Re: [Wien] Problem in Bandstructure plot

[Dr. K. C. Bhamu]

Hii,
Did you shift the the bands with Fermi?

Use case.insp and put FERMI in this file at  and run lapw2 and
spaghetti again.

Regards
Bhamu



On Sat, Nov 23, 2019, 10:46 Peeyush kumar kamlesh <
peeyush.physik@gmail.com> wrote:

> Dear Users,
> Greetings!
> Generally when we plot bandstructure in WIEN2k, it shows fermi level (Ef =
> 0) just above the valance band. But when i am plotting it for my compound
> then fermi level is crossing the valance band. Can you please explain the
> problem and suggest a solution?
>
> Regards
> Peeyush Kumar Kamlesh
> ___
> 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

[Wien] Problem in Bandstructure plot

[Peeyush kumar kamlesh]

Dear Users,
Greetings!
Generally when we plot bandstructure in WIEN2k, it shows fermi level (Ef =
0) just above the valance band. But when i am plotting it for my compound
then fermi level is crossing the valance band. Can you please explain the
problem and suggest a solution?

Regards
Peeyush Kumar Kamlesh
___
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] Beween non spin polarized and spin polarized calculations

[djamel slamnia]

 thanks for every things i would like to use PBE0 on-site hybrid can you 
explain it to me please step by step and i will be gratful 

Le vendredi 22 novembre 2019 à 14:50:45 UTC+1, Xavier Rocquefelte 
 a écrit :  
 
  
I am working on similar compounds but not with Mn. 
 
 
I would recommend to use PBE+U or PBE0 on-site hybrid. 
 
 
In the first case (PBE+U) you must define an appropriate value for Ueff of 
Mn-3d states. 
 
 
In the second case, use alpha = 0.25 for only M-3d states. 
 
 
It should be a not to bad approximation. However, it can be tricky because of 
the metallic nature of the system. 
 
 
Thus, the best would be to have experimental data to compare. 
 
 
Regards
 
 Xavier
 

 
 Le 22/11/2019 à 11:52, djamel slamnia a écrit :
  
 
 the chemical composition are Cu2MnZrSe4    Cu2MnZrS4   
   
 ___
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] Beween non spin polarized and spin polarized calculations

[Xavier Rocquefelte]

I am working on similar compounds but not with Mn.

I would recommend to use PBE+U or PBE0 on-site hybrid.

In the first case (PBE+U) you must define an appropriate value for Ueff 
of Mn-3d states.


In the second case, use alpha = 0.25 for only M-3d states.

It should be a not to bad approximation. However, it can be tricky 
because of the metallic nature of the system.


Thus, the best would be to have experimental data to compare.

Regards

Xavier


Le 22/11/2019 à 11:52, djamel slamnia a écrit :

the chemical composition are Cu2MnZrSe4    Cu2MnZrS4



___
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] Beween non spin polarized and spin polarized calculations

[djamel slamnia]

 the chemical composition are Cu2MnZrSe4    Cu2MnZrS4  
Le vendredi 22 novembre 2019 à 09:33:40 UTC+1, Xavier Rocquefelte 
 a écrit :  
 
  
OK so you have transition metal elements surrounded by ligands (S, Se) ...
 
For copper, it depends if its oxidation state is +2 or +1, leading to 10 and 9 
electrons in the Cu-3d states, respectively. 
 
 
For Manganese you have to use DFT+U or DFT+onsite hybrid functional or full 
hybrid if not metallic. 
 
 
For zirconium, you should have no difficulties. 
 
 
I cannot say more  without the chemical composition ... 
 
 Le 22/11/2019 à 09:13, djamel slamnia a écrit :
  
 
 thanks Mr for replay  
  my compounds one of them is metalic and the second is half metalic. both 
compounds containing d-Orbitals of transition metal (Cu ,Mn , Zr , S , Se) i 
thought that -hf module can be the best methode  
  What do you suggest  ??    thanks in advance  
  Le vendredi 22 novembre 2019 à 08:00:02 UTC+1, Peter Blaha 
 a écrit :  
  
   Please: Do all the examples from our wien2k-workshop (from 
 www.wien2k.at) first !  Learn how to run WIEN2k properly on small examples.
 
