If you have problems during the execution of MRCC, please attach the output with an adequate description of your case as well as the followings:
  • the way mrcc was invoked
  • the way build.mrcc was invoked
  • the output of build.mrcc
  • compiler version (for example: ifort -V, gfortran -v)
  • blas/lapack versions
  • as well as gcc and glibc versions

This information really helps us during troubleshooting :)

Possible bug in properties module ?

  • SBehnle
  • Topic Author
  • Offline
  • New Member
  • New Member
More
6 years 4 months ago #467 by SBehnle
Possible bug in properties module ? was created by SBehnle
Dear all,
i tried to compute an analytical geometry optimization at the CCSD(T)/cc-pVDZ level for the tetrahydroborate anion (see the attached input and output). Unfortunately, the calculation didn't go as expected. The problems started when prop was called the first time. The SCF moments are all fine but the CC dipole moment is unequal to zero which is unexpected for a perfectly tetrahedral input geometry. The computed nuclear gradients are also weird, specifically, the gradient at the boron atom is unequal to zero. The molecule subsequently gets distorted (which shouldn't happen) and the optimization doesn't converge within 100 cycles.
The problem occurs irrespecive whether symmetry is used or not. Setting OMP_NUM_THREADS to 1 doesn't solve the problem.
On the other hand, at the B2PLYP/cc-pVDZ level, the tetrahedral structure is a minimum (not the input structure but the tetrahedron).
All occupied orbital energies are below 0 so IMHO there is no obvious reason why this shouldn't work.
Used MRCC version: precompiled binaries from 2017/09/25
glibc: 2.11.3 on Debian squeeze
What am i doing wrong?
Best regards,
Stefan

Please Log in or Create an account to join the conversation.

  • kallay
  • Offline
  • Administrator
  • Administrator
  • Mihaly Kallay
More
6 years 4 months ago #468 by kallay
Replied by kallay on topic Possible bug in properties module ?
Dear Stefan,
Unfortunately, we do not have CCSD(T) gradients (please see Sect. 6.2 of the manual for the list of available methods). Of course, the program should stop in this case with an error message, we will fix this.

Best regards,
Mihaly Kallay
The following user(s) said Thank You: SBehnle

Please Log in or Create an account to join the conversation.

  • SBehnle
  • Topic Author
  • Offline
  • New Member
  • New Member
More
6 years 4 months ago #469 by SBehnle
Replied by SBehnle on topic Possible bug in properties module ?
Dear Prof. Kallay,
thank you very much for the quick reply. I erroneously thought that this includes perturbatively corrected methods (as they are not explicitly excluded; my bad). In this case, i have to stick to numerical gradients.
Anyway: Great program!
Best regards,
Stefan

Please Log in or Create an account to join the conversation.

Time to create page: 0.040 seconds
Powered by Kunena Forum