Talk:Electronic correlation

Latest comment: 9 years ago by Bduke in topic Static and dynamic correlation


Term "correlation energy" not invented by Lowdin

edit

In the article it is said that "The difference is called the correlation energy, a term coined by Löwdin.[1]".

Even though Lowdin is very often quoted as the inventor of the term "correlation energy", I don't think this is historically correct. At most, he made the term very popular in the chemistry community, but "correlation energy" as an expression was first used by Wigner and Seitz in 1934 in the context of studies of the solid state (Phys Rev 509, 1934 ; Phys Rev 1002, 1934). I also seem to remember that Lowdin in his famous review never claims he invented the term himself. I'll check the matter in more detail if I have more time. —Preceding unsigned comment added by 128.40.5.101 (talk) 13:11, 21 July 2009 (UTC)Reply

Coupled cluster

edit

Coupled cluster should probably be added to the methods list, according to that article it is used quite often. Regretably I don't know enough about it to feel comfortable adding it a.t.m. Mverleg (talk) 04:13, 3 August 2012 (UTC)Reply

Qualitatively emerge.

edit

The sentence "In them, interactions plays such an important role that qualitatively new phenomena emerge" doesn't make sense, not really. Should it read something along the lines of "In them, interactions become so significant that new phenomena are observed."? Tomásdearg92 (talk) 00:46, 23 April 2013 (UTC)Reply

Moved from article added by User:147.210.60.13

edit

NOTE (added by Dr S.F. Matar Bordeaux Dec'13) : You need to take into account the DFT (density functional theory) treatment of exchange correlation at the level of the electron ε_{xc} such as in its genuine exchange-correlation functional LDA based on the homogeneous electron gas. LDA is the local density approximation. In a non interacting electron system for kinetic and Hartree terms, information on the interacting systems is embedded in the exchange correlation energy: Exc[ρ(r)]=∫ρ ε_xc(r)d3r (LDA) + gradient terms(*GGA). Adding gradient terms (*) leads to generalized gradient approximation GGA. This involves considering 'quasi-particles' in that sense of an electron surrounded by its Fermi (exchange) and Coulomb (correlation) hole. In spite of large successes of DFT-LDA and DFT-GGA with several modern computational methods built around DFT and its approximations, several drawbacks could be traced out and new developments are active in the scientific community of quantum mechanics.

Static and dynamic correlation

edit

" The multi-configurational self-consistent field (MCSCF) method takes account of this static correlation but not dynamical correlation."

That is an inaccurate statement. Both dynamic and static correlation effects are taken into account for the electrons and orbitals used in the MCSCF calculation. Those that are excluded from the MCSCF calculation will not be included in the correlation effects. — Preceding unsigned comment added by 68.188.178.22 (talk) 16:15, 23 January 2015 (UTC)Reply

The problem is that the terms static and dynamic correlation are not well defined. I like a definition that says that static correlation is the minimum required for the bond to dissociate correctly. Thus for the hydrogen molecule MCSCF at the CASSCF(2,2) level would be static correlation, not dynamic correlation. Of course using a larger basis set, CASSCF(2,n) with 2 greater than 2 would be including dynamic correlation. In practice most CASSCF calculations are essentially at the static correlation level. Methods such as coupled-cluster are used to get dynamic correlation. For example CASSCF(6,6) for the pi electrons in benzene is usually considered as getting the static correlation only. Getting good sources for whatever we say will not be easy as different people have different ideas about what constitutes dynamic and static correlation. --Bduke (Discussion) 21:01, 23 January 2015 (UTC)Reply