QUANTUM PHYSICS · EXPERIMENT 004

Quantum tunnelling

Launch a quantum wave at the walls and find the settings that let it pass through.

Ready to launcht = 0.0
A wave packet moving through potential barriers
Transmitted76%Reflected 24% · expected result

The quantum wave reflects many times between the walls. At certain distances, those reflections help it pass through.

The line shows |ψ|². Percentages include the wave packet’s finite momentum spread.
Theory behind the experimentHow a quantum particle gets through a barrier and why two barriers sometimes let it through especially wellPlain explanation, equations and how the model works

In plain words

Picture a particle heading toward a region where the potential energy is higher than its total energy. In classical mechanics it can’t get in; it simply bounces back. In quantum mechanics it still has a chance to turn up on the other side.

The reason is that a quantum state is described by a wave function. In the region a classical particle can’t enter, the wave function doesn’t stop dead; it fades quickly. If the barrier is thin enough, some of the wave is left on the far side. That’s tunnelling. The particle doesn’t punch a hole or “borrow energy”: the barrier stays the same and energy is conserved.

On screen you see a wave packet, a blob of probability. When it reaches the barrier it splits into a reflected and a transmitted part. These aren’t halves of a particle. Measure where the particle is and you find all of it in one place; the picture shows the odds of what you’d get over many identical runs.

Below its height, a thick barrier usually lets less through than a thin one. Two barriers are more interesting: the wave bounces back and forth between them, and those bounces can add up so that more gets through. A chain of barriers shows bands of strong and weak transmission. What matters is not just the barriers themselves but the interference of waves between them.

Try it yourself

  1. In single-barrier mode keep the energy below the barrier height and change the thickness. Watch the transmission and how the packet splits.
  2. In double-barrier mode change the gap and find the peak on the chart. A second barrier doesn’t necessarily just get in the way.
  3. In the periodic structure change the period. The same energy lands in a band of strong or weak transmission. The chain is finite, so these are its transmission bands, not the exact band gaps of an infinite crystal.

Physics and equations

How the wave function changes in time in one dimension is set by the Schrödinger equation:

iℏ ∂ψ∂t=−ℏ22m ∂2ψ∂x2+V(x) ψi\hbar\,\frac{\partial\psi}{\partial t}=-\frac{\hbar^2}{2m}\,\frac{\partial^2\psi}{\partial x^2}+V(x)\,\psi

Here ψ(x,t)\psi(x,t) is the wave function, V(x)V(x) the potential energy, mm the particle’s mass, ℏ\hbar Planck’s constant divided by 2π2\pi, and ii the imaginary unit. The probability of finding the particle in a small interval dxdx is ∣ψ∣2dx|\psi|^2dx:

ρ(x,t)=∣ψ(x,t)∣2,∫−∞∞ρ dx=1\rho(x,t)=|\psi(x,t)|^2,\qquad \int_{-\infty}^{\infty}\rho\,dx=1

The line on screen is this density ρ\rho. Its height at one point isn’t the transmission percentage; the percentage refers to the whole part of the packet that ends up past the barrier.

Inside a rectangular barrier of height V0V_0 with energy E<V0E<V_0, the solution contains exponentials with the scale

κ=2m (V0−E)ℏ\kappa=\frac{\sqrt{2m\,(V_0-E)}}{\hbar}

The larger κ\kappa, the faster the wave fades inside the barrier. For a barrier of width aa, the transmitted fraction of a wave with one exact energy 0<E<V00<E<V_0 is

T(E)=[1+V02sinh⁡2(κa)4E (V0−E)]−1\mathcal T(E)=\left[1+\frac{V_0^2\sinh^2(\kappa a)}{4E\,(V_0-E)}\right]^{-1}

The formula holds when the potential is zero on both sides of the barrier. For large κa\kappa a the thickness dependence is mainly exponential: T∝e−2κa\mathcal T\propto e^{-2\kappa a}. That’s an approximation for a thick barrier and doesn’t work for a system of several barriers.

For two barriers, the phase the wave picks up travelling between them matters. At certain gaps you get resonant transmission, so the chance of getting through doesn’t have to drop as the gap grows.

A packet is made of waves with different wave numbers kk, and so different energies. The overall transmission is an average:

Tpacket=∫P(k) T ⁣(ℏ2k22m)dk\mathcal T_{\text{packet}}=\int P(k)\,\mathcal T\!\left(\frac{\hbar^2k^2}{2m}\right)dk
∫P(k) dk=1\int P(k)\,dk=1

P(k)P(k) is the normalised distribution of wave numbers in the packet, and kk has units of inverse length. Averaging smooths out narrow resonances, so the packet’s peaks are lower than for a wave with a single exact energy.

How the model works

The model is one-dimensional with fixed rectangular barriers, in dimensionless units where ℏ=m=1\hbar=m=1. Don’t read the slider values as electronvolts or nanometres; that would need physical scales chosen separately.

The packet’s motion is computed numerically with the split-step Fourier method. The transmission percentage and the chart are computed separately: transmission at each energy comes from transfer matrices and is averaged over the packet’s momentum distribution. So the percentage isn’t a count of particles that have just crossed the barrier in the picture.

The packet has a finite width in both position and momentum. Its position standard deviation is 6 model units. The calculation runs on a finite grid for a limited time. It illustrates wave-packet scattering; it isn’t a model of a specific material or device.

Raise the energy above the barrier and you get over-barrier scattering, which isn’t called tunnelling. Partial reflection can still happen because the potential changes abruptly.

Sources and further reading

HyperPhysics: Barrier Penetration ↗: tunnelling and the transmission formula for a rectangular barrier.