0% found this document useful (0 votes)
21 views

Chapter 3

This document discusses the interaction of light with condensed matter from a quantum perspective. It outlines how the quantum electromagnetic field interacts with atomic quantum states, covering classical electrodynamics, quantum electrodynamics in the dipole approximation, and calculations of spontaneous emission, absorption, and stimulated emission rates. The total Hamiltonian is derived that includes terms for the free particle, its interaction with electromagnetic fields via the vector and scalar potentials, and the free photon Hamiltonian. Perturbation theory is used to analyze processes involving single photon creation and annihilation.

Uploaded by

teshomedebela
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
21 views

Chapter 3

This document discusses the interaction of light with condensed matter from a quantum perspective. It outlines how the quantum electromagnetic field interacts with atomic quantum states, covering classical electrodynamics, quantum electrodynamics in the dipole approximation, and calculations of spontaneous emission, absorption, and stimulated emission rates. The total Hamiltonian is derived that includes terms for the free particle, its interaction with electromagnetic fields via the vector and scalar potentials, and the free photon Hamiltonian. Perturbation theory is used to analyze processes involving single photon creation and annihilation.

Uploaded by

teshomedebela
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 43

Home Page

Absorption and Emission of


Title Page Radiation
Contents

JJ II

J I
By L. Salasnich
Page 1 of 43

Go Back

Full Screen Interaction of Light with Condensed Matter (Phys 843)


Close

Quit Department of Physics, AAU


Addis Ababa, Ethiopia

October 24, 2023


OUTLINE

1. Introduction

Home Page 2. The Total and Interaction Hamiltonians


Title Page
3. Classical Electrodynamics
Contents
4. Quantum Electrodynamics in the Dipole Approximation
JJ II

J I
5.Spontaneous Emission

Page 2 of 43 6. Absorption

Go Back
7. Stimulated Emission
Full Screen
8. Selection Rules
Close
9. Einstein Coefficients
Quit
1. Introduction

I In this chapter, we investigate the role of the quantum electromagnetic


field on the spontaneous and stimulated transitions between two atomic
Home Page
quantum states.
Title Page

I First, we review some properties of classical electrodynamics, and then


Contents
proceed to analyze the quantum electrodynamics within the dipole ap-
JJ II
proximation.
J I
I Later, employing the quantum mechanical approach, we calculate the
Page 3 of 43
rate of spontaneous emission, absorption, and stimulated emission and
Go Back connect them with the transition coefficients introduced by Einstein.

Full Screen I Finally, we discuss the lifetime of an atomic state and the line width of
Close an electromagnetic transition.

Quit
2. The Total and Interaction Hamiltonians

I Electromagnetic (EM) fields are always coupled to charged particles.


Here, coupling of the quantized EM field to nonrelativistic charges is con-
Home Page
sidered. It is expected that this is a situation where a perturbation ap-
Title Page
proach can be applied.
Contents
I In particular, we want to see the emission of light that is expected, ei-
JJ II
ther spontaneously due to fluctuations in the EM field, or stimulated by
J I the EM field.
Page 4 of 43
• The usual way to investigate light-matter interactions is to include
Go Back
interactions via modifying the four-momentum of the particle of charge
Full Screen q, according to the effects of the field it experiences.
Close
• That is, the Hamiltonian for light-matter interaction can be obtained
Quit by starting from the classical field-free relation E = \vec{p}2 /2m be-
tween the energy E and momentum p~ of the particle, making in it
the following substitutions:

~
p~ → p~ − q A, and ~ →E
E ~ − qφ, (1)
~ is the vector potential and φ is the scalar potential. Recall
where A

that the energy operator is E = i~ ∂t and the momentum operator is
p~ = −i~∇.

Home Page • The original problem in the absence of the fields is a Schrodinger
Title Page
problem HΨ = EΨ,
" #
Contents p~2 ∂Ψ(~r, t)
+ V (~r) Ψ(~r, t) = i~ . (2)
2m ∂t
JJ II
where V (~r) given by
J I
N N
X 1X qi q j
Page 5 of 43 V (~r) = qi ϕ i = . (3)
i=1
2 i,j=1 4π0 | ~ri − ~rj |
Go Back

is the Hamiltonian of the Coulomb interaction.


