Dear all vasp users,
I am new user of VASP. I ran the SCF calculation to relax my structure (hexagonal). My question is that how do we know that my structure is stable or not. And this stability is measured with respect to which parameter(s). Can anyone explain this clearly.
Thanks,
Regards,
Vipin
Hi,
in principle you might test the stability of your relaxed structure (once it has reached the convergence and a finite total energy) changing the lattice parameter of your starting geometrical model for the hexagonal structure and running the RELAXATION calculations for all these different values of the lattice parameter.
Plotting the total energy of each relaxed structure once converged as a function of the lattice parameter, you can evaluate if there is a minimum as a function of the lattice parameter. After that you can assume that the relaxed structure with the lattice param. value which minimizes the total energy is the stablest configuration for your system.
Best,
A.
If you really want to make sure your calculations are good, you'll want to make sure your basis set is properly defined. While VASP is fairly robust with its suggested wave-function energy cut-off, it's not a bad idea to test convergence based on the energy cut-off of the plane-wave basis and k-points. The higher the energy cut-off for the plane waves, the larger the basis set you have, and with more k-points you have a better description of the energy bands formed and the interactions with "other" cells around it.
You can do several calculations with different energy cut-offs and k-points independently of each other, and plot the total energy of the cell (or lattice parameter if you relax the cell parameters) as a function of the energy cut-off or number of k-points. The energy as a function of cut-off or k-points should converge to a point and will tell you how much energy cut-off or kpoints you need for numeric accuracy.
If you relax the cell parameters, this also affects the basis set as the plane waves used are defined a multiples of the reciprocal lattice vectors constrained by a specific plane wave energy cut-off. You may want to restart a relaxed cell parameters calculation a few times after it has converged to make sure you do not have to much Pulay Stress. If cell relaxation converges within 5 steps or less, you are usually good. However, If you made sure your energy cut-off is high enough, you really shouldn't have to do this much.
What Angelo Pidatella is referring to is making sure your relaxed volume or cell parameter is the lowest value. You can restrict the cell parameters, and calculate the energy at different variations of you cell parameter(s) or volume. When you plot the energy of the cell as a function of volume or cell parameter, it should have a parabolic shape. If your cell is properly relaxed it should be the minimum of the parabola.
one more suggestion to the well detailed above: with your relaxed structure converged using the best basis set you found, compute the phonon density of states (or bands). If you see imaginary (negative) frequencies, your structure is not stable, while if all frequencies are real (positive) then your structure is dynamically stable and could be possible to synthesize it experimentally.
Hello all,
I get two things described above- (1) Energy vs energy cut-ooff or k-points. (2) phonon density of states or band structure. Regarding cell parameter means I have to change the unit cell size but how can we change the size of unit cell it is fixed. I didn't understand it. Can you please give me some hint.
Thanking you all to your replies to give me proper information.
Regards,
Vipin
You'll have to do more than one calculation, each with a different volume or lattice parameter.
http://personales.unican.es/junqueraj/JavierJunquera_files/Metodos/Structuralproperties/Lattice-constant-si/Exercise-structure-si.pdf
- Badges
- Science topic
- Similar topics
- Chemistry
- Theoretical Chemistry