Fixed atoms in geometry optimization

[Dune André]

Dear CP2K  users,

I would like to run a geometry optimization while keeping all atoms except hydrogen atoms fixed.

I used the following sections in the input file with a list of around 750 atoms:

 &MOTION
   &CONSTRAINT
     &FIXED_ATOMS
       LIST 1 3 4 ...

     &END FIXED_ATOMS
   &END CONSTRAINT
 &END MOTION
             

But it seems that just a part of the atoms I gave in the list were frozen. Is there a maximal number of atoms I can give in the list ; or is there a specific way to give to cp2k a long list of atoms?

Besides, how can I check in the output file which atoms have been fixed or not?

Thanks for your help,

Dune

[fabia...@gmail.com]

Dear Dune,

There is no cap on the number of atoms that can be fixed. But there is a limited number of characters per line that cp2k reads in. So if you list is too long, the last atoms won't be fixed. I don't remember the limit by heart, but in general I don't put more than 200 characters per line, even if it only improves readability.

Two hints to help you:

 - use .. for consecutive numbers, e.g.

LIST 1..10

 - LIST can be repeated, e.g.

LIST 2..10 12..20

LIST 22..30 32..40

Cheers,

Fabian

[Dune André]

Thanks Fabian for your answer.

Another question related to this topic is the following: How does the information of the fixed atoms appears in the output file?

Reading the list of atoms in the .out file, I did not find the info of whether atoms were fixed or not.

Regards,

Dune

[Martin Konôpka]

Dear Dune,

There is no such explicit info in the output file at least with  PRINT_LEVEL  LOW. As for the higher print levels, I do not know. My practice is just to carefully prepare the input files and then anyway (at least in first run of a series) to check some picked-up constrained atoms (or coordinates) if they really keep fixed.

Best

Martin.

st 19. 7. 2023 o 15:30 Dune André <dune....@gmail.com> napísal(a):

--
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/4ba1beac-c61f-4ff7-9e42-680e8d6b66d0n%40googlegroups.com.

[Dune André]

Thanks Martin for your answer, I'll check in this way then.

Best,

Dune