Input Files

Overview

The  VASP manual  is the "holy bible" of VASP. Read it with care when you have questions. Also check out our  😎⁠⁠Assorted Wisdom⁠  section for some tips/tricks.
There are four key input files needed to run a VASP calculation:
The  INCAR : this contains most of the input flags.
The  POSCAR : this contains the input structure.
The  KPOINTS : this contains details about the k-point grid.
The  POTCAR : this contains the pseudopotential details.

INCAR


INCAR
179Bytes, Uploaded 8 months ago

There are a  wide assortment  of input arguments that VASP can take in the  INCAR  file, and there are many subtle rules for when to say which flags to what value. The group's software infrastructure will automatically make sure that your  INCAR  settings are compatible for the most part.
Please do not try to memorize all the INCAR flags. This is merely for ease of reference. That said, after some time using VASP, you will end up knowing more about these flags than you would ever wish.

Green Circle = "Very important; must learn for running your first calculations."
Yellow Square = "Rather important but not needed right way; learn after you've got the basics down."
Diamond Suit = "A bit more specialized. Come back to this later."

Level of Theory

Green Circle  ENCUT : this sets the plane-wave kinetic energy cutoff. The higher the value, the more precise your calculations will be at the expense of higher computational cost. At the top of each element's  POTCAR  file is a parameter named  ENMAX . The value of  ENCUT  , if not explicitly set, is the largest  ENMAX  value across the elements in your system. In practice, you should set  ENCUT  to at least 1.3 times this maximum  ENMAX  value for any structure relaxation involving changes to the cell volume in order to minimize artificial stresses.
Rule of thumb: Try at least 520 eV to start. In practice, you should ensure this is converged.
Green Circle  GGA : this sets the exchange-correlation functional (i.e. the "level of theory"). This has the largest impact on the results of the calculation.
Rule of thumb: Try  GGA = PE  first (i.e. the PBE functional), particularly if you are just learning how to use VASP.
Yellow Square  IVDW : this sets any van der Waals correction scheme. Grimme's DFT-D methods are common in molecular systems where vdW corrections are particularly important.
Rule of thumb: For typical bulk solids, don't set it unless you have reason to. For systems with notable vdW interactions, try  IVDW = 12  first.
Diamond Suit  LDAU /  LDAUJ  /  LDAUL  /  LDAUPRINT  /  LDAUTYPE  /  LDAUU : these arguments specify details related to the DFT+U method, which corrects so-called "self-interaction errors" of GGA functionals. If you are new to DFT, do not worry about these parameters yet.
Diamond Suit  LHFCALC  /  HFSCREEN : these methods specify details related to the use of hybrid functionals. Like DFT+U, this can also correct common "self-interaction errors" of GGA functionals.
Diamond Suit  METAGGA : this parameter defines a particular meta-GGA functional, which is generally more accurate than a GGA functional but less accurate than a hybrid functional.

SCF

Green Circle  PREC : specifies the precision used.
Rule of thumb: Generally set this to  Accurate .
Green Circle  EDIFF : this sets the SCF convergence tolerance. The lower the value, the more precise your energies and forces will be.
Rule of thumb: Try  1e-5  to start.
Yellow Square  ALGO : this sets the SCF convergence algorithm.
Rule of thumb: Try  Fast  to start. If you have problems, switch to  All  .
Yellow Square  ISMEAR : this sets how the partial occupancies are treated (i.e. the smearing method) to smooth out discontinuities near the Fermi level. There are a lot of subtle rules about what value to set for what material, which our software stack will try and set for you where appropriate.
Rule of thumb: When in doubt,  ISMEAR = 0  is a good bet with a small value of  SIGMA  (0.01 - 0.05).
Yellow Square  LASPH : this defines whether non-spherical contributions are included in the energy and forces. Do not mix calculations with  True  and  False  for  LASPH .
Rule of thumb: Always set this to  True  to make your life easier.
Yellow Square  LREAL : setting this value to  Auto  can speed up calculations for large systems but can reduce the quality of the energy and should always be set back to  False  for computed properties of interest.
Rule of thumb: Always set this to  False  (i.e. the default) to make your life easier, even when VASP complains.
Yellow Square  SIGMA : specifies the smearing parameter (e.g. width). Like  ISMEAR , this is very material-dependent.
Rule of thumb: When in doubt, set this to a small value (0.01 - 0.05) along with  ISMEAR = 0 .

Geometry Optimization

Green Circle  EDIFFG : this sets the maximum net force for a structure to be considered converged during a geometry optimization. The lower the value, the closer to the true local minimum your structure will be. Note that a negative sign should be used by convention (e.g.  EDIFFG = -0.03 ).
Rule of thumb: Use a value of  -0.03  or tighter (e.g.  -0.01 ).
Green Circle  IBRION : this is the geometry optimization algorithm.
Rule of thumb: Try  IBRION = 2  (conjugate gradient) to start.
Green Circle  ISIF : this defines which degrees of freedom should be modified during the structure relaxation.  ISIF = 2  indicates that the positions are relaxed at fixed cell volume.  ISIF = 3  indicates that the positions and cell shape/volume are simultaneously relaxed.
Rule of thumb: Use  ISIF = 3  for cell relaxations and  ISIF = 2  otherwise.
Green Circle  NSW : specifies the maximum number of geometry optimization steps to consider.
Rule of thumb: Set to an arbitrarily large value, e.g. 200.
Yellow Square  ISYM : this defines whether symmetry constraints should be accounted for.
Rule of thumb: Set to  0  for geometry optimizations to ensure the symmetry is allowed to change, unless you are intentionally looking to model a given symmetry.

Magnetization

Green Circle  ISPIN : this sets whether the calculation should be run without or with spin-polarization. If the system has unpaired electrons, it should be run with spin-polarization.
Rule of thumb: When in doubt, start with  ISPIN = 2  to see if there are favorable magnetic states. If there aren't, the default value of  ISPIN = 1  is fine.
Green Circle  LORBIT : this parameter specifies how much information related to magnetization is written out to the  OUTCAR  file.
Rule of thumb: Always set this to  LORBIT = 11  to make your life easier.
Green Circle  MAGMOM : this parameter specifies details related to the initial guess for the magnetic moments.
Rule of thumb: Initialize each transition metal with a high-spin magnetic configuration as a first attempt. Then go back and try other magnetic configurations, if applicable.
Diamond Suit  NUPDOWN : forces a particular net number of unpaired electrons.

File I/O

Yellow Square  ICHARG : whether to write out the charge density to the filesystem. This is useful for post-processing but is quite large.
Rule of thumb: No need to change the default value in most cases.
Yellow Square  ISTART : this defines if the VASP calculation should restart from a pre-existing wavefunction. If a  WAVECAR  is present in the same directory, VASP will automatically try to restart from it.
Rule of thumb: No need to change the default value in most cases.
Yellow Square  LWAVE : this parameter specifies whether the wavefunction is written out to the filesystem. This is especially useful for restarting a calculation but is quite large.
Rule of thumb: No need to change the default value in most cases.
Diamond Suit  LAECHG : this specifies whether  AECCAR  files are written out, which are used by certain post-processing codes. Never use  LAECHG  if  NSW  is not set to 0.
Diamond Suit  NEDOS : describes how many points are written out in the density of states output file ( DOSCAR ).

Miscellaneous

Green Circle  EFERMI : VASP recommends generally just setting this to  MIDGAP . That will yield slightly more consistent Fermi energies (and, as a result, band gaps) but otherwise changes nothing.
Rule of thumb: Always set this to  MIDGAP  to make your life easier if you are using VASP 6.4+.
Yellow Square  NCORE : this parameter can be tuned for a specific machine and system to speed up calculations by parallelizing over orbitals. The only way to know which value is best is to try them out! It should generally be used in place of  NPAR .
Yellow Square  KPAR : similarly, this parameter tells VASP how to parallelize the k-points. It only makes sense to use if there are many k-points, as a result.

POSCAR


POSCAR
7KB, Uploaded 8 months ago

The  POSCAR  file contains all information needed to define the input structure. The easiest way to generate a  POSCAR  file is using a program like  ⭐⁠⁠ASE⁠ ,  🧞⁠⁠Pymatgen⁠ , or a visualization program like VESTA. A code-snippet to convert a crystal structure in  .cif  format to a  POSCAR  file is shown below:
from ase.io import read, write

atoms = read("/path/to/my.cif")
write("POSCAR", atoms, sort=True)
Alternatively, with Pymatgen:
from pymatgen.core import Structure

structure = Structure.from_file("/path/to/my.cif")
structure.to(filename="POSCAR")

KPOINTS


KPOINTS
69Bytes, Uploaded 8 months ago

The  KPOINTS  file encodes the density of the k-point grid. A greater number of k-points typically indicates a more precise calculation. The number of required k-points to reach convergence is typically smaller as a given lattice dimension increases (i.e. large unit cells require fewer total k-points, on average).

There are many schemes in the  🧞⁠⁠Pymatgen⁠  code for automatically setting appropriate k-point values. We will often use the  pymatgen.io.vasp.inputs.Kpoints  class and the  Kpoints.automatic_density_by_vol()  function. An example is shown below:
from pymatgen.io.vasp.inputs import Kpoints
from pymatgen.core import Structure

structure = Structure.from_file("POSCAR")
kpts = Kpoints.automatic_density_by_vol(structure, 100)
print(kpts)
Rule of thumb: Use  Kpoints.automatic_density_by_vol()  with a density of at least 100 for  kppvol . In practice, this is also a parameter you should ensure is converged.

POTCAR

VASP ships with several different variations of pseudopotentials to use. Except in unusual circumstances, you can generally want to use the most up-to-date version of the  potpaw_PBE  pseudopotentials. Even if you are not using the PBE functional, these are generally still the pseudopotentials you should use.

VASP has a  recommended set  of pseudopotentials to use. The recommendations, for the most part, are reasonable.

The VASP POTCAR files are proprietary and should never be made publicly available (e.g. on GitHub, Zenodo, or anywhere else).