Work function on metal oxide surface

[Jiaye Li]

Dear All

Recently, I have met eith a problem about calculating the workfunction of metal oxide surface. I took Indium Oxide for example.

TO get the coverged vaccum potential, I tried different vacum space (10A-50A). Unluckily, the vacum potetial didn't converge at all. Is there something wrong in my calculation?

My INCAR:

SYSTEM = IO
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
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 = 0 * # 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
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


KPOINTS:

IO
0
G
1 2 2
0 0 0



POSCAR:

IO
1.00000000000000
57.00 0.0000000 0.0000000
0.0000000 10.117000 0.0000000
0.0000000 0.0000000 10.11700
24 36
Selective dynamics
Direct
0.11495 0.24998 0.24998 F F F ! In
0.11495 0.74993 0.74993 F F F ! In
0.11495 0.46496 0 F F F ! In
0.11495 0.53504 0.50005 F F F ! In
0.11495 0.03509 0.49995 F F F ! In
0.11495 0.96501 0 F F F ! In
0.115 0.25007 0.75002 F F F ! In
0.115 0.75002 0.25007 F F F ! In
0.21382 0 0.24998 F F F ! In
0.21382 0.49995 0.75002 F F F ! In
0.22991 0.24998 0.03509 F F F ! In
0.22991 0.75002 0.46496 F F F ! In
0.22995 0.24998 0.53504 F F F ! In
0.22995 0.75002 0.96491 F F F ! In
0.24605 0.50005 0.24998 F F F ! In
0.24605 0 0.75002 F F F ! In
0.35758 0.76336 0.24012 T T T ! In
0.35758 0.26339 0.76005 T T T ! In
0.35758 0.73663 0.73998 T T T ! In
0.35759 0.23674 0.25984 T T T ! In
0.35708 0.02273 0.01491 T T T ! In
0.35708 0.97733 0.48504 T T T ! In
0.35705 0.52249 -0.01511 T T T ! In
0.35704 0.47755 0.51506 T T T ! In
0.04905 0.84482 0.8802 F F F ! O
0.04905 0.15518 0.61985 F F F ! O
0.04905 0.34477 0.1198 F F F ! O
0.04905 0.65523 0.38015 F F F ! O
0.05509 0.6067 0.65523 F F F ! O
0.05509 0.3933 0.84482 F F F ! O
0.05509 0.10665 0.34477 F F F ! O
0.05509 0.89335 0.15518 F F F ! O
0.07136 0.38015 0.3933 F F F ! O
0.07136 0.61985 0.10665 F F F ! O
0.07136 0.1198 0.89335 F F F ! O
0.07136 0.8801 0.6067 F F F ! O
0.15855 0.1198 0.10665 F F F ! O
0.15859 0.38015 0.6067 F F F ! O
0.15859 0.61985 0.89335 F F F ! O
0.15859 0.8801 0.3933 F F F ! O
0.17482 0.3933 0.15518 F F F ! O
0.17482 0.6067 0.34487 F F F ! O
0.17482 0.89335 0.84482 F F F ! O
0.17482 0.10665 0.65523 F F F ! O
0.18086 0.15518 0.38015 F F F ! O
0.18086 0.84482 0.1198 F F F ! O
0.18086 0.65523 0.61985 F F F ! O
0.18086 0.34487 0.8801 F F F ! O
0.292 0.35055 0.38513 T T T ! O
0.29203 0.64949 0.11481 T T T ! O
0.29195 0.85052 0.61489 T T T ! O
0.29192 0.1495 0.8851 T T T ! O
0.28969 0.10648 0.15871 T T T ! O
0.28968 0.89355 0.34129 T T T ! O
0.28969 0.60649 0.84124 T T T ! O
0.28969 0.39351 0.65872 T T T ! O
0.31274 0.87238 0.8924 T T T ! O
0.31274 0.12764 0.60759 T T T ! O
0.3126 0.62764 0.39243 T T T ! O
0.31262 0.37242 0.10748 T T T ! O

I used the same computational paramters to calculate oure metal surface(ie. In), which give rise to a accurate result.

Thanks in advance!

[admin]

please set ALGO=Fast instead of IALGO=48 (this is slightly slower, but safer)
it looks to me as if the system has a dipole moment along the surface normal (which is along x, am I correct?) because the left side is O-terninated whereas the right side is In-terminated In this case please set IDIPOL=1 to account for the polarity of the surface
if the number of valence electrons is odd, please set ISPIN=2

[Jiaye Li]

Thanks!

Yes, the surface normal is along x axis. The surface slab is constructed by alternating three layers of In and three layers of O.

I have set ALGO=Fast and IDIPOL=1. The system should be close shell so I keep ISPIN=1. I will follow the work to see what will come out. Thanks again!

[Jiaye Li]

I tried IDIPOL=1, but the vaccum energy doesn't converge. I also use a vaccum layer of more than 50 A, it didn't work. Is it reasonable that increasing the vaccum layer to a as large as possible value is able to get the right and converged vaccum energy, ie. 100A?

[alex]

Polar surfaces are messy. Try to saturate one side with e.g. H atoms or some other approaches from literature. And think about if your polar surface might exist at all. Mostly reconstruction appears to minimize dipole related forces ...
Since dipoles converge with 1/r you'll observe _very_ slow convergence ...

alex