[QE-users] Phonons for a metal with DFT+U at finite q-points
Iurii TIMROV
iurii.timrov at epfl.ch
Fri Nov 26 16:08:57 CET 2021
Dear Akkarach,
Try to converge better the ground-state calculation: use conv_thr in the range from 10^{-10} to 10^{-15}.
Indeed, as Mahmoud suggested try to lower alpha_mix(1) to 0.1 or so.
In general, do you observe significant changes in the band structure of SrCoO3 when you use a +U correction? Is it a "true metal" (i.e. experimentally it is known to be metallic)? From my experience, converging phonons+U even for bulk Ni (which is a true metal) with U>2 eV is very problematic.
HTH
Iurii
--
Dr. Iurii TIMROV
Senior Research Scientist
Theory and Simulation of Materials (THEOS)
Swiss Federal Institute of Technology Lausanne (EPFL)
CH-1015 Lausanne, Switzerland
+41 21 69 34 881
http://people.epfl.ch/265334
________________________________
From: users <users-bounces at lists.quantum-espresso.org> on behalf of Mahmoud Payami Shabestari via users <users at lists.quantum-espresso.org>
Sent: Friday, November 26, 2021 2:48:16 PM
To: asukserm at ictp.it; Quantum ESPRESSO users Forum
Subject: Re: [QE-users] Phonons for a metal with DFT+U at finite q-points
Dear Akkarach,
You may play with "alpha_mix(i)", where i is the iteration number.
HTH
Mahmoud
---------------------------------------
Mahmoud Payami
NSTRI, AEOI, Tehran, Iran
Email: mpayami at aeoi.org.ir
Phone: +98 (0)21 82066504
--------------------------------------------------------
From: Akkarach Sukserm via users <users at lists.quantum-espresso.org>
To: users at lists.quantum-espresso.org
Date: Fri, 26 Nov 2021 13:55:59 +0100
Subject: [QE-users] Phonons for a metal with DFT+U at finite q-points
Dear QE developer and users:
I am trying to calculate phonons for a metal with DFT+U at finite q-points
(different from Gamma) with the ph.x code. With a few exceptions (very low
U), the self-consistent iteration of the phonon calculation does not
converge.
My system is a SrCoO3 which is a metal and ferromagnetic material. I use a
DFT+U calculation for considering the magnetic atom (Co) including the
smearing method for the metal system. I have tried to vary the parameters,
such as U-parameter, k-point grid = 4x4x4 and 8x8x8, degauss of smearing,
and pseudopotential types.
A representative set of SCF and PH inputs are below.
Do you have a suggestion or a method to solve this problem?
Thank you in advance.
Akkarach Sukserm
ICTP, Italy
asukserm at ictp.it<mailto:asukserm%40ictp.it>
----------------------------------------------------------------------
&CONTROL
calculation = "scf"
forc_conv_thr = 1.00000e-04
nstep = 100
outdir = "./temp/"
prefix = "prefix"
pseudo_dir = "./PP_PAW_PBE"
/
&SYSTEM
degauss = 1.00D-02 # vary 0.01, 0.02, 0.05, 0.08
ecutrho = 7.20D+02
ecutwfc = 8.00D+01
ibrav = 0, nat = 5
nspin = 2, ntyp = 3
occupations = "smearing" , smearing = "marzari-vanderbilt"
starting_magnetization(2) = 0.50D0
lda_plus_u = .TRUE.
hubbard_u(2) = 3.00D0 # vary U= 1.25 and 3.00 eV
/
&ELECTRONS
conv_thr = 1.0e-08
diago_david_ndim = 4
diagonalization = "david"
mixing_beta = 0.7D0
mixing_mode = "plain", startingpot = "atomic", startingwfc =
"atomic+random"
/
K_POINTS {automatic}
8 8 8 0 0 0 # vary kp-mesh = 4x4x4 and 8x8x8
CELL_PARAMETERS {bohr}
7.216934286 0.000000000 0.000000000
0.000000000 7.216934286 0.000000000
0.000000000 0.000000000 7.216934286
ATOMIC_SPECIES # vary (1) PAW-PBE, (2) PAW-PBEsol, (3) USPP-PBEsol
Sr 87.62000 Sr.pbe-spn-kjpaw_psl.1.0.0.UPF
Co 58.93320 Co.pbe-spn-kjpaw_psl.0.3.1.UPF
O 15.99940 O.pbe-n-kjpaw_psl.1.0.0.UPF
ATOMIC_POSITIONS {crystal}
Sr 0.000000 0.000000 0.000000
Co 0.500000 0.500000 0.500000
O 0.500000 0.500000 0.000000
O 0.500000 0.000000 0.500000
O 0.000000 0.500000 0.000000
----------------------------------------------------------------------
This is my ph.x input:
----------------------------------------------------------------------
Phonon of SrCoO3 at q = 0, 0, 0.25 #NAME TAG
&inputph
tr2_ph=1.0d-12,
amass(1)=87.62000,
amass(2)=58.93320,
amass(3)=15.99940,
prefix='prefix',
alpha_mix(1)=0.7D0 # Vary 0.5 and 0.7
outdir='./temp/'
fildyn='./ SrCoO3_q.dynq',
/
0.0 0.0 0.25
----------------------------------------------------------------------
This is my ph.x output:
---------------------------------- PH.out ----------------------------
Representation # 1 mode # 1
Self-consistent Calculation
iter # 1 total cpu time : 2300.3 secs av.it.: 17.2
thresh= 1.000E-02 alpha_mix = 0.700 |ddv_scf|^2 = 9.086E-03
iter # 2 total cpu time : 2384.5 secs av.it.: 29.1
thresh= 9.532E-03 alpha_mix = 0.700 |ddv_scf|^2 = 1.768E+01
iter # 3 total cpu time : 2469.9 secs av.it.: 29.9
thresh= 1.000E-02 alpha_mix = 0.700 |ddv_scf|^2 = 1.807E+01
iter # 4 total cpu time : 2533.8 secs av.it.: 20.9
thresh= 1.000E-02 alpha_mix = 0.700 |ddv_scf|^2 = 4.797E+01
iter # 5 total cpu time : 2607.1 secs av.it.: 24.9
thresh= 1.000E-02 alpha_mix = 0.700 |ddv_scf|^2 = 1.311E+02
…
iter # 98 total cpu time : 21889.7 secs av.it.: 138.5
thresh= 1.000E-02 alpha_mix = 0.700 |ddv_scf|^2 = 1.431E+49
iter # 99 total cpu time : 22224.4 secs av.it.: 139.3
thresh= 1.000E-02 alpha_mix = 0.700 |ddv_scf|^2 = 4.764E+49
iter # 100 total cpu time : 22563.4 secs av.it.: 140.9
thresh= 1.000E-02 alpha_mix = 0.700 |ddv_scf|^2 = 1.497E+50
End of self-consistent calculation
No convergence has been achieved
------------------------- DONE --------------------------
_______________________________________________
Quantum ESPRESSO is supported by MaX (www.max-centre.eu<http://www.max-centre.eu>)
users mailing list users at lists.quantum-espresso.org<mailto:users%40lists.quantum-espresso.org>
https://lists.quantum-espresso.org/mailman/listinfo/users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20211126/fffbd331/attachment.html>
More information about the users
mailing list