Problems with minimum energy pathways calculated by the graph.popt tool.

[René Schliemann]

Hello,

I would like to ask for help concerning the post processing module for 
metadynamic simulations included in the CP2K code graph.popt.
I am investigating dissolution processes at the mineral surfaces in water. 
So far I did some metadynamics calculations with cp2k using three 
collective variables (CVs), each of which are coordination numbers of one 
specific atom at the surface towards surrounding species on the surface and 
OH-groups. 
Now the one thing I am wondering about is the following: 
With graph.popt module I obtain the FES and respective minima like this:
*graph.popt -ndim 3 -ndw 1 2 3 -periodic 1 2 3 -file *-1.restart 
-find-minima -cp2k *
Then I sort the minima by free energy value, from lowest to highest. 
After sorting them I search for two minima having specific CV combinations, 
one being close to the configuration at the start of the simulation
(tolerance of up to 0.2) and the second is the CV configuration where the 
coordination number to surrounding species at the surface becomes
close to zero (again 0.2 tolerance), which means it is dissolved. With the 
found CV points from the minima file I run the following:
*graph.popt -ndim 3 -ndw 1 2 3 -periodic 1 2 3 -file *-1.restart -find-path 
-point-a CV1(start) CV2(start) CV3(start) -point-b CV1(diss) CV2(diss) 
CV3(diss) -cp2k*
What I don’t quite understand is now in the mep.data file the path 
sometimes go through a point which is lower in free energy than the global
minimum found by the minima search. How is this possible, can I somehow 
correct for this or do I have to change sth. in the CV setup?
If you need anymore insight or maybe some files just tell me.
I would appreciate your help. Thank you in advance.

Best regards,
René Schliemann

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20180417/160006ab/attachment.htm>