Re: [Wien] cohesive energy

2016-05-13 Thread Zsolt Rak 
Thank you Prof. Blaha,

Your explanation makes sense. Of course an alternative method to calculate
formation and cohesive energies  would be to combine experimental data with
DFT energies, such as the method described in Phys. Rev. B 85, 115104
(2012). This way one would avoid calculations of metallic/gaseous phases
and of free atoms.

Thanks,
Zsolt

On Fri, May 13, 2016 at 2:22 PM, Peter Blaha 
wrote:

> I would answer it differently:
>
> No, it does not really make sense to use LDA+U for an atom;  but there are
> not many other chances to calculate formation or cohesive energies.
>
> For such quantities one MUST do all calculations with exactly the same
> calculational parameters and of course also with identical
> DFT-approximations.
>
> In all cases, where for one or the other system a certain approximation is
> not very good, the corresponding E-differences are of course very
> problematic.
>
> So if you compare a TM oxide with free TM and O atoms, all approximations
> must be the same.
>
> Of course for a TMO GGA+U is probably much better, but I would NOT say
> this for a free atom.
> You can test the sensitivity of your results be repeating the same
> E-difference using GGA (or even better a meta-GGA), i.e. you treat the atom
> more meaningful, but may even have a "metallic" TMO.
>
> The situation is similar for a formation energy of a TMO with respect to
> the metallic element and O2. GGA+U for bcc Fe with a sizable U gives
> magnetic moments up to 3 mB !
>
>
>
> Am 13.05.2016 um 19:42 schrieb Zsolt Rak:
>
>> Does it make sense to perform LDA+U/GGA+U calculation on free, isolated
>> atom?
>>
>> On Thu, May 12, 2016 at 11:47 AM, Laurence Marks
>> > wrote:
>>
>> To obtain reasonable results you have to be fully consistent with
>> everything, e.g. RMT, RKMAX, U's etc.
>>
>> On Thu, May 12, 2016 at 10:37 AM, Zsolt Rak > > wrote:
>>
>> ​​
>> Dear users,
>>
>> I am calculating the cohesive energy of a series of transition
>> metal (TM) oxides and I have the following questions: because
>> the energy of the TMO is calculated within the LDA+U framework
>> should I use the same LDA+U for the free atoms? Does it make
>> sense to subtract LDA+U/GGA+U energies of oxides from LDA/GGA
>> energies of free atoms? Does it make sense to use LDA+U/GGA+U
>> for free atom calculations?
>> Thanks,
>> Zsolt
>>
>>
>>
>>
>> --
>> 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.northwestern.edu
>>  ; Corrosion in 4D:
>> MURI4D.numis.northwestern.edu 
>> Partner of the CFW 100% program for gender
>> equity, www.cfw.org/100-percent 
>> Co-Editor, Acta Cryst A
>>
>> ___
>> Wien mailing list
>> Wien@zeus.theochem.tuwien.ac.at > 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
>>
>>
> --
> --
> 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
>
___
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] cohesive energy

2016-05-13 Thread Peter Blaha 

I would answer it differently:

No, it does not really make sense to use LDA+U for an atom;  but there 
are not many other chances to calculate formation or cohesive energies.


For such quantities one MUST do all calculations with exactly the same 
calculational parameters and of course also with identical 
DFT-approximations.


In all cases, where for one or the other system a certain approximation 
is not very good, the corresponding E-differences are of course very 
problematic.


So if you compare a TM oxide with free TM and O atoms, all 
approximations must be the same.


Of course for a TMO GGA+U is probably much better, but I would NOT say 
this for a free atom.
You can test the sensitivity of your results be repeating the same 
E-difference using GGA (or even better a meta-GGA), i.e. you treat the 
atom more meaningful, but may even have a "metallic" TMO.


The situation is similar for a formation energy of a TMO with respect to 
the metallic element and O2. GGA+U for bcc Fe with a sizable U gives 
magnetic moments up to 3 mB !




Am 13.05.2016 um 19:42 schrieb Zsolt Rak:

Does it make sense to perform LDA+U/GGA+U calculation on free, isolated
atom?

On Thu, May 12, 2016 at 11:47 AM, Laurence Marks
> wrote:

To obtain reasonable results you have to be fully consistent with
everything, e.g. RMT, RKMAX, U's etc.

On Thu, May 12, 2016 at 10:37 AM, Zsolt Rak > wrote:

​​
Dear users,

I am calculating the cohesive energy of a series of transition
metal (TM) oxides and I have the following questions: because
the energy of the TMO is calculated within the LDA+U framework
should I use the same LDA+U for the free atoms? Does it make
sense to subtract LDA+U/GGA+U energies of oxides from LDA/GGA
energies of free atoms? Does it make sense to use LDA+U/GGA+U
for free atom calculations?
Thanks,
Zsolt




--
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.northwestern.edu
 ; Corrosion in 4D:
MURI4D.numis.northwestern.edu 
Partner of the CFW 100% program for gender
equity, www.cfw.org/100-percent 
Co-Editor, Acta Cryst A

___
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



--
--
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

Re: [Wien] cohesive energy

2016-05-13 Thread Zsolt Rak 
Does it make sense to perform LDA+U/GGA+U calculation on free, isolated
atom?

On Thu, May 12, 2016 at 11:47 AM, Laurence Marks 
wrote:

> To obtain reasonable results you have to be fully consistent with
> everything, e.g. RMT, RKMAX, U's etc.
>
> On Thu, May 12, 2016 at 10:37 AM, Zsolt Rak  wrote:
>
>> ​​
>> Dear users,
>>
>> I am calculating the cohesive energy of a series of transition metal (TM)
>> oxides and I have the following questions: because the energy of the TMO is
>> calculated within the LDA+U framework should I use the same LDA+U for the
>> free atoms? Does it make sense to subtract LDA+U/GGA+U energies of oxides
>> from LDA/GGA energies of free atoms? Does it make sense to use LDA+U/GGA+U
>> for free atom calculations?
>> Thanks,
>> Zsolt
>>
>
>
>
> --
> 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.northwestern.edu ; Corrosion in 4D:
> MURI4D.numis.northwestern.edu
> Partner of the CFW 100% program for gender equity, www.cfw.org/100-percent
> Co-Editor, Acta Cryst A
>
> ___
> 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] cohesive energy

2016-05-13 Thread Laurence Marks 
Yes

On Fri, May 13, 2016 at 12:42 PM, Zsolt Rak  wrote:

> Does it make sense to perform LDA+U/GGA+U calculation on free, isolated
> atom?
>
> On Thu, May 12, 2016 at 11:47 AM, Laurence Marks  > wrote:
>
>> To obtain reasonable results you have to be fully consistent with
>> everything, e.g. RMT, RKMAX, U's etc.
>>
>> On Thu, May 12, 2016 at 10:37 AM, Zsolt Rak  wrote:
>>
>>> ​​
>>> Dear users,
>>>
>>> I am calculating the cohesive energy of a series of transition metal
>>> (TM) oxides and I have the following questions: because the energy of the
>>> TMO is calculated within the LDA+U framework should I use the same LDA+U
>>> for the free atoms? Does it make sense to subtract LDA+U/GGA+U energies of
>>> oxides from LDA/GGA energies of free atoms? Does it make sense to use
>>> LDA+U/GGA+U for free atom calculations?
>>> Thanks,
>>> Zsolt
>>>
>>
>>
>>
>> --
>> 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.northwestern.edu ; Corrosion in 4D:
>> MURI4D.numis.northwestern.edu
>> Partner of the CFW 100% program for gender equity,
>> www.cfw.org/100-percent
>> Co-Editor, Acta Cryst A
>>
>> ___
>> 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
>>
>>
>


-- 
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.northwestern.edu ; Corrosion in 4D: MURI4D.numis.northwestern.edu
Partner of the CFW 100% program for gender equity, www.cfw.org/100-percent
Co-Editor, Acta Cryst A
___
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] cohesive energy

2016-05-12 Thread Laurence Marks 
To obtain reasonable results you have to be fully consistent with
everything, e.g. RMT, RKMAX, U's etc.

On Thu, May 12, 2016 at 10:37 AM, Zsolt Rak  wrote:

> ​​
> Dear users,
>
> I am calculating the cohesive energy of a series of transition metal (TM)
> oxides and I have the following questions: because the energy of the TMO is
> calculated within the LDA+U framework should I use the same LDA+U for the
> free atoms? Does it make sense to subtract LDA+U/GGA+U energies of oxides
> from LDA/GGA energies of free atoms? Does it make sense to use LDA+U/GGA+U
> for free atom calculations?
> Thanks,
> Zsolt
>



-- 
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.northwestern.edu ; Corrosion in 4D: MURI4D.numis.northwestern.edu
Partner of the CFW 100% program for gender equity, www.cfw.org/100-percent
Co-Editor, Acta Cryst A
___
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] cohesive-energy

2015-08-17 Thread tran 

Hi,

Yes you can choose P instead of F. Beside this, it is important
to reduce the symmetry (e.g., from cubic to orthorhombic) for
atoms with an open shell (most free atoms are magnetic) in order to reach
the proper electronic ground state. I attached the struct file (P) that I
used for a free Na atom. Usually, a box of 12x12x12 angstrom is large
enough except for the heavy alkali and alkali-earth atoms.

For RMT and RKMAX, you have to test. It is important to choose the same
parameters for the bulk and free atom.

F. Tran

On Mon, 17 Aug 2015, Seyyed Amir Abbas Emami wrote:


​Dear users.

I am trying to calculate cohesive energy. As it is mentioned in FAQ, we must 
create FCC supercell and also use identical RKmax and RMT. Now i have two 
question?

1- Can i choose Cubic (P) supercell instead of FCC. In other words, is there 
any reason for that choosing?
2- what is the good or best value for RKmax or RMT. (actually i automatically 
chose RMT for one element equal to 2.5 and consequently RKmax also 2.5. Is this 
value valid?)


Thank you in advance.


Na 
P   LATTICE,NONEQUIV.ATOMS:  1 
MODE OF CALC=RELA unit=ang 
 22.676723 22.771209 22.865695 90.00 90.00 90.00   
ATOM  -1: X=0. Y=0. Z=0.
  MULT= 1  ISPLIT= 8
Na NPT=  781  R0=0.0001 RMT=2.   Z: 11.0   
LOCAL ROT MATRIX:1.000 0.000 0.000
 0.000 1.000 0.000
 0.000 0.000 1.000
   8  NUMBER OF SYMMETRY OPERATIONS
-1 0 0 0.
 0-1 0 0.
 0 0-1 0.
   1
-1 0 0 0.
 0-1 0 0.
 0 0 1 0.
   2
-1 0 0 0.
 0 1 0 0.
 0 0-1 0.
   3
-1 0 0 0.
 0 1 0 0.
 0 0 1 0.
   4
 1 0 0 0.
 0-1 0 0.
 0 0-1 0.
   5
 1 0 0 0.
 0-1 0 0.
 0 0 1 0.
   6
 1 0 0 0.
 0 1 0 0.
 0 0-1 0.
   7
 1 0 0 0.
 0 1 0 0.
 0 0 1 0.
   8
___
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] cohesive-energy

2015-08-17 Thread tran 

Orthorhombic means that the lattice constants a, b and c are not
equal. P means primitive, which is one of the lattice types
(see Table 4.4 in users's guide). The unit cell for Na has
orthorhombic symmetry and is of lattice type P.

The wave function of the electronic state with the lowest total
energy can eventually have a symmetry which is lower than cubic.
If this is the case, then you need to reduce the symmetry to
obtain this state. The use of an orthorhombic symmetry should
be enough in every case.

F. Tran

If the symmetry is too high (e.g., cubic)

On Mon, 17 Aug 2015, Seyyed Amir Abbas Emami wrote:


​Thank you dear Tran.

It was completely helpful.But what is the role of symmetry in calculating 
cohesive energy as you said reduce cubic to orthorhombic for open shell 
atoms.Also for
Na which is open sell, you use P instead of orthorhombic.


___
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] cohesive-energy

2015-08-17 Thread Seyyed Amir Abbas Emami 
​Thank you dear Tran.

It was completely helpful.But what is the role of symmetry in calculating 
cohesive energy as you said reduce cubic to orthorhombic for open shell 
atoms.Also for Na which is open sell, you use P instead of orthorhombic.

___
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