Year: 2026
Type: dissertation
Abstract: This work explores novel approaches for circuit design by leveraging new materials, devices and computing paradigms. Using as central theme the design of a flexible system capable of detecting and classifying insect species as either harmful or beneficial, three distinct Beyond Moore approaches are investigated: metal-oxide thin film transistor (TFT)-based circuit design, organic devices and neuromorphic computing. These address complementary aspects of the system from device-level optimization to bio-inspired data processing. Following preliminary studies to determine the most suitable insect features for classification, a flexible sensing structure and transduction technique are chosen and implemented. For this, a unipolar metal-oxide TFT technology is employed, and novel design strategies are introduced to mitigate its inherent limitations. Additionally, several RF communication blocks for the signal post-processing, including amplifiers, frequency mixers and voltage controlled oscillators (VCOs), are designed, implemented and characterized using TFTs. These blocks are monolithically integrated, demonstrating the technology’s maturity for having complex circuits that can operate in the open Industrial Scientific and Medical (ISM) 13.56 MHz band. As an alternative to metal-oxide TFTs, organic field effect transistors (OFETs) are investigated as potential candidates for active devices. A comprehensive study is conducted to compare different device structures, fabrication processes, materials and configurations, to assess their suitability for circuit design. This involves static characterization, s-parameter measurement and noise analysis. The latter provides insights into the device microstructure, allowing to study the semiconductor’s interface electrically, and to optimize the devices fabrication. The versatility of OFETs is further demonstrated by designing a circuit with devices tailored for high-voltage operation for insect pest deactivation. Building on the exploration of organic devices, a novel organic device presenting memcapacitive properties is introduced: the pinMOS. By means of an extensive electrical characterization, it is demonstrated that the pinMOS meets the required criteria for an artificial synapse. A model is proposed to understand its working mechanism, and optimal electrical parameters for neuromorphic computing are identified. Using this device, two different neural network architectures are evaluated: a multi-layer perceptron (MLP) and a reservoir computing (ReCo) system. Various crossbar structures are proposed for a pinMOS-based MLP, showing that a high accuracy in a benchmark handwritten number classification task can be achieved for multiple device per synapse. However, to address the challenge of requiring a large number of devices for the MLP implementation, ReCo is employed. With only six pinMOS devices, a system is implemented able to classify insect’s feeding behaviour with high accuracy, showcasing the device’s potential for time-signal classification in energy-constrained environments. Overall, this work contributes to the advancement of Beyond Moore circuit design, by demonstrating the potential of new materials, devices, and computing paradigms for the development of flexible and sustainable electronics.:Erklärung zur Autorenschaft III Acknowledgements V Abstract/Zusammenfassung VII Symbols and Abbreviations XI Previous Publications XVII Notation XIX 1. Introduction 1 1.1. Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2. Proposed System . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.3. Dissertation Structure . . . . . . . . . . . . . . . . . . . . . . 5 2. Metal-Oxide Devices 9 2.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 2.2. Technology Overview . . . . . . . . . . . . . . . . . . . . . . . 12 2.3. Charge-Based Capacitive Sensor . . . . . . . . . . . . . . . . 12 2.4. Voltage Controlled Oscillator . . . . . . . . . . . . . . . . . . 22 2.5. Mixers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 2.5.1. Ring Mixer . . . . . . . . . . . . . . . . . . . . . . . . 30 2.5.2. Differential Pair Active Mixer . . . . . . . . . . . . . . 34 2.5.3. Gilbert Cell . . . . . . . . . . . . . . . . . . . . . . . . 37 2.5.4. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . 41 2.6. Transmitter front-end . . . . . . . . . . . . . . . . . . . . . . 42 2.7. Digital Circuits . . . . . . . . . . . . . . . . . . . . . . . . . . 47 2.8. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 3. Organic Devices 57 3.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 3.2. Solution-sheared OFETs . . . . . . . . . . . . . . . . . . . . . 60 3.2.1. Device Structure . . . . . . . . . . . . . . . . . . . . . 60 3.2.2. Electrical Characterization . . . . . . . . . . . . . . . 61 3.2.3. Noise . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 3.2.4. Flicker Noise Analysis . . . . . . . . . . . . . . . . . . 68 3.3. Novel Approaches in OFETs . . . . . . . . . . . . . . . . . . 77 3.3.1. Device Structure . . . . . . . . . . . . . . . . . . . . . 78 3.3.2. Electrical Characterization . . . . . . . . . . . . . . . 80 3.4. High-voltage OFETs . . . . . . . . . . . . . . . . . . . . . . . 85 3.4.1. Device Structure . . . . . . . . . . . . . . . . . . . . . 85 3.4.2. Application Example . . . . . . . . . . . . . . . . . . . 86 3.5. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 4. Synaptic Devices 91 4.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 4.2. Mem-devices . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 4.3. The pinMOS . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 4.3.1. Operation . . . . . . . . . . . . . . . . . . . . . . . . . 97 4.3.2. Neuromorphic properties . . . . . . . . . . . . . . . . 101 4.4. Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 4.4.1. Rationale . . . . . . . . . . . . . . . . . . . . . . . . . 107 4.4.2. Charging Mechanism . . . . . . . . . . . . . . . . . . . 111 4.5. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 5. Neuromorphic Computing 117 5.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 5.2. Feed-forward Neural Network: Multi-layer Perceptron . . . . 119 5.3. Recursive Neural Network: Reservoir Computing . . . . . . . 128 5.4. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 6. Conclusion 141 6.1. Contributions to the State-of-the-Art . . . . . . . . . . . . . . 142 6.2. Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 A. Insect Characterization 147 A.1. Impedance sensing . . . . . . . . . . . . . . . . . . . . . . . . 147 A.2. Piezoelectric sensing . . . . . . . . . . . . . . . . . . . . . . . 151 Bibliography 153 Own Publications – Periodicals . . . . . . . . . . . . . . . . . . . . 153 Own Publications – Conference Proceedings . . . . . . . . . . . . . 154 Co-Authored Publications . . . . . . . . . . . . . . . . . . . . . . . 154 Other References . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
Author: Petrauskas, Lautaro Nicolás
Language: English
Subfield: Electrical and Electronic Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7225957446
