Ground-state energy
★☆☆An atom vibrates in a crystal at the angular frequency \(\omega = 1.0\times10^{14}\ \mathrm{rad/s}\). What is its zero-point energy, in eV?
Chapter 6 · Lectures 6 h + tutorials 3 h
A mass on a spring, a vibrating molecule, a mode of the electromagnetic field: the same Hamiltonian, the same energy ladder with equally spaced rungs. Two operators are enough to move along it.
A mass on a spring, a vibrating molecule, a mode of the electromagnetic field: all of them obey the same Hamiltonian.
We solve it with two ladder operators. The creation operator, a dagger, moves up one rung; the annihilation operator, a, moves down one. Their commutator equals one. The energies are then h-bar omega times n plus one half: equally spaced levels.
The lowest level is not zero. It equals one half h-bar omega. This is the zero-point energy, imposed by the uncertainty relation. A particle at rest exactly at the bottom of the well would have a perfectly known position and momentum.
The wave functions of the Fock states have n nodes, and they spill slightly into the regions forbidden to a classical particle.
Coherent states are the closest to classical motion. Their wave packet oscillates in the well without changing shape. The number of quanta follows a Poisson distribution, whose mean is the modulus of alpha, squared. This is the state of laser light. Gravitational-wave detectors go further: they inject squeezed light, which reduces the noise on one quantity at the cost of the other, without ever violating Heisenberg.
Classically, a particle of mass \(m\) pulled back by a force \(F = -kx\) oscillates at the angular frequency \(\omega = \sqrt{k/m}\). Its energy can take any positive value. It moves back and forth between two turning points \(\pm x_0\), and it is near these edges, where it slows down, that we are most likely to find it.
With the length \(\ell = \sqrt{\hbar/m\omega}\), we set \(\hat X = X/\ell\) and \(\hat P = \ell P/\hbar\), which satisfy \([\hat X, \hat P] = i\). Then
We find \([a, a^\dagger] = 1\) and \(H = \hbar\omega\left(N + \tfrac12\right)\) with the Hermitian number operator \(N = a^\dagger a\). \(a\) and \(a^\dagger\) are not Hermitian: they are not observables.
The energies are \(E_n = \hbar\omega\left(n + \tfrac12\right)\), \(n = 0, 1, 2,\dots\): equally spaced rungs, separated by \(\hbar\omega\). The lowest one, \(E_0 = \tfrac12\hbar\omega\), is the zero-point energy.
Before running the lab: write down your prediction, run it, compare, then explain the gap in one sentence. What do \(a\ket0\) and \(a^\dagger\ket0\) give? What factor do you expect for \(a\ket n\)?
Start from \(\ket3\) and compare \(a\,a^\dagger\ket3 = 4\ket3\) with \(a^\dagger a\ket3 = 3\ket3\): the difference is 1, which is \([a, a^\dagger] = 1\). Then apply \(a\) to the vacuum: you get the zero vector, not the state \(\ket0\).
The lecture notes attribute it to a “principle of minimal action”. The name is misleading: the principle of least action is a classical variational principle, unrelated to this. The correct argument uses the Heisenberg inequality. In an eigenstate, \(\langle X\rangle = \langle P\rangle = 0\), so
the minimum being reached for \((\Delta X)^2 = \hbar/2m\omega\), that is, by the Gaussian \(\psi_0\).
In 1900, Planck derived the black-body law by assuming that the material oscillators of the cavity wall exchange energy only in multiples of \(h\nu\). At first he saw this as a calculational device. It was Einstein, in 1905, who assigned quanta to light itself. Planck’s levels were \(nh\nu\): the half-quantum of the zero point appears only with the quantum mechanics of 1925–1926.
An atom vibrates in a crystal at the angular frequency \(\omega = 1.0\times10^{14}\ \mathrm{rad/s}\). What is its zero-point energy, in eV?
The HCl molecule absorbs in the infrared at the wavenumber \(\tilde\nu = 2886\ \mathrm{cm^{-1}}\), which corresponds to the \(n = 0 \to 1\) jump of its vibration. What is the level spacing \(\hbar\omega\), in eV?
