Practical advice for metadynamics

Anthony Debellis adebe... at
Sat Mar 11 15:09:05 UTC 2017

I am a new cp2k user who would like to run some metadynamics (DFT) 
simulations.  I have defined a collective variable of the bond length 
difference for 3 atoms.  After running unbiased dynamics for 15ps I can see 
that the CV fluctuates between 3-5 bohr.  I would like some advice on 
setting the parameters for a biased metadynamics run.  For example, what 
are reasonable values for the hill height (I expect the barrier of the 
relevant process to be ~35kj/mol), the hill width, the # of steps between 
hill growth, the langevin parameters of spring constant and particle mass?  
What is the benefit of running the langevin method versus direct 
metadynamics?  I would greatly appreciate any help offered.  Please include 
keywords and units!


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <>

More information about the CP2K-user mailing list