Chapter 4

Electrostatic Fields in Matter

Problem 4.1

E = V/x = 500/10-3 = 5x 105. Table 4.1: a/47r€0 = 0.66x 10-30, so a = 47r(8.85x 10-12)(0.66x 10-30) =
7.34X10-41. p = aE = ed ~ d = aE/e = (7.34x 10-41)(5 x 105)/(1.6 x 10-19) = 2.29 X 10-16 m.
d/R = (2.29 x 10-16)/(0.5 x 10-10) = 14.6x 10-6.1 To ionize, say d = R. Then R = aE/e = aV/ex ~ V =
Rex/a.= (0.5 x 10-10)(1.6 x 10-19)(10-3)/(7.34 x 10-41) = 1108v.1
Problem 4.2

First find the field, at radius r, using Gauss' law: J E.da = E~Qenc,or E = 4;<0~Qenc.
r
a - a2
Qenc = l 0
r
pdT =-
47rq
7ra3 0 l r -
e-2r/ar2dr = -
a2 a2
4q
a3 [ 2
--e-2r/a
(r2 + ar +-
r
2 )]l 0
r2
2q
= - a2 [ e-2r/a (r2 + ar +"2 ) -"2 ] = q [ 1 - e-2r/a 1 + 2~ + 2 a2 ( )] .
[Note:Qenc(r --+ 00) =
q.] So the field of the electron cloud is Ee = 4;<0 [1 - e-2r/a(1+ 2~+ 2~)]. ~ The
protonwill be shifted from r =
0 to the point d where Ee = E (the external field):

1 q 2d/ a d ~
E=-- 1-e- 1+2-+2-.
47r€0d2 [ ( a a2 )]
Expandingin powers of (d/a):

(~a ) - ~3 (a~) +...

2d 2d - .!. 2d
= 1- !2 =1-
e-2d/a
() ()a
+
a
2
3! a ( ) +...
3 2~ + 2
a
2 3

d d2
(
1- e-2d/a 1 + 2- + 2-
a a2 ) = 1- (1-2~+2(~r -~(~r +..-) (1+2~+2~)
d d
= r - r - 2ta - 2:+
cP. cP. d3 cP. d3 4~
d2 + 2ta + 4:+
d2 + 4:+
d3 - 2:+ d3 + -3 -a3 + .. .
d2 - 4:+
3
4 d
= 3 ~ ( ) + higher order terms.
73
74 CHAPTER 4. ELECTROSTATIC FIELDS IN MATTER

1 q 4 d3
E = --
471"€0
dl- 3 a3 ( )
-- = --(qd)
1 3a3
471"€04 = -po 1.
371"€oa3 I
a = 311"!:oa3 . I

[Not so different from the uniform sphere model of Ex. 4.1 (see Eq. 4.2). Note that this result predicts
4;EOa = !a3 = ! (0.5 X 10-10)3 = 0.09 X 10-30 m3, compared with an experimental value (Table 4.1) of
0.66 x 10-30 m3. Ironically the "classical" formula (Eq. 4.2) is slightly closer to the empirical value.]
Problem 4.3

per) = Ar. Electricfield (by Gauss's Law): §E.da = E (471"r2)

= -!oQenc = EloJ;Ar471"r2
dr, or E =
~ €oA r4
471"r 471" 4 = Ar2
4€0 . This "internal" field balances the external field E when nucleus is "off-center" an amount
d: ad2/4€0 =E ~ d = V4€oE/A. So the induced dipole moment is p = ed = 2ev€0/AVE. Evidently
I p is proportional to El/2.1
For Eq. 4.1 to hold in the weak-field limit, E must be proportional to r, for small r, which means that p
must go to a constant (not zero) at the origin: I p(O) :/; 0 (nor infinite). I

Problem 4.4
. r
0
Field of q:
Q A
~ ~
1I"EO r
f. Induced dipole moment of atom: P =aE =
q 411"E:r2 r.
Field of this dipole, at location of q (0 = 71",
inEq. 3.103): E = _4 1 13
7I"€0r ( 2aq 2
471"€or ) (to the right).

= 2a -47I"€0 2 r13 I(attractive).

q
Force on q due to this field: IF
( )
Problem 4.5
Field of PI at P2 (0 = 71"/2in Eq. 3.103): E1 = 47I"€or
PI 39 (points down).

Field of P2 at PI (0 = 71" in Eq. 3.103): E2 = 47I"€or

P2 3 (-2f) (points to the right).

2PIP2 . .
Torque on PI: N1 = PI X E2 = I-47I"€or3 I(pomts mto the page).

Problem 4.6
(a)
Use image dipole as shown in Fig. (a). Redraw, placing Pi at the origin, Fig. (b).

E-- P (
. - 471"€0(2z)32cosOf+sinO9);
P = pcosOf + psinO9.

N = P X Ei = 471"€:(2Z )3 [(cos 0 f + sin 0 9) x (2cos0 f + sin0 9)]

(b) p2 A A

= 4r.€0(2z)3 [cosOsinO4J + 2sinOcosO(- 4J) ]

p2 sin 0 cos 0
10 ~o
A

= 471"€0(2z)3 (-4J) (out of the page).

