Thermal transport properties of MoS₂and MoSe₂monolayers

Abstract

The isolation of single-to few-layer transition metal dichalcogenides opens new directions in the application of two-dimensional materials to nanoelectronics. The characterization of thermal transport in these new low-dimensional materials is needed for their efficient implementation, either for general overheating issues or specific applications in thermoelectric devices. In this study, the lattice thermal conductivities of single-layer MoS 2 and MoSe 2 are evaluated using classical molecular dynamics methods. The interactions between atoms are defined by StillingerWeber-type empirical potentials that are developed to represent the structural, mechanical, and vibrational properties of the given materials. In the parameterization of the potentials, a stochastic optimization algorithm, namely particle swarm optimization, is utilized. The final parameter sets produce quite consistent results with density functional theory in terms of lattice parameters, bond distances, elastic constants, and vibrational properties of both single-layer MoS 2 and MoSe 2 . The predicted thermal properties of both materials are in very good agreement with earlier firstprinciples calculations. The discrepancies between the calculations and experimental measurements are most probably caused by the pristine nature of the structures in our simulations.S Online supplementary data available from stacks.iop.org/nano/27/055703/mmedia

Thermal transport in MoS2 from molecular dynamics using different empirical potentials

“…The large thermal conductivity values computed using the SW13 and SW16 potentials are clearly unphysical as compared to experimental data, while that computed using the REBO-LJ potential is very reasonable. Note that our HNEMD predictions using the SW potentials differ significantly from those from previous works [23][24][25][26]; see Table II. We give the detailed comparisons next.…”

“…An evidence for its high transferability comes from a very satisfactory reproduction [66] of the formation energies of point defects in MoS 2 , as compared to those obtained using DFT calculations [67]. In contrast, the SW potentials [25,33] were fitted by considering only some equilibrium properties such as bond lengths and elastic constants. The implementations of these SW potentials also involve an ad hoc modification to the LAMMPS source code, which is problematic when the structure is away from the equilibrium MoS 2 structure; the intention of this modification was to exclude the interactions in triplets such as Mo-S 1 -S 5 in Fig.…”

“…Kandemir et al [25] developed another SW potential for MoS 2 in 2016, which we call the SW16 potential. This potential differs from the SW13 potential in that the Mo-S-S three-body interaction, for the case where the two S atoms are in the same sublayer (such as the Mo-S 1 -S 2 triplet in Fig.…”

Thermal transport in van der Waals solids from first-principles calculations

“…1), calculations have usually been restricted to monolayers. [20][21][22][23][24][25][26] This is at least in part due to the fact that computational studies of bulk systems 27,28 require taking into account the vdW forces that mediate interlayer binding. These interactions are, however, not captured by common semi-local exchange-correlation (XC) functionals, 29 includ-ing widely popular functionals such as PBE 30 and PBEsol.…”

“…Furthermore, since vdW forces are rather weak and computational noise can blur anharmonic effects, both the choice of the XC functional and the convergence of the computational parameters require special care. 20 (c), 21 (d), 22 (e), 28 (f), 23 (g), 24 (h), 10 (i), 33 (j), 34 (k), 35 (l), 25 (m), 36 (n), 26 This perspective motivates the present study, in which we have carefully evaluated both the in-plane and out-of-plane arXiv:1605.07067v1 [cond-mat.mtrl-sci] 23 May 2016 thermal conductivities of Mo and W-based TMDs. To this end, we employ a combination of density functional and Boltzmann transport theory calculations based on the vdW density functional method 37 in combination with a recently formulated consistent-exchange part, 29,38 which has already been found to work very well for e.g., WSe 2 .…”