Given a delta function $\alpha\delta(x+a)$ and an infinite energy potential barrier at $[0,\infty)$, calculate the scattered state, calculate the probability of reflection as a function of $\alpha$, momentum of the packet and energy. Also calculate the probability of finding the particle between the two barriers.

I start by setting up the standard equations for the wave function:

$$\begin{align}\psi_I &= Ae^{ikx}+Be^{-ikx} &&\text{when } x<-a, \\ \psi_{II} &= Ce^{ikx}+De^{-ikx} &&\text{when } -a<x<0\end{align}$$

The requirement for continuity at $x=-a$ means


Then the requirement for specific discontinuity of the derivative at $x=-a$ gives

$$ik(-Ce^{-ika}+De^{ika}+Ae^{-ika}-Be^{ika}) = -\frac{2m\alpha}{\hbar^2}(Ae^{-ika}+Be^{ika})$$

At this point I set $A = 1$ (for a single wave packet) and set $D=0$ to calculate reflection and transmission probabilities. After a great deal of algebra I arrive at

$$\begin{align}B &= \frac{\gamma e^{-ika}}{-\gamma e^{ika} - 2ike^{ika}} & C &= \frac{2e^{-ika}}{\gamma e^{-ika} - 2ike^{-ika}}\end{align}$$

(where $\gamma = -\frac{2m\alpha}{\hbar^2}$) and so reflection prob. $R=\frac{\gamma^2}{\gamma^2+4}$ and transmission prob. $T=\frac{4}{\gamma^2+4}$.

Here's where I run into the trouble of figuring out the probability of finding the particle between the 2 barriers. Since the barrier at $0$ is infinite the only leak could be over the delta function barrier at $-a$. Would I want to use the previous conditions but this time set $A=1$ and $C=D$ due to the total reflection of the barrier at $0$ and then calculate $D^*D$?

  • 1
    $\begingroup$ Hi Hippie_Eater, and welcome to Physics Stack Exchange! Excellent question :-) I hope you don't mind that I made some of the equations display style to aid readability. $\endgroup$
    – David Z
    Sep 20, 2012 at 17:33
  • $\begingroup$ Thank you, that's much better - I am still polishing my TeK-Fu so I hope I'll be making it look all sexyfine like this in the futue. $\endgroup$ Sep 20, 2012 at 17:37

2 Answers 2


Hints to the question(v5):

  1. OP correctly imposes two conditions because of the delta function potential at $x=-a$, but OP should also impose the boundary condition $\psi(x\!=\!0)=0$ because of the infinite potential barrier at $x\geq 0$.

  2. There is zero probability of transmission because of the infinite potential barrier at $x\geq 0$. (Recall that transmission would imply that the particle could be found at $x\to \infty$, which is impossible.)

  3. Hence there is a 100 percent probability of reflection, cf. the unitarity of the $S$-matrix. See also this Phys.SE answer.

  4. As OP writes, away from the two obstacles, one has simply a free solution to the time-independent Schrödinger equation, namely a linear combination of the two oscillatory exponentials $e^{\pm ikx}$. This solution is non-normalizable over a non-compact interval $x\in ]-\infty,0]$.

  5. To make the wave function normalizable, let us truncate space for $x< -K$, where $K>0$ is a very large constant. So now $x\in [-K,0]$. One may then define and calculate the probability $P(-a \leq x\leq 0)$ of finding the particle between the two barriers via the usual probabilistic interpretation of the square of the wave function.

  6. If we now let the truncation parameter $K\to \infty$, then we can deduce without calculation that this probability $P(-a \leq x\leq 0)\to 0$ goes to zero.

  • $\begingroup$ I updated the answer. $\endgroup$
    – Qmechanic
    Sep 20, 2012 at 21:14

The probability of finding a particle in an interval $a<x<b$ is given by the integral $$\int_a^b \psi^* \psi \, dx ,$$ assuming that your wave function is properly normalised.

So in your case, you should calculate $$\frac{\int_{-a}^0 \psi_{II}^* \psi_{II} \,dx}{\int_{-\infty}^{-a} \psi_{I}^* \psi_{I} \,dx+\int_{-a}^0 \psi_{II}^* \psi_{II} \,dx} . $$

The numerator is the region you are interested in, the denominator takes care of the normalisation so that the probability will come out between 0 and 1. I'll leave it to you to calculate the integrals.

  • $\begingroup$ Thank you, I do have a tendency to over-complicate things. But that raises the question, what conditions should I use to figure out $A, B, C, D$? Presumably it would be the standard two regarding the continuity in of $\psi$ in $-a$ and discontinuity of $\psi'$ in $-a$ but I think I'll need more than that? Am I correct in thinking that the barrier at $0$ reflects completely and thus $C=D$? $\endgroup$ Sep 20, 2012 at 20:58
  • $\begingroup$ OK, so you have four unknowns ($A,B,C,D$). You already have two conditions, you need two more. As #Qmechanic states, one of the conditions should be $\psi(0)=0$. The other can be got from normalisation ($\int_{-\infty}^0 \psi^* \psi \, dx =1$). $\endgroup$ Sep 20, 2012 at 21:06
  • $\begingroup$ Note that the scattering wave function $\psi(x)$ is not normalizable. $\endgroup$
    – Qmechanic
    Sep 20, 2012 at 23:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.