[Pw_forum] Slab optimization with -24kbar stress!!!

Serge Nakhmanson smn at ims.uconn.edu
Fri Dec 5 14:54:27 CET 2014


On 12/5/14 5:11 AM, David Foster wrote:
> 1- Cyrille, I fixed both in-plane and z-direction lattice constants. I need to fix the 15A vacuum in the z to prevent periodic interactions for studying adsorption.

Dear David,

Just my 2 cents to add to this.

(a) Residual surface stress is a perfectly "legal" quantity. Some people hunt 
for it (as well as for the surface elastic constants)
specifically with DFT. You can check it out here:

http://dx.doi.org/10.1103/PhysRevB.71.094104

or here

http://dx.doi.org/10.1016/j.cma.2010.09.007

It is a lot of fun! :)


(b) You have to be *extremely careful* when attaching something (a molecule?) to 
your surface, as this will open a
new can of worms. You will need to check if your combined system is likely to 
have a dipole moment pointing perpendicularly to
the slab -- and it might. Then you may have to keep an xy mirror plane running 
through the middle of your slab [adding many more
atoms to the calculation] or apply a dipole correction [makes calculation 
convergence much slower]
to get rid of it. If you don't, you run a chance of having a spurious electric 
field in vacuum due to interaction between the
polarized slab and is periodic images.

There is an old paper that comments on the situation:

http://dx.doi.org/10.1103/PhysRevB.63.205426

but much more had been published since then.

The really graceful way of studying slabs involves using a code that can do 2D 
PBC in plane and break periodicity along the
perpendicular direction. But then you have to abandon the PW basis -- i.e., 
codes like VASP and QE -- and use something else.

HTH,

Serge


-- 
Serge M. Nakhmanson
Dep-t of Materials Science & Engineering, and Institute of Materials Science
University of Connecticut
Storrs, CT 06269-3136
Phone: (860) 486-5252
http://satori.ims.uconn.edu/

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20141205/38bad426/attachment.html>


More information about the users mailing list