Effects of the surface on double surface Fermi arcs in a realistic tight-binding model of Na3Bi (100)

The 30-second take

• What: First-principles tight-binding models of two Na3Bi (100) terminations show textbook double surface Fermi arcs on the stoichiometric face, and either arcs or two loops that keep arc-like local physics on a non-stoichiometric face.
• Why it matters: Topological surface states are a scarce materials primitive for possible spintronics. Knowing which arc properties survive dirty or reconstructed surfaces is a step toward usable defaults, not just ideal crystals.
• Who should care: Topological-materials theorists, ARPES and STM experimentalists, and spintronics materials scouts.

What the paper actually did

Na3Bi is a topological Dirac semimetal that can host double surface Fermi arcs (DSFAs), important both for classifying the material and for possible spintronic use. How real (100) surfaces affect those arcs has been unclear.

The authors build first-principles-derived tight-binding models of two Na3Bi (100) terminations. On the stoichiometric termination they find DSFAs as expected. On the non-stoichiometric structure they find DSFAs on one termination and, on the other, two loops that still keep many DSFA properties.

They conclude that local properties such as spin-momentum locking and hybridization with the Dirac points look more robust than global connectivity: the surface can reshape the arc fingerprint without destroying the arc-like physics.

What makes this disruptive

Classification arguments often treat Fermi-arc connectivity as the fingerprint. If a realistic surface can turn arcs into loops while spin-momentum locking and Dirac-point hybridization remain, the scarce capability—usable topological surface transport—may be more robust than the cartoon, and the cartoon may be the fragile part.

Why it matters (outside the lab)

Abundance lens: certain secure-signal and spintronic primitives are elite-only materials physics. Understanding which surface-state features survive real terminations is a long-horizon step toward those primitives becoming engineerable.

Not a device date. Near-term: a termination-resolved theory result.

Limitations & open questions

This is a tight-binding study of two modeled (100) terminations, not a full device measurement. “Two loops that retain many DSFA properties” is not the same as proving transport equivalence. Spintronic usefulness remains potential in the abstract. Preprint ≠ a Na3Bi chip. Abundance is not automatic.