HomeArticlesQuantum Physics

Quantum Tunnelling: The WKB Approximation and the Gamow Factor

Real barriers aren't rectangular. The WKB approximation turns exponentially suppressed leakage through any smoothly varying barrier into a formula you can actually evaluate.

mysimulator teamUpdated July 2026≈ 9 min read▶ Open the simulation

The exact rectangular barrier

For a particle of energy E < V₀ incident on a rectangular barrier of height V₀ and width L, matching the wavefunction and its derivative at both boundaries gives a closed-form transmission coefficient. Three regimes emerge depending on how large κL is:

κ = √(2m(V₀−E)) / ħ
T = [1 + V₀²·sinh²(κL) / (4E(V₀−E))]⁻¹
κL ≫ 1 (thick/tall barrier):  T ≈ 16·E(V₀−E)/V₀² · e^(−2κL)

The thick-barrier row is the key result: T falls off exponentially with κL — a factor of e in T per extra unit of width — which is why tunnelling is negligible for macroscopic objects but dominant for electrons and light nuclei.

live demo · transmission through an adjustable potential barrier● LIVE

WKB: generalising to any barrier shape

Real barriers — the Coulomb tail around a nucleus, the vacuum gap under an STM tip — aren't rectangular. The Wentzel–Kramers–Brillouin (WKB) approximation treats V(x) as locally rectangular at each point, defining a position-dependent decay constant κ(x), then integrates across the full classically forbidden region between turning points a and b (where E = V(a) = V(b)):

κ(x) = √(2m(V(x)−E)) / ħ
T ≈ exp[ −2∫ₐᵇ κ(x) dx ]     ← the Gamow tunnelling integral

George Gamow applied exactly this formula in 1928 to explain alpha decay: an alpha particle trapped by the strong nuclear force still faces a long-range Coulomb barrier on the way out. The Gamow integral through that barrier reproduces the Geiger–Nuttall law spanning over twenty orders of magnitude in half-life — for a 5 MeV alpha, T can be as small as 10⁻³⁸, yet the particle "attempts" the barrier ~10²¹ times per second, so decay still happens on human timescales for many isotopes.

Where WKB breaks down

WKB assumes the local de Broglie wavelength varies slowly. It fails right at the turning points, where κ(x) → 0 and the naive WKB wavefunction diverges — fixed with Airy-function connection formulas — and for very thin or low barriers (κL ≲ 1), where the exact solution should be used instead. Despite these caveats, WKB is accurate to a few percent for essentially every barrier relevant to nuclear and semiconductor physics: alpha decay, STM tunnelling currents ∝ e^(−2κd), Fowler-Nordheim tunnelling in flash memory, and negative-differential-resistance tunnel diodes all rest on this single integral.

Frequently asked questions

What is the WKB approximation used for?

The WKB (Wentzel–Kramers–Brillouin) approximation extends the exact rectangular-barrier tunnelling result to arbitrarily shaped potentials V(x). It treats the barrier as locally rectangular, defining a position-dependent decay constant κ(x) and integrating it across the classically forbidden region to get the transmission probability.

What is the Gamow tunnelling integral?

T ≈ exp[−2∫ₐᵇ κ(x)dx] between classical turning points a and b. George Gamow used exactly this formula in 1928 to explain alpha decay, correctly predicting the Geiger–Nuttall relationship between decay half-life and alpha particle energy across more than twenty orders of magnitude.

When does the WKB approximation break down?

WKB fails near the classical turning points, where κ(x) → 0 and the approximate wavefunction diverges — fixed with Airy-function connection formulas — and for very thin or low barriers (κL ≲ 1), where the exponential shortcut loses accuracy and the exact or numerical solution should be used instead.

Try it live

Everything above runs in your browser — open Quantum Tunnelling and compute exact transmission coefficients for square, double-barrier and step potentials with the transfer matrix method. Nothing is installed, nothing is uploaded.

▶ Open Quantum Tunnelling simulation

What did you find?

Add reproduction steps (optional)