Lets look at Euler angles, a representation of rotations
๐ง Remember: the unitary transformation for a rotation about the vector angle $\vec \theta=(\theta_1,\theta_2,\theta_3)$ is
$$ \hat U_{\vec \theta}=e^{-i\vec \theta \cdot \hat L/\hbar}=e^{-i(\theta_x \hat L_x + \theta_y \hat L_y + \theta_z \hat L_z)/\hbar} $$
where $\hat L$ is the orbital angular momentum operator meaning that $[\hat L_i,\hat L_j]\ne 0$ for $i\ne j$
The operators are none commutative meaning that we canโt do
$$ e^{-\tfrac{i}{\hbar}\,(\theta_1 \hat{L}_x + \theta_2 \hat{L}_y + \theta_3 \hat{L}_z)}\ne e^{-\tfrac{i}{\hbar}\,\theta_1 \hat{L}_x}\,e^{-\tfrac{i}{\hbar}\,\theta_2 \hat{L}_y}\, e^{-\tfrac{i}{\hbar}\,\theta_3 \hat{L}_z} $$
Making the mathematics impractical so instead we will try to represent it in terms of $\hat L^2$ and $\hat L_z$
๐๏ธ Note: while we can choose a coordinate system which makes the rotation 1 dimensional simplifying the math. This would not help as our measuring apparatus would likely not be in this arbitrary direction and so to find the eigenvalues in a meaningful direction (say $z$) we would need to worry about the commutations of $\hat L$ when we decompose.
In parametrisation we write a rotation about a fixed axes as three consecutive rotations
If we perform this sequence the wavefunction transforms as
$$
\begin{aligned}
\psi (\vec{r}) \to \psi'(\vec{r}) &= \psi \left( [R_z(\gamma) R_x(\beta) R_z(\alpha)]^{-1} \vec{r} \right)
\\ &= \psi \left( R_z(-\alpha) R_x(-\beta) R_z(-\gamma) \vec{r} \right) = \hat{U}_{\alpha, \beta, \gamma} \psi (\vec{r})
\end{aligned}
$$
Now we write down $\hat U_{\alpha,\beta,\gamma}$
$$ \begin{aligned} \hat{U}_{\alpha, \beta, \gamma} \psi(\vec{r}) &= \psi \big( R_z(-\alpha) R_x(-\beta) R_z(-\gamma) \vec{r} \big) \\ &= e^{-i \alpha \hat{L}_z / \hbar} \psi \big( R_x(-\beta) R_z(-\gamma) \vec{r} \big) \\ &= e^{-i \alpha \hat{L}_z / \hbar} e^{-i \beta \hat{L}_x / \hbar} \psi \big( R_z(-\gamma) \vec{r} \big) \\ &= e^{-i \alpha \hat{L}_z / \hbar} e^{-i \beta \hat{L}_x / \hbar} e^{-i \gamma \hat{L}_z / \hbar} \psi(\vec{r}) \end{aligned} $$
๐๏ธ Note: this looks super similar to what we said was not allowed previously, the only difference is that here we rotate by each direction one after the other
Which thus gives us
$$ \boxed{\hat{U}_{\alpha, \beta, \gamma}=e^{-i \alpha \hat{L}_z / \hbar} e^{-i \beta \hat{L}_x / \hbar} e^{-i \gamma \hat{L}_z / \hbar}} $$
Finally we can show that $\hat U _{\alpha,\beta,\gamma}$ is unitary
$$ \hat{U}{\alpha, \beta, \gamma} \hat{U}{\alpha, \beta, \gamma}^{\dagger} = e^{-i\gamma \hat{L}_z / \hbar} e^{-i\beta \hat{L}_x / \hbar} e^{-i\alpha \hat{L}_z / \hbar} e^{i\alpha \hat{L}_z / \hbar} e^{i\beta \hat{L}_x / \hbar} e^{i\gamma \hat{L}_z / \hbar} = \hat{I} $$
๐ผ Case: lets consider the basis $\ket{l,m}$ such that