Anomalous strain-dependent thermal conductivity in superelastic screw-dislocated graphites

TL;DR

This paper addresses the challenge of achieving strain‑stable or strain‑enhanced thermal transport in nanostructured carbon: screw‑dislocated graphites (SDGs) are engineered as 3D topological carbon allotropes whose cross‑plane conductivity can be tuned by strain and dislocation density. Using NEP‑C machine‑learning potentials and non‑equilibrium MD, the authors demonstrate an anomalous increase in cross‑plane thermal conductivity under both tensile and compressive elastic strains, surpassing multilayer graphene by over an order of magnitude, with tensile enhancements exceeding 100% up to 80% strain and compressive enhancements over 700% up to 30% strain. They develop an analytic model linking κ to dislocation numbers $N_{ m x}$ and $N_{ m y}$ and strain, incorporating a coupling factor γ that captures enhanced phonon transport at higher dislocation densities. This work provides a design framework for strain‑tunable thermal management in flexible and wearable electronics, highlighting SDGs as a platform combining robust topological electronic states with tunable thermal transport properties.

Abstract

The design of strain-stable, or even strain-enhanced thermal transport materials is critical for stable operation of high-performance electronic devices. However, most nanomaterials suffer from strain-induced degradation, with even minor tensile strains markedly reducing thermal conductivity. Here, we demonstrate that screw-dislocated graphites (SDGs), recently identified as topological semimetals, display an unusual increase in cross-plane thermal conductivity under both tensile and compressive strains, revealed by high-accuracy machine-learning-potential-driven non-equilibrium molecular dynamics. Notably, SDGs exhibit over 100% enhancement under tensile strains up to 80% along the dislocation axis, arising from strain-induced increase in dislocation interface tilt angle that elongates the effective heat transfer paths. Their thermal conductivity surpasses multilayer graphene by an order of magnitude. An analytical model is further derived linking thermal conductivity to dislocation number and strain, offering a predictive framework for designing strain-tunable screwdislocated structures. These findings highlight SDGs as a promising platform for high-performance electronic and wearable devices with tunable thermal properties.

Figures (7)

• Figure 1: (Color online) Structures of screw-dislocated graphites (SDGs). (a) The high-resolution transmission electron microscopy images of screw dislocations in graphite, reproduced from Ref. 25, © 2023 The Authors. Published by Elsevier Ltd., distributed under the terms of the Creative Commons CC-BY license. (b) Schematics of SDGs, featuring paired dislocations with the same chirality (marked in blue) along the $x$-direction and paired dislocations with complementary chirality (marked in red) along the $y$-direction in graphene. The smallest repeat unit is given as marked by red dashed lines (left). Structures with different dislocation spacings along the $x$- (middle) and $y$-directions (right). (c) Set-up for the NEMD method. Heat flux flows from the hot region to the cold region.
• Figure 2: (Color online) Effect of screw dislocation number along $x$-direction $N_{\rm x}$ on the cross-plane thermal conductivity of SDGs. (a) Comparison of per-atom heat current distributions in SDGs with different $N_{\rm x}$, and only a small portion of the system (24576 atoms in total) is shown for clarity. Atom colors indicate the normalized magnitude of the heat current along the transport direction ($z$-direction) . Arrows represent both the magnitude and direction of the per-atom heat current. (b) Thermal conductivity and (c) fraction of covalent bonds along the dislocation axis as functions of dislocation number $N_{\rm x}$. The gray and red line are the theoretical results from equations (2) and (3), respectively.
• Figure 3: (Color online) Phonon density of states calculation. (a) Schematic of the partitioning of SDGs with different screw dislocation numbers $N_{\rm x}$. (b) Vibrational density of states of carbon atoms in regions 1 and 2 for SDGs with different $N_{\rm x}$.
• Figure 4: (Color online) The impact of screw dislocation number along $y$-direction $N_{\rm y}$ on the cross-plane thermal conductivity of SDGs. (a) Schematics of representative SDGs with different dislocation number $N_{\rm y}$, and only a small part of the system (36864 atoms in total) is shown for clarity. Variation of (b) thermal conductivity and (c) width of graphene ramp with $N_{\rm y}$. The line is the theoretical result from equation (3), respectively.
• Figure 5: (Color online) Comparison between the theoretical value of the cross-plane thermal conductivity of SDGs, calculated from equation (3), and the corresponding simulation results.