Spin systems are powerful mathematical models widely used and studied in Statistical Physics and Computer Science. This thesis focuses the study of spin systems on colorings and weighted independent sets (the hard-core model). In many spin systems, there exist phase transition phenomena: there is a threshold value of a parameter such that when the parameter is on one side of the threshold, the system exhibits the so-called spatial decay of correlation, i.e., the influence from a set of vertices to another set of vertices diminishes as the distance between the two sets grows; when the parameter is on the other side, long range correlations persist. The uniqueness problem and the reconstruction problem are two major threshold problems that are concerned with the decay of correlations in the Gibbs measure from different perspectives. In Computer Science, the study of spin systems mainly focused on finding an efficient algorithm that samples the configurations from a distribution that is very close to the Gibbs measure. Glauber dynamics is a typical Markov chain algorithm for performing sampling. In many systems, the convergence time of the Glauber dynamics also exhibits a threshold behavior: the speed of convergence experiences a dramatic change around the threshold of the parameter. The first two parts of this thesis focus on making connections between the phase transition of the convergence time of the dynamics and the phase transition of the reconstruction phenomenon in both colorings and the hard-core model on regular trees. A relatively sharp threshold is established for the change of the convergence time, which coincides with the reconstruction threshold. A general technique of upper bounding the conductance of the dynamics via analyzing the sensitivity of the reconstruction algorithm is proposed and proven to be very effective for lower bounding the convergence time of the dynamics. The third part of the thesis provides an innovative analytical method for establishing a strong version of the decay of correlation of the Gibbs distributions for many two spin systems on various classes of graphs. In particular, the method is applied to the hard-core model on the square lattice, a very important graph that is of great interest in both Statistical Physics and Computer Science. As a result, we significantly improve the lower bound of the uniqueness threshold on the square lattice and extend the range of parameter where the Glauber dynamics is rapidly mixing.

Neste trabalho estudamos as transições de fases cinéticas e o comportamento crítico da produção de entropia em modelos de spins com interação entre primeiros vizinhos e sujeitos a duas dinâmicas de Glauber, as quais simulam dois banhos térmicos a diferentes temperaturas. Para tanto, é admitido que o sistema corresponde a um processo markoviano contínuo no tempo o qual obedece a uma equação mestra. Dessa forma, o sistema atinge naturalmente estados estacionários, que podem ser de equilíbrio ou de não-equilíbrio. O primeiro corresponde exatamente ao modelo de Ising, que ocorre quando o sistema se encontra em contato com apenas um dos reservatórios. Dessa forma, há uma transição de fase na temperatura de Curie e o balanceamento detalhado é seguramente satisfeito. No segundo caso, os dois banhos térmicos são responsáveis por uma corrente de probabilidade que só existe visto que a reversibilidade microscópica não é mais verificada. Como conseqüência, nesse regime de não-equilíbrio o sistema apresenta uma produção de entropia não nula. Para avaliarmos os diagramas de fase e a produção de entropia utilizamos as aproximações de pares e as simulações de Monte Carlo. Além disso, admitimos que a teoria de escala finita pode ser aplicada no modelo. Esses métodos foram capazes de preverem as transições de fases sofridas pelo sistema. Os expoentes e os pontos críticos foram estimados através dos resultados numéricos. Para a magnetização e a susceptibilidade obtemos = 0,124(1) e = 1,76(1), o que nos permite concluir que o nosso modelo pertence à mesma classe de Ising. Esse resultado refere-se ao princípio da universalidade do ponto crítico, que é verificado devido o nosso modelo apresentar a mesma simetria de inversão que a do modelo de Ising. Além disso, as aproximações de pares também mostraram uma singularidade na derivada da produção de entropia no ponto crítico. E as simulações de Monte Carlo nos permitem sugerir que tal comportamento é uma divergência logarítmica cujo expoente crítico associado vale 1. / We study kinetic phase transitions and the critical behavior of the entropy production in spin models with nearest neighbor interactions subject to two Glauber dynamics, which simulate two thermal baths at different temperatures. In this way, it is assumed that the system corresponds to a continuous time Markov process which obeys the master equation. Thus, the system naturally reaches steady states, which can be equilibrium or nonequilibrium. The former corresponds exactly to the