The document provides an overview of the WKB approximation method. It is used to find approximate solutions to the time-independent Schrodinger equation for problems with spatially varying potentials. The WKB method works by assuming the wavefunction can be written as an oscillating term multiplied by a slowly varying amplitude. It is valid when the potential does not change rapidly over the wavelength of the particle. The document outlines the derivation of the WKB approximation and discusses its application to problems with classical and non-classical regions, as well as the connection formulae used at turning points.
1. Introduction:
• Generally, the WKB approximation or WKB
method is a method for finding approximate
solutions to linear differential equations with
spatially varying coefficients.
• In Quantum Mechanics it is used to obtain
approximate solutions to the time-
independent Schrodinger equation in one
dimension.
2. Applications in Quantum Mechanics
In quantum mechanics it is useful in
1. Calculating bound state energies (Whenever
the particle cannot move to infinity)
2. Transmission probability through potential
barriers.
These are given in next slides.
3. Main idea:
• If potential 𝑽(𝒙) is constant and energy 𝐸 of the particle is 𝑬 > 𝑽,
then the particle wave function has the form
𝜓 𝑥 = 𝐴𝑒±𝑖𝑘𝑥 where 𝑘 = √(2𝑚(𝐸−𝑉))
ℏ
(+) sign indicates : particle travelling to right
(-) sign indicates : particle travelling to left
• General solution : Linear superposition of the two.
• The wave function is oscillatory, with fixed wavelength, 𝜆 =
2𝜋
𝑘
.
• The amplitude (A) is fixed.
4. • If V (x) is not constant, but varies slow in comparison
with the wavelength λ in a way that it is essentially
constant over many λ, then the wave function is
practically sinusoidal, but wavelength and amplitude
slowly change with x. This is the inspiration behind
WKB approximation. In effect, it identifies two
different levels of x-dependence :- rapid oscillations,
modulated by gradual variation in amplitude and
wavelength.
5. • If E<V and V is constant, then wave function is
𝜓 𝑥 = 𝐴𝑒±𝜅𝑥
where κ = √(2𝑚(𝑉−𝐸))
ℏ
If 𝑽(𝒙) is not constant, but varies slowly in
comparison with 1
𝜅 , the solution remain
practically exponential, except that 𝐴 and 𝜅 are
now slowly-varying function of 𝑥.
6. Failure of this idea
There is one place where this whole program is
bound to fail, and that is in the immediate vicinity
of a classical turning point, where 𝐸 ≈ 𝑉. For here
𝜆(𝑜𝑟 1
𝜅)) goes to infinity and 𝑉(𝑥) can hardly be
said to vary “slowly” in comparison. A proper
handling of the turning points is the most difficult
aspect of the WKB approximation, though the final
results are simple to state and easy to implement.
The diagram showing turning points is given in next
slide.
8. The Classical Region
• Let's now solve the Schrödinger equation using WKB
approximation
−ℏ2
2𝑚
𝑑2 𝜓
𝑑𝑥2 + 𝑉 𝑥 𝜓 = 𝐸𝜓 can be rewritten in the
following way:
𝑑2 𝜓
𝑑𝑥2 = −
𝑝2
ℏ2 𝜓 ; 𝑝 𝑥 = 2𝑚(𝐸 − 𝑉(𝑥)) is the
classical formula for the momentum of a particle with
total energy 𝐸 and potential energy 𝑉(𝑥). Let’s
assume 𝐸 > 𝑉(𝑥), so that 𝑝(𝑥) is real. This is the
classical region , as classically the particle is confined
to this range of 𝑥. The classical and non-classical region
is shown in the diagram on the next slide.
