TURBOMOLE Users Forum
TURBOMOLE Modules => Jobex: Structure Optimization and Molecular Dynamics => Topic started by: golden on April 16, 2012, 05:15:49 PM
-
Dear All,
When I run a single point calculation the HOMO-LUMO gap is a negative value.
HOMO-LUMO Separation
HOMO: 929. a 233 ag +0.04893055 H = +1.33147 eV
LUMO: 927. a 234 ag +0.04779418 H = +1.30055 eV
Gap : -0.00113637 H = -0.03092 eV
!! WARNING: HOMO-LUMO Gap is negativ !!
Number of MOs= 2618, Electrons= 927.00, Symmetry: ci
and the occupation is like;
where it has a unoccupied orbitals in-between some occupied orbitals.
931. a 235 ag +0.075774 H = +2.062 eV
930. b 234 ag +0.051708 H = +1.407 eV
929. a 233 ag 1.000 +0.048931 H = +1.331 eV
928. b 233 ag +0.048597 H = +1.322 eV
927. a 234 ag +0.047794 H = +1.301 eV
926. b 231 au 1.000 +0.005601 H = +0.152 eV
925. b 230 au 1.000 +0.004848 H = +0.132 eV
924. a 231 au 1.000 +0.004691 H = +0.128 eV
923. a 230 au 1.000 +0.004060 H = +0.110 eV
and in the ridft file it give a warning like;
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
! !
! WARNING : ORBITAL ENERGIES SHOULD BE NEGATIVE FOR OCCUPIED ORBITALS !
! !
! orbital energy ( 0.04893) of occupied MO 233ag is > 0 !
! orbital energy ( 0.00012) of occupied MO 229au is > 0 !
! orbital energy ( 0.00406) of occupied MO 230au is > 0 !
! orbital energy ( 0.00469) of occupied MO 231au is > 0 !
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
! !
! WARNING : ORBITAL ENERGIES SHOULD BE NEGATIVE FOR OCCUPIED ORBITALS !
! !
! orbital energy ( 0.00100) of occupied MO 229au is > 0 !
! orbital energy ( 0.00485) of occupied MO 230au is > 0 !
! orbital energy ( 0.00560) of occupied MO 231au is > 0 !
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
This is after not using fermi , and if fermi is used, I am getting a partial occupation in the orbitals;
931. a 235 ag +0.075928 H = +2.066 eV
930. b 234 ag 0.075 +0.049910 H = +1.358 eV
929. b 233 ag 0.076 +0.049899 H = +1.358 eV
928. a 234 ag 0.420 +0.048656 H = +1.324 eV
927. a 233 ag 0.429 +0.048634 H = +1.323 eV
926. b 231 au 1.000 +0.005095 H = +0.139 eV
925. b 230 au 1.000 +0.005070 H = +0.138 eV
924. a 231 au 1.000 +0.004507 H = +0.123 eV
923. a 230 au 1.000 +0.004482 H = +0.122 eV
So I have 2 questions.
1) Is the negative HOMO-LUMO gap possible?
2) Is partial occupation possible solution?
partial occupation occurs due to the fermi key word, but as suggested in the fermi, although it smears the electron occupation it should at the end should give a ground state occupation of the electrons, but it seems not to be working for the calculation.
thanks for the help given in advance.
-
Having the HOMO higher in energy than the LUMO is not a valid ground state, and is almost certainly an artifact of the SCF optimisation attempting to force integer occupations.
You'll see in the Fermi-smeared calculation that there are two sets of two orbitals with almost the same energy. While a partial occupation isn't a legitimate physical ground state either (though it's often good enough for an intermediate optimisation step), I think it indicates that your system in the current geometry has a multiconfigurational ground state, which DFT can't model properly. Reducing the end Fermi temperature will probably move those 0.075 parts down into the lower orbitals to give you two 0.5 occupations, but to get better than that I think you're going to have to change the geometry or at least the symmetry restrictions, to break the symmetry of those two orbitals, or else use a more sophisticated method that can model the multiconfigurational character.
I'm a little concerned about the orbitals having positive energies, which could point to additional problems in your setup. Are you performing calculations on an anion, and if so, have you included augmented functions in the basis set?
-
Hi,
thank you very much for the helpful comments.
I was using def2-SV(P) on all atoms and RI method to calculate the energy. Do I need to add anything to represent the augmented functions?
I was changing the occupation and then the fermi temperatures to get something reasonable, and I tried lot of combinations and when I tried
$fermi tmstrt=50.00 tmend= 10.00 tmfac=1.000 hlcrt=1.0E-01 stop=1.0E-03 nue= 1
I got a reasonable HOMO-LUMO gap and a positive gap of ;
HOMO-LUMO Separation
HOMO: 927. a 233 ag +0.04651231 H = +1.26566 eV
LUMO: 928. b 233 ag +0.04979731 H = +1.35505 eV
Gap : +0.00328501 H = +0.08939 eV
Number of MOs= 2618, Electrons= 927.00, Symmetry: ci
931. a 235 ag +0.075836 H = +2.064 eV
930. b 234 ag +0.050579 H = +1.376 eV
929. a 234 ag +0.050283 H = +1.368 eV
928. b 233 ag +0.049797 H = +1.355 eV
927. a 233 ag 1.000 +0.046512 H = +1.266 eV
926. b 231 au 1.000 +0.005351 H = +0.146 eV
925. b 230 au 1.000 +0.005116 H = +0.139 eV
924. a 231 au 1.000 +0.004493 H = +0.122 eV
923. a 230 au 1.000 +0.004287 H = +0.117 eV
922. b 229 au 1.000 +0.001081 H = +0.029 eV
921. a 229 au 1.000 +0.000180 H = +0.005 eV
920. b 232 ag 1.000 -0.024338 H = -0.662 eV
But still the energy of the 921 - 927 is positive (occupied orbitals) which is confusing.
Surely there should be a far better approach than trial & error method, for changing the occupations and fermi temperatures to get a reasonable answer. :-\ Still I am not quite certain to make of this positive energy values.
thank you very much.
-
Hi,
Fermi smearing is usually very effective in finding the ground state configuration, you just have to use a low enough temperature. And concerning the positive orbital energies: like Ian said, these are typical for anionic species (with DFT). Typical ways to get negative orbital energies instead are 1) using the COSMO solvation model; 2) using point charges around the system. COSMO is definitely more straightforward, but without any more details on your system it's difficult to say whether this is a reasonable choice. Adding diffuse basis functions probably won't help (you could test def2-SVPD, but the calculation will also be more time-consuming due to the diffuse functions)
Best,
Antti
-
Hi,
but without any more details on your system it's difficult to say whether this is a reasonable choice
My system consists of Au25(SCH3)18 , where the SCH3 groups are like a mono layer around the gold core.
So is the positive energy values a unreasonable answer ?
thank you very much for the help.
-
Hi,
In this case the positive orbital energies do sound rather strange and it's probably better to fix the issue. Is the geometry reasonable? Also, def2-TZVP might be better for the Au atoms (although more expensive).
Antti