1
$\begingroup$

Consider a particle in simple harmonic oscillator potential. Suppose the eigenstate of the operator

$$Q=\mu a+\nu a^\dagger$$ with eigenvalue $\alpha$ where $\mu,\nu $ and $\alpha$ are three complex numbers satisfying $|\mu|^2-|\nu|^2=1$. I'm trying to find uncertainty in this state $|\alpha\rangle $.

To find $|\alpha\rangle$ $$Q|\alpha\rangle =\alpha |\alpha\rangle $$ writing $|\alpha\rangle =\sum_n c_n|n\rangle $ and putting in above. I get $$\mu c_1=\alpha c_0,\ \ \ \&\ \ \ \ \mu c_{n+1} \sqrt{n+1}+\nu c_{n-1}\sqrt{n}=\alpha c_n,\ \ \ n=1,2,3,\ldots$$ I don't understand what to do now. Since it doesn't give an explicit solution for $c_n$. Please help me with this.


Edit:

\begin{align*} X&= \sqrt{\frac{\hbar}{2m\omega}}(a+a^\dagger) \\ P&= -i\frac{\sqrt{m\omega \hbar}}{2}(a-a^\dagger) \end{align*} Furthermore, $$Q=\mu a+\nu a^\dagger,\ \ \ \ Q^\dagger =\mu^*a^\dagger +\nu^* a$$ $$\rightsquigarrow a=\mu^* Q-\nu Q^\dagger,\ \ \ \ a^\dagger=\mu Q^\dagger -\nu^*Q$$ We can solve these for $X$ and $P$ in term of $Q$ and $Q^\dagger$. We have \begin{align*} X&=\sqrt{\frac{\hbar}{2m\omega}}[(\mu^*-\nu^*)Q+(\mu-\nu)Q^\dagger] \\ P&= -i\frac{\sqrt{m\omega \hbar}}{2}[(\mu^*+\nu*)Q-(\mu+\nu)Q^\dagger] \end{align*}

We would like to calculate $\langle X\rangle $ in state $|\alpha\rangle $. \begin{align*} \langle X\rangle &= \langle \alpha |X|\alpha\rangle \\ &= \sqrt{\frac{\hbar}{2m\omega}}\left[(\mu^*-\nu^*)\langle \alpha|Q|\alpha\rangle +(\mu-\nu)\langle \alpha| Q^\dagger|\alpha\rangle \right] \\ &= \sqrt{\frac{\hbar}{2m\omega}}\left[\alpha(\mu^*-\nu^*)+\alpha^*(\mu-\nu)\right] \end{align*}

The operator $X^2$ given by

$$X^2=\frac{\hbar}{2m\omega}\left[(\mu^*-\nu^*)^2Q^2+(\mu-\nu)^2(Q^\dagger)^2+(\mu^*-\nu^*)(\mu-\nu)(QQ^\dagger+Q^\dagger Q)\right]$$

Now we can find the expectation value of the same in state $|\alpha\rangle $. We can explicitly see $$\langle \alpha |Q^\dagger Q|\alpha\rangle =\alpha^*\alpha\langle \alpha|\alpha\rangle =|\alpha|^2$$ Similarly, $$\langle \alpha |Q Q^\dagger|\alpha\rangle =\langle \alpha|(Q^\dagger Q)^\dagger|\alpha\rangle =\langle \alpha|Q^\dagger Q|\alpha\rangle^*= |\alpha|^2$$ Therefore, We have $$\langle \alpha |X^2|\alpha \rangle =\frac{\hbar}{2m\omega}\left[(\mu^*-\nu^*)^2(\alpha^2)+(\mu-\nu)^2(\alpha^*)^2+2|\alpha|^2(\mu^*-\nu^*)(\mu-\nu)\right]=\frac{\hbar}{2m\omega}\left[(\mu^*-\nu^*)\alpha+(\mu-\nu)\alpha^*\right]^2=\langle X\rangle ^2$$
But this means, $\Delta X=0$, How is it so?

$\endgroup$
1
2
$\begingroup$

You are working in the oscillator algebra $[a,a^\dagger]=1$ in a different basis, since $$ [Q,Q^\dagger]=1 . $$ So you are inspecting its action on its normalized coherent states. $$ \langle \alpha| Q|\alpha\rangle =\alpha~~~\leadsto \\ \langle \alpha| Q^2|\alpha\rangle -\langle \alpha| Q|\alpha\rangle ^2 = \alpha^2 -\alpha^2 =0. $$

The Fock space you should be working in has a vacuum annihilated by Q, not a, $$ Q|0\rangle=0. $$


Edit in response to question edit

After the retargeting of your question, $$\langle QQ^\dagger+ Q^\dagger Q\rangle = \langle 2Q^\dagger Q + 1\rangle=2|\alpha|^2+1, $$ so, as you computed, $$ \langle X^2\rangle = \langle X\rangle ^2 + \left ({\hbar\over 2m\omega} \right ) |\mu-\nu|^2. $$ As you might expect, for μ=ν, Χ collapses to 0, so small wonder its variance vanishes there.

(Irrelevant to this, but this might be as good a place as any to be exposed to Dirac's trick.)

$\endgroup$
4
  • $\begingroup$ I added a little bit of calculation, and I'm getting $\Delta X$ to be zero in $|\alpha\rangle $ state, But the uncertainty shouldn't be zero, Where is the flaw? $\endgroup$ Oct 16 '21 at 5:07
  • 1
    $\begingroup$ Commute Q and it’s hc. Your “similarly” expression is wrong. $\endgroup$ Oct 16 '21 at 5:32
  • $\begingroup$ Got that, I can calculate that from commutation relation. $\endgroup$ Oct 16 '21 at 5:49
  • $\begingroup$ Got it, thanks for the explanation $\endgroup$ Oct 20 '21 at 9:27

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.