Homework 5
Homework 5
2
L
sin
_
x
L
_
sin
_
x
2L
_
dx
_
2
=
_
2
L
_
2
2L
3
__
2
=
_
8
3
_
2
gomez@physics.rutgers.edu
1
Integration made in Mathematica. Wolfram Alpha LLC. 2012. WolframAlpha.
http://www.wolframalpha.com (access October 15, 2012).
1
Shankar Ex. 5.2.4 Consider a particle of mass m in the state |n of a box of length
L. Find the force F = E/L encountered when the walls are slowly pushed in,
assuming the particle remains in the nth state of the box as its size changes. Consider
a classical particle of energy E
n
in this box. Find its velocity, the frequency of collision
on a given wall, the momentum transfer per collision, and hence the average force.
Compare it to E/L compute above.
The energy of a particle of mass m in a box of length L is:
E =
2
2
n
2
2mL
2
, n = 1, 2, 3, ...
Next, lets calculate the force:
F =
E
L
=
L
_
2
n
2
2mL
2
_
=
2
2
n
2
mL
3
=
2E
L
On the other hand, classically, we know that the energy is:
E
n
=
p
2
2m
p =
_
2mE
n
v =
2mE
n
m
In addition, the time that one particle needs to collide again on a given wall, i.e., its
frequency is:
t =
2L
v
=
2Lm
2mE
n
The change in momentum, in an elastic collision, when the particle hits the wall
would be:
P = p (p) = 2p = 2
_
2mE
n
Finally, using all the previous computed results, the average force of the particle is:
F =
P
t
=
2(2mE
n
)
2mL
=
2E
n
L
Shankar Ex. 5.2.2 Hint: The spirit of this problem is to see how the kinetic and potential
energies scale as 0 (i.e., for a very wide Gaussian) and show that the (positive)
kinetic energy drops to zero faster than the (negative) potential energy.
(a) Show that for any normalized |, |H| E
0
, where E
0
is the lowest-energy
eigenvalue. (Hint: Expand | in the eigenbasis of H.)
Lets dene |n as the orthonormal eigenbasis of H, i.e., H|n = E
n
|n and
n
|nn| = 1. Then, we can compute:
|H| =
m,n
|mm|H|nn|
=
m,n
E
n
|mm|nn|
=
m,n
E
n
|mn|
m,n
=
n
E
n
|n||
2
= E
0
|0||
2
+E
1
|1||
2
+E
2
|2||
2
+...
E
0
2
(b) Prove the following theorem: Every attractive potential in one dimension has at
least one bound state. Hint: Since V is attractive, if we dene V () = 0, it
follows that V (x) = |V (x)| for all x. To show that there exists a bound state
with E < 0, consider
a
(x) =
_
a
_
1/4
e
ax
2
/2
and calculate
E() =
|H|
, H =
2
2m
d
2
dx
2
|V (x)|
Show that E() can be made negative by suitable choice of . The desired result
follows from the application of the theorem proved above.
Lets calculate the value of E(a):
E(a) =
a
|H|
a
=
_
_
a
_
1/2
e
ax
2
2
_
2
2m
d
2
dx
2
|V (x)|
_
e
ax
2
2
dx
Lets calculate separately the value of H:
_
2
2m
d
2
dx
2
|V (x)|
_
e
ax
2
2
=
2
2m
d
2
dx
2
_
e
ax
2
2
_
|V (x)|e
ax
2
2
=
2
2m
d
dx
_
axe
ax
2
2
_
|V (x)|e
ax
2
2
=
2
2m
_
a
2
x
2
e
ax
2
2
axe
ax
2
2
_
|V (x)|e
ax
2
2
=
_
2
2m
_
a
2
x
2
ax
_
|V (x)|
_
e
ax
2
2
Returning to the value of E(a):
E(a) =
_
_
a
_
1/2
e
ax
2
2
_
2
2m
_
a
2
x
2
ax
_
|V (x)|
_
e
ax
2
2
dx
=
_
a
__
2
a
2m
e
ax
2
dx
_
2
a
2
2m
x
2
e
ax
2
dx
_
|V (x)|e
ax
2
dx
_
=
_
a
2
a
2m
_
a
0
_
|V (x)|e
ax
2
dx
_
=
a
2
2m
_
a
|V (x)|e
ax
2
dx E
0
from (a)
where here the rst term corresponds to the kinetic energy and the second to
the potential energy. In addition, we can see that when a 0 the kinetic term
decrease faster than the second term due to the exponent of a in each term.
The theorem can be prove using part (a), where we can see that the expression
for E(a) is greater or equal than the lowest-energy eigenvalue. Therefore, in a
one dimensional problem there is at least one bound state.
3
Shankar Ex. 5.2.3 Consider V (x) = aV
0
(x). Show that it admits a bound state of en-
ergy E = ma
2
V
2
0
/2
2
. Are there any other bound state?. Hint: Solve Schrodingers
equation outside the potential E < 0, and keep only the solution that has the right
behavior at innity and is continuous at x = 0. Draw the wave function and see how
there is a cusp, or a discontinuous change of slope at x = 0. Calculate the change in
slope and equate it to
_
+
_
d
2
dx
2
_
dx
(where is innitesimal) determined from Schr odingers equation.
Lets calculate the Schrodinger equation with the potential V (x):
2
2m
d
2
dx
2
+aV
0
(x) = E
d
2
dx
2
+
2m
2
(E aV
0
(x)) = 0
Now, lets integrate this relation between an innitesimal change:
_
+
_
d
2
dx
2
_
dx =
_
+
_
2m
2
(E aV
0
(x))
_
dx
_
+
d
dx
_
d
dx
_
dx =
_
+
2mE
2
dx +
_
+
2m
2
aV
0
(x))dx
d
dx
= 0 +
2maV
0
2
(0)
d()
dx
d()
dx
=
2maV
0
2
(0)
Figure 1: Diagram of a particle moving in a delta function potential, according to Shankar,
Exercise 5.2.3.
On the other hand, the equation of motion of a particle moving in a V = 0 potential
is:
2
2m
d
2
dx
2
= E
4
where if we separate the cases for in this exercise, shown in Figure 1, the wavefunctions
of a particle are:
I
(x) = Ae
kx
+Be
kx
= Ae
kx
II
(x) = Ce
kx
+De
kx
= De
kx
where k
2
= 2mE/ and, B and C vanish due to boundary conditions at both
innities. Applying the boundary conditions when x = 0:
I
(0) =
II
A = D
Finally, lets calculate the discontinuous change of slope at x = 0 using the previous
results:
d
II
(0)
dx
d
I
(0)
dx
=
2maV
0
2
(0)
kD kA =
2maV
0
2
(0)
2kD =
2maV
0
2
D taking (0) =
II
(0)
k =
2maV
0
2
k
2
=
4m
2
a
2
V
2
0
2mE
2
=
4m
2
a
2
V
2
0
4
E =
ma
2
V
2
0
2
2
In this case, there is no other bound state because E and the wavefunction do not
depend on any integer.
5) In class we derived Shankar (5.2.23)
k tan ka =
for the even solutions in a nite well of depth V
0
. Repeat for the odd solutions and
show that satisfy:
k cot ka =
We can separate this problem in three parts, described in Figure 2, according to the
value of the potential. In the case when V = V
0
:
2
2m
d
2
dx
2
+V
0
= E
(x) = k
2
where
=
2
/x
2
, and k
2
=
2m
2
(V
0
E). When V = 0:
2
2m
d
2
dx
2
= E
(x) =
2
5
Figure 2: Sketch of the square well potential.
where =
2mE
2
. Therefore, following the notation
I
as the wavefunction of the part
(I) in Figure 2, we found:
I
(x) = Ae
kx
+Be
kx
= Ae
kx
II
(x) = C cos(x) +Dsin(x)
III
(x) = Ee
kx
+Fe
kx
= Fe
kx
Here, we can quickly said that B and E must vanish due to conditions in very large
and small values of x. Next, we need to satisfy boundary conditions:
I
(a) =
II
(a)
Ae
ka
= C cos(x) +Dsin(a)
Ae
ka
= C cos(x) Dsin(a) (1)
II
(a) =
III
(a)
Fe
ka
= C cos(x) +Dsin(a) (2)
I
(a) =
II
(a)
Ake
ka
= Csin(x) +Dcos(a)
Ake
ika
= Csin(x) +Dcos(a) (3)
II
(a) =
III
(a)
Fke
ka
= Csin(x) +Dcos(a) (4)
Solving this equations:
(1) (2) (AF)e
ka
= 2Dsin(a) (5)
(3) + (4) (AF)e
ka
= 2Dcos(a) (6)
(5)/(6)
1
k
=
tan(a)
k = cot(a)
6
6) a) Show that
j =
2mi
(
)
for the current density can be rewritten
j =
m
where (r) and (r) are the magnitude and phase of (r), respectively, i.e.,
= e
i
with and real.
Dening = e
i
, then
= e
i
, and the expression inside the brakets is:
= e
i
_
e
i
_
e
i
_
e
i
_
=
2
e
i
ie
i
r
+
2
e
i
ie
i
r
= 2
2
i
Replacing in the expression above:
j =
2mi
_
2
2
i
_
=
m
b) Use the result of (a) to do Shankar Ex. 5.3.2.: Convince yourself that if = c
,
where c is constant (real or complex) and
is real, the corresponding j vanishes.
Using the result of (a):
j =
2mi
(
)
=
2mi
(c
_
c
_
c
_
c
_
)
=
2mi
(c
cc
)
=
2mi
|c|
2
(
)
= 0 if
R
c) Use the result of (a) to show that the probability current density of a plane wave
(r) =
0
e
ikr
is what you would expect, based on its probability density and the
momentum (and thus velocity) of the particles.
Using the result of (a) with (r) =
0
e
ikr
:
j =
m
2
0
(k r) =
m
2
0
k =
p
m
2
0
7