Hello, dear Ahsan.
I am not aware if CP2K has a devoted instrumentary.
For such a task,
I would prepare input files for geometries where distance between the two atoms is systematically varied
to compute potential energies for the geometries using single point
RUN_TYPE ENERGY.
After plotting the potential energies on distances,
if the suggested range of distances allows capturing the repel, the binding and the dissociation,
I think one would need to write a fitting program for the Lennard-Jones interaction to account sigma and epslon
in proper units.
Some time ago, I did this using Mathematica.
For point-like structural elements this works.
It becomes more "exciting" if you address interactions of larger systems.
Victor