Converged properties of clean metal surfaces by all-electron first-principles calculations

Atomistic origins of the phase transition mechanism in Ge2Sb2Te5

“…33,34 Thus, we found deviations of about 2.7%, which is typical of DFT-PBE calculations for semiconductor and metal systems. 72,73 The 3Ge-Te-3Sb rule leads to the intermixing of Ge and Sb atoms in the same ͑111͒ planes and the formation of two distinct bond lengths for Ge-Te ͑2.87 and 3.24 Å, ⌬ b = 0.37 Å͒ and Sb-Te ͑2.96 and 3.30 Å, ⌬ b = 0.34 Å͒, which are important finger prints to prove that our structure model for m-GST provides a correct description of the experimental findings. For example, Kolobov et al 37 employing EXAFS identified two bond lengths in m-GST, i.e., 2.83Ϯ 0.01 and 3.2Ϯ 0.3 Å ͑⌬ b = 0.37 Å͒ for Ge-Te, and 2.91Ϯ 0.01 and 3.2Ϯ 0.3 Å ͑⌬ b = 0.29 Å͒ for Sb-Te.…”

The three-dimensional equilibrium crystal shape of Pb: Recent results of theory and experiment

“…The surface energies calculated by Vitos et al [13] and Feibelman [14,56] used density functional theory in the general gradient approximation (GGA) while Mansfield and Needs [8] as well as Yu et al [16,18] used DFT in the local density approximation (LDA). Surface energies for Pb and other metals calculated with GGA are about 30% lower than those with LDA [16,20]. The LDA yields better agreement with experiment, which is believed to be due to a better error cancellation of the surface exchange and correlation energy [57,[80][81][82].…”

“…More recently, full potential first principles calculations have produced surface energies of many crystalline materials [7][8][9][10][11][12][13][14][15][16][17][18] mostly for low-index orientations. Although there can be considerable scatter in theoretical values for a single material and orientation [16,19], it has been suggested by u Fax: +49 2461 61 2950, E-mail: h.bonzel@fz-juelich.de some authors that the lack of experimental surface energy data for well-defined orientations may be replaced by ab initio calculations [15,17,20].…”

“…More recently, full potential first principles calculations have produced surface energies of many crystalline materials [7][8][9][10][11][12][13][14][15][16][17][18] mostly for low-index orientations. Although there can be considerable scatter in theoretical values for a single material and orientation [16,19], it has been suggested by u Fax: +49 2461 61 2950, E-mail: h.bonzel@fz-juelich.de some authors that the lack of experimental surface energy data for well-defined orientations may be replaced by ab initio calculations [15,17,20].Theoretical attempts of describing the more or less complete anisotropy of the surface free energy of crystals by invoking attractive pairwise atomic interaction potentials have been numerous, beginning in 1931 with the work by Stranski et al [21][22][23][24][25][26] followed by others [27][28][29][30]. A different phenomenological description of the anisotropy and to some extent of the temperature dependence of the surface free energy was based on the terrace-step-kink model of a crystalline surface [31] assuming the existence of steps of monoatomic height, their thermal formation, interaction and roughening [32,33].…”

Of interfaces and spin transport across nanoscale layers

“…As for electronic properties, the work functions of metals, and the ionization potentials of insulators are typically given within a few tenths of an eV of the corresponding experimental values [29], with a slight trend to overestimation of these quantities. LDA is less successful, however, when calculating electronic band gaps, where it is known to underestimate the size of the gap on the scale of 50% [30].…”