MetaDynamics MD for transition

Message

tanlx · #1 Post by **tanlx** » Mon May 20, 2013 10:54 pm

Hi Everyone,
Is there anyone know how to setup the metaDynamics to simulate the alloy transition? It is very appreciated if you could help this.
I found the metaDynamics in the website (http://cms.mpi.univie.ac.at/vasp/vasp/M ... ics_I.html) as following:

For the brief overview of the method, see Sec. 6.62.1 and references therein. The computational setup for metadynamics should contain the following items:
1. set the standard MD-related flags: IBRION=0, TEBEG, POTIM, NSW
2. set MDALGO to 11 or 21 for metadynamics with an Andersen, or Nose-Hoover thermostat, respectively (check description of ANDERSEN_PROB and SMASS for the thermostat-specific setting)
3. define the parameters HILLS_H, HILLS_W, HILLS_BIN, see Sec. 6.62.3.
4. define collective variables in the file ICONST (see Sec. 6.62.4), the input parameter STATUS for collective variables must be set to 5
5. if needed, define bias potential in file PENALTYPOT, see Sec. 6.62.4.

I used the MDALGo = 11, but don't know how to setup the ICONST. Can any body help me? Email: tanliuxi@gmail.com.
Thanks very muck for view!

#2 Post by **admin** » Tue May 21, 2013 1:08 pm

ICONST defines constraints. These depend on your system and on the collective variable. A set of different collective variables can be used to sample the same reaction coordinate.

tanlx · #3 Post by **tanlx** » Thu May 23, 2013 6:33 pm

Hi Head admin,
It is very appreciated. Could you show me some sample of the ICONST file?

#4 Post by **admin** » Fri May 24, 2013 11:20 am

In this example
http://cms.mpi.univie.ac.at/vasp/vasp/P ... penaltypot
ICONST file defines two interatomic distances (between atoms 1 - 5 and 1 - 6) in order to make constraints during the metadynamics run (the distance 1 -5 is not allowed to become larger than 1.6 A). This is achieved using the bias potential defined in PENALTYPOT file (seven lines stand for seven Gaussians located as defined by first two columns, each Gaussian is 1.0 eV high with the halfwidth of 0.2 eV).

tanlx · #5 Post by **tanlx** » Mon Jun 03, 2013 7:14 pm

Thanks.

kmrana · #6 Post by **kmrana** » Fri Jul 11, 2014 2:09 am

In metadynamics calculation with vasp, however, I found there is no clear description how to mention different intermediates along the reaction profile. Should I need to perform separate run for each intermediate or they can be put together in the same run? If you please clarify this I would greatly appreciate. Thank you for your help!

#7 Post by **admin** » Fri Jul 11, 2014 9:06 am

How large system to take and how long simulation to run depends on your computer resources. The ideal case is the simulation of the complete story reactant-intermediate-product.

kmrana · #8 Post by **kmrana** » Sat Jul 12, 2014 12:27 am

Is there any example how to define reactant-intermediate-product (in ideal case) in the same input of metadynamics? If you describe it little bit in more details it would be of great help. Thank you very much for your reply!

My system contains a total of 102 atoms, however, 64 of those are fixed. (Simulation Box: ~10 A x 10 A x 15 A)

#9 Post by **admin** » Tue Jul 15, 2014 8:25 am

Have a look at a simple reaction described in JP-CM 20(2007)064211, eventually contact the author of the paper.

kmrana · #10 Post by **kmrana** » Sat Jul 19, 2014 6:06 am

Thank you very much for your suggestion and help!