[QE-users] Broken kramers degeneracy in antiferromagnet
Dear users, I have a problem computing the electrons band structure of BiFeO3, antiferromagnet (no SOC), R3c space gorup, with Hubbard corrections on Fe. I define different starting_magnetizations for the two sublattices of spin centers, as described in the Input File Description. When I compute the electronic bands I see that the degeneracy of the spin-up and spin-down states is lifted only along certain directions. Since the time-reversal symmetry is present but the inversion symmetry is broken I would expect the bands to respect Kramers degeneracy: E_up (k) = E_dn (-k) . This is not what I find, as you can see from the picture attached. The energy difference can be as large as 0.2 eV, so this is not numerical noise. I tried different solutions, including using an explicit k-points mesh, use nosym=true, starting from different initial magnetizations, but I always obtain the same odd behaviour. I also tried FeO3, same procedure, with an automatically generated k-mesh, and the spin-polarized bands are degenerate as they should. I paste below the input file of the self consistent calculation, I'm using version 7.0, with PAW pseudopotentials. I appreciate any suggestion, thank you! Francesco Delodovici, CentraleSupelec, Universitè Paris-Saclay. ## &control calculation='scf' restart_mode='from_scratch', pseudo_dir = './pseudop/', outdir='./' prefix='BFO' tstress = .TRUE. tprnfor = .TRUE. / &system ibrav = 0, celldm(1)=10.660942, nat= 10, ntyp=4, ecutwfc =120.0, ecutrho=600.0 occupations='smearing', smearing='gaussian', degauss=0.02 nspin = 2, starting_magnetization(2) = -1 starting_magnetization(3) = 1 lda_plus_u=.true. lda_plus_u_kind = 0 Hubbard_U(2) = 4.5 Hubbard_U(3) = 4.5 / &electrons conv_thr = 1.0e-9 / ATOMIC_SPECIES Bi 208.98 Bi.upf Fe1 55.85 Fe.upf Fe2 55.85 Fe.upf O 15.999 O.upf CELL_PARAMETERS (alat= 10.66094200) 0.496963620 -0.286922080 0.819002251 -0.0 0.573844160 0.819002251 -0.496963620 -0.286922080 0.819002251 ATOMIC_POSITIONS (crystal) Bi -0.0064211942 -0.0064211938 -0.0064211942 Bi 0.4935787307 0.4935787311 0.4935787307 Fe1 0.2189690501 0.2189690504 0.2189690501 Fe2 0.7189689110 0.7189689114 0.7189689110 O 0.3893335266 0.5347428617 0.9390758847 O 0.9390758845 0.3893335266 0.5347428617 O 0.5347428621 0.9390758842 0.3893335268 O 0.0347428643 0.8893334799 0.4390758846 O 0.8893334801 0.4390758843 0.0347428647 O 0.4390758848 0.0347428643 0.8893334799 K_POINTS (automatic) 6 6 6 0 0 0 bands.pdf Description: Adobe PDF document ___ The Quantum ESPRESSO community stands by the Ukrainian people and expresses its concerns about the devastating effects that the Russian military offensive has on their country and on the free and peaceful scientific, cultural, and economic cooperation amongst peoples ___ Quantum ESPRESSO is supported by MaX (www.max-centre.eu) users mailing list users@lists.quantum-espresso.org https://lists.quantum-espresso.org/mailman/listinfo/users
[QE-users] Parallelization nscf calculation
Dear users, I need to use wannier90 on top of quantum espresso to deal with a multilayer BiFeO3-LaFeO3 with 60 atoms overall. I run the scf calculation to compute the Kohn-Sham orbitals on a 6x6x4 k-mesh, and then I need to refine the k-mesh to 18x18x10 with a nscf calculation, asking for 220 bands. The calculation is spin-polarized to deal with antiferromagnetism in BFO with the Hubbard corrections to the Fe d states included. I employ the following parallelization parameters over 360 processors: -nk 36 -nb 10 without OMP parallelization. The problem here is that the nscf calculation does not complete within 24 hours, so I wonder: can I change the parallelization options in order to obtain a better scaling? What would be a better setting? I appreciate any suggestion, even though I know that parallelization does not perform miracles. Thank you, Francesco Delodovici, CentraleSupelec Paris-Saclay Université ___ The Quantum ESPRESSO community stands by the Ukrainian people and expresses its concerns about the devastating effects that the Russian military offensive has on their country and on the free and peaceful scientific, cultural, and economic cooperation amongst peoples ___ Quantum ESPRESSO is supported by MaX (www.max-centre.eu) users mailing list users@lists.quantum-espresso.org https://lists.quantum-espresso.org/mailman/listinfo/users
