NEB problem: the intermediate points become totally disordered

[Jiaye Li]

Dear all

I feel very disappointed with NEB. I am going to find the transition state of a surface reaction. The surface is a periodic model, I considered the boundary. 1. When I used nebmake.pl to generate the intermediate structures, those newly obtained structures are totally disordered! Some surface atoms fly everywhere. 2. I have to generate those intermediate points by hand, however, after several hours those points become disorder again, although the inital structures (generated points by hand) look alright. These problem really caused me headache. Could anybody help me?

This is my INCAR:
SYSTEM = Silicon
Startparameter for this Run:
NWRITE = 2 verbosity write-flag
LPETIM= F write-flag & timer #
ISTART = 0 0-new 1-cont 2-samecut job
ISPIN = 1 1-no, 2-yes
INIWAV = 1 0-jellium WF; 1-random numbers (recomm.)

Electronic Relaxation
! ENCUT = 400.00 #Ecut (eV)
PREC = Medium * #precission: #Low/Medium/High
NELM = 150 max. # of electronic SC steps
NELMDL = 5 # of non-consistent steps at the beggining:>0 or <0
EDIFF = 1.0e-4 stopping-criterion for electronic upd.
VOSKOWN = 1
ISMEAR = -5
SIGMA = 0.1
AMIX = 0.2
BMIX = 0.0001

Ionic relaxation
EDIFFG = 1.0e-3 stopping-conditions for ionic relax.loop (EDIFF*10)
NSW = 250 * # of steps for ionic update (def:0)
IBRION = 2 * ionic relax: -1:no move; 0-MD;1-Newton;2-CG;3-damped;4-?
ISIF = 2 * calc.stress:0-no;1-tr;2-7 yes; 3-change vol.4-change shape
! ISYM = 2 symmetry:0-OFF, 1-ON,2-ON+memory conserving for CD
NBLOCK = 1
POTIM = 0.3 if IBRION=1,2,3:scaling constant forces;IBRION=0,time st.


IALGO = 48 * algorithm: use only 8 (CG) or 48 (RMM-DIIS)
LDIAG = .TRUE. sub-space diagonalisation
GGA = 91 * xc-type: PB, PW (86), LM, 91 (Perdew-Wang 91)

LWAVE = .T. * write WAVECAR
LCHARG = .T. * write CHGCAR and CHG
LVTOT = .T. * write the local potential LOCPOT
LELF = .F. create ELFCAR file
LORBIT = .F. create PROOUT
LREAL = Auto


! NEB

! ICHAIN = 0 !Indicates which method to run. NEB (ICHAIN=0) is the default
IMAGES = 4 ! Number of images in the band, excluding endpoints
SPRING = -5.0 eV/A2 ! Spring force between images; negative value turns on nudging
! LCLIMB = .FALSE. !Flag to turn on the climbing image algorithm
! LTANGENTOLD = .FALSE. ! Flag to turn on the old central difference tangent
! LDNEB = .FALSE. !Flag to turn on modified doubble nudging


POSCAR1
Si
1.00000000000000
8.5096 0 0
0 9.8260 0
0 0 13.6095
4 20 2 20 9
Selective dynamics
Direct
0.70375 0.5622 0.44489 T T T ! C
0.42122 0.58428 0.45392 T T T ! C
0.27292 0.26957 0.5008 T T T ! C
0.48999 0.11117 0.51786 T T T ! C
0.55038 0.54164 0.29673 F F F ! H
0.7066 0.26256 0.39225 T T T ! H
0.80121 0.50473 0.4123 T T T ! H
0.30879 0.55691 0.4197 T T T ! H
0.71588 0.66981 0.42346 T T T ! H
0.43793 0.69428 0.44262 T T T ! H
0.22306 0.35138 0.45582 T T T ! H
0.60751 0.08451 0.49053 T T T ! H
0.19186 0.18225 0.49565 T T T ! H
0.41215 0.02297 0.50548 T T T ! H
0.71293 0.55468 0.52555 T T T ! H
0.41287 0.56414 0.53347 T T T ! H
0.27488 0.30403 0.57805 T T T ! H
0.49868 0.12882 0.59773 T T T ! H
0.01499 0.32895 0.23151 T T T ! H
0.48954 0.09665 0.23287 T T T ! H
0.01336 0.83488 0.23517 T T T ! H
0.44788 0.69375 0.24526 T T T ! H
0.84444 0.9702 0.25198 T T T ! H
0.02325 0.51302 0.30535 T T T ! H
0.55238 0.50871 0.4092 T T T ! N
0.4299 0.23192 0.46763 T T T ! N
0.0888 0.35098 0 F F F ! O
0.0888 0.85098 0 F F F ! O
0.5888 0.10098 0 F F F ! O
0.5888 0.60098 0 F F F ! O
0.3564 0.85098 0.09471 F F F ! O
0.3564 0.35098 0.09471 F F F ! O
0.8564 0.60098 0.09471 F F F ! O
0.8564 0.10098 0.09471 F F F ! O
0.18238 0.58991 0.09742 T T T ! O
0.67941 0.83111 0.10927 T T T ! O
0.67641 0.33035 0.11045 T T T ! O
0.17623 0.07449 0.11671 T T T ! O
0.56005 0.02825 0.20482 T T T ! O
0.52748 0.53276 0.20679 T T T ! O
0.0453 0.73617 0.237 T T T ! O
0.00307 0.48335 0.23842 T T T ! O
0.95848 0.9763 0.24525 T T T ! O
0.02274 0.22958 0.24599 T T T ! O
0.47852 0.78679 0.26378 T T T ! O
0.48482 0.28554 0.26732 T T T ! O
0.54263 0.31073 0.38201 F F F ! Si
0.2226 0.94613 0.04736 F F F ! Si
0.2226 0.44613 0.04736 F F F ! Si
0.7226 0.69613 0.04736 F F F ! Si
0.7226 0.19613 0.04736 F F F ! Si
0.0185 0.61471 0.16126 T T T ! Si
0.51361 0.87725 0.1656 T T T ! Si
0.51202 0.37777 0.1659 T T T ! Si
0.00586 0.10544 0.17143 T T T ! Si

