Re: [Wien] Diference in EFG values

2015-09-16 Thread pieper 

Hello Stefaan and Muhammad

From the Wien2k UG, chapter 7.1 on lapw0, I take it that Wien2k 
calculates the potential, and from that the EFG, from the TOTAL electron 
density. For lapw0 explicitely including interstitials, for the 
decomposition in lapw2 (chapter 7.7) explicitely only the electron 
density in the atomic sphere. So as a reminder to Muhammad there may 
appear (minor) differences within Wien2k between the two methods.


But in my understanding both calculations in Wien2k always include core, 
semi-core, and valence electrons (within the atomic sphere)? The 
difference to VASP can then occure because that code only considers core 
states for the EFG?


Best regards,

Martin


---
Dr. Martin Pieper
Karl-Franzens University
Institute of Physics
Universitätsplatz 5
A-8010 Graz
Austria
Tel.: +43-(0)316-380-8564


Am 15.09.2015 13:55, schrieb Stefaan Cottenier:
According to my understanding, core-state with tag F are valence 
states.

AM I right? If yes then why it is taking 3P states as valence states?
The valence electrons for Co are 3d7, 4s2.


You use the words 'valence electrons' in the traditional 'chemical'
way as states outside the noble gas core. In the context of the DFT,
the term 'valence electrons' refers to all states that are not forced
to behave as in a free atom (i.e. they allow to feel the presence of
the surrounding crystal).

You cannot avoid to use Co-3p as valence states with wien2k, the
numerics would break down if you didn't.

Your output shows that wien2k has Co-3s and Co-3p as valence states,
whereas vasp considers these as core states (i.e. they will not
contribute to the EFG in vasp). Co-3s will never contribute to the EFG
(spherical), whereas Co-3p could (albeit not very likely).

First consider the suggestion by Peter Blaha: make sure you have
identical XC-functionals in both codes, and inspect whether all
magnetic moments and the DOS and band structure pictures agree for
both codes.

If that is all right, then use the procedure to which I refered in my
previous post to isolate the EFG contribution of Co-3p. If that is
nearly zero, I don't know what is going on. If it would be equal to
the vasp-wien2k difference, then you have found the origin of your
problem.

Stefaan

___
Wien mailing list
Wien@zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
SEARCH the MAILING-LIST at:
http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html

___
Wien mailing list
Wien@zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
SEARCH the MAILING-LIST at:  
http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html

Re: [Wien] Diference in EFG values

2015-09-16 Thread Stefaan Cottenier 



 From the Wien2k UG, chapter 7.1 on lapw0, I take it that Wien2k
calculates the potential, and from that the EFG, from the TOTAL electron
density. For lapw0 explicitely including interstitials, for the
decomposition in lapw2 (chapter 7.7) explicitely only the electron
density in the atomic sphere. So as a reminder to Muhammad there may
appear (minor) differences within Wien2k between the two methods.


Exactly. This is the difference between the :EFGxxx (total) and :VZZxxx 
(in sphere) labels.



But in my understanding both calculations in Wien2k always include core,
semi-core, and valence electrons (within the atomic sphere)?


Yes, that's what I think too.


The difference to VASP can then occure because that code only considers core
states for the EFG?


The opposite. VASP uses only valence states for the EFG (I think). If 
there would be a sizeable 3p contribution for Co, then VASP would miss 
it. We observed such sizeable 5p core contributions to the EFG in 
lanthanides, where the core anisotropy was induced by the 4f anisotropy 
(http://dx.doi.org/10.1103/PhysRevB.77.155101, Table II). No idea 
whether a related mechanism could be at work for Co in this crystal (I 
doubt), but if all obvious other explanations fail it is worth trying 
(@Muhammad : there is not much more to explain, take the pdf quoted 
before, understand the example on page 9, and apply it to your own case).


Stefaan
___
Wien mailing list
Wien@zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
SEARCH the MAILING-LIST at:  
http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html

Re: [Wien] Diference in EFG values

2015-09-15 Thread Muhammad Sajjad 
Dear Prof. Stefaan and Blaha
Thank you for your reply.
Actually I have to reduce separation energy to -8.5 Ry because core leakage
issue was appearing with Ge.

