In the above paper, published in September 2009, the authors presented diagrams in Figs. 1, 2, and 7, presenting electrostatic potential in the GaN slab representing GaN(0001) surface, derived from density functional theory (DFT) calculation using VASP and SIESTA codes. In addition, the difference of the Fermi and valence band maximum energy in function of electric field was presented in Fig. 12. This interpretation turned out to be incorrect, the correct identification is that the diagrams in Figs. 1, 2, and 7 present electron energy (in eV). In Fig. 12, the electric field should be reversed. Accordingly, in the text of the paper, n-type conductivity should be replaced by p-type and vice versa. Furthermore, surface quantum states types (donors vs. acceptors) should be also exchanged. We regret for publishing the incorrect data in the above paper.
Cross-sectional view of the electric potential (left) and the electron density (right) in the region close to InCl molecule, obtained from VASP(LOCPOT file—top), SIESTA (SystemLabel. VH file—middle) and Dmol (SystemLabel_ potential.grd—bottom) calculations. These lines due to cylindrical symmetry of the system present the equipotential and equidensity closed surfaces.
Cross-sectional view of the electric potential (left) and the electron density (right) in the region close to InCl molecule, obtained from VASP(LOCPOT file—top), SIESTA (SystemLabel. VH file—middle) and Dmol (SystemLabel_ potential.grd—bottom) calculations. These lines due to cylindrical symmetry of the system present the equipotential and equidensity closed surfaces.
VASP manual describes the following: “The LOCPOT file contains the total local potential (in eV)”.1 Inside SIESTA manual, the following information was given “SaveElectrostaticPotential (logical): Instructs to write the total electrostatic potential, defined as the sum of the hartree potential plus the local pseudopotential, at the mesh used by DHSCF, in file SystemLabel.VH.”2 These statements, by the two independent sources were taken erroneously by the authors as guarantee that the output is electrostatic potential, and accordingly the LOCPOT and SystemLabel.VH output files data were published in the above paper. The authors claim that they are solely responsible for the error as the responsibility of the authors is to verify the integrity of the data published in the paper.
It is likely however that such error may happen to other authors. Therefore, in order to prevent that the following test was devised by the authors which identifies the physical quantity. An example is presented below in which InCl molecule was simulated using the DFT codes: VASP, SIESTA, and Dmol.3,4 The crossectional view of the output electrostatic potential and electronic density are presented in Fig. 1. In accordance to Gauss law, the equipotential closed surfaces surround regions containing unbalanced electric charge. Such regions, on the external side of Cl atom depicted in Fig. 1, encompass no atomic nuclei but contain the electronic (i.e., negative) charge and accordingly surround the electrostatic potential minima. As shown in Fig. 1, these regions are erroneously denoted as maxima in VASP and SIESTA output, while in DMol output, the region is correctly depicted as minimum. Therefore, the correct interpretation is that the output from VASP and SIESTA represents electron energy (in eV) while the output from Dmol—the electrostatic potential (in V).
We regret for publishing the incorrect data in the above paper using identification based on the description in VASP and SIESTA manuals. As such misleading information can be possibly found in other manuals the above test is presented, allowing the researchers to find proper interpretation, independent of the written word, and to avoid the error we have made.
