Author Topic: IRC in turbomole  (Read 1895 times)

inozerox

  • Full Member
  • ***
  • Posts: 29
  • Karma: +0/-0
IRC in turbomole
« on: June 06, 2017, 03:11:28 am »
Hello everyone

It's me again!! I have another question

How Can I perfom an IRC calculation in turbomole?

Thanks in advance

uwe

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 425
  • Karma: +0/-0
Re: IRC in turbomole
« Reply #1 on: June 07, 2017, 11:20:11 am »
Hi,

the module which does IRC is called DRC. To get its options (especially the one to use IRC) please just call

DRC -h

on the command line. Basically all you have to do is to start DRC in the directory where you have successfully optimized your transition state and checked with a frequency calculation that you are at the TS.

Regards,

Uwe

inozerox

  • Full Member
  • ***
  • Posts: 29
  • Karma: +0/-0
Re: IRC in turbomole
« Reply #2 on: August 17, 2017, 06:36:22 pm »
Thanks uwe for aswering me

Cheers

inozerox

  • Full Member
  • ***
  • Posts: 29
  • Karma: +0/-0
Re: IRC in turbomole
« Reply #3 on: January 01, 2018, 08:24:25 am »
Hello everyone

I need help again  :(

I am trying to perform an IRC calculation using the command DRC -i -c (number of cycles) to study the acroleine system (TURBOMOLE tutorial).

However, I am trying to find some kind of output file. But, I can't find it.

Do I have to put some manual option in control fille to add the option to check the results?

Cheers

Arnim

  • Developers
  • Sr. Member
  • *
  • Posts: 224
  • Karma: +0/-0
Re: IRC in turbomole
« Reply #4 on: January 08, 2018, 02:18:42 pm »
Hello,

the energy and gradient calculations of the last cycle can be found in job.last. The other cycles are collected in the energy, gradient, and mdlog.x files.
You can use the AIMD tools log2x, log2egy, log2int, ... to extract data from the mdlog.x files.
Also, the evalgrad tool can be quite helpful. That will analyse the gradient file and print out the energy versus an internal coordinate. E.g. if your reaction coordinate is a bending angle connecting the atoms 1,2,3, you can run
evalgrad 1 2 3 > plot.me
and then
gnuplot -e "plot 'plot.me' u 3:2 w lp"
to look at the potential energy curve.

Cheers,

Arnim

inozerox

  • Full Member
  • ***
  • Posts: 29
  • Karma: +0/-0
Re: IRC in turbomole
« Reply #5 on: January 11, 2018, 07:58:28 pm »
Thank you Arnim for answering me

I am going to tell you what I obtained.

When I applied the command DRC -i -c (number of cycles) I obtained two new folders.

when I checked the file with Tmolex. I saw the gradient information and I could find the results. But the program made a graphical energy versus cycle step (like every gradient graphical)

I have some questions related with the things you answered me in your reply.

Can I use those tools when I used cartesians coordinates?

Does it existe any form to obtain a graphical Energy Versus IRC (Intrinsic Reaction Coordinate)?

Cheers

Arnim

  • Developers
  • Sr. Member
  • *
  • Posts: 224
  • Karma: +0/-0
Re: IRC in turbomole
« Reply #6 on: January 12, 2018, 10:27:08 am »

> Can I use those tools when I used cartesians coordinates?

I don't know what that should mean.

> Does it existe any form to obtain a graphical Energy Versus IRC (Intrinsic Reaction Coordinate)?

Yes. E.g. with the evalgrad and gnuplot commands I mentioned.
You could also use Excel/LibreOffice, if you are not familiar the gnuplot.

inozerox

  • Full Member
  • ***
  • Posts: 29
  • Karma: +0/-0
Re: IRC in turbomole
« Reply #7 on: January 12, 2018, 06:19:17 pm »
Thanks for answering again Arnim.

> Can I use those tools when I used cartesians coordinates?

In that question I meant I am not using internal coordinates (bond length, bond angles, dihedral angles) in my work. I am using x,y,z coordinates instead of internal ones.

About the second answer. Thank you for the piece of advice you gave me. But, I have a question related with that. is it necessary to select some couple of atoms to do that? (I guess you select some couple of atoms to see the change of some internal coordinates when the reaction is working)

what does it happend if I do not specify any atom? like Gaussian programm.

Cheers and thank you again




inozerox

  • Full Member
  • ***
  • Posts: 29
  • Karma: +0/-0
Re: IRC in turbomole
« Reply #8 on: January 12, 2018, 08:54:29 pm »
I meant I want to obtain a graphical like this



Cheers

Arnim

  • Developers
  • Sr. Member
  • *
  • Posts: 224
  • Karma: +0/-0
Re: IRC in turbomole
« Reply #9 on: January 15, 2018, 01:54:24 pm »
You can also plot the energy versus the time step. If you run log2egy, you will get these values. You have to reorder the data for displaced_plus and displaced_minus, however, to get a continuous plot.

But selecting and plotting internal coordinate should be straightforward and would be more informative in a discussion.