If I focus on Co, the useful information is:

*Magnetic moment *

 VASPWIEN2K

Co1 2.530:MMI001:  = 2.35929

Co2 -2.530   :MMI002:  = -2.35900

*Core states *

VASP
   WIEN2K

VRHFIN = Co: d8 s1
E-up(Ry) E-dn(Ry) Occupancy q/sphere
   core-state


1S -557.761897  -557.761629   1.00  1.00
1.T


2S -66.048914 -65.9348801.00  1.00
  1.T


2P* -57.314278-57.2287991.00  1.00
  1.T


2P  -56.225866 -56.1379022.00  2.00
  1.T


3S  -7.605100   -7.363843  1.00  1.00
0.9968 F


3P* -5.003328   -4.767091  1.00  1.00
0.9917 F


3P  -4.866118   -4.631354   2.00  2.00
  0.9909F


3D* -0.699000  -0.490964   2.00  2.00
 0.8948F

 3D
  -0.684578  -0.477338   3.00  0.00
   0.9053F

 4S
  -0.423921  -0.381223   1.00  1.00
0.1768F


If I use -6 Ry as the separation energy then core states are


E-up(Ry) E-dn(Ry) Occupancy q/sphere
   core-state


1S -557.761897  -557.761629   1.00  1.00
1.T


2S -66.048914 -65.9348801.00  1.00
  1.T


2P* -57.314278-57.2287991.00  1.00
  1.T


2P  -56.225866 -56.1379022.00  2.00
  1.T


3S  -7.605100   -7.363843  1.00  1.00
0.9968 T


3P* -5.003328   -4.767091  1.00  1.00
0.9917 F


3P  -4.866118   -4.631354   2.00  2.00
  0.9909F


3D* -0.699000  -0.490964   2.00  2.00
 0.8948F

 3D
  -0.684578  -0.477338   3.00  0.00
   0.9053F

 4S
  -0.423921  -0.381223   1.00  1.00
0.1768F

According to my understanding, core-state with tag F are valence states. AM
I right? If yes then why it is taking 3P states as valence states? The
valence electrons for Co are 3d7, 4s2.

Kind Regards

On Mon, Sep 14, 2015 at 5:16 PM, Peter Blaha 
wrote:

> And of course, make sure that the same DFT functionals are used (not
> comparing DFT+U and DFT alone. This could explain the differences on Co.
>
> PS: In addition, I'd make sure that the basic "electronic structure" is
> identical (magnetic moments, DOS, bands, ...)
>
> On 09/14/2015 04:07 PM, Stefaan Cottenier wrote:
>
>>
>> First guess (assuming everything is numerically converged): do you
>> consider the same electrons as valence electrons both in vasp and wien2k
>> ? It could happen that a 'semicore' state is taken as valence in wien2k
>> and core in vasp. As long as the EFG contribution of these states are
>> small, wien2k and vasp will show no differences. But if you hit a
>> crystal structure where these states have a larger EFG contribution,
>> then the differences will show up.
>>
>> You can find out whether or not this is the case by (1) explicitly
>> taking the same core/valence assignment in both codes, or (2) examining
>> the contribution to the EFG of all different oribitals (or regions of
>> the DOS). See http://www.wien2k.at/reg_user/faq/efg2.pdf (top of page 9)
>> for the procedure to follow in wien2k.
>>
>> No warranty, just a guess...
>>
>> Stefaan
>>
>>
>> Op 14/09/2015 om 15:53 schreef Muhammad Sajjad:
>>
>>> Dear Users
>>>
>>> I run some test calculations for EFG values for different compounds
>>> (Sc2O3, In, SmCo5). Their EFG values (computed with VASP) are in
>>> agreement with that of previous values. Then I computed the EFG values
>>> with WIEN2K and are in strong agreement with previous as well as VASP
>>> values.
>>>
>>>
>>> But the EFG values for anti-ferromagnetic Ba2CoGe2O7 do not agree with
>>> that of VASP values (no previous study

Re: [Wien] Diference in EFG values

2015-09-15 Thread Stefaan Cottenier 

According to my understanding, core-state with tag F are valence states.
AM I right? If yes then why it is taking 3P states as valence states?
The valence electrons for Co are 3d7, 4s2.


You use the words 'valence electrons' in the traditional 'chemical' way 
as states outside the noble gas core. In the context of the DFT, the 
term 'valence electrons' refers to all states that are not forced to 
behave as in a free atom (i.e. they allow to feel the presence of the 
surrounding crystal).


You cannot avoid to use Co-3p as valence states with wien2k, the 
numerics would break down if you didn't.