Full Screen

Close
• When the EM fields are now turned on, this changes to
" #
Quit 1 ~ 2 + V (~r) + qφ Ψ(~r, t) = i~ ∂Ψ(~r, t) ,
(~p − q A) (4)
2m ∂t

so that the effective Hamiltonian for the perturbed problem is just


1 ~ 2 + V (~r) + qφ.
H= (~p − q A) (5)
2m
• Let us consider the interaction EM field with a single particle of charge.
In addition, there is supposed to be the Hamiltonian for the free EM
(quantized) fields themselves. So, the total effective Hamiltonian for
a particle in the field takes the form:
Home Page

1 ~ 2 + V (~r) + qφ + Hrad ,
Title Page H = (~p − q A) (6)
2m
Contents p~2 q ~ 2 2
~ + q A + qφ + Hrad .
= + V (~r) − (A · p~ + p~ · A)
2m 2m 2m
JJ II
- The first two terms are the free particle Hamiltonian,
J I
- the next four terms are due to the particle’s interaction with the
Page 6 of 43
EM fields, and
Go Back
- the last term is the free-photon Hamiltonian, whose eigenstates
Full Screen
are states with certain numbers of photons of given wave vec-
Close tors and polarizations.

Quit • Also, applying the radiation (Coulomb) gauge, where ∇ · A~ = 0 and


~=A
φ = 0, as we used for the free photons, we obtain p~ · A ~ · p~. That
is,

~ · p~)ψ = A
(A ~ · (−i~∇ψ), and ~ = −i~∇ · (Aψ).
(~p · A)ψ ~
But,
~ = (∇ · A)ψ
∇ · Aψ ~ +A~ · ∇ψ = A
~ · ∇ψ,

~ = 0 (applying the Lorentz gauge). Thus,


since ∇ · A
Home Page
~ · p~ = p~ · A
A ~ = −i~A
~ · ∇.
Title Page

Contents • Accordingly, Eq. (6) becomes

JJ II p~2 q ~ q 2 A2
H= + V (~r) − (A · p~) + + Hrad . (7)
J I
2m m 2m

Page 7 of 43 • Note that there are only two interaction terms - one that is linear in
~ and one that is quadratic in A.
A ~
Go Back

Full Screen • The linear terms will involve individual photon creation and annihi-
lation terms, corresponding to single-photon processes that can be
Close
analyzed in first order perturbation theory (PT).
Quit

• The quadratic terms correspond to two-photon processes even in


first order PT, and probably to get those correct would actually re-
quire second order PT.
• However, the quadratic term should be smaller compared with the
linear term so that the leading perturbation is due to only the single-
photon term, approximated by
N 
Home Page X q ~ q 2 A2  q ~
Ĥint = − A · p~ + '− A · p~, (8)
Title Page i=1
m 2m m

Contents where Ĥint describes the interaction between the atomic electrons
JJ II and the radiation field.

J I ~ operator is due to the EM field (creation/annihilation


• Note that in Eq. (8) the A
Page 8 of 43
while the p~ operator is associated with the charged particle momen-
tum only. These would act on composite bra or kets involving both
Go Back
the EM field state and the particle state.
Full Screen
• Then, taking into account Eq. (8), the total Hamiltonian of the sys-
Close
tem reduces to:
Quit
Ĥ = Ĥmat + Ĥrad + Ĥint = Ĥ0 + Hint , (9)

where Hint is given by Eq. (9) and Ĥ0 is


p~2
Ĥ0 = + V̂ (~r) + Ĥrad , (10)
2m
with Hrad given by
Z 
1 1 2
Ĥrad = 0 E 2 +
B dV (11)
2 µ0
Z "  ~ 2
#
Home Page 1 ∂A 1 ~ 2 dV .
= 0 − + (∇ × A)
2 ∂t µ0
Title Page

Contents
At this point, it is worth noting the following facts:

JJ II – In a basic process where the final state has one less photon than
J I the initial state, a photon was absorbed by the matter. On the
other hand, if the final state has one more photon than the ini-
Page 9 of 43
tial state, a photon was emitted by the matter.
Go Back
– In absorption, the absorbed photon could come from any di-
Full Screen
rection dΩ. In emission, the emitted photon can go out in any
Close direction dΩ.

Quit
In either case, some density of states is involved in the calcula-
tion of the rate for the process, consistent with conservation of
energy and other variables.
3. Classical Electrodynamics

Let us consider a classical system composed of N particles with masses


mi , electric charges qi , positions ~ri and linear momenta p~i , under the pres-
Home Page
ence of an electromagnetic field.
Title Page

Contents • The Hamiltonian of free matter is given by


JJ II N
X p~2i
Hf ree = , (12)
J I i=1
2mi

Page 10 of 43
where p~2i /(2mi ) = p2i /(2mi ) is the kinetic energy of i-th particle.
Go Back
• The presence of the electromagnetic field is modelled by the Hamil-
Full Screen
tonian
Close H = Hshif t + Hrad , (13)
Quit where
N ~ i )2
X (~pi − qi A
Hshif t = + qi φi , (14)
i=1
2mi
~i = A
with A ~ i (~ri , t), φi = φi (~ri , t), where φ(~r, t) and A(~
~ r, t) are the elec-
tromagnetic potentials and
Z " #
 0 ~2 1 ~ 2 (~r, t) .
Ĥrad = d3~r E (~r, t) + B (15)
2 T 2µ0
Home Page is the radiation Hamiltonian.
Title Page
• We observe that by using the Hamilton equations
Contents ∂Hshif t
~r˙i = . (16)
∂~pi
JJ II

J I ∂Hshif t
p~˙i = − . (17)
∂~ri
Page 11 of 43
on the shift Hamiltonian (14) it is straightforward to derive Newton
Go Back equations with the Lorentz force acting on the i-th particle as
Full Screen
h ∂A~i i
¨
mi~ri = qi ∇i φi − ~ ~ i + ~vi × B
+ ~vi × (∇i × Ai ) = qi (E ~ i) (18)
∂t
Close
where ~vi = ~r˙i and, as always, the electric field E
~ is obtained as
Quit
~
~ = −∇φ − ∂ A .
E (19)
∂t
~ can be written as
while the magnetic field B
~ = ∇ × A.
B ~ (20)
~ T which, in the
• In Eq. (15) it appears the transverse electric field E
Coulomb gauge ∇ · A ~ = 0, is related to the total electric field E
~ by
~ L and transverse E
the following decomposition into longitudinal E ~T
fields
Home Page
~ =E
E ~L + E
~T (21)
Title Page
such that
~
Contents ~ L = ∇φ,
E ~ T = − ∂A .
E (22)
∂t
JJ II
~ L gives rise to the instan-
• Remarkably, the longitudinal electric field E
J I
taneous electrostatic interaction between the charges. Indeed, Eq.
Page 12 of 43 (14) can be rewritten as
N ~ i )2
Go Back X (~pi − qi A
Hshif t = + HC , (23)
Full Screen i=1
2mi
where
Close N N
X 1X qi qj
HC = qi φi = . (24)
Quit
i=1
2 i,j=1 4π0 | ~ri − ~rj |
is the Hamiltonian of the Coulomb interaction.

• The Hamiltonian Hshif t can be further decomposed as follows

Hshif t = Hmat + Hint , (25)


where
Hmat = Hf ree + HC , (26)

is the Hamiltonian of the self-interacting matter and


Home Page N 
X qi ~ q 2 A2 
Ĥint = − Ai · p~i + i i , (27)
Title Page
i=1
mi 2mi
Contents
is the interaction Hamiltonian between matter and radiation.
JJ II
• Combining the above results, we obtain the complete Hamiltonian
J I

Page 13 of 43 H = Hshif t + Hrad = Hmat + Hint + Hrad . (28)

Go Back
• Notice, however, that this Hamiltonian does not take into account
Full Screen
the possible spin of particles.
Close

Quit
4. Quantum Electrodynamics in the Dipole Ap-
proximation

Home Page
• The quantization of electrodynamics is obtained promoting the clas-
sical Hamiltonian of the system to a quantum Hamiltonian.
Title Page

Contents
• For simplicity we consider the hydrogen atom with Hamiltonian
p~2 e2
JJ II Ĥmatt = − , (29)
2m 4π0 | ~r |
J I
where p~ = −i~∇ is the linear momentum operator of the electron in
Page 14 of 43 the state | p~i, in the presence of the radiation field with Hamiltonian
Go Back XX
Ĥrad = ~ωk â†ks âks , (30)
Full Screen s
k

Close where âks and â†ks are the annihilation and creation operators of the
photon in the state | ksi.
Quit

• Here, −e is the electric charge of the electron, e = 1.60 × 10−19 C is


the electric charge of the proton, and m = me mp /(me + mp ) ' me is
the reduced mass of the electron-proton system, with me = 9.11 ×
10−31 kg the electron mass and mp = 1.67 × 10−27 kg the proton mass.
• We take into account the matter-radiation interaction by using the
so-called dipolar Hamiltonian
e ~
ĤD = A(0, t) · p~. (31)
m
Home Page

• With respect to the complete interaction Hamiltonian HI , given in


Title Page
our case by Eq. (??) with N = 1, we neglect the term e2 A2 /(2m) of Eq.
Contents
(??) which represents only a tiny perturbation in the atomic system.
JJ II
• In addition, we also neglect the spatial variations in the vector po-
J I
tential operator
Page 15 of 43 r
XX ~ h ~ ~
i
~ r, t) =
A(~ âks ei(k·~r−ωk t) + â†ks e−i(k·~r−ωk t) ûks . (32)
Go Back
k s
20 ωk V
Full Screen
This dipolar approximation, which corresponds to
Close
~ 1
e±ik·~r = 1 ± i~k · ~r + (i~k · ~r)2 ± ... ' 1, (33)
Quit 2
is reliable if ~k · ~r  1, namely if the electromagnetic radiation has a
wavelength λ = 2π/ | ~k | very large compared to the linear dimen-
sion R of the atom. Indeed, the approximation is fully justified in
atomic physics where λ ' 10−7 m and R ' 10−10 m.
• Notice, however, that for the γ electromagnetic transitions of atomic
nuclei the dipolar approximation is not good and it is more con-
~ r, t) into vector
venient to expand the vector potential operator A(~
spherical harmonics, i.e., photons of definite angular momentum.
Home Page

Title Page • The total Hamiltonian of our system is then given by

Contents
H = H 0 + HD , (34)
JJ II
where
J I
H = H0 + Hrad , (35)
Page 16 of 43
is the unperturbed Hamiltonian, whose eigenstates are of the form
Go Back
| âi | ...nks ...i =| âi⊗ | ...nks ...i, (36)
Full Screen

Close where | âi is the eigenstate of Ĥmatt with eigenvalue Ea and | ...nks ...i
P
is the eigenstate of Ĥrad with eigenvalue ks ~ωk nks , i.e.,
Quit

Ĥ0 | âi | ...n̂ks ...i = (Ĥmatt + Ĥrad ) | âi | ...nks ...i (37)
 X 
= Ea + ~ωk nks | âi | ...nks ...i.
ks
The time-dependent perturbing Hamiltonian is instead given by
r
e X ~ h i
ĤD (t) = âks e−iωk t + â†ks eiωk t ûks · p~ˆ. (38)
m ks 20 ωk V
Home Page
Fermi golden rule: Given the initial | ii and final | f i eigenstates
Title Page
of the unperturbed Hamiltonian Ĥ0 under the presence to the time-
Contents dependent perturbing Hamiltonian ĤD , the probability per unit time
JJ II
of the transition from | ii to | f i is given by

J I Wif = | h| f | ĤD (0) | ii |2 δ(Ei − Ef ), (39)
~
Page 17 of 43
with the constraint of energy conservation.
Go Back

• This is the so-called Fermi golden rule, derived in 1926 by Paul Dirac
Full Screen
on the basis of the first order time-dependent perturbation theory,
Close
and named “golden rule” few years later by Enrico Fermi.
Quit
5. Spontaneous Emission

• Let us now apply the Fermi golden rule to the very interesting case
of the hydrogen atom in the state | bi and the radiation field in the
Home Page
vacuum state | 0i.
Title Page

• We are thus supposing that the initial state is


Contents

JJ II | ii =| bi | 0i, (40)

J I
• Notice that, because we are considering the hydrogen atom, one has
Page 18 of 43

Go Back
Ĥmatt | bi = Eb | bi, (41)

Full Screen
where
mc2 α2 13.6 eV
Eb = − 2
'− , (42)
Close 2nb n2b

Quit
is the well-known quantization formula of the nonrelativistic hydro-
gen atom with quantum number nb = 1, 2, 3, ... and
e2 1
α= ' . (43)
(4π0 )~c 137
is the fine-structure constant.
Home Page

Title Page

Contents

JJ II

J I Figure 1: Spontaneous emission.


Page 19 of 43

Go Back
• In addition we suppose that the final state is
Full Screen
| f i =| ai | nks i, (44)
Close

Quit
i.e., the final atomic state is | ai and the final photon state is | ksi =|
1ks i = â†ks | 0i.

• Obviously, we have
Ĥmatt | ai = Ea | ai, (45)
where
mc2 α2 13.6 eV
Ea = − 2
'− , (46)
2na n2a
with na = 1, 2, 3, ....
Home Page
• In this process, where Eb > Ea , there is the spontaneous production
Title Page
of a photon from the vacuum: a phenomenon strictly related to the
Contents quantization of the electromagnetic field. From Eqs. (38) and (39)
JJ II one finds
2π  e 2  ~ 
J I spo
Wba = | ûks · ha | p~ˆ | bi |2 δ(Eb − Ea − ~ωk ), (47)
~ m 20 ωk V
Page 20 of 43
because
Go Back
âk0 s0 | ii = âk0 s0 | bi | 0i =| biâk0 s0 | 0i = 0, (48)
Full Screen
while
Close

â†k0 s0 | f i = â†k0 s0 | bi | 0i =| biâ†k0 s0 | 0i =| bi | k 0 s0 i, (49)


Quit

and consequently

hf | p~ˆâk0 s0 | ii = 0, hf | p~ˆâ†k0 s0 | ii = ha | p~ˆ | biδk0 ,k δs0 ,s . (50)


• Remember that p~ˆ acts on a different Hilbert space with respect to
âk0 s0 and â†k0 s0 .

• On the basis of Heisenberg equation of motion of the linear momen-


Home Page tum operator p~ˆ of the electron is given by
Title Page
p~ˆ d~r 1
= = [~r, Ĥmatt ]. (51)
Contents m dt i~

JJ II • This can also be shown by recalling the commutation relation be-


J I tween p~2 and ~r, which is given by
m
Page 21 of 43
[~r, p~2 ] = 2i~~p → p~ = [~r, Ĥmatt ]. (52)
i~
Go Back

• Employing Eq. (52, the matrix element becomes


Full Screen

m m
Close ha | p~ˆ | bi = ha | [~r, Ĥmatt ] | bi = ha | [~rĤmatt − Ĥmatt~r] | bi (53)
i~ i~
m
Quit = (Eb − Ea )hb | ~r | ai = −imωba ha | ~r | bi,
i~
where ωba = (Eb − Ea )/~, and consequently
2
spo πωba
Wba = | ûks · ha | e~r | bi |2 δ(~ωba − ~ωk ). (54)
V 0 ωk
• The delta function is eliminated by integrating over the final photon
states. The photon states allowed by the periodic boundary condi-
tion form a continuous energy spectrum as the normalization be-
comes infinite.
Home Page

Title Page
• For a photon emitted at a solid angle dΩ, the number of allowed
states in the energy interval [~ω, ~(ω+dω)] can be written as ρ~ω,dΩ d(~ω),
Contents
where
JJ II V | ~k |2 d | ~k | dΩ V ω2
ρ~ω,dΩ = = dΩ.
(2π)3 d(~ω) 8π 3 ~c3
J I
• Consequently, the probability of spontaneous emission in the solid
Page 22 of 43
angle dΩ becomes
V k 2 (ωba )
Go Back X X sp Z X sp
spo
Wba = Wba,ks = 3)
dΩ Wba,ks , (55)
Full Screen k s
(2π s=1,2

Close where dΩ is the differential solid angle.

Quit • Because ûk1 , ûk2 , and n̂ = ~k/k form a orthonormal system of vectors,
setting ~rab = ha | ~r | bi one finds

| ~rab |2 = | ûk1 · ~rab |2 + | ûk2 · ~rab |2 + | n̂ · ~rab |2 (56)


X
= | ûks · ~rab |2 + | ~rab |2 cos2 (θ),
s=1,2
where θ is the angle between ~rba and n̂. It follows immediately
X
| ûks · ~rab |2 =| ~rab |2 (1 − cos2 (θ)) =| ~rab |2 sin2 (θ). (57)
s=1,2

Home Page
namely
X
Title Page | ûks · ha | ~r | bi |2 =| ha | ~r | bi |2 sin2 (θ). (58)
s=1,2
Contents

• In addition, in spherical coordinates one can choose dΩ = sin(θ)dθdφ,


JJ II
with θ ∈ [0, π] the zenith angle of colatitude and φ ∈ [0, 2π] the az-
J I
imuth angle of longitude, and then
Page 23 of 43 Z 2π Z
Z π
2 8π
dΩ sin (θ) = dφ dθ sin3 (θ) = , (59)
Go Back
0 0 3
Full Screen where Z π Z +1
3 4
Close sin (θ)dθ = sin2 (θ)d cos(θ) = .
0 −1 3
Quit
• Expressing k in terms of ω, where k(ωba ) = ωba /c, we finally obtain
the integrated transition probability of spontaneous emission to be:
3
ωba
spo
Wba = | ha | d~ | bi |2 , (60)
3π0 ~c3
where the d~ = −e~r is the classical electric dipole momentum of the
hydrogen atom, i.e., the dipole of the electron-proton system where
~r is the position of the electron of charge −e < 0 with respect to the
proton of charge e > 0, and ha | d~ | bi = −ha | e~r | bi is the so-called
Home Page
dipole transition element.
Title Page

Contents

JJ II

J I

Page 24 of 43

Go Back

Full Screen

Close

Quit
6. Absorption

• We now consider the excitation from the atomic state | ai to the


atomic state | bi due to the absorption of one photon.
Home Page

Title Page
• Thus we suppose that the initial state is

Contents
| ii =| bi | nks i, (61)

JJ II while the final state is

J I | f i =| ai | nks − 1i, (62)

Page 25 of 43 where Ea < Eb . From Eqs. (38) and (39) one finds
2π  e 2  ~ 
Go Back
abs
Wab = nks | ûks ·hb | p~ˆ | ai |2 δ(Ea +~ωk −Eb ), (63)
~ m 20 ωk V
Full Screen because

Close hf | p~ˆâk0 s0 | ii = nks hb | p~ˆ | aiδk0 ,k δs0 ,s , hf | p~ˆâ†k0 s0 | ii = 0. (64)
Quit
• Note that with respect to Eq. (47) in Eq. (63) there is also the multi-
plicative term nks . We can follow the procedure of the previous sec-
tion to get
2
abs πωba
Wab = nks | ûks · hb | e~r | ai |2 δ(~ωba − ~ωk ). (65)
V 0 ωk
• Again the delta function can be eliminated by integrating over the
final photon states but here one must choose the functional depen-
dence of nks . We simply set

Home Page nks = n(ωk ), (66)

Title Page and after integration over k and s, from Eq. (65) we get
3
ωba
Contents
abs
Wab = 3
| hb | d~ | ai |2 n(ωba ) = Wba
spo
n(ωba ). (67)
3π0 ~c
JJ II

J I • The photon states allowed by the periodic boundary condition form


Page 26 of 43 a continuous energy spectrum as the normalization becomes infi-
nite.
Go Back

Full Screen • For a photon emitted at a solid angle dΩ, the number of allowed
states in the energy interval [~ω, ~(ω+dω)] can be written as ρ~ω,dΩ d(~ω),
Close
where
Quit V | ~k |2 d | ~k | dΩ
ρ~ω,dΩ = .
(2π)3 d(~ω)
Equivalently, this can be written in terms of the frequency interval
[ω, ω + dω] to be
V ω 2 dΩ
ρω,dΩ = .
8π 3 c3
Home Page

Title Page

Contents

JJ II

J I
Figure 2: Optical absorption.

Page 27 of 43

so that the energy density (density of allowed states per unit volume,
Go Back
V ) of the states, n(ω), with energy ~ω is obtained by integrating over
Full Screen
the solid angle, dΩ. That is, (taking the spin into account)
ω 2 [~ωn(ω)] ~ω 3
Close
Z
ρ(ω) = 2 × dΩ = n(ω). (68)
Quit
8π 3 c3 π 2 c3
where
1
n(ω) = ,
e~ω/(kB T )
−1
is the Fermi-Dirac distribution function with where kB is the Boltz-
mann constant and T the absolute temperature.
• Note that Eq. (68) represents the thermal distribution of photons,
with ρ(ω) the energy density per unit of angular frequency - the thermal-
equilibrium Planck formula.

Home Page • Consequently, the integrated transition probability for absorption


Title Page can be written as:
2 3
π spo π c
Contents abs
Wab =| hb | d~ | ai |2 ρ(ωba ) = W ba 3
ρ(ωba ). (69)
30 ~2 ~ωba
JJ II

J I

Page 28 of 43

Go Back

Full Screen

Close

Quit
7. Stimulated Emission

• Finally, we consider the stimulated emission of a photon from the


atomic state | bi to the atomic state | ai.
Home Page

Title Page
• Thus, we suppose that the initial state is

Contents | ii =| bi | nks i, (70)

JJ II while the final state is


J I | f i =| ai | nks + 1i, (71)
Page 29 of 43
where Eb > Ea . From Eqs. (38) and (39) one finds
Go Back 2π  e 2  ~ 
sti
Wba = (nks + 1) | ûks · ha | p~ˆ | bi |2 δ(Eb − Ea − ~ωk ),
Full Screen
~ m 20 ωk V
(72)
Close because
Quit

âk0 s0 | ii = âk0 s0 | bi | nks i =| biâk0 s0 | nks i = nks | bi | nks − 1i, (73)

while

â†k0 s0 | ii = â†k0 s0 | bi | nks i =| biâ†k0 s0 | nks i = nks + 1 | bi | nks + 1i,
(74)
Home Page

Title Page

Contents

JJ II
Figure 3: Stimulated emission.
J I

Page 30 of 43
and consequently
Go Back √
hf | p~ˆâk0 s0 | ii = 0, hf | p~ˆâ†k0 s0 | ii = nks + 1ha | p~ˆ | biδk0 ,k δs0 ,s .
Full Screen
(75)
Close

Quit
• Note that with respect to Eq. (64) in Eq. (72) there is the factor nks + 1
instead of nks . It is straightforward to follow the procedure of the
previous section to obtain
3
ωba
sti
Wba = 3
| ha | d~ | bi |2 [n(ωba ) + 1] = Wab
abs spo
+ Wba , (76)
3π0 ~c
sti
which shows that the probability Wba of stimulated emission re-
spo
duces to the spontaneous one Wba when n(ωba ) = 0. It is then useful
to introduce
sti sti spo
W̃ba = Wba − Wba , (77)
Home Page

which is the effective stimulated emission, i.e., the stimulated emis-


Title Page
sion without the contribution due to the spontaneous emission. Clearly,
Contents sti
for a very large number of photons (n(ωba )  1) one gets W̃ba '
sti
JJ II Wba .
J I • Moreover, for a thermal distribution of photons, with the energy den-
Page 31 of 43 sity per unit of angular frequency ρ(ω), we can also write
2 3
Go Back 1 spo π c
sti
W̃ba abs
= Wba =| ha | d~ | bi |2 ρ(ωba ) = W ba 3
ρ(ωba ). (78)
30 ~2 ~ωba
Full Screen

Close • Remarkably, the probability of stimulated emission is different from


zero only if the emitted photon is in the same single-particle state
Quit
| ksi of the stimulating ones (apart when the stimulating light is the
vacuum | 0i).

• In the stimulated emission the emitted photon is said to be “coher-


ent” with the stimulating ones, having the same frequency and the
same direction (Fig. 3).

Home Page

Title Page

Contents

JJ II

J I

Page 32 of 43

Go Back

Full Screen

Close

Quit
8. Selection Rules

• It is clear that, within the dipolar approximation, no electromagnetic


transition, either spontaneous or stimulated, will occur between the
Home Page
atomic states | ai and | bi unless at least one component of the dipole
transition matrix element hb | d~ | ai is nonzero.
Title Page

Contents
• It is possible to show that the matrix elements are zero for certain
JJ II
pairs of states. If so, the transition is not allowed (at least in the
J I electric dipole approximation), and the results can be summarized
Page 33 of 43
in terms of simple selection rules governing the allowed changes in
quantum numbers in transitions.
Go Back

Full Screen
• Since the electric dipole d~ = −e~r changes sign under parity (~r → −~r),
the matrix element hb || d~ | ai is zero if the states | ai and | bi have
Close
the same parity. Therefore
Quit
– the parity of the wavefunction must change in an electric dipole
transition.

• This means that if ψa (~r) and ψb (~r) are the eigenfunctions of the states
| ai and | bi and, for instance, both eigenfunctions are even, i.e.,

ψa (−~r) = ψa (~r) and ψb (−~r) = ψb (~r), (79)

Home Page then the dipole matrix element is such that


Z
Title Page ~
hb | d | ai = −e ψb∗ (~r)~rψa (~r)d3~r = 0. (80)
Contents
R R
JJ II • In fact, under the transformation ~r → −~r one finds −d3~r = d3~r
while hb | d~ | ai = −hb | d~ | ai, and then it follows hb | d~ | ai = 0.
J I

Page 34 of 43
• Let us recall that in our calculations a generic eigenstate | ai =| nlmi
of the matter Hamiltonian Ĥmat is such that
Go Back

Full Screen ψa (~r) = ψnlm (~r) = h~r | ai = h~r | nlmi = Rnl (r)Ylm (θ, φ), (81)

Close
where Ylm (θ, φ) is the spherical harmonic function.
Quit
• The spherical harmonics satisfy the orthonormalization condition
Z π Z 2π
sin θdθ Ylast
0 m0 (θ, φ)Ylm (θ, φ)dφ = δl,l0 δm,m0 . (82)
0 0


where Yl,m (θ, φ) = (−1)m Yl,−m (θ, φ).
• Moreover under parity transformation one finds

ψnlm (−~r) = Rnl (r)Ylm (π−θ, φ+π) = (−1)l Rnl (r)Ylm (θ, φ) = (−1)l ψnlm (~r),
(83)
Home Page
thus the radial part Rnl (r) of the wave function is unchanged and the
Title Page parity of the state is fully determined from the angular part.

Contents
• The generic dipole matrix element is given by
JJ II Z ∞ Z π Z 2π
0 0 0 ~ 3
hn l m | d | nlmi = −e r Rn0 l0 (r)Rnl (r)dr sin θdθ Ylast
0 m0 (θ, φ)
J I 0 0 0
× (cos θ sin φ, sin θ sin φ, cos φ)Ylm (θ, φ)dφ. (84)
Page 35 of 43

Go Back with ~r = r(cos θ sin φ, sin θ sin φ, cos φ).

Full Screen
• Setting ∆l = l − l0 and ∆m = m − m0 , by using the properties of
Close the spherical harmonics it is possible to prove that the dipole matrix
element hn0 l0 m0 | d~ | nlmi is different from zero if only if ∆l = ±1 and
Quit
∆m = 0, ± 1.

