Author Topic: JOBBSSE and constrained optimization?  (Read 18856 times)

dbk80

  • Newbie
  • *
  • Posts: 7
  • Karma: +0/-0
JOBBSSE and constrained optimization?
« on: June 06, 2016, 10:37:42 AM »
Hi,

I'm trying to optimize a structure while taking into account the BSSE and therefore using the JOBBSSE script. I specified on top some geometrical constraints but it seems that they are not imposed during the simulation.
1- Is it possible to do constrained geometry optimization with counterpoise correction using JOBBSSE?
2- At the moment the constraints are in the coord file, no other reference to them in my *.sh file nor in my control file. Is this where my bug is?

Thanks,
D.

inozerox

  • Full Member
  • ***
  • Posts: 29
  • Karma: +0/-0
Re: JOBBSSE and constrained optimization?
« Reply #1 on: June 11, 2016, 11:21:15 AM »
Hello.

I have a question for you.

Are you using Cartesian coordinates? If you answer is yes, you have to be careful the Cartesian coordinates is actived instead of the internal ones. By default, the jobbsse script activate both internal coordinates and redundant coordinates. This option don't have problems for working with jobbsse script because this one only makes a serial of singlepoints calculations.

In order to do a geometry optimization you have two options:

-Make a jobbsse calculation and after you have to take the imputs, which was built in n°_ghosts folders (e.g 1_ghost). Here, you have to perform a geometry optimization by using jobex script (you must remember to check what kind of  coordinates are activated in control file).

-The another option is you have to build the input file manually. You can do this by editing the control file. When you want to assign the group of atoms like ghost atoms, you must add the line charge = 0.00000 below of the basis line.

I adjunct the next example:



For my own experience, I advice you don't do a geometry optimization by using ghost basis. I don't recommended that because these kind of calculation may have convergence problems (I lived that hehehehe).

I recommend you perform a normal jobbsse calculation and after you take the corrected energy value of complete system. you may optimize the monomers molecules without BSSE correction and you may compare the energies with the energy corrected.

I hope to help you with this post

Cheers

 
« Last Edit: June 13, 2016, 11:57:29 AM by inozerox »

sumitagrawal

  • Newbie
  • *
  • Posts: 1
  • Karma: +0/-0
Re: JOBBSSE and constrained optimization?
« Reply #2 on: February 01, 2020, 12:22:38 PM »
Hi everyone,

I am using Turbomole V7.3. Is it possible to do cluster calculation with BSSE corrected? In the manual, it is written that more than 3 fragments jobbbsse script will not work. But when I submitted the calculation for four fragments with jobbsse script it is running, and I also get the bsse corrected energy. Am I doing something wrong?

gunnar.schmitz

  • Developers
  • Jr. Member
  • *
  • Posts: 13
  • Karma: +0/-0
Re: JOBBSSE and constrained optimization?
« Reply #3 on: February 04, 2020, 10:04:33 AM »
Dear sumitagrawal,

if you are using Turbomole 7.3 you can specify as many fragments as you want and get CP corrected energies and gradients. The limitation to 3 fragments is an old limitation. In any case you can expect the results "by hand". Turbomole should save the outputs from the monomer and ghost calculations and you can check via the outputs if it did the subtraction of the energies correctly.

For later questions in the Forum consider to open a new discussion and not to append your question to an older one.

Good luck with your calculations