On Mon, 2012-09-03 at 10:16 +0800, florence liu wrote: > I thought that it might be reasonable to use a PAW which is generated > for a functional based on RW86, is that right? I would try the setting > dft='NOX-PW-PW96-(anything else?)'
note that the definition of the vdW-DF2 functional will change in the next release: SLA-PW-RW86, i.e. th eSlater term is no longer contained into the rPW86 functional (the latter contains only the gradient corrections) P. -- Paolo Giannozzi, IOM-Democritos and University of Udine, Italy