• Thus, in the dipole approximation, the orbital angular momentum l


of the atomic state and its third component m = −l, −l+1, ..., l−1, l
must satisfy the selection rules

4l = ±1 and 4m = 0, ±1. (85)

Home Page • This means that in the electric dipole transitions the photon carries
off (or brings in) one unit of angular momentum.
Title Page

Contents • It is important to stress that the above selection rules are obtained
within the dipole approximation. This means that they can be vio-
JJ II
lated by rare electromagnetic transitions involving higher multipo-
J I
larities.
Page 36 of 43
• Concluding this section we observe that in the general case of N par-
Go Back
ticles with positions ~ri and electric charges qi , i = 1, 2, ..., N , the elec-
Full Screen tric dipole momentum is defined
Close N
X
d~ = qi~ri . (86)
Quit i=1

• Moreover, due to Eq. (3.42), the dipolar interaction Hamiltonian


(3.20) can be effectively written as

~ˆ D = −d~ · E,
H ~ˆ (87)
which is the familiar expression of the coupling between the electric

dipole d~ and the electric field E.

Home Page

Title Page

Contents

JJ II

J I

Page 37 of 43

Go Back

Full Screen

Close

Quit
9. Einstein Coefficients

• In 1919 Albert Einstein observed that, given an ensemble of N atoms


in two possible atomic states | ai and | bi, with Na (t) the number of
Home Page
atoms in the state | ai at time t and Nb (t) the number of atoms in the
Title Page
state | bi at time t, it must be
Contents
N = Na (t) + Nb (t), (88)
JJ II

J I and consequently
dNa dNb
=− . (89)
Page 38 of 43 dt dt
Go Back • Einstein suggested that, if the atoms are exposed to an electromag-
Full Screen
netic radiation of energy density per unit of frequency ρ(ω), the rate
of change of atoms in the state | ai must be
Close

dNa
Quit = Aba Nb + Bba ρ(ωba )Nb − Bab ρ(ωba )Na , (90)
dt
where the parameters Aba , Bba , and Bab are known as Einstein coeffi-
cients.
- The first two terms in this equation describe the increase of the
number of atoms in | ai due to spontaneous and stimulated
transitions from | bi,

Home Page - while the third term takes into account the decrease of the num-
ber of atoms in | ai due to absorption with consequent transi-
Title Page
tion to | bi.
Contents

• Einstein was able to determine the coefficients Aba , Bba , and Bab by
JJ II
supposing that the two rates in Eqs. (89) and (90) must be equal to
J I zero at thermal equilibrium, i.e.,
Page 39 of 43
dNa dNb
=− = 0. (91)
Go Back dt dt

Full Screen • In this way Einstein found


Close Nb  Nb 
Aba = ρ(ωba ) Bab − Bba . (92)
Na Na
Quit

• Because the relative population of the atomic states | ai and | bi is


given by a Boltzmann factor

Nb e−βEb
= −βEa = e−β(Eb −Ea ) = e−β~ωba . (93)
Na e
Einstein got
Aba
ρ(ωba ) = . (94)
Bab eβ~ωba − Bba
• At thermal equilibrium we know that
Home Page
3
~ωba 1
Title Page ρ(ωba ) = 2 3 β~ω
. (95)
π c e ba −1
Contents
• It follows that 3
JJ II ~ωba
Aba = Bba , Bab = Bba . (96)
π 2 c3
J I
• Notice that in this way Einstein obtained the coefficient Aba of spon-
Page 40 of 43
taneous decay by simply calculating the coefficient of stimulated de-
Go Back
cay Bba .
Full Screen
• Actually, by using the results we have obtained in the previous sec-
Close tion, it is clear that
Quit 3
ωba
spo
Aba = Wba = | ha | d~ | bi |2 , (97)
3π0 ~c3

sti
W̃ba π 2 c3
Bba = = 3
Aba , (98)
ρ(ωba ) ~ωba
abs
Wab π 2 c3
Bab = = 3
Aba . (99)
ρ(ωba ) ~ωba

Home Page

Title Page

Contents

JJ II

J I

Page 41 of 43

Go Back

Full Screen

Close

Quit
Main References:

• Luca Salasnich, Quantum Physics of Light and matter: Photons, Atoms,


and Strongly Correlated Systems, Second Edition, Springer Interna-
Home Page
tional Publishing (2017).
Title Page

Contents

JJ II

J I

Page 42 of 43

Go Back

Full Screen

Close

Quit
Home Page

Title Page

Contents

JJ II

J I Thank you !
Page 43 of 43

Go Back

Full Screen

Close

Quit

You might also like