1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Homework Help: Sakurai 2.17 - More elegant solution help?

  1. Dec 24, 2008 #1

    Show for the one-dimensional simple harmonic oscillator

    [tex]\langle 0 | e^{ikx} | 0 \rangle = \exp{[-k^2 \langle 0 | x^2 | 0 \rangle / 2]} [/tex]

    where [tex]x[/tex] is the position operator (here, [tex]k[/tex] is a number, not an operator, with dimensions 1/length).

    My Solution

    Well, I already know how to do this problem, but my solution isn't as clean. I was searching for a more elegant solution. Here's the outline to my solution:

    1. We know that [tex]e^{ikx}|p'\rangle = |p' + \hbar k\rangle[/tex] (pretty simple to prove).
    2. We can show that
    [tex]\langle 0 |e^{ikx}| 0 \rangle = \int dp' \langle 0|p' \rangle \langle p'-\hbar k | 0 \rangle[/tex] ​
    by putting everything in the momentum basis.
    3. We then just need to find the harmonic oscillator's ground state in the momentum representation, which is just a Fourier transform of the ground state in the position representation.
    4. Substitute into the thing got in #2 and compute away, leading to the equality. (and this works, after expanding everything)

    As we can see, computing the Fourier transform in step 3 and computing everything in step 4 takes some time, and isn't exactly the "winning" solution.:smile: So would someone help me find or post a more elegant solution?
  2. jcsd
  3. Dec 24, 2008 #2


    User Avatar
    Science Advisor

    I don't know if this is more elegant or not, but ...

    1) Expand in powers of k.
    2) Note that expectation value of odd powers of x is zero, because the ground-state wave function is an even function of x.
    3) Prove that

    [tex]\langle x^{2n}\rangle = (2n{-}1)!!\,\langle x^2\rangle^n.[/tex]

    Useful identify for the proof:

    [tex]\int_{-\infty}^{+\infty}dx\;x^{2n}e^{-cx^2}=\left(-{d\over dc}\right)^{\!n}\int_{-\infty}^{+\infty}dx\;e^{-cx^2}.[/tex]

    4) Resum the series.
  4. Dec 24, 2008 #3
    Wow... number three seems to be pretty useful! This was actually my initial first approach... I found a really neat analog to the Catalan numbers, but wasn't able to complete it. I'll go and try out this approach. Thanks a lot for the help :smile:
  5. Dec 24, 2008 #4


    User Avatar
    Science Advisor

    In quantum field theory #3 is known as Wick's theorem.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook