Dear all,
I know that this topic has been treated several times in this forum. I have looked at most of the answers but got confused with the examples. My aim is to get the nearest neighbour Jij exchange parameters for a magnetic alloy say of the form Ax B1-x where 0<x<1. I am only interested in nearest neighbour interactions and such that I only have three parameters JAA, JBB and JAB. I also have interest in calculating the magnonic dispersion across specific directions in the Brillouin zone.
From the various comments I read, I understood that I have to use the spin spiral states. However got confused with examples. which example shall I follow to do this. The Fe with spin spiral (no angles are defined in inout) or the Ni magnon energy.
In some kind of talk about non collinear magnetism in Elk code given in the cecam workshop 2011, there was a graph showing dE versus q for LaOFeAs. what does this graph serve for other than applying ot for magnetic excitations. Another graph was about Fe showing (E(theta)-EFM) versus theta for a given point q. Why is theta changing. to calculate Jij do we have to change theta or keep it the same for all qs. How do we know which theta to use. Why do we fix the spin moment sometimes (as I have seen in the input file)
Please if anyone can guide me of what to exactly use as first steps to get the Jij(q).
Thank you
Elie Moujaes
University of Rondonia
Brazil
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
if you are only interested in J_1^AA ...--> you don't really need to consider the spin-spiral stuff instead you may just take energy differences as a first approximations. Consequently, you write your Heisenberg Hamiltonian such as
H= J1 S1.S2 + C,
and then cycle through different magnetic arrangements S1=1 S2=1 & S1=1 S2=-1 to resolve C and J. Please note that here C contains all other Energy contributions. A more sophistacted approach is also given in Ref. [1] ( H. J. Xiang, E. J. Kan, S.-H. Wei, M. H. Whangbo, and X. G. Gong, Phys. Rev. B 84, 224429 (2011).) in particular eqn. 2 and 3, which are a more advanced version than mine introduced above.
Finally, please note that this is always just a model mapping. Hence you should pay attention if for example S changes it values once you go to another magnetic configuration. In this case the usage of a Heisenberg model may not be appropiate.
best regards
Michael
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Thanks for your reply. In fact I also want to calculate the magnonic energy so I have an interest in learning about the spin spiral method. I am trying first to do it for bcc Fe to understand it following an example in elk folder. I have two questions:
1- why is the bfieldc defined that way. It says that it must be along q but with perpendicular component only. I didnt get that. if the q point is 0.5 0.5 0 (xy plane) shouldnt the bfieldc be defined with a z component only . In the input of Fe-spiral example it only has an x component. Why is that. How does this affect things
2- I am trying to run the input but getting:
Info(elk): current task : 0
Info(checkmt): reduced muffin-tin radius of species 1 (Fe) from 2.4000 to 2.1801
Error(eveqnz): diagonalisation failed
ZHEEVD returned INFO = 77
Error(eveqnz): diagonalisation failed
ZHEEVD returned INFO = 77
Dear all,
I know that this topic has been treated several times in this forum. I have looked at most of the answers but got confused with the examples. My aim is to get the nearest neighbour Jij exchange parameters for a magnetic alloy say of the form Ax B1-x where 0<x<1. I am only interested in nearest neighbour interactions and such that I only have three parameters JAA, JBB and JAB. I also have interest in calculating the magnonic dispersion across specific directions in the Brillouin zone.
From the various comments I read, I understood that I have to use the spin spiral states. However got confused with examples. which example shall I follow to do this. The Fe with spin spiral (no angles are defined in inout) or the Ni magnon energy.
In some kind of talk about non collinear magnetism in Elk code given in the cecam workshop 2011, there was a graph showing dE versus q for LaOFeAs. what does this graph serve for other than applying ot for magnetic excitations. Another graph was about Fe showing (E(theta)-EFM) versus theta for a given point q. Why is theta changing. to calculate Jij do we have to change theta or keep it the same for all qs. How do we know which theta to use. Why do we fix the spin moment sometimes (as I have seen in the input file)
Please if anyone can guide me of what to exactly use as first steps to get the Jij(q).
Thank you
Elie Moujaes
University of Rondonia
Brazil
Dear Elie,
if you are only interested in J_1^AA ...--> you don't really need to consider the spin-spiral stuff instead you may just take energy differences as a first approximations. Consequently, you write your Heisenberg Hamiltonian such as
H= J1 S1.S2 + C,
and then cycle through different magnetic arrangements S1=1 S2=1 & S1=1 S2=-1 to resolve C and J. Please note that here C contains all other Energy contributions. A more sophistacted approach is also given in Ref. [1] ( H. J. Xiang, E. J. Kan, S.-H. Wei, M. H. Whangbo, and X. G. Gong, Phys. Rev. B 84, 224429 (2011).) in particular eqn. 2 and 3, which are a more advanced version than mine introduced above.
Finally, please note that this is always just a model mapping. Hence you should pay attention if for example S changes it values once you go to another magnetic configuration. In this case the usage of a Heisenberg model may not be appropiate.
best regards
Michael
Dear Michael,
Thanks for your reply. In fact I also want to calculate the magnonic energy so I have an interest in learning about the spin spiral method. I am trying first to do it for bcc Fe to understand it following an example in elk folder. I have two questions:
1- why is the bfieldc defined that way. It says that it must be along q but with perpendicular component only. I didnt get that. if the q point is 0.5 0.5 0 (xy plane) shouldnt the bfieldc be defined with a z component only . In the input of Fe-spiral example it only has an x component. Why is that. How does this affect things
2- I am trying to run the input but getting:
Info(elk): current task : 0
Info(checkmt): reduced muffin-tin radius of species 1 (Fe) from 2.4000 to 2.1801
Error(eveqnz): diagonalisation failed
ZHEEVD returned INFO = 77
Error(eveqnz): diagonalisation failed
ZHEEVD returned INFO = 77
Why is that. Please find the input below:
tasks
0
spinsprl
.true.
spinpol
.true.
vqlss
0.1 0.1 0.0
mixtype
1
xctype
3
epsengy
1.D-5
epspot
1.D-7
epsforce
2.D-4
nempty
10
rgkmax
8.5
gmaxvr
16
lmaxapw
10
lmaxmat
9
bfieldc
0.0 0.1 0.0
avec
1.0 1.0 -1.0
1.0 -1.0 1.0
-1.0 1.0 1.0
scale
2.54628
sppath
'/........../species/'
atoms
1 : nspecies
'Fe.in' : spfname
1 : natoms
0.0 0.0 0.0 0.0 0.0 0.0 : atposl, bfcmt
ngridk
10 10 10
!vkloff
!0.5 0.5 0.5
Regards