Volume 10, Issue 5, May – 2025 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25may414

Monte Carlo-Based Modeling of 2-D Ising

Systems Using Metropolis Algorithm, Simulation
Techniques, Thermodynamic Behavior and
Magnetization Patterns
Adama Gaye1; Omolola Dorcas, Atanda2
1
Food Safety Quality Analyst, SFC Global Supply Chain Inc (Schwan’s), Florence, KY, USA
2
Division of Mathematics, Frontier Schools, 6800 Corporate Drive, Kamsas City, MO, United States

Publication Date: 2025/05/19

Abstract: This study presents a comprehensive Monte Carlo simulation of the two-dimensional (2-D) Ising model using the
Metropolis algorithm to investigate critical phenomena and thermodynamic behavior in spin-lattice systems. The model,
implemented in MATLAB with periodic boundary conditions, explores equilibrium properties such as magnetization,
internal energy, specific heat, and susceptibility across a range of temperatures. By employing various initial spin
configurations—ordered and random—the simulations demonstrate the system's ergodicity and convergence to thermal
equilibrium. Key results include a sharp decline in magnetization and a pronounced peak in specific heat near the critical
temperature, consistent with second-order phase transition behavior. The simulation captures microscopic domain
evolution, highlighting the transition from ferromagnetic to paramagnetic phases as thermal fluctuations increase. The study
further evaluates algorithmic efficiency, discusses the impact of lattice size on statistical accuracy, and proposes
improvements using advanced cluster algorithms and parallel computing frameworks. The findings validate theoretical
predictions from Onsager's solution and underscore the versatility of Monte Carlo techniques in modeling collective
behavior in magnetic systems. The simulation framework offers a robust foundation for analyzing critical dynamics and
extends its relevance to broader applications in material science, computational physics, and complex systems modeling.

Keywords: Monte Carlo, Modeling, 2-D, Ising Systems, Metropolis Algorithm, Simulation Techniques, Thermodynamic Behavior,
Magnetization Patterns.

How to Cite: Adama Gaye. (2025). Monte Carlo-Based Modeling of 2-D Ising Systems Using Metropolis Algorithm, Simulation
Techniques, Thermodynamic Behavior and Magnetization Patterns. International Journal of Innovative Science and
Research Technology, 10(5), 511-526. https://doi.org/10.38124/ijisrt/25may414.

I. INTRODUCTION solution in 1944 established that the 2-D Ising model on a

square lattice exhibits a second-order phase transition at a
A. Background to the Ising Model and Kinetic Theory critical temperature, beyond which long-range magnetic
The Ising model has long served as a foundational ordering vanishes (Onsager, 1944). This property renders it
paradigm in statistical mechanics and condensed matter an ideal testbed for kinetic theories and numerical simulations
physics, particularly in the study of magnetic phase aiming to replicate and analyze emergent behaviors in
transitions in spin-lattice systems. Originally introduced by complex systems.
Wilhelm Lenz and rigorously explored by Ernst Ising in 1925,
the model describes a lattice of discrete variables (spins) that Modern computational methods, especially Monte
interact with their neighbors and can adopt one of two Carlo simulations using the Metropolis-Hastings algorithm,
possible states: +1 or –1 (Ising, 1925). The kinetic theory in have enabled the dynamic simulation of the Ising model,
this context offers a mesoscopic perspective, connecting capturing the stochastic nature of spin flipping and allowing
microscopic interactions to macroscopic thermodynamic the system to evolve toward thermodynamic equilibrium
quantities such as magnetization and internal energy (Landau & Binder, 2021). These simulations rely on the
(Azonuche, & Enyejo, 2024). Boltzmann probability distribution to determine the
likelihood of energy state transitions, effectively linking
In the two-dimensional (2-D) case, the Ising model statistical thermodynamics to computational models.
becomes particularly significant due to its non-trivial phase
transition behavior at a finite temperature. Onsager's exact

IJISRT25MAY414 www.ijisrt.com 511

Volume 10, Issue 5, May – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25may414
𝑁
Furthermore, the kinetic approach embedded in these 1
simulations enables the analysis of non-equilibrium processes 𝑀 = ∑ 𝜎𝑖
such as nucleation, relaxation, and critical slowing down near 𝑁
𝑖=1
phase transitions. By evaluating spin-lattice configurations at
different temperatures, researchers can extract key and the internal energy per site as:
thermodynamic parameters, including specific heat,
susceptibility, and correlation lengths (Newman & Barkema, 𝐽
𝑈=− ∑ 𝜎𝑖 𝜎𝑗
1999). The use of periodic boundary conditions, as applied in 𝑁
⟨𝑖,𝑗⟩
contemporary lattice-based kinetic models, minimizes edge
effects and mimics an infinite system, enhancing the accuracy
Where 𝜎𝑖 = ±1 represents the spin state at site 𝑖, 𝐽 is the
of numerical predictions (Wolff, 1989).
exchange interaction parameter, and the summation is over
The combination of kinetic theory and Ising-based nearest-neighbor pairs ⟨𝑖, 𝑗⟩.
modeling thus represents a powerful framework for exploring
Modern applications of MC simulations extend beyond
critical phenomena in low-dimensional systems. It provides
magnetic systems. They are employed in lattice gauge
insight not only into the fundamental behavior of magnetic
materials but also into broader domains such as neural theories, protein folding, and quantum many-body physics
networks, social behavior modeling, and quantum computing (Katzgraber, 2010). For the Ising model, they are particularly
useful in mapping phase diagrams, characterizing critical
analogues (Goldenfeld, 2018).
phenomena, and computing critical exponents with high
precision (Ferrenberg et al., 2018). Notably, MC methods
B. Significance of Monte Carlo Simulations in Statistical
Physics enable finite-size scaling analysis, which facilitates the
Monte Carlo (MC) simulations have become extrapolation of bulk properties from simulations on finite
indispensable in statistical physics, providing robust lattices—a critical advancement for validating universality
frameworks for solving problems involving stochastic hypotheses.
processes, high-dimensional integrals, and thermodynamic
Further refinement through cluster algorithms such as
averaging. Unlike deterministic methods, MC simulations
the Wolff or Swendsen-Wang methods has reduced critical
rely on probabilistic sampling to explore configuration spaces
slowing down near phase transitions, enhancing convergence
and estimate equilibrium properties of many-body systems.
speed and statistical accuracy (Wolff, 1989). These
These techniques are especially vital for studying systems
innovations underscore the transformative impact of Monte
with vast microstate ensembles, such as the Ising model,
where analytical solutions are limited or intractable in higher Carlo methods in computational statistical physics.
dimensions (Landau & Binder, 2021).
C. Objectives of the Study
A pivotal aspect of MC methods in statistical mechanics The primary objective of this study is to implement and
lies in their connection to the canonical ensemble, where the analyze the two-dimensional (2-D) Ising model using a
Monte Carlo simulation framework based on the Metropolis
probability 𝑃𝑖 of a system occupying a microstate 𝑖 with
algorithm to investigate critical thermodynamic behavior in
energy 𝐸𝑖 at temperature 𝑇 is governed by the Boltzmann
lattice-based spin systems. The simulation aims to examine
distribution:
how varying temperature regimes influence microscopic
configurations, total energy, magnetization, and phase
𝑒 −𝐸𝑖/𝑘𝐵𝑇
𝑃𝑖 = transition phenomena in square lattice structures subjected to
𝑍 periodic boundary conditions.
Here, 𝑘𝐵 is the Boltzmann constant and 𝑍 is the partition A key goal of the research is to derive numerical
function, defined as: estimates for the mean energy and magnetization at
equilibrium under different initial spin configurations—
𝑍 = ∑ 𝑒 −𝐸𝑗/𝑘𝐵𝑇 positive, negative, and random—across controlled
𝑗 temperature intervals. This investigation is also directed at
exploring the emergence of ferromagnetic and paramagnetic
Calculating 𝑍 directly is often infeasible for large phases and how temperature thresholds govern spin
systems due to the exponential growth in the number of alignment and fluctuations. The study further seeks to
microstates. Monte Carlo algorithms, such as the Metropolis- evaluate the system's susceptibility and specific heat capacity,
Hastings method, circumvent this by generating a Markov which are essential indicators of criticality and fluctuations
chain of states that sample from the equilibrium distribution, near phase transition points.
using acceptance criteria based on energy differences 𝛥𝐸
(Metropolis et al., 1953). Additionally, the simulation model is intended to
validate known theoretical predictions regarding critical
In particular, for the Ising model, each spin flip trial temperature behavior and to illustrate how stochastic
contributes to constructing time-averaged estimates of techniques can be effectively leveraged to capture emergent
macroscopic observables like magnetization 𝑀 and internal macroscopic properties from microscopic dynamics. A
energy 𝑈. The magnetization per site is defined as: secondary objective is to assess the efficiency and

IJISRT25MAY414 www.ijisrt.com 512

Volume 10, Issue 5, May – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25may414
convergence behavior of the implemented Metropolis Monte second-order phase transition at a finite critical temperature
Carlo algorithm, including its scalability with respect to 𝑇𝑐 (Onsager, 1944). The critical temperature for a square
lattice size and computation time. The findings are expected lattice with nearest-neighbor interaction strength 𝐽 is defined
to provide insights applicable to a broader class of statistical by:
mechanical systems and inform improvements in
computational physics methodologies. 2𝐽
𝑘𝐵 𝑇𝑐 =
ln(1 + √2)
D. Scope and Structure of the Paper
This research paper is confined to the computational
Where 𝑘𝐵 is the Boltzmann constant. This exact result
modeling and thermodynamic analysis of the two-
was pivotal, establishing a benchmark for both analytical and
dimensional Ising model on a finite square lattice using the numerical investigations into critical phenomena and
Metropolis Monte Carlo method. The scope specifically spontaneous symmetry breaking in lattice systems.
focuses on spin-½ systems with nearest-neighbor interactions
and periodic boundary conditions, enabling an approximation The Hamiltonian governing the Ising model in its
of infinite system behavior within a tractable numerical classical form is expressed as:
framework. The study does not consider external magnetic
fields or interactions beyond the nearest-neighbor
approximation, thereby isolating intrinsic temperature- ℋ = −𝐽 ∑ 𝜎𝑖 𝜎𝑗 − ℎ ∑ 𝜎𝑖
dependent effects on system energetics and magnetization. ⟨𝑖,𝑗⟩ 𝑖

The analysis is restricted to lattice dimensions favoring Where ⟨𝑖, 𝑗⟩ indicates summation over nearest-neighbor
computational feasibility while maintaining statistical pairs, ℎ is the external magnetic field, and 𝜎𝑖 denotes the spin
relevance, with simulations conducted across a range of sub- at site 𝑖. In the absence of an external field (ℎ = 0), the model
critical and super-critical temperatures. Only canonical simplifies, and the system's behavior becomes purely
ensemble dynamics are evaluated, and quantum mechanical temperature-driven.
effects, three-dimensional lattice extensions, and alternative
spin models fall outside the boundaries of this investigation. The Ising model’s success lies in its abstraction and
versatility. Beyond ferromagnetism, it has been extended to
The remainder of this paper is structured as follows: model phenomena in fields ranging from neuroscience and
Section 2 presents a review of relevant literature on the Ising sociophysics to computational biology and quantum
model, Monte Carlo simulation techniques, and information. Despite its simplicity, the model encapsulates
thermodynamic observables. Section 3 details the simulation criticality, universality, and collective behavior, making it a
methodology, including initialization protocols, algorithmic canonical example for exploring phase transitions in
design, and model assumptions. Section 4 discusses the equilibrium statistical mechanics (Yeomans, 1992).
numerical results, providing visual and statistical
interpretations of energy fluctuations, magnetization trends, B. Applications of the Metropolis Algorithm in Physics
and phase transitions. Finally, Section 5 offers conclusions Simulations
and targeted recommendations for future extensions of the The Metropolis algorithm remains one of the most
work, including potential improvements in algorithmic influential stochastic methods in computational physics,
efficiency and applicability to more complex physical particularly in simulating systems governed by statistical
systems. mechanics. Originally developed by Metropolis et al. (1953)
for simulating the behavior of particles in a fluid, the
II. LITERATURE REVIEW algorithm introduced a probabilistic rule for accepting or
rejecting state transitions based on energy differences. This
A. Historical Development of the Ising Model method has since been adapted to lattice models like the Ising
The Ising model, one of the cornerstones of modern model to simulate thermal equilibrium configurations at a
statistical mechanics, originated from the early attempts to fixed temperature 𝑇.
understand ferromagnetism at the microscopic level.
Introduced by Wilhelm Lenz in 1920 and further developed At the heart of the Metropolis algorithm lies the
by his student Ernst Ising in 1925, the model proposed a principle of detailed balance, which ensures that the Markov
lattice-based system of binary variables (spins), each taking a chain converges toward the Boltzmann distribution. For any
value 𝜎𝑖 = ±1, to mimic magnetic dipole moments in a two microstates 𝑖 and 𝑗, the detailed balance condition is
crystalline solid (Ising, 1925). Although Ising initially satisfied if:
analyzed the one-dimensional (1D) system and concluded
that it exhibited no phase transition at non-zero temperature, 𝑃(𝑗 → 𝑖) 𝐸𝑖 − 𝐸𝑗
this limitation was soon overcome by higher-dimensional = exp (− )
𝑃(𝑖 → 𝑗) 𝑘𝐵 𝑇
generalizations.
Where 𝐸𝑖 and 𝐸𝑗 are the energies of states 𝑖 and 𝑗,
The two-dimensional (2D) variant of the Ising model respectively. The transition probability is defined such that
gained prominence with the exact solution by Lars Onsager moves to lower energy states are always accepted, while
in 1944. Onsager demonstrated that the 2D Ising model on a
square lattice with zero external magnetic field undergoes a

IJISRT25MAY414 www.ijisrt.com 513

Volume 10, Issue 5, May – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25may414
higher energy moves are accepted with probability Where 𝝂 is the critical exponent characterizing the
exp(−𝛥𝐸/𝑘𝐵 𝑇), where 𝛥𝐸 = 𝐸𝑖 − 𝐸𝑗 . divergence. Accompanying this divergence are singularities
in other thermodynamic observables, such as the specific heat
In the context of the Ising model, each Monte Carlo step 𝑪𝒗 , magnetic susceptibility 𝝌, and magnetization 𝑴, each
involves selecting a random lattice site, flipping the spin, and described by their respective critical exponents 𝜶, 𝜸, and 𝜷:
evaluating the energy difference 𝛥𝐸. If the flip leads to a
decrease in the system's energy, it is accepted; otherwise, it is 𝑪𝒗 ∼∣ 𝑻 − 𝑻𝒄 ∣−𝜶 , 𝝌 ∼∣ 𝑻 − 𝑻𝒄 ∣−𝜸 , 𝑴 ∼ (𝑻𝒄 − 𝑻)𝜷
accepted with the aforementioned Boltzmann-weighted
probability. This facilitates an efficient exploration of the Extensive Monte Carlo simulations have been employed
system’s configuration space while respecting to validate these power-law relations and extract critical
thermodynamic constraints (Landau & Binder, 2021). exponents with high numerical precision (Ferrenberg et al.,
2018). Finite-size scaling theory has further enhanced the
The Metropolis algorithm has been pivotal in studying critical accuracy of these estimates by accounting for system-size-
behavior, where observables such as specific heat 𝑪𝒗 , dependent rounding and shifting of transition signatures.
magnetic susceptibility 𝝌, and order parameters exhibit According to finite-size scaling, the peak value of the
divergence or sharp transitions. For instance, specific heat is susceptibility 𝝌𝒎𝒂𝒙 scales with lattice size 𝑳 as:
computed as:
𝝌𝒎𝒂𝒙 (𝑳) ∼ 𝑳𝜸/𝝂
𝟏
𝑪𝒗 = (⟨𝑬𝟐 ⟩ − ⟨𝑬⟩𝟐 )
𝒌𝑩 𝑻𝟐 Similarly, the temperature at which the peak occurs
shifts as:
and magnetic susceptibility is given by:
𝑻𝒄 (𝑳) = 𝑻𝒄 + 𝒂𝑳−𝟏/𝝂
𝟏
𝝌= (⟨𝑴𝟐 ⟩ − ⟨𝑴⟩𝟐 ) Such methodologies have been systematically applied to
𝒌𝑩 𝑻
not only the Ising model but also the XY, Potts, and
These fluctuations provide insight into second-order phase Heisenberg models, each revealing unique universality
transitions and are extracted from ensemble averages over characteristics while adhering to scaling hypotheses (Cardy,
millions of MC steps. 1996).

Beyond magnetic systems, the Metropolis algorithm has Moreover, research by Binder and Luijten (2001) has
found applications in polymer dynamics, quantum Monte emphasized the importance of order parameter distributions
Carlo simulations, and optimization problems such as and Binder cumulants in pinpointing 𝑻𝒄 , especially in
simulated annealing. Its flexibility and ease of simulations of finite systems. These cumulants provide
implementation have made it a cornerstone in diverse fields dimensionless quantities that intersect at 𝑻𝒄 for different
such as materials science, lattice gauge theory, and biophysics system sizes, offering a robust estimator free from non-
(Kalos & Whitlock, 2008; Newman & Barkema, 1999). universal amplitudes.

C. Prior Research on Phase Transitions in Lattice-Based Modern computational studies have also explored more
Models complex interactions such as long-range coupling, quenched
The study of phase transitions in lattice-based systems disorder, and anisotropy, extending the understanding of
has been a central theme in statistical physics, offering criticality into disordered and frustrated systems (Anyibama,
fundamental insights into critical behavior and universality et al., 2025). Notably, 3-D extensions and quantum lattice
classes. The Ising model, in particular, has served as a models have uncovered novel quantum phase transitions,
primary vehicle for analyzing second-order (continuous) where quantum fluctuations dominate thermal ones (Sachdev,
phase transitions due to its analytically solvable 2-D case and 2011).
rich phenomenology. Onsager’s exact solution for the 2-D
Ising model revealed that the system undergoes a spontaneous D. Thermodynamic Properties of Magnetic Systems in 2-D
symmetry-breaking transition at a finite critical temperature Models
𝑻𝒄 , marking the boundary between ferromagnetic and The thermodynamic behavior of magnetic systems in
paramagnetic phases (Onsager, 1944). two-dimensional (2-D) lattice models, particularly the Ising
model, has been extensively explored to elucidate phenomena
One of the hallmark features of continuous phase such as spontaneous magnetization, energy fluctuations, and
transitions in such systems is the divergence of correlation critical behavior. In these systems, key macroscopic
length 𝝃 as the temperature approaches 𝑻𝒄 . Near the critical observables—internal energy 𝑼, magnetization 𝑴, specific
point, the correlation length exhibits a power-law behavior heat 𝑪𝒗 , and magnetic susceptibility 𝝌—are derived from
given by: microscopic spin interactions governed by a Hamiltonian of
the form:
𝝃 ∼∣ 𝑻 − 𝑻𝒄 ∣−𝝂
𝓗 = −𝑱 ∑ 𝝈𝒊 𝝈𝒋 − 𝒉 ∑ 𝝈𝒊
⟨𝒊,𝒋⟩ 𝒊

IJISRT25MAY414 www.ijisrt.com 514

Volume 10, Issue 5, May – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25may414
Where 𝝈𝒊 = ±𝟏 represents the spin at site 𝒊, 𝑱 is the critical exponents. Its thermodynamic richness continues to
coupling constant, 𝒉 is the external magnetic field, and the inform broader studies on collective behavior in spin glasses,
summation is over nearest-neighbor interactions. In the quantum spin systems, and topological phases (Domb, 1996).
absence of an external field (𝒉 = 𝟎), the Hamiltonian
simplifies, allowing a direct analysis of thermal fluctuations E. Computational Strategies in Stochastic Simulations
and collective behavior. Stochastic simulations, particularly those used to model
thermodynamic systems, rely heavily on efficient
The internal energy per site, 𝒖 = 𝑼/𝑵, where 𝑵 is the computational strategies to ensure both statistical accuracy
total number of lattice sites, can be computed as the average and numerical stability. The Metropolis Monte Carlo
energy over Monte Carlo configurations. Similarly, the mean algorithm, despite its conceptual simplicity, demands
magnetization per site is given by: significant computational optimization to effectively sample
the configuration space of large lattice systems, especially
𝟏 near criticality where correlation lengths diverge and
𝒎= ⟨∑ 𝝈𝒊 ⟩ autocorrelation times increase (Landau & Binder, 2021).
𝑵
𝒊
A fundamental element of computational strategy in
This quantity serves as an order parameter, such simulations is the construction of an efficient Markov
distinguishing between ordered (ferromagnetic) and Chain Monte Carlo (MCMC) sequence that satisfies
disordered (paramagnetic) phases. At low temperatures, ergodicity and detailed balance. The Metropolis-Hastings
spontaneous symmetry breaking leads to non-zero method, widely used in Ising-type models, proposes a new
magnetization, while at high temperatures, thermal agitation state 𝒔′ from a current state 𝒔 and accepts it with a probability
randomizes spin orientations, resulting in 𝒎 → 𝟎 (McCoy & 𝒑 given by:
Wu, 2014).
𝝅(𝒔′ )𝒒(𝒔 ∣ 𝒔′ )
Thermal response functions such as the specific heat and 𝒑 = 𝐦𝐢𝐧 (𝟏, )
𝝅(𝒔)𝒒(𝒔′ ∣ 𝒔)
magnetic susceptibility provide deeper insight into critical
behavior. The specific heat is defined via the fluctuation-
Where 𝝅 is the target distribution (Boltzmann
dissipation theorem as:
distribution in physics applications) and 𝒒 is the proposal
𝟏 distribution. In the original Metropolis formulation with
𝑪𝒗 = (⟨𝑬𝟐 ⟩ − ⟨𝑬⟩𝟐 ) symmetric proposals, this reduces to:
𝒌𝑩 𝑻𝟐
𝒑 = 𝐦𝐢𝐧(𝟏, 𝒆−𝜟𝑬/𝒌𝑩𝑻 )
and exhibits a logarithmic divergence in the 2-D Ising
model near the critical temperature 𝑻𝒄 , a hallmark of second-
order phase transitions (Baxter, 2016). The magnetic An important optimization in spin systems is
susceptibility 𝝌, which quantifies the response of the precomputing the Boltzmann factors 𝒆−𝜟𝑬/𝒌𝑩 𝑻 for all possible
magnetization to an infinitesimal field, is similarly expressed energy changes 𝜟𝑬, which greatly reduces runtime overhead.
as: Moreover, the use of bit-coded spin states and lookup tables
further accelerates spin updates (Newman & Barkema, 1999).
𝟏
𝝌= (⟨𝑴𝟐 ⟩ − ⟨𝑴⟩𝟐 ) To overcome critical slowing down near phase
𝒌𝑩 𝑻
transitions, where the autocorrelation time 𝝉 diverges as:
Near 𝑻𝒄 , both 𝑪𝒗 and 𝝌 show divergent or peak-like
𝝉 ∼ 𝝃𝒛
behavior that becomes sharper as the system size increases,
consistent with finite-size scaling predictions (Pelissetto &
Vicari, 2002). with 𝝃 being the correlation length and 𝒛 the dynamic
critical exponent, advanced algorithms such as cluster
Another important thermodynamic aspect is the free updates—including the Swendsen-Wang and Wolff
algorithms—are employed. These non-local methods flip
energy 𝑭, which encapsulates all equilibrium properties.
Although it cannot be directly computed in Monte Carlo entire correlated clusters of spins rather than individual ones,
thereby dramatically improving decorrelation rates (Wolff,
simulations, it is indirectly estimated via thermodynamic
1989; Swendsen & Wang, 1987).
integration or by studying derivatives such as entropy and
internal energy across temperature sweeps (Binder, 1981).
Another crucial component is finite-size scaling
These thermodynamic observables not only characterize
analysis, which provides a framework to extrapolate infinite-
equilibrium phases but also allow for the identification of
system behavior from simulations on finite lattices. The
phase boundaries and universality classes within broader
statistical mechanical systems. scaling relations for observables such as magnetization 𝑴 and
susceptibility 𝝌 as a function of lattice size 𝑳 near the critical
The 2-D Ising model remains a paradigmatic system for temperature 𝑻𝒄 are given by:
testing theoretical predictions against numerical simulations,
providing exact benchmarks for energy, magnetization, and 𝑴 ∼ 𝑳−𝜷/𝝂 , 𝝌 ∼ 𝑳𝜸/𝝂

IJISRT25MAY414 www.ijisrt.com 515

Volume 10, Issue 5, May – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25may414
Parallelization techniques, including domain Due to the exponential growth in the number of possible
decomposition and replica parallelism, have also become spin configurations (𝟐𝑳×𝑳 ), exact evaluation of 𝒁 is
standard for accelerating large-scale simulations. Modern computationally intractable for large systems. This
implementations utilize GPU acceleration and multi-threaded necessitates the use of stochastic numerical techniques, such
architectures to simultaneously evolve multiple replicas or as Monte Carlo simulations, to sample the configuration
lattice subregions, thus enhancing sampling and reducing space and evaluate ensemble averages of observables like
convergence time (Preis et al., 2009). internal energy 𝑼, magnetization 𝑴, specific heat 𝑪𝒗 , and
susceptibility 𝝌 (Newman & Barkema, 1999).
Moreover, post-simulation techniques such as Jackknife
resampling and binning analysis are used to quantify To initialize the model, three types of spin
statistical uncertainty, eliminate autocorrelation bias, and configurations are considered: all-up (𝝈𝒊 = +𝟏), all-down
compute reliable estimates of observables and their variances (𝝈𝒊 = −𝟏), and random (𝝈𝒊 = ±𝟏 with equal probability).
(Azonuche, & Enyejo, 2024). These computational strategies These initial states evolve dynamically through spin-flip trials
form a backbone for accurate and efficient stochastic governed by the Metropolis algorithm, allowing the system to
modeling in condensed matter and statistical physics. reach thermal equilibrium across a range of temperatures
(Vojta, 2003).
III. METHODS
This Hamiltonian-based formulation, while
A. Model Description and Hamiltonian Formulation conceptually simple, captures the essential physics of order-
The two-dimensional (2-D) Ising model forms the basis disorder transitions and facilitates the numerical exploration
for understanding phase transitions and critical phenomena in of criticality, making it a foundational construct in statistical
discrete spin systems. In this study, the model is implemented mechanics.
on a square lattice of linear size 𝑳, with each site 𝒊 occupied
by a spin variable 𝝈𝒊 ∈ {+𝟏, −𝟏}. The spins interact with B. Monte Carlo Simulation Setup and Assumptions
their nearest neighbors, and the system is constrained using Monte Carlo (MC) simulations offer a probabilistic
periodic boundary conditions, effectively mapping the lattice approach to studying thermodynamic behavior in spin lattice
onto a torus. This eliminates boundary artifacts and preserves systems where analytical solutions are limited. The
translational symmetry, enabling a more accurate implementation in this study utilizes the Metropolis algorithm
representation of bulk behavior (Landau & Binder, 2021). to simulate a two-dimensional (2-D) Ising lattice of
dimension 𝑳 × 𝑳, with spin variables 𝝈𝒊 ∈ {+𝟏, −𝟏} assigned
The system’s energy configuration is described by the to each lattice site. The simulation domain is constructed
Ising Hamiltonian, which accounts for pairwise interactions using periodic boundary conditions, effectively simulating an
between neighboring spins and optionally an external infinite system by wrapping edges such that each site has four
magnetic field 𝒉. The general form of the Hamiltonian is: neighbors regardless of its location on the grid (Landau &
Binder, 2021).
𝓗 = −𝑱 ∑ 𝝈𝒊 𝝈𝒋 − 𝒉 ∑ 𝝈𝒊
⟨𝒊,𝒋⟩ 𝒊
The simulation evolves through a sequence of Monte
Carlo steps per spin (MCSS). One MCSS constitutes 𝑳𝟐
Here, 𝑱 represents the interaction strength between attempted spin flips, ensuring each spin has, on average, one
adjacent spins. For ferromagnetic coupling, 𝑱 > 𝟎, opportunity to flip per step. The probability of accepting a
encouraging parallel alignment of spins. The summation ⟨𝒊, 𝒋⟩ proposed spin flip is derived from the Metropolis acceptance
is taken over all nearest-neighbor spin pairs on the lattice. The criterion:
second term, involving 𝒉, introduces a global magnetic field
that biases spin alignment, but in this study, 𝒉 is set to zero to 𝑷accept = 𝐦𝐢𝐧(𝟏, 𝒆−𝜟𝑬/𝒌𝑩𝑻 )
isolate temperature-driven ordering effects.
Where 𝜟𝑬 is the change in energy resulting from the
In thermal equilibrium, the probability of the system spin flip, 𝑻 is the absolute temperature, and 𝒌𝑩 is the
being in a particular spin configuration {𝝈𝒊 } is governed by Boltzmann constant. If 𝜟𝑬 ≤ 𝟎, the move is always accepted.
the Boltzmann distribution: Otherwise, it is accepted with the probability above, which
ensures detailed balance and convergence to the Boltzmann
𝟏 𝓗({𝝈𝒊 }) distribution (Metropolis et al., 1953).
𝑷({𝝈𝒊 }) = 𝐞𝐱𝐩 (− )
𝒁 𝒌𝑩 𝑻
 Initial Spin Configurations are Randomized or Fully
Aligned Depending on the Experimental Condition Under
Where 𝒌𝑩 is the Boltzmann constant, 𝑻 is the absolute
Investigation:
temperature, and 𝒁 is the partition function:
 Positive initialization (𝝈𝒊 = +𝟏)
𝓗({𝝈𝒊 })
𝒁 = ∑ 𝐞𝐱𝐩 (− )  Negative initialization (𝝈𝒊 = −𝟏)
𝒌𝑩 𝑻  Random initialization (𝝈𝒊 = ±𝟏 with equal probability)
{𝝈𝒊 }

IJISRT25MAY414 www.ijisrt.com 516

Volume 10, Issue 5, May – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25may414
These initial states enable the exploration of ergodicity The implementation proceeds by initializing a spin
and the system’s ability to evolve towards equilibrium lattice 𝝈 ∈ {−𝟏, +𝟏}𝑳×𝑳 , where 𝑳 is the linear size of the
regardless of starting configuration. Simulations are lattice. The system is updated iteratively by proposing spin
performed across a temperature sweep 𝑻 ∈ [𝟏. 𝟎, 𝟒. 𝟎], with flips at randomly selected lattice sites. The change in energy
fine temperature increments (e.g., 𝜟𝑻 = 𝟎. 𝟏) to accurately 𝜟𝑬 associated with a proposed spin flip is calculated using
capture thermodynamic transitions, especially near the the Ising model Hamiltonian (no external field):
known critical point 𝑻𝒄 ≈ 𝟐. 𝟐𝟔𝟗 for the 2-D Ising model
without external fields (Onsager, 1944). 𝜟𝑬 = 𝟐𝑱𝝈𝒊,𝒋 (𝝈𝒊+𝟏,𝒋 + 𝝈𝒊−𝟏,𝒋 + 𝝈𝒊,𝒋+𝟏 + 𝝈𝒊,𝒋−𝟏 )

To enhance statistical robustness, thermal averaging is Here, periodic boundary conditions are applied to
applied after an equilibration phase, typically consisting of manage edge interactions using modular arithmetic. The spin
𝟏𝟎𝟓 MCSS, beyond which observable quantities such as 𝝈𝒊,𝒋 is flipped with probability:
magnetization 𝑴, energy 𝑬, specific heat 𝑪𝒗 , and
susceptibility 𝝌 are averaged over an additional 𝟏𝟎𝟔 MCSS. 𝟏 if 𝜟𝑬 ≤ 𝟎
The magnetization per spin is computed as: 𝑷accept = { 𝜟𝑬
𝐞𝐱𝐩 (− ) if 𝜟𝑬 > 𝟎
𝒌𝑩 𝑻
𝑳𝟐
𝟏
𝑴 = 𝟐 ∑ 𝝈𝒊 This condition preserves detailed balance and ensures
𝑳
𝒊=𝟏 convergence to the canonical ensemble distribution
(Metropolis et al., 1953; Binder & Heermann, 2010).
While the energy per spin is:
In the MATLAB implementation, the random spin site
𝑱 is selected using randi, and neighbors are retrieved with a
𝑬 = − 𝟐 ∑ 𝝈𝒊 𝝈𝒋
𝑳 user-defined Neighbor function that maps lattice coordinates
⟨𝒊,𝒋⟩
via periodic indices. The energy and magnetization are
updated dynamically using global variables to track state
Fluctuation-based estimates of thermodynamic
quantities across iterations. Energy per spin is computed as:
quantities follow:

𝟏 𝟏 𝑱
(⟨𝑬𝟐 ⟩ − ⟨𝑬⟩𝟐 ), (⟨𝑴𝟐 ⟩ − ⟨𝑴⟩𝟐 ) 𝑬=− ∑ 𝝈𝒊 𝝈𝒋
𝑪𝒗 = 𝝌= 𝑳𝟐
𝒌𝑩 𝑻𝟐 𝒌𝑩 𝑻 ⟨𝒊,𝒋⟩

The random number generator used in the Metropolis and the total magnetization as:
algorithm is selected for uniformity and long periodicity,
ensuring statistically independent configurations. 𝑳𝟐
Furthermore, binning analysis is employed to estimate the 𝟏
𝑴 = 𝟐 ∑ 𝝈𝒊
variance of computed quantities and eliminate autocorrelation 𝑳
𝒊=𝟏
effects (Newman & Barkema, 1999).
The MATLAB code incorporates control logic to
Advanced implementations incorporate temperature differentiate between the equilibration phase and the
parallelization, where different processors simulate measurement phase. During equilibration (typically 𝟏𝟎𝟓
independent replicas at distinct temperatures, enabling Monte Carlo steps), observable quantities are not recorded to
simultaneous data generation and more efficient computation allow the system to stabilize. In the measurement phase,
(Preis et al., 2009). All simulations are conducted in statistical quantities such as ⟨𝑴⟩, ⟨𝑴𝟐 ⟩, ⟨𝑬⟩, and ⟨𝑬𝟐 ⟩ are
MATLAB, utilizing matrix-vectorized operations for calculated over successive iterations to compute specific heat
efficient handling of spin arrays and periodic indexing. 𝑪𝒗 and susceptibility 𝝌 via fluctuation formulas (Ferrenberg
et al., 2018).
C. Metropolis Algorithm Implementation in MATLAB
The Metropolis algorithm is one of the most widely To enhance computational efficiency, precomputed
adopted Monte Carlo methods for simulating statistical tables of 𝐞𝐱𝐩(−𝜟𝑬/𝒌𝑩 𝑻) values for all possible 𝜟𝑬 ∈
systems in equilibrium. It provides an efficient mechanism {−𝟖𝑱, −𝟒𝑱, 𝟎, 𝟒𝑱, 𝟖𝑱} are stored and referenced in real time.
for sampling the Boltzmann distribution in high-dimensional This avoids repeated exponential evaluations and speeds up
configuration spaces, such as those encountered in the 2-D the simulation significantly (Newman & Barkema, 1999).
Ising model. In this study, the algorithm is implemented in
MATLAB due to its matrix computation capabilities and ease Furthermore, graphical output functions such as
of managing lattice-based structures. MATLAB’s inherent imagesc and plot are used to render spin configurations and
support for vectorized operations and graphical rendering thermodynamic observables in real time. MATLAB's
further facilitates efficient iteration, data analysis, and vectorized matrix operations also allow for the simultaneous
visualization (Higham & Higham, 2017). updating of lattice elements when parallelizing operations
across temperature sweeps or initial conditions (Preis et al.,
2009).

