Quantum Tunneling
Through Classically Forbidden Barriers
Forbidden Passage
Roll a ball toward a hill. If it does not have enough energy to reach the top, it slows, stops, and rolls back. Every time. Classical mechanics is absolute on this point: a particle without sufficient energy cannot cross a potential barrier. It is forbidden. End of story.
Quantum mechanics disagrees. A particle approaching a barrier it classically cannot surmount does not simply bounce back. Its wave function, that mathematical object encoding everything knowable about a quantum system, does something remarkable. It decays exponentially inside barrier but maintains a nonzero amplitude on the far side. There is a small but real probability that a particle detected on one side of a wall will next be detected on the other side, having never had enough energy to climb over it. This is quantum tunneling, and it is not a loophole or a trick. It is a direct consequence of wave mechanics. Probability of transmission depends on barrier width and height relative to particle energy. Narrow, low barriers allow substantial tunneling. Wide, tall barriers suppress it exponentially. But probability never reaches exactly zero. In quantum mechanics, forbidden does not mean impossible. It means unlikely.
What makes tunneling so counterintuitive is that particle does not burrow through barrier in any physical sense. It does not find a hole. It does not borrow energy temporarily. Inside barrier, wave function amplitude drops off exponentially, meaning probability of finding particle there decreases sharply with depth. But on the far side, a diminished wave function emerges. Particle either appears beyond barrier or reflects back. It is never caught halfway inside.
Wave Function at a Wall
To see why tunneling happens, consider what a barrier looks like to a wave function. Outside barrier, wave function oscillates freely as a traveling wave. Energy of particle exceeds potential energy of its surroundings, so oscillation is sustained. But inside barrier, potential energy exceeds particle energy. Schrödinger's equation demands that wave function change character here. Instead of oscillating, it decays exponentially. Amplitude drops exponentially with distance into barrier, shrinking by a fixed fraction for every additional nanometer of depth. That decay rate depends on how far particle energy falls below barrier height. Larger energy deficit means steeper decay.
If barrier is thin enough, wave function has not decayed to negligible amplitude by the time it reaches the far edge. On the other side, where potential drops back down, wave function resumes oscillating, now as a transmitted wave with reduced amplitude. Square that reduced amplitude and you get transmission probability. Double barrier width and transmission probability does not halve. It drops by a factor that is itself exponential. This extraordinary sensitivity means tunneling is overwhelmingly important at atomic scales and completely negligible at everyday ones.
Nothing about this requires particle to "exist" inside barrier in a classical sense. Particle is not a little ball traveling through a forbidden zone with some mysterious extra energy. Wave function is the complete description. Inside barrier, it is an evanescent wave, present in the mathematics but not representing a classical trajectory. Tunneling is not a particle cheating classical rules. It is quantum mechanics being itself.
Powering Stars
Deep inside Sun's core, temperature reaches about 15 million kelvin. That sounds extreme, and it is by everyday standards. But it is nowhere near enough. For two protons to fuse, they must overcome their mutual electromagnetic repulsion, the Coulomb barrier. At 15 million kelvin, average thermal energy of a proton is roughly 1 keV. Coulomb barrier between two protons sits around 550 keV. Protons in Sun's core carry less than one five-hundredth of energy needed to overcome repulsion through classical means alone.
Tunneling makes fusion possible. Each proton-proton encounter has a vanishingly small chance of actually fusing, because two things have to happen at once: the protons must tunnel through their Coulomb barrier, and then a weak-force interaction must convert one proton into a neutron before they separate again. Combined probability per close encounter is something like one in 1026. But Sun's core contains roughly 1056 protons, each undergoing on the order of 108 close encounters every second. Multiply that astronomical number of attempts by an astronomically small probability and you get steady, sustained fusion at roughly 1038 events per second across the entire Sun. About 3.8 × 1026 watts of power, continuously, for billions of years. Without tunneling, protons in Sun's core would simply bounce off each other forever. Stars would not shine. Planets would not warm. Chemistry that supports life would never begin. Tunneling is not a minor quantum curiosity. It is what lights universe.
Escaping a Nuclear Prison
Tunneling works in the other direction too. Inside a heavy nucleus, an alpha particle (two protons and two neutrons bound together) rattles around in a potential well created by strong nuclear force. That well is deep, roughly 30 MeV below the surrounding potential landscape. But beyond the range of strong force, electromagnetic repulsion between alpha particle and remaining nucleus creates a barrier. Alpha particle does not have enough energy to climb over it classically.
