Eigenfunctions and Vector Space

Wavefunctions are analogous to vectors in 3D space. The unit vectors of our vector space are eigenstates.

In normal 3D space, we represent a vector by its components.

\begin{displaymath}\bgroup\color{black} \vec{r}= x\hat{x}+y\hat{y}+z\hat{z} = \sum\limits_{i=1}^3r_i\hat{u}_i \egroup\end{displaymath}

The unit vectors \bgroup\color{black}$\hat{u}_i$\egroup are orthonormal,

\begin{displaymath}\bgroup\color{black} \hat{u}_i\cdot\hat{u}_j=\delta_{ij} \egroup\end{displaymath}

where \bgroup\color{black}$\delta_{ij}$\egroup is the usual Kroneker delta, equal to 1 if \bgroup\color{black}$i=j$\egroup and otherwise equal to zero.

Eigenfunctions - the unit vectors of our space - are orthonormal.

\begin{displaymath}\bgroup\color{black}\langle\psi_i\vert\psi_j\rangle=\delta_{ij}\egroup\end{displaymath}

We represent our wavefunctions - the vectors in our space - as linear combinations of the eigenstates (unit vectors).

\begin{displaymath}\bgroup\color{black}\psi=\sum\limits_{i=1}^\infty \alpha_i\psi_i\egroup\end{displaymath}


\begin{displaymath}\bgroup\color{black}\phi=\sum\limits_{j=1}^\infty \beta_j\psi_j\egroup\end{displaymath}

In normal 3D space, we can compute the dot product between two vectors using the components.

\begin{displaymath}\bgroup\color{black}\vec{r}_1\cdot\vec{r}_2=x_1x_2+y_1y_2+z_1z_2 \egroup\end{displaymath}

In our vector space, we define the dot product to be

\begin{eqnarray*}
\langle\psi\vert\phi\rangle
&=&\langle\sum\limits_{i=1}^\inft...
...eta_j\delta_{ij}
=\sum\limits_{i=1}^\infty\alpha^*_i\beta_i \\
\end{eqnarray*}


We also can compute the dot product from the components of the vectors. Our vector space is a little bit different because of the complex conjugate involved in the definition of our dot product.

From a more mathematical point of view, the square integrable functions form a (vector) Hilbert Space. The scalar product is defined as above.

\begin{displaymath}\bgroup\color{black} \langle\phi\vert\psi\rangle=\int\limits_{-\infty}^\infty d^3r \phi^*\psi \egroup\end{displaymath}

The properties of the scalar product are easy to derive from the integral.

\begin{displaymath}\bgroup\color{black} \langle\phi\vert\psi\rangle=\langle\psi\vert\phi\rangle^* \egroup\end{displaymath}


\begin{displaymath}\bgroup\color{black} \langle\phi\vert\lambda_1\psi_1+\lambda_...
...ert\psi_1\rangle+\lambda_2\langle\phi\vert\psi_2\rangle \egroup\end{displaymath}


\begin{displaymath}\bgroup\color{black} \langle\lambda_1\phi_1+\lambda_2\phi_2\v...
...ert\psi\rangle+\lambda_2^*\langle\phi_2\vert\psi\rangle \egroup\end{displaymath}

\bgroup\color{black}$\langle\psi\vert\psi\rangle$\egroup is real and greater than 0. It equals zero iff \bgroup\color{black}$\psi=0$\egroup. We may also derive the Schwartz inequality.

\begin{displaymath}\bgroup\color{black} \langle\psi_1\vert\psi_2\rangle\leq\sqrt...
...psi_1\vert\psi_1\rangle\langle\psi_2\vert\psi_2\rangle} \egroup\end{displaymath}

Linear operators take vectors in the space into other vectors.

\begin{displaymath}\bgroup\color{black} \psi'=A\psi \egroup\end{displaymath}

Jim Branson 2013-04-22