Electric eld and charge at the corner or edge

Masahiro Yamamoto
This manuscript is modied on February 21, 2009 1 : 17 pm

For simplicity Poisson equation is considered here, i.e. no charge except at the interface between the
electrode and the medium.
(

=
2

2
2
+
x2 y 2

=0

(1)

In the present work we will consider the potential and charge distribution at an edge or corner of the
two conducting(metal) plate. We just follow the discussion done by J. D. Jacksons book on Classical
Electrodynamics, 3rd edition, Chap. 2, Section 11. In the Fig.1 the corner between the two conducting
plate is shown.

Figure 1:

It is convenient to use cylindrical coordinate r, , z. In the z direction the potential and charges are the
same, then we will consider the r, problem.

r

=
+
y
y r y

x
x
r =
x2 + y 2 , cos = = 2
r
x + y2
1 2
x
r
y
r
=
[x + y 2 ]1/2 2x = = cos ,
= = sin
x
2
r
y
r
cos
x
sin
y

r

+
,
x r x

(2)
(3)
(4)

cos

1
1 1
1
x2
= sin
= +x
2x
=
1

x
x
r
2 r3
r
r2
(

=
=

sin2
,
r

sin
=
x
r

sin

1
1 1
1
y2
cos2

cos
= cos
= +y
2y
=
1

=
,
=
3
2
y
y
r
2 r
r
r
r
y
r
r

sin

cos
+
= cos

,
=
+
= sin
+
x r x
r
r
y
y r y
r
r
1

(5)
(6)
(7)

2
x2

2
y 2

2
2
+
x2 y 2

)(

sin

sin
2
cos sin
cos sin 2
=
cos

cos

= cos2 2 +

r
r
r
r
r
r2

r
r
2
2
2
2
sin
sin cos
sin cos
sin
+

+
+
(8)
r r
r
r
r2

r2 2
(
)(
)

cos

cos
2
cos sin
cos sin 2
=
sin
+
sin
+
= sin2 2
+
r
r
r
r
r
r2

r
r
2
2
2
2
sin cos
sin cos
cos
cos
+

+
(9)
+
2
r r
r
r
r

r2 2
2
1
1 2
=
+
+
(10)
r2 r r r2 2
(

(r, ) =
2

1
2
1 2
+
+
r2 r r r2 2

(r, ) = 0

(11)

The radial direction er and the angular directon erd as shown in Fig.1 is perpendicular, then the gradient
of the potential can be dened
= er

1
+ erd
r
r

(12)

If we can separate the solution in r and variables,

(r, ) = R(r)()
(

2 (r, ) =
r2
R

2 R 1 R
+
r2
r r

2
1 2
1
+
+
r2 r r r2 2

(r, ) =

2 R 1 R
+
r2
r r

1 2
= 2
2

(13)

R 2
=0
r2 2

(14)
(15)

The general solutions for = 0 are

R(r) = ar + br

(16)

() = c cos() + d sin()

(17)

R(r) = + ln r

(18)

() = +

(19)

and for = 0

From the boundary conditions, i.e. (r, = 0) = V, (r, = 0 ) = V

c = 0, [sin 0 = 0, sin(0 ) = 0],

=0

(20)

r = 0 is included, then
b = 0,

=0

(21)

and
sin(0 ) = 0,

n
,
0

n = 1, 2, 3, ....

(22)

Then we have
(r, ) = V +

an rn/0 sin(n/0 )

(23)

n=1

Near r = 0 the n = 1 term is important, then we can write

(r, ) V + a1 r/0 sin(/0 )
2

(24)

The electric eld near r = 0 is given by

1
E = = er
erd
r
r
/0 1
Er a1 r
sin(/0 )
0

Erd = a1 r/0 1 cos(/0 )

0

(25)
(26)
(27)

From the Gauss theorem at the surface, we can write with the unit vector n form metal to vacuum
n [Dvac Dmetal ] = metal|vac ,

Dmetal = 0

The surface charge density at = 0 and = 0 are equal and can be approximated
(r) = Dvac = 0 Erd = a1

0 /0 1
r
0

(28)

When 0 is small, the power of r becomes large and no charge accumulation at the corner r = 0. For
at surface 0 = , there is no dependence on r. When 0 > the power /0 1 of r becomes negative,
and at the edge the surface charge density become singular.

3
2

7
4

Figure 2: coner and edge with 0 = /4, /2, 3/2, 7/4.

Table 1: r sigularity at the coner or edge

0
0
/4
/2

3/2
7/4
2

/0 1
+
3
1
0
-1/3
-3/7
-1/2

r/0 1
r3
r
1
r1/3
r3/7
r1/2