28
Part One · Quantum mechanics

Dirac notation

The state of one qubit, the state of an electron in a hydrogen atom, and the state of every field in the universe can all be written with the same three symbols: a vertical bar, a name, and an angle bracket. Paul Dirac introduced this notation in 1939, and it has never been replaced, because it lets you calculate without first deciding what kind of space you are in.

This chapter teaches you to read and write it fluently. Each idea is shown three ways at once on the plate: as a ket, as a column of numbers, and as a picture.

Kets and bras

A state is a ket, ∣ψ⟩|\psi\rangle. In a chosen basis it is a column of complex numbers. Its partner, the bra ⟨ψ∣\langle\psi|, is the row of complex conjugates. Put a bra and a ket together and you get a number, the inner product ⟨ϕ∣ψ⟩\langle\phi|\psi\rangle: a “bra-ket”, which is where the names come from.

Take a qubit at angle 20 °, ∣ψ⟩=cos⁡θ∣0⟩+sin⁡θ∣1⟩|\psi\rangle = \cos\theta|0\rangle + \sin\theta|1\rangle. The plate shows the ket, its column of numbers, and the arrow in the plane.

Projectors and the identity

Put them the other way round, ket then bra, and you get not a number but an operator: ∣a⟩⟨a∣|a\rangle\langle a|, which projects any state onto ∣a⟩|a\rangle. Add the projectors for a complete set of perpendicular basis states and you get the identity, the operator that does nothing:

∑a∣a⟩⟨a∣=1\sum_a |a\rangle\langle a| = \mathbb 1

This completeness relation is the most useful line in quantum mechanics. Insert it anywhere and you have expanded in a basis: ∣ψ⟩=∑a∣a⟩⟨a∣ψ⟩|\psi\rangle = \sum_a |a\rangle\langle a|\psi\rangle, with the numbers ⟨a∣ψ⟩\langle a|\psi\rangle as the components.

Changing basis

The same state has different components in different bases. Switch from the basis {∣0⟩,∣1⟩}\{|0\rangle, |1\rangle\} to , ∣±⟩=(∣0⟩±∣1⟩)/2|\pm\rangle = (|0\rangle \pm |1\rangle)/\sqrt2, or . The arrow does not move; the axes do, and with them the numbers. For your state, ∣⟨+∣ψ⟩∣2|\langle +|\psi\rangle|^2 = 0.82.

The matrix that converts components from one basis to the other is unitary, here the Hadamard matrix H=12(111−1)H = \tfrac{1}{\sqrt2}\begin{pmatrix}1&1\\1&-1\end{pmatrix}, which you will meet again as a quantum gate.

Continuous bases

A particle’s position can take any real value, so its basis is continuous: one ket ∣x⟩|x\rangle for every point. Sums become integrals, and the wavefunction is simply the list of components in this basis:

ψ(x)=⟨x∣ψ⟩,∫∣x⟩⟨x∣ dx=1,⟨x∣x′⟩=δ(x−x′)\psi(x) = \langle x|\psi\rangle, \qquad \int |x\rangle\langle x|\,dx = \mathbb 1, \qquad \langle x|x'\rangle = \delta(x - x')

The same state written in the momentum basis has components ψ~(p)=⟨p∣ψ⟩\tilde\psi(p) = \langle p|\psi\rangle. The bridge between the two is a single amplitude:

⟨x∣p⟩=12πℏ eipx/ℏ\langle x|p\rangle = \frac{1}{\sqrt{2\pi\hbar}}\,e^{ipx/\hbar}

So going from position to momentum components is a Fourier transform (Chapter 6). The plate shows a narrow packet in xx and its wide partner in pp.

DeeperThe momentum eigenstate obeys ⟨x∣p^∣p⟩=−iℏ ∂x⟨x∣p⟩=p ⟨x∣p⟩\langle x|\hat p|p\rangle = -i\hbar\,\partial_x\langle x|p\rangle = p\,\langle x|p\rangle, so ⟨x∣p⟩=Ceipx/ℏ\langle x|p\rangle = Ce^{ipx/\hbar}. Requiring ⟨p∣p′⟩=∫⟨p∣x⟩⟨x∣p′⟩dx=∣C∣2 2πℏ δ(p−p′)\langle p|p'\rangle = \int \langle p|x\rangle\langle x|p'\rangle dx = |C|^2\,2\pi\hbar\,\delta(p-p') to equal δ(p−p′)\delta(p - p') fixes C=1/2πℏC = 1/\sqrt{2\pi\hbar}.

Operators as matrices

Sandwich an operator between basis states and you get its matrix elements, Aab=⟨a∣A^∣b⟩A_{ab} = \langle a|\hat A|b\rangle. In the {∣0⟩,∣1⟩}\{|0\rangle,|1\rangle\} basis, σx\sigma_x has matrix (0110)\begin{pmatrix}0&1\\1&0\end{pmatrix}; in the ∣±⟩|\pm\rangle basis the same operator is diagonal, (100−1)\begin{pmatrix}1&0\\0&-1\end{pmatrix}, because ∣±⟩|\pm\rangle are its eigenvectors. Choosing the basis in which an operator is diagonal is how most quantum problems are solved.

What you now have

Kets, bras, inner and outer products, projectors, completeness, change of basis, continuous bases with ⟨x∣p⟩\langle x|p\rangle, and matrix elements.

· · ·

Three questions to carry with you.

What is ψ(x) in Dirac notation?tap
The component of the state along the position eigenstate: ψ(x) = ⟨x|ψ⟩.
What does Σ|a⟩⟨a| equal for a complete orthonormal basis?tap
The identity operator.
How are position and momentum components related?tap
By a Fourier transform, with kernel ⟨x|p⟩ = e^(ipx/ħ)/√(2πħ).

And one to work out. For ∣ψ⟩=35∣0⟩+45∣1⟩|\psi\rangle = \tfrac35|0\rangle + \tfrac45|1\rangle, what is ∣⟨1∣ψ⟩∣2|\langle 1|\psi\rangle|^2?

P =

Sources: P. A. M. Dirac, Math. Proc. Camb. Phil. Soc. 35, 416 (1939). R. Shankar, Principles of Quantum Mechanics, ch. 1. J. J. Sakurai and J. Napolitano, Modern Quantum Mechanics, 3rd ed. Checked by code: discrete ⟨x|p⟩ matrix is unitary; H² = 1.

Next · Chapter 29
Operators, observables, commutators

Position then momentum is not the same as momentum then position. The difference is exactly iħ.

Turn the page