How to choose sigma when the fluctuation range of the CV in the reactant region is very different from that in the product region

[qi pang]

Hi everyone,

I’m using metadynamics in PLUMED to study a simple gas-phase association reaction between two molecules.

One of my CVs is the distance between the two atoms that form the new bond. The bond length is about 1.68 Å.

In the reactant basin, the unbiased fluctuation range of this distance CV is about 3 Å, while in the product basin it is only about 0.28 Å.

In this situation, how should I choose σ?

I’m thinking of transforming this distance CV into a coordination-number-like CV to rescale its fluctuation range, so that the ranges in the reactant and product basins become more similar.

For the interatomic distance d, I set d₀ = 1.9 Å and define:

New CV = (1 − (d/d₀)³) / (1 − (d/d₀)⁶)

With this definition, the fluctuation ranges in the reactant and product basins become about 0.12 and 0.20, respectively, which makes σ easier to choose. Is this approach reasonable?

Thank you!