bug of excitation enetgy calculation by CC(2)-LRT

[zork]

Dear Developers,

In the following test input, the ^1Sigma^+ and ^1Pi excited states of the HF molecule are calculated by the CC(2)-LRT method.

Code: Select all

ATOMBASIS
HF
test
Atomtypes=2 Angstrom
Charge=1.0  Atoms=1      Basis=3-21G
H          0.0000000000        0.0000000000        0.0000000000
Charge=9.0  Atoms=1      Basis=3-21G
F          0.0000000000        0.0000000000        0.9500000000

**DALTON INPUT
.RUN WAVE FUNCTIONS
**WAVE FUNCTIONS
.CC
*CC INPUT
.CC(2)
*CCEXCI
.NCCEXC
  1 1 1 0
**END OF DALTON INPUT

The printed CCS-LRT results are okey:
^1A1 16.91703 eV
^1B1 11.27095 eV
^1B2 11.27095 eV
but the CC(2)-LRT results are strange:
^1A1 26.98359 eV
^1B1 3.13472 eV
^1B2 10.48748 eV

But if only one excited state is specified after .NCCEXC, reasonable CC(2)-LRT results can be obtained:
^1A1 16.36293 eV
^1B1 10.48748 eV
^1B2 10.48748 eV

This problem was not found for CCS, CC2, CCSD, CC3, and CCSDR(3) with LRT calculations. The 2015.1 Version of Dalton was used. Maybe it has been solved in the 2016 version? Thanks!

[xiongyan21]

I repeated your calculation and noticed that CC(2) excitation energies are not in a table but in front of the CCS table.
If the input includes this, i.e.,
.NCCEXC
1 1 1 0
the results of CC(2) energies are
^1A1 16.362925 eV
^1B1 10.487474 eV
^1B2 10.487474 eV
based on the CCS excitation energies.

...

*******************************************************************
* *
* *
*---------- CALCULATION OF PERTURBATIONAL CORRECTIONS >---------*
* *
*******************************************************************



======================================================================
### ###
### ###
### Perturbational Corrections to Excitation energies. ###
### ###
### ###
### Calculating doubles corrected CCS excitation energies. ###
### ###
### ###
======================================================================




CC-model Excitation no. Hartree eV
---------------------------- -------- --------
CCS 1 0.621689 16.917028
---------------------------- -------- --------
...
---------------------------- -------- --------
CC(2) 1 0.601326 16.362925
============================ ======== ========


...


The results you find strange are CIS(D) excitation energies.




For this calculation, the output of Dalton2015.1 is identical to that of Dalton2016.0.

Very Best Regards!