10
$\begingroup$

I have built a CG model of a surface on which I want to place a water droplet (some water molecules), and calculate the contact angle. Has anyone done similar work with gromacs (MD simulation)? First, I just want to place some water molecules in a sphere on the surface. Any suggestions?

Improve this question
$\endgroup$
1
  • $\begingroup$ Isn’t it easier to calculate it from the surface tension ? $\endgroup$
    – Greg
    Mar 29, 2022 at 1:33

1 Answer 1

Reset to default
12
$\begingroup$

For generating the droplet, there are two ways. The 'dumb but it works' way is to just generate a box of water adjacent to the surface, and let it equilibrate in vacuum for a time. The box will naturally adhere to the surface and form a spherical droplet after a few 10's of nanoseconds. This has the advantage that you can keep the number of waters equivalent across simulations with different surfaces - just use the same waterbox while swapping out the surface.

The slightly the more complex way is to first generate a regular box on top of the surface. Then, choose a point on the surface to act as the centre of a droplet, and delete any waters that are greater than some distance R (where R is the radius of a sphere). I suggest using VMD or mdanalysis for this task but I'm sure lots of programs can achieve it.

Finally, a trick to improve correspondence with experiment: nano-scale waterdroplets may not accurately represent the contact angle observed in experiment. This is due to the ratio between the distance of the water/surface/vacuum contact line vs the surface area of the interface. To avoid this, one can make the surface periodic with a small cell distance in the x direction (or y), say 1-2nm, so that the droplet forms a cylinder. This makes the contact line linear and removes the effect of so-called 'line tension'. This is described in "On the Water-Carbon Interaction for Use in Molecular Dynamics Simulations of Graphite and Carbon Nanotubes", DOI 10.1021/jp0268112

Improve this answer
$\endgroup$
7
  • 1
    $\begingroup$ Here's another example: doi.org/10.1021/jp063896o $\endgroup$
    – WaterMolecule
    Mar 28, 2022 at 19:47
  • $\begingroup$ Thanks a lot for the answer. I was just wondering, is it a good idea of having a spherical water droplet instead of a water box from the beginning? $\endgroup$
    – Roshan Shrestha
    Jul 20, 2022 at 14:46
  • $\begingroup$ If you're really strapped for time, then yes - because the droplet is then closer to being equilibrated. Even if it's square, if will adhere to the surface and form a spherical droplet after a few 10's of nanoseconds anyway $\endgroup$
    – lewiso1
    Jul 21, 2022 at 3:38
  • 1
    $\begingroup$ You may need to post a new question to provide the details - for example, how (and why) are you calculating the density? Is the z direction parallel to the surface, or normal to it? In any case, if you're in NVT, just divide the number of waters by the volume of the periodic cell to get the number density of water. If it's NVT then that won't be changing. $\endgroup$
    – lewiso1
    Sep 1, 2022 at 5:22
  • 1
    $\begingroup$ perhaps looking through some publications will help. Here's one I found by googling 'water contact angle molecular dynamics cylindrical': core.ac.uk/download/pdf/199431488.pdf . Title is 'Generalized line tension of water nanodroplets' if the link goes down $\endgroup$
    – lewiso1
    Dec 4, 2022 at 22:10

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .