Dipolar interactions and origin of spin ice in ising pyrochlore magnets

Recent experiments suggest that the Ising pyrochlore magnets Ho2Ti2O7 and Dy2Ti2O7 display qualitative properties of the nearest-neighbor "spin ice" model. We discuss the dipolar energy scale present in both these materials and discuss how spin-ice behavior can occur despite the presence of long-range dipolar interactions. We present results of numerical simulations and a mean field analysis of Ising pyrochlore systems. Based on our quantitative theory, we suggest that the spin-ice behavior in these systems is due to long-range dipolar interactions, and that the nearest-neighbor exchange in Dy2Ti2O7 is antiferromagnetic.     

A microscopic derivation of the classical nucleation equation

The Becker-Döring equation of classical nucleation theory is derived in the framework of spin-flip dynamics of the Ising model. The derivation is based on a microscopic analysis of droplet geometry at low temperature.     

Nucleation near the spinodal in long-range Ising models

Properties of metastable long-range Ising models (LRIMs) are studied for deep quenches near the mean-field spinodal with Monte Carlo simulations using Glauber dynamics. The theory of spinodal-assisted nucleation is found to agree well with the data. Nucleating droplets are shown to have the same structure as large clusters in random long-range bond percolation.     

Predicting initial nucleation events occurred in a metastable nanodroplet during acoustic droplet vaporization

Acoustic droplet vaporization (ADV) capable of converting liquid perfluorocarbon (PFC) micro/nanodroplets into gaseous microbubbles has gained much attention due to its medical potentials. However, its physical mechanisms for nanodroplets have not been well understood due to the disappeared superharmonic focusing effect and the prominent Laplace pressure compared to microdroplets, especially for the initial ADV nucleation occurring in a metastable PFC nanodroplet. The classical nucleation theory (CNT) was modified to describe the ADV nucleation via combining the phase-change thermodynamics of perfluoropentane (PFP) and the Laplace pressure effect on PFP nanodroplets. The thermodynamics was exactly predicted by the Redlich–Kwong equation of state (EoS) rather than the van der Waals EoS, based on which the surface tension of the vapor nucleus as a crucial parameter in the CNT was successfully obtained to modify the CNT. Compared to the CNT, the modified CNT eliminated the intrinsic limitations of the CNT, and it predicted a larger nucleation rate and a lower ADV nucleation threshold, which agree much better with experimental results. Furthermore, it indicated that the nanodroplet properties exert very strong influences on the nucleation threshold instead of the acoustic parameters, providing a potential strategy with an appropriate droplet design to reduce the ADV nucleation threshold. This study may contribute to further understanding the ADV mechanism for PFC nanodroplets and promoting its potential theranostic applications in clinical practice.     

Microcanonical relation between continuous and discrete spin models

A relation between a class of stationary points of the energy landscape of continuous spin models on a lattice and the configurations of an Ising model defined on the same lattice suggests an approximate expression for the microcanonical density of states. Based on this approximation we conjecture that if a O(n) model with ferromagnetic interactions on a lattice has a phase transition, its critical energy density is equal to that of the n=1 case, i.e., an Ising system with the same interactions. The conjecture holds true in the case of long-range interactions. For nearest-neighbor interactions, numerical results are consistent with the conjecture for n=2 and n=3 in three dimensions. For n=2 in two dimensions (XY model) the conjecture yields a prediction for the critical energy of the Bere?inskij-Kosterlitz-Thouless transition, which would be equal to that of the two-dimensional Ising model. We discuss available numerical data in this respect.     

Monte Carlo computer experiments on critical phenomena and metastable states

Ising and Heisenberg models are studied by the Monte Carlo method. Several hundred up to 60 000 spins located at two- and three-dimensional lattices are treated and various boundary conditions used to elucidate various aspects of phase transitions. Using free boundaries the finite size scaling theory is tested and surface properties are derived, while the periodic boundary condition or the effective field-like ‘self-consistent’ boundary condition are used to derive bulk critical properties. Since Monte Carlo averages can be interpreted as time averages of a stochastic model, ‘critical slowing down of convergence’ occurs. The critical dynamics is investigated in the case of the single spin-flip kinetic Ising model. Also non-equilibrium relaxation processes are treated, e.g. switching on small negative fields the magnetization reversal and nucleation processes are studied. The metastable states found can be understood in terms of a scaling theory and the droplet model. Using a spin exchange model the phase separation kinetics of a binary alloy is simulated.     

Solution of an Ising model with two- and four-spin interactions

The Ising model on a hexagonal lattice with nearest-neighbor and certain four-spin interactions is solved exactly. The critical behaviors are the same as those of the nearest-neighbor Ising models.     

Anisotropy of the interface tension of the three-dimensional Ising model

We determine the interface tension for the 100, 110 and 111 interface of the simple cubic Ising model with nearest-neighbour interaction using novel simulation methods. To overcome the droplet/strip transition and the droplet nucleation barrier we use a newly developed combination of the multimagnetic algorithm with the parallel tempering method. We investigate a large range of inverse temperatures to study the anisotropy of the interface tension in detail.     

Simulations of spinodal nucleation in systems with elastic interactions

Systems with long-range interactions quenched into a metastable state near the pseudospinodal exhibit nucleation that is qualitatively different from classical nucleation near the coexistence curve. We observe nucleation droplets in Langevin simulations of a two-dimensional model of martensitic transformations and determine that the structure of the nucleating droplet differs from the stable martensite structure. Our results, together with experimental measurements of the phonon dispersion curve, allow us to predict the nature of the droplet. The results have implications for nucleation in many solid-solid transitions and the structure of the final state.     

Elastic Ising-like model for the nucleation and domain formation in spin crossover molecular solids

We investigate the nucleation, domain formation and propagation mechanisms observed in Spin Crossover materials, in the framework of an Ising-like model taking into account the elastic nature of the interactions. In Spin Crossover materials, the intermolecular coupling originates from a volume difference between the High Spin and the Low Spin molecular states and is simulated by anharmonic interaction potentials whose strengths are molecular-state-dependent. Using Monte Carlo methods, the phase diagram has been established. We show that the model contains both Ising short-range couplings and long-range elastic interactions. In particular, the results of long-range elastic models are reproduced. The introduction of lattice dynamics leads to the existence of spatial distributions of interaction energy and crystal field, corresponding to a local definition of physical properties. The nucleation process becomes highly dependent on the structural inhomogeneities induced by the spin transition. In this approach, connections strength between neighboring molecules are no more equivalent and have different ability to propagate domains. The presence of short-range Ising couplings gives rise to the occurrence of strong bonds forming a volume in which domains of the daughter phase can grow; in this case a macroscopic phase separation appears during the first order transition, even in a system with periodic boundary conditions. By contrast, in the case of a model with only long-range elastic interactions; strong bonds are uniformly spread in the lattice and a homogeneous phase transformation is observed, in good agreement with previous theoretical investigations.     

Dimensional reduction by Monte Carlo

A Monte Carlo method for mapping a field theoretical or statistical system to a new theory embedded in a space-time of lesser dimensionality is presented. Typically, the critical properties of the dimensionally reduced system depend upon the details of the mapping. As an example, the two-dimensional Ising model is mapped to a one-dimensional Ising model with long-range forces and a phase transition. Systems with long-range interactions and known exponents can thus be constructed with this procedure.     

Effective-field theory of the Ising model with three alternative layers on the honeycomb and square lattices

The Ising model with three alternative layers on the honeycomb and square lattices is studied by using the effective-field theory with correlations. We consider that the nearest-neighbor spins of each layer are coupled ferromagnetically and the adjacent spins of the nearest-neighbor layers are coupled either ferromagnetically or anti-ferromagnetically depending on the sign of the bilinear exchange interactions. We investigate the thermal variations of the magnetizations and present the phase diagrams. The phase diagrams contain the paramagnetic, ferromagnetic and anti-ferromagnetic phases, and the system also exhibits a tricritical behavior.     

Nonclassical behavior of the van der Waals gas

The partition function of the van der Waals gas is represented by a functional integral which is evaluated by summing the value of the integrand over its absolute and all of its secondary maxima. This leads to a one-to-one correspondence with the Ising model with nearest-neighbor interactions only. Whereas the classical behavior of the van der Waals gas is due to the absolute maximum in function space, the nonclassical behavior is shown to derive from the combined contribution of all the secondary maxima. The relation of this work to inverse range expansions and to the droplet model of condensation is discussed.     

Universal critical behavior of curvature-dependent interfacial tension

From the analysis of Monte Carlo simulations of a binary Lennard-Jones mixture in the coexistence region, we provide evidence that the curvature dependence of the interfacial tension can be described by a simple theoretical function σ(R)ξ(2)=C(1)/[1+C(2)(ξ/R)(2)], where ξ is the correlation length and R is the droplet radius. The universal constants C(1) and C(2) are estimated. In the model, a Tolman length is strictly absent, but, since its critical behavior is believed to be much weaker than ξ, we argue that it only provides a correction to scaling and does not affect the leading critical behavior, which should be described by the above function for any system in the Ising universality class. The large value of C(2)?32 implies that conventional nucleation theory becomes inaccurate even for a significantly large droplet radius.     

The low-temperature properties of the one-dimensional fluid

The low-temperature properties of a one-dimensional fluid with hard core attractive nearest-neighbor interactions have been investigated. The fluid exhibits a critical behavior nearT=0 which is, in some respects, analogous to that of the one-dimensional Ising models. With the proper choice of scaling variables the singular part of the appropriate thermodynamic potential has the same homogeneous scaling form as the Ising model. The correlations in the scaling region have a more complex structure than in the Ising model but do have a long-ranged part of scaling form. TheT=0 limit in the scaling region gives states of low density and zero pressure whose correlations are those of a two-phase state, of which one component is a perfect crystal phase and the other is a zero density phase. The positive pressureT=0 states are single-phase perfect crystal states whose long-range order develops continuously asT approaches zero. Those Fourier components of the correlations, which correspond to reciprocal lattice vectors, diverge asT approaches zero; hence, the transition is second-order, unlike higher-dimensional systems.     

Kinetic Monte Carlo simulations of electrodeposition: Crossover from continuous to instantaneous homogeneous nucleation within Avrami’s law

The influence of lateral adsorbate diffusion on the dynamics of the first-order phase transition in a two-dimensional Ising lattice gas with attractive nearest-neighbor interactions is investigated by means of kinetic Monte Carlo simulations. For example, electrochemical underpotential deposition proceeds by this mechanism. One major difference from adsorption in vacuum surface science is that under control of the electrode potential and in the absence of mass-transport limitations, local adsorption equilibrium is approximately established. We analyze our results using the theory of Kolmogorov, Johnson and Mehl, and Avrami (KJMA), which we extend to an exponentially decaying nucleation rate. Such a decay may occur due to a suppression of nucleation around existing clusters in the presence of lateral adsorbate diffusion. Correlation functions prove the existence of such exclusion zones. By comparison with microscopic results for the nucleation rate I and the interface velocity of the growing clusters v, we can show that the KJMA theory yields the correct order of magnitude for Iv². This is true even though the spatial correlations mediated by diffusion are neglected. The decaying nucleation rate causes a gradual crossover from continuous to instantaneous nucleation, which is complete when the decay of the nucleation rate is very fast on the time scale of the phase transformation. Hence, instantaneous nucleation can be homogeneous, producing negative minima in the two-point correlation functions. We also present in this paper an n-fold way Monte Carlo algorithm for a square lattice gas with adsorption/desorption and lateral diffusion.     

Stark resonances in disordered systems

We study further the metastable behavior of Metropolis dynamics for the two-dimensional nearest neighbor ferromagnetic Ising model, with positive and small external field, in the limit as the temperature vanishes (see [NS]). We focus on the typical features of the escape (nucleation) from the (metastable) configuration with all spins –1, to the (stable) configuration with all spins +1. Using the reversibility of the process as the main tool, we prove (for the discrete time version of the model) that the first step of a typical escaping path is the time reverse of a typical time evolution of a shrinking subcritical rectangular droplet, which is one slice smaller than a critical droplet. This subcritical droplet then evolves in a time of order 1 to a critical droplet, which finally grows with features described in [NS].Work partially supported by the Brazilian CNPq and by the American NSF, under grant DMS91-00725     

Critical properties of the antiferromagnetic layered Ising model on a cubic lattice with competing interactions

The critical properties of the antiferromagnetic layered Ising model on a cubic lattice with regard to the nearest-neighbor and next-nearest-neighbor interactions are investigated by the Monte Carlo method using the replica algorithm. The investigations are carried out for the ratios of exchange nearest-neighbor and next-nearest-neighbor interactions r = J ₂/J ₁ in the range of 0 ≤ r ≤ 1.0. Using the finite-size scaling theory, the static critical indices of specific heat α, order parameter β, susceptibility γ, correlation radius ν, and Fisher index η are calculated. It is shown that the universality class of the critical behavior of this model is retained in the range of 0 ≤ r ≤ 0.4. It is established that the change in the next-nearest-neighbor interaction value in this model in the range of r > 0.8 leads to the same universality class as the three-dimensional fully frustrated Ising model on the cubic lattice.

     

Wang-Landau study of the 3D Ising model with bond disorder

We implement a two-stage approach of the Wang-Landau algorithm to investigate the critical properties of the 3D Ising model with quenched bond randomness. In particular, we consider the case where disorder couples to the nearest-neighbor ferromagnetic interaction, in terms of a bimodal distribution of strong versus weak bonds. Our simulations are carried out for large ensembles of disorder realizations and lattices with linear sizes L in the range L=8-64L=8{-}64. We apply well-established finite-size scaling techniques and concepts from the scaling theory of disordered systems to describe the nature of the phase transition of the disordered model, departing gradually from the fixed point of the pure system. Our analysis (based on the determination of the critical exponents) shows that the 3D random-bond Ising model belongs to the same universality class with the site- and bond-dilution models, providing a single universality class for the 3D Ising model with these three types of quenched uncorrelated disorder.