# Converging the potential of a dilute binary alloy system using the Coherent Potential Approximation. What troubleshooting options could I take?

I'm trying to converge a dilute fcc alloy consisting of Cu and U, with the Uranium concentration at 0.01% and the lattice parameter of Cu, using Hubert Ebert's Spin Polarized Relativistic Korringa-Kohn-Rostoker (SPR-KKR) program.

However it seems that after one iteration the energy of the lower limit of the semi-core states shoots above the upper energy limit of the core states. This problem only started to arise when I introduced the valence of the Cu and U atoms as parameters. I set the valence of U as 12 and the valence of Cu as 10, but have also tried 1 for the Cu valence. I have tried lowering the energy range of my calculation but that also doesn't seem to work, now I'm stuck and I'm not sure what other troubleshooting options to do.

• Maybe I can write to Hubert Ebert and see if we can get an answer that way. – Nike Dattani Sep 10 '20 at 3:16
• I've sent the email now to Prof. Dr. Ebert. Let's see if he answers it here now. – Nike Dattani Oct 6 '20 at 23:01
• Hi Nike, I think the issue I had was actually with my starting potential, I find that starting with a pure U structure in a Cu lattice and then converging the potentials from previous whilst increasing the Cu concentration until eventually I'm at 99% Cu conc. solves the issue. – LUPHYS Oct 8 '20 at 14:00
• please write an answer then. Unamswered questions can be harmful to our site and community. – Nike Dattani Oct 8 '20 at 14:04

I managed to get the converged U$$_{0.01}$$Cu$$_{0.99}$$ potential by converging a pure Uranium potential in a Cu lattice. Then by increasing the concentration of Cu and using the old converged potential as a starting potential you can increase the concentration of Cu in steps until finally reaching a converged U$$_{0.01}$$Cu$$_{0.99}$$ potential. This is a good method to try in any binary alloy system if you have a point that doesn't seem to converge.