Suppose a half-integer $l$-value, say $\frac{1}{2}$, were allowed for orbital angular momentum. From
$$
L_{+} Y_{1 / 2}^{1 / 2}(\theta, \phi)=0,
$$
we may deduce, as usual,
$$
Y_{1 / 2}^{1 / 2}(\theta, \phi) \propto e^{j \phi / 2} \sqrt{\sin \theta} .
$$
Now try to construct $Y_{1 / 2}^{-1 / 2}(\theta, \phi)$ (a) by applying $L_{-}$to $Y_{1 / 2}^{1 / 2}(\theta, \phi)$ and (b) using $L_{-} Y_{1 / 2}^{-1 / 2}(\theta, \phi)=0$. Show that the two procedures lead to contradictory results. (This gives an argument against half-integer $l$-values for orbital angular momentum.)