First-Principles Study of the Structural, Electronic, Dynamic, and Mechanical Properties of HOPG and Diamond: Influence of Exchange–Correlation Functionals and Dispersion Interactions

Abstract

Various properties of two polymorphs of carbon, highly oriented pyrolytic graphite (HOPG) and diamond, were investigated at the ab initio level using different Hamiltonians with all-electron Gaussian-type functions (GTF) and projector augmented wave (PAW) basis sets. Their equilibrium lattice parameters, cohesive and interlayer interaction energies, band structures, vibrational frequencies, and elastic constants were evaluated. The calculations were performed at the Hartree–Fock, density functional theory (DFT), and hybrid (B3LYP and PBE0) levels. As regards DFT, the local density and generalized gradient (PBE and PBEsol) approximations were used. For GTF, the influence of the basis set superposition error (BSSE) was assessed. Since these approaches do not take dispersion interactions correctly into account, two different versions of Grimme’s dispersion correction, D2 and D3, were evaluated. The D2 and D3 corrections were reparameterized in order to reproduce the experimental structure of HOPG. For the properties depending on the description of the covalent bonds, such as the cohesive energy, C 11 and C 12, the dispersion corrections have a negligible influence. The best agreement is found for B3LYP-GTF-D3, PBE providing close results. For the properties depending on the description of the interaction between different layers, such as the interlayer interaction energies, C 33, and C 44, using the dispersion corrections improves the results. In general, D3 performs better than D2. PBE overall provides better results than the other functionals. As regards the basis sets, PAW describes the vibrational frequencies better than GTF due to the BSSE, whereas GTF provides a better description of the bandwidths and interband separations, elastic constants, and thermodynamical data. On the basis of the overall results, PBE-GTF-D3 appears to be a good compromise for an accurate description of the properties of graphite.

Ambient aerosol composition by infrared spectroscopy and partial least squares in the chemical speciation network: Multilevel modeling for elemental carbon

“…Absorption spanning 1600-1591 cm ¡1 is plausibly assigned to graphitic C D C stretches and has been identified previously in the spectra of combustion soot and model elemental carbons (Akhter et al 1985b;Kosti c et al 2009;Lechner et al 2016). However, a few qualifications regarding this assignment are necessary.…”

“…Second, the C D C atoms confined to a graphitic carbon lattice (e.g., as stacked graphene planes) have very few symmetry-allowed transitions in the mid-infrared (Kim et al 2005;Kosti c et al 2009;Kuhlmann et al 1998). Moreover, even in the perturbed ("turbostratic") morphology of soot carbon, it is unlikely that the symmetry conditions are relaxed such that this particular lattice vibration at »1585 cm ¡1 is experimentally observed in transmission spectra (Andreae and Gelencs er 2006;Lechner et al 2016). For example, only a very weak and broad C D C stretch is observed in the transmission spectra of graphite (ground under vacuum) even with areal densities are on the order of »1000 mg-EC/cm 2 (Smith et al 1975).…”

Error estimates for density-functional theory predictions of surface energy and work function

“…In the case of the p metals, the overestimation of the exchange energy also corrects for the absence of dispersion contributions. A well-known example where this correction is relevant is the c/a of graphite, which is predicted very accurately by LDA but is overestimated by PBE [82,83]. When, as a test, all materials from group 12 onwards (those with a full d shell or a partially occupied p shell) and noble metals Pt, Ag, and Au are excluded from the test set, PBE and LDA perform equally well (Table II).…”

Adsorption of atomic hydrogen on defect sites of graphite: Influence of surface reconstruction and irradiation damage

“…The calculations were performed with the periodic CRYSTAL09 [60,61] code in which the most recent dispersion correction D3 of Grimme [62] was implemented [63]. The PBE (for Perdew-Burke-Ernzerhof exchangecorrelation) functional [64] combined with the D3 dispersion correction [62] as parametrized in reference [63] was used to describe the properties of graphite. Regarding the parameterization of the D3 correction the zero damping was used: the s 6 , s 8 , s r,6 and s r,8 scaling parameters were set to 1.0, 0.410, 1.217 and 1.0, respectively.…”