Author Topic: TD-DFT Excited state optimization failed  (Read 6908 times)

e.berardo

  • Jr. Member
  • **
  • Posts: 13
  • Karma: +0/-0
TD-DFT Excited state optimization failed
« on: May 01, 2012, 03:18:02 PM »
Hi all,
I am trying to optimize the excited state of a cluster, but the optimization fails in job.13, after the dscf and escf converge, with this error message:

    ----------------------------------------
     S+T+V CONTRIBUTIONS TO ENERGY GRADIENT
    ----------------------------------------

 1e-integral 1st. derivatives will be neglected if expon. factor < 0.100000E-10


 gradient run cancelled due to user signal in  (i|STV|j)`      :    2   1   0   0 (nshell= 259) !
 
 MODTRACE: no modules on stack

  *** this job must be restarted ! ***
 egrad ended abnormally
error in gradient step (1)
(END)


I am right now restarting the optimization with the new coordinates, but I would like to understand why this error occurred. Can anybody help me?
Thanks
Enrico
« Last Edit: May 01, 2012, 04:28:24 PM by e.berardo »

Arnim

  • Developers
  • Sr. Member
  • *
  • Posts: 253
  • Karma: +0/-0
Re: TD-DFT Excited state optimization failed
« Reply #1 on: May 02, 2012, 09:52:26 AM »
Hi Enrico,

that means a file called 'STOP' for a soft break was found in the working directory.

Cheers,

Arnim

e.berardo

  • Jr. Member
  • **
  • Posts: 13
  • Karma: +0/-0
Re: TD-DFT Excited state optimization failed
« Reply #2 on: May 02, 2012, 12:35:06 PM »
Thanks for the fast response Arnim,
I did put a stop file in the directory of the first calculation, but then I ran a restart on it and the same error came up (this time no STOP file was in the directory). I am now running a clean restart (with the opt coord and starting define again) and see what happens. I will keep you updated.

Regards,

Enrico

e.berardo

  • Jr. Member
  • **
  • Posts: 13
  • Karma: +0/-0
Re: TD-DFT Excited state optimization failed
« Reply #3 on: May 08, 2012, 03:27:25 PM »
Hi to everybody,
doing a restart [of jobex -ex] by copying all the files in the new folder didn't work and the calculation crashed again, but with a clean restart (just using the coord file and calling define again) the opt reached convergency with no problems.
Cheers,

Enrico