I want to calculate time dependent interaction energy for water droplet and atom substrate. To accelerate the computation speed I used neigh_modify exclude group carbon carbon. However, I doubted about the effect of this exclusion on calculated interaction energy between water and carbon atom.
This command will accelerate computation speed only because you are excluding the atoms from common neighbor lists. This will also remove the computation of forces and energy between them. I don’t know what you expect from it, nor do I know which atoms the carbon group contain, but I highly doubt this leads to meaningful and physical results in a production simulation.
The rest of your input is over-engineering compared to what I understand you want to do and I am not sure you are computing what you say you want to compute, but only a profile of the potential energy of your left_drop group.
Why not simply use compute group/group between your droplet and your surface?
For more clarification, I have some water drops (30A radius) on the carbon substrate surface and hereby want to compute the interaction energy for one of these water drops with substrate surface.
So, I follow your suggestion and use compute group/group as follows:
moreover, concerning the neigh_modify exclude group carbon carbon: I think it has no effects on above output data, since it has only neglect the carbon carbon interactions.
Well you will not compute any interaction between atoms in the carbon group at all. This means that if you perform any molecular dynamics, be it NVE, NVT or whatever, they will not see one another nor will they interact during the simulation. It’s up to you to consider if this correspond to your simulation settings, but from a physics point of view, it makes little sense considering the system you describe.
One possibility is to save a set of configurations from an actual MD simulation where all the forces are duly computed as prescribed in the force field and then you can use this set of configurations as input to a rerun (rerun command — LAMMPS documentation). This would allow you to turn off whatever itneractions you want to turn off using the neigh_modify exclude command. Then you can output the global values of energy concerning the interactions you want as well as possibly local data to build a plot.
On what concerns defining the surface atoms in the substract: I suppose you would need a criterion to decide which atoms belong to the water/substrate interface. Speaking from a different context: in the case of some MOF/polymer interfaces, it has been observed that some thickness of “pure” polymer phase (i.e., outside the region of overlap between with the MOF) to be required before you are able to converge to the density prescribed for the bulk material. I suppose this is one way of defining the length of an interface: how big the thickness before converging to bulk properties is. But also I suppose you could opt for a simplistic approach where you consider the interactions of the bottom of the water droplet and carbon atoms lying within a range of r <= X from it (X being a value set by you).
One more important note is to convert interaction energy data from LAMMPS real unit, kcal/(mol.Ă…) into the electronvolts. what is the mole here? Is it water plus carbon moles? If it is true, it should be complicated, Since the drop interacts with a apart of the substarte.
I hope you saw Stan Moore’s post on your other thread. Your current input means that the group left_drop contains molecules 1 AND 1747. Only those two molecules.
To select a molecule range you need to specify 1:1747. Note the colon.
I am not sure what you want to do with your potential energy calculation. MD potential energies aren’t particularly useful unless measured relative to some sensible reference configuration. (As a simple example – the energy of “the water drop on the surface” includes both the water-carbon adhesion and the water-water deformation, since presumably some surface water molecules will be geometrically frustrated relative to their ideal bulk solvation environment.)
