[Wien] Si vacancy

[fabella]

Hello,

I have completed calculations of the neutral Si vacancy in Si.

I have results for both the ideal (w/o vacancy), not relaxed, and 
relaxed.  The cohesive energies are good (580Ry/Si atom), energy and 
force convergence good.

But while comparing (with literature) the DOS for both not 
relaxed/relaxed cases are slightly different.  For example, we'd expect 
A1 and T2 symmetry (after relaxation?), there respected peaks show in 
the DOS.  In my calculation of the not relaxed case it appears that the 
initial degeneracy is already lifted (A1 and T2) and are deep centers.  
In addition, it appears that in the relaxed case the peak at about -8 
eV (belonging to A1) disappears (present in not relaxed case and 
literature), while there remain states in the gap, but are shifted 
toward the valence band.

Is there something special about the basis that Wien2k is using for 
this vacancy calculation?  The plane waves are represented about this 
local "isolated" defect, does it prematurely form bonds within the 
vacancy (not relaxed case)? and therefore in relaxation the DOS is just 
shifted?

Has anyone calculated the Silicon Vacancy or have had similar issues to 
deal with in a typical Vacancy problem?  Any information would be 
helpful...

Thank You

Wesley J. Fabella