[Wien] Reg: Fixed spin moment calculations

Discussion:

Peram sreenivasa reddy

2014-03-29 12:17:10 UTC

Dear WIEN2k,

I am working on magnetic compound. For my system i got
total magnetic moment 1.95 bohr magneton. Now i want to do fixed spin
moment calculations.

My quires are
1). To do fixed spin moment calculations, first i have to initiate the
calculations in spin polarized case after that i have to give the run
command like "runfsm_lapw -m 1.95 -ec 0.000001".
Is it correct way or give any suggestions?

2). To check weather this magnetic moment is correct or not i have to take
different magnetic moment values like 0.5, 1.0, 1.5, 2.0, 2.5, 3.0. For all
the values i have to run scf and check the total energy for each fsm value.
After this i have to plot total energy verses magnetic moment values. In
this plot where i get the minimum energy for a particular fsm value, that
magnetic moment value i should have to consider.
Is this correct procedure?

Please suggest me further.

Thank you very much in advance.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20140329/f74eb4d0/attachment.htm>

tran

2014-03-29 12:25:47 UTC

Permalink

Yes, this is the correct procedure.

Post by Peram sreenivasa reddy
Dear WIEN2k,
??????????????????? I am working on magnetic compound. For my system i got total magnetic moment
1.95 bohr magneton. Now i want to do fixed spin moment calculations.
????????? My quires are
1). To do fixed spin moment calculations, first i have to initiate the calculations in spin
polarized case after that i have to give the run command like "runfsm_lapw -m 1.95 -ec 0.000001".
Is it correct way or give any suggestions?
2). To check weather this magnetic moment is correct or not i have to take different magnetic moment
values like 0.5, 1.0, 1.5, 2.0, 2.5, 3.0. For all the values i have to run scf and check the total
energy for each fsm value. After this i have to plot total energy verses magnetic moment values. In
this plot where i get the minimum energy for a particular fsm value, that magnetic moment value i
should have to consider.
Is this correct procedure?
Please suggest me further.
Thank you very much in advance.

Peram sreenivasa reddy

2014-03-29 13:03:50 UTC

Permalink

Dear Sir,

Thank you very much for your replay.

Post by tran
Yes, this is the correct procedure.
Dear WIEN2k,

Post by Peram sreenivasa reddy
I am working on magnetic compound. For my system i
got total magnetic moment
1.95 bohr magneton. Now i want to do fixed spin moment calculations.
My quires are
1). To do fixed spin moment calculations, first i have to initiate the
calculations in spin
polarized case after that i have to give the run command like
"runfsm_lapw -m 1.95 -ec 0.000001".
Is it correct way or give any suggestions?
2). To check weather this magnetic moment is correct or not i have to
take different magnetic moment
values like 0.5, 1.0, 1.5, 2.0, 2.5, 3.0. For all the values i have to
run scf and check the total
energy for each fsm value. After this i have to plot total energy verses
magnetic moment values. In
this plot where i get the minimum energy for a particular fsm value, that
magnetic moment value i
should have to consider.
Is this correct procedure?
Please suggest me further.
Thank you very much in advance.

_______________________________________________
Wien mailing list
Wien at zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
http://www.mail-archive.com/wien at zeus.theochem.tuwien.ac.at/index.html

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20140329/c88f5886/attachment.htm>

Peter Blaha

2014-03-31 05:50:48 UTC

Permalink

Step 2) is the correct thing, but at least you should also add a total
energy for M=0.
And if the normal runsp calculation gives you the lowest total energy
compared to all your fsm calculations, this is the proper solution.

However: do you know WHY you want to make a fsm calculation ?

The FSM method is used in 2 cases:
a) The total energy of more than ONE magnetic/non-magnetic calculations
is almost identical. You may see this in a normal scf calculation that
it converges very slowly (or not at all) or the calculations of runsp_c
and runsp (zero and finite moments) give almost the same energy.
b) You obtained some solution with runsp, but you want to search for
another (eg. a high-spin or low-spin) solution. In this case you have to
estimate what would be the total moment of such an solution and search
the corresponding M-range.

