The ABINIT Users Mailing List ( CLOSED )

[Masayoshi Mikami <mmikami@rc.m-kagaku.co.jp>]

Dear Jiang Fangdan,

Please, read the Netiquette before you send your question.
http://www.abinit.org/community/?text=netiquette

I forward your question to the forum ML. I expect you subscribe to it.
(You would have more chances to gain more answers from the forum,
 not from an individual who might not be available...)

If you want to treat the CuInSe2, for example,
http://www.ccp14.ac.uk/ccp/ccp14/ftp-mirror/powdcell/Powder_Cell/ 
structure_files/CUINSE2.TXT

You might want to set:
acell  5.781 5.781 11.642  angstrom # (or "2*5.781 11.642 angstrom")
spgroup 122
brvltt -1 # see the manual. check also "chkprim" in passing
natrd 3
typat 1 2 3
xred
 0.     0.      0.      #Cu
 0.     0.      1/2   # In
 0.226   1/4     1/8  # Se
znucl 29.0 49.0 34.0

Then you should see the output(log) to understand "rprim".
It would be a good idea to try also with "brvltt 2", as your homework.

You might want to consult with the help files on the WEB site,
e.g.
http://www.abinit.org/Infos_v4.6/keyhr.html
http://www.abinit.org/Infos_v4.6/vargeo.html#spgroup
http://www.abinit.org/Infos_v4.6/vargeo.html#brvltt
http://www.abinit.org/Infos_v4.6/spacegrouphelpfile.html

Everything in the ABINIT tar.gz archive ! Have fun !

Good continuation,
Masayoshi

On 2006/10/18, at 22:19, jfd03@mails.tsinghua.edu.cn wrote:

> Dear Masayoshi Mikami,
>     I'm a new abiniter. Now, I've problem in setting up the  
> structural model of CuInSe2 (space group: I-42D, #122). As  
> mentioned in the ABINIT help files, rprim give the three  
> dimentionless primitive vectors in real space. For instance, the  
> rprim should be like:
> rprim
>      -0.5   0.5   0.5
>       0.5  -0.5   0.5
>       0.5   0.5  -0.5
>     Then, multiplied by acell (X,Y,Z), the dimentional primitive  
> vectors rprimd should be like:
> rprimd
>     X*-0.5  X*0.5   X*0.5
>     Y*0.5   Y*-0.5  Y*0.5
>     Z*0.5   Z*0.0   Z*-0.5
>     So, if I want to set up a structure with lattice parameter  
> a=b=c=10 Bohr, set acell (10, 10, 10) is ok. #122 space group is  
> recognized. However, if I want to set up a structure with lattice  
> parameter a=b=10, c=20 Bohr, how could I do?
> Case 1:
>     Set acell (10, 10, 20), the output shows that rprimd is like:
> rprimd
>     -5.0   5.0   5.0
>      5.0  -5.0   5.0
>     10.0  10.0 -10.0
>     And #40 space group is recognized. Not the one I want.
> Case 2:
>     Revise rprimd to be:
> rprim
>      -0.5   0.5   1.0
>       0.5  -0.5   1.0
>       0.5   0.5  -1.0
>     In this case, rprimd remains the same as for acell (10, 10,  
> 10). Still, #122 space group is recognized. But, how could I do  
> when a=b=11, c=10? Keep on revising rprim?
>     Apparently, I've some mis-understanding about acell and rprim,  
> and even xred. And I do appreciate if you could give me some  
> thought on these things. Show me a way to solve the problem above  
> is thankful.
>
> Best regards,
>
> Jiang Fangdan