[Wien] more about tisi2

[Peter Blaha]

Apparently you already realized that spacegroup Fddd (70) has two 
different settings, and WIEN requires the setting WITH an INVERSION center.

It seems you have a version of w2web, where there is a problem with 
a,b,c inputs and wrong information will be printed into the struct file.
The next update will fix this, in the meantime you must put a.b.c. 
directly into the struct file in an editor.

This is an F-centered structure, thus the reciprocal lattice is 
B-centered and all 3 B-vectors contain 2pi/a, 2pi/b and 2pi/c, thus have 
the same length.

The problem with lapw0 might also result from a bug with your version.
Edit case.in0 and put   0 0 0 2.0  in the last line.

Vicki Keast schrieb:
> Dear All,
> 
> Further to my previous email about having trouble getting TiSi2. The 
> generation of additional and wrong atoms was probably due to us entering 
> the structure incorrectly. However we are still having trouble getting 
> this structure to work.
> 
> One unusual characteristic is that although the y lattice parameter is 
> different to the x lattice parameter, and stays this way in the the text 
> of the structure file, when we look at the structure file through w2web, 
> they are the same. And if you then save it through w2web, they are 
> overwritten as the same.
> 
> Another thing I don't understand is that when we run kgen, the number of 
> k-points in all three directions is the same, even though the y-lattice 
> parameter is nearly half the length of the other two.
> 
> Finally the scf cycle will not run and lapw0.error has the following 
> information.
> 
> 'SETFF1'- ifft too small in xcpot3
> 'SETFF1' - 2*(KKK+1) LARGER THAN IFFT PARAMETERS IN XCPOT3
> 'SETFF1' - KKK= 0 0 30
> 'SETFF1' - IIx= 0 0 30
> 'SETFF1' - IFFT=  64  64  60 
> 
> 
> We have also tried directly putting in all the atom positions and 
> letting wien sort out the space group and symmetry, which it does as we 
> would expect except for a shift of the origin. In this case the lattice 
> parameters at least stay as they should be, but we still get an 
> identical error to the one above when we try the scf cycle.
> 
> The structure is 70 Fddd (a=8.27, b = 4.80, c=8.55) Ti at 1/8, 1/8, 1/8 
> and Si at 0.4615, 1/8, 1/8.
> 
> As I only receive the weekly digest, could any responses be sent 
> directly to me as well.
> 
> Thanks and best regards,
> Vicki

-- 
-----------------------------------------
Peter Blaha
Inst. Materials Chemistry, TU Vienna
Getreidemarkt 9, A-1060 Vienna, Austria
Tel: +43-1-5880115671
Fax: +43-1-5880115698
email: pblaha at theochem.tuwien.ac.at
-----------------------------------------