 I guess you do all wrong. A wrong command (even one soingle letter) can 
 spoil the calculations.
 
 a) If your compound is metallic, it is very doubtful that a hybrid 
 calculation is good. Hybrid-DFT is not good for metals.
 
 b) runsp_lapw -hf -ec -f 1 0.0001 -cc 0.001 -NI
 
 This command has several errors. What should that be:
 -ec -f 1 0.0001    
 And I doubt that a non-parallel (without -p) calculation runs 
 meaningfully for 200 cycles with -hf. This would run VERY long, unless 
 your k-mesh is very small.
 
 c) You said you did geometry optimization "successfully" ! ?
 
 did you use    runsp -min -fc 1 ...    for this purpose ???
 
 Did you check that the forces were small at the end in your case.scf 
 file ???
 
 
 Am 21.11.2019 um 20:51 schrieb djamel slamnia:
 > I am very grateful to communicate with you
 > 
 > my compound is quaternary all atoms are d-Orbitals of  transition metal
 > 
 > i did the geometry optimisation successfuly
 > Now i'm trying to do minimisation of position with hf
 > after i intialisation init_hf_lapw  i filled nband i did this  ( 
 > runsp_lapw -hf -ec -f 1 0.0001 -cc 0.001 -NI)
 > but with more then 200 iterations never get convegence ???
 > 
 > Le jeudi 21 novembre 2019 à 18:27:12 UTC+1, Xavier Rocquefelte 
 >  a écrit :
 > 
 > 
 > If you want our help we need to have details about your system.
 > 
 > Then we can advice you concerning the best strategy.
 > 
 > Best Regards
 > 
 > Xavier
 > 
 > Le 21/11/2019 à 18:02, djamel slamnia a écrit :
 > thanks a lot Mr XAVIER for
 >   i'm traying hard to do option 2.a minimisation position with hybrid  
 > functionel but the process did not converged after more than 200 
 > iteration ??
 > can you Suggests me solutio
 > Le jeudi 21 novembre 2019 à 16:51:45 UTC+1, Xavier Rocquefelte 
 >  
 >  a écrit :
 > 
 > 
 > Hummm ... Here are many options ... choose the one you like ;)
 > 
 > 1/ You consider the importance to have hybrid functional to properly 
 > describe the atomic structure. If not, optimize using regular DFT and 
 > then estimate the property of interest with hybrid functional using to 
 > optimized structure in regular DFT.
 > 
 > 2/ You really want to optimize the geometry in hybrid because your 
 > system requires such a level of accuracy:
 > 
 > 2.a. you implement the optimization using hybrid functional in WIEN2k 
 > and all the community will be happy of such a contribution
 > 
 > 2.b. you use a plane-wave code (the one you like) and you optimize in 
 > hybrid. Then you use this optimized structure in WIEN2k to simulate a 
 > specific property or to do an accurate analysis.
 > 
 > Regards,
 > 
 > Xavier
 > 
 > Le 21/11/2019 à 16:41, djamel slamnia a écrit :
 > so what  i have to do now ??? thanks in advance
 > 
 > Le mercredi 20 novembre 2019 à 15:42:59 UTC+1, Tran, Fabien 
 >   a écrit :
 > 
 > 
 > If you are trying to do an optimization of atom positions, then you can 
 > not do it with hybrid functional (-hf option) because the forces are not 
 > implemented for hybrid functionals.
 > 
 > 
 > 
 > *From:* Wien  
 >  on behalf of djamel 
 > slamnia  
 > *Sent:* Wednesday, November 20, 2019 2:41 PM
 > *To:* A Mailing list for WIEN2k users
 > *Subject:* Re: [Wien] Beween non spin polarized and spin polarized 
 > calculations
 > i'm  using hf hybrid functional i tryed to do minimisation MSR1 but i 
 > cant get convergence
 > what is the solution in your opinion ??   thank in advance
 > 
 > Le dimanche 17 novembre 2019 à 18:36:12 UTC+1, Abderrahmane Reggad 
 >   a écrit :
 > 
 > 
 > Thanks xavier for the interesting article and useful information
 > 
 > My ne

Re: [Wien] Beween non spin polarized and spin polarized calculations

[Xavier Rocquefelte]

OK so you have transition metal elements surrounded by ligands (S, Se) ...

For copper, it depends if its oxidation state is +2 or +1, leading to 10 
and 9 electrons in the Cu-3d states, respectively.


For Manganese you have to use DFT+U or DFT+onsite hybrid functional or 
full hybrid if not metallic.


For zirconium, you should have no difficulties.

I cannot say more  without the chemical composition ...