Post by Peram sreenivasa reddy
Dear WIEN2k,
I am working on magnetic compound. For my system i
got total magnetic moment 1.95 bohr magneton. Now i want to do fixed
spin moment calculations.
My quires are
1). To do fixed spin moment calculations, first i have to initiate the
calculations in spin polarized case after that i have to give the run
command like "runfsm_lapw -m 1.95 -ec 0.000001".
Is it correct way or give any suggestions?
2). To check weather this magnetic moment is correct or not i have to
take different magnetic moment values like 0.5, 1.0, 1.5, 2.0, 2.5, 3.0.
For all the values i have to run scf and check the total energy for each
fsm value. After this i have to plot total energy verses magnetic moment
values. In this plot where i get the minimum energy for a particular fsm
value, that magnetic moment value i should have to consider.
Is this correct procedure?
Please suggest me further.
Thank you very much in advance.
_______________________________________________
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 at zeus.theochem.tuwien.ac.at/index.html

Peram sreenivasa reddy

2014-03-31 06:55:51 UTC

Permalink

Dear Peter Blaha Sir,

Thank you very much for your replay.

For my system there is no experimental or theoretical proof as it is a
magnet. But in my calculations i got minimum energy in magnetic case with
magnetic moment 1.95 bohr magneton.

I want to confirm that this magnetic moment is correct or not. For that i
am using the fsm calculations.

I followed the step 2 procedure and i got the results like as bellow.

Magnetic moment (bohr magneton) Energy (Ry)
0.0 -16590.18461593
0.5 -16590.18612148
1.0 -16590.18900235
1.5 -16590.19082779
1.9 -16590.19142097
1.95 -16590.19142434
2.0 -16590.19141282
2.05 -16590.19138454
2.10 -16590.19133936
2.5 -16590.19027153
3.0 -16590.18587074
3.5 -16590.17627193
4.0 -16590.16408747

this shows with 1.95 magnetic moment i got minimum energy -16590.19142434
Ry.

In my normal scf calculations with "runsp_lapw -ec 0.000001" command i got
minimum energy -16590.19143942 Ry.

Is this correct?

Please give me suggestions. Is there any another procedure?. If it is
please suggest me.

If i have to add a total energy in M=0. How to add it?.

Thank you very much.

On Mon, Mar 31, 2014 at 11:20 AM, Peter Blaha

Post by Peter Blaha
Step 2) is the correct thing, but at least you should also add a total
energy for M=0.
And if the normal runsp calculation gives you the lowest total energy
compared to all your fsm calculations, this is the proper solution.
However: do you know WHY you want to make a fsm calculation ?
a) The total energy of more than ONE magnetic/non-magnetic calculations is
almost identical. You may see this in a normal scf calculation that it
converges very slowly (or not at all) or the calculations of runsp_c and
runsp (zero and finite moments) give almost the same energy.
b) You obtained some solution with runsp, but you want to search for
another (eg. a high-spin or low-spin) solution. In this case you have to
estimate what would be the total moment of such an solution and search the
corresponding M-range.

--
P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-165300 FAX: +43-1-58801-165982
Email: blaha at theochem.tuwien.ac.at WWW: http://info.tuwien.ac.at/
theochem/
--------------------------------------------------------------------------
_______________________________________________
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 at zeus.theochem.tuwien.ac.at/index.html

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20140331/3831358d/attachment.htm>

pluto

2014-03-31 12:13:02 UTC

Permalink

Dear All,

When doing only SOC calculation, one defines the spin quantization axis
when running initso.

But how about FM + SOC? All is good as long as one plots spin projected on
the magnetization easy axis. But how to e.g. define magnetization along
001 and plot spin projected on 100?

I hope you can provide me some good hint!

