Surface tension

GROMACS version: 2018
GROMACS modification: Yes/No
Dear Justin,
I have a simulated system containing one fullerene and 4000 water molecule. I’m confused, how can I get surface tension of this system? which command will give me surface tension parameter?
Best regards,

Surface Tension Coupling equation 52 is how you calculate the surface tension.

So you need to run gmx energy get the appropriate diagonal pressure tensors, box length, and then plug into the above equation.

1 Like

Dear Dr_DBW,
Of course, I did it but my surface tension was achieved negative. I think negative value is not true. I calculated surface tension for pure water by this equation (52), again negative value was obtained. what must I do? what is wrong?
Best regards,

Dear Dr_DBW

When we run gmx energy, as u mentioned we can get the ppropriate diagonal pressure tensors, box length, and then plug into the above equation and get the surface tension.

However, when we run gmx energy there is option to directly calculate surface tension, It should give us the same value or it has some different function?

Provided you have interfaces with the normal parallel to the z-axis, mdrun computes the surface tension and it can be extracted from the energy file with gmx energy.

But you haven’t provided details on your system. Do you actually have interfaces or are you simulating bulk liquid without a vacuum/gas phase?

I was simulating a bulk liquid without any interfaces. After reading manual I think it won’t work here. I have to provide and interface with any gas or vacuum? right? ( As I think a surface is needed to calculate surface/interfacial tension)

Yes, you need an interface/surface. The surface tension can’t be “magically” computed from data from a bulk simulation.

yes… actually the confusion is due to that gmx energy is giving you option to calculat, even if there is no interface and thats why a lot of new user is getting negative value. Anyway thanks for the clarification.

Ah, indeed. That is confusing and I realized that before. But the other solution is to add a boolean in the mdp file to control this output, but that adds an option and requires the user setting it. Another solution is to post-compute it in gmx energy, but that can strictly speaking not be done with the same accuracy, as instantaneous pressure differences are multiplied with instantaneous areas. In practice this difference is probably always smaller than the statistical accuracy, so maybe that is the best solution.

1 Like

Hi Can you clear on doubt.
I have seen some research paper where they are creating 2 interfaces like this
I believe they want to consider the effect of long range interaction and PBC more accurately. But can’t we have single interface with sufficient widths to address the same?

Dear hess,
I have a fullerene in the bulk of water (4000 molecules of water), and I want to calculate the surface tension at the interface of fullerene/water. Would you please help me? How can I do it?

How can you arrange a single interface in a simulation box, which has PBC?

The geometry is of it such that you have to have two interfaces.

Dear Dr_DBW,
I used PBC in three dimension (x,y,z), and I want to obtain surface tension of this system. Is it possible?

You can indeed not have only one interface.
To use get the surface tension from directly from gmx energy you should have the normal of the interface parallel to the z-axis. In other orientations you need to manually compute the surface tensions using the different pressure components and the area.

Dear hess,

Extending the z axis to create an interface and using the below surface tension coupling .mdp for water.

The surface tension value produced using gmx energy # surface tension, is then taking into account the two interfaces or do I need to halve the value again?

Here I have used ref_p 1440 as the reference surface tension * number of surfaces


;using the beresden pressure coupling

Pcoupl = Berendsen
Pcoupltype = surface-tension

;Surface tension coupling for surfaces parallel to the xy-plane. Uses normal pressure coupling for the z-direction, while the surface tension is coupled to the x/y dimensions of
;the box.

;The first ref-p (page 215) value is the reference surface tension times the number of surfaces bar nm,
;the second value is the reference z-pressure bar.

;The two compressibility (page 215) values are the compressibility in the x/y and z direction respectively. ;The value for the z-compressibility should be reasonably accurate since
;it influences the convergence of the surface-tension, it can also be set to zero to have a box
;with constant height.

ref_p = 1440.0 1.0

compressibility = 0.0 4.5e-5

;time constant for pressure coupling
tau_p = 5.0


gmx energy gives the value “#Surf*SurfaceTension” which isn’t entirely obvious, but “#Surf” stands for the number of surfaces along your Z-plane.

So, you need to divide by that number to get the surface tension.

To be precise, what gmx energy presents is the integral of the pressure tensor fluctuations along Z (eq. 55 at this page in the manual). But since it cannot know how many surfaces are present it cannot divide by that number.


1 Like

I have one question here. I have made an NVT run of my simulation. Do I have to make an NPT run to determine the surface tension? If that is so, what should I use as pcoupletype, also what values of compressibility should I use?
I am simulating the effect of surfactant on the IFT of oil and water, wherein the oil is adsorbed over a silica surface. Here I have placed the slab in the xy plane, so the normal to the interface should be parallel to z axis.