Thank you very much Abraham for your EXTENSIVE and useful answer.

Also Thanks Yun-Peng, it is an interesting software.


-Abbas


On Thu, Sep 29, 2011 at 9:36 PM, Abraham Hmiel <[email protected]> wrote:

> Hello Abbas,
>
> This is usually a multi-step process. At a general level, you need to look
> at the quantities derived from orbital populations. Projected density of
> states is an important one. If you use the pdos utility or a special script
> you write, you can separate the PDOS data file into PDOS aggregated onto a
> specific atom, group of atoms, orbital, or selected orbitals on selected
> atoms. You should re-read the section in the manual about PDOS, how to
> create it, etc. The other thing you can look at is the crystal orbital
> overlap population by running the COOP utility mprop. Look elsewhere on the
> list for information in how to set that data file, basically, you select
> pairwise atoms/orbitals and a range of distances and it will calculate COOP.
> You will have to manually prepare a lot of data and look at it side by side
> to determine which orbitals are interacting strongly at what energies.
>
> From that, you can see a general energy range of where to look for the
> local (spatial) electronic density of states (plotted with grid2cube), which
> can be supplied as a range of energies in the fdf. The other way to do is is
> to plot a lot of wavefunctions and use the denchar utility to make .cube
> files then plot isosurfaces of the complex wavefunction at a particular grid
> density using a program of your choosing. If you look at them side by side
> in sequence, you can discern characteristics that mark the spatial patterns
> of what you're looking for. Unfortunately, sometimes these visualization
> files can get quite large and cumbersome for some computers to handle, but
> either way you should then have a pretty good idea of "which band" since
> they are numbered sequentially up from the lowest energy state. It's not
> going to tell you "BEEP: THIS BAND IS PI*" but it is a judgment you will
> have to make.
>
> And then have the band structure handy for reference & to put everything in
> perspective. You should write a script or something that can identify an
> energy band in the k-range you specify in a crystal. Combined, these things
> are all ingredients of good convincing arguments in a publication.
>
> Best of luck,
>
> --
> Abraham Hmiel
> Katherine Belz Groves Fellow in Nanoscience
> Xue Group, College of Nanoscale Science and Engineering at SUNY Albany
>
>
> On Thu, Sep 29, 2011 at 9:54 PM, Abbas Ebnonnasir <
> [email protected]> wrote:
>
>> Dear SIESTA users,
>>
>> I'm trying to find out how I can identify each band in band structure
>> plot, resulted from SIESTA, i.e, I need to know which band belongs to which
>> molecular/crystal orbital (which one is Pi, which one is Pi*, Sigma....).
>>
>> Any comment will be appreciated.
>> Thank you in advance.
>>
>> All the wishes,
>>
>> -Abbas
>>
>>
>>
>
>

Responder a