Examensarbeten för masterexamen
Länka till denna samling:
Browse
Senast publicerade
Visar 1 - 5 av 260
- PostDesign and Analysis of a Narrowband Stepped Digital Elliptic Filter at 5 GHz(2024) Svanberg, Hans; Chalmers tekniska högskola / Institutionen för mikroteknologi och nanovetenskap (MC2); Chalmers University of Technology / Department of Microtechnology and Nanoscience (MC2); Stake, Jan; Deleniv, Anatoli; Pereira, PatriciaThe increasing frequency and complexity of modern radio frequency (RF) systems makes more compact components essential, particularly in antenna arrays, where numerous electronic components must fit within the small spaces between antenna elements. Filters are a crucial component to every RF system, making the development of more compact filter solutions necessary. Furthermore, new and precise microfabrication methods are emerging where the cost of production is influenced by the physical size of the components. In 1969, a narrowband bandpass filter structure, the stepped digital elliptic filter, was introduced. It has received little to no attention in subsequent research due to difficulties with tuning and fabrication. To the best of the author’s knowledge, this is the most compact higher-order elliptic function bandpass filter structure consisting solely of distributed elements capable of achieving n − 1 transmission zeros. The unique structure also allows the resonators to be suspended without the use of dielectrics, enabling high Q values. Considering advances in microfabrication technology and the need for more compact solutions, it is now an opportune time to revisit this filter structure. Starting from network theory and progressing to electromagnetic simulations, a filter prototype with a fractional bandwidth of 11%, a ripple level of 10%, and a minimum stopband attenuation of 40 dB was designed at 5 GHz, demonstrating close agreement between the simulated and theoretical results, with a return loss above 15 dB in the passband. The filter prototype was shown to be sensitive to fabrication errors, with tolerances greater than 5 μm severely degrading the return loss. Two different designs are proposed suitable for fabrication with the mentioned specifications. In addition, filter characteristics and capacitance matrices (relating directly to the physical dimensions of the filter) for stopband attenuation levels of 30 dB, 40 dB and 60 dB, return losses of 28 dB and 14 dB, and fractional bandwidths (FBW) between 1% and 30% are presented, as well as the structure’s dependence on scaling of the capacitance matrices. Finally, the filter structure has been shown to be suitable for design using microfabrication, with almost any size and filter characteristics with FBWs below 30% possible with sufficiently precise manufacturing tolerances.
- PostChemical vapor deposition of graphene on prepatterned catalystLin, Xiaotian; Chalmers tekniska högskola / Institutionen för mikroteknologi och nanovetenskap (MC2); Chalmers University of Technology / Department of Microtechnology and Nanoscience (MC2); Yurgens, August; Khan, MunisThis thesis explores the growth of high-mobility monolayer graphene on copper foils using Chemical Vapor Deposition (CVD), focusing on the influence of surface morphology and grain size of the copper substrate. Graphene’s exceptional electrical, thermal, and mechanical properties make it a promising material for various advanced electronic applications, particularly in the development of graphene field-effect transistors (GFETs). However, achieving high-quality graphene with uniformity and minimal defects remains a significant challenge, primarily due to the variations in the copper substrate’s surface characteristics. In this work, we investigate the correlation between the copper foil’s surface morphology and grain size and the quality of the graphene produced. The study employs low-pressure CVD (LPCVD) with methane as the carbon source, and hot-wall thermal CVD to precisely control the deposition environment. Key factors such as the copper substrate’s crystallinity, smoothness, and catalytic properties are analyzed to optimize the growth process. The self-limiting nature of graphene growth on copper, facilitated by copper’s low carbon solubility, is leveraged to achieve uniform monolayer graphene. This project also includes the prepatterning method as one of the ways of controlling the alignment of nuclei, which is done based on maskless photo-lithography. By using prepatterning, patterned oxidation on copper foils can be achieved, the nucleation of graphene during CVD growth can only happen on the open areas where there’s no oxide layer, which controls the position of nuclei. Furthermore, this research aims to enhance the copper substrate preparation methods to improve graphene’s mobility and overall quality, thus making it more viable for large-scale production. The outcomes of this study could contribute to advancing the scalability of high-performance graphene for industrial applications, particularly in the field of nanoelectronics.
