Organic phototransistors (OPTs) can enable essential applications, such as nonvolatile memory, artificial synapses, and photosensors in next-generation optical communication and wearable electronics. Among these applications, nonvolatile OPT memories are particularly promising, as they can retain captured visual information for extended periods, making them valuable for data storage, image and video processing applications. The capability of storing multi-bit information, which provides a low-cost way to increase the memory density per unit cell area, is one of the most critical challenges of memory products. In this work, we explore different solution-processible electrets to obtain highly sensitive phototransistor memory devices. Different planar heterojunctions, including small molecule/small molecule and small molecule/polymer, are used to fabricate OPT memories. Additionally, we explore the feasibility of producing polymer/polymer planar heterojunctions through printing processes. Firstly, OPT memories that can be programmed with white light and erased by applying a negative voltage are fabricated with a planar heterojunction of a nonconductive nanographene layer and a semiconducting layer of 2,9-didecyldinaphtho[2,3-b:2’,3’-f]thieno[3,2-b]thiophene (C10-DNTT). We systematically study the optical and memory characteristics of devices with an 8 nm nanographene (NG) layer. The photosensitivity of such devices can be as high as 3.4×105. The memory also shows quite good endurance and data-storing stability; an endurance of 100 write-read-erase-read (WRER) cycles and 1.5×105 s retention time are obtained. The thickness of the NG layer has a considerable influence on the performance of fabricated devices. The results suggest that devices with a thicker NG layer are more sensitive to weak light. In comparison, devices with a relatively thin NG layer are found to be promising for multi-bit photo memory devices. Secondly, we fabricate OPT memories by replacing the nanographene layer with a commercially available semiconducting polymer, namely Poly(2,5-bis(2-octyldodecyl)-3,6-di(pyridin-2-yl)-pyrrolo[3,4-c]pyrrole-1,4(2H,5H)-dione-alt-2,2’-bithiophene) (PDBPyBT). This polymer possesses a narrow bandgap and exhibits a broad range of light absorption, spanning from ultraviolet (UV) to red light wavelengths. As a result, the fabricated devices are capable of responding to a broad spectrum of light colors. The light response of these devices is investigated in terms of their reaction to different colors of light. Also, devices with varying thicknesses of the PDBPyBT layer are fabricated and studied. The results indicate that all of the fabricated devices demonstrate multi-bit programming properties, and the devices incorporating a thin, ribbon-structured PDBPyBT layer are particularly well-suited for applications as light dosimeters. Moreover, the results highlight that both the C10-DNTT and the PDBPyBT layer function as photo exciton generation and charge-trapping layers. Last, we seek to fabricate cost-effective organic multilayer devices through a solution-processing approach, eliminating the need for orthogonal solvents. We observe a crosslinking effect in the thin films caused by thermal annealing without using any crosslinker. Remarkably, this effect is found to be universal for several commercial semiconducting polymers investigated in our study. Following annealing at 200 ºC or higher temperatures, the thin films exhibit enhanced stability against the original solvent. Various analytical techniques are employed to examine the thin films to gain insights into the microstructural changes. Our results suggest that the observed crosslinking effect is predominantly attributed to a physical transformation, whereby the films became more crystalline after annealing at relatively high temperatures. To further explore the feasibility of fabricating multilayer devices, we simulate the construction of multilayer devices by top-gate-bottom-contact (TGBC) devices using the same solvent for the polymer dielectric layer and the semiconducting layer. We also fabricated planar polymer/polymer heterojunction via this method. Encouragingly, this approach demonstrated that thermal annealing could work as a straightforward and promising method for producing cost-effective organic multilayer devices, e.g., fully solution-processed diodes, functional transistors, and solar cells.:Abstract iii Contents vii 1 Introduction 1 1.1 Motivation 1 1.2 Organic semiconductor 2 1.2.1 Atom orbitals and molecular orbitals 2 1.2.2 Energy levels in solid 5 1.2.3 Fermi level 6 1.2.4 Band bending 7 1.2.5 From orbital to states 8 1.2.6 Organic semiconductor materials 9 1.2.7 Nanographene 10 1.2.8 Charge carrier transport in organic semiconductors 11 1.3 Organic field-effect transistors (OFET) 11 1.3.1 OFET architectures 12 1.3.2 OFET operation principle 12 1.3.3 OFET performance parameters 14 1.3.4 OFET memory 17 1.4 Optical electronics 20 1.4.1 Exciton pair generation. 