forum@abinit.org

Subject: The ABINIT Users Mailing List ( CLOSED )

List archive

[abinit-forum] Strange result for Pb Z* in PTO using PAW

From: Eamonn Murray <eamonn.murray@gmail.com>
To: forum@abinit.org
Subject: [abinit-forum] Strange result for Pb Z* in PTO using PAW
Date: Tue, 10 Nov 2009 21:39:40 -0500
Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:date:message-id:subject:from:to:content-type :content-transfer-encoding; b=l+E1oWnltVCKjWMNjkwdaExOPHGNDlgM+b8zspEZ6hweseplLmzIKPATf914pzAQ7W Tdk0C/c0RFjIyEkwvA4bT2A0H3sGMA2KfglVEb1eGBTSn6xiV8vCi54r8wMcYeX7IN29 MvncuJ1ZIE54K8kETGAPIyU+zDmFW+qHEps2w=

Hello,

I've been doing some Born effective charge calculations and I notice
some very wrong looking results when using PAW.
I'm using the PAW data from the USPP table, and an input file such as follows
----
ndtset 2
kptopt1 1
kptopt2 3
getwfk2 1
berryopt2 -1
rfdir2 0 0 1
prtwf2 0

ngkpt 2 2 10
ecut 20.0
pawecutdg 40.0

ixc 1 # LDA
nstep 100
tolvrs 1.0d-18

nband 22
natom 5
ntypat 3
typat 1 2 3 3 3
znucl 82.0 22.0 8.0
acell 3*7.350
xred 0.00 0.00 0.001
0.50 0.50 0.500
0.50 0.50 0.000
0.50 0.00 0.500
0.00 0.50 0.500
-----

The published calculated value for Pb in cubic PTO is 3.90.

Here, I get a the total phase= 0.0104, giving a Z* for lead of 10.4 in
the 001 direction. This is fairly far off. I've tried with higher
energy cut-off & kpoint density - but it's still around 9. Also
switching the Pb PAW to the one from the "User contributed table"
doesn't change much.

Doing the same calculation with, say the HGH pseudopotentials, I get a
Z* of 4.1 - so much more reasonable.

Incidentally, I've also tried with BTO, and using PAW here, I find a
Z* of 1.8, with a published value of 2.75 - so not so bad.

If anyone has any ideas whether I've made some mistake or what else
could be going wrong here, I'd appreciate it.

thanks,
Éamonn

[abinit-forum] Strange result for Pb Z* in PTO using PAW, Eamonn Murray, 11/11/2009