I am unable to achieve convergence in my VC Relax calculation for a doped Bismuth structure using Quantum Espresso. When I ran the calculation for only Bismuth it converged but after I doped it with Cu, It has failed to converge. With some of the input files, I get the error: "Error in routine bfgs (1): dE0s is positive which should never happen", while with other similar doped structures the calculation terminates upon reaching the maximum number of steps. What could be the reason for this? I have attached here the input file and output file of Bi that achieved relaxation/convergence as well as the doped Bi (Bi and Cu) that did not achieve convergence. Please let me know your thoughts.
Please consider using smearing = gaussian and decreasing the mixing beta to 0.1. You can also use degauss = 0.01.
You are welcome, I am happy that the calculation has been successful. Good luck
S. Muhammady
I was facing the same error problem when I ran the input file for doping Pt on a specific site of SnO2 surface. So as said I resubmitted the calculation by setting mixing beta 0.1 (earlier it was 0.7). Smearing is Gaussian and degauss is 0.01 as before. I hope my issue gets resolved now.
- Badges
- Science topic
- Similar topics
- Filing