Author Topic: PEECM calculation with ECP shells in Turbomole 7.5  (Read 102 times)

dsahsah

  • Newbie
  • *
  • Posts: 1
  • Karma: +0/-0
PEECM calculation with ECP shells in Turbomole 7.5
« on: December 15, 2021, 12:19:56 AM »
Greeting everyone,
I am trying to run a PEECM calculation in Turbomole 7.5 and my calculations are failing at the early stages of setting up the PEECM calculation. Particularly, the error occurs immediately after calculating "number of PCs in first FMM Zone" (please refer to attachment: error_ridft.txt).  The error messages reads :"Cell mismatch in nfpcstf (3)" (see attachment:mismatch_PEECM.txt).
 The system at hand is a metal QM metal cluster embedded in a field of point charges defined as one hundredth of standard TIP3P partial charge, since I have 100 water conformations in one embedded file. Point charges at a certain proximity to the surface are assigned an ECP in order to construct an isolating shell to overcome potential electron spill-out.  (ECP shell is constructed out of water molecules with Ne and He ECPs assigned for O and H, respectively).

Another user reported a similar problem and they mentioned that they have found success in defining their box size using a maximum of eight decimal places in the header of embedded (link to post: https://forum.turbomole.org/index.php?topic=1032.0). However, I already use eight decimal places to define my box size.

I also would like to ask about the appropriate way to assign charges to the molecules that make up the ECP shell. Should those molecules have charges assigned in embedded and control or in control only or in embedded file only ?

I would really appreciate any recommendation or suggestions
Thanks in advance!