NEB and geometry relaxation with Electric field

[santanumahapatra]

Following is the INCAR file I am using:

Code: Select all

SYSTEM  = 2H-MoS2
ISTART=0
ENCUT=400
ALGO=Fast
IBRION=3
ISIF=2
NSW=200
EDIFF=1.0e-04
ISYM=0
LREAL=.FALSE.
ISMEAR=0
SIGMA=0.1
PREC=Med
ADDGRID=.TRUE.
NWRITE=1
LCHARG=.FALSE.
LWAVE=.FALSE.
LCLIMB = .TRUE.
IMAGES= 4
SPRING= -5.0
ICHAIN= 0
#LDIPOL = .TRUE.
#IDIPOL = 3   # for Z direction
#EFIELD = -5.0*1E19   # eV/A**
#DIPOL = 0.0 0.0 0.5 
#LNEBCELL = .TRUE

[ahampel]

I fear that performing NEB to predict structural phase transitions, i.e. changing lattice structure, is something not really supported in VASP, or at least not specifically. This method is also known as "variable-cell NEB" (https://www.sciencedirect.com/science/a ... 5513001392) and requires special algorithm I believe. The paper describes a code published here: https://uspex-team.org/online_utilities ... t0030.html that can be used in connection with VASP to perform such phase transition NEB calculation. I highly suggest to use these optimizers and intermediate image generation methods to perform the calculation you want to do.

I will discuss the topic with the other developers later but I am fairly certain that this is not covered VASP internally. If so we will make notes on the wiki to make this more clear.

[santanumahapatra]

Thank you for your response. I have one more question. When I run the VASP calculation for NSW=0 in the presence of electric field, I get very high energy value with the below INCAR:

Code: Select all

PREC=Accurate
EDIFF=1e-5
EDIFFG=-1e-2
ISTART=0
ENCUT=600.0
NSW=0
ISIF=2
IBRION=2
POTIM=0.02
ISMEAR=1
SIGMA=0.05
LDIPOL = .TRUE.
IDIPOL = 3   # for Z direction
EFIELD = 0.1   # eV/A**
DIPOL = 0.0 0.0 0.5 
#DIPOL = 0.4273  0.3434  0.4888
#LNEBCELL = .TRUE
#EFIELD_PEAD=0.0 0.0 0.5

But if I only use the PEAD tag and setting off the other Efield tags, the energy I get for single step calculation is the same as that in the absence of Efield.
Does this mean that I should not use NSW=0 when I am using the EFIELD tags and instead do a full relaxation of the structure? I am a bit confused here.

I have tested with different DIPOL values and I get different values of high energy. I believe this has something to do with the proper setting of DIPOL, howver, I am not very sure about how to calculate it accurately.

Below is the test POSCAR:

Code: Select all

EA0     4.078  5.542 18.703 95.19 88.97 91.64 Sym.group:    0
1.0
    4.077563     0.000000     0.073334
   -0.149723     5.517480    -0.501556
    0.000000     0.000000    18.703023
  Mo  Te
   2   4
Direct
    0.486712     0.643803     0.988911 
    0.000000     0.000000     0.000000 
    0.003467     0.661042     0.886774 
    0.971906     0.500467     0.081632 
    0.522218     0.154394     0.909593 
    0.483926     0.984842     0.100588 

Kindly suggest