4-State Energy Mapping Workflow for 2D GdI₂

ASE-based workflow to study variation of nearest-neighbor Heisenberg magnetic exchange tensor components of 2D GdI₂ across:

• Different supercell sizes (SC size convergence check)
• Gd–Gd bonds
• Direction of moments in the surrounding bath

…for the 4 states using the 4-state energy mapping method.

The workflow initializes VASP calculations via ASE and executes them with a bash script.

Variations

#	Variation	Type	Example
1	Tensor component	Input <string>	xy, zz
2	One of 4 states	Optional input <string>	ud, dd
3	Direction of bath moments	Input <string>	x, y
4	2 selected atoms (J1/J2/J3…)	Optional input <2 ints>	3 1, 5 2
5	2D supercell to create	Input <2 ints>	4 5, 3 4

Files

• prepare-4state-noncol.sh — main script; run with the arguments above. Minimum requirement: two integers for SC size.
• POSCAR.3atom.transZ.cart.vasp — unit cell structure (POSCAR format), used to build supercells.
• supercell.py — supercell builder.
• run_leo.sh — submits jobs for execution.
• GdI2_4state-noncol.py — main ASE workflow that initializes the calculation.
• toten.sh — gathers energies and computes J values.

Requirements

• ASE (Python)
• Environment variable: VASP_PP_PATH