Quantum Bath Puts Distant Qubit Entanglement on Autopilot

Physicists at the Institute of Science and Technology Austria (ISTA) have demonstrated a fully autonomous method to entangle distant quantum bits using a “quantum bath” of correlated microwave photons — confirming a theoretical prediction first proposed more than two decades ago. Published in Physical Review X in July 2026, the experiment eliminates the need for active control pulses and repeated measurements that have defined entanglement generation since the field’s inception.

The work, led by PhD student Alejandro Andrés-Juanes and professor Johannes Fink in collaboration with an international team, represents the first experimental realization of steady-state distributed entanglement through a non-local squeezed reservoir.

Why this matters

Distributed entanglement — correlations between physically separated qubits that exceed anything classical physics allows — is the backbone of scalable quantum computers and future quantum networks. Until now, creating such connections has followed one of two paths: sending a single actively controlled photon from one qubit to another, or having each qubit emit a photon and matching the results through post-selection. The latter approach earned the 2022 Nobel Prize in Physics but still depends on repeated measurements and does not always succeed.

The ISTA team’s quantum bath flips this paradigm. Instead of choreographing entanglement with precisely timed control pulses, the qubits’ environment itself becomes the source of entanglement. A parametric amplifier generates a continuous stream of entangled photon pairs, with individual photons from each pair sent to the two distant qubits through separate waveguides. Because the photons are inherently correlated, emission events at one qubit become quantum-mechanically linked to events at the other. Through nonlocal destructive interference, these emission pathways cancel out, trapping the qubits in a stable “dark state” that protects entanglement from decay.

“This way, the entangled qubit state is stabilized, even beyond the qubits’ own ‘lifetime’, and remains always available as a resource for further quantum processing,” said Fink. The entanglement persists continuously — not as a fleeting resource that must be consumed immediately, but as a persistent ground state the qubits inhabit.

Bridging two worlds of entanglement

A central challenge the team addressed was the mismatch between two forms of entanglement. Continuous-variable entangled states — analogous to a pendulum where position and momentum vary smoothly — can be generated efficiently but are not directly useful for most quantum computing tasks. Discrete-variable entanglement, the “all-or-nothing” correlations that stationary qubits rely on, is what practical applications demand.

“In this work, we aimed to overcome this mismatch between the readily available and the practically useful forms of entanglement,” said Andrés-Juanes. The quantum bath bridges this gap by using the readily available squeezed (continuous-variable) light to stabilize the discrete-variable entanglement that superconducting qubits need.

The experiment

The prototype uses superconducting transmon qubits — the workhorse of modern quantum processors — embedded in separate three-dimensional microwave cavities cooled to millikelvin temperatures. The cavities couple not to each other but to a common transmission line carrying correlated microwave photons. The qubits were tuned to different frequencies to prevent direct resonant energy exchange.

To verify entanglement, the team employed quantum tomography, reconstructing the qubits’ quantum states from many rapid measurement “slices” lasting only 20–80 nanoseconds each. The analysis confirmed that the two isolated qubits maintained synchronized, entangled states inside the quantum bath — the first experimental validation of the 20-year-old theory.

Efficiency and scalability

The method currently transfers approximately 10% of the quantum bath’s available entanglement, making it less efficient than approaches that actively control qubit states. “Our method currently transfers about 10% of the bath’s available entanglement,” Andrés-Juanes acknowledged. However, the researchers emphasize that the setup is relatively simple and could be scaled to synchronize multiple distant qubits — a critical requirement for fault-tolerant quantum computation.

Fink noted that the original theory was proposed under idealized conditions, which may explain why a functional quantum bath took two decades to realize. “Our experiments helped us reveal several factors that may have prevented scientists from designing a functional quantum bath using a single source of correlated photons for distributed entanglement,” he said.

A parallel breakthrough

The ISTA work was published alongside a complementary demonstration by a team at the University of Illinois Urbana-Champaign (UIUC), which achieved steady-state remote entanglement through a different mechanism — a unidirectional quantum system using a microwave circulator. Both approaches exploit dark states protected by quantum interference between emission pathways, and both were featured in an APS Physics commentary by MIT postdoctoral researcher Aziza Almanakly.

Whether steady-state entanglement can remain “always-on” during actual computational operations remains an open question. In practice, entanglement must eventually be transferred to other qubits for consumption by an algorithm — a process that a continuous active drive might disrupt. Nonetheless, both demonstrations serve as valuable proofs of principle for a new class of quantum interconnects.

Looking ahead

The ISTA prototype opens avenues for quantum-optics experiments and for scaling quantum processors toward fault-tolerant operation. By providing on-demand, long-lived entanglement without complex feedback loops, the quantum bath approach could eventually simplify the architecture of modular quantum computers — where connecting distant modules through autonomous entanglement may become as fundamental as wiring is to classical computing.

The research was funded by the Austrian Science Fund (FWF), the European Research Council, and the EU Horizon Europe program, among others.