Chapter 4: One-Dimensional Potentials

Up to now we discussed the meaning of the wave function as a solution of Schrodinger

equation, and how we obtain the observed quantities from the wave function. We apply
the Schrödinger equation to one-dimensional problems. One dimensional problems are
interesting since there exists many physical phenomena whose motion is in one
dimensional. The application of Schrodinger equation to one dimensional problem
enables us to compare the predictions of classical and quantum mechanics in a simple
setting.

The behavior of a particle of mass m in one-dimensional time-independent potential V(x) is

described by the Schrodinger equation
2
d 2 ( x)
− + V ( x) ( x) = E ( x)
2m dx 2
where E is the total energy of the particle. We need to specify V(x) and the boundary
conditions to solve this differential equation. Solutions of this equation yield the allowed
energy eigenvalues E and the corresponding wave functions that contains information
about the spatial state of the particle.

In this chapter, the eigenfunctions, eigenvalues and wave functions will be obtained by
solving the time-independent Schrodinger equation for different forms of the potential
energy function V(x). We then interpret the physical significance of these quantities.

The Potential Step

a) When a free neutron with an external kinetic energy K enters a nucleus, it
experiences a potential due to the strong nuclear force. Suppose a neutron emitted
with K=5 MeV by a fission event strikes a nucleus. What is the probability that the
neutron will be absorbed, thus initiating another fission event?

b) While a charged particle moves along the axis of two cylindrical electrodes held at
different voltages, its potential energy changes very rapidly when passing from
one to the other. Potential energy function can be approximated by a step potential.

1
For the step potential, the x-axis breaks up into two regions

0 x0
V( x ) = 
V0 x0

Since the potential is time-independent, problem is to solve the time-independent

Schrodinger equation. The Schrodinger equations in two regions ( x  0 and x  0 ) are
solved separately. They satisfy the requirements that the the eigenfunctions and their first
order derivatives are everywhere finite, single-valued and continuous.

There are two cases, E  V0 and E  V0

For the case E  V0 :

The incident beam with energy E  V0 comes from the left to the right

To construct u(x) we divide the x-axis into two domain: region I ( x  0 where V ( x) = 0 )
and region II ( x  0 where V ( x) = V0 ). Schrodinger equations for both regions become

 2 d 2 u I (x)
- = Eu I (x) for x  0
2m dx 2
 2 d 2 u II (x)
- + V0 u II (x) = Eu II (x) for x  0
2m dx 2

They are simplified as

d 2 u I (x) 2mE d 2 u I (x)

+ 2 u I (x) = 0 + k 2 u I (x) = 0
dx 2 dx 2
or
d 2 u II (x) 2m(E-V0 ) d 2 u II (x)
+ u II (x) = 0 + q 2 u II (x) = 0
dx 2 2
dx 2

2mE 2m(E - V0 )
where k1  k 2 = and k 2 = q 2 =
2 2
. These are second order linear
 2
2
differential equations with constant coefficients k and q. They have two linearly
independent solutions. The general solution becomes their superposition.

2
For region I ( x  0 )

d 2 u I (x)
k 2 u I (x) 0
dx 2
D2 k2 0 D2 k2 D1,2 k 1= ik

Linearly independent solutions are eikx and e-ikx , and the general solution becomes

u I (x) Aeikx Be-ikx

where eikx is the incoming wave (moving positive x-direction) with the amplitude A and
e ikx represents the reflected wave (moving negative x-direction) with the amplitude B.
We can take A=1 and B=R. Thus the eigenfunction for region I ( x  0 ) is
uI ( x) eikx Re ikx .

A net flux moving in the positive x-direction is obtained as

du I du I*
j u I* ( x) u I ( x)
2im dx dx
ikx *
eikx Re ikeikx ik Re ikx
ike ikx
ik R *eikx eikx Re ikx

2im
2 ikx 2 2
ik ikR e ikR* e 2ikx ik R ik ikR e 2 ikx
ikR* e 2ikx ik R
2im
k 2 2
1 R vI 1 R
m

k
where v I is the speed of the particle in region I. The term eikx with the flux is written
m
for the incoming wave. In fact if V(x) were zero everywhere, we would take eikx as the
solution for a free particle. The step potential causes a reflected wave Re ikx with the flux
k 2
R .
m

For region II ( x  0 ), the Schrodinger equation

d 2 u II (x)
2
+ q 2 u II (x) = 0
dx
D + q 2 = 0  D1,2 = iq
2

Linearly independent solutions are eiqx and e iqx

, and the general solution becomes

u II (x) Ceiqx De-iqx

3
Where eiqx is the transmitted wave (moving positive x-direction) with the amplitude C
(we use C=T). The coefficient D is called the amplitude of reflected wave (moving
negative x-direction). Since the term De − iqx represents a wave moving from the right
( x =  ) to the left, there is no such a wave in x  0 and the constant D therefore must
vanish. Thus the eigenfunction for region II ( x  0 ) becomes u II (x) = Teiqx + Deiqx .

The flux for the transmitted wave is obtained as

duII duII*
j uII* ( x) u II ( x)
2im dx dx

T *e iqx
iqTeiqx ( iq )T *e iqx
Teiqx
2im
q 2 2
T v II T
m

where v II is the speed of the particle in region II. Teiqx represents the transmitted wave
q 2
function with the flux T .
m

As a result, the physically acceptable solutions are obtained as

u I (x) = eikx + Re-ikx for x  0

u II (x) = Teiqx for x  0

2mE 2m(E - V0 )
where k1  k 2 = and k 2 = q 2 =
2 2
. Using the relation between the wave
 2
2
number and the wave length as k1 k 2 / I and k2 q 2 / II , we obtain that
k q I II . The form of the waves in x  0 and x  0 regions are given below

Similarly we may explain the speed of the particle ( v I in region x  0 and v II in region
x  0 ) as

4
 pI mv I k 2 / I

pII mv I q 2 / II

We see that the speed of the particle is inversely proportional to the wave length. We thus
find v II v I . It implies that particle slows down due to the step potential and certain
fraction of the incident particles is reflected from this potential.

Since there is no time-dependence in the problem, the conservation law given by

 
P ( x, t ) + j ( x, t ) = 0 implies that j ( x) is constant (independent of x). Hence the flux
t x
on the left must be equal to the flux on the right,

k 2 q 2
1 R T
m m

At the point x=0, the continuity property of the wave functions requires (Since any
wavefunction and its first derivative are continuous)

u I (x)|x =0 = u I (x)|x =0  1 + R = T
du I (x) du (x)
|x =0 = II |x =0  ik (1 − R) = iqT
dx dx

We can solve these equations for T and R to obtain

2k
T=
k+q
V0
1- 1−
k-q E
R= =
k+q V
1+ 1− 0
E

Thus, the reflected and transmitted fluxes become

2 2 2
2
k 2 k k q k k 1 q/k k 1 1 V0 / E
R 2
m m k q m k2 1 q / k m 1 1 V0 / E
2 2
q 2 q 2k k 4kq k 2 k k 1 1 V0 / E
T 2
1 R
m m k q m k q m m m 1 1 V0 / E

5
Notes:

• According to the classical mechanics, when E  V0 the particle does not reflected
dV ( x)
but slows down due to an impulsive force F = − . Quantum mechanically a
dx
certain fraction of the incident particles is reflected. This is the wave property of
the particle (partial reflection of light from an interface between two media is a
familiar phenomena).
• For E  V0 , we have k → q and the ratio
k 2
R 2
m reflected flux 2 k q
R 0 as q k
k incident flux k q
m
Reflection probability | R |2 is not zero for E  V0 . Only tends to zero in high
energy limit, E  V0 (correspondence principle again). Notice that reflection and
transmission probabilities satisfy the relation | R |2 + | T |2 = 1 . This is a simple
example of a scattering problem.
• For E = V0 or q / k = 0 , R = 1 and T = 0 that corresponds total reflection and zero
transmission.

Example:

Consider a neutron incident upon a nucleus with an external kinetic energy K=5 MeV
(which is typical for a neutron that has just been emitted from a nuclear fission).
Estimate the probability that the neutron will be reflected at the nuclear surface (the
surface potential is about 50 MeV).

E = V0 + K = 50MeV + 5MeV = 55MeV

2 2
1 − 1 − V0 / E1 − 1 − 50 / 55
R = =  0.29
2

1 + 1 − V0 / E 1 + 1 − 50 / 55

This estimates a correct idea of the reflection phenomena when low-energy neutron
collides with nuclei. But the numerical value we have obtained for the reflection
coefficient is not accurate since the actual neutron-nucleus potential does not drop
quite as rapidly at the nuclear surface as a step potential.

6
For the case E  V0 :
If the energy of incident beam is less than V0 , we then have

Classical mechanics says that the particle cannot enter the region x  0 . The reason is
that E = V0 + K  K = E − V0  0 . The kinetic energy would be negative in the region
x  0 , which would lead to an imaginary value for the linear momentum p in the region
( K = p 2 / 2m  0  p → ip ). Neither is allowed in classical mechanics.

To determine the motion of the particle according to quantum mechanics, we must find
the wave function which is solution to Schrodinger equation with E  V0 .

Schrodinger equations for both regions are

 2 d 2 u I (x)
- = Eu I (x) for x  0
2m dx 2
 2 d 2 u II (x)
- + V0 u II (x) = Eu II (x) for x  0
2m dx 2

They are simplified as

d 2 u I (x) 2mE d 2 u I (x)

+ 2 u I (x) = 0 + k 2 u I (x) = 0
dx 2 dx 2
or
d 2 u II (x) 2m(V0 − E) 2
d u II (x)
− u II (x) = 0 − q '2 u II (x) = 0
dx 2 2
dx 2

2mE 2m(V0 − E)
where k 2 = 2
and q '2 = 2
 0 are positive constants. Here, we use

V0 − E  0 since E  V0 .

For region I ( x  0 ), the general solution becomes

d 2 u I (x)
k 2 u I (x) 0 u I (x) eikx Re-ikx
dx 2

For region II ( x  0 ), the Schrodinger equation

7
 d 2 u II (x)
− q '2 u II (x) = 0
dx 2
D2 − q '2 = 0  D1,2 =  q '

Linearly independent solutions are e q ' x and e q'x

, and the general solution becomes

uII ( x) Ceq ' x De q'x

q'x
Since the wave function must be finite everywhere, and since lim e , the
x

constant C has to be zero. Thus the eigenfunction for region II ( x  0 ) becomes.

uII ( x) = Ceq ' x + Te− q ' x .

where Te −q ' x represents the transmitted wave into x  0 region. The physically
acceptable solutions are found as

u I ( x) = eikx + Re − ikx for x  0

u II ( x) = Te − q ' x for x  0

2mE 2m(V0 − E)
where k12  k 2 = 2
and k22  q '2 = 2
 0 . The form of the waves in x  0
and x  0 regions are given below

At the point x=0, the continuity property of the wave functions requires

u I (x)|x =0 = u I (x)|x =0  1 + R = T
du I (x) du (x)
|x =0 = II |x =0  ik (1 − R) = −q 'T
dx dx

We can solve these equations for T and R to obtain

2k k - iq'
T= and R =
k + iq' k + iq'

Te −q ' x illustrates an important difference between classical and quantum physics. While
none of the particles can be found classically in the region x>0, quantum mechanically
there is a nonzero probability density that requires the wave function penetrates this

8
classically forbidden region (the wave function decreases exponentially in the classical
forbidden region). The reflection probability is

2 *
2 k iq ' k iq ' k iq ' k iq ' k iq ' k2 q '2
R 1
k iq ' k iq ' k iq ' k iq ' k iq ' k2 q '2

This means that a particle incident upon the potential step, with E  V0 , has the
probability one being reflected (it is always reflected). This is in agreement with the
predictions of classical mechanics.

While none of the particles can be found classically in the region x  0 , quantum
mechanically there is a nonzero probability that the wave function penetrates this
classical forbidden region with the probability density

4k 2
P( x ) =| u II ( x ) |2 =| T |2 e -2q' x = e -2q' x
k 2 + q' 2

This shows that the probability decreases rapidly with increasing x and this shows a finite
probability of finding the particle in the region x  0 .

All classical particles are reflected at the boundary. The quantum mechanical wave
packet, on the other hand, can penetrate a short distance into the forbidden region. The
classical particle is never directly observed in that region since E  V0 and therefore the
kinetic energy would be negative in region II ( x  0 region).

The penetration distance Δx at which the probability density is

1
P( x x0 x) P ( x x0 )
e
2 1 2 q'x 2 1 q'x 2
uII ( x) x x0 x
u II ( x ) x x0
Te Te
e x x0 x e x x0

2 2 q '( x0 x) 1 2 q ' x0 1 1
T e T e e 2q ' x 2q ' x 1 x
e e 2q '
1 1
x
2q ' 2 2m(V0 E ) / 2 2 2m(V0 E)

In classical limit, the term 2m(V0 − E) is so large compared to , thus Δx is

immeasurably small.

From Heisenberg uncertainty relation, we obtain

x p 1 1
2 x x min
2 2m(V0 E) 2q ' 2q '
p 2m(V0 E)

9
The probability density | u II ( x ) |2 for a group wave function incident on the step with
E  V0 . The group moves up to the step, penetrates slightly into the classically forbidden
region, and then is completely reflected from the step.

• There is tunneling of
particles into classically
forbidden region even for
energies below step height.
• Tunneling depth depends on
energy difference V0 − E ,
1 
Δx = =
2q' 2 2m(V 0 − E)
• But no transmitted particle
flux, 100% reflection, like
classical case.

Example:
An electron with a kinetic energy of 6 eV at large negative values of x is moving from
left to the right along x-axis. The potential step is given by
0 if x  0 : Region I
V(x) = 
10 eV if x  0 : Region II
( Use c = 197.3  10 −9 eV  m , me c 2 = 0.511 106 eV and wave number k = 2 /  )
Calculate the probability for finding the electron near x = 10−10 m .

2 q'x 2 2 2 q 'x
P uII ( x) Te T e
2
2 2k 4k 2 4(2mE / 2 ) 4E 4 6eV
T 2.4
k iq ' k 2 q '2 2mE / 2 2m(V0 E ) / 2
V0 10eV
2m(V0 E) 2mc2 (V0 E) 2(0.511 106 eV)(10 6) eV 1
q' 2
1010
( c) 2 (197.3 10 9 eV.m) 2 m
2 2 q 'x 2 1010 10 10
P T e 2.4 e 0.11 11%

10
Example:
Estimate the penetration distance x for a very small dust particle, of radius r 10 6 m and
density 104 kg / m3 moving at the very low velocity v 10 2 m / s , if the particle
incidents on a potential step with E V0 / 2 in the region to the left of the step.

4 3 18
m r 4 10 m3 104 kg / m3 4 10 14
kg
3
1 2 1 14
K mv 4 10 kg 10 4 m 2 / s 2 2 10 18
J
2 2
34
10 J .s 19
x 10 m
14 18
2 2m(V0 E) 2 2mK 2 2 4 10 kg 2 10 J

This is many orders of magnitude smaller than could be detected in any possible
measurement. For the more massive particles and higher energies typically considered in
classical mechanics, x is even smaller.

Example:
A conduction electron moving through a block of Cu at total energy E under a potential
V0 E ; i.e. measurements of the work function show that V0 E 4eV . From these data
estimate the distance x that the electron can penetrate the classical forbidden region
outside the block.

19
1.6 10 J 19
V0 E 4 eV 6 10 J
1eV
34
10 J .s 10
x 0.5 10 m
31 19
2 2m(V0 E) 2 2 9 10 kg 6 10 J

The penetration distance is of the order of atomic dimensions. Therefore, the effects can be
of consequence in atomic systems.

11
The Potential Well

We now consider some examples of scattering and bound states. The energy quantization
can be expected only for potentials which are capable of binding a particle. The square
well potential is often used in quantum mechanics to represent a situation in which a
particle moves in a restricted region of space under the influence of forces.

 0 x  -a region I

V( x ) = − V0 - a  x  a region II
 0 x  a region III


a) The motion of a neutron in a nucleus can be approximated by assuming that the

particle is in a square well potential with a depth about 50 MeV.
b) A square well potential results from superimposing the potential acting on a
conducting electron in a metal.

If the energy is greater than zero, this represents a scattering problem and the states correspond
unbound states in which total energy E is continuous (not quantized). If the energy is less
than zero, the states correspond the bound states that are expected discrete states.

12
The scattering solutions (E > 0):

When E 0 , the particle is unconfined and this case corresponds to the scattering
problems. There are unbound states with a continuum range of energies.

 0 x  -a region I

V( x ) = − V0 - a  x  a region II
 0 x  a region III


Classically, if the particle is initially incident from left with constant momentum 2mE ,
it will speed up to 2m( E V0 ) between a x a , and then slow down to its initial
momentum in the region x a . All the particles that come from the left will be
transmitted, none will be reflected back, therefore T=1 and R=0. Quantum mechanically,
we get a finite reflection coefficient ( R 0 ). The wave functions will display as
oscillating pattern in all three regions.

Let us write the Schrodinger equation in these three regions.

2
d 2 u I (x)
− + 0u I (x) = Eu I (x) ; x  -a
2m dx 2
2
d 2 u II (x)
− − V0 u II (x) = Eu II (x) ; - a  x  a
2m dx 2
2
d 2 u III (x)
− + 0u III (x) = Eu III (x) ; xa
2m dx 2

The accepted solutions become

d 2 u I (x) 2mE
2
+ 2 u I (x) = 0  u I (x) = eikx + Re-ikx ; x  -a
dx
d u II (x) 2m(E + V0 )
2

2
+ 2
u II (x) = 0  u II (x) = Aeiqx + Be -iqx ; -a  x  a
dx
2
d u III (x) 2mE
2
+ 2 u III (x) = 0  u III (x) = Teikx + Ge -ikx ; xa
dx

2mE 2m(E + V0 )
where k 2 = 2
and q 2 = 2
. Since the term Ge −ikx represents a wave moving
from the right ( x =  ) to the left in x  a , there is no such a reflected wave in x  a and
the constant G therefore must vanish. Thus the eigenfunction for region III ( x  a )
becomes u III (x) = Teikx + Ge-ikx .

13
 k
The eigenfunctions correspond to an incoming flux from the left, a reflected flux
m
k 2 k 2
R , and a transmitted flux T to the right. According to the flux conservation
m m
we have

k
m
(
1− R =
2 q
m
)
A −B =
2 2 k
m
T ( ) 2

At the point x = − a , the continuity property of the wave functions requires

u I (x)|x =-a = u I (x)|x =-a  e-ika + Reika = Ae-iqa + Beiqa

du I (x) du (x)
|x =− a = II |x =− a  ik (e-ika − Reika ) = iq( Ae-iqa − Beiqa )
dx dx

At the point x = a , the continuity property of the wave functions requires

u II (x)|x =a = u III (x)|x =a  Aeiqa + Be-iqa = Teika
du II (x) du (x)
|x = a = III |x = a  iq( Aeiqa − Be-iqa ) = ikTeika
dx dx
1 du(x)
It is useful to combine the conditions into a singlet set as at x =  a . But we
u(x) dx
will not use this in our derivations.

To find A and B in terms of T we use the above relations as

e-ika + Reika = Ae-iqa + Beiqa  q − k i (k +q )a

 B = e T
ik (e -ika
− Re ) = iq( Ae
ika -iqa
− Be ) 
iqa
2q
Aeiqa + Be −iqa = Teika  q + k i ( k −q ) a
 A= e T
iq( Aeiqa − Be−iqa ) = ikTeika  2q

Let us use these expressions into the first expression obtained at the point x = − a

e − ika + Reika = Ae − iqa + Beiqa

q + k i ( k − q ) a −iqa q − k i ( k + q ) a iqa
= e Te + e Te
2q 2q
eikaT  −2 iqa

=  (q + k ) e + (q − k ) e 2iqa 
2q  cos 2 qa −i sin 2 qa cos 2 qa + i sin 2 qa 

eikaT
e − ika + Reika =  2q cos 2qa − 2ik sin 2qa 
2q
T
e −2ika + R =  2q cos 2qa − 2ik sin 2qa  ( Eq. 1)
2q

14
and into the second expression obtained at the point x = − a

ik (e −ika − Reika ) = iq ( Ae −iqa − Beiqa )

q + k i ( k − q ) a -iqa q − k i ( k + q ) a iqa
= iq e Te − iq e Te
2q 2q
iT ika  
= e  (q + k ) e −2iqa − (q − k ) e 2iqa 
2  cos 2 qa −i sin 2 qa cos 2 qa + i sin 2 qa 