IJISRT25MAY414 www.ijisrt.com 517

Volume 10, Issue 5, May – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25may414
𝒏 𝒏
This rigorous implementation of the Metropolis 𝟏 𝟏
algorithm in MATLAB provides a scalable and modular ⟨𝑬⟩ = ∑ 𝑬𝒕 , ⟨𝑬 ⟩ = ∑ 𝑬𝟐𝒕
𝟐
𝒏 𝒏
platform for simulating the dynamics and equilibrium 𝒕=𝟏 𝒕=𝟏
properties of spin lattice systems with high statistical
accuracy. Correct initialization of these global variables ensures
accurate tracking of time-averaged observables and facilitates
D. Initialization Parameters and Boundary Conditions post-simulation computation of specific heat and magnetic
The initialization of spin configurations and the susceptibility using fluctuation-dissipation relations (Binder
imposition of boundary conditions are critical for achieving & Heermann, 2010).
statistical reliability and physical realism in lattice-based
simulations. In the context of the two-dimensional (2-D) Ising The consistent application of these initialization
model, initialization parameters dictate the system's entry parameters and boundary conditions establishes a robust
point into the phase space, while boundary conditions foundation for exploring thermodynamic behavior and phase
determine how local interactions are treated at the edges of transitions in the 2-D Ising lattice under various temperature
the lattice (Landau & Binder, 2021). regimes and interaction scenarios (Wang & Swendsen, 1990;
Preis et al., 2009).
Three distinct initialization strategies are employed in
this study to investigate the influence of initial order on the E. Temperature Variation Scenarios and Sampling Process
system's evolution toward equilibrium: Temperature plays a central role in determining the
equilibrium properties and phase behavior of spin systems. In
 Ordered Positive Initialization: 𝝈𝒊,𝒋 = +𝟏 for all 𝒊, 𝒋, Monte Carlo simulations of the 2-D Ising model, temperature
representing a fully magnetized ferromagnetic state. variation is employed to drive the system through different
thermodynamic regimes—ranging from ordered
 Ordered Negative Initialization: 𝝈𝒊,𝒋 = −𝟏 for all 𝒊, 𝒋,
(ferromagnetic) to disordered (paramagnetic) phases. The
simulating an anti-ferromagnetic ground state.
simulation's ability to capture critical behavior hinges on
 Random Initialization: 𝝈𝒊,𝒋 ∈ {−𝟏, +𝟏} with equal precise temperature control and sufficient statistical sampling
probability 𝑷(𝝈 = +𝟏) = 𝑷(𝝈 = −𝟏) = 𝟎. 𝟓, producing across each temperature point (Landau & Binder, 2021).
a disordered paramagnetic state.
In this study, simulations are conducted over a
The diversity of these initial conditions enables temperature range 𝑻 ∈ [𝟏. 𝟎, 𝟒. 𝟎], with increments 𝜟𝑻 =
verification of ergodicity, whereby the long-time averages of 𝟎. 𝟏, to investigate both low-temperature ordering and high-
observables are independent of the starting configuration, temperature disordering regimes. The critical temperature 𝑻𝒄
given sufficient thermalization (Newman & Barkema, 1999). for the 2-D Ising model on a square lattice, in the absence of
an external magnetic field, is known analytically from
To eliminate finite-size edge effects, periodic boundary Onsager’s solution:
conditions (PBCs) are applied to all four lattice edges. In this
configuration, the lattice is topologically transformed into a 𝟐𝑱
torus, allowing the spins at the boundaries to interact with 𝑻𝒄 = ≈ 𝟐. 𝟐𝟔𝟗 (for 𝑱 = 𝟏, 𝒌𝑩 = 𝟏)
those on the opposite edges. Mathematically, PBCs are 𝒌𝑩 𝐥𝐧(𝟏 + √𝟐)
defined as:
This allows for accurate comparison of simulation data
𝝈𝒊,𝑳+𝟏 = 𝝈𝒊,𝟏 , 𝝈𝑳+𝟏,𝒋 = 𝝈𝟏,𝒋 , 𝝈𝒊,𝟎 = 𝝈𝒊,𝑳 , 𝝈𝟎,𝒋 = 𝝈𝑳,𝒋 with theoretical predictions (Onsager, 1944; Baxter, 2016).

To ensure thermodynamic consistency, each simulation

This approach maintains spatial homogeneity and
at a given temperature begins with an equilibration phase,
translational symmetry, which are crucial for accurate
during which the system is allowed to evolve without
computation of bulk thermodynamic properties such as
energy per spin 𝑬, magnetization 𝑴, and susceptibility 𝝌 recording observables. This phase typically lasts for 𝟏𝟎𝟓 to
(Ferrenberg et al., 2018). 𝟏𝟎𝟔 Monte Carlo steps per spin (MCSS), allowing the system
to relax into a temperature-dependent equilibrium state.
The lattice size 𝑳 significantly influences the Convergence is assessed by monitoring time series of global
simulation's resolution of critical phenomena. In this study, observables such as energy and magnetization until they
values of 𝑳 = 𝟓𝟎, 𝟏𝟎𝟎, 𝟐𝟎𝟎 are tested to observe finite-size exhibit stable fluctuations around mean values (Newman &
effects and facilitate scaling analysis. The total number of Barkema, 1999).
spins in the system is 𝑵 = 𝑳𝟐 , and the simulation time is
Following equilibration, a sampling phase is initiated,
expressed in terms of Monte Carlo steps per spin (MCSS),
during which the following thermodynamic quantities are
with typical values ranging from 𝟏𝟎𝟓 to 𝟏𝟎𝟕 to ensure
recorded over 𝟏𝟎𝟔 MCSS for statistical averaging:
convergence.

Initialization also includes setting global observables to

zero before the simulation loop begins. These include energy
𝑬 = 𝟎, magnetization 𝑴 = 𝟎, and their corresponding mean
and squared values used in calculating fluctuations:

IJISRT25MAY414 www.ijisrt.com 518

Volume 10, Issue 5, May – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25may414
Average energy per spin: To mitigate the effect of autocorrelation, binning
analysis and Jackknife resampling are employed to compute
𝒏
𝟏 statistical errors in measured observables. This ensures
⟨𝑬⟩ = ∑ 𝑬𝒕 reliable variance estimation, especially important when
𝒏
𝒕=𝟏 measuring quantities derived from higher moments such as
𝑪𝒗 and 𝝌 (Ferrenberg et al., 2018).
Average magnetization per spin:
Finally, the temperature-dependent results are
𝒏
𝟏 aggregated into plots of thermodynamic quantities versus 𝑻.
⟨𝑴⟩ = ∑ 𝑴𝒕 These include ⟨𝑴(𝑻)⟩, ⟨𝑬(𝑻)⟩, 𝑪𝒗 (𝑻), and 𝝌(𝑻), from which
𝒏
𝒕=𝟏 critical behavior, symmetry breaking, and scaling laws are
Energy and magnetization fluctuations: inferred (Binder & Heermann, 2010).
𝟏 𝟏
𝑪𝒗 = (⟨𝑬𝟐 ⟩ − ⟨𝑬⟩𝟐 ), 𝝌= (⟨𝑴𝟐 ⟩ − ⟨𝑴⟩𝟐 ) IV. RESULTS AND DISCUSSION
𝒌𝑩 𝑻𝟐 𝒌𝑩 𝑻
A. Microscopic Configuration at T = 2.0 and T = 2.5
These fluctuations are used to estimate specific heat 𝑪𝒗 The microscopic configuration of spin systems at fixed
and magnetic susceptibility 𝝌, respectively, which are temperatures reveals the local ordering and dynamical
expected to peak near 𝑻𝒄 due to critical fluctuations. alignment behavior of the lattice. In the case of the two-
dimensional Ising model, examining snapshots at
Temperature sweep simulations are conducted serially representative temperatures—specifically below and near the
for each 𝑻, with each instance using either a freshly initialized critical temperature (𝑻𝒄 ≈ 𝟐. 𝟐𝟔𝟗)—can elucidate critical
lattice or the final configuration of the previous temperature transition dynamics. For this study, temperatures 𝑻 = 𝟐. 𝟎
as a starting point (hot-start or cold-start). The hot-start and 𝑻 = 𝟐. 𝟓 are selected to represent a quasi-ordered
protocol, where configurations from higher 𝑻 are used for (ferromagnetic) and a quasi-disordered (paramagnetic)
lower 𝑻, is particularly useful in traversing the vicinity of 𝑻𝒄 regime, respectively.
due to reduced correlation times (Wolff, 1989).

Fig 1: Microscopic Configuration Analysis at Varying Temperatures

At 𝑻 = 𝟐. 𝟎, the system exhibits strong spontaneous Conversely, at 𝑻 = 𝟐. 𝟓, the thermal agitation surpasses
magnetization, as seen in the high average magnetization per the ordering tendency of spin interactions. As a result, the
spin of approximately 0.98. The majority of the lattice spins magnetization drops significantly to 0.12, and the lattice
are aligned, forming large coherent spin domains. This shows randomly oriented spin domains with high
phenomenon is indicative of the system residing below the configurational entropy. The increase in system energy to -
critical threshold, where thermal fluctuations are insufficient 1.21 confirms the destabilization of cooperative behavior
to disrupt the ferromagnetic order. The energy per spin is also among spins, reflecting the onset of paramagnetic
low, measured at approximately -1.95, consistent with a disordering. This state is characterized by frequent spin flips
minimized interaction energy resulting from aligned and domain fragmentation, in line with the statistical
neighboring spins. suppression of long-range order above 𝑻𝒄 .