Le 22/11/2019 à 09:13, djamel slamnia a écrit :

thanks Mr for replay

my compounds one of them is metalic and the second is half metalic. 
both compounds containing d-Orbitals of transition metal (Cu ,Mn , Zr 
, S , Se)

i thought that -hf module can be the best methode

What do you suggest  ??
thanks in advance

Le vendredi 22 novembre 2019 à 08:00:02 UTC+1, Peter Blaha 
 a écrit :



Please: Do all the examples from our wien2k-workshop (from
www.wien2k.at) first !  Learn how to run WIEN2k properly on small 
examples.


I guess you do all wrong. A wrong command (even one soingle letter) can
spoil the calculations.

a) If your compound is metallic, it is very doubtful that a hybrid
calculation is good. Hybrid-DFT is not good for metals.

b) runsp_lapw -hf -ec -f 1 0.0001 -cc 0.001 -NI

This command has several errors. What should that be:
-ec -f 1 0.0001    
And I doubt that a non-parallel (without -p) calculation runs
meaningfully for 200 cycles with -hf. This would run VERY long, unless
your k-mesh is very small.

c) You said you did geometry optimization "successfully" ! ?

did you use    runsp -min -fc 1 ...    for this purpose ???

Did you check that the forces were small at the end in your case.scf
file ???


Am 21.11.2019 um 20:51 schrieb djamel slamnia:
> I am very grateful to communicate with you
>
> my compound is quaternary all atoms are d-Orbitals of  transition metal
>
> i did the geometry optimisation successfuly
> Now i'm trying to do minimisation of position with hf
> after i intialisation init_hf_lapw  i filled nband i did this  (
> runsp_lapw -hf -ec -f 1 0.0001 -cc 0.001 -NI)
> but with more then 200 iterations never get convegence ???
>
> Le jeudi 21 novembre 2019 à 18:27:12 UTC+1, Xavier Rocquefelte
> > a écrit :

>
>
> If you want our help we need to have details about your system.
>
> Then we can advice you concerning the best strategy.
>
> Best Regards
>
> Xavier
>
> Le 21/11/2019 à 18:02, djamel slamnia a écrit :
> thanks a lot Mr XAVIER for
>   i'm traying hard to do option 2.a minimisation position with hybrid
> functionel but the process did not converged after more than 200
> iteration ??
> can you Suggests me solutio
> Le jeudi 21 novembre 2019 à 16:51:45 UTC+1, Xavier Rocquefelte
> >
> > a écrit :

>
>
> Hummm ... Here are many options ... choose the one you like ;)
>
> 1/ You consider the importance to have hybrid functional to properly
> describe the atomic structure. If not, optimize using regular DFT and
> then estimate the property of interest with hybrid functional using to
> optimized structure in regular DFT.
>
> 2/ You really want to optimize the geometry in hybrid because your
> system requires such a level of accuracy:
>
> 2.a. you implement the optimization using hybrid functional in WIEN2k
> and all the community will be happy of such a contribution
>
> 2.b. you use a plane-wave code (the one you like) and you optimize in
> hybrid. Then you use this optimized structure in WIEN2k to simulate a
> specific property or to do an accurate analysis.
>
> Regards,
>
> Xavier
>
> Le 21/11/2019 à 16:41, djamel slamnia a écrit :
> so what  i have to do now ??? thanks in advance
>
> Le mercredi 20 novembre 2019 à 15:42:59 UTC+1, Tran, Fabien
> mailto:fabien.t...@tuwien.ac.at>> 
> a 
écrit :

>
>
> If you are trying to do an optimization of atom positions, then you can
> not do it with hybrid functional (-hf option) because the forces are 
not

> implemented for hybrid functionals.
>
>
> 
> *From:* Wien >
> > on behalf of djamel
> slamnia mailto:djamel4...@yahoo.fr>> 
>

> *Sent:* Wednesday, November 20, 2019 2:41 PM
> *To:* A Mailing list for WIEN2k users
> *Subject:* Re: [Wien] Beween non spin polarized and spin polarized
> calculations
> i'm  using hf hybrid functional i tryed to do minimisation MSR1 but i
> cant get convergence
> what is the solution in your opinion ??   thank in advance
>
> Le dimanche 17 novembre 2019 à 18:36:12 UTC+1, Abderrahmane Reggad
> mailto:abde.reg...@gmail.com>> 
> a écrit :

Re: [Wien] Beween non spin polarized and spin polarized calculations

[djamel slamnia]

 thanks Mr for replay 
my compounds one of them is metalic and the second is half metalic. both 
compounds containing d-Orbitals of transition metal (Cu ,Mn , Zr , S , Se)i 
thought that -hf module can be the best methode 
What do you suggest  ??   thanks in advance 
Le vendredi 22 novembre 2019 à 08:00:02 UTC+1, Peter Blaha 
 a écrit :  
 
 Please: Do all the examples from our wien2k-workshop (from 
www.wien2k.at) first !  Learn how to run WIEN2k properly on small examples.

