Dear Wien2k users,
It seems that I have been able to solve the problem by
increasing Rmt of Si. If the problem still persists I would report the same.
with regards,
On Sat, 25 Jun 2022 at 15:59, shamik chakrabarti
wrote:
> Dear Wien2k users,
>
> I hav
Thank you, Madam. I understand.
On Sun, 22 May 2022 at 21:16, Lyudmila Dobysheva via Wien <
wien@zeus.theochem.tuwien.ac.at> wrote:
> 21.05.2022 09:46, shamik chakrabarti wrote:
> > I have obtained converged structure (Volume vs. E) of
> > LiCoSbO6 by using spin polarization only
21.05.2022 09:46, shamik chakrabarti wrote:
I have obtained converged structure (Volume vs. E) of
LiCoSbO6 by using spin polarization only for Co. However, while doing
a,b,c, Gamma variation for monoclinic lattice I have again encountered
Ghost Band error. In this case I have f
Dear Prof. Blaha,
I have obtained converged structure (Volume vs. E) of
LiCoSbO6 by using spin polarization only for Co. However, while doing
a,b,c, Gamma variation for monoclinic lattice I have again encountered
Ghost Band error. In this case I have found that keeping the spin
pol
Dear Prof. Blaha,
Thank you for your advice Sir. Its running fine now.
with regards,
On Thu, 19 May 2022 at 21:55, Peter Blaha
wrote:
> If you do it spin-polarized, create a proper case.inst file with
> non-magnetic Li, Sb and O, and only Co should be magnetic.
>
> instge
If you do it spin-polarized, create a proper case.inst file with
non-magnetic Li, Sb and O, and only Co should be magnetic.
instgen -ask
init -sp -b ...
Am 19.05.2022 um 10:56 schrieb shamik chakrabarti:
Dear Wien2k users,
I am trying to simulate the material as attached
Dear Prof. Gavin,
The problem is solved by changing Rmt of P to 1.45.
Thank you so much!
with regards,
On Sat, 18 Sept 2021 at 20:27, Gavin Abo wrote:
> You might want to check what value you used for -ecut as I noticed a
> WARNING from lstart of WIEN2k 21.1 if the default val
You might want to check what value you used for -ecut as I noticed a
WARNING from lstart of WIEN2k 21.1 if the default value of -6.0 is used:
username@computername:~/Desktop/test/LFP_opt$ ls
LFP_opt.struct
username@computername:~/Desktop/test/LFP_opt$ init_lapw -b -ecut -6.0
...
next is lstart
With such small RMTs it is clear that the structure is wrong.
Am 28.08.2020 um 18:01 schrieb shamik chakrabarti:
Dear Wien2k users,
I am obtaining Ghost band error at the first
cycle during running anatase TiO2 (struct file attached). I have changed
the Rmt of Ti to
Your structure is completely wrong -- the Ti-O distance is 1.41073
Angstroms! You probably made a mistake in the cif or positions (I have not
searched this, you can).
On Fri, Aug 28, 2020 at 11:02 AM shamik chakrabarti <
shamik15041...@gmail.com> wrote:
> Dear Wien2k users,
>
>
Dear Prof, Marks & Prof, Gavin,
Thank you so much for your reply.
It works fine with Rmt (P) = 1.5 & Rmt (O) =1,4. The calculation
converges smoothly.
Thanks once again.
with regards,
On Mon, 13 Jul 2020 at 00:02, Gavin Abo wrote:
> I would suggest refe
I would suggest referring to previous advice given in the mailings list.
For example, there are different ways in which core leakage might need
to be addressed:
https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg19495.html
You might try swapping the RMT of O and P atoms:
https:/
>From a quick test, the issue has nothing to do with the core leakage,
rather the small P RMTs. One approach (probably not the only one) is to use
a larger RMT for the P and smaller for the O.
On Sun, Jul 12, 2020 at 12:25 PM shamik chakrabarti <
shamik15041...@gmail.com> wrote:
> Dear Prof. Gavi
Dear Prof. Gavin,
I am sorry for the previous reply. Yes, after,
grep WARNING *.outputs the following warnings appeared
WARNING: 0.038 PCORE electrons leak out of MT-sphere
:WARNING: 0.038 PCORE electrons leak out of MT-sphere
:WARNING: touch .lco
I am using w2web. After following your suggestion, I got the
following message in w2web page.
Commandline: : grep WARNING *.outputst
Program input is: ""
Execute another command line:
Type of execution:
Also, in case.outputst there is no warning message displayed.
with regards,
On Sun, 12
It might only show that in a terminal. Are you using w2web, are the
warnings in the Li2NiPO4F_check.outputst file?
1. In w2web, click on "command line" under Execution in the left menu.
2. In the "Execute a command line" box, type: grep WARNING *.outputst
3. Click the Run button
Does it show
Dear Prof. Gavin,
I am using wien 2k 19.1 & surprisingly I am not
getting any error regarding core-leakage. (I have kept initial reduction of
RMT to 0%) The output is as given below.
Commandline: x lstart -up
Program input is: "13 -6.0 "
SELECT XCPOT:
recommended: PBE
Probably you need to mention how you handled the core leakage issue
during initialization (init_lapw) of WIEN2k 19.2 since if that was not
addressed it might lead to the QTL error later during the scf:
username@computername:~/wiendata/Li2NiPO4F_check$ ls
Li2NiPO4F_check.struct
username@computer
Dear Sir,
Thank you so much for your solution. It converges smoothly.
with kind regards,
On Mon, 25 May 2020 at 19:08, Peter Blaha
wrote:
> No.
>
> The proper solution is to look into case.scf1 and case.scf2
>
> In case.scf2 you can see that the intended EF is:
>
> :FER : F E R M
No.
The proper solution is to look into case.scf1 and case.scf2
In case.scf2 you can see that the intended EF is:
:FER : F E R M I - ENERGY(TETRAH.M.)= -0.2835569073
Edit case.in1 and put instead of 0.3 --> -0.2 for EF (1st line)
Then do another run_lapw and it converges quickly.
Am 2
Unfortunately the initial density and parameters are sometimes not very
good and leads to ghost bands in the 1st iteration. If you reduce the RMTs
to 2.0 (setrmt case -a O:2.0,Na:2.0 ; cp case*setrmt case.struct) and then
run there will be a warning in the 1st iteration, but for me it ran OK.
Alter
Dear Sir,
I have started to optimize the structural parameter with non-local
vanderwaals functionals. However when I edit case.in0 to put
XC_GGA_X_B86_R EC_LDA VC_LDA, lapw0 error appeared with LIBXC not
installed. I am using wien2k 19.1. Is it that we have to install the XC
functional se
Dear Prof. Laurence, Prof. Gerhard & Prof. Peter,
Thank you so much for your valuable
advices. I will try all the options & will let you know the results within
a few days.
with best regards,
On Wed, 29 Jan 2020 at 00:59, Peter Blaha
wrote:
> First of all, the
First of all, the VASP gaps of bulk SnS2 are 1.91 eV, not 2.1 as you are
writing ?
Then there are other VASP papers, which report different results, in
particular different structural parameters:
(Phys. Chem. Chem. Phys., 2016, 18, 318)
a=3.693 c=11.680 gap: 1.91
https://doi.org/10.
Just to explain the mechanism how a U for "semicore" 3d states (Sn, but
also Zn in ZnO or Ge, ...):
Usually, U shifts occupied d-states down by U/2 and unoccupied d-states
up by U/2. This is the mechanism eg. in NiO.
However, for these fully occupied states, the effect of U is much
smaller,
[wien-boun...@zeus.theochem.tuwien.ac.at] im Auftrag von Laurence
Marks [laurence.ma...@gmail.com]
Gesendet: Dienstag, 28. Januar 2020 16:45
An: A Mailing list for WIEN2k users
Betreff: Re: [Wien] Ghost band error in SnS2 continued
I suggest looking into:
a) Whether experimentally this is a direct or ind
An important point Peter pointed out to me some years ago.
Remember that electrons are dumb. They go where they want; we think about
them as s,p,d etc. I suspect that your Sn d-states should be semi-core
(-1.5Ryd) and largely uneffected by +U
What you (and perhaps your cited paper) are probably r
N.B., and, as I have said before, use runsp_c which is for spin-polarized
but no net magnetic moment; it is twice as fast as runsp and often more
stable.
On Tue, Jan 28, 2020 at 9:45 AM Laurence Marks
wrote:
> I suggest looking into:
>
> a) Whether experimentally this is a direct or indirect ban
I suggest looking into:
a) Whether experimentally this is a direct or indirect band-gap material.
If it is an indirect gap, what that value is. The number in case.scf2* is
the "minimum gap", not the sometimes larger direct gap. Trust experiment,
do not just follow what someone else did.
b) Look a
I want to simulate Li intercalation voltage in SnS2. But for that I need to
simulate pristine SnS2 properly
On Tue, Jan 28, 2020, 19:46 Laurence Marks wrote:
> Reducing the RMT by 30% is somewhat large.
>
> The key question is what are you trying to do? Reproducing a result with
> Vasp is no
Reducing the RMT by 30% is somewhat large.
The key question is what are you trying to do? Reproducing a result with
Vasp is not good science if that is your purpose. Without that information
I doubt that anyone can provide useful advice
_
Professor Laurence Marks
"Research is to see what ever
1) Use runsp_c.
2) First converge with U=7, then increase.
3) Reduce your RMTs by a further 5-10%.
4) Look at the PBE/mBJ DOS to see what is going wrong, don't just look at
the gap.
_
Professor Laurence Marks
"Research is to see what everybody else has seen, and to think what nobody
else has t
On Mon, 27 Jan 2020 at 16:23, Laurence Marks
wrote:
> Having a metal rather than an insulator is more of an issue than the gap.
