[molpro-user] left-hand EOM-CCSD unconverged

[Tatiana Korona]

Dear Shaohong,

It seems that the LIIS procedure worsens the convergence instead of improving it 
in this case, so please try

local,liis=0

command somewhere at the beginning of the input.

Best wishes,

Tatiana

On Tue, 22 Sep 2015, Shaohong Louis Li wrote:

> Dear Molpro users,
>
> I encountered a convergence problem in the left-hand EOM-CCSD program
> (Molpro 2012.1). The system is Ni(CO)4. The output is as follows. I found a
> similar problem in the archive of this mailing list where Dr. Korona
> suggested using startt1=32 option in eom; I tried that but it did not work.
> The input is attached. Any help will be appreciated.
>
> Best,
> Shaohong
>
> ...
> Start left-hand EOM-CCSD program.    Author: T. Korona (2001)
>
> Record  5555.2 for saving EOM-CCSD vectors of length  26517513 reserved
>
>
> Record  4500.8 for saving EOM-CCSD vectors of length  26517513 reserved
>
>
> First produce intermediates, this can take some time ...
>
>
> Solve a set of linear equations for Lambda
> <Lambda|[e-T H eT,exc.] = -<0|[e-T H eT,exc.]
>
> Iteration   Residuum      time/iteration   time/total
>
>    1       0.86E+00            8.7            8.7
>    2       0.82E+00            9.3           17.9
>    3       0.12E+01            9.3           27.3 liis
>    4       0.19E+01            9.3           36.6 liis
>    5       0.29E+01            9.3           45.8 liis
> ...
>   50       0.17E+10            9.3          462.4 liis
>
> Convergence not achieved in  50 iterations, norm of residuum
> 0.1704047254E+10
>

Dr. Tatiana Korona http://tiger.chem.uw.edu.pl/staff/tania/index.html
Quantum Chemistry Laboratory, University of Warsaw, Pasteura 1, PL-02-093 Warsaw, POLAND