Re^4: Input data of SiF4(in English) ---It seems a bug of AkaiKKR.

Posted on : March 14, 2003 (Fri) 18:12:30

by takao kotani

I observed the same output you mentioned.
I changed 0.33 to 0.5 (higher symmetry). Then it works.
But 0.49 would not work. Maybe because of a bug in a routine to check the crystal symmetry.
Anyway the solids does not have good spece filling and requires empty sphere as below.

takao



===========================================================
I recommend you to install XtalEdit to Windows(is better) anyway.
You need to install python2.2 or above and Numerical python.
Not so difficult for Windows.
After XtalEdit started, put your input to EditorWindow
and push imort button of AkaiKKR.
I did it and it was well translated to csy form (standard format in XtalEdit).
Then push rasmol button. I could see the crystal structure.
(add
CV1=a*EX
CV2=a*EY
CV3=a*EZ
so as to see the cell by conventional vectors).

I also push Gen 0 so as to check the crystal symmetry.
It seem OK. The neibor pair tabel shown in ResultWindow is
ib jb cl1 cl2 Pos(jb)-Pos(ib) Dist sumrs Ovlp %
1 1 Si Si 4.800 -4.800 -4.800 8.314 0.000 -8.31-100.0
1 2 Si F 1.584 1.584 1.584 2.744 0.000 -2.74-100.0
1 3 Si F 1.584 -1.584 -1.584 2.744 0.000 -2.74-100.0
1 4 Si F -1.584 1.584 -1.584 2.744 0.000 -2.74-100.0
1 5 Si F -1.584 -1.584 1.584 2.744 0.000 -2.74-100.0
...
See Dist, which means distance. The crystal structure is OK?
It seems to have very huge empty space, so I think it will be better to add empty spheres as
0.000000000000a 0.000000000000b 0.000000000000c Si
0.330000000000a 0.330000000000b 0.330000000000c F
-0.330000000000a 0.000000000000b 0.000000000000c F
0.000000000000a -0.330000000000b 0.000000000000c F
0.000000000000a 0.000000000000b -0.330000000000c F
0.000000000000a 0.500000000000b 0.500000000000c Vc
0.500000000000a 0.000000000000b 0.500000000000c Vc
-0.500000000000a -0.500000000000b 0.000000000000c Vc
Vc means empty sphere. Anyway AkaiKKR might be not so suitable to calculate such materials.
It dependes on what quantities you want to calculate...
Anyway the input does not work now...
I have to report it to Prof.Akai.