Les architectures d’ordinateur classiques sont optimisées pour le traitement déterministe d’informations pré-formatées et ont donc des difficultés avec des données naturelles bruitées (images, sons, etc.). Comme celles-ci deviennent nombreuses, de nouveaux circuits neuromorphiques (inspirés par le cerveau) tels que les réseaux de neurones émergent. Des nano-dispositifs, appelés memristors, pourraient permettre leur implémentation sur puce avec une haute efficacité énergétique et en s’approchant de la haute connectivité synaptique du cerveau.Dans ce travail, nous étudions des memristors basés sur des jonctions tunnel ferroélectriques qui sont composées d’une couche ferroélectrique ultramince entre deux électrodes métalliques. Nous montrons que le renversement de la polarisation de BiFeO3 induit des changements de résistance de quatre ordres de grandeurs et établissons un lien direct entre les états de domaines mixtes et les niveaux de résistance intermédiaires.En alternant les matériaux des électrodes, nous révélons leur influence sur la barrière électrostatique et les propriétés dynamiques des memristors. Des expériences d’impulsion unique de tension montrent un retournement de polarisation ultra-rapide. Nous approfondissons l’étude de cette dynamique par des mesures d’impulsions cumulées. La combinaison de leur analyse avec de l’imagerie par microscopie à force piézoélectrique nous permet d’établir un modèle dynamique du memristor. Suite à la démonstration de la spike-timing-dependent plasticity, une règle d’apprentissage importante, nous pouvons prédire le comportement de notre synapse artificielle. Ceci représente une avance majeure vers la réalisation de réseaux de neurones sur puce dotés d’un auto-apprentissage non-supervisé. / Classical computer architectures are optimized to process pre-formatted information in a deterministic way and therefore struggle to treat unorganized natural data (images, sounds, etc.). As these become more and more important, the brain inspires new, neuromorphic computer circuits such as neural networks. Their energy-efficient hardware implementations will greatly benefit from nanodevices, called memristors, whose small size could enable the high synaptic connectivity degree observed in the brain.In this work, we concentrate on memristors based on ferroelectric tunnel junctions that are composed of an ultrathin ferroelectric film between two metallic electrodes. We show that the polarization reversal in BiFeO3 films can induce resistance contrasts as high as 10^4 and how mixed domain states are connected to intermediate resistance levels.Changing the electrode materials provides insights into their influence on the electrostatic barrier and dynamic properties of these memristors. Single-shot switching experiments reveal very fast polarization switching which we further investigate in cumulative measurements. Their analysis in combination with piezoresponse force microscopy finally allows us to establish a model describing the memristor dynamics under arbitrary voltage signals. After the demonstration of an important learning rule for neural networks, called spike-timing-dependent plasticity, we successfully predict new, previously unexplored learning curves. This constitutes an important step towards the realization of unsupervised self-learning hardware neural networks.