I guess you do all wrong. A wrong command (even one soingle letter) can 
spoil the calculations.

a) If your compound is metallic, it is very doubtful that a hybrid 
calculation is good. Hybrid-DFT is not good for metals.

b) runsp_lapw -hf -ec -f 1 0.0001 -cc 0.001 -NI

This command has several errors. What should that be:
-ec -f 1 0.0001    
And I doubt that a non-parallel (without -p) calculation runs 
meaningfully for 200 cycles with -hf. This would run VERY long, unless 
your k-mesh is very small.

c) You said you did geometry optimization "successfully" ! ?

did you use    runsp -min -fc 1 ...    for this purpose ???

Did you check that the forces were small at the end in your case.scf 
file ???


Am 21.11.2019 um 20:51 schrieb djamel slamnia:
> I am very grateful to communicate with you
> 
> my compound is quaternary all atoms are d-Orbitals of  transition metal
> 
> i did the geometry optimisation successfuly
> Now i'm trying to do minimisation of position with hf
> after i intialisation init_hf_lapw  i filled nband i did this  ( 
> runsp_lapw -hf -ec -f 1 0.0001 -cc 0.001 -NI)
> but with more then 200 iterations never get convegence ???
> 
> Le jeudi 21 novembre 2019 à 18:27:12 UTC+1, Xavier Rocquefelte 
>  a écrit :
> 
> 
> If you want our help we need to have details about your system.
> 
> Then we can advice you concerning the best strategy.
> 
> Best Regards
> 
> Xavier
> 
> Le 21/11/2019 à 18:02, djamel slamnia a écrit :
> thanks a lot Mr XAVIER for
>   i'm traying hard to do option 2.a minimisation position with hybrid  
> functionel but the process did not converged after more than 200 
> iteration ??
> can you Suggests me solutio
> Le jeudi 21 novembre 2019 à 16:51:45 UTC+1, Xavier Rocquefelte 
>  
>  a écrit :
> 
> 
> Hummm ... Here are many options ... choose the one you like ;)
> 
> 1/ You consider the importance to have hybrid functional to properly 
> describe the atomic structure. If not, optimize using regular DFT and 
> then estimate the property of interest with hybrid functional using to 
> optimized structure in regular DFT.
> 
> 2/ You really want to optimize the geometry in hybrid because your 
> system requires such a level of accuracy:
> 
> 2.a. you implement the optimization using hybrid functional in WIEN2k 
> and all the community will be happy of such a contribution
> 
> 2.b. you use a plane-wave code (the one you like) and you optimize in 
> hybrid. Then you use this optimized structure in WIEN2k to simulate a 
> specific property or to do an accurate analysis.
> 
> Regards,
> 
> Xavier
> 
> Le 21/11/2019 à 16:41, djamel slamnia a écrit :
> so what  i have to do now ??? thanks in advance
> 
> Le mercredi 20 novembre 2019 à 15:42:59 UTC+1, Tran, Fabien 
>   a écrit :
> 
> 
> If you are trying to do an optimization of atom positions, then you can 
> not do it with hybrid functional (-hf option) because the forces are not 
> implemented for hybrid functionals.
> 
> 
> 
> *From:* Wien  
>  on behalf of djamel 
> slamnia  
> *Sent:* Wednesday, November 20, 2019 2:41 PM
> *To:* A Mailing list for WIEN2k users
> *Subject:* Re: [Wien] Beween non spin polarized and spin polarized 
> calculations
> i'm  using hf hybrid functional i tryed to do minimisation MSR1 but i 
> cant get convergence
> what is the solution in your opinion ??   thank in advance
> 
> Le dimanche 17 novembre 2019 à 18:36:12 UTC+1, Abderrahmane Reggad 
>   a écrit :
> 
> 
> Thanks xavier for the interesting article and useful information
> 
> My new question is : what's the relationship between the unpaired 
> electrons and the orbital contribution. I think that the unpaired 
> electrons are always related to the spin contribution.
> 
> Best regards
> 
> -- 
> Dr. Abderrahmane Reggad
> Engineering Physics Laboratory
> Faculty of Material Sciences, Ibn Khaldoun University, Tiaret, 14000, 
> Algeria
> Tel: +213(0)561861963 - Algeria
> ___
> 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
> 
> _