[Wien] Promoting Ge 3s shell to valence

[Priyanka Seth]

Dear all,

I am trying to do a calculation on a Ge compound where I would like to 
promote the core 3s states of Ge (configuration [Ar] 3s2 3p2) to 
semicore and hence be included in the valence states. I am hoping this 
will have some convergence problems I had in the scf cycles.

However, I am not able to select an energy cutoff during lstart that 
allows me to do this:

  SELECT ENERGY to separate core and valence states:
   recommended: -6.0 Ry (check how much core charge leaks out of MT-sphere)
   ALTERNATIVELY: specify charge localization (between 0.97 and 1.0) to 
select core state
-12.5
   SELECT ENERGY to separate core and valence states:
   recommended: -6.0 Ry (check how much core charge leaks out of MT-sphere)
   ALTERNATIVELY: specify charge localization (between 0.97 and 1.0) to 
select core state
-12.0
STOP LSTART ENDS
0.266u 0.004s 11:47.23 0.0%    0+0k 0+1384io 0pf+0w

where the energy of the states are as follows:

           E-up(Ry)      E-dn(Ry)   Occupancy   q/sphere  core-state
   1S    -803.903411   -803.898620  1.00  1.00    1.0000  T
   1S    -803.903411   -803.898620  1.00  1.00    1.0000  T
   2S    -100.595366   -100.593339  1.00  1.00    1.0000  T
   2S    -100.595366   -100.593339  1.00  1.00    1.0000  T
   2P*    -89.575687    -89.572019  1.00  1.00    1.0000  T
   2P*    -89.575687    -89.572019  1.00  1.00    1.0000  T
   2P     -87.237957    -87.234392  2.00  2.00    1.0000  T
   2P     -87.237957    -87.234392  2.00  2.00    1.0000  T
   3S     -12.310095    -12.304838  1.00  1.00    1.0000  T
   3S     -12.310095    -12.304838  1.00  1.00    1.0000  T
   3P*     -8.676791     -8.670481  1.00  1.00    0.9999  F
   3P*     -8.676791     -8.670481  1.00  1.00    0.9999  F
   3P      -8.361586     -8.355077  2.00  2.00    0.9999  F
   3P      -8.361586     -8.355077  2.00  2.00    0.9999  F
   3D*     -2.187019     -2.174623  2.00  2.00    0.9944  F
   3D*     -2.187019     -2.174623  2.00  2.00    0.9944  F
   3D      -2.144032     -2.131383  3.00  3.00    0.9941  F
   3D      -2.144032     -2.131383  3.00  3.00    0.9941  F
   4S      -0.907245     -0.814618  1.00  1.00    0.6202  F
   4S      -0.907245     -0.814618  1.00  1.00    0.6202  F
   4P*     -0.336907     -0.246809  1.00  0.00    0.3584  F
   4P*     -0.336907     -0.246809  1.00  0.00    0.3584  F
   4P      -0.322467     -0.233001  1.00  0.00    0.3414  F
   4P      -0.322467     -0.233001  1.00  0.00    0.3414  F

I found an earlier question on the mailing list where the user wished to 
use a very low cutoff (~ -30 Ry) here: 
http://www.mail-archive.com/wien%40zeus.theochem.tuwien.ac.at/msg05396.html. 
What I am trying to do is far less extreme, I am only promoting one 
shell to the valence. Why can I not do this in lstart, and is there 
another way around this problem?

Thank you in advance for your help.

Best wishes,
Priyanka Seth