L'objectif de cette thèse était de développer un magnétomètre à force de Faraday pour mesurer des valeurs absolues de l'aimantation à très basse température (30 mK) et fort champ magnétique (16 T) avec une haute sensibilité (10-5 emu). Ce magnétomètre sera principalement utilisé pour sonder les propriétés induites sous champ dans les systèmes magnétiques frustrés.Dans une première partie, je détaille le développement de ce magnétomètre. Cela comprend l'optimisation du réfrigérateur à dilution et de bobines de gradient de champ ainsi que le développement d'une détection capacitive avec un étage amplificateur à froid pour améliorer la sensibilité de la mesure. Les mesures magnétiques préliminaires sont présentées. Les changements nécessaires pour rendre le magnétomètre opérationnel sont discutés.Dans une seconde partie, je présente l'étude expérimentale de systèmes magnétiques frustrés: des grenats de Gadolinium, décrits par des spins classiques, et des composés moléculaires frustrés à base de Cuivre, pour sonder les effets quantiques. Ces mesures ont été réalisées entre 70 mK et 300 K sur les magnétomètres existant à l'Institut Néel.Dans Gd3Ga5O12, nous avons complété le diagramme de phases H-T. Nous avons montré la robustesse de ce diagramme de phases par notre étude sur le composé isomorphe GGd3Al5O12. Nous avons mis en évidence la convergence de toutes les phases observées en un unique point dans les deux systèmes.Dans les systèmes quantiques, nous avons réalisé des études préliminaires sur des clusters Cu44 à base de tétraèdres, et un système triangulaire Cu3. Bien que prometteuses, nous n'avons pas approfondi ces études à cause de problèmes d'échantillons. / The objective of this work was to develop a Faraday force magnetometer to measureabsolute values of the magnetization at very low temperatures (30 mK) andhigh applied magnetic fields (16 T) with a high sensitivity (10-5 emu). This magnetometer will be especially dedicated to the study of the field induced properties of frustrated magnets.In a first part, I present the development of this magnetometer. It involves the optimization of the dilution refrigerator and field gradient coils, and the development of a capacitive detection with a cold amplifier stage to improve the measurement sensitivity. Preliminary magnetic measurements are shown. The changes in the design required to make the magnetometer operational are discussed.In a second part, I focus on experimental studies of frustrated magnets: Gadolinium garnets, described by classical spins, and Copper based frustrated molecular compounds, to probe quantum effects. These measurements were performed from 70 mK to 300 K, with the existing magnetometers at the Institut Neel.In Gd3Ga5O12, we have complemented the H-T phase diagram. This phase diagram was proven to be robust by our study on the isomorphous compound, Gd3Al5O12. We evidence the convergence of all the observed phases to a unique point in both samples.In quantum systems, we performed preliminary studies on Cu44 clusters with tetrahedral motives, and on a triangular system Cu3. Although promising, these studies were not pursued due to sample problems.

The decline of Moore's law has catalyzed a significant effort to identify beyond-CMOS devices and architectures for the coming decades. A multitude of classical and quantum systems have been proposed to address this challenge, and spintronics has emerged as a promising approach for these post-Moore systems. Many of these architectures are tailored specifically for applications in combinatorial optimization and machine learning. Here we propose the use of spintronics for such applications by exploring two distinct but related computing paradigms. First, the use of spin-currents to manipulate and control quantum information is investigated with demonstrated high-fidelity gate operation. This control is accomplished through repeated entanglement and measurement of a stationary qubit with a flying-spin through spin-torque like effects. Secondly, by transitioning from single-spin quantum bits to larger spin ensembles, we then explore the use of stochastic nanomagnets to realize a probabilistic system that is intrinsically governed by Boltzmann statistics. The nanomagnets explore the search space at rapid speeds and can be used in a wide-range of applications including optimization and quantum emulation by encoding the solution to a given problem as the ground state of the equivalent Boltzmann machine. These applications are demonstrated through hardware emulation using an all-digital autonomous probabilistic circuit.

