2
$\begingroup$

I have been trying for find a closed form solution, or at least something neat for the commutation relation $$[e^{-x^{2}},e^{\alpha i p}] = ?$$ (where $[x,p] = i\mathbb{I}$) but have had little luck. I have tried using BHC theorem but this does not get me very far. I think that there must be some simple relation that I am over looking.

$\endgroup$
8
  • $\begingroup$ Related: physics.stackexchange.com/q/98372 $\endgroup$
    – Photon
    Jun 15, 2022 at 11:58
  • $\begingroup$ Does this answer your question? Commutators involving functions $\endgroup$ Jun 15, 2022 at 12:06
  • $\begingroup$ Oops I made a mistake. Let me fix the commutator. As it is prior to the changes indeed that post that you shared would suffice. $\endgroup$
    – Hldngpk
    Jun 15, 2022 at 12:25
  • $\begingroup$ @Photon there, this is the commutator I was interested in. I accidentaly wrote down something else. $\endgroup$
    – Hldngpk
    Jun 15, 2022 at 12:27
  • $\begingroup$ Here I have used BCH theorem and them I must compute interations of nested commutators involving the operator $p$. Using the stackexchange post that you shared I end up with a messy series of operators. I was hoping that there would be a nice closed form for the commutator posted above that does not need BCH theorem. $\endgroup$
    – Hldngpk
    Jun 15, 2022 at 12:29

1 Answer 1

8
$\begingroup$

You have set $\hbar=1$, so $p= -i\partial_x$ in the coordinate representation, so one of your operators is a bland Lagrange shift operator, and hence $$ [e^{\alpha i p}, f(x) ] = (f(x+\alpha)-f(x)) e^{\alpha i p} ~~~~\leadsto \\ [e^{-x^2}, e^{\alpha i p}]= - (e^{-(x+\alpha)^2}-e^{x^2}) e^{\alpha i p} . $$

(With a tip of the hat to @thedude 's comment! The linked WP article reminds you that $e^{i\alpha p} f(x) e^{-i\alpha p}= f(x+\alpha) $, in operator calculus language; when it acts on a constant, it reduces to just $e^{i\alpha p} f(x)= f(x+\alpha) $.)

Make sure to confirm for small α.

$\endgroup$
7
  • $\begingroup$ This is phenomenal !!! Thank you very much! $\endgroup$
    – Hldngpk
    Jun 15, 2022 at 13:22
  • $\begingroup$ Actually $[e^{\alpha ip},f(x)]g(x)=e^{\alpha ip}f(x)g(x)-f(x)e^{\alpha ip}g(x)$, so this answer is in the right direction but not correct $\endgroup$
    – thedude
    Jun 15, 2022 at 13:22
  • $\begingroup$ Do you know where I might find a proof of these wonderful result. i.e. $[e^{\alpha i p},f(x)] = f(x+\alpha)$ where $x$ and $p$ are the position and momentum operators respectively. $\endgroup$
    – Hldngpk
    Jun 15, 2022 at 13:23
  • 2
    $\begingroup$ After the edit, this is now correct $\endgroup$
    – thedude
    Jun 15, 2022 at 13:34
  • $\begingroup$ Fantastic! I appreciate all of your comments and help. Special thanks to @CosmasZachos $\endgroup$
    – Hldngpk
    Jun 15, 2022 at 13:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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