different basis set for the same element

[Axel Kohlmeyer]

On Sep 26, 2:42 pm, Nguyen Ngoc Ha <h.... at hnue.edu.vn> wrote:
>  Dear Axel,
> Thanks for the tips, maybe the code requires a Z number with the
> corresponding name.

no.

> If I only use a modiffied name in both of BASIS_SET and POTENTIAL database,
> I'm not sure that this is a right way.

you don't need to change any of those files either.

> I think that problems are in files: periodic_table.F,
> semi_empirical_parameters.F,...

no, this is a case of PEBCAC. you didn't get what i was suggesting.

> It's simple to add 'new elements' started from 107 with the same
> propeties,... of the original elements into the files.

this is a _very_ ugly hack. have a closer look at the syntax
in the &force_eval/subsys section instead.

if you provide coordinates with a &COORD section or though some
other file, you have to use a "label" to link the atom at
a given position to a corresponding &kind section. by default and
because of convenience, people usually take the element name, but
this is not at all required.

the input syntax of cp2k is meant to accommodate not only
quantum chemistry but also classical models, where it is quite
possible to have multiple atom "types" (or labels) for the same
element, there has to be a mechanism to have multiple "kinds" for
the same element. in fact, the label can be _anything_. so please
consider the following input segment for two water molecules with
two different basis sets:

  &SUBSYS
    &CELL
      ABC 10.0 10.0 10.0
    &END CELL
    &COORD
     OA        -1.5574991518        5.5615110942        1.6017385036
     HA        -1.1071155404        4.9123357898        1.0055969642
     HA        -2.4841650102        5.5320929304        1.2848208725
     OB        -0.0936672407        3.6972834592       -0.0942506957
     HB         0.3514665955        3.1856287764        0.6144874465
     HB         0.6019083928        4.3150027824       -0.4056961815
    &END COORD
    &KIND OA
      BASIS_SET TZV2P-GTH
      POTENTIAL GTH-PADE-q6
      ELEMENT O
    &END KIND
    &KIND HA
      BASIS_SET TZV2P-GTH
      POTENTIAL GTH-PADE-q1
      ELEMENT H
    &END KIND
    &KIND OB
      BASIS_SET SZV-GTH
      POTENTIAL GTH-PADE-q6
      ELEMENT O
    &END KIND
    &KIND HB
      BASIS_SET SZV-GTH
      POTENTIAL GTH-PADE-q1
      ELEMENT H
    &END KIND
  &END SUBSYS

that approach should allow you to do what you want in a clean and
consistent way and without any ugly modifications to either the
sources or the potential or basis set files.

cheers,
    axel.

> Best regards,
>
> --
> Nguyen, Ngoc Ha
> Hanoi National University of Education
> Faculty of Chemistry
> Department of Physical Chemistry
> Tel: Office: 04/38330842
> Home: 04/37891674
> Mobile: 0989133436
> E-mai... at hnue.edu.vn <E-mail... at hnue.edu.vn>