Ising model, which occurs since the system is in contact with only one of the reservoirs. In this case, there is a phase transition at the Curie temperature and the detailed balance surely holds. In the second case, the two thermal baths create a non trivial probability current only when microscopic reversibility is not verified. As a consequence, there is a positive entropy production in a non-equilibrium steady state. Pair approximations and Monte Carlo simulations are employed to evaluate the phase diagrams and the entropy production. Furthermore, we assume that the finite-size scaling theory can be applied to the model. These methods were able to predict the phase transitions undergone by the system. The exponents and the critical points were estimated by the numerical results. Our best estimates of critical exponents to the magnetization and susceptibility are = 0,124 (1) and = 1,76 (1), which allows us to conclude that our model belongs to the same class of Ising. This result refers to the principle of universality of the critical point, which is checked because our model has the same inversion symmetry of the Ising model. Moreover, the pair approximation also showed a singularity in the derivative of the entropy production at the critical point. And Monte Carlo simulations allow us to suggest that the divergence at the critical point is of the logarithmic type whose critical exponent is 1

Nous avons étudié la dynamique de Glauber sur les pavages de domaines finies du plan par des losanges ou par des dominos de taille 2 × 1. Ces pavages sont naturellement associés à des surfaces de R^3, qui peuvent être vues comme des interfaces dans des modèles de physique statistique. En particulier les pavages par des losanges correspondent au modèle d'Ising tridimensionnel à température nulle. Plus précisément les pavages d'un domaine sont en bijection avec les configurations d'Ising vérifiant certaines conditions au bord (dépendant du domaine pavé). Ces conditions forcent la coexistence des phases + et - ainsi que la position du bord de l'interface. Dans la limite thermodynamique où L, la longueur caractéristique du système, tend vers l'infini, ces interfaces obéissent à une loi des grand nombre et convergent vers une forme limite déterministe ne dépendant que des conditions aux bord. Dans le cas où la forme limite est planaire et pour les losanges, Caputo, Martinelli et Toninelli [CMT12] ont montré que le temps de mélange Tmix de la dynamique est d'ordre O(L^{2+o(1)}) (scaling diffusif). Nous avons généralisé ce résultat aux pavages par des dominos, toujours dans le cas d'une forme limite planaire. Nous avons aussi prouvé une borne inférieure Tmix ≥ cL^2 qui améliore d'un facteur log le résultat de [CMT12]. Dans le cas où la forme limite n'est pas planaire, elle peut être analytique ou bien contenir des parties “gelées” où elle est en un sens dégénérée. Dans le cas où elle n'a pas de telle partie gelée, et pour les pavages par des losanges, nous avons montré que la dynamique de Glauber devient “macroscopiquement proche” de l'équilibre en un temps L^{2+o(1)} / We studied the Glauber dynamics on tilings of finite regions of the plane by lozenges or 2 × 1 dominoes. These tilings are naturally associated with surfaces of R^3, which can be seen as interfaces in statistical physics models. In particular, lozenge tilings correspond to three dimensional Ising model at zero temperature. More precisely, tilings of a finite regions are in bijection with Ising configurations with some boundary conditions (depending on the tiled domain). These boundary conditions impose the coexistence of the + and - phases, together with the position of the boundary of the interface. In the thermodynamic limit where L, the characteristic length of the system, tends toward infinity, these interface follow a law of large number and converge to a deterministic limit shape depending only on the boundary condition. When the limit shape is planar and for lozenge tilings, Caputo, Martinelli and Toninelli [CMT12] showed that the mixing time of the dynamics is of order (L^{2+o(1)}) (diffusive scaling). We generalized this result to domino tilings, always in the case of a planar limit shape. We also proved a lower bound Tmix ≥ cL^2 which improve on the result of [CMT12] by a log factor. When the limit shape is not planar, it can either be analytic or have some “frozen” domains where it is degenerated in a sense. When it does not have such frozen region, and for lozenge tilings, we showed that the Glauber dynamics becomes “macroscopically close” to equilibrium in a time L^{2+o(1)}

