Particles can pass through barriers they classically shouldn’t be able to cross. This isn’t science fiction — it’s the reason the sun shines, transistors switch, and radioactive atoms decay.
Consider a ball rolled toward a hill. Classically, if the ball doesn’t have enough energy to get over the hill, it bounces back. End of story.
Quantum mechanically, the ball’s wavefunction doesn’t stop at the hill’s edge. It decays exponentially inside the barrier but doesn’t vanish. If the hill is thin enough, the wavefunction emerges on the other side — diminished but nonzero.
The particle doesn’t go “over” the barrier. It doesn’t go “around” it. Its probability amplitude seeps through it.
Tunneling is not the particle disappearing from one side and appearing on the other. The wavefunction is continuous — it extends through the barrier, decaying exponentially with distance. There’s no gap, no discontinuity.
Consider sound leaking through a wall. The sound wave doesn’t teleport — its amplitude drops inside the wall material, but some makes it through to the other side. Tunneling works the same way, except the “sound” is probability amplitude and the “wall” is a potential energy barrier.
The key factors governing tunneling probability are intuitive:
Electrons tunnel readily. Protons tunnel occasionally. Baseballs never tunnel through walls — the probability is nonzero but on the order of 10⁻³⁰ or worse.
Tunneling probability falls off roughly as e^(-2κL), where κ depends on the barrier height and particle mass, and L is the barrier width. This exponential dependence is extreme.
Double the barrier width and you don’t halve the probability — you square the suppression factor. This is why tunneling matters at atomic scales (barriers of nanometers) but is utterly negligible at human scales (barriers of centimeters).
The exponential sensitivity also explains why scanning tunneling microscopes achieve sub-angstrom resolution — tiny changes in distance produce large changes in tunneling current.
Tunneling isn’t a thought experiment. It’s the mechanism behind:
Bloom filters offer a surprisingly apt parallel. A Bloom filter says “definitely not in the set” or “probably in the set” — queries can leak through as false positives. The probability of leakage depends on the filter’s size (barrier width) and the number of hash functions (barrier height).
Make the filter larger and false positives drop exponentially — just like tunneling probability drops exponentially with barrier width.
Consider NAT traversal in networking. A firewall creates a “potential barrier” — packets from outside shouldn’t reach internal hosts. But techniques like hole punching exploit timing windows where packets get through the barrier by finding paths that shouldn’t classically exist.
The packet doesn’t break the firewall. It finds a configuration where the barrier is thin enough to traverse. Same energy, different effective barrier width — just like a particle finding a thinner region of a potential barrier.
Tunneling enables quantum annealing — a computational approach used by systems like D-Wave. The idea: an optimization problem defines an energy landscape with many local minima. A classical optimizer can get stuck in a local minimum, unable to climb over barriers to find better solutions.
A quantum annealer exploits tunneling to pass through energy barriers between local minima, potentially reaching the global minimum without climbing over every hill.
This is not the same as gate-based quantum computing. It’s a distinct paradigm, and its advantage over classical approaches remains an active area of research.
This lesson establishes:
Next: Quantum Spin