[Wien] chemical bonding

[Rocquefelte]

Just to add one more comment on this topic. Bond-valence analysis is 
very powerful and will also works for other compounds (beyond oxides).
It will be more tricky if you have hydrogen-bonds, but also in this case 
there is possibilities to extract the valence of the metals and ligands 
form their coordination shell.

About Fe2+ and Fe3+, one other simple method could be used to extract 
the "FORMAL CHARGE" which is not REAL, because it is based on a 
democratic (covalent) distribution of the bonding electrons between the 
different partners. You can look at the partial DOS and see where the 
Fermi energy cross the 3d bands:

- if 6 electrons are in the 3d bands --> Fe2+
- if 5 electrons are in the 3d bands --> Fe3+

The charge in the RMT of iron differ by small numbers although if the 
DOS clearly shows a difference between Fe2+ and Fe3+. It is the reason 
why the Bader analysis is better to have access to the "REAL" charge of 
the atoms, however as mentioned by Laurence these charges will be 
different from the "FORMAL" values. One iron will tend to be 2+ and the 
other one will tend to be 3+ (just a tendency) ...

But definitely the bond-valence analysis is a very easy method which 
allows to extract the FORMAL CHARGE based on the geometry of the ligands 
around the iron metal.

Regards

Xavier


On 03/06/2012 05:58 AM, Laurence Marks wrote:
> Charge can be estimated using (as a conventional measure) the Bader
> volume around an atom, for instance by using the aim command in
> Wien2k.
>
> However, this is not the same as "Fe2+" which has a strict meaning
> that the iron atom is in a 2+ valence state. This is not the same (but
> alas often confused with) the effective charge, which I prefer to
> write as +X, as against valence as X+. You can blame Linus Pauling for
> this as he used the word "charge" in his original papers, when
> referring to what is really valence.
>
> As an example, strontium titanate has a valence of Sr2+ Ti4+ (O2-)3.
> However, the "charges" from a Bader analysis are closer to Sr+1.6
> Ti+2.2 (O-1.2)3 . (The numbers are right, there is a slight rounding
> error in the charges).
>
> In the limit of completely ionic bonding the two numbers are the same.
> However, in general they are not and a source of endless confusion.
> Depending upon exactly what functional you use the charges change, for
> instance PBE tends to be too covalent and hybrids and/or LDA+U
> improves on this (for d-electrons in Wien2k) making oxides more ionic.
>
> For an oxide, you can readily get the valence from a bond-valence
> analysis. I am not sure about other cases.
>
> And, just to make life more interesting, many spectroscopies (e.g.
> XPS, EELS) have been calibrated versus valence.
>
> 2012/3/5 Yundi Quan<quan at ms.physics.ucdavis.edu>:
>> Dear Professor Laurence Marks,
>> Charge ordering has long been a generator of controversies. I'm wondering
>> whether charge as measure from DFT can be used to identify charge ordering.
>>
>> Thanks
>> Yundi
>>
>>
>> 2012/3/5 Laurence Marks<L-marks at northwestern.edu>
>>> Charge, as measured from DFT, and valence (chemical bonding) are very,
>>> very different.
>>>
>>> Without more information as to what you are asking, that is about the only
>>> answer available.
>>>
>>> ---------------------------
>>> Professor Laurence Marks
>>> Department of Materials Science and Engineering
>>> Northwestern University
>>> www.numis.northwestern.edu 1-847-491-3996
>>> "Research is to see what everybody else has seen, and to think what nobody
>>> else has thought"
>>> Albert Szent-Gyorgi
>>>
>>> On Mar 5, 2012 7:11 PM, "Yundi Quan"<quanyundi at gmail.com>  wrote:
>>>> Does LAPW method always underestimate the chemical bonding? For example,
>>>> Fe2+ and Fe3+ have the same charge around them.
>>>> _______________________________________________
>>>> Wien mailing list
>>>> Wien at zeus.theochem.tuwien.ac.at
>>>> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>>>>
>>> _______________________________________________
>>> Wien mailing list
>>> Wien at zeus.theochem.tuwien.ac.at
>>> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>>>
>>
>> _______________________________________________
>> Wien mailing list
>> Wien at zeus.theochem.tuwien.ac.at
>> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>>
>
>