Author Topic: x2t convergence factor  (Read 6521 times)

klickklack

  • Newbie
  • *
  • Posts: 4
  • Karma: +0/-0
x2t convergence factor
« on: February 25, 2009, 10:27:04 am »
Hello,

I noticed that the convergence factor from Angstrom to Bohr in x2t is 0.52917706 at least in version 5.10.
The NIST value is 0.529177208 (http://physics.nist.gov/cgi-bin/cuu/Value?bohrrada0|search_for=Bohr+radius).
I know, this is a small difference, but for large systems it matters, especially when it comes to comparisons.

What are the internal convergence factors (Bohr->Angstrom; Angstrom->Bohr)?

klickklack


uwe

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 477
  • Karma: +0/-0
Re: x2t convergence factor
« Reply #1 on: February 25, 2009, 03:17:22 pm »
Hello,

the t2x script is probably as old as Turbomole is, so at least 20 years. And I guess that the Bohr radius in the t2x script has not been changed ever since. Feel free to change the script in your local version, we will change that in the official version.

I guess your systems must be very large if you can see a significant difference, and it is good to know that Turbomole also works for such huge molecules ;)

Internally, the Turbomole programs do everything in atomic units, and since the input is also in atomic units, no conversions are needed. For outputs in different units, a value of 0.5291772083 is used.

Uwe