Вы находитесь на странице: 1из 19

A Summary of Quantum Mechanics

Oliver Nikolic
oliver.nikolic1@hotmail.com
2015

Contents
2.9 Conservation of Probability . . .
1 Mathematical Formulation of Quantum Mechanics
2
2.10 The Harmonic Oscilator . . . . .
1.1 Inner product . . . . . . . . . . . . 2
1.2 Operators . . . . . . . . . . . . . . 2 3 Theory of Angular Momentum
1.3 Heisenbergs Uncertainty Principle
3
3.1 Orbital Angular Momentum . . .
1.4 Completeness . . . . . . . . . . . . 3
3.1.1 Spherical Harmonics . . .
1.5 Matrix Representation of Operators 4
3.2 Spin Angular Momentum . . . .
1.6 Continuous Bases . . . . . . . . . . 4
3.2.1 Spin-1/2 Particles . . . .
1.7 Momentum-Space Wave function . 4
3.3 Total Angular Momentum . . . .
3.4 Addition of Angular Momentum
2 Quantum Dynamics
5
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8

The Schrodinger Equation . . . .


Time Evolution of Expectation
Values . . . . . . . . . . . . . . .
The Schrodinger Picture . . . . .
The Heisenberg Picture . . . . .
The Interaction Picture . . . . .
Symmetry and Conservation . . .
Commuting Operators . . . . . .
Path Integral Formulation . . . .
2.8.1 Connection to Statistical
Mechanics . . . . . . . . .

.
.
.
.
.
.
.
.
.

. 9
. 10

.
.
.
.
.
.

11
11
12
12
13
14
14

4 Identical Particles and Manyparticle Systems


15
6
4.1
Fermions
and
Bosons
.
.
.
.
.
.
.
.
16
7
4.2
Systems
of
Identical
Non7
interacting Particles . . . . . . . . 16
7
8
17
8 5 Perturbation Theory
5.1
Degenerate
Perturbation
Theory
.
17
8
5.2 Time Dependent Perturbation
9
Theory . . . . . . . . . . . . . . . . 18

MATHEMATICAL FORMULATION OF QUANTUM MECHANICS

Mathematical Formulation of Quantum Mechanics

In Paul Diracs formulation of quantum mechanics1 ; a physical state is fully described by an abstract
vector, |i, called ket vector2 , and is defined in a so called Hilbert space. A Hilbert space H is a
linear complex vector space of dimension corresponding to the number of states that the physical
system can be in, and may be infinite. Each ket vector has its corresponding bra vector, h|, in
a dual Hilbert space Hd . It is useful to think of ket vectors as column vectors and bra vectors
as complex conjugated row vectors, where these are related through the Hermitian adjoint an
operation composed of complex conjugation and vector transpose, and is denoted with a dagger
as |i = h| [5, pp. 84-85].

1.1

Inner product

The inner product of |i and |i forms a bra ket, and can be represented in an orthonormal basis
as

1
n
2 X


h|i = 1 , 2 , , n . =
i i .
(1.1)
..
i=1

n
The ket vectors |i and |i are orthogonal if h|i = 0. Furthermore, its convenient to work with
normalized ket vectors, such that for any ket vector |i,
h|i =

n
X

i i =

i=1

1.2

n
X

|i |2 = 1.

(1.2)

i=1

Operators

Physical quantities that can be measured are called observables. Each observable has its corresponding operator which produces a new ket vector when acting on a ket vector. For example,

consider the operator A acting on |i to produce |i, i.e. A|i


= |i. The corresponding Hermitian
adjoin relation becomes h| = h|A .
When an operator acting on a ket vector produces the same ket vector multiplied with a constant,
then the ket vector is an eigenvector3 , and the constant is the corresponding eigenvalue.

A|i
=

|i
|{z}

a
|{z}

a C.

(1.3)

eigenvalue eigenvector

The eigenvalues of an Hermitian operator are


An operator A is said to be Hermitian if A = A.
real and the eigenvectors corresponding to different eigenvalues are orthogonal.
An operator A is said to be unitary if the inverted operator is equal to its Hermitian adjoint,
where I is the unit
i.e. A = A1 . The inverse of the operator is defined in the relation A1 A = I,
operator which leaves a ket vector unchanged when acting on it [5, pp. 91,98,100].
1 See for example: Paul A.M. Dirac. The Principles of Quantum Mechanics, Oxford University Press, 4th edition,
1967.
2 A ket vector can also be called a state vector.
3 An eigenvector may also be referd to as an eigenstate in this context.

MATHEMATICAL FORMULATION OF QUANTUM MECHANICS

Since quantum mechanics is probabilistic, we are forced to deal with probabilities of different
the expectation value
outcomes of measurements. For an observable represented by the operator A,

of the corresponding observable in the state |i is given by hAi = h|A|i [6, pp. 24].