At 300 K, \(k_BT \approx 0.026\ \mathrm{eV}\), about 14 times less than \(\hbar\omega\). The Boltzmann factor \(e^{-14}\approx 10^{-6}\) shows that almost all molecules are in \(\ket0\) at room temperature: the vibration is “frozen”. The harmonic approximation holds for the first few levels; higher up, the real levels get closer together.
We write \((a^\dagger)^2\ket1 = c\,\ket3\). What is \(c\)? (decimal or exact form, e.g. sqrt(5))
What is \(a\ket0\)?
One is tempted to look for a direct meaning, as for \(X\) and \(P\). But \(a^\dagger\neq a\): they are not Hermitian, so they represent no measurable quantity. They are tools that move up or down the ladder of levels. The physical quantities built from them are Hermitian: \(H=\hbar\omega(a^\dagger a+\tfrac12)\), \(N=a^\dagger a\).
With the reduced variable \(u = x/\ell\), the operator \(a\) becomes \(\tfrac{1}{\sqrt2}\left(u + \tfrac{d}{du}\right)\). The condition \(a\ket0 = 0\) is then a first-order equation, \(\psi_0' = -u\,\psi_0\), whose solution is a Gaussian. The excited states are obtained by applying \(a^\dagger\).
with \(H_0 = 1\), \(H_1 = 2u\) and the recurrence
\(\psi_n\) has \(n\) nodes and the parity of \(n\): \(\psi_n(-u) = (-1)^n\psi_n(u)\). The normalisation is with respect to \(u\); in the variable \(x\), divide by \(\sqrt\ell\).
The lecture notes state the Hermite recurrence “for \(n \ge 2\)”. It holds from \(n = 1\) onwards: it gives \(H_2 = 2u\cdot 2u - 2 = 4u^2 - 2\), which could not be obtained otherwise from \(H_0\) and \(H_1\).
quantumclassical density
Tap a level in the diagram to select it.
The hatched regions are classically forbidden: \(|x| > x_0\), with \(\tfrac12 m\omega^2x_0^2 = E_n\), that is \(u_0 = \sqrt{2n+1}\). For \(n = 0\), the particle is found there with probability \(\mathrm{erfc}(1) \approx 15.7\%\). As \(n\) grows, \(|\psi_n|^2\) oscillates around the classical density \(\rho(x) = 1/\big(\pi\sqrt{x_0^2 - x^2}\big)\): this is the correspondence principle.
What can be said about \(\psi_7\) at the centre of the well, and about \(\langle x\rangle\) in the state \(\ket7\)?
Compute \(\Delta x\,\Delta p\) in the state \(\ket2\), in units of \(\hbar\). Only \(\ket0\) saturates the inequality.
Write \(x = \tfrac{\ell}{\sqrt2}(a + a^\dagger)\). In \(\langle n|(a+a^\dagger)^2|n\rangle\), only the terms \(aa^\dagger\) and \(a^\dagger a\) survive: \((n+1) + n\).
A Fock state is stationary: its density does not move. A superposition of two levels, on the other hand, evolves. Each component rotates at its own frequency \(E_n/\hbar\), and only the difference matters:
|ψ(x,t)|²average over one period
\(T = 2\pi/\omega\). With \(\ket0\) and \(\ket1\), the cloud sways from left to right and \(\langle x\rangle\) oscillates at the angular frequency \(\omega\) of the classical spring. With \(\ket0\) and \(\ket2\), the density “breathes” at \(2\omega\) but stays symmetric: \(\langle x\rangle = 0\), because \(x\) only connects neighbouring levels.
We prepare \((\ket1 + \ket4)/\sqrt2\). At what angular frequency does the probability density oscillate, and what does \(\langle x\rangle\) do?
In every Fock state, \(\langle X\rangle = 0\). Yet a laser beam, or a pendulum, oscillates visibly. We therefore need states that mix several \(\ket n\) with well-tuned phases. Schrödinger wrote them down in 1926; Glauber made them the foundation of quantum optics in the 1960s (Nobel Prize 2005).
\(\ket\alpha\) is an eigenstate of the annihilation operator, \(a\ket\alpha = \alpha\ket\alpha\) with \(\alpha\in\mathbb C\):
The displacement operator \(D(\alpha) = e^{\alpha a^\dagger - \alpha^* a}\) is unitary. \(\ket\alpha\) is therefore the vacuum, displaced without deformation to the point \(\big(\langle\hat X\rangle, \langle\hat P\rangle\big) = \sqrt2\,(\mathrm{Re}\,\alpha, \mathrm{Im}\,\alpha)\) of phase space.
