NAGITIVE PHONON FREQUENCIES

[asgharali_jan]

Hi Can anyone Kindly tell me about the factors on which phonon calculation results depend?
Which factors can i take into account while performing DFPT? and still if after taking into account all tags required and correct inputs, we get nagitive frequencies then how can we stable a strucuture? KIndly need a step by step hint
Thanks in advance

[michael_wolloch]

Hi,

we have quite a lot of resources on our Wiki concerning phonons. Definitely check out the theory page for some background. We also provide a how-to page for step-by-step instructions and specific details about the finite differences method and density-functional-perturbation theory.

Another good resource would be our tutorials on phonons which you can use for hands-on training using py4vasp.

If you encounter imaginary frequencies (plotted as negative) your structure is dynamically unstable as you correctly stated. Obtaining a stable structure is a bit more advanced, but the main idea is to displace the atoms according to the collective movement at the unstable wave vector in reciprocal space by a bit and then carefully relax the positions and the cell. Note that such instabilities can depend on supercell size and computational details like exchange-correlation functional or pseudopotential (see e.g. Phys. Rev. B 94, 174435).

You can also check out systematic approaches to tackling the issue of finding stable structures given imaginary phonon frequencies. Phys. Rev. B 87, 184104 and works citing it are a good starting point.

[asgharali_jan]

michael_wolloch
Can you please tell me why someone gets this error
-----------------------------------------------------------------------------
| |
| EEEEEEE RRRRRR RRRRRR OOOOOOO RRRRRR ### ### ### |
| E R R R R O O R R ### ### ### |
| E R R R R O O R R ### ### ### |
| EEEEE RRRRRR RRRRRR O O RRRRRR # # # |
| E R R R R O O R R |
| E R R R R O O R R ### ### ### |
| EEEEEEE R R R R OOOOOOO R R ### ### ### |
| |
| ZBRENT: fatal error in bracketing |
| please rerun with smaller EDIFF, or copy CONTCAR |
| to POSCAR and continue |
| |
| ----> I REFUSE TO CONTINUE WITH THIS SICK JOB ... BYE!!! <---- |
| |
-----------------------------------------------------------------------------


i am using a monolayer and 27 27 1 kpoints mesh
below is incar

GNU nano 4.8 INCAR
ISTART = 0
ICHARG = 2

PREC = Accurate
ISIF = 5
IBRION = 2
EDIFF = 0.00000001
EDIFFG = -0.0005
NSW = 200
NELM = 100
ENCUT = 500
ISMEAR = 1
SIGMA = 0.005
ISYM = 2
LWAVE = .False
LCHARG = .False
LMAXMIX = 4
ISPIN = 2

[michael_wolloch]

Dear asgharali_jan,

this is a bit off-topic, but your conjugate gradient algorithm for relaxing your ions is failing. Your convergence criterion for the forces is exceedingly tight. I assume that your cell shape is already well converged, and the algorithm cannot determine a good gradient in the very flat potential energy surface close to the minimum.

If you are trying to find highly accurate in-plane lattice parameters of your monolayer, instead of running relaxations, there is a better way:
Vary the lattice parameter, make static calculations, fit an equation of state, and determine the minimum energy minimum that way.

All the best, Michael

[asgharali_jan]

ok thanks for bearing with me