[Pw_forum] The number of bands increases with c-axial

[Stefano Baroni]

Dear all:


On Aug 29, 2011, at 5:11 PM, sreekar guddeti wrote:

>  Respected Mr. Huang
> 
> 
> When I computed the band structure of bi-layer graphene.
> 
> I found that if I fix the displacement of two layers (3.3A) and other
> parameters. Change the c-axial (celldm(3)) only.
> 
> The number of bands will be different.

... as it should. It's just a matter of state counting, I believe. Let's consider a simple example, where you compare two structures, one having the c parameter twice as large as the other (say, c and 2c), and let's concentrate on what happens along the z axis. k=0 (kz=0, i should say) of the first structure will correspond to TWO wavevectors (k=0 and k=pi/c) of the second. Hence, when you calculate the band structure of the first, for any value of kz you will obtain twice as many states as in the second (corresponding to kz and kz+pi/c)

> The only possible way I can interpret this (I may be wrong) is the variation in the d-like bands (clutter of bands just above the Fermi level). I did a quick(ecut 30,360 Ry) calculation using your 'c' parameters and I got a similar band structure (I restricted #of bands to 20). I think you may not have scaled the z-coordinates accordingly. To confirm the d-character of the clutter-bands, perform k-resolved DOS calculations... the output is really soothing to the eyes :P. If you intend to perform a precise calculation (of order of 1 meV as your ecut suggests), I may suggest convergence with degauss.

As said, I think that the explanation is simpler than this, but I may have misunderstood the problem in the first place ...

>  Finally, as I too am learning the code, I would like to ask if there is any issue to use other Bravais lattice instead of ibrav=0?

I do not think so, but ... why not ibrav=0?

> 24.62 A    equals to the result from VASP (should be the correct one)
> why should VASP be 'the correct one'... I wouldnt see it as an authority ;)

Here I agree, but I would also argue that the two codes (vasp and pwscf) should result in the same energy bands, when given the same input structure. Is this the case?

Hope this helps
Stefano B

---
Stefano Baroni - SISSA  &  DEMOCRITOS National Simulation Center - Trieste
http://stefano.baroni.me [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype)

La morale est une logique de l'action comme la logique est une morale de la pensée - Jean Piaget

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20110829/78aa1d81/attachment.html>