Recent advances in molecular biology and nanotechnology have enabled scientists to study biological systems at molecular and atomic scales. This level of sophistication demands for new technologies to emerge for providing the necessary sensing tools and equipment. Recent studies have shown that terahertz technology can provide revolutionary sensing techniques for organic and non-organic materials with unprecedented accuracy and sensitivity. This is due to the fact that most of the macromolecules have vibrational and/or rotational resonance signatures in terahertz range. To further increase the sensitivity, terahertz radiation is generated and interacted with the bio-sample on a miniaturized test site or the so-called biochip. From the view point of generation and manipulation of terahertz radiation, the biochip is designed based on the same rules as in high frequency electronic chips or integrated circuits (IC). By increasing the frequency toward terahertz range, the conventional IC design methodologies and analysis tools fail to perform accurately. Therefore, development of new design methodologies and analysis tools is of paramount importance for future terahertz integrated circuits (TIC) in general and terahertz biochips in particular. In this thesis, several advancements are made in design methodology, analysis tool and architecture of terahertz and millimeter-wave integrated circuits when used as a biochip. A global and geometry independent approach for design and analysis of the travelling-wave terahertz photomixer sources, as the core component in a TIC, is discussed in details. Three solvers based on photonic, semiconductor and electromagnetic theories are developed and combined as a unified analysis tool. Using the developed terahertz photomixer source, a resonance-based biochip structure is proposed, and its operation principle, based on resonance perturbation method, is explained. A planar metallic resonator acting as a sample holder and transducer is designed, and its performance in terms of sensitivity and selectivity is studied through simulations. The concept of surface impedance for electromagnetic modeling of DNA self-assembled monolayer on a metal surface is proposed, and its effectiveness is discussed based on the available data in the literature. To overcome the loss challenge, Whispering Gallery Mode (WGM) dielectric resonators with high Q factor are studied as an alternative for metallic resonator. The metallic loss becomes very high at terahertz frequencies, and as a result of that planar metallic resonators do not exhibit high Q factor. Reduced Q factor results in a low sensitivity for any sensor using such resonators. Theoretical models for axially and radially layered dielectric resonators acting on WGM are presented, and the analytical results are compared with the measured data. Excitation of WGM through dielectric waveguide is proposed, and the critical coupling condition is explained through analytical formulation. The possibility of selecting one resonance among many for sensing application is also studied both theoretically and experimentally. A high sensitivity sensor based on WGM resonance in mm-wave and terahertz is proposed, and its sensitivity is studied in details. The performance of the proposed sensor is tested for sensing drug tablets and also liquid droplets through various measurements in mm-wave range. The comprehensive sensitivity analysis shows the ability of the proposed sensor to detect small changes in the order of 10−4 in the sample dielectric constant. The results of various experiments carried out on drug tablets are reported to demonstrate the potential multifunctional capabilities of the sensor in moisture sensing, counterfeit drug detection, and contamination screening. The measurement and simulation results obtained in mm-wave hold promise for WGM to be used for sensing biological solutions in terahertz range with very high sensitivity.

Recent advances in molecular biology and nanotechnology have enabled scientists to study biological systems at molecular and atomic scales. This level of sophistication demands for new technologies to emerge for providing the necessary sensing tools and equipment. Recent studies have shown that terahertz technology can provide revolutionary sensing techniques for organic and non-organic materials with unprecedented accuracy and sensitivity. This is due to the fact that most of the macromolecules have vibrational and/or rotational resonance signatures in terahertz range. To further increase the sensitivity, terahertz radiation is generated and interacted with the bio-sample on a miniaturized test site or the so-called biochip. From the view point of generation and manipulation of terahertz radiation, the biochip is designed based on the same rules as in high frequency electronic chips or integrated circuits (IC). By increasing the frequency toward terahertz range, the conventional IC design methodologies and analysis tools fail to perform accurately. Therefore, development of new design methodologies and analysis tools is of paramount importance for future terahertz integrated circuits (TIC) in general and terahertz biochips in particular. In this thesis, several advancements are made in design methodology, analysis tool and architecture of terahertz and millimeter-wave integrated circuits when used as a biochip. A global and geometry independent approach for design and analysis of the travelling-wave terahertz photomixer sources, as the core component in a TIC, is discussed in details. Three solvers based on photonic, semiconductor and electromagnetic theories are developed and combined as a unified analysis tool. Using the developed terahertz photomixer source, a resonance-based biochip structure is proposed, and its operation principle, based on resonance perturbation method, is explained. A planar metallic resonator acting as a sample holder and transducer is designed, and its performance in terms of sensitivity and selectivity is studied through simulations. The concept of surface impedance for electromagnetic modeling of DNA self-assembled monolayer on a metal surface is proposed, and its effectiveness is discussed based on the available data in the literature. To overcome the loss challenge, Whispering Gallery Mode (WGM) dielectric resonators with high Q factor are studied as an alternative for metallic resonator. The metallic loss becomes very high at terahertz frequencies, and as a result of that planar metallic resonators do not exhibit high Q factor. Reduced Q factor results in a low sensitivity for any sensor using such resonators. Theoretical models for axially and radially layered dielectric resonators acting on WGM are presented, and the analytical results are compared with the measured data. Excitation of WGM through dielectric waveguide is proposed, and the critical coupling condition is explained through analytical formulation. The possibility of selecting one resonance among many for sensing application is also studied both theoretically and experimentally. A high sensitivity sensor based on WGM resonance in mm-wave and terahertz is proposed, and its sensitivity is studied in details. The performance of the proposed sensor is tested for sensing drug tablets and also liquid droplets through various measurements in mm-wave range. The comprehensive sensitivity analysis shows the ability of the proposed sensor to detect small changes in the order of 10−4 in the sample dielectric constant. The results of various experiments carried out on drug tablets are reported to demonstrate the potential multifunctional capabilities of the sensor in moisture sensing, counterfeit drug detection, and contamination screening. The measurement and simulation results obtained in mm-wave hold promise for WGM to be used for sensing biological solutions in terahertz range with very high sensitivity.

