Projected amorphous topological insulators

The 30-second take

• What: The authors define projected amorphous topological insulators (PATIs): keep a random disconnected fraction x of a parent crystal, integrate out the rest, and still find a quantized global invariant (strong PATI) or only local markers on a few sites (fragile PATI), plus in-gap boundary modes.
• Abundance angle: today, topological protection is an elite, lattice-perfect materials luxury. Showing topology can survive a random, incomplete projection is a step toward cheaper default topological devices if experiments follow this theory (mid-horizon: manufacturing and measurement still decide).
• Who should care: Condensed-matter theorists of amorphous and disordered topology, and materials groups wondering whether crystalline order is required for a strong invariant.

What the paper actually did

The authors introduce projected amorphous topological insulators (PATIs). Even though the system contains only a fraction x of otherwise randomly selected disconnected sites of a parent crystal, it can still carry a quantized global topological invariant (a strong PATI) when the effective Hamiltonian is built by integrating out the leftover sites of the original lattice.

Inside the parent model’s topological regime there is also a critical x below which the system is only a fragile PATI: a quantized local topological marker lives on a small fraction of sites. Both strong and fragile PATIs host in-gap modes near the boundary. In the entire trivial parameter regime of the lattice model, any x yields a normal insulator with no gapless boundary modes and vanishing global invariant and local marker.

They demonstrate these outcomes, which they call possibly generic, starting from a parent square-lattice model of time-reversal-symmetry-breaking insulators. The strong-to-fragile PATI transition, tuned by x, is characterized by a mean correlation-length exponent ν in (1.00, 1.46).

What makes this disruptive

The scarce capability is topological insulation that does not demand a perfect crystal. If a random subset of sites, after integrating out the rest, still supports a strong quantized invariant and edge modes, “amorphous” stops meaning “topology optional.”

The strong versus fragile distinction, the trivial-side control (always normal), and a reported ν window for the x-tuned transition give the idea a phase-diagram shape, not only a slogan.

This is a theory paper on a square-lattice parent. Treat “possibly generic” as their hope, not a measured material.

Why it matters (outside the lab)

Abundance lens: topological devices and protected edge modes are still boutique crystals. If projection-and-integrate-out topology is real beyond this model, more disordered or sparse assemblies might carry protection that is closer to a default materials trick than an ultra-clean lattice luxury.

Near-term, the preprint is a theoretical phase distinction. Medium-term, other parent models, interactions, and experiments decide whether PATIs become a materials program.

No year. A Hamiltonian construction does not ship a chip.

Limitations & open questions

Theory preprint; no experimental PATI is reported. The demonstration uses a square-lattice, time-reversal-breaking parent. “Possibly generic” is not a proof for all topological classes. We have not reproduced the invariant calculations or the ν range.

The abstract does not specify the parent Hamiltonian, how x is sampled, or the numerical system sizes. Fragile PATIs only show local markers on a small fraction of sites — easy to miss experimentally. Integrating out residual sites is a theoretical effective-Hamiltonian step, not a lab recipe.

Abundance is not automatic: a new insulator class on paper does not cheapen topological hardware.