🔬 materials science

Anisotropic and isotropic elasticity and thermal transport in monolayer C $_{24}$ networks from machine-learning molecular dynamics

Using a machine-learned neuroevolution potential, this study demonstrates that the mechanical stiffness and thermal transport of monolayer $\text{C}_{24}$ networks are governed by their bonding topology, where the quasi-tetragonal phase exhibits isotropic properties and the quasi-hexagonal phase shows pronounced anisotropy driven by low-frequency acoustic phonons.

Original authors: Qing Li, Haikuan Dong, Penghua Ying, Zheyong Fan

Published 2026-02-12

📖 3 min read☕ Coffee break read

CC BY 4.0

Original authors: Qing Li, Haikuan Dong, Penghua Ying, Zheyong Fan

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

Imagine you are building a massive, high-tech playground using nothing but tiny, carbon-based LEGO bricks. This paper is essentially a "blueprinting and testing" report for a new kind of ultra-strong, super-conductive material made from these tiny carbon clusters.

Here is the breakdown of what the scientists did, using some everyday analogies.

1. The "LEGO" Bricks: C24 vs. C60

For a long time, scientists have played with C60—which looks like a tiny soccer ball made of carbon. But these researchers decided to go smaller. They used C24, which is like switching from a standard soccer ball to a much smaller, denser marble.

Because these "marbles" are smaller, you can pack them much closer together. Think of it like this: if you try to build a wall out of basketballs, there are huge gaps between them. If you build a wall out of marbles, the wall is much tighter, denser, and harder to break. This is why the researchers found that C24 is much stiffer and stronger than the older C60 materials.

2. The Two "Architectural Styles" (qHP vs. qTP)

The researchers looked at two different ways to snap these C24 marbles together:

The "Chain-Link Fence" Style (qHP): Imagine laying out long, straight chains of marbles and then connecting them with a few diagonal wires. Because the structure is "stretched" in one direction, it behaves differently depending on which way you pull it. If you pull it along the chains, it’s strong; if you pull it across the wires, it’s different. This is called anisotropy.
The "Checkerboard" Style (qTP): Imagine a perfectly square grid of marbles, all connected equally in every direction. This structure is very balanced. No matter which way you pull it, it reacts almost exactly the same. This is called isotropy.

3. The "Heat Highway" (Thermal Transport)

The researchers also wanted to know how heat moves through these structures. Heat in these materials moves via phonons, which you can think of as "vibrational waves." Imagine a long line of people holding hands; if the person at one end shakes, the wave travels down the line.

In the "Chain-Link" (qHP) material: The heat travels like a car on a highway. It zooms very fast along the "main roads" (the strong covalent bonds) but struggles to move through the "side streets" (the weaker connections). This makes it a "directional" heat conductor.
In the "Checkerboard" (qTP) material: The heat moves like water spreading through a sponge—it goes pretty much the same way in every direction.

4. The "Digital Twin" (Machine Learning)

Simulating these tiny atoms is incredibly difficult and takes a massive amount of computer power. It’s like trying to predict the movement of every single drop of water in a crashing ocean wave.

To solve this, the scientists used Machine Learning (NEP-C24). Instead of calculating every single mathematical interaction from scratch (which is slow), they "taught" a computer program what carbon atoms usually do. The computer created a "Digital Twin" of the material. This allowed them to run massive, complex simulations in a fraction of the time, with incredible accuracy.

Why does this matter?

By understanding how the shape of the connections (the topology) changes how strong the material is and how fast heat moves, scientists can now "design" materials on purpose.

If you need a material for a computer chip that stays cool by pulling heat away in one specific direction, or a super-strong coating for a spacecraft, you can use these "blueprints" to pick the exact carbon arrangement you need.

Technical Summary: Anisotropic and Isotropic Elasticity and Thermal Transport in Monolayer $\text{C}_{24}$ Networks

1. Problem Statement

While 2D fullerene networks (such as $\text{C}_{60}$ networks) have gained attention for their mechanical robustness and diverse bonding topologies, the properties of smaller fullerene networks remain largely unexplored. Specifically, for the $\text{C}_{24}$ monolayer—the smallest stable fullerene molecule—there is a fundamental lack of understanding regarding its mechanical stability, the interplay between its bonding topology and anisotropic transport, and its phonon-mediated heat conduction. Traditional empirical interatomic potentials (e.g., Tersoff) struggle to accurately describe these systems due to the complex mixture of $\text{sp}^2$ and $\text{sp}^3$ hybridization.

2. Methodology

The authors employ a multi-scale computational approach to bridge the gap between quantum mechanics and large-scale dynamics:

Machine-Learned Potential (NEP-C24): To overcome the limitations of empirical potentials, the authors developed a new Neuroevolution Potential (NEP) specifically for $\text{C}_{24}$ . This model was trained on a dataset of 500 structures (400 for training, 100 for testing) generated via molecular dynamics and refined using Density Functional Theory (DFT) reference data (PBE functional with many-body dispersion corrections).
Structural Phases: The study investigates two distinct phases: the quasi-hexagonal phase (qHP), characterized by misaligned 1D chains, and the quasi-tetragonal phase (qTP), which possesses four-fold symmetry.
Mechanical Analysis: Elastic constants ( $C_{ij}$ ), Young’s modulus ( $E$ ), shear modulus ( $G$ ), and Poisson’s ratio ( $\nu$ ) were calculated using both DFT and NEP-C24 to assess stability and anisotropy.
Thermal Transport: The Homogeneous Nonequilibrium Molecular Dynamics (HNEMD) method was used to calculate lattice thermal conductivity ( $\kappa$ ). The authors also performed spectral decomposition to analyze the frequency-dependent contributions of phonons and calculated phonon group velocities and mean free paths (MFPs).

3. Key Contributions

Development of NEP-C24: A highly accurate machine-learned potential that reproduces DFT-level energy, force, and virial data, enabling large-scale, high-fidelity MD simulations of $\text{C}_{24}$ systems.
Structure-Property Mapping: The work establishes a direct link between the intermolecular bonding topology (the arrangement of covalent bonds between fullerene cages) and the resulting macroscopic mechanical and thermal anisotropy.
Identification of the "Small-is-Stronger" Trend: The study demonstrates that reducing the fullerene size from $\text{C}_{60}$ to $\text{C}_{24}$ significantly enhances both stiffness and thermal conductivity due to increased bond density.

4. Results

Mechanical Properties:
- qHP Phase: Exhibits pronounced anisotropy. The stiffness is higher along the $y$ -direction ( $E_y \approx 273.2$ N/m) than the $x$ -direction ( $E_x \approx 223.1$ N/m) due to the dense concentration of inter-fullerene bonds along the $y$ -axis.
- qTP Phase: Displays nearly isotropic elastic properties due to its four-fold symmetry. It also shows potential as a zero Poisson’s ratio material at specific orientations ( $\theta = 45^\circ$ ).
- Comparison: Both $\text{C}_{24}$ phases are significantly stiffer (approx. 1.5 times) than $\text{C}_{60}$ networks.
Thermal Transport:
- qHP Phase: Shows significant thermal anisotropy ( $\kappa_y \approx 341$ W m⁻¹K⁻¹ vs. $\kappa_x \approx 233$ W m⁻¹K⁻¹). This is driven by higher phonon group velocities and longer mean free paths along the $y$ -direction.
- qTP Phase: Exhibits isotropic thermal conductivity ( $\kappa \approx 272$ W m⁻¹K⁻¹).
- Phonon Mechanism: Heat transport is dominated by low-frequency acoustic phonons (< 5 THz). Real-space visualizations confirm that heat is carried by strong inter-fullerene covalent bonds rather than weak van der Waals forces.
- Magnitude: Thermal conductivities of $\text{C}_{24}$ are roughly twice those of $\text{C}_{60}$ networks, placing them in the same regime as biphenylene but below graphene.

5. Significance

This research provides a theoretical framework for the rational design of 2D fullerene-based materials. By tuning the fullerene size and the topology of the covalent inter-cage connections, engineers can customize the mechanical strength and directional heat conduction of these materials. This has potential applications in direction-specific mechanical sensors, actuators, and thermal management components in nanoelectronics.