Nobel Prize in Physics 2025: Quantum Tunneling

The 2025 Nobel Prize in Physics will be awarded to three American researchers for their discovery of the phenomenon of quantum tunneling at the macroscopic scale.The prize will be shared by John Clarke, Michel Devoret, and John Martinis of the University of California, recognizing work that began nearly forty years ago.

In a series of experiments conducted at UC Berkeley in 1984 and 1985, the three researchers demonstrated that in a circuit built from superconductors separated by an insulating material, the particles inside behaved collectively as if they were a single, large particle. This was the first demonstration that a system composed of large particles could, under certain conditions, display distinctly quantum behavior, exhibiting phenomena such as quantum tunneling, in which particles act like waves and pass through a physical barrier. Their discovery laid the foundation for later advances in encryption systems, high-precision sensors, and more recently, the development of quantum computers.

A Question of Scale

Physics today is generally divided into two major theories, each describing very different kinds of phenomena. Classical physics explains the behavior of large objects—such as balls, satellites, and even stars, whose motion can be described by Newton’s three laws of motion, which form the foundation of classical mechanics.   In contrast, at the subatomic scale, the laws of quantum mechanics and its derivatives apply. This is a strange and counterintuitive world in which particles sometimes behave like waves, giving rise to phenomena such as interference, tunneling, and diffraction.

This year’s prize recognizes research on quantum tunneling and the extension of quantum physics to much larger systems—to objects large enough to see with the naked eye or even hold in one’s hand. To get a sense of what tunneling means, imagine a trainee on a military obstacle course facing a wall that’s simply too high to climb. After repeated failed attempts, instead of scaling it, he pulls out a hammer, makes a hole, and walks straight through to the other side. It’s not the expected solution, but it gets the job done.

In the quantum world—the realm of tiny particles—these particles can behave in surprisingly unexpected ways. This happens when electrons encounter a barrier,  such as the electric force binding them to an atom’s nucleus and preventing them from escaping into free space. At first glance, the electron has no chance of crossing the barrier, since its energy is lower than what would be required to overcome the electric force. Yet, because of its quantum, probabilistic nature, it can partially pass through the barrier, so that a fraction of its presence leaks out of the atom. Most of the electron is still stopped by the wall, but some portion manages to “tunnel” into the world beyond the barrier.

But if the difference between classical and quantum physics is so great—macroscopic versus microscopic—where exactly is the boundary between them? At what point does a collection of quantum particles become a ball that obeys Newton’s laws? At what point does a collection of quantum particles become a ball that obeys Newton’s laws? These questions are closely tied to when a system’s energy shifts from being quantized (discrete) to effectively continuous. In the macroscopic world, energy is continuous. When we bounce a ball or run down the street, in principle there is no limit to the precise amount of energy we can assign to the system at any given moment. Any real value is possible, aside from limits imposed by the speed of light. A rolling ball might carry 100 joules, 50 joules, 2 joules, but also half a joule, 0.000027 joules, or any other real number. Quantum mechanics, by contrast, tells us that energy comes in indivisible packets—quanta—that cannot be subdivided. A system cannot take on just any value, but only specific discrete ones. An electron bound within an atom, for example, is restricted to certain energy levels; all other values are forbidden. This is analogous to a guitar string fixed at both ends, which can vibrate only at certain frequencies and therefore produce only specific notes.

Phase in a Ring

John Clarke was born in Britain in 1942 and earned his Ph.D. in physics from the University of Cambridge in 1968. He later became a researcher at the University of California, Berkeley. Michel Devoret, born in France, completed his doctorate in physics at Paris-Sud University in 1982 before moving to Berkeley for postdoctoral training under Clarke. John Martinis, born in the United States in 1958, was at that time Clarke’s doctoral student; the research that would eventually earn him the Nobel Prize formed the basis of his Ph.D. thesis.

In a series of experiments conducted in 1984 and 1985, the three built an electronic circuit composed of superconductors—materials capable of conducting electric current with zero resistance. These superconductors only function at extremely low temperatures. The superconducting elements were separated by a thin insulating layer, forming a Josephson junction, a device that had itself been recognized with the Nobel Prize in Physics in 1973.