T
e −2ika − R =  2k cos 2qa − 2iq sin 2qa  ( Eq. 2 )
2k

We use Eq 1 and Eq 2 to find R in terms of T,

 2q cos 2qa − 2ik sin 2qa 

T
e −2ika + R =
2q  q2 − k 2
  R = iT sin 2qa
2 kq
 2k cos 2qa − 2iq sin 2qa 
T
e −2ika − R =
2k 

This relation is used into Eq. 1, we then obtain T as

T
e −2ika + R =  2q cos 2qa − 2ik sin 2qa 
2q
q2 − k 2 T
e −2ika + T sin 2qa =  2q cos 2qa − 2ik sin 2qa 
2kq 2q
2kq
T = e-2ika
2kq cos 2qa − i (q 2 + k 2 ) sin 2qa
and R as
q2 − k 2
R = iT sin 2qa
2kq
2kq q2 − k 2
R = ie-2ika sin 2qa
2kq cos 2qa − i (q 2 + k 2 ) sin 2qa 2kq
(q 2 − k 2 ) sin 2qa
R = ie-2ika
2kq cos 2qa − i (q 2 + k 2 ) sin 2qa

As a result, we obtain the transmission and reflection coefficients as

2kq
T = e-2ika
2kq cos 2qa − i (q 2 + k 2 ) sin 2qa
(q 2 − k 2 ) sin 2qa
R = ie-2ika
2kq cos 2qa − i (q 2 + k 2 ) sin 2qa

To find the transition probability, let us use the above expression

15
 T = T *T
2

 2kq   −2ika 2kq 

= e 2ika   2kq cos 2qa − i (q + k ) sin 2qa 
e
 2kq cos 2qa + i (q + k ) sin 2qa  
2 2 2 2

(2kq) 2
=
(2kq cos 2qa) 2 + [(q 2 + k 2 ) sin 2qa ]2
(2kq) 2
=
(2kq) 2 cos 2 2qa + (q 4 + 2k 2 q 2 + k 4 ) sin 2 2qa
1−sin 2 2 qa

1
=
(q − k 2 ) 2
2
1+ sin 2qa
(2kq) 2
1
T =
2
2
V0
1+ sin 2 (2qa)
4 E ( E + V0 )

2m( E + V0 ) 2mE 2mV0 2mE 2m( E + V0 )

where q 2 − k 2 = 2
− 2
= 2
and k 2 q 2 = 2 4
.

Notes

a) If E V0 ,
2m( E + V0 ) 2mE 2mV0 
q2 − k 2 = 2
− 2 = 2 
 q −k 2kq  R  0
2 2

+
= 2 E ( E + V0 ) 
2mE 2 m ( E V ) 4m
2kq = 2 0
2 2 
There is no reflection

2mE
b) As E → 0, k 2 = 2
→ 0 and T = 0
sin 2qa = 0  R = 0 no reflection
2qa = n , n = 1, 2,3,...
c) 2m( E + V0 ) 2m( E + V0 ) 2 2
2a 2
= n  4a 2
2
= n   E = −V0 +
2 2
2
n2
8ma
E = −V0 + n E1 , 2
n = 1, 2,3,...

This is called the transmission resonance (Ramsaver-Townsend). The resonance

condition (perfect transmission occurs) 2qa = n may be written in the form
2 4a
= = that corresponds to transmission without reflection in optics.
q n

16
Transmission probability is given by

17
Example:

A particle of mass m and energy E 0 encounters a potential well of width a and depth
V0
0 x a region I
V ( x) V0 -a x a region II
0 x a region III
For which values of a will the transmission be complete if the particle is an electron of
energy 1eV and V0 4eV ?
2kq
T = e-2ika
2kq cos 2qa − i (q 2 + k 2 ) sin 2qa
( 2kq ) ( 2kq )
2 2

T = =
2

( 2kq )
2
cos 2 2qa + (q 2 + k 2 ) 2 sin 2 2qa ( 2kq )
2
(1 − sin 2
2qa ) + ( q 4 + k 4 + 2q 2 k 2 ) sin 2 2qa

( 2kq ) ( 2kq )
2 2

= =
( 2kq ) − 4q 2 k 2 sin 2 2qa + (q 4 + k 4 + 2q 2 k 2 ) sin 2 2qa ( 2kq ) + (q 4 + k 4 − 2q 2 k 2 ) sin 2 2qa
2 2

1
=
(q − k )
2 2 2
1+ sin 2 2qa
( 2kq )
2

2m ( E + V0 ) 2mE 2mV0
q2 − k 2 = − =2 2 2

2m ( E + V0 ) 2mE 4m 2 ( E + V0 ) V0
q2k 2 = 2 2
= 2

1 1
T = =
2
2 2
 2mV0  1+
V 0
sin 2 (2qa)
 2  4 E ( E + V0 )
1+   sin 2 2qa
4m 2 ( E + V0 ) V0
4 2

1 V02
T =  1  1 = 1+ sin 2 (2qa)  sin 2 (2 qa) = 0
2

V02 4 E ( E + V0 )
1+ sin 2 (2qa)
4 E ( E + V0 )
2qa = n , n = 1, 2,...
2m ( E + V0 )
2a 2
= n

n n n n
2a = = = =
2m ( E + V0 ) 2mc 2 ( E + V0 ) 2 ( 0.5MeV ) ( 5 10−6 MeV ) 11.3 10−6 fm −1
( c) (197.3MeV  fm )
2 2 2

2a = 0.28 106 fm = 0.28  106  10−15 m = 2.8 10−10 m

18
The Potential Barrier

The potential barrier is a region in which particles (as alpha particles, photoelectrons,
etc.) are decelerated or stopped by a repulsive force.

 0 x  -a region I

V( x ) =  V0 - a  x  a region II
 0 x  a region III


According to classical theory, a particle can pass the barrier if its energy is much than the
height of the potential barrier. However, in quantum theory, there is a finite probability
that a particle with less energy will pass through the barrier (tunnel effect). A particle that
behaves as a matter wave has a non-zero probability of penetrating the barrier and
continuing its travel as a wave on the other side.

Examples:
• A potential barrier surrounds the atomic nucleus
• A barrier exists at the interface between semiconductors and metals and between
differently doped semiconductors. These barriers are important in the design of
electronic devices.

19
The case E  V0 (scattering problems):

Potential barrier represents a repulsive interaction potential. A particle of energy E  V0

interacts with a repulsive potential in a given region interval, and it is free outside this
interval. For example, electron-electron scattering (Coulomb interaction between
electrons is repulsive).

According to classical mechanics, the particles of the total energy E that approach the
barrier from the left at constant momentum (velocity) p1 = 2mE will show down to a
momentum during they enter the region − a  x  a , and they then pass beyond the
barrier with the momentum p3 = 2mE . The particles have enough energy and they will
emerge on the right side (total transmission). According to quantum mechanics, there will
be some refection from the barrier.

