The Angular Momentum Eigenfunctions

The angular momentum eigenstates are eigenstates of two operators.

\bgroup\color{black}$\displaystyle L_z Y_{\ell m}(\theta,\phi)=m\hbar Y_{\ell m}(\theta,\phi) $\egroup
\bgroup\color{black}$\displaystyle L^2 Y_{\ell m}(\theta,\phi)=\ell(\ell+1)\hbar^2 Y_{\ell m}(\theta,\phi) $\egroup
All we know about the states are the two quantum numbers \bgroup\color{black}$\ell$\egroup and \bgroup\color{black}$m$\egroup. We have no additional knowledge about \bgroup\color{black}$L_x$\egroup and \bgroup\color{black}$L_y$\egroup since these operators don't commute with \bgroup\color{black}$L_z$\egroup. The raising and lowering operators \bgroup\color{black}$L_\pm=L_x\pm iL_y$\egroup raise or lower \bgroup\color{black}$m$\egroup, leaving \bgroup\color{black}$\ell$\egroup unchanged.
\bgroup\color{black}$\displaystyle L_\pm Y_{\ell m}=\hbar\sqrt{\ell(\ell+1)-m(m\pm 1)}Y_{\ell(m\pm 1)}$\egroup

The differential operators take some work to derive.

\bgroup\color{black}$\displaystyle L_z={\hbar\over i}{\partial\over\partial\phi}$\egroup
\bgroup\color{black}$\displaystyle L_\pm=\hbar e^{\pm i\phi}\left(\pm{\partial\over\partial\theta}+i\cot\theta{\partial\over\partial\phi}\right)$\egroup

Its easy to find functions that give the eigenvalue of \bgroup\color{black}$L_z$\egroup.

\begin{eqnarray*}
Y_{\ell m}(\theta,\phi)=\Theta(\theta)\Phi(\phi)=\Theta(\thet...
... i}im\Theta(\theta)e^{im\phi}=m\hbar Y_{\ell m}(\theta,\phi)\\
\end{eqnarray*}


To find the \bgroup\color{black}$\theta$\egroup dependence, we will use the fact that there are limits on \bgroup\color{black}$m$\egroup. The state with maximum \bgroup\color{black}$m$\egroup must give zero when raised.

\begin{displaymath}\bgroup\color{black} L_+ Y_{\ell\ell}=\hbar e^{i\phi}\left({\...
...r\partial\phi}\right)\Theta_\ell(\theta)e^{i\ell\phi}=0 \egroup\end{displaymath}

This gives us a differential equation for that state.

\begin{eqnarray*}
{d\Theta(\theta)\over d\theta}+i\Theta(\theta)\cot\theta(i\el...
...{d\Theta(\theta)\over d\theta}=\ell\cot\theta\Theta(\theta) \\
\end{eqnarray*}


The solution is

\begin{displaymath}\bgroup\color{black} \Theta(\theta)=C\sin^\ell\theta .\egroup\end{displaymath}

Check the solution.

\begin{displaymath}\bgroup\color{black} {d\Theta\over d\theta}=C\ell\cos\theta\sin^{\ell-1}\theta=\ell\cot\theta\Theta \egroup\end{displaymath}

Its correct.

Here we should note that only the integer value of \bgroup\color{black}$\ell$\egroup work for these solutions. If we were to use half-integers, the wave functions would not be single valued, for example at \bgroup\color{black}$\phi=0$\egroup and \bgroup\color{black}$\phi=2\pi$\egroup. Even though the probability may be single valued, discontinuities in the amplitude would lead to infinities in the Schrödinger equation. We will find later that the half-integer angular momentum states are used for internal angular momentum (spin), for which no \bgroup\color{black}$\theta$\egroup or \bgroup\color{black}$\phi$\egroup coordinates exist.

Therefore, the eigenstate $Y_{\ell\ell}$ is.

\bgroup\color{black}$\displaystyle Y_{\ell\ell}=C\sin^\ell(\theta)e^{i\ell\phi}$\egroup
We can compute the next state down by operating with \bgroup\color{black}$L_-$\egroup.

\begin{displaymath}\bgroup\color{black} Y_{\ell(\ell-1)}=C L_-Y_{\ell\ell} \egroup\end{displaymath}

We can continue to lower \bgroup\color{black}$m$\egroup to get all of the eigenfunctions.

We call these eigenstates the Spherical Harmonics. The spherical harmonics are normalized.

\begin{eqnarray*}
\int\limits_{-1}^{1}d(\cos\theta)\int\limits_{0}^{2\pi}d\phi\...
...m}Y_{\ell m}=1 \\
\int d\Omega\; Y^*_{\ell m}Y_{\ell m}=1 \\
\end{eqnarray*}


Since they are eigenfunctions of Hermitian operators, they are orthogonal.
\bgroup\color{black}$\displaystyle \int d\Omega\; Y^*_{\ell m}Y_{\ell' m'}=\delta_{\ell\ell'}\delta_{mm'}$\egroup

We will use the actual function in some problems.

\begin{eqnarray*}
Y_{00} &=& {1\over \sqrt{4\pi}}\\
Y_{11} &=& -\sqrt{3\over ...
...heta\\
Y_{20} &=& \sqrt{5\over 16\pi}\; (3 \cos^2\theta-1)\\
\end{eqnarray*}


The spherical harmonics with negative \bgroup\color{black}$m$\egroup can be easily compute from those with positive \bgroup\color{black}$m$\egroup.
\bgroup\color{black}$\displaystyle Y_{\ell(-m)} = (-1)^m Y_{\ell m}^*$\egroup

Any function of \bgroup\color{black}$\theta$\egroup and \bgroup\color{black}$\phi$\egroup can be expanded in the spherical harmonics.

\begin{displaymath}\bgroup\color{black} f(\theta,\phi)=\sum\limits_{\ell=0}^\inf...
...mits_{m=-\ell}^\ell
C_{\ell m}Y_{\ell m}(\theta,\phi) \egroup\end{displaymath}

The spherical harmonics form a complete set.

\begin{displaymath}\bgroup\color{black} \sum\limits_{\ell=0}^{\infty}\sum\limits...
...{m=-\ell}^\ell \vert\ell m\rangle \langle\ell m\vert=1 \egroup\end{displaymath}

When using bra-ket notation, \bgroup\color{black}$\vert\ell m\rangle$\egroup is sufficient to identify the state.

The spherical harmonics are related to the Legendre polynomials which are functions of \bgroup\color{black}$\theta$\egroup.

\begin{eqnarray*}
Y_{\ell 0}(\theta,\phi)&=&\left({2\ell+1\over 4\pi}\right)^{1...
...ell+m)!}\right]^{1\over 2}
P_\ell^m(\cos\theta)e^{im\phi} \\
\end{eqnarray*}




Subsections
Jim Branson 2013-04-22