<p>Conventional computing is based on the concept of bits which are classical entities that are either 0 or 1 and can be represented by stable magnets. The field of quantum computing relies on qubits which are a complex linear combination of 0 and 1. Recently, the concept of probabilistic computing with probabilistic (<em>p-</em>)bits was introduced where <em>p-</em>bits are robust classical entities that fluctuate between 0 and 1. <em>P-</em>bits can be naturally represented by low-barrier nanomagnets. Probabilistic computers (<em>p-</em>computers) based on <em>p-</em>bits are domain-based hardware accelerators for Monte Carlo algorithms that can efficiently address probabilistic tasks like sampling, optimization and machine learning. </p> <p>In this dissertation, starting from the intrinsic physics of nanomagnets, we show that a compact hardware implementation of a <em>p-</em>bit based on stochastic magnetic tunnel junctions (s-MTJs) can operate at high-speeds in the order of nanoseconds, a prediction that has recently received experimental support.</p> <p>We then move to the system level and illustrate by simulation and by experiment how multiple interconnected <em>p-</em>bits can be utilized to train a Boltzmann machine built with hardware <em>p-</em>bits. We observe that even non-ideal s-MTJs can be utilized for probabilistic computing when combined with hardware-aware learning.</p> <p>Finally, we show how to build a <em>p-</em>computer to accelerate a wide variety of problems ranging from optimization and sampling to quantum computing and machine learning. The common theme for all these applications is the underlying Monte Carlo and Markov chain Monte Carlo algorithms and their parallelism enabled by a unique <em>p-</em>computer architecture.</p>

<p dir="ltr">Antiferromagnetic spintronics promises next-generation information processing devices with ultra-fast speeds and ultra-low power consumption. Inspired by the recent demonstration of signatures of Tunnel Magnetoresistance (TMR) in non-colinear chiral antiferromagnets of the Mn₃X family, we study the thermal stability of such magnets in both low and high barrier limits. A stochastic Landau-Lifshitz-Gilbert (s-LLG) based numerical assessment of the dynamics reveals that strong exchange fields in Mn₃Sn could lead to thermally-driven rapid fluctuations of the order parameter, viz., octupole moment. However, distinct Random Telegraph Noise (RTN)-like signals distinguish the high barrier limit from the low barrier limit - suggesting different physical phenomena in the two regimes. To that end, the correlation time for thermal fluctuations has been explored analytically following an approach inspired by Langer's theory in the high barrier limit and dephasing mechanisms in the low barrier limit. It has been shown that the dynamics in chiral antiferromagnetic nanoparticles in both regimes are an order of magnitude faster than easy plane ferromagnetic particles. The thermal instability of chiral antiferromagnets could lead to picosecond-scale random number generation in probabilistic bits -- paving the path toward ultra-fast probabilistic computation. </p>

Submitted by Renata Lopes (renatasil82@gmail.com) on 2017-06-28T14:29:11Z No. of bitstreams: 1 damiaodesousavieirajunior.pdf: 10265456 bytes, checksum: 8b1ceaeb4c4be0e91a46c2d30add7349 (MD5) / Approved for entry into archive by Adriana Oliveira (adriana.oliveira@ufjf.edu.br) on 2017-08-07T21:17:55Z (GMT) No. of bitstreams: 1 damiaodesousavieirajunior.pdf: 10265456 bytes, checksum: 8b1ceaeb4c4be0e91a46c2d30add7349 (MD5) / Made available in DSpace on 2017-08-07T21:17:55Z (GMT). No. of bitstreams: 1 damiaodesousavieirajunior.pdf: 10265456 bytes, checksum: 8b1ceaeb4c4be0e91a46c2d30add7349 (MD5) Previous issue date: 2016-02-03 / CAPES - Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / O contínuo desenvolvimento das técnicas de fabricação de estruturas em escala nanométrica, com considerável precisão e reprodutibilidade, tem permitido e estimulado a investigação científica em torno das propriedades básicas e novas aplicações tecnológicas desses sistemas. Especialmente a partir dos anos 90, é crescente o interesse da comunidade científica no comportamento de sistemas magnéticos nano-estruturados. Nestes, a quebra da simetria espacial devido às pequenas dimensões faz com que exibam comportamentos completamente distintos dos observados em amostras macroscópicas. A anisotropia de forma resultante das interações clássicas entre os dipolos magnéticos permite a formação de estruturas magnéticas exóticas em nanomagnetos como vórtices, skyrmions, paredes de domínio individuais e, até mesmo, excitações topológicas similares a monopolos magnéticos. A compreensão e controle do comportamento magnético estático e dinâmico dessas estruturas é fundamental para o desenvolvimento de novos dispositivos tecnológicos baseados em spintrônica. Neste trabalho foram estudadas nanopartículas planares, alongadas na forma elíptica, de material ferromagnético macio, especificamente o Permalloy-79. Tais nanopartículas tem atraído atenção devido ao seu potencial de aplicação prática no desenvolvimento de novos sensores, dispositivos de lógica, mídias de armazenamento de dados de alta densidade e dispositivos MRAM (Magnetic Random Access Memory). Pelo viés do interesse científico básico, tais nano-ilhas ferromagnéticas são a unidade fundamental em arranjos magnéticos bidimensionais geometricamente frustrados, como sistemas de gelo de spin artificiais. Nestes sistemas o arranjo geométrico das ilhas quebra a degenerescência do estado fundamental da rede, caracterizando um estado de frustração geométrica que permite excitações de comportamento análogo ao de monopólos magnéticos. Sob tais aspectos, é essencial caracterizar as configurações magnéticas no estado fundamental e os processos de reversão da magnetização em nanopartículas individuais. A forma elíptica planar gera uma forte anisotropia magnética, definindo duas configurações fundamentais para a magnetização do estado fundamental das nanopartículas: o estado de vórtice ou o estado alinhado ao longo do maior eixo — estado tipo C. A partir de uma razão de aspecto limite, a magnetização do estado fundamental é confinada no plano e ao longo do eixo maior de cada nano ilha, definindo um nanomagneto monodomínio com dois estados degenerados de magnetização, útil às aplicações previamente descritas. Partindo desse intuito estudamos inicialmente, através de simulação por dinâmica de spin, a competição entre os estados de vórtice e os estados alinhados tipo C como uma função da forma de cada nano-ilha elíptica, construindo um diagrama de fases de estados vórtice - tipo C. Cada nanopartícula magnética é modelada por momentos magnéticos que interagem via interação de troca entre primeiros vizinhos e por interação dipolar clássica de longo alcance. Nossos resultados mostram que é possível fabricar nano-ilhas alongadas com estado fundamental alinhado tipo C em razões de aspecto menores que dois. Este é um resultado interessante do ponto de vista tecnológico, pois permite usar ilhas menores que as atuais em pesquisas com gelos de spin e MRAM. Geralmente, os arranjos experimentais são feitos com nanopartículas de razão de aspecto próximas a três para garantir o estado fundamental alinhado da magnetização. Acrescentando ao modelo um termo de interação Zeeman com um campo magnético externo, estudamos o comportamento da reversão da magnetização nas nanopartículas. Consideramos espessuras diferentes e duas razões de aspecto distintas: uma do tamanho experimental usual e outra menor proposta a partir de nossos resultados. Aplicando campo magnético senoidal em diferentes frequências e em direções distintas no plano das nanoilhas, observou-se a dependência dos processos de reversão em função da espessura das partículas e com a direção e frequência do campo aplicado. Os resultados permitem traçar linhas gerais acerca do comportamento da reversão da magnetização nas nanopartículas individuais sob campo magnético externo. Evidentemente para o desenvolvimento das possíveis aplicações tecnológicas, inclusive o controle de excitações como monopólos magnéticos em gelos de spin, é crucial entender os processos ultra rápidos de reversão da magnetização, o que envolve a aplicação de campo externo de alta frequência em direções cuidadosamente definidas. Com esse objetivo, também estudamos a reversão da magnetização nas nano-ilhas por pulsos curtos de campo magnético (da ordem de nanosegundos) aplicados em diferentes direções. Observamos uma forte dependência da coerência da reversão da magnetização com a direção do campo aplicado e uma significante diferença na dependência angular da coercividade em relação ao observado em trabalhos prévios para campos aplicados na condição quase-estática. Finalmente, baseado em nossos resultados, propomos um método para o controle da reversão coerente da magnetização de nanopartículas individuais em matrizes quadradas de gelos de spin artificiais. Acreditamos que nossos resultados poderão ser úteis no desenvolvimento ulterior de arranjos magnéticos artificiais geometricamente frustrados e no controle das excitações topológicas destes sistemas. / The continuous development of structures fabrication techniques at the nanometer scale with considerable precision and reproducibility has allowed and encouraged scientific research around the basic properties and new technological applications of these systems. Especially from the 90's, there is growing interest of the scientific community in the behavior of nanostructured magnetic systems. In these, the breaking of spatial symmetry due to small dimensionality causes quite different behaviors from those observed in the bulk. The resulting shape anisotropy of the classical interaction between magnetic dipoles allows the formation of exotic magnetic structures in nanomagnets as vortices, skyrmions, single domain walls and even topological excitations similar to magnetic monopoles. The understanding and control of static and dynamic magnetic behavior of these structures is essential for the development of new technological devices based on spintronics. In this work we studied