[QE-users] Energy of isolated atom of CaO
Pietro Delugas
pdelugas at sissa.it
Fri Jun 26 08:28:11 CEST 2020
Hello
In ATOMIC_POSITIONS you need to specify the units that you are using for the coordinates, if you don’t put anything he program assumes alat units and puts the 2 atoms very far apart.
If you actually wanted compute the two separated atoms specifying startingwfc=’random’ in the &electrons name list may help. In this case I would also chose calculation=’scf’ in &control.
Regards - Pietro
Sent from Mail for Windows 10
From: Pooja Vyas
Sent: Friday, June 26, 2020 8:12 AM
To: Quantum ESPRESSO users Forum
Subject: [QE-users] Energy of isolated atom of CaO
Dear users,
I want to calculate cohesive energy of CaO. I'm using cluster_example/PW which uses assume_isolated = 'martyna_tuckerman' . Following is my input script:
#!/bin/sh
for a in 30
do
cat > ${a}.in << EOF
&CONTROL
calculation = 'relax'
prefix = "${a}",
pseudo_dir = "/home/user/pv/cohesive/pseudo/",
outdir = "/home/user/pv/cohesive/",
/
&SYSTEM
ibrav = 1,
celldm(1) = $a,
nat = 2,
ntyp = 2,
ecutwfc = 100.D0,
assume_isolated = 'martyna-tuckerman'
/
&ELECTRONS
conv_thr = 1.D-8,
mixing_beta = 0.7D0,
/
&IONS
/
ATOMIC_SPECIES
Ca 40.078 Ca.pbe-nsp-van.UPF
O 15.999 O.pbe-van_ak.UPF
ATOMIC_POSITIONS
Ca 0.0 0.0 0.0
O 0.5 0.5 0.5
K_POINTS Gamma
EOF
mpirun -np 40 -machinefile x1 /apps/codes/qe/6.4/bin/pw.x < ${a}.in > ${a}.out
done
done
My lattice constant is around 8 a.u , so I assumed the box size to be 3 times of it. But when I run the script, the convergence was not achieved even after 100 iterations. Then I tried to take small box size of about 16 bohr. That too didn't work. Then I even tried reducing ecutwfc and mixing_beta..but same error was obtained. Can anyone suggest what can I do to resolve the issue?
Any kind of help is appreciated.
Thanks and regards.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20200626/9cab07bb/attachment.html>
More information about the users
mailing list