Hadamard Get the app

Macroscopic Quantum Tunnelling Explained: 1985 Chip to 2025 Nobel

Macroscopic Quantum Tunnelling Explained: 1985 Chip to 2025 Nobel — Hadamard

Macroscopic quantum tunnelling happens when a large system, in this case a whole superconducting electric circuit, crosses an energy barrier it could not cross under classical physics, acting as a single quantum particle. John Clarke, Michel Devoret and John Martinis demonstrated it in 1984 and 1985 with a circuit on a chip about one centimetre across, and the work won them the 2025 Nobel Prize in Physics.

This article explains macroscopic quantum tunnelling step by step. It covers what tunnelling is, how a Josephson junction works, what the two 1985 papers measured, and how that chip connects to the superconducting qubits used in some quantum computers today.

What is quantum tunnelling, in plain terms?

Classical physics says a ball rolling toward a hill can’t get over it without enough energy. It rolls back every time. A barrier you lack the energy to climb works like a wall.

Quantum mechanics treats particles differently. A particle is described by a wave that gives the probability of finding it in different places. That wave doesn’t stop dead at a barrier. It fades inside the barrier, and if the barrier is thin enough, a small part of it comes out the other side. So there is some chance of finding the particle beyond a barrier it could never climb classically. That is quantum tunnelling.

Tunnelling is familiar for single tiny particles such as electrons. The open question for decades was whether it also applies to something large: a system containing a huge number of particles, built in a lab and wired to ordinary measuring equipment. Everyday objects don’t tunnel through walls. Was that just a matter of probability, or does quantum behaviour disappear at large scales? Other quantum effects raise similar puzzles. Our explainers on the EPR paradox and Bell tests and the Aharonov-Bohm effect cover two of them.

Josephson junction explained: two superconductors and a thin barrier

The laureates worked with a device called a Josephson junction: two pieces of superconductor separated by a thin layer that doesn’t conduct electricity. In a superconductor, electrons pair up into what physicists call Cooper pairs, and these pairs can move together without resistance.

According to Physics World’s report on the prize, Brian Josephson predicted in 1962 that Cooper pairs could tunnel across the insulating barrier. The prediction was confirmed in 1963. That is tunnelling by individual pairs of particles.

In the late 1970s, Anthony Leggett proposed something bolder: in such a junction, the whole macroscopic quantum state could tunnel, not just particles one at a time. The Royal Swedish Academy of Sciences’ press release describes what makes this possible: the charged particles in the superconductor behave together like a single particle that fills the entire circuit.

This is the key idea. The quantity that tunnels isn’t one electron. It is a collective property of the whole circuit. In the 1985 papers that property is the phase difference across the junction, a single variable that describes the shared state of all the Cooper pairs.

Macroscopic quantum tunnelling explained: the Clarke, Devoret and Martinis experiment

The experiments ran at UC Berkeley. Physics World notes that Devoret was then a postdoc and Martinis a graduate student, both in Clarke’s group.

The trapped state

The circuit starts in a zero-voltage state. A current flows through the junction but no voltage appears across it. The Academy describes the system in this state as trapped behind a barrier. Classically it could only escape with enough energy to get over the barrier, for example from heat.

Quantum mechanics allows another way out: tunnelling through the barrier. Whichever way the system escapes, the result is easy to see, because a voltage appears across the junction.

How they measured it

The method, as summarised by Physics World, had three steps:

  1. Slowly raise the current through the junction.
  2. Record the current at which a voltage suddenly appears, which marks the moment of escape.
  3. Repeat many times and average the results, because each escape is random.

The tell-tale sign is temperature. If heat is pushing the system over the barrier, the escape behaviour changes with temperature. At very low temperatures the team found that the average stopped depending on temperature. Something other than heat was getting the circuit out, which pointed to tunnelling.

Why noise was the main obstacle

The hardest part was noise. Stray electrical disturbances reaching the junction can kick the system out of its trapped state. As Physics World points out, that kind of noise can mimic tunnelling. A temperature-independent escape rate only means something if outside disturbances have been ruled out, so much of the work went into isolating the circuit.

What the paper reported

The results appeared as “Measurements of Macroscopic Quantum Tunneling out of the Zero-Voltage State of a Current-Biased Josephson Junction” by Devoret, Martinis and Clarke in Physical Review Letters 55, 1908, published on 28 October 1985. According to the abstract:

  • The team measured the rate at which an underdamped (Q≈30), current-biased junction escaped the zero-voltage state.
  • They measured the junction’s parameters in situ, using how the escape rate depended on bias current and a technique called resonant activation with microwaves.
  • At low temperatures the escape rate became temperature-independent and matched the zero-temperature quantum tunnelling prediction with no adjustable parameters.

That last point matters most. The team didn’t fit the theory to the data afterwards. They measured the circuit’s properties independently, plugged them into the quantum prediction, and the prediction matched.

Energy quantization in an electric circuit

The second half of the Nobel citation concerns energy quantisation. In quantum mechanics, a bound system such as an electron in an atom can only hold certain discrete amounts of energy. It absorbs or emits energy in specific steps, never in between.

The Academy’s press release says the laureates showed the same behaviour in their circuit: it absorbs or emits only specific amounts of energy. The result appeared in “Energy-Level Quantization in the Zero-Voltage State of a Current-Biased Josephson Junction” by Martinis, Devoret and Clarke, Physical Review Letters 55, 1543. It was received on 14 June 1985 and published on 7 October 1985.

