My aim is to prove the restriction for the orbital momentum quantum number $-\ell \leq m \leq \ell$. My professor was giving me the hint, that I should use the norm of the state $|| L_+ |\ell,m\rangle ||$ with the state vector $|\ell, m\rangle$.

So, I start with the norm

$$|| L_+ |\ell,m\rangle ||^2 = \langle\ell,m| L_-L_+ |\ell,m\rangle \geq 0.$$

First of all, I calculate the operator product:

$$L_-L_+ = (L_x - iL_y)(L_x + i L_y) = L_x^2 + L_y^2 + i[L_x, L_y] = \vec{L}^2 - L_z^2 - \hbar L_z.$$

Using the eigenvalues of the operator, I get

$$|| L_+ |\ell,m\rangle || ^2 = \underbrace{\langle\ell, m|\ell,m\rangle}_{=1} \, \hbar^2 \big( \ell(\ell+1) - m(m+1) \big) \geq 0.$$

Finally I have

$$\ell(\ell + 1) \geq m(m+1).$$

Calculating now the norm $|| L_- |\ell, m\rangle ||$ leads me to a similar inequality:

$$\ell(\ell + 1) \geq m(m-1).$$

My question now is, how do get from these two inequalities to the restriction $-\ell \leq m \leq \ell$?

  • 1
    $\begingroup$ There is a standard notation for kets and bras. Why not use it? $\endgroup$ – G. Smith Oct 14 '20 at 20:16
  • $\begingroup$ I tried \bra{.} and \ket{.} they didnt seem to work, I don't know another command. $\endgroup$ – Octavius Oct 14 '20 at 20:16
  • 2
    $\begingroup$ You can use \langle, \mid, and \rangle. $\endgroup$ – G. Smith Oct 14 '20 at 20:17
  • 1
    $\begingroup$ Using | for \mid works as well. You can also do \right > or \left < as well. So some options for ket of $\psi$. \mid\psi\rangle gives $\mid\psi\rangle$, \left|\psi\right> also gives $\left|\psi\right>$. I usually just do |\psi\rangle : $|\psi\rangle$ $\endgroup$ – BioPhysicist Oct 14 '20 at 21:09

Your inequality can be read as

$$f(\ell)\geq f(m)$$ where $f(x)=x(x+1)$. This can be reduced to $\ell\geq m$ by applying $f^{-1}$ to both sides. But there's a caveat. Our function $f$ is not an injection since multiple x-values can map to the same number. This means we can't define the inverse easily and we have to do more work.

Our function is a parabola with a minimum at $x=-1/2$. We can make this inversion a little easier by shifting to this minimum using $\tilde \ell=\ell+1/2,\ \tilde m=m+1/2$. \begin{align} \ell(\ell+1)&\ge m(m+1)\\ (\tilde \ell-1/2)(\tilde\ell+1/2)&\ge(m-1/2)(m+1/2)\\ \tilde\ell^2-1/4&\geq \tilde m^2-1/4\\ \tilde \ell^2&\geq \tilde m^2\\ |\tilde \ell|&\ge|\tilde m| \end{align} This last line means $-\tilde \ell\le\tilde m\leq\tilde l\implies-\ell\le m\le \ell$ where I assumed $\ell$ to be positive.

  • 1
    $\begingroup$ Clear and complete. Thank you very much for your answer. $\endgroup$ – Octavius Oct 14 '20 at 23:47

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.