Poisson P(n)measured frequencies
Top: \(|\psi(x,t)|^2 = \frac{1}{\sqrt\pi\,\ell}\exp\!\big[-(x - x_c(t))^2/\ell^2\big]\) with \(x_c(t) = \sqrt2\,|\alpha|\,\ell\cos\omega t\): the width never changes. Bottom: the same state in phase space, an uncertainty disc rotating on the classical circle. At \(|\alpha| = 0\), we recover the motionless vacuum.
An oscillator is in the coherent state \(\ket\alpha\) with \(\alpha = 3\,e^{i\pi/4}\). What is \(\langle N\rangle\)? And \(\Delta N\) (check it in lab 6.4)?
For \(\langle N\rangle = 2\), what is the probability that a measurement of \(N\) gives \(0\)? (decimal answer)
Schrödinger introduced coherent states as early as 1926, but they long remained a mathematical curiosity. In the 1960s Roy Glauber showed that they describe the light of an ideal laser and set it apart from the chaotic light of a bulb, a statistical mixture. This work founded modern quantum optics and earned him a share of the 2005 Nobel Prize in Physics.
Imagined dialogue: how does a wave emerge from the vacuum? The lines summarise an argument; they are not quotations.
The sceptic: “My laser emits billions of photons per second; their number is well defined. Why a state where it is uncertain?”
Glauber: “In a state \(\ket n\) the mean field is zero, yet your beam visibly oscillates. To get an oscillation with a stable phase you must superpose states \(\ket n\), and so give up a definite photon number.”
The sceptic: “A compromise, then?”
Glauber: “Yes: we lose certainty about the number and gain a stable phase. That is a coherent state.”
One mode of the electric field is written \(E = i\xi\left(a\,e^{-i\omega t} - a^\dagger e^{i\omega t}\right)\), with \(\xi\) real. (1) Show that \(\langle E\rangle = 0\) in every state \(\ket n\). (2) Compute \(\langle E\rangle\) in \(\ket\alpha\) with \(\alpha = |\alpha|e^{i\theta}\). (3) Show that \((\Delta E)^2 = \xi^2\) in \(\ket\alpha\), whatever \(\alpha\), and comment on \(\Delta E/|\langle E\rangle|_{\max}\) for an intense laser.
For (1), \(\langle n|a|n\rangle = \langle n|a^\dagger|n\rangle = 0\). For (3), rewrite \(\langle\alpha|a^\dagger a|\alpha\rangle\) and \(\langle\alpha|aa^\dagger|\alpha\rangle = |\alpha|^2 + 1\).
(1) \(a\) and \(a^\dagger\) change \(n\) by one unit, so their mean values in \(\ket n\) are zero. A Fock state has a definite amplitude but no phase: the mean field is zero.
(2) \(\langle E\rangle = i\xi\left(\alpha e^{-i\omega t} - \alpha^* e^{i\omega t}\right) = 2\xi|\alpha|\sin(\omega t - \theta)\): a sinusoidal wave of amplitude \(2\xi|\alpha|\), with its phase set by \(\theta\).
(3) \(\langle E^2\rangle = -\xi^2\langle(ae^{-i\omega t} - a^\dagger e^{i\omega t})^2\rangle = \xi^2\left[2|\alpha|^2 + 1 - \alpha^2e^{-2i\omega t} - \alpha^{*2}e^{2i\omega t}\right]\), and subtracting \(\langle E\rangle^2\) removes everything except \(\xi^2\). Hence \(\Delta E = \xi\), the same value as in the vacuum, and \(\Delta E/(2\xi|\alpha|) = 1/(2\sqrt{\langle N\rangle})\to 0\) for an intense beam: the field becomes classical.
The lecture notes describe the vacuum as filled with “virtual pairs that appear and disappear”. This is a popular-science image, not an observation. The calculation rests on the sum of the energies \(\tfrac12\hbar\omega\) of the modes allowed between the plates. In 1D, \(E(L) = -\pi\hbar c/(24L)\) gives \(F = -dE/dL = -\pi\hbar c/(24L^2)\): an attractive force.
Analogy: “a steady rhythm played without interruption carries a whole group along (gnawa, vodun); likewise a coherent state keeps a stable phase”. Which assessment is correct?
The ground-state energy of the oscillator is \(\hbar\omega/2\), not zero. Why? Discuss in pairs before answering.
A coherent state has \(\langle N\rangle = 9\). What is \(\Delta N\)? Discuss in pairs before answering.
An ideal single-mode laser pulse is described by a coherent state \(\ket\alpha\) with \(\langle N\rangle=|\alpha|^2=4\). A perfect photodetector (100 % efficiency) records no photon at all with what probability? Discuss in pairs before answering.
Choose one analogy from this chapter. With the person’s consent, who may be an elder, a practitioner, a craftsperson or a musician, ask: (1) how they present this notion, in what language and with what words; (2) under what conditions this knowledge is passed on; (3) what, in their view, does not translate into physics. Write down the words in the local language, return your file to the person, then present both systems, each coherent by its own criteria, without concluding that one proves the other.
Main ideas.
Formulas.
Pitfalls. Believing Planck proved light quanta in 1900 (Einstein postulates them in 1905); confusing a coherent state, a pure state with a definite phase, with a thermal mixture; believing a squeezed state violates Heisenberg (it redistributes the uncertainty between two quadratures).
| Term | Terme français | Definition |
|---|---|---|
| Zero-point energy | Énergie du point zéro | \(E_0=\tfrac12\hbar\omega\), the minimum set by the uncertainty relation. |
| Coherent states | États cohérents | Eigenstates of the annihilation operator, \(a\ket\alpha=\alpha\ket\alpha\), with minimum uncertainty. |
| Squeezed states | États comprimés | Reduced fluctuations in one quadrature, increased in the conjugate one. |
| Fock states | États de Fock | Eigenstates of \(N\) with a fixed number of quanta: \(N\ket n=n\ket n\). |
| Ladder operators | Opérateurs d’échelle | \(a\) and \(a^\dagger\), with \([a,a^\dagger]=1\). |
| Hermite polynomials | Polynômes de Hermite | Polynomial part of the eigenfunctions \(\psi_n(x)\). |
Squeezed states of light, seen here as oscillator states, are injected into gravitational-wave detectors to get below shot noise. Single-photon detectors make it possible to count weak light fluxes and are used in some QKD systems.
John Clarke, Michel Devoret and John Martinis were awarded the prize “for the discovery of macroscopic quantum mechanical tunnelling and energy quantisation in an electric circuit”. In their experiments of the 1980s, a superconducting circuit with a Josephson junction shows discrete energy levels, like the oscillator of this chapter. The junction makes the oscillator anharmonic: the rungs are no longer equally spaced, which makes it possible to address only the first two. This is the principle of today’s superconducting qubits.
Press release of the Nobel Committee, 7 October 2025 — nobelprize.orgThe shot noise of a laser is that of the coherent state: \(\Delta N = \sqrt{\langle N\rangle}\). Since May 2023, LIGO has been injecting into its interferometers a squeezed vacuum whose noise-ellipse orientation depends on frequency, thanks to a 300 m filter cavity. Quantum noise drops across the whole band: shot noise at high frequency, radiation pressure at low frequency. The expected gain is about 60% more mergers detected.
LIGO, press release of 23 October 2023; Phys. Rev. X (2023)The fourth observing run of LIGO, Virgo and KAGRA ended on 18 November 2025, after two and a half years. About 250 candidate signals were identified in real time, and the analysis of the first third has already yielded 128 significant events, compared with 90 for the three previous runs combined. The signal GW250114, received on 14 January 2025, has a signal-to-noise ratio of 80 and made it possible to test Hawking’s area theorem. An intermediate six-month run is planned from November 2026, a schedule that may change.
LIGO, 18 November 2025; Phys. Rev. Lett., September 2025; observing.docs.ligo.org/planAt ETH Zurich, a 16 µg mechanical resonator, about \(10^{17}\) atoms, was prepared in superpositions of motional states, known as “Schrödinger cat” states. This resonator is a macroscopic quantum harmonic oscillator: it displays the Fock states and coherent states of this chapter, and their superpositions.
Bild et al., Science (2023), ETH Zurich