1,721,152 research outputs found
Active Defects in 4H–SiC MOS Devices
No full textThe research findings presented in this thesis have provided several key contributions towards a better understanding of the SiC–SiO2 interface in SiC MOS structures. The electrically active defects directly responsible for degrading the channel-carrier mobility in 4H–SiC MOSFETs have been identified and a novel technique to detect these defects in 4H–SiC MOS capacitors has been proposed and experimentally demonstrated. With a better understanding of defects at the SiC–SiO2 interface two alternative gate oxide growth processes have been proposed to overcome the practical limitations associated with current NO-nitridation techniques in high-volume, production based oxidation furnaces. This work therefore contributes to the wider research effort towards improving the performance of SiC MOSFETs in several ways. The following paragraphs summarise the key conclusions that have been obtained as a result of this study.
Electrically Active Defects and the Channel-Carrier Mobility (Chapter 3)
A critical review of defects at the SiC–SiO2 interface exposed a few key discrepancies in both the current understanding of the dominant defects responsible for channel-carrier mobility degradation in 4H–SiC MOSFETs and in the current approach to characterise and evaluate the SiC–SiO2 interface. Firstly, it was recognised that the Shockley-Read-Hall statistical model, based on thermally activated transport for traps spatially located at the semiconductor-oxide interface, cannot be directly applied to describe the transfer mechanism between free conduction band electrons and the shallow NITs near EC. This implication tends to suggest that the NITs near EC in SiC MOS structures cannot be accurately examined using traditional MOS characterisation techniques that are based on this statistical model. Secondly, in accordance with the studies conducted by Saks et. al. [1-3], it was realized that channel-carrier mobility degradation in 4H–SiC MOSFETs is primarily due to the significantly reduced free electron density in the inversion channel. In light of this understanding, the interfacial defects that actively trap channel electrons under strong inversion conditions were considered to be dominant in these devices as opposed to the NITs near EC that are typically examined using conventional MOS characterisation techniques on N-type MOS capacitors in depletion. To further support this hypothesis, a theoretical analysis of the inversion carrier concentration using the charge sheet model was conducted to demonstrate that the NITs with energy levels corresponding to strong inversion are of key importance to the channel-carrier mobility.Thesis (PhD Doctorate)Doctor of Philosophy (PhD)Griffith School of EngineeringScience, Environment, Engineering and TechnologyFull Tex
On Memristive Threshold Logic Memory Networks
No full textThe ability of the human brain to provide different logic functions using physically similar cell structures is at the heart of its plasticity and ability to generalize. The principle of firing of neurons and connectivity of neuron networks gives rise to its generalized logical ability required to perform different cognitive tasks such as face recognition, object detection, identification and categorisation, rapidly, effortlessly and in real-time. These are computationally difficult tasks when performed in machine but the brain needs very low power and no pre-defined software.
Inspired by the principle of firing of neurons in the primate brain, and the ability of similar neuronal structures to perform different cognitive tasks, this thesis explores pro- grammable (mem)resistance-based threshold logic memory networks that are shown to perform basic logic functions to highly involved cognitive tasks such as face recognition and fast moving object detection, in a parallel, scalable hardware architecture.
