Methods to deal with Electron correlations

[leoant]

Dear all,

I am here again with problems

I am now trying to get some feeling of electron correlations in solid, following some literature, but it seems like that ALL confusing things lie here---there are s o many methods, and usually different people insist on their own one; Moreover, reviews are usually concentrated on comparison of results.

I think this is a rather good playground for me (of course, everyone) to learn something--since all these methods are important. this is the reason why I have some questions about different methods:

1. Time-dependent perturbation theory(TDPT), which is derived from usual perturbation , is extensively used in, say, P. Noziere's papers and Pines' book(elementary excitation), but what's the relation between this and the so-called "Many body perturbation theory"?

2. Another very important method is of course green's function, or field theory, So how this is related to the TDPT, since in both method diagrammatic methods to trace interaction is used?

3. It seems like that there're so many versions of the Field theory, such as that by von Hove, and that formally derived from Feynman's QED theory, so what's the relation between them?

I think it's better if you can refer me to some literature(better if focused on different methods), since these question can not be easily answered with several words. Anyway, I appreciate you for any comments. Thanks in advance.

Hongbin

 

[quetzalcoatl9]

DFT makes use of an exchange-correlation functional to include these effects in an approximate way

there are also the correlation-consistent basis sets for HF parameterized to reproduce the correlation energy (aug-cc-pVDZ, etc.)

 

[Fizol]

DFT makes use of an exchange-correlation functional to include these effects in an approximate way

But this method works AFAIR only for ground state of many-body system.

 

[quetzalcoatl9]

But this method works AFAIR only for ground state of many-body system.

yes that's true

if you want excited states it is hard to beat TDDFT

 

[marlon]

But this method works AFAIR only for ground state of many-body system.

Yes but if you know the crystal structure at some temperature, your problem is solved. It all depends of what exactly you want to study. LDA/GGA works just fine for solids (especially metals) except for the bandgap calculations. They latter, require GW approxiamations to get the correct values.


marlon