An ab initio study on liquid silicon carbide. (February 2020)

• Record Type: Journal Article
• Authors: Saiz, Fernan
• Is Part Of: Journal of physics and chemistry of solids. Volume 137(2020)
• Publication Date: 2020-02
• Subjects: Materials under extreme conditions -- Density functional theory -- Ab initio molecular dynamics -- Phase transitions -- Silicon carbide; Solids -- Periodicals; Solides -- Périodiques; Solids; Periodicals; 530.41
• DOI: 10.1016/j.jpcs.2019.109204
• Languages: English
• ISSNs: 0022-3697

Abstract

Abstract: This work presents for the first time the properties of the liquid phase of silicon carbide using ab initio molecular dynamics simulations based on density-functional theory (DFT). Our DFT scheme employs a plane-wave basis to expand the atomic orbitals, pseudopotentials built with the projector augmented wave method, and the local-density approximation to describe the exchange–correlation interactions. With this approach we we determine a melting temperature of the zinc-blend phase of 2678.54 ( ± 41.67) K with a pressure of 0.25 ( ± 0.40) GPa and a density of 3.06 g/cm 3 in good agreement with the experimental normal melting point of 2818.00 ( ± 40.00) K. At these conditions, the diffusion coefficient of the melt is 6.86 x 10 −3 nm 2 /ps which compares well with the estimated value of 2.46 x 10 −3 nm 2 /ps in the experiments done at atmospheric pressure. Finally, our model shows that silicon carbide has a negative melting curve that qualitatively agrees with experiments, with a slope of -36.93 K/GPa with pressures between 2.56 and 6.48 GPa, which compares well with the -44 K/GPa reported from the laboratory carried out with pressures of up to 7.7 GPa. This work provides a straightforward methodology based on the popular 'Z-method' to produce liquid systems of silicon carbide, from which amorphous systems can easily be then produced by quenching. Highlights: The structural, thermodynamic, dynamic properties of silicon carbide are predicted with ab initio methods. Diffusion coefficients and melting temperatures are calculated as a function of pressure. Comparison with experimental shows the accuracy once more of density functional theory.