In this dissertation I present advances in the studies of artificial lattices with honeycomb topology, called artificial graphene (AG), in nano-patterned GaAs quantum wells (QWs). AG lattices with very small lattice constants as low as 40 nm are achieved for the first time in GaAs. The high quality AG lattices are created by optimized electron-beam (E-beam) lithography followed by inductively coupled plasma reactive-ion etching (ICP-RIE) process. E-beam lithography is used to define a honeycomb lattice etch mask on the surface of the GaAs QW sample and the optimized anisotropic ICP-RIE process is developed to transfer the pattern into the sample and create the AG lattices. Such high-resolution AG lattices with small lattice constants are essential to form AG miniband structures and create well-developed Dirac cones. Characterization of electron states in the nanofabricated artificial lattices is by optical experiments. Optical emission (photoluminescence) yields a determination of the Fermi energy of the electrons. A significant reduction of the Fermi energy is due to the nano-patterning process. Resonant inelastic light scattering (RILS) spectra reveal novel transitions related to the electron band structures of the AG lattices. These transitions exhibit a remarkable agreement with the predicted joint density of states (JDOS) based on the band structure calculation for the honeycomb topology. I calculate the electron band structures of AG lattices in nano-patterned GaAs QWs using a periodic muffin-tin potential model. The evaluations predict linear energy-momentum dispersion and Dirac cones, where the massless Dirac fermions (MDFs) appear, occur in the band structures. Requirements of the parameters of the AG potential to achieve isolated and well-developed Dirac cones are discussed. Density of states (DOS) and JDOS from AG band structures are calculated, which provide a basis to interpret quantitatively observed transitions of electrons involving AG bands. RILS of intersubband transitions reveal intriguing satellite peaks that are not present in the as-grown QWs. These additional peaks are interpreted as combined intersubband transitions with simultaneous change of QW subband and AG band index. The calculated JDOS for the electron transitions within the AG lattice model provide a remarkably accurate description of the combined intersubband excitations. Novel low-lying excitation peaks in RILS spectra, interpreted as direct transitions between AG bands without change in QW subband, provide a more direct insight on the AG band structures. We discovered that RILS transitions around the Dirac cones are resonantly enhanced by varying the incident photon energies. The spectral lineshape of these transitions provides insights into the formation of Dirac cones that are characteristic of the honeycomb symmetry of the AG lattices. The results confirm the formation of AG miniband structures and well-developed Dirac cones. The realization of AG lattices in a nanofabricated high mobility semiconductor offers the advantage of tunability through methods suitable for device scalability and integration. The last part of this thesis describes the growth of nanocrystalline single layer and bilayer graphene on sapphire substrates by molecular beam epitaxy (MBE) with a solid carbon source. Raman spectroscopy reveals that fabrication of single layer, bilayer or multilayer graphene crucially depends on MBE growth conditions. Etch pits revealed by atomic force microscopy indicate a removal mechanism of carbon by reduction of sapphire. Tuning the interplay between carbon deposition and its removal, by varying the incident carbon flux and substrate temperature, should enable the growth of high quality graphene layers on large area sapphire substrates.

Seja K/Q uma extensÃo abeliana de grau primo Ãmpar ρ e condutor n, onde ρ nÃo se ramifica em K/Q. As principais contribuiÃÃes deste trabalho sÃo: 1) caracterizaÃÃo de ideais Ok em cuja fatoraÃÃo constam apenas ideais primos ramificados K/Q; 2) cÃlculo da densidade de centro da representaÃÃo geomÃtrica de Z-mÃdulos em Ok caracterizados por uma equaÃÃo modular (para ρ = 3,5 e 7, parametriza-se o algoritmo que otimiza a densidade de centro destes reticulados). AlÃm disso, os seguintes resultados sÃo tambÃm descritos: 1) FamÃlias de reticulados associados a polinÃmios em Z[x] de grau dois e trÃs; 2) uma prova alternativa da finitude do grupo das classes de um corpo nÃmeros baseada somente em empacotamentos esfÃricos. / Let K/Q be an Abelian extension of ood degree ρ and conductor n, where ρ does not ramify in K/Q. The main contributions of this work are: 1) characterization of ideals of Ok whose factorization includes only prime ramified ideals K/Q; 2) calculation of the center density of the geometric representation of Z-modules in Ok characterized by a modular equation (for ρ = 3.5, and 7, the algorithm that is used to optimize the center density of those lattices is parametrized). Besides, the following results are also described: 1) Families of lattices associated to polynomials in Z[x] of degree two and three; 2) an alternative proof of the finiteness of the class group of a number field based solely on sphere packings.