L’objectif de ces travaux est le développement en technologie intégrée standard d’une topologiede ligne de propagation optimisée en termes de pertes, d’encombrement et de facteur de qualitéaux fréquences millimétriques. Cette topologie nommée S-CPW (Shielded CoPlanarWaveguide) utilise le phénomène d’ondes lentes afin de miniaturiser longitudinalement la ligned’un facteur compris entre 1,3 et 3,2 par rapport à des topologies classiques. Disposantégalement de faibles pertes, les lignes développées présentent un facteur de qualité élevé parfoissupérieur à 40, à 60 GHz. A partir de l’étude du champ électromagnétique dans la structure, unmodèle électrique a été développé. C’est le premier modèle dans la littérature prenant en compteles pertes dans ce type de guide d’onde. Plusieurs dispositifs passifs intégrés réalisés avec deslignes S-CPW dans différentes technologies CMOS ont été caractérisés jusqu’à 110GHz. Lacompacité et les faibles pertes d’insertion obtenues pour la mesure de filtres à stubs et dediviseurs de puissance permettent de réussir l’intégration de circuits passifs compacts entechnologie microélectronique CMOS standard aux fréquences millimétriques. / This work focuses on high-performance S-CPW (Shielded CoPlanar Waveguide) transmissionlines in classical CMOS integrated technologies for the millimeter-wave frequency band.Thanks to an important slow-wave phenomenon, the physical length of S-CPW decreases by afactor from 1.3 to 3.2 compared with classical transmission lines. Presenting also lowattenuation loss, the developed transmission lines show very high quality factor (higher than 40at 60 GHz). The precise study of the electromagnetism field leads to an electrical model forS-CPWs. This is the first model that takes the losses in this topology into account. Then, somebasic passive circuits designed with S-CPWs and characterized up to 110 GHz are presented invarious CMOS technologies. The low insertion losses and relative low surfaces of a powerdivider and a passband filter show the great interest of S-CPW to integrate compact passivecircuits in classical CMOS technologies at millimeter-wave frequencies.

