You are on page 1of 34

Contents

1 Hilbert Space, the Uncertainty Relation and Basis Changes 3


1.1 Hilbert Space and Dual Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Uncertainty Relation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 Unitary Operators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.4 Basis Change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.5 A Central Theorem of Quantum Mechanics . . . . . . . . . . . . . . . . . . . . . . 4

2 Operators with Continuous Spectrum and Momentum in relations to spatial


Translation 5
2.1 Operators with Continuous Spectrum . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2 Translation and Momentum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.3 Wave Functions in Momentum- and Position-Space . . . . . . . . . . . . . . . . . . 6
2.4 Fourier Transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

3 Unitary Time Evolution and Different Pictures in Quantum Mechanics 7


3.1 Time-Evolution Operator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.2 The Schrödinger and the Heisenberg Picture . . . . . . . . . . . . . . . . . . . . . . 8

4 Rotation Groups and Angular Momentum 9


4.1 Rotation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4.2 Spectrum of Angular Momentum states . . . . . . . . . . . . . . . . . . . . . . . . 10

5 Coupling of Angular Momenta and Tensor Operators 12


5.1 Coupling of Angular Momenta . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
5.1.1 Examples of Coupling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
5.1.2 Formal Theory of Coupling . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
5.2 Clebsch-Gordon coefficients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
5.3 Tensor Operators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

6 Time Dependent Perturbation Theory and Fermi’s Golden Rule 15


6.1 Statement of the Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
6.2 The Interaction Picture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
6.3 Differential Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
6.4 Time Dependent Perturbation Theory . . . . . . . . . . . . . . . . . . . . . . . . . 16
6.5 Fermi’s Golden Rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
6.6 Constant Pertubation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

7 Adiabatic Approximation and Geometric Phase 17


7.1 Adiabatic Approximation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
7.2 Berry’s Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
7.3 Examples of Spin 21 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

8 Discrete Symmetries in Quantum Mechanics 19


8.1 Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
8.2 Discrete Symmetries in General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
8.3 Parity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
8.4 Lattice Translation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
8.5 Time Reversal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
8.5.1 Digression on Symmetry Operation . . . . . . . . . . . . . . . . . . . . . . . 20
8.5.2 Time-Reversal Operator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
8.6 Time Reversal of Spin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
8.7 Kramers Degeneracy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
8.8 Symmetry Breaking of Magnetic Field . . . . . . . . . . . . . . . . . . . . . . . . . 21

1
9 Scattering Theory and Born Approximation 22
9.1 Scattering as Time-Dependent Pertubation . . . . . . . . . . . . . . . . . . . . . . 22
9.2 The Born Approximation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
9.3 Box Potential . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

10 Low Energy Scattering Theory and Scattering Length 25


10.1 Partial Waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
10.2 Hard Sphere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
10.3 Low Energy Scattering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

11 Identical Particels and Second Quantization 28


11.1 Identical Particles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
11.2 Symmetrization Postulate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
11.3 Multiparticle States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
11.4 Second Quantization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
11.5 Interaction among Particles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

12 Quantization of Electromagnetic Waves 31


12.1 Electrodynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
12.2 Finding the Hamilton . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
12.3 Zero Point Energy and Casimir Effect . . . . . . . . . . . . . . . . . . . . . . . . . 31
12.4 Quantum states of the electric field . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
12.5 Coherent States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

13 Dirac Equation 33
13.1 Relativistic Quantum Mechanics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
13.2 The Dirac Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
13.3 Dirac Sea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

2
1 Hilbert Space, the Uncertainty Relation and Basis Changes
Describe the concept of Hilbert space and dual space, operators acting on states, including uncer-
tainty relations, unitary operations and changes of basis.

1.1 Hilbert Space and Dual Space


A vector space H over the complex field C is said to be a Hilbert space, if it is equipped with an
inner product h · | · i. An element |αi in H is called a ket.
Similarly we define the dual space H∗ whose elements hα| are called bras. There is a one-to-one
correspondence between the two spaces, called the Dual Correspondence, that obeys
DC
|αi ←−→ hα| (1.1)
DC
X̂|αi ←−→ hα|X̂ † (1.2)
DC
cα |αi ←−→ c∗α hα| (1.3)

Some operators are Hermitian


A = A† (1.4)
They have real eigenvalues.
Â|a(i) i = a(i) |a(i) i (1.5)

1.2 Uncertainty Relation


Define an operator ∆ =  − hÂi, where (∆Â)2 is called the dispersion of A. We have the
variance or mean square deviation:

h(∆Â)2 i = hÂ2 − 2ÂhÂi + hÂi2 i = hÂ2 i − hÂi2 (1.6)

We can now state the uncertainty relation


1 h i
h(∆Â)2 ih(∆B̂)2 i ≥ |h Â, B̂ i|2 (1.7)
4
If two operators do not commute then their observable cannot be simultaneously diagonalised
(known). Like x̂ and p̂.

1.3 Unitary Operators


Unitary operators are norm preserving operators

hα|U † U |αi = hα|αi (1.8)

They must satisfy that U † U = 1

U= (k)
iha(k) | is an operator that changes basis from |a(i) i to |b(i) i by
P
k |b

|b(i) i = Û |a(i) i (1.9)

U is also unitary as can be seen


! !
X X
Û Û =

|a(k) (k)
ihb | |b(l) iha(l) | (1.10)
k l
 
X
=  |a(k) iδkl ha(l) | (1.11)
k,l

=1 (1.12)

It is unitary.

3
1.4 Basis Change
The expansion coefficients in |b(i) i can be found to be
X X
hb(i) |αi = hb(i) |a(j) iha(j) |αi = hai |Û † |a(j) iha(j) |αi (1.13)
j j

That is in matrix notation the column vector in the new basis can be found by applying an
appropriate unitary transformation the the column vector in the old basis.
Similarly for operators we have
XX
hb(k) |X|b(l) i = hb(k) |a(m) iha(m) |X|a(n) iha(n) |b(l) i (1.14)
m n
XX
= ha(k) |Û † |a(m) iha(m) |X|a(n) iha(n) |Û |a(l) i (1.15)
m n

1.5 A Central Theorem of Quantum Mechanics


Consider
P (i) an observable  and two basis |a(i) i and |b(i) i and the unitary transformation Û =
k |b iha | between the sets
(i)

Then  and Û ÂÛ −1 have the same eigenvalues. The eigenvalue equation in |a(i) i basis is given
by
Â|a(i) i = ai |a(i) i (1.16)
Multiply by Û on the left and insert identity I.
h ih i
Û ÂÛ −1 Û |a(i) i = Û ÂÛ −1 Û |a(i) i = ai Û |a(i) i = ai |b(i) i (1.17)

They have identical spectra.