To build neuromorphic systems in hardware, we use memristors, which are highly dense, non-volatile, and naturally amenable to be operated analogously to the biological synapse, to build a threshold logic cell. This cell is programmable and is the building block of memory networks that can be reconfigured to perform the cognitive tasks. The said memory networks also serve to shift focus from memory being a storage-only unit to being a computational unit as well, just as in the human brain.Thesis (PhD Doctorate)Doctor of Philosophy (PhD)School of Information and Communication TechnologyScience, Environment, Engineering and TechnologyFull Tex
Silicon Carbide as the Nonvolatile-Dynamic-Memory Material
No full textThis thesis consists of three main parts, starting with the use of improved nitridation processes to grow acceptable quality gate oxides on silicon carbide (SiC)[1]–[7], to the comprehensive investigation of basic electron-hole generation process in 4H SiC-based metal–oxide–semiconductor (MOS) capacitors [8], [9], and concluding with the experimental demonstration and analysis of nonvolatile characteristics of 4H SiC-based memory devices [10]–[15]. In the first part of the thesis, two improved versions of nitridation techniques have been introduced to alleviate oxide-growth rate and toxicity problems. Using a combination of nitridation and oxidation processes, a sandwich technique (nitridation–oxidation–nitridation) has been proposed and verified to solve the lengthy and expensive oxide-growing process in direct nitric oxide (NO) gas [1]. The nitrogen source from the toxic-NO gas has been replaced by using a nontoxic nitrous oxide (N2O) gas. The best combination of process parameters in this gas is oxide-growing temperature at 1300oC with 10% N2O [2], [3]. The quality of nitrided gate oxides obtained by this technique is lower than the sandwich technique [6], [13]. Using 4H SiC-based MOS with nitrided gate oxides grown by either of the abovementioned nitridation techniques, the fundamentals of electron-hole generation have been investigated using high-temperature capacitance–transient measurements. The contributions of carrier generation, occurring at room temperature, in the bulk and at the SiC–SiO2 interface are evaluated and compared using a newly developed method [8], [9]. The effective bulk-generation rates are approximately equal for both types of nitrided oxides, whereas the effective surface-generation rates have been shown to exhibit very strong dependencies on the methods of producing the nitrided gate oxide. Based on analysis, the prevailing generation component in a SiC-based MOS capacitor with nitrided gate oxide is at SiC–SiO2 interface located below the gate. Utilizing the understanding of electron-hole generation in SiC, the nonvolatile characteristics of memory device fabricated on SiC have been explored. The potential of developing a SiC-based one-transistor one-capacitor (1T/1C) nonvolatile-dynamic memory (NDM) has been analyzed using SiC-based MOS capacitors as storage elements or test structures. Three possible leakage mechanisms have been evaluated [10]–[16]: (1) leakage via MOS capacitor dielectric, (2) leakage due to electron-hole generation in a depleted MOS capacitor, and (3) junction leakage due to generation current occurred at a reverse-biased pn junction surrounding the drain region of a select metal–oxide– semiconductor field–effect–transistor (MOSFET). Among them, leakage through capacitor oxide remains an important factor that could affect the nonvolatile property in the proposed device, whereas others leakage mechanisms are insignificant. Based on the overall results, the potential of developing a SiC-based 1T/1C NDM is encouraging.Thesis (PhD Doctorate)Doctor of Philosophy (PhD)School of Microelectronic EngineeringFaculty of Engineering and Information TechnologyFull Tex
Pose-invariant Face Recognition through 3D Reconstructions
No full textPose invariance is a key ability for face recognition to achieve its advantages of being non-intrusive over other biometric techniques requiring cooperative subjects such as fingerprint recognition and iris recognition. Due to the complex 3D structures and various surface reflectivities of human faces, however, pose variations bring serious challenges to current face recognition systems. The image variations of human faces under 3D transformations are larger than that existing face recognition can tolerate. This research attempts to achieve pose-invariant face recognition through 3D reconstructions, which inversely estimates 3D shape and texture information of human faces from 2D face images. This extracted information is intrinsic features useful for face recognition which is invariable to pose changes. The proposed framework reconstructs personalised 3D face models from images of known people in a database (or gallery views) and generates virtual views in possible poses for face recognition algorithms to match the captured image (or probe view). In particular, three different scenarios of gallery views have been scrutinised: 1) when multiple face images from a fixed viewpoint under different illumination conditions are used as gallery views; 2) when a police mug shot consisting of a frontal view and a side view per person is available as gallery views; and 3) when a single frontal face image per person is used as gallery view. These three scenarios provide the system different amount of information and cover a wide range of situations which a face recognition system will encounter. Three novel 3D reconstruction approaches have then been proposed according to these three scenarios, which are 1) Heterogeneous Specular and Diffuse (HSD) face modelling, 2) Multilevel Quadratic Variation Minimisation (MQVM), and 3) Automatic Facial Texture Synthesis (AFTS), respectively. Experimental results show that these three proposed approaches can effectively improve the performance of face recognition across pose...Thesis (PhD Doctorate)Doctor of Philosophy (PhD)School of EngineeringScience, Environment, Engineering and TechnologyFull Tex
Design and Application of SiC Power MOSFET
No full textThis thesis focuses on the design of high voltage MOSFET on SiC and its application in power electronic systems. Parameters extraction for 4H SiC MOS devices is the main focus of the first topic developed in this thesis. Calibration of two-dimensional (2-D) device and circuit simulators (MEDICI and SPICE) with state-of-the-art 4H SiC MOSFETs data are performed, which includes the mobility parameter extraction. The experimental data were obtained from lateral N-channel 4H SiC MOSFETs with nitrided oxide-semiconductor interfaces, exhibiting normal mobility behavior. The presence of increasing interface-trap density (Dit) toward the edge of the conduction band is included during the 2-D device simulation. Using measured distribution of interface-trap density for simulation of the transfer characteristics leads to good agreement with the experimental transfer characteristic. The results demonstrate that both MEDICI and SPICE simulators can be used for design and optimization of 4H SiC MOSFETs and the circuits utilizing these MOSFETs. Based on critical review of SiC power MOSFETs, a new structure of SiC accumulation-mode MOSFET (ACCUFET) designed to address most of the open issues related to MOS interface is proposed. Detailed analysis of the important design parameters of the novel structure is performed using MEDICI with the parameter set used in the calibration process. The novel structure was also compared to alternative ACCUFET approaches, specifically planar and trench-gate ACCUFETs. The comparison shows that the novel structure provides the highest figure of merit for power devices. The analysis of circuit advantages enabled by the novel SiC ACCUFET is given in the final part of this thesis. The results from circuit simulation show that by utilizing the novel SiC ACCUFET the operating frequency of the circuit can be increased 10 times for the same power efficiency of the system. This leads to dramatic improvements in size, weight, cost and thermal management of power electronic systems.Thesis (PhD Doctorate)Doctor of Philosophy (PhD)School of Microelectronic EngineeringFull Tex
A Memory Based Face Recognition Method
No full textThe human brain exhibits robustness against natural variability occurring in
face images, yet the commonly attempted algorithms for face recognition are not
modular and do not apply the principle of binary decisions made by the firing of
neurons. This thesis presents a memory based face recognition method based on
the concepts of local binary decisions and spatial change features. Local binary
decisions are inspired from the binary conversions done by firing of neurons
while spatial change features are inspired from the retinal processing of the
human visual system. Applying these principles and by using the principle of
modularity in a hierarchical manner, a class of memory based face recognition
algorithms is formed.
These algorithms when applied to difficult testing conditions show high recognition
performance. This high recognition performance is enabled by (1) local
binary decisions and (2) spatial change detection. The baseline algorithm formed
by using these two concepts is called local binary decisions on similarity (LBDS)
algorithm. An analysis is performed using the LBDS algorithm to optimize the
parameters, and to study the relative effect of spatial change features, local binary
decisions, normalization of features, normalization of similarity measure, use
of color, localization error compensation and resolution on recognition performance.
From the insights gained through the analysis, the LBDS algorithm is
further improved by incorporating various preprocessing spatial filter operations to extract more spatial information. The inclusion of preprocessing step helps
to achieve even higher recognition performance and robustness to difficult tasks.
This improved algorithm is called enhanced local binary decisions on similarity
(ELBDS) algorithm. The ELBDS algorithm is further used to incorporate the
multiple training images per person in the gallery, and is called an exemplar
based face recognition method.
The following is the overall recognition performance when using single gallery
image per person: 97% on AR, 100% on YALE, 97% on EYALE , 97% on
CALTECH, 98% on FERET(FaFb), 94% on FERET(FaFc), 74% on FERET
(FaDup1) and 76% on FERET(FaDup2). When using multiple training samples
per person, following recognition accuracies are achieved, 99.0% on AR, 99.5%
on FERET, 99.5% on ORL, 99.3% on EYALE, 100.0% on YALE and 100.0% on
CALTECH face databases.Thesis (PhD Doctorate)Doctor of Philosophy (PhD)Griffith School of EngineeringScience, Environment, Engineering and TechnologyFull Tex
Electrical Properties of 3C-SiC Devices on Si under Mechanical Stress