Pif/ Z
 75

. p2 sin 20
But sin 0 cos 0 = (1/2) sm 20, so N I = 4m:o(16z3) (out of the page).
For0 < 0 < '!r/2, N tends to rotate p counterclockwise;for '!r/2 < 0 < '!r,N rotates p clockwise. Thus the
stableorientation is perpendicular to the surface-either t or ..t..
Problem 4,7
Say the field is uniform and points in the y direction. First slide p
y
in from infinity along the x axis-this takes no work, since F is J.. dl.
(If E is not uniform, slide p in along a trajectory J.. the field.) Now
tE rotate (counterclockwise) into final position. The torque exerted by
O.
P E is N = pxE =
pEsinOz. The torque we exert is N = pEsinO
p x clockwise, and dO is counterclockwise, so the net work done by us is
negative:
U = J:/2 pE sin OdO = pE (- cosO) 1~/2 = -pE (cosO- cos~) = -pE cos 0 = -p,E. Qed
Problem 4,8

U = -pI,E2, but E2 = ~-!:r [3(p2,f) f - P2]. SOU = ~-!:r [PI'P2- 3 (pI,f) (p2,f)]. Qed
Problem 4,9

(a) F = (p . V)E (Eq. 4.5); E =_

4'!rEOr~ r = _
1 q ~ q xx+yy+zz
4'!rEO(2x + y 2 + z 2)3/2'

8 8 8 q x
=
Fx
(Px - + P - + pz-
8x Y8y
1
8z ) -
4'!rEO(X2 + y2 + Z2)3/2
3 2x
.

q 3 2y
- 4'!rEO{ Px [ (x2 + y2 + Z2)3/2 - 2x (X2 + y2 + Z2)5/2 ] + py [ -2x (X2 + y2 + Z2)5/2 ]
3 2z q Px 3x q p 3r(p.r)
+ pz [ -2x (X2 + y2 + Z2)5/2]} = 4'!rEO[ r3 - ;:s(Pxx + Pyy + pzz) ] = 4'!rEOr3 [ - r5 ] x'
.
F = I4 1_ '!rEor~ [p - 3(p . f) f] .

(b) E =_ 1 r-;.{3[p.
4'!rEO (-f)]( -f) - p} =_ 1 r13[3(p.
4'!rEO f) f - p]. (This is from Eq. 3.104; the minus signs
are because r points toward p, in this problem.)

F = qE = 1-4'!rEO
1 rq3[3(p . f) f - p] ,

[Notethat the forces are equal and opposite, as you would expect from Newton's third law.]
Problem 4,10
~ 1 8 2 1 2 ~
(a) Ub = P,n = §]
kR; Pb = -V.p = -3"-
r 8r (r kr) = -~3kr
r =~
(b) For r < R, E = 3~oprf (Prob. 2.12), so E = I-(k/EO) r.1

For r > R, same as if all charge at center; but Qtot = (kR)(4'!rR2) + (-3k)(t'!rR3) = 0, so IE = 0.1
76 CHAPTER 4. ELECTROSTATIC FIELDS IN MATTER

Problem 4.11
Pb = 0; ab = P.il = :!:P (plus sign at one end-the one P points toward; minus sign at the other-the one
P points away from).

(i) L » a. Then the ends look like point charges, and the whole thing is like a physical dipole, Qflength Land
charge P-rra2. See Fig. (a).

(ii) L « a. Then it's like a circular parallel-plate capacitor. Field is nearly uniform inside; nonuniform "fringing
field" at the edges. See Fig. (b).

(iii) L ~ a. See Fig. (c).

p p p

(c)
(a) Like a dipole (b) Like a parallel-plate capacitor

Problem 4.12

J { J
v = 4';EO I;jdT = p. 4';EO ~dT }. But the term in curly brackets is precisely the field of a uniformly
charged sphere, divided by p. The integral was done explicitly in Prob. 2.7 and 2.8:

R3
I (4/3);R3Pf, (r>R), I 3€or2P.r=
A

I
R3 P cos B
3€or2 '
I ( > R)
r ,
1 .t. dT -- 1 411"€0 r So V(r,B) =

4W<O!
.' - p{ ~ R3 r, (r < R). }
411"€0(4/3)wR'p I ...!...P.r
3€0 ~ I" 'owoo, 1 (r < R).

Problem 4.13
Think of it as two cylinders of opposite uniform charge density :!:p. Inside, the field at a distance s from
the axis of a uniformly charge cylinder is given by Gauss's law: E211"se= -:OP1l"S2e :::} E = (p/2€0)s. For
two such cylinders, one plus and one minus, the net field (inside) is E = E+ + E- = (p/2fO) (s+ - s_). But
s+ - s- = -d, so E = l-pd/(2€0),where d is the vector from the negative axis to positive axis. In this case
I

the total dipole moment of a chunk of length e is P (1I"a2e)= (p7ra2e) d. So pd = P, and IE = -P /(2€0), I for
s < a.
 77

Outside, Gauss's law gives E27r8£ = .1...p7ra2£

<0 :::}E = 1!!£.
2a2!,
<0 s for one cylinder. For the combination, E =
a2 !:t. - iL
E+ + E- = 1!!£. 2<0 (s+
, where s- )
d
s:!: = S T -j
2
12 -1 -1
1 d s.d 1 d s.d
-S:i: = 2 a-
8?t ( d)(
ST-
2
8 +-Ts.d
4 ) ~-
82 ( )(
ST-
2
IT-
82 ) ~-
82 ( )(
ST-
2
1:f:-
82 )
1 (s . d) d . .

