I have been working with a system that has a transition metal (TM) and imidazole ligand. When ORCA comes to solve the CP-SCF equation, it performs the solution until a certain interaction but after that fails in convergence. I have changed the converge accelarator from DIIS to KDIIS and even supplemented with Newton-Raphson methodology but unfortunetely the procedure is freazed by a whyle and after that is broken. Any suggestion to faced this problems with TM o maybe I am omiting something detail? Thank you in advance for you responses
Alexis Antoinette Ann Delgado I have, I used slowconv keyword
Hi Jhon Elmer Bustamante ,
CP-SCF equations often fails to converge for systems containing TM due to numerical precision "problems" in SCF or numerical noise from the grids. Sometimes I face similar problems and proceed by changing these settings:
1) tightening SCF convergence (TightSCF/VeryTightSCF);
2) increase the numerical precision of DFT integration grids (gridN/FinalGridN);
3) increase the accuracy of RI approaches (specially if RIJCOSX is used).
Please check:
https://sites.google.com/site/orcainputlibrary/numerical-precision
Hope it helps!
Felippe Mariano Colombari Appreciate your guides I will try them
- Badges
- Science topic
- Similar topics
- Biological Science
- Biochemistry
- Enzymology
- Enzymes