next up previous contents index
Next: CHGCAR file Up: Files used by VASP Previous: CONTCAR file   Contents   Index

N.B. This document is no longer maintained, please visit our wiki.


This file is not required in VASP.3.2 and VASP.4.X, because the required tables are calculated by VASP directly. Use the EXHCAR file only with caution. If the file exists it must contain a table for the exchange-correlation energy of the homogeneous electron gas as a function of the charge density in the interval [0,RHO(2)]. This file can be generated with the program

>   setexch
setexch is distributed with the package, but it must be created separately, by typing
>   make setexch
in the VASP directory.

If you execute setexch you are asked for several parameters, enter similar values as given below:

 1   Perdew and Zunger, PHYS. REV. B23, 5048 (1982)
 2   Vosko, Wilk and Nusair, CAN. J. PHYS. 58, 1200 (1980)
 3   Gunnarson and Lundqvist
 4   Hedin and Lundqvist, J. PHYS. C4, 2064 (1971)
 5   Barth and Hedin
 6   Wigner-interpolation
1    <<< choose xc-type
 Relativistic corrections? (.T. of .F.)
.T.  <<< should be .T. for scalar rel. PP
 Interpolation type from para- to ferromagnetic corr.
 0   exchange-like 'standard interpolation'
 1   Vosko-type function (CAN. J. PHYS. 58, 1200 (1980)
0    <<< we recommend  0
 maximal small electron density RHO(1) ?
 number of points N(1) between 0 and RHO(1) ?
 maximal electron density RHO(2) ?
To get a good accuracy in the interpolation, the table is splitted in two regions, a low density region (0... "maximal small electron density RHO(1) ?") and a high density region (" maximal electron density RHO(2) ?"). This allows an accurate interpolation for atoms and molecules. As a crude guideline RHO(2) should not exceed 200, for transition metals this value was sufficient, and we generally recommend this setting for all materials. For 'simple' elements of the main group a value around 10 is sufficient. The correlation type selected should be the same as used for the pseudopotential generation (usually Ceperley-Alder as parameterized by Perdew and Zunger with relativistic corrections, i.e. switch '1').

Starting from version 3.2 VASP generates the EXHCAR file internally, in this case the parameters (given in the example session above) are used to create the table, only the first parameter is adopted to the pseudopotential.

N.B. Requests for support are to be addressed to: