[Wien] AFM calculations for YBCO6

[pieper]

Good evening,

Sorry, I commented first without remembering the structure correctly.

The antiferromagnetic and superconducting phase diagram of Y123O(6+x) 
has been studied experimentally very extensively in the 90's with - in 
my opinion - very clear results on the exact AF phases from neutron 
diffraction and NMR. I don't want to drown anyone in literature, so 
being an NMR guy I strongly suggest to look into the NMR work:

E. Brecht and W. W. Schmahl and H. Fuess and S. Schmenn and H. 
L{\"u}tgemeier and N. H. Andersen and B. Lebech and Th. Wolf,
Neutron diffraction and NQR study of the intermediate turn angle phase 
formed during AFI to AFII reordering in 
{YBa$_2$Cu$_{3-x}$Al$_x$O$_{6+\delta}$},
Physical Review B, 1997, 56, no.2, p 940

Allow me to suggest in addition my own work

K. Nehrke and M. W. Pieper and T. Wolf,
Local magnetic properties of PrBa2Cu3O6+x: {NMR} and {NQR} in crystals,
Phys. Rev. B, 1996, 53, 1

Perhaps you can work from there through the references for details and 
additional studies.

There are two crystalografic different Cu sites in the perovskite, the 
so-called chain- and twice as many plane-sites. The variable oxygen 
content is between the Cu chain sites. In accord with what Prof. 
Delamora noted, the Cu on chains never carrys a magnetic moment. Cu in 
the planes forms a (well studied) spin-1/2 2D Heisenberg square lattice 
AF (nearest neighbors antiparallel) at low oxygen concentration (the 
Y123O6 end of the phase diagram), and the famous high-temperature 
superconductor when the oxygen sites between chain-Cu are filled, doping 
the electron holes into the planes.

If the system orders AF the planes ALWAYS form a square lattice spin-1/2 
AF with nearest neighbors antiparallel. So you need to assign numbers 
like Cu1 (for up) and Cu2 (for down) for Cu in the planes. The chain Cu 
is on a different crystalografic site and doesn't need numbering.

What I did not remember correctly is that since you have two plane site 
Cu even in the primitive cell you can, in fact, get away with a 1*1*1 
cell. BUT look in XCRYSDEN or something like it to make sure that your 
numbering makes nearest neighbors in the planes antiparallel! A 
configuration with parallel spins along the a- or b-axis should have 
significantly higher E_tot.

If you consider not only Y123O6 but introduce oxygen between Cu 
chain-sits you loose the 4-fold symmetry of the c-axis even in the 
non-magnetic case. This is why, as far as I recall, at least a 2*2*1 
supercell is necessary. To describe the different stacking sequences 
along the c-axis you will have to at least double the cell along the 
c-axis as well. See the above mentioned Phys Revs and the literature 
cited therein for the variety of AF phases that come about by different 
stacking of these planes along the c-axis depending on the oxygen 
content, or on the rare earth element.

As an additional comment: To my knowledge all evidence indicates that 
there is NO coexistence of AF-order and superconductivity. At 
intermediate oxygen content a phase separation occures into 
superconducting and AF phases.

Good luck with your calculations,

Martin Pieper

---
Dr. Martin Pieper
Karl-Franzens University
Institute of Physics
Universitätsplatz 5
A-8010 Graz
Austria
Tel.: +43-(0)316-380-8564


Am 16.05.2015 21:15, schrieb delamora:
> This is a problem I had studied before, but non magnetic, so I was
> curious to try the magnetic case; w
> With YBa2Cu3O7 you have magnetism and superconductivty! braking one of
> Matthias rules
> 
> The result I got for, YBa2Cu3O6, was that Cu1 MM=0.0007 or "non
> magnetic" as far as the calculations show.
> Cu2=0.5873 with U=4eV
> So, you have to study only Cu2 for magnetism.
> From my point of view you the most and I would guess only ordering of
> the inplane spins is like a chess board, alternating un and dn.
> the interplanar Cu2-Cu2 ordering can be up-up or up-dn, but since
> there are no oxygen atoms between the planes the difference should be
> quite small.
> 
> I did the calculation with RxK=7 and 100 k-points with 4 parallel
> cores, it took half an hour.
> 
> In the supercell output you have to delete all the numbers
> Ba 1 => Ba, O 1 => O and only number 1 and 2 the Cu as you want to 
> order them
> I did;
> Cu 00x and 1/2,1/2,x as Cu 1 and later as 'up'
> and for 0,1/2,x and 1/2,0,x as Cu 2 and later as 'dn'
> Now you can have 00,-x as 'dn' and 0,1/2,-x as up and this gives
> another ordering
> ________________________________________
> De: wien-bounces at zeus.theochem.tuwien.ac.at
> <wien-bounces at zeus.theochem.tuwien.ac.at> en nombre de Madesis
> Ioannis(John) <imadesis at physics.uoc.gr>
> Enviado: sábado, 16 de mayo de 2015 01:09 p. m.
> Para: Wien
> Asunto: Re: [Wien] AFM calculations for YBCO6
> 
> Mr. Delamora, first of all thank you for your dedication, and thorough
> examination of my problem.
> 
> I haven't fully tested your solution, however, there are plenty of AFM
> orderings that I wish to test such as A, G and C type, all of which are
> different combinations of in-plane antiferromagnetism. From first 
> trials
> of your solution, what troubles me is the fact that the
> case_super.struct file does not include any numbers such as Cu1, O3.
> However, I have noticed that some of the atoms have "MULT= 2" or "MULT=
> 4" and present additional 1 or 3 positions. So I thought of the
> following:
> Keep the 1 x 1 x 1 cell, for the sake of calculations, and split by
> myself the Cu1 atoms from 1 atom of MULT= 4, to 2 groups of MULT= 2, 
> and
> use your solution of naming appropriately each atom. I know that you
> suggested something particular, and thank you, but I need to keep the
> computational cost low. Last, I change the space group to "1". After I
> do what I said, I run an initialization, and a proper space group is
> defined. How is that for a solution?
> 
> --
> Ioannis Madesis (Μαδέσης Ιωάννης)
> PhD Student
> Atomic and Molecular Physics
> Department of Physics
> University of Crete
> (0030)-210-6503598
> _______________________________________________
> Wien mailing list
> Wien at 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 at 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