0

Whenever I get stuck in quantum mechanics, I refer back to things I know well. In this case, I referred back to energy quantization.

In the case of quantization of energy, I observed that quantization of energy differs from one state to another, which I understand perfectly. However, it seems that angular momentum is always quantized as $\hbar ^2 \ell(\ell+1)$. In the case of hydrogen atom, I get where this comes from. Can there be a system where angular momentum is quantized as $\hbar^2 \ell^2$? Or a system with angular momentum $\hbar^2 / \ell^2$?

I don't know why angular momentum is quantized like that except in the hydrogen atom case. If there is an explanation that doesn't need advanced quantum mechanical course and group theory, I'll be glad to hear it.

Qmechanic
  • 201,751
Habouz
  • 1,312
  • 1
  • 19
  • 2
    I am curious what you think is special about the case of the hydrogen atom that doesn't apply to angular momentum in general. – d_b Nov 12 '22 at 22:41
  • Possible duplicates: https://physics.stackexchange.com/q/174018/2451 and links therein. – Qmechanic Nov 13 '22 at 02:20

2 Answers2

6

The eigenvalues of $\mathbf{J}^2$ can be derived in an algebraic way solely based on the commutation relations $$\begin{align} [J_x,J_y]=i\hbar J_z \\ [J_y,J_z]=i\hbar J_x \\ [J_z,J_x]=i\hbar J_y \end{align}$$ The derivation can be found in every advanced text on quantum mechanics. The result are the eigenvalues $$J^2=\hbar^2 j(j+1) \quad\text{with }j=0,\frac 12, 1, \frac 32, 2, ...$$

3

Yes, eigenvalues of $J^2$ are always of the form $\hbar^2 \ell(\ell+1)$ with $\ell\in \frac{1}{2}\mathbb{Z}_+$ and this can be shown just using the angular momentum algebra. We shall first establish that they are of the form $\hbar^2 \ell(\ell+1)$ for some $\ell \in [0,+\infty)$ and then show that $\ell\in \frac{1}{2}\mathbb{Z}_+$.

First point, $\hbar$ has units of angular momentum. In that case $\hbar^2$ has units of angular momentm squared. In that case if $\mathbf{J}=(J_1,J_2,J_3)$ is an arbitrary angular momentum operator and if $\lambda$ is any eigenvalue of $J^2$ then we must always be able to write $\lambda = \hbar^2 \widetilde{\lambda}$. There is absolutely no loss of generality in doing so, we are just factoring out the dimension.

Second, $J^2$ must have non-negative eigenvalues because it is a sum of three squared Hermitian operators. Since $\hbar^2\geq 0$ then necessarily $\widetilde{\lambda}\geq 0$. But then, any non-negative number can be written as $\widetilde{\lambda}=\ell(\ell+1)$ for some $\ell$.

So we see any possible eigenvalue of $J^2$ for any possible angular momentum will take the form $\hbar^2 \ell(\ell+1)$. The fact that $\ell$ must be integer or half-integer can be shown by studying the angular momentum algebra. Let me outline the proof (I encourage you to fill in the details of the computations):

  1. Define the ladder operators $J_\pm = J_1\pm i J_2$. It then follows from the angular momentum algebra $[J_i,J_j]=i\hbar \epsilon_{ijk}J_k$ that \begin{eqnarray} [J_3,J_\pm] &=& \pm \hbar J_\pm\tag{1}. \end{eqnarray}

  2. Now let $|\ell,m,k\rangle$ be the common eigenvectors of $J_3$ and $J^2$, with $k$ an additional degeneracy label. These satisfy $$J^2|\ell,m,k\rangle=\hbar^2 \ell(\ell+1)|\ell,m,k\rangle,\quad J_3|\ell,m,k\rangle = m\hbar |\ell,m,k\rangle\tag{2}.$$ It follows immediately from (1) that $J_\pm$ raises/lowers $m$ by one unit, in other words: $$J_3 J_\pm |j,m,k\rangle = (m\pm 1)\hbar J_\pm |j,m,k\rangle\tag{3}$$

  3. The above observation means that $J_\pm |\ell,m,k\rangle = \alpha |\ell,m\pm 1,k\rangle$. Demanding the state to be normalized we may determine $\alpha$ using the commutation relations. Carrying out this computation $$J_\pm |\ell,m,k\rangle =\sqrt{\ell(\ell+1)-m(m\pm 1)}\hbar |\ell,m\pm 1,k\rangle.\tag{4}$$

  4. Observe that $J^2=J_1^2+J_2^2+J_3^2$. In that case $J^2-J_3^2= J_1^2+J_2^2$ and again, since $J^2-J_3^2$ is a sum of squared Hermitian operators it must have non-negative eigenvalues. This means that the states $|\ell,m,k\rangle$ must satisfy $$m^2\leq \ell(\ell+1)\tag{5}$$

  5. The above observation implies that there must be one $m_{\rm min}$ and one $m_{\rm max}$ such that $m_{\rm min}\leq m\leq m_{\rm max}$ for the allowed states. Since $J_-$ lowers $m$ and $J_+$ raises $m$ it follows that we must necessarily have $J_-|\ell,m_{\rm min},k\rangle=0$ and $J_+|\ell,m_{\rm max},k\rangle=0$, since otherwise the application of these operators would create states which are not allowed in the representation.

  6. This means that $m_{\rm min}$ and $m_{\rm max}$ can be found by these conditions. Inspecting (4) we immediately observe that $m_{\rm min}=-\ell$ and $m_{\rm max}=\ell$. In that case we may construct all the states by applying powers of $J_+$ to $|\ell,-\ell,k\rangle$. What we find is that the allowed values of $m$ are $$m=-\ell,-\ell+1,\dots, \ell-1,\ell\tag{6}$$ and that $\ell$ itself must be integer or half-integer. Indeed to see this last point, observe that there must be an integer $N$ such that $(J_+)^N |\ell,-\ell,k\rangle \propto |\ell,\ell,k\rangle$. Since we know that $(J_+)^N |\ell,-\ell,k\rangle \propto |\ell,-\ell+N,k\rangle$ it is necessary that there be such an integer $N$ with $\ell=N-\ell$ which implies that $2\ell$ must be an integer.

Gold
  • 35,872