Error basis set for actinides in Turbmole 8

[arturosauza]

Hello Community,

I am trying to run geometry optimizations on some actinide-containing molecules but I am struggling when using define to select the basis functions of the atoms in the molecule. After loading the geometrical data, I get the following warning in the terminal:

Code: [Select]

SUPPLYING BASIS SETS TO  131 ATOMS
 #                                                                       
 #                      BASIS SET LIBRARY FOR Uran                       
 #                      last update : 08/04/2022                         
 #                                                                       

  THERE ARE NO DATA SETS CATALOGUED IN FILE
 /Shared/chemistry_software/TURBOMOLE-80/basen/u
  CORRESPONDING TO NICKNAME u def2-SV(P)                                                           


  USE ONE OF THE FOLLOWING OPTIONS :
  nick    - REPEAT NICKNAME INPUT
  file    - SWITCH TO ANOTHER FILE
  lib     - SWITCH TO THE STANDARD LIBRARY
  newlib  - SPECIFY NEW STANDARD LIBRARY
  list    - PRINT A SURVEY OF CATALOGUED DATA
  filnam  - PRINT ACTUAL FILENAME
  libnam  - PRINT NAME OF ACTUAL STANDARD LIBRARY
  DEFAULT - NO FURTHER ACTION

I am new in Turbomole, but previously I was using Turbomole 7.8 and I never received such warning when trying to assign default basis functions to the actinide centers. What can I do to solve this issue?

I was able to assign the def-TZVP basis to the U atom, but later when tryng to generate the initial guess for my calculation, the EHT routine failes and I cannot continue.
This is what I get in the terminal after defining the basis functions for my system:

Code: [Select]

there are 1 real representations :   a   

 OCCUPATION NUMBER & MOLECULAR ORBITAL DEFINITION MENU

 CHOOSE COMMAND
 infsao     : OUTPUT SAO INFORMATION
 atb        : Switch for writing MOs in ASCII or binary format
 eht        : PROVIDE MOS && OCCUPATION NUMBERS FROM EXTENDED HUECKEL GUESS
 use <file> : SUPPLY MO INFORMATION USING DATA FROM <file>
 man        : MANUAL SPECIFICATION OF OCCUPATION NUMBERS
 hcore      : HAMILTON CORE GUESS FOR MOS
 atoms      : SUPERPOSITION OF ATOMIC DENSITIES
 flip       : FLIP SPIN OF A SELECTED ATOM
 &          : MOVE BACK TO THE ATOMIC ATTRIBUTES MENU
 THE COMMANDS  use  OR  eht  OR  *  OR q(uit) TERMINATE THIS MENU !!!
 FOR EXPLANATIONS APPEND A QUESTION MARK (?) TO ANY COMMAND

eht

 PROVIDING EHT AOS FOR THE FOLLOWING SET OF ATOMS :
     1 u       

 NO MORE DATA AVAILABLE FOR EHT !!
 FOR A SECOND CHANCE HIT  >return<  OR
 ENTER  q  TO QUIT THE EHT SECTION

q

 SOME PROBLEMS AROSE IN PROCESSING ATOMIC ORBITALS OF u
 WHICH ARE NEEDED FOR EHT START ORBITAL GENERATION !
 CHECK IF DATA BLOCK $ehtdef IS IN PLACE IN LIBRARY FILE
/Shared/chemistry_software/TURBOMOLE-80/basen/u ! MOREOVER CHECK IF DATA HAVE CORRECT SYNTAX !


  basget ABEND : basis set with nickname
  is not contained within file /Shared/chemistry_software/TURBOMOLE-80/basen/u


  basget ABEND : basis set with nickname
  is not contained within file /Shared/chemistry_software/TURBOMOLE-80/basen/u


========================
 internal module stack:
------------------------
    define
    mobspy
    ehtinp
========================

  fatal fault in basget called by ehtinp
 define ended abnormally
 define ended abnormally
program stopped.

Thanks in advance for your time and help.

[yannickf]

Hello,

Turbomole 7.8 used the def family as default basis sets. These are the old ones of Ahlrichs and co-workers for H-Rn. For 5f, they correspond to the Dolg basis sets published with their ECPs. The def2 basis sets generally perform better but they are not available for 5f. You can check that in the basen directory of your installation. You could add an alias there, i.e. insert def2-XVP with X=S,TZ,AZ below the line with def-XVP.

The eht failure looks like the ECP is not assigned. Could you please confirm that the def-ecp or a given mwb/mdf ECP was assigned for U. Right after changing the basis set to def-TZVP, you can use "ecpl" in define to list the assigned ECPs. If not assigned yet, you can assign it with the "ecpb" option similar to the basis set.

Anyway, this is a bug (automatic assignment fails) and it can be fixed. The above method is just a workaround.

Best,

Yannick

PS: if you want to use X2C: https://forum.turbomole.org/index.php/topic,1356.0.html you don't have EHT vectors so you need another guess.

[arturosauza]

Hello Yanninck,
I appreciate all your help. I confirm that the ECP was not assigned for the U atom. I followed your suggestions, and now I was able to set the calculation properly. Thank you so much for your time and help.


Arturo.

[yannickf]

Hello Arturo,

Thanks, and I am glad that we could solve it. This issue was also fixed in the code of define. So, you will get the def basis sets for 5f elements by default. This will be made available in the next bugfix release (I don't know about the schedule though).

Best,

Yannick

[arturosauza]

I'll be looking for the next software release.

Thank you.