[Wien] LDA+U

[Stefaan Cottenier]

> But I have another question, that is for which system should I use
> LDA+U and for which use GGA+U?

They are usually equivalent, just the value of U is a bit (0.2 Ry)
smaller for GG+U.

> Prof. Schwarz ever told me that if there is no
> Oxygen in system we need not use LDA+U. My studying system is 
> monoatomic iron wires, so I need not use LDA+U?

LDA or GGA is the method of choice for 'common' situations. Only if more
than moderate correlations are present (typically in f-electron
materials and transition metal oxides, where respectively the f- and the
d-electrons are rather localized and therefore correlated), LDA+U should
be considered.

Stefaan


Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm