VASP Forum

Dear All,

I am trying to obtain simulated STM images for a perovskite system. I have obtained the DOS. Please see the atachements.
I want to see the Iodine atoms in the system. Below is one example of my INCAR file. My question is to observe this I atoms which potentail window I should select.? Is it 2.5 to 3.
Please advise me on this.

SYSTEM = R STM partial charge density EINT 0.0 to 0.5
ISTART = 1
ICHARG = 11
ENCUT  = 450
PREC   = Accurate
LREAL  = Auto
LORBIT = 11
NWRITE = 2
ADDGRID = .TRUE.
LASPH   = .TRUE.
ISMEAR = 0
SIGMA  = 0.1
NELM   = 100
NELMIN = 6
EDIFF  = 1E-07
ALGO   = Normal
NSW    = 0
IBRION = -1
ISYM   = 0
IVDW   = 11
LVHAR  = .TRUE.
LWAVE  = .FALSE.
LCHARG = .FALSE.
LPARD = .TRUE.
LSEPK = .FALSE.
LSEPB = .FALSE.
NBMOD = -3
EINT = 0.0 0.5

Dear gayani_nadeerapallewela

Unfortunately, STM simulations are usually a bit of a trial-and-error procedure.
Please read our tutorial on STM and the wiki article on Partial charge densities and STM simulations carefully.

As mentioned in the tutorial, I would advise you to plot the DOS of the pz shell of iodine, because this will likely contribute the most to the STM picture and help you to select the correct energy window.
For the data you show in the DOS now, I would start with EINT= -1.5 -0.5 (assuming NBMOD=-3) to sample the largest peak of iodine in the occupied states. But this also depends on the bias voltage (positive or negative) from the experiment.

In practice, you often have to try a bunch of settings until you see something that resembles the experimental picture. Then you can start to analyze the features.
STM simulations in the Tersoff-Hamann approximations are very seldom predictive, but they can help to understand and interpret the experimental results.

Cheers, Michael

Dear Michael,

Thank you for your advise.
I will plot the pz of Iodine.

Can you please expalin what do you mean by this "But this also depends on the bias voltage (positive or negative) from the experiment."

Experimentally they have observed Iodine around 2.5V bias (positive bias)

Thank you
Gayani

Dear Gayani,

A positive bias voltage means the experiment probes the unoccupied states. However, you don't get much DOS there from iodine in your plot, so I suggested looking at the occupied states, but I was not aware that your experimental data were from a positive bias voltage.
So you might indeed scan carefully with a positive bias. But make the window not too large, and don't expect the bias from the experiment to match the one you need to set in EINT to get the matching picture. As I said, trial-and-error is usually required.

Cheers, Michael