Repository files navigation

Calculating Phonon may be simple compared to other calculating process.

More details coule be got at Click🔗 Here

Now I show a simple example for calculting Phonon based onVASP.5.4.4+Phonopy.2.3.2.

It is easy and you can refer this article.click🔗 HERE.

It is worth noting that the calculation of the phonon spectrum needs to be fully optimised with high precision for the original cell structure, otherwisefalse frequencies can easily occur.

I show a example which containINCAR file

SYSTEM = 2D_InSeISTART = 0NWRITE = 2   PREC   = AccurateENCUT  = 500GGA    = PENSW    = 200ISIF   = 3ISYM   = 2NBLOCK = 1   KBLOCK = 1IBRION = 2NELM   = 80     EDIFF  = 1E-08   EDIFFG = -0.001 # Perhaps such high accuracy cannot be converged in one step, at which point if an error is reported, try distribution optimisation to make it convergeALGO   = NormalLDIAG  = .TRUE.LREAL  = .FALSE.ISMEAR = 0       SIGMA  = 0.02ICHARG = 2LPLANE = .TRUE.#NPAR  = 4   # If you are not using supercomputers, do not add this parameter as it may cause an error to be reported！      LSCALU = .FALSE.NSIM   = 4LWAVE  = .FALSE.LCHARG = .FALSE.ICORELEVEL =  1

For 2d materials,we often add this.Like follwing:

100110000

Rename Output file（CONTCAR） of the previous step to POSCAR

>cp CONTCAR POSCAR

>phonopy -d --dim="a a 1"

>cp POSCAR POSCAR-unitcell>cp SPOSCAR POSCAR

ISTART = 0NWRITE = 2IBRION = 8    IALGO  = 38NELM   = 200EDIFF  = 1E-07EDIFFG = -0.001ISMEAR = 0   SIGMA  = 0.02ENCUT  = 500PREC   = AccurateLREAL  = .FALSE.LWAVE  = .FALSE.LCHARG = .FALSE.ADDGRID = .TRUE.

If your machines or Servers'RAM is totally enough,try to increase KPOINTS may be a better choice!

It is a wise Initiative to estimate the consume RAM and decrease Number of cores properly.

When entering the operation step of the mechanics matrix, the memory will suddenly increase and this may crash your program. So please allow sufficient running memory when entering the second step.

Phonopy will generateFORCE_CONSTRAINS which is based onvasprun.xml.

>phonopy --fc vasprun.xml

You should generate by yourself and modify it.

>touch band.conf>vi band.confATOM_NAME = DIM = a a 1 BAND = 0.5 0.0 0.0  0.0 0.0 0.0  0.333333 0.333333 0.0  0.5 0.0 0.0 BAND_LABELS = M G K MFORCE_CONSTANTS = READ>ENTER->"ESC" and :wq!

Pay attention to "DIM = a a 1"!It depends on "phonopy -d --dim="a a 1"

>phonopy --dim="a a 1" -c POSCAR-unitcell band.conf

And the last process is:

>phonopy-bandplot  --gnuplot> PBAND.dat

Then you will getPBAND.dat and you can draw figures by Origin Gnuplot Matlab etc.(The Phonopy‘s default is51 Points between twohigh symmetry points)

More details could be seen at Click🔗 HERE

If thefalse frequencies still exists,change the "phonopy -d --dim="a a 1"".Whether to take thelarger or the smaller often depends on experience！

Sometimes it is unstable in thenon-magnetic state, so we need to verify and illustrate the phonon spectrum in theferromagnetic andantiferromagnetic cases！Don't forget to add the parameterRWIGS andMAGMOM!

For 2D materials,if you useVASP+Phonopy for calculation,It may have littte false frequencies.It is not a great problem,you can show it in your paper!

The simple tutorial is ending and thanks forYong He !

You can refer to his tutorials!Click🔗 HERE.