= (
82 S :f: S--;2 T "2) (keepmg only 1st order terms in d).

8+ - L = ~ - ~ - s-s~ =~
(s+ 8- ) 82 [(
s+s~
82 ) (
2 82
+ ~
2 )] 82 (2S(S.d) 82
-d
).
a2 1
E(s) = -- [2(P. 8) § - P] ' for 8 > a.
2fO 82

Problem 4.14

Total charge on the dielectric is Qtot = is O"b

da + Iv Pbdr = is P . da - Iv V.p dr. But the divergence
theorem says is p. da = Iv V.p dr, so Qenc = O. qed
Problem 4.15

(a)Pb=-v,p=-~~
r2 or ( r ) =-~; r2
r2~ O"b=P.ii= +P.~=k/b
{ -P . r = -k/a
(atr=b),
(at r = a). }
Gauss's law:::} E = 4:<0Q;~cr. For r < a, Qenc = 0, so IE = 0.1 For r > b, Qenc = 0 (Prob. 4.14), so E I = 0.1
Fora < r < b, Qenc = (~k) (47ra2)+ I: (~) 47rf2dr = -47rka - 47rk(r- a) = -47rkr; so E I = -(k/for) r.1

(b) fD.da = Qfenc = O:::}D = 0 everywhere. D = foE + P = O:::}E = (-l/fo)P, so

IE = 0 (for r < a and r > b);j IE = -(k/fOr) r (for a < r < b). I

Problem 4.16

(a) Same as Eo minus the field at the center of a sphere with uniform polarization P. The latter (Eq. 4.14)

is-P/3fO. SoIE = Eo + ~p.1 D = foE= foEo+ ~P = Do- P + ~P, so D = Do - ~p.1 I

(b) Same as Eo minus the field of:f: charges at the two ends of the "needle"-but these are small, and far
away,so ! E = Eo.J D = foE = foEo = Do - P, SO I D = Do - P .1

(c) Same as Eo minus the field of a parallel-plate capacitor with upper plate at 0" = P. The latter is

-(l/fo)P, so IE = Eo + !op.1 D = foE = foEo + P, so ID = Do.!

78 CHAPTER 4. ELECTROSTATIC FIELDS IN MATTER

Problem 4.17

P
E D
(uniform) (same as E outside, but lines
(field of two circular plates)
continuous, since V.D = 0)
Problem 4.18

(a) Apply D . da J =
Q/enc to the gaussian surface shown. DA = aA ::} I D = a.1 (Note:D = 0 insidethe
metal plate.) This is true in both slabs; D points down.

~ 2+u

(b) D = fE ::} E = a/fl in slab 1, E = a/f2 in slab 2. But f = tofT, so fl = 2fo; f2 = ~fO' lEI = a/2fo, I

I E2 = 2a/3fo.1
(c) P = foXeE, so P = foXed/(fOfr) = (Xe/fr)a; Xe = fr -I::} P = (1- f;l)a. IPl=a/2,IIP2=a/3.1
(d) V = E1a + E2a = (aa/6fo)(3 + 4) = 17aa/6fo,!
ab = +P1 at bottom of slab (1) = a/2, ab = + P2 at bottom of slab (2) = a /3,
(e)Pb=O;
ab= -PI at top of slab (1) = -a/2; ab = -P2 at top of slab (2) = -a /3.

total surface charge above: a - (a/2) = a/2, E - ~ ..(

(f) In slab 1: { total surface charge below: (a/2) - (a/3) + (a/3) - a = -a/2, } ==> 1 - 2fO'
total surface charge above: a - (a/2) + (a/2) - (a/3) = 2a/3, E - 2a ..(
In slab 2: { total surface charge below:(a/3) - a = -2a/3, } ==> 2 - 3fO'
]+u
-u/2
(!)
+u/2
-u/3
@
+u/3
]-u
Problem 4.19
With no dielectric, Co = Afo/d (Eq. 2.54).
In configuration (a), with +a on upper plate, -a on lower, D = a between the plates.
E = a/ fO(in air) and E = a / f (in dielectric). So V = {;; ~ + 7~ = 2~~ (1 + ~) .
- ~ - ~ 2
Ca - V - d 1+1/(r (
==> Ca )
Co - 1 2fr+ fr .
I I

In configuration (b), with potential difference V: E = V /d, so a = foE = fO V / d (in air).

 79

P = EOXeE= EOXeV/d (in dielectric), so O'b= -EOXeV/d (at top surface of dielectric).
O"tot = EoV/d= O'f + = O'f-
EoXeV/d,so O'f= EoV(l + Xe)/d = EOErV/d(on top plate above dielectric).
O'b
Q 1 A AV V AEO A
1 + Er
=? Cb = V = V (
+ O'f2" = 2V Eod + EOdEr = d ~
0'2" ) Cb = ~.
. Co 1 + Er ( ) ( ) I I

[Which is greater?. ~Co - ~Co = l+<r.-

2
~l+<r = {l+<r)2-4<r
2(1+<r)
= 1+2<r+4<~-4<r
2(1+<r)
= 2(1+<r)
(1-<r)2> 0. So Cb > Ca.]
If the x axis points down:
I ~ E D p ] O'b (top surface) ~ O'f (top plate) I
< - .2<r - 0 0 Y
(a) air «r+1) d x «rH) d X «r+l) d
(a) dielectric
2 v
«r+l) d X
- «rH) d
x
<r+l) d
-
X
-
«r+l) d
-
-
(b) air Yx
d d x 0 0 7 (left)
(b) dielectric Yx
d Er7x (Er - 1)7 X -(Er - 1)7 Er7 (right)

Problem 4.20

ID.da = Qfenc => D41fr2 = p!1fr3 => D = lpr => E = (pr/3E) r, for r < R; D41fr2 = p!1fR3 => D =
pR3/3r2=>E = (pR3/3Eor2)r, for r > R.

fO pR3 1 P fo pR2 P R2 pR2 1

V = - } E. dl = 00 3EO -:;. 00 - 3E1
R
}R rdr = 3EO + 3E ""2 = I 3EO ( + ).
1 2Er

Problem 4.21

Let Q be the charge on a length £ of the inner conductor.

f D . da = D21fs£ = Q => D
Q
_
= 21fS~0; E
Q
= 2-1fEOS~ 0 (a < s < b), E = -21fES~0 (b < r < c).
Q

a b C
b

C
V - -
f C
E.dl-
- I(a
-Q
21fEO£ ) l( )
dS
-+
S b
-Q
21fd
dS
---
S -
Q
21fEO£ [
In -
(a )
EO -
+-In ( E
C

b')]

£ - Q
V£ = I
In(b/a) + 21fEO
(l/Er) In(cjb)" I

Problem 4.22

Same method as Ex. 4.7: solve Laplace's equation for V;n(s, tj)) (s < a) and Vout(s, tj)) (s > a), subject to
the boundary conditions x
(i) V;n = Vout at s = a,
(ii) E8~n = EO8~~ut at s = a, Eot
{ (Hi) Vout -+ - Eos cos tj) for S » a.
y

FromProb. 3.23 (invoking boundary condition (Hi)):

00 00

V;n(s, tj)) = 2::>k(ak cosktj)+ bksinktj)), Vout(s, tj))= -Eoscostj) +

k=l
L
k=l
s-k(Ck cosk</> + dk sink</».
80 CHAPTER 4. ELECTROSTATIC FIELDS IN MATTER

(1 eliminated the constant terms by setting V = 0 on the y z plane.) Condition (i) says

I::ak(ak cosk</J+ bk sin k</J)= -Eoscos</J + L a-k(ck cosk</J + dk sink</J),

while (ii) says

f.r L kak-l(ak cosk</J + bk sink</J) = -Eo cos</J- L ka-k-l(ck cosk4> + dk sink</J).

Evidently bk = dk = 0 for all k, ak = Ck = 0 unless k = 1, whereasfor k = 1,
aal = -Eoa + a-ICI, f.ral = -Eo - a-2CI'

Solving for aI,

Eo Eo
-x
Eo scos</J=
al = - (1 + Xe/2)' so ~n(S, </J)= - (1 + Xe/2) (1 + Xe/2) ,

and hence Ein(s, </J)= -~ x =~2).1 As in the spherical case (Ex. 4.7), the field inside is uniform.
Problem 4.23

Po = f.oXeEo; EI = --31~ Po = --Xe3 Eo; PI = f.oXeEI = --f.OX~

3 Eo; E2 = --PI
~
1
= X~
-Eo;
9

Evidently En = (- ~e )n Eo, so

E=Eo+EI+E2+"'= [~(_~er] Eo.

The geometric series can be summed explicitly:

00 1
'"' n - -, so 1 Eo,
~x-l-x
n=O E = (1 + Xe/3)

which agrees with Eq. 4.49. [Curiously, this method formally requires that Xe < 3 (else the infinite series
diverges), yet the result is subject to no such restriction, since we can also get it by the method of Ex. 4.7.]
Problem 4.24
Potentials:
Vout(r,O) = -EorcosO+): r~lPI(COSO), (r > b);
Vrned(r,O)= L(Alrl+r~l)1't(cosO), (a < r < b);
{ ~n(r,O) = 0, (r < a).
Boundary Conditions:
(i) Vout = Vrned, (r = b)j
(ii) f.~ or = f.0~ or ' (r = b)j
{ (Hi) Vrned = 0, (r = a).
 81

" BI " I hi
(i) ~
.-Eob cos0 + ~ bl+l ~(cosO) = ~ (
Alb + bl+l ) ~(cosO)j

" I 1 hi "BI
(ii) ~ €r ~ [ lAlb - ]
- (l + 1)bl+2 ~(cosO) = -Eo cosO- ~(l + 1) bl+2 PI (cos 0);
(Hi) ~ A al + hi = 0 ~ h = -a2l+1 A
I al+l I I.
Fori =f.1 :

.
(i) BI
bl+1
=
( A bl - a21+1AI
I bl+l
2l+1
AI
) ~ B = A
I
BI
I
(b2l+1 - a21+1 ),
l
1-1 a - - 21+1 21+1 --
(ii)
[
€r lAlb + (l + 1) bl+2 ] - -(l + 1)bl+2 ~ Bl - -€rAI [( l +1 b ) +a ] ~ Al - Bl - O.
Fori = 1 :

B1 a3Al 3 3 3
(i) -Eob+b2=A1b-~ ~ BI-Eob =A12 (b -a );
.. a3 At B1 3 3 3
(ll
) €r ( Al + 2~ ) = -Eo - 2b3 ~ -2B1 - Eob = €rAI (b + 2a ) .
3 3 3 3 3 -3Eo
So -3Eob = Al
[2 (b - a
)
+ €r
(
b + 2a )] ; Al = 2[1 - (afb)3] + €r[1 + 2(afb)3].
-3Eo a3

Vrned(r,O) = 2[1- (afb)3] + €r[1 + 2(afb)3] ( -;:2 )

r cosO,

E(r,O) = -VVrned = 12[1- (afb)3] ~E:r[1 + 2(afb)3] { (1+ 2r~3)cosOi'- (1- ;:) sinoo}.
Problem 4.25

There are four charges involved: (i) q, (ii) polarization charge surrounding q, (Hi) surface charge (CTb)on
the top surface of the lower dielectric, (iv) surface charge (CT~)on the lower surface of the upper dielectric.
In view of Eq. 4.39, the bound charge (ii) is qp = -q(X~f(1 + X~), so the total (point) charge at (0,0, d) is
qt = q.+ qp = qf(1 + X~) = qf€~. As in Ex. 4.8,

-1 qdf€~ CTb CT~

(a) CTb = €oXe _ 3 -_
2 - - (here CTb
~

= P.n = +Pz = €oXeEz)j

[ 47I"€o(r2 + cF):2 €o 2€o]
CTb - _
(b)
I
CTb = €oXe 4
I
_ 1 qdf€~
- _
2 3
CT~
2 ( here CTb= -Pz = -€OXeEz )
I
.
[ 7I"€o(r2 + cF)2 €o €o]
Solvefor CTb,CT~:first divide by Xe and X~ (respectively) and subtract:

CT~ CTb 1 qdf€~ I I CTb 1 qdf€~

X~ - Xe = 271"
(r2 + cF)~ ~ CTb= Xe [ Xe + 271"
(r2 + cF)~ ] .
82 CHAPTER 4. ELECTROSTATIC FIELDS IN MATTER

Plug this into (a) and solve for O"b,using €~ = 1 + X~:

-1 qd/€~ ' O"b ( '

)
-1 qd Xe
O"b - - ( )
";!Xel+Xe--Xe+Xe,soO"b=- 3 , ;
41T (r2 + d2)2 2 41T(r2 + d2)2 1 + Xe + Xe)/]2
[ (

, - , -I qd 1 1 qd/€~ , 1 qd €rX~/€~
O"b-
Xe { 41T (r2 + d2) ~ [1 + (Xe + X~)/2] + 21T(r2 + d2) ~ } ' so O"b= 41T(r2 + d2) ~ [1 + (Xe + X~)/2]'

X x. ,
The total bound surface charge is O"t O"b + O"~ = 417r =
(r2+d2)
~- + )/ 2 (which vanishes, as it should, when
qd ~ E'~ l + ( ", x.
X~ = Xe)' The total bound charge is (compare Eq. 4.51):

= €~ - €r q
qt (X~ - Xe )q
2€~ [1 + (Xe + X~)/2] -
_
I ( €~ + €r ) €~' I and hence

V(r) =~ q/€~ + qt (for z I > 0). \

41T€0{ ";X2 + y2 + (z - d)2 ";x2 + y2 + (z + d)2 } I
I
I
I
q €~ - €r - ~ 1 [2q/(€~ + €r)] (for z < 0). I
Meanwhile, ~ + qt
since €r = 7"r [ 1 + €'r + €r ] - €~ + €r ' V(r) = 41T€0 ";X2 + y2 + (z d)2

Problem 4.26 \

From Ex. 4.5:

0, (r < a)
Q
-r,
~

D = O'Q (r < a) (a < r < b) .

{ 41Tr2r, (r > a) } , E= 41Ttt ~

(r > b) }
{ ~r,
41T€or

00
I 1 I -1 -1
w = -1
2 I 1 Q2
D.EdT=--41T
2 (41T)2 {€
-
l b

a r2 r2
1 1
--r2dr+-
1
€o b1
00
-dr
r2 }
=-Q2
81T { €
- -
( )r I
b

a
1
+--
€o ( )
r 1 b }
Q2 1 1 1 1Q2 1 Xe
+ Xe) ( ~ + b ) .
= 81T€0{ (1 + Xe) ( ~ - b) + b } = 81T€0(1
 83

Problem 4.27

Using Eq. 4.55: W = !f J E2 dr. From Ex. 4.2 and Eq. 3.103,

-I
(r < R)
E = 3102
P z, , so
{ R ~(2cosBf+sinBO),
3for (r > R) }
fa
Wr<R =
2 ( ) =
.£-. 2 ~11'R3
3100 3
2
211'P2R3.
27 fa
fa R3 P . 1 . 2 2 2
Wr>R = -
2 ( )J
- 6"
3100
(4cos B+ sm
1r
B)
r r smBdrdBd<jJ

= (R3 p)2
l
1 8fa 211' a (1 + 3 COS2
00

B) sin BdB R 4"

1
r dr 1 =-
11'(R3 p)2
9100 (-COSB-COS3B)I~(-3~3)1:

(~ ) =
11'(R3P)2 411'R3p2.
9100 3R3 27100

211'R3 p2
Wtot =
9100

This is the correct electrostatic energy of the configuration, but it is not the "total work necessary to assemble
the system," because it leaves out the mechanical energy involved i~ polarizing the molecules.

Using Eq. 4.58: W = ~ J D.E dr. For r <: R, D = foE, so this contribution is the same as before.
Forr < R, D = foE + P = -!p + P = jp = -2fOE, so ~D.E = -2!fE2, and this contributionis
now(-2) (~~ p:~3) =-
~~ R::2, exactly cancelling the exterior term. Conclusion: Wtot = 0.1 This is not I

surprising, since the derivation in Sect. 4.4.3 calculates the work done on the free charge, and in this problem
there is no free charge in sight. Since this is a nonlinear dielectric, however, the result cannot be interpreted as
the "work necessary to assemble the configuration" -the latter would depend entirely on how you assemble it.
Problem 4.28

First find the capacitance, as a function of h:

Air part: E = -bL

41rfOS
==> V = ..1L
41rfO'
In(bJa) >..
==> -fa = -j>'"10 >..'
10
= fa->..= lOr>".
OilPart: D = 41rs
~ ==>E = 41rfS 2>"In(bJa) ' }
2>" ==>V = 41rf

Q = >..'h + >"(f - h) = fr>"h - >"h+ >"f = >..[(fr - l)h + f) = >"(Xeh+ f), where f is the total height.

Q >"(Xeh+ f) (Xeh + f)
C =V = 2>"ln(bJa) = 211'100In(bJa)
411'100 .

Th t d l' .. b E 4 64 ' F - 1 V 2 dC - 1 V 2 21rfOXe V2

e ne upwar LOrCeIS gIven y q. . . - 2" dh - 2" In(bfa) . h = fOXe .
The gravitational force down is F = mg = p11'(b2 - a2)gh. } I p(b2 - a2)g In(bJa)
84 CHAPTER 4. ELECTROSTATIC FIELDS IN MATTER

Problem 4.29
8
(a) Eq. 4.5
.
:::} F2 = (P2 . V) EI = P2~uy (Ed;
. PI
Eq. 3.1O3:::} EI = 4m:or
~
A PI
() = -- 41rfoY3 z. Therefore
A
~y
?jr
= _ -PIP2
d 1
( Y )] z = _
F2 - A 3PIP2 A 3PIP2 A

I(upward).
4 d 1rfO [ y 3" 41rfoY4 Z, or I
F2 = ~z
41rfor
z
.
To calculate F I, put P2 at the origin, pointing in the z direction; then PI I

y
is at -r z, and it points in the -y direction. So FI = (PI' V) E2 =
I

- - z--r
-PI 8:2y x-y-O, - ; we need E2 as a function of x, y, and z.
1

ll 3(P2' r)r
From E q. 3.104: E2 = -3" ?
h
- P] , were
A A

r = xx + yy + ZZ, P2 = -P2Y, and hence

A A

41rfOr [ r-
P2 . r = -P2Y'

E2 = ~ -3Y(XX + yy + zz) + (x2 +y2 + Z2)y =~ -3XYX + (x2 - 2y2 + z2)y - 3YZZ
41rfO [ (x2 + y2 + z2)5/2 ] 41rfO [ (x2 + y2 + z2)5/2 ]

8E2 = ~ -~~2Y[-3XYX + (x2- 2y2+ Z2)y - 3yzz] + ~(-3xx - 4yy - 3ZZ) ;

8y ~~ { 2~ ~ }
8E2 = ~ -3z = -PI ~41rfO 3r Z =
8y 1
(0,0) 41rfO r5
z; FI
r5( ) - 3PIP2 Z.
41rfor4

These results are consistent with Newton's third law: FI = -F2.

(b) From page 165, N2 = (P2 x EI) + (r x F2). The first term was calculated in Frob. 4.5; the second we
get from (a), using r = r y:

= (ry-) x -
(_ ) = _ _
3PIP2 3PIP2 2PIP2
P2 X EI = 41rfor3
PIP2 (
A

-X)j
r x F2
41rfor 4 Z 41rfor 3 x; so
N2
A

I
= 41rfor 3 X.
A

This is equal and opposite to the torque on PI due to P2, with respect to the center of PI (see Frob. 4.5).
Problem 4.30
Net force is to the right (see diagram). Note that the field lines must bulge to the right, as shown, because
I I

E is perpendicular to the surface of each conductor.

E
 85

Problem 4.31

P = kr = k(xx + yy + zz) =? Pb= -V.p = -k(l + 1 + 1) = 1-3k.1

Total volume bound charge: I Qvol = -3ka3.1
(Jb = P.il. At top surface, il = z, z = a/2j so O"b
= ka/2. Clearly, I O"b = ka/21 on all six surfaces.
T~tal surface bound charge: I Qsurf= 6(ka/2)a2 = 3ka3.1 Total bound charge is zero. if
Problem 4.32

f D.da = Qfonc::}D = 4- q 7rr

2 rj
~

E = -D
10
1
=
q f
4 7rfO(1 + Xe) 2";
r P = fOXeE =
qXe
4 7r(1 + Xe) 2"'
r
f

Pb= -V.p = 47r(~~Xe) (V. )

~ = -q 1 ~eXe 83(r) (Eq. 1.99)j O"b P.f = = 47r(1~X~e)R2;

Qsurf= (Jb(47rR2) = q 1 +XeXe .1 The compensating negative charge is at the center:

I

j PbdT = - l qXe j 83(r)dT = -q-1 Xe .