20 1.4.2 Photoelectronic devices 21 1.4.3 Phototransistor devices 22 1.5 Phototransistor memories 23 1.5.1 Working mechanism of phototransistor memories 23 1.5.2 Phototransistor memory architecture 24 1.5.3 State-of-the-art organic phototransistor memory 25 1.6 Objective and outline 27 2 Materials and methods 29 2.1 Materials 29 2.2 Device fabrication 30 2.2.1 Substrate cleaning 30 2.2.2 Solution shearing 30 2.2.3 Thermal vapor deposition 31 2.3 Characterization 31 2.3.1 Thin film characterization 31 2.3.2 Current voltage characteristics 35 2.3.3 Capacitance 36 3 C10-DNTT/NG planar heterojunction phototransistor memories 37 3.1 Introduction 37 3.2 Thin films 39 3.2.1 Film and device fabrication 39 3.2.2 Characterization of thin films 39 3.3 Transfer characteristics under light 41 3.3.1 Writing process 41 3.3.2 Erasing process 48 3.3.3 C10-DNTT-only devices 51 3.4 Summary of working principle 52 3.5 Output characteristics and evaluation of the optical properties 52 3.6 Memory properties of NG-based OPT memory devices 55 3.7 Devices with different NG thicknesses 56 3.7.1 The impact of NG thickness 56 3.7.2 Devices fabricated from 0.05 mg ml-1 NG solution 60 3.8 Conclusion 64 4 C10-DNTT/PDBPyBT heterojunction phototransistor memories 67 4.1 Introduction 67 4.2 Device Architecture 68 4.3 Physical characterization of PDBPyBT and C10-DNTT thin films 69 4.4 Performance of devices with a thick PDBPyBT layer 72 4.4.1 Erasing and programming process 72 4.4.2 Response to different colors of light 78 4.5 Variation of PDBPyBT thickness 80 4.5.1 Transfer characteristics 80 4.5.2 Morphology of C10-DNTT 85 4.5.3 Output characteristics 86 4.5.4 Multi-level programming test 86 4.6 Comparison of the devices 92 4.7 Summary 93 5 Organic multilayer devices fabricated via thermal annealing 95 5.1 Introduction 95 5.2 Film Fabrication 97 5.3 Study on thin films 97 5.3.1 Thickness changes 97 5.3.2 Characterization of the thin films 99 5.3.3 Impact of re-annealing 107 5.3.4 Other semiconducting polymers 108 5.4 Discussion of the working mechanism 110 5.5 Impact of thermal annealing on devices’ performance 111 5.5.1 BGTC devices fabrication 111 5.5.2 TGBC devices fabrication 113 5.6 Planar heterojunction devices via solution processing 116 5.7 Conclusion 117 6 Conclusions and outlook 119 6.1 Conclusions 119 6.2 Outlook 120 Bibliography 123 List of Figures 143 List of Tables 155 List of abbreviations 157 Appendix A 159 Appendix B 165 B1.1 Introduction 165 B1.2 Devices with a 7 nm shear coated Al2O3 dielectric 166 B1.2.1 Normal-sized channel devices 166 B1.2.2 Ultra-wide channel devices 167 B1.3 Devices with a 30 nm ALD Al2O3 dielectric 169 B1.3.1 Normal-sized channel devices 169 B1.3.2 Ultra-wide channel devices 170 B1.4 Ferroelectric organic phototransistor devices 172 B1.4.1 Dielectric layer 172 B1.4.2 Devices with 10 nm HZO 173 B1.4.3 Devices with 30 nm HZO 175 Conclusion 176 Publications 177 Acknowledgment 179

Knowledge of the end winding inductance of electrical machines is decisive for calculating their operating performance. In this article, two different approaches to analytically calculate the stator end winding inductance of large induction machines are discussed. The first method is based on the exact replication of the 3D conductor geometry using serially connected straight filaments, where the inductances are calculated by solving Neumann’s integral. In the second method, the end winding flux is resolved into components excited by the axial and circumferential end winding magnetomotive force, resulting in a far simpler geometrical model. In both cases, end face effects are taken into account by adopting the method of images. The analytical approaches are compared to the known analytical calculation method proposed by Alger [1]. In addition, the stator end winding inductance is computed by means of 3D finite-element analysis. Using experimental validation, it is shown that both the analytical and numerical results reasonably correlate with removed rotor inductance measurements taken for several induction machines with different rated powers and frame sizes, if the permeability of the laminated core is taken into consideration.

