In the search for low power operation of microelectronic devices, spin-based solutions have attracted undeniable increasing interest due to their intrinsic magnetic nonvolatility. The ability to electrically manipulate the magnetic order using spin-orbit interaction, associated with the recent emergence of topological spintronics with its promise of highly efficient charge-to-spin conversion in solid state, offer alluring opportunities in terms of system design. Although the related technology is still at its infancy, this thesis intends to contribute to this engaging field by investigating the nature of the charge and spin transport in spin-orbit coupled and topological systems using quantum transport methods. We identified three promising building blocks for next-generation technology, three classes of systems that possibly enhance the spin and charge transport efficiency: (i)- topological insulators, (ii)- spin-orbit coupled magnonic systems, (iii)- topological magnetic textures (skyrmions and 3Q magnetic state). Chapter 2 reviews the basics and essential concepts used throughout the thesis: the spin-orbit coupling, the mathematical notion of topology and its importance in condensed matter physics, then topological magnetism and a zest of magnonics. In Chapter 3, we study the spin-orbit torques at the magnetized interfaces of 3D topological insulators. We demonstrated that their peculiar form, compared to other spin-orbit torques, have important repercussions in terms of magnetization reversal, charge pumping and anisotropic damping. In Chapter 4, we showed that the interplay between magnon current jm and magnetization m in homogeneous ferromagnets with Dzyaloshinskii-Moriya (DM) interaction, produces a field-like torque as well as a damping-like torque. These DM torques mediated by spin wave can tilt the imeaveraged magnetization direction and are similar to Rashba torques for electronic systems. Moreover, the DM torque is more efficient when magnons are thermally driven. Chapters 5 and 6 carry throughout tight-binding studies on the topological charge-spin transport in two-dimensional lattices with ferromagnetic skyrmions and 3Q magnetic structure. We use the Landauer-Buttiker formalism and evaluate the robustness of the topological signals. For the 3Q state, a spin-polarized quantum anomalous Hall state with chiral edge modes, unaffected by deformation and disorder, is reachable in zero net magnetization. We finish with concluding remarks and perspectives.

