advanced28 min studyLesson 386 in Physics

Tunneling splits a doublet without creating a decay rate

Compare forbidden-region actions, coherent doublet oscillation, and Euclidean instanton estimates.

On this page

What you’ll understand

  • Distinguish a tunneling amplitude exponent from a transmission probability exponent.
  • Compute a symmetric-doublet splitting and its coherent transfer time.
  • State the role of instanton action and fluctuation prefactors.

A particle localized in one well of a symmetric double well is generally not an exact stationary state. Weak overlap through the barrier combines left and right states into a symmetric and an antisymmetric pair with slightly different energies. Their splitting can be exponentially small even when each separate well looks almost classical. We will distinguish this coherent doublet physics from an irreversible escape rate and explain how WKB and imaginary-time methods estimate the small coupling.

A two-state projection exposes the splitting

Assume two identical wells have one low-lying localized state each, with common approximate energy E_w. If their overlap is weak and higher local excitations are well separated, project the full Hamiltonian onto those two states, orthonormalizing them as needed. Choose their phases so the effective off-diagonal element is negative with magnitude J. The symmetric combination then has lower energy E_w−J, while the antisymmetric one has E_w+J. The doublet splitting is twice the coupling, not the coupling itself.

An initially localized left state is an equal superposition of these two energy eigenstates. Their phases evolve at slightly different rates, causing coherent transfer between the wells. The right-well probability is a sine squared whose argument contains Jt/ħ, and the first complete transfer occurs after πħ/(2J). This closed two-state model predicts recurring oscillations. Calling that inverse time a universal tunneling decay rate would discard the coherence and the recurrence without specifying a bath, an absorbing region, or another irreversible mechanism.

Heff=(Ew−J−JEw),ΔE=2J,PR(t)=sin⁡2 ⁣(Jtℏ)H_{\mathrm{eff}}=\begin{pmatrix}E_w&-J\\-J&E_w\end{pmatrix},\qquad \Delta E=2J,\qquad P_R(t)=\sin^2\!\left(\frac{Jt}{\hbar}\right)
E_w and J>0 have energy units. The probability assumes the initial state is localized left in an isolated symmetric doublet, with higher levels and overlap corrections controlled.↔ Scroll wide equations horizontally.

The forbidden action controls an exponential amplitude

In a smooth barrier with two relevant turning points, the decaying WKB tail carries an amplitude suppression exp(−S/ħ), where S is the integral of the positive forbidden momentum between the points at the localized-state energy. The overlap coupling and doublet splitting inherit an exponential of this kind. A transmission probability through a similar barrier instead squares an amplitude and has the leading exponential exp(−2S/ħ). Confusing the two can change an estimate by exponentially many orders of magnitude.

The prefactor multiplying the splitting has energy units and contains normalization, well curvature, and matching information. It is not determined by the exponent alone. Low-lying ground doublets require particularly careful matching near quadratic well minima, and textbook prefactors can fail even when the leading logarithm is correct. Near the barrier top or when the wells are strongly coupled, the assumption of an isolated localized doublet and a large forbidden action can both break down.

S(Ew)=∫x1x22m(V(x)−Ew) dx,ΔE∼Ae−S(Ew)/ℏ,T∼e−2S/ℏS(E_w)=\int_{x_1}^{x_2}\sqrt{2m(V(x)-E_w)}\,dx,\qquad \Delta E\sim A e^{-S(E_w)/\hbar},\qquad \mathcal T\sim e^{-2S/\hbar}
The integral is over the forbidden barrier, so V>E_w; S has action units and A has energy units. These are leading semiclassical exponential estimates, not universal prefactor formulas.↔ Scroll wide equations horizontally.

An instanton is an imaginary-time saddle

In a path integral continued to imaginary time, the dominant tunneling path connects the two minima through a Euclidean classical solution. If the potential minimum is set to zero, its zero-Euclidean-energy action is the integral of √(2mV) from one minimum to the other. This path is not a real-time trajectory of a particle climbing an energetically forbidden hill. Its action organizes a saddle-point approximation, while fluctuations around it, including a translation zero mode, determine additional factors in the splitting.

