[Wien] Qestion about DOS results

[Lyudmila Dobysheva]

11.12.2018 14:14, Fecher, Gerhard wrote:
> it means, if I change the structure and the type of atoms the result becomes different,
> but it means nothing if I don't know the structure and the elements

Yes, details are important. If it is interesting, a part of that work 
was published in
Journal of Magnetism and Magnetic Materials 320 (2008) 1904–1908
Physics of the Solid State, 2008, Vol. 50, No. 11, pp. 2095–2101

Briefly, we had obtained monotonous dependence of M_Fe on the number of 
Al atoms in NN environment:
:MMI001: in 1= 2.21855 N_Al=0
:MMI002: in 2= 2.07591 N_Al=0
:MMI003: in 3=-0.00358 NM
:MMI004: in 4= 0.00406 NM
:MMI005: in 5= 0.66093 N_Al=4
:MMI006: in 6=-0.08633 N_Al=6 <---!
:MMI007: in 7=-0.44927 N_Al=8
:MMI008: in 8=-0.74394 N_Al=8

> I also wonder what "ferromagnetic and nonmagnetic elements" should be
> and why you have a ferromagnetic order in cases where the magnetic moments
> have opposite signs (atoms 1/2 and 7/8 in your example).

When I used "ferromagnetic element" I meant that it is inclined to 
ferromagnetically order when pure and under normal conditions. I had Fe, 
but the topic starter may have another. As his M is rather small 
(MMI002=0.71935), maybe it is Co, or Ni. And I can imagine that AFM 
elements, like Cr, in compounds with NM element can have something like 
the data of the topic.

Of course, the problem of the topic could be simply an error of the ini 
files that are not shown to us...

> seems there is some confusion with physical concepts this is a ferrimagnet

Yes, that order is ferrimagnetic.

> The behaviour of the magnetic moments of alloys is very often in accordance with the so called Slater Pauling rule
> that is, the total moment depends on the number of valence electrons and the individual moments are adjusted to  result in the correct total moment.
> For localized moment systems it is caused by a trapping of the Fermi energy in a minimum of the density of states for minority  (or majority) electrons.
> Maybe this explains your observation. It's not a miracle, it's physics.

In that alloys, at some lattice parameters and concentrations, we had 
obtained two collinear solutions AFM (we called it so, in fact ferriM) 
and FM (also with monotonous dependence of M_Fe, all positive in that 
case). So, Slater-Pauling can work on only one of them.
For us the two solutions were an indication that a non-collinear order 
may appear.

Best regards
Lyudmila Dobysheva

>   > Am 09.12.2018 um 22:55 schrieb shaymlal dayananda:
> :MMI001: in 1=-0.00322
> :MMI002: in 2= 0.71935
> :MMI003: in 3=-0.00020 NM
> ...
> 10.12.2018 10:43, Peter Blaha wrote:
>> I don't know the details, but for me it seems very unlikely, that the
>> calculations are ok with these moments .....
> 
> Details are important.
> Once I had a study with an idea of the magnetic moment dependence on the
> environment. You can see result for a system with a ferromagnetic and a
> nonmagnetic elements. Number 3 and 4 are NM, rest are FM:
> :MMI001: in 1= 2.69759
> :MMI002: in 2= 2.60331
> :MMI003: in 3=-0.02591 NM
> :MMI004: in 4=-0.01019 NM
> :MMI005: in 5= 1.12279
> :MMI006: in 6= 0.34180
> :MMI007: in 7=-0.05635 <---!
> :MMI008: in 8=-0.07717 <---!
> The same structure and lattice parameter and another NM element:
> :MMI001: in 1= 2.21855
> :MMI002: in 2= 2.07591
> :MMI003: in 3=-0.00358 NM
> :MMI004: in 4= 0.00406 NM
> :MMI005: in 5= 0.66093
> :MMI006: in 6=-0.08633 <---!
> :MMI007: in 7=-0.44927
> :MMI008: in 8=-0.74394
Lyudmila Dobysheva
------------------
Physics-Techn.Institute,
Udmurt Federal Research Center, Ural Br. of Rus.Ac.Sci.
426000 Izhevsk Kirov str. 132
Russia
---
Tel. +7 (34I2)43-24-59 (office), +7 (9I2)OI9-795O (home)
Skype: lyuka18 (office), lyuka17 (home)
E-mail: lyuka17 at mail.ru (office), lyuka17 at gmail.com (home)