Der Elektronentransport durch Nanostrukturen bietet eine Perspektive auf interessante Anwendungen und neue Einsichten in die Nichtgleichgewichtsdynamik von Elektronen in komplexen Umgebungen. Quantendotsysteme erlauben im Speziellen ein hohes Maß an Kontrolle ihrer Eigenschaften und ermöglichen damit detaillierte Untersuchungen. Das wachsende Interesse an zeitaufgelöstem Elektronentransport in diesen Systemen erklärt sich vor allem durch die rasanten Fortschritte bei der experimentellen Realisierung von pulsinduziertem Transport. Zur Beschreibung und Interpretation dieser Experimente bedarf es der Entwicklung neuer theoretischer Zugänge und Berechnungsverfahren. In dieser Arbeit werden zwei Propagationsmethoden zur numerischen Beschreibung von zeitaufgelöstem Elektronentransport entwickelt. Hierbei wird einerseits von einer Einteilchenbeschreibung mit Nichtgleichgewichts-Green-Funktionen (NEGF) und andererseits von einer Vielteilchenbeschreibung, basierend auf verallgemeinerten Quantenmastergleichungen für die reduzierte Vielteilchendichtematrix, ausgegangen. Das Konzept ist in beiden Fällen ähnlich: Im ersten Schritt der Herleitung werden Hilfsgrößen eingeführt und gleichberechtigt zum reduzierten Zustand des Systems behandelt. Eine Hilfsmodenentwicklung der Fermi-Funktion ermöglicht im zweiten Schritt die numerische Berechnung mit den hergeleiteten Bewegungsgleichungen. Mit Hilfe einer Partialbruchzerlegung wird eine Entwicklung der Fermi-Funktion abgeleitet, die sich durch eine wesentlich verbesserte Konvergenz gegenüber bisher bekannten Entwicklungen auszeichnet. Diese Zerlegung erweist sich für die Propagation als effizienter Zugang und kann darüber hinaus bei Berechnungen zur Elektronenstruktur angewendet werden. Obwohl der NEGF-Formalismus eines der Standardverfahren für die Behandlung von Transportdynamik in Nanostrukturen darstellt, ist die Auswahl an numerischen Implementierungen verschwindend gering. Die in dieser Arbeit entwickelte Propagationsmethode stellt eine neue Herangehensweise dar, die im Vergleich zu den bisherigen Zugängen ein günstigeres Skalierungsverhalten aufweist. Anhand von zwei Beispielen wird demonstriert, dass die Methode sowohl auf stochastisch getriebene Systeme als auch auf Situationen mit realistischen Spannungspulsen anwendbar ist. Eine Erweiterung auf wechselwirkende Elektronen wird ausgehend von der Methode der Bewegungsgleichungen abgeleitet. Im Rahmen der Vielteilchenbeschreibung durch die verallgemeinerten Quantenmastergleichungen wird insbesondere der Einfluss von Termen höherer Ordnung untersucht. Hierzu wird, neben der üblichen Quantenmastergleichung zweiter Ordnung, explizit die vierte Ordnung berechnet. Ein Vergleich mit dem NEGF-Formalismus zeigt die Notwendigkeit höhere Ordnungen, zumindest partiell, zu berücksichtigen, da erst hierdurch die Verbreiterung der Energieniveaus aufgrund der Tunnelkopplung an die Reservoirs konsistent beschrieben wird. Dieser Befund wird am Beispiel des stationären und transienten Elektronentransports durch einen Doppelquantendot untermauert. Auf der Basis von numerischen Berechnungen und einem analytisch lösbaren Modell werden die Resultate eines aktuellen Pump-Probe-Experiments zur kohärenten Kontrolle von Ladungs-Qubits in Doppelquantendots interpretiert. Die Anwendungsmöglichkeiten der entwickelten Propagationsmethoden gehen weit über die in der Arbeit betrachteten Beispiele hinaus. Sie erlauben die Beschreibung von neuartigen Transportkonzepten und ermöglichen einen erweiterten Einblick in die Nichtgleichgewichtsdynamik von Elektronen in Nanostrukturen.

Atomically thin 2D layered semiconductor materials such as Transition Metal Di-Chalcogenides (TMDCs) have great potential for use as flexible, ultra-thin photovoltaic materials in solar cells due to their favorable photon absorption and electronic transport properties. In this dissertation, the electronic properties, such as band structure and bandgap, and optical absorption properties of a TMDC known as Tungsten Disulfide (WS2) were obtained from Density Functional Theory (DFT) calculations to design conventional and unconventional solar cells. Using these properties, a 1 μm thick heterojunction solar cell based on monolayer and bulk WS2 together with amorphous silicon (a-Si) was modeled using numerical calculations and simulations. The maximum efficiency of this cell is 23.3% with Voc = 0.84 V and Jsc = 33.5 mA/cm2 under the AM1.5G terrestrial solar spectrum. Next, a similar but even thinner solar cell with a thickness of 200 nm, together with a light trapping structure and an anti-reflection coating layer, was modeled under the AM0 space solar spectrum; similar device performance efficiencies around 21-23% were obtained. The performance of these solar cell models is comparable to many commercial cells in both terrestrial and space photovoltaics. As conventional photovoltaics approach the Shockley-Queisser limit, the need for unconventional materials and approaches has become more apparent. Hybrid alloys of TMDCs exhibit tunable direct bandgaps and significant dipole moments. Dark state protection induced by dipole-dipole interactions forms new bright and dark states in the conduction band that reduce radiative recombination and enhance photon-to-electron conversion, leading to significantly higher photocurrents. In our work, current enhancement of up to 35% has been demonstrated by modeling dark state protection in a solar cell composed of Tungsten Diselenide (WSe2) and Tungsten Sulfo-Selenide (WSeS), with the potential to exceed the Shockley-Queisser limit under ideal conditions.

<p>Copper (Cu) has been used as the main conductor in interconnects due to its low resistivity. However, because of its high diffusivity, diffusion barriers/liners (tantalum nitride/tantalum; TaN/Ta) must be incorporated to surround Cu wires. Otherwise, Cu ions/atoms will drift/diffuse through the inter-metal dielectric (IMD) that separates two distinct interconnects, resulting in circuit shorting and chip failures. The scaling limit of conventional Cu diffusion barriers/liners has become the bottleneck for interconnect technology, which in turn limits the IC performance. The interconnect half-pitch size will reach ~20 nm in the coming sub-5 nm technology nodes. Meanwhile, the TaN/Ta (barrier/liner) bilayer stack has to be > 4 nm to ensure acceptable liner and diffusion barrier properties. Since TaN/Ta occupy a significant portion of the interconnect cross-section and they are much more resistive than Cu, the effective conductance of an ultra-scaled interconnect will be compromised by the thick bilayer. Therefore, two dimensional (2D) layered materials have been explored as diffusion barrier alternatives owing to their atomically thin body thicknesses. However, many of the proposed 2D barriers are prepared at too high temperatures to be compatible with the back-end-of-line (BEOL) technology. In addition, as important as the diffusion barrier properties, the liner properties of 2D materials must be evaluated, which has not yet been pursued. </p> The objective of the thesis is to develop a 2D barrier/liner that overcomes the issues mentioned. Therefore, we first visit various 2D layered materials to understand their fundamental capability as barrier candidates through theoretical calculations. Among the candidates, hexagonal-boron-nitride (h-BN) and molybdenum disulfide (MoS₂) are selected for experimental studies. In addition to studying their fundamental properties to know their potential, we have also developed techniques that can realize low-temperature-grown 2D layered materials. Metal-organic chemical vapor deposition (MOCVD) is adopted for the synthesis of BEOL-compatible MoS₂. The electrical test results demonstrate the promises of integrating 2D layered materials to the state-of-the-art interconnect technology. Furthermore, by considering not only diffusion barrier properties but also liner properties, we develop another 2D layered material, tantalum sulfide (TaS_x), using plasma-enhanced chemical vapor deposition (PECVD). The TaS_x is promising in both barrier and liner aspects and is BEOL-compatible. Therefore, we believed that the conventional TaN/Ta bilayer stack can be replaced with an ultra-thin TaS_x layer to maximize the Cu volume for ultra-scaled interconnects and improve the performance. Furthermore, Since via resistance has become the bottleneck for overall interconnect performance, we study the vertical conduction of TaS_x. Both the intrinsic and extrinsic properties of this material are investigated and engineering approaches to improve the vertical conduction are also tested. Finally, we explore the possibilities of benefiting from 2D materials in other applications and propose directions for future studies.