Neste trabalho estudamos as transições de fases cinéticas e o comportamento crítico da produção de entropia em modelos de spins com interação entre primeiros vizinhos e sujeitos a duas dinâmicas de Glauber, as quais simulam dois banhos térmicos a diferentes temperaturas. Para tanto, é admitido que o sistema corresponde a um processo markoviano contínuo no tempo o qual obedece a uma equação mestra. Dessa forma, o sistema atinge naturalmente estados estacionários, que podem ser de equilíbrio ou de não-equilíbrio. O primeiro corresponde exatamente ao modelo de Ising, que ocorre quando o sistema se encontra em contato com apenas um dos reservatórios. Dessa forma, há uma transição de fase na temperatura de Curie e o balanceamento detalhado é seguramente satisfeito. No segundo caso, os dois banhos térmicos são responsáveis por uma corrente de probabilidade que só existe visto que a reversibilidade microscópica não é mais verificada. Como conseqüência, nesse regime de não-equilíbrio o sistema apresenta uma produção de entropia não nula. Para avaliarmos os diagramas de fase e a produção de entropia utilizamos as aproximações de pares e as simulações de Monte Carlo. Além disso, admitimos que a teoria de escala finita pode ser aplicada no modelo. Esses métodos foram capazes de preverem as transições de fases sofridas pelo sistema. Os expoentes e os pontos críticos foram estimados através dos resultados numéricos. Para a magnetização e a susceptibilidade obtemos = 0,124(1) e = 1,76(1), o que nos permite concluir que o nosso modelo pertence à mesma classe de Ising. Esse resultado refere-se ao princípio da universalidade do ponto crítico, que é verificado devido o nosso modelo apresentar a mesma simetria de inversão que a do modelo de Ising. Além disso, as aproximações de pares também mostraram uma singularidade na derivada da produção de entropia no ponto crítico. E as simulações de Monte Carlo nos permitem sugerir que tal comportamento é uma divergência logarítmica cujo expoente crítico associado vale 1. / We study kinetic phase transitions and the critical behavior of the entropy production in spin models with nearest neighbor interactions subject to two Glauber dynamics, which simulate two thermal baths at different temperatures. In this way, it is assumed that the system corresponds to a continuous time Markov process which obeys the master equation. Thus, the system naturally reaches steady states, which can be equilibrium or nonequilibrium. The former corresponds exactly to the Ising model, which occurs since the system is in contact with only one of the reservoirs. In this case, there is a phase transition at the Curie temperature and the detailed balance surely holds. In the second case, the two thermal baths create a non trivial probability current only when microscopic reversibility is not verified. As a consequence, there is a positive entropy production in a non-equilibrium steady state. Pair approximations and Monte Carlo simulations are employed to evaluate the phase diagrams and the entropy production. Furthermore, we assume that the finite-size scaling theory can be applied to the model. These methods were able to predict the phase transitions undergone by the system. The exponents and the critical points were estimated by the numerical results. Our best estimates of critical exponents to the magnetization and susceptibility are = 0,124 (1) and = 1,76 (1), which allows us to conclude that our model belongs to the same class of Ising. This result refers to the principle of universality of the critical point, which is checked because our model has the same inversion symmetry of the Ising model. Moreover, the pair approximation also showed a singularity in the derivative of the entropy production at the critical point. And Monte Carlo simulations allow us to suggest that the divergence at the critical point is of the logarithmic type whose critical exponent is 1

In this work we provide several improvements in the study of phase transitions of interacting particle systems: - We determine a quantitative relation between non-extremality of the limiting Gibbs measure of a tree-based spin system, and the temporal mixing of the Glauber Dynamics over its finite projections. We define the concept of 'sensitivity' of a reconstruction scheme to establish such a relation. In particular, we focus on the independent sets model, determining a phase transition for the mixing time of the Glauber dynamics at the same location of the extremality threshold of the simple invariant Gibbs version of the model. - We develop the technical analysis of the so-called spatial mixing conditions for interacting particle systems to account for the connectivity structure of the underlying graph. This analysis leads to improvements regarding the location of the uniqueness/non-uniqueness phase transition for the independent sets model over amenable graphs; among them, the elusive hard-square model in lattice statistics, which has received attention since Baxter's solution of the analogous hard-hexagon in 1980. - We build on the work of Montanari and Gerschenfeld to determine the existence of correlations for the coloring model in sparse random graphs. In particular, we prove that correlations exist above the 'clustering' threshold of such a model; thus providing further evidence for the conjectural algorithmic 'hardness' occurring at such a point.