METALORGANIC CHEMICAL VAPOR DEPOSITION OF ALGAN AND GAN NANOSTRUCTURES
Open Access
- Author:
- Mirabito, Timothy
- Graduate Program:
- Materials Science and Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- December 12, 2024
- Committee Members:
- John Mauro, Program Head/Chair
Suzanne Mohney, Major Field Member
Joan Redwing, Chair & Dissertation Advisor
David Snyder, Major Field Member
Luke Lyle, Outside Unit & Field Member - Keywords:
- MOCVD
III-Nitrides - Abstract:
- Group-III nitrides (GaN, AlN, and their ternary alloys) are an exciting and emerging class of materials in the field of electronics and optoelectronics. With bandgap ranges that span from 3.4 eV for GaN to 6.1 eV for AlN, these materials have been utilized in deep ultraviolet laser diodes (LDs), light emitting diodes (LEDs), RF electronics, and high-power devices. Beyond its application to current technology, it is also essential to forecast its utilization in the next generation of devices. The process development necessary to achieve those goals therefore starts with the growth synthesis to target application specific performance metrics, such as the electronic or morphological properties. The studies within this dissertation have the goal of developing new methods to improve the quality of Ge-doped AlxGa1-xN epilayers and GaN nanostructures using metal-organic chemical vapor deposition towards planar and non-planar field emitter devices. The work is structured into three parts which are described in the sections below. The first part of the thesis will focus on investigating the stress mechanisms of an AlxGa1-xN epilayer grown on an AlN on sapphire template and how they can be modified using a pre-metallization growth step. The heteroepitaxial growth of III-Nitrides is of interest for developing the contact layer in planar field emitter structures, but also finds applications in solid state lighting in the blue emission range. When n-type doping AlxGa1-xN for contact layers, one of the common side effects of heavy doping is the loss of quality with frequent v-pitting across the surface. This becomes more significant as an increase in Al fraction causes Si to become the only shallow n-type dopant but has been shown to increase tensile strain within the film. In this work, AlxGa1-xN epilayers were fabricated via MOCVD using a short pre-metallization step prior to growth. In situ reflectance measurements were used to monitor the growth rate and stress of the AlN and AlxGa1-xN during the MOCVD process. It was found that the pre-metallization step resulted in a reduction in surface roughness in the AlxGa1-xN and lead to tensile stress in the films compared to compressive stress which is typical for AlxGa1-xN growth on AlN. The tensile stress was correlated to increased TD density which arises due to the presence of a carbon-rich layer at the AlxGa1-xN /AlN interface resulting from pre-metallization. By reducing the growth rate of Ge-doped AlxGa1-xN and hence the amount of metal deposited on the growth surface, the carbon interfacial layer was largely eliminated enabling a reduction in surface roughness and V-pits without increased TD density. The second part of the thesis will explore the role of trimethylindium (TMIn) as a surfactant during the growth of n-doped AlxGa1-xN on various AlN/sapphire substrates. N-type doping in the III-Nitride material system has been widely explored due to its importance in heavily utilized device architectures such as high electron mobility transistors (HEMTs) and LEDs. Despite this prevalence, the growth of high quality AlxGa1-xN remains difficult owing to the short diffusion lengths of Al adatoms especially in lower temperature ranges (< 1100° C), which in turn limits the process growth window. In addition, n-type doping of AlxGa1-xN is problematic as the most commonly investigated donor impurity Silicon (Si), introduces tensile stress that can lead to the formation of channeling cracks. Aside from the structural complexities, the demand for high electrical performance requires high electron concentrations without compromising film quality. To achieve higher electrical performance, it has been suggested that increasing the formation energy of the VIII-n complexes can lead to improvements in conductivity, which surfactants such as Indium can enable. Supplementing this, Ge as an n-type dopant has been shown to produce AlxGa1-xN epilayers without cracking at very high carrier concentrations. Despite initial reports of Indium-silicon co-doping in high Al-fraction AlxGa1-xN leading to decreased sheet resistance the nature of that improvement remains unclear. In the dissertation, the impact of introducing TMIn with Ge and Si doping on the structural quality and electrical performance of AlxGa1-xN was investigated using two different AlN substrates. It was found that TMIn can improve the carrier concentration, sheet resistance, and crystal quality of Al0.3Ga0.7N when grown on HVPE AlN/sapphire templates (SX). The third part of the dissertation moves away from the planar AlxGa1-xN material system, towards GaN nanostructures. In particular, this focuses on the process development of selective area epitaxy of GaN nanostructures grown on both Ga-polar and N-polar GaN substrates towards field emitting devices. Traditionally the application of GaN nanostructures has been towards optoelectronic devices such as LEDs, as one of the key features of III-nitrides is both a direct bandgap and the broad spectral range covering the UV and the whole visible spectrum. Nanoscale field emission devices require a different design consideration than the optically focused devices typically encountered with III-N nanostructures. Nanoscale field emitters aim to take advantage of the smaller structure size to generate higher electric fields, leading to lower turn on voltages for emission. To date however, cathode degradation has been a central issue with current field emitter (FE) technology. With the small structures typically employed, large emission currents can lead to high thermal heating causing deterioration. To help mitigate this issue, a low density of structures aggregated into a field emitter array is critical. This differs from nanostructure LED designs where a high density of structures is preferable to overcome the limited active volume. This difference manifests in the need for a selective area design where the growth dynamics are highly dependent on the geometric configuration. Thus, this thesis carried out extensive studies on the impact of growth parameters, pattern geometry and GaN polarity on the resultant morphology. Subsequent characterization of the nanostructures was carried out by CL to highlight the differences in both form and polarity. Finally, it was demonstrated that these structures could produce field emission.