Comprendre la dynamique des parois magnétiques est essentiel pour le développement de technologies comme les mémoires magnétiques à haut densité. D'un point de vue fondamental, les parois de domaines peuvent être décrites comme des interfaces élastiques qui se déplacent dans un faible désordre d’ancrage. Leur dynamique, dite de reptation, présente des comportements universels qui sont caractérisés par des exposants critiques dépendant notamment de la dimensionnalité. Plus généralement, les comportements universels sont observés dans différents phénomènes aussi diverses que la propagation des fractures en solides, des fronts de combustion, des parois de domaines ferroélectriques... La première partie de la thèse propose une analyse les comportements universels de la transition de dépiégeage de parois de domaines sous champ magnétique. La dynamique de paroi a été étudiée sur une large gamme de températures, dans une couche ferromagnétique ultramince de Pt/Co/Pt. Nous avons comparé nos résultats avec ceux obtenus pour d’autres matériaux. Nous avons pu mettre en évidence la fonction universelle de la transition de dépiégeage qui rend compte des effets de champ magnétiques et des effets thermiques. La deuxième partie présente une étude des effets de taille finie sur les régimes de reptation. Nous avons mesuré les exposants critiques pour des couches de (Ga,Mn)(As,P) de différentes épaisseurs. Nous avons pu observer une discontinuité des exposants dits de rugosité et de reptation. Cette discontinuité est la signature d’un changement criticité de la dynamique qui est associé à une transition de dimensionnalité. Au-dessous d’un champ critique dépendant de l’épaisseur de la couche et de la température, une paroi se comporte comme une ligne élastique (d = 1) se déplaçant dans un milieu 2D. Au-dessus du champ critique, le mouvement de paroi correspond à celui d’une surface (d = 2) dans un milieu 3D. Dans la dernière partie, nous analysons la criticité du mouvement de paroi déplacé par courant électrique dans une couche mince de (Ga,Mn)(As,P). Nous étudions comment varie la dynamique de paroi avec l’angle entre la direction du courant et la normale à la paroi. Pour un courant perpendiculaire à la paroi, les exposants critiques de rugosité et de reptation mesurés sont similaires à ceux obtenus sous champ magnétique. Cela indique une compatibilité avec la classe d’universalité dite quenched Edward-Wilkinson. Pour un angle non-nul, la croissance de facettes révèle une compatibilité avec la classe d’universalité quenched Kardar-Parisi-Zhang négative. / Understanding magnetic domain walls dynamics (DW) is crucial in order to develop technological applications like high density memories in ferromagnets. From the fundamental point of view domain walls are described as interfaces moving in a weak pinning potential. Below the depinning threshold, DWs move in what it is known as the creep regime, which exhibit universal behaviors characterized by critical exponents who depend, among other things on the dimensionality and geometry of the interface. More generally, these universal behaviors are shared by different physical systems as diverse as propagation of fractures in solids, combustion fronts, ferroelectric domain walls... In the first part of the thesis, we address the universal behavior of the depinning transition in domain walls driven by magnetic field. For this purpose we measure the DW velocity as a function of magnetic field in an ultrathin Pt/Co/Pt film and then compare our results with other materials. We reveal a universal scaling function and obtain a consistent description for both the depinning transition and the thermally activated creep regime. In the second part of the manuscript, we study the sample size effects on the critical exponents within the creep regime. We use ferromagnetic (Ga, Mn)(As,P) films of different thicknesses. We observe a discontinuity in the roughness ζ and the creep µ exponents. This discontinuity evidences a dimensional crossover and a change in criticality in the quenched Edward-Wilkinson model. Below a certain critical field Hc, the DW behaves as an elastic line (d = 1) moving in a 2D medium. Above Hc the DW motion corresponds to an elastic interface (d = 2) moving in a 3D medium. In the last part, we compare the thermally activated creep dynamics in domain walls driven by magnetic field and by electric current separately in a (Ga,Mn)(As,P) thin film. We study the DW dynamical response with the angle between the current and the normal to the DW. When the angle between the current and the normal to the DW is sufficiently small, the critical exponents measured for current induced DW motion are very similar than for field induced DW motion. This result indicates agreement with the quenched Edward-Wilkinson universality class. When the angle between the DW and the current is not negligible, the DW faceting reveals compatibility with the quenched negative Kardar-Parisi-Zhang universality class.