4
2 Operators with Continuous Spectrum and Momentum in
relations to spatial Translation
Describe the concept of an operator with continuous spectrum, including momentum operator and
its relation to translations.

2.1 Operators with Continuous Spectrum


Say we have an observable with a continuous spectrum and corresponding operator x̂. Examples
could be x̂ or p̂. We write the eigenvalue equation as:

x̂|x(i) i = x(i) |x(i) i, where x(i) lies in a continuous range (2.1)

Now we have an infinite amount of different eigenvalues and eigenkets.


We let Z
X

− and δij →− δ(a(i) − a(j) ) (2.2)

The following identities appears:

hx(i) |x(j) i = δ(x(i) − x(j) ) (Orthonormality) (2.3)


Z
dx|xihx| = 1 (Completness) (2.4)
Z
2
dx khx|αik = 1 (Normalization) (2.5)
Z
|αi = dx|xihx|αi (Use of completness on ket) (2.6)
Z
hβ|αi = dxhβ|xihx|αi (Use of completness in braket) (2.7)

hx(i) |x̂|x(j) i = x(j) δ(x(i) − x(j) ) (Use of orthonomality in sandwich) (2.8)

The probability of finding a particle in a given interval [a, b] is given by


Z b
P (α ∈ [a, b]) = dx0 |hx0 |αi|2 (2.9)
a

Notice the probability of finding the particle at a particular point is zero. Unless the wave function
is a dirac-delta function.

2.2 Translation and Momentum


Consider an operator that translates by an infinitesimal displacement denoted by J (dx) such that

J (dx)|xi = |x + dxi (2.10)

We demand of J (dx) that the transformed ket be normalized. This condition is guaranteed by
demanding the operation to be unitary.

J † (dx)J (dx) = 1 (2.11)


J (dx )J (dx ) = J (dx + dx )
00 0 00 0
(2.12)
J (−dx) = J −1
(dx) (2.13)
lim J (dx) = 1 (2.14)
dx→
−0
This can be satisfied by
ip · dx
J (dx) = 1 − (2.15)
~

5
Compounding N infinitesimal translation, each of which is characterized by a spatial displacement
∆x/N in the x-direction, we obtain
N
ipx ∆x0

J (∆xx̂) = lim 1 − (2.16)
N→
−∞ N~
ipx ∆x0
 
= exp − (2.17)
~
 
d
= exp −∆x (alternativ) (2.18)
dx

This can also be found by considering the effect on an analytical function on the alternativ above.

2.3 Wave Functions in Momentum- and Position-Space


The wave function of an abitrary space can be written in either the momentum space or position
space as such:

ψα (x0 ) = hx0 |αi (2.19)


φα (p0 ) = hp0 |αi (2.20)

Consider the effect of translation on an arbitrary ket

ip̂∆x0
 Z
J (dx)|αi = 1 − dx|xihx|αi (2.21)
~
Z
= dx|x + dxihx|αi (2.22)
Z
= dx|xihx − dx|αi (2.23)
Z  

= dx|xi hx |αi − ∆x 0 hx |αi
0 0
(2.24)
∂x

Comparing we get that


∂ ∂
hx0 |p̂|αi = −i~ hx0 |αi = −i~ 0 ψα (x0 ) (2.25)
∂x0 ∂x

2.4 Fourier Transformation


Let find the transform between ψα (x0 ) and φα (p0 )

ψα (x0 ) = hx0 |αi (2.26)


Z
= dp0 hx0 |p0 ihp0 |αi (2.27)
1
Z
p0 x0
=√ dp0 ei ~ φα (p0 ) (2.28)
2π~
The overlap hx0 |p0 i is a plane wave and the transformation is the Fourier Transformation. Where
we have solved the differential equation of hx|p̂|pi.

6
3 Unitary Time Evolution and Different Pictures in Quan-
tum Mechanics
Outline the concept of unitary time evolution. Describe the different pictures of quantum mechanics.

3.1 Time-Evolution Operator


Our basic task is to study time evolution of a state ket such that:

|α, ti = U(t, t0 )|α, t0 i, (3.1)

where U(t, t0 ) is the time-evolution operator.

Requirement 1 : We demand that the transformed ket be normalized. This requires the time-
evolution operator to be unitary.

U † (t, t0 )U(t, t0 ) = 1 and U(t, t0 )U † (t, t0 ) = 1 or U † (t, t0 ) = U −1 (t, t0 ) (3.2)

Requirement 2 : Another feature we require of the U operator is the composition property:

U(t2 , t0 ) = U(t2 , t1 )U(t1 , t0 ) (3.3)

Requirement 3 : Consider an infinitesimal time-evolution we require that

lim U(t0 + dt, t0 ) = 1 (3.4)


dt→
−0
All these requirements are satisfied by. The Hamiltonian is the generator of time evolution so to
first order, we may take the ansatz
iHdt
U(t0 + dt, t0 ) = 1 − , (3.5)
~
where H is the Hamiltonian. That is the Hamiltonian is the generator of time-evolution.
Using the composition property we get the Schrödinger Equation for the time evolution
operator.
 
iHdt
U(t + dt, t0 ) = U(t + dt, t)U(t, t0 ) = 1 − U(t, t0 ) (3.6)
~
iHdt
U(t + dt, t0 ) − U(t, t0 ) = − U(t, t0 ) (3.7)
~

i~ U(t, t0 ) = HU(t, t0 ) (3.8)
∂t
This can be solved under three cases.
Case 1. The Hamiltonian operator is indepedent of time. The solution is

−iH(t − t0 )
 
U(t, t0 ) = exp (3.9)
~

Case 2. The Hamiltonian operator H is time dependent but the H’s at different times commute

−i t 0
 Z 
U(t, t0 ) = exp dt H(t )
0
(3.10)
~ t0
Case 3. The general case is solved using the Dyson series
∞  n Z t Z t1 Z tn−1
X i
U(t, t0 ) = 1 + dt1 dt2 · · · dtn H(t1 )H(t2 ) · · · H(tn ) (3.11)
n=1
~ t0 t0 t0

7
3.2 The Schrödinger and the Heisenberg Picture
In the Schrödinger picture the time-evolution operator affects state kets, whereas in the Heisen-
berg picture the observables are affected. One can make the distinction

(hβ|U † ) · X · (U |αi) = hβ|U † XU |αi = hβ|(U † XU )|αi (3.12)

In classical mechanics we not talk about particle changing state rather there observables like x and
L changing. The connection is therefore better seen in the Heisenberg picture.
We define observables in the Heisenberg picture as being dependent on time

A(H) (t) ≡ U † (t)A(S) U(t). (3.13)

The state kets are frozen in time in the Heisenberg picture in contrast to the Schrödinger picture
where they do evolve

|α, tiH = |α, t0 i (3.14)


|α, tiS = U(t, t0 )|α, t0 i (3.15)

The base kets do not evolve in the Schrödinger picture but they do in the Heisenber picture.

|a0 , tiH = U † |a0 , t0 i (3.16)


|a0 , tiS = |a0 , t0 i (3.17)

This can be seen from the fact that in the Schrödinger Picture eigenvalue equation is

A(S) |a0 iS = a0 |a0 iS (3.18)

and in the Heisenberg picture it can be written as

A(H) |a0 , tiH = U † (t)A(S) U(t)U † |a0 , t0 i = a0 U † |a0 , t0 i = a0 |a0 , tiH . (3.19)

So U † |a0 , t0 i is an eigenket and so a baseket.


But how is the state kets stationary when the base kets are not as they one can be expanded
by the other as X
|αi = |a0 iha0 |αi (3.20)
a0

Because the coefficient also change.


The expectation values are obviously the same in both pictures

S hα, t|A
(S)
|α, tiS = hα, t0 |U † A(S) U|α, t0 i =H hα, t|A(H) |α, tiH (3.21)

By evaluating the derivative of an observable that does not depend explicitly on time in the Heisen-
berg Picture we find Heisenberg equation of motion and using the fact that U † HU = H because
they commute.

dA(H) d  † 
= U (t)A(S) U(t) (definition of operator in Heisenberg) (3.22)
dt dt
   
d † d
= U (t) A(S) U(t) − U † (t)A(S) U(t) (product rule) (3.23)
dt dt
i 
= HA(S) − A(S) H (differentiation of U) (3.24)
~
1 h i
= A(H) , H (definition of commutator) (3.25)
i~
Base kets in the Heisenberg picture evolves as

|a0 , tiH = U † |a0 i (3.26)

8
4 Rotation Groups and Angular Momentum
Outline how the rotation group works as an operator on quantum mechanical states, and discuss
the angular momentum algebra including the spectrum of angular momentum states.

4.1 Rotation
Given a rotationoperation R, characterized by a 3 × 3 orthogonal matrix classically. We associate
an operator in QM as D(R) such that

|αiR = D(R)|αi (4.1)

The dimension of D(R) depends on the size of the particular ket space. We further postulate that
D(R) has the same group properties as R:

Identity: D(R) · 1 = D(R) (4.2)


Closure: D(R1 ) · D(R2 ) = D(R3 ) (4.3)
Inverses: D(R) · D(R) −1
=1 (4.4)
Associativity: [D(R1 ) · D(R2 )] · D(R3 ) = D(R1 ) · [D(R2 ) · D(R3 )] (4.5)

The generator of rotation is angular momentum. We therefore define the infinitesimal-rotation


operator to be  
J · n̂
D(n̂, dφ) = 1 − i dφ, (4.6)
~
which is clearly unitary when higher order terms is ignored and we also have that limdφ→ −0D =1
A finite rotation can be obtatined by compunding successively infinitesimal rotations about the
same axis.
    N
Jz φ
Dz (φ) = lim 1 − i (4.7)
N→
−∞ ~ N
 
−iJz φ
= exp (4.8)
~

The matrix representation is given by


 
(j) −iJ · nφ
Dm0 m (R) = hj, m0 | exp |j, mi (4.9)
~

Notice that the same j appear since a rotation does not change the eigenvalue of J2 . When more
j’s are considerede one can write the matrix on a block diagonal form. Where each block has a
length of 2j + 1.

Figure 1: Each block cannot be reduced further

9
4.2 Spectrum of Angular Momentum states
Fundamental Commutation Relations of Angular Momentum

[Ji , Jj ] = i~ijk Jk (4.10)

We define a new operator Ĵ2 ≡ Jˆx Jˆx + Jˆy Jˆy + Jˆz Jˆz . One can easily show that

[Ĵ2 , Jk ] = 0, (k = 1, 2, 3). (4.11)

By firstly checking that [Ĵ2 , Jz ] = 0 and then using a cyclic argument. Because Jx , Jy and Jz do
not commute with each other, we can only choose one to be simultaneously eigenvalues with Ĵ2 .
By convention we choose Jz . Such that.

Ĵ2 |a, bi = ~2 a|a, bi (4.12)


Jˆz |a, bi = ~b|a, bi (4.13)

To determine the allowed values for a and b, it is conveniant to work with non-Hermitian ladder
operators
Jˆ± ≡ Jˆx ± iJˆy (4.14)
They satisfy the following commutation relation:

Jˆz , Jˆ± = ±~Jˆ±


h i
(4.15)

The physical meaning of can be seen by

Jˆz (Jˆ± |a, bi) = ~(b ± 1)(Jˆ± |a, bi) (4.16)

Even though Jˆ± changes the eigenvalue of Jˆz by one unit of ~, it does not change the eigenvalue
of Ĵ2 :

Ĵ2 (Jˆ± |a, bi) = Jˆ± Ĵ2 |a, bi using Ĵ2 , Jˆ± = 0
 h i 
(4.17)
= ~2 a(Jˆ± |a, bi) (4.18)

It turns out that the exist an upper limit to b for a given a:


Let us find the expectation value
1 † ˆ
ha, b|Ĵ2 − Jˆz2 |a, bi = ha, b| (Jˆ− J− + Jˆ+
† ˆ
J+ )|a, bi (4.19)
2
1 
2 2

= kJ+ |a, bik + kJ− |a, bik (4.20)
2
= ~2 (a − b2 ) ≥ 0 (4.21)

Meaning:
a ≥ b2 (4.22)
If we define the maximum value of b to be j ≡ bmax . Then we must have that

Jˆ+ |a, ji = 0 and hence that Jˆ− Jˆ+ |a, ji = 0 (4.23)

But we also have that

(Jˆ− Jˆ+ )|a, ji = (Ĵ2 − Jˆz2 − ~Jˆz )|a, ji = ~2 a − ~2 j 2 − ~2 j = 0 (4.24)

Given us
a = j(j + 1) (4.25)
Similarly there is a lowest value of b denoted by bmin

bmin = −bmax (4.26)

10
We now define m ≡ b and we find the spectrum of m, to be

m = −j, j + 1, · · · , j − 1, j (2j + 1 states) (4.27)

and the basic eigenvalue equations now becomes by using j and m instead of a and b

Ĵ2 |j, mi = j(j + 1)~2 |j, mi (4.28)


Jˆz |j, mi = m~|j, mi (4.29)

11
5 Coupling of Angular Momenta and Tensor Operators
Describe how the coupling of angular momenta works in quantum mechanics, and discuss the
concept of tensor operators.

5.1 Coupling of Angular Momenta


5.1.1 Examples of Coupling
A realistic description of a particle with both spin and space degrees of freedom is necessary. This
gives rise to addition of angular momentum. The generator of rotation now becomes

Ĵ = L̂ + Ŝ (5.1)

Instead of using |x0 i as the base kets for the space part, we may use |n, l, mi, that are eigenkets of
L̂2 and L̂z with eigenvalues ~2 l(l + 1) and ml ~ respectively.

5.1.2 Formal Theory of Coupling


A formal theory is needed. Consider two angular momentum operators Ĵ1 and Ĵ2 , that each
satisfies the usual angular momentum commutation relations:

Jˆ1i , Jˆ1j = i~ijk Jˆ1k and Jˆ2i , Jˆ2j = i~ijk Jˆ2k


h i h i
(5.2)

However, we have
Jˆ1m , Jˆ2l = 0,
h i
(for all m and l) (5.3)

between any pair of operators from different subspaces. We define the total angular momentum
by
Ĵ = Ĵ1 ⊗ 1 + 1 ⊗ Ĵ2 (5.4)
Note that Ĵ also satifies the usual angular momentum commutation relation

Jˆi , Jˆj = i~ijk Jˆk ,


h i
(5.5)

as a consequence of Equation 5.2. So Ĵ is an angular momentum in the same sense.


We have two options for the choice of base kets that are simultaneously diagonalizable for a
total of four operators that all commute. One we will call the product basis and the other we
will call the coupled basis.

Product Basis: The base kets are simultaneously eigenkets of Ĵ21 , Ĵ22 , Jˆ1z and Jˆ2z with the base
ket |j1 , j2 ; m1 m2 i.

Coupled Basis: The base kets are simultaneously eigenkets of Ĵ2 , Ĵ21 , Ĵ22 and Jˆz with the base
ket |j1 , j2 ; jmi.

with the appropriate eigenvalues.

5.2 Clebsch-Gordon coefficients


Let us change basis from the product basis to the coupled basis by using the identity twice
X
|j1 , j2 ; jmi = |j1 j2 ; m1 m2 ihj1 j2 ; m1 m2 |j1 j2 , jmi, (5.6)
m1 ,m2

where we have used the completeness relation of the product base kets. The elements of this trans-
formation matrix are called Clebsch-Gordan coefficients. There are many important properties
of the Clebsch-Gordan coefficients:

12
Property 1 : The coefficients vanish unless

m = m1 + m2 . (5.7)

This can be proven by noticing that

hj1 j2 ; m1 m2 |Jz − J1z − J2z |j1 j2 ; jmi = 0, (5.8)

because Jz − J1z − J2z = 0. This is also equal to

(m − m1 − m2 )hj1 j2 ; m1 m2 |j1 j2 ; jmi = 0 (5.9)

which proves our assertion.

Property 2 :
|j1 − j2 | ≤ j ≤ j1 + j2 (5.10)
This may appear obvious from the vector model of angular momentum addition. Let check that
the size of the bases are the same.
Product Basis: There are 2j1 + 1 states for a given j1 and similarly for j2 , therefore we get

N = (2j1 + 1)(2j2 + 1). (5.11)

Coupled Basis: There are 2j + 1 states for a given j and j itself runs according to Equations 5.10.
Without the loss generality we assume that j1 ≥ j2 . We therefore obtain:
jX
1 +j2

N= (2j + 1) (5.12)
j=j1 −j2
1
= [2((j1 − j2 ) + 1) + 2((j1 + j2 ) + 1)](2j2 + 1) (Gauss’ Sum) (5.13)
2
= (2j1 + 1)(2j2 + 1) (5.14)

So they indeed do match.

Property 3 : The Clebsch Gordon Coefficients form a unitary matrix (they preserves the norm),
furthermore, the matrix elements are taken to be real by convention. As as consequence the matrix
is orthogonal and the inverse coefficients are the same as the transpose hj1 j2 ; jm|j1 j2 ; m1 m2 i =
hj1 j2 ; m1 m2 |j1 j2 ; jmi.

These coefficients can be found be ladder operators and by using orthogonality, or recur-
sion relation. A general formula does also exist and tables can be found.

5.3 Tensor Operators


Spherical Tensor are Irreducible Tensor in contrast to a Cartesian Tensor.

Tqk = Ykq (V) (5.15)

They have follow the following algebra


h i
Jz , Tq(k) = ~qTq(k) (5.16)

and h i
(k)
J± , Tq(k) = ~ (k ± q)(k ± q + 1)Tq±1 (5.17)
p

The Wigner-Eckart Theorem.

hj 0 ||T (k) ||ji


hj 0 , m0 |Tq(k) |jmi = hjk; mq|jk; j 0 m0 i √ (5.18)
2j + 1

13
where the double-bar matrix element is independent of m and m0 , and q.

The dipole operator er is a tensor of rank 1. Gives us the selection rules

∆m = 0, ±1 (5.19)
∆j = 0, ±1 (not from 0 to 0 m) (5.20)

14
6 Time Dependent Perturbation Theory and Fermi’s Golden
Rule
Outline how time-dependent perturbation theory works and discuss Fermi’s golden rule.

6.1 Statement of the Problem


Consider a Hamiltonian H that can be split into two parts
H = H0 + V (t), (6.1)
where H0 is not time dependent. Assume that the energy eigenkets and energy eigenvalues for
H0 |ni = En |ni (6.2)
is completely known.
We wish to find cn (t) at a future time such that
X
|α, tiS = cn (t)e−iEn t/~ |ni. (6.3)
n

The exponential factor is simply the time evolution of the base kets. This factor is present also
without V (t). So this is a way of separating the two.

6.2 The Interaction Picture


We define
|α, tiI = eiH0 t/~ |α, tiS (6.4)
This is the state ket in the interaction picture. If use equation 6.3 and 6.4 we get the expansion of
the interaction ket
X
|α, tiI = cn (t)|ni. (6.5)
n

Observables in the interaction picture is defined by


AI ≡ eiH0 t/~ AS e−iH0 t/~ (6.6)
This is comparison to the Heisenberg Picture
|α, tiH = eiHt/~ |α, tiS (6.7)
AH ≡ e iHt/~
AS e −iHt/~
(6.8)
We now derive the fundmaental differential equation that characterizes the time evolution of a
state ket in the interaction picture. We start by taking the time derivative of 6.4.

i~ |α, tiI = VI |α, tiI , (6.9)
∂t
which is a Schrödinger-like equation.

6.3 Differential Equation


To find the differential equation of cn (t) we multiply both sides of the Schrödinger like equation
6.9 by hn|

i~ hn|α, tiI = hn|VI |α, tiI (6.10)
∂t X
= hn|eiH0 t/~ V (t)e−iH0 t/~ |mihm|α, tiI (6.11)
m
X
= Vnm (t)e(En −Em )t/~ hm|α, tiI . (6.12)
m

15
This in turn becomes
∂ X
i~ cn (t) = Vnm (t)eiωnm t cm (t), (6.13)
∂t m

where we have used that cn (t) = hn|α, tiI and that ωnm = En −Em
~ .

6.4 Time Dependent Perturbation Theory


Solution to this equation is usually not possible. Instead we expand the coefficients

cn (t) = c(0)
n + cn + cn + · · ·
(1) (2)
(6.14)

Say we are in some initial state |ii that coinciding in all pictures. Then cn (t) = hn|UI (t, t0 )|ii UI
can be found from the Schrödinger equation of the time-evolution operator in the interaction
picture. I can then be expanded using the Dyson Series and compared to the order of cn

n = δni
c(0) (6.15)
Z t
i 0
n =−
c(1) eiωni t Vni (t0 )dt0 (6.16)
~ t0

To zeroth order the different states do not couple.

To first order they do couple between each other.

To second order they can couple to each other and then again to others.

6.5 Fermi’s Golden Rule


The transition rate from an initial state i to some final state n is given by Fermi’s Golden Rule

−f =
ωi→ |Vf i |2 ρ(Ef ) (6.17)
~
provided the first-order time-dependent pertubation theory is valid that is En ' Ei .

6.6 Constant Pertubation


V (t) = V for t ≥ 0 (6.18)
Then we get

n = δni
c(0) (6.19)
Z t
i 0
n = − Vni
c(1) eiωni t dt0 (6.20)
~ 0
Vni 0
= (1 − eiωni t ) (6.21)
ωni ~

|Vni |2
 
ωni t
|c(1) |2
∝ sin 2
(6.22)
n 2
ωni 2

16
7 Adiabatic Approximation and Geometric Phase
Discuss the adiabatic approximation in quantum mechanics and the concept of geometric phase.

7.1 Adiabatic Approximation


We begin with the eigenvalue equation

H(t)|n, ti = En (t)|n, ti (7.1)

Looking for general solution for the Schrödinger Equation



i~ |α, ti = H(t)|α, ti, (7.2)
∂t
then we can write X
|α, ti = cn (t)eiθn |n, ti, (7.3)
n
where
1 t
Z
θn (t) = − En (t0 )dt0 (Dynamical Phase Factor) (7.4)
~ 0
Insert into the Schrödinger equation and project onto hm, t|.
  X
∂ hm; t|Ḣ|n; ti
ċm (t) = −cm (t)hm; t| |m; ti − cn (t)ei(θn (t)−θm (t)) (7.5)
∂t En − Em
n6=m

The adiabatic approximation amounts to neglecting the second term. When we do so the differ-
ential equation can be solved as:
cn (t) = eiγ(t) cn (0) (7.6)
where Z t  

γ(t) = i hn; t| |n; t i dt0 . (Geometric Phase Factor)
0
(7.7)
0 ∂t0
A genereal state ket can then be written as
X
|α, ti = cn (0)eiγn (t) eiθn (t) |n, ti. (7.8)
n

For H 6= H(t) we get that γn (t) = −θn (t) so they cancel and we get our stationary states.

7.2 Berry’s Phase


Assume the Hamiltonian can be represented be a vector parameter R(t). Then |n; ti = |n(R(t))i.
Applying the chain-rule to express the Geometric Phase in a general parameter space yields.
Inserting this into the geometric phase yields

dR
Z T
γn (T ) = i hn; t|[∇R |n; ti] · dt (7.9)
0 dt
Z R(T )
=i hn; t|[∇R |n; ti] · dR (7.10)
R(0)

When T represents a full cycle, this becomes a loop integral


I
γn (C) = i hn; t|[∇R |n; ti] · dR (7.11)
C

We define

An (R) ≡ ihn; t|[∇R |n; ti] (Berry Connection) (7.12)


Bn (R) ≡ ∇R × An (R) (Berry Curvature) (7.13)

17
By using stokes theorem we get
I Z Z
γn (C) = An (R) · dR = [∇R × An (R)] · da = Bn (R) · da (7.14)
C

This vanishes when the parameter space is one dimensional.

Notice the gauge transformations

An (R) −→ An (R) − ∇R δ(R), (7.15)

leaves the Berry Phase unchanged because ∇ × (∇f ) = 0 for a general field f .

The final result becomes of Berrys works is


Z
γn (C) = Bn (R) · da (7.16)
X hn; t|∇R H|m; ti × hm; t|∇R H|n; ti
Bn (R) = i (7.17)
(En − Em )2
m6=n

1
7.3 Examples of Spin 2
Time varying magnetic field B(t), so we parameterize using the magnetic field

H(t) = H(R(t)) = − S · R(t) (7.18)
~

∇R H = − S (7.19)
~
1
γ± (C) = ∓ Ω (7.20)
2
This has been measured.

18
8 Discrete Symmetries in Quantum Mechanics
Describe the concept and formalism of discrete symmetries in quantum mechanics, in particular,
time reversal symmetry.

8.1 Definition
A symmetry is an operation that leaves the Hamiltonian invariant.

8.2 Discrete Symmetries in General


Discrete symmetries cannot be obtained by applying successively infinitesimal symmetry opera-
tions.

The three most common discrete symmetries are parity, lattice translation. and time re-
versal.

8.3 Parity
Consider the parity operator π̂ applied to some state

|αi →
− π̂|αi, (8.1)

where we require of the expectation value of x̂ that

hα|π̂ † x̂π̂|αi = −hα|x̂|αi (8.2)

This can be accomplished if


π̂ † x̂π̂ = −x̂ (8.3)
or
{x̂, π̂} = 0, (8.4)
in other words x̂ and π̂ must anticommute.

How does an eigenket of the position operator transform under parity? We claim that

π̂|x0 i = |−x0 i (8.5)

So π̂|x0 i is an eigenket of x̂ with eigenvalue −x0 , so must be same as a position eigenket |−x0 i up
to some phase factor.
π̂ 2 |x0 i = |x0 i (8.6)
that is π̂ 2 = 1. From this we can conclude that the eigenvalues must be 1 or −1.

8.4 Lattice Translation


Consider a periodic potential in one dimension, where V (x±a) = V (x). In generel, the Hamiltonian
is not invariant under a translation represented by the lattice-translation operator τ (l), where
l is some distance.
When l coincides with the lattice spacing a, we do have that

τ̂ † (a)V (x)τ̂ (a) = V (x + a) = V (x) (8.7)

and hence that


τ̂ † (a)Ĥ τ̂ (a) = H (8.8)
because the kinetic energy part is invariant. In conclusion we have that
h i
Ĥ, τ̂ (a) = 0, (8.9)

19
that is they can be simultaneously diagonalized in some basis. This basis does not consist of states
localized in some n’th lattice since these states would not be an eigenstate of the lattice-translation
operator but might be of the Hamiltonian.

τ̂ (a)|ni = |n + 1i and Ĥ|ni = E0 |ni. (8.10)

This is handled by forming a linear combination and then finding the overlap withhx0 |τ (a) We then
get Bloch’ Theorem 0
hx0 |θi = eikx uk (x0 ) (8.11)
That is a plane wave and a periodic function.

8.5 Time Reversal


Better described as reversal of motion
Consider the Schödinger wave equation
∂ψ
i~ = Hψ. (8.12)
∂t
Suppose ψ(x, t) is solution. We can easily verify that ψ(x, −t) is not a solution because we get an
extra sign from the first order time derivative. ψ(x, −t)∗ is a solution though.
∗
∂ψ ∗

∂ψ
i~ = (Hψ)∗ ⇔ − i~ = Hψ ∗ , (8.13)
∂t ∂t
where we have used that H is hermitian.

8.5.1 Digression on Symmetry Operation


In general we require that for symmetry operator that the inner product of two states be preserved
after a transformation of both states such that

hα|βi = hα̃|β̃i (8.14)

This arises from the fact that the corresponding symmetry operators are unitary (U † U = 1). For
the time reversal operator this requirement is too restrictive. Where we instead require it to be
antiunitary. That is
hα|βi = hα̃|β̃i∗ (8.15)
and
= θ(c1 |αi + c2 |2i) = c∗1 θ|αi + c∗2 θ|2i (8.16)
An antiunitary operator can be written as

θ̂ = Û K̂, (8.17)

where Û is an unitary operator and K̂ is the complex-conjugate operator

8.5.2 Time-Reversal Operator


U (t0 + δt, t0 )Θ|αi = ΘU (t0 − δt , t0 )|αi (8.18)
or written out    
iH iH
1− δt Θ|αi = Θ 1 − (−δt) |αi (8.19)
~ ~
Time reversing and then propagating forwards is the same as propagating backwards and then
time reversing.
− iHΘ| i = ΘiH| i (8.20)
Suppose Θ is unitary then the i cancels

− iHΘ = ΘiH = iΘH ⇒ {Θ, H} = 0 (8.21)

20
Meaning that the time reversed state would be an eigenket of the Hamilton with the eigenvalue
−En , which is nonsensical.

Suppose Θ is anti-unitary then the i get a sign and cancels.

− iHΘ = ΘiH = −iΘH ⇒ [Θ, H] = 0 (8.22)

Meaning that the time reversed state would be an eigenket of the Hamilton with the eigenvalue En .

Hermitian operators are either even or odd under time reversal.

ΘxΘ−1 = x (8.23)

ΘpΘ−1 = −p (8.24)
ΘJΘ−1 = −J (8.25)
¨

8.6 Time Reversal of Spin


For a spin-j systems we have that

Θ2 |j, mi = (−1)2j |j, mi = (−1)N |j, mi (8.26)

8.7 Kramers Degeneracy


Now we know if [H, Θ] = 0, so |ni and Θ|ni must have the same energy eigenvalues. But are they
the same state?
Θ|ni = eiδ |ni (8.27)
Doing it again we should get back.
Θ2 |ni = +|ni (8.28)
If we have N electrons then the total j is either integer or half integer. There is a degeneracy when
N is odd.

8.8 Symmetry Breaking of Magnetic Field


Magnetic field breaks time reversal symmetry that is [Θ, H] 6= 0 for example H ∝ S · B. If we
consider ΘB = −B then it does not break the symmetry.

21
9 Scattering Theory and Born Approximation
Describe the basic idea of scattering theory and the treatment of scattering problems within the
Born approximation.

9.1 Scattering as Time-Dependent Pertubation


Scattering theory is necessary to measure a quantum mechanical system

Assuming the distance the measurement point is much greater than the size of the potential
|x|  |x0 |, and the potential is assumed to be local hx0 |V |x00 i = V (x0 )δ (3) (x0 − x00 ), we get

1 eikr
 
hx|ψ (+) i = 3/2 eik·x + f (k0 , k) (9.1)
L r

We refer to f (k, k0 ) to scattering amplitude. The differential cross-section is given by


dσ kσtot
= |f (k0 , k)|2 Imf (θ = 0) = (9.2)
dΩ 4π
where the final part is the optical theorem.

Our Hamilton of some localized potential is generally given by

p2
H = H0 + V (r), H0 = (9.3)
2m
~2 k 2
We have solved the case for H0 , where the eigenket are planewaves and the eigenvalues are 2m .

The electron sees the potential as time dependent.

Moving on we find the Lippmann-Schwinger equation by


• Using interaction picture

i~ UI (t, t0 ) = VI (t)UI (t, t0 ) (9.4)
∂t
• Transition amplitude T = hf |UI (t, t0 )|ii
• Box Quantization 1
L3/2
eik·x L→
− ∞
• Limit for t and t0
We get the Lippmann-Schwinger equation
1
|ψ (±) i = |ii + V |ψ (±) i (9.5)
Ei − H0 + i~

22
1
T =V +V T = V + V GT (9.6)
Ei − H0 + i~
Notice it is recursive.

Where the scattering amplitude is given by

mL3 0
f (k0 , k = − hk |V |ψ + i (9.7)
2π~

9.2 The Born Approximation


Starting from equation 9.6 we see that it can be solved iteratively. To first order we simply get

|ψ (±) i = |ii, (9.8)

T (1) = V, (9.9)
which is known as the first Born approximation. The second Born approximation is achieved by
inserting T (1) into the right hand side of equation 9.6, yielding

T (2) = V + V GV. (9.10)

The physical interpretation of the first Born approximation is that a particle comes in, scatters,
and then leaves the potential. In the second Born approximation the particle comes in, scatters,
then it propagates as described by the Greens function, scatters again, and then leaves. This can
be seen on figure 9.2. To the third order, the particle would propagate twice before leaving and so
on.

In the first Born approximation, the scattering amplitude is given by the fourier transform to
q = k − k0 space Z
m 0 0
f (k , k) = −
(1) 0
d3 x0 ei(k−k )·x V (x0 ), (9.11)
2π~2
which can then be used to find the differential cross section for a given potential such as the
spherically symmetric infinite well, the Yukawa potential, etc. Note that f (1) is always real so the
optical theorem cannot be applied.41
In general, the Born approximation gets better at higher energies.

Wont necessarily converge for lower energies with more terms

23
9.3 Box Potential
Spherically Symmetric Potential
V (r) = V0 r≤a (9.12)

where q = |k − k0 | so the zeros can be used to determine the radius of the potential.

24
10 Low Energy Scattering Theory and Scattering Length
Outline the theoretical description of low-energy scattering and describe the physical meaning of
the scattering length.

10.1 Partial Waves


Scattering theory is necessary to measure a quantum mechanical system

Assuming the distance the measurement point is much greater than the size of the potential
|x|  |x0 |, and the potential is assumed to be local hx0 |V |x00 i = V (x0 )δ (3) (x0 − x00 ), we get

1 eikr
 
hx|ψ i = 3/2 e
(+) ik·x
+ f (k , k)
0
(10.1)
L r

We refer to f (k, k0 ) to scattering amplitude.


We can find the scattering amplitude using Wigner-Eckart theorem tensor of rank 0 and
by expanding in terms of partial waves

mL3 0
f (k0 , k) = − hk |V |ψ (+) i (10.2)
2π~2
We refer to f (k, k0 ) to scattering amplitude. The differential cross-section is given by

dσ kσtot
= |f (k0 , k)|2 Imf (θ = 0) = (10.3)
dΩ 4π
where the final part is the optical theorem.

Change to using partial waves |E, l, mi that are eigenkets of H0 , L2 and Lz instead of k as
planewaves |ki.

Assuming a rotational invariance of the potential and using probability conservation we


can write the scattering amplitude as
1X
f (θ) = (2l + 1)eiδl sin δl Pl (cos(θ)) (10.4)
k
l=0

The only change to the wave function at long range is a phase shift of the outgoing wave. We know
how the waves is at r > a and we can solve it for the case inside and then comparing boundaries.

Al = Bl (jl (kr) − tan(δl )nl (kr)) (10.5)

Neumann diverges in zero so this only defined outside the potential.

25
10.2 Hard Sphere
(
∞ for r > R
V = (10.6)
0 for r < R

sin(k(r − R))
A0 (r) = (10.7)
kr
That is a spherical outgoing wave with a phase shift

δ0 = −kR (10.8)

10.3 Low Energy Scattering


At extremly low energies where (k ' 0). For low energies only l = 0 is relevant because higher l
are because a particle cannot penetrate the centrifugal barrier.

l(l + 1)
Veff = V (r) + (10.9)
r2
For r > R and for l = 0, the outside radial wave function is just a straight line.

d2 u
=0→
− u(r) = constant(r − a) (10.10)
dr2
Low-Energy Scattering off any short-range potential can be described by hard-sphere scattering
with a scattering length a.
tan(δ0 )
a = lim − (10.11)
k→
−0 k

26
a >> R we have bound states in attractive potential

~2
E' (10.12)
2ma2

27
11 Identical Particels and Second Quantization
Describe the concept of identical particles in quantum mechanics, its implication for multi-particle
quantum states, and its relation to the formalism of second quantization.

11.1 Identical Particles


In classical mechanics we can label particles. In quantum mechanics we cannot distinguishes two
identical particles.

The state ket of two identical particles can be written as a product as


c1 |k 0 i|k 00 i + c2 |k 00 i|k 0 i, (11.1)
where c1 and c2 has to be normalised. All kets of this form lead to an identical set of eigenvalues
also know as exchange degeneracy.

We define the permutation operator P12 that


P12 |k 0 i|k 00 i = P12 |k 00 i|k 0 i (11.2)
Clearly
2
P12 =1 (11.3)
The eigenvalues of P12 must either be 1 or -1.

For all observables we must have that [A, P ] = 0 for identical particles otherwise we could measure
a difference and they would not be identical.
If the Hamiltonian commutes with the permutation operator [P12 , H] then any (anti)symmetric
state remains (anti)symmetric.

11.2 Symmetrization Postulate


Nature only consist of totally symmetric states or antisymmetric states

Bosons - (B-E) statistics or Fermions - (Fermi-Dirac)-statistics


Pij | i = ±| i (11.4)
This is an assumption in non-relativistic quantum mechanics

Fermions follow Pauli Exclusion Principle Because a state like |k 0 i|k 0 i which is symmetric
and is not allowed
For two identical particles a dramatically different number of possible states up to some phase

Bosons: 3 state
k0
|k 0 i|k 0 i, |k 00 i|k 00 i, √ (|k 0 i|k 00 i + |k 00 i|k 0 i) (11.5)
2
Fermions: 1 state
1
√ (|k 0 i|k 00 i − |k 00 i|k 0 i) (11.6)
2

28
Classical Particles: 4 states

|k 0 i|k 00 i, |k 00 i|k 0 i, |k 0 i|k 0 i, |k 00 i|k 00 i (11.7)

11.3 Multiparticle States


This can be generalised to N -particle states.

|k (1) i|k (2) i · · · |k (i) i · · · |k (N ) i (11.8)

We now have many permutation operators.

[Pij , Pkl ] 6= 0, (11.9)

which is generally not equal to zero although many times it is.

