I want to show that $$[\hat{x}^n,f(\hat{p})] = i\hbar\left(n\hat{x}^{n-1}\frac{\partial f(\hat{p})}{\partial \hat{p}}\right) + \mathcal{0}(\hbar^2)$$ where the last term means of order $\hbar^2$.

I have already shown that $[\hat{x},f(\hat{p})] = i\hbar \frac{\partial f(\hat{p})}{\partial \hat{p}}$

I think what I can do first is to expand the commutator like so: $[\hat{x}^n,f(\hat{p})] = \hat{x}^{n-1}[\hat{x},f(\hat{p})] + [\hat{x}^{n-1},f(\hat{p})]\hat{x} $

Then I expand the 2nd term. However, if do this some more times, I don't seem to get a closed form expression so I am stuck here.

I would be grateful for any advice


We will show the sought for expression by induction. Assume that $$[\hat{x}^n,f(\hat{p})] = i\hbar\left(n\hat{x}^{n-1}\frac{\partial f(\hat{p})}{\partial \hat{p}}\right) + \mathcal{0}(\hbar^2)$$ which is true for $n=1$.

The $(n+1)$-th case is $$[\hat{x}^{n+1},f(\hat{p})] = \hat{x}^{n}[\hat{x},f(\hat{p})] + [\hat{x}^{n},f(\hat{p})]\hat{x}\\ =i\hbar x^n \frac{\partial f(\hat{p})}{\partial \hat{p}}+i\hbar\left(n\hat{x}^{n-1}\frac{\partial f(\hat{p})}{\partial \hat{p}}\right)\hat x + \mathcal{0}(\hbar^2) .$$ But using the same relation, $\frac{\partial f(\hat{p})}{\partial \hat{p}}\hat x=\hat x\frac{\partial f(\hat{p})}{\partial \hat{p}}-\left[\hat x,\frac{\partial f(\hat{p})}{\partial \hat{p}}\right] =\hat x\frac{\partial f(\hat{p})}{\partial \hat{p}}+ \mathcal{0}(\hbar)$, and thus $$[\hat{x}^{n+1},f(\hat{p})] = i\hbar\left((n+1)\hat{x}^{n}\frac{\partial f(\hat{p})}{\partial \hat{p}}\right) + \mathcal{0}(\hbar^2),$$ which completes the proof.

| cite | improve this answer | |
  • $\begingroup$ Didn't even think about induction, yet it is so elegant. Thanks alot! But just for case n=1: I know it is true if the $\mathcal{0}(\hbar^2)$ is missing, but why is it still true with that term? $\endgroup$ – Eren May 17 '17 at 16:04

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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