Electron in double square well

1. Aug 15, 2014

QualTime

1. The problem statement, all variables and given/known data

Consider an electron subject to the following 1-D potential:
$$U(x) = -U_0 \left( \delta(x+a) + \delta(x-a) \right)$$
where U_0 and a are positive reals.

(a) Find the ground state of the system, its normalized spatial wavefunction and the parameter κ related to the ground state energy.

(b) Write the transcendental equation satisfied by κ.

Ok, so I'm a bit confused first by what κ is here. This was from an old qual and I don't quite remember the exact phrasing, and it is possible that my memory has failed me and that k (the wavevector?) was meant here instead of kappa. In any event, I don't quite see what kind of transcendental equation it is supposed to satisfy, exept perhaps by matching the three spatial wavefunction and their first derivative at x=-a and x=a?

How would go about solving the Schrodinger equation with a potential given with delta functions? Any help would be appreciated.

2. Aug 15, 2014

ShayanJ

You just solve the equation for all places except x=a and x=-a.
Then you should impose boundary conditions.
The first one is that the wave-functions should be equal at x=a and x=-a.The second is the discontinuity of the first spatial derivative of the wave-function and the particular value of $\psi'(a+\varepsilon)-\psi'(a-\varepsilon)$ should be derived by integrating Schrodinger equation with delta potentials with respect to position from $a-\varepsilon$ to $a+\varepsilon$. These should be done for x=-a too.

3. Aug 16, 2014

QualTime

Thanks. I found out it was a problem from Griffith and was able to fill in the details.

One more question if I may: I remember the last part was asking to consider the case where a second electron was being subjected to the same potential, in addition to the resulting Coulomb repulsion. The question asked to discuss how a 2 electron molecule may form and how spin matters.

With the result from the previous part I can show that a bound state exist provided certain restriction on U_0 and m. Would the spin be relevant because of exchange force / symmetrization of the wavefunction (I'm assuming they would be indistinguishable so one would need an antisymmetric wf)? Always been a bit shaky on this so any hint is welcome.

4. Aug 17, 2014

ShayanJ

Whether you consider spin or not, the wave function of a number of fermions should be antisymmetric under the exchange of particles. Lets analyse the two cases.
First,without spin: The two particle antisymmetric wave function in this case is $\frac{1}{\sqrt{2}}(\phi(x_1)\psi(x_2)-\phi(x_2)\psi(x_1))$ where $\phi$ and $\psi$ are any two normalized states for the specific problem.
Second,with spin: in this case the wave function has also a spin part. So the spatial and spin part combined make the wave function of the system and so this combination should be antisymmetric under the exchange of particles. So if one part is antisymmetric, the other should be symmetric.
This and this may help too.

Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted