[Wien] electron transfer - RMT spheres

[Hua Wu]

You may have to set an equal muffin-tin size, to say ~2 a.u. for the same 
element Co at inequivalent sites. When talking about charge transfer or 
charge ordering, it is better to look at the orbital occupation number rather 
than the total charge (within MT sphere), since the former can tell you the
formal valence state. Normally, covalency reduces the nominal charge
difference and thus leads to a smaller difference of the 1s core levels of the 
Co1 and Co2 than expected from a pure ionic model. But I guess the core
level energy difference should be still observable in the calculations.  

regards -- H. Wu  
 
On Thursday 27 August 2009 18:26, Gheorghe P wrote:
>      Dear Wien users,
>
>      I am studying a system which has a structural transition; above the
> structural transition all the Co ions are in equivalent positions having an
> octahedral environment with oxygen ions in the corners of the octahedra
> (distance Co-O = 1.95); below the structural transition the oxygen ions
> move such that there will be two inequivalent sites for Co after the
> distortion; Co1 will be inside a contracted octahedra (Co-O = 1.91) and Co2
> will be inside an expanded octahedra (Co-O = 2.02). From some measurements
> I expect that at the structural transition there is a electron transfer
> from the Co1 site to the Co2 site; to test this scenario I did an
> experiment using resonant X-ray scattering at the Co K edge; during the
> scattering an electron from the 1s level is moved into an empty 4p state
> (for a very short time); so in my case I will have (for both Co1 and Co2) a
> shift of the 1s level due to the electron transfer and in the same time I
> will have a shift of the 4p states due to the hybridization of the Co 4p
> states with O 2p states.
>
> I would like to calculate these shifts of the 1s and 4p level for both Co1
> and Co2 and to obtain a "quantitatively" value for the electron transfer
> using WIEN2k but i have a question about how to set up the RMT spheres
> around the Co ions during the calculations.
>
> I read so far in the emailing list (see below) that if I want to obtain a
> approximate value for the electron transfer I have to look at the charge in
> the RMT spheres.
>
> My question is about how to set the RMT sphere for the two ions Co1 and
> Co2?????
>
> I did a non-magnetic calculation using the w2web interface; if I use the
> default program from the w2web graphical interface, the calculated spheres
> for Co1 and Co2 are identical and there is no electron transfer between the
> two spheres; but in the reality if there is an electron transfer the ionic
> radius of Co1 and Co2 would be different.
>
> So how do I have to set up the sphere such that the calculations model as
> good as possible my system???
>
> Thank you very much in advance for any help you can give me to understand
> how to model this problem.
>
> Best wishes,
> Lucian
>
>
>
>
>
> ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
> Jorissen Kevin
>     Kevin.Jorissen at ua.ac.be
>
>
>     Tue Nov  4 13:43:42 CET 2003
>     Previous message: [Wien] Large no of K-points in the IBZ
> Next message: [Wien] Error in nn of a supercell
>  Messages sorted by:
>               [ date ]
>               [ thread ]
>               [ subject ]
>               [ author ]
>
>
> Some possibilities :
> * draw charge density maps and difference density maps
> * do Bader analysis
> * look at the charges in the MT
> Eg. in PRB67 075102 you'll find these three methods
>  used and compared (APW calculations of Cu2O).
>
> As there is no unambiguous definition of charge transfer,
> different methods will yield different quantitative values.
> But you can hope to discover trends.
>
> Kevin.++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++