Post Hartree-Fock HOMO-LUMO gap
Clicked A Few Times
7:14:27 AM PDT - Mon, Jun 2nd 2014
Hello NWChem community,

I am attempting to estimate HOMO-LUMO gaps (and UV absorption wavelength) with NWChem.

However, the module does not print HOMO-LUMO after the SCF cycle (for example, MP2, DFT, and CCSD(T) on H2 molecule).

How can I get NWChem to print out the HOMO-LUMO gap post-Hartree Fock energy for H2?

Also,is there a way for NWChem to print out whether HOMO-LUMO is sigma-sigma* or pi-pi* or anything like that?

Here is my input (last line modified for CCSD or SCF)

Quote:hhg
echo

start molecule

title "Cytosine optimization"
charge 0

geometry units angstroms print xyz autosym
H 0.0 0.0 0.0
H 0.0 0.0 0.74
end

basis
* library aug-cc-pVQZ
end

scratch_dir scratch

ccsd
nodisk
end

dft
noio

direct

xc ssb-d
end

task dft energy

Forum Regular
6:53:11 PM PDT - Tue, Jul 1st 2014
Hi,

Indeed the code does not print the HOMO-LUMO gap explicitly. However, at the end of the SCF or DFT calculation it will print the orbital analysis. For H2 (I have used SCF instead of DFT so I could simply use "task ccsd energy", and I used the cc-pvdz basis set) I got:

Vector    1  Occ=2.000000D+00  E=-5.924110D-01  Symmetry=a1g

             MO Center=  0.0D+00,  0.0D+00, -4.0D-17, r^2= 7.1D-01

  Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function

 ----- ------------  ---------------      ----- ------------  ---------------

    1      0.403665  1 H  s                  6      0.403665  2 H  s

    2      0.173763  1 H  s                  7      0.173763  2 H  s



Vector    2  Occ=0.000000D+00  E= 1.974400D-01  Symmetry=a2u

             MO Center=  3.3D-17,  6.2D-17, -9.0D-16, r^2= 2.7D+00

  Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function

 ----- ------------  ---------------      ----- ------------  ---------------

    2      1.958656  1 H  s                  7     -1.958656  2 H  s

    1      0.150244  1 H  s                  6     -0.150244  2 H  s

In this table "Occ" refers to the orbital occupation and "E" to the corresponding orbital energy. So it is relatively easy to identify vector 1 as the HOMO and vector 2 as the LUMO, giving a HOMO-LUMO gap of 0.19744 - -0.59241 = 0.78985 Hartree. Of course if you want you can write a little script to extract that information. I am more than happy to put that in our contrib directory, so you have it in future.

Best wishes, Huub

Just Got Here
8:41:25 AM PDT - Sat, Apr 23rd 2016
What should you do if you want gaps that are 0.1 hartree or less?

Just Got Here
8:41:37 AM PDT - Sat, Apr 23rd 2016
What should you do if you want gaps that are 0.1 hartree or less?


Forum >> NWChem's corner >> NWChem functionality