No full textSemiconductor devices can experience various kinds of stresses which can be induced during device fabrication, packaging and during application in thermo-mechanical or harsh environment. While the stress induced effects in semiconductors can be useful for developing mechanical sensors but these effects can also cause long term instability and unwanted errors in the device performance. These stresses in worst case can result in permanent damage of the device. Therefore, it is very important to investigate the effect of various stresses induced to the semiconductor devices for analyzing the stress
sensing capability of the device as well as from the device performance point of view. Although stress induced electrical changes in silicon (Si) based devices have been studied for years and are well understood but a complete understanding of the stress induced changes in silicon carbide (SiC) based devices is still missing. SiC is a promising material for stress sensing applications in harsh environment. It is the purpose of this study to investigate the effects of stress on SiC based devices. The main focus of the study is 3C-SiC/Si based heterojunction devices and 3C-SiC Hall devices. The stress induced
piezojunction effects in SiC/Si heterojunctions investigated in this study indicate that the stress can signicantly alter the heterojunction characteristics. Additionally, the pseudo-Hall effect in p-type and n-type 3C-SiC Hall devices observed in this study shows that the 3C-SiC Hall devices are promising candidates for stress sensing applications. The piezo-Hall effect has also been analyzed in both p-type and n-type 3C-SiC Hall devices and it has been observed that the piezo-Hall effect can signicantly change the magnetic eld sensitivity of the 3C-SiC Hall devices. The fundamental piezo-Hall coefficients for both n-type and p-type 3C-SiC are calculated in this study which are the basic design parameters to compensate the stress dependent offset voltage drifts in Hall devices.Thesis (PhD Doctorate)Doctor of Philosophy (PhD)Griffith School of EngineeringScience, Environment, Engineering and TechnologyFull Tex
Characterization of Active Defects in SiC MOSFETs
No full textIn the recent years, SiC has become a popular material for power semiconductor devices, after decades long dominance of Si. SiC metal‒oxide‒semiconductor field-effect transistors (MOSFETs) are now commercially available and performing beyond the theoretical limits of Si MOSFETs, however, they are still far from the theoretical limits of SiC. One of the major issues in the commercial SiC MOSFETs is the low channel-carrier mobility, which is attributed to the high density of defects at or near the SiC/SiO2 interface. Consequently, it is necessary to characterize the SiC/SiO2 interface appropriately for the future development of SiC power devices.
This thesis begins with a critical review of conventional characterization techniques for SiC metal‒oxide‒semiconductor (MOS) devices, which are directly adapted from techniques developed for Si. To address the challenges in characterizing SiC/SiO2 interface, a new characterization technique is proposed, which measures the effect of near-interface traps (NITs) in the strong-accumulation region of N-type SiC MOS capacitors. The technique measures the current through a SiC MOS capacitor in strong-accumulation and compares it to the trap-free current. With this technique, an active defect with energy levels localized between 0.13 eV to 0.23 eV, above the bottom of conduction band, is identified. As the effect of NITs is observed only at high frequencies, it is expected to be located very close to the SiC/SiO2 interface. To further investigate the effect of high temperature and positive bias stress on the identified NITs, the MOS capacitors are measured at high temperatures with positive bias stress. No significant difference is observed between measurements performed before and after high temperature bias stress. This led to the conclusion that the temperature independent tunneling is responsible for the trapping and de-trapping of channel electrons to and from the NITs. For the first time in this work we have demonstrated the NITs with response time in tens of ns. A detailed explanation of trapping/de-trapping mechanism of NITs, localized in energy, is also presented in this thesis. The developed technique is further used to perform a comparative analysis of NITs in as-grown and nitrided gate oxides. The density of NITs in nitrided gate oxide is localized in energy whereas it tends to decrease with increasing energy levels in as-grown gate oxide. It is experimentally shown that the nitridation helps to eliminate NITs further away from the SiC/SiO2 interface.Thesis (PhD Doctorate)Doctor of Philosophy (PhD)School of Eng & Built EnvFull Tex
Deposition and Characterisation of Amorphous and Nanocrystalline SiC