planar elongated nanoparticles in the elliptical shape of soft ferromagnetic material, specifically the Permalloy-79. Such nanoparticles have attracted attention because of their potential to practical application in the development of new sensors, logic devices, high density data storage media and MRAM (Magnetic Random Access Memory) devices. By the bias of basic scientific interest, such ferromagnetic nano-islands are the fundamental unit in two-dimensional magnetic arrangements geometrically frustrated as artificial spin ice systems. In these systems, the geometric arrangement of islands break the degeneracy of the network ground state featuring a state of geometrical frustration that allows excitations with analogous behavior of magnetic monopoles. Under these aspects, it is essential to characterize the magnetic configurations in the ground state and the magnetization reversal processes in individual nanoparticles. The elliptical planar shape generates a strong magnetic anisotropy which defines two basic configurations for the magnetization of the ground state of the nanoparticles: the vortex state or the aligned state along the major axis - type C state. As from an aspect ratio limit value, the magnetization of the ground state is confined in the plane and along the major axis of each nano-island defining mono-domain nanomagnet with two degenerate states of magnetization, useful for the applications previously described. Starting from this purpose we study initially, through simulation by spin dynamics, the competition between the vortex states and aligned type C states as a function of the shape of each elliptical nano-island to build a states diagram. Each magnetic nanoparticle is modeled by magnetic moments that interact by exchange interaction between nearest neighbors and by the classical long-range dipolar interaction. Our theoretical results indicate the possibility to manufacture elongated nano-islands with ground state like aligned C state for aspect ratios less than two. This is an interesting result from the technological point of view because it will be possible to use smaller islands in researches on spin ice and MRAM. Generally, the experimental arrangements are made with nanoparticles of aspect ratio close to three to ensure aligned magnetization in the ground state. Adding to the model a Zeeman interaction term between the magnetic moments and an external magnetic field we study the behavior of the magnetization reversal in nanoparticles. We consider different thickness and two different aspect ratios: one in the usual experimental size and a smaller proposed from our results. Applying sinusoidal magnetic field at different frequencies along the anisotropy axis in directions of ten and forty-five degrees from this, we observed the dependence of the reversal processes on the thickness of the particles and with the direction and frequency of the applied field. The results allow to establish general guidelines about the magnetization reversal behavior of the individual nanoparticles under external magnetic field. Evidently, for the development of possible technological applications, including the control of excitation like magnetic monopoles in spin ice, it is crucial to understand the ultrafast magnetization reversal processes which involves the application of high frequency magnetic fields in carefully defined directions. With this aim, we also studied the magnetization reversal of the nano-islands by short pulses of magnetic field (of the nanosecond order) applied in different directions. We observed a strong dependence on the coherence of the magnetization reversal with the direction of the applied field and a significant difference in the angular dependence of the coercivity compared to those seen in previous studies with applied magnetic fields in quasistatic conditions. Finally, based on our results we propose a method for the control of the coherent magnetization reversal of individual nanoparticles in square artificial spin ice arrays. We believe that our results may be useful in further developments of geometrically frustrated magnetic artificial arrangements and in the control of the topological excitations of these systems.

In this work, low dimensional iron group clusters have been studied by application of high magnetic fields. The magnetization has been probed with an MPMS as function of temperature and field. The combination with pulse field measurements up to 52\,T allowed determination of the magnetic exchange coupling parameters, and to probing the effective spin of the ground state. The main focus was on tunable high-field/high-frequency (tHF) ESR in static fields &amp;lt; 17 T and pulse field ESR up to 36 T. This magnetic resonance method has been used for the characterization of the local magnetic properties: The detailed analysis of the field dependence of dedicated spin states allowed to determine the magnetic anisotropy and g-factors. The results were analyzed in the framework of the appropriate effective spin Hamiltonians in terms of magnetization fits and ESR spectrum simulations.

