Author Topic: heavy atoms without ecp  (Read 3140 times)

fru14909

  • Newbie
  • *
  • Posts: 2
  • Karma: +0/-0
heavy atoms without ecp
« on: September 22, 2014, 09:52:43 am »
Hello,

how can I avoid the use of ecps for heavy atoms such as Pb?
Whenever I optimize the geometry of my molecule (consisting of Pb and Se) with jobex, the molecule is completely distorted. I removed the ecp, but then eht did not work anymore.
What can I do?

Thanks in advance

Arnim

  • Developers
  • Sr. Member
  • *
  • Posts: 224
  • Karma: +0/-0
Re: heavy atoms without ecp
« Reply #1 on: September 28, 2014, 01:14:40 pm »
Hi,

it is in general not a good idea to remove ECP and then just use the remaing basis functions. That will get something completely unbalanced. You would have to use an all electron basis set, if you want abstain from using an ECP.

However, ECPs usually work very well. I would think that there might be other obstacles in your system.

Cheers,

Arnim

fru14909

  • Newbie
  • *
  • Posts: 2
  • Karma: +0/-0
Re: heavy atoms without ecp
« Reply #2 on: October 01, 2014, 01:08:01 pm »
Hi,

for more cage-like systems the geometry optimization worked well with ecps. This molecule is built like ammonia, but geometry optimization gives a (unrealistic) planar transition form.
I already tried an all electron basis set, but then the eht guess does not work, although I added $ehtdef and $ehtao paragraphs in the basis file.

Cheers

Arnim

  • Developers
  • Sr. Member
  • *
  • Posts: 224
  • Karma: +0/-0
Re: heavy atoms without ecp
« Reply #3 on: October 07, 2014, 04:31:31 pm »
Hi,

you can try to find an all electron basis set here:
https://bse.pnl.gov/bse/portal
Download it in Turbomole format and copy it in basis.
Remove the $ecp from control and ecp in $atoms. (In older version remove also $rundimensions)
Then run hcore and start dscf. (For ridft you would have to create an appropriate auxbasis first)

That should do the trick,

Arnim