In this case, the particle interacts with the potential in the interval − a  x  a . In the
external regions the particle is free.

Let us write the Schrodinger equation in these three regions.

2
d 2 u I (x)
− + 0u I (x) = Eu I (x) ; x  -a
2m dx 2
2
d 2 u II (x)
− +V0 u II (x) = Eu II (x) ; -a  x  a
2m dx 2
2
d 2 u III (x)
− + 0u III (x) = Eu III (x) ; xa
2m dx 2

The accepted solutions become

d 2 u I (x) 2mE
+ 2 u I (x) = 0  u I (x) = eikx + Re-ikx ; x  -a
dx 2
d 2 u II (x) 2m(E-V0 )
2
+ 2
u II (x) = 0  u II (x) = Aeiqx + Be -iqx ; -a  x  a
dx
2
d u III (x) 2mE
+ 2 u III (x) = 0  u III (x) = Teikx + Ge-ikx ; xa
dx 2

2mE 2m(E-V0 )
where k 2 = 2
and q 2 = 2
. Since the term Ge −ikx represents a wave moving
from the right ( x =  ) to the left in x  a , there is no such a reflected wave in x  a and

20
the constant G therefore must vanish. Thus the eigenfunction for region III ( x  a )
becomes u III (x) = Teikx + Ge-ikx .

At the point x = − a , the continuity property of the wave functions requires

u I (x)|x =-a = u I (x)|x =-a  e-ika + Reika = Ae-iqa + Beiqa

du I (x) du (x)
|x =− a = II |x =− a  ik (e-ika − Reika ) = iq( Ae-iqa − Beiqa )
dx dx

At the point x = a , the continuity property of the wave functions requires

u II (x)|x =a = u III (x)|x =a  Aeiqa + Be-iqa = Teika
du II (x) du (x)
|x = a = III |x = a  iq( Aeiqa − Be-iqa ) = ikTeika
dx dx

These expression are the same that was found in the potential well problem with
2m(E-V0 ) 2m(E+V0 )
q2 = 2
→ 2
. Therefore we find the following expressions for R and T

2kq
T = e-2ika
2kq cos 2qa − i (q 2 + k 2 ) sin 2qa
(q 2 − k 2 ) sin 2qa
R = ie-2ika
2kq cos 2qa − i (q 2 + k 2 ) sin 2qa

We finally obtain the transition probability as

1
T =
2

(q − k 2 ) 2
2
1+ sin 2qa
(2kq) 2
1
= 2
V0
1+ sin 2 (2qa)
4 E ( E − V0 )

Special cases:

21
 2 2
a) If E V0 , T 1 and R 0 . This implies that at very high energies and weak
potential barrier, the particles would not feel the effect of the barrier and there is no
reflection (there is totally transmission).

b) We also have total transmission when sin 2qa = 0  R = 0 no reflection . This

requires 2qa = n , n = 1, 2,3,...

2qa = n
2m( E − V0 ) 2m( E − V0 ) 2 2
2a 2
= n  4a 2 2
= n 2 2  E = V0 + 2
n2
8ma
E = V0 + n E1 ,
2
n = 1, 2,3,...

E
c) In the limit → 1, sin2qa 2qa
V0
1 1 1
T = → =
2

V0 2
V0 2
V 4a 2 m( E − V0 ) /
2 2 2
1+ sin 2 (2qa) 1 + (2qa) 2 1 + 0
4 E ( E − V0 ) 4 E ( E − V0 ) 4 E ( E − V0 )
2 1
T
2 m V0 a 2
1+ 2

22
 The case E  V0 :

Classically, we would expect total reflection. Every particle that arrives at the barrier
( x  − a ) will reflected back, no particle penetrate the barrier. Quantum mechanical
predictions differ sharply from their classical counterparts. There is a certain probability
that the particle will be transmitted through the barrier into the region x  a . The effect
of transitioning through a classical forbidden region is called tunneling.

Examples:

i) Emission of α particles from radioactive nuclei.

α–particles are held within the nucleus by a potential barrier which consists of an
attractive part (due to nuclear forces), and a repulsive part (due to the electrostatic
repulsion between the α–particle and the residual nucleus). The α–particle can
tunnel out of its potential well and escape from the nucleus.

ii) Fusion process

It would be possible to induce nuclear reactions using low–energy protons (about
0.5 MeV) without sufficient energy to overcome the electrostatic repulsion. The
same process is involved in the nuclear reactions that supply the Sun’s energy. The
principal reaction here involves the eventual formation of helium by the fusion of
hydrogen nuclei (protons). In laboratory fusion experiments the same principle
allows nuclear fusion at reachable temperatures.

iii) Tunnel diode

Quantum tunneling occurs when an electric current passes through a junction made
by twisting two copper wires together. in a tunnel diode, the height of the potential
barrier between two semiconductors is controlled externally, and the physical size
of the junction is so small that the flow of electrons can be turned on or off very
rapidly, within a few picoseconds.

iv) Cold emission electrons

The scanning tunneling electron microscope was developed by Binnig and Rohrer,
who were awarded the Nobel Prize in 1986. This remarkable device works by the
tunneling of electrons from the surface of a sample to the tip of a very fine needle.
It is so sensitive that the detail of the surface can be resolved to distances much
less than an atomic radius.

23
Let us write the Schrodinger equation for the energies E  V0 in these three regions.

2
d 2 u I (x)
− + 0u I (x) = Eu I (x) ; x  -a : region I
2m dx 2
2
d 2 u II (x)
− +V0 u II (x) = Eu II (x) ; - a  x  a : region II
2m dx 2
2
d 2 u III (x)
− + 0u III (x) = Eu III (x) ; xa : region III
2m dx 2

The Schrodinger equations and its solution are obtained as

d 2 u I (x) 2mE
Re gion I : 2
+ 2 u I (x) = 0  u I (x) = eikx + Re -ikx
dx
d u II (x) 2m(V0 − E )
2
Re gion II : − u II (x) = 0  u II (x) = Ae - x + Be x
dx 2 2

d 2 u III (x) 2mE

Re gion III : 2
+ 2 u III (x) = 0  u III (x) = Teikx + Ge-ikx
dx

2m(V0 − E )
where  2 = 2
 0 is defined as a positive constant. At the point x = − a , the
continuity property of the wave functions requires

u I (x)|x =-a = u II (x)|x =-a  e-ika + Reika = Ae a + Be − a

du I (x) du (x)
|x =− a = II |x =− a  ik (e-ika − Reika ) =  (− Ae a + Be − a )
dx dx
and at the point x = a , the continuity property of the wave functions requires
u II (x)|x =a = u III (x)|x =a  Ae − a + Be a = Teika
du II (x) du (x)
|x = a = III |x = a   (− Ae − a + Be a ) = ikTeika
dx dx

From these four equations, we can find the transmission and reflections amplitudes. This
can be obtained from T and R results for E  V0 by using q → i = i 2m(V0 − E ) / .