IJISRT25MAY414 www.ijisrt.com 519

Volume 10, Issue 5, May – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25may414
The line plot illustrates the sharp reduction in both Please refer to the table titled Microscopic
magnetization and energy as the temperature increases from Configuration Metrics for summarized values of average
2.0 to 2.5, marking a qualitative change in system behavior. magnetization and energy per spin at 𝑻 = 𝟐. 𝟎 and 𝑻 = 𝟐. 𝟓.
Additionally, the tabulated data below presents the computed
mean values from simulation runs across these temperature B. Energy Variation and Equilibration Patterns
points. The marked contrast in these metrics further The energy landscape of the two-dimensional Ising
corroborates the onset of critical behavior in the model provides critical insight into the system’s
neighborhood of 𝑻𝒄 , as captured through visual and numerical thermodynamic stability and convergence behavior.
observation of microscopic states. Monitoring the energy per spin over time allows for assessing
equilibration dynamics, thermal fluctuations, and critical
energy regimes as the system evolves toward steady state
under different thermal conditions.

Fig 2: Energy Variation Over Time

At lower temperatures such as 𝑻 = 𝟐. 𝟎, the system confirms these thermal dynamics. The stable, narrow
resides in a near-minimum energy state due to strong spin fluctuation band at 𝑻 = 𝟐. 𝟎 contrasts with the broader,
alignment. As depicted in the accompanying line graph, the irregular trajectory at 𝑻 = 𝟐. 𝟓. This temporal evolution
energy per spin stabilizes around an average value of reveals that the system below 𝑻𝒄 rapidly relaxes into a
approximately -2.0, with minimal fluctuations. This behavior metastable energy basin, while above 𝑻𝒄 , the system explores
is indicative of a thermodynamically favorable, low-entropy a wider set of microstates with comparable energy, resulting
ferromagnetic phase. The convergence to equilibrium is in dynamic fluctuation behavior.
rapid, and the system maintains coherence with only minor
perturbations arising from local spin flips. These patterns are essential in determining the sampling
period post-equilibration, as they influence the statistical
In contrast, at 𝑻 = 𝟐. 𝟓, the energy landscape reflects reliability of observables such as specific heat and magnetic
increased thermal disorder. The energy per spin fluctuates susceptibility derived from energy fluctuations. The high
around -1.32 with greater amplitude, revealing the influence stability at 𝑻 = 𝟐. 𝟎 allows for shorter equilibration times,
of frequent spin reversals. This is consistent with the onset of whereas the noisy energy signature at 𝑻 = 𝟐. 𝟓 necessitates
paramagnetic behavior, where thermal agitation overcomes extended simulation durations to ensure accurate ensemble
the cooperative interaction energy 𝑱 between neighboring averages.
spins. The energy profile demonstrates slower convergence
and larger variance, typical of systems near or above the The table titled Energy Variation Across Time provides
critical temperature 𝑻𝒄 , where long-range order diminishes a snapshot of energy per spin values at distinct time intervals
and configurational entropy increases. during the simulation. These values align well with expected
thermodynamic behavior, reinforcing the system’s physical
Quantitatively, the energy time series plotted across realism and validating the simulation protocol.
Monte Carlo time steps (expressed in units of 𝟏𝟎𝟑 MCSS)

IJISRT25MAY414 www.ijisrt.com 520

Volume 10, Issue 5, May – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25may414
C. Magnetization Behavior and Susceptibility Trends spontaneous symmetry breaking and phase transition
The magnetization profile of a spin system reflects the phenomena. Its temporal evolution, especially across
degree of collective spin alignment and serves as a different thermal regimes, reveals the interplay between
fundamental order parameter for distinguishing between temperature-induced fluctuations and cooperative spin
thermodynamic phases. In the context of the 2-D Ising model, interactions.
magnetization per spin 𝑴 provides critical insight into

Fig 3: Magnetization Variation Over Time

At 𝑻 = 𝟐. 𝟎, well below the critical temperature 𝑻𝒄 , the This formulation links the observed fluctuations in 𝑴 to
system stabilizes rapidly into a strongly magnetized state, the system’s response to an external perturbation (e.g., a
characterized by a mean magnetization per spin close to 0.98. magnetic field). Although no external field is applied in this
As shown in the plotted magnetization trajectory, fluctuations simulation, the intrinsic susceptibility peaks near the phase
remain minimal over time, indicating that most spins remain transition due to enhanced fluctuations. As 𝑻 → 𝑻𝒄 , 𝝌
coherently aligned due to dominant ferromagnetic diverges in the thermodynamic limit, confirming the onset of
interactions. This ordered state is thermodynamically criticality.
favorable at low thermal energy and marks the persistence of
long-range spin correlation across the lattice. The high stability of magnetization at 𝑻 = 𝟐. 𝟎 allows
for precise susceptibility calculations with low variance,
Conversely, the magnetization at 𝑻 = 𝟐. 𝟓 shows a while the erratic magnetization at 𝑻 = 𝟐. 𝟓 necessitates larger
marked departure from stability. The plotted magnetization statistical sampling and binning correction to suppress
curve for this temperature exhibits wide fluctuations and a autocorrelation effects. The magnetization values also guide
significant decrease in mean value, hovering around 0.06– phase classification—systems with ⟨𝑴⟩ → 𝟏 are deemed
0.13 across time steps. This behavior is symptomatic of ferromagnetic, while those with ⟨𝑴⟩ ≈ 𝟎 are indicative of
paramagnetic disorder, where thermal agitation disrupts spin paramagnetic states.
alignment, resulting in transient and spatially incoherent
magnetization domains. The loss of macroscopic order above The figure titled Magnetization Variation Over Time
𝑻𝒄 is thus statistically captured in the suppression of 𝑴 over clearly demonstrates the contrasting behaviors across
time. temperature regimes, while the table titled Magnetization
Variation Across Time quantifies magnetization at discrete
From a thermodynamic standpoint, the magnetic time intervals. This comparative analysis substantiates the
susceptibility 𝝌 is defined via the fluctuation–dissipation critical transition from ordered to disordered phases,
relation: emphasizing the predictive power of 𝑴(𝒕) trajectories in
detecting equilibrium phase states.
𝟏
𝝌= (⟨𝑴𝟐 ⟩ − ⟨𝑴⟩𝟐 )
𝒌𝑩 𝑻

IJISRT25MAY414 www.ijisrt.com 521

Volume 10, Issue 5, May – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25may414
D. Mean Energy and Magnetization Analysis impact of momentary fluctuations inherent in Monte Carlo
The mean values of energy and magnetization per spin simulations. In the Ising model, their temperature dependence
are essential thermodynamic indicators in the study of phase captures the transformation from ferromagnetic order to
transitions in lattice-based spin systems. These ensemble- paramagnetic disorder as the system crosses the critical
averaged quantities provide a smoothed and stable temperature 𝑻𝒄 .
characterization of equilibrium properties, minimizing the

Fig 4: Mean Energy and Magnetization vs Temperature

The mean energy per spin ⟨𝑬⟩ quantifies the average critical sensitivity of magnetization to thermal excitation
interaction energy between neighboring spins and evolves compared to energy.
with temperature due to changing spin configurations. As
shown in the plotted graph, ⟨𝑬⟩ begins at approximately Furthermore, the tabular data under Mean Energy and
−𝟏. 𝟖𝟕 at 𝑻 = 𝟏. 𝟓, indicative of a low-entropy, energy- Magnetization vs Temperature provides discrete numerical
minimized state where spins are predominantly aligned. As values that trace the trajectory of these observables across the
temperature increases, ⟨𝑬⟩ smoothly increases (less negative) sampled thermal regime. This dual analysis, combining visual
due to enhanced spin flipping, reaching approximately −𝟏. 𝟏 and statistical perspectives, confirms that both observables
near 𝑻 = 𝟑. 𝟎. This transition is continuous and reflects the serve as effective thermodynamic signatures of the Ising
thermally driven destruction of spin coherence. phase transition, with ⟨𝑴⟩ functioning as a sharper indicator
of criticality.
The mean magnetization per spin ⟨𝑴⟩, on the other
hand, exhibits a sharper transition. At low temperatures (𝑻 < E. Interpretation of Phase Transition Characteristics
𝟐. 𝟎), ⟨𝑴⟩ remains close to unity, reflecting robust long-range The transition from an ordered to a disordered state in
ferromagnetic order. However, as the system approaches the the two-dimensional Ising model is a hallmark of second-
critical point (𝑻 ≈ 𝟐. 𝟐𝟕), ⟨𝑴⟩ rapidly decays due to order phase transitions. This transformation is driven by
intensified thermal agitation and critical fluctuations. Beyond thermal fluctuations and characterized by sharp changes in
𝑻𝒄 , ⟨𝑴⟩ trends toward zero, signifying the loss of global spin macroscopic observables. In this context, two primary
alignment in the paramagnetic phase. This steep decay is indicators of phase transition—namely the order parameter
characteristic of a second-order phase transition and reflects (mean magnetization ⟨M⟩) and the specific heat 𝑪𝒗 —are
spontaneous symmetry breaking in the order parameter. evaluated to dissect the critical dynamics near the Curie point
𝑻𝒄 ≈ 𝟐. 𝟐𝟕.
The line plots in the figure titled Mean Energy and The order parameter, defined as:
Magnetization vs Temperature visually capture these trends, 𝑵
𝟏
with sigmoid-like transitions centered around 𝑻𝒄 . The smooth ⟨𝑴⟩ = ⟨∣ ∑ 𝝈𝒊 ∣⟩
rise in ⟨𝑬⟩ contrasts with the steep fall in ⟨𝑴⟩, underlining the 𝑵
𝒊=𝟏

IJISRT25MAY414 www.ijisrt.com 522

Volume 10, Issue 5, May – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25may414

Fig 5: Phase Transition Indicators vs Temperature

quantifies the degree of spontaneous magnetization. As specific heat across the simulated temperature range. These
shown in the plot, ⟨M⟩ remains close to 1.0 at low values consolidate the graphical trends and demonstrate the
temperatures, indicating coherent spin alignment and strong system’s critical behavior with high fidelity.
ferromagnetic ordering. However, as temperature increases
and approaches 𝑻𝒄 , ⟨M⟩ experiences a sharp decline, V. CONCLUSION AND RECOMMENDATIONS
asymptotically approaching zero. This reflects the breakdown
of long-range order due to enhanced thermal agitation and the A. Summary of Key Findings
emergence of symmetric spin distributions. The sigmoid This study rigorously explored the thermodynamic
decay of ⟨M⟩ mirrors the mean-field prediction and behavior of the two-dimensional Ising model using
substantiates the role of ⟨M⟩ as an effective symmetry- Metropolis-based Monte Carlo simulations implemented in
breaking order parameter. MATLAB. Key physical observables, including energy,
magnetization, specific heat, and magnetic susceptibility,
In parallel, the specific heat 𝑪𝒗 , defined via energy were computed across a finely sampled temperature range to
fluctuations as: characterize the model’s critical phenomena.

𝟏 At subcritical temperatures (e.g., 𝑻 = 𝟐. 𝟎), the system

𝑪𝒗 = (⟨𝑬𝟐 ⟩ − ⟨𝑬⟩𝟐 ) exhibited strong ferromagnetic order, evidenced by high and
𝒌𝑩 𝑻𝟐
stable magnetization and low energy per spin. These
captures the sensitivity of the internal energy to configurations were characterized by minimal spin
temperature changes. The plotted data reveal a pronounced fluctuation and rapid equilibration, indicative of a highly
peak in 𝑪𝒗 around 𝑻𝒄 , consistent with diverging energy ordered phase. As temperature increased toward and beyond
fluctuations typical of critical phenomena. This peak is finite the critical point (𝑻𝒄 ≈ 𝟐. 𝟐𝟕), the system displayed
due to the finite lattice size, but its sharpness and location are significant energy and magnetization fluctuations.
indicative of the underlying singularity in the thermodynamic Magnetization sharply declined while energy per spin
limit. Above 𝑻𝒄 , the decline in 𝑪𝒗 reflects the saturation of increased, signaling the loss of long-range spin order and the
energy disorder and the system’s approach toward high- emergence of the disordered paramagnetic phase.
entropy paramagnetic behavior.
The susceptibility and specific heat functions peaked
Together, these observables delineate the critical region prominently near the critical temperature, confirming the
and confirm the presence of a continuous (second-order) presence of a second-order phase transition. The order
phase transition. The smooth but rapid transition in ⟨M⟩ and parameter exhibited a continuous, yet abrupt transition, while
the peak in 𝑪𝒗 are fundamental signatures predicted by the the energy followed a smoother trajectory, aligning with
Landau theory of phase transitions and validated by theoretical expectations. Time-resolved analyses showed that
Onsager’s exact solution for the 2-D Ising model. equilibrium was reached faster at lower temperatures and that
systems at or above the critical point required extended
Refer to the table titled Phase Transition Characteristics sampling due to longer autocorrelation times.
for detailed numerical values of the order parameter and

IJISRT25MAY414 www.ijisrt.com 523

Volume 10, Issue 5, May – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25may414
These findings validate the 2-D Ising model’s predictive C. Recommendations for Future Work
capabilities for phase transition dynamics and reinforce the To enhance the resolution, scalability, and analytical
effectiveness of the Metropolis Monte Carlo approach for depth of Ising model simulations, several strategic
probing equilibrium thermodynamics in spin-lattice systems. improvements are recommended for future investigations.
The results also provide a strong numerical foundation for Foremost among these is the adoption of advanced Monte
extracting critical exponents and conducting finite-size Carlo techniques, such as the Wolff and Swendsen–Wang
scaling in future extended studies. cluster algorithms, which offer substantial improvements in
sampling efficiency by reducing critical slowing down
B. Limitations and Computational Constraints through non-local spin updates. These methods are
While the Monte Carlo simulation framework applied in particularly advantageous near the critical temperature, where
this study provides an effective mechanism for investigating large correlated domains emerge and single-spin-flip
equilibrium properties of the two-dimensional Ising model, algorithms become computationally inefficient.
several inherent limitations and computational constraints
influence the scope and precision of the results. One Incorporating finite-size scaling methodologies will
fundamental limitation is the reliance on finite lattice sizes, enable more rigorous extraction of critical exponents and
which introduces finite-size effects that can obscure the true universal scaling functions. Simulations conducted across
singularities of thermodynamic observables near the critical multiple lattice sizes can be systematically analyzed to
temperature. In particular, phenomena such as divergence of extrapolate thermodynamic quantities to the infinite-volume
specific heat or susceptibility are smoothed out and shifted limit. This will facilitate quantitative validation of theoretical
due to the finite number of degrees of freedom, limiting the predictions and enable accurate mapping of phase boundaries
ability to directly extrapolate to the thermodynamic limit and scaling relations.
without implementing rigorous finite-size scaling analysis.
The implementation of histogram reweighting and
Another constraint arises from the autocorrelation multihistogram analysis is also recommended to improve the
inherent in sequential Markov chain sampling. Near the precision of thermodynamic measurements across a
critical point, the phenomenon of critical slowing down continuous temperature range. These techniques allow post-
significantly extends the autocorrelation time, thereby simulation interpolation of observable quantities, thereby
reducing the statistical independence of sampled reducing the need for densely spaced simulations and
configurations. This necessitates longer simulation times and enabling higher-resolution studies of critical behavior
careful statistical treatment, including binning and error without additional computational cost.
estimation techniques, to achieve reliable ensemble averages.
The single-spin update nature of the standard Metropolis For computational efficiency, migrating the simulation
algorithm also contributes to inefficiency in sampling platform to a compiled language such as C++ or integrating
correlated domains, particularly in large systems or near GPU acceleration via CUDA or OpenCL will significantly
criticality where collective behavior dominates. enhance performance, particularly for large-scale simulations
or ensemble-based parallel tempering methods. Employing
The computational implementation in MATLAB, while massively parallel architectures can reduce equilibration
well-suited for prototyping and visualization, imposes times, increase sampling throughput, and make it feasible to
performance limitations due to its interpreted nature and simulate systems at much larger spatial resolutions.
memory overhead associated with matrix operations on large-
scale lattices. Although vectorization and pre-computed Finally, extending the current model to include external
lookup tables were employed to optimize performance, the magnetic fields, anisotropic couplings, or quenched disorder
absence of low-level parallelism or GPU acceleration limits would provide broader insights into more complex statistical
scalability, especially for high-resolution lattice systems or systems and real-world materials. Such extensions would
multi-replica simulations across temperature ranges. allow for the exploration of rich phenomena including
hysteresis, Griffiths phases, and spin-glass behavior, further
Additionally, the simulation framework does not expanding the utility of the Ising framework in statistical
incorporate alternative update schemes such as cluster mechanics and computational condensed matter physics.
algorithms, which are known to mitigate critical slowing
down and improve convergence properties near 𝑻𝒄 . The D. Application of Findings in Real-World Magnetic Systems
absence of reweighting techniques, such as histogram or The insights derived from the computational analysis of
multicanonical methods, also constrains the resolution of the two-dimensional Ising model have direct applicability in
phase transition characteristics within narrow temperature the characterization and modeling of real-world magnetic
intervals. systems, particularly those exhibiting discrete spin behavior
and phase transition dynamics. The fundamental mechanisms
Overall, while the current computational setup yields of spontaneous magnetization, domain formation, and critical
qualitatively accurate and theoretically consistent results, the fluctuations captured by the Ising framework serve as
aforementioned limitations highlight the necessity for prototypical representations of ferromagnetic ordering
methodological enhancements and computational observed in crystalline solids, thin films, and low-
refinements in future studies aimed at achieving higher dimensional nanostructures.
precision, broader scalability, and deeper insights into critical
phenomena.

IJISRT25MAY414 www.ijisrt.com 524

Volume 10, Issue 5, May – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25may414
In real materials, such as transition metal ferromagnets Despite the inherent simplifications of the Ising
and magnetic semiconductors, the collective behavior of model—such as binary spin states, short-range interactions,
localized magnetic moments is influenced by short-range and the absence of quantum effects—it encapsulates essential
exchange interactions akin to those modeled in the Ising features of collective behavior observed in real-world
Hamiltonian. The simulation outcomes, particularly the physical systems. Its success in replicating critical signatures,
temperature-dependent behavior of energy, magnetization, such as divergences in susceptibility and specific heat,
and specific heat, align with experimental observations of underscores its foundational role in bridging computational
Curie transitions, enabling predictive modeling of material methods with thermodynamic theory. The methodological
response under varying thermal conditions. The ability to framework established here not only confirms known
capture critical exponents and response functions such as theoretical results but also lays the groundwork for future
susceptibility also supports the development of scaling investigations into higher-dimensional models, disordered
theories and universality classifications applicable to diverse systems, and technologically relevant spin-based devices.
magnetic compounds.
In a broader scientific context, the findings from this
The application of this model is particularly significant research exemplify how abstract mathematical constructs,
in the context of two-dimensional magnetic systems, when paired with robust numerical methods, can yield
including monolayer ferromagnets, spintronic predictive models with far-reaching implications. As
heterostructures, and artificial spin ices, where reduced computational resources and algorithms continue to advance,
dimensionality and finite-size effects are prominent. The the precision and scope of such simulations will expand,
results inform the design of nanoscale magnetic devices by enabling even deeper exploration into the rich landscape of
predicting thermal stability, coercivity, and critical operating statistical mechanics and its interdisciplinary applications.
conditions. Additionally, the simulation framework can be
extended to model magnetocaloric effects, enabling the REFERENCES
optimization of magnetic refrigeration materials through
accurate prediction of entropy and heat capacity near phase [1]. Anyibama, B. J., Orjinta, K. K., Omisogbon, T. O.,
transitions. Atalor, S. I., Daniels, E. O., Fadipe, E. & Galadima,
D. A. (2025). Modern Agricultural Technologies for
Beyond conventional ferromagnetic applications, the Sustainable Food Production: A Comprehensive
principles extracted from the Ising model extend to broader Review of Technological Innovations and Their
systems characterized by binary-state interactions. These Impact on Global Food Systems. International
include magnetic memory devices based on bistable spin Journal of Innovative Science and Research
configurations, probabilistic logic units in neuromorphic Technology ISSN No:-2456-2165 Volume 10, Issue 2
architectures, and even social or biological systems modeled https://doi.org/10.5281/zenodo.14964384
through analogous spin-lattice formalisms. The rigorous [2]. Azonuche, T. I., & Enyejo, J. O. (2024). Evaluating
quantification of phase stability and criticality provides a the Impact of Agile Scaling Frameworks on
foundation for evaluating robustness, adaptability, and Productivity and Quality in Large-Scale Fintech
collective behavior in such complex networks. Software Development. International Journal of
Scientific Research and Modern Technology, 3(6),
Consequently, the methodological rigor and theoretical 57–69. https://doi.org/10.38124/ijsrmt.v3i6.449
insights obtained from this study establish a scalable [3]. Azonuche, T. I., & Enyejo, J. O. (2024). Exploring AI-
computational foundation for the analysis, design, and Powered Sprint Planning Optimization Using
optimization of physical systems governed by collective spin Machine Learning for Dynamic Backlog Prioritization
dynamics, contributing to advancements in material science, and Risk Mitigation. International Journal of
nanotechnology, and interdisciplinary modeling. Scientific Research and Modern Technology, 3(8),
40–57. https://doi.org/10.38124/ijsrmt.v3i8.448.
E. Final Thought [4]. Baxter, R. J. (2016). Exactly solved models in
The two-dimensional Ising model remains a cornerstone statistical mechanics. Dover Publications.
in the theoretical and computational study of critical [5]. Binder, K. (1981). Finite size scaling analysis of Ising
phenomena, offering profound insights into the macroscopic model block distribution functions. Zeitschrift für
implications of microscopic interactions. This study has Physik B Condensed Matter, 43(2), 119–140.
reaffirmed its utility as a paradigmatic system for probing https://doi.org/10.1007/BF01293604
equilibrium thermodynamics, emergent order, and phase [6]. Binder, K., & Heermann, D. W. (2010). Monte Carlo
transition dynamics in discrete spin systems. Through the simulation in statistical physics: An introduction (5th
application of Monte Carlo simulation techniques, ed.). Springer. https://doi.org/10.1007/978-3-642-
particularly the Metropolis algorithm, it was possible to 03163-2
accurately capture the system’s statistical behavior across a [7]. Binder, K., & Luijten, E. (2001). Monte Carlo tests of
wide range of temperatures, highlighting the mechanisms of renormalization-group predictions for critical
symmetry breaking, fluctuation-driven transitions, and phenomena in Ising models. Physics Reports, 344(4–
critical scaling. 6), 179–253. https://doi.org/10.1016/S0370-
1573(00)00127-4
[8]. Cardy, J. (1996). Scaling and renormalization in
statistical physics. Cambridge University Press.

IJISRT25MAY414 www.ijisrt.com 525

Volume 10, Issue 5, May – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25may414
[9]. Domb, C. (1996). The critical point: A historical [27]. Wolff, U. (1989). Collective Monte Carlo updating for
introduction to the modern theory of critical spin systems. Physical Review Letters, 62(4), 361–
phenomena. CRC Press. 364. https://doi.org/10.1103/PhysRevLett.62.361
[10]. Ferrenberg, A. M., Xu, J., & Landau, D. P. (2018). [28]. Yeomans, J. M. (1992). Statistical mechanics of phase
Pushing the limits of Monte Carlo simulations for the transitions. Oxford University Press
three-dimensional Ising model. Physical Review E,
97(4), 043301.
https://doi.org/10.1103/PhysRevE.97.043301
[11]. Goldenfeld, N. (2018). Lectures on phase transitions
and the renormalization group. CRC Press.
[12]. Higham, D. J., & Higham, N. J. (2017). MATLAB
guide (3rd ed.). SIAM.
[13]. Ising, E. (1925). Beitrag zur Theorie des
Ferromagnetismus. Zeitschrift für Physik, 31(1), 253–
258. https://doi.org/10.1007/BF02980577
[14]. Kalos, M. H., & Whitlock, P. A. (2008). Monte Carlo
methods (2nd ed.). Wiley-VCH.
https://doi.org/10.1002/9783527632535
[15]. Katzgraber, H. G. (2010). Introduction to Monte Carlo
methods. In J. Moreno (Ed.), New optimization
algorithms in physics (pp. 1–20). Wiley-VCH.
https://doi.org/10.1002/9783527632580.ch1
[16]. Landau, D. P., & Binder, K. (2021). A guide to Monte
Carlo simulations in statistical physics (5th ed.).
Cambridge University Press.
[17]. McCoy, B. M., & Wu, T. T. (2014). The two-
dimensional Ising model. Courier Corporation.
[18]. Metropolis, N., Rosenbluth, A. W., Rosenbluth, M.
N., Teller, A. H., & Teller, E. (1953). Equation of state
calculations by fast computing machines. The Journal
of Chemical Physics, 21(6), 1087–1092.
https://doi.org/10.1063/1.1699114
[19]. Newman, M. E. J., & Barkema, G. T. (1999). Monte
Carlo methods in statistical physics. Oxford
University Press.
[20]. Onsager, L. (1944). Crystal statistics. I. A two-
dimensional model with an order-disorder transition.
Physical Review, 65(3–4), 117–149.
https://doi.org/10.1103/PhysRev.65.117
[21]. Pelissetto, A., & Vicari, E. (2002). Critical
phenomena and renormalization-group theory.
Physics Reports, 368(6), 549–727.
https://doi.org/10.1016/S0370-1573(02)00219-3
[22]. Preis, T., Virnau, P., Paul, W., & Schneider, J. J.
(2009). GPU accelerated Monte Carlo simulation of
the 2D and 3D Ising model. Journal of Computational
Physics, 228(12), 4468–4477.
https://doi.org/10.1016/j.jcp.2009.03.018
[23]. Sachdev, S. (2011). Quantum phase transitions (2nd
ed.). Cambridge University Press.
[24]. Swendsen, R. H., & Wang, J. S. (1987). Nonuniversal
critical dynamics in Monte Carlo simulations.
Physical Review Letters, 58(2), 86–88.
https://doi.org/10.1103/PhysRevLett.58.86
[25]. Vojta, T. (2003). Disorder-induced rounding of
certain quantum phase transitions. Physical Review
Letters, 90(10), 107202.
https://doi.org/10.1103/PhysRevLett.90.107202
[26]. Wang, J. S., & Swendsen, R. H. (1990). Low-
temperature properties of the Ising model from Monte
Carlo simulations. Physical Review B, 42(4), 2465–
2474. https://doi.org/10.1103/PhysRevB.42.2465

IJISRT25MAY414 www.ijisrt.com 526