Kitaev interaction

[HANIF YUANDI WIDYANDARU]

Dear expert,

I notice there is the Kitaev function in the file. How do you implement that? Does it require an explicit K or the use of the exchange tensor? 

Thanks

[gabo...@gmail.com]

If a kitaev-constant is specified in the material file, I have noticed that after the calculation the log file contains a line with "Calculating Kitaev interactions":

username@computername:~/Desktop/test$ grep -i kitaev input
exchange:kitaev-cutoff-range=2.6 !A
username@computername:~/Desktop/test$ grep -i kitaev test.mat
material[1]:kitaev-constant=1.0e-18

username@computername:~/Desktop/test$ vampire-serial
...
                      Version 7.0.0 Dec 27 2023 21:29:33

             Git commit: fd9de813cab95bc2b2837122bd03f0238a5d2d76
...
username@computername:~/Desktop/test$ grep -i kitaev log
06-11-2025 [00:04:01] Calculating Kitaev interactions

Hope that helps.

Kind Regards,

Gavin

VAMPIRE user

[HANIF YUANDI WIDYANDARU]

Thank you for the answer. I plan to input the Kitaev interaction in tensorial form. Furthermore, I am currently generating a .ucf file with full tensorial exchange interactions and would like to verify the correct input format regarding matrix symmetry and unit cell translation.

1. Matrix Transposition If I define a tensorial interaction for a bond 0→1 with matrix J01​, does VAMPIRE require me to explicitly define the reverse bond 1→0 as the transpose (J10​=J01T​), or does the code automatically symmetrize the input?

I am currently assuming manual transposition is required. For example:

• Bond 0→1: 0 1 0 0 0 Jxx Jxy Jxz ...

• Bond 1→0: 1 0 0 0 0 Jxx Jyx Jzx ... (where off-diagonals are swapped)

2. Intra-unit cell bonds For atoms 0 and 1 located within the same unit cell, I have set the translation vectors (columns 3, 4, and 5) to 0 0 0 for both the forward and reverse bonds.

Could someone confirm if this explicit definition of the reverse bond with the transposed matrix and zero translation is the correct approach?

Thank you for your help.

[gabo...@gmail.com]

Yes, you need to provide in the unit cell file (ucf) the interactions both ways: i -> j and j -> i.  If not, the VAMPIRE program should give you the dreaded error message:

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

That was seen in a past thread of posts at [1].

As seen on slide 14 at [2], the tensorial columns are:

IID i j dx dy dz Jxx Jxy Jxz Jyx Jyy Jyz Jzx Jzy Jzz

From page 35 (pdf page 36) of the VAMPIRE 7.0 manual at [3]:

Column 1 - IID: Interaction ID that is only used for accounting purposes and starts at 0

Column 2 - i: The i atom number of atom in the local unit cell

Column 3 - j: The j atom number of atom in the local/remote unit cell

Columns 4, 5, and 6 - dx, dy, dz: Relative integer coordinates of unit cell for atom j

Columns 7 - 15 -  Jxx through Jzz: The nine exchange values in Joules

Based on that, column 3 seems to be one of the atom numbers for j and not x of a translation vector. 

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

[2] https://vampire.york.ac.uk/resources/workshop2017_day3.pdf

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

Kind regards and hope that helps,

Gavin

VAMPIRE user