Abnormal strain-dependent thermal conductivity in biphenylene monolayer using machine learning interatomic potential

Abstract

Applying tensile strain on an intrinsic lattice always results in the reduction in thermal conductivity due to the red-shift of phonon frequency and enhanced phonon anharmonicity. However, in this work, we explored an unexpected strain-enhanced thermal conductivity of a planar biphenylene network (BPN) in the frame of a Boltzmann transport equation combined with the machine learning interatomic potential. Under 5% biaxial tensile strain, the room temperature thermal conductivity of BPN reaches to about 4–5 times of that in an intrinsic sample. This phenomenon can be understood by considering a mirror symmetry induced phonon selection rule. This work highlights the significant effect of the selection rule on thermal transport and enriches the understanding of the thermal conductivity regulation in strained two-dimensional materials.

Complex role of strain engineering of lattice thermal conductivity in hydrogenated graphene-like borophene induced by high-order phonon anharmonicity

“…Due to the large phonon bandgap in BH, it is difficult for the phonons in the low frequency region (<30 THz) to be scattered with the phonon in the high frequency region (>44 THz). In addition, the previous study reported the strain-induced red-shift in phonon dispersion can weaken the three-phonon scattering [55,61]. The second part is the higher order phonon scattering (mainly four-phonon scattering).…”

“…The phonon group velocity and the linear coefficient in the nonequilibrium phonon distribution function are represented by n l a and l b F , respectively. The second-and third-order interatomic force constants (IFCs) applied to solve the PBTE are calculated from MTP [39,41,46,55]. A q mesh of 51 × 51 × 1 was used in ShengBTE for the calculation of lattice thermal conductivity with the cutoff radius of 5.87 Å, which is sufficiently large according to the previous study [13].…”

Overcoming the barrier: designing novel thermally robust shape memory vitrimers by establishing a new machine learning framework

“…In a subsequent cutting-edge development, Yan and colleagues introduced a novel self-enhanced deep neural network (SDNN) by first incorporating a self-attention algorithm for the prediction of flame retardancy, resulting in a highly efficacious ML model. 36 In the field of metals and alloys, Lee et al designed a MVAE framework that is able to generate 10 new alloy entries with higher yield strength and ultimate tensile strength. 37 In the field of metamaterials, Adithya et al designed new cellular structures and lattice structures with improved properties (load-carrying capacity, natural frequency, energy absorption capabilities) using a generative adversarial network (GAN).…”

Two-Dimensional Biphenylene-Based Carbon Allotrope Family with High Potassium Storage Ability

“…It exhibits typical metallic properties, different from that of semimetallicity for graphene and semiconductivity for GD and 2D-poly-C 60 . By cutting it into nanoribbons or performing surface hydrogenation, the BPN can be made semiconducting. , Its unique atomic arrangement imparts various anisotropic properties, such as electrical conductivity (including spin current), thermal conductivity, , and mechanical anisotropy, in addition to negative differential resistance, negative thermal expansion ratio, etc. It has great application potential in conductive materials, topological superconductors, lithium storage, hydrogen storage, electrocatalytic hydrogen production and oxygen reduction, , and anticorrosion coatings .…”