[CP2K:5226] Re: Implementing new functionals

[Michael Banck]

Hi,

On Thu, May 01, 2014 at 01:10:42AM -0700, August Melcher wrote:
> So should I use a different solver and rerun? Just zooming in on the
> plot doesn't help.

First of all, anything beyond 5-7 Angstrom is probably not very 
interesting, and the latest plot you sent (b97s.jpg) is again unreadable
due to the artefacts beyond 15 Angstrom.

What are the units in the second plot (the binding one, image.png), and
how did you make it?

I ran a scan with the PSI4 (Beta5) package using the (probably similar?) 
wB97X-D functional and a aug-cc-pVQZ basis.  This is the potential curve
I get:

        R (Ang)         E (a.u.)
-----------------------------------------------------
        2.50            -257.90034199
        2.75            -257.90174824
        3.00            -257.90232610
        3.25            -257.90253327
        3.50            -257.90257489
        3.75            -257.90255286
        4.00            -257.90252022
        4.25            -257.90249566
        4.50            -257.90247981
        4.75            -257.90246941
        5.00            -257.90246231
        5.50            -257.90245369
        6.00            -257.90244979
        6.50            -257.90244770
        7.00            -257.90244637
        7.50            -257.90244526

This translates to a binding energy of 0.08 kcal/mol which (assuming
identical units) is slightly higher than what your second plot shows for
B97.


Michael