The first focus of this thesis is a basic scientific research in the field of super-regenerative receivers (SRRs) and, in particular, super-regenerative oscillators (SROs). In chapter 3, a theoretical analysis of SROs is performed and its nonlinear behavior is studied in both time and frequency domains. The cross-coupled architecture is studied as a special case, however, widely used topology in the SRO design. The start-up and decay envelopes of the oscillator output are studied in relation to the input. The SRO start-up time and the maximal achievable quenching frequency are investigated. For phase modulation purposes, the relation between the initial phases of the input and output signals is investigated. In addition, a frequency-domain analysis is performed to ease the characterization of test circuits at high frequencies where time-domain measurements are not possible. The analytical results are verified by circuit-level simulations and measurements of a 2.4-GHz SRO. This study provides design guidelines for the design of SROs and helps in determining the optimal system parameters when targeting both amplitude and phase modulations. The SRO concept and its phase sampling capability are further studied in chapter 4. For this purpose, two SRO circuits operating around 60GHz are investigated, designed − in both CMOS and SiGe BiCMOS technologies − and characterized. Different circuit techniques are studied in order to improve the energy efficiency and maximize the switching speed. For instance, a novel quenching technique to maximize the power efficiency is presented. This circuit approach includes a pulse generator, which generates the quench signal allowing to switch the SRO with a minimum duty cycle. In fact, the SRO input is sampled at the onset of oscillation and, therefore, it is unnecessary for the oscillator to run for a long time in its steady state. This allows to minimize the on-time of the SRO and to considerably reduce its power consumption. Compared to the state of the art, the circuit achieves by far the highest power efficiency of 21.7%. In addition, a novel switching scheme is presented in order to maximize the quenching speed. This approach enables to greatly reduce the switching time constant, allowing to improve the SRO quench rate by approximately a factor of 3 compared to the state of the art. A record switching frequency of up to 10GHz is achieved. The second focus of this thesis is the investigation of on-off-keying (OOK) receivers operating at a carrier frequency of 60GHz. OOK modulation schemes have the advantage of low circuit complexity and, therefore, low power consumption and small chip area. The drawback of low spectral efficiency can be mitigated by the large bandwidth at mm-wave frequencies, in particular, at the 60-GHz ISM band, 9GHz from 57 to 66GHz, which allows to achieve high data rates. The goals of this study are enhancing the speed, maximizing the energy efficiency and improving the sensitivity. For this purpose, several circuit approaches are studied. A highly efficient envelope detector and a novel limiting amplifier architecture with simultaneously large bandwidth and high gain ensure high-speed low-power operation. A high-gain low noise amplifier is employed to increase the sensitivity. A feedforward dc-offset cancellation technique is used to ensure the proper operation of the receiver. As a proof of concept, the chips are implemented in a 130-nm SiGe BiCMOS technology. The OOK receiver achieves a record data rate of up to 20Gb/s at a bit-error-rate less than 10^(-12). The low power consumption and the ultra high speed capability allow a highly energy-efficient operation of only 2.2pJ/bit. Finally, this thesis studies the design of 60-GHz amplifiers. In fact, one of the main challenges of transmitting data at mm-wave frequencies is the high free-space path loss, which limits the transmission range. To alleviate this issue, amplifiers with high output power at the transmitter side and with low noise figure and high gain at the receiver side are required. Chapter 6 investigates the design of a low noise amplifier (LNA), with the study focus being achieving simultaneously large bandwidth, high gain, low noise figure and low power consumption in order to enable a high-speed receiver with high power efficiency and high range. The LNA achieves a bandwidth of 23GHz, a gain of 23.8dB and an average noise figure of 3.2dB and consumes only 8mW, remarkably improving the state of the art. In addition, this chapter presents the investigation and design of a 60-GHz power amplifier with an output power of up to 14.7dBm and a small chip area to ensure a sufficient communication range at the transmitter side.

Optoelektrische Baugruppen spielen eine entscheidende Rolle in der modernen Welt und finden in einer Vielzahl von Anwendungen Verwendung. Diese Baugruppen kombinieren die Optik mit der Elektronik und ermöglichen die Konvertierung von optischen Signalen in elektrische Signale und umgekehrt. Insbesondere in der Automotivbranche müssen diese Komponenten zusätzlich hohe Zuverlässigkeitsstandards erreichen, um beispielweise beim Autonomen Fahren eingesetzt zu werden. Ein wesentlicher Einflussfaktor ist die Gewährleistung einer kontaminationsfreien Oberfläche der aktiven Bereiche der optoelektrischen Baugruppen. Üblicherweise wird dies durch die Anwendung einer Verkapselung in Verbindung mit einer Glasabdeckung erreicht. In gängigen kommerziellen Verfahren erfolgt diese Verkapselung in der Regel am Ende des Fertigungsprozesses. Dadurch kann es während der vorherigen Fertigungsschritte leicht zu einer Verschmutzung der aktiven Bereiche kommen. In dieser Arbeit werden optoelektrische Baugruppen verwendet, wobei bereits im Wafer- Level-Verbund eine Glaskappe auf den aktiven Bereich montiert wird, die sie vor einer Kontamination der anschließenden Fertigungsschritte schützt. Aufgrund der frühzeitigen Montage der Glaskappe bleiben nachfolgend hergestellte Anschlussleitungen wie Drahtbonds ungeschützt. Diese Arbeit konzentriert sich auf die Entwicklung eines Hausungsverfahrens, um die Anschlussleitungen derartig hergestellter optoelektrischen Baugruppen zu schützen. Ein vielversprechender Lösungsansatz ergibt sich aus der Entwicklung eines stereolithographischen Prozesses und der Verwendung eines angepassten Hausungsmaterials. Besonders die Auswahl der Bestandteile des Hausungsmaterials ist entscheidend, um die geforderten Zuverlässigkeitsstandards zu erfüllen. Aus diesem Grund beinhaltet diese Arbeit eine gründliche Untersuchung des Hausungsmaterials, seiner Bestandteile und Eigenschaften. Erste Zuverlässigkeitsuntersuchungen von gehausten optoelektrischen Baugruppen zeigen vielversprechende Ergebnisse für das entwickelte Hausungsverfahren und das angepasste Hausungsmaterial.

