Hello!I have tried both suggestions-- increased the convergence threshold to 1e-8 and used the converged values of CELL_PARAMETERS (so in the phonon calculation it was less than experimental one:
CELL_PARAMETERS (alat= 10.69207046) -0.496242253 -0.000000000 0.496242253 0.000000000 0.496242253 0.496242253 -0.496242253 0.496242253 -0.000000000 )
The result is better (\omega_max=298 cm^-1) but still is not perfect (thick magenta curve in the attached figure). It seems that the germanium have to be squeezed further to fit experimental DOS.
Sincerely yours, Mikhail Kondrin On 24.12.2020 12:39, mkondrin wrote:
Hi, Paolo! Thank you for the hints. I will try all of them. Sincerely yours, Mikhail Kondrin On 23.12.2020 20:09, Paolo Giannozzi wrote:On Wed, Dec 23, 2020 at 1:46 PM mkondrin <[email protected]> wrote:&SYSTEMibrav = 0 A = 5.658A is the experimental lattice parameter and would be fine for ibrav=2 (fcc), but then you set your lattice vectors: CELL_PARAMETERS (alat= 10.69207046)to a larger cell (they are in "alat" unit, that is, the value specified as-0.510237966 0.000000000 0.510237966 -0.000000000 0.510237966 0.510237966 -0.510237966 0.510237966 0.000000000"A" or "celldm(1)", NOT the value in the card). With ibrav=2 you getsomething like 290 cm^-1. Note that the phonons are typically computed atthe theoretical equilibrium lattice parameter, that in LDA is typically smaller than the experimental, leading to higher frequencies. Also note that the convergence threshold conv_thr=1.e-5 in the scfcalculation is too large: reduce it to at least 1.0e-8 (it costs close tonothing with respect to the phonon calculation) Paolo _______________________________________________ Quantum ESPRESSO is supported by MaX (www.max-centre.eu) users mailing list [email protected] https://lists.quantum-espresso.org/mailman/listinfo/users_______________________________________________ Quantum ESPRESSO is supported by MaX (www.max-centre.eu) users mailing list [email protected] https://lists.quantum-espresso.org/mailman/listinfo/users
RPlots.pdf
Description: Adobe PDF document
_______________________________________________ Quantum ESPRESSO is supported by MaX (www.max-centre.eu) users mailing list [email protected] https://lists.quantum-espresso.org/mailman/listinfo/users
