I am currently performing coarse-grained MD simulations in LAMMPS using mW potential for water. I initialize the water model at 273K, equilibrate for 50ns followed by a quenching process from 273K to 200K at a rate of 0.5K/ns for a total of 146ns. At the end I observe what looks like a nucleated structure. But how can I be certain that homogeneous nucleation has indeed taken place?
Can a specific property be extracted from LAMMPS or any visualization technique be used to confirm that the structure has indeed nucleated?
Any insight is appreciated.
Write a script to calculate the radial distribution function of your system along a predefined reaction coordinate between the interface and the buried atoms. Check the coordination number and the behavior of the RDF to get insights about nucleation.
Do you know if your time/length scales are appropriate for nucleation process ? What is the critical nucleus size? Your periodic box should be larger than Rc to avoid artifacts.
1. You can compute the RDF using LAMMPS if the system is isotropic. You can also compute the RDF as a function of the distance vector, as opposed to the distance magnitude that LAMMPS will use for RDF computation. Even for a stable liquid, the RDF will have fluctuations at every frame, which can be smoothened out by averaging over a number of frames.
2. As a sanity check, along with the RDF it can be helpful to check the variation of the pressure/volume depending on the ensemble in question.
3. You might also want to run a few steps of equilibriation after your configuration reaches 200K.
@shaswat: Ty for the detailed answer, especially the advice about equilibration post quenching. I will try and incorporate it into the simulations. I think it'd be important for the overall structural stability.
- Badges
- Science topic
- Similar topics
- Chemistry
- Theoretical Chemistry
- Computational Chemistry
- Molecular Dynamics