10. The function 𝜓
In general, 𝜓 is some complex function; we can
express it in terms of its amplitude 𝐴(𝑥), and its
phase, 𝜙(𝑥) – both of which are real :
𝜓 𝑥 = 𝐴(𝑥)𝑒 𝑖𝜙(𝑥)
11. Solving the Schrödinger equation
𝜓 𝑥 = 𝐴(𝑥)𝑒 𝑖𝜙(𝑥)
Using prime to denote the derivative with respect to
𝑥, we find:
𝑑𝜓
𝑑𝑥
= (𝐴′+𝑖𝐴𝜙′)𝑒 𝑖𝜙
and
𝑑2 𝜓
𝑑𝑥2
= [𝐴′′
+2𝑖𝐴′
𝜙′
+ 𝑖𝐴𝜙′′
− 𝐴 𝜙′ 2
]𝑒 𝑖𝜙
Putting all these into
𝑑2 𝜓
𝑑𝑥2 = −
𝑝2
ℏ2 𝜓 (From Schrödinger
equation ) , we get
12. 𝐴′′
+ 2𝑖𝐴′
𝜙′
+ 𝑖𝐴𝜙′′
− 𝐴 𝜙′ 2
= −
𝑝2
ℏ2
𝐴
Solving for real and imaginary parts we get,
𝐴′′ − 𝐴 𝜙′ 2 = −
𝑝2
ℏ2
𝐴
𝐴′′ = 𝐴 (∅′)2−
𝑝2
ℏ2
𝐴′′
𝐴
= (∅′)2−
𝑝2
ℏ2
The above equation cannot be solved in general, so we use WKB
approximation: we assume amplitude A varies slowly, so that the
A’’ term is negligible. We assume that 𝐴′′
𝐴 << [(∅′)2−
𝑝2
ℏ2].
Therefore, we drop that part and we get
𝜙 𝑥 = ±
1
ℏ
𝑝 𝑥 𝑑𝑥
13. And from second equation, we get
2𝐴′
𝜙′
+ 𝐴𝜙′′
= 0
𝐴2
𝜙′ ′
= 0
𝐴 =
𝐶
𝜙′
Where C is real constant.
14. Thus from the previous slides from the equations
and making ‘C’ a complex constant, we get
𝜓(𝑥) ≅
𝐶
𝑝(𝑥)
𝑒±
𝑖
ℏ
𝑝 𝑥 𝑑𝑥
And the general solution can be written as
𝜓(𝑥) ≅
1
𝑝(𝑥)
𝐶1 𝑒
𝑖
ℏ
𝑝 𝑥 𝑑𝑥
+ 𝐶2 𝑒
−𝑖
ℏ
𝑝 𝑥 𝑑𝑥
where 𝐶1 and 𝐶2 are constants.
15. Alternate approach
In this approach, the wave function is expanded
in powers of ℏ .
Let, the wave function be:
𝜓 = 𝐴𝑒
𝑖
ℏ
𝑆(𝑥)
Using this in:
𝑑2 𝜓
𝑑𝑥2 + 𝑘2
𝜓 = 0, we get
𝑑𝑆
𝑑𝑥
2
− 𝑖ℏ
𝑑2 𝑆
𝑑𝑥2 − 𝑘2
ℏ2
= 0 --------(1)
Expanding S(x) in powers of ℏ2
:
𝑆 𝑥 = 𝑆0 𝑥 + 𝑆1 𝑥 ℏ + 𝑆2 𝑥
ℏ2
2
+ ⋯
16. ∴ 1 ⟹
𝑑𝑆0
𝑑𝑥
2
− 𝑘2
ℏ 2
+ 2
𝑑𝑆0
𝑑𝑥
𝑑𝑆1
𝑑𝑥
− 𝑖
𝑑2 𝑆0
𝑑𝑥2 ℏ = 0 ;
(neglecting higher powers of ℏ).
where 𝑘2
ℏ 2
= 2𝑚[𝐸 − 𝑉(𝑥)]
For the above equation to be valid, the coefficient
of each power of ℏ must vanish separately,
𝑑𝑆0
𝑑𝑥
2
− 𝑘2
ℏ 2
= 0 𝑜𝑟
𝑑𝑆0
𝑑𝑥
= ±𝑘ℏ
⟹ 𝑆0 x = ± 𝑘ℏ𝑑𝑥
17. and, 2
𝑑𝑆0
𝑑𝑥
𝑑𝑆1
𝑑𝑥
− 𝑖
𝑑2 𝑆0
𝑑𝑥2 = 0
𝑜𝑟,
𝑑𝑆1
𝑑𝑥
=
𝑖
2𝑘
𝑑𝑘
𝑑𝑥
; using the value of 𝑆0(x)
𝑜𝑟, 𝑆1 =
𝑖
2
𝑙𝑛𝑘 or, 𝑒 𝑖𝑆1 = 𝑘−1 2
∴ 𝜓 = 𝐴𝑒
𝑖
ℏ
𝑆(𝑥)
= 𝐴𝑒
𝑖
ℏ
𝑆0(𝑥)+𝑆1(𝑥)ℏ
= 𝐴𝑒
𝑖
ℏ
𝑆0
𝑒 𝑖𝑆1
= 𝐴
1
𝑘
𝑒±𝑖 𝑘𝑑𝑥
𝑜𝑟 𝜓 =
𝐴
𝑘
𝑒±𝑖 𝑘𝑑𝑥
where A is a normalization
constant.
18. For 𝐸 < 𝑉 𝑥 : the Schrodinger equation is ,
𝑑2 𝜓
𝑑𝑥2 − 𝜅2
𝜓 = 0 ; 𝜅2
=
2𝑚[𝑉 𝑥 −𝐸]
ℏ2
Proceeding in a similar fashion , the solution can be
obtained as
𝜓 =
𝐵
𝜅
exp(± 𝜅𝑑𝑥) ; where B is a
normalization constant.
20. But we are interested in solving the following eqn.
𝑑2 𝜓0
𝑑𝑥2 + 𝑘2
𝜓0 = 0
Hence,
𝑑𝑘
𝑑𝑥
<< 𝑘2
⟹
1
𝑘
𝑑𝑘
𝑑𝑥
≪ 𝑘(𝑥) i.e. k(x) should not vary so rapidly
This is the Validity condition for WKB
approximation.
22. The WKB approximation
E V x
( )
( )
i
p x dx
WKB
C
x e
p x
h
E V x
Excluding the turning points:
1
( )
( )
p x dx
WKB
C
x e
p x
h
23. Patching region
The WKB approximation
V(x)
E
Classical region (E>V)
Non-classical
region (E<V)
( )
( )
i
p x dxC
x e
p x
h
1
( )
( )
p x dxD
x e
p x
h
30. Potential Square well with a Bumpy
Surface
Suppose we have an infinite square well with a
bumpy bottom as shown in figure:
and 𝑉 𝑥 =
𝑠𝑜𝑚𝑒 𝑠𝑝𝑒𝑐𝑖𝑓𝑖𝑒𝑑 𝑓𝑢𝑛𝑐𝑡𝑖𝑜𝑛, 𝑖𝑓 0 < 𝑥 < 𝑎
∞, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
𝑽 𝒙
𝒙𝟎
31. Inside the well, (𝑎𝑠𝑠𝑢𝑚𝑖𝑛𝑔 𝐸 > 𝑉 𝑥 𝑡ℎ𝑟𝑜𝑢𝑔ℎ𝑜𝑢𝑡),
we have
or,
𝜓 𝑥 ≅
1
𝑝(𝑥)
𝐶1 𝑠𝑖𝑛
1
ℏ 0
𝑥
𝑝 𝑥 𝑑𝑥 + 𝐶2 𝑐𝑜𝑠
1
ℏ 0
𝑥
𝑝 𝑥 𝑑𝑥
𝜓(𝑥) must go to zero at 𝑥 = 0 and 𝑥 = 𝑎. So, putting the
values we get respectively,
𝐶2 = 0 and 0
𝑎
𝑝 𝑥 𝑑𝑥 = 𝑛𝜋 ℏ ,this
quantization condition determines the allowed energies.
𝜓(𝑥) ≅
1
𝑝(𝑥)
𝐶1 𝑒
𝑖
ℏ 0
𝑥
𝑝 𝑥 𝑑𝑥
+ 𝐶2 𝑒
−𝑖
ℏ 0
𝑥
𝑝 𝑥 𝑑𝑥
32. Special Case:
If the well has a flat bottom i.e. (𝑉 𝑥 = 0), then
𝑝 𝑥 = 2𝑚𝐸
and from quantization equation, we get
𝑝𝑎 = 𝑛𝜋ℏ.
Solving these, we get value of :
𝐸 𝑛 =
𝑛2 𝜋2ℏ2
2𝑚𝑎2
which is the formula for the
discrete energy levels of the infinite square well.
34. In the non-classical region (𝐸 < 𝑉),
𝜓(𝑥) ≅
𝐶
𝑝(𝑥)
𝑒±
1
ℏ 0
𝑥
𝑝(𝑥) 𝑑𝑥
; 𝑝(𝑥) is complex
Let us consider the following example : problem of
scattering from a rectangular barrier also called
tunneling.
35. To the left of the barrier (𝑥 < 0),
𝜓 𝑥 = 𝐴𝑒 𝑖𝑘𝑥
+ 𝐵𝑒−𝑖𝑘𝑥
where 𝐴 is the incident amplitude and 𝐵 is the
reflected amplitude, and 𝑘 =
2𝑚𝐸
ℏ
.
To the right of the barrier (𝑥 > 𝑎),
𝜓 𝑥 = 𝐹𝑒 𝑖𝑘𝑥
where 𝐹 is the transmitted amplitude.
In the tunneling region (0 ≤ 𝑥 ≤ 𝑎) WKB
approximation gives,
𝜓 𝑥 ≅
𝐶
𝑝 𝑥
𝑒
1
ℏ 0
𝑥
𝑝 𝑥 𝑑𝑥
+
𝐷
𝑝(𝑥)
𝑒
−1
ℏ 0
𝑥
𝑝(𝑥) 𝑑𝑥
36. The transmission probability is
𝑇 =
𝐹 2
𝐴 2
∵
𝐹
𝐴
~ 𝑒−
1
ℎ 0
𝑎
𝑝 𝑥 𝑑𝑥
∴ 𝑇 ≅ 𝑒−
2
ℎ 0
𝑎
𝑝 𝑥 𝑑𝑥
where T is transmission probability.
38. Eigen value equation for Bound State
Here, a and b are the classical turning points.
𝜓𝐼 =
𝐴
𝜅
𝑒− 𝑥
𝑏
𝜅 𝑥 𝑑𝑥
;
𝜅2
= −𝑘2
=
2𝑚
ℏ2 𝑉 𝑥 − 𝐸
Using connection formula at 𝑥 = 𝑎
𝜓𝐼𝐼 =
2𝐴
𝑘
𝑠𝑖𝑛
𝑎
𝑥
𝑘(𝑥)𝑑𝑥 +
𝜋
4
41. 𝜃 = 𝑛 +
1
2
𝜋 ; 𝑛 = 0,1,2,3 …
𝑎
𝑏
𝑘𝑑𝑥 = 𝑛 +
1
2
𝜋
This is the Eigen value equation for
bound state using WKB approximation.
Now, using this equation the energy
Eigen values for Linear Harmonic Oscillator (LHO)
can be calculated as shown in next slide :
42. LHO energy Eigen values
For LHO potential is
𝑉 𝑥 =
1
2
𝑚𝜔2
𝑥2
∴ 𝑘2
𝑥 =
2𝑚
ℏ2 𝐸 −
1
2
𝑚𝜔2
𝑥2
=
2𝑚
ℏ2
1
2
𝑚𝜔2
𝑥0
2
− 𝑥2
where, 𝑥0
2
=
2𝐸
𝑚𝜔2 is the classical
amplitude or turning points.
43. 𝑜𝑟, 𝑘2
𝑥 =
𝑚2
𝜔2
ℏ2
𝑥0
2
− 𝑥2
Using this value of 𝑘(𝑥) in Eigen value equation
for bound state,
−𝑥0
+𝑥0 𝑚𝜔
ℏ
𝑥0
2 − 𝑥2 𝑑𝑥 = (𝑛 +
1
2
)𝜋
On solving the above equation,
𝐸
ℏ𝜔
= (𝑛 +
1
2
)
or, 𝐸 𝑛 = (𝑛 +
1
2
)ℏ𝜔 ; 𝑛 = 0,1,2, …
- which are the energy Eigen values
for a Linear Harmonic Oscillator.