L’objectif de cette thèse est la réalisation expérimentale de calcul bio-inspiré en utilisant la dynamique transitoire d’oscillateurs magnétique nanométriques.Pour bien des tâches telle que la reconnaissance vocale, le cerveau fonctionne bien plus efficacement en terme d’énergie qu’un ordinateur classique. Le développement de puces neuro-inspirées offre donc la perspective de surmonter les limitations des processeurs actuels et de gagner plusieurs ordres de grandeurs sur la consommation énergétique du traitement de données. L’efficacité du cerveau à traiter des données est due à son architecture, qui est particulièrement adaptée à la reconnaissance de motifs. Les briques de base de cette architecture sont les neurones biologiques. Ceux-ci peuvent être vus comme des oscillateurs non linéaires qui interagissent et génèrent des cascades spatiales d’activations en réponse à une excitation. Cependant le cerveau comprend cent milliards de neurones et le développement d’une puce neuro-inspiré requerrait des oscillateurs de très petite dimension. Les oscillateurs à transfert de spin (STNO) sont de taille nanométrique, ont une réponse rapide (de l’ordre de la nanoseconde), sont fortement non-linéaires et leur réponse dépendante du couple de transfert de spin est aisément ajustable (par exemple par l’application d’un courant continu ou d’un champ magnétique). Ils fonctionnent à température ambiante, ont un très faible bruit thermique, et sont compatible avec les technologies CMOS. Ces caractéristiques en font d’excellents candidats pour la réalisation de réseaux artificiels de neurones compatibles avec un ordinateur classique.Dans cette thèse, nous avons utilisé un unique STNO pour générer le comportement d’un réseau de neurones. Ainsi l’oscillateur joue à tour de rôle chaque neurone. Une cascade temporelle remplace donc la cascade spatiale d’un réseau de neurones biologiques. En particulier nous avons utilisé la relaxation et la dépendance non-linéaire de l’amplitude des oscillations afin de réaliser du calcul neuromorphique. L’un des résultats principaux de cette thèse est la réalisation de reconnaissance vocale (reconnaissance de chiffres dits par 5 locuteurs différents) en obtenant un taux de reconnaissance à l’état de l’art de 99.6%. Nous avons pu montrer que les performances de la reconnaissance sont étroitement dépendantes des propriétés physiques du STNO tel que l’évolution de la largeur de raie, la puissance d’émission, ou la fréquence d’émission. Nous avons donc optimisé les conditions expérimentales (champs magnétiques et courant continu appliqués, fréquence du signal à traiter) afin de pouvoir utiliser au mieux les propriétés physiques du STNO pour la reconnaissance.  Les signaux vocaux requièrent d’être transformés du domaine temporel au domaine fréquentiel, avant de pouvoir être traités, et cette étape est réalisée numériquement en amont de l’expérience. Nous avons étudié l’influence de différents prétraitements sur la reconnaissance et mis en évidence le rôle majeur de la non-linéarité de ces derniers. Enfin, afin de pouvoir traiter des problèmes requérant de la mémoire, tel que par exemple des signaux sous forme de séquences temporelles, nous avons mesuré la mémoire que possède intrinsèquement un STNO, du fait de sa relaxation. Nous avons aussi augmenté cette mémoire à l’aide d’une boucle à retard. Ce dispositif a permis d’accroître la plage de mémoire de quelques centaines de nanosecondes à plus d’une dizaine de microsecondes. L’ajout de cette mémoire extrinsèque a permis de supprimer jusqu’à 99% des erreurs sur une tâche de reconnaissance de motifs temporels (reconnaissance de signaux sinusoïdaux et carrés). / This thesis studies experimentally the transient dynamics of magnetic nano-oscillators for brain-inspired computing.For pattern recognition tasks such as speech or visual recognition, the brain is much more energy efficient than classical computers. Developing brain-inspired chips opens the path to overcome the limitations of present processors and to win several orders of magnitude in the energy consumption of data processing. The efficiency of the brain originates from its architecture particularly well adapted for pattern recognition. The building blocks of this architecture are the biological neurons, which can be seen as interacting non-linear oscillators generating spatial chain reactions of activations. Nevertheless, the brain has one hundred billion neurons and a brain-inspired chip would require extremely small dimension oscillators. The spin-transfer torque oscillators (STNO) have nanometric size, they are fast (nanosecond time-scales), highly non-linear and their spin-torque dependent response is easily tunable (for instance by applying an external magnetic field or a d.c. current). They work at room temperature, they have a low thermal noise and they are compatible with CMOS technologies. Because of these features, they are excellent candidates for building hardware neural networks, which are compatible with the standard computers.In this thesis, we used a single STNO to emulate the behavior of a whole neural network. In this time multiplexed approach, the oscillator emulates sequentially each neuron and a temporal chain reaction replace the spatial chain reaction of a biological neural network. In particular, we used the relaxation and the non-linear dependence of the oscillation amplitude with the applied current to perform neuromorphic computing. One of the main results of this thesis is the demonstration of speech recognition (digits said by different speakers) with a state-of-the-art recognition rate of 99.6%. We show that the recognition performance is highly dependent on the physical properties of the STNO, such as the linewidth, the emission power or the frequency. We thus optimized the experimental bias conditions (external applied magnetic field, d.c. current and rate of the input) in order to leverage adequately the physical properties of the STNO for recognition. Voice waveforms require a time-to-frequency transformation before being processed, and this step is performed numerically before the experiment. We studied the influence of different time-to-frequency transformations on the final recognition rate, shading light on the critical role of their non-linear behavior. Finally, in order to solve problems requiring memory, such as temporal sequence analysis, we measured the intrinsic memory of a STNO, which comes from the relaxation of the oscillation amplitude. We also increased this memory, using a delayed feedback loop. This feedback improved the range of memory from a few hundreds of nanoseconds to more than ten microseconds. This feedback memory allows suppressing up to 99% of the errors on a temporal pattern recognition task (discrimination of sine and square waveforms).