No full textIn this thesis, the deposition of unintentionally doped amorphous and nanocrystalline SiC was demonstrated using a standard hot-wall low-pressure chemical vapour deposition reactor in a substrate temperature range of 600 to 850 oC, with methylsilane used as the single precursor. The pressure of methylsilane varied from 0.006 to 0.54 mbar. The Arrhenius activation energy of SiC deposition varied with methylsilane pressure, in a range of 2 to 2.8 eV. Surface reaction was found to be the rate-determining process at temperatures below 700 ºC; both surface reaction and mass transport processes control the rate-determining process in the temperature range of 700 to 850 ºC. The deposition rate of SiC was relatively insensitive to the substrate surface conditions. The crystallinity of deposited SiC improved with reduced methylsilane pressure and at elevated substrate temperature based on the results from x-ray diffraction, high-resolution transmission electron microscopy, selected area electron diffraction, and the Fourier transform infrared spectroscopy. The deposition of Al-doped SiC films was demonstrated in a temperature range of 600 to 750 ºC, with methylsilane and trimethylaluminium used as precursors. The incorporation of trimethylaluminium caused in situ crystallisation of a-SiC film deposited on Si substrate at 600 ºC, which is much lower than the crystallisation temperatures usually required by other techniques. Hydrogen concentration ranged from 17 at. % to less than 1 at. % based on the secondary ion mass spectroscopy depth profile analyses, demonstrating that hydrogen concentration decreased both at an elevated substrate temperature and with increasing methylsilane pressure. The Si–C absorption band was the major peak found in the Fourier transform infrared spectra in the range of 400 to 4000 cm-1 for all SiC films. According to the chemical bonding analyses conducted by x-ray photoelectron spectroscopy, the fraction of Si and C atoms incorporated into the tetrahedral Si–C bonds ranged from 60 to 70 % in all a-SiC films. No contribution by Si–H/Si–Si bonds was identified. Si to C ratio was in the range of 0.88 to 0.99 in the unintentionally doped a-SiC. The presence of sp2 C–C/C–H bonds in the a-SiC indicates the network of a-SiC is neither fully chemically ordered nor does it completely follow the tetrahedral structure. By comparing samples deposited at 650 ºC and 600 ºC at a constant methylsilane pressure, it was seen that a-SiC deposited at higher temperatures had a higher Si–C bond percentage and less oxygen concentration, and its chemical composition was less sensitive to the substrate type. The Al concentration increased when the sample was placed further away from the gas inlet, increasing from about 4.1 at. % to 6.3 at. %. Al–Al metallic bonds were detected only in SiC film deposited on quartz substrate, which might indicate that Al concentration is beyond its solubility limit. The incorporation of Al reduced the fraction of sp2 C–C/C–H bonds and increased the fraction of sp3 C–C/C–H bonds. The optical energy gap and absorption coefficient were derived based on optical transmittance and reflectance measurements. The optical energy gap of unintentionally doped SiC was in the range of 1.6 to 4.2 eV, with the absorption coefficient varying in the range of 103 to 105 cm-1. For SiC films deposited at 600 ºC, the incorporation of Al not only narrowed the optical energy gap by 1.0 eV but also reduced the absorption coefficient by one order of magnitude. Both a decrease in methylsilane pressure and an increase in substrate temperature widen the optical energy gap of deposited SiC. The unintentionally doped SiC was n type conductive and Al-doped SiC was p type conductive, which were determined by hot-probe and capacitance-voltage techniques. The conductivity was derived using a modified four-point probe technique. It ranged from 2.7 × 10-7 S cm-1 to 2.1 × 10-3 S cm-1 for the unintentionally doped a-SiC, from 8.1 × 10-5 to 1.5 S cm-1 for the unintentionally doped nc-SiC, and from 7.0 × 10-3 S cm-1 to 1.0 × 101 S cm-1 for the Al-doped SiC. The linear I-V characteristics of the Al-doped SiC demonstrate that conduction was by a drift of holes in the valence band, and the temperature dependence of the conductivity in Al-doped sample could be modelled by acceptors 0.20 eV above the valence band edge with the mobility limited by ionic impurity scattering.Thesis (PhD Doctorate)Doctor of Philosophy (PhD)Griffith School of EngineeringScience, Environment, Engineering and TechnologyFull Tex
Design and Fabrication of Sic Micro-Transducers With Large Q-Factors for High Resolution Sensing
No full textGravimetric sensing with microresonators is the most accurate means for molecular recognition applications, i.e. specific molecule sensing. The molecular recognition sensitivity is determined by frequencyquality factor (fQ) figure of merit, which is influenced by resonator’s type, geometry, material, and damping parameters. Cubic silicon carbide has outstanding mechanical and chemical properties, which make it excellent for resonant sensing applications. We have fabricated epitaxial silicon carbide on silicon resonators using silicon surface micromachining. We demonstrate that outstanding fxQ products could be achieved on string resonators: by increasing the length and the tensile stress, by high vacuum operation, and by improvement of crystal quality and clamping condition. We have achieved fxQ products in the order of ~1012 Hz, which are better than the state-of-the-art silicon nitride strings.
We also show the reduction of metal damping by growing graphene overlayer on epitaxial silicon carbide membranes, through a novel transfer-free alloy-mediated approach, for electrical transduction purposes. We report that graphene results in much smaller quality factor reduction (factor of 2) as compared to conventional metals overlayer (an order of magnitude). This is the highest transfer-free quality graphene reported so far on large silicon substrates; based on solid source growth from epitaxial silicon carbide.Thesis (PhD Doctorate)Doctor of Philosophy (PhD)Griffith School of EngineeringScience, Environment, Engineering and TechnologyFull Tex
- …