Neste trabalho estudamos o comportamento crítico do modelo de Potts p-estados na árvore de Cayley, através das propriedades do conjunto de zeros de Yang-Lee da função de partição. Tratando a transformação do grupo de renormalização como um mapeamento racional na esfera de Riemann utiliza-se alguns resultados da teoria de Julia e Fatou para obter-se uma descrição geométrica do comportamento crítico do modelo. Mostra-se de que forma o conjunto de zeros de Yang-Lee se relaciona com o conjunto de Julia do mapa do grupo de renormalização, e calculam-se alguns parâmetros geométricos desse conjunto que descrevem o comportamento não universal do modelo. / We study the critical behavior of the p-state Potts model on a Cayley tree, looking for the properties of the Yang-Lee zeros set of the partition function. We treated the renormalization group transformation as a rational mapping on the Riemann sphere, and use some results from the Julia and Fatou theory to obtain a geometrical description of the critical properties of the model. We show how the Yang-Lee zeros set is associated with the Julia set of the renormalization group map, and we also calculate some geometrical parameters of this set which describes the non-universal behavior of the model.

By exploiting the versatility of ultracold atoms experiments, a variety of many-body phe- nomena can be studied. Ferromagnetism in a trapped ultracold gas of repulsively interacting fermions is considered within a local-density approximation and beyond, where energetic sig- natures indicate its onset. Transport of a Bose-Einstein condensate is explored experimentally in a tunable double-well potential, and a crossover from hydrodynamic to Josephson transport is observed as the barrier between condensates is raised. To add a degree of freedom for the manipulation of ultracold gases, two schemes for species-specific optical lattices are developed theoretically.

Two framework oxide materials of the MO��� network type have been synthesized and structurally characterized by synchrotron and X-ray powder diffraction and the Rietveld method in the temperature range 25~500 K. The results show one of them to be a low thermal expansion material. Theoretical studies of negative thermal expansion (NTE) in framework oxides were conducted with two methods, geometrical modeling by Rigid Unit Mode (RUM) method and lattice dynamic calculations by free energy minimization (FEM) method, the results are compared with each other as well as with experimental observations. RUM analysis of all five types of framework oxide structures negates any simple and direct correlation between presence or absence of RUMs in a structure and the sign of its thermal expansion. Instead, results suggest that NTE of a crystalline solid can not be explained by pure geometrical considerations over its structure alone, and for a better understanding of structure-relationship in negative thermal expansion structures, specific interatomic interactions present in each one must be brought in explicitly. FEM calculation of two negative thermal expansion structures indicates on a structure by structure basis NTE could be predicted and understood within the Gruneisen model, which attributes NTE of a structure to special vibration modes in a structure that softens when the lattice shrinks. The soft NTE modes are, however, not necessarily RUM or RUM like vibration motions. / Graduation date: 2003

