Band-decomposed charge densities: Difference between revisions

Revision as of 12:11, 5 February 2024

The partial (band-decomposed) charge density can be used to analyze the contributions of different orbitals or energy ranges to a specific region in real space. It helps in gaining insight and visualizing electronic, magnetic, or transport properties, and is especially important when simulating scanning tunneling microscopy (STM) pictures. In VASP, the calculation of partial charges is a quick post-processing step that is selected by setting LPARD = .TRUE. in the INCAR file. It is necessary to provide a WAVECAR file from a converged ground state calculation as an input file. To select the contributing k points and bands, various options exist, which can be selected via the NBMOD, IBAND, EINT, and KPUSE tags.

Mind: All charge densities, including the band-decomposed charge densities, are symmetrized using both the space and point group symmetries. However, in certain cases, such as when calculating partial charge from selected k-points, this can lead to wrong results. In that case, the symmetry must be turned off during the initial ground state calculation from which the WAVECAR is generated, as well as during the subsequent band-decomposed charge density calculation.

Input tags and modes of selecting the partial charges

LPARD: Evaluate partial (band and/or k-point) decomposed charge densities. We want to stress again, that the orbitals read from WAVECAR must be converged in a separate prior run. If only LPARD is set (and none of the tags discussed below), the total charge density is evaluated from the orbitals and written to CHGCAR.

There are several ways how to specify for which bands the charge density is evaluated: In general the input lines with IBAND, EINT and NBMOD control this aspect of the routine.

IBAND: Calculate the partial charge density for all bands specified in the array IBAND. If IBAND is specified in the INCAR file and NBMOD is not given, NBMOD is set automatically to the size of the array. If IBAND is for instance

IBAND = 20 21 22 23

the charge density will be calculated for bands 20 to 23.

EINT: Specifies the energy range of the bands that are used for the evaluation of the partial charge density. Two real values should be given, if only one value is specified, the second one is set to $\epsilon _{f}$ . If EINT is given and NBMOD is not specified, NBMOD is set automatically to -2.

NBMOD: This integer variable can take the following values

>0: Number of values in the array IBAND. If IBAND is specified, NBMOD is set automatically to the correct value (in that case NBMOD should not be set manually in the INCAR file)

0: Take all bands to calculate the charge density, even unoccupied bands are taken into account.

-1: Calculate the total charge density as usual. This is the default value if nothing else is given.

-2: Calculate the partial charge density for electrons with the eigenvalues in the range specified by EINT.

-3: The same as before, but the energy range is given vs. the Fermi energy.

KPUSE: Specifies which k-points are used in the evaluation of the partial DOS. KPUSE is an array of integer values.

KPUSE= 1 2 3 4

means that the charge density is evaluated and summed for the first four k-points. Be careful: VASP changes the k-point weights if KPUSE is specified.

LSEPB: Specifies whether the charge density is calculated for every band separately and written to a file PARCHG.nb.* (LSEPB=.TRUE.) or whether charge density is merged for all selected bands and written to the file PARCHG.ALLB.* or PARCHG. Default is LSEPB=.FALSE..

LSEPK: Specifies whether the charge density of every k-point is write to the files PARCHG.*.nk (LSEPK=.TRUE.) or whether it is merged (LSEPK=.FALSE.) to a single file. If the merged file is written, then the weight of each k-point is determined from the KPOINTS file, otherwise the kpoints weights of one are chosen.

Output files

The partial valence charge density is written in the PARCHG file. If you want to separate the output by k points or bands, setting LSEPB and/or LSEPK allows you to write it to multiple PARCHG.*.* files.

Mind: {{{2}}}

@@ Line 41: / Line 41: @@
 The partial valence charge density is written in the PARCHG file. If you want to separate the output by k points or bands, setting LSEPB and/or LSEPK allows you to write it to multiple PARCHG.*.* files.
-{{NB|mind|For spin-polarized calculations, the PARCHG and its variants hold the total density and the magnetization density. For instance, if the 4th band is selected ({{TAG|IBAND}}=4) the first data set in the PARCHG file corresponds to the summed density of the 4th spin up and 4th spin down orbital, whereas the second data set holds the difference between the 4th spin-up and 4th spin-down orbital (magnetization density). Hence, to obtain the charge density corresponding to a specific orbital of a specific spin channel some post-processing of the {{TAG|PARCHG}} file might be required (building differences or sums). A simple workaround is to use  {{TAG|EINT}} and specify sufficient digits to select only one orbital from either the spin-up or spin-down channel.
+{{NB|mind|For spin-polarized calculations, the PARCHG and its variants hold the total density and the magnetization density. For instance, if the 4th band is selected ({{TAG|IBAND}}=4) the first data set in the PARCHG file corresponds to the summed density of the 4th spin up and 4th spin down orbital, whereas the second data set holds the difference between the 4th spin-up and 4th spin-down orbital (magnetization density). Hence, to obtain the charge density corresponding to a specific orbital of a specific spin channel some post-processing of the {{TAG|PARCHG}} file might be required (building differences or sums). A simple workaround is to use  {{TAG|EINT}} and specify sufficient digits to select only one orbital from either the spin-up or spin-down channel.}}
 == Related tags and articles ==