[Wien] -Isolated molecules of benzene - my approach has not led me further than "dstart"

[Andrei Postnikov]

Alexei,

I would sugest that, "to start with", it would be good enough
to use the hexagonal symmetry, unless you'll have special reasons
(in your subsequent calculations) to make the atoms non-equivalent.

This would imply all C equivalent and all H equivalent,
AND a hexagonal unit cell.
Figure out its reasonable dimensions - large enough to keep
repeated molecules reasonably apart, but not too large.
Empty space and low symmetry both come very expensive in LAPW.
Your cubic cell is anyway inefficient because the benzene molecule
is flat, and you don't need as long C as A.

You'll have enough troubles with small H size and large cutoffs,
so first play around with the smallest cell size and highest possible
symmetry, then you can start to make things worse and closer to reality.

Good luck,

Andrei

 +-- Dr. habil. Andrei Postnikov ----- Tel. +49-541-969.2377 -- Fax .2351 ---+
 | Universitaet Osnabrueck - Fachbereich Physik, D-49069 Osnabrueck, Germany |
 +-- apostnik at uos.de --------- http://www.home.uni-osnabrueck.de/apostnik/ --+

On Fri, 27 Aug 2004, Komolov wrote:

| Dear experts in calculations of electronic structure of solids by means of Wien2K,
|
|
|
| I am trying to start Wien2K calculations on conjugated organic molecules and I chose isolated molecules of benzene C6H6 to start with, because I thought it would be a technically simple case, good for beginners.
|
|
|
| I have tried to rerun the examples (TiC, TiO2) enclosed to the software - it worked, with small troubles, though.
|
|
|
| I am asking you for an advice or a critical comment on a procedure I used to run the calculations on benzene, which have not led me further than "dstart" within the initialization of the calculations, reporting an error.
|
|
|
| I will describe my input to the calculation procedure below and in the .htm and .gif file attached.
|
| -Isolated molecules of benzene - Alexei>>W2K-
|
| I supposed the unit cell to be cubic with 15 Angstr side length - to make sure the distance between adjacent molecules is long enough.
|
| I used conventional literature data on the relevant bond lengths and angles (detailes are attached)
|
| I used 1.2 and 0.6 RMT for C and H respectively.
|
| I typed the necessary positions of the 12 inequivalent atoms as the fractions of the unit cell.
|
| --With this the programs "sgroup" and "symmetry" did not agree: they suggested to make 6 couples of inequivalent atoms so as to couple the left and the right sides of benzene molecule (according to the figure in the attachment), the programs suggested that I, in fact, had a lattice type 6_m_m, - I tried both to accept and not to accept that, and as soon as I accepted it they suggested a change again (that the lattice type should actually be 25_(p_m) or so). As soon as I accepted these symmetries - "lstart" would indicate inconsistency with Z-atomic number.
|
| I also checked the cell composition as 6 equivalent carbons and 6 equivalent hydrogens.
|
| I also tried to introduce distortions to the structure as manually typing 5-10% misplacements in the atom positions.
|
| I also tried to type different indexes to the atoms.
|
| In any case "dstart" would give an error with no diagnostics or an error: ROTDEF - (no symmetry found)
|
|
|
| Please write me a few words with a clue or a comment on what was wrong with my approach.
|
|
|
| Thank you in advance for your time
|
| Alexei Komolov
|
| St.Petersburg State University
|