calculation of core excitation energy

Dear MRCC Developers,

I am interested in calculating core excitation energies of small molecules in EOMCC methods.
As I understand from the manual, the ciguess keyword should be used for that. But I have unable to find any examples. A sample input file, lets say for water, will be very helpful. Another question is can linear response MRCC schemes can used for the purpose?

With best regards,
achintya

Dear Achintya,
This is a LR-CCSD example for water where the initial guess is a single excitation from the core orbital to the LUMO:

basis=cc-pVDZ
calc=CCSD
mem=5000MB
nsing=2
core=corr
ciguess=on
1
1 1 6 1 1.0
diag=follow

geom=xyz
3

H -0.756940 -0.520501 0.000000
H 0.756940 -0.520501 0.000000
O 0.000000 0.065584 0.000000


LR-MRCC is also available. In addition to the above input, you only need to specify the active space (see keywords nactv, nacto, or active).

Dear Prof Kallay,

Both the LR-CCSD and LR-MRCCSD is working fine.
However, in big basis set the excited state calculations are facing convergence problem, which I expected from my previous experience with EOMCC. Is there any way to change the default convergence or increase the number of iterations in Davidson's method?

With best regards,
achintya

Dear Achintya,
The max. number of iterations is hardwired and cannot be controlled by keywords. To change it edit file MRCCCOMMON and set the value of parameter maxit.
A more convenient solution is to restart the calculation from the unconverged wave function, see keyword rest.

Dear Prof Kallay,

I found the first option more convenient, obviously at the expense of more memory requirement.
But its working fine.
Thank you for the kind help.

With best regards,
achintya