I understand that the momentum operator is Hermitian (thanks to this proof), as demonstrated by verifying the inner product relation. However, I am unsure how to prove that the angular momentum operator, specifically its $L_x$ component, is Hermitian and thus observable.
The $L_x$ operator can be expressed as:
$$
L_x = y p_z - z p_y,
$$
where $p_z = -i\hbar \frac{\partial}{\partial z}$ and $p_y = -i\hbar \frac{\partial}{\partial y}$ are the momentum operators in the $z$- and $y$-directions, respectively.
Alternatively, in spherical coordinates, $L_x$ is expressed as:
$$
L_x = -i\hbar \left( -\sin\phi \frac{\partial}{\partial \theta} - \cos\phi \cot\theta \frac{\partial}{\partial \phi} \right).
$$
To prove $L_x$ is Hermitian, I believe we need to check whether it satisfies the Hermitian condition:
$$
\langle \psi_1 | L_x \psi_2 \rangle = \langle L_x \psi_1 | \psi_2 \rangle^*,
$$
for arbitrary wavefunctions $\psi_1$ and $\psi_2$ in a suitable Hilbert space.
Attempting the proof:
I substitute $L_x$ into the inner product expression:
$$ \langle \psi_1 | L_x \psi_2 \rangle = \int \psi_1^* \, (y p_z - z p_y) \psi_2 \, dV. $$Compute the $y p_z$ and $z p_y$ terms explicitly:
$$ \langle \psi_1 | y p_z \psi_2 \rangle = \int \psi_1^* y \left( -i\hbar \frac{\partial}{\partial z} \psi_2 \right) dV, $$
$$ \langle \psi_1 | -z p_y \psi_2 \rangle = \int \psi_1^* \left( -z \left( -i\hbar \frac{\partial}{\partial y} \psi_2 \right) \right) dV. $$Use integration by parts to see if the boundary terms vanish for well-behaved wavefunctions.
My questions:
- Do both terms satisfy the Hermitian condition?
- Is this enough to confirm that $L_x$ is Hermitian and thus observable?
- Are there specific pitfalls in extending this analysis to spherical coordinates?
Could someone confirm if my approach is correct and point out any errors or assumptions I may have overlooked?