<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html; charset=UTF-8" http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
Some time ago I did some calculations for cubic HfO2, but did not find
any serious problems. In the beginning there was a Ghost-band warning,
but the QTL-B value is not as high as that. I got rid of the warning by
reducing the RMT value a little bit. So it might be related to the RMT
value. In my case, I used RMT(Hf)=2.0 and RMT(O)=1.80. <br>
<br>
<br>
On 08/17/2009 12:43 PM, Laurence Marks wrote:
<blockquote
cite="mid:876512660908162143o13b7f055m2acae84b8b6e12b5@mail.gmail.com"
type="cite">
<pre wrap="">If this is occurring in the very first cycle it is a bit harder to set
the case.in1 to avoid this. Maybe someone has done HfO2 so has a
specific suggestion; two that might work:
a) Switch the L=1 energies so the lower one is an lo, the higher
APW+lo, i.e. use
1 0.30 0.000 CONT 1
1 -2.16 0.002 CONT 1
b) Run just a single HfO2 cell (not a superstructure) and use this to
guide what the linearization energies should be.
c) Redo the initialization and make sure that you did not make a
mistake somewhere.
2009/8/16 Matthew Field <a class="moz-txt-link-rfc2396E" href="mailto:matthew.field@rmit.edu.au"><matthew.field@rmit.edu.au></a>:
</pre>
<blockquote type="cite">
<pre wrap="">Dear Dr. Blaha and Wien2k users,
I am running WIEN2k version 09.1 on an Australian supercomputer cluster VPAC.
Ultimately I wish to use WIEN2k to calculate ELNES spectra for a variety of oxides using the TELNES package, but I am currently having problems with the scf cycle, which constantly returns QTL-B errors with ghost bands. I have tried solving the problem with the FAQ and with a few suggestions from the mailing list, but this hasn't worked.
Attached to this mail are the following files:
case.dayfile
case.in1
case.scf2
case.struct
lapw2.error
I am currently trying to get a cubic HfO2 supercell structure to work. The calculation stops after a single scf cycle, with an error in lapw2. The lapw2.error file reads:
'l2main' - QTL-B.GT.15., Ghostbands, check scf files
The message at the bottom of the case.scf2 file reads:
QTL-B VALUE .EQ. 50.63333 in Band of energy -7.99540 ATOM= 2 L= 1
Check for ghostbands or EIGENVALUES BELOW XX messages
Adjust your Energy-parameters or use -in1new switch, check RMTs !!!
:WARN : QTL-B value eq. 50.63 in Band of energy -7.99540 ATOM= 2 L= 1
:WARN : You should change the E-parameter in case.in1 or use -in1new switch
According to my structure, this would be the second Hafnium atom in the list.