1.3

Heisenbergs Uncertainty Principle

Heisenbergs uncertainty principle is a generalization of Cauchy-Schwarz inequality for two observ which states
ables A and B and the associated operators A and B,
1
|h[A, B]i|,
(1.4)
2
q
2 , and analogue for B [5, pp. 95where A is the standard deviation in A, i.e. A = hA2 i hAi
96]. This inequality implies an uncertainty in the observables whos operators do not commute.
As an example, consider the one-dimensional momentum operator in position representation, p =
i~/x, and the position operator x
= x. The commutator of x
and p (multiplied with a ket
vector in this case) becomes
AB

[
x, p]|i = i~x

|i + i~ x|i = i~x |i + i~|i i~ |i = i~|i,


x
x
x
x

(1.5)

from which one concludes that


[
x, p] = i~,
and hence, Eq. (1.4) implies that
xp

1.4

~
.
2

(1.6)

Completeness

Any ket vector can be expanded in a complete orthonormal basis {|ni} as


X
|i =
cn |ni.

(1.7)

Multiplying Eq. (1.7) with hn0 | from the left gives that cn = hn|i due to the orthonormality of
the base vectors: hn0 |ni = n,n0 , where is the Kronecker delta function. Inserting the expression
for cn back in Eq. (1.7) gives
X
X
|i =
hn|i|ni =
|nihn|i,
(1.8)
n

where one can concludes that


X

|nihn| = I,

(1.9)

where I is just the identity operator which leaves a ket vector unchanged when acting on it. Equation
(1.9) is the so called completeness relation, and is frequently used.

1.5

MATHEMATICAL FORMULATION OF QUANTUM MECHANICS

Matrix Representation of Operators

we can express an operator as a matrix


By inserting Eq. (1.9) twice, on each side of an operator A,
XX
0 ihn0 |,
A =
|nihn|A|n
(1.10)
n

n0

i corresponds to the matrix elements with


where hn|A|n
n, n0 {1, 2, 3, . . . , k}, then we can express the operator A

h1|A|1i h1|A|2i

h2|A|1i

h2|A|2i

A = .
..
..
..
.
.

hk|A|1i
hk|A|2i
...
0

1.6

row index n and column index m. If


as

h1|A|ki

h2|A|ki

(1.11)
.. .
.

hk|A|ki

Continuous Bases

So far we have only considered representations in discrete bases. In continuous orthonormal bases,
the previously discussed relations changes slightly. Here, we will denote discrete complete orthonormal bases as {|ni} or {|n0 i} and denote the continuous complete orthonormal bases {|i}. The
orthonormality between ket vectors changes as
hn0 |ni = n0 ,n hn0 |n i = (n0 n),

(1.12)

where (m n) is the Dirac delta function. The completeness relation (Eq. (1.9)) changes from a
sum to an integral as
Z
X

|nihn| = I d|ih| = I.
(1.13)
n

The inner product (Eq. (1.1)), changes as


Z
X
h|i =
h|nihn|i dh|ih|i.

(1.14)

By expressing a ket vector in the position basis {|xi}, which is obvious a continuous basis, we
obtain the familiar wave function as hx|i = (x). The inner product of |i and |i may then be
expressed in terms of wave functions as
Z
Z
h|i = dxh|xihx|i = dx (x)(x),
(1.15)
where it has been used that ha|bi = hb|ai .

1.7

Momentum-Space Wave function

Consider the eigenstates |xi and |pi of the operators x


and p, i.e. x
|xi = x|xi and p|pi = p|pi. The
momentum
and
position
representations
of
the
state
|i
can
be
expressed,
using the completeness
R
R
relations dx|xihx| = 1 and dp|pihp| = 1 as
Z
Z
(x) = hx|i = dphx|pihp|i = dphx|pi(p),
(1.16)
4

Z
(p) = hp|i =

QUANTUM DYNAMICS

Z
dxhp|xihx|i =

dxhx|pi(x).

(1.17)

The function hx|pi can be found by considering


hx|p|pi = phx|pi = i~

hx|pi,
x

(1.18)

ipx

where the solution is easily obtained as hx|pi = Ae ~ . The normalization constant can be found,
for example, by requiring that the ket vectors |xi are orthogonal, i.e.
Z
Z
Z
ip
0
0
0
0
0

2
(1.19)
hx |xi = (x x ) = dphx |pihp|xi = dphx |pihx|pi = |A|
dpe ~ (x x) .
|
{z
}
2~(xx0 )

This shows that A = 1/ 2~. The wave functions (1.16) and (1.17) becomes then
Z
ipx
1
(x) =
dp(p)e ~ ,
2~
Z
ipx
1
(p) =
dx(x)e ~ .
2~

(1.20)
(1.21)

The wave function (x) can be viewed as the Fourier transform of (p), and (p) can be viewed
as the inverse Fourier transform of (x), or vice versa depending on the sign convention used in
defining the Fourier transform.

Quantum Dynamics

(t, t0 ) = ei(tt0 )H/~


When a systems Hamiltonian is time independent, the time evolution operator U
,
takes a state |(t0 )i into the state |(t)i when acting on it, where t t0 [6, p. 70-72]. If t = t0 , its
(t0 , t0 ) = I.
Furthermore, one can easy verify that the time evolution operator
easy to verify that U

is unitary since U = U , but not Hermitian.


As an example, consider an electron in a magnetic field modeled with the time-independent
B = Sz Bz , where S
= (Sx , Sy , Sz ) is the spin vector operator, B is the
= S
Hamiltonian H
magnetic flux density applied in the z-direction and is a constant. If the initial state is |, t0 i =
z with eigenvalues ~/2 and ~/2
1 |+i + 1 |i, where |+i and |i are the eigenvectors of S
2
2
respectively, the state at a later time t becomes

(t, t0 )|, t0 i = 1 ei(tt0 )(Bz )/2 |+i + 1 ei(tt0 )(Bz )/2 |i.
|, ti = U
2
2

2.1

(2.1)

The Schrodinger Equation

It is easy verify that the time evolution operator is satisfying the fundamental equation governing
(t, t0 )
the dynamics of quantum mechanical systems; the Schrodinger equation4 . Differentiation of U
4 Schrodingers original development of this equation can be viewed in: Schrdinger E. Quantisierung als Eigenwertproblem, Annalen der Physik, 79(6), 1926.

QUANTUM DYNAMICS

gives
1

U (t, t0 ) = H
U (t, t0 ).
t
i~

(2.2)

(t, t0 )|(t0 )i = |(t)i


Multiplying (2.2) with the ket vector |(t0 )i on the right and using that U
gives

|(t)i = H|(t)i.
(2.3)
t
The position representation of (2.3) gives the wave function Schrodinger equation. If the Hamilto = p2 + V (
nian is describing a particle experiencing a potential V (
x), i.e., H
x), then
2m
i~

hx|i~

p2
p2
|(t)i = hx|
+ V (
x) |(t)i = hx|
|(t)i + hx|V (
x)|(t)i
t
2m
2m

~2 2

(x, t) + V (x)(x, t) = i~ (x, t)


2
2m x
t
2

(2.4)

(2.5)

~
x) = V (x)hx|, since p2 and V (
x) are
where we have used that hx|
p2 = 2m
x2 hx| and hx|V (
necessarily Hermitian. This approach to obtain the time-dependent wave Schrodinger equation
(2.5) can easily be generalized to obtain the multidimensional time-dependent wave Schrodinger
equation as

~2  2
2
2 

+
+
(x, y, z, t) + V (x, y, z)(x, y, z, t) = i~ (x, y, z, t),
2
2
2
2m x
y
z
t

(2.6)

or in short-hand notation

2.2

~2 2

(r, t) + V (r)(r, t) = i~ (r, t).


2m
t

(2.7)

Time Evolution of Expectation Values

The expectation value of an observable A in the state |(t)i (putting t0 = 0, without loss of
generality) is
= h(t)|A|(t)i

(t)AU
(t)|(0)i.
hAi
= h(0)|U
(2.8)
d
d
It is easy to work out an expression for dt
hAi by using the Schrodinger equation twice; i~ dt
U (t) =
d

(t) and i~ U
(t) = U (t)H
to obtain
HU
dt


d
(t) iH AU
(t) U
A iH U
(t) + U
(t) A U
(t) |(0)i
hAi = h(0)| U
dt
~
~
t

d
1
A
hAi = h[A,
H]i + h
i.
dt
i~
t

(2.9)

(2.10)

2.3

QUANTUM DYNAMICS

The Schrodinger Picture

So far, we have assigned the time dependence to the state vectors, whereas operators have been
assumed to have no time dependence. This representation of quantum dynamics is known as the
Schrodinger picture [5, p. 572].
(t, t0 )|(t0 )i.
|(t)i = U
(2.11)
Differentiation Eq. (2.11) with respect to time and multiplying with i~ resembles the Schrodinger
equation
d

(2.12)
i~ |(t)i = H|(t)i.
dt

2.4

The Heisenberg Picture

In the Heisenberg picture, the situation is reversed, i.e. state vectors does not depend on time but
operators do [5, pp. 572-573]. In this picture we will write AH (t) and |iH to denote that the
operator and the ket is in the Heisenberg picture. To get the time-independent ket vector in the
Heisenberg picture |iH from the Schrodinger picture, we have to apply the inverse time translation
(t), i.e
operator U
(t)|(t)i = U
(t)U
(t)|(t)i = |(0)i.
|iH = U
(2.13)
The expectation value of an operator in the Heisenberg picture is

H h|AH (t)|iH

(t)AU
(t)|(0)i = h(t)|A|(t)i,

= h(0)|U

(2.14)

where the operator AH in the Heisenberg picture is given as


(t)AU
(t).
AH (t) = U

(2.15)

An important remark is that the expectation values are equal in both the Schrodinger and the
Heisenberg picture.
d H
A can easily be verify to become
The expression for dt
d
1

AH = [AH , H],
dt
i~

(2.16)

which is the so called Heisenberg equation of motion [5, pp. 572-573].

2.5

The Interaction Picture

In the interaction picture, both ket vectors and operators possess time dependence [5, pp. 573-574].
= H
0 + V (t),
This picture is very useful when the Hamiltonian can be split into two terms; H
where V (t) carries the total time dependence. The state vectors |(t)iI in the interaction pictures
are defined as
(t)|(0)i.
|(t)iI = U
(2.17)
The time evolution of the state vectors in this picture are governed by
i~

d
|(t)iI = VI (t)|(t)iI .
dt

(2.18)

QUANTUM DYNAMICS

The representation of an operator is


(t)AU
(t),
AI (t) = U

(2.19)

1
d
0 ].
AI (t) = [AI , H
dt
i~

(2.20)

and its dynamics are governed by

[5, pp. 573-574].

2.6

Symmetry and Conservation

Whenever an operator A commutes with the Hamiltonian and is explicitly not depending on time;
H]
= 0, and /tA = 0, then the observable A is a constant of motion as time evolves. If the
[A,
commutes with
operator A commutes with the Hamiltonian, then any function of the operator f (A)
the Hamiltonian.
In classical mechanics, conservation of energy, momentum and angular momentum are consequences of homogenity of time, homogenity of space and isotropy of space respectively. In analogy,
in quantum mechanics whenever the Hamiltonian is invariant under a unitary transformation, the
generator of transformation is conserved [5, pp. 182-183].

t (t) = eiHt/~
Consider the time translation operator U
, the position translation operator

x (x) = eixp/~

U
and the rotation operator U () = einJ which rotates the system around n
is the angular momentum operator (3.18). These operators are clearly unitary. The correwhere J
respectively
p and J
sponding generators of transformation for time, position and rotation are H,
[5, pp. 183-187].

2.7

Commuting Operators

When operators commute; the corresponding observables can be measured simultaneously to arbitrary accuracy. If a system has a complete set of commuting operators (CSCO), then it is possible
to construct simultaneous eigenstates of these set of operators [5, pp. 175-177].

2.8

Path Integral Formulation

In Feynmans path integral formulation of quantum mechanics5 , the probability amplitude of being
in the state |x(t)i at a time t, given that the system earlier was in the state |x0 (t)i at the time t0 is
given by the propagator [6, p. 128]

(t, t0 )|x0 i =
K(x, t; x0 , t0 ) = hx|U

x0

x0

exp
x0

hZ

t0

i
i
L(q, q,
t)dt dx1 dx2 dxn ,
~

(2.21)

where L is the Lagrangian, and the integrals are performed aver all possible connections between
the initial state and the final state in the time-space plane. By introducing D[x(t)] = dx1 dx2 ...dxn ,
5 See Feynmans original development of the Path integral formulation in: R.P. Feynman. Space-Time Approach
to Non-Relativistic Quantum Mechanics, Reviews of Modern Physics, 20(2), 1948.

QUANTUM DYNAMICS

we may express Eq. (2.21) more as


(t, t0 )|x0 i =
K(x, t; x , t ) = hx|U
0

D[x(t)]exp

hZ

x0

t0

i
i
L(q, q,
t)dt .
~

(2.22)

The propagator can also be interpreted as the Greens function for the Schrodinger equation [6,
p. 119], satisfying


K(x, t; x0 , t0 ) = (x x0 )(t t0 ).
(2.23)
i~ H
t
When the propagator is acting on an initial wave function, it produces the final wave function
as an integral transform according to [6, p. 118]
Z
(x, t) = dx0 K(x, t; x0 , t0 )(x0 , t0 ).
(2.24)
2.8.1

Connection to Statistical Mechanics

Theres a close relation between the propagator and the partition function in statistical mechanics.
Consider the propagator
X

(t, 0)|x0 i = hx|eiHt/~


K(x, t; x0 , 0) = hx|U
|x0 i =
hx|eiHt/~ |nihn|x0 i =
(2.25)
n

eiEn t/~ hx|nihn|x0 i =

eiEn t/~ hn|x0 ihx|ni,

where the relation (1.9) has been used for the eigenstates |ni of the Hamiltonian. If we now require
that x0 = x and perform a Wick rotation, i.e. it/~ = 1/Kb T , we then find that
Z
X
X
dx
eiEn t/~ hn|xihx|ni =
eEn ,
(2.26)
n

and hence the partition function Z is given by


Z
~
; x, 0).
Z = dxK(x,
i

2.9

(2.27)

Conservation of Probability

The probability density function is in position representation (x, t) = (x, t)(x, t) = |(x, t)|2 .
We start by consider


(x, t)(x, t) = i~(x, t) (x, t) + i~ (x, t) (x, t).


t
t

Now, consider the Schrodinger equation and its complex conjugated equation,
i~

i~

i~

(x, t)
~2 2 (x, t)
=
+ V (x, t)(x, t),
t
2m x2

(x, t)
~2 2 (x, t)
=
+ V (x, t) (x, t).
t
2m
x2
9

(2.28)

(2.29)

(2.30)

QUANTUM DYNAMICS

Multipling Eq.(2.29) with (x, t) and multiplying Eq.(2.30) with (x, t) and subtracting one of
these equations from the other gives
i~ (x, t)

2 (x, t) 
(x, t)
(x, t)
~2 
2 (x, t)
(x, t)
(x, t)
. (2.31)
+ i~(x, t)
=
2
t
t
2m
x
x2

Combining Eq. (2.9) and Eq. (2.31) gives then the conservation law
(x, t)

+
J(x, t) = 0,
t
x

(2.32)

where the probability current density is [5, pp. 181-182]


J(x, t) =

i~ 
(x, t)
(x, t) 
(x, t)
.
(x, t)
2m
x
x

(2.33)

This approach can easily be generalized to the tree-dimensional case to obtain


(r, t)
+ J(r, t) = 0,
t
where
J(r, t) =


i~ 
(r, t) (r, t) (r, t)(r, t) .
2m

(2.34)

(2.35)

As an example, consider the plane wave wave function (x, t) = Aei(kxt) . Equation (2.33)
gives after simplifications that the probability current density becomes J(x, t) = |A|2 ~k
m = v,
since (x, t) = (x, t)(x, t) = |A|2 . This can be connected to the classical current density
Jclassical = nqv, where q is the electron charge as Jclassical = qJ.

2.10

The Harmonic Oscilator

In this section, we are going to solve the harmonic oscillator problem in an algebrac way using
creation- and annihilation operators, see for example [5, pp. 239-245] for details. The Hamiltonian
for the harmonic oscillator is
2
= p + 1 m 2 x
H
2 .
(2.36)
2m 2
p

By introducing x
0 = x
m/~ and p0 = p/( m~), we may rewrite the Hamiltonian (2.36) as
= ~ (p0 2 + x0 2 ).
H
2

(2.37)

Furthermore, introducing the annihilation operator a


= 12 (
x0 + i
p0 ) and the creation operator
a
= 12 (
x0 i
p0 ), and noting that the commutator a
a
can be calculated as
a
a
=

1 0 2 0 2
i 0 0
(p + x ) + [
x , p ],
2
2

(2.38)

where we easy can find out that [


x0 , p0 ] = i if [
x, p] = i~. Hence we get that Eq. (2.36) can be
rewritten as



1
= ~ a
+1 ,
H
a +
= ~ N
(2.39)
2
2
10

THEORY OF ANGULAR MOMENTUM

=a
and N
commute, hence these operators
where N
a
is the number operator. Its obvious that H
possess common eigenstates |ni such that
|ni = n|ni,
N

(2.40)


1

~|ni.
(2.41)
H|ni
= En |ni = n +
2
It can be shown that n is a positive number. The creation- and annihilation operators acts on the
states |ni as

a
|ni = n + 1|n + 1i,
(2.42)
a
|ni =

n|n 1i.

(2.43)

The eigenstates can be obtained by acting with the creation operator a


on the vacuum state |0i,
2
1
1

|0i and so forth. Generally,


|1i = 2 a
i.e., |1i = a
|0i, |2i = 2 a
1  n
|ni =
a

|0i.
n!

(2.44)

We will now represent the eigenstates |ni in position space in order to obtain
the wave
p
functions.
Recall that the creation operator is defined as a
= 12 (
x0 i
p0 ) = 12 (
x m/~ i
p/ m~).
 n
1  m  n2 
~ d n
1
|0i =
x
0 (x),
n (x) = hx|ni = hx| a
m dx
n!
n! 2~

(2.45)

where 0 (x) is the ground state wave function, which can be determined by requiring that n|0i = 0
1/4 mx2
R
e 2~ .
and using the normalization condition |0 (x)|2 dx = 1 to become 0 (x) = m
~

Theory of Angular Momentum

When dealing with angular momentum, we are forced to work in multi-dimensional space since
angular momentum doesnt exist in one-dimensional space. We will discuss orbital angular momentum and spin angular momentum, and how to add these together to obtain the total anuglar
momentum.

3.1

Orbital Angular Momentum

= (L
= r P,
where
x, L
y, L
z ) can be written as L
The orbital angular momentum operator L

r is the position operator and P is the momentum operator, hence L = i~r in position
are related to its components as L
2 = L
2 + L
2 + L
2 , where the components are
representation. L
x
y
z
easily found to be

x = i~ y z ,
L
z
y


y = i~ z x
L

,
x
z

11


z = i~ x
L

y
.
y
x

(3.1)

THEORY OF ANGULAR MOMENTUM

i, L
j ] = i~ijk L
k , where ijk is the antisymThese components obeys the commutation relations [L
2
commutes with any component L
i , hence its possible to find a
metric tensor. The operator L
common basis of eigenvectors |ml , li of these operators. The rule for the qunatum numbers defining
2 and L
z
the eigenvectors is that ml = l, l + 1, ..., l 1, l for a given value of l. The actions of L
on the common eigenstates |ml , li are
2 |ml , li = ~2 l(l + 1)|ml , li,
z |ml , li = ~ml |ml , li.
L
L
(3.2)
+ and the lowering operator L
can be written L
= L
x iL
y . These
The raising operator L
operators changes the quantum number ml by 1 and acts as
p
|ml , li = ~ l(l + 1) ml (ml 1)|ml 1, li.
(3.3)
L
See for example [5, pp. 283-285, 301] for derivations.
3.1.1

Spherical Harmonics

2 can be represented in spherical coordinates as


z and L
The operators L



2 
2 = ~2 1 sin + 1
z = i~ .
L
,
L
2
2
sin

sin

(3.4)

2 can be represented as a wave function, which clearly only


z and L
The common eigenstates of L
depends on the angels and , i.e.
h, |ml , li = Yl,ml (, ),

(3.5)

where Yl,ml (, ) are the so called normalized spherical harmonics. These are the solutions to the
eigenvalue-eigenvector equations:
2 Yl,m (, ) = ~2 l(l + 1)Yl,m (, ),
L
l
l

(3.6)

z Yl,m (, ) = ~ml Yl,m (, ),


L
l
l

(3.7)

and
which can be shown to be [5, pp. 303-306]
s
(1)l 2l + 1 (l ml )! iml 1
dlml
Yl,ml (, ) = l
e
(sin )2l ,
m
2 l!
4 (l + ml )!
sin d(cos )lml

3.2

m 0.

(3.8)

Spin Angular Momentum

Spin is an intrinsic angular momentum that particles possess and is a purely quantum mechanical
= (Sx , Sy , Sz ) is related to its components as S
2 = S2 + S2 + S2
effect. The spin operator S
x
y
z
and obeys the same commutation relations as the orbital angular momentum; [Si , Sj ] = i~ijk .

Furthermore, the actions of the respective operators when acting on the common eigenstates of S
and Sz are also similar,
2 |ms , si = ~2 s(s + 1)|ms , si,
S
12

Sz |ms , si = ~ms |ms , si

(3.9)

THEORY OF ANGULAR MOMENTUM

S = Sx iSy ,

(3.10)

p
S |ms , si = ~ s(s + 1) ms (ms 1)|ms 1, si.

(3.11)

And as previously, ms = s, s + 1, ..., s 1, s for a given value of the spin quantum number6 s.
3.2.1

Spin-1/2 Particles

For a spin-1/2 particle (s = 1/2) ms = {1/2, 1/2}. The two different quantum numbers ms
corresponds to two different spin states
complete basis in a two  which forms an
 orthonormal

1
0
1 1
1 1
dimensional Hilbert space: | 2 , 2 i =
and | 2 , 2 i =
.
0
1
Recall the section of matrix representation of operators, which shows that we can represent an
operator as a matrix according to (1.11). In the basis {|ms , si} = {| 21 , 12 i, | 12 , 21 i}, Sz can be
represented as



 1 1
~ 1 0
h 12 , 12 |Sz | 12 , 21 i
h 2 , 2 |Sz | 21 , 12 i

=
.
(3.12)
Sz =
2 0 1
h 12 , 12 |Sz | 12 , 12 i h 12 , 12 |Sz | 12 , 12 i
Now we want to find the raising- and lowering operators. Since the raising operator S+ increases
ms with 1 and ms is restricted to ms = {1/2, 1/2}, we must have that S+ | 21 , 12 i = 0. In analogy
for S , one concludes that S | 12 , 12 i = 0. We get then, using (1.11) again and (3.11) that
S+ =
and
S =

h 12 , 12 |S+ | 12 , 12 i
h 21 , 12 |S+ | 21 , 12 i

h 12 , 12 |S+ | 12 , 12 i
h 12 , 21 |S+ | 12 , 12 i

h 12 , 12 |S | 21 , 12 i
h 12 , 12 |S | 12 , 12 i

h 12 , 12 |S | 21 , 12 i
h 12 , 12 |S | 12 , 12 i


=~


0
0


1
,
0

(3.13)

=~


0
1


0
.
0

(3.14)

Now, using equation (3.10) allows us to write Sx = 12 (S+ + S ) and Sy = 2i (S S+ ). Hence,


we get the other spin operator components as




~ 0 1
~ 0 i

Sx =
,
Sy =
.
(3.15)
2 1 0
2 i 0
This is the representation of the spin operator components in the frame of the z-component, which
is convention. The total spin operator vector can be related to a vector of matrix components as
~
= ~

S
i , where
2 , where each component is related as Si = 2






0 1
0 i
1 0

x =
,

y =
,

z =
.
(3.16)
1 0
i 0
0 1
These are the Pauli spin matrices for spin-1/2 particles [5, pp. 297-301]. To obtan the matrix
2 , one can follow the previous approaches and calculate the matrix
representation of the operator S
6 When

refering to spin (for example, an electron has spin 1/2), we mean the value of s.

13

THEORY OF ANGULAR MOMENTUM

= S2 + S2 + S2 since we have already calculated


elements. Its also possible to use the fact that S
x
y
z
the components of the spin operator. One gets with the latter approach
2
2 = ~
S
4

3.3

0
1


1
0
0
1

 2
~ 0
1
+
0
4 i

i
0


0
i

 2
~ 1
i
+
0
4 0

0
1


1
0



3~2 1
0
=
0
1
4


0
. (3.17)
1

Total Angular Momentum

= (Jx , Jy , Jz ) is the sum of the orbit and spin contributions, i.e.


The total angular momentum J
obeys the same commutator relations as L
and S:
[Ji , Jj ] = i~ijk Jk
J = L + S. The operator J
2
, Ji ] = 0,
and [J
2 = J2 + J2 + J2 .
J
x
y
z

(3.18)

2 and Jz on their common eigenstates |mj , ji are


The actions of J
2 |mj , ji = ~2 j(j + 1)|mj , ji
J

Jz |mj , ji = ~mj |mj , ji.

(3.19)

The raising- and lowering operators are defined as


J = Jx iJy ,
and the action on the common eigenstates are
q
J |mj , ji = ~ j(j + 1) mj (mj 1)|mj 1, ji,

(3.20)

(3.21)

where mj = j, j + 1, ..., j 1, j [5, pp. 285-290].

3.4

Addition of Angular Momentum

1 and J
2 , which are operating on different
Consider two different angular momentum operators, J

subspaces and hence commute with each other, i.e. [J1 , J2 ] = 0. Individually, each operator
obey the usual commutation relations as previously. The problem in adding angular momenta, i.e.
= J
1 + J
2 = J
1 I + I J
2 is in finding the eigenvalues and eigenvectors of J
2 and
finding J
2
2
, J1 , J2 all commute, its possible to find a common basis
,J
Jz . However, since the operators J
2
1
z
z
of eigenvectors where these operators can jointly be diagonalized. We will denote this common
eigenstate |m1 , m2 , j1 , j2 i and can be thought of as a product of the eigenstates in each subspace,
|m1 , j1 i and |m2 , j2 i,
|m1 , m2 , j1 , j2 i = |m1 , j1 i |m2 , j2 i,

|m1 , m2 , j1 , j2 i H = H1 H2 .

|m1 , m2 , j1 , j2 i H = H1 H2 .

(3.22)
(3.23)

The actions of the relevant operators on the common eigenstates are (see for example [6, pp. 223-226]
or [5, pp. 403-409])
2 |m1 , m2 , j1 , j2 i = ~2 j1 (j1 +1)|m1 , m2 , j1 , j2 i,
J
1

J1z |m1 , m2 , j1 , j2 i = ~m1 |m1 , m2 , j1 , j2 i, (3.24)


14

IDENTICAL PARTICLES AND MANY-PARTICLE SYSTEMS

2 |m1 , m2 , j1 , j2 i = ~2 j2 (j2 +1)|m1 , m2 , j1 , j2 i,


J
2

J2z |m1 , m2 , j1 , j2 i = ~m2 |m1 , m2 , j1 , j2 i. (3.25)

Furhtermore, the raising and lowering operators are related as J = J1 +J2 , where each individual
operator acts as
p
(3.26)
J1 |m1 , m2 , j1 , j2 i = ~ (j1 m1 )(j1 m1 + 1)|m1 1, m2 , j1 , j2 i,
p
J2 |m1 , m2 , j1 , j2 i = ~ (j2 m2 )(j2 m2 + 1)|m1 , m2 1, j1 , j2 i.

(3.27)

The basis {|m1 , m2 , j1 , j2 i} is complete;


j1
X

j2
X

|m1 , m2 , j1 , j2 i hm1 , m2 , j1 , j2 | = 1,

(3.28)

m1 =j1 m2 =j2

and orthonormal;
hm1 , m2 , j1 , j2 |m1 , m2 , j1 , j2 i = j10 ,j1 j20 ,j2 m01 ,m1 m02 ,m2 .

(3.29)

We want to find the state |m, ji in the basis {|m1 , m2 , j1 , j2 i}. Inserting the identity gives the
relation in terms of the Clebsch-Gordan coefficients
|m, ji =

j1
 X

j2
X


|m1 , m2 , j1 , j2 i hm1 , m2 , j1 , j2 | |m, ji =

(3.30)

m1 =j1 m2 =j2

j1
X

j2
X

m1 =j1 m2 =j2

hm1 , m2 , j1 , j2 |m, ji |m1 , m2 , j1 .j2 i.


{z
}
|

Clebsch-Gordan coefficients

The Clebsch-Gordan coefficients (CGC) are non-zero when the condistions: m = m1 + m2 and
|j1 j2 | j j1 + j2 are met simultaneously. These rules are the selection rules for the CGC. The
problem in adding angular momenta is thus to find the CGC [5, pp. 406-411].

Identical Particles and Many-particle Systems

So far, we have been treating particles as if we can distinguish them. In classical mechanics, this
is the proper way to treat the problem in terms of distinguishibility where its possible to track
every single particles trajectory in phase space. When describing many particle systems, we will
have to construct symmetric- or antisymmetric states depending on the type of particles that we
are considering, takes into account the fact that we can not tell which particle is in which state [5,
pp. 455-463].
Consider a two-particle system and suppose we have obtained an normalized asymmetric wave
function (x1 , x2 ). It is then possible to construct a symmetric wave function as s (x1 , x2 ) as

1 
s (x1 , x2 ) = (x1 , x2 ) + (x2 , x1 ) ,
2
15

(4.1)

IDENTICAL PARTICLES AND MANY-PARTICLE SYSTEMS

and an antisymmetric wave function as



1 
s (x1 , x2 ) = (x1 , x2 ) (x2 , x1 ) .
2

(4.2)

This principle can be generalized to a system of N identical particles [5, p. 464].

4.1

Fermions and Bosons

A fermion is a particle with half-odd integral spin, such as the electrons, neutrons and quarks.
In constrast, bosons are particles with integral spin, such as photons and pions, to name a few.
The wave function that is describing bosons must be symmetric, and remain unchanged upon
interchaning two particle. For fermions there must be the reversed; the wave function is asymmetric
upon interchanging two particles [5, pp. 460-464].

4.2

Systems of Identical Non-interacting Particles

For a system of N non-interacting particles, the Schrodinger equation is reduced to N single-particle


equations


~2 2
i + V (
xi ) i (xi ) = Ei i (xi ),
i = i...N
(4.3)

2m
For a system of two identical particles the symmetric wave function s (x1 , x2 ) and the antisymmetric wave function a (x1 , x2 ) can be constructed from the single-particle wave functions as

1 
s (x1 , x2 ) = 1 (x1 )2 (x2 ) + 1 (x2 )2 (x1 )
2

1 
a (x1 , x2 ) = 1 (x1 )2 (x2 ) 1 (x2 )2 (x1 )
2

(4.4)
(4.5)

so when the permutation operator P which interchanges x1 with x2 , acts on these states, one obtains
P s (x1 , x2 ) = s (x1 , x2 )

(4.6)

P a (x1 , x2 ) = a (x1 , x2 ).

(4.7)

This is fine as long as the particles are in different states, i.e. at least one of the quantum numbers
defining the states 1 (x) differs from 2 (x). If these the particles states however are equal, the
antisymmetric wave function becomes zero, which can be considered to be a consequence of the
Pauli exclusion principle. The symmetric wave function however, needs a more general formula for
taking this into account. The appropriate formulas for a system of N identical particles with N1
particles in state 1, N2 particles in state 2 and so on becomes
s
N1 !N2 ! . . . NN ! X

P 1 (x1 )2 (x2 ) . . . N (xN ),


(4.8)
s (x1 , x2 , ..., xN ) =
N!
P
1 X
a (x1 , x2 , ..., xN ) =
(1)P 1 (x1 )2 (x2 ) . . . N (xN ),
N! P
16

(4.9)

PERTURBATION THEORY

P
where P is over the terms which gives distinct terms.
When including spin degree of freedom, we can interpret the spin state as a wave function
(s). For bosons, this implies that for a N particle system, (x1 , x2 , ..., xN )(s1 ,2 , ..., sN ) must by
symmetric, and for fermions (x1 , x2 , ..., xN )(s1 ,2 , ..., sN ) must be antisymmetric [5, pp. 467-469].

Perturbation Theory

Perturbation theory can be applied when a weak perturbation is described by an extra term in
the Hamiltonian, whereas the the solutions is know or can be obtained for the unperturbed part of
Hamiltonian.
First we consider the time-independent non-degenerate case. Assuming that we can split the
into a term of a known solution H
0 and one small perturbation term H
p as
Hamiltonian H
=H
0 + H
p = H
0 + V ,
H

 1,

(5.1)

where we know the solutions En0 and |n i to


0 |n i = E 0 |n i.
H
n

(5.2)

is then in a first order approximation


The eigenfunctions |n i of the perturbed Hamiltonian H
given as
X hm |V |n i
|m i,
(5.3)
|n i = |n i +
En Em
m6=n

and the eigenvalues En is in a second order approximation given as


p |n i +
En = En0 + hn |H

X |hm |V |n i|2
.
En Em

(5.4)

m6=n

See for example [5, pp. 490-493].

5.1

Degenerate Perturbation Theory

which can be split into a term of known solutions H


0 and one
We consider again a Hamiltonian H

small perturbation term Hp as


n i = (H
0 + H
p )|n i = En |n i,
H|

(5.5)

where we know the solutions: En0 and the f-fold degenerate eigenvectors |n i = 1, 2, ..., f to
0 |n i = En0 |n i.
H

The methodology to follow is to calculate

p |n i
hn1 |H
1
hn |H
p |n i
2
1

..
Hp =
.

h |H

nf
p |n1 i
.

(5.6)

the following matrix for each f -fold degenerate level,


p |n i hn |H
p |n i
hn1 |H
2
1
f
p |n i hn |H
p |n i
hn2 |H
2
2
f

..
..
..
(5.7)

.
.
.

p |n i . . . hn |H
p |f i
hnf |H
2
1
f

17

PERTURBATION THEORY

After diagonalizing the matrix, the eigenvalues En1 and the corresponding eigenvectors a =
(a,1 , a,1 , ..., a,f ) are found. The energy eigenvalues are given to first order by
En = En0 + En1 ,

(5.8)

and the corresponding eigenvectors are given to zeroth order by


|n i =

f
X

a, |n i.

(5.9)

=1

See for example [5, pp. 496-498].

5.2

Time Dependent Perturbation Theory

Consider a system described by a Hamiltonian with a time-dependent perturbation term V (t) and
0 with known solutions:
a time-independent part H
=H
0 + V (t),
H

(5.10)

0 |n i = En |n i.
H

(5.11)

The probability for a transition to occur from an initial state |i i to a final state |f i is then given
by
2
iZ t
0


hf |V (t0 )|i ieif i t dt0 ,
(5.12)
Pif (t) =
~ 0
where
f i =



Ef Ei
0 |f i hi |H
0 |i i .
= hf |H
~

(5.13)

In the case when V is time-independent, Eq. (5.12) can be simplified as


Pif (t) =



|hf |V |i i|2
2 f i t
sin
.
2
~2 f i
2

(5.14)

Furthermore, a simplification of Eq. (5.12) can be madein the long time limit, i.e. when t , as
2t
|hf |V |i i|2 (Ef Ei ).
(5.15)
~
When we are considering a transition to a continuum of states, the transition probability is given
as
2t
|hf |V |i i|2 (Ei ),
(5.16)
Pif (t) =
~
where (Ei ) is the number of states per unit energy at Ei [5, pp. 576-579].
Pif (t) =

18

REFERENCES

References
[1] Paul A.M. Dirac. The Principles of Quantum Mechanics, Oxford University Press, 4th edition,
1967.
[2] Schrdinger E. Quantisierung als Eigenwertproblem, Annalen der Physik, 79(6), 1926.
[3] R.P. Feynman. Space-Time Approach to Non-Relativistic Quantum Mechanics, Reviews of Modern Physics, 20(2), 1948.
[4] R. Shankar. Principles of Quantum Mechanics, 2nd edition, 1994.
[5] N. Zettili. Quantum Mechanics: Concepts and Applications, 2nd edition, 2009.
[6] J.J. Sakurai, Jim J. Napolitano. Modern Quantum Mehcanics, 2nd edition, 2014.

19

Вам также может понравиться