Tyndall National Institute - Doctoral Theses
Permanent URI for this collection
Browse
Recent Submissions
Item AI-enabled chipless RFID sensing system for reliable IoT applications(University College Cork, 2024) Rather, Nadeem; O'Flynn, Brendan; Buckley, John; Tedesco, Salvatore; Simorangkir, Roy B.V.B.; Science Foundation IrelandThe Internet of Things (IoT) is growing rapidly, driving the need for innovative and sustainable solutions for wireless identification and environmental monitoring. Passive Radio Frequency Identification technology (RFID) has been a key wireless communication technology enabling IoT. Recent advances have paved the way for battery-less, chipless RFID (CRFID), which eliminates the need for an integrated circuit (IC) component on the tag. This PhD thesis presents a new design strategy for developing concentric rings-based polarization-insensitive CRFID sensing tags. The proposed tag design approach of exponential spacing results in an 88.2% higher tag data encoding capacity than conventional designs which incorporate uniform spacing of the resonant rings. This is coupled with the idea of using the innermost ring for capacitive sensing. The concept of using RCS nulls for data encoding is implemented to enable convenient and accurate sensing by the innermost ring. This is made possible by adding an extra ring at the tag’s outermost edge. To enable robust detection of these tags, Artificial Intelligence (AI) is integrated on the reader side, employing both machine learning (ML) and deep learning (DL) techniques for decoding RCS EM signatures. In this research, ML and DL regression modelling techniques are applied to a dataset of measured RCS data derived from large-scale automated measurements of custom-designed, 4-bit CRFID sensor tags. The robotic measurement system is implemented using the first-of-its-kind automated data acquisition method using an industry-standard robot. The results show that all the ML/DL models were able to generalize well, that is, the ability of a model to perform accurately on new, previously unseen data. However, the 1D-CNN DL models outperformed the conventional ML models in the detection of ID and sensing values. In another contribution, a 3-bit depolarizing CRFID tag is developed and enabled for surface and shape robust detection using AI. For the first time reported, the system was trained on a dataset of 12,600 EM signatures, capturing varying surface permittivity, tilt angles, read ranges, and tag bend scenarios. The AI models using 1D-CNN are trained and validated, resulting in a low RMSE of 0.040 (0.66%) for tag ID detection. On the same dataset, for the first time, DL models were evaluated with Bidirectional Long Short-Term Memory (Bi-LSTM) and attention mechanism, further reducing the RMSE to 0.029 (0.48%). The outcomes of this thesis contribute significantly towards state-of-the-art of AI-enabled CRFID systems for robust and reliable real-world IoT applications.Item Surface dynamics in III-V epitaxy and its device implications(University College Cork, 2024) Ozcan Atar, Ayse; Pelucchi, Emanuele; Juska, Gediminas; Science Foundation IrelandMetalorganic vapour phase epitaxy (MOVPE) is a well-established, industry-compatible, compound semiconductor crystal growth technique, allowing for efficient and controllable material deposition. A wide range of semiconductor devices, both from III-V and nitrides families, are commonly fabricated for a broad range of applications in photonics, electronics and related fields, due to the reproducibility, scalability and overall excellent control over the growth process. Nevertheless, despite the technique’s popularity, there persists a large number of unresolved issues (mostly related to growth process/dynamic understanding) effectively hindering some of potential developments of III-V devices. A major unresolved technological issue, is the reported long range “leakage” of the dopant Zn into intrinsic layers during (and post) epitaxy, including the device processing steps. Zn is used as a typical p-type dopant for III-V materials and devices, but it is reportedly highly diffusive and historically very problematic. To bypass the Zn related problems, the large majority of InP based photonic devices, such as lasers and modulators, are fabricated with an n-i-p design, using p-type dopant at the top of the device. This approach essentially limits the design degree of freedom and stands in the way of the novel advanced stacked device architectures. This work reevaluated the Zn doping issues with unprecedented and surprising findings on Zn dopant behavior. The secondary ion mass spectrometry (SIMS) experiments show that Zn (or its precursors) can behave as a