[CPMD-list] about pade functional

[Juerg Hutter]

Hi
>
> I'd like to get lattice constant of Pt bulk
>
> Among PPs provided in CPMD, I could find Pt PP only at SG PPs with pade.
>
> There 2 kinds of PP - Pt-q10 & Pt-q18.
>
> My questions are:
>
> First, on pade functional. How to describe it in FUNCTIONAL keyword? Is
> it PADE or LDA?

FUNCTIONAL LDA

defaults to the PADE representation of the local density approximation.

>
> Second, which of two are more reliable? Is there anyone who tried the
> same thing - getting lattice constant with this PPs?

Usually, the more electrons the more reliable the PP. However, you
pay for it with a higher cutoff and more electronic states.
I would assume that a 10 electron PP should be good enough for the
lattice constant.

>
> Third, is there any better PPs of Pt with PBE functional or else?
>
Have a look at www.abinit.org and www.pwscf.org

There you can download other Pt pseudos. However, they are not in the
correct format for CPMD. Nvertheless you get information on how
these PP were constructed and it should be easy to generate your
own PP with one of the common codes.

> Fourth, is it ok to mix up functionals, i.e. PBE with PADE(LDA), or PPs,
> i.e. SG with TM with same functional?
>

It is common practice to only use PP that have been generated using
the same functionals as will be used in the actual calculation.

regards

Juerg Hutter

>
>
> Thanks in advance
>
>
>
> HONG W.K.
>
>
>
>