Dear all: I'm using QE-5.0.1 to calculate phonons of a hypothetical crystal. It is known to all that an imaginary phonon frequency at non-Gamma point means the crystal structure is not stable. My question is how to adjust the crystal structure to eliminate the imaginary phonon frequency? My method is as follows.
(1) calculate the dynamical matrix at the q-point which has imaginary phonon frequency; (2) diagonalize the dynamical matrix by dynmat.x; (3) plot the imaginary phonon modes by Xcrysden to see the movements of atoms; (4) adjust the inner-coordinations of atoms towards the direction of motion, corresponding to the imaginary phonon modes. Relax this new structure and recalculate phonons to see whether there is some improvement or not. Is the procedures correct? But after relaxation, atoms do not form a stable structure and go back to the initial positions, which has imaginary phonon frequency. Any suggestions? Any help will be appreciated. Best Miao Gao --------------------------------------- Ph.D. student of Department of Physics, Renmin University of China. --------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: http://pwscf.org/pipermail/pw_forum/attachments/20130415/5e8a14b2/attachment.html
