Free Energy Distribution and Relaxation Dynamics Near the First-Order Transition Line
Ranran Guo, Xiaobing Li, Mingmei Xu, Jinghua Fu, Yuanfang Wu · 2025
Read Free Energy Distribution and Relaxation Dynamics Near the First-Order Transition Line by Ranran Guo, Xiaobing Li, Mingmei Xu, Jinghua Fu, Yuanfang Wu (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-14878.txt.
Introduction
Ranran Guo and co-authors study how a magnet-like model settles after a sudden change in a 2025 paper. The model is the three-dimensional kinetic Ising model, a grid of interacting spins used to study phase changes. The team trace the free energy landscape along the whole line where the system jumps between phases.
Near the change point, the landscape matches classic Landau-Ginzburg theory. Far below it, the landscape grows fine traps that catch random starting states and delay settling badly, an effect the authors name ultra-slow relaxation. They propose a signature for it: the spread of settling times blows up on its own. They offer that signature as a new hallmark of first-order transitions.
Catalog record
| CREATOR | Ranran Guo, Xiaobing Li, Mingmei Xu, Jinghua Fu, Yuanfang Wu |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 27 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-14878
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Free Energy Distribution and Relaxation Dynamics Near the First-Order Transition Line Ranran Guo, Xiaobing Li, Mingmei Xu, Jinghua Fu, Yuanfang Wu Abstract Using the three-dimensional kinetic Ising model with Metropolis algorithm, we calculate the free energy in the whole phase boundary, particularly near the first phase transition line (1st-PTL).The results show that along the 1st-PTL, as the temperature decreases, the energy barrier between the two coexisting phases diverges. This results in more difficulty to reach the equilibrium, i.e., ultra-slow relaxation, which has been recently demonstrated. Meanwhile, we exam the randomness of the equilibrium time. It is found that near the 1st-PTL the equilibrium time is self-diverging, in contrast to the non-self-averaging near the critical point. I Introduction Free energy is a fundamental thermodynamic state function that describes the energy available to perform work under specified conditions. It plays a crucial role in the relaxation dynamics of the system and contributes to understanding the stability, phase transition behavior, and critical phenomena of the system in equilibrium Xu and Wu 2023 ; Mori 2021 ; Kozlovskii and Romanik 2011 ; Abalmasov 2024 ; KOLESIK et al. 2003 . The principle of free energy is also widely applied in various fields, including intelligence Jose and Simeone 2021 ; Friston et al. 2023 , neuroscience Ramstead et al. 2018 ; Hansen and van Gunsteren 2014 , and biological systems Colombo and Palacios 2021 , making it an essential tool for interdisciplinary research. The calculation of free energy is primarily determined by a suitable model or Hamiltonian, a sampling protocol, and an estimator of free energy Mezei and Beveridge 1986 ; Christ et al. 2010 . To accurately capture the behavior of the system, the model or Hamiltonian must reflect the interactions between particles, external fields, and temperature dependencies. The choice of sampling protocol is crucial to generate a representative set of system configurations, ensuring that all relevant microstates are taken into account in the calculation. A variety of methodologies have been employed to compute the free energy in different Ising models, such as mean field approximation (MFA), Monte Carlo (MC) simulations, and methods based on collective variables, each offering different levels of accuracy and computational efficiency. The MFA primarily provides the distribution of the free energy per spin near the critical point and in small external fields Griffiths et al. 1966 . In the absence of an external field and below the phase transition temperature, the per spin free energy exhibits a double-well structure. Under a small positive external field, both minima shift to the right, with the right one deepening, and the left one becoming shallower. The deeper minimum corresponds to the stable state, while the shallower one represents the metastable state. Within the MC method for finite-size samples Binder 2007 , at a temperature close to zero, the restricted free energy is determined by states with minimal energy, and the barrier between its two minima with opposite polarization corresponds to states with two large domains, the interface energy of which is proportional to the sample size L . By extending the results of the collective variables method KII 1989 to the 3D Ising-like model with an external field Kozlovskii and Romanik 2011 , the Gibbs free energy and Helmholtz free energy can be calculated and establish the regions of stability, metastability, and unstability for the system under investigation. While the general properties of the free energy are well understood, a comprehensive and detailed study of the free energy characteristics across the entire phase boundary still remains. In this paper, using finite-size samples generated by the Metropolis algorithm, we demonstrate the characteristics of the free energy in the 3D Ising model under a uniform external field at different temperatures. We provide a presentation of the functional distributions of free energy and magnetization, as well as the changes in the energy barriers between the minima, at different evolution time of the 3D Ising model. The existence of coexistence states and metastable states leads to an increase in the randomness and instability of the system. In our previous work on nonequilibrium evolution Li et al. 2022 , the ultra-slow relaxation along the first-order phase transition line (1st-PTL) also reflects this characteristic. Therefore, we further investigated the distribution of free energy at different evolution times to reflect the relaxation characteristics of the system. To evaluate the influence of \tau_{eq} on the randomness of the system, we provide the self-averaging principle of \tau_{eq} along the first-order phase transition line and reveal the impact of the free energy on the self-averaging properties of the \tau_{eq} . This paper is structured as follows: Section II introduces the kinetic Ising model and calculate method for free energy. Sections III and IV present the simulation results. Section V provides a summary and discussion. II Model and method The three-dimensional Ising model considers a simple cubic lattice composed of N=L^{3} spins, where L is called the system size. The state of the system can be represented by a series of spins, i.e., k=\{s_{i}\},s_{i}=\pm 1,i=1,2,\ldots N. The total energy of the system in a uniform external field H is E_{k}=E_{\{s_{i}\}}=-J\mathop{}\sum_{<ij>}s_{i}s_{j}-H\sum_{i=1}^{N}s_{i}, (1) where J is known as the nearest-neighbor interaction constant between two spins s_{i} and s_{j} . For a series of Monte Carlo simulations of the model, {E_{k}} represents the total energy of the k^{th} configuration. The energy per spin is \varepsilon_{k}=\frac{E_{k}}{N}. (2) The partition function is Z\left(T,H\right)=\sum_{k}exp\left(-E_{k}/k_{B}T\right), (3) where k_{B} is Boltzman’s constant. The free energy is evaluated by F=-k_{B}TlnZ. (4) In the 3D Ising model, the large values of E_{k} cause the exponential function of Z to diverge as E_{k} increases. Therefore, we choose to use the per-spin average energy \varepsilon_{k} instead of the total energy E_{k} to calculate the partition function, that is Z\left(T,H\right)=\sum_{k}exp\left(-\varepsilon_{k}/k_{B}T\right). (5) And the free energy is F=-k_{B}Tln\sum_{k}exp\left(-\varepsilon_{k}/k_{B}T\right). (6) The average total magnetization is M=-\left(\frac{\partial F}{\partial H}\right)_{T}=~\left\langle\sum_{i=1}^{N}s_{i}\right\rangle, (7) and the per-spin magnetization is m=\frac{1}{N}\sum_{i=1}^{N}s_{i}. (8) It serves as the order parameter of the continuous phase transition at the critical temperature T_{c} . The most recent shows a high-precision estimate of the critical coupling for the 3D Ising model as K_{c}=J/k_{B}T_{c}=0.221654631(8) Ferrenberg et al. 2018 ; Hou et al. 2019 . Usually, J and k_{B} are set to 1 and the phase boundary is depicted by the line H=0 . Metropolis algorithm, introduced by Nicolas Metropolis Metropolis et al. 1953 , is widely used to simulate the process of a system evolving from non-equilibrium to equilibrium. Starting from an initial configuration, the Metropolis algorithm flips one single spin at each step. Whether the chosen spin flips depends on the acceptance probability A(\bm{u}\rightarrow\bm{\upsilon}) , which is given by A\left(\bm{u}\rightarrow\bm{\upsilon}\right)=\begin{cases}e^{-\left(E_{\bm{\upsilon}}-E_{\bm{u}}\right)/k_{B}T}&if~E_{\bm{\upsilon}}-E_{\bm{u}}>0,\\ 1&otherwise.\end{cases} (9) u and v represent the state of the system before and after flipping this spin. If A(\bm{u}\rightarrow\bm{\upsilon})=1 , the spin is flipped. If A(\bm{u}\rightarrow\bm{\upsilon})<1 ,a random numberr (0<r<1) is generated. If A(\bm{u}\rightarrow\bm{\upsilon})>r , the spin is flipped, otherwise, the spin keeps its original state. A Monte Carlo step refers to a single spin-flip attempt, while a sweep represents the completion of one full update cycle involving all N spins in the lattice. The system configuration is updated after each sweep. In subsequent calculations, time t is defined as the number of sweeps. The stable value of magnetization in