Brillouin Zone Route and Bandstructure Problem

[Juan]

Hello,

        I'm trying to replicate the bandstructure of the BiFeO3 of the following paper,

http://folk.uio.no/ravi/allpapers/85-BiFeO3-prb.pdf

Taking all the computational details in consideration, the parameters I need to build the crystal are on page 5 (I'm doing the R3c structure). When I put all the parameters and go to the brillouin zone route points, it only lets me put G,X when in the paper, the bandstructure is around F, G, Z, F, L, G brillouin points (page 9 fig #8 b).

What I tried was to change the lattice type to triclinic using fractional option. This way the structure stayed almost the same and when putting Brillouin zone route, it gives more variables. Even though I had to use the points T, G, Z, T, Y, G, when checking the bandstructure, it appeared equal although the region of 0 to 6 eV is not equal to the one on the paper.

Even putting the parameters given on the paper, Im not getting the results. Is there something I'm missing on the general parameter calculations? Any recommendation will be great! Thanks!

[zh]

Have you tried the use of "Rhombohedral" lattice to do the calculations?

For the Brilliouin zone of a rhombohedral lattice:
i)  a > sqrt(2) *c :  Gamma =(0,0,0),  Z= (1/2, 1/2, 1/2), L = (0, 1/2, 0), F = (0,1/2, 1/2).
ii) sqrt(2) * > a:   Gamma =(0,0,0), Z= (1/2, 1/2, 1/2), L = (0, 1/2, 0), F = (1/2, 1/2, 0).

According to the above definition of high-symmetry k points, you can generate the  k points in the desired k-path and then do the band structure calculations.

[Juan]

        I've used the trigonal R3c spacegroup (rhombohedral settings). When I change the "Lattice parameters", it is already on "Rhombohedral". All I do is change the a and gamma angle as in the paper (and the atoms with corresponding distances). Once that, I put "New Calculator" and put K points to 10 in each and change to GGA (no SCF iteration). When I go to "bandsctructure" it only lets me put G and X as Brillouin zone route.

[Anders Blom]

See discussion on exactly the same topic here: http://quantumwise.com/forum/index.php?topic=2893.msg13483#msg13483

[Juan]

Hi again,

       Thank you very much for that info! Even though I've been doing now what I've seen in those, it appears there is an "IndexError: list index out of range". I found a post where you explain that it's because the NC file does not contain the object I'm trying to read. I still don't understand that means. I'm wondering if that has something to do with the values I'm putting to override the symmetry points? For the Brillouin zone, I used the points zh gave me:

i)  a > sqrt(2) *c :  Gamma =(0,0,0),  Z= (1/2, 1/2, 1/2), L = (0, 1/2, 0), F = (0,1/2, 1/2).
ii) sqrt(2) * > a:   Gamma =(0,0,0), Z= (1/2, 1/2, 1/2), L = (0, 1/2, 0), F = (1/2, 1/2, 0).

I have use both of them to test it. Same error message appears.

[HIMANI]

Directly go to the editor file of your script and change the brillioun route there.
But for this you need to define the additional brillioun zone points first. For that you need to know the (x,y,z) coordinates of each of the brillioun zone point you are using.

[Jess Wellendorff]

So I had allok at your script. The IndexError occurs because you try to read the BandStructure object before you have actually calculated and saved it. I have rearranged the script a bit and got the attached bandstructure plot. The modified script is also attached,

[Juan]

Thank you so much for the help!!