Thesis (Ph. D.)--University of Texas at Austin, 2000. / Vita. Includes bibliographical references (leaves 111-119). Available also in a digital version from Dissertation Abstracts.

Thesis (Ph.D)--Electrical and Computer Engineering, Georgia Institute of Technology, 2008. / Committee Chair: Honsberg, Christiana; Committee Co-Chair: Ferguson, Ian; Committee Member: Citrin, David; Committee Member: Klein, Benjamin; Committee Member: Rohatgi, Ajeet; Committee Member: Snyder, Robert. Part of the SMARTech Electronic Thesis and Dissertation Collection.

Gallium nitride (GaN) based high electron mobility transistors (HEMTs) offer numerous benefits for both direct current (DC) and radio frequency (RF) power technology due to their combination of large band gap, high electrical breakdown field, high peak and saturation carrier velocity, and good stability at high temperatures. In particular, AlGaN/GaN heterostructures are of great interest because of the unique conduction channel that develops as a result of the spontaneous and piezoelectric polarization that occurs in these layers. This channel is a vertically confined plane of free carriers that is often called a 2 dimensional electron gas (or 2DEG). Although these devices have shown an improvement in performance over previous heterostructures, reliability issues are a concern because of the high temperatures and electric fields that develop during operation. Therefore, characterizing electrical and thermal profiles within AlGaN/GaN HEMTs is critical for understanding the various factors that contribute to device failures. Little research has been performed to model and characterize these devices under RF bias conditions, and is therefore of great interest. Under pulsed conditions, a single cycle consists of an “on-state” period where power is supplied to the device and self-heating occurs, followed by an “off-state” period where no power is supplied to the device and the device cools. The percentage of a single cycle in which the device is powered is called the duty cycle. In this work, we present a coupled electro-thermo-mechanical finite-element model for describing the development of temperature, stress, and strain profiles within AlGaN/GaN HEMTs under DC and AC power conditions for various duty cycles. It is found that bias conditions including source-to-drain voltage, source-to-gate voltage, and pulsing frequency directly contribute to the electro-thermo-mechanical response of the device, which is known to effect device performance and reliability. The model is validated by comparing numerical simulations to experimental electrical curves (Ids-Vds) and experimental strain measurements performed using scanning joule expansion microscopy (SJEM). In addition, we show how the operating conditions (bias applied and AC duty cycle) impact the thermal profiles of the device and outline how the stress in the device changes through a pulsed cycle due to the changing thermal and electrical profiles. Qualitatively, the numerical model has good agreement across a broad range of bias conditions, further validating the model as a tool to better understand device performance and reliability.

Electrowetting is a phenomenon that controls the wettability of liquids on solid surfaces by the application of electric potential. It is an interesting method to handle tiny amounts of liquid on solid surfaces. In recent times, researchers have been investigating this phenomenon and have reported some unexplained behavior and degradation in the Electrowetting system performance. Electrowetting systems include the presence of electric field and different materials from metals to dielectrics and electrolytes that create an environment in which corrosion processes play a very important role. With the small dimensions of the electrodes, corrosion can cause failure quickly when the dielectric fails. In this work, commonly used dielectric films such as silicon dioxide and silicon nitride were deposited using Plasma Enhanced Chemical Vapor Deposition and characterized on the basis of thickness uniformity, etch rate measurements, Dry current – voltage measurements and Wet current – voltage measurements. Sputtered silicon dioxide films were also characterized using the same methods. The correlation between Dry I – V and Wet I – V measurements was studied and a comparison of dielectric quality of films based on these measurements is presented. Also, impact of different liquids on the dielectric quality of films was studied.

published_or_final_version / Physics / Master / Master of Philosophy

In this thesis, Gallium Nitride (GaN) micro- and nanostructures were fabricated based on focused ion beam (FIB) milling. The starting wafer is an epitaxial structure containing InGaN/GaN multi-quantum wells. High crystal quality structures such as the nano-cone, nanopillar array and single pillar were fabricated based on the FIB method. During the fabrication process, various approaches were designed to minimize FIB damage caused by Gallium ion bombardment. The fabrication process for nano-cone is a combination of mask preparation by FIB with subsequent reactive ion etching (RIE). For fabricating nanopillar arrays, the nanopillars were patterned directly using FIB with an optimized beam current followed by wet etching process to remove the damage. On the other hand, the single pillar is achieved by gradually decreasing the ion beam current as the diameter of the pillar becomes smaller. The first order Raman spectra for the nanopillar array reveal a strong additional peak when the diameter of the nanopillars is less than 220 nm. This peak can also be observed in GaN pillars without MQW and is clearly assigned to the surface optical (SO) mode originated from the A1 phonon in wurtzite GaN. The frequency of this SO mode is found to be sensitive with the diameter and surface roughness of the nanopillars. Temperature-variable photoluminescence (PL) measurements show that a broadband emission in the as-grown sample split into the two well-resolved bands for nanopillars and the emission band at the higher energy side quickly thermally quenched. Room temperature PL measurements on the single pillars exhibit an increasing blue-shift of the peak emission with the decreasing of the pillar diameter. Additional simulation data and excitation power dependent PL studies confirm the observation of strain relaxation in the pillar’s MQW due to FIB fabrication. The temperature variable PL on the single pillar shows a monotonous blue shift as the temperature arises to 300 K. / published_or_final_version / Electrical and Electronic Engineering / Master / Master of Philosophy

Wide band gap based nanostructures have being attracting much research interest because of their promise for application in optoelectronic devices. Among those wide band gap semiconductors, gallium nitride (GaN) and zinc oxide (ZnO) are the most commonly studied and optoelectronic devices based on GaN and ZnO have been widely investigated. This thesis concentrates on the growth, optical and electrical properties of GaN and ZnO nanostructures, plus their application in solar cells and light emitting diodes (LEDs). GaN-nanowire based dye sensitized solar cells were studied. Different post-growth treatments such as annealing and coating with a TiOx shell were applied to enhance dye absorption. It was found that TiOx increased the dye absorption and the performance of the dye sensitized solar cell. ZnO nanorods were synthesized by vapor deposition and electrodeposition. Post-growth treatments such as annealing and hydrothermal processing were used to modify the defect chemistry and optical properties. LEDs based on GaN/ZnO heterojunctions were studied. The influence of ZnO seed layers on GaN/ZnO LEDs was investigated. GaN/ZnO LEDs based on ZnO nanorods with MgO and TiOx shells were also prepared in order to modify the LED performance. The coating condition of the shell was found to influence the current-voltage (I-V) characteristics and device performance. Moreover, high brightness LEDs based on GaN with InGaN multiple quantum wells were also fabricated. The origin of the emission from GaN/ZnO LEDs was studied using different kinds of GaN substrates. Direct metal contacts on bare GaN substrates were also employed to investigate the optical emission and electrical properties. It is found that the emission from the GaN/ZnO LEDs probably originated from the GaN substrate. GaN/ZnO LEDs with MgO as an interlayer were also fabricated. The MgO layer was expected to modify the band alignment between the GaN and the ZnO. It was shown that GaN/MgO/ZnO heterojunctions (using both ZnO nanorods and ZnO films) have quite different emission performance under forward bias compared to those that have no MgO interlayer. An emission peak was around 400 nm could originate from ZnO. Nitrogen doped ZnO nanorods on n-type GaN have been prepared by electrodeposition. Zinc nitrate and zinc acetate were used as ZnO precursors and NH4NO3 was used as a nitrogen precursor. Only the ZnO nanorods made using zinc nitrate showed obvious evidence of doping and coherent I-V characteristics. Cerium doped ZnO based LEDs were fabricated and showed an emission that depended on the cerium precursor that was employed. This indicates that the choice of precursor influences the growth, the materials properties and the optical properties of ZnO. / published_or_final_version / Physics / Doctoral / Doctor of Philosophy