Mit fortschreitender Energiewende wandelt sich sowohl die Erzeuger- als auch die Verbraucherstruktur des Elektrizitätsversorgungssystems. Als Folge resultieren in den Elektrizitätsversorgungsnetzen (Netz) bidirektionale witterungsabhängige Leistungsflüsse und eine höhere Auslastung der Betriebsmittel. Zur Vermeidung möglicher Überlastungen der Betriebsmittel erfolgen daher temporäre Schaltzustandsänderungen sowie dauerhafte Netzverstärkungs- oder Netzausbaumaßnahmen. Dabei müssen zur Gewährleistung einer weiterhin hohen Versorgungssicherheit durch die Verteilnetzbetreiber die Schutzkonzepte kontinuierlich hinsichtlich einer Einhaltung der allgemeinen Anforderungen Selektivität, Schnelligkeit und Zuverlässigkeit überprüft werden. In Anbetracht zunehmender Komplexität und Anzahl unterschiedlicher Netzzustandsszenarien müssen für die Schutzkoordination der im Netz eingesetzten Schutzsystemen neue Verfahren entwickelt werden. In vermascht betriebenen Netzen wird in der Regel der Distanzschutz als Schutzsystem zur Sicherstellung der Selektivität bei angemessener Schnelligkeit eingesetzt. Von zentraler Bedeutung sind dabei die Distanzzonen jedes Distanzschutzes als wesentlicher Baustein eines Selektivschutzkonzeptes. Die vorliegende Arbeit stellt ein deterministisches regelbasiertes Verfahren vor mit dem eine automatisierte Koordination von Distanzzonen mit den in einem Netzberechnungsprogramm ermittelten Schutzreaktionen möglich ist. Durch das Verfahren werden dabei sowohl anwenderdefinierte Netzzustandsszenarien als auch die in den Vorgaben der Schutzkoordination definierten allgemeinen Anforderungen an Selektivität, Schnelligkeit und Zuverlässigkeit berücksichtigt. Die Struktur des Verfahrens umfasst die fünf Teilfunktionen der Analyse eines Selektivschutzkonzeptes, Ermittlung der Schutzreaktionen, Bewertung und Anpassung der Parameter von Distanzzonen sowie eine Ablaufsteuerung. Als wesentliche Aspekte des Verfahrens sind für die Bewertung der Parameter von Distanzzonen Methoden entwickelt worden, um sowohl eine zeitlich kaskadierte Abschaltung eines Kurzschlusses durch mehrere Distanzschutzgeräte als auch eine negative Steigung in der Kennlinie einer gemessenen Fehlerreaktanz zu berücksichtigen. Weiterhin wurde eine qualitative und quantitative Bewertung der Koordinationsergebnisse hinsichtlich der Einhaltung der allgemeinen Anforderungen vorgeschlagen. Als Grundlage einer Erprobung des Verfahrens wird in der Arbeit zudem ein Datensatz zur Nachbildung einzelner und kombinierter Einflussfaktoren auf die Koordination von Distanzzonen vorgeschlagen. Als Einflussfaktoren werden zum einen netzbedingte Unschärfen einer Impedanzmessung aber auch mögliche Eigenschaften und Anordnungen der Schutzzonen, die als Netzabschnitte von Distanzschutzgeräten überwacht werden, verstanden, da durch diese die Parameter von Distanzzonen beschränkt und somit das Koordinationsergebnis beeinflusst wird. Durch die Auswahl von realitätsnahen Modellierungsparametern der Netze und Netzzustandsszenarien aus der Literatur ist ein Rückschluss auf die Anwendbarkeit des Verfahrens in realen Elektrizitätsversorgungsnetzen möglich. Durch Anwendung im vorgeschlagenen Datensatz und in einem realitätsnahen synthetischen 110-kV-Verteilnetz wird das Verfahren im letzten Teil der Arbeit erprobt. Dabei wird neben der Funktionsfähigkeit des Verfahrens auch dessen Grenze der Anwendbarkeit sowie Vor- und Nachteile aufgezeigt.

