[Wien] Supercells and molecules

[John Appleton]

Dear all,
  the WIEN2k supercell method is still not obvious to me. StructGen gives 
  all these options for lattice structures (P, F, H, e.t.c). How does one build a
   supercell starting from StructGen. The FAQ is not explicit on that, and I think a 
  step-by-step example for supercell calculations would help those of us who 
  are beginners and doing independent calculations.
   
  I posted a mail with regards to computing the total energy of 
  isolated CO2 molecule. Someone suggested the supercell method. 
  The truth is that I simply don't get how to build a supercell. 
  The answer to the supercell FAQ does not elucidate on how to build supercells. 
  So, I tried the FCC lattice method to computing Cohesive energies of 
  atoms outlined in the FAQ ( note that I used the P option for the 
  lattice instead of F because of warnings in "x sgroup").
   
  A couple of other Questions concerning RMT and bond lengths:
   
  (1) Should I position O and C such the distance is close to 
       the experimental C=O bondlength?
   
  (2) If I decide to implement point (1) above then I'm forced to choose the RMTs 
       such that 
   
      RMT(C) + RMT(O)  less than or equal to C-O bond-length
  otherwise I get an error in the nn list.
   
  (3) These choices of RMT in point (2) are small so I have to decrease RKMAX and increase GMAX. Otherwise I get all manner of warning (especially RKMAX 
  reduced because of NNTMAX).  What is an appropriate RKMAX and GMAX in
   such a situation? 
   
   Can anyone suggest a nice, simple and smooth way to compute the energy 
  of the isolated molecules. 
   
  Thanks!
   
  J. Appleton
   
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20060125/a89a1cb1/attachment.html