Regarding Curie temperature calculation for Full Heusler alloy(Co2MnAl)

[Karunakaran M (RC2133002011015)]

Dear Vampire users,

I'm trying to calculate the Curie temperature of the Co₂MnAl full Heusler alloy. However, during the calculation, the process gets stuck at the “Verifying exchange interactions” step and displays the following error:

“Error! Exchange interaction list in unit cell file vampire.UCF contains the following asymmetric interactions:”

I'm familiar with calculating the Curie temperature for quaternary Heusler alloys using the Vampire package, as they involve four distinct atom types occupying four different lattice sites. However, in the case of a full Heusler alloy, there are only three different atom types across four sites, with the X(Co) atom occupying two of the positions.

I have attached the necessary input files and the log file related to the Vampire calculations. Could someone kindly review them and guide me on how to properly calculate the Curie temperature for full Heusler alloys?

If any additional files or details are required, I will be glad to provide them. Thank you in advance for your support.

Best regards,
Karunakaran M

[الأستاذ لطرش‎]

  Dear  Karunakaran M

1 - There is a problem in the exchange interaction list and the atomic coordinates, i think it comes from the generation of the .ucf file

# Unit cell size (Angstrom):
5.742 5.742 5.742
# Unit cell vectors:
1.0 0.0 0.0
0.0 1.0 0.0
0.0 0.0 1.0
# Atoms num, id cx cy cz mat lc hc
4 3
 0 0.00 0.00 0.00 0 0 0
 1 0.50 0.50 0.50 1 0 0
 2 0.25 0.25 0.25 2 0 0
 3   0.75   0.75   0.75   3  0 0
# Interactions; J values from SPR-KKR multiplied by 2; DMI not yet

2 - There is a problem in your .mat file , you should have 4 materials instead of 3 and the Co atom must be repeated twice.

It should look like this 

material:num-materials = 4
#---------------------------------------------------
# Material 1
#---------------------------------------------------
material[1]:material-name=Al
material[1]:unit-cell-category = 1
material[1]:damping-constant=1.0
material[1]:atomic-spin-moment=0.13750000 !muB
material[1]:uniaxial-anisotropy-constant=0.00000000e+00
material[1]:material-element=Al
material[1]:initial-spin-direction = 0.0,0.0,-1.0
material[1]:uniaxial-anisotropy-direction = 0.0 , 0.0, 1.0
#---------------------------------------------------
#---------------------------------------------------
# Material 2
#---------------------------------------------------
material[2]:material-name=Mn
material[2]:unit-cell-category = 2
material[2]:damping-constant=1.0
material[2]:atomic-spin-moment=2.92420000 !muB
material[2]:uniaxial-anisotropy-constant=0.00000000e+00
material[2]:material-element=Mn
material[2]:initial-spin-direction = 0.0,0.0,1.0
material[2]:uniaxial-anisotropy-direction = 0.0 , 0.0, 1.0
#---------------------------------------------------
#---------------------------------------------------
# Material 3
#---------------------------------------------------
material[3]:material-name=Co
material[3]:unit-cell-category = 3
material[3]:damping-constant=1.0
material[3]:atomic-spin-moment=0.66370000 !muB
material[3]:uniaxial-anisotropy-constant=0.00000000e+00
material[3]:material-element=Co
material[3]:initial-spin-direction = 0.0,0.0,1.0
material[3]:uniaxial-anisotropy-direction = 0.0 , 0.0, 1.0
#---------------------------------------------------
# Material 4
#---------------------------------------------------
material[3]:material-name=Co
material[3]:unit-cell-category = 4
material[3]:damping-constant=1.0
material[3]:atomic-spin-moment=0.66370000 !muB
material[3]:uniaxial-anisotropy-constant=0.00000000e+00
material[3]:material-element=Co
material[3]:initial-spin-direction = 0.0,0.0,1.0
material[3]:uniaxial-anisotropy-direction = 0.0 , 0.0, 1.0
#---------------------------------------------------

3 - Can i ask you how you did calculate the exchange coupling parameters , i'm working on a quaternary heusler similar to yours "CoFeMnAl" but the Curie temperature i got was 150K lower than the experimental and Mean fieled approximation value if i calculate J interaction using SPRKKR and its 350K higher if i calculate J using Wannier and TB2J.