Los solitones ópticos son paquetes de luz (haces y/o pulsos) que no se dispersan gracias al balance entre difracción/dispersión y no linealidad. Al propagarse e interactuar los unos con los otros muestran propiedades que normalmente se asocian a partículas. Las propiedades de los solitones ópticos en fibras ópticas y cristales han sido investigadas en profundidad durante las últimas dos décadas. Sin embargo, los solitones en mallas, o redes, ópticas, que podrían ser usados para procesado y direccionamiento totalmente óptico de señales, se han convertido en una nueva área de investigación. El principal objetivo de esta tesis es el estudio de nuevas técnicas para controlar solitotes en medios no lineales en mallas ópticas.El capítulo 2 se centra en ciertas propiedades de los solitones ópticos en medios no lineales cuadráticos. La primera sección presenta en detalle la existencia y estabilidad de tres familias representativas de solitones espacio temporales en dos dimensiones en series de frentes de onda cuadráticos no lineales. Se asume, además de la dispersión temporal del pulso, la combinación de difracción discreta que surge debido al acoplamiento débil entre frentes de onda vecinos. La otra sección da cuenta de la existencia y estabilidad de vórtices de solitones multicolores en retículo, consistentes en cuatro jorobas principales dispuestas en una configuración cuadrada. También se investiga la posibilidad de generarlos dinámicamente a partir de haces de entrada Gaussianos con vórtices anidados. La técnica de inducción de mallas ópticas ofrece un sinfín de posibilidades para la creación de configuraciones de guía de ondas con varios haces de luz no difractantes. El capítulo 3 presenta el concepto de estructuras reconfigurables ópticamente inducidas por haces no difractantes de Bessel mutuamente incoherentes en medios no lineales de tipo Kerr. Los acopladores de dos nucleos son introducidos y se muestra cómo calibrar las propiedades de conmutación de estas estructuras variando la intensidad de los haces de Bessel. El capítulo también discute varios escenarios de conmutación para solitones lanzados al interior de acopladores direccionales multinucleares ópticamente inducidos por apropiadas series de haces de Bessel. Es más, la propagación de solitones es investigada en redes reconfigurables bidimensionales inducidas ópticamente por series de haces de Bessel no difractantes. Se muestra que los haces anchos de solitones pueden moverse a través de redes con diferentes topologías casi sin pérdidas por radiación. Finalmente, se estudian las propiedades de las uniones X, que se crean a partir de dos haces de Bessel intersectantes. La respuesta no local de los medios no lineales puede jugar un papel importante en las propiedades de los solitones. El capítulo 4 trata el impacto de la no localidad en las características físicas exhibidas por los solitones que permiten los medios no lineales de tipo Kerr con una retícula óptica integrada. El capítulo investiga propiedades de diferentes familias de solitones en mallas en medios no lineales no locales. Se muestra que la no localidad de la respuesta no lineal puede afectar profundamente la movilidad de los solitones. Las propiedades de los solitones de gap también se discuten en el caso de cristales fotorefractivos con una respuesta de difusión no local asimétrica y en presencia de una malla inducida.El capítulo 5 trata del impacto de la no localidad en la estabilidad de complejos de solitones en medios no lineales de tipo Kerr uniformes. En primer lugar, se muestra que la diferente respuesta no local de los materiales tiene distinta influencia en la estabilidad de los complejos de solitones en el caso escalar. En segundo lugar, se da cuenta de una serie de resultados experimentales sobre solitones multipolares escalares en medios no lineales fuertemente no locales en 2D, incluyendo solitones dipolares, tripolares y de tipo pajarita, organizados en series de puntos brillantes fuera de fase. Finalmente, el capítulo estudia la interacción entre la no linealidad no local y el acoplamiento vectorial, enfatizando especialmente la estabilización de efectos vectoriales en complejos de solitones en medios no lineales no locales.Por último, el capítulo 6 resume los principales resultados obtenidos en la tesis y discute algunas cuestiones abiertas. / Optical solitons are light packets (beams and/or pulses) that do not broaden because of the proper balance between diffraction/dispersion and nonlinearity. They propagate and interact with one another while displaying properties that are normally associated with real particles. The properties of optical solitons in optical fibers and crystals have been investigated comprehensively during the last two decades. However, solitons in optical lattices, which might be used for all-optical signal processing and routing have recently emerged a new area of research. The main objective of this thesis is the investigation of new techniques for soliton control in nonlinear media with/without an imprinted optical lattice. Chapter 2 focuses on properties of optical solitons in quadratic nonlinear media. The first section presents in detail the existence and stability of three representative families of two-dimensional spatiotemporal solitons in quadratic nonlinear waveguide arrays. It is assumed in addition to the temporal dispersion of the pulse, the combination of discrete diffraction that arises because of the weak coupling between neighboring waveguides. The other section reports on the existence and stability of multicolor lattice vortex solitons, which comprise four main humps arranged in a square configuration. It is also investigated the possibility of their dynamical generation from Gaussian-type input beams with nested vortices. The technique of optical lattice induction opens a wealth of opportunities for creation of waveguiding configurations with various nondiffracting light beams. Chapter 3 puts forward the concept of reconfigurable structures optically induced by mutually incoherent nondiffracting Bessel beams in Kerr-type nonlinear media. Two-core couplers are introduced and it is shown how to tune the switching properties of such structures by varying the intensity of the Bessel beams. The chapter also discusses various switching scenarios for solitons launched into the multi core directional couplers optically-induced by suitable arrays of Bessel beams. Furthermore, propagation of solitons is investigated in reconfigurable two-dimensional networks induced optically by arrays of nondiffracting Bessel beams. It is shown that broad soliton beams can move across networks with different topologies almost without radiation losses. Finally, properties of X-junctions are studied, which are created with two intersecting Bessel beams.Nonlocal response of nonlinear media can play an important role in properties of solitons. Chapter 4 treats the impact of nonlocality in the physical features exhibited by solitons supported by Kerr-type nonlinear media with an imprinted optical lattice. The chapter investigates properties of different families of lattice solitons in nonlocal nonlinear media. It is shown that the nonlocality of the nonlinear response can profoundly affect the soliton mobility. The properties of gap solitons are also discussed for photorefractive crystals with an asymmetric nonlocal diffusion response and in the presence of an imprinted optical lattice.Chapter 5 is devoted to the impact of nonlocality on the stability of soliton complexes in uniform nonlocal Kerr-type nonlinear media. First, it is shown that the different nonlocal response of materials has different influence on the stability of soliton complexes in scalar case. Second, experimental work is reported on scalar multi-pole solitons in 2D highly nonlocal nonlinear media, including dipole, tripole, and necklace-type solitons, organized as arrays of out-of-phase bright spots. Finally, the chapter addresses the interplay between nonlocal nonlinearity and vectoral coupling, specially emphasizing the stabilization of vector effects on soliton complexes in nonlocal nonlinear media.Finally, Chapter 6 summarizes the main results obtained in the thesis and discusses some open prospects.

