PHY103AA
PHYSICS II - Electrodynamics
Lecture 5_A_1
Displacement Vector 𝐷
Y N Mohapatra
Deptt. of Physics
Materials Science Programme
National Centre for Flexible Electronics
Samtel Centre for Display Technologies
IIT Kanpur
24 April 2021 YNM/IITK/PHY103AA 1
Motivation:
Introducing a new Character - The 𝐷 Vector
We can control the free charge or potential that we can put on a body or material.
A conductor responds by finding a charge distribution that makes the electric field inside it zero.
In the process, 𝐸 outside gets changed accordingly.
For a dielectric, we have seen that it responds by producing bound charges either on the surface or
in volume of the body.
The net field is determined by both ‘free’ charges and ‘bound charges’
+- + - Bound charges depend on the properties of the dielectric and its shape as pe boundary conditions.
+ -
+- + -
+ -
+- If I can’t control 𝐸 , can I get a vector that is primarily due to FREE charges ?
+ -
+ -
In presence of a dielectric, how can I estimate
Polarization 𝑃 ,
Bound Charges 𝜎𝑏 𝑎𝑛𝑑 𝜌𝑏
& hence Electric Field 𝐸 . 2
Gauss’s Law in presence of Dielectric
How to deal with bound charges?
In a dielectric we may have both free charges (e.g. electrons, ions), and also bound charges due to Polarization.
Both determine E. Total charge density is where 𝜌𝑓 is free charge density,
𝜌 = 𝜌𝑓 + 𝜌𝑏 and 𝜌𝑏 is bound charge density.
Now, define the displacement vector as
In that case, Gauss’ Law for Electric Field is 𝜀0 ∇ ∙ 𝐸 = 𝜌𝑓 + 𝜌𝑏 𝐷 ≡ 𝜀0 𝐸 + 𝑃
Rewrite in terms of Polarization: Differntial Form of Gauss’s Law is then
𝜀0 ∇ ∙ 𝐸 = 𝜌𝑓 − ∇ ∙ 𝑃
∇ ∙ 𝐷 = 𝜌𝑓
Collecting Divergence Terms together Integral form of Gauss’s Law is then
(for convenience)
𝜀0 ∇ ∙ 𝐸 + ∇ ∙ 𝑃 = 𝜌𝑓
ර 𝐷 ∙ 𝑑𝑎 = 𝑄𝑓 𝑒𝑛𝑐
Combining Divergence Terms: ∇ ∙ 𝜀0 𝐸 + 𝑃 = 𝜌𝑓
D has units of surface charge as C/m2 .
𝐷 vector is invaluable in presence of dielectrics and bound charges.
Why is it so named? Answer in the next half when we discuss Ampere’s Law.
Dielectric Materials
What property measures how ‘polarizable’ they are?
• 𝑃 = 𝜒𝑒 𝜀0 𝐸 𝑃 produced in the material is a function of 𝐸 at the location.
• This is called the Constitutive Relation.
• 𝜒𝑒 is a dimensionless material property called the Electric Susceptibility of the medium.
• If 𝑃 ∝ 𝐸 i.e. 𝜒𝑒 is a constant, then the material is called a linear Dielectric P
𝐷 = 𝜀0 𝐸 + 𝑃 = 𝜀0 𝐸 + 𝜒𝑒 𝜀0 𝐸
E
= 1 + 𝜒𝑒 𝜀0 𝐸
𝐷 = 𝜀𝑟 𝜀0 𝐸 = 𝜀 𝐸
P vs E for linear dielectrics
For common semiconductors (such as Si, GaAs) 𝜀𝑟 ≅10-12,
For SiO2 𝜀𝑟 ≅ 4 (typical of good insulators)
Diamond: ≅ 5.6 𝜀 = 1 + 𝜒𝑒 𝜀0
𝜀
Water: ≅ 80.1 𝜀𝑟 ≡ 1 + 𝜒𝑒 =
𝜀0
If 𝜒𝑒 is itself a function of 𝐸 , then it is called a non-linear dielectric as in Ferroelectrics, Piezoelectric,
Pyroelectric materials and other non-linear crystals. In this course we deal with linear dielectric only.
Ex.5A-1 : 𝐷, 𝑃, 𝐸 for a Parallel Plate capacitor with a dielectric
+ - -
Assume charge on the plates are given i.e. 𝜎𝑓 =Q/A, and +
𝜎𝑓 + -
the relative dielectric constant is 𝜀𝑟 . - −𝜎𝑓
+ - +
+ -
+ - -
𝐷 = 𝜀0 𝐸 + 𝑃 = 1 + 𝜒𝑒 𝜀0 𝐸 +
+
-
• Source of D are free charges. D ර 𝐷 ∙ 𝑑𝑎 = 𝑄𝑓 𝑒𝑛𝑐
• Source of P are bound charges
P 1
• Source of E are both free and bound charges. 𝑃 = 1− 𝐷
𝜀𝑟
• In this case the direction of D and P are the same.
x
1 𝜖0 E
𝜎𝑏 = 1 − 𝜎𝑓 𝜌𝑏 = −∇ ∙ 𝑃 = 0
𝜀𝑟
𝜀0 𝐸 = 𝐷 − 𝑃
Argue on physical grounds why capacitance is larger with
dielectric inside. Larger the dielectric constant larger is the
x
capacitance.
Next Example : A Coaxial Cable
Oliver Heaviside FRS[1]
(18 May 1850 – 3 February 1925)
Example 5A.2: Capacitance of a long Coaxial cable
A coaxial cable consists of a copper core of radius a, surrounded by a c
concentric copper tube of inner radius c. The space in-between is partially b
filled with material of dielectric constant 𝜖𝑟 , as shown in the Figure.
a
Find the capacitance per unit length.
Strategy : Put a free charge Q on the inner cylinder.
Obtain D first, symmetry allows use of Gauss Law
From there obtain E everywhere,
and thus potential in terms of Q. 𝑄
For 𝑎 < 𝑠 < 𝑏 𝐸=
𝑄 2𝜋𝜀0 𝑠𝑙
ර 𝐷 ∙ 𝑑𝑎 = 𝐷 2𝜋𝑠 𝑙 = 𝑄 𝐷= 𝑄
2𝜋𝑠𝑙 For b< 𝑠 < 𝑐 𝐸=
2𝜋 𝜀𝑠𝑙
𝑏 𝑐
𝑎 Notice how D alone let us
𝑄 𝑄
𝑉 = − න 𝐸 ∙ 𝑑𝑙 = න 𝑑𝑠 + න 𝑑𝑠 obtain E in all regions !
𝑐 2𝜋𝜀0 𝑠𝑙 2𝜋 𝜀𝑠𝑙
𝑎 𝑏
𝑄 𝑏 𝜀0 𝑐 𝐶 𝑄 2𝜋𝜀0
= 𝑙𝑛 + 𝑙𝑛 = =
2𝜋𝜀0 𝑙 𝑎 𝜀 𝑏 𝑙 𝑉𝑙 𝑙𝑛 𝑏 + 1 𝑙𝑛 𝑐
𝑎 𝜀𝑟 𝑏
PHY103AA
PHYSICS II - Electrodynamics
Lecture 5_A_2
Dielectrics : Boundary Conditions
Y N Mohapatra
Deptt. of Physics
Materials Science Programme
National Centre for Flexible Electronics
Samtel Centre for Display Technologies
IIT Kanpur
24 April 2021 YNM/IITK/PHY103AA 8
Dielectric Boundary Conditions : Electrostatics
∇ ∙ 𝐷 = 𝜌𝑓 ර 𝐷 ∙ 𝑑𝑎 = 𝑄𝑓,𝑒𝑛𝑐
⊥ ⊥
𝐷𝑎𝑏𝑜𝑣𝑒 − 𝐷𝑏𝑒𝑙𝑜𝑤 = 𝜎𝑓
∥ ∥ ∥ ∥
𝐷𝑎𝑏𝑜𝑣𝑒 − 𝐷𝑏𝑒𝑙𝑜𝑤 = 𝑃𝑎𝑏𝑜𝑣𝑒 − 𝑃𝑏𝑒𝑙𝑜𝑤
⊥ ⊥
𝜎𝑓 + 𝜎𝑏
𝐸𝑎𝑏𝑜𝑣𝑒 − 𝐸𝑏𝑒𝑙𝑜𝑤 =
𝜀0
Just as we argued in the case of E earlier
∥ ∥
𝐸𝑎𝑏𝑜𝑣𝑒 − 𝐸𝑏𝑒𝑙𝑜𝑤 =0
Obvious, but very useful to remember:
At any interface in absence of free charge, the normal component of D is continuous even if there are bound charges there.
𝐸 𝑎𝑛𝑑 𝐷 are Different : Be careful
∇ ∙ 𝐷 = 𝜌𝑓 Similarity can be deceptive -→ 𝜀0 ∇ ∙ 𝐸 = 𝜌
∇×𝐸 =0
• Though both obey similar looking Gauss’s Law, there is NO
Coulomb’s Law for 𝐷. But, NOT 𝐷
∇×𝐷 ≠0
• 𝐷 is NOT conservative, hence there is NO corresponding potential.
∇ × 𝐷 = 𝜀0 ∇ × 𝐸 + ∇ × 𝑃 = ∇ × 𝑃 ≠ 0
Therefore, you cannot assume that D is determined exclusively by free
charges. A discontinuity in P can also give rise to D in absence of free
charges. (Think of a cylindrical bar with ‘frozen-in’ Polarization) 𝑃
• For computing 𝑫. , First look for symmetry. If you can use Gauss’s Law go ahead.
• If symmetry is absent, look for other means such as Boundary Conditions.
𝐸 and 𝐷 are parallel in Linear Dielectrics
𝜒𝑒 𝜒𝑒
𝜌𝑏 = −∇ ∙ 𝑃=−∇ ∙ 𝜀0 𝐷 =− 𝜌𝑓 Shielding Effect
+
𝜀 1+𝜒𝑒
• In a homogenous, isotropic linear dielectric, the bound charge density 𝜌𝑏 is -
proportional to free charge density 𝜌𝑓 . (This does not apply to surface bound
charge, because it is of course at the boundary.) + - +q - +
• Unless free charge is embedded in the material, 𝜌 = 0, any net charge must
-
reside at the surface.
• Within such a dielectric, then the potential obeys Laplace’s Equation.
+
• If a free charge is embedded in a large dielectric, its effect becomes weaker, because P shields the charge.
• In general, the dielectric responds to cancel the effect of the external field as much as possible, like an imperfect
conductor. (Example : the field due to polarization is against the external field in parallel pale capacitor.)
Example 5A.3 : Boundary Condition at interface of two Dielectrics
At the interface of two dielectrics (if there are no free charges), let 𝜃 𝑏𝑒 𝑡ℎ𝑒 𝑎𝑛𝑔𝑙𝑒 𝑏𝑒𝑡𝑤𝑒𝑒𝑛 𝐸
and normal to the interface,
then show that 𝒌𝟏 𝐜𝐨𝐭 𝜽𝟏 = 𝒌𝟐 𝐜𝐨𝐭 𝜽𝟐 , 𝑤ℎ𝑒𝑟𝑒 𝑘1 and 𝑘1 are the respective dielectric constants.
Assumed No free charges.
We know : 𝐷1 ∙ 𝑛ො1 = 𝐷2 ∙ 𝑛ො 2 𝐷1 cos 𝜃1 = 𝐷2 cos 𝜃2
𝜃1 𝐸1
𝐸1 × 𝑛ො1 = 𝐸2 × 𝑛ො 2 𝐸1 sin 𝜃1 = 𝐸2 sin 𝜃2
𝐷1 = 𝑘1 𝐸1 𝐷2 = 𝑘2 𝐸2 𝐸2
𝜃2
𝒌𝟏 𝐜𝐨𝐭 𝜽𝟏 = 𝒌𝟐 𝐜𝐨𝐭 𝜽𝟐
Is Dielectric Lensing of Electric Field Possible? To Focus E, would we need convex or concave geometries ?
Example 5A.4 : Point Charge on a huge slab of Dielectric
A point charge q is placed at a height d above large linear dielectric slab (for all z<0) with susceptibility
𝜒𝑒 . Calculate the force on the point charge.
z
Since it is a homogeneous linear dielectric slab, no volume bound charge i.e.𝜌𝑏 = 0.
q
𝑃 ∙ 𝑛ො = 𝜎𝑏 = 𝑃𝑧 = 𝜀0 𝜒𝑒 𝐸𝑧 Where 𝐸𝑧 is the vertical component at a point
just inside the slab at z=0 at a point r. d 𝜃
r y
This field is combination of both that due to q and the bound charge at that point.
The contribution due to q is
1 𝑞 1 𝑞𝑑 x
− 2 2
cos 𝜃 = −
4𝜋𝜀0 𝑟 + 𝑑 4𝜋𝜀0 𝑑2 + 𝑟 2 3/2
𝜎
But, the contribution due to induced charge at that location is − 2𝜀𝑏 .
0
Therefore 1 𝑞𝑑 𝜎𝑏
𝜎𝑏 = 𝜀0 𝜒𝑒 − −
4𝜋𝜀0 2 2 32 2𝜀0
𝑑 +𝑟
Solving the above for 𝜎𝑏 : One can obtain 𝐸 from direct integration of this
1 𝜒𝑒 𝑞𝑑
𝜎𝑏 = − charge density. But image method provides a
2𝜋 𝜒𝑒 + 2 𝑑2 + 𝑟 2 3/2
shorter route to get V, 𝐸 and Force.
Example 5A.4 : (Contd.) Use of IMAGE Method : Point Charge on a huge slab of Dielectric z
A point charge q is placed at a height d above large linear dielectric slab (for all z<0) with q
susceptibility
𝜒𝑒 . Calculate the force on the point charge. d𝜃
One can integrate the bound charge distribution to obtain the total bound charge: r y
2𝜋 ∞
𝜒𝑒 1 𝜒𝑒 𝑞𝑑 x
𝑞𝑏 = න 𝜎𝑏 𝑑𝑎 = − 𝑞 ඵ − 𝑟𝑑𝑟𝑑𝜑
𝜒𝑒 + 2 2𝜋 𝜒𝑒 + 2 𝑑2 + 𝑟 2 3/2
00
1 𝑞 𝑞𝑏
For solution z >0, replace the dielectric by 𝑞𝑏 at (0,0,-d) : 𝑉 = 4𝜋𝜀 +
0 𝑥2 + 𝑦2 + 𝑧−𝑑 2 𝑥2 + 𝑦2 + 𝑧 + 𝑑 2
But, for z<0, place 𝑞 + 𝑞𝑏 at (0,0,d) : 1 𝑞 + 𝑞𝑏
𝑉=
4𝜋𝜀0 𝑥2 + 𝑦2 + 𝑧 − 𝑑 2
Taken together both regions provide a complete solution
𝜕𝑉 𝜕𝑉 1 𝜒𝑒 𝑞𝑑
since boundary conditions match and we have not −𝜀0 ቤ − ቤ =−
violated the charge distribution in one region while 𝜕𝑧 𝑧=0+ 𝜕𝑧 𝑧=0− 2𝜋 𝜒𝑒 + 2 𝑑2 + 𝑟 2 3/2
calculating for the other one.
The attractive Force is then obtained easily as :
1 𝜒𝑒 𝑞2 Proof of the ‘pudding’ ?
𝐹Ԧ = − 2
𝑘 It is consistent, & it works !
2𝜋 𝜒𝑒 + 2 4𝑑
Example 5A.5 : Point charge +q embedded at the center of linear dielectric sphere
of radius R and susceptibility 𝜒𝑒 .
Find the electric field, the polarization, and all bound charges.
Strategy : Proceed from D→ E→ P and then bound charges and total charges !
𝑞
ර 𝐷 ∙ 𝑑𝑎 = 𝑄𝑓,𝑒𝑛𝑐 = 𝑞 𝐷=
4𝜋𝑟 2
𝑞
𝐷= 𝑟Ƹ 𝐷 𝑞
4𝜋𝑟 2 For 𝑟 < 𝑅 ∶ 𝐸= = 𝑟Ƹ
𝜖 4𝜋𝜀0 1 + 𝜒𝑒 𝑟 2
𝜒𝑒 𝑞 𝜒𝑒 𝑞
𝐹𝑜𝑟 𝑟 < 𝑅: 𝑃 = 𝜀0 𝜒𝑒 𝐸 = 𝑟Ƹ 𝜎𝑏 = 𝑃 ∙ 𝑟Ƹ =
4𝜋 1+𝜒𝑒 𝑟 2 4𝜋 1 + 𝜒𝑒 𝑅2
1 𝜒𝑒 𝑞 𝑟Ƹ 𝜒𝑒 𝜒𝑒
𝜌𝑏 = −∇ ∙ 𝑃 = − ∇ ∙ 2 = −𝑞 𝛿 3 𝑟Ԧ Total surface charge : 𝜎𝑏 4𝜋𝑅 2 = 𝑞
4𝜋 1 + 𝜒𝑒 𝑟 1 + 𝜒𝑒 1+𝜒𝑒
Its +ive.
Where is the compensating –ive charge?
𝜒𝑒 3
𝜒𝑒
න 𝜌𝑏 𝑑𝜏 = −𝑞 න 𝛿 𝑟Ԧ 𝑑𝜏 = −𝑞
1 + 𝜒𝑒 1 + 𝜒𝑒 That is the compensating –ive charge at the centre.
