Hello,
I am trying to simulate wetting of water droplet on graphite. In my simulation, the droplet doesn’t maintain its shape and it gradually spreads over surface as if it’s evaporating. In the other words, it doesn’t form a meniscus. Does anyone has any experience with simulation of wetting? Could anyone please give me any advice how to set up the simulation such that I can model wetting properly?
I used SPCE model for water and rebo potential for graphite and Lennard Jones for interaction between graphite and water. I applied periodic boundary conditions in x and y directions and fixed boundary condition in the direction normal to graphite plane. I placed waterbox on top of graphite close to surface and ran the simulation under NVE ensemble. Temp/rescale was used to adjust the temperature. “move” command was used to freeze the substrate. To equilibrate the system, I gradually increase the temperature from 10 K to 298 K. When temperature passes 100 K it starts spreading over surface and water molecules move far away from each other. I ran this simulation also under NVT ensemble and I qualitatively got similar distribution of water molecules. I have attached the structure data and input file and a snapshot of the simulation.
I will greatly appreciate any help.
Farshad
mixed-graph-water.dat.zip (174 KB)
graphite.zip (471 KB)
Hello,
I am trying to simulate wetting of water droplet on graphite. In my
simulation, the droplet doesn't maintain its shape and it gradually spreads
over surface as if it's evaporating. In the other words, it doesn't form a
meniscus. Does anyone has any experience with simulation of wetting? Could
anyone please give me any advice how to set up the simulation such that I
can model wetting properly?
I used SPCE model for water and rebo potential for graphite and Lennard
Jones for interaction between graphite and water. I applied periodic
boundary conditions in x and y directions and fixed boundary condition in
the direction normal to graphite plane. I placed waterbox on top of graphite
close to surface and ran the simulation under NVE ensemble. Temp/rescale
was used to adjust the temperature. "move" command was used to freeze the
substrate. To equilibrate the system, I gradually increase the temperature
from 10 K to 298 K. When temperature passes 100 K it starts spreading over
surface and water molecules move far away from each other. I ran this
simulation also under NVT ensemble and I qualitatively got similar
distribution of water molecules. I have attached the structure data and
input file and a snapshot of the simulation.
I will greatly appreciate any help.
you have an extremely small droplet. why should it behave like a
macroscopic drop?
axel.
Dear Farshad,
I have only rudimentary knowledge of this as I am currently trying out the same droplet on substrate simulation as you are.
As far as my own experience goes, I had to equilibrate the droplet extremely slowly, on the order of 500ps, using fix nvt from 1K to 300K and further equilibration at 300K, to get a fully stable, well equilibrated configuration. Initial configuration of the water was generated with Packmol. Some papers report much faster equilibration (order of 200ps with 2000 water molecules) but I couldn’t achieve that (probably need to read more), but in any case my two cents. And the timestep need not be so incredibly small, 1fs should be sufficient as most papers report.
Also your water molecules are initially only 1A away from graphite, might be too small resulting in (rather) violent kinetics upon starting the simulation. Maybe you could try 3A.
If anyone has a better idea kindly let me know as well as I would very much like to learn more regarding water droplet simulation as well.
YMMV and HTH.
Best Regards,
Jingjie Yeo
Ph.D. Student
School of Mechanical and Aerospace Engineering
Nanyang Technological University, Singapore
Thank you all for your response.
I am currently running a much larger system and hopefully this time the system will respond more realistically to the applied dynamics.
Farshad
Thank you all for your response.
I am currently running a much larger system and hopefully this time the
system will respond more realistically to the applied dynamics.
i suspect you are making a mistake in thinking here. when something
behaves different from your expectations, that doesn't automatically
mean that it is not realistic. it may very well behave according to
the boundary conditions that you have imposed.
remember we are not in the business of making movies where it only
matters that something looks the way it is expected.
axel.