2kq q =i 2ki

T = e-2ika ⎯⎯⎯ → e-2ika
2kq cos 2qa − i (q + k ) sin 2qa
2 2
2ki cos(2i a) − i ((i ) 2 + k 2 ) sin(2i a)
(q 2 − k 2 ) sin 2qa q =i ((i ) 2 − k 2 ) sin(2i a)
R = ie-2ika ⎯⎯⎯ → i e -2ika

2kq cos 2qa − i (q 2 + k 2 ) sin 2qa 2ki cos(2i a) − i((i ) 2 + k 2 ) sin(2i a)

Let us use
eiy − e−iy   ei y − e − i y e − y − e y
2 2
1
siny =  sin(iy ) = = = − sinh y  sinh y = −i sin(iy )
2i   2i 2i i
−y 

e −e  
y −i2 y
e −e i2 y
e −e
y −y
1
sinhy = sin(−iy ) = = = sinh y  sinh y = i sin(−iy )
2    2i 2i i
24
Similarly
eiy + e−iy   ei y + e − i y e − y + e y
2 2

cosy =   cos(iy ) = = = cosh y  cosh y = cos(iy)

2 2 2

e y + e− y   e −i2 y
+ e i2 y
e y + e− y
coshy = cos( −iy ) = = = cosh y  cosh y = cos(−iy)
2   2 2

A→iB
sin 2 A + cos 2 A = 1 ⎯⎯⎯ → sin 2 (iB) + cos 2 (iB) = 1
sin(iB) + cos(iB) = 1
2 2

i sinh(B) + cosh( B) = 1  cosh( B)  − sinh( B)  = 1

2 2 2 2

We then find

2 k 2 k
T = e-2ika = e-2ika
2k cos(2i a) − (− 2 + k 2 ) sin(2i a) 2k cosh(2 a) − i(k 2 −  2 ) sinh(2 a)
cosh(2 a ) i sinh(2 a )

(− − k ) i sin(2i a)
2 2
( 2 + k 2 ) sin(2 a)
R = ie-2ika = e-2ika
2ki cos(2i a) − i(− + k ) sin(2i a)
2 2
2ik cosh(2 a) + (k 2 −  2 ) sinh(2 a)
cosh(2 a ) i sinh(2 a )

Transition probability (the probability of penetrating the barrier) is found as

(2k ) 2 (2k ) 2
T = =
2

(k 2 −  2 ) 2 sinh 2 (2 a) + (2k ) 2 cosh 2 (2 a) (k 2 +  2 ) sinh 2 (2 a) + (2 k ) 2

(k 2 + 2 )2 − (2 k ) 2 1+ sinh 2 (2 a )

1 1
= =
(k +  )  2mE + 2m(V0 − E )  /
2 2 2 2 4
1+ sinh 2
(2 a ) 1+ sinh 2 (2 a)
(2k ) 2 4 2mE (V − E) / 4 0

1
=
V sinh 2 (2 a)
2
1+ 0
4 E (V0 − E)

This means that there is transmission even though the energy lies below the potential
barrier. This is called tunneling of particles. Tunnel effect is quite common in atomic,
nuclear and solid state physics. For the energies E  V0 , no penetration of the barrier
would occur in classical physics.

25
Note that reflection probability is obtained as

(2k ) 2 (k 2 +  2 ) 2 sinh 2 (2 a)

R = 1− T = 1− =
2 2

(k 2 +  2 ) sinh 2 (2 a) + (2k ) 2 (k 2 +  2 ) 2 sinh 2 (2 a) + (2 k ) 2

1 1
= =
(2k ) 2
4 E (V0 − E)
1+ 2 1 +
(k +  2 ) 2 sinh 2 (2 a) V0 sinh 2 (2 a)
2

or

2
( 2 + k 2 ) sinh(2 a)  ( 2 + k 2 ) 2 sinh 2 (2 a)
R = =
2

 2k cosh(2 a) + (k 2 −  2 ) sinh(2 a)  (2k ) cosh (2 a) + (k −  ) sinh (2 a)
2 2 2 2 2 2 2 2

1+ sinh 2 (2 a ) ( 2 + k 2 ) − (2 k )2

(k 2 +  2 ) 2 sinh 2 (2 a) 1 1
= = =
(k 2 +  2 ) 2 sinh 2 (2 a) + (2k ) 2 (2k ) 2
4 E (V0 − E)
1+ 2 2 2 1+ 2
(k +  ) sinh (2 a)
2
V0 sinh 2 (2 a)

Classically we would expect | T |2 = 0 (the particle is not permitted to enter the forbidden
region). The quantum wave can penetrate the barrier and give a nonzero probability to
find a particle beyond the barrier.

For E  V0 , the wave function uII ( x) does not vanish inside the barrier, and thus there
appears to be some probability of finding a particle with negative kinetic energy
( E = K + V  K = E − V where E  V ). How can this make sense?
An experiment to study the particle inside the potential barrier must be able to localize it
with an accuracy x 2a . This measurement will transfer to the particle’s momentum
with an accuracy
xp   p
2a
Which corresponds to a transfer of energy

26
 (p)2
2
1   2 2
E = =  E
2m 2m  2a  8ma 2 8ma 2
In order to observe the negative kinetic energy, this uncertainty must be less than
E − V0 , so that
2

E 2 
8ma  
2 2 2
E E − V0   E  2 a 1
2m 8ma 2
2m( E − V0 ) 
 =
2
2


When κa  1 ,

2 m(V0 − E)
2
a2 1 E V0

e 2 a − e −2 a 1
sinh(2 a) =  e 2 a
2 2

1 1
| T |2 = ⎯
⎯κa ⎯1→
1+
(k 2
+κ )
2 2
sinh (2κκ)
2
1+
(k 2
+ κ2 )
2
1 4 κa
e
(2kκ )2 (2kκ )2 4


(4kκ )2 e −4 κa =
16E 
1 −
E  −( 4 a /  )
e 2 m ( V0 − E )

(k 2
+κ )
2 2 V0  V0 

The fact that | T |2 does not vanish for E  V0 , but it is very small which is a purely
quantum mechanical result. | T |2 is extremely sensitive function of the width of barrier
and V0 − E . This result shows that the transmission coefficient is not zero, but has a finite
value. So, quantum mechanically, there is a finite tunneling beyond the barrier, x  0 .

 T 2 → 0
In the classical limit, →0 2
 R → 1

27
Example:

An electron with energy E = 6 eV is incident from left upon a rectengular barrier of

potential energy V0 = 10 eV , and it then penetrates beyond the barrier.

Using the transmission probability as

16E  E  −( 4 a /  ) 2 m ( V0 − E )
| T |2 = 1 − e
V0  V0 
find the width of the barrier so that the transmission probability will be 1  10 -3 .
(Hint: use c = 197.3 MeV  fm , m e c 2 = 0.511 MeV and 1fm = 10 −13 cm ).

16 E E − 4a 2 mc 2 (V0 − E )
T = (1 − )e c
2

V0 V0
16  6eV
2d
6eV − (197.3MeV . fm ) 20.511MeV (V 10 − 6)10−6 MeV
= (1 − )e  1.0 10 −3
10eV 10eV fm−1
e −2.05 d
3.84

−1 1.0 10−3
e −2.05 d fm =
3.84
−2.05 d fm −1 1.0 10−3
ln e = ln  −2.05d fm −1 = −8.25  d = 4.02 fm = 4.02  10−10 m
3.84

28
Example:
A particle with energy E is incident from left upon a rectengular barrier of potential
energy V0 .
0 if x  0 : Region I

V(x) = V0 if 0  x  a : Region II
0 if x  a
 : Region III
Consider the special case when the particle energy is equal to the barrier height, E = V0 .
a. Write down the Schrodinger equations for all regions.
b. Solve the equations in part (a) defining wavenumbers and find the physically
acceptable solutions as
u I ( x ) = e ikx + Ae − ikx
u II ( x ) = Bx + C
u III ( x ) = De ikx
where k 2 = 2mE /  2 and A, B, C and D are constants.
c. State the continuity conditions that must be satisfy by the wave function at x = 0
2
and x = a . Show that the reflection coefficient for the barrier R = A is given by
−1
 4 
R = 1 + 2 2 
 k a 
2
d. What would be the transmission coefficient T = D for this barrier if the incident
particles were electrons of energy 0.5 eV incident on a barrier of width
2  10 −10 m ?
(Hint: use c = 197.3 MeV  fm , m e c 2 = 0.511 MeV and 1fm = 10 −15 m ).

a. The Schrodinger equations :

2
d 2u I ( x )
Region I: − = EuI ( x)
2m dx 2
2
d 2uII ( x)
Region II: − + V0uII ( x) = Eu II ( x)
2m dx 2
2
d 2uIII ( x)
Region III: − = EuIII ( x)
2m dx 2

b. The solutions of Schrodinger equations :

d 2 u I (x) 2mE
Region I : 2
+ 2 u I (x) = 0  u I (x) = eikx + Ae -ikx
dx
2
d u II (x) 2m
Region II : + 2 (E − V0 ) u II (x) = 0  u II (x) = Bx + C
dx 2
=0
2
d u III (x) 2mE
Region III: 2
+ 2 u III (x) = 0  u III (x) = Deikx + Ge-ikx
dx
29
 c. At the point x = 0 , the continuity property of the wave functions requires
u I (x)|x =0 = u II (x)|x =0  1 + A = C 
 B
du I (x) du II (x) C = 2−
|x = 0 = |x =0  ik (1 − A) = B  ik
dx dx 
and at the point x = a , the continuity property of the wave functions requires
u II (x)|x =a = u III (x)|x =a  Ba + C = Deika 


B
2− 2ik
ik
 B = −
du II (x) du (x)  ika − 2
|x =a = III |x =a  B = ikDeika 
dx dx

B 1  2ik  2 ika − 2 + 2 ka
ik (1 − A) = B  A = 1 − = 1−  − ika − 2  = 1 + ika − 2 = ika − 2  A = ka + 2i
ik ik
 ka   ka   ka   ka   (ka ) 
* 2
1
A = = = =
2
        
 ka + 2i   ka + 2i   ka − 2i   ka + 2i   (ka ) + 4  1 + 4
2

k 2a2
−1
 4 
R  A = 1 + 2 2 
2

 k a 

d.nn
B 1 −2ik 2
B = ikDeika  D = e − ika = e − ika = e − ika
ik ik ika − 2 2 − ika
*
 2   − ika 2   + ika 2   − ika 2  4
T = D = e − ika = e =
2
  e    e 
 2 − ika   2 − ika   2 + ika   2 − ika  4 + (ka ) 2
4 4 4 4
T=  = −6
=
4 + 2 a 2 4 + 2mc 2E a 2 4 + 2  (0.511MeV )  (0.5 10 4 + 0.525
2
2mE MeV )
2
(2 105 fm) 2
( c) (197.3Mev.fm)
T = 0.884

Note that

−1 −1
 4   4 
R = 1 + 2 2  = 1 + = 0.116  T + R = 0.884 + 0.116 = 1
 k a   0.525 

30
Application of Tunneling
In general, the barriers that occur in physical phenomena are not square. For an irregular
shaped barrier. The theory of Wentzel, Kramers, and Brillouin (WKB) method provides
one of the most approximations methods.

The transmission probability is an extremely sensitive function of the width of the barrier
and of V0 − E . When  a is large (  a 1 or E V0 ), T is found as
2

 4k  −4 a 16 E
2
E − 4 ca 2 mc 2 (V0 − E )
T  2 = (1 − )e
2
e
 k +  
2
V0 V0
where  a = 2ma 2 (V0 − E ) / 2
. It will be
 (4kκ )2   4kκ 
ln | T |2 = ln e −4 κa  = −( 2κκ)(2a) + 2 ln 2
(
 k 2 + κ 2 )
2
  k +κ (
2 
)
The first term dominates the second one for any reasonable size of κa .

For a smooth potential barrier

If the transmission probability of each barrier is small, the overall probability is the
product of the individual ones. Equivalently, the overall transmission coefficient is a
product of the transition coefficients of the individual barriers

T  T1 T2 ...... TN
2 2 2 2

ln T  ln T1 + ln T2 + ...... + ln TN = 
2 2 2 2
ln Tpartial
partial barriers
barriers

 ( −2 
n
n
xn ) = −2 xn 
n
n

where xn is the width and  n

is the average value of  for that partial barrier.
2 2 2
T 2  T T ...... T
1 2 N
 
ln T 2 = lim  −2 xn  n  = −2  dx 2m  V(x) − E  /
2
x → 0  n 

−2 dx 2m  V(x) − E  / 2
T 2 = Ce
where the integration is over the region in which the square root is real. Note that in the
classical limit, → 0  T → e−1/0 = e− = 0 so that no tunneling exists in classical
2

physics.
31
Example 1: Cold emission electrons ((Nobel Prize, 1986)

Electrons needs a minimum energy (called work function) to escape from an atom. There
are some methods to emits electrons from a metal:
• Photoelectric effect
• By heating the metal
• Cold mission

Electrons in a metal can be removed at a room temperature by application of an external

electric field ε . This phenomena is called as cold emission. The electric field will change
the work function as W → W − eε x . The change creates a barrier of finite width and
electron can tunnel through it.

