single point DFT, SCF does not converge

[Egor Bulavko]

Hello!

I am experiencing a very strange thing while trying to perform single point DFT calculations. According to information from the out file, my convergence criteria is 10e-05 for energy and 10e-04 for density. Nevertheless, after dozens of tries I got the energy and density values that do not change at all while iterating (for instance, energy changes in the eleventh decimal place), but calculations do not finish, and finally, after reaching the maximum number of iterations it says that convergence did not achieved. Any ideas about it?

I attach my inp and out files

Thanks in advance,

Egor

[Edoardo Aprà]

The input posted uses an all-electron ZORA style basis set but it is missing the the ZORA relativistic input section https://nwchemgit.github.io/Relativistic-All-electron-Approximations.html#zeroth-order-regular-approximation-zora

Here are the input modifications that should make your calculations converge.

basis spherical
* library 6-31G* except U
U library SARC-ZORA
end

dft
 convergence fast
 maxiter 99
 xc b3lyp
 DISP vdw 4
 direct
 tolerances acccoul 13
 mulliken
end
relativistic
  zora on
  zora:cutoff 1d-30
  modelpotential  1
end
set quickguess t
task dft

[Egor Bulavko]

Dear Edoardo,

Thank you very much, it works.

Nevertheless, the single point calculation is not my final goal. Also I would like to perform QMMM optimization for a protein with uranium. I tried to initialize these calculations, but there is an error:

fock_txs: no room for scr            10040560
In documentation it is mentioned that ZORA can be accessed only via DFT and SO-DFT modules. Does it mean that it will not work for QMMM optimizaion? 

I attach all files.

Sincerily,

Egor

[Edoardo Aprà]

Your job ran out of memory.

To fix, add the following memory line at the beginning of your input file

start 3odc

memory stack 700 mb heap 100 mb global 300 mb

charge -1

[Egor Bulavko]

Thank yot again! Now everything is okey!