It is sufficient to know the occupation number ni , where


X
N= ni (11.10)
i

Bosons ni ∈ {0, 1, 2, . . .} and Fermions ni ∈ {0, 1} We can generalize the anti(symmetrization)


operator
(±) 1 X
SN = (±)p P (11.11)
N! p

11.4 Second Quantization


Define a multiparticle state vector as
1
 

|n1 , n1 , . . . , ni , . . .i = Πi √ (ai )ni |0i, (Fock State) (11.12)
ni

where we have assumed that each particle does not interact. Now everything can be expressed in
terms of creation a†i and annihilation operators ai .
P
n
a†j |n1 , n1 , . . . , ni , . . .i = nj + 1(±1) k<j k |n1 , n1 , . . . , ni , . . .i (11.13)
p

If we label the particles in the order the are created that is the ”first” particle and the ”second”
particle we see from permutation for fermions and bosons that

a†i a†j |0i = ±a†j a†i |0i (11.14)

From this we get the commutation relations


h i
a†i , a†j = 0 Bosons (11.15)
n o
a†i , a†j = 0 Fermions (11.16)

29
Notice the Pauli Exclusion Principle is build into the relations a†i a†i = 0.

We can also define the very useful counting operator


X †
N= ai ai (11.17)
i

How do we express operators in Fock Space?

ki a†i ai
X X
K= k i Ni = (11.18)
i i

Suppose our Fock state is written in a basis |li with the creation and annihilation operator b†i and
bi that are not eigenstates for K. It makes sense to postulate that

a†i =
X †
bj hlj |ki i (11.19)
j
X
ai = bj hkj |li i (11.20)
j

From this by insterting we get X


K= b†m bn hlm |K|ln i (11.21)
mn

11.5 Interaction among Particles


Many particles system can interact among themselves
1X
V= Vij (Ni Nj − Ni δij ) (11.22)
2 ij

The last factor is just counting the number of interactions without selfinteraction. This can be
rewritten to a different basis again.

30
12 Quantization of Electromagnetic Waves
Discuss the quantization of electromagnetic waves and the properties of simple quantum states of
the electromagnetic field.

12.1 Electrodynamics
Wave equation for magnetic potential in free space in the Coulomb gauge, ∇ · A = 0, useful when
source free as will be assumed.
1 ∂2A
∇2 A − 2 2 = 0 (12.1)
c ∂t
where you can find the fields as

B=∇×A (12.2)
1 ∂A
E=− (12.3)
c ∂t
We also have hat k · A = 0 from the Coulomb gauge The general solution to the wave equation is
X
A(x, t) = êkλ Ak,λ e−i(ωk t−k·x) + ê∗kλ A∗k,λ ei(ωk t−k·x) , (12.4)
k,λ

where êkλ are two unit vectors, which will turn out to be the polarization vectors, and Ak,λ (x, t)
can be written as

12.2 Finding the Hamilton


Classically the energy of the field is given by
1
Z
E= |E| + |B|2 d3 x (12.5)
 2 
8π V
V X 2h † i
= k Ak,λ Ak,λ + Ak,λ A†k,λ (12.6)

k,λ

where we have found E and B in terms of A and we sum over some finite volume V , which limits
L (nx , ny , nz ).
k = 2π

If we let A†k,λ be the creation and annihilation operators ak,λ we get.


X1 h
† †
i
E= ~ωk ak,λ ak,λ + ak,λ ak,λ (12.7)
2
k,λ

where ωk =  k/c. Now photons are bosons with spin 1, so we know from second quantization
that â, ↠= 1
1
 

X
E= ~ωk ak,λ ak,λ + (12.8)
2
k,λ

So just a sum of harmonic oscillators.

12.3 Zero Point Energy and Casimir Effect


The field now has energy without photons
X1
E= ~ωk (12.9)
2
k,λ

Gives rise to the Casimir effect. Limits the k inside gives a positive pressure inwards.

31
12.4 Quantum states of the electric field
Single mode Fock state
(a†k,λ )
|nk,λ i = √ |0i (12.10)
nk,λ
Using the general solution of the A field, one can determine the E field. We can use this to
determine the dispersion
hnk,λ |Ê|nk,λ i = 0 (12.11)
Find the derivative of  to find Ê.

hnk,λ ||Ê|2 |nk,λ i =


6 0 (12.12)

So the dispersion is positive. This is called quantum fluctuations.

h∆Ei = hE 2 i − hEi2 (12.13)

12.5 Coherent States


The electric field contains two terms, one involving the annihilation operator E(+) , and one in-
volving the creation operator E(−) . The E(+) leads to coherent states, which are minimum
uncertainty states that are poisson distributed in the limit gauss distributed.

They are eigenstates for the lowering operator

32
13 Dirac Equation
Discuss the solutions of the Dirac equation for free particles and their physical interpretation.

13.1 Relativistic Quantum Mechanics


The energy in relativity is given by p
E = + p2 + m2 (13.1)
The square root is problematic. The Klein Gordon Equation tries to solve this by considering
E 2 or H 2 and using canonical quantization of →
− −i~∇ and E → ∂
− i~ ∂t on

H 2 ψ = E 2 ψ = (p2 + m2 )ψ (13.2)

Cannot take spin into account and requires negative energy interpretation.

13.2 The Dirac Equation


The Dirac equation can be written as

(iγ µ ∂µ − m)Ψ(x) = 0, (13.3)

where γ µ are four 4x4 matrices. We choose the Dirac representation, where they have the form
1 0 0
   
σi
γ0 = , γi = , (13.4)
0 −1 −σi 0

Where σi are the Pauli spin matrices. This satisfies E 2 = p2 + m2 , hence negative energy solution
are allowed. Follows the continuity equation and has a great positive probability interpretation.

This has the general solution of the form

ψ = u(p)ei(p·r−Et)/~ (13.5)

Setting p = 0, where only the first term in Einstein sum reamins. We get the equation:
Ψ1 Ψ1
   
∂ Ψ2   = m  Ψ2  ,
i  (13.6)
 
∂t Ψ 3  −Ψ 3
Ψ4 −Ψ4

which have the solutions


1 0 0 0
       

−imt 0 −imt 1 imt 0 imt 0


Ψ1 = e Ψ2 = e Ψ3 = e   , Ψ4 = e   . (13.7)
       
0 , 0 , 1 0
0 0 0 1

Each entry is interpreted as spin and particle-antiparticle. Notice the phase is opposite on the
antiparticle which usually lead to the interpretation that they moves backwards in time.

This can also be seen by operating with the spin operator


~ σ 0
 
~
S= Σ= , (13.8)
2 2 0 σ
When we consider a non zero momentum in p = pẑ we get
1 p
   
− E(p)+m
(+)  0  0
uR =   , u(+) =  (13.9)
 
p ,
 E(p)+m  R  1 
0 0

33
13.3 Dirac Sea
The negative energy solution are problematic has no stable ground state unless we consider the
Dirac sea.
(i) Vacuum is filled with negative energy electrons. That are not observables.
(ii) Absorbtion can promote them to positive energy electrons

(iii) The hole is the positron with some mass, but opposite charge and momentum.

34

You might also like