QuantumATK Forum

QuantumATK => General Questions and Answers => Topic started by: ramkrishna on October 6, 2012, 07:42

Title: Ge Huckel basis set
Post by: ramkrishna on October 6, 2012, 07:42

Dear Sir,
        To calculate the band structure of Si (/SiNW/ SiC )by using extended huckel method, we have to set the basis as Cerda.Silicon(GW diamond or SiC) and it gives the band structure properly but Ge (or GeC like SiC)  does not have Cerda basis set. It shows Hoffmann/Muller basis set and that is not giving proper band structure. Does Ge (or GeC) have Cerda basis set? How can I modify my Script generator for that? Please help.

I want to one more thing. Can you please tell me the procedure to build GRAPHANE having armchair, zigzag, boat configuration ? I am trying to do from Graphene, but it is not coming. Please help.

Thanking you,
Tamkrishna

Title: Re: Ge Huckel basis set
Post by: Anders Blom on October 6, 2012, 22:26

No, there are no parameters of Cerda's website for Ge. Probably you are better off anyway using either DFTB or some other tight-binding model.

As for the graphane question, I find this provides an interesting educational case for how to set up a new geometry in ATK from basic information, so let me tell you how I would approach it.

• First of all, always remember that Google is your friend. So, step 1: search "graphane boat chair".
• The third hit (in my case, it may differ for you) looks promising: a paper on arXiv: http://arxiv.org/abs/cond-mat/0606704
• Opening the paper, we find it is exactly what we need - on page 4 is a complete description of both the chair and boat configurations of graphane
• The information given here is precisely what goes into a CIF file - so let's make one. If you don't have a template for a CIF file, you can always make one from VNL - just choose a simple system (like - graphene!) from the Database, right-click the Stash item and choose Export and make a CIF file. It will look like this:

Code

data_global
_cell_length_a 2.4612
_cell_length_b 2.4612
_cell_length_c 6.709
_cell_angle_alpha 90
_cell_angle_beta 90
_cell_angle_gamma 120
_symmetry_space_group_name_H-M 'P -1'
loop_
_symmetry_equiv_pos_as_xyz
  'x,y,z'
loop_
_atom_site_label
_atom_site_fract_x
_atom_site_fract_y
_atom_site_fract_z
C 0 0 0
C 0.333333 0.666667 0

(Yes, I know this is not a proper CIF file for graphene - we are working on exporting the symmetries too. But it works, and is sufficient for our present purpose.)

• We just need to replace the information with the relevant numbers for graphane.
• Doing that, and also inserting the correct Hermann-Mauguin symbols, we end up with the attached files, where I also extended the notation somewhat for future purposes to include the Wyckoff symbol in the CIF file.
• NOTE: We must also remove the symmetry operations, since they apply only to P1. If no symmetry operations are given in a CIF file on import, VNL uses a set of built-in operations if it can match the H-M symbol, which it will be able to do in this case.
• Drop the files on the Builder and presto!

I should probably convert this answer into a mini-tutorial :)