33
Part One · Quantum mechanics

Infinite and finite square wells

A vial of tiny crystals, all made of the same material, glows red, orange, green or blue depending only on how big the crystals are. Nothing about their chemistry differs; only their size. The reason is the simplest problem in quantum mechanics, a particle trapped in a box, and it won the 2023 Nobel Prize in Chemistry.

On the plate: the allowed states of a particle in a box. There are only certain energies, and between them, none.

Why energy comes in steps

Put a particle between two impenetrable walls a distance LL apart. Inside, there is no force, so the time-independent Schrödinger equation (Chapter 31) is −ℏ22mϕ′′=Eϕ-\frac{\hbar^2}{2m}\phi'' = E\phi, solved by sines and cosines. At the walls the wavefunction must vanish, so only waves that fit a whole number of half-wavelengths survive, exactly like a guitar string:

ϕn(x)=2Lsin⁡nπxL,En=n2h28mL2,n=1,2,3,…\phi_n(x) = \sqrt{\tfrac{2}{L}}\sin\frac{n\pi x}{L}, \qquad E_n = \frac{n^2h^2}{8mL^2}, \quad n = 1, 2, 3, \dots

Step through 1. Each step adds a node, and the energy climbs as n2n^2. Quantization here is not imposed; it is forced by the boundary conditions. Established

The zero-point energy

There is no state with n=0n = 0: that would be ϕ=0\phi = 0, no particle at all. The lowest energy is E1>0E_1 > 0. A confined particle can never sit still, because confining it to LL forces a momentum spread of order ℏ/L\hbar/L (Chapter 30).

Make it real. An electron in a box 1.0 nm wide has E1=E_1 = 0.376 eV. The jump from n=1n = 1 to n=2n = 2 absorbs light of wavelength 1100 nm. Atoms and molecules are a few tenths of a nanometre to a few nanometres across, which is why their energies are electron-volts and their light is visible or near it.

Walls that are not infinite

Real wells have finite depth. Set the depth to 10 and the width to 1.0 (half-width, in units with ℏ=m=1\hbar = m = 1). The plate solves for the bound states exactly.

Three things change. The wavefunction leaks into the walls, decaying as e−κ∣x∣e^{-\kappa|x|} with κ=2m(V0−E)/ℏ\kappa = \sqrt{2m(V_0 - E)}/\hbar: the particle has a real chance of being found where classically it has too little energy to go. The levels sit lower than in the infinite box, because the wave has more room. And there are only finitely many bound states, but always at least one in one dimension, however shallow the well.

DeeperMatch ϕ\phi and ϕ′\phi' at x=ax = a. With z=kaz = ka and z0=a2mV0/ℏz_0 = a\sqrt{2mV_0}/\hbar, even states satisfy ztan⁡z=z02−z2z\tan z = \sqrt{z_0^2 - z^2} and odd states −zcot⁡z=z02−z2-z\cot z = \sqrt{z_0^2 - z^2}. The number of bound states is the number of half-multiples of π\pi below z0z_0, rounded up. For depth 10 and half-width 1, z0=4.47z_0 = 4.47 and there are three.

Colours from size

A quantum dot is a crystal of semiconductor a few nanometres across. An electron and the hole it leaves behind are both confined, and the energy of the light they emit when they recombine is the bulk band gap plus their confinement energies, which grow as 1/R21/R^2. Choose a cadmium selenide dot of radius , or .

Smaller dots, bluer light. Louis Brus worked out this model in 1983, Alexei Ekimov had seen the effect in coloured glass, and Moungi Bawendi learned in 1993 to make the dots with precise sizes. They shared the 2023 Nobel Prize in Chemistry; quantum dots now set the colours of some television screens and are being developed as bright labels for imaging tissue. Established

NoteThe plate uses Brus’s simple formula, with effective masses for the electron and hole and a correction for their attraction. It gets the trend and the rough colours right; accurate predictions for small dots need more detailed models.

A box made of a molecule

In the late 1940s Hans Kuhn treated the electrons of a long dye molecule as particles in a box the length of its chain. Fill the levels two electrons at a time; the colour is the jump from the highest filled level to the first empty one, ΔE=(N+1) h2/8mL2\Delta E = (N + 1)\,h^2/8mL^2 for NN electrons. Try 10 electrons, with the chain about 0.14 nm longer per electron.

Longer chains, redder absorption. Ten electrons over 1.4 nm predicts about 590 nm, and real dyes of this family absorb in that region. A biomedical engineer meets this in every fluorescent label: the cyanine dyes used in imaging, such as Cy3 and Cy5, differ by a two-carbon extension of the chain, and their colours shift accordingly.

What you now have

Quantization from boundary conditions, En∝n2/L2E_n \propto n^2/L^2, zero-point energy, penetration into finite walls, at least one bound state in any one-dimensional well, quantum-dot colours and Kuhn’s dye model.

· · ·

Three questions to carry with you.

Why is there no n = 0 state in a box?tap
It would be ψ = 0 everywhere, meaning no particle. Confinement forces a minimum kinetic energy.
Why do smaller quantum dots emit bluer light?tap
Confinement energy grows as 1/R², raising the energy of the emitted photon.
Can a particle be found inside the wall of a finite well?tap
Yes. The wavefunction decays exponentially into the wall rather than stopping, so there is a nonzero probability there.

And one to work out. The electron in a 1 nm box has E1=0.376E_1 = 0.376 eV. What is E1E_1 for a 0.5 nm box, in eV?

E₁ =eV

Sources: L. E. Brus, J. Chem. Phys. 79, 5566 (1983); C. B. Murray, D. J. Norris and M. G. Bawendi, J. Am. Chem. Soc. 115, 8706 (1993); Nobel Prize in Chemistry 2023 (Bawendi, Brus, Ekimov). H. Kuhn, J. Chem. Phys. 17, 1198 (1949). Checked by code: E₁(1 nm) = 0.376 eV; finite well (depth 10, half-width 1) has 3 bound states at −9.18, −6.78, −3.05; Kuhn N = 10, L = 1.4 nm → 2.11 eV, 587 nm; Brus CdSe radii 1.5, 2.5, 4 nm → about 380, 550, 650 nm.

Next · Chapter 34
Tunneling and barriers

A particle walks through a wall it doesn't have the energy to climb. Your phone's flash memory does this all the time.

Turn the page