The authors describe it as the first observation of quantized energy levels for a macroscopic variable, namely the phase difference across the junction. The positions of the energy levels agreed quantitatively with a quantum mechanical calculation that used junction parameters measured in the classical regime.

The publication date has a neat coincidence. The paper appeared on 7 October 1985, and the Nobel Prize was announced on 7 October 2025, exactly 40 years to the day later.

Paper Journal reference Published Main finding
Energy-Level Quantization in the Zero-Voltage State… PRL 55, 1543 7 October 1985 Discrete energy levels for the junction’s phase difference, matching quantum theory
Measurements of Macroscopic Quantum Tunneling… PRL 55, 1908 28 October 1985 Temperature-independent escape at low temperature, matching tunnelling theory with no adjustable parameters

From a 1985 chip to superconducting qubits

Taken together, the two results showed that a circuit you can design, build and wire up behaves like an artificial atom. It has discrete energy levels, and its collective state follows quantum mechanics. Vladimir Manucharyan of EPFL put it this way in EPFL’s news coverage: the work showed that “electrical currents and voltages can themselves behave as quantum mechanical variables.” He said it led to “circuit quantum electrodynamics and superconducting qubits.”

Physics World makes the same link. The noise-reduction techniques and the multi-level behaviour of the junction led to superconducting qubits, which are the basis of some early quantum computers. A qubit needs two well-separated energy levels that can be controlled and read out. The 1985 energy-level paper showed that a Josephson circuit has such levels, and the noise work showed how to protect them.

Clarke was direct about the connection: “The basis of quantum computing relies to quite an extent on our discovery,” he said, as quoted by Physics World. The later careers of his co-laureates follow the same path. Devoret became chief scientist at Google Quantum AI in 2024. Martinis joined Google in 2014 and co-founded Qolab in 2022, where he is CTO.

Quantum computers built on this foundation also matter for cryptography. A sufficiently large, error-corrected quantum computer would threaten the elliptic-curve signatures (ECDSA) that secure most blockchains today. That is why Hadamard builds BB84, a Layer 2 rollup on Ethereum that uses hash-based one-time addresses instead of ECDSA keys. Our post-quantum migration tracker follows how blockchains, browsers and standards bodies are preparing.

The 2025 Nobel Prize: who won, and why it took 40 years

On 7 October 2025, the Royal Swedish Academy of Sciences awarded the 2025 Nobel Prize in Physics to John Clarke, Michel H. Devoret and John M. Martinis “for the discovery of macroscopic quantum mechanical tunnelling and energy quantisation in an electric circuit.” The prize of 11 million Swedish kronor is shared equally. Physics World puts that at about US$1.2 million and notes that the award ceremony takes place in Stockholm on 10 December.

The laureates

  • John Clarke: born in 1942 in Cambridge, UK, and on the UC Berkeley faculty since 1969. Affiliated with UC Berkeley at the time of the prize.
  • Michel H. Devoret: born in 1953 in Paris. Led a group at CEA-Saclay from 1984, moved to Yale in 2002, then in 2024 moved to UC Santa Barbara and became chief scientist at Google Quantum AI. Affiliated with Yale, UC Santa Barbara and Google Quantum AI.
  • John M. Martinis: born in 1958 in the US. Moved to UC Santa Barbara in 2004, joined Google in 2014, joined Silicon Quantum Computing in Australia in 2020, and co-founded Qolab in 2022. Affiliated with UC Santa Barbara and Qolab.

Clarke’s reaction to the news, quoted by Physics World: “To put it mildly, it was the surprise of my life.” Olle Eriksson of Uppsala University, chair of the Nobel Committee for Physics, gave the broader view: “There is no advanced technology today that does not rely on quantum mechanics.” In the Academy’s release he added: “It is wonderful to be able to celebrate the way that century-old quantum mechanics continually offers new surprises.”

Why the wait?

The sources don’t give an official reason for the timing. One reasonable reading is that the full significance of the 1985 results only became clear as superconducting qubits developed in the decades that followed. The citation itself names the physics discovery, not the technology. For another case of a physics prize explained from experiment to impact, see our piece on IceCube and ghost particles.

FAQ

What is quantum tunnelling in simple terms?

It is the ability of a quantum system to cross an energy barrier it doesn’t have enough energy to climb. Quantum objects are described by probability waves that leak through thin barriers, so there is some chance of finding the system on the far side.

How can a whole circuit with a huge number of particles behave like a single quantum particle?

In a superconductor, the charged particles act together in one shared quantum state. According to the Royal Swedish Academy, they behave like a single particle filling the whole circuit. The variable that tunnels is a collective one, the phase difference across the Josephson junction, rather than any individual electron.

Did this experiment lead directly to quantum computers?

It laid the groundwork. Physics World reports that the noise-reduction techniques and multi-level behaviour from these experiments led to superconducting qubits, the basis of some early quantum computers. Turning them into working machines took decades of further work by many groups.

Why did the 1985 result take 40 years to win a Nobel Prize?

No official reason has been given. The prize was announced on 7 October 2025, exactly 40 years after the energy-quantisation paper was published, by which time superconducting circuits had become a leading platform for quantum computing.

Sources