Regards,
Lukasz

Robert Laskowski

2014-04-01 00:47:07 UTC

Permalink

Hi,

for that you need to compute whole spin density matrix, from it you can extract all components of the spin magnetization.
However spin collinear code computes only diagonal part of the matrix, therefore only z component can be determined.
To get x and y you would have to go for non collinear code.

regards

Robert

Post by pluto
Dear All,
When doing only SOC calculation, one defines the spin quantization axis
when running initso.
But how about FM + SOC? All is good as long as one plots spin projected on
the magnetization easy axis. But how to e.g. define magnetization along
001 and plot spin projected on 100?
I hope you can provide me some good hint!
Regards,
Lukasz
_______________________________________________
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 at zeus.theochem.tuwien.ac.at/index.html

--
==================================================
Dr. Robert Laskowski

Senior Scientist, Materials Science & Engineering Department
Institute of High Performance Computing, A*STAR
1 Fusionopolis Way, #16-16, Connexis, Singapore 138632
Tel(Off): +65. 64191493 Fax: +65. 64632536
=================================================

________________________________

This email is confidential and may be privileged. If you are not the intended recipient, please delete it and notify us immediately. Please do not copy or use it for any purpose, or disclose its contents to any other person. Thank you.

pluto

2014-04-01 07:15:20 UTC

Permalink

Hello Robert,

Thank you for the quick answer.

Does it mean, that the standard version of Wien2k cannot do this?

Regards,
Lukasz

Post by Robert Laskowski
Hi,
for that you need to compute whole spin density matrix, from it you can
extract all components of the spin magnetization.
However spin collinear code computes only diagonal part of the matrix,
therefore only z component can be determined.
To get x and y you would have to go for non collinear code.
regards
Robert

Post by pluto
Dear All,
When doing only SOC calculation, one defines the spin quantization axis
when running initso.
But how about FM + SOC? All is good as long as one plots spin projected on
the magnetization easy axis. But how to e.g. define magnetization along
001 and plot spin projected on 100?
I hope you can provide me some good hint!
Regards,
Lukasz
_______________________________________________
Wien mailing list
Wien at zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
http://www.mail-archive.com/wien at zeus.theochem.tuwien.ac.at/index.html

--
==================================================
Dr. Robert Laskowski
Senior Scientist, Materials Science & Engineering Department
Institute of High Performance Computing, A*STAR
1 Fusionopolis Way, #16-16, Connexis, Singapore 138632
Tel(Off): +65. 64191493 Fax: +65. 64632536
=================================================
________________________________
This email is confidential and may be privileged. If you are not the
intended recipient, please delete it and notify us immediately. Please do
not copy or use it for any purpose, or disclose its contents to any other
person. Thank you.
_______________________________________________
Wien mailing list
Wien at zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
http://www.mail-archive.com/wien at zeus.theochem.tuwien.ac.at/index.html

Robert Laskowski

2014-04-01 08:02:18 UTC

Permalink

Hi,

correct, you may try to use non-collinear port WIENNCM, but it is not mpi parallel,
so it is practical with smaller cases only, and has not been touched for long time.

regards

Robert

Post by pluto
Hello Robert,
Thank you for the quick answer.
Does it mean, that the standard version of Wien2k cannot do this?
Regards,
Lukasz

Post by pluto
Dear All,
When doing only SOC calculation, one defines the spin quantization axis
when running initso.
But how about FM + SOC? All is good as long as one plots spin projected on
the magnetization easy axis. But how to e.g. define magnetization along
001 and plot spin projected on 100?
I hope you can provide me some good hint!
Regards,
Lukasz
_______________________________________________
Wien mailing list
Wien at zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
http://www.mail-archive.com/wien at zeus.theochem.tuwien.ac.at/index.html

--
==================================================
Dr. Robert Laskowski
Senior Scientist, Materials Science & Engineering Department
Institute of High Performance Computing, A*STAR
1 Fusionopolis Way, #16-16, Connexis, Singapore 138632
Tel(Off): +65. 64191493 Fax: +65. 64632536
=================================================
________________________________
This email is confidential and may be privileged. If you are not the
intended recipient, please delete it and notify us immediately. Please do
not copy or use it for any purpose, or disclose its contents to any other
person. Thank you.
_______________________________________________
Wien mailing list
Wien at zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
http://www.mail-archive.com/wien at 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 at zeus.theochem.tuwien.ac.at/index.html

Peter Blaha

2014-04-01 08:45:03 UTC

Permalink

I guess in principle the lapwdm program of standard WIEN2k has all the
information, but it prints only the integrated quantities summed over
the BZ.
(create a case.indmc file with "d"-for your atom(s) and run
x lapwdm -up -so
You can see in case.scfdmup the total moments as vector.

Since most likely you want the individual components for each k-point
and eigenvalues, one would need to check where they are in the code and
print them.

PS: In case you want the 3D-magnetization density of all components, you
should clearly go for the WIENncm code....

Post by Robert Laskowski
Hi,
correct, you may try to use non-collinear port WIENNCM, but it is not mpi parallel,
so it is practical with smaller cases only, and has not been touched for long time.
regards
Robert

Post by pluto
Hello Robert,
Thank you for the quick answer.
Does it mean, that the standard version of Wien2k cannot do this?
Regards,
Lukasz

Post by pluto
Dear All,
When doing only SOC calculation, one defines the spin quantization axis
when running initso.
But how about FM + SOC? All is good as long as one plots spin projected on
the magnetization easy axis. But how to e.g. define magnetization along
001 and plot spin projected on 100?
I hope you can provide me some good hint!
Regards,
Lukasz
_______________________________________________
Wien mailing list
Wien at zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
http://www.mail-archive.com/wien at zeus.theochem.tuwien.ac.at/index.html

--
==================================================
Dr. Robert Laskowski
Senior Scientist, Materials Science & Engineering Department
Institute of High Performance Computing, A*STAR
1 Fusionopolis Way, #16-16, Connexis, Singapore 138632
Tel(Off): +65. 64191493 Fax: +65. 64632536
=================================================
________________________________
This email is confidential and may be privileged. If you are not the
intended recipient, please delete it and notify us immediately. Please do
not copy or use it for any purpose, or disclose its contents to any other
person. Thank you.
_______________________________________________
Wien mailing list
Wien at zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
http://www.mail-archive.com/wien at zeus.theochem.tuwien.ac.at/index.html

--
==================================================
Dr. Robert Laskowski
Senior Scientist, Materials Science & Engineering Department
Institute of High Performance Computing, A*STAR
1 Fusionopolis Way, #16-16, Connexis, Singapore 138632
Tel(Off): +65. 64191493 Fax: +65. 64632536
=================================================
________________________________
This email is confidential and may be privileged. If you are not the intended recipient, please delete it and notify us immediately. Please do not copy or use it for any purpose, or disclose its contents to any other person. Thank you.
_______________________________________________
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 at zeus.theochem.tuwien.ac.at/index.html

pluto

2014-04-01 14:36:48 UTC

Permalink

Dear Robert, Peter,

Thank you for quick and helpful comments. I will not be able to modify the
code anytime soon, but this is ok :-)

I need it for bulk bcc Fe, so everything is very quick.

Can I maybe use some trick? Will it e.g. work if I do one extra scf
iteration with a single k-point, and then do the lapwdm?

Regards,
Lukasz

Post by Peter Blaha
I guess in principle the lapwdm program of standard WIEN2k has all the

information, but it prints only the integrated quantities summed over
the BZ.

Post by Peter Blaha
(create a case.indmc file with "d"-for your atom(s) and run
x lapwdm -up -so
You can see in case.scfdmup the total moments as vector.
Since most likely you want the individual components for each k-point

and eigenvalues, one would need to check where they are in the code and
print them.

Post by Peter Blaha
PS: In case you want the 3D-magnetization density of all components, you

should clearly go for the WIENncm code....