Your output shows that wien2k has Co-3s and Co-3p as valence states, 
whereas vasp considers these as core states (i.e. they will not 
contribute to the EFG in vasp). Co-3s will never contribute to the EFG 
(spherical), whereas Co-3p could (albeit not very likely).


First consider the suggestion by Peter Blaha: make sure you have 
identical XC-functionals in both codes, and inspect whether all magnetic 
moments and the DOS and band structure pictures agree for both codes.


If that is all right, then use the procedure to which I refered in my 
previous post to isolate the EFG contribution of Co-3p. If that is 
nearly zero, I don't know what is going on. If it would be equal to the 
vasp-wien2k difference, then you have found the origin of your problem.


Stefaan

___
Wien mailing list
Wien@zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
SEARCH the MAILING-LIST at:  
http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html

Re: [Wien] Diference in EFG values

2015-09-15 Thread Muhammad Sajjad 
Right
I forgot to mention about XC potentials. Identical Potentials were used. I
already read the document related to computing EFG by WIEN2K. How to  isolate
the EFG contribution of Co-3p?


On Tue, Sep 15, 2015 at 2:55 PM, Stefaan Cottenier <
stefaan.cotten...@ugent.be> wrote:

> According to my understanding, core-state with tag F are valence states.
>> AM I right? If yes then why it is taking 3P states as valence states?
>> The valence electrons for Co are 3d7, 4s2.
>>
>
> You use the words 'valence electrons' in the traditional 'chemical' way as
> states outside the noble gas core. In the context of the DFT, the term
> 'valence electrons' refers to all states that are not forced to behave as
> in a free atom (i.e. they allow to feel the presence of the surrounding
> crystal).
>
> You cannot avoid to use Co-3p as valence states with wien2k, the numerics
> would break down if you didn't.
>
> Your output shows that wien2k has Co-3s and Co-3p as valence states,
> whereas vasp considers these as core states (i.e. they will not contribute
> to the EFG in vasp). Co-3s will never contribute to the EFG (spherical),
> whereas Co-3p could (albeit not very likely).
>
> First consider the suggestion by Peter Blaha: make sure you have identical
> XC-functionals in both codes, and inspect whether all magnetic moments and
> the DOS and band structure pictures agree for both codes.
>
> If that is all right, then use the procedure to which I refered in my
> previous post to isolate the EFG contribution of Co-3p. If that is nearly
> zero, I don't know what is going on. If it would be equal to the
> vasp-wien2k difference, then you have found the origin of your problem.
>
> Stefaan
>
>
> ___
> Wien mailing list
> Wien@zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> SEARCH the MAILING-LIST at:
> http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
>



-- 
Kind Regards
Muhammad Sajjad
Post Doctoral Fellow
KAUST, KSA.
___
Wien mailing list
Wien@zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
SEARCH the MAILING-LIST at:  
http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html

Re: [Wien] Diference in EFG values

2015-09-14 Thread Stefaan Cottenier 


First guess (assuming everything is numerically converged): do you 
consider the same electrons as valence electrons both in vasp and wien2k 
? It could happen that a 'semicore' state is taken as valence in wien2k 
and core in vasp. As long as the EFG contribution of these states are 
small, wien2k and vasp will show no differences. But if you hit a 
crystal structure where these states have a larger EFG contribution, 
then the differences will show up.


You can find out whether or not this is the case by (1) explicitly 
taking the same core/valence assignment in both codes, or (2) examining 
the contribution to the EFG of all different oribitals (or regions of 
the DOS). See http://www.wien2k.at/reg_user/faq/efg2.pdf (top of page 9) 
for the procedure to follow in wien2k.


No warranty, just a guess...

Stefaan


Op 14/09/2015 om 15:53 schreef Muhammad Sajjad:

Dear Users

I run some test calculations for EFG values for different compounds 
(Sc2O3, In, SmCo5). Their EFG values (computed with VASP) are in 
agreement with that of previous values. Then I computed the EFG values 
with WIEN2K and are in strong agreement with previous as well as VASP 
values.



But the EFG values for anti-ferromagnetic Ba2CoGe2O7 do not agree with 
that of VASP values (no previous study available). Actually for Ge and 
O they agree *but for Co and Ba WIEN2K computed values are almost 
double*. Could you please guide me where the problem is? I am drawing 
a table containing the EFG values and also attaching the structure file.