Reconfigurable computing architectures combine the flexibility of software with the performance of custom hardware. Such architectures are of particular interest at the nanoscale level. We argue that a bottom-up self-assembled fabric of nodes will be easier and cheaper to manufacture, however, one has to make compromises with regards to the device regularity, homogeneity, and reliability. The goal of this thesis is to evaluate the performance and cost of a self-configurable computing architecture composed of simple reconfigurable nodes for unstructured and unknown fabrics. We built a software and hardware framework for this purpose. The framework enables creating an irregular network of compute nodes where each node can be configured as a simple 2-input, 4-bit logic gate. The compute nodes are organized hierarchically by sending a packet through a top anchor node that recruits compute nodes with a chemically-inspired algorithm. The nodes are then self-configured by means of a gate-level netlist describing any digital logic circuit. A topology-agnostic optimization algorithm inspired by simulated annealing is then initiated to self-optimize the circuit for latency. Latency comparisons between non-optimized, brute-force optimized and our optimization algorithm are made. We further implement the architecture in VHDL and evaluate hardware cost, area, and energy consumption. The simple on-chip topology-agnostic optimization algorithm we propose results in a significant (up to 50\%) performance improvement compared to the non-optimized circuits. Our findings are of particular interest for emerging nano and molecular-scale circuits.

The exceptional self-assembly properties of DNA as well as its ability to interact with different kinds of chemical compounds and biological structures make this biomolecule to an interesting object for the fabrication of artificial nanostructures. In this work several methods for a DNA-based self-assembly of electronic nanocircuitry are explored. For this, four basic steps, which turned out to be essential within a circuit assembly process, are addressed: (i) The formation of multi-branched DNA junctions by a simple building-block procedure. (ii) The site-specific attachment of nanoobjects (gold colloids) at the center of DNA junctions. (iii) The integration of DNA into microstructured gold electrode arrays, in particular the stretching of single DNA molecules between two electrodes. For this a simple, but reliable methods for the functionalization of gold electrodes by using aminoethanethiol was developed, which enables end-specific attachment of the DNA but does not require DNA modification. (iv) The metallization of DNA. A synthesis procedure was developed, which results in the formation of continuous chains of 5nm platinum clusters along the DNA. The metal deposition process turned out to take place exclusively at the DNA while background metallization is completely suppressed.

Indiana University-Purdue University Indianapolis (IUPUI) / Nanoscale materials have become an active area of research due to the enhanced mechanical properties of the nanomaterials in comparison to their respective bulk materials. The effect that the size and shape of a nanomaterial has on its mechanical properties is important to understand if these materials are to be used in engineering applications. This thesis presents the results of molecular dynamics (MD) simulations on copper, gold, nickel, palladium, platinum, and silver nanowires of three cross-sectional shapes and four diameters. The cross-sectional shapes investigated were square, circular, and octagonal while the diameters varied from one to eight nanometers. Due to a high surface area to volume ratio, nanowires do not have the same atomic spacing as bulk materials. To account for this difference, prior to tensile loading, a minimization procedure was applied to find the equilibrium strain for each structure size and shape. Through visualization of the atomic energy before and after minimization, it was found that there are more than two energetically distinct areas within the nanowires. In addition, a correlation between the anisotropy of a material and its equilibrium strain was found. The wires were then subjected to a uniaxial tensile load in the [100] direction at a strain rate of 108 s-1 with a simulation temperature of 300 K. The embedded-atom method (EAM) was employed using the Foiles potential to simulate the stretching of the wires. The wires were stretched to failure, and the corresponding stress-strain curves were produced. From these curves, mechanical properties including the elastic modulus, yield stress and strain, and ultimate strain were calculated. In addition to the MD approach, an energy method was applied to calculate the elastic modulus of each nanowire through exponential fitting of an energy function. Both methods used to calculate Young’s modulus qualitatively gave similar results indicating that as diameter decreases, Young’s modulus decreases. The MD simulations were also visualized to investigate the deformation and yield behavior of each nanowire. Through the visualization, most nanowires were found to yield and fail through partial dislocation nucleation and propagation leading to {111} slip. However, the 5 nm diameter octagonal platinum nanowire was found to yield through reconstruction of the {011} surfaces into the more energetically favorable {021} surfaces.

