[CP2K-user] Multipoles and core-shell model

[Pierre-André Cazade]

Dear CP2K users,

I am checking the DIPOLE and QUADRUPOLE subsections in 
FORCE_EVAL%SUBSYS%MULTIPOLES. I am puzzled by the absence of specification 
of a local framework. To my experience, the latter is necessary as 
multipoles are not defined in a univoque way when the previous terms are 
not null and are therefore framework dependant. How does CP2K account for 
this? Does CP2K define automatically the framework based on connectivity?

Is the multipolar interactions subjected to the exclusion list (no 
nonbonded interactions between 1-2 and 1-3 atoms)?

Finally, when using a core-shell model to model polarisation, is there a 
way to exclude bonded atoms (1-2 and 1-3) from the interactions aside from 
the SHELL_CUTOFF in FORCE_EVAL%MM%FORCEFIELD%SHELL?

Regards,
Pierre
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20200611/0eb9e175/attachment.htm>