WIEN2K calculated values



VASP calculated values

 V_zz (V/m^2 )

Co1 13.22

Co2 13.20

Ba  -15.02

Ge   7.49

O1   9.32

O2   9.93

O3   9.42



  V_zz (V/m^2 )

Co1 5.97

Co2 5.97

Ba  -8.55

Ge   7.54

O1   9.73

O2 10.35

O3 9.00




Kind Regards
Muhammad Sajjad
Post Doctoral Fellow
KAUST, KSA.


___
Wien mailing list
Wien@zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
SEARCH the MAILING-LIST at:  
http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html


--
Stefaan Cottenier
Center for Molecular Modeling (CMM) &
Department of Materials Science and Engineering (DMSE)
Ghent University
Technologiepark 903
BE-9052 Zwijnaarde
Belgium

http://molmod.ugent.be
http://www.ugent.be/ea/dmse/en
email: stefaan.cotten...@ugent.be

my conference talks on Youtube: http://goo.gl/P2b1Hs

___
Wien mailing list
Wien@zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
SEARCH the MAILING-LIST at:  
http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html

Re: [Wien] Diference in EFG values

2015-09-14 Thread Peter Blaha 
And of course, make sure that the same DFT functionals are used (not 
comparing DFT+U and DFT alone. This could explain the differences on Co.


PS: In addition, I'd make sure that the basic "electronic structure" is 
identical (magnetic moments, DOS, bands, ...)


On 09/14/2015 04:07 PM, Stefaan Cottenier wrote:


First guess (assuming everything is numerically converged): do you
consider the same electrons as valence electrons both in vasp and wien2k
? It could happen that a 'semicore' state is taken as valence in wien2k
and core in vasp. As long as the EFG contribution of these states are
small, wien2k and vasp will show no differences. But if you hit a
crystal structure where these states have a larger EFG contribution,
then the differences will show up.

You can find out whether or not this is the case by (1) explicitly
taking the same core/valence assignment in both codes, or (2) examining
the contribution to the EFG of all different oribitals (or regions of
the DOS). See http://www.wien2k.at/reg_user/faq/efg2.pdf (top of page 9)
for the procedure to follow in wien2k.

No warranty, just a guess...

Stefaan


Op 14/09/2015 om 15:53 schreef Muhammad Sajjad:

Dear Users

I run some test calculations for EFG values for different compounds
(Sc2O3, In, SmCo5). Their EFG values (computed with VASP) are in
agreement with that of previous values. Then I computed the EFG values
with WIEN2K and are in strong agreement with previous as well as VASP
values.


But the EFG values for anti-ferromagnetic Ba2CoGe2O7 do not agree with
that of VASP values (no previous study available). Actually for Ge and
O they agree *but for Co and Ba WIEN2K computed values are almost
double*. Could you please guide me where the problem is? I am drawing
a table containing the EFG values and also attaching the structure file.

WIEN2K calculated values



VASP calculated values

 V_zz (V/m^2 )

Co1 13.22

Co2 13.20

Ba  -15.02

Ge   7.49

O1   9.32

O2   9.93

O3   9.42



  V_zz (V/m^2 )

Co1 5.97

Co2 5.97

Ba  -8.55

Ge   7.54

O1   9.73

O2 10.35

O3 9.00




Kind Regards
Muhammad Sajjad
Post Doctoral Fellow
KAUST, KSA.


___
Wien mailing list
Wien@zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
SEARCH the MAILING-LIST 
at:http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html


--
Stefaan Cottenier
Center for Molecular Modeling (CMM) &
Department of Materials Science and Engineering (DMSE)
Ghent University
Technologiepark 903
BE-9052 Zwijnaarde
Belgium

http://molmod.ugent.be
http://www.ugent.be/ea/dmse/en
email:stefaan.cotten...@ugent.be

my conference talks on Youtube:http://goo.gl/P2b1Hs



___
Wien mailing list
Wien@zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
SEARCH the MAILING-LIST at:  
http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html



--

  P.Blaha
--
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-165300 FAX: +43-1-58801-165982
Email: bl...@theochem.tuwien.ac.atWIEN2k: http://www.wien2k.at
WWW:   http://www.imc.tuwien.ac.at/staff/tc_group_e.php
--
___
Wien mailing list
Wien@zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
SEARCH the MAILING-LIST at:  
http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html