In the Laboratory for Topological Matter at the University of Cologne, physicists have for the first time directly observed the quantum states that theory predicted back in 2008. The device — a Josephson junction on the surface of a topological insulator — exhibited precisely those highly conducting Andreev states that should arise from the interaction of superconductivity with topological surface electrons.
The team, led by Professor Yoichi Ando and with the participation of Dr. Jakob Schluck and Ella Nikodem, used tunneling spectroscopy. They measured the spectrum of states inside the junction by varying the phase difference between the two superconductors and applying a magnetic field. The results matched the calculations of Fabian Hassler's group at RWTH Aachen University and the model of Liang Fu and Charles Kane.
A topological insulator conducts current only along its surface, while remaining an insulator in its interior. When this surface is coupled to a superconductor, special electron-hole pairs — Andreev states — arise at the boundary. In an ordinary material they behave differently than on a topological surface: here they remain highly conducting over a broad range of carrier density and localize in specific regions of the junction under the action of a magnetic field.
This observation strengthens the theoretical foundation of topological quantum computing. In such a system, Majorana zero modes could store information nonlocally, protecting it from local disturbances and errors. The transition to topological superconductivity — the closing and reopening of the energy gap — has not yet been recorded, but the data obtained so far agree with the predictions to high precision.
The results were published on 2 November 2026 in the journal Science Advances. The researchers emphasize that the next step is to increase the energy and temporal resolution of the experiment in order to catch the very moment of the topological phase transition.
The real experiment in Cologne has shown that the surface of a topological insulator behaves exactly as needed for future error-resistant qubits, and this is already changing the map of possible paths toward a practical topological quantum computer.
