Since LDA does not require the derivatives of the electron density,
maybe this leads to more stable and smoother SCF/optimization.

On Mon, 25 Jan 2016, Luis Ogando wrote:

Dear Wien2k community,

   I am working with III-V semiconductors (bulks and some 
combinations/heterostructures of their different compounds).
   I have noticed that sometimes LDA is better than PBE for convergence in the 
structural optimizations (in other words, convergence is easier with LDA).
   I would like to know if there is a kind of "general tendency" for III-V 
semiconductors or it will depend on the atomic species (or anything else) ?
   All the best,

Wien mailing list

Reply via email to