Re: [Pw_forum] Phonons in 2D boron nitride
Thank all of you very much for your help. Concerning what prof. Sitangshu suggestion: the raising of the cutoff over plane waves works, now I can recover the right (semi)quadratic dispersion for the acoustic out-of-plane phonons. I used smeared occupation because if not effective charges are computed by ph.x, giving rise to the LO-TO splitting at long wavelength (it is just a workaround). I will try, as a double check, to use the lattice parameter of the bulk system, as prof. Paulatto suggested. Regarding the residual strain, of course it vanishes when I increase the cut off performing vc-relax from scratch. I guess this is why my results was not as I expected. Finally my question about the possibility of obtain the correct LO behaviour at low q was related to the article that prof. Marzari reported. I'm glad to know that the corrections will be implemented in the new version of the code. Thank you very much again! ___ Pw_forum mailing list Pw_forum@pwscf.org http://pwscf.org/mailman/listinfo/pw_forum
[Pw_forum] Phonons in 2D boron nitride
Hello all! I have some troubles with phonons in two dimensional hexagonal Boron Nitride, and I am looking for some hints: 1) the out-of-plane acoustic branch at low wave vector is not quadratic, istead is linear. At a first glance, I thought this could be related to some residual forces, but I performed two successive vc-relax calculations and they are 0.00 according to the output. There is some residual pressure on z axis, but I suppouse it is related to the constrain on the c primitive vector (through the flag cell_dofree = '2Dxy' ). I doubt that this is due to some periodic images interaction, because the cell is 20 A in the direction perpendicular to the layer plane. I also increased the q-point mesh from 6x6x1 up to 10x10x1 in the ph.x input. In the latter case the out-of-plane acoustic branch turns out to have a parabolic-like behaviour around Gamma but with opposite concavity. 2) Since I don't include the flag epsil='true' there is no LO-TO splitting, but the slope of the LO branch is vanishing at Gamma and it shouldn't. Any chance that some option exists to obtain correct behaviour of otpical phonons in 2D for vanishing wave vector? I'm using quantum espersso v6.0. Find attached below the input file for vc-relax and phonons. Any suggestion will be appreciated, thank you! Francesco Delodovici, Phd Student Università degli studi di Milano VC RELAX ### &CONTROL title = 'hbn' , calculation = 'vc-relax' , restart_mode = 'from_scratch' , outdir = '/localdisk1/delodovici/' , pseudo_dir = '~/Pseudop/' , prefix = 'hbn' , forc_conv_thr = 1D-8 , etot_conv_thr = 1D-8 , wf_collect = .false. , / &SYSTEM ibrav = 4 , celldm(1) = 4.695926777902829 , celldm(3) = 8.048362859199854 , nat = 2 , ntyp = 2 , ecutwfc = 50 , ecutrho = 200 , occupations = 'smearing' , smearing = 'gaussian' , degauss = 2e-4 , nspin = 1 , / &ELECTRONS conv_thr = 1.0d-9 / &IONS ion_dynamics = 'bfgs' / &CELL cell_dynamics = 'bfgs' cell_dofree = '2Dxy' / ATOMIC_SPECIES N 14.007 N.pz-rrkjus.UPF B 10.810 B.pz-n-rrkjus_psl.0.1.UPF ATOMIC_POSITIONS crystal B 0.00 0.00 0.00 N 0.33 0.66 0.00 K_POINTS automatic 20 20 1 0 0 0 ### PHONONS ## phonons of hBN &inputph tr2_ph=1.0d-14, prefix='hbn', amass(1)=14.007, amass(2)=10.810, outdir='/localdisk1/delodovici', fildyn='phonons.dyn', ldisp=.true., nq1=10, nq2=10, nq3=1, / ___ Pw_forum mailing list Pw_forum@pwscf.org http://pwscf.org/mailman/listinfo/pw_forum
