBN-13:

Recently, nonpolar InGaN/GaN optoelectronic structures have been widely studied for applications in ultrafast communication, solid-state lighting, solar cell, sensing, photonic integrated circuits and quantum cryptography. When grown in a core-shell architecture (where the nonpolar, multiple disk active region is radially grown on the sidewall of a hexagonal GaN nanowire), these devices exhibit superior properties that mainly arise from the availability of a larger active region. Recently, the viability of using such architectures in electrically injected, low-threshold single-nanowire laser operating at room temperature has been experimentally demonstrated. In contrast, axially (or expitaxially) grown disk-in-wire structures suffer from a smaller gain-volume and, thus, have failed to produce optically pumped lasing emissions. From fundamental physics point of view, the benefits of using nonpolar m-axis and a-axis oriented InGaN/GaN in the active region are as follows: a) lesser degree of lattice mismatch, resulting in a weaker strain field; b) absence of spontaneous (pyroelectric) polarization; c) smaller piezoelectric polarization, induced internal potential, and electric field in the carrier transport direction; d) stronger overlap of conduction electron and valence hole wavefunctions; e) elimination or reduction of quantum-confined stark effect (QCSE); f) higher transition probability (emission probability) and quantum efficiency; g) higher degree of polarized emission with spectral stability; and h) higher injection efficiency by reducing carrier overflow in a thicker active region. Nevertheless, nonpolar structures exhibit a small internal potential, which mainly arise from non-zero off-diagonal strain components. In addition, even when the active region is completely relaxed in such structures, there remains a small degree of anisotropy that originates from the fundamental symmetry lowering at the material interfaces. In this dissertation, we make efforts to: a) investigate the effects of atomistic strain distributions in realistic multiple dot-in-nanowire In0.08Ga0.92N/GaN structures, as reported in some recent experiments; b) compare the emission characteristics of c-axis and m-axis oriented optical structures (i.e. laser structure); c) explore possibility of improving optical transition probability (rate) via engineering the optical cavity spacer dot size, aspect ratio, Indium mole fraction, and crystal growth direction for precise control over nanowire geometry and high material quality, d) numerically investigate and demonstrate lasing from nonpolar p-i-n core−shell InGaN/GaN multiple quantum dots in nanowires under electrical injection at room temperature, e) carry out detailed numerical investigation with a goal to optimize optical gain, lasing threshold, dynamic response, and device performance of these ultrafast laser structures, and f) explore viability of nonpolar architecture for nanolaser for providing a route forward for integrable, electrically injected nanowire laser for novel nanophotonic applications. The core simulations are performed with an augmented version of the open-source NEMO 3-D software that uses a fully-atomistic valence force-field (VFF) for strain distributions and empirical sp3s*-spin tight-binding model to compute the electronic structure. Both linear and nonlinear components of internal polarization field have been included using a recently proposed first-principles based polarization model. When compared to conventional c-plane based polar structures, the nonpolar device, overall, exhibits a much weaker (yet non-zero) internal potential and improved emission characteristics. In particular, we have found that the m-plane structure exhibits a much smaller (peak ~18.5 mV) internal potential than the c-plane counterpart (peak ~242 mV). However, the fundamental atomicity in the active region results in pronounced anisotropy in the emission characteristic. The energy bandgap is found to be little larger (3.24 eV) in the m-plane structure than in the c-plane device (3.15 eV). With a stronger wavefunction overlap, m-plane clearly offers a higher optical transition probability. Yet, the overall yield in these nonpolar structures suffers from the presence of a strong localization of wavefunctions, which confines the carriers (electrons and holes) in just one (lowest) quantum disk. As for design optimization, it is found that increasing the spacer size (i.e. disk separation) leads to a higher transition rate. Furthermore, detailed analysis has been presented comparing the performance of c-plane, m-plane and the a-plane based InGaN disk-in-wire structures as they show promise in novel optoelectronic applications. It is found that the magnitude of the net polarization potential in the non-polar m-plane and a-plane structures is much smaller (~5 mV) than the polar c-plane counterpart (~129 mV). This particular finding eventually leads to the formation of strongly localized wavefunctions and higher optical transition probabilities in non-polar wurtzite structures. As for the terminal device characteristics, it is found that the disk-in-wire LED in the a-plane orientation offers the highest internal quantum efficiency (IQE) as well as the smallest efficiency droop characteristics.