POSCAR2
Si
1.00000000000000
8.5096 0 0
0 9.8260 0
0 0 13.6095
4 20 2 20 9
Selective dynamics
Direct
0.26925 0.42519 0.44698 T T T ! C
0.72212 0.74157 0.49094 T T T ! C
0.44579 0.78716 0.51852 T T T ! C
0.45052 0.26262 0.5238 T T T ! C
0.01276 0.32777 0.23309 T T T ! H
0.50304 0.62869 0.23545 T T T ! H
0.00459 0.83434 0.23702 T T T ! H
0.51660 0.07416 0.23709 T T T ! H
0.83277 0.96911 0.25161 T T T ! H
0.00449 0.50964 0.30786 T T T ! H
0.69986 0.31650 0.37576 T T T ! H
0.26230 0.50203 0.38864 T T T ! H
0.54600 0.72784 0.39 T T T ! H
0.80471 0.68904 0.44166 T T T ! H
0.16534 0.35806 0.44309 T T T ! H
0.74904 0.85223 0.48775 T T T ! H
0.32797 0.76860 0.48832 T T T ! H
0.56402 0.21075 0.51419 T T T ! H
0.46772 0.89866 0.51616 T T T ! H
0.26743 0.47935 0.51781 T T T ! H
0.36028 0.18334 0.5338 T T T ! H
0.74616 0.70712 0.56644 T T T ! H
0.45740 0.32441 0.5912 T T T ! H
0.44561 0.75628 0.59626 T T T ! H
0.41496 0.34848 0.43861 T T T ! N
0.56124 0.70860 0.46344 T T T ! N
0.08880 0.35098 0 F F F ! O
0.08880 0.85098 0 F F F ! O
0.58880 0.10098 0 F F F ! O
0.58880 0.60098 0 F F F ! O
0.35640 0.85098 0.09471 F F F ! O
0.35640 0.35098 0.09471 F F F ! O
0.85640 0.60098 0.09471 F F F ! O
0.85640 0.10098 0.09471 F F F ! O
0.68161 0.83705 0.10018 T T T ! O
0.18253 0.58707 0.10106 T T T ! O
0.67930 0.33667 0.10368 T T T ! O
0.17291 0.07139 0.11913 T T T ! O
0.04236 0.73687 0.23734 T T T ! O
0.55468 0.97865 0.2376 T T T ! O
0.99745 0.48202 0.23941 T T T ! O
0.49974 0.72831 0.2438 T T T ! O
0.94764 0.97515 0.2465 T T T ! O
0.02091 0.22825 0.24718 T T T ! O
0.49680 0.22962 0.24896 T T T ! O
0.52530 0.46555 0.25339 T T T ! O
0.22260 0.94613 0.04736 F F F ! Si
0.22260 0.44613 0.04736 F F F ! Si
0.72260 0.69613 0.04736 F F F ! Si
0.72260 0.19613 0.04736 F F F ! Si
0.01647 0.61388 0.16347 T T T ! Si
0.51200 0.35014 0.16353 T T T ! Si
0.51526 0.85235 0.16659 T T T ! Si
0.00295 0.10276 0.17401 T T T ! Si
0.53732 0.34230 0.34235 T T T ! Si

Thanks

[Jiaye Li]

Problem solved.
That's due to the mismatching of atoms.

[admin]

probably (one of the) atoms re-entered the unit cell from the other side (like -0.01 --> 0.99)

[dano]

An atom leaving an reentering the box would not effect this calculation because the neb code takes into account periodic boundary conditions when calculating distances.

I was looking at you files and the issue is that the ordering of the atoms changes from POSCAR1 to POSCAR2. The script is working correctly but the way it works is to linearly interpolate the first atom listed to the position of the first atom listed in file 2 and so on.

For example atom 47 in POSCAR1 has the same position and bonding configuration as atom 55 in POSCAR2.

Even when you rotate the atoms manually so the structures look correct the neb still connects atom 1 to atom 1 and thus the big mess after several steps.

Also I notice that you have LCLIMB=.FALSE. if you are really interested in a correct saddle I recommend this flag be set to TRUE.

One more thing is that the VASP group does not support the vtstscripts or NEB package added by Henkelman. You should post questions related to these at Henkelman's forum http://theory.cm.utexas.edu/forum/index.php

[admin]

the problem with the re-entering may occur if a linear interpolation script is used to generate the images:
it dos make a difference for the intermediate positions if you move an atom from 0.1 --> -0.1 or from 0.1 --> 0.9
Of course these 2 structures are equivalent, but the interpolated 0.0 is different from 0.5