>
> 1) Are you using runsp_c ? The system should not be magnetic, so it will
> be both faster and more stable.
>
No...I am using spin polarization
> 2) First conv
With GGA+U with U = 7eV & -1.56 Ry for Sn d in case.in1 we are not getting
any Ghostband error...however, with U=8 or 9 eV we are getting the
Ghostband error. Again, with U=7 eV we are getting 1.43 eV band gap which
is an underestimation with respect to earlier calculation in VASP...
On Mo
Having a metal rather than an insulator is more of an issue than the gap.
1) Are you using runsp_c ? The system should not be magnetic, so it will be
both faster and more stable.
2) First converge with PBE, then turn +U on. Maybe only use 5 eV first.
Test versus the U value.
3) Reduce your RMTs by
Again, in that case we are not getting accurate band gap. With GGA it is
metallic while the reported value is 2.2 eV with VASPI have also tried
mbj, however it gives 0.278 eV
On Mon, 27 Jan 2020 at 16:01, Laurence Marks
wrote:
> Remove the U, I suspect it is unphysical. Probably it is not ac
Remove the U, I suspect it is unphysical. Probably it is not acting on the
full Sn d semicore states but in fact on the tails of the sulphur.
_
Professor Laurence Marks
"Research is to see what everybody else has seen, and to think what nobody
else has thought", Albert Szent-Gyorgi
www.numis.n
But in that case also we are getting ghostband error...with GGA it is
running fine, while with GGA+U with U=9 eV for 4d of Snthe ghostband
error appears with -1.56 ry for Sn-d in case.in1...what to do?
On Mon, 27 Jan 2020 at 15:42, Laurence Marks
wrote:
> The ghostbands almost certainly aris
The ghostbands almost certainly arise because of the change you made to the
d-orbital linearization energy.
I strongly recommend that you stay with the default linearization energies
and don't adjust them yourself.
_
Professor Laurence Marks
"Research is to see what everybody else has seen, a
.ac.at] im Auftrag von Laurence
Marks [l-ma...@northwestern.edu]
Gesendet: Donnerstag, 13. Oktober 2016 16:43
An: A Mailing list for WIEN2k users
Betreff: Re: [Wien] Ghost band error
Did you distort the cell, or is this in the first iteration of the cubic cell
you attached.
To look at tetragonal d
Dear Shamik,
I am also facing the same issue.
I am trying to solve it.
I found some useful links:
http://www.mail-archive.com/wien%40zeus.theochem.tuwien.ac.at/msg10561.html
http://www.mail-archive.com/wien%40zeus.theochem.tuwien.ac.at/msg09836.html
http://www.wien2k.at/reg_user/faq/qtlb.html
Pl
Did you distort the cell, or is this in the first iteration of the cubic
cell you attached.
To look at tetragonal distortions you should probably first switch to the
tetragonal cell (make for instance c different, e.g. 10.862701) and accept
the change to the cell in init_lapw (I get I4/mmm).
On
Your structure is wrong. Such small RMTs are unphysical.
Mixed Ang and bohr ??
Am 27.07.2016 um 16:31 schrieb shamik chakrabarti:
Dear wien2k users,
I have tried to optimize the volume of TiO2
(anatase) by starting with -10% to 10% with a difference of 5% of the
initial v
Dear Tran & wien2k users,
After running almost 140 iterations when I
switch the mixing parameter from 0.05 (initial mixing param.) to 0.2, (as
the charge convergence was oscillating within 0.01 for almost 30 odd
iterations), ghost band appears at the same cycle (at
Thanks Tran. Its working.
On Fri, Apr 1, 2016 at 9:16 PM, wrote:
> Hi,
>
> you can try to start the mBJ calculation with the electron density
> from the GGA+U calculation (before executing runsp_lapw for mBJ,
> copy case.clm/sum/up/dn from the GGA+U calculations). Maybe the
> QTL-B value at
Hi,
you can try to start the mBJ calculation with the electron density
from the GGA+U calculation (before executing runsp_lapw for mBJ,
copy case.clm/sum/up/dn from the GGA+U calculations). Maybe the
QTL-B value at the 1st iteration is smaller such that the
calculation does not stop. This can hap
Dear Laurence,
I have converted a 14 atomic Fd3m cell to an 1x1x1
supercell with target lattice primitive. This has transformed the cell to a
56 atomic cell with all 56 atoms inequivalent. Rmt (Ti)=1.8, Rmt(Fe)=1.77,
Rmt (Li) = 1.54, Rmt (O)=1.53. I have also changed the MSR1 mixi
I have said this many times; reducing the mixing parameter (correctly
called the greed https://en.wikipedia.org/wiki/Greedy_algorithm) is not a
clever thing to do. While the general written and unwritten literature says
that this is an appropriate thing to do, it is WRONG and a complete
misundersta
48 matches
Mail list logo