Yet every so often, wave function of alpha particle extends through that barrier and emerges on the other side. When this happens, alpha particle flies away and nucleus has undergone alpha decay. George Gamow worked out this explanation in 1928, one of the earliest triumphs of quantum mechanics applied to nuclear physics. What makes this particularly striking is how sensitive tunneling probability is to barrier shape. Uranium-238 has a half-life of 4.5 billion years. Polonium-212 has a half-life of 0.3 microseconds. Both decay by the same mechanism, alpha tunneling through an electromagnetic barrier. Difference in half-life spans roughly 23 orders of magnitude. Yet barrier heights and widths differ by modest amounts. Exponential sensitivity of tunneling to barrier geometry turns small differences in nuclear structure into wildly different lifetimes. Same physics, same mechanism, dramatically different outcomes.
Technology Built on Tunneling
In 1981, Gerd Binnig and Heinrich Rohrer built a device that exploits tunneling's exponential sensitivity to distance. A scanning tunneling microscope (STM) holds a sharp metal tip just a few angstroms above a conductive surface. At that distance, electron wave functions from tip and surface overlap, and a tunneling current flows. Move tip one angstrom farther away and current drops by roughly an order of magnitude. This extreme sensitivity allows STM to map surface topography with sub-atomic resolution. Individual atoms become visible, not as photographs but as contours of electron probability. Binnig and Rohrer received Nobel Prize in 1986 for this work.
Flash memory in your phone and computer relies on tunneling too. Data is stored by trapping electrons on a floating gate, a tiny conductor surrounded by insulating oxide. To write data, a voltage is applied that allows electrons to tunnel through oxide layer onto floating gate. To erase, a reverse voltage lets them tunnel back out. Oxide barrier is thin enough for tunneling to occur at practical voltages but thick enough to retain charge for years without power. Every photo, message, and application on your device sits on electrons that tunneled into place.
Tunnel diodes, invented by Leo Esaki in 1957, use tunneling to create a component with negative resistance: a region where increasing voltage actually decreases current. This happens because tunneling probability through a thin semiconductor junction changes non-monotonically with applied voltage. Tunnel diodes switch faster than any conventional transistor of their era and remain useful in high-frequency circuits. Esaki shared Nobel Prize in 1973 for this discovery.
Tunneling You Can Hold
Everything so far has been a single particle slipping through a barrier - one electron, one proton. In the early 1980s a group at Berkeley asked a stranger question: could something you can hold in your hand tunnel as a whole? They built a circuit around a Josephson junction - two superconductors separated by an insulating gap a few atoms thick - and cooled it near absolute zero. In a superconductor, vast numbers of electrons lock into one coherent quantum state, so the electrical state of the circuit is not the story of any single electron. It is one collective quantum variable describing the whole device at once.
That single variable behaved exactly like the particle at the barrier. Held in one configuration by an energy hill it classically could not cross, the circuit tunneled out anyway - the entire macroscopic state jumped to the other side. And when John Clarke, Michel Devoret, and John Martinis measured the circuit's energy, it did not vary smoothly. It climbed in discrete steps, quantized into a ladder of allowed levels, the same kind of ladder an electron has inside an atom. Quantum mechanics, normally the rulebook of the very small, was running an object large enough to wire onto a chip.
This became the foundation of a technology. Take the lowest two rungs of that energy ladder, call them 0 and 1, and you have a superconducting qubit - the building block of the quantum processors Google, IBM, and others operate today. The 2025 Nobel Prize in Physics went to Clarke, Devoret, and Martinis for the discovery: macroscopic quantum tunneling and energy quantization in an electric circuit. The reason it works, and the reason you still cannot walk through a wall, is coherence. The circuit tunnels as one object only because all its electrons share a single quantum state. Ordinary matter does not.
Limits of Tunneling
Tunneling is not teleportation. It does not let particles appear at arbitrary distances or pass through barriers of any size. Transmission probability drops exponentially with barrier width. Double the width and probability does not halve; it drops by an exponential factor. For a barrier ten times wider, probability drops by that exponential factor raised to the tenth power. This compounding is ruthless.
Consider what this means for everyday objects. You are made of roughly 1028 atoms. For you to tunnel through a wall, every one of those atoms would need to tunnel simultaneously through a barrier that is, by atomic standards, astronomically wide. Probability of this happening is not merely small. It is something like 10 raised to the negative 1030. That number has more zeros after the decimal point than there are particles in observable universe. You could wait longer than any cosmological timescale and it would still not happen. Not once.
Tunneling is a quantum effect that thrives at quantum scales: individual particles, thin barriers measured in nanometers, energy differences measured in electron volts. It powers stars, enables technology, and shapes nuclear lifetimes. But it does not scale up. Exponential suppression sees to that. Quantum mechanics is not magic. It is precise, predictable, and bounded by its own mathematics. Tunneling respects those bounds completely.