This dissertation describes the development of a process for the precise positioning of semiconductor nanoparticles grown by hot wire chemical vapor deposition and thermal chemical vapor deposition on amorphous dielectrics, and it presents two studies that demonstrate the process. The studies entailed growth and characterization using surface science techniques and scanning electron microscopy. The two systems, Ge nanoparticles on HfO₂ and Si nanoparticles on Si₃N₄, are of interest because their electronic properties show potential in flash memory devices. The positioning technique resulted in nanoparticles deposited within 20 nm diameter feature arrays having a 6x10¹⁰ cm⁻² feature density. Self-assembling diblock copolymer poly(styrene-b-methyl methacrylate) thin films served as the patterning soft mask. The diblock copolymer features were transferred using a CHF₃/O₂ reactive ion etch chemistry into a thin film SiO₂ hard mask to expose the desired HfO₂ or Si₃N₄ deposition surface underneath. Selective deposition upon exposed pore bottoms was performed at conditions where adatom accumulation occurred on the HfO₂ or Si₃N₄ surfaces and not upon the SiO₂ mask template. The selective deposition temperatures for the Ge/HfO₂ and Si/Si₃N₄ systems were 700 to 800 K and 900 to 1025 K, respectively. Germanium nucleation on HfO₂ is limited from hot wire chemical vapor deposition by depositing nanoparticles within 67% of the available features. Unity filling of features with Ge nanoparticles was achieved using room temperature adatom seeding before deposition. Nanoparticle shape and size are regulated through the Ge interactions with the SiO₂ feature sidewalls with the adatom removal rate from the features being a function of temperature. The SiO₂ mask limited Ge nanoparticle growth laterally to within ~5 nm of the hard mask at 800 K. Silicon deposition on patterned Si₃N₄ has multiple nanoparticles, up to four, within individual 20 nm features resulting from the highly reactive Si₃N₄ deposition surface. Silicon nucleation and continued nanoparticle growth is a linear function of deposition flux and an inverse function of sample temperature. Diblock copolymer organization can be directed into continuous crystalline domains having ordered minority phases in a process known as graphoepitaxy. In graphoepitaxy forced alignment within microscopic features occurs provided certain dimensional constraints are satisfied. Graphoepitaxy was attempted to precisely locate 20 nm diameter features for selective Ge or Si deposition and initial studies are presented. In addition to precise nanoparticle positioning studies, kinetic studies were performed using the Ge/HfO₂ material system. Germanium hot wire chemical vapor deposition on unpatterned HfO₂ surfaces was interpreted within the mathematical framework of mean-field nucleation theory. A critical cluster size of zero and critical cluster activation energy of 0.4 to 0.6 eV were estimated. Restricting HfO₂ deposition area to a 200 nm to 100 [mu]m feature-width range using SiO₂ decreases nanoparticle density compared to unpatterned surfaces. The studies reveal the activation energies for surface diffusion, nucleation, and Ge etching of SiO₂ are similar in magnitude. Comparable activation energies for Ge desorption, surface diffusion and cluster formation obscure the change with temperature an individual process rate has on nanoparticle growth characteristics as the feature size changes. / text

No description available.

Microelectronics has enjoyed great success in the past century. As the technology node progresses, the complementary metal-oxide-semiconductor scaling has already reached a wall, and serious challenges in high-bandwidth interconnects and fast-speed signal processing arise. The incorporation of photonics to microelectronics provides potential solutions. The theme of this thesis is focused on the novel applications of travelling-wave microresonators such as microdisks and microrings for the on-chip optical interconnects and signal processing. Challenges arising from these applications including theoretical and experimental ones are addressed. On the theoretical aspect, a modified version of coupled mode theory is offered for the TM-polarization in high index contrast material systems. Through numerical comparisons, it is shown that our modified coupled mode theory is more accurate than all the existing ones. The coupling-induced phase responses are also studied, which is of critical importance to coupled-resonator structures. Different coupling structures are studied by a customized numerical code, revealing that the phase response of symmetric couplers with the symmetry about the wave propagating direction can be simply estimated while the one of asymmetric couplers is more complicated. Mode splitting and scattering loss, which are two important features commonly observed in the spectrum of high-Q microresonators, are also investigated. Our review of the existing analytical approaches shows that they have only achieved partial success. Especially, different models have been proposed for several distinct regimes and cannot be reconciled. In this thesis, a unified approach is developed for the general case to achieve a complete understanding of these two effects. On the experimental aspect, we first develop a new fabrication recipe with a focus on the accurate dimensional control and low-loss performance. HSQ is employed as the electron-beam resist, and the lithography and plasma etching steps are both optimized to achieve vertical and smooth sidewalls. A third-order temperature-insensitive coupled-resonator filter is designed and demonstrated in the silicon-on-insulator (SOI) platform, which serves as a critical building block element in terabit/s on-chip networks. Two design challenges, i.e., a broadband flat-band response and a temperature-insensitive design, are coherently addressed by employing the redundant bandwidth of the filter channel caused by the dispersion as thermal guard band. As a result, the filter can accommodate 21 WDM channels with a data rate up to 100 gigabit/s per wavelength channel, while providing a sufficient thermal guard band to tolerate more than ±15°C temperature fluctuations in the on-chip environment. In this thesis, high-Q microdisk resonators are also proposed to be used as low-loss delay lines for narrowband filters. Pulley coupling scheme is used to selectively couple to one of the radial modes of the microdisk and also to achieve a strong coupling. A first-order tunable narrowband filter based on the microdisk-based delay line is experimentally demonstrated in an SOI platform, which shows a tunable bandwidth from 4.1 GHz to 0.47 GHz with an overall size of 0.05 mm². Finally, to address the challenges for the resonator-based delay lines encountered in the SOI platform, we propose to vertically integrate silicon nitride to the SOI platform, which can potentially have significantly lower propagation loss and higher power handling capability. High-Q silicon nitride microresonators are demonstrated; especially, microresonators with a 16 million intrinsic Q and a moderate size of 240 µm radius are realized, which is one order of magnitude improvement compared to what can be achieved in the SOI platform using the same fabrication technology. We have also successfully grown silicon nitride on top of SOI and a good coupling has been achieved between the silicon nitride and the silicon layers.