# Bound states of Dirac Delta function in infinite well

If there is a potential of $-\alpha\delta(x)$ for $-a<x<a$ and $\infty$ elsewhere, and the energy of the system is less than 0, then I'm trying to find the wave function. From the Schrodinger Equation, I've worked that wave function is

\begin{align} &-\frac{\hbar}{2m}\frac{d^2}{dx^2}\psi = E\psi \\ &k^2 = \frac{-2mE}{\hbar^2} \\ &\frac{d^2}{dx^2}\psi - k^2\psi = 0 \\ \end{align}

Which I solved to two parts

$A e^{kx}+ B e^{-kx}$ for $-a<x<0$

$C e^{kx} + D e^{-kx}$ for $a>x>0$.

From here, I tried the boundary conditions at $x=0,a,-a$:

\begin{align} &x=0: A+B=C+D \\ &x=-a: A e^{-ka} + B e^{ka} = 0 \\ &x=a: C e^{ka} + B e^{-ka} = 0 \\ \end{align}

But I'm not sure where to go from here. I can show a relationship between the boundary conditions at x=-a and x=a, but I don't think that would give me anything valuable.

You should also match the derivatives at $x=0$ so that they took into account the $\delta$-function. If you take smooth well $V_\epsilon(x)$ and consider small region near zero $(-\epsilon,\epsilon)$ where the "meat" of the well is concentrated you may then integrate the Schrodinger equation at that region, $$-\frac{\hbar^2}{2m}\int\limits_{-\epsilon}^{+\epsilon}\psi'' dx+\int\limits_{-\epsilon}^{+\epsilon}V_\epsilon\psi dx=E\int\limits_{-\epsilon}^{+\epsilon}\psi dx$$. Consider now the limit of $\epsilon\rightarrow 0$. The rhs will disappear and the first integral will give the difference between the derivatives from the right and the left. So if $V_\epsilon(x)\rightarrow -\alpha\delta(x)$ you get the needed matching condition, $$-\frac{\hbar^2}{2m}\Big(\psi'|_{0+}-\psi'|_{0-}\Big)=\alpha\psi(0)$$
• Which gives me $k C - k D - k A + k B = \frac{-2 m \alpha}{\hbar^2}(A+B)$ Which I still am not sure where to go with that. Nov 24, 2015 at 13:01
• Where to go? Solve the system of equations on $A$,$B$,$C$,$D$ and $k$, that's it. By the way, check the condition for $x=-a$
• Is there continuity of the derivative at $x=\pm a$ also? I don't see a way to solve the system of equations with only the information I've found so far. Nov 24, 2015 at 14:21
• Ok, I don't know what happened with my last comments but you miss the option $A=D$,$C=B$.
• @TheStrangeQuark For an infinite potential well, the derivative of the wave function is not continuous at the boundary (consider a finite potential and take the limit to infinity). You have 3 equations from the continuity of the wave function, one for the discontinuity of the derivative at the delta peak and one for the normalization. In total, this gives 5 equations for 4 constants and the bound state momentum. For large enough $a$ you should recover the bound state of the 1D delta well $E = -\frac{m \alpha^2}{2 \hbar^2}$ Nov 25, 2015 at 21:21