The system was initially prepared in a classical state, meaning it could, in principle, assume any energy. But once the experiment began, the charge carriers flowing through the Josephson junction exhibited remarkable behavior: they acted collectively as if they were a single particle “filling” the entire circuit. Quantum mechanics provides the explanation for this phenomenon, often described as the “particle in a ring.” In this case, however, it is more accurate to call it a “phase in a ring.” This marked the first demonstration of a macroscopic system displaying purely quantum behavior.

Measuring this unusual behavior of particles was made possible by a phenomenon known as Cooper pairs—pairs of electrons that move together on opposite sides of a physical barrier. The phase difference between the electrons in such pairs—that is, the angle describing their relative quantum states—generates a measurable electric current.  This provided direct evidence that tunneling can occur not only at the microscopic level but also at the macroscopic scaleת involving large particles.

Encryption and Quantum Computers

In the years since, the semiconductor industry has created new opportunities for building quantum devices—including encryption systems, sensors, and, more recently, quantum computers—many of which would not have been possible without the discoveries of Martinis, Devoret, and Clarke. Quantum sensors designed to measure magnetic fields, for instance, use phase measurements in superconducting circuits and are far more sensitive than their classical counterparts. Moreover, the ability to control quantum states on a macroscopic scale makes it possible to implement encryption protocols based on quantum principles. In such systems, information can be transmitted using superposition—where a system exists in several states at once until it is measured—or entanglement, where multiple particles are linked so that the state of one particle instantly affects the state of another.

The ability to control and monitor a system’s quantum state led to a series of follow-up experiments by Martinis, in which he demonstrated that the electronic circuit he had built with his colleagues functioned as a macroscopic “artificial atom.”  This artificial atom mimicked the properties of a real atom—a quantum system with discrete energy levels—precisely what is required for building a quantum computer.

From then on, Martinis’s research focused on that goal: using Josephson devices for the construction of quantum computers. In 2004, he moved to the University of California, Santa Barbara, and a decade later joined Google, which set out to build a practical quantum computer.  In 2019, he and his team published a study claiming “quantum supremacy”—a clear advantage of a quantum computer over its classical counterpart—but their claim was later challenged and refuted.

Today, quantum computers exist and are capable of carrying out complex calculations. However, they remain plagued by high error rates and have not yet surpassed the performance of classical computers—the ones we use daily at home, at work, and in our pockets. Much of today’s research in the field of quantum computing, in both academia and industry is devoted to quantum error correction and improving the reliability of quantum computation.

Clarke, now 83, is professor emeritus but continues as an active researcher at the University of California, Berkeley. In 2004 he received the Hughes Medal of the Royal Society for his development of sensors based on superconductivity. His current research includes, among other things, the use of superconducting devices as possible detectors of the axion, a hypothetical particle  that, if it exists, could be a fundamental component of dark matter.

Michel Devoret remained in the United States and is now a professor at Yale University and at the University of California, Santa Barbara. His research focuses on developing methods of quantum error correction and ensuring that quantum computers can deliver reliable results despite the inherent limitations of current technologies

Last year, the Nobel Prize in Physics highlighted advances in artificial intelligence and was awarded jointly to John Joseph Hopfield (Princeton University) and Geoffrey Everest Hinton (University of Toronto) for developing computational tools that simulate the activity of the nervous system and laid the groundwork for modern AI systems.

In 2023, the prize was awarded to three researchers who contributed to the development of methods for producing ultrashort flashes of light—lasting only a billionth of a billionth of a second: Pierre Agostini (Ohio State University, USA), Ferenc Krausz (Max Planck Institute of Quantum Optics, Munich, Germany), and Anne L’Huillier (Lund University, Sweden).

In 2022, the Nobel Prize in Physics was awarded to three scientists who demonstrated that quantum entanglement is a real physical phenomenon: Alain Aspect, John Clauser, and Anton Zeilinger.