The zero-energy instanton action and the WKB action evaluated at the localized-state energy use different endpoints and energies. Their exponential and prefactor organization must be matched consistently before comparing precise formulas. The quantum lab can illustrate a coherent doublet state: Reset and set both controls to θ=90° and φ=0°, predicting Bloch x=1; Reset and set θ=90° and φ=180°, predicting x=−1. If the basis labels the parity eigenstates, these opposite x directions represent localized combinations. Playback adds a fixed phase rate; it does not calculate a barrier action, choose a physical splitting, or produce irreversible decay.

A coherent transfer time from a measured coupling

In an isolated symmetric doublet, take J=0.02 energy unit and ħ=1 energy-unit times time-unit. Starting localized in the left well, find the energy splitting, the first time of complete right-well transfer, and the population recurrence period.

  1. The symmetric and antisymmetric energies differ by 2J=0.04 energy unit. A common E_w shifts both energies and does not change the relative-phase evolution.

  2. The first maximum of sin²(Jt/ħ) occurs at Jt/ħ=π/2. Thus the first complete transfer is at t=π/(0.04)=25π time units, approximately 78.5.

  3. The sine squared repeats when its argument increases by π, so the population recurrence period is πħ/J=50π time units. At that time the system has returned to its initial localized population.

The result

The splitting is 0.04 energy unit, the first transfer time is 25π time units, and the population period is 50π time units.

Explore the model

Interactive lab

Interactive lab · quantum

Localized combinations of parity states

Reset and set BOTH controls to θ=90° and φ=0°, predicting Bloch x=1. Reset and set θ=90° and φ=180°, predicting x=−1. Interpreting the basis as parity states makes these opposite localized combinations; no barrier or splitting is calculated.

Localized combinations of parity statesA normalized two-level pure state, not an interacting field simulation. Playback adds one radian per second to the relative phase; basis probabilities stay fixed. Current readings: Probability |0⟩: 71.1%; Probability |1⟩: 28.9%; Bloch (x, y, z): 0.78, 0.45, 0.42. All parameters are adjustable with the labeled controls below.xyz|0⟩|1⟩Basis probabilitiesP(0)71.1%P(1)28.9%
Bloch vector / P(0)P(1)Phase latitude
Probability |0⟩
71.1%
Probability |1⟩
28.9%
Bloch (x, y, z)
0.78, 0.45, 0.42
t = 0 s

A normalized two-level pure state, not an interacting field simulation. Playback adds one radian per second to the relative phase; basis probabilities stay fixed. Parameter changes pause playback and recompute the model at the current elapsed time. Reset restores both parameters and returns time to zero. Playback starts paused. Adjust either parameter, then choose Play or Step.

Look forA two-level phase difference can encode coherent localization, but the lab has only its fixed phase-precession rate.

Check your understanding

Which leading exponential usually describes a weak doublet splitting generated by a smooth forbidden barrier action S?

Try it yourself

Two high barriers have actions S and S+2ħ, with equal splitting prefactors and otherwise identical isolated-doublet assumptions. Find the ratio of the second splitting to the first, the ratio of first-transfer times, and the corresponding leading transmission-probability ratio.

Take a moment to make a prediction or work it out on paper. Open a hint when you need one.

Give me a hint

The splitting is amplitude-like, transfer time is inversely proportional to splitting, and transmission probability squares the amplitude.

Compare with the solution

The splitting ratio is exp(−2)≈0.135. The first-transfer time ratio is its reciprocal exp(2)≈7.39. The leading transmission-probability ratio is exp(−4)≈0.0183. Equal prefactors are an explicit comparison assumption; changing barrier shape can also change them, so the exponential ratios are not a general exact formula.

Take it with you

  • A symmetric isolated doublet has splitting 2J and coherent recurring transfer.
  • Splittings and transmission probabilities have different leading exponential factors.
  • Instanton actions require imaginary-time interpretation and consistent fluctuation prefactors.

Read further

Original sources and deeper explanations to follow at your own pace.

Optional progress, saved in this browser when available. Revisit any lesson at any time.

Lesson not marked complete.