[Wien] multiplicity

[Stefaan Cottenier]

> Just another question:
> If I now make 2*2*2 supercell of such unit cell, will I then get in
> band structure 8* more bands for O-2p for instance?


For a supercell, 2*2*2 refers to the conventional cell. Therefore, the
volume of the supercell is

8*2=16 times larger than the primitive cell if you select P for the supercell
8*2/2=8 times larger if you select B
8*2/4=4 times larger if you select F

In any case, the case.struct of your supercell will have for all atoms
the correct multiplicity in the primitive cell of the supercell, and
you can right away get the number of bands from this.

Stefaan


>
> ----- Original Message ----- From: "Stefaan Cottenier"
> <Stefaan.Cottenier at fys.kuleuven.be>
> To: <wien at zeus.theochem.tuwien.ac.at>
> Sent: Sunday, September 30, 2007 9:27 AM
> Subject: Re: [Wien] multiplicity
>
>
>>
>> There is nothing wrong. The multiplicity in the Wyckoff label (i.e.
>> the 8 in 8h) is the multiplicity in the conventional cell (a cube in
>> your case), while MULT in case.struct gives the multiplicity in the
>> primitive cell. For space groups of the I-type (as your nr. 87) the
>> conventional cell is twice as large as the primitive cell.
>>
>> Stefaan
>>
>> Quoting Igor Djerdj <Igor.Djerdj at mat.ethz.ch>:
>>
>>> Dear all,
>>> I am dealing with the structure VO2, Z=8 and space group 87. In my
>>> structure there are V in 8h, O1 in 8h and O2 in 8h. General
>>> multiplicity of s. g. is 16. When I insert my data in structgen, the
>>> programe generates for instance for V 4 equivalent atoms
>>> (multiplicity 4), instead of 8. The same occured for oxygens. Then,
>>> when I plot the band structure for O2p, I have 24 bands instead of
>>> 48. When I visualise the structure with Crysden, everything is
>>> perfect, i.e. the structure looks like as it should be. I am using
>>> spin-polarized calculation, so I obtain two band structure for each
>>> spin. Where is the problem, if any?
>>> Regards,
>>> Igor


Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm