I have a question about the structure relaxation:
usually using the IBRON=1/2, the algrithm will try to optimize the atom position and get to the global/system minimum energy. However i am thinking whether there is a way to find a local minimum energy. WHich means that the structure relaxation may stop at some energy valley, but not the minimum system energy.
thanks
local minimum energy
Moderators: Global Moderator, Moderator
-
Gu Chenjie
- Newbie
- Posts: 18
- Joined: Thu Nov 25, 2010 5:41 am
local minimum energy
Last edited by Gu Chenjie on Mon Mar 04, 2013 3:11 am, edited 1 time in total.
-
hipertrofia
- Newbie
- Posts: 27
- Joined: Mon Mar 07, 2011 5:24 pm
- License Nr.: 5-216
- Location: Espoo, Finland
local minimum energy
The optimization will usually stop at a local minimum every time. It will only give you a global minimum if you started the calculation with a good guess. For instance, if you do a calculation for GaAs with your atoms positioned initially close to what would be a wurtzite lattice, then the calculation will converge to the wurtzite arrangement, even though the most stable structure for GaAs is zinc-blende.
In short, your ability to find a local minimum depends on how close your starting configuration is to that minimum.
In short, your ability to find a local minimum depends on how close your starting configuration is to that minimum.
Last edited by hipertrofia on Wed Mar 06, 2013 7:32 pm, edited 1 time in total.