yes / This is the tenth workshop to be organised under the postgraduate programmes in electrical and electronic engineering (EEE). Over the past ten years, the Research Seminar Series has provided a snapshot of the research agenda. Early Proceedings addressed issues such as third-generation (3G) mobile and GPS satellite navigation, while in this issue, the importance of the green agenda and the influence of broadband mobile communications, smartphones and the World Wide Web are in evidence. In total, forty-five papers have been selected for the Proceedings.

Hocheffiziente Solarzellen beruhen auf der exzellenten Oberflächenpassivierung, die minimale Rekombinationsverluste gewährleistet. Innerhalb des letzten Jahrzehnts wurde Al2O3 in der Photovoltaikindustrie zum bevorzugten Material für p-leitendes Si. Unterschiedliche Abscheidetechnologien erreichten Passivierungen mit effektiven Minoritätsladungsträgerlebensdauern nahe der AUGER–Grenze. Die ausgezeichnete Passivierungswirkung von Al2O3wird zwei Effekten zugeschrieben: Einerseits werden Si−SiO2-grenzflächennahe Rekombinationszentren passiviert, wenn Wasserstoff, beispielsweise aus der Al2O3-Schicht, offene Bindungen absättigt. Bedingt durch die hohe Konzentration intrinsischer negativer Ladungen an der SiO2-Grenzfläche weist Al2O3 andererseits einen charakteristischen Feldeffekt auf. Das resultierende elektrische Feld hält Elektronen von Oberflächenrekombinationszentren fern. Negative Ladungen im Al2O3 werden generell als fest bezeichnet. Allerdings hat Al2O3 zusätzlich eine hohe Dichte an Haftstellen, die von Elektronen besetzt werden können. Die Dichte negativer Ladungen im Al2O3-Passivierungsschichten hängt vom elektrischen Feld und der Bestrahlungsintensität ab. Ziel dieser Arbeit ist die systematische Untersuchung dielektrischer Passivierungsschichtstapel für die Anwendung auf Si-Solarzellen. Der Qualität und Dicke der SiO2-Grenzschicht kommt in diesem Kontext eine besondere Rolle zu, da sie Ladungsträgertunneln ermöglicht. Der Elektronentransport ist eine Funktion der Oxiddicke und das Optimum zwischen Ladungseinfang und -haltung liegt bei etwa 2 nm SiO2. Vier relevante Al2O3-Abscheidetechnologien werden untersucht: Atomlagenabscheidung, Kathodenzerstäubung, Sprühpyrolyse und Rotationsbeschichtung, wobei die erstgenannte dominiert. Es werden Nanolaminate verglichen, die aus Al2O3 und TiO2, HfO2 oder SiO2 mit subnanometerdicken Zwischenschichten bestehen. Während letztgenannte die Oberflächenrekombination nicht vermindern, beeinflussen TiO2- und HfO2-Nanolaminate die Passivierungswirkung. Ein dynamisches Wachstumsmodell, das initiale und stationäre Wachstumsraten der beteiligten Metalloxide berücksichtigt, beschreibt die physikalischen Parameter. Schichtsysteme mit 0,2 % TiO2 oder 7 % HfO2 sind konventionellen Al2O3-Schichten überlegen. In beiden Fällen überwiegt die veränderte Feldeffekt- der chemischen Passivierung, die mit einer Grenzflächenzustandsdichte von maximal 5·1010 eV−1·cm−2 unverändert auf hohem Niveau verbleibt. Die Ladungsdichte beider Schichtsysteme wird entweder über die Änderung ihrer Polarität der festen Ladungen oder der Fähigkeit zum Ladungseinfang bestimmt. Das Tunneln von Elektronen wird durch ein mathematisches Modell erklärt, dass eine bewegliche Ladungsfront innerhalb der Oxidschicht beschreibt. / High-efficiency solar cells rely on excellent passivation of the surface to ensure minimal recombination losses. In the last decade, Al2O3 became the material of choice for p-type Si in the photovoltaic industry. A remarkable surface passivation with effective minority carrier lifetimes close to the AUGER–limit was demonstrated with different deposition techniques. The excellent passivation effect of Al2O3 is attributed to two effects: Firstly, recombination centers at the Si−SiO2 interface get chemically passivated when hydrogen, for instance from the Al2O3 layer, saturates dangling bonds. Secondly, Al2O3 presents an outstanding level of field effect passivation due to its high concentration of intrinsic negative charges close to the SiO2 interface. The generated electrical field effectively repels electrons from surface recombination centers. Negative charges in Al2O3 are generally termed fixed charges. However, Al2O3 incorporates a high density of trap sites, too, that can be occupied by electrons. It was shown that the negative charge density in Al2O3 passivation layers depends on the electrical field and on the illumination intensity. The goal of this work is to systematically investigate dielectric passivation layer stacks for application on Si solar cells. The SiO2 interface quality and thickness plays a major role in this context, enabling or inhibiting carrier tunneling. Since the electron transport is a function of the oxide thickness, the balance between charge trapping and retention is achieved with approximately 2 nm of SiO2. Additionally, four relevant Al2O3 deposition techniques are compared: atomic layer deposition, sputtering, spray pyrolysis and spin–on coating, whereas the former is predominant. Using its flexibility, laminates comprising of Al2O3 and TiO2, HfO2 or SiO2 with subnanometer layers are compared. Although the latter do not show decreased surface recombination, nanolaminates with TiO2 and HfO2 contribute to the passivation. Their physical properties are described with a dynamic growth model that considers initial and steady–state growth rates for the involved metal oxides. Thin films with 0.2 % TiO2 or 7 % HfO2 are superior to conventional Al2O3 layers. In both cases, the modification of the field effect prevails the chemical effect, that is, however, virtually unchanged on a very high level with a density of interface traps of 5·1010 eV−1·cm−2 and below. The density of charges in both systems is changed via modifying either the polarity of intrinsic fixed charges or the ability of trapping charges within the layers. The observations of electron tunneling are explained by means of a mathematical model, describing a charging front, which moves through the dielectric layer.

Electron Beam Induced Deposition (EBID) is an emerging additive nanomanufacturing tool which enables growth of complex 3-D parts from a variety of materials with nanoscale resolution. Fundamentals of EBID and its application to making a robust, low-contact-resistance electromechanical junction between a Multiwall Carbon Nanotube (MWNT) and a metal electrode are investigated in this thesis research. MWNTs are promising candidates for next generation electrical and electronic devices, and one of the main challenges in MWNT utilization is a high intrinsic contact resistance of the MWNT-metal electrode junction interface. EBID of an amorphous carbon interface has previously been demonstrated to simultaneously lower the electrical contact resistance and to improve mechanical characteristics of the MWNT-electrode junction. In this work, factors contributing to the EBID formation of the carbon joint between a MWNT and an electrode are systematically explored via complimentary experimental and theoretical investigations. A comprehensive dynamic model of EBID using residual hydrocarbons as a precursor molecule is developed by coupling the precursor mass transport, electron transport and scattering, and surface deposition reaction. The model is validated by comparison with experiments and is used to identify different EBID growth regimes and the growth rates and shapes of EBID deposits for each regime. In addition, the impact of MWNT properties, the electron beam impingement location and energy on the EBID-made carbon joint between the MWNT and the metal electrode is critically evaluated. Lastly, the dominant factors contributing to the overall electrical resistance of the MWNT-based electrical interconnect and relative importance of the mechanical contact area of the EBID-made carbon joint to MWNT vs. that to the metal electrode are determined using carefully designed experiments.