You are on page 1of 140

8

-2

-4

-6

-8 -8 -6 -4 -2 0 rotating dipole: t = 2 2 4 6 8

Contents
1 Maxwells electrodynamics 1.1 Dynamical variables of electromagnetism 1.2 Maxwells equations and the Lorentz force 1.3 Conservation of charge 1.4 Conservation of energy 1.5 Conservation of momentum 1.6 Junction conditions 1.7 Potentials 1.8 Greens function for Poissons equation 1.9 Greens function for the wave equation 1.10 Summary 1.11 Problems 2 Electrostatics 2.1 Equations of electrostatics 2.2 Point charge 2.3 Dipole 2.4 Boundary-value problems: Greens theorem 2.5 Laplaces equation in spherical coordinates 2.6 Greens function in the absence of boundaries 2.7 Dirichlet Greens function for a spherical inner boundary 2.8 Point charge outside a grounded, spherical conductor 2.9 Ring of charge outside a grounded, spherical conductor 2.10 Multipole expansion of the electric eld 2.11 Multipolar elds 2.12 Problems 3 Magnetostatics 3.1 Equations of magnetostatics 3.2 Circular current loop 3.3 Spinning charged sphere 3.4 Multipole expansion of the magnetic eld 3.5 Problems 4 Electromagnetic waves in matter 4.1 Macroscopic form of Maxwells equations 4.1.1 Microscopic and macroscopic quantities 4.1.2 Macroscopic smoothing 4.1.3 Macroscopic averaging of the charge density 4.1.4 Macroscopic averaging of the current density 4.1.5 Summary macroscopic Maxwell equations 4.2 Maxwells equations in the frequency domain 4.3 Dielectric constant i 1 1 3 4 4 6 7 10 12 14 16 17 21 21 21 22 24 28 30 34 35 38 40 42 45 47 47 47 49 51 53 55 55 55 56 57 60 61 62 63

ii 4.4 4.5 Propagation of plane, monochromatic waves Propagation of wave packets 4.5.1 Description of wave packets 4.5.2 Propagation without dispersion 4.5.3 Propagation with dispersion 4.5.4 Gaussian wave packet Problems

Contents 65 68 68 69 69 71 73 75 75 76 78 78 79 81 83 84 84 85 85 87 88 90 90 91 94 95 96 98 99 100 101 102 103 103 105 109 109 112 114 115 116 120 121 122 122 123 124 126 127 131 132 135

4.6

5 Electromagnetic radiation from slowly moving sources 5.1 Equations of electrodynamics 5.2 Plane waves 5.3 Spherical waves the oscillating dipole 5.3.1 Plane versus spherical waves 5.3.2 Potentials of an oscillating dipole 5.3.3 Wave-zone elds of an oscillating dipole 5.3.4 Poynting vector of an oscillating dipole 5.3.5 Summary oscillating dipole 5.4 Electric dipole radiation 5.4.1 Slow-motion approximation; near and wave zones 5.4.2 Scalar potential 5.4.3 Vector potential 5.4.4 Wave-zone elds 5.4.5 Energy radiated 5.4.6 Summary electric dipole radiation 5.5 Centre-fed linear antenna 5.6 Classical atom 5.7 Magnetic-dipole and electric-quadrupole radiation 5.7.1 Wave-zone elds 5.7.2 Charge-conservation identities 5.7.3 Vector potential in the wave zone 5.7.4 Radiated power (magnetic-dipole radiation) 5.7.5 Radiated power (electric-quadrupole radiation) 5.7.6 Total radiated power 5.7.7 Angular integrations 5.8 Pulsar spin-down 5.9 Problems 6 Electrodynamics of point charges 6.1 Lorentz transformations 6.2 Fields of a uniformly moving charge 6.3 Fields of an arbitrarily moving charge 6.3.1 Potentials 6.3.2 Fields 6.3.3 Uniform motion 6.3.4 Summary 6.4 Radiation from an accelerated charge 6.4.1 Angular prole of radiated power 6.4.2 Slow motion: Larmor formula 6.4.3 Linear motion 6.4.4 Circular motion 6.4.5 Synchrotron radiation 6.4.6 Extremely relativistic motion 6.5 Radiation reaction 6.6 Problems

Chapter 1 Maxwells electrodynamics


1.1 Dynamical variables of electromagnetism

The classical theory of the electromagnetic eld, as formulated by Maxwell, involves the vector elds E (t, x) electric eld at position x and time t and B (t, x) magnetic eld at position x and time t. (1.1.2) We have two vectors at each position of space and at each moment of time. The dynamical system is therefore much more complicated than in mechanics, in which there is a nite number of degrees of freedom. Here the number of degrees of freedom is innite. The electric and magnetic elds are produced by charges and currents. In a classical theory these are best described in terms of a uid picture in which the charge and current distributions are imagined to be continuous (and not made of pointlike charge carriers). Although this is not a true picture of reality, these continuous distributions t naturally within a classical treatment of electrodynamics. This will be our point of view here, but we shall see that the formalism is robust enough to also allow for a description in terms of point particles. An element of charge is a macroscopically small (but microscopically large) portion of matter that contains a net charge. An element of charge is located at position x and has a volume dV . It moves with a velocity v that depends on time and on position. The volume of a charge element must be suciently large that it contains a macroscopic number of elementary charges, but suciently small that the density of charge within the volume is uniform to a high degree of accuracy. In the mathematical description, an element of charge at position x is idealized as the point x itself. The quantities that describe the charge and current distributions are (t, x) v (t, x) j (t, x) density of charge at position x and time t, (1.1.3) velocity of an element of charge at position x and time t,(1.1.4) current density at position x and time t. (1.1.5) (1.1.1)

We shall now establish the important relation j = v . The current density j is dened by the statement j da current crossing an element of area da, 1 (1.1.7) (1.1.6)

Maxwells electrodynamics

(v t )cos
Figure 1.1: Point charges hitting a small surface. so that j is the current per unit area. The current owing across a surface S is then j da.

On the other hand, current is dened as the quantity of charge crossing a surface per unit time. Let us then calculate the current associated with a distribution of charge with density and velocity v . Figure 1.1 shows a number of charged particles approaching with speed v a small surface of area a; there is an angle between the direction of the velocity vector v and the normal n to the surface. The particles within the dashed box will all hit the surface within a time t. The volume of this box is (v t cos )a = v n ta. The current crossing the surface is then calculated as current = = = (charge crossing the surface)/t volume of box 1 (total charge) total volume t (density)(volume of box)/t

= (v n ta)/t = v (n a) = v a . Comparing this with the expression given previously, current = j x, we nd that indeed, v is the current density. As was mentioned previously, the uid description still allows for the existence of point charges. For these, the density is zero everywhere except at the position of a charge, where it is innite. This situation can be described by a -function. Suppose that we have a point charge q at a position r . Its charge density can be written as (x) = q (x r ), where (x) (x) (y ) (z ) is a three-dimensional -function. If the charge is moving with velocity v , then j (x) = q v (x r )

1.2

Maxwells equations and the Lorentz force

is the current density. More generally, let us have a collection of point charges qA at positions rA (t), moving with velocities vA (t) = drA /dt. Then the charge and current densities of the charge distribution are given by (t, x) =
A

qA x rA (t)

(1.1.8)

and j (t, x) =
A

qA vA (t) x rA (t) .

(1.1.9)

Each qA is the integral of (t, x) over a volume VA that encloses this charge but no other: qA =
VA

(t, x) d3 x,

(1.1.10)

where d3 x dxdydz . The total charge of the distribution is obtained by integrating the density over all space: Q (t, x) d3 x =
A

qA .

(1.1.11)

1.2

Maxwells equations and the Lorentz force

The four Maxwell equations determine the electromagnetic eld once the charge and current distributions are specied. They are given by E B E B 1 , 0 = 0, B , = t = = 0 j + 0 0 E . t (1.2.1) (1.2.2) (1.2.3) (1.2.4)

Here, 0 and 0 are constants, and = (x , y , z ) is the gradient operator of vector calculus (with the obvious notation x f = f /x for any function f ). The Maxwell equations state that the electric eld is produced by charges and time-varying magnetic elds, while the magnetic eld is produced by currents and time-varying electric elds. Maxwells equations can also be presented in integral form, by invoking the Gauss and Stokes theorems of vector calculus. The Lorentz-force law determines the motion of the charges once the electromagnetic eld is specied. Let f (t, x) force density at position x and time t, (1.2.5)

where the force density is dened to be the net force acting on an element of charge at x divided by the volume of this charge element. The statement of the Lorentzforce law is then f = E + j B = E + v B . (1.2.6) The net force F acting on a volume V of the charge distribution is the integral of the force density over this volume: F (t, V ) =
V

f (t, x) d3 x.

(1.2.7)

Maxwells electrodynamics

For a single point charge we have = q (x r ), j = q v (x r ), and the total force becomes F = q E (r ) + v B (r ) . Here the elds are evaluated at the charges position. This is the usual expression for the Lorentz force, but the denition of Eq. (1.2.6) is more general. Taken together, the Maxwell equations and the Lorentz-force law determine the behaviour of the charge and current distributions, and the evolution of the electric and magnetic elds. Those ve equations summarize the complete theory of classical electrodynamics. Every conceivable phenomenon involving electromagnetism can be predicted from them.

1.3

Conservation of charge

One of the most fundamental consequences of Maxwells equations is that charge is locally conserved: charge can move around but it cannot be destroyed nor created. Consider a volume V bounded by a two-dimensional surface S . Charge conservation means that the rate of decrease of charge within V must be equal to the total current owing across S : d dt d3 x =
V S

j da.

(1.3.1)

Local charge conservation means that this statement must be true for any volume V , however small or large. We may use Gauss theorem to turn Eq. (1.3.1) into a dierential statement. For any smooth vector eld u within V we have u d3 x = u da.

The right-hand side of Eq. (1.3.1) can thus be written as V j d3 x, while the left-hand side is equal to V (/t) d3 x. Equality of the two sides for arbitrary volumes V implies + j = 0. (1.3.2) t This is the dierential statement of local charge conservation, and we would like to prove that this comes as a consequence of Maxwells equations. To establish this we rst dierentiate Eq. (1.2.1) with respect to time to obtain E = 0 , t t where we have exchanged the order with which we take derivatives of the electric eld. If we now use Eq. (1.2.4) to eliminate the electric eld, we get 1 = ( B ) j . t 0 The rst term on the right-hand side vanishes identically (the divergence of a curl is always zero) and we arrive at Eq. (1.3.2). We have therefore established that local charge conservation is a consequence of Maxwells equations.

1.4

Conservation of energy

[The material presented in this section is also covered in Sec. 6.7 of Jacksons text.]

1.4

Conservation of energy

Other conservation statements follow from Maxwells equations and the Lorentzforce law. In this section we formulate and derive a statement of energy conservation. In the next section we will consider the conservation of linear momentum. It is also possible to prove conservation of angular momentum, but we shall not pursue this here. Conservation of energy is one of the most fundamental principle of physics, and Maxwells electrodynamics must be compatible with it. A statement of local energy conservation can be patterned after our previous statement of charge conservation. Let (t, x) S (t, x) da electromagnetic eld energy density, eld energy crossing an element of surface da per unit time. (1.4.1) (1.4.2)

So is analogous to charge density, and S (which is known as Poyntings vector), is analogous to current density. If electromagnetic eld energy were locally conserved, we would write the statement d d3 x = S da, dt V S which is analogous to Eq. (1.3.1). But we should not expect eld energy to be conserved, and this statement is not correct. The reason is that the eld does work on the charge distribution, and this takes energy away from the eld. This energy goes to the charge distribution, and total energy is conserved. A correct statement of energy conservation would therefore be (eld energy leaving V per unit time) = (eld energy crossing S per unit time) + (work done on charges within V per unit time). To gure out what the work term should be, consider an element of charge within V . It moves with a velocity v and the net force acting on it is f dV = (E + v B ) dV. While the element undergoes a displacement dx, the force does a quantity of work equal to f dx dV . The work done per unit time is then f v dV = E v dV = j E dV . Integrating this over V gives the total work done on the charges contained in V , per unit time. The correct statement of energy conservation must therefore have the form d d3 x = S da + j E d3 x. (1.4.3) dt V S V In dierential form, this is + S = j E . t (1.4.4)

This last equation must be derivable from Maxwells equations, which have not yet been involved. And indeed, the derivation should provide expressions for the quantities and S . What we have at this stage is an educated guess for a correct statement of energy conservation, but Eq. (1.4.4) has not yet been derived nor the the quantities and S properly dened in terms of eld variables. Now the hard work begins. We start with the right-hand side of Eq. (1.4.4) and eliminate j in favour of eld variables using Eq. (1.2.4). This gives j E = E 1 E ( B ) + 0 E . 0 t

Maxwells electrodynamics

We use a vector-calculus identity to replace E (B ) with B (E )(E B ), and in this we replace E with B /t, using Eq. (1.2.3). All of this gives us j E = = B E 1 1 B + 0 E + (E B ) 0 t t 0 1 1 1 0 E 2 + B2 + EB . t 2 20 0

We have arrived at an equation of the same form as Eq. (1.4.4), and this shows that we do indeed have energy conservation as a consequence of Maxwells equations and the Lorentz-force law. The calculation also gives us denitions for the elds energy density, 1 1 0 E 2 + B2, 2 20 (1.4.5)

and for the elds energy ux (Poynting) vector, S 1 E B. 0 (1.4.6)

It is an important consequence of Maxwells theory that the electromagnetic eld carries its own energy.

1.5

Conservation of momentum

[The material presented in this section is also covered in Sec. 6.7 of Jacksons text.] We should expect a statement of momentum conservation to take a form similar to our previous statement of energy conservation. In that case we had a scalar quantity representing the density of energy, a vectorial quantity Sa representing the ux of energy, and the conservation statement took the form of + a Sa = work term. t We have introduced an explicit component notation for vectors, and summation over a repeated index is understood. For momentum conservation we will need a vectorial quantity a to represent the density of momentum, and a tensorial quantity Tab to represent the ux of momentum (one index for the momentum component, another index for the ux direction). So let a (t, x) Tab (t, x) dab density of a-component of eld momentum, a-component of eld momentum crossing an element of surface da per unit time. (1.5.1) (1.5.2)

The minus sign in front of Tab is introduced by convention. The statement of momentum conservation is (eld momentum leaving V per unit time) = (eld momentum crossing S per unit time) + (rate at which momentum is communicated to the charges). Because a rate of change of momentum is a force, we have the integral statement d dt a d 3 x = Tab dab +
S V

fa d3 x,

(1.5.3)

or the equivalent dierential statement a + b Tab = fa = Ea + (j B )a . t (1.5.4)

1.6

Junction conditions

We now would like to derive this from Maxwells equations, and discover the identities of the vector a and the tensor Tab . We rst use Eqs. (1.2.1) and (1.2.4) to eliminate and j from Eq. (1.5.4): f = (0 E )E + E 1 B B 0 0 t 1 1 ( B )B + ( B ) B = 0 ( E )E + 0 0 B 0 (E B ) + 0 E , t t

where to go from the rst to the second line we have inserted a term involving B = 0 and allowed the time derivative to operate on E B . The next step is to use Eq. (1.2.3) to replace B /t with E . Collecting terms, this gives f = 0 1 (E B ) + 0 ( E )E E ( E ) + ( B )B B ( B ) . t 0

1 To proceed we invoke the vector-calculus identity 2 (u u) = (u )u + u ( u), where u stands for any vector eld. We use this to clean up the quantities within square brackets. For example,

1 ( E )E E ( E ) = ( E )E + (E )E (E E ). 2 In components, the right-hand side reads 1 (b Eb )Ea + (Eb b )Ea a E 2 2 1 = Ea b Eb + Eb b Ea ab b E 2 2 1 2 = b Ea Eb ab E . 2

Doing the same for the bracketed terms involving the magnetic eld, we arrive at fa = 0 1 (E B )a + 0 b Ea Eb ab E 2 t 2 + 1 1 b Ba Bb ab B 2 . 0 2

This has the form of Eq. (1.5.4) and we conclude that momentum conservation does indeed follow from Maxwells equations and the Lorentz-force law. Our calculation also tells us that the elds momentum density is given by a 0 (E B )a , and that the elds momentum ux tensor is 1 Tab = 0 Ea Eb ab E 2 2 + 1 1 Ba Bb ab B 2 . 0 2 (1.5.6) (1.5.5)

Notice that the momentum density is proportional to the Poynting vector: = 0 0 S . This means that the elds momentum points in the same direction as the ow of energy, and that these quantities are related by a constant factor of 0 0 .

1.6

Junction conditions

[The material presented in this section is also covered in Sec. I.5 of Jacksons text.] Maxwells equations determine how the electric and magnetic elds must be joined at an interface. The interface might be just a mathematical boundary (in

Maxwells electrodynamics

which case nothing special should happen), or it might support a surface layer of charge and/or current. Our task in this section is to formulate these junction conditions. To begin, we keep things simple by supposing that the interface is the x-y plane at z = 0 a nice, at surface. If the interface supports a charge distribution, then the charge density must have the form (t, x) = + (t, x)(z ) + (t, x)(z ) + (t, x, y ) (z ), (1.6.1)

where + is the charge density above the interface (in the region z > 0), the charge density below the interface (in the region z < 0), and is the surface density of charge (charge per unit surface area) supported by the interface. We have introduced the Heaviside step function (z ), which is equal to 1 if z > 0 and 0 otherwise, and the Dirac -function, which is such that f (z ) (z ) dz = f (0) for any smooth function f (z ). These distributions (also known as generalized functions) are related by the identities d (z ) = (z ), dz d (z ) = (z ). dz (1.6.2)

As for any distributional identity, these relations can be established by integrating both sides against a test function f (z ). (Test functions are required to be smooth and to fall o suciently rapidly as z .) For example,

d(z ) f (z ) dz dz

= =

(z )

df (z ) dz dz

df
0

= f (0), and this shows that d(z )/dz is indeed distributionally equal to (z ). You can show similarly that f (z ) (z ) = f (0) (z ) and f (z ) (z ) = f (0) (z ) f (0) (z ) are valid distributional identities (a prime indicates dierentiation with respect to z ). If the interface at z = 0 also supports a current distribution, then the current density must have the form j (t, x) = j+ (t, x)(z ) + j (t, x)(z ) + K (t, x, y ) (z ), (1.6.3)

where j+ is the current density above the interface, j the current density below the interface, and K is the surface density of current (charge per unit time and unit length) supported by the interface. We now would like to determine how E and B behave across the interface. We can express the elds as E (t, x) B (t, x) = E+ (t, x)(z ) + E (t, x)(z ), = B+ (t, x)(z ) + B (t, x)(z ),

(1.6.4)

in an obvious notation; for example, E+ is the electric eld above the interface. We will see that these elds are solutions to Maxwells equations with sources given by Eqs. (1.6.1) and (1.6.3). To prove this we need simply substitute Eq. (1.6.4) into Eqs. (1.2.1)(1.2.4). Dierentiating the coecients of (z ) is straightforward, but we must also dierentiate the step functions. For this we use Eq. (1.6.2) and write (z ) = (z )z ,

1.6

Junction conditions

where z is a unit vector that points in the z direction. A straightforward calculation gives E = = = = E+ (z ) + E (z ) + z E+ E (z ),

B+ (z ) + B (z ) + z B+ B (z ).

B+ (z ) + B (z ) + z B+ B (z ),

E+ (z ) + E (z ) + z E+ E (z ),

Substituting Eq. (1.6.1) and the expression for E into the rst of Maxwells equations, E = /0 , reveals that E is produced by and that the surface charge distribution creates a discontinuity in the normal component of the electric eld: 1 (1.6.5) z E+ E z=0 = . 0 Substituting the expression for B into the second of Maxwells equations, B = 0, reveals that B satises this equation on both sides of the interface and that the normal component of the magnetic eld must be continuous: z B+ B
z =0

= 0.

(1.6.6)

Substituting Eq. (1.6.4) and the expression for E into the third of Maxwells equations, E = B /t, reveals that E and B satisfy this equation on both sides of the interface and that the tangential (x and y ) components of the electric must also be continuous: z E+ E
z =0

= 0.

(1.6.7)

Finally, substituting Eqs. (1.6.3), (1.6.4), and the expression for B into the fourth of Maxwells equations, B = 0 j + 0 0 E /t, reveals that B is produced by j and that the surface current creates a discontinuity in the tangential components of the magnetic eld: z B+ B
z =0

= 0 K .

(1.6.8)

Equations (1.6.5)(1.6.8) tell us how the elds E and B are to be joined at a planar interface. It is not dicult to generalize this discussion to an interface of arbitrary shape. All that is required is to change the argument of the step and functions from z to s(x), where s is the distance from the interface in the direction normal to the interface; this is positive above the interface and negative below the interface. Then s = n , the interfaces unit normal, replaces z in the preceding equations. If E+ , B+ are the elds just above the interface, and E , B the elds just below the interface, then the general junction conditions are n E+ E n E+ E n B+ B = = 1 , 0 0, (1.6.9) (1.6.10) (1.6.11) (1.6.12)

= 0, = 0 K .

n B+ B

Here, is the surface charge density on the general interface, and K is the surface current density. It is understood that the normal vector points from the minus side to the plus side of the interface.

10

Maxwells electrodynamics

1.7

Potentials

Suppose that B is expressed in this form. Then the other inhomogeneous Maxwell equation, E + B /t = 0, can be cast in the form E+ A t = 0.

[The material presented in this section is also covered in Secs. 6.2 and 6.3 of Jacksons text.] The Maxwell equation B = 0 and the mathematical identity ( A) = 0 imply that the magnetic eld can be expressed as the curl of a vector potential A(t, x): B = A. (1.7.1)

This, together with the mathematical identity () = 0, imply that E + A/t can be expressed as the divergence of a scalar potential (t, x): E= A ; t (1.7.2)

the minus sign in front of is conventional. Introducing the scalar and vector potentials eliminates two of the four Maxwell equations. The remaining two equations will give us equations to be satised by the potentials. Solving these is often much simpler than solving the original equations. An issue that arises is whether the scalar and vector potentials are unique: While E and B can both be obtained uniquely from and A, is the converse true? The answer is in the negative the potentials are not unique. Consider the following gauge transformation of the potentials: A new = Anew f , t = A + f, (1.7.3) (1.7.4)

in which f (t, x) is an arbitrary function of space and time. We wish to show that this transformation leaves the elds unchanged: Enew = E , Bnew = B . (1.7.5)

This implies that the potentials are not unique: they can be redened at will by a gauge transformation. The new electric eld is given by Enew Anew new t f = A + f t t f = E f + ; t t =

the last two terms cancel out and we have established the invariance of the electric eld. On the other hand, the new magnetic eld is given by Bnew = Anew = A + f

= B + (f );

the last term is identically zero and we have established the invariance of the magnetic eld. We will use the gauge invariance of the electromagnetic eld to simplify the equations to be satised by the scalar and vector potentials.

1.7

Potentials To derive these we start with Eq. (1.2.1) and cast it in the form A 1 =E = 0 t = A 2 , t

11

or

1 2 A + 0 0 , = 0 0 2 2 0 t t t

(1.7.6)

where, for reasons that will become clear, we have added and subtracted a term 0 0 2 /t2 . Similarly, we re-express Eq. (1.2.4) in terms of the potentials: 0 j = B 0 0 E t

A t t 2A , = ( A) 2 A + 0 0 2 + 0 0 t t = ( A) 0 0 where we have used a vector-calculus identity to eliminate ( A). We have obtained 2A . (1.7.7) 0 j = 0 0 2 2 A + A + 0 0 t t Equations (1.7.6) and (1.7.7) govern the behaviour of the potentials; they are equivalent to the two Maxwell equations that remain after imposing Eqs. (1.7.1) and (1.7.2). These equations would be much simplied if we could demand that the potentials satisfy the supplementary condition A + 0 0 = 0. t (1.7.8)

This is known as the Lorenz gauge condition, and as we shall see below, it can always be imposed by redening the potentials according to Eqs. (1.7.3) and (1.7.4) with a specic choice of function f (t, x). When Eq. (1.7.8) holds we observe that the equations for and A decouple from one another, and that they both take the form of a wave equation: 1 2 + 2 (t, x) c2 t2 1 2 2 2 + 2 A(t, x) c t where c 1 0 0 = 1 (t, x), 0 (1.7.9) (1.7.10)

= 0 j (t, x),

(1.7.11)

is the speed with which the waves propagate; this is numerically equal to the speed of light in vacuum. To see that the Lorenz gauge condition can always be imposed, imagine that we are given potentials old and Aold that do not satisfy the gauge condition. We know that we can transform them according to old = old f /t and Aold A = Aold + f , and we ask whether it is possible to nd a function f (t, x) such that the new potentials and A will satisfy Eq. (1.7.8). The answer is in the armative: If Eq. (1.7.8) is true then 0 = A + 0 0 t

12

Maxwells electrodynamics f old t t 2 f 0 0 2 + 2 f, t

= Aold + f + 0 0 = Aold + 0 0 old t

and we see that the Lorenz gauge condition is satised if f (t, x) is a solution to the wave equation old 1 2 + 2 f = Aold 0 0 . 2 2 c t t

Such an equation always admits a solution, and we conclude that the Lorenz gauge condition can always be imposed. (Notice that we do not need to solve the wave equation for f ; all we need to know is that solutions exist.) To summarize, we have found that the original set of four Maxwell equations for the elds E and B can be reduced to the two wave equations (1.7.9) and (1.7.10) for the potentials and A, supplemented by the Lorenz gauge condition of Eq. (1.7.8). Once solutions to these equations have been found, the elds can be constructed with Eqs. (1.7.1) and (1.7.2). Introducing the potentials has therefore dramatically reduced the complexity of the equations, and correspondingly increased the ease of nding solutions. In time-independent situations, the elds and potentials no longer depend on t, and the foregoing equations reduce to 2 (x) = 1 (x), 0 2 A(x) = 0 j (x), (1.7.12) (1.7.13)

as well as E = and B = A. The Lorenz gauge condition reduces to A = 0 (also known as the Coulomb gauge condition), and the potentials now satisfy Poissons equation.

1.8

Greens function for Poissons equation

[The material presented in this section is also covered in Sec. 1.7 of Jacksons text.] We have seen in the preceding section that the task of solving Maxwells equations can be reduced to the simpler task of solving two wave equations. For timeindependent situations, the wave equation becomes Poissons equation. In this and the next section we will develop some of the mathematical tools needed to solve these equations. We begin in this section with the mathematical problem of solving a generic Poisson equation of the form 2 (x) = 4f (x), (1.8.1)

where (x) is the potential, f (x) a prescribed source function, and where a factor of 4 was inserted for later convenience. To construct the general solution to this equation we shall rst nd a Greens function G(x, x ) that satises a specialized form of Poissons equation: 2 G(x, x ) = 4 (x x ), (1.8.2)

where (x x ) is a three-dimensional Dirac -function; the source point x is arbitrary. It is easy to see that if we have such a Greens function at our disposal, then the general solution to Eq. (1.8.1) can be expressed as (x) = 0 (x) + G(x, x )f (x ) d3 x , (1.8.3)

1.8

Greens function for Poissons equation

13

where 0 (x) is a solution to the inhomogeneous (Laplace) equation, 2 0 (x) = 0. (1.8.4)

This assertion is proved by substituting Eq. (1.8.3) into the left-hand side of Eq. (1.8.1), and using Eqs. (1.8.2) and (1.8.4) to show that the result is indeed equal to 4f (x). In Eq. (1.8.3), the role of the integral is to account for the source term in Eq. (1.8.1). The role of 0 (x) is to enforce boundary conditions that we might wish to impose on the potential (x). In the absence of such boundary conditions, 0 can simply be set equal to zero. Methods to solve boundary-value problems will be introduced in the next chapter. To construct the Greens function we rst argue that G(x, x ) can depend only on the dierence x x ; this follows from the fact that the source term depends only on x x , and the requirement that the Greens function must be invariant under a translation of the coordinate system. We then express G(x, x ) as the Fourier (k): transform of a function G G(x, x ) = 1 (2 )3 (k)eik(xx ) d3 k, G (1.8.5)

where k is a vector in reciprocal space; this expression incorporates our assumption that G(x, x ) depends only on x x . Recalling that the three-dimensional function can be represented as (x x ) = 1 (2 )3 eik(xx ) d3 k,

(1.8.6)

(k) = 4 , so that Eq. (1.8.6) becomes we see that Eq. (1.8.2) implies k2 G 4 G(x, x ) = (2 )3

eik(xx ) 3 d k. k2

(1.8.7)

We must now evaluate this integral. To do this it is convenient to switch to spherical coordinates in reciprocal space, dening a radius k and angles and by the relations kx = k sin cos , ky = k sin sin , and kz = k cos . In these coordinates the volume element is d3 k = k 2 sin dkdd. For convenience we orient the coordinate system so that the kz axis points in the direction of R x x . We then have k R = kR cos , where R |R| is the length of the vector R. All this gives G(x, x ) = 1 2 2
2

dk
0 0

d
0

d eikR cos sin .

Integration over d is trivial, and the integral can easily be evaluated by using = cos as an integration variable:
1

eikR cos sin d =


0 1

eikR d =

2 sin kR . kR

We now have G(x, x ) = 2


0

sin kR 2 dk = kR R

sin d,

where we have switched to = kR. The remaining integral can easily be evaluated by contour integration (or simply by looking it up in tables of integrals); it is equal to /2.

14 Our nal result is therefore G(x, x ) = 1 ; |x x |

Maxwells electrodynamics

(1.8.8)

the Greens function for Poissons equation is simply the reciprocal of the distance between x and x . The general solution to Poissons equation is then (x) = 0 (x) + f (x ) 3 d x. |x x | (1.8.9)

The meaning of this equation is that apart from the term 0 , the potential at x is built by summing over source contributions from all relevant points x and dividing by the distance between x and x . Notice that in Eq. (1.8.8) we have recovered a familiar result: the potential of a point charge at x goes like 1/R.

1.9

Greens function for the wave equation

[The material presented in this section is also covered in Sec. 6.4 of Jacksons text.] We now turn to the mathematical problem of solving the wave equation, (t, x) = 4f (t, x), (1.9.1)

for a time-dependent potential produced by a prescribed source f ; we have introduced the wave, or dAlembertian, dierential operator 1 2 + 2 . c2 t2 (1.9.2)

For this purpose we seek a Greens function G(t, x; t , x ) that satises G(t, x; t , x ) = 4 (t t ) (x x ). In terms of this the general solution to Eq. (1.9.2) can be expressed as (t, x) = 0 (t, x) + G(t, x; t , x )f (t , x ) dt d3 x , (1.9.4) (1.9.3)

where 0 (t, x) is a solution to the homogeneous wave equation, 0 (t, x) = 0. (1.9.5)

That the potential of Eq. (1.9.4) does indeed solve Eq. (1.9.1) can be veried by direct substitution. To construct the Greens function we Fourier transform it with respect to time, G(t, x; t , x ) = 1 2 ( ; x, x )ei(tt ) d, G (1.9.6)

and we represent the time -function as (t t ) = 1 2 ei(tt ) d.

(1.9.7)

Substituting these expressions into Eq. (1.9.3) yields ( ; x, x ) = 4 (x x ), 2 + (/c)2 G (1.9.8)

which is a generalized form of Eq. (1.8.2). From this comparison we learn that (0; x, x ) = 1/|x x |. G

1.9

Greens function for the wave equation

15

( ; x, x ) At this stage we might proceed as in Sec. 1.8 and Fourier transform G with respect to the spatial variables. This would eventually lead to ( ; x, x ) = G 4 (2 )3 eik(xx ) 3 d k, k2 (/c)2

which is a generalized form of Eq. (1.8.7). This integral, however, is not dened, because of the singularity at k2 = (/c)2 . We shall therefore reject this method of solution. (There are ways of regularizing the integral so as to obtain a meaningful answer. One method involves deforming the contour of the integration so as to avoid the poles. This is described in Sec. 12.11 of Jacksons text.) will be of the form We can anticipate that for = 0, G
( ; x, x ) = g (, |x x |) , G |x x |

(1.9.9)

with g representing a function that stays nonsingular when the second argument, should depend on the spatial variables R |x x |, approaches zero. That G through R only can be motivated on the grounds that three-dimensional space is can only depend on the vector R x x ) and both homogeneous (so that G isotropic (so that only the length of the vector matters, and not its direction). That should behave as 1/R when R is small is justied by the following discussion. G Take Eq. (1.9.8) and integrate both sides over a sphere of small radius centered = G , we can use Gauss theorem to get at x . Since 2 G
R =

R da + (/c)2 G

R<

d3 x = 4, G

3 and it = R/R. In this equation, the volume integral is of order G where R contributes nothing in the limit 0, unless G happens to be as singular as 1/3 . The surface integral, on the other hand, is equal to 42 dG dR .
R =

were to behave as 1/3 , then dG/dR If G would be of order 1/4 , the surface integral 2 would contribute a term of order 1/ , and the left-hand side would never give rise cannot be so singular, and that the leftto the required 4 . So we conclude that G must be of order hand side is dominated by the surface integral. This implies that G 1/, as was expressed by Eq. (1.9.9). Setting G = g/R returns 4g (, ) + O() for the surface integral, and this gives us the condition g (, 0) = 1. (1.9.10)

We also recall that g (0, R) = 1. We may now safely take R = 0 and substitute Eq. (1.9.9) into Eq. (1.9.8), taking depends on x only through R, the Laplacian its right-hand side to be zero. Since G operator becomes 1 d 2 d 2 2 R . R dR dR = g/R yields g /R and Eq. (1.9.8) becomes Acting with this on G g + (/c)2 g = 0, (1.9.11)

with a prime indicating dierentiation with respect to R. With the boundary condition of Eq. (1.9.10), two possible solutions to this equation are g (, R) = ei(/c)R . (1.9.12)

16

Maxwells electrodynamics

Substituting this into Eq. (1.9.9), and that into Eq. (1.9.6), we obtain G (t, x; t , x ) = or G (t, x; t , x ) = 1 2 1 ei(/c)R i(tt ) e d = R 2R ei(tt R/c) d,

These are the two fundamental solutions to Eq. (1.9.3). The function G+ (t, x; t , x ), which is nonzero when t t = +R/c, is known as the retarded Greens function; the function G (t, x; t , x ), which is nonzero when t t = R/c, is known as the advanced Greens function. Substituting the Greens functions into Eq. (1.9.4) gives (t, x) = 0 (t, x) + t t |x x |/c f (t , x ) dt d3 x . |x x |

t t |x x |/c . |x x |

(1.9.13)

The time integration can be immediately carried out, and we obtain (t, x) = 0 (t, x) + f t |x x |/c, x 3 d x |x x | (1.9.14)

for the fundamental solutions to the wave equation. Except for the shifted time dependence, Eq. (1.9.14) looks very similar to Eq. (1.8.9), which gives the general solution to Poissons equation. The time translation, however, is very important. It means that the potential at time t depends on the conditions at the source at a shifted time t = t |x x |/c. The second term on the right-hand side is the time required by light to travel the distance between the source point x and the eld point x; it encodes the property that information about the source travels to x at the speed of light. The + solution, + (t, x), depends on the behaviour of the source at an earlier time t R/c there is a delay between the time the information leaves the source and the time it reaches the observer. This solution to the wave equation is known as the retarded solution, and it properly enforces causality: the cause (source) precedes the eect (potential). On the other hand, (t, x) depends on the behaviour of the source at a later time t + R/c; here the behaviour is anti-causal, and this unphysical solution to the wave equation is known as the advanced solution. Although both solutions to the wave equation are mathematically acceptable, causality clearly dictates that we should adopt the retarded solution + (t, x) as the only physically acceptable solution. So we take (t, x) = 0 (t, x) + f t |x x |/c, x 3 d x |x x | (1.9.15)

to be the solution to the wave equation (1.9.1). Recall that 0 (t, x) is a solution to the homogeneous equation, 0 = 0.

1.10

Summary

The scalar potential and the vector potential A satisfy the wave equations 1 1 2 + 2 (t, x) = (t, x) c2 t2 0 (1.10.1)

1.11 and

Problems

17

1 2 + 2 A(t, x) = 0 j (t, x) c2 t2 when the Lorenz gauge condition A+

(1.10.2)

1 =0 (1.10.3) c2 t is imposed. The speed of propagation of the waves, c = 1/ 0 0 , is numerically equal to the speed of light in vacuum. The retarded solutions to the wave equations are (t, x) = 0 (t, x) + and 1 40 t |x x |/c, x 3 d x |x x | (1.10.4)

j t |x x |/c, x 3 0 d x. (1.10.5) 4 |x x | In static situations, the time dependence of the sources and potentials can be dropped. Once the scalar and vector potentials have been obtained, the electric and magnetic elds are recovered by straightforward dierential operations: A(t, x) = A0 (t, x) + A , B = A. (1.10.6) t This ecient reformulation of the equations of electrodynamics is completely equivalent to the original presentation of Maxwells equations. It will be the starting point of most of our subsequent investigations. E=

1.11

Problems

1. We have seen that the charge density of a point charge q located at r (t) is given by (x) = q x r (t) , and that its current density is j (x) = q v (t) x r (t) , where v = dr /dt is the charges velocity. Prove that and j satisfy the statement of local charge conservation, + j = 0. t 2. In this problem we explore the formulation of electrodynamics in the Coulomb gauge, an alternative to the Lorenz gauge adopted in the text. The Coulomb gauge is especially useful in the formulation of a quantum theory of electrodynamics. a) Prove that the Coulomb gauge condition, A = 0, can always be imposed on the vector potential. Then show that in this gauge, the scalar potential satises Poissons equation, 2 = /0 , so that it can be expressed as (t, x) = 1 40 (t, x ) 3 d x. |x x | . t

Show also that the vector potential satises the wave equation A = 0 j + 0 0

18 b) Dene the longitudinal current by jl (t, x) = 1 4

Maxwells electrodynamics

j (t, x ) 3 d x, |x x |

c) Prove that the transverse current can also be dened by jt (t, x) = 1 4

and show that this satises jl = 0. Then prove that the wave equation for the vector potential can be rewritten as A = 0 jt , where jt j jl is the transverse current. j (t, x ) 3 d x , |x x |

d) The labels longitudinal and transverse can be made more meaningful by formulating the results of part b) and c) in reciprocal space instead of ordinary space. For this purpose, introduce the Fourier transform j (t, k) of the current density, such that j (t, x) = 1 (2 )3 j (t, k)eikx d3 k.

and that it satises jt = 0.

Then prove that the Fourier transform of the longitudinal current is j (t, k) k , jl (t, k) = k = k/|k| is a unit vector aligned with the reciprocal position where k vector k; thus, the longitudinal current has a component in the direction of k only. Similarly, prove that the Fourier transform of the transverse current is j (t, k) k ; jt (t, k) = k thus, the transverse current has components in the directions orthogonal to k only. For a concrete illustration, assume that k points in the direction of the z axis. Then show that jl = jz z and jt = jx x + jy y . 3. Prove that G(x, x ) = is a solution to 1 ik|xx | e 2ik

d2 + k 2 G(x, x ) = (x x ), dx2

where k is a constant. This shows that G(x, x ) is a Greens function for the inhomogeneous Helmholtz equation in one dimension, (d2 /dx2 + k 2 ) (x) = f (x), where is the potential and f the source. 4. The dierential equation d2 x + 2 x = f (t) dt2 governs the motion of a simple harmonic oscillator of unit mass and natural frequency driven by an arbitrary external force f (t). It is supposed that the force starts acting at t = 0, and that prior to t = 0 the oscillator was at rest, so that x(0) = x (0) = 0, with an overdot indicating dierentiation with respect to t.

1.11

Problems

19

Find the retarded Greens function G(t, t ) for this dierential equation, which must be a solution to d2 G + 2 G = (t t ). dt2 Then nd the solution x(t) to the dierential equation (in the form of an integral) that satises the stated initial conditions. Check your results by verifying that if f (t) = (t) cos(t), so that the oscillator is driven at resonance, then x(t) = (2 )1 t sin t. 5. In this problem (adapted from Jacksons problem 6.1) we construct solutions to the inhomogeneous wave equation = 4f in a few simple situations. A particular solution to this equation is (t, x) = where G(t, x; t , x ) = is the retarded Greens function. a) Take the source of the wave to be a momentary point source, described by f (t , x ) = (x ) (y ) (z ) (t ). Show that the solution to the wave equation in this case is given by (t, x, y, z ) = (t r/c) , r G(t, x; t , x )f (t , x ) dt d3 x ,

(t t |x x |/c) |x x |

where r x2 + y 2 + z 2 . The wave is therefore a concentrated wavefront expanding outward at the speed of light. b) Now take the source of the wave to be a line source described by f (t , x ) = (x ) (y ) (t ). (Notice that this also describes a point source in a space of two dimensions.) Show that the solution to the wave equation in this case is given by 2c(t /c) , (t, x, y ) = (ct)2 2 where x2 + y 2 and (s) is the Heaviside step function. Notice that the wave is no longer well localized, but that the wavefront still travels outward at the speed of light. c) Finally, take the source to be a sheet source described by f (t , x ) = (x ) (t ). (Notice that this also describes a point source in a space of a single dimension.) Show that the solution to the wave equation in this case is given by (t, x) = 2c(t |x|/c). Notice that apart from a sharp cuto at |x| = ct, the wave is now completely uniform. 6. In the quantum version of Maxwells theory, the electromagnetic interaction is mediated by a massless photon. In this problem we consider a modied

20

Maxwells electrodynamics classical theory of electromagnetism that would lead, upon quantization, to a massive photon. It is based on the following set of eld equations: E + 2 B E B + 2 A 1 , 0 = 0, B = , t = = 0 j + 0 0 E , t

where is a new constant related to the photon mass m by = mc/ , and where the elds E and B are related in the usual way to the potentials and A. The eld equations are supplemented by the same Lorentz-force law, f = E + j B , as in the original theory. a) Prove that the modied theory enforces charge conservation provided that the potentials are linked by A + 0 0 = 0. t

The Lorenz gauge condition must therefore always be imposed in the modied theory. b) Find the modied wave equations that are satised by the potentials and A. c) Verify that in modied electrostatics, the scalar potential outside a point charge q at x = 0 is given by the Yukawa form = q er . 40 r

d) Prove that the modied theory enforces energy conservation by deriving an equation of the form + S = j E . t Find the new expressions for and S ; they should reduce to the old expressions when 0. In the original theory the potentials have no direct physical meaning; is this true also in the modied theory?

Chapter 2 Electrostatics
2.1 Equations of electrostatics

For time-independent situations, the equations of electromagnetism decouple into a set of equations for the electric eld alone electrostatics and another set for the magnetic eld alone magnetostatics. The equations of magnetostatics will be considered in Chapter 3. The topic of this chapter is electrostatics. The equations of electrostatics are Poissons equation for the scalar potential, 2 (x) = 1 (x), 0 (2.1.1)

and the relation between potential and eld, E = . The general solution to Poissons equation is (x) = 0 (x) + 1 40 (x ) 3 d x, |x x | (2.1.3) (2.1.2)

where 0 (x) satises Laplaces equation, 2 0 (x) = 0. (2.1.4)

The role of 0 (x) is to enforce boundary conditions that we might wish to impose on the scalar potential.

2.2

Point charge

The simplest situation involves a point charge q located at a xed point b. The charge density is (x) = q (x b), (2.2.1) and in the absence of boundaries, Eq. (2.1.3) gives (x) = 1 q . 40 |x b| (2.2.2)

To compute the electric eld we rst calculate the gradient of R |x b|, the distance between the eld point x and the charge. We have R2 = (x bx )2 + (y by )2 + (z bz )2 and dierentiating both sides with respect to, say, x gives 2RR/x = 2(x bx ), or R/x = (x bx )/R. Derivatives of R with respect 21

22

Electrostatics

z p

+q

y x q

Figure 2.1: Geometry of a dipole. to y and z can be computed in a similar way, and we have established the useful vectorial relation xb . (2.2.3) |x b| = |x b| Notice that R is a unit vector that points in the direction of R x b. The calculation of the electric eld involves R1 = R2 R, or This gives E (x) = xb q . 40 |x b|3 (2.2.5) xb 1 = . |x b| |x b|3 (2.2.4)

The electric eld goes as q/R2 and points in the direction of the vector R = x b. This well-known result is known as Coulombs law.

2.3

Dipole

Another elementary situation is that of two equal charges, one positive, the other negative, separated by a distance d 2. We align the charges along the direction of the unit vector p , and place the origin of the coordinate system at the middle point; see Fig. 2.1. The charge density of this distribution is given by (x) = q (x p ) q (x + p ), and the potential is (x) = q 40 1 1 . |x p | |x + p | (2.3.2) (2.3.1)

This is an exact expression. We can simplify it if we take the observation point x to be at a large distance away from the dipole, r |x| . (2.3.3)

2.3

Dipole

23

We can then approximate |x p | = = (x p ) (x p ) r2 2p x + 2

= r 1 p x/r2 + O(2 /r2 ) , or 1 1 = 1 p x/r2 + O(2 /r2 ) . |x p | r ) x 1 (2qp . 40 r3

Keeping terms of order only, the potential becomes =

The vector p 2qp is the dipole moment of the charge distribution. In general, for a collection of several charges, this is dened as p=
A

qA x A ,

(2.3.4)

where xA is the position vector of the charge qA . In the present situation the denition implies p = (+q )(p ) + (q )(p ) = 2qp , as was stated previously. The potential of a dipole is then (x) = 1 px , 40 r3 (2.3.5)

when r satises Eq. (2.3.3). It should be noticed that here, the potential falls o as 1/r2 , faster than in the case of a single point charge. This has to do with the fact that here, the charge distribution has a vanishing total charge. It is instructive to rederive this result by directly expanding the charge density in powers of . For any function f of a vector x we have f (x + p ) = f (x) + p f (x) + O(2 ), and we can extend this result to Diracs distribution: (x p ) = (x) p (x) + O(2 ). The charge density of Eq. (2.3.1) then becomes (x) = 2qp (x) + O(2 ), or (x) = p (x). This is the charge density of a point dipole. Substituting this into Eq. (2.1.3) and setting 0 = 0 yields (x) = pa 40
a (x ) 3 d x, |x x |

(2.3.6)

where summation over the vectorial index a is understood. The standard procedure when dealing with derivatives of a -function is to integrate by parts. The integral is then (x ) a 1 d3 x = |x x | (x ) xa (x x )a 3 d x = 3, |x x |3 r

and we have recovered Eq. (2.3.5). Notice that in these manipulations we have inserted a result analogous to Eq. (2.2.4); the minus sign does not appear because we are now dierentiating with respect to the primed variables.

24

Electrostatics

n Sout V S in n
Figure 2.2: An inner boundary Sin on which is specied, an outer boundary Sout on which is specied, and the region V in between that contains a charge distribution . The outward unit normal to both boundaries is denoted n . From the potential of Eq. (2.3.5) we can calculate the electric eld of a point dipole. Taking a derivative with respect to x gives x = 1 x (p x) 3(p x) x r . 40 r3 r4

We have x (p x) = px and according to Eq. (2.2.3), x r = x/r. So x = 1 px 3(p x)x , 40 r3 r5

and a similar calculation can be carried out for the y and z components of . Introducing the unit vector r = x/r = (x/r, y/r, z/r), we nd that the electric eld is given by E (x) = 1 3(p r )r p . 3 40 r (2.3.8) (2.3.7)

This is an exact expression for the electric eld of a point dipole, for which the charge density is given by Eq. (2.3.6). For a physical dipole of nite size, this expression is only an approximation of the actual electric eld; the approximation is good for r .

2.4

Boundary-value problems: Greens theorem

[The material presented in this section is also covered in Secs. 1.8 and 1.10 of Jacksons text.] After the warmup exercise of the preceding two sections we are ready to face the more serious challenge of solving boundary-value problems. A typical situation in electrostatics features a charge distribution (x) between two boundaries on which the potential (x) is specied (see Fig. 2.2) the potential is then said to satisfy Dirichlet boundary conditions. A concrete situation might involve an inner

2.4

Boundary-value problems: Greens theorem

25

da

Sout V

da

S in

Figure 2.3: The union of the inner boundary Sin , the outer boundary Sout , and the narrow channel between them forms a closed surface S that encloses the region V. boundary Sin that is also a grounded conducting surface (on which = 0) and an outer boundary Sout that is pushed to innity (so that = 0 there also). We would like to derive an expression for (x) in the region between the boundaries; this should account for both the charge density and the boundary data. To arrive at this expression we will need to introduce a new type of Greens function, which we will call a Dirichlet Greens function and denote GD (x, x ); its properties will be identied along the way. To get there we will make use of Greens identity, which is essentially an application of Gauss theorem. Gauss theorem states that for a volume V enclosed by a surface S , b d3 x = b da,

where b is any vector eld dened within V and da is an outward surface element on S . The theorem, as stated, says nothing about a region V bounded by two boundaries Sin and Sout , the situation that interests us. But we can make our situation t the formulation of the theorem by digging a narrow channel from Sout to Sin , as depicted in Fig. 2.3. The union of Sin , Sout , and the channel is a closed surface, and integrating b over both sides of the channel produces a zero result because da points in opposite directions. We therefore have b d3 x = b da + b da,

Sout

Sin

and noting the opposite orientations of da and n on Sin (see Fig. 2.3), we write this as b d3 x = bn da bn da

Sout

Sin

Sout Sin

bn da.

It is important to notice that the unit vector n points out of both Sin and Sout . We now choose the vector b to be b = ,

26

Electrostatics

where (x) and (x) are two arbitrary functions. We have b = 2 2 and Gauss theorem gives 2 2 d3 x = n da. (2.4.1)

Sout Sin

This is Greens identity. For convenience we write it in the equivalent form 2 2 d3 x = n da ,

Sout Sin

where all elds are now expressed in terms of the new variables x . We now pick (x ) (x ) scalar potential and (x ) GD (x, x ) Dirichlet Greens function, where the properties of the Dirichlet Greens function will be described in detail below. For now we assume that it satises the equation 2 GD (x, x ) = 4 (x x ), (2.4.2)

while the scalar potential satises Poissons equation, 2 (x ) = (x )/0 . Making these substitutions in the equation following Eq. (2.4.1) leads to 4 (x) + 1 0 GD (x, x )(x ) d3 x
V

=
Sout Sin

(x ) GD (x, x ) GD (x, x ) (x ) n da ,

or (x) = 1 40 1 4 GD (x, x )(x ) d3 x


V

Sout Sin

(x ) GD (x, x ) GD (x, x ) (x ) n da .

Apart from a remaining simplication, this is the kind of expression we were seeking. The volume integral takes care of the charge distribution within V , and the two surface integrals take care of the boundary conditions. But there is one problem: While the value of the scalar potential is specied on the boundaries, so that the functions (x ) are known inside the surface integrals, we are given no information about n (x ), its normal derivative, which also appears within the surface integrals. It is not clear, therefore, how we might go about evaluating these integrals: the problem does not seem to be well posed. To sidestep this problem we demand that the Dirichlet Greens function satisfy the condition GD (x, x ) = 0 when x is on the boundaries. (2.4.3)

Then the boundary terms involving n (x ) simply go away, because they are multiplied by GD (x, x ) which vanishes on the boundaries. With this property we arrive at (x) = 1 40 1 4 GD (x, x )(x ) d3 x
V

Sout Sin

(x )n GD (x, x )da .

(2.4.4)

2.4

Boundary-value problems: Greens theorem

27

This is our nal expression for the scalar potential in the region V . The problem of nding this potential has been reduced to that of nding a Dirichlet Greens function that satises Eqs. (2.4.2) and (2.4.3). Notice that this is a specialized form of our original problem: GD (x, x ) is the potential produced by a point charge of strength 40 located at x, and its boundary values on Sin and Sout are zero. If we can solve this problem and obtain the Dirichlet Greens function, then Eq. (2.4.4) gives us the means to calculate (x) in very general circumstances. We already know that GD (x, x ) satises Eq. (2.4.2) and vanishes when x lies on Sin and Sout . We now use Greens identity to show that the Dirichlet Greens function is symmetric in its arguments: GD (x , x) = GD (x, x ). (2.4.5)

This implies that it is also a solution to the standard equation satised by a Greens function, 2 GD (x, x ) = 4 (x x ), (2.4.6) which is identical to Eq. (1.8.2). To establish Eq. (2.4.5) we take y to be the integration variables in Greens identity,
2 3 2 y y d y =

Sout Sin

y y n day ,

and we pick (y ) GD (x, y ) and (y ) GD (x , y ). Then according to Eq. (2.4.2) 2 we have 2 y = 4 (x y ) and y = 4 (x y ). The left-hand side of Greens identity gives 4 [GD (x, x ) GD (x , x)] after integration over d3 y . The right-hand side, on the other hand, gives zero because both and are zero when y is on Sin and Sout . This produces Eq. (2.4.5). To summarize, we have found that the Dirichlet Greens function satises Eq. (2.4.6), 2 GD (x, x ) = 4 (x x ), possesses the symmetry property of Eq. (2.4.5), GD (x , x) = GD (x, x ), and satises the boundary conditions of Eq. (2.4.3), GD (x, x ) = 0 when x or x is on the boundaries.

Notice that the boundary conditions have been generalized in accordance to Eq. (2.4.5): the Dirichlet Greens function vanishes whenever x or x happens to lie on the boundaries. Once the Dirichlet Greens function has been found, the potential produced by a charge distribution (x) in a region V between two boundaries Sin and Sout on which (x) is specied can be obtained by evaluating the integrals of Eq. (2.4.4). The task of nding a suitable Dirichlet Greens function can only be completed once the boundaries Sin and Sout are fully specied; the form of the Greens function will depend on the shapes and locations of these boundaries. In the absence of boundaries, GD (x, x ) reduces to the Greens function constructed in Sec. 1.8, G(x, x ) = |x x |1 . In the following sections we will consider a restricted class of boundary-value problems, for which there is only an inner boundary Sin (Sout is pushed to innity and does not need to be considered); the inner boundary is spherical (Sin is described by the statement |x | = R, where R is the surfaces radius);

28

Electrostatics the inner boundary is the surface of a grounded conductor ( vanishes on Sin ).

With these restrictions (introduced only for simplicity), Eq. (2.4.4) reduces to (x) = 1 40 GD (x, x )(x ) d3 x .
|x |>R

(2.4.7)

In these situations, the electric eld vanishes inside the conductor and there is an induced distribution of charge on the surface. The surface charge density is given by Eq. (1.6.9), = 0 E n , (2.4.8) where E is the electric eld just above Sin . This is given by E = (|x| = R). To proceed we will need to nd a concrete expression for the Dirichlet Greens function. For the spherical boundary considered here, this is done by solving Eq. (2.4.6) in spherical coordinates.

2.5

Laplaces equation in spherical coordinates

[The material presented in this section is also covered in Secs. 3.1, 3.2, and 3.5 of Jacksons text.] When formulated in spherical coordinates (r, , ), Laplaces equation takes the form 2 = 1 r2 r2 r r + r2 1 sin sin + r2 1 2 2 2 = 0. sin (2.5.1)

To solve this equation we separate the variables according to (r, , ) = R(r)Y (, ), and we obtain the decoupled equations dR d r2 dr dr and 1 Y sin sin = ( + 1)R (2.5.3) (2.5.2)

1 2Y = ( + 1)Y, sin2 2

(2.5.4)

where ( + 1) is a separation constant. The solutions to the angular equation (2.5.4) are the spherical harmonics Ym (, ), which are labeled by the two integers and m. While ranges from zero to innity, m is limited to the values , + 1, , 1, . For m 0, Ym (, ) = 2 + 1 4 ( m)! m P (cos )eim , ( + m)! (2.5.5)

where Pm (cos ) are the associated Legendre polynomials. For m < 0 we use
Ym = (1)m Y, m ,

(2.5.6)

where an asterisk indicates complex conjugation. Particular spherical harmonics are Y00 = 1 , 4

2.5

Laplaces equation in spherical coordinates 3 sin ei , 8 3 cos , 4 15 sin2 e2i , 2 15 sin cos ei , 8 5 (3 cos2 1). 4

29

Y11 Y10 Y22 Y21 Y20

= = = 1 4

= = 1 2

Two fundamental properties of the spherical harmonics are that they are orthonormal functions, and that they form a complete set of functions of the angles and . The statement of orthonormality is Y m (, )Ym (, ) d = mm , (2.5.7)

where d = sin dd is an element of solid angle. The statement of completeness is that any function f (, ) can be represented as a sum over spherical harmonics:

f (, ) =
=0 m=

fm Ym (, )

(2.5.8)

for some coecients fm . By virtue of Eq. (2.5.7), these can in fact be calculated as fm = f (, )Ym (, ) d. (2.5.9) Equation (2.5.8) means that the spherical harmonics form a complete set of basis functions on the sphere. It is interesting to see what happens when Eq. (2.5.9) is substituted into Eq. (2.5.8). To avoid confusion we change the variables of integration to and : f (, ) =
m

Ym (, ) f ( , )
m

f ( , )Ym ( , ) d

Ym ( , )Ym (, ) d .

The quantity within the large square brackets is such that when it is multiplied by f ( , ) and integrated over the primed angles, it returns f (, ). This must therefore be a product of two -functions, one for and the other for . More precisely stated,
Ym ( , )Ym (, ) = =0 m=

( ) ( ) , sin

(2.5.10)

where the factor of 1/ sin was inserted to compensate for the factor of sin in d (the -function is enforcing the condition = ). Equation (2.5.10) is known as the completeness relation for the spherical harmonics. This is analogous to a well-known identity,
1 eikx 2

1 eikx dk = (x x ), 2

30

Electrostatics

in which the integral over dk replaces the discrete summation over and m; the basis functions (2 )1/2 eikx are then analogous to the spherical harmonics. The solutions to the radial equation (2.5.3) are power laws, R r or R (+1) r . The general solution is Rm (r) = am r + bm r(+1) , (2.5.11)

where am and bm are constants. The general solution to Laplaces equation is obtained by combining Eqs. (2.5.2), (2.5.5), (2.5.11) and summing over all possible values of and m:

(r, , ) =
=0 m=

am r + bm r(+1) Ym (, ).

(2.5.12)

The coecients am and bm are determined by the boundary conditions imposed on (r, , ). If, for example, is to be well behaved at r = 0, then bm 0. If, on the other hand, is to vanish at r = , then am 0. From these observations we deduce that the only solution to Laplaces equation that is well behaved at the origin and vanishes at innity is the trivial solution = 0.

2.6

Greens function in the absence of boundaries

[The material presented in this section is also covered in Secs. 3.6, and 3.9 of Jacksons text.] Keeping in mind that our ultimate goal is to obtain the Dirichlet Greens function GD (x, x ) for a spherical inner boundary, in this section we attempt something simpler and solve 2 G(x, x ) = 4 (x x ) (2.6.1) in the absence of boundaries. We already know the answer: In Sec. 1.8 we found that 1 . (2.6.2) G(x, x ) = |x x | We will obtain an alternative expression for this, one that can be generalized to account for the presence of spherical boundaries. The -function on the right-hand side of Eq. (2.6.1) can be represented as (x x ) = = (r r ) ( ) ( ) r2 sin (r r ) Ym ( , )Ym (, ), r2
m

(2.6.3)

where we have involved Eq. (2.5.10). The factor of 1/(r2 sin ) was inserted because in spherical coordinates, the volume element is d3 x = r2 sin drdd, and the Jacobian factor of r2 sin must be compensated for. We then have, for example, f (x ) = = = f (x) (x x ) d3 x f (r, , ) (r r ) ( ) ( ) 2 r sin drdd r2 sin

f (r, , ) (r r ) ( ) ( ) drdd

= f (r , , ),

2.6

Greens function in the absence of boundaries

31

as we should. Inspired by Eq. (2.6.3) we expand the Greens function as G(x, x ) =


m using the spherical harmonics Ym (, ) for angular functions and gm (r, r )Ym ( , ) for radial functions. For convenience we have made the radial function depend on x through the variables r , , and ; these are treated as constant parameters when substituting G(x, x ) into Eq. (2.6.1). Applying the Laplacian operator of Eq. (2.5.1) on the Greens function yields gm (r, r )Ym ( , )Ym (, ),

(2.6.4)

2 G(x, x ) =

dgm 1 d r2 r2 dr dr

( + 1) ( , )Ym (, ), gm Ym r2

and setting this equal to 4 (x x ) as expressed in Eq. (2.6.3) produces an ordinary dierential equation for gm (r, r ): dgm d r2 dr dr ( + 1)gm = 4 (r r ). (2.6.5)

This equation implies that the radial function depends only on , and not on m: gm (r, r ) g (r, r ). To simplify the notation we will now omit the label on the radial function. To solve Eq. (2.6.5) we rst observe that when r = r , the right-hand side of the equation is zero and g is a solution to Eq. (2.5.3). We let g< (r, r ) be the solution to the left of r (for r < r ) and g> (r, r ) be the solution to the right of r (for r > r ). We then assume that g (r, r ) can be obtained by joining the solutions at r = r : g (r, r ) = g< (r, r )(r r) + g> (r, r )(r r ),

(2.6.6)

where (r r) and (r r ) are step functions. Dierentiating this equation with respect to r gives
g (r, r ) = g< (r, r )(r r) + g> (r, r )(r r ) + g> (r , r ) g< (r , r ) (r r ),

where a prime on g indicates dierentiation. We have involved the distributional identities (r r ) = (r r ), (r r) = (r r ), and f (r) (r r ) = f (r ) (r r ); these were rst encountered in Sec. 1.6. Multiplying by r2 yields
r2 g (r, r ) = r2 g< (r, r )(r r) + r2 g> (r, r )(r r ) + r2 g> (r , r ) g< (r , r ) (r r )

and taking another derivative brings r2 g (r, r ) =


+ r2 g> (r , r ) g< (r , r ) (r r ) (r, r ) (r r ) r2 g< (r, r ) (r r) + r2 g>

+ r2 g> (r , r ) g< (r , r ) (r r ). d dg< r2 ( + 1)g< (r r) dr dr d dg> + r2 ( + 1)g> (r r ) dr dr dg> dg< + r 2 (r , r ) (r , r ) (r r ) dr dr d (r r ). + r2 g> (r , r ) g< (r , r ) dr

Substituting this and Eq. (2.6.6) back into Eq. (2.6.5), we obtain 4 (r r ) =

32

Electrostatics

For the right-hand side to match the left-hand side we must demand that g< (r, r ) and g> (r, r ) be solutions to the homogeneous version of Eq. (2.6.5), as was already understood, and that these solutions be matched according to g> (r , r ) g< (r , r ) = 0 and (2.6.7)

dg< 4 dg> (r , r ) (r , r ) = 2 . (2.6.8) dr dr r The function g (r, r ) is therefore continuous at r = r , but its rst derivative must be discontinuous to account for the -function on the right-hand side of Eq. (2.6.5). Because g< is a solution to Eq. (2.5.3), according to Eq. (2.5.11) it must be a linear combination of terms proportional to r and 1/r+1 . But only the rst term is well behaved at r = 0, and we set g< (r, r ) = a(r )r . For g> we select instead g> (r, r ) = b(r )/r+1 ,

which is well behaved at r = . It does not matter that g< diverges at r = because this function is restricted to the domain r < r < ; similarly, it does not matter that g> diverges at r = 0 because it is restricted to the domain r > r > 0. To determine the constants a and b we use the matching conditions. First, Eq. (2.6.7) gives b = ar2+1 . Second, Eq. (2.6.8) gives a = 4r(+1) /(2 + 1). So g< (r, r ) = and 4 r 2 + 1 r+1 (2.6.9)

4 r . (2.6.10) 2 + 1 r+1 We observe a nice symmetry between these two results. Combining Eqs. (2.6.6), (2.6.9), and (2.6.10), we nd that the radial function can be expressed in the compact form 4 r< g (r, r ) = , (2.6.11) +1 2 + 1 r> g> (r, r ) = where r< denotes the lesser of r and r , while r> is the greater of r and r . Thus, if r < r then r< = r, r> = r , and Eq. (2.6.11) gives g (r, r ) = g< (r, r ), as it should. Similarly, when r > r we have r< = r , r> = r, and Eq. (2.6.11) gives g (r, r ) = g> (r, r ). Substituting Eq. (2.6.11) into Eq. (2.6.4) and taking into account Eq. (2.6.2), we arrive at 1 = |x x |
4 r< Ym ( , )Ym (, ). +1 2 + 1 r>

(2.6.12)

=0 m=

This very useful identity is known as the addition theorem for spherical harmonics. At rst sight the usefulness of this identity might seem doubtful. After all, we have turned something simple like an inverse distance into something horrible involving special functions and an innite double sum. But we will see that when 1/|x x | appears inside an integral, the representation of Eq. (2.6.12) can be very useful indeed. Not relying on this identity would mean having to evaluate an integral that involves 1 = |x x | 1 r2 2rr [sin sin cos( ) + cos cos ] + r2 ,

2.6

Greens function in the absence of boundaries

33

and such integrals typically cannot be evaluated directly: they are just too complicated. Let us consider an example to illustrate the power of the addition theorem. We want to calculate the electrostatic potential both inside and outside a spherical distribution of charge with density (x ) = (r ). There are no boundaries in this problem, but the distribution is conned to a sphere of radius R. After substituting Eq. (2.6.12) into Eq. (2.1.3) and writing d3 x = r2 dr d , we obtain (x) = 1 40 4 Ym (, ) 2 + 1
R 0 r< (r )r2 dr +1 r> Ym ( , ) d .

To evaluate the angular integral we multiply and divide by (4 )1/2 Y00 ( , ):


Ym ( , ) d =

Ym ( , )Y00 ( , ) d =

4,0 m,0 ,

where we have involved Eq. (2.5.7). At this stage we happily realize that the innite sum over and m involves but a single term. We therefore have (x) = 1 4 4Y00 (, ) 40
R 0

r 2 1 (r ) dr = 4 r> 40

R 0

r 2 (r ) dr , r>

and as we might have expected from the spherical symmetry of the problem, the potential depends on r only. To evaluate the radial integral we must be careful with the meaning of r> max(r, r ). Suppose rst that x is outside the charge distribution, so that r > R. Then r > r , r> = r, and we have
R

4
0

1 r 2 (r ) dr = 4 r> r
R 0

(r )r2 dr =
0

Q , r

where Q (x ) d3 x = 4 In this case we have

(r )r2 dr is the total charge of the distribution. 1 Q , 40 r (2.6.13)

out (x) =

the same result as in Eq. (2.2.2). Suppose next that x is inside the charge distribution, so that r < R. Then part of the integration covers the interval 0 r < r and the remaining part covers r r R. In the rst part r is smaller than r and r> = r; in the second part r is larger than r and r> = r . So
R

4
0

1 r 2 (r ) dr = 4 r> r

(r )r2 dr + 4
0 r

(r )r dr .

In the rst term we recognize


r

q (r)

(x ) d3 x = 4
|x |r 0

(r )r2 dr ,

(2.6.14)

the charge enclosed by a sphere of radius r; if r > R then q (r) = Q. The internal potential can thus be expressed as in (x) = 1 q (r) + 4 40 r
R

(r )r dr .
r

(2.6.15)

While the expression (2.6.15) for the internal potential depends on the details of the charge distribution the particular form of the function (r ) a nicer

34

Electrostatics

conclusion applies to the electric eld. The eld is obtained by dierentiating in with respect to r, and 1 q 1 dq din = 2+ 4r . dr 40 r r dr But since dq/dr = 4r2 according to Eq. (2.6.14), we see that the last two terms cancel out. The electric eld is then E (x) = 1 q (r) r , 40 r2 (2.6.16)

both inside and outside the charge distribution; recall that when r > R, q (r) becomes Q, the total charge. We have recovered the well-known result that for any spherical distribution of charge, the electric eld at a radius r depends only on the charge q (r) enclosed by a sphere of that radius; and the eld is the same as if all this charge were concentrated in a point at r = 0.

2.7

Dirichlet Greens function for a spherical inner boundary

[The material presented in this section is also covered in Sec. 3.9 of Jacksons text.] We now return to the task of nding the Dirichlet Greens function GD (x, x ) for the specic situation described near the end of Sec. 2.4, in which we have a spherical inner boundary at |x| = R and no outer boundary. It will be a simple matter to modify the treatment presented in Sec. 2.6 to account for the inner boundary. As in Eq. (2.6.4), the Dirichlet Greens function can be expanded as GD (x, x ) =
m g (r, r )Ym ( , )Ym (, ).

(2.7.1)

As in Eq. (2.6.6), the radial function can be expressed as g (r, r ) = g< (r, r )(r r) + g> (r, r )(r r ), (2.7.2)

where g< and g> are solutions to Eq. (2.5.3) that must also satisfy the matching conditions of Eqs. (2.6.7) and (2.6.8), g> (r , r ) g< (r , r ) = 0 and (2.7.3)

dg< 4 dg> (2.7.4) (r , r ) (r , r ) = 2 . dr dr r The only dierence with respect to what was done in Sec. 2.6 is that now, according to Eq. (2.4.3), the Greens function must vanish on the inner boundary. This gives the additional conditions g (r = R, r ) = g (r, r = R) = 0, or g< (r = R, r ) = 0, g> (r, r = R) = 0; (2.7.5)

recall that R is now the smallest possible value of both r and r . Since g< (r, r ) is no longer required to be well behaved at r = 0, we must set it equal to g< (r, r ) = a(r , R)r + (r , R)/r+1 , the general solution to Eq. (2.5.3). For g> (r, r ) we choose instead g> (r, r ) = b(r , R)/r+1 ,

2.8

Point charge outside a grounded, spherical conductor

35

which is well behaved at r = . Solving Eqs. (2.7.3)(2.7.5) for the constants produces a = 4r(+1) /(2 + 1), = aR2+1 , and b = a(r2+1 R2+1 ). Collecting these results gives g< (r, r ) = and g> (r, r ) = 1 R2+1 4 r +1 +1 2 + 1 r r R2+1 4 1 r +1 . +1 2 + 1 r r (2.7.6)

(2.7.7)

Combining Eqs. (2.7.2), (2.7.6), and (2.7.7), we nd that the radial function can be expressed in the compact form g (r, r ) = 4 R2+1 1 r< +1 +1 2 + 1 r> r< =
2+1 4 r< R2+1 , 2 + 1 (rr )+1

(2.7.8)

where r< denotes the lesser of r and r , while r> is the greater of r and r (notice that r< r> rr ). We see that g (r, r ) vanishes on the inner boundary: When r = R it must be that r < r , so that r< = r = R, and the terms within the brackets vanish; on the other hand, when r = R it must be that r < r, so that r< = r = R, and the same conclusion applies. The Dirichlet Greens function is obtained by substituting Eq. (2.7.8) into Eq. (2.7.1): GD (x, x ) =
m

1 4 R2+1 r< +1 Ym ( , )Ym (, ). +1 2 + 1 r> r<

(2.7.9)

This is slightly more complicated than Eq. (2.6.12), but still manageable. Notice that the Dirichlet Greens function reduces to 1/|x x | when R = 0. It would be straightforward to generalize this derivation to account also for the presence of an outer boundary; we shall not pursue this here.

2.8

Point charge outside a grounded, spherical conductor

[The material presented in this section is also covered in Sec. 2.2 of Jacksons text.] The electrostatic potential outside a grounded, spherical conductor of radius R is given by Eq. (2.4.7), (x) = 1 40 GD (x, x )(x ) d3 x ,
|x |>R

(2.8.1)

and the Dirichlet Greens function was obtained as GD (x, x ) =


m

R2+1 1 4 r Ym ( , )Ym (, ) < +1 +1 2 + 1 r> r<

(2.8.2)

in the preceding section. The charge density induced on the surface of the conductor is given by Eq. (2.4.8), = 0 E n , where E is the electric eld just above the conductor and n r is the surfaces unit normal. Because E = , this works out to be . (2.8.3) = 0 r r=R We place a point charge q on the z axis, at a distance z0 > R from the centre of the conducting sphere. The charges position vector is z0 = z0 z and its spherical coordinates are r0 = z0 , 0 = 0, and 0 is undetermined. The charge density is (x ) = q (x z0 ). (2.8.4)

36 Substituting this into Eq. (2.8.1) gives (x) = =

Electrostatics

q GD (x, z0 ) 40 r< 4 R2+1 q +1 +1 Ym (0, 0 )Ym (, ), +1 40 2 + 1 r> r r < > m

where r< now stands for the lesser of r and z0 , while r> stands for the greater of the two. It is a property of the spherical harmonics that Ym (0, 0 ) = 2 + 1 m,0 , 4

so that the potential is actually independent of 0 and involves a single sum over . It is another property of the spherical harmonics that for m = 0, they reduce to ordinary Legendre polynomials, Y,0 (, ) = 2 + 1 P (cos ). 4

After cleaning up the algebra, the nal expression for the potential is (x) = q 40
+1 r< R2+1 /z0 P (cos ). +1 r+1 r>

(2.8.5)

=0

We see that the potential depends on r and only; the fact that it does not depend on reects the axial symmetry of the problem. Before moving on we pause and recall the main properties of the Legendre polynomials. These are ordinary polynomials of order ; if is even the function P (x) contains only even terms, while if is odd it contains only odd terms. For example, P0 (x) = 1, P1 (x) = x, 1 P2 (x) = (3x2 1), 2 1 (5x3 3x), P3 (x) = 2 1 P4 (x) = (35x4 30x2 + 3). 8 The Legendre polynomials are orthogonal functions,
1

P (x)P (x) dx =
1

2 , 2 + 1

(2.8.6)

and they possess the special values P2n (0) = (1)n (2n 1)!! , (2n)!! P2n+1 (0) = 0, P (1) = 1, (2.8.7)

where the double factorial notation means (2n)!! = (2n)(2n 2)(2n 4) (2) or (2n 1)!! = (2n 1)(2n 3)(2n 5) (1). The potential of Eq. (2.8.5) is expressed as an innite sum over . We will evaluate this sum in a moment, but for now we calculate the charge density ()

2.8

Point charge outside a grounded, spherical conductor

37

30

25

20

15

10

0.5

1.5 theta

2.5

Figure 2.4: Surface charge induced by a point charge above a spherical conductor. Plotted is 4R2 /q as a function of for z0 /R = 1.3 (upper curve), z0 /R = 1.6 (middle curve), and z0 /R = 1.9 (lower curve). The charge density is peaked at = 0 and its width increases with increasing values of z0 /R: the closer the charge is to the surface the narrower is the charge distribution. on the surface of the conductor. For r close to R we have r < z0 , r< = r, r> = z0 , and Eq. (2.8.5) becomes (x) =
2+1

q 40

1
+1 z0

R2+1 P (cos ). r+1

Let f (r) = r R /r be the function of r that appears inside the brackets. Its derivative is f (r) = r1 + ( + 1)R2+1 /r+2 and evaluating this at r = R gives f (R) = (2 + 1)R1 . From Eq. (2.8.3) we then obtain = or () = where will be shown presently to be the total induced charge on the surface of the conductor. Plots of the charge density are presented in Fig. 2.4 for selected values of R/z0 . That q is truly the total surface charge can be shown by evaluating the integral of the charge density over the surface of the conductor; we should recover q = da. To evaluate this we let da = R2 d = R2 sin dd, which reduces to da = 2R2 sin d after integration over . Then 1 da = q 2

+1

q 4

(2 + 1)

R1 P (cos ), +1 z0 (2.8.8)

q 4R2

(2 + 1)(R/z0 ) P (cos ),
=0

q q (R/z0 )

(2.8.9)

(2 + 1)(R/z0 )
=0 0

P (cos ) sin d.

38
1

Electrostatics

The integral over is equal to 1 P (x)P0 (x) dx = 2,0 by virtue of Eq. (2.8.6), and we obtain da = q , as required. The potential of Eq. (2.8.5) can be decomposed as (x) = q (x) + q (x), where q (x) = q 40
r< P (cos ) +1 r>

(2.8.10)

(2.8.11)

=0

is the potential that would be obtained in the absence of a boundary, and q (x) = q 40
) (z0 P (cos ) r+1

(2.8.12)

=0

is the modication introduced by the boundary condition (r = R) = 0; here q = q (R/z0 ) was dened in Eq. (2.8.9) and
z0 R2 /z0 .

(2.8.13)

< R r. /R = R/z0 < 1, so that z0 Notice that z0 The sums over can now be evaluated. For q we already know that Eq. (2.8.11) is equivalent to q 1 , (2.8.14) q (x) = 40 |x z0 | because q is the potential produced by a charge q at a position z0 = z0 z in the < r we have that Eq. (2.8.12) is absence of boundaries. By analogy, because z0 equivalent to 1 q q (x) = (2.8.15) |, 40 |x z0

where z0 = z0 z . This is the potential of a ctitious image charge q at a position z0 inside the conductor. While in reality the charge q is distributed on the surface of the conductor, the potential q is the same as if all this charge were concentrated at z0 . This observation is the basis for the method of images, which consists of nding solutions for the potential in the presence of conducting surfaces by guessing the correct positions of the correct number of image charges. It is a powerful, but tricky, method.

2.9

Ring of charge outside a grounded, spherical conductor

In this section we consider a uniformly charged ring of radius a > R. If the ring is placed in the x-y plane, its charge density is described by q (r a) ( (2.9.1) (x ) = 2 ), 2a2 where q = (x ) d3 x is the rings total charge. The -function in r indicates that the ring is innitely thin, while the -function in tells us that it is placed in the x-y plane, as was previously stated. Substituting Eq. (2.9.1) into Eq. (2.8.1) along with Eq. (2.8.2) gives (x) = q 1 40 2a2 4 Ym (, ) 2 + 1
0 r< R2+1 (r a)r2 dr +1 +1 +1 r> r> r<

Ym ( , ) ( 2 ) d ,

2.9

Ring of charge outside a grounded, spherical conductor

39

where d = sin d d, r< = min(r, r ), and r> = max(r, r ). To perform the integration we recall that Ym eim , so that integrating gives zero unless m = 0. So
2 0 Ym ( , ) d = 2Y,0 ( )m,0 .

Integration over then gives 2Y,0 ( 2 )m,0 = 2 2 + 1 P (0)m,0 . 4

On the other hand, the radial integration gives a2


r< R2+1 +1 +1 +1 r> r> r<

where r< now stands for the lesser of r and a, while r> stands for the greater of the two. With these results the potential becomes (x) = q 40
r< 4 R2+1 Y,0 () +1 +1 +1 2 + 1 r> r< r>

2 + 1 P (0). 4

Invoking once more the relation between Y,0 () and P (cos ), we arrive at (x) = q 40

P (0)
=0

r< R2+1 /a+1 P (cos ). +1 r+1 r>

(2.9.2)

Once more the potential is independent of , and is expressed as an innite sum over . Once again we will see that the sum over can be evaluated. We rst calculate the charge density on the surface of the conductor. For r close to R and therefore smaller than a, we have r< = r, r> = a, and Eq. (2.9.2) becomes (x) = q 40

P (0)
=0

f (r) P (cos ), a+1

where the function f (r) = r R2+1 /r+1 was previously encountered in Sec. 2.8. We recall that f (R) = (2 + 1)R1 , and Eq. (2.8.3) gives = or () = where q q (R/a) (2.9.4) is the total charge on the conducting surface. Plots of the surface charge density are presented in Fig. 2.5 for selected values of R/a. As in the previous section the potential of Eq. (2.9.2) can be decomposed as (x) = q (x) + q (x), (2.9.5) q 4 (2 + 1)P (0)

R1 P (cos ), a+1

q 4R2

(2 + 1)P (0)(R/a) P (cos ),


=0

(2.9.3)

where q is the potential of a ring of total charge q and radius a, while q is the potential of a ctitious image ring of total charge q and radius a = R2 /a. (You should work through the details and make sure that these statements are true.)

40

Electrostatics

2.5

1.5

0.5

0.5

1.5 theta

2.5

Figure 2.5: Surface charge induced by a ring of charge outside a spherical conductor. Plotted is 4R2 /q as a function of for a/R = 1.3 (upper curve), a/R = 1.6 (middle curve), and a/R = 1.9 (lower curve). The charge density is peaked at = 2 and its width increases with increasing values of a/R: the closer the ring is to the surface the narrower is the charge distribution.

2.10

Multipole expansion of the electric eld

[The material presented in this section is also covered in Sec. 4.1 of Jacksons text.] We now leave boundary-value problems behind, and derive the fact that in the absence of boundaries, the electrostatic potential outside a bounded distribution of charge can be characterized by a (potentially innite) number of multipole moments. Among them are Q pa Qab (x) d3 x total charge monopole moment, (x)xa d3 x dipole moment vector, (2.10.1) (2.10.2)

(x) 3xa xb r2 ab d3 x quadrupole moment tensor.(2.10.3)

Additional multipole moments can be constructed in a similar way, up to an innite number of tensorial indices. But to proceed along this road quickly becomes cumbersome, and we shall nd a better way of packaging the components of these multipole-moment tensors. Notice that the quadrupole moment is a symmetric and tracefree tensor: Qba = Qab and Qaa = 0 (summation over the repeated index is understood); it therefore possesses 5 independent components. Consider the complex quantities qm
(x)r Ym (, ) d3 x.

(2.10.4)

By virtue of Eq. (2.5.6), they satisfy the identity


q,m = (1)m qm .

(2.10.5)

2.10

Multipole expansion of the electric eld

41

Because m ranges from to , for a given there are 2 + 1 independent real quantities within the qm s. For = 0 we have only one, and we will see that q00 is proportional to the total charge. For = 1 we have three independent quantities, and we will see that the q1m s give the components of the dipole moment vector. For = 2 we have ve independent quantities, and we will see that the q2m s give the components of the quadrupole moment tensor. Thus, the quantities of Eq. (2.10.4) give a convenient packaging of the multipole moments, to all orders. To evaluate Eq. (2.10.4) for = 0 we recall that Y00 = (4 )1/2 , so that 1 q00 = Q. 4 (2.10.6)

We see that indeed, q00 is directly related to the total charge. For = 1 we have rY11 = (3/8 )1/2 (r sin )ei = (3/8 )1/2 (x iy ) and rY10 = (3/4 )1/2 r cos = 1/ 2 (3/4 ) z . This gives q11 = 3 px ipy , 8 q10 = 3 pz , 4 (2.10.7)

and we see that indeed, the q1m s are directly related to the dipole moment vec = (15/32 )1/2 (r sin ei )2 = (15/32 )1/2 (x iy )2 , tor. For = 2 we have r2 Y22 2 1/ 2 i r Y21 = (15/8 ) (r sin e )(r cos ) = (15/8 )1/2 (x iy )z , and r2 Y20 = 1/ 2 2 2 1/ 2 2 2 (5/16 ) r (3 cos 1) = (5/16 ) (3z r ). This gives q22 q21 q20 15 1 Qxx Qyy 2iQxy , 12 2 1 15 Qxz iQyz , = 6 2 5 1 Qzz , = 2 4 =

(2.10.8)

and we see that indeed, the q2m s are directly related to the quadrupole moment tensor. For higher values of it is best to stick with the denition of Eq. (2.10.4). How are the multipole moments qm related to the potential? The answer comes from Eq. (2.1.3) in which we set 0 (x) = 0 (because we are no longer considering boundaries) and substitute Eq. (2.6.12): (x) = 1 40 4 Ym (, ) 2 + 1 (x )
r< Y ( , ) d3 x . +1 m r>

Because we are looking at the potential outside the charge distribution, r is larger than r and we can set r< = r , r> = r. This gives (x) = 1 40 4 Ym (, ) 2 + 1 r+1
(x )r Ym ( , ) d3 x .

We recognize the integral over the primed variables as the denition of the qm s, and we arrive at (r, , ) = 1 40

=0 m=

4 Ym (, ) . qm 2 + 1 r+1

(2.10.9)

This is an expansion of the potential in inverse powers of r, and the multipole moments play the role of expansion coecients.

42

Electrostatics

The leading term in Eq. (2.10.9) comes from the monopole moment q00 , or total charge Q: 1 Q monopole (x) = . (2.10.10) 40 r The next term comes from the dipole moment. Using Eq. (2.10.7) and recalling the explicit expressions for the spherical harmonics of degree = 1, we have
1

q1,m Y1,m =
m=1

3 px x + py y + pz z 4 1 px , 40 r3

and dipole (x) =

(2.10.11)

as in Eq. (2.3.5). The next term in the expansion of the electrostatic potential comes from the quadrupole moment. Using Eq. (2.10.8) and the explicit expressions for the spherical harmonics of degree = 2, it is easy to show that quadrupole (x) = xa xb 1 1 Qab 5 . 40 2 r (2.10.12)

In a situation in which Q = 0, the electrostatic potential far outside the charge distribution is well approximated by monopole ; in this case the potential is characterized by a single quantity, the total charge Q. When the total charge vanishes, however, the potential is approximated by dipole , and in this case we need three quantities, the components of p, the dipole moment vector. When this vanishes also, the potential is more complicated and is approximated by quadrupole ; in this case the characterization of the potential involves the ve independent components of the quadrupole moment tensor. It is important to be aware of the fact that the multipole moments qm depend on the choice of origin for the coordinate system. For example, a point charge q at x = 0 has a charge density described by (x) = q (x) and multipole moments given by =0 q/ 4 . qm = 0 otherwise As expected, only the monopole moment is nonzero. But the same point charge placed at another position x0 possesses a very dierent set of multipole moments. In this case the density is (x) = q (x x0 ) and Eq. (2.10.4) gives
qm = qr0 Ym (0 , 0 ),

where (r0 , 0 , 0 ) are the spherical coordinates of the point x0 . We have agreement only for q00 , the lowest nonvanishing multipole moment. This is a special case of a general result: Only the lowest nonvanishing multipole moments have values that are independent of the choice of origin for the coordinate system; all other moments change under a translation of the coordinates. We shall not provide a proof of this statement.

2.11

Multipolar elds

To conclude this chapter we construct plots of electric eld lines that would be produced by a pure multipole of order . We shall consider only axisymmetric situations and set m = 0.

2.11

Multipolar elds

43

The potential for such a multipole is obtained from Eq. (2.10.9), in which we set [4/(2 + 1)]q0 Y0 (, ) = q P (cos ). This gives (r, ) = and the corresponding electric eld is E (r, ) = 1 q , ( + 1)P (cos )r + sin P (cos ) 40 r +2 (2.11.2) q P (cos ) , 40 r+1 (2.11.1)

in which a prime indicates dierentiation with respect to the argument. The electric eld lines are represented by parametric curves x() for which is an arbitrary parameter. These curves are integral curves of the vector eld E (x), which means that the tangent vector to the curves is dened to be everywhere equal to E . Thus, the eld lines are determined by solving the dierential equations dx = E (x). d (2.11.3)

In the case of an axisymmetric eld the curves can depend on r and only, and we have dx x dr x d dr d . = + = r + r d r d d d d This shows that dr/d = Er and rd/d = E , and using Eq. (2.11.2) produces dr ( + 1)P and rd sin P , or ( + 1)P (cos ) ( + 1)P () dr = d = d, r sin P (cos ) (1 2 )P () (2.11.4)

where = cos . The eld lines are therefore determined by integrating Eq. (2.11.4). To accomplish this it is useful to recall the dierential equation satised by the Legendre polynomials, dP d (1 2 ) + ( + 1)P = 0. d d We solve this algebraically for P and substitute into Eq. (2.11.4) to get (1 2 )P dr = d. r (1 2 )P ()

This can be integrated at once, and we obtain ln r = ln (1 2 )P + constant, or r() = r0 sin2 P (cos )
1/

(2.11.5)

where r0 is constant on each curve, but varies from one curve to the next. Equation (2.11.5) gives the mathematical description of the electric eld lines of an axisymmetric multipole of degree . The eld lines are plotted for selected values of in Fig. 2.6.

44

Electrostatics

Figure 2.6: Electric eld lines of an axisymmetric multipole, for selected values of the multipole order .

2.12

Problems

45

2.12

Problems

1. Calculate the electrostatic potential (x) inside a charge distribution whose density is given by 3 0 (r/r0 )2 sin2 for r < r0 , 8 where 0 and r0 are constants. For r > r0 the density is zero. There are no boundary surfaces in this problem. (x) = 2. (Jacksons Problem 1.12) Prove Greens reciprocation theorem: If 1 is the potential due to a volume charge density 1 within a volume V and a surface charge density 1 on the conducting surface S bounding the volume V , while 2 is the potential due to another charge distribution 2 and 2 , then 1 2 d3 x +
V S

1 2 da =
V

2 1 d3 x +
S

2 1 da.

3. A point charge q is brought to a distance z0 > R measured from the centre of a conducting sphere of radius R (as in Sec. 2.8). The conductor, however, is not grounded, but kept in careful isolation; as it initially supports a zero net charge on its surface, its nal net surface charge must also vanish. The surface of the conductor is therefore at a constant potential V = 0. Find V as a function of q , z0 , and R. 4. Suppose that at a spherical boundary surface S of radius R, the electrostatic potential is assigned the values (R, , ) = V sin cos cos , where V is a constant. Suppose that there are no charges outside this surface, and that S is the only boundary present in the problem. Find the potential for r > R. 5. In this problem we consider the electrostatic potential due to a uniformly charged disk located outside a grounded, conducting sphere. The conductor has a radius R and the potential over its surface is zero. The disk has an inner radius a > R and an outer radius b > a. It is convenient to place the disk in the equatorial plane of the spherical coordinate system, so that its charge density is described by (x) = 3q ( 2) , 3 2 b a3 a < r < b,

with q denoting the disks total charge.

a) Verify that q is indeed the disks total charge. b) Calculate the potential in the region R < r < a. c) Calculate the surface charge density () on the surface of the conducting sphere, and provide a plot of this function. (Use the values q = 1, R = 1, a = 1.2, and b = 2; make sure to include a sucient number of terms in the innite sum over .) d) Show that the total charge on the surface of the conductor is given by q = 3q (a + b)R . 2 a2 + ab + b2

6. We generalize the situation considered in Sec. 2.8 by placing an additional charge q at position z0 = z0 z below the conducting sphere. The rst charge +q is still at +z0 = z0 z , the conducting sphere still has a radius R, and the sphere is still grounded. Calculate the vector p, the total electric dipole moment of this charge distribution. Express your result in terms of q , z0 , and R.

