[Wien] RMT

[Stefaan Cottenier]

> 1. The RMT should be usually reduced for some percent in surface
> relaxation(MSR1a) as we know, however, for a correct comparison of some
> physical properties(surface energy, work function...) between a relaxed
> and unrelaxed surface(to see the relaxation effects), if a new scf
> should be again done for the relaxed one with original RMT?

It is least trouble/most efficient if you determine first the smallest 
RMT you'll need, and then use those consistently for all your calculations.

> 2. I was wonering if the physical properties (surface energy ...) of two
> different surfaces of a solid, let say Ni3B(110) and Ni3B(101), are
> comparable if different RMT have been used in their calculation?

Probably yes, provided both where converged w.r.t. basis set size.

Stefaan