🔬 materials science

Analysis of some solid amorphous inorganic structures and the boson peak phenomenon with a computational random graph approach

This study proposes a new computational random graph algorithm that unifies low-temperature bosonic and high-temperature crystalline paradigms to analytically model solid amorphous inorganic structures, successfully explaining the boson peak phenomenon and validating its results against experimental neutronography data without requiring melting simulations.

Original authors: A. Berezner, M. Rybakov, M. Sidlyar, V. Fedorov

Published 2026-02-02

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: A. Berezner, M. Rybakov, M. Sidlyar, V. Fedorov

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: Building a "Glass" City Without a Blueprint

Imagine you are trying to build a city. Usually, architects start with a perfect, orderly grid of streets (a crystal) and then, if they want to make something chaotic, they melt the buildings down and let them cool randomly. This is how most scientists currently simulate amorphous alloys (metallic glasses)—materials that are strong and flexible because their atoms are jumbled up, not lined up in neat rows.

However, the authors of this paper say, "Why melt the city first? Why not just build the messy city from scratch?"

They propose a new computer program that assembles these disordered metal structures without ever simulating a melting crystal. Instead, they use a "random graph" approach. Think of it like throwing a million Lego bricks into a box and then asking the computer to snap them together based on which ones fit best energetically, rather than following a pre-drawn map.

The Core Problem: The "Boson Peak" Mystery

There is a weird phenomenon in these metallic glasses called the Boson Peak.

The Analogy: Imagine a choir. In a perfect crystal, everyone sings the exact same note in perfect harmony. In a messy amorphous glass, the singers are out of sync. At very cold temperatures, this "messy choir" suddenly starts humming a specific, extra loud note that doesn't exist in the perfect choir. Scientists call this the Boson Peak.
The Issue: We know this peak exists, but we don't have a single mathematical model that explains how the atoms behave from freezing cold all the way up to room temperature. Current models are like having two different rulebooks: one for the cold (physics of pairs) and one for the hot (physics of crowds), but they don't talk to each other.

The Solution: A New Algorithm

The authors created a Python program to solve this. Here is how their "magic" works:

The Random Scatter: They start by randomly placing points in a digital box. These points represent atoms (Iron, Nickel, and Chromium) in the exact proportions found in a real alloy called AMAG-225.
The "Dating" Game: The program measures the distance between every single point. It then asks: "If these two atoms were to hold hands (bond), how much energy would it cost?"
The Energy Minimization: The program looks for the "cheapest" bonds. It pairs up atoms that have the lowest energy cost to stick together. It's like a matchmaking service that only introduces people who are perfectly compatible, ignoring everyone else.
The Graph Theory Twist: They treat the atoms as "vertices" (dots) and the bonds as "edges" (lines). By analyzing this random web of connections, they can mathematically prove that the resulting structure behaves like a real metallic glass.

The Results: Does It Work?

The team ran their simulation on a supercomputer. Here is what they found:

It Matches Reality: When they compared their computer-generated "city" to real-world data from neutron scattering experiments (which is like taking an X-ray of the atoms), the shapes matched almost perfectly. The correlation was 99%.
It Explains the Peak: Their math shows that at low temperatures, the "Boson Peak" is caused by these specific, frozen pairs of atoms holding hands tightly. As the temperature rises, these pairs start interacting with the whole crowd, turning the "duet" into a "symphony," which explains why the peak disappears at higher temperatures.
Speed: The original code was slow (like a single person sorting a deck of cards). They optimized it to run on many processors at once (like a whole team sorting cards). This made the simulation 19 times faster, allowing them to simulate up to 10,000 particles instead of just 2,000.

The "Glass-Forming" Test

One cool feature of their code is that it can tell you if a mixture of metals will actually turn into glass or if it will accidentally turn into a crystal.

The Analogy: If you throw the Lego bricks together and the computer finds that two different pairs of bricks have the exact same distance and energy, it throws an error. This is a warning sign: "Hey, these atoms are too organized! They are trying to form a crystal!"
If the code runs without errors, it means the mixture is "glass-forming" (it stays messy and disordered).

Summary

In short, the authors built a new digital tool that assembles metallic glasses atom-by-atom using a "best-fit" energy strategy rather than melting crystals. They proved that this random, graph-based approach accurately predicts real-world behavior, explains the mysterious "Boson Peak," and runs much faster than traditional methods. They didn't just simulate the structure; they provided a mathematical bridge connecting the cold, frozen state of the material to its warmer, more fluid state.

Technical Summary: Analysis of Solid Amorphous Inorganic Structures and the Boson Peak via a Computational Random Graph Approach

Problem Statement
Current computational models for amorphous metallic alloys (metallic glasses) face significant limitations. Traditional approaches, such as molecular dynamics with melting and cooling cycles, rely heavily on a "crystalline basement" or pre-existing ordered phases. This dependency complicates the simulation of multicomponent systems and fails to provide a unified model connecting low-temperature bosonic statistics (responsible for the boson peak) with high-temperature collective particle interactions. Furthermore, existing methods often struggle to explain the physical origin of the boson peak—a feature in thermal capacity plots at cryogenic temperatures absent in crystalline materials—and the transition of the system's statistical behavior from bosonic to normal distributions as temperature rises. There is a recognized need for a novel algorithmic approach that simulates amorphous assembly independently of prior crystalline structures while accounting for pair potentials and energy minimization.

Methodology
The authors propose a new algorithmic model that treats amorphous structure assembly as a random graph problem. Instead of melting crystals, the method begins with a random distribution of point sets (potential atomic centers) within a defined spatial "box," with particle counts proportional to the specific atomic composition of the target alloy (specifically the AMAG-225 alloy: Fe-73.5%, Ni-25%, Cr-1.5%).

The core workflow involves:

Random Generation: Generating random coordinate pairs (in 2D) for the specified number of particles.
Distance Calculation: Computing interatomic distances between all possible point pairs.
Energy Minimization: Substituting these distances into Finnis-Sinclair pair potentials (analytically defined for Cr-Cr, Fe-Fe, Ni-Ni, and mixed pairs) to determine binding energies.
Filtering and Sorting: The algorithm filters distances (removing duplicates and those below a minimum atomic radius threshold) and selects pairs that correspond to the minimal binding energy.
Graph Matching: The process is interpreted as matching vertices in a random graph. If duplicate distances or energies are detected during sorting, the code flags a potential local crystallization, serving as a metric for glass-forming ability.
Optimization: To handle the polynomial time complexity associated with nested loops and list searching, the authors implemented parallel computing using MPI (Message Passing Interface) on high-performance hardware (Intel Xeon processors), significantly reducing calculation time compared to isolated CPU execution.

Key Results

Structural Simulation: The model successfully generated stable cyclic operations for the AMAG-225 alloy. The calculated radial distribution function, $U(r)$ , showed a linear correlation coefficient of 0.99 with experimental neutron scattering data, with an absolute error of ±0.011 eV.
Boson Peak Explanation: By analyzing the mean energy function through the lens of random graph theory, the authors derived that the system's thermal capacity follows an expression proportional to $\exp(-E/kT)/T^2$ . This mathematical form yields a local minimum at low temperatures, theoretically explaining the emergence of the boson peak as a result of "frozen" atomic pairs (dipoles) in the amorphous matrix.
Statistical Transition: The study demonstrates a transition in the system's behavior: at cryogenic temperatures, the system is dominated by isolated pair contributions (bosonic statistics), while at room and higher temperatures, the behavior shifts toward a collective interaction modeled by a completely connected graph, resembling a biatomic ideal gas described by Johnson-Mehl-Avrami-Kolmogorov (JMAK) and Kohlrausch-Williams-Watts (KWW) models.
Performance: The implementation of MPI parallelization increased calculation speed by a factor of 19 or more, allowing the simulation of up to $10^4$ particles, an expansion from the previous limit of 2,000 particles.

Significance and Claims
The paper claims to offer a unified computational framework that bridges the gap between low-temperature bosonic phenomena and room-temperature structural properties without requiring the simulation of crystal melting. Key contributions include:

Independence from Crystalline Prehistory: The algorithm assembles amorphous structures based solely on atomic composition and energy minimization, removing the need for a "crystalline basement" or complex multi-phase melting simulations.
Glass-Forming Ability Estimation: The detection of duplicate distances/energies serves as a diagnostic tool to estimate the tendency of a specific atomic composition to crystallize rather than form a stable glass.
Theoretical Unification: The work provides a theoretical link between random graph theory and the physical origin of the boson peak, suggesting that the peak arises from the specific statistical distribution of frozen atomic pairs in the low-temperature limit.
Computational Efficiency: The study outlines effective optimization strategies, including parallel processing and potential porting to C-language, to enhance the efficiency of simulating large-scale amorphous systems.

The authors conclude that while the current model is robust for the tested alloy and offers a unified description across temperature regimes, comprehensive data on atomic bonding in rare-earth elements remains incomplete, and further performance enhancements (such as C-language porting) are recommended.