surfactant; accumulating on the sample surface during the growth of intentional doping layer and gradually incorporating into the nominally undoped layers even after the Zn source is shut off. Experimental findings are modelled by combining the surfactant and diffusion behavior with good qualitative agreement. Also, we demonstrated that this phenomenology can be suppressed and controlled either by introducing growth interruption steps or, even more effectively, introducing a competing surfactant species (Sb or its precursors). Our results highlight the relevance of (often overlooked) multi-faceted surface processes during MOVPE epitaxy, and help the implementation of robust solutions for novel device designs, crucially enabling next-generation integrated III-V applications.Item Modelling of thin film oxide growth and etching(University College Cork, 2023) Mullins, Rita; Nolan, Michael; Lam Research; Science Foundation IrelandAs integrated circuit technology follows Moore's Law and continues to shrink, conventional methods for depositing and etching thin films encounter numerous challenges. Moreover, the challenges intensify due to the expected widespread adoption of 'more-than-Moore' devices designed for non-conventional computing applications. Traditional deposition methods will struggle to deliver continuous films at increasingly thinner levels and on complex 3D structures. Moreover, traditional continuous-wave plasma etching faces problems in achieving precise critical dimension control, enhanced selectivity, and with minimal plasma damage, especially below the 10 nm scale. In response to these limitations, atomic-scale processing techniques have emerged; producing thin films tailored to meet the demands of smaller and more intricate structures crucial for future semiconductor devices. Atomic layer deposition (ALD) and atomic layer etching (ALE) offer uniform and conformal processing with precise thickness control and can be achieved using sequential, self-limiting thermal surface reactions. These are used in various applications, such as employing ALE for etching high dielectric metal oxides necessary for gate dielectrics in complex transistor structures such as Gate All-Around or Complementary FET, and utilizing ALD to deposit barrier/liner thin films within interconnects. Presently, ALD is widely used in the semiconductor industry whereas thermal ALE is in the early stage of development and is an emerging and promising frontier in thin film processing. It is difficult to investigate ALD and ALE reactions directly using experimental techniques. First principles density functional theory (DFT) can give deep insights into precursor chemistry and reaction mechanisms of ALD and ALE processes. In my thesis I studied the hydrogen fluoride (HF) pulse as the first step in thermal ALE of high dielectric metal oxides. A thermodynamic analysis is used to predict the temperature at which the targeted self-limiting (SL) reactions are favoured over continuous spontaneous etching (SE) in an ALE cycle. Furthermore, calculations of HF adsorption are performed on the oxide surfaces to understand the mechanistic details of the HF pulse and calculate theoretical etch rates. The results are compared between the metal oxides studied: monoclinic HfO2 and ZrO2, orthorhombic HfO2, amorphous HfO2 and ZnO. HCl is examined as an alternative to HF for crystalline HfO2, ZrO2 and ZnO. The second step in thermal ALE, the ligand exchange reaction is examined for crystalline HfO2 using HF and SiCl4 allowing us to determine how the target Hf species can be chlorinated before being eliminated as volatile Cl-containing species. A combined barrier and liner material incorporating Ru or Co into TaN has been proposed to replace the tri-layer stack of TaN/Ta/Cu for advanced interconnect technology. This will accommodate high aspect ratio trench structures with the continued miniaturization of devices and extend the use of Cu in interconnects for the next generation of electronic devices. Ruthenium and cobalt are also potential replacements for Cu in next-generation interconnects. In this thesis, DFT calculations are also used to explore the nature of N2/H2 terminated TaN surfaces that are produced after a plasma pulse (with H2 or N2/H2 plasma) in Plasma Enhanced ALD to incorporate Ru or Co into TaN. The reactivity of Ru and Co precursors is studied on stable NHx-terminated TaN surfaces. This work will help guide experimental PEALD for the incorporation of ruthenium or cobalt into TaN as a combined barrier and liner material. Finally the mechanism of Ru ALD using the novel precursor RuO4 and molecular H2 was investigated to elucidate the role of both reactants.Item Modelling and growth of boron containing alloys of III-Nitrides for their application in the ultraviolet range(University College Cork, 2023) O'Connor, Thomas; Parbrook, Peter James; Schulz, Stefan; Science Foundation IrelandIII-Nitride semiconductor materials such as AlN, GaN, InN and their alloys, can emit light from the infrared to the deep ultraviolet region. Strong polarisation fields and significant differences in the optimum growth conditions for these binary compounds, make it difficult to understand these materials. BN is a relatively new material in this group which has the potential to improve strain engineering, increase the flexibility of bandgap engineering along with reducing the overall polarisation charge when alloyed with other III-Nitride materials. In this work, the Schrödinger equation was solved for a single quantum well system consisting of BxGa1-xN/ AlyGa1-yN and to understand the impact BN had on the wavefunction overlap along with the emission energy. Incorporation of wz-BN with GaN thin layers appears to prevent plastic relaxation of these layers with respect to their substrate and at higher temperatures resulting in phase separation of the material. A mechanism for recognising this clustering of boron atoms in the material is proposed using X-ray diffraction is presented. B(Al)GaN/AlGaN multiple quantum wells (MQWs) were grown and excited by photoluminescence (PL), and emission wavelengths between 328-349 nm were obtained. The incorporation of the lowest amounts of wurtzite-BN appears to result in a redshift and an improvement in the PL emission intensity of the material. However, this benefit comes at the cost of nanovoids/ nanopits forming in the material under the growth conditions used. Nanomasking effects dominate for the smallest levels of BN incorporation, with higher degrees of disorder, propagating into the barrier regions, being observed as the B/III ratio increased, as well as a reduction in the PL intensity. To our knowledge, this is the first report of both a ternary and quaternary B(Al)GaN/AlGaN MQW stacks grown by MOCVD on a c-plane AlN/sapphire templates for UV emission.Item Diffuse reflectance spectroscopy in the identification of oral potentially malignant disorders(University College Cork, 2024) Fahy, Edward; Ni Riordain, Richeal; Burke, RayAims: Diffuse reflectance spectroscopy (DRS) examines the composition of tissue by analysing light reflected from inside the tissue. DRS has been applied to cancer diagnostics in liver, brain, breast and others. The aims of this study were to (i) to formulate a clinical protocol for use of DRS in diagnosis of oral potentially malignant disorder (OPMD), (ii) to explore the clinical utility for this probe in the mouth and (iii) to review the protocol and assess its potential use in clinical practice. Materials and Methods: A clinical protocol for use of the DRS probe was formulated based on literature review and clinical experience. A translational clinical research study with two groups, one with histologically confirmed OPMD (n=53) and a control group (n=27) were enrolled. All participants received DRS of mucosal surfaces, including areas of OPMD, in the mouth. The readings were then investigated to find reliable biomarkers and their accuracy to differentiate the two groups. Repeatability and reproducibility with two users were examined. Interviews were carried out with oral surgery clinicians after reading the protocol. Results: The protocol produced good results, which were grouped based on the site of acquisition. Our analysis found accuracy figures of 89% and 87% respectively for distinguishing OPMD and normal mucosa in the buccal mucosa and ventral tongue. This OPMD group was made up of mostly oral lichen planus (OLP), with small numbers of other diagnoses. (OLP n=41, oral leukoplakia n=5, others n=5) Accuracy figures for the dorsum of the tongue were poor. DRS was easy to use, quick and acceptable to patients. Repeatability was shown to be good, however reproducibility was fair. This may be due to inter-user pressure differences in the mouth. Clinicians were sceptical of this new technology, in part due to lack of evidence of clinical utility compared with biopsy. Conclusion: DRS has shown a new role in identifying oral lichen planus in the buccal mucosa and ventral tongue. This device may have poor accuracy in identifying oral lichen planus on the dorsal surface of the tongue. It has potential as a replacement for biopsy as a screening tool, however observed difficulties in clinical use preclude widespread implementation. Observed inter-user variability in the mouth has not been thus far explored in the DRS literature. Further studies are needed to confirm this observation.