warning: cosmo sigma out of range

[Amin]

Hello,

I am running a TURBOMOLE calculation with dCOSMO-RS for water and I am seeing repeated warnings like this in the output:
warning: cosmo sigma out of range      0.9092574(max.      0.1280000)
warning: cosmo sigma out of range      1.8546818(max.      0.1280000)
warning: cosmo sigma out of range      0.4208319(max.      0.1280000)
warning: cosmo sigma out of range      0.3297463(max.      0.1280000)
warning: cosmo sigma out of range      1.2903717(max.      0.1280000)
ediel:   -553.2556768 ediel RS:   -544.0457994

The SCF appears to continue, for example:

ITERATION  ENERGY          1e-ENERGY        2e-ENERGY     NORM[dD(SAO)]  TOL
  3  -334.45580029077    -18713.953846     11250.919684    0.000D+00 0.670D-12
                           Exc = -252.445625550     Coul =  11596.9336979
                           exK = -93.5683885342
                             N = 360.00014289
                         Ediel = -544.04579941

For the setup, I only added the following lines to the control file:

$cosmo
$cosmo_out file=out.cosmo
$dcosmo_rs file=h2o_25.pot

My question is: is this the correct way to set up a dCOSMO-RS calculation for water, or is there something else I was supposed to do in addition to adding these keywords?
I am wondering whether these warnings indicate that I have not set up the solvent model properly, or whether this is a known/acceptable warning in some cases.

The documentation was a little unclear to me, so I would really appreciate any guidance on whether this setup is complete and whether these warnings are a sign of a real problem.

Thanks in advance.

[uwe]

Hello,

the input is correct, the warnings are printed because you are (most likely) using direct COSMO-RS approach on a charged and/or very polar molecule. Is that correct?

direct COSMO-RS adds a potential that stems from the solvent (water at 25 deg Celsius according to your input) and multiplies this potential with the surface charges to get an energy contribution. The dCOSMO-RS potential was generated by the COSMOtherm program and uses COSMO-RS statistical thermodynamics theory which works very well for a certain range of surface charges. If the charges on the cavity of your input molecule are (far) larger than the range COSMO-RS has been fitted to, you are using the potential at values which have been extrapolated rather than fitted. So the expected accuracy will be lower than usual - and that's why there is a warning. 

[Amin]

Thank you! This address my question.