CCL:G: TDDFT in gaussian 09

 Sent to CCL by: Lars Goerigk [lars.goerigk|*|uni-muenster.de]
 Hello Mike,
 running TD calculations with double-hybrids works somewhat differently than
 e.g. with GGAs or hybrids. The reason for that is that the DFT part of the
 functional is treated like standard linear response TDDFT, but that for the
 perturbative part a CIS(D) type correction has to be carried out. For an
 explanation of how that works, you can have a look at the following
 references:
 Grimme and Neese, J. Chem. Phys. 127, 154116 (2007).
 Goerigk and Grimme, J. Phys. Chem. A 113, 767 (2009).
 Goerigk, Moellmann and Grimme, Phys. Chem. Chem. Phys. 11, 4611 (2009).
 So for running TD-B2PLYP as described in these papers is not implemented into
 the standard quantum chemistry packages. However, ORCA, which is available for
 free, can do TDA-B2PLYP calculations (Tamm-Dancoff- Approximation). Using TDA
 just for obtaining excitation energies is OK, though, and usually it does not
 differ much from the TD result.
 If you just run a TDDFT calculation after having specified "b2plyp" in
 Gaussian, this actually gives you "TD-B2LYP" excitation energies, i.e.
 just
 for the DFT part of the functional. As the amount of Fock exchange is usually
 responsible for the specific value of an excitation energy, TD-B2LYP is very
 similar to TD-BHLYP (53% vs. 50% of Fock exchange).
 I hope that I could help clarifiying this issue.
 Cheers,
 Lars
 On Thursday 08 April 2010 10:08:08 pm Mike F Falcetta mffalcetta-x-gcc.edu
 wrote:
 > Sent to CCL by: "Mike F Falcetta" [mffalcetta * gcc.edu]
 > I am running some trial TDDFT calculations with gaussian 09 using the
 >  b2plyp and wB97xD functionals.  My question is which parts of these
 >  functionals actually affect the excitation energies.  My guess is that the
 >  long-range correction for the exchange does affect them but the dispersion
 >  correction in the wb97xD and and perturbative correlation corrections in
 >  b2bplyp do not.
 >
 > IF anyone can either confirm or correct my assumption and/or point me to a
 >  further discussion on the matter I woudl greatly appreciate it.
 >
 > Thanks in advance
 >
 > Mike>