This thesis presents the experimental study on the two-spin component, strongly interacting 6 Li Fermi gases in 3D and 1D traps. The interaction strength is tuned from the molecular BEC regime to the BCS regime using a Feshbach resonance. The trap dimension can be tuned from 3D to 1D with the implementation of optical lattice. The evaporation of imbalanced Fermi gases in 3D trap is studied. The anisotropic and fast evaporation is the cause of the deformation observed in the 2006 Rice experiment. In a balanced Fermi system, the fraction of correlated states is measured as a function of interaction and temperature. At unitarity, the fraction of correlated states is ∼85% and exists above T c . The one-body-like photoexcitation rate can be related to the contact quantity. Lastly, the spin-imbalance in a one-dimensional Fermi gas is studied. The 1D phase diagram is mapped out. The result agrees well with the 1D theory, in which the partially polarized regime is predicted to be a FFLO phase, an exotic superfluid with pairs carrying finite center-of-mass momentum proposed almost 50 years ago.

The theory of lattices --- a mathematical approach for representing infinite discrete points in Euclidean space, has become a powerful tool to analyze many point-to-point digital and wireless communication systems, particularly, communication systems that can be well-described by the linear Gaussian vector channel model. This is mainly due to the three facts about channel codes constructed using lattices: they have simple structure, their ability to achieve the fundamental limits (the capacity) of the channel, and most importantly, they can be decoded using efficient decoders called lattice decoders. Since its introduction to multiple-input multiple-output (MIMO) wireless communication systems, sphere decoders has become an attractive efficient implementation of lattice decoders, especially for small signal dimensions and/or moderate to large signal-to-noise ratios (SNRs). In the first part of this dissertation, we consider sphere decoding algorithms that describe lattice decoding. The exact complexity analysis of the basic sphere decoder for general space-time codes applied to MIMO wireless channel is known to be difficult. Characterizing and understanding the complexity distribution is important, especially when the sphere decoder is used under practically relevant runtime constraints. In this work, we shed the light on the (average) computational complexity of sphere decoding for the quasi-static, LAttice Space-Time (LAST) coded MIMO channel. Sphere decoders are only efficient in the high SNR regime and low signal dimensions, and exhibits exponential (average) complexity for low-to-moderate SNR and large signal dimensions. On the other extreme, linear and non-linear receivers such as minimum mean-square error (MMSE), and MMSE decision-feedback equalization (DFE) are considered attractive alternatives to sphere decoders in MIMO channels. Unfortunately, the very low decoding complexity advantage that these decoders can provide comes at the expense of poor performance, especially for large signal dimensions. The problem of designing low complexity receivers for the MIMO channel that achieve near-optimal performance is considered a challenging problem and has driven much research in the past years. The problem can solved through the use of lattice sequential decoding that is capable of bridging the gap between sphere decoders and low complexity linear decoders (e.g., MMSE-DFE decoder). In the second part of this thesis, the asymptotic performance of the lattice sequential decoder for LAST coded MIMO channel is analyzed. We determine the rates achievable by lattice coding and sequential decoding applied to such a channel. The diversity-multiplexing tradeoff under such a decoder is derived as a function of its parameter--- the bias term. In this work, we analyze both the computational complexity distribution and the average complexity of such a decoder in the high SNR regime. We show that there exists a cut-off multiplexing gain for which the average computational complexity of the decoder remains bounded. Our analysis reveals that there exists a finite probability that the number of computations performed by the decoder may become excessive, even at high SNR, during high channel noise. This probability is usually referred to as the probability of a decoding failure. Such probability limits the performance of the lattice sequential decoder, especially for a one-way communication system. For a two-way communication system, such as in MIMO Automatic Repeat reQuest (ARQ) system, the feedback channel can be used to eliminate the decoding failure probability. In this work, we modify the lattice sequential decoder for the MIMO ARQ channel, to predict in advance the occurrence of decoding failure to avoid wasting the time trying to decode the message. This would result in a huge saving in decoding complexity. In particular, we will study the throughput-performance-complexity tradeoffs in sequential decoding algorithms and the effect of preprocessing and termination strategies. We show, analytically and via simulation, that using the lattice sequential decoder that implements a simple yet efficient time-out algorithm for joint error detection and correction, the optimal tradeoff of the MIMO ARQ channel can be achieved with significant reduction in decoding complexity.

We describe a fast method for the evaluation of an arbitrary high-dimensional multivariate algebraic polynomial in Chebyshev form at the nodes of an arbitrary rank-1 Chebyshev lattice. Our main focus is on conditions on rank-1 Chebyshev lattices allowing for the exact reconstruction of such polynomials from samples along such lattices and we present an algorithm for constructing suitable rank-1 Chebyshev lattices based on a component-by-component approach. Moreover, we give a method for the fast, exact and stable reconstruction.

The theory of lattices --- a mathematical approach for representing infinite discrete points in Euclidean space, has become a powerful tool to analyze many point-to-point digital and wireless communication systems, particularly, communication systems that can be well-described by the linear Gaussian vector channel model. This is mainly due to the three facts about channel codes constructed using lattices: they have simple structure, their ability to achieve the fundamental limits (the capacity) of the channel, and most importantly, they can be decoded using efficient decoders called lattice decoders. Since its introduction to multiple-input multiple-output (MIMO) wireless communication systems, sphere decoders has become an attractive efficient implementation of lattice decoders, especially for small signal dimensions and/or moderate to large signal-to-noise ratios (SNRs). In the first part of this dissertation, we consider sphere decoding algorithms that describe lattice decoding. The exact complexity analysis of the basic sphere decoder for general space-time codes applied to MIMO wireless channel is known to be difficult. Characterizing and understanding the complexity distribution is important, especially when the sphere decoder is used under practically relevant runtime constraints. In this work, we shed the light on the (average) computational complexity of sphere decoding for the quasi-static, LAttice Space-Time (LAST) coded MIMO channel. Sphere decoders are only efficient in the high SNR regime and low signal dimensions, and exhibits exponential (average) complexity for low-to-moderate SNR and large signal dimensions. On the other extreme, linear and non-linear receivers such as minimum mean-square error (MMSE), and MMSE decision-feedback equalization (DFE) are considered attractive alternatives to sphere decoders in MIMO channels. Unfortunately, the very low decoding complexity advantage that these decoders can provide comes at the expense of poor performance, especially for large signal dimensions. The problem of designing low complexity receivers for the MIMO channel that achieve near-optimal performance is considered a challenging problem and has driven much research in the past years. The problem can solved through the use of lattice sequential decoding that is capable of bridging the gap between sphere decoders and low complexity linear decoders (e.g., MMSE-DFE decoder). In the second part of this thesis, the asymptotic performance of the lattice sequential decoder for LAST coded MIMO channel is analyzed. We determine the rates achievable by lattice coding and sequential decoding applied to such a channel. The diversity-multiplexing tradeoff under such a decoder is derived as a function of its parameter--- the bias term. In this work, we analyze both the computational complexity distribution and the average complexity of such a decoder in the high SNR regime. We show that there exists a cut-off multiplexing gain for which the average computational complexity of the decoder remains bounded. Our analysis reveals that there exists a finite probability that the number of computations performed by the decoder may become excessive, even at high SNR, during high channel noise. This probability is usually referred to as the probability of a decoding failure. Such probability limits the performance of the lattice sequential decoder, especially for a one-way communication system. For a two-way communication system, such as in MIMO Automatic Repeat reQuest (ARQ) system, the feedback channel can be used to eliminate the decoding failure probability. In this work, we modify the lattice sequential decoder for the MIMO ARQ channel, to predict in advance the occurrence of decoding failure to avoid wasting the time trying to decode the message. This would result in a huge saving in decoding complexity. In particular, we will study the throughput-performance-complexity tradeoffs in sequential decoding algorithms and the effect of preprocessing and termination strategies. We show, analytically and via simulation, that using the lattice sequential decoder that implements a simple yet efficient time-out algorithm for joint error detection and correction, the optimal tradeoff of the MIMO ARQ channel can be achieved with significant reduction in decoding complexity.