We tend to think quantum tunnelling is a subatomic curiosity, but in 1984 and 1985 physicists at UC Berkeley showed that a superconducting electrical circuit big enough to hold in your hand could tunnel from one state to another — a discovery that won the 2025 Nobel Prize in Physics.

Quantum tunnelling is usually introduced through the smallest things physics can describe: an alpha particle escaping a nucleus, an electron crossing a barrier, a particle appearing where ordinary intuition says it should not be. The language makes it sound like a private trick of the subatomic world.

The 2025 Nobel Prize in Physics recognised something stranger and more useful. In 1984 and 1985, John Clarke, Michel H. Devoret and John M. Martinis showed that a superconducting electrical circuit could display the same kind of behaviour on a scale large enough to be held in the hand. The Nobel Prize in Physics 2025 was awarded to the three physicists “for the discovery of macroscopic quantum mechanical tunnelling and energy quantisation in an electric circuit.”

The phrase needs care. The circuit did not vanish from one side of a table and reappear on the other. What tunnelled was not a metal loop as a visible object, but a collective electrical state inside a superconducting circuit. In the experiment, billions of paired electrons in the superconductor behaved together as one quantum system. That system could move from a zero-voltage state to a voltage state even though, in classical terms, it did not have enough energy to cross the barrier between them.

That is why the experiment mattered. It did not make quantum mechanics larger by metaphor. It made a macroscopic circuit behave in a way that had previously belonged mostly to atoms, nuclei and individual particles.

A circuit cold enough to behave as one

The Berkeley experiment used superconductors, materials that can carry electrical current without resistance when cooled to extremely low temperatures. In a superconductor, electrons pair up into Cooper pairs. These pairs can move in a coordinated way, and under the right conditions the whole superconducting state can be described by a shared quantum wave function.

Clarke, Devoret and Martinis built a circuit containing a Josephson junction: two superconductors separated by a very thin insulating barrier. A Josephson junction is already a deeply quantum device, because paired electrons can tunnel through the barrier and create a supercurrent. The prize-winning work asked a more ambitious question. Could a whole collective variable of the circuit, not just an individual particle, show quantum behaviour?

The relevant variable was the phase difference across the junction. This is not a familiar everyday quantity, but in a superconducting circuit it can act like the coordinate of a particle moving in an energy landscape. In the zero-voltage state, that effective particle sits in a valley. Classical physics says it should stay there unless it receives enough energy to climb out. Quantum mechanics allows another possibility: it can tunnel through the barrier.

This is the image behind the Nobel explanation, but it should not be taken too literally. The circuit was not a ball passing through a wall. It was a manufactured electrical system whose collective state followed the mathematics of a quantum particle trapped behind an energy barrier.

Why the 1984 and 1985 dates matter

The Nobel committee’s popular science background places the key experiments in 1984 and 1985 at the University of California, Berkeley. Clarke was a professor there. Devoret had joined his group as a postdoctoral researcher. Martinis was a doctoral student.

The group had to solve a difficult experimental problem. If the circuit switched out of its zero-voltage state, how could they know the cause was quantum tunnelling rather than heat, electrical noise or some flaw in the setup? At very low temperatures, the thermal explanation should fade. But noise in the laboratory could imitate the result if not carefully controlled.

So the work depended on precision as much as imagination. The team cooled the circuit, controlled the current through the Josephson junction, and repeatedly measured how long the system stayed in the zero-voltage state before switching. Quantum tunnelling is probabilistic, so a single switch would not prove much. Many repeated measurements could reveal whether the escape rate followed the pattern predicted by quantum mechanics.

The result was that the system escaped in a way consistent with macroscopic quantum tunnelling. It behaved as if the collective phase of the junction, representing a large number of particles acting together, had passed through a barrier it could not classically cross.

The second part was energy quantisation

The Nobel citation has two parts: tunnelling and energy quantisation. The second part is just as important.

Quantum systems do not take on any energy value they like. Their allowed energies can come in discrete levels. That idea is familiar from atoms, where electrons occupy specific energy states. Clarke, Devoret and Martinis showed a related effect in their superconducting circuit.

They introduced microwaves while the circuit was in its zero-voltage state. At particular microwave frequencies, the system absorbed energy and moved to a higher level. That made it more likely to escape the zero-voltage state. The pattern showed that the circuit had quantised energy levels, not a continuous spread of possibilities.

This was the second sign that the circuit was not merely an electrical component behaving oddly. It was a macroscopic electrical system showing recognisable quantum structure.

Big enough to hold, not ordinary enough to see

The phrase “big enough to hold in your hand” can easily mislead. It does not mean the circuit looked like a household object displaying quantum behaviour in plain sight. The active structures were on a chip about a centimetre in size, held inside a carefully controlled low-temperature apparatus. The relevant state was measured through voltage and switching statistics, not watched directly.

Still, the scale matters. Earlier demonstrations of tunnelling were tied to microscopic systems. Here the circuit involved billions of Cooper pairs filling the superconductor on the chip. The important coordinate belonged to the system as a whole. That is why the result is described as macroscopic quantum tunnelling.

The experiment also helps explain why the boundary between the quantum and classical worlds is not a simple size line. Large systems usually lose visible quantum behaviour because they interact with their surroundings, pick up noise and become effectively classical. But if a system is built carefully, cooled deeply and isolated well enough, collective quantum behaviour can survive in an engineered circuit.

Why it mattered decades later

The Berkeley work was not a quantum-computing product launch. It was a physics experiment testing whether quantum mechanics could be observed in a collective circuit variable. But the later relevance is clear. Superconducting qubits, one of the major approaches to quantum computing, rely on circuits whose energy levels can be treated quantum mechanically. A qubit needs two controllable quantum states, and superconducting circuits use Josephson junctions to create the needed nonlinearity.

Martinis later became one of the major figures in superconducting quantum computing. Devoret’s work continued to shape circuit quantum electrodynamics and superconducting quantum devices. Clarke’s Berkeley group had already built a long record in superconducting electronics and sensitive measurement. The Nobel recognition connects these later technologies to an earlier conceptual step: showing that an electrical circuit could be a quantum object in a measurable way.

That does not mean the 1980s experiment directly delivered today’s machines. Scientific lineages are rarely that clean. Many other theorists, experimentalists, materials advances and engineering choices sit between the Berkeley circuit and modern quantum processors. But the experiment helped legitimise a central idea: quantum mechanics could be designed into electrical circuits, not only found in isolated atoms.

The useful shock of the result

The reason tunnelling feels unintuitive is that everyday objects do not behave this way. A ball thrown at a wall does not pass through it. A switch does not turn itself on without a mechanism. Classical experience teaches us that barriers are barriers.

Quantum mechanics is different. It describes systems through probabilities, wave functions and allowed states. If the barrier is finite and the system is quantum, there can be a nonzero chance of appearing on the other side. With a single particle, that idea is already difficult. With a superconducting circuit involving billions of paired electrons acting together, it becomes a more pointed challenge to the old habit of separating the small quantum world from the large engineered world.

The 2025 Nobel Prize in Physics honoured that challenge. Clarke, Devoret and Martinis did not show that everything around us is visibly quantum in daily life. They showed that under exacting conditions, a human-made circuit could be pushed into a regime where the collective state of many particles still obeyed quantum rules.

That is the quiet force of the experiment. Quantum tunnelling was not only a subatomic curiosity. In a cold superconducting circuit on a Berkeley chip, it became an electrical transition that could be measured, repeated and eventually built into the language of modern quantum technology.