I am trying to perform geometry optimization of a Yttrium compound with orca 4.2.1 using the following:
!BP86 ZORA ZORA-def2-TZVP SARC/J TightOpt D3BJ RIJCOSX TIGHTSCF KDIIS GRID7 NOFINALGRID GRIDX9
However the calculation is getting terminated with the error in Cholesky inversion of numerical overlap. I did same calculations for similar Zirconium compound also which did not result in any error. What could be the origin of this error and why it is occurring for Y and not for Zr?
I suggest you follow https://www.researchgate.net/post/Cause_of_ORCA_crashing_at_the_Cholesky_decomposition_of_the_V-matrix_step
http://www.esqc.org/lectures/ESQC17-Neese-Lecture2-Approximations.pdf
Hello Somnath, for this error you can directly refer to the orca manual and search for "Symptoms of convergence issues".
This error basically arises when the chosen grid faces difficulties in representing your basis sets.
As your system contains a heavier atom which is heavier then Kr, you might be missing something related to the ECPs or the all electron approach used with ZORA basis sets.
As from H-Kr in the periodic table this basis sets used by you is valid but from Elements Rb-I you have to add the basis block to reproduce the old ORCA-behaviour keywords as in ORCA 4.0 and new versions there is no valid ZORA-def2-TZVP keyword. This issue mentioned by me is an extra one which you can take as a measure to prevent your calculations from future usage. (Refer. relativistic approximations and effective core potentials of orca google site)
Hope this will help...
Dhirendra Kumar Mishra I am using old-ZORA-TZVP as basis set for Y in the basis set block. The problem persists even if an enlarged grid is used. I have attached the input file.
- Badges
- Science topic