In this work, low dimensional iron group clusters have been studied by application of high magnetic fields. The magnetization has been probed with an MPMS as function of temperature and field. The combination with pulse field measurements up to 52\,T allowed determination of the magnetic exchange coupling parameters, and to probing the effective spin of the ground state. The main focus was on tunable high-field/high-frequency (tHF) ESR in static fields &amp;lt; 17 T and pulse field ESR up to 36 T. This magnetic resonance method has been used for the characterization of the local magnetic properties: The detailed analysis of the field dependence of dedicated spin states allowed to determine the magnetic anisotropy and g-factors. The results were analyzed in the framework of the appropriate effective spin Hamiltonians in terms of magnetization fits and ESR spectrum simulations.

<p dir="ltr">Probabilistic computing is a novel computing scheme that offers a more efficient approach than conventional complimentary metal-oxide-semiconductor (CMOS)-based logic in a variety of applications ranging from Bayesian inference to combinatorial optimization, and invertible Boolean logic. These applications, which have found use in the rapidly growing fields of machine learning and artificial intelligence, are traditionally computationally-intensive and so make the push for novel computing schemes that are intrinsically low-power and scalable all the more urgent.</p><p dir="ltr">The probabilistic bit (or p-bit, the base unit of probabilistic computing) is a naturally fluctuating entity that requires <i>tunable </i>stochasticity; low-barrier nanomagnets, in which the magnetic moment fluctuates randomly and continuously due to the presence of thermal energy, are a natural vehicle for providing the core functionality required. This dissertation describes the work done in mining the rich field of spintronics to produce devices that can act as natural hardware accelerators for probabilistic computing algorithms.</p><p dir="ltr">First, experiments exploring Fe₃O₄ nanoparticles as naturally stochastic systems are presented. Using NV center measurements on an array of such nanoparticles, it is shown that they fluctuate intrinsically at GHz frequencies at room temperature; these fluctuations could be harnessed to act as a stochastic noise source, and would, in principle, enable fast computation.</p><p dir="ltr">The focus then shifts to the development of a platform that allows for easier <i>electrical</i> readout: the low-barrier magnetic tunnel junction (MTJ). We show the work done in the development and characterization of these devices, how they respond to non-ideal environments, such as elevated temperatures and exposure to high-energy electromagnetic radiation, how their intrinsic stochasticity might be tuned with electrical currents and external magnetic fields, and then how these might be integrated with a simple transistor circuit to produce a compact low-energy implementation of a p-bit.</p><p dir="ltr">Next, by integrating our stochastic MTJs with 2D-MoS₂field-effect transistors (FETs), the first <i>on-chip </i>realization of a key p-bit building block, displaying voltage-controllable stochasticity, is demonstrated. This is followed by another key demonstration through the fabrication of stochastic MTJs directly on top of an integrated circuit platform, where the transistor circuitry is provided by 180nm-node CMOS technology.</p><p dir="ltr">In addition, supported by circuit simulations, this work provides a careful device-level analysis of the three transistor-one magnetic tunnel junction (3T-1MTJ) p-bit design, evaluating how the characteristics of each component can influence the overall p-bit’s output. In particular, we show that – against common wisdom – a large tunnel magnetoresistance (TMR) is not the best choice for p-bits; bimodal telegraphic fluctuations are highly undesirable and are a sign of a slow device; and an ideal inverter with a large gain is unsuitable for p-bit applications due to the higher likelihood of unwanted plateaus in the resulting p-bit’s output.</p><p dir="ltr">This analysis is extended to consider the impact of such non-ideal p-bits when used to construct probabilistic circuits, with the focus on the emulation of the Boolean logic AND gate through a three p-bit correlated system. It is found that a probabilistic circuit made with ideal p-bits can accurately emulate the function of an AND gate, while the non-ideal p-circuits suffer from an increased error rate in emulating the AND gate’s truth table.</p><p dir="ltr">The understanding gained at the individual device level, in what makes a good or bad MTJ, to how the different components of the 3T-1MTJ p-bit can affect its output, and subsequently how non-ideal p-bits can impact circuit performance, can be important for the future realization of scaled on-chip p-bit networks.</p>