Highly dense Multiple-Input Multiple-Output (MIMO) radars provide robust detection at a high angular resolution in automotive applications. However, these systems require extensive parallel processing, higher off-chip communication data rates, and higher power consumption as a result of denser arrays to tackle. In this work, concepts, algorithms and circuits for the first-in-literature System-on-Chip (SoC) performing on-chip 5D localization of targets are researched and developed. The targeted system features a scalable homogeneous architecture for fully integrated signal processing of Frequency Modulated Continuous Wave (FMCW) signals and near-sensor smart capabilities for the identification of critical targets (e.g., pedestrians) in an automotive context. To achieve this functionality at a low-power budget, the interconnection of locally constrained processing nodes each with low-area and low-power accelerators operating at low frequencies is developed. The design of efficient algorithms, data flow, and cognitive concepts is also researched and presented to provide an unconventional holistic co-design approach. As a result of the research carried out in this thesis, a 12-ADC multi-core Digital Signal Processor (DSP) in 22FDX GLOBALFOUNDRIES (GF) using Adaptive Body Biasing (ABB) at 0.6V to enable smart on-chip classification with 25 processing elements (PEs) is proposed, providing a low power consumption of only 52.6mW that is at least 90x lower than the state-of-the-art commercial MIMO Radar DSPs.:Abstract Zusammenfassung List of Figures List of Tables Nomenclature 1. Introduction 1.1. Introduction 1.2. Motivation 1.3. Aims and Objectives 1.4. Contributions 1.5. Publications 1.6. Thesis Outline 2. Signal Processing for 5D Radars 2.1. Signal Processing Overview 2.2. Velocity Disambiguation Proposals for MIMO Radars 2.3. Cognitive Principles 2.4. Proposed Clustering for Constrained Processors 2.5. Target Classification 2.6. Summary of the Proposed Contributions 3. Hardware for 5D Radars: A Prototype Implementation 3.1. Multiprocessor Systems-on-Chip 3.2. FFT Acceleration 3.3. Silicon Implementation 3.4. Proposed Testchip 3.5. Proposed Implementation of the Processing Stages 3.6. Summary of Contributions 4. Conclusion and Outlook 4.1. Summary 4.2. Applications 4.3. Further Work 4.4. Conclusion A. Appendix A.1. FMCW Equations A.2. Dataset for Velocity Disambiguation A.3. Dataset for Machine Learning A.4. Dataset for Cognitive Radar Mode A.5. Reliability Proof for the Ambiguity Detector A.6. Measurement Setup A.7. EDA Tools Used in this Work Publications Bibliography / Hochdichte MIMO-Radar-Systeme bieten robuste Detektion mit hoher Winkelauflösung in Automobilanwendungen. Diese Systeme erfordern jedoch aufgrund dichterer Arrays umfangreiche parallele Verarbeitung, höhere Datenraten für die Off-Chip-Kommunikation und einen höheren Stromverbrauch. In dieser Arbeit werden Konzepte, Algorithmen und Schaltungen für den erstmaligen SoC-Ansatz zur On-Chip-Lokalisierung von Zielen in fünf Dimensionen untersucht und entwickelt. Das angestrebte System zeichnet sich durch eine skalierbare homogene Architektur für die vollständig integrierte Signalverarbeitung von FMCW-Signalen aus sowie durch intelligente Fähigkeiten in der Nähe des Sensors zur Identifizierung kritischer Ziele (z. B. Fußgänger) im automobilen Kontext. Um diese Funktionalität bei geringem Energieverbrauch zu erreichen, wird die Verbindung von lokal begrenzten Verarbeitungsknoten mit niedrigflächigen und energieeffizienten Beschleunigern entwickelt, die bei niedrigen Frequenzen arbeiten. Des Weiteren werden effiziente Algorithmen, Datenflüsse und kognitive Konzepte erforscht und präsentiert, um einen unkonventionellen ganzheitlichen Co-Design-Ansatz zu bieten. Als Ergebnis der in dieser Arbeit durchgeführten Forschung wird ein 12-ADC-Multi-Core-DSP in 22FDX GF mit ABB bei 0,6V vorgeschlagen, um eine intelligente On-Chip-Klassifizierung mit 25 Verarbeitungselementen (PEs) zu ermöglichen. Dieser Prozessor weist einen geringen Stromverbrauch von lediglich 52,6mW auf, der mindestens 90-mal niedriger ist als bei den modernsten kommerziellen MIMO-Radar-DSPs.:Abstract Zusammenfassung List of Figures List of Tables Nomenclature 1. Introduction 1.1. Introduction 1.2. Motivation 1.3. Aims and Objectives 1.4. Contributions 1.5. Publications 1.6. Thesis Outline 2. Signal Processing for 5D Radars 2.1. Signal Processing Overview 2.2. Velocity Disambiguation Proposals for MIMO Radars 2.3. Cognitive Principles 2.4. Proposed Clustering for Constrained Processors 2.5. Target Classification 2.6. Summary of the Proposed Contributions 3. Hardware for 5D Radars: A Prototype Implementation 3.1. Multiprocessor Systems-on-Chip 3.2. FFT Acceleration 3.3. Silicon Implementation 3.4. Proposed Testchip 3.5. Proposed Implementation of the Processing Stages 3.6. Summary of Contributions 4. Conclusion and Outlook 4.1. Summary 4.2. Applications 4.3. Further Work 4.4. Conclusion A. Appendix A.1. FMCW Equations A.2. Dataset for Velocity Disambiguation A.3. Dataset for Machine Learning A.4. Dataset for Cognitive Radar Mode A.5. Reliability Proof for the Ambiguity Detector A.6. Measurement Setup A.7. EDA Tools Used in this Work Publications Bibliography