Les oscillateurs à transfert de spin se distinguent des autres oscillateurs électroniqueshyperfréquences notamment par leurs grandes non-linéarité et agilité en fréquence.Cependant quoi que les principes fondamentaux de ces systèmes soient bien compris,leurs performances en termes de puissance de sortie et de largeur de raie en limitent lesapplications au stade de prototype. Pour y remédier, une des stratégies est celle de lasynchronisation mutuelle, qui devrait améliorer les caractéristiques de ces systèmes enaugmentant le volume oscillant, mais aussi permettre la réalisation de structures pluscomplexes. Bien que le mécanisme fondamental a été démontré, toutes ses implicationsne sont pas encore parfaitement comprises. C’est dans cette perspective que nous étudionsle réseau minimal d’oscillateurs non-linéaires constitué par une paire d’oscillateursà transfert de spin mutuellement couplés via leur rayonnement dipolaire. L’originalitéde ce travail réside dans l’introduction d’une antenne, qui peut générer un signal hyperfréquence et agir comme troisième oscillateur "idéal" pour explorer la riche dynamique du système, qui présente des intérêts aussi bien fondamentaux qu’applicatifs. / Spin torque oscillators have driven interest among other electronic microwaveoscillators notably for their high nonlinearity and agility. However although the fundamentalprinciples of those systems are well-understood, these are limited to the realisationof prototypes due to their poor performances in terms of emitted power and linewidth.One strategy to deal with those limitations consists in mutually synchronising several suchoscillators, which would increase the oscillating volume, thereby improving these characteristics and allowing the realisation of more complex structures. Despite the fact thatthe fundamental principle has been demonstrated, its implication are still far from beingperfectly understood. In this perspective we study the minimalistic network consisting ina pair of spin torque oscillators mutually coupled via their magneto-dipolar interaction.The originality of this work lies in the introduction of a microstrip antenna, which enablesthe generation a microwave signal and acts as a third "ideal" oscillator to probe the richdynamics of this system, which displays fundamental as well as applicative interests.

This work is dedicated to the study of two antiferromagnetic materials that are suitable for use in spintronic devices. In series of FeRh samples we studied the transition temperature between the antiferromagnetic and ferromagnetic phases. We developed a method based on material optical response for a quick determination of this temperature, which enabled us to study with a spatial resolution of 1 μm a magnetic inhomogeneity of prepared samples.We also developed a method for a determination of the Néel temperature and the magnetization easy axis position in thin films prepared from compensated antiferromagnetic metal. We successfully applied this method on an uniaxial sample of CuMnAs and we discussed its applicability for a research of samples with a biaxial magnetic anisotropy.

This thesis is focused on two open problems of antiferromagnetic (AFM) spintronics: manipulation of AFM coupled moments and development of new materials combining AFM and semiconductor properties. We present three particular methods enabling AFM moments manipulation. The rst method, based on the exchange spring effect in an AFM/FM double layer, strongly de- pends on the AFM layer thickness and temperature. We systematically vary these two parameters and identify the conditions when AFM moments can be manip- ulated. By the second method, cooling an AFM in a magnetic eld through the critical temperature, we prove the concept of a fully AFM-based (containing no FM) spintronic device. The last studied method is based on current induced effects in nanostructures containing an AFM. By systematic study of samples with and without AFM we demonstrate the ability of AFM moments to absorb a current induced torque. Relying neither on a FM nor on cooling in magnetic eld, this method represents an elegant way of AFM moments manipulation. In the second experimental part new materials for AFM spintronics are discussed, and one representative example, CuMnAs, is studied in detail. Characterization of bulk and epitaxial CuMnAs is presented and rst spintronic functionality is shown. Powered by TCPDF (www.tcpdf.org)