The overall transmission coefficient becomes

−2 dx 2m  V(x) − E  / 2 −2  dx 2m  W − e  x  / 2
T 2 = Ce → Ce

W
where V(x = a) = 0  W− e  a  a = . Let us use the following integration for the above
e
expression
1 1 ( A + Bx)3/ 2 ( A + Bx)3/ 2
 dx A + Bx =
B  dxB ( A + Bx)1/ 2 =
B 3/ 2
=
3B / 2

a
1 (2 mW − 2m e  x)3/ 2
a a
1
I =  dx 2m  W − e  x  / 2 =  dx 2mW −2m e  x =
0 0 A B 3(−2m e  ) / 2 0

 
 
 3/ 2 
a
1 1 2 1 1
=  dx 2mW −2m e  x = (2 mW − 2m e  a ) − (2 mW − 0)  = − 3m e   −(2 mW) 
3/ 2 3/ 2

3( −2 m e  )
0 A B  W
e

 
 =0 
1 2 2 W  2 2
=   mW = mWa 2 / 2

3 e  3
a

We finally find

 a  −
4 2
mWa 2 / 2

T 2 = exp  −2  dx 2m  W − e  x  / 2  = e 3 “Fowler-Nordheim formula”

 0 
32
Cold emission has an important application in the scanning tunneling microscope that has
been used for a study of surfaces of metals and some semiconductors. A conducting
probe with a very sharp tip is brought close to a metal. Electrons tunnel through the
empty space to the tip. Tunneling current is so sensitive to the metal (prob) distance
(barrier width) that even individual atoms can be mapped.

Example 2: Emission of α -particle from radioactive nucleus

The idea of tunneling was first used to explain α–decay in radioactive nuclei (in heavy
nuclei, α -particles may be formed with enough energy to escape). They are held within
the nucleus by a potential barrier which consists of an attractive part (due to nuclear
forces), and a repulsive part (due to the electrostatic repulsion between the α–particle and
the residual nucleus). Notice that the α–particle can tunnel out of its potential well and
escape from the nucleus.

Classically, the -particle cannot

enter the region r>R 0 . Quantum
mechanically, there is a tunneling
through such a barrier

0 for r  R

V(r) =  Z Z e2
k
1 2
for r  R
 r

The transmission probability (the probability of penetrating the barrier) is given by

 b 2m  Z1Z 2e 2
P  T  e −G = exp −2  dr
2
2 V ( r ) − E   E = V (r = b ) = k
 R  b

 2m 
1/2
b Z1Z 2e 2 2mZ1Z 2e 2  −1  R 1/2  R 1/2  R 1/2 
G = 2 2   dr k −E =2 k b cos   −   1 −  
  b  b   b  
2
R r 
where Z  = 2 and Z D = Z daughter  Z . At low energies, b  R .
 R  R 2
Cos -1 ( x) = − x + ....  Cos -1  − and R
  →0
2 b 2 b b

mα Ze 2 π R R  R  
2
mα Ze 2 π R
G=2 b − − −   =2 b  − 2 
πε 0 2
2 b b b  πε 0 2 2 b
 

33
 1 2Ze2 1 2Ze 2
Eα = mα vα2 =  b= 2
2 πε 0 2 b πε 0 mα vα2
mα Ze2 2Ze2 π mα Ze2 R Ze2 mα Ze2
G=2 2 −4 b = −4 R
πε 0 2
πε 0 mα vα2 2 πε 0 2 b πε 0 v πε 0 2

Thus the probability of penetrating the barrieras an function of Eα becomes

 Ze2 mα Ze2   Z 
P  e-G = exp  − +8 R  = exp  C1 − C 2 
 ε 0 v 4πε 0 2   E α 

where C1 and C2 are constants. We then find

Z
lnP = C1 − C 2
Eα
−G
The probability of escape at each collisions is e . Thus the probability of emission per
unit time is  v 0 e −G and hence the life time (mean life) of the parent nucleus is about
 2R 
 2R  G
τ =   e
 v0 
The emission of 4.2 MeV alpha-particles from 23892 U→ 23490Th + 42 He , the transmission
probability (the probability of penetrating the barrier) is obtained as

P  T = e−90  10−39
2

which is so small and correspondingly

1  v0 
λ= =  P=1.7  1021  10-39 s −1 =1.7  10-18 s −1
τ  2R 
 = 1/1.7 10-18 s = 4.11017 s  1.3 1010 y

The experimental half-life is 0.45 1010 y that is a remarkable agreement.

34
Example 3. Fusion occurs by tunneling nuclei through the Coulomb barrier

As an example, the fusion reaction of two deuterons ( 12 H ) is consider as

2
1 H + 12 H → 13 H + p + 4.0MeV (D-D reaction)
which relates to the tunneling process shown in the figure.

The Coulomb repulsive force between two charged nuclei must be overcome before they
can fuse. The electrical repulsion produces a barrier to this process called a Coulomb
barrier. The Coulomb barrier classically prevents low-energy particles to approach each
other. When the energy is not high, the probability of tunneling of the nuclei through the
Coulomb barrier is extremely small according to WKB. So only high energy nuclei can
fuse.

The fundamental problem is to give two nuclei enough kinetic energy to overcome this
repulsive force. For this, a sufficiently high temperature (perhaps 107 K) is required. Such
high temperatures are not easy to obtain in laboratory or a power planet.

The necessary energy to initiate fusion is the energy equivalent to the Coulomb barrier.
We can estimate the Coulomb repulsion barrier by calculating the electrostatic repulsion
of two light nuclei when they are just touching. If Z1q1 and Z2q2 are the charges of two
nuclei with radii r1 and r2, the Coulomb repulsion barrier becomes

1 ( Z1q1 )( Z2 q 2 ) ZZ
VC  = 1.44MeV . fm 1 2
4π ε 0 r1 + r2 r1 + r2

e2 1
where the value of the fine structure constant  = = gives
4π ε 0 c 137
e2
= 1.44 MeV . fm and the radius is given by r = r0 A1/3 = 1.2 A1/3 fm .
4π ε 0

35
This reaction ( 12 H + 12 H → 13 H + p + 4.0MeV ) can be performed in the laboratory by
2
accelerating a beam of deuterons 1
H onto a deuteron target. The radius of 21 H is
about r = 1.2 A1/3 fm = 1.2  ( 2 )
1/3
 1.5 fm and the electrostatic energy of the two
charges separated by r1 + r2 becomes

Z1Z2 1 1
VC  1.44 = 1.44 = 0.48 MeV  0.5 MeV
r1 + r2 (1.5 + 1.5) fm
As a result, a deuteron with 0.5 MeV of kinetic energy can over to Coulomb repulsion
and initiate a reaction. The temperature for this energy becomes

( )
E = k BT = 0.5 MeV  T = 0.5  106 / 0.862  10−4 eV/ K = 0.6  1010 K

where Boltzman constant is k B = 1.38  10−23 J / K = 0.862  10−4 eV/ K .

Even at the core of the sun, the temperature is about 107 Kelvin (the energy of this
temperature is about is kT → 1keV ), which is about a thousand times less than the
electrical potential energy ( VC MeV  T = 0.6 1010 K ). However, quantum
mechanical tunneling through Coulomb barrier occurs for T 106 − 107 K . Therefore
nuclear fusion reactions can only occur inside the sun, or in any other normal star,
through the process of quantum-mechanical tunneling. To get the probability of a
reasonable value one has to use high energy nuclei with the energy of the order of 107 K
that is less than the barrier height.

36