Abstract
With the increasing applications of Phosphorene in nano/optoelectronics and thermoelectrics, a comprehensive study on its thermal transport properties is necessary. It has been concluded from previous studies that there exist vast differences and uncertainty in the theoretically predicted thermal conductivity of single-layer phosphorene, which is generally attributed to the selection of XC functionals. However, even though the selection of functional groups in the first-principles calculations is particularly essential for predicting the thermal conductivity of phosphorene, there is no comprehensive investigation on this issue, which is unclear and is a gap in the field. Thus, the goal of this study is to investigate the effects of different XC functions on the thermal transport properties of phosphorene.
Introduction
Phosphorene, a monolayer of black phosphorus, has emerged as a significant material for nano/optoelectronics and thermoelectrics due to its high hole mobility and unique "hinged" honeycomb structure. However, a major challenge in its development is the inconsistency in theoretical predictions of its thermal conductivity ($\kappa$). Reported values for the zigzag direction vary significantly, ranging from 15 to 152 $Wm^{-1}K^{-1}$. This discrepancy is primarily attributed to the choice of exchange-correlation (XC) functionals in Density Functional Theory (DFT) calculations. This study systematically benchmarks 12 different XC functionals to provide a reliable reference for phosphorene's thermal transport properties.
Methods and Validation
The researchers employed a first-principles framework combining DFT with the Boltzmann Transport Equation (BTE). Structural optimizations and electronic properties were calculated using the Vienna ab initio simulation package (VASP), while phonon transport equations were solved using the ShengBTE package. Twelve XC functionals were evaluated, including standard LDA/GGA approximations (PBE, PW91), Ultrasoft Pseudopotentials (USPP_LDA, USPP_GGA), and van der Waals (vdW)-inclusive functionals (optB88, optB86b, optPBE, vdW_DF2, vdW_DF22).
Validation of these functionals was performed by comparing calculated phonon dispersion curves against experimental Raman scattering data for bulk black phosphorus. The phonon branches and density of states (DOS) vary across functionals, directly impacting the accuracy of subsequent thermal conductivity calculations.
Key Results and Discussion
The choice of XC functional has a profound impact on the predicted thermal conductivity, with values varying by up to a factor of 60. Functionals that incorporate vdW interactions, such as the "opt" series, generally predict higher $\kappa$ values. This is attributed to phosphorene's hinge-like structure, where long-range interactions between non-bonded atoms—often missed by standard functionals—play a crucial role in phonon transport.
A consistent finding across all functionals is the high degree of anisotropy in thermal transport. The zigzag direction consistently exhibits higher thermal conductivity than the armchair direction, with anisotropy ratios ranging from 3.4 to 6.9. This behavior is rooted in the electronic structure: electrons are localized along the zigzag direction but delocalized along the armchair direction, leading to higher phonon group velocities in the former.
The study also identified that the USPP_GGA functional produces anomalously low $\kappa$ values and should be avoided for phosphorene studies. The "optB86b" functional was found to provide a balanced description of the material's structural and thermal properties, showing the smallest Grüneisen parameter and aligning well with structural expectations.
Conclusion
This comprehensive benchmarking highlights the critical importance of XC functional selection in predicting the thermal properties of 2D materials with complex geometries. While experimental data for monolayer phosphorene is still needed for absolute validation, this study provides a vital framework for future theoretical investigations into phosphorene and similar anisotropic 2D materials like silicene or borophene.
Summary updated