This dissertation presents the design, fabrication, and characterization of low-voltage capacitive RF MEMS switches. Although, RF MEMS switches have shown superior performance as compared to the existing solid-state semiconductor switches and are viable alternate to the present and the future communication systems, not been able to match the commercial standards due to their poor reliability. Dielectric charging due high actuation is one of the major concerns that limit the reliability of these switches. Hence, the focus of this thesis is on the development of low actuation voltage RF MEMS switches without compromising much on their RF and dynamic performances i.e., low insertion loss and high isolation. Four different switch topologies are studied and discussed. Electromechanical and electromagnetic modelling is presented to study the effect of various components that comprise a MEMS switch on the transient and the RF behaviour. The analytical expressions for switching and release times are established in order to estimate the switching and release times. An in-house developed surface micromachining process is adapted for the micro fabrication. This process eliminates the need for an extra mask used for the anchors and restricts the overall process to four-masks only. These switches are fabricated on 500 µm thick glass substrate. A 0.5 µm thick gold film is used as the structural material. For the final release of the switch, chemical wet etching technique is employed. The fabricated MEMS switches are characterized mechanically and electrically by measuring mechanical resonant frequency, quality factor, pull-in, and pull-up voltages. Since, low actuation voltage switches have slow response time. One of the key objectives of this thesis is to realize switches with fast response time at low actuation voltage. Measurements are performed to estimate the switching and release times. The measured Q-factors of switches are found to be in between 1.1 -1.4 which is the recommended value for Q in MEMS switches for a suppressed oscillation after the release. Furthermore, the effect of hole size on the switching dynamics is addressed. RF measurements are carried out to measure the S-parameters in order to quantify the RF performance. The measured results demonstrate that these switches need low actuation voltage in range of 4.5 V to 8.5 V for the actuation. The measured insertion loss less than -0.8 dB and isolation better than 30 dB up to 40 GHz is reported. In addition, the robustness of realized switches is tested using in-house developed Lab View-based automated measurement test set-up. The reliability test analysis shows no degradation in the RF performance even after 10 millions of switching cycles. Overall yield of 70 -80% is estimated in the present work. Finally, the experimentally measured results presented in this work prove the successful development of low actuation voltage capacitive RF MEMS switches and also offers that even with 0.5 µm thick gold film better reliability for MEMS switches can be achieved.

Le sujet de thèse abordé dans ce mémoire s'inscrit dans la thématique du LE²P sur l'autonomie énergétique des réseaux de capteurs. Ce travail est axé sur la partie réception et redressement du transfert de l'énergie sans fil pour l'apport d'énergie à des capteurs nomades. Ce procédé n'est pas nouveau et prend son origine dans les années 1950. Les connaissances dans l'appréhension de ce processus sont nombreuses pour certains guides d'onde tels que le microruban. Mais la nécessité de perçages dans ces structures de guide d'onde peut être contraignante et causer des disparités dans une chaîne de construction. Ceci a motivé les travaux présentés dans ce mémoire qui utilise une technologie de guide d'onde coplanaire (CPW) peu exploitée. Ainsi, la conception d'un tel dispositif passe par la maîtrise d'un point de vue conceptuel et expérimental de cette technologie. La démarche consiste à utiliser ce guide d'onde coplanaire en minimisant les effets négatifs que peut engendrer ce dernier, pour s'abroger du besoin de perçage et faciliter la réalisation des dispositifs de redressement en limitant le nombre d'interactions humaines. / The thesis subject dealt in this report lies in the LE²P framework on the energy sustainability of wireless sensor network. This work is dedicated to the reception and rectifying part of wireless power transfer to give energy sustainability to nodes in a sensor network. This process is not new and originate from the years 1950. The behavior of this process is since well-known in several waveguide such technology as microstrip. But the need of drill in those waveguide circuit may be inconvenient and lead to discrepancy from one circuit to another. This was the motivational keystone to the work address in this report which uses coplanar waveguide (CPW) over microstrip. The conception of such devices goes through a good conceptual and experimental understanding of the waveguide technology. The approach in this document consists of using coplanar waveguide while minimizing its drawbacks, in order to avoid drilling in the substrate and ease the realization of the rectifying part by limiting the human interaction.

The largest 21st century radio telescope, the Square Kilometre Array (SKA) is now being planned, and the first phase of construction is estimated to commence in the year 2016. Phased array technology, the key feature of the SKA, requires the use of a tremendous number of receivers, estimated at approximately 37 million. Therefore, in the context of this project, the Low Noise Amplifier (LNA) located at the front end of the receiver chain remains the critical block. The demanding specifications in terms of bandwidth, low power consumption, low cost and low noise characteristics make the LNA topologies and their design methodologies one of the most challenging tasks for the realisation of the SKA. The LNA design is a compromise between the topology selection, wideband matching for a low noise figure, low power consumption and linearity. Considering these critical issues, this thesis describes the procedure for designing a monolithic microwave integrated circuit (MMIC) LNA for operation in the mid frequency band (400 MHz to 1.4 GHz) of the SKA. The main focus of this work is to investigate the potential of MMIC LNA designs based on a novel InGaAs/InAlAs/InP pHEMT developed for 1 µm gate length transistors, fabricated at The University of Manchester. An accurate technique for the extraction of empirical linear and nonlinear models for the fabricated active devices has been developed. In addition to the linear and nonlinear model of the transistors, precise models for passive devices have also been obtained and incorporated in the design of the amplifiers. The models show excellent agreement between measured and modelled DC and RF data. These models have been used in designing single, double and differential stage MMIC LNAs. The LNAs were designed for a 50 Ω input and output impedance. The excellent fits between the measured and modelled S-parameters for single and double stage single-ended LNAs reflects the accurate models that have been developed. The single stage LNA achieved a gain ranging from 9 to 13 dB over the band of operation. The gain was increased between 27 dB and 36 dB for the double stage and differential LNA designs. The measured noise figures obtained were higher by ~0.3 to ~0.8 dB when compared to the simulated figures. This is due to several factors which are discussed in this thesis. The single stage design consumes only a third of the power (47 mW) of that required for the double stage design, when driven from a 3 V supply. All designs were unconditionally stable. The chip sizes of the fabricated MMIC LNAs were 1.5 x 1.5 mm2 and 1.6 x 2.5 mm2 for the single and double stage designs respectively. Significantly, a series of differential input to single-ended output LNAs became of interest for use in the Square Kilometre Array (SKA), as it utilises differential output antennas in some of its configurations. The single-ended output is preferable for interfacing to the subsequent stages in the analogue chain. A noise figure of less than 0.9 dB with a power consumption of 180 mW is expected for these designs.

