Physics Stack Exchange is a question and answer site for active researchers, academics and students of physics. It's 100% free.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

How do we normalize a wavefunction that's a linear combination of sines and cosines (or of $Ae^{ikx}+Be^{-ikx}$ -- they're the same, right)? One you square it, wouldn't the integrand be oscillating through all space, and thus infinite?

share|cite|improve this question
Could you please specify which "things" are supposed to cancel? The norm surely can't cancel because it is a sum of positively semidefinite terms. Your function is a combination of plane waves so much like a single wave, its normal can't be one. One could normalize similar combinations to a wave function if the $A/B$ ratio were determined as a function of $k$ and if it were continuous. In that case, $|A|^2+|B|^2$ would be normalized exactly in the same way as $|A|^2$ is normalized if there's only one term, one plane wave. – Luboš Motl May 13 '11 at 4:45
You wouldn't normalize it over all space, the wavefunction would be zero, or something else, outside of a finite range. – Ramashalanka May 13 '11 at 4:46
@Luboš I'm sorry, I was forgetting to square it. But now that I remember that, it seems psi^2 is infinite? – wrongusername May 13 '11 at 4:48
@Ram I thought that for scattering states in a finite square well, psi to the left of the well is A e^ikx + B e^-ikx, which would be some infinite area to integrate over. – wrongusername May 13 '11 at 4:50
@wrongusername: scattering states aren't normalized, instead we look for reflection and transmission ratios. Bound states in a finite square well will decay outside the well like $e^{-k x}$ (with real $k$). – Ramashalanka May 13 '11 at 4:54
up vote 8 down vote accepted

You don't. Sinusoidal wavefunctions like these, a.k.a. plane waves, are non-normalizable, because the integral which defines the norm does not converge.

$$\langle\psi|\psi\rangle = \int_{-\infty}^{\infty} |A e^{ikx} + B e^{-ikx}|^2\mathrm{d}x = \infty$$

(EDIT: just thought I should mention that $\ldots = \infty$ doesn't mean the integral literally equals infinity, it's a notation for "does not converge".)

People tend to talk about these plane waves because it's easy to figure out how they behave, but in reality, a wavefunction is never just a pure plane wave. It's always some linear combination of plane waves with different frequencies/wavelengths, such that the overall wavefunction goes to zero at large distances quickly enough to make the integral converge. For example, you might have a plane wave confined within a potential well, so that the wavefunction is zero everywhere outside the well, or you could have a Gaussian wavepacket.

share|cite|improve this answer
So when Marder (Condensed Matter Physics, 2nd) writes that solutions of the free-particle Schrödinger equation are $$\Psi_{\vec{k}} = \frac{1}{\sqrt{V}}\mathrm{e}^{\mathrm{i}\vec{k}\cdot\vec{r}}$$ he implies that we'll be using LCs of them? Or is it an error? $$\left[\frac{\vec{r}}{V}\right]_{-\infty}^\infty$$ is either indefinite or infinity. He states that the plane wave solutions are normalized, which bothers me... maybe I'm missing something in his argument. – CHM Sep 29 '13 at 18:32
You always use a normalizable linear combination of plane waves, i.e. a wavepacket, to represent a real particle. (NB technically an infinite sum is not a linear combination, but it's close enough for physics purposes.) If $V=0$ everywhere, then a plane wave is not normalizable in the sense I describe in the answer, but in a bounded space it can be normalized. Maybe that's what the book is talking about. Alternatively, Marder could be using a different norm. (Or it could be an error. I can't tell without seeing the context.) – David Z Sep 29 '13 at 19:28

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.