Robots are machines, often resembling living creatures or parts of them, capable of autonomously performing complex tasks. They are becoming increasingly widespread because of their many possible applications. Soft robotics is the branch of robotics that studies devices made with soft and compliant materials. The aim is to more efficiently execute all those tasks that are hard to carry out with traditional robots: manipulation of delicate objects, exploration in harsh terrains, and interaction with humans in close contact are just some examples. While many soft actuators have been presented in the literature, they are usually integrated into a non-soft structure composed of sensors, electrical circuits, and other bulky parts such as pumps, motors, and valves. This work investigates the possibility of achieving entirely soft robots independent of external stiff and bulky components. Electroactive polymers (EAPs) are a good prospect in soft robotics because of their properties. They are a class of lightweight and soft materials that change shape or size when stimulated by an electric field. Thus, they do not need motors or pumps to generate movement. This contribution focuses on dielectric elastomers (DEs), a sub-class of EAPs. Due to their fast response time, low static energy consumption, and low elastic modulus, dielectric elastomer actuators (DEAs) are sometimes referred to as 'artificial muscles' and are used as such in soft robotic structures. Soft sensors, generators, and circuits can be produced with the same technology. Therefore, all the main, required components of robots can be built with the same soft materials, which makes DEs a potential candidate for realizing entirely soft autonomous robots. In particular, this work investigates the use of DEs for making electronic circuits and their integration into multi-functional structures. The focus is on the dielectric elastomer transistor (DET), defined as the standard cell to realize DE circuitry. A consistent method to design circuits with DEs is demonstrated in this thesis, stemming from the parallelism between DETs and conventional transistors. Furthermore, autonomous robotic devices comprising DE circuits are built and tested. DE circuits can control the movement of DEAs and can be equipped with DE sensors that allow the structure to respond to external stimuli. DEs perform actuation, sensing, and signal processing functions inside these smart structures. Thus, they are largely independent of external components. An example is given by the Venus flytrap robot that automatically catches objects, realized during this study. The final goal of the work is to illustrate the potentiality of autonomous control, through DE electronics, of entirely soft robots based on DEs.

In the last decades, ferroelectric properties have been observed in thin films based on hafnia and zirconia. These fluorite-structured thin films have attracted significant interest due to their excellent compatibility with complementary metal-oxide-semiconductor technology. Undoped zirconia has a lower crystallization temperature compared to hafnia-based or HfxZr1-xO2 thin films, making it more attractive for back-end-of-line processes. However, it is commonly reported that undoped zirconia thin films exhibit antiferroelectric properties. When ferroelectric properties are observed in undoped zirconia thin films, they are not as remarkable as those of hafnia-based or HfxZr1-xO2 thin films. In this work, in order to better understand the crucial factors that stabilize the ferroelectric properties in zirconia-based thin films, various process parameters were varied, and physical and electrical characterizations were carried out. The results indicate that the in-plane tensile strain is a crucial factor in stabilizing the ferroelectric orthorhombic phase of these films. The ferroelectric properties of zirconia films are lost when a non-polar monoclinic or tetragonal phase is stabilized due to an in-plane strain that is too small or too large, respectively. The best ferroelectric behavior is achieved when the polar orthorhombic phase can be stabilized with an in-plane tensile strain of 0.4-0.6%. After optimizing the interfaces between the ZrO2 and electrodes, the ferroelectric properties can be further improved in the ZrO2 thin films to be comparable to the HfO2-based ferroelectric thin films.:1 Motivation 2 Fundamentals 3 Experimental Setup 4 Effect of Process Parameters 5 Crucial Factors for Ferroelectric Zirconia 6 Conclusion

The discovery of ferroelectricity in the CMOS-compatible material HfO2 at nanometer scales has enabled the realization of low power, high speed, and dense nonvolatile memories. Even though this discovery can be exploited for memory applications, the reliability needs to be thoroughly examined and improved to meet the standard requirements. HfO2 has an intrinsic coercive field of ≈ 1 MV/cm, field-cycling with such a high field leads to defect generation and, subsequently, hard-breakdown of the material. The defect generation with field cycling is evident from the leakage current increase. Prior to the discovery of ferroelectricity, HfO2 was used for high-k applications, where it was experimentally proven that doping it with La suppresses the leakage current through the dielectric. Motivated by this experiment, this thesis was initiated through the additional doping of La into a mixed oxide of Hf0.5Zr0.5O2 to subsequently