I am doing ferromagnetic calculations on Mn And C co-doped ZnO using quantum espresso and I want an idea about the values of starting magnetization to use.
I used 0.5 and -0.4 for Zn and Mn respectively.. But the relax calculation runs the first Scf for 100 iterations and does not converge.
You need to specify which atom type the starting magnetization is for. I will recommend Mn. Secondly, from your input file it seems you have a 2x2x2 supercell and am sure you want your end doped compound to remain in the wurtzite structure. In realm space, the lattice constant a is about twice the lattice constant c. This means that in reciprocal spaces, c is about twice a. In choosing the k-points, you need to follow the same analogy ie x=y, z=x/2. Having said that, the k-points in your supercell should be what ever value that converged the unit cell divided by 2 if the supercell is 2x2x2. U should also consider increasing your smearing width to say 0.02. You can start with a magnetization of 1 and increase to see the effect till you obtain what you desire. Good luck
Possibly 0.5 (in the unit QE take) would be good starting, still. You may decrease Force convergence threshold to 1.0d-04 and electron to 1.0d-08. I am afraid that your values look bit unreasonable.
- Badges
- Science topic
- Similar topics
- Biological Science
- Kinesiology
- Running