ABACUS variable-cell DFT-relaxed CIF
What do the magnetic moments action say about this CIF?
This was I4/mmm - think the DFT is correct or is there something wrong?
The collinear-PBE Magnetic moments action converged this 3-atom Fe₂B cell to an almost compensated state:
Fe sites: −0.0144 μB each (predominantly Fe- )
B: +0.0021 μB
Net cell moment: −0.0266 μB (absolute total reported as 0.0266 μB)
Reported saturation magnetization: 9.65 × 10³ A/m = 0.0121 T
So this run's observation is effectively nonmagnetic, despite is_magnetic: true indicating that it used spin-polarized (nspin=2) DFT. For Fe₂B, that near-zero result warrants a sanity check before interpreting it as a material property: it may be a low-spin/compensated SCF solution rather than the ferromagnetic state one would expect to test. The route used PBE, DZP, 50 Ry, and 0.3 Å⁻¹ k-spacing.
Compute the converged total energy, total and site-projected magnetic moments (Mulliken), site charges, and saturation magnetization (A/m, T = μ₀ Ms, emu/cm³) when available. Returns a magCIF with atomsitemoment so the local moments can be visualized. Run FM- and AFM-seeded requests at identical settings to compare their total energies. Set soc_axis to include spin–orbit coupling and site orbital moments, which carry a large share of the moment in Co and rare-earth compounds.
I would not treat this as a correct Fe₂B magnetic ground state yet. The calculation converged numerically, but it converged to the wrong physical branch or, more likely, started from the wrong structural prototype.
The input’s provenance says it requested I4/mcm (#140) but was generated as I4/mmm (#139). That is material here, not just a setting label: established CuAl₂-type Fe₂B is tetragonal I4/mcm, with B on 4a and Fe on 8h. For that phase, Materials Project reports a 3.70 μB/f.u. final moment (mp-1915), and experiment/relativistic DFT put it near 3.8–4.0 μB/f.u. and about 1.5–1.6 T. Our result, 0.027 μB/f.u. and 0.012 T, is two orders of magnitude away.
The relaxation itself also raised the same warning: moments nearly vanished during SCF. “Converged” therefore means forces/stress and the SCF met tolerances for this nonmagnetic I4/mmm 3-atom solution. It does not establish that it is lower in energy than the ferromagnetic I4/mcm Fe₂B state.
The decisive validation is to rebuild from a vetted I4/mcm Fe₂B CIF, seed parallel Fe moments around +2 μB/Fe, then compare its energy and moment with the nonmagnetic solution. An AFM test needs the 2×2×2 cell because the 3-atom cell cannot represent distinct Fe sublattices. Until those energy comparisons are made, I would flag this current DFT relaxation as an invalid magnetic reference rather than evidence that Fe₂B is nonmagnetic.