improve endurance. The experiment was successful but at the cost of non-polar t-phase stabilization, which induced a decrease in pristine Pr, a strengthening of the undesired wake-up effect, and a degradation of data retention. As shown in the literature, the Hf/Zr ratio can potentially change the phase composition of Hf1−xZrxO2. As a next step, the concentration of Hf/Zr in the ternary oxide of La: Hf1−xZrxO2 is altered to roll back the degradation while maintaining the benefits of La. The experimental results deviated a bit from the expectation, derived by the different origins of structural changes caused by the La and Hf/Zr ratio. After understanding the origin of La-induced structural changes, a different strategy was employed instead of doping La into Hf0.5Zr0.5O2. Here, films comprising a non-mixed La mono-layer and an in-series ferroelectric Hf0.5Zr0.5O2 are fabricated and measured, improving field-cycling endurance while successfully rolling back the cons of La doping. Hence, this thesis concludes with an overall improvement in material properties while avoiding any degradation.:Declaration of Authorship iii Abstract vii Acknowledgements ix 1 Introduction 1 2 Fundamentals 5 2.1 Models explaining ferroelectricity . . . . . . . . . . . . . . . . . . . . . . 6 2.1.1 Thermodynamic model . . . . . . . . . . . . . . . . . . . . . . . 6 2.1.2 Preisach model of hysteresis . . . . . . . . . . . . . . . . . . . . 8 2.1.3 Switching kinetic models . . . . . . . . . . . . . . . . . . . . . . 8 2.2 Ferroelectric materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 2.2.1 Prior to ferroelectric H fO2: Perovskite structure ferroelectric materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 2.2.2 Ferroelectricity in H fO2 : A scalable fluorite structure ferroelectric material . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 2.3 Ferroelectric Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 2.3.1 One transistor one capacitor FeRAM . . . . . . . . . . . . . . . . 22 2.3.2 Ferroelectric Tunnel Junction . . . . . . . . . . . . . . . . . . . . 23 2.3.3 Ferroelectric Field Effect Transistor . . . . . . . . . . . . . . . . . 24 2.4 Dielectric leakage current mechanisms . . . . . . . . . . . . . . . . . . . 26 2.4.1 Direct tunneling . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 2.4.2 Schottky emission . . . . . . . . . . . . . . . . . . . . . . . . . . 28 2.4.3 Fowler-Nordheim tunneling . . . . . . . . . . . . . . . . . . . . 29 2.4.4 Poole-Frenkel conduction . . . . . . . . . . . . . . . . . . . . . . 29 2.4.5 Trap assisted tunneling (hopping) . . . . . . . . . . . . . . . . . 29 3 Experimental 31 3.1 Processes for device fabrication . . . . . . . . . . . . . . . . . . . . . . . 31 3.1.1 Atomic layer deposition . . . . . . . . . . . . . . . . . . . . . . . 31 3.1.2 Physical vapor deposition . . . . . . . . . . . . . . . . . . . . . . 32 3.1.3 Rapid thermal annealing . . . . . . . . . . . . . . . . . . . . . . . 34 3.1.4 Chemical patterning of electrodes . . . . . . . . . . . . . . . . . 34 3.2 Test structure fabrication process . . . . . . . . . . . . . . . . . . . . . . 34 3.3 Characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 3.3.1 Structural characterization . . . . . . . . . . . . . . . . . . . . . 37 3.3.2 Electrical characterization . . . . . . . . . . . . . . . . . . . . . . 37 4 Results and Analysis 43 4.1 Doping Hf0.5Zr0.5O2 with La2O3 . . . . . . . . . . . . . . . . . . . . . . . 43 4.1.1 Structural characterization . . . . . . . . . . . . . . . . . . . . . 45 4.1.2 Electrical characterization . . . . . . . . . . . . . . . . . . . . . . 46 4.1.3 Summary and conclusion . . . . . . . . . . . . . . . . . . . . . . 55 4.2 Doping of Hf1−xZrxO2 with 2.3 mol% La2O3 . . . . . . . . . . . . . . . 57 4.2.1 Structural characterization . . . . . . . . . . . . . . . . . . . . . 57 4.2.2 Electrical characterization . . . . . . . . . . . . . . . . . . . . . . 59 4.2.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 4.3 The strong wake-up effect in La-doped films . . . . . . . . . . . . . . . 61 4.3.1 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 4.3.2 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 4.3.3 Summary and conclusions . . . . . . . . . . . . . . . . . . . . . . 68 4.4 Metal-ferroelectric interface engineering through La2O3 and Al2O3 . . 68 4.4.1 Structural characterization . . . . . . . . . . . . . . . . . . . . . 70 4.4.2 Electrical characterization . . . . . . . . . . . . . . . . . . . . . . 71 4.4.3 Summary and conclusions . . . . . . . . . . . . . . . . . . . . . . 79 5 Summary and outlook 85 A The strong wake-up effect in La doped films 89 A.1 Switching current peaks assignment . . . . . . . . . . . . . . . . . . . . 89 A.2 Derivation of depolarization field equation . . . . . . . . . . . . . . . . 90 A.3 Impact of polarization on leakage current measurements . . . . . . . . 90 Bibliography

