Metadynamics using cp2k with CNs as CV

[Meili Liu]

Dear all,

I am a newer, I learn to use cp2k to run metadynamics with one CV (Si atom coordinated with bridge-Oxygen), When I analyzed the FES, I don't understand the phenomenon that CN is 0 (showed in fig. 1, and second list in fes_2.5ps.dat), it is not consistent with my structures (3 Si-O(bridge Oxygen)). And the attached are my input file and fes_2.5ps.dat (obtained from graph.popt -ndim 1 -ndw 1 -file Meta_CV_5000.restart -cp2k). Could someone give me any suggestions ?

Thanks a lot.

Best,

Meili

[Marcella Iannuzzi]

Dear Meili

This is the value calculated from the definition of COLVAR in the input and from the coordinates.

Have you check the structure, apparently the defined CN is at the beginning larger than two, but it quickly goes to zero (in less than 0.5 ps)

and there it remains. You should be able to identify the corresponding structural change.

Have you monitored the dynamics of the CN? 

I cannot judge whether  this is the correct behaviour or whether the CN is correctly defined. 

For sure it is not optimal to run MTD with only one CV. Moreover hills are spawned every 7.5 fs and the height of the hill is 1.36 eV, which is a lot. 

Probably something went wrong in the sampling. 

Regards

Marcella

[Meili Liu]

Dear Marcella,

Thanks very much for you opinions. I checked the structures and monitored the dynamics of the CN. Firstly, structures look normal.

but the changes of bridge-Oxygen are very fast that like what you said (from 2 to 0  during less than 0.5ps). I don't understand why is it

so fast? and I want to make sure how to know the value of 1.36 eV for the height of the hill? 

Many thanks, 

Meili