Talk:Hybrid functional

Semiempirical hybrids?
Are there semiempirical hybrid functionals? Having just written the article and looked in the Gaussian manual list, I'm not so sure. Maybe it doesn't make sense since you set the parameters in an empirical sense. WilliamDParker 16:03, 28 March 2007 (UTC)


 * Having recently reviewed some of Perdew's articles, it turns out that LYP correlation is empirical in that it is fit with four parameters, and any mixing of Hartree-Fock exchange with more than one paramters is semiempirical because it is fit to some set of experimental outcomes WilliamDParker 20:17, 7 May 2007 (UTC)


 * Continuing on this, remember Adamo and Barone implemented the Purdue NON-empirical formulation as the 'B1LYP' functional in Chem Phys Lett. 274 (1997). . This is available in Both Gaussian and Neese's ORCA, as well as many other packages.RabidDutchman (talk) 11:30, 18 August 2008 (UTC)

B3LYP is a semi-empirical functional, because the three parameters were obtained by fitting the results on atoms and small molecules to those given by a more accurate method, which I believe is Pople's G1 or G2 method.

Chibibrain (talk) 02:53, 8 May 2009 (UTC)

French version
Hi ! I would like to signale that I have written a version of hybrid functional in french that is more complete and detailed than this english version. Maybe kind translators can use it to improve the page. Faithfully yours, Grimlock (talk) 14:07, 11 August 2010 (UTC)