Diese Arbeit stellt eine Untersuchung und experimentelle Analyse von Empfänger- (RX) und Sender- (TX) Frontends vor, die die Herausforderungen des MIMO-Radars in Kombination mit einer extremen Bandbreite von 4 GHz bis 40 GHz angehen. Diese Bandbreite wurde gewählt, um die Machbarkeit der Implementierung eines hochgradig rekonfigurierbaren MIMO-Abbildungsradarsystems zu untersuchen. Die RX- und TX-Frontends wurden so untersucht, dass sie in einem frequenzmodulierten Dauerstrichmodus (FMCW) arbeiten, so dass die vorgesehene Radaranwendung Objekte im Nahbereich genau messen kann. Es wurden neuartige Komponententopologien und Implementierungstechniken eingesetzt, um die Machbarkeit einer ultrabreiten Bandbreite in Kombination mit einer hohen Eingangsleistungsfestigkeit, einem hohen Dynamikbereich und einem hohen Integrationsgrad zu untersuchen. Die experimentelle Analyse der hergestellten IC-Komponenten hat gezeigt, dass eine extreme Bandbreite, die nicht nur 4 GHz bis 40 GHz erreicht, sondern übersteigt, in Kombination mit einem hohen Dynamikbereich, einem hohen Integrationsgrad und anderen für das MIMO-Radar relevanten Leistungen mit der Silizium-Germanium-Technologie (SiGe) nach dem Stand der Technik (SotA) erreicht werden kann. Darüber hinaus wurde gezeigt, dass mit speziellen IC-Topologien eine hohe Eingangsleistungsfestigkeit des RX-Frontends erreicht werden kann. Die untersuchten RX-Frontends können bis zu 25 dBm Eingangsleistung im Low-Band und bis zu 20 dBm Eingangsleistung im High-Band verkraften. Dies wurde durch die Verwendung der neuartigen Leistungsbegrenzertopologie und spezieller rauscharmer Wanderwellenverstärker erreicht. Diese Arbeit zeigte auch, dass mit der modernen SiGe-Technologie ein Dynamikbereich erreicht werden kann, der dem der wesentlich größeren Galliumarsenid (GaAs)-Technologie entspricht. Dies wurde durch die Untersuchung verschiedener rauscharmer Verstärker- und Mischertopologien erreicht. Bei den rauscharmen Verstärkern wurden gemeinsame Basis-, differenzielle verteilte und schaltbare Rückkopplungstopologien untersucht, um die beste Linearität und die niedrigste Rauschzahl zu erreichen. Bei den Mischern wurden Multi-Tanh-Triplet- und modifizierte Gilbert-Zellen-Topologien untersucht, um die Belastbarkeit zu erhöhen und die Rauschzahl zu minimieren. Die separaten RX- und TX-Komponenten wurden zu einzelnen RX- und TX-Front-End-ICs zusammengefügt und dann für die abschließende Verifizierung auf speziellen, kundenspezifisch implementierten Hochfrequenzplatinen integriert. Die abschließenden Messergebnisse bestätigten die Durchführbarkeit der Arbeit und belegen die Hypothese, dass eine Kombination aus ultrabreiter Bandbreite von 4 GHz bis 40 GHz, hoher Robustheit der Eingangsleistung, hohem Dynamikbereich und hohem Integrationsgrad in RX- und TX-Frontends für ein MIMO-Radar möglich ist. Diese Forschungsarbeit hat die Machbarkeit und das Potenzial für hochleistungsfähige, hochintegrierte und kosteneffiziente Bildgebungsradare bewiesen, die in Zukunft gebaut werden sollen. / This work presents an investigation and experimental analysis of Receiver (RX) and Transmitter (TX) front-ends tackling the MIMO radar related challenges in combination with an extreme 4 GHz to 40 GHz bandwidth, which was chosen to investigate the feasibility of implementing a highly re-configurable MIMO imaging radar system. The RX and TX front-ends were investigated to operate in a Frequency Modulated Continuous Wave (FMCW) mode, so the foreseen radar application can accurately measure objects at close range. Novel component topologies and implementation techniques were employed to investigate the feasibility of achieving ultra-wide bandwidth in combination with, high input power ruggedness, high dynamic range and high degree of integration. The experimental analysis of fabricated IC components demonstrated that extreme bandwidth, not just reaching, but exceeding 4 GHz to 40 GHz in combination with high dynamic range, high degree of integration and other MIMO radar relevant performances can be achieved with State of the Art (SotA) Silicon Germanium (SiGe) technology. Furthermore, it was demonstrated that with dedicated IC topologies, high input power ruggedness of the RX front-end can be achieved. The investigated RX front-ends could withstand up to 25 dBm input power in case of low band and up to 20 dBm of input power in case of high band respectively. This was achieved by using the novel power limiter topology and dedicated travelling wave low noise amplifier designs. This work also showed that with the state of the art SiGe technology the dynamic range equivalent to the much larger in size, Gallium Arsenide (GaAs) technologies is achievable. This was achieved by investigating various low noise amplifier and mixer topologies. In case of low noise amplifiers, common base, differential distributed and switchable feedback topologies were assessed to achieve the best linearity versus lowest noise figure performance. In case of the mixers, multi-tanh triplet and modified Gilbert cell topologies were investigated to increase the power handling and to minimize the noise figure. The separate RX and TX components were assembled into single RX and single TX front-end ICs and then integrated on dedicated, custom implemented Radio Frequency (RF) boards for the final verification. The final measurement results confirmed the feasibility of the work and prove the hypothesis of combining ultra-wide 4 GHz to 40 GHz bandwidth, high input power ruggedness, high dynamic range and high degree of integration in RX and TX front-ends dedicated for a MIMO radar. This research has proven the feasibility and potential for high performance, highly integrated and cost effective imaging radars to be built in future.