One of the most remarkable achievements of modern condensed matter physics is the discovery of topological phases of matter. Materials in a non-trivial topological phase or the topological insulators can be distinguished by their unique electronic and transport properties which are indifferent to different types of perturbations and thus open new routes towards the dissipationless transport. Explaining their properties requires proper involvement of relativistic approach as well as topological analysis. Among different classes of topological insulators, the Z2 topological insulators have drawn special attention due to their strong spin-orbit coupling which makes them a promising candidate for spintronics application, especially for magnetic memory devices. Due to their inherent strong spin-orbit coupling, they provide an efficient way to manipulate electronic spin with an applied electric field via spin orbit torque. The topological insulators have been found to be far more superior in manipulating the magnetic order parameter of a ferromagnet compared to the conventional heavy metals like platinum or tantalum. Another milestone in magnetic memory devices is marked by the introduction of antiferromagnetic memory devices which has not drawn any attention for long time as they cannot be controlled by an applied magnetic field. Recently it has been found that in case of a non-centrosymmetric antiferromagnet, the magnetic order parameter can be manipulated by with spin-orbit torque which also have been verified experimentally. The advantages of antiferromagnetic devices over ferromagnetic devices are that they allow faster switching speed and they are immune to an external magneticfield which are two highly solicited properties for next generation spintronic devices. This thesis is focused on understanding the transport properties in topologically nontrivial materials and their interface with different magnetic material. We use simplified continuum model as well as tight binding models to capture the salient features of these systems. Using non-equilibrium Green's function we explore their transport properties as well as spin-charge conversion mechanism. Our finding would provide a better understanding of these new class of materials and thus would be instrumental to discover new mechanisms to manipulate their properties.

Energy enables basic and innovative services to reach a seemingly ever-growing population and when its generation costs are reduced or when its usage is optimized it has the greatest impact on the reduction of poverty. Furthermore, there is a pressing need to decouple energy generation from non-renewable and carbon-heavy sources which has led mayor economies to increase research efforts in these areas. This thesis discusses research on water oxidation using nanostructured iron oxide electrodes and current-induced magnetic domain wall motion in nickel/cobalt bi-segmented nanowires. These two fields may seem disparate at first glance, but are linked by such common theme: materials for energy, and more precisely, materials for energy conversion and economy. The work presented in this document aims also to reflect this theme by using widely available materials like iron and aluminum, and optimizing the methods to produce the final samples using the least resources possible. All samples were prepared by electroplating metals (iron, cobalt and nickel) into anodized alumina templates fabricated inhouse. For water oxidation, iron nanorods were integrated into an electrode and annealed in air, while nickel/cobalt nanowires were isolated and contacted individually to test for spintronics-related effects. Spintronic-based devices aim to reduce energy usage in nowadays microelectronic devices. The nanostructured iron oxide electrode showed its usefulness for water oxidation in a laboratory environment, making it an appropriate complement to other electrodes specially designed for water reduction in a photoelectrochemical cell. This two-electrode design, allows for hydrogen and oxygen to be produced at each electrode and therefore eases their separate collection for, e.g., fuel or fertilizers. On the other hand, this work presents one of the first experimental demonstration of current-induced domain wall motion in soft/hard cylindrical magnetic nanowires at zero applied external magnetic field. These kinds of experiments are expected to be the first of many which will allow researchers in the field to test for spintronic-relevant properties and interactions in cylindrical magnetic nanowires.

京都大学 / 0048 / 新制・課程博士 / 博士(工学) / 甲第21109号 / 工博第4473号 / 新制||工||1695(附属図書館) / 京都大学大学院工学研究科電子工学専攻 / (主査)教授 白石 誠司, 教授 木本 恒暢, 教授 山田 啓文 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

付記する学位プログラム名: 京都大学卓越大学院プログラム「先端光・電子デバイス創成学」 / 京都大学 / 新制・課程博士 / 博士(理学) / 甲第24396号 / 理博第4895号 / 新制||理||1699(附属図書館) / 京都大学大学院理学研究科物理学・宇宙物理学専攻 / (主査)教授 柳瀬 陽一, 教授 石田 憲二, 准教授 池田 隆介 / 学位規則第4条第1項該当 / Doctor of Science / Kyoto University / DGAM