1 Finite Square Well
1 Finite Square Well
We shall now see the problem of the finite square well in brief. The potential is finite and constant
within an interval, and zero outside, as shown in Fig. 1. The potential V (x) is thus
(
−V0 , if −a ≤ x ≤ a
V (x) =
0, if |x| > a
V(x)
-a a
x
-V0
We will consider the bound states of the system, obtained when the energy of the particle is E < 0.
For x < −a, V (x) = 0 and the TISE can be written as
−h̄2 d2 ψ
= Eψ
2m dx2
√
d2 ψ 2 −2mE
2
= κ ψ , where κ = >0
dx h̄
The solution is ψ(x) = Ae−κx + Beκx .
As usual, we throw out the e−κx term since it blows up as x → −∞.
So for x < −a, the solution is ψ(x) = Beκx .
Similarly the solution for x > a is ψ(x) = F e−κx .
1
Now for −a ≤ x ≤ a, V (x) = −V0 . The TISE is
−h̄2 d2 ψ
− V0 ψ = Eψ
2m dx2
q
dψ2 2m(E + V0 )
= −l2 ψ , where l = >0
dx2 h̄
Note that this potential is an even function. So the solutions inside the well will be even and odd.
So the general solution can be written as
ψ(x) = C sin lx + D cos lx for −a ≤ x ≤ a.
We can now impose our boundary conditions of continuity of ψ and dψ/dx at x = ±a. We can
impose the boundary conditions on one side, they will be imposed on the other side automatically
since ψ(−x) = ±ψ(x).
Let us consider the even solutions (the cos terms).
F e−κx ,
if x > a
ψ(x) = D cos lx, if 0 ≤ x ≤ a
ψ(−x), if x < 0
and dψ/dx is continous at x = a, thus −κF e−κa = −lD sin la which gives us κ = l tan la.
√
Let us now write this down in a different form. Define z = la, and z0 = a 2mV0 /h̄. Note that z0
depends only on the properties of the well (larger z0 is a deeper well). We can thus write
q
z02 − z 2 = κa = la tan la = z tan z
giving us s
2
z0
tan z = −1
z
2
r 2
z0
This is a transcendental equation, and we can solve it graphically by plotting tan z and z
−1
and looking for intersections. I have done this in Fig 2 for z0 = 2, 5, 10, 15 respectively.
We can consider two limiting cases of the finite square well based on z0 which is measure of the
size of the well.
1. Wide deep well: z0 is large. For very deep well, the intersections will occur at zn = nπ/2
with odd n, n = 1, 3, 5, ...
n2 h̄2 π 2
With some manipulation we will get ln = nπ/2a, i.e. En + V0 = 2m(2a) 2
and the energies will be those of the infinite square well as V0 → ∞. Of course the even
solutions will give the n = 2, 4, 6, ...
2. Shallow narrow well: z0 is small. As z0 decreases, we get fewer and fewer bound states,
until eventually only one state remains (an even state). This is what we saw in the case of
the δ function well.
3
Now let us consider the scattering states, with E > 0.
√
For x < −a, we have ψ(x) = Aeikx + Be−ikx , where k = 2mE
h̄
.
√
2m(E+V0 )
Inside the well (−a < x < a), we have ψ(x) = C sin la + D cos la, with l = h̄
and for x > a, we have ψ(x) = F eikx , where we have already put in that there is no wave incoming
from right.
Thus as in the previous delta function well, A represents the incoming amplitude, B the reflected
amplitude, and F the transmitted amplitude.
Let us use the continuity of ψ(x) at −a and a,
and the continuity of dψ/dx at −a and a to write down
We have here four equations, which we can manipulate to write C, D, F in terms of A. A lot
of algebra later, we can get an expression for the transmission coefficient T = |F |2 /|A|2 as (see
Griffiths)
V02 2 2a
q
−1
T =1+ sin 2m(E + V0 )
4E(E + V0 ) h̄
Here its interesting to see that T does become unity when the argument for the sin2 vanishes, i.e.
n2 π 2 h̄2
when (En + V0 ) = 2m(2a) 2 , which are precisely the allowed energies for the infinite square well.
This is resonant transmission (Ramsauer-Townsend scattering), which happens when the width of
the well corresponds to an integral number of half-wavelengths of the incoming wave.
4
Figure 4: Schematic of the operation of a scanning tunneling microscope
(fig credit:Michael Schmid and Grzegorz Pietrzak,
see https://en.wikipedia.org/wiki/Scanning_tunneling_microscope )