+ Xe + Xe
Problem 4.33

Ell is continuous (Eq. 4.29); Dl. is continuous (Eq. 4.26, with O"f = 0). So EXl =-EX2' DYl = DY2 ::}
E1EYI = f2EY2'and hence

tan02 = EX2/EY2= EYl = E2. Qed

tan 01 EXl / EYl EY2 El

If 1 is air and 2 is dielectric, tan O2/ tan 01 = E2/ EO> 1, and the field lines bend away from the normal. This is
the opposite of light rays, so a convex "lens" would defocus the field lines.
Problem 4.34

In view of Eq. 4.39, the net dipole moment at the center is pi =P - 1~~e p = I';Xo p = tp. We want the
potentialproduced by pi (at the center) and O"b(at R). Use separation of variables:

00 B
Outside: V(r,O)= L rl:l
1=0
Pz(cosO) (Eq.3.72)

1
.
lnstde: V(r,O)= 4 _ 1 pcosO
7rEO Err
~
00
+ LAlrIPI(cosO)
1=0
(Eqs.3.66,3.102)
)
.

R~L = AIRI, or BI = R2/HAI (l ¥' 1)

V continuous at R ::} .
B1 - 1 P - p- 3
{ R2 - 47rEoErR2 + AIR, or B1 - 47rfOf~+ AIR }
8V av
8r R+I ar R- i
= - ""'
L)l
BI
+ 1)R l+2Pz(cosO)+ 47rEO _ 1 2pcosO
Er R3 -
""'
LlAIR
I 1
- PI(cosO) = --O"b
EO
1

av
= --p.
1
r
~

= --1 (EoXeE.r) = Xe _
a
~

= Xe { --47rEO
I 2pcosO ""' I 1
- .
EO EO r I R- lOrR3 + ~lAIR PI(cosO) }
86 CHAPTER 4. ELECTROSTATIC FIELDS IN MATTER

-(I + 1) R~~2 - lAIRI-I = XelAIRI-I (l ¥- 1); or - (2l + l)AIRI-I = XelAIRI-I =>Al = 0 (£ ¥- 1).
BI 1 2p 1 2P P AIR3 1 XeP AIR3
Forl=l: -2-+---AI=Xe
R3 411"fO frR3 ( ---+A1
411"fOfrR3 ) -BI+---=---+Xe-;
411"fOfr 2 411"fo fr 2

-~ - AIR3 + ~ - AIR3 = _~XeP + XeAIR3 => AIR3 (3 + Xe) = ~XeP.

411"fOfr 411"fOfr 2 411"fOfr 2 2 411"fOfr

=> Al = ~ 2XeP = ~ 2(fr -l)p;

411"foR3fr(3 + Xe) 411"foRafr(fr + 2)
BI = ~
411"fOfr [
1 + 2(fr - 1)
(fr + 2) ]
= ~~.
411"fOfrfr + 2

)(~ + 2)
p COSO
V(r,O)=
(
411"for2 fr
(r ~ R).

Meanwhile, for r::; R, V(r, 0) = ~P cosO + 1 prcos(} 2(fr -1)

411"fo frr2 R3
411"fO freEr + 2)

P cos () fr - 1 r3
= I 41I"for 2 fr [ 1 + 2 -fr + 2
( ) ] (r::;
R3 R).

Problem 4.35
Given two solutions, VI (and EI = -VVI, DI = fEd and V2 (E2 = -VV2, D2 = fE2), define V3 ==V2- VI
(E3 = E2 - EI' Da = D2 - DI).
Iv V.(VaD3) dr = Is V3Da' da = 0, (Va= ° on S), so I(VV3) . D3 dr + I V3(V.D3) dr = 0.
But V.D3 = V.D2 - V.DI = PI- PI = 0, and VV3 = VV2 - VVI = -E2 + EI = -E3, so IE3' D3 dr = O.
I
But D3 = D2 - DI = fE2 - EEl = fE3, so f(E3)2 dr = 0. But f > 0, so E3 = 0, so V2 - VI = constant. But
at surface, V2 = VI, so V2 = VI everywhere. qed
Problem 4.36

(a) Proposed potential: I VCr) = Vo~.1 If so, then IE = - VV = VO~ f, in which case P I = fOXe Vo ~ f,
;
in the region z < 0. (P = o for z > 0, of course.) Then CTb= fOXeVo (f.ft) = 1- fO~ Vo .1 (Note: ft points out
of dielectric => ft = -f.) This CTbis on the surface at r = R. The flat surface z = 0 carries no bound charge,
since ft = z 1..f. Nor is there any volume bound charge (Eq. 4.39). If V is to have the required spherical
symmetry, the net charge must be uniform:
CTtot411"R2 = Qtot = 411"fORVo(since Vo = Qtot/411"foR), so CTtot= foVo/R. Therefore

- (fOVo/ R), on northern hemisphere

CT
I - { (fOVol R)(l + Xe), on southern hemisphere' }

(b) By construction, CTtot CTb+CTI = =

foVo/R is uniform (on the northern hemisphere CTb= 0, CTI foVo/R; =
on the southern hemisphere CTb= -foXeVo/R, so CTI= fVo/R). The potential of a uniformly charged sphere is

Vo = Qtot = CTtot(411"R2) = fOVo R2 = VoR. ./

411"for 411"fOr R for r
(c) Since everything is consistent, and the boundary conditions (V = Vo at r = R, V -+ 0 at 00) are met,
Prob. 4.35 guarantees that this is the solution.
 87

(d) Figure (b) works the same way, but Fig. (a) does not: on the flat surface, P is not perpendicular to ft,
so we'd get bound charge on this surface, spoiling the symmetry.
Problem 4.37

~
Eext = 27r€08 8. Since the sphere is tiny, this is essentially constant, Xe / 3 Eext (Ex. 4.7).
and hence P = 1 +€oXe
2
F -
- J (1 + Xe!3
€oXe
) (~
) ( ) (
27r€08
~ ~
d8 27r€08
8dr -
-
€oXe
1 + Xe/3 )( ) (! ) ( ) J
~
2no 8
-1
82
8 dr

(~ ) 2-~7rR38 - - (~ )
= -Xe )..2R3 8
1 + Xe/3 47r2€0 833 - 3 + Xe 7r€083 .

Problem 4.38

The density of atoms is N = (4/3)7rR3'The macroscopic field E is Eself + Eelse, where Eself is the average
fieldover the sphere due to the atom itself.

p = o:Eelse =} P = No:Eelse.

[Actually,it is the field at the center, not the average over the sphere, that belongs here, but the two are in
fact equal, as we found in Prob. 3.41d.] Now
1 p
Eself = - 47r€0R3
(Eq. 3.105), so
1 0: 0: NO:
E=- 47r€0 R3 Eelse + Eelse = ( 1- 47r€oR3 ) Eelse = (
1- 3€0 ) Eelse.
So
P= No:
(1 - N o:/3€0) E = €oXeE,
and hence
No:/€o
Xe = (1 - No:/3€0)'
Solving for a:
No: No: No: Xe
Xe - -3 Xe= - =} - (1+ _3 ) = Xe,
€o €o €o
or
€o Xe 3€0 Xe 3€0 €r - 1
a = N (1 + Xe!3) = Ii (3 + Xe' But Xe = €r - 1, so 0:= Ii (z+2 ) . qed
Problem 4.39

Foran ideal gas, N = Avagadro's number/22.4liters = (6.02 x 1023)/(22.4x 10-3) = 2.7 X 1025. No:/€o =
(2.7x 1O25)(47r€0x 1O-3O),8/€0= 3.4 X 10-4,8, where,8 is the number listed in Table 4.1.
H: (3 = 0.667, No:/€o = (3.4 x 10-4)(0.67) = 2.3 x 10-4, Xe = 2.5 X 10-4
He: (3= 0.205, No:/€o = (3.4 x 10-4)(0.21) = 7.1 x 10-5, Xe = 6.5 X 10-5 ..
Ne: (3 = 0.396, No:/€o= (3.4x 10-4)(0.40)= 1.4x 10-4, Xe= 1.3X10-4 agreementISqUItegood.
Ar: (3 = 1.64, N 0:/ €o = (3.4 X 10-4)(1.64) = 5.6 x 10-4, Xe = 5.2 X 10-4 }
88 CHAPTER 4. ELECTROSTATIC FIELDS IN MATTER

Problem 4.40
J~:E ue-u/kT du - (kT)2e-u/kT [-(u/kT) - l]I~~E
(a) (u) - pE
J-pEPE e-u/kT du - -kTe-U/kT l -pE

= kT [e-pE/kT - ePE/kT] + [(pE/kT)e-pE/kT + (pE/kT)ePE/kT]

{
e-pE/kT - epE/kT
}
ePE/kT + e-PE/kT PE
= kT - pE [ epE/kT - e-pE/kT ] = kT - pE coth ( kT ) .
A - - -(u) PE kT
P = N(p); p = (pcosO)E = (p. E)(E/E) = -(u)(E/E); P = Np pE = I Np { coth (
kT ) - pE } .
Lety ==P/Np, x ==pE/kT. Theny = cothx-1/x.
Asx --+0, y = (~+ f - ~; +... )-~ = f-~; +... ~
0, so the graph starts at the origin, with an initial slope of 1/3. As x --+00, y --+coth(oo) = 1, so the graph
goes asymptotically to y = 1 (see Figure).
.E...
np'
11 """""""""""""'" .

pe/kT

(b) For small x, y :::::::

kx, so;; p :::::::
-f!-r, or P :::::::
~E = €oXeE => P is proportional to E, and Xe = ~~Np2 .
For water at 20° = 293 K , p = 6.1 X 10-30 em' 'volume
N = molecules= molecules
mole X molesX
gram !\rams.
volume
1 6- 0 33 1029. - (O.33Xl029)(6.1Xl0-30)2 - j"1;)l .
N -- (6 .0 X 1023) X ( 18) x-. (10 ) X , Xe - (3)(8.85xl0-12)(1.38XlO-23)(293)
- ~
12 T bl 4 2
a e . gives an
experimental value of 79, so it's pretty far off.
For water vapor at 100° = 373 K, treated as an ideal gas, v~~r::e= (22.4 X 10-3) X (~~~) = 2.85 X 10-2 m3.

(2.11 x 1025)(6.1x 10-30)2 -

N = 6.0 X 1023 ,
2.85 X 10-2 = 2.11 X 1025. Xe = (3)(8.85 x 10-12)(1.38x 10-23)(373)= 15.7x 10 3.1

Table 4.2 gives 5.9 x 10-3, so this time the agreement is quite good.