It is said that the D2/D3 dispersion scheme by Grimme, and the spin-component scaled version of MP2 (SCS-MP2) when they discuss the relevance of density functionals. Unlike the M06-2X that is fitted to give short- and medium-range correlation approximately correctly, these two methods (B3LYP-D3, SCS-MP2) would probably be better suited to study the stacked systems.
Are you asking from the technical point of view? If so, it's simple - you just specify B3LYP and then add the keyword EmpiricalDispersion=GD3 (http://www.gaussian.com/g_tech/g_ur/k_dft.htm). Make sure that you have one of the latest g09 revisions, since it was introduced into gaussian quite recently.
You can also use the standalone version of DFT-D3 (http://www.thch.uni-bonn.de/tc/index.php?section=downloads&subsection=getd3&lang=english), which will only calculate the D3 correction for a specific system.
Are you asking from the technical point of view? If so, it's simple - you just specify B3LYP and then add the keyword EmpiricalDispersion=GD3 (http://www.gaussian.com/g_tech/g_ur/k_dft.htm). Make sure that you have one of the latest g09 revisions, since it was introduced into gaussian quite recently.
You can also use the standalone version of DFT-D3 (http://www.thch.uni-bonn.de/tc/index.php?section=downloads&subsection=getd3&lang=english), which will only calculate the D3 correction for a specific system.
Thanks
Yes, I asked a technical point of view.Our g09 version is A-02.
I want to compare the results with M062X method.
As far as I know, Gaussian09 does not include SCS-MP2, but Gaussian may print direct and exchange parts of correlation energy in a MP2 calculation, after that you can calculate SCS-MP2 energies:
SCS-MP2= SCF + {corr = αE^(2)_corr,direct + βE^(2)_corr,exchange}
If you have A-02 revision you can use IOp(3/124=30) keyword.
I'm not sure about the g09 A-02 revision, but if I remember correctly it DOES NOT include D3 correction. You can easily check it by running test job with EmpiricalDispersion=GD3 and see if gaussian complains about the unknown keyword.
If this is the case, do a standard B3LYP in gaussian and then use the mentioned DFT-D3 program to get the correction.
Simply MP2 and use IOp(3/125=0333312000).
You can refer to http://sobereva.com/93. This webpage is written in Chinese, what a pity! The owner of this blog creates Multiwfn.
Dispersion corrections were introduced since G09 revision D-01. This information comes directly from Gaussian Inc..
- Badges
- Science topic