In this thesis is discussed a filter design of a band-pass substrate integrated waveguide filter on the frequency of 5 GHz. It also includes design theory, filter synthesis, filter analysis in the program CST Microwave Studio® and the results of measurement on a vector network analyzer. The analysis results are compared with the measured results. Good agreement between simulated and measured results is observed.

Les ondes de spin sont une des voies technologiques proposées pour surmonter les obstacles que rencontre la miniaturisation des complementary metal-oxide-semiconductor (CMOS) dans la gamme du nanomètre, comme en témoignent les derniers développements en matière de dispositifs logiques à ondes de spin. Cependant, l'attrait industriel de ces preuves de concept est conditionné par leur intégration évolutive à la technologie CMOS. Ici, nous présentons des pistes ultrafines de Ta/CoFeB/MgO utilisées comme guides d'ondes de spin. Ce système a été choisi pour sa compatibilité avec les procédés CMOS, son anisotropie magnétique perpendiculaire et ses fortes interactions spin-orbite. Ces derniers sont intéressants pour manipuler les ondes de spin et ont été caractérisés par résonance ferromagnétique à couple de spin où il est démontré que l'effet Hall de spin inverse est responsable de la détection de la dynamique de magnétisation. Ensuite, nous utilisons des guides d'ondes coplanaires nanométriques intégrés pour exciter localement des ondes de spin dans une large gamme de vecteurs d'ondes. La comparaison du spectre d'ondes de spin mesuré avec les calculs analytiques montre que l'effet Hall de spin inverse permet la détection des ondes de spin indépendamment de leur vecteur d'onde avec des longueurs d'onde allant jusqu'à 150 nm. Des expériences complémentaires de diffusion de la lumière de Brillouin révèlent que les ondes de spin dans le guide d'ondes de spin ultra-mince à anisotropie magnétique perpendiculaire ont des longueurs de propagation étonnamment élevées compte tenu de l'amortissement relativement élevé des systèmes Ta/CoFeB/MgO. Ces résultats ouvrent la voie à des dispositifs à ondes de spin ultraminces compatibles CMOS avec des techniques d'excitation et de détection évolutives jusqu'à l'ordre du nanomètre, avec la perspective de contrôler les ondes de spin via des couples spin-orbite. / Spin-waves have been proposed as a possible technological path to overcome the hurdles encountered by the miniaturization of complementary metal-oxide-semiconductor (CMOS) into the nanometer range, demonstrated by recent developments in spin-wave-based logic devices. However the industrial appeal of these proofs-of-concept is conditional upon their scalable integration with CMOS technology. Here, we report on ultrathin Ta/CoFeB/MgO wires used as spin-wave waveguides. This system is chosen for its compability with CMOS processes, its perpendicular magnetic anisotropy and strong spin-orbit interactions. The latter are of interest for manipulating spin waves and are characterized via spin-torque ferromagnetic resonance where it is shown that the inverse spin Hall effect is responsible for the detection of magnetization dynamics. Following this, we use integrated nanometric coplanar waveguides to locally excite spin-waves in a broad range of wavevectors. Comparison of the measured spin-wave spectrum with analytical calculations show that the inverse spin Hall effect allows the wavevector-independent detection of spin-waves with wavelengths down to 150 nm. Complementary Brillouin light scattering experiments reveal that spin-waves in the ultrathin spin-wave waveguide with perpendicular magnetic anisotropy have unexpectedly high propagation lengths considering the relatively high damping in Ta/CoFeB/MgO systems. These findings pave the way for ultrathin CMOS-compatible spin-wave devices with excitation and detection techniques that are scalable into the nanometer range, with the prospect of controlling spin-waves via spin-orbit torques.

No description available.

No description available.