Best regards.

‫في الجمعة، 22 أغسطس 2025 في 6:39 م تمت كتابة ما يلي بواسطة ‪Karunakaran M (RC2133002011015)‬‏ <‪km0...@srmist.edu.in‬‏>:‬

--
You received this message because you are subscribed to the Google Groups "Vampire Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to vampire-user...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/vampire-users/6e4b1f5a-8afd-4963-a05b-6d36a9272c92n%40googlegroups.com.

[Karunakaran M (RC2133002011015)]

Dear Vampire users,

First of all, thank you for your response.

I apologize for the earlier error. The correct values are: number of atoms = 4 and number of materials = 3.

# Unit cell size (Angstrom):
5.742 5.742 5.742
# Unit cell vectors:
1.0 0.0 0.0
0.0 1.0 0.0
0.0 0.0 1.0
# Atoms num, id cx cy cz mat lc hc
4 3
0 0.00 0.00 0.00 0 0 0
1 0.50 0.50 0.50 1 0 0
2 0.25 0.25 0.25 2 0 0

3 0.75 0.75 0.75 2 0 0

# Interactions; J values from SPR-KKR multiplied by 2; DMI not yet

The earlier mistake occurred while I was testing whether the files would run after modifying the number of materials to 4 in the .UCF and .mat files, as you suggested. However, technically, full Heusler alloys have 3 different types of atoms occupying 4 distinct atomic sites. My earlier upload contained an incorrectly modified .UCF file, which was my error. Now, I have attached the corrected files of full Heusler alloys for your reference.
I am still encountering the following error:

“Error! Exchange interaction list in unit cell file vampire.UCF contains the following asymmetric interactions.”

Could anyone please assist me on how to resolve this issue?

With respect to your queries:

(1) I am calculating exchange coupling parameters using the KKR-CPA method as implemented in the SPRKKR package.

(2) As you mentioned, sometimes the Monte Carlo results are underestimated and the mean field approximation results are overestimated compared to the experimental values. This overestimation by the mean-field approximation is well-documented.

Based on my experience, if you are targeting more accurate Monte Carlo results, I would recommend increasing the number of Monte Carlo steps and applying a field to the system can improve your results.

I hope this information will be useful for your calculations. Please feel free to reach out if you have any further questions.

Best regards,
Karunakaran M

--

With Regards,

Karunakaran M

Ph.D. scholar,

Dept. of Physics and Nanotechnology,

SRMIST KTR.

Ph. No: 9092931036.

[gabo...@gmail.com]

In your log file, I see the following error list:

23-08-2025 [10:20:06] Error! Exchange interaction list in unit cell file vampire.UCF contains the following assymetric interactions:
0 1 3 -1 1 0
1 1 3 0 0 0
4 1 3 0 1 -1

                       ...

In your vampire.UCF file, I see:

5508 isotropic
0 1 3 -1 1 0 2.288484445464e-21
1 1 3 0 0 0 2.288484445464e-21
2 1 2 0 0 0 2.288484445464e-21
3 2 1 0 0 0 2.288484445464e-21
4 1 3 0 1 -1 2.288484445464e-21

...

The VAMPIRE 7.0 manual under section "6 Unit Cell Files" on page 34 has:

13 num_interactions [exchange_type]
14 IID i j dx dy dz | Jij

As you might have read at [2] or [3], that assymetric error comes from the i and j values.  For the exchange line for IID of 0, you have i of 1 and j of 3.  To remove the error by making it symmetric, you also need to add an exchange line for i of 3 and j of 1.  You will also have to add opposite i and j exchange lines for the other IIDs 1, 4, 6, 12, and so on that your seeing in the log file error list.

[1] https://vampire.york.ac.uk/resources/vampire-manual.pdf

[2] https://groups.google.com/g/vampire-users/c/yZ7cykl7tGo/m/Q58k-18rBgAJ

[3] https://groups.google.com/g/vampire-users/c/jOGUDJS8Br4/m/1HVyvJZCAgAJ

Kind Regards,

Gavin

VAMPIRE user