[Pw_forum] why total energy vs. lattice parameter is not a parabola?

[Stefano Baroni]

Dear Vu Ongphuong,

the explanation goes back to very early days of plane-wave total  
energy calculations. The reason is, in order to have a consistent  
accuracy at different values of the lattice parameter, a0, a same  
kinetic-energy cutoff has to be used for all of them. The consequence  
is that the number of plane waves will depend on the lattice  
parameter. Being the number of plane waves an integer number, it will  
display jumps in correspondence to some values of a0. If the cutoff  
is large enough, the jumps will be small and the energy-vs-volume  
curve will look like a parabola. If the cutoff is not large enough  
_for the given set of pseudopotentials that you are using_, then the  
curve will rather look like a sawtooth ... Two fixes are possible: 1)  
use a larger cutoff (or a softer pseudopotential), or 2) use some  
phenomenological equation of state (such as Birtch's or Murnaghan's)  
to fit the sawtooth ...

Stefano

On Oct 20, 2006, at 7:21 AM, vu ongphuong wrote:

> Dear all users,
>
> I calculate the lattice canstant of SrTiO3. When I plot the total  
> energy vs. lattice parameter, the shape is not a parabola even the   
> lattice parameter is taken in very near-minimum-region. The the  
> lattice canstant obtained is 3.851Angstrong (the  experimental  
> value 3.905)
>
> When I calculate with VASP package the result is 3.86 and the total  
> energy vs. lattice parameter is really a parabola.
>
> Anybody can explain me?
>
> Do you Yahoo!?
> Everyone is raving about the all-new Yahoo! Mail.

---
Stefano Baroni - SISSA  &  DEMOCRITOS National Simulation Center -  
Trieste
[+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype)

Please, if possible, don't  send me MS Word or PowerPoint attachments
Why? See:  http://www.gnu.org/philosophy/no-word-attachments.html



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20061020/e6c4743e/attachment.html>