- PostBenchmarking the design, fabrication and shielding of superconducting qubits(2024) Jakobsson, Halldór; Chalmers tekniska högskola / Institutionen för mikroteknologi och nanovetenskap (MC2); Chalmers University of Technology / Department of Microtechnology and Nanoscience (MC2); Bylander, Jonas; Tancredi, Giovanna; Osman, AmrSuperconducting qubits interact with their environment, which inherently limits their coherence. In this thesis, we describe measurements and present experimental results for characterising the qubits interaction with its environment. We performed flux pulse distortion measurements which allowed us to correct the flux pulse with 0.2% error of the target pulse. A two-level-system (TLS) defect spectroscopy measurement showed a TLS density of 14 ± 2 TLS/GHz, with TLS-qubit coupling between 50 kHz and 650 kHz, and TLS lifetime less than 200 ns. Repeating the TLS spectroscopy measurement every hour for twelve hours showed TLS drifting and telegraphic switching between frequencies with a timescale on the order of hours. Using a parity selective Ramsey sequence, we measured the qubit parity switching rate due to quasiparticles (QPs) as 500 ± 100 Hz, which randomly varied between 200 Hz to 800 Hz over 24 hours. We designed and measured a device where the qubit frequency could be tuned through the resonator. To this end, we measured the qubit decay rate as a function of frequency and found the single mode Purcell approximation to be negligible compared to TLS and QP losses in this device. In addition, the TLS loss was measured as F δTLS = (3±1)· 10−6 , and the quasiparticle density normalised to the density of Cooper pairs as xQP = (6±1)· 10−7 . In equilibrium, this quasiparticle density would corresponds to a qubit temperature of 147±2 mK, since the qubit is cooled down to 10 mK, we interpret this as the presence of non-equilibrium quasiparticles.
- PostEffect of adhesives on characteristics of CVD graphene(2024) Elvind, Charlotta; Chalmers tekniska högskola / Institutionen för mikroteknologi och nanovetenskap (MC2); Chalmers University of Technology / Department of Microtechnology and Nanoscience (MC2); Yurgens, August; Khan, MunisCharacteristics of graphene when transferred using various adhesives have been investigated. The primary objective was to identify an adhesive capable of withstanding elevated temperatures common in photolithography (100-150°C) without compromising graphene properties. The project consisted of three parts. First, the graphene growth by chemical vapor deposition (CVD) was optimized by adjusting gas flow rates during the process and input parameters such as time and temperature. The quality of the graphene was examined at different regions of the copper foil used as a catalyst in CVD. Second, the impact of different adhesives on graphene doping and charge-carrier mobility was explored. Finally, the role of multilayer patches in graphene’s interaction with adhesives was discovered by applying different adhesives to both sides of the graphene.
- PostPost-fabrication frequency tuning in superconducting transmon qubitsToselli, Maurizio; Chalmers tekniska högskola / Institutionen för mikroteknologi och nanovetenskap (MC2); Chalmers University of Technology / Department of Microtechnology and Nanoscience (MC2); Bylander, Jonas; Nylander, Andreas; Tancredi, GiovannaThis thesis presents a study on the manipulation of the normal state resistance RN of Al/AlOx/Al Josephson junctions using electrical DC signals. The goal is to develop a post-fabrication method for tuning the frequency of transmon qubits in superconducting quantum processors. This method aims to correct initial fabrication uncertainties and mitigate frequency collisions to scale up the number of qubits. The project initially focused on a precise, non-intrusive technique to measure RN at room temperature, which is directly related to the qubit plasma frequency f01 through the Ambegaokar-Baratoff relation. Subsequently, the natural aging of two types of devices, thin-oxide and thick-oxide junctions, was studied. Thin-oxide junctions proved less stable, exhibiting a resistance increase of more than 30% within the first two weeks after fabrication, compared to the 3% to 4% observed in thickoxide junctions. A procedure was then developed to deliberately increase RN at room temperature using high DC voltage biases, achieving increases of nearly 20% for thin-oxide devices and about 10% for thick-oxide ones. This demonstrates the potential for correcting fabrication variations on a wafer-level scale. An essential finding was the delayed resistance increase after manipulation, suggesting the need for further studies to better understand and control this effect. Theoretical modeling and simulations also revealed that the applied voltage plays a crucial role beyond simply delivering localized heat to the junction. Finally, a technology demonstration performed with a new experimental setup specifically designed to address individual qubits validated this method on a working quantum processor, providing a proof of concept. The results showed frequency shifts of several hundred megahertz after manipulation, without compromising qubit lifetimes. Despite a systematic overestimation of the final frequencies, likely due to the delayed resistance increase, this thesis successfully demonstrates a promising frequency tuning method and suggests future optimization and implementation directions.