|q⟩ Bad Qubits

beginner · Physics · Observables, Operators & Measurement

The Position and Momentum Operators

In wave mechanics every physical quantity is represented by a linear operator acting on the wave function. The two most fundamental operators are position and momentum; every other observable in non-relativistic quantum mechanics is built from them.

The position operator

In one dimension the position operator x^\hat{x} simply multiplies the wave function by the coordinate xx:

x^ψ(x)=xψ(x).\hat{x}\,\psi(x) = x\,\psi(x).

In three dimensions each component acts the same way: x^iψ(r)=xiψ(r)\hat{x}_i\,\psi(\mathbf{r}) = x_i\,\psi(\mathbf{r}). The operator is said to act by multiplication; the wave function at each point is scaled by the value of the coordinate at that point. Because multiplication by a real number is its own adjoint, x^\hat{x} is Hermitian, which is required for any observable.

The momentum operator

The momentum operator is less obvious — it cannot be a simple multiplication in position space. The physical requirement comes from the de Broglie relation: a state of definite momentum pp should correspond to a plane wave eipx/e^{ipx/\hbar}. For this wave to be an eigenstate of the momentum operator with eigenvalue pp, the operator must generate a spatial derivative. One can show (by demanding consistency with the commutation relation and Hermiticity) that the only valid choice is

p^=ix.\hat{p} = -i\hbar\,\frac{\partial}{\partial x}.

Verification in three dimensions

In three dimensions the momentum operator generalises to the gradient:

p^=i.\hat{\mathbf{p}} = -i\hbar\,\nabla.

Acting on a plane wave ψ(r)=eikr\psi(\mathbf{r}) = e^{i\mathbf{k}\cdot\mathbf{r}} gives

p^eikr=i(ik)eikr=keikr,\hat{\mathbf{p}}\,e^{i\mathbf{k}\cdot\mathbf{r}} = -i\hbar\,(i\mathbf{k})\,e^{i\mathbf{k}\cdot\mathbf{r}} = \hbar\mathbf{k}\,e^{i\mathbf{k}\cdot\mathbf{r}},

so the eigenvalue is k\hbar\mathbf{k}, exactly the de Broglie momentum. This self-consistency is the derivation that fixes the form of p^\hat{p}.

Expectation values

Given a normalised wave function ψ(x)\psi(x), the expectation (mean) value of position is

x=ψ(x)xψ(x)dx=xψ(x)2dx,\langle x \rangle = \int_{-\infty}^{\infty} \psi^*(x)\,x\,\psi(x)\,dx = \int_{-\infty}^{\infty} x\,|\psi(x)|^2\,dx,

which is the probability-weighted average of xx over all space. For momentum,

p=ψ(x)(ix)ψ(x)dx.\langle p \rangle = \int_{-\infty}^{\infty} \psi^*(x)\left(-i\hbar\frac{\partial}{\partial x}\right)\psi(x)\,dx.

Note that the derivative acts on ψ\psi, not on ψ\psi^*; reversing the order would give the wrong (complex) answer.

Summary

| Observable | Operator in position space | |---|---| | Position xx | x^=x\hat{x} = x (multiply by xx) | | Momentum pp | p^=i/x\hat{p} = -i\hbar\,\partial/\partial x |

These two operators are the building blocks of quantum mechanics: kinetic energy is p^2/(2m)\hat{p}^2/(2m), the harmonic oscillator Hamiltonian is p^2/(2m)+mω2x^2/2\hat{p}^2/(2m) + m\omega^2\hat{x}^2/2, and their commutator [x^,p^]=i[\hat{x},\hat{p}] = i\hbar encodes the Heisenberg uncertainty principle.

Sign in on the full site to ask questions and join the discussion.