EMF waves in lossless, lossy and

conducting media

1
• Plane Wave in Lossless simple Media
• Plane Wave in good conductor
• Plane Wave in Lossy simple Media
 EMF In Media – Chapter 3
We restrict our discussion to LIH media, that is to say:
ത
ഥ=𝐵
ഥ = ε𝐸ത and 𝐻
• Linear, i.e. linear dependencies: 𝐷
𝜇

ത 𝐸,
• Isotropic, i.e. indifferent of the directions of 𝑘, ത 𝐷
ഥ , 𝐵,
ത 𝑎𝑛𝑑 𝐻,
ഥ and

• Homogeneous, i.e. ε and 𝜇, representative macroscopic characteristics of the media, are indifferent
to position (do not vary from point to point) within the media.

• Within regions in the media where there are NO free charges and/or free currents, i.e. 𝑞 = 𝐽 = 0, the
Maxwell Equations differ from the vacuum analogs only in replacement of 𝝁𝟎 𝜺𝟎 with 𝝁𝜺, becoming:
𝜕𝐵
3.1.1 𝛻 × 𝐸 = - 3.1.2 𝛻 ∙ 𝐸 = 0
𝜕𝑡

𝜕𝐸
3.1.3 𝛻 × 𝐵= 𝝁𝜺 3.1.4 𝛻 ∙ 𝐵 = 0
𝜕𝑡
Though the mathematical observation is pretty trivial, the physical implication is astonishing:
As the wave propagates through the media, the fields polarize and magnetize all the atoms/molecules
and the resulting oscillating dipoles generate their own electric fields.
These induced fields combine with the original fields in such a way as to create a single wave with
the same frequency but a different speed. This is responsible for the phenomenon of transparency.

2/27/2020 Electromagnetism 3
 Performing the procedures on equations 3.1.1 and 3.1.3 as done with equations 1.1.1 and
1.1.3 in Chapter 2, results in the wave equations in LIH media:

2 𝜕2 𝐸 2 𝜕2 𝐵
𝛻 𝐸= 𝜇𝜀 𝜕𝑡 2 and 𝛻 𝐵= 𝜇𝜀 𝜕𝑡 2 [4]
𝜇0 𝜀0 𝜇𝜀
Multiplying the right side of the two equations by yields 𝜇0 𝜀0 = 𝜇0 𝜀0 ∙ 𝜇𝑟 𝜀𝑟 =
𝜇0 𝜀0 𝜇0 𝜀0
1 1 Τ𝑣 2 1 𝑐2 𝑛2
= = = 2
𝑐 2 1 Τ𝑐 2 𝑐 2 𝑣2 𝑐

2
1 2
1 𝜇 𝜀 𝑐2 2
𝒄
𝐻𝑒𝑛𝑐𝑒: 𝑐 = ; 𝑣 = ; 𝜇𝑟 = ; 𝜀𝑟 = ; 𝜇𝑟 𝜀𝑟 = 2 = 𝑛 ⟹ 𝒏 = = 𝝁𝒓 𝜺𝒓
𝜇0 𝜀0 𝜇𝜀 𝜇0 𝜀0 𝑣 𝒗
𝑛2 𝜕 2 𝐸 𝑛2 𝜕 2 𝐵
Introducing the expressions into 4 : 𝛻2 𝐸 = and 𝛻2𝐵 = [5]
𝑐 2 𝜕𝑡 2 𝑐 2 𝜕𝑡 2

ҧത 𝑛2
Introducing the solution 𝐸 𝑥, 𝑦, 𝑧, = 𝐸0 𝑒 𝑖(𝑘∙𝑟−𝑘𝑐𝑡) into [5] gives: (𝑖𝑘)2 −
𝑐2
(𝑖ω)2 𝐸=0⇒

2 𝑛2 2 2𝜋𝑓 2𝜋 𝝀𝟎
⇒ 𝑘𝑖𝑛 𝑚𝑒𝑑𝑖𝑎 = 2 ω ⟹ 𝒇 = 𝒇𝟎 ⟹ 𝒌 = 𝑛 =𝑛 = 𝒏𝒌𝟎 ⟹ 𝝀 =
𝑐 𝑐 𝜆0 𝒏
Why?
In dielectrics and biological tissues 𝜇𝑟 ≅ 1 and hence 𝑛 = 𝜀𝑟
 ഥ and H
The relation between E ഥ in unbound, simple media
𝜕𝐸
Recalling Maxwell’s 4th equation: 𝛻 × 𝐵 = μ𝜀 𝜕𝑡 , and introducing into them the full
expression of the fields: ത ҧ ത ҧ
𝐸ത = 𝐸ത0 𝑒 𝑖(𝑘∙𝑟−𝜔𝑡) 𝑎𝑛𝑑 𝐵ത = 𝐵ത0 𝑒 𝑖(𝑘∙𝑟−𝜔𝑡) we get:
× 𝑘ത
𝑖 𝑘ത × 𝐵= 𝜇𝜀(−𝑖𝜔)𝐸 / ⟹ 𝑘ത × 𝑘ത × 𝐻=−𝜀𝜔𝑘ത × 𝐸 [6]
𝜇
Applying (from vector analysis): 𝐴ҧ × 𝐵ത × 𝐶 = 𝐴ҧ ∙ 𝐶ҧ 𝐵ത − 𝐴ҧ ∙ 𝐵ത 𝐶ҧ onto [6], one gets: 𝑬
= 0
𝑘ത × 𝑘ത × 𝐻= 𝑘ത ∙ 𝐻
ഥ 𝑘ത − 𝑘 2 𝐻
ഥ = −𝜀𝜔𝑘ത × 𝐸 ⟹
෡
𝒌
𝜔 1 𝜀 𝑘෠ × 𝐸
⟹ 𝐻=𝜀 𝑘෠ × 𝐸 = 𝜀𝑣 𝑘෠ × 𝐸 = 𝜀 𝑘෠ × 𝐸 = ෠
𝑘×𝐸 =
𝑘 𝜇𝜀 𝜇 𝜂
𝑯
෡×𝑬
𝒌
⟹ 𝑯= ෡×𝑯
𝑎𝑛𝑑 𝑠𝑖𝑚𝑖𝑙𝑎𝑟𝑙𝑦 𝑤𝑒 𝑐𝑎𝑛 𝑠ℎ𝑜𝑤 𝑡ℎ𝑎𝑡 𝑬 = −𝜼𝒌 [7]
𝜼

Where 𝜼= 𝝁𝜺 𝑖𝑠 a quantity that has units of impedance (ohms) and is defined as the intrinsic
impedance of the medium. Thus for uniform plane waves in a simple lossless medium:

the ratio of the electric and magnetic fields is 𝜂 and is determined only by the material
properties of the medium, i.e., 𝜂, 𝜀.
As for the units of 𝜂 and its value in vacuum:

𝜇0 𝜇0 𝜀0 1 1 1
𝜂0 = = = 2 = = =
𝜀0 𝜀0 2 𝑐 2 𝜀0 𝑐𝜀0 3 ∙ 108 𝑚𝑠𝑒𝑐 −1 8.85 ∙ 10−12 𝐶 2 𝑁𝑒𝑤 −1 𝑚−2

𝑁𝑒𝑤 ∙ 𝑚 𝐽𝑜𝑢𝑙𝑒
1 = 𝑉(𝑜𝑙𝑡)
= 𝐶 = 377 𝐶 = 377Ω ≅ 120𝜋Ω
3 ∙ 8.85 ∙ 10−4 𝐶 𝐼(𝐴𝑚𝑝𝑒𝑟)
𝑆𝑒𝑐
𝑉 ෡
𝒌×𝑬 𝑬
Analogy to Ohm's law: 𝐼 = ; From [7] 𝑯 = = 𝜼 ⟶ 𝑤ℎ𝑒𝑟𝑒
𝑅 𝜼

𝑯 𝑖𝑠 𝑎𝑛𝑎𝑙𝑜𝑔𝑜𝑢𝑠 𝑡𝑜 𝑡ℎ𝑒 𝑐𝑢𝑟𝑟𝑒𝑛𝑡 𝑰, 𝑬 𝑡𝑜 𝑡ℎ𝑒 𝑣𝑜𝑙𝑡𝑎𝑔𝑒 𝑽 𝑎𝑛𝑑 𝜼 𝑡𝑜 𝑟𝑒𝑠𝑖𝑠𝑡𝑎𝑛𝑐𝑒 𝑹.

Summary: As 𝜀 𝑎𝑛𝑑 𝜂 in simple LIH media are indifferent to time and position,
namely they are constant, then:

The electric and magnetic vectors of plane wave fields are perpendicular to one another. In
lossless media 𝑬 𝑎𝑛𝑑 𝑯 are in phase and both proportional to 𝒑 𝒛 − 𝒗𝒑 𝒕 , both propagate
in harmony along 𝑧 , reaching their maxima and minima at the same points in space and at
the same times.

The orientation of 𝑬 𝑎𝑛𝑑 𝑯 𝑖𝑠 𝑠𝑢𝑐ℎ 𝑡ℎ𝑎𝑡 𝑬 × 𝑯 𝑖𝑠 𝑖𝑛 𝑡ℎ𝑒 𝑑𝑖𝑟𝑒𝑐𝑡𝑖𝑜𝑛 𝑜𝑓 + 𝒛, which is

the direction of propagation of the wave.
Plane Wave in Lossy Media
Electromagnetic applications involve the interaction between electric and magnetic
fields and matter.

The important parameters in the macroscopic levels are 𝜀, 𝜇 𝑎𝑛𝑑 𝜎.

Most media exhbit nonzero conductivity, 𝜎, or complex permativuty, 𝑎𝑛𝑑 ℎ𝑒𝑛𝑐𝑒

𝑐𝑎𝑛 𝑎𝑏𝑠𝑜𝑟𝑏 𝐸𝑀 𝑒𝑛𝑒𝑟𝑔𝑦, 𝑟𝑒𝑠𝑢𝑙𝑡𝑖𝑛𝑔 𝑖𝑛 𝑎𝑡𝑡𝑒𝑛𝑢𝑎𝑡𝑖𝑜𝑛 of 𝐸𝑀 𝑤𝑎𝑣𝑒 𝑤ℎ𝑖𝑙𝑒 𝑝𝑟𝑜𝑝𝑜𝑔𝑎𝑡𝑖𝑛𝑔
𝑡ℎ𝑟𝑢𝑜𝑔ℎ 𝑡ℎ𝑒 𝑚𝑒𝑑𝑖𝑢𝑚.

The loss of energy is frequency dependent and hence determines its application
range. For instance: air is quite transparent over the radio and microwave ranges,
yet it is highly lossy medium at optical frequencies.

The electric field of a propagating wave within a conductive media induces the conductive
currents 𝑱𝑐 = 𝜎𝑬 .

𝑇ℎ𝑒𝑠𝑒 𝑐𝑢𝑟𝑒𝑛𝑡𝑠 𝑎𝑟𝑒 𝑖𝑛 𝑝ℎ𝑎𝑠𝑒 𝑤𝑖𝑡ℎ 𝑡ℎ𝑒 𝑤𝑎𝑣𝑒 𝑒𝑙𝑒𝑐𝑟𝑖𝑐 𝑓𝑖𝑒𝑙𝑑, 𝑎𝑛𝑑 𝑐𝑎𝑢𝑠𝑒 𝑑𝑖𝑠𝑠𝑎𝑝𝑎𝑡𝑖𝑜𝑛
of some of the wave enery as heat 𝑤𝑖𝑡ℎ𝑖𝑛
𝑡ℎ𝑒 𝑚𝑎𝑡𝑒𝑟𝑖𝑎𝑙 𝑖𝑛 𝑎 𝑟𝑎𝑡𝑒 𝑝𝑜𝑤𝑒𝑟 𝑝𝑒𝑟 𝑢𝑛𝑖𝑡𝑒 𝑣𝑜𝑙𝑢𝑚𝑒 𝑔𝑖𝑣𝑒𝑛 𝑏𝑦 𝑬 ∙ 𝑱.
 Uniform Plane Wave Propagation in a good Conductor

A. Reminder: we show that for a plane wave in a simple, source free, and lossless
medium (𝜌 = 𝑗 = 𝜎 = 0) that (4th Max. Eq.):

𝜕𝐸 𝑬 𝑬
𝛻×𝐻=𝜀 = −𝑖𝜔 𝜀𝐸; 𝑤ℎ𝑒𝑟𝑒 𝜀 𝑖𝑠 𝑎 𝑟𝑒𝑎𝑙 𝑞𝑢𝑎𝑛𝑡𝑖𝑡𝑦, 𝑖. 𝑒. 𝜀 = 𝜀 𝑅𝑒𝑎𝑙 ;𝐻 = 𝜼 ; 𝑩 = 𝒄 and
𝜕𝑡
hence there is no phase difference between 𝐻 , 𝐵, 𝐸 𝑎𝑛𝑑 𝐷. However,

B. In source-free conductive media, 𝜌 = 0, 𝑗𝑐 ≠ 0, 𝑎𝑛𝑑 𝜎 ≠ 0. 𝑇ℎ𝑒𝑛 𝑤𝑒 ℎ𝑎𝑣𝑒:

𝜕𝐸 𝜎
𝛻 × 𝐻 = 𝐽 ҧ + 𝜀 𝜕𝑡 = 𝜎𝐸 − 𝑖𝜔𝜀𝐸 = 𝜎 − 𝑖𝜔𝜀 𝐸 = −𝑖𝜔 𝜀 𝑅 + 𝑖 𝜔 𝐸;

𝐼𝑚 𝜎
𝜀 𝑒𝑓𝑓 ⟶ 𝜀𝑒𝑓𝑓 =
𝜔
 1
The reciprocal of the medium impedance (𝜂) would then be:

2 1/4
1/2 1/2 𝜀𝐼 𝑖𝜙
𝜎 2 1/4 𝑖𝜙
𝜀𝑅 2 + 𝜀𝐼 2 ∙ 𝑒 𝑖𝜙 𝜀𝑅 ∙ 1 + 𝑅 ∙𝑒2 𝜀𝑅 ∙ 1+ 𝑅 ∙𝑒2
1 𝜀 𝑒𝑓𝑓 𝜀 𝝈 𝜀 𝜔
= = = =⟶ 𝜺𝑰 = =
𝜂 𝜇 𝜇 𝜇 𝝎 𝜇

𝜎
In a good conductor (𝜌 = 0, 𝑗𝑐 ≠ 0, 𝑎𝑛𝑑 𝜎 ≠ 0): 𝜀𝑒𝑓𝑓 = 𝜀 + 𝑖 = 𝜀 𝑅 + 𝑖𝜀 𝐼 ,
𝜔
𝑜𝑢𝑡 𝑜𝑓 𝑤ℎ𝑖𝑐ℎ the loss thangent of the medium is:
𝜀𝐼 𝜎
𝑡𝑎𝑛𝜙𝑙𝑜𝑠𝑠 = 𝑅 = 𝑅
𝜀 𝜀 𝜔

𝝈
𝑁𝑜𝑡𝑒: 𝑖𝑛 𝑎 𝑔𝑜𝑜𝑑 𝑐𝑜𝑛𝑑𝑢𝑡𝑜𝑟 𝑎𝑛𝑑 𝑎𝑡 𝑓𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑖𝑒𝑠 𝑜𝑓 𝑣𝑖𝑠𝑖𝑏𝑒𝑙𝑒 𝑙𝑖𝑔ℎ𝑡 ≫ 𝟏 𝑎𝑛𝑑 ℎ𝑒𝑛𝑐𝑒
𝝎
𝜋
𝜙𝑙𝑜𝑠𝑠 →
2

1 𝜀𝑅 𝜎 1/2 𝜋
= ∙ ∙ 𝑒𝑖4
𝜂 𝜇 𝜀𝑅𝜔
 And the relation between ഥ ഥ in that case is:
E and H

1 1
𝐸 𝜀𝑅 𝜎 2 𝑖
𝜋 𝜎 2
𝑖
𝜋
⟹
𝐻= = ∙ 𝑅 ∙ 𝐸𝑒 4 = ∙ 𝐸𝑒 4
𝜂 𝜇 𝜀 𝜔 𝜇𝜔
>>1
= 𝜂(𝜎 = 𝐽 = 0)

𝑎. 𝐻, 𝐵 > 𝐸 𝑖𝑛 𝑐𝑜𝑛𝑡𝑟𝑎𝑠𝑡 𝑡𝑜 𝐻, 𝐵 < 𝐸, 𝑎𝑠 𝑖𝑠 𝑖𝑛 𝑎 𝑙𝑜𝑠𝑠𝑙𝑒𝑠𝑠 𝑚𝑒𝑑𝑖𝑢𝑚 𝑙𝑚 .

b. 𝐼𝑛 𝑙𝑜𝑠𝑠𝑦 𝑐𝑜𝑛𝑑𝑢𝑐𝑡𝑜𝑟 𝑚𝑒𝑑𝑖𝑢𝑚 𝑡ℎ𝑒 𝑚𝑎𝑔𝑛𝑒𝑡𝑖𝑐 𝑓𝑖𝑒𝑙𝑑 𝑙𝑎𝑔𝑠 𝑡ℎ𝑒 𝑒𝑙𝑒𝑐𝑡𝑟𝑖𝑐

𝜋
𝑓𝑖𝑒𝑙𝑑 𝑏𝑦 𝑎 𝑝ℎ𝑎𝑠𝑒 𝑑𝑖𝑓𝑓𝑒𝑟𝑒𝑛𝑐𝑒 𝑤ℎ𝑖𝑐ℎ, 𝑖𝑛 𝑎 𝑔𝑜𝑜𝑑 𝑐𝑜𝑛𝑑𝑢𝑐𝑡𝑜𝑟, 𝑚𝑎𝑦 𝑟𝑒𝑎𝑐ℎ ~ 4 .
EMF in good conductor (Copper, Aluminum, Silver, Etc.): Approximation approach 𝜔2
Dielectric component = 2
𝜎 𝜕2 𝐸 𝑣
We showed that In exhalent conductor, >>1. Applying that cond𝑖𝑡𝑖𝑜𝑛 𝑖𝑛𝑡𝑜 𝛻 2 𝐸 = 𝜇𝜀 :
𝜀𝜔 𝜕𝑡 2

𝜎
⟹ (𝑖𝑘)2 𝐸 = [𝜇𝜀𝑒𝑓𝑓 (−𝑖𝜔)2 ]𝐸 ⟹ 𝑘 → 𝑘෨ ⟹ 𝑘෨ 2 = 𝜇𝜀𝑒𝑓𝑓 𝜔2 = 𝜇 𝜀 𝑅 + 𝑖 𝜔 𝜔2 = 𝜇𝜀 𝑅 𝜔2 + 𝑖𝜇𝜎𝜔 [10]

𝜎
The ratio between the two components in the right side of Eq. 10 in the case of >>1 is:
𝜀𝑅 𝜔

𝜇𝜀 𝑅 𝜔2 𝜀 𝑅 𝜔
= ≪ 1 𝑎𝑛𝑑 ℎ𝑒𝑛𝑐𝑒 𝑡ℎ𝑒 𝑒𝑥𝑝𝑟𝑒𝑠𝑠𝑖𝑜𝑛 𝜇𝜀 𝑅 𝜔2 𝑐𝑎𝑛 𝑏𝑒 𝑛𝑒𝑔𝑙𝑒𝑐𝑡𝑒𝑑 𝑖𝑛 𝐸𝑞. 10, 𝑦𝑖𝑒𝑙𝑑𝑖𝑛𝑔:
𝜇𝜎𝜔 𝜎

𝑖𝜋 𝑖𝜋
𝜋 𝜋 𝝁𝝈𝝎 𝟏
𝑘෨ 2 = 𝑖𝜇𝜎𝜔 ⟹ 𝑘෨ = 𝜇𝜎𝜔 ∙ 𝑖 1/2 = 𝜇𝜎𝜔 ∙ e 2 = 𝜇𝜎𝜔 ∙ e 4 = 𝜇𝜎𝜔 𝑐𝑜𝑠 + 𝑖𝑠𝑖𝑛 = 1+𝑖 ≡ 1+𝑖 ⟹
4 4 𝟐 𝜹

𝑘෨ = 𝑘෨ 𝑒 𝑖𝜙 , 𝑤ℎ𝑒𝑟𝑒 𝑘෨ = 𝑎 + 𝑖𝑏, 𝑏𝑢𝑡 𝑖𝑛 𝑎 𝑔𝑜𝑜𝑑 𝑐𝑜𝑛𝑑𝑢𝑐𝑡𝑜𝑟:

𝟏 𝝁𝝈𝝎 𝝁𝝈𝟐𝝅𝒇 𝝅
𝒂=𝒃= = = = 𝝁𝝈𝝅𝒇 𝒂𝒏𝒅 𝝓 = [𝟏𝟎𝒂]
𝜹 𝟐 𝟐 𝟒
 𝐼𝑛 𝑎 𝑔𝑜𝑜𝑑 𝑐𝑜𝑛𝑑𝑢𝑐𝑡𝑜𝑟, 𝑡ℎ𝑒 𝑒𝑥𝑝𝑟𝑒𝑠𝑠𝑖𝑜𝑛 𝑓𝑜𝑟 𝐸 𝑜𝑓 𝑎 𝑤𝑎𝑣𝑒 𝑝𝑟𝑜𝑝𝑎𝑔𝑎𝑡𝑖𝑛𝑔 𝑎𝑙𝑜𝑛𝑔 𝑡ℎ𝑒 𝑧 𝑎𝑥𝑠𝑒𝑖𝑠
𝑎𝑛𝑑 𝑣𝑖𝑏𝑟𝑒𝑎𝑡𝑒𝑠 𝑎𝑙𝑜𝑛𝑔 𝑡ℎ𝑒 𝑥 𝑎𝑥𝑠𝑒𝑖𝑠, 𝑤𝑜𝑢𝑙𝑑 𝑏𝑒:

𝒛 𝒛
ഥ 𝒛, 𝒕 = 𝐸ത0 𝑒 −𝑏𝑧 𝑒 𝑖
𝑬 𝑎𝑧−𝜔𝑡 ഥ 𝟎 𝒆− 𝜹 𝒆𝒊
𝑥ො = 𝑬 𝜹
− 𝝎𝒕
𝑥ො [11]
Thus [11] presents an attenuated wave, i.e. the amplitude of the wave decreasing with increasing z.

The distance it takes to reduce the amplitude by 𝑒 −1 𝑖𝑠 𝑐𝑎𝑙𝑙𝑒𝑑 𝑡ℎ𝑒

𝟏 𝟐 𝟏
𝒔𝒌𝒊𝒏 𝒅𝒆𝒑𝒕𝒉 𝜹 ≡ → 𝒊𝒏 𝒐𝒖𝒓 𝒄𝒂𝒔𝒆 = =
𝒃 𝝁𝝈𝝎 𝝁𝝈𝝅𝒇
𝟐𝝅 𝟐𝝅 𝝅
𝑓𝑟𝑜𝑚 11 𝑤𝑒 𝑚𝑎𝑦 𝑐𝑜𝑚𝑝𝑟𝑒ℎ𝑒𝑛𝑑 𝑡ℎ𝑎𝑡 𝑡ℎ𝑒 𝑝𝑟𝑎𝑐𝑡𝑖𝑐𝑎𝑙 𝑤𝑎𝑣𝑒 # 𝑖𝑛 𝑐𝑜𝑛𝑑𝑢𝑐𝑡𝑜𝑟 𝑖𝑠 𝒂 = 𝒐𝒓 𝝀𝒄 = = 𝟐𝝅𝜹 = 𝟐
𝝀𝒄 𝒂 𝝁𝝈𝒇

So, regarding good conductors, we may say that the phase velocity 𝒗𝒑.𝒄𝒐𝒏 and the wavelength 𝝀𝒄 in it are:

𝜔 𝟐 𝟐𝝎
𝒗𝒑.𝒄𝒐𝒏 = =𝜔∙ = ; 𝝀𝒄 = 𝟐 𝝅/𝝁𝝈𝒇
𝑎 𝝁𝝈𝝎 𝝁𝝈
Electric and magnetic fields in a plane EM wave in a conductor. The wave propagate
in the +𝑧 𝑑𝑖𝑟𝑒𝑐𝑡𝑖𝑜𝑛.
N: Poynting Vector in a good conductor
Energy per unit volume of electromagnetic wave in lossless and lossy (good conductor) media

𝑢
In lossless media relation 𝑢 𝑒 is:
𝑚

2
1 ത ഥ 𝜀𝐸 2 → 𝜀𝜇 = 1 = 𝐸
𝑢𝑒 𝐸∙𝐷 𝑣2 𝜇𝑣 2
= 2 =
𝑢𝑚 1 𝐵ത ∙ 𝐻 1 𝐸 𝐸 2 =1
ഥ 𝐵 2 → 𝐵 =
2 𝜇 𝑣 = 𝜇𝑣 2
1
𝜋
𝜎 2 𝑖4
In good conducting media (gcm) we showed that H = 𝐸𝑒 . Hence:
𝜇𝜔

1ത ഥ
𝑢𝑒 2 𝐸∙𝐷 𝜀 𝐸ത ∙ 𝐸ത ∗ 𝜀𝐸 2 𝜔𝜀 𝒖𝒎 𝝈
= = = 𝜎 = ⟹ = ≫𝟏
𝑢𝑚 1 𝐵ത ∙ 𝐻
ഥ 𝜇 𝐻ഥ ∙ 𝐻ഥ ∗
𝜇 𝜇𝜔 𝐸 2 𝜎 𝒖 𝒆 𝝎𝜺
2
𝒖𝒎
What is the physical mechanism which explains the result: ≫𝟏
𝒖𝒆
Dielectric dissipative media: general treatment (The relation between
ഥ ഥ in unbuond, simple Lossy media)
E and H
Suppose we have a media with both conductance 𝜎 and imaginary permittivity ෥ε = 𝜀 𝑅 + 𝑖𝜀 𝐼 , Hence:

𝜕𝐸 𝜎
ഥ = 𝐽 ҧ + ε෤
𝛻×𝐻 = 𝜎 − 𝑖𝜔෤ε 𝐸 = 𝜎 − 𝑖𝜔(𝜀 𝑅 + 𝑖𝜀 𝐼 ) 𝐸 = −𝑖𝜔 𝜀 𝑅 + 𝑖 + 𝜀𝐼 𝐸⟹
𝜕𝑡 𝜔

𝜎 𝐼
𝜀𝑒𝑓𝑓
ε෤ = 𝜀𝜀 ∗ = [(𝜀 𝑅 )2 + ( + 𝜀 𝐼 )2 ]1/2
𝜔
⟹ 𝜀𝑒𝑓𝑓
ǁ
𝜎
𝜔 + 𝜀𝐼
𝑡𝑎𝑛𝜙𝑙𝑜𝑠𝑠 =
𝜀𝑅

1 εǁ
Therefore again: ε෤ = ε෤ 𝑒 𝑖𝜙 and consequently 𝐻= 𝐸= 𝑒 𝑖 𝜙 Τ2 𝐸
𝜂 𝜇

𝜋
Obviously, for this medium 𝜙 may differ from
2
 Next, the influence of the dissipative media upon the traveling electromagnetic wave media, is realized through the impact
of its complex permittivity 𝜀ǁ on the wave # 𝑘ത (𝑡ℎ𝑒 𝑑𝑖𝑠𝑝𝑒𝑟𝑠𝑖𝑜𝑛 𝑟𝑎𝑡𝑖𝑜) as follows:

𝜕2 𝐸 𝑤 2 𝑤
We recall in vacuum and lossless media that: 𝛻 2 𝐸 = (𝑖𝑘)2 𝐸 = 𝜇𝜀 = 𝜇𝜀(−𝑖𝜔)2 𝐸 ⟹ 𝑘 2 = 𝜇𝜀𝜔2 = ⟶𝑘=
𝜕𝑡 2 𝑣 𝑣

However, in lossy media 𝜀𝑒𝑓𝑓 = 𝜀 𝑅 + 𝑖𝜀 𝐼 ⟹ 𝑘 2 = 𝜇(𝜀 𝑅 +𝑖𝜀 𝐼 )𝜔2 = 𝜇𝜀 𝑅 𝜔2 + 𝜇𝑖𝜀 𝐼 𝜔2 ෩ 𝟐

≡ (𝒌)
lossless lossy
Out of which: 1 Component of ෩
𝑘2 Component of
1 ෩
𝑘2
𝜀𝐼 𝜀𝐼 2 𝜀𝐼
𝑘෨ 2 = 𝜇𝜀 𝑅 𝜔2 1 + 𝑖 𝑅 ⟹ 𝑘෨ = 𝜇𝜀 𝑅 𝜔 2 1 + 𝑖 𝑅 = 𝓴 1 + 𝑖𝓹 2 = 𝒂 + 𝒊𝒃, 𝑤ℎ𝑒𝑟𝑒: 𝓀 ≡ 𝜇𝜀 𝑅 𝜔 2 𝑎𝑛𝑑 𝓅 ≡ . 𝑁𝑒𝑥𝑡:
𝜀 𝜀 𝜀𝑅

1. 𝑘෨ 𝑘෨ ∗ = 𝑎2 + 𝑏2 = 𝓀2 (1 + 𝓅2 )1/2
2. 𝑘 2 = 𝑎2 − 𝑏2 + 2𝑖𝑎𝑏 = 𝓀2 1 + 𝑖𝓅 ⟹ 𝑎2 − 𝑏2 = 𝓀2 𝑎𝑛𝑑 2𝑎𝑏 = 𝓀2 𝓅
1 1 1
𝐹𝑟𝑜𝑚 (1 + 2) ⟹ 2𝑎2 + 𝑖𝓀2 𝓅 = 𝓀2 1 + 𝓅2 2 + 𝓀2 1 + 𝑖𝓅 = 𝓀2 1 + 𝓅2 2 + 𝓀2 + 𝑖𝓀2 𝓅. = 𝓀2 1 + 𝓅2 2 +1

1 1/2 1 1/2
1 + 𝓅2 2 +1 1+ 𝜀 𝐼 Τ𝜀 𝑅 2 2 +1
𝑎=𝓀 = 𝜔 𝜇𝜀 𝑅 ; 𝑎𝑛𝑑 𝑓𝑟𝑜𝑚 1 − 2 𝑏
2 2 ⟹ 𝑡𝑎𝑛𝜙𝑙𝑜𝑠𝑠 =
1/2 1/2
1+
1
𝓅2 2 −1 1+
1
𝜀 𝐼 Τ𝜀 𝑅 2 2 −1
𝑎
𝑏=𝓀 = 𝜔 𝜇𝜀 𝑅
2 2
Then , for wave propagating in Z direction (this time the wave # 𝒌 is imaginary) :
෨
𝐸ത = 𝐸ത0 𝑒 𝑖(𝑘𝑧−𝜔𝑡) = 𝑬
ഥ 𝟎 𝑒 𝑖((𝑎+𝑖𝑏)𝑧−𝜔𝑡) = 𝑬
ഥ 𝟎 𝒆−𝒃𝒛 𝑒 −𝑖(𝑎𝑧−𝜔𝑡) [8]

The right side expression is that of an attenuated wave, i.e. the amplitude of the wave decrease with increasing z.

The distance it takes to reduce the amplitude by 𝑒 −1 𝑖𝑠 𝑐𝑎𝑙𝑙𝑒𝑑 𝑡ℎ𝑒

𝟏
𝒔𝒌𝒊𝒏 (𝒑𝒆𝒏𝒆𝒕𝒓𝒂𝒕𝒊𝒐𝒏) 𝒅𝒆𝒑𝒕𝒉 𝓵 ≡ .
𝒃
𝓵 is a measure of how far the wave penetrates into the conductor deissipative media . Meanwhile, the real part of
𝑘, 𝑖. 𝑒. 𝒂 , determines the wavelength, the propagation speed and the refraction index, in the media in the usual way:

2𝜋 𝜔 𝑐 𝑐𝑎
𝜆= , 𝑣= , 𝑛= =
𝑎 𝑎 𝑣 𝜔

Like Equation [8] it is trivial to show that the same procedure works for 𝐵ത as well:
෨ 𝑧−𝜔𝑡) ෩
𝒌 ෩
𝒌
ത
𝐵 = 𝐵0 𝑒 ത 𝑖( 𝑘 ഥ
= 𝝎 𝑬𝟎 𝑒 𝑖((𝑎+𝑖𝑏)𝑧−𝜔𝑡)
= 𝝎𝑬 ഥ 𝟎 𝒆−𝒃𝒛 𝑒 −𝑖(𝑎𝑧−𝜔𝑡) [9]
Loss tangent (𝑡𝑎𝑛𝛿𝑐 = 𝜎Τ𝜔𝜖) versus frequency
For typical good conductors,
both 𝜎 𝑎𝑛𝑑 𝜖 are nearly
independent of frequencies
below the optical range.

Relative permittivity and conductivity used in the

above Figure.
Examples:
Microwave heating of milk: the dielectric properties of milk with 7.3% moisture content at
20° 𝐶 𝑎𝑛𝑑 3𝐺𝐻𝑧 are 𝜖𝑟′ = 51 𝑎𝑛𝑑 𝑡𝑎𝑛 𝛿𝑐 = 0.59. Calculate (a) 𝜖𝑟′′ and (b) the average
dissipated power per unit volume if the peak electric field inside the dielectric is 30𝑘𝑉𝑚−1 .
Solution:
(a) From 𝑡𝑎𝑛 𝛿𝑐 = 0.59 = 𝜖𝑟′′ Τ𝜖𝑟′ = 𝜖𝑟′′ Τ51 ⟹ 𝜖𝑟′′ = 51 ∙ 0.59 ≅ 30.1

Recall that 𝑢𝑒𝑙𝑜𝑠𝑠 = 𝐽𝑐ҧ ∙ 𝐸ത = 𝜎𝐸ത ∙ 𝐸ത = 𝜎𝐸 2 . However, when 𝜔𝜖 ′′ 𝑖𝑠 high an alternating

current density 𝐽𝐷ҧ = 𝜔𝜖 ′′ 𝐸ത flows within the dielectric, yielding an instantaneous loss
𝑢𝑒𝑙𝑜𝑠𝑠 = 𝐽𝐷ҧ ∙ 𝐸ത = 𝜔𝜖 ′′ 𝐸ത ∙ 𝐸ത = 𝜔𝜖 ′′ 𝐸 2 .

Consequently, the average power dissipated per unite volume for time harmonic electric
1 2
field 𝐸 = 𝐸𝑝𝑒𝑎𝑘 ∙ cos(𝜔𝑡) would be 𝑢𝑒𝑙𝑜𝑠𝑠 = 2 𝜔𝜖 ′′ 𝐸𝑝𝑒𝑎𝑘 . Hence we have:

1 ′′ 𝐸 2 1
(b) 𝜔𝜖 𝑝𝑒𝑎𝑘 ≅ 2 (2𝜋 × 3 × 109 𝑟𝑎𝑑/ sec)(30.1 × 8.85 × 10−12 𝐹/𝑚)(30𝑘𝑉/𝑚)2 ≅
2

109 𝑊
≅ 2.26 × 3 = 2.26𝑊/𝑚3
𝑚
 ෩
𝒌
𝐸ത = 𝑬
ഥ 𝟎 𝒆−𝒃𝒛 𝑒 −𝑖(𝑎𝑧−𝜔𝑡) [8] 𝐵ത = ഥ 𝟎 𝒆−𝒃𝒛 𝑒 −𝑖(𝑎𝑧−𝜔𝑡)
𝑬 [9]
𝝎

𝜕𝐵 3.1.2 𝛻∙𝐸=0
3.1.1 𝛻 × 𝐸= - 𝜕𝑡
The attenuated plane waves [8, 9] satisfy the modified equations: 𝜕𝐸
3.1.3 𝛻 × 𝐵= 𝜇𝜀 𝜕𝑡 + 𝜇𝜎𝐸 3.1.4 𝛻∙𝐵=0
For any 𝐸ത0 and 𝐵ത0 .

Yet these Maxwell’s equations impose further constraints, which serve to determine the relative amplitudes, phase
and polarization of 𝐸ത 𝑎𝑛𝑑 𝐵ഥ . Obviously, 𝛻 ∙ 𝐸 = 𝛻 ∙ 𝐵 = 0, as shown previously, rule out any 𝑧 components, i.e. the
fields are transverse.
For instance, lets choose 𝐸ത to be polarized (vibrate) along the x direction:

ത 𝑡) = 𝐸ത0 𝑒 −𝑏𝑧 𝑒 −𝑖(𝑎𝑧−𝜔𝑡) 𝑥ො ,

𝐸(𝑧, then applying 3.1.3 above yields: ത 𝑡) = 𝐵ത0 𝑒 −𝑏𝑧 𝑒 −𝑖(𝑎𝑧−𝜔𝑡) 𝑦ො
𝐵(𝑧,
So once again, 𝐸ത ⊥ 𝐵. ෨ as any complex number can be expressed in terms of its modulus and
ത On the other hand, 𝑘,
phase: 1 1
2 2
𝜀𝐼 𝜎 2 2
𝑘෨ = 𝑘෨ 𝑒 , 𝑤ℎ𝑒𝑟𝑒 𝑘෨ =
𝑖𝜙
𝑎2 + 𝑏2 = 𝜇𝜀 𝑅 𝜔 2 1+ 𝑅 ↦ 𝑐𝑜𝑛𝑑𝑢𝑐𝑡𝑜𝑟 ↦ = 𝜔 𝜇𝜀 𝑅 1+
𝜀 𝜀𝜔

Summary: According to Equations [8, 9] the complex amplitudes 𝐸෨ 0 = 𝐸0 𝑒 𝑖𝛿𝐸 𝑎𝑛𝑑 𝐵෨ 0 = 𝐵0 𝑒 𝑖𝛿𝐵 are no longer in
phase; and they are related by:
 ෨ 𝑒 𝑖𝜙
𝑘
𝑖𝜑𝐵
𝐵0 𝑒 = 𝐸0 𝑒 𝑖𝜑𝐸 ; 𝑤ℎ𝑒𝑟𝑒 𝜑𝐵 − 𝜑𝐸 = 𝜙
𝜔

The magnetic field lags behind the electric field. The (real) amplitudes of 𝐸ത and 𝐵ത are related by:

𝐸0 𝑘෨ 𝜎 2
= = 𝜀𝜇 1 +
𝐵0 𝜔 𝜀𝜔

Hence, the (real) magnetic and electric fields are:

𝒛 𝒛
𝑬 ഥ 𝟎 𝒆− 𝜹 𝒆𝒊
ഥ 𝒛, 𝒕 = 𝑬 𝜹
−𝝎𝒕+𝝋𝑬
𝑥ො
𝒛 𝒛
𝑩 ഥ 𝟎 𝒆− 𝜹 𝒆𝒊
ഥ 𝒛, 𝒕 = 𝑩 𝜹
−𝝎𝒕+𝝋𝑬 +𝜙
𝑦ො
Since both 𝒂 𝑎𝑛𝑑 𝒃 are proportional to 𝝈𝟏/𝟐 and 𝝈 is large, it appears that uniform plane waves not only are
attenuated heavily but also undergo a significant phase shift per unite length as they propagate in a good conductor.

On the other hand, since the phase velocity 𝒗𝒑.𝒄𝒐𝒏 and the wavelength 𝝀𝒄 are both proportional to 𝝈−𝟏/𝟐 , they are
both significantly smaller than the corresponding values in free space, i.e. 𝒗𝒑.𝒗𝒂𝒄 𝑎𝑛𝑑 𝝀𝒗𝒂𝒄 .

For instance, for copper (𝜎 = 5.8 ∙ 107 𝑆𝑚−1 ), at 𝟑𝟎𝟎 𝑴𝑯𝒛, 𝑤𝑒 𝑔𝑒𝑡:

𝒗𝒑.𝒄𝒐𝒑𝒑 ≅ 𝟕𝟏𝟗𝟐𝒎𝒔𝒆𝒄−𝟏 [𝑣𝑣𝑎𝑐 = 𝑐 = 3 ∙ 108 𝑚𝑠𝑒𝑐 −1 ] and 𝝀𝒄𝒐𝒑𝒑 = 𝟎. 𝟎𝟐𝟒𝒎𝒎 [𝜆𝑣𝑎𝑐 = 1 𝑚 ]

For 𝟔𝟎 𝑯𝒛 instance, for copper, the values are more dramatic:

𝒗𝒑.𝒄𝒐𝒑𝒑 ≅ 𝟑. 𝟐𝟐𝒎𝒔𝒆𝒄−𝟏 [𝑣𝑣𝑎𝑐 = 𝑐 = 3 ∙ 108 𝑚𝑠𝑒𝑐 −1 ] and 𝝀𝒄𝒐𝒑𝒑 = 𝟓𝟑. 𝟔𝒎 [𝜆𝑣𝑎𝑐 ~5000𝑘𝑚 ]

As an example of a nonmetallic conductor, for seawater (𝜀𝑟 = 81, 𝜎 = 4𝑆𝑚−1 ), at 𝟏𝟎 𝒌𝑯𝒛 one gets:

𝒗𝒑.𝑺𝑾 ≅ 𝟏. 𝟓𝟖 ∙ 𝟏𝟎𝟓 𝒎𝒔𝒆𝒄−𝟏 and 𝝀𝑺𝑾 ≅ 𝟏𝟓. 𝟖𝒎 [𝜆𝑣𝑎𝑐 ~30𝑘𝑚 ]

 Dispersion:

The frequency dependence of the permittivity

(and refractive index)
Hecht
Maxwell’s Theory treats matter as continuous, represented
by the constants 𝜀 𝑎𝑛𝑑 𝜇 with resulting 𝑛, unrealistically,
independent of the frequency of the EMF. To understand of
the dependency 𝑛 𝑓 , 𝑖. 𝑒. 𝑑𝑖𝑠𝑝𝑒𝑟𝑠𝑖𝑜𝑛, the
atomic/molecular aspect of matter must be considered.
When a dielectric is subjected to an electric field the
atomic/molecular charge distribution within it is distorted,
inducing of electric dipole moments, which in turn modifies
the total internal electric field.

There are permanent polar molecules (Figure 3.34):

Thermal agitation keeps the dipoles randomly oriented,

yielding overall zero polarization.

Introduction of electric field cause dipoles alignment, i.e.

causing orientational polarization

And in many cases ionic/crystal polarization

Polarization due to EMF
When a dielectric experiences EMF, the charges of its atoms/molecules is subjected to time varying forces/torques,
ഥ 𝒕 of the wave. Due to molecule inertia (and bounds), the higher the 𝑓 is, the
proportional to the electric field 𝑬
ഥ 𝒕 ) and consequently 𝜀 will markedly decrease.
lesser the molecule response will be (aligning with 𝑬

For instance, for water 𝜀𝑟 = 81 up to 1010 𝐻𝑧, after which it drops quite rapidly.

ഥ 𝒕 even at optical frequencies

In contrast, due to their little inertia, electrons are relatively highly responsive to 𝑬
(~1014 𝐻𝑧). 𝑇ℎ𝑖𝑠 𝑚𝑎𝑘𝑒𝑠 the permittivity and the refraction index frequency-dependent, i.e. :

𝜀 → 𝜀(𝜔) and 𝑛 → 𝑛(𝜔)

For the sake of simplicity, within the frame of a reasonable approximation and
small displacement 𝑥, we relate to the attracting electrical force acting between
the positive nucleus and the electrons within an atom/molecule, as being
spring-like i.e. picture an electron as attached to the end of an imaginary spring,
with force
𝐹𝑟𝑒𝑠𝑡𝑜𝑟𝑖𝑛𝑔 = −𝑘𝑠𝑝𝑟𝑖𝑛𝑔 ∙ 𝑥 = −𝑚𝜔02 𝑥
constant 𝑘𝑠𝑝𝑟𝑖𝑛𝑔 :

𝑘
Where 𝜔0 = is the bound electron natural or resonant oscillation frequency.
𝑚

‫להכניס הדמיה‬
 So we have: (1) 𝐹𝑟𝑒𝑠𝑡𝑜𝑟𝑖𝑛𝑔 = −𝑘𝑠𝑝𝑟𝑖𝑛𝑔 ∙ 𝑥 = −𝑚𝜔02 𝑥 and

In that simple model the damping force will be velocity (𝑥)

ሶ dependent, namely
(2) 𝐹𝑑𝑚𝑝𝑖𝑛𝑔 = −𝑚𝛾𝑥ሶ (the actual cause for damping is radiation by the oscillating electron. We will learn this).
𝑥
The deriving EMF is the harmonic: ഥ 𝒕 =𝑬
𝑬 ഥ 𝟎 𝑒 𝑖(2𝜋𝜆−𝜔𝑡)
ഥ 𝟎 𝑒 𝑖(𝑘𝑥−𝜔𝑡) = 𝑬
𝑥
However, since 𝜆𝐸𝑀𝐹 ≫ 𝑎𝑡𝑜𝑚 𝑠𝑖𝑧𝑒, 𝑖. 𝑒. ∆𝑥, the special wave component 2𝜋 can be neglected, leaving the EMF to be
𝜆
ഥ ഥ ഥ
just time dependent: 𝑬 = 𝑬 𝒕 = 𝑬𝟎 𝑒 −𝑖(𝜔𝑡)
and hence the driving force to be:

3 ഥ 𝒅𝒓𝒊𝒗𝒊𝒏𝒈 = 𝑒𝑬
𝑭 ഥ 𝒕 = 𝑒𝑬
ഥ 𝟎 𝑒 𝑖𝜔𝑡

ഥ 𝟎 the is amplitude of the EM wave. Putting all this into Newton’s second law yields:
Where 𝒆 is the electron charge and 𝑬
𝑑2𝑥
𝑚 = 𝐹𝑟𝑒𝑠𝑡𝑜𝑟𝑖𝑛𝑔 + 𝐹𝑑𝑎𝑚𝑝𝑖𝑛𝑔 + 𝐹𝑑𝑟𝑖𝑣𝑖𝑛𝑔 , 𝑎𝑛𝑑 𝑖𝑑𝑛𝑡𝑟𝑜𝑑𝑢𝑐𝑖𝑛𝑔 1, 2, 3 𝑎𝑛𝑑 𝑑𝑖𝑣𝑖𝑑𝑖𝑛𝑔 𝑏𝑦 𝑡ℎ𝑒 𝑚𝑎𝑠𝑠 𝒎 𝑤𝑒 𝑔𝑒𝑡:
𝑑𝑡

𝒆
𝒙ሷ + 𝟐𝜸𝒙ሶ + 𝝎𝟐𝟎 𝒙 = 𝑬ത 𝟎 𝒆𝒊𝝎𝒕 [𝟐𝟎]
𝒎
Where 𝒎 is the electron mass, and 𝜸 is the damping coefficient (the coefficient 2 was chosen for convenience). To satisfy
Eq. 20, 𝒙 must be a function whose second derivative is similar to itself. Furthermore, it is anticipated that the electron will
follow the deriving force and hence have the same frequency. So we guess that:
 𝑥 𝑡 = 𝐶𝑒 𝑖𝜔𝑡 [21] and introducing [21] into [20] gives:
𝑒
−𝜔2 + 2𝛾 𝑖𝜔 + 𝜔02 𝐶𝑒 𝑖𝜔𝑡 = 𝐸𝑒 𝑖𝜔𝑡 out of which we get that:
𝑚

𝑒𝐸/𝑚
𝐶 = 𝜔2 −𝜔2+2𝑖𝛾𝜔
0

ሚ
multiplying both the numerator and denominator by the latter yields the complex number 𝐶:
2
𝑒𝐸 /𝑚(𝜔 −𝜔 2 −2𝑖𝛾𝜔)
𝐶ሚ ≡ 𝐶 𝑅 + 𝑖𝐶 𝐼 = 𝑜 2 20 2 and hence [22]
(𝜔0 −𝜔 ) +4𝛾2 𝜔 2

𝐶ሚ = 𝐶ሚ 𝐶ሚ ∗ =
1
𝑒𝐸𝑜 /𝑚 2 (𝜔02
− 𝜔2 − 2𝑖𝛾𝜔)(𝜔02 − 𝜔2 + 2𝑖𝛾𝜔) 2 𝑒𝐸𝑜 /𝑚
= = = 𝐶ሚ 𝑒 ±𝑖𝜑
(𝜔02 − 𝜔 2 )2 +4𝛾 2 𝜔 2 2 (𝜔02 − 𝜔 2 )2 +4𝛾 2 𝜔 2 1/2

𝐶𝐼 −2𝛾𝜔
Where: 𝑡𝑎𝑛𝜑 = =
𝐶𝑅 𝜔02 −𝜔2
Introducing 𝐶ሚ 𝑒 ±𝑖𝜑 into [21], i.e. into 𝑥 𝑡 = 𝐶𝑒 𝑖𝜔𝑡 one gets [22]:

𝑒𝐸𝑜 /𝑚𝑒
𝑥 𝑡 = 𝐶ሚ 𝑒 ±𝑖𝜑 𝑒 𝑖𝜔𝑡 = 𝐶ሚ 𝑒 𝑖 𝜔𝑡±𝜑 = 2 1/2
𝑒 𝑖 𝜔𝑡±𝜑 [22]
(𝜔02 −𝜔2 )2 +4𝛾 2 𝜔
𝜔2
When 20 ≪ 1, elastic scattering occurs, i.e. no energy loss (from the EMF within the substance) happened.
𝜔
Since 𝜔02 − 𝜔2 >> 𝛾𝜔, 𝑡ℎ𝑒 𝑑𝑎𝑚𝑝𝑖𝑛𝑔 𝑒𝑙𝑒𝑚𝑒𝑛𝑡 𝑐𝑎𝑛 𝑏𝑒 𝑛𝑒𝑔𝑙𝑒𝑐𝑡𝑒𝑑. The closer 𝜔 is to 𝜔𝑜 the larger 𝑥 𝑡 is, and
absorption occurs.

Polarizability (𝛼), Electrical Susceptibility (𝜒𝑒 ) and the refractive index (𝑛) in LIH medium
The atomic electrical dipole moment 𝑝 is the product of its positive charge times the distance (displacement) 𝑥 between
the positive and negative charges. As said, when the dipole is subjected to varying EMF, both the displacement and the
dipole becomes time dependent:

𝑒 2 𝐸𝑜 /𝑚𝑒
𝑝(𝑡)
ҧ = 𝑒 ∙ 𝑥ҧ 𝑡 = 𝑒𝑖 𝜔𝑡±𝜑
(𝜔02 − 𝜔 2 )2 +4𝛾 2 𝜔 2 1/2

The polarizability is defined as the induced polarization per electric field, that is to say:

𝑝(𝑡) 𝑒 2 /𝑚𝑒
𝛼≡ =
𝐸(𝑡) (𝜔02 − 𝜔 2 )2 +4𝛾 2 𝜔 2 1/2
 In a simple LIH medium, for the sake of simplicity, we assume that the dipoles follow the electric field and the
overall susceptibility 𝜒𝑒 of a unit volume containing 𝑁 atoms would be 𝑁 times the atomic polarizability:

𝑁𝑝(𝑡) 𝑁𝑒 2 /𝑚𝑒 𝑃(𝑡)

𝜒𝑒 = 𝑁𝛼 = = =
𝐸(𝑡) (𝜔02 − 𝜔 2 )2 +4𝛾 2 𝜔 2 1/2 𝐸(𝑡)

Where 𝑃(𝑡) is the macroscopic electrical polarization of the medium.

𝑃(𝑡) 𝜀0 𝐸ത = 𝐷
ഥ − 𝑃ത = 𝜀𝐸ത − 𝑃ത ⟹
Next, remembering that = 𝜀 − 𝜀0 ⟹ 𝑃ത = (𝜀 − 𝜀0 )𝐸ത
𝐸(𝑡)

𝑃(𝑡) 𝑁𝑒 2 /𝑚𝑒 2
𝜀 𝑁𝑒 2
𝑝𝑒𝑟𝑚𝑖𝑡𝑡𝑖𝑣𝑖𝑡𝑦 𝜀 = 𝜀0 + = 𝜀0 + ⟹ 𝑛 = 𝜀𝑟 ≡ =1+ ⟹
𝐸(𝑡) (𝜔02 − 𝜔 2 )2 +4𝛾 2 𝜔 2 1/2 𝜀0 𝜀0 𝑚𝑒 (𝜔02 − 𝜔 2 )2 +4𝛾 2 𝜔 2 1/2

𝑁𝑒 2
𝑛= 1+
𝜀0 𝑚𝑒 (𝜔02 − 𝜔 2 )2 +4𝛾 2 𝜔 2 1/2
 𝑑2𝑥
𝑚 = 𝐹𝑟𝑒𝑠𝑡𝑜𝑟𝑖𝑛𝑔 + 𝐹𝑑𝑎𝑚𝑝𝑖𝑛𝑔 + 𝐹𝑑𝑟𝑖𝑣𝑖𝑛𝑔 , 𝑎𝑛𝑑 𝑖𝑑𝑛𝑡𝑟𝑜𝑑𝑢𝑐𝑖𝑛𝑔 1, 2, 3 𝑤𝑒 𝑔𝑒𝑡:
𝑑𝑡

𝒅𝟐 𝒙
𝒎 + 𝒎𝜸𝒙ሶ + 𝒎𝝎𝟐𝟎 𝒙 = 𝒒𝑬𝒆−𝒊𝝎𝒕 [𝟐𝟎]
𝒅𝒕
Dividing by 𝒎 we get:
𝒅𝟐 𝒙 𝒒
+ 𝜸𝒙ሶ + 𝝎𝟐𝟎 𝒙 = 𝑬𝒆−𝒊𝝎𝒕
𝒅𝒕 𝒎

𝑥 𝑡 = 𝑥0 𝑒 −𝑖(𝜔𝑡+𝛼) [21]
To evaluate the amplitude 𝑥0 , we introduce [21] into [20]:

2 𝑒 −𝑖(𝜔𝑡+𝛼) +𝜔2 𝑥 𝑒 −𝑖(𝜔𝑡+𝛼) 𝑞 𝑞

𝑥0 𝑖𝜔 0 0 = 𝐸𝑒 −𝑖𝜔𝑡 /dividing by 𝑒 −𝑖(𝜔𝑡+𝛼) one get: 𝑥0 𝜔02 − 𝜔2 = 𝐸𝑒 𝑖𝛼 ⟹
𝑚 𝑚

𝒒𝑬𝒆𝒊𝜶
𝒙𝟎 =
𝒎 𝝎𝟐𝟎 − 𝝎𝟐