46

Electrostatics 7. Suppose that the upper hemisphere of a conducting sphere is maintained at a potential +V , while its lower hemisphere is maintained at a potential V . (The two hemispheres are separated by a thin layer of insulating material, but there is no need to include this into the mathematical model.) The conducting surface has a radius R, and there are no charges outside the conductor. Show that the potential outside the conducting sphere can be expressed as

(r, ) = V
=1

a (R/r)+1 P (cos ),

where the numerical coecients a are zero if is even, and equal to (2 + 1 1) 0 P (x) dx if is odd. Calculate a explicitly for = (1, 3, 5). 8. A conducting sphere of radius R is maintained at a potential V . It is surrounded by a thin spherical shell of radius a > R on which there is a surface charge density proportional to cos . The volume density of charge is given by 3p cos (r a), (r, ) = 4a3 where p is a constant it is the dipole moment of the charge distribution on the thin shell. a) Verify that the shells total charge is zero, and that p = pz is the dipole moment vector of this charge distribution. b) Calculate the potential in the region R < r < a. c) Calculate the potential in the region r > a. d) What is the total charge of this conguration? What is the total dipole moment vector? (The total charge conguration includes the free charge on the spherical shell and the induced charge on the surface of the conductor.) e) Calculate the density of charge () on the surface of the conductor. What is the total charge on the conductor? 9. A dielectric sphere of radius R (the inner boundary) is maintained at a potential = V0 cos . It is surrounded by a thin spherical shell of radius a > R with a density of charge given by = where p is a constant. Find the value of V0 which makes the scalar potential (x) vanish for r > a. 10. Calculate the electrostatic potential (x) outside a thin, hollow sphere of radius R on which there is a surface density of charge proportional to 1 + 3 cos(2). You may assume that there are no boundaries, and that the volume charge density is given by (r, ) = where is a constant. What are the components of the quadrupole moment tensor for this charge distribution? 5 1 + 3 cos(2) (r R), 8R4 3p cos (r a), 4a3

Chapter 3 Magnetostatics
3.1 Equations of magnetostatics

For time-independent situations, the part of Maxwells equations that involve the magnetic eld are decoupled from those involving the electric eld, and they can be dealt with separately. The equations of magnetostatics are Poissons equation for the vector potential, 2 A(x) = 0 j (x), (3.1.1) the Lorenz (or Coulomb) gauge condition A = 0, and the relation between potential and eld, B = A. (3.1.3) (3.1.2)

We recall that the current density must satisfy the statement of charge conservation, j = 0. Because Eq. (3.1.1) is very similar to the electrostatic equation (2.1.1), the tools introduced in Chapter 2 to solve Poissons equation can be taken over directly to magnetostatics. And because we already have gained much experience solving such equations, it will suce to keep our discussion of magnetostatics quite brief. In particular, it will be sucient to solve Eq. (3.1.1) in the absence of boundaries. We therefore write the solution as A(x) = recalling the results of Sec. 1.8. 0 4 j (x ) 3 d x, |x x | (3.1.4)

3.2

Circular current loop

[The material presented in this section is also covered in Sec. 5.5 of Jacksons text.] As our rst application we calculate the magnetic eld produced by a current I owing inside a circular loop of radius R. We place the loop in the x-y plane, at z = 0. To evaluate the integral of Eq. (3.1.4) for this situation we will use spherical coordinates and express the Greens function as in Eq. (2.6.12), 1 = |x x |
4 r< Y ( , )Ym (, ), +1 m 2 + 1 r>

(3.2.1)

=0 m=

where r< is the lesser of r and r , while r> is the greater of the two. 47

48

Magnetostatics The density of the negative charge carriers inside the wire is given by (x) = (r R) ( 2 ), R (3.2.2)

where is the magnitude of the linear charge density (minus the wires total charge per unit length); the total charge is given by (x) d3 x = 2R. The charge carriers move in the negative direction so as to produce a positive current. Their velocity vector is = v sin x v = v + cos y , (3.2.3)

is a unit vector that points in the direction of increasing . Its expression in where terms of the constant vectors x and y is obtained by noting that the position vector of a point in the plane is given by r = r cos x + r sin y , so that a vector pointing in the direction is r / = r sin x + r cos y ; dividing by r produces the . From Eqs. (3.2.2) and (3.2.3) we obtain the current density, normalized vector j (x) = v = I (r R) ( 2 ), R (3.2.4)

where I v is the current. To obtain the vector potential we substitute Eqs. (3.2.1) and (3.2.4) into Eq. (3.1.4). For concreteness we take x to be outside the loop, so that r > r , r< = r , and is not a constant vector: its r> = r. We must pay attention to the fact that expression in terms of x and y must be substituted inside the integral before integration is attempted. It is convenient to consider the complex combination Ay iAx , which is generated by jy ijx = I i I (r R) ( (cos + i sin ) (r R) ( 2 ) = Re 2 ); R

we have now expressed this in terms of the primed variables. We notice that the dependence of the current density is contained entirely in the exponential factor ei . After evaluating the r and integrals the vector potential becomes Ay iAx = 0 I 4 R 4 R+2 Ym (, ) 2 + 1 r+1
2 0

2 0

Ym ( 2 , )ei d .

The integral evaluates to Ym ( 2 , 0) the potential simplies to Ay iAx = 0 I 4

eim ei d = 2Ym ( 2 , 0)m,1 , and


+1

=1

R 8 2 , 0) Y,1 ( 2 2 + 1 r

Y,1 (, );

the sum begins at = 1 because there is no spherical harmonic Y01 . The sphericalharmonic functions can then be expressed as Y,1 (, ) = Y,1 (, 0)ei , and this allows us to extract the real and imaginary parts of the right-hand side. We obtain Ax = and Ay = 0 I 4 0 I 4

=1

R 8 2 , 0) Y,1 ( 2 2 + 1 r

+1

Y,1 (, 0) sin

=1

R 8 2 Y,1 ( 2 , 0) r 2 + 1

+1

Y,1 (, 0) cos ,

and these combine to form the vector 0 A(x) = I 4 8 2 R Y,1 ( 2 , 0) r 2 + 1


+1

Y,1 (, 0).

(3.2.5)

=1

3.3

Spinning charged sphere

49

We see that the vector potential is expressed as an expansion in powers of R/r. When r R the vector potential is well approximated by the leading term of the expansion, given by = 1. In this case we have Y11 (, 0) = 3/(8 ) sin and Eq. (3.2.5) becomes 0 1 . A(x) (R2 I ) 2 sin (3.2.6) 4 r We will write this in a dierent form. First we dene the vector m = (R2 I )z , (3.2.7)

which points in the z direction and whose magnitude is the product of the current I and the area R2 enclosed by the loop; this vector is known as the magnetic moment of the current distribution. Then we notice that z x/r = (y/r)x + (x/r)y = . So Eq. (3.2.6) can be written as sin sin x + sin cos y = sin A(x) 0 m x , 4 r3 (3.2.8)

and this shall be our nal expression. The curl of A gives the magnetic eld. Apart from a common factor of (0 /4 )|m|, its x component is equal to z (x/r3 ) = 3xz/r5 , its y component is equal to z (y/r3 ) = 3yz/r5 , and nally, its z component is equal to x (x/r3 )+y (y/r3 ) = 2/r3 3x5 /r5 3y 2 /r5 = 1/r3 + 3z 2 /r5 . These results combine to form the vector 3z x/r5 z /r3 , and multiplying by |m| gives 3(m x)x/r5 m/r3 . The magnetic eld is therefore )r m 0 3(m r , (3.2.9) B (x) 4 r3 where r = x/r is a unit vector that points in the radial direction. Notice that this expression for the magnetic eld is very similar to what was obtained in Sec. 2.3 for the electric eld of a dipole: Edipole = 1 3(p r )r p . 3 40 r

So the magnetic eld far outside a current loop has a dipolar behaviour, with m playing the role of dipole moment.

3.3

Spinning charged sphere

As a second application we calculate the magnetic eld produced by a hollow sphere of radius R, uniformly charged, that is rotating with an angular velocity = z . Here the charge density is given by (x) = q (r R), 4R2 (3.3.1)

where q is the spheres total charge. To calculate the velocity vector v of a point x on the surface of the sphere, we consider a narrow strip at an angle with respect to the z axis (see Fig. 3.1). A point on this strip moves in the direction; in a time dt the angle increases by d = dt. Since this point is at a distance R sin from the z axis, the length traveled is dl = (R sin ) d. The speed is therefore = R sin , or v = dl/dt = (R sin ), and the velocity vector is v = v v (x) = x. q (r R). sin 4R (3.3.2)

This is the velocity of the charge distribution, and the current density is given by j (x) = (3.3.3)

50

Magnetostatics

R sin

Figure 3.1: A hollow sphere rotating at an angular velocity around the z axis. = sin x We recall that + cos y . Again it is convenient to consider the complex combination Ay iAx , which is generated by q sin ei (r R). jy ijx = 4R The angular function happens to be equal to 8/3Y11 ( , ). Substituting this and Eq. (3.2.1) into the expression for Ay iAx that comes from Eq. (3.1.4), we nd that the angular integration is nonzero if and only if = m = 1; the double sum over and m therefore contains a single term. The radial integration reduces to r< r r<R 2 , 3 2 2 (r R)r dr = R /r r>R r > 0 and we obtain Ay iAx = 0 q 4 4 4R 3 8 Y11 (, ) 3 r R3 /r2 .

The angular function within square brackets is sin ei , and the vector potential is A= 0 q sin sin x + sin cos y 4 3R r R3 /r2 .

After multiplication by r the vector within brackets becomes y x + xy =z x. If we let 1 1 m q R 2 z = qR2 , (3.3.4) 3 3 we nd that the vector potential inside the sphere is given by Ain (x) = 0 m x 4 R3 (r < R), (3.3.5)

while the potential outside the sphere is Aout (x) = 0 m x 4 r3 (r > R). (3.3.6)

3.4

Multipole expansion of the magnetic eld

51

Notice that these are exact results no approximations were involved in the calculation of the vector potential. Notice also that Aout takes the same form as in Eq. (3.2.8), except for the fact that here, the magnetic moment m is a dierent vector. From this we conclude that the magnetic eld Bout outside the sphere has a dipolar behaviour, and that it takes the form of Eq. (3.2.9). To calculate the magnetic eld inside the sphere we express the vector potential as Ain = (0 /4 )(m/R3 )(y x + xy ) and take its curl. Its x and y components are both zero, and its z component is proportional to x x + y y = 2. We arrive at Bin (x) =

0 2m . (3.3.7) 4 R3 This states that inside the sphere, the magnetic eld is constant and points in the same direction as the magnetic moment (which corresponds to the rotation axis).

3.4

Multipole expansion of the magnetic eld

[The material presented in this section is also covered in Sec. 5.6 of Jacksons text.] For both the current loop (Sec. 3.2) and the rotating sphere (Sec. 3.3) we found that the vector potential at a large distance from the current distribution was given by 0 m x , (3.4.1) A(x) = 4 r3 where the vector m, known as the magnetic moment, characterizes the distribution of current. And in both cases we found the magnetic eld to be described by 0 3(m r )r m , (3.4.2) 3 4 r where r = x/r; this is a dipolar eld. Here we will show that these expressions are in fact very general: they are the leading terms in an expansion of the vector potential and the magnetic eld in powers of 1/r. This multipole expansion is valid for points x that lie outside of the current distribution. Let the distribution of current be conned to a volume V bounded by a surface S . We take x to be outside of V , while x is necessarily inside. We examine the vector potential at a large distance r r , starting from the exact expression B (x) = A(x) = 0 4 j (x ) 3 d x. |x x |

We recall that the current density satises j = 0, and we recognize that there is no current owing across S ; this implies that j da = 0, where da is the surface element on S . As we shall show below, these equations imply the integral statements j (x ) d3 x = 0
V

(3.4.3)

1 x j (x ) d3 x . (3.4.4) r x j (x ) d3 x = r 2 V V To simplify the vector potential we expand 1/|x x | in powers of 1/r. We could do this very systematically by involving the addition theorem for spherical harmonics (Sec. 2.6), but it will be good enough to keep only a small number of terms. For this we proceed directly: |x x | = = (x x ) (x x ) r2 2x x r2

and

x /r + = r 1 2r = r 1r x /r + ,

52 so that

Magnetostatics

1 1 = 1+r x /r + . |x x | r 0 1 4 r 1 r2

Substituting this into the exact expression for A(x) gives A(x) = j (x ) d3 x +
V

r x j (x ) d3 x + .

The rst term on the right-hand side vanishes by virtue of Eq. (3.4.3), and the second term is altered by Eq. (3.4.4). We have A(x) = 1 0 1 r 4 r2 2 x j (x ) d3 x + ,

and if we dene the magnetic moment of the current distribution by m= 1 2 x j (x ) d3 x , (3.4.5)

then we recover Eq. (3.4.1). We see that indeed, A(x) = 1 0 m r +O 3 4 r2 r

is a very general expression, and that the magnetic moment of an arbitrary distribution of current is given by Eq. (3.4.5). It is a straightforward exercise to show that for the current loop of Sec. 3.2, Eq. (3.4.5) reproduces Eq. (3.2.7). Similarly, Eq. (3.4.5) reduces to Eq. (3.3.4) for the rotating sphere of Sec. 3.3. We still have to establish the identities of Eqs. (3.4.3) and (3.4.4). We begin with the z component of Eq. (3.4.3), in which we drop the primes on the integration variables. Consider (z j ) = z ( j ) + j z = jz , which follows because j = 0 and z = z . Integrating this equation over V gives jz d3 x =
V V

(z j ) d3 x =

z j da,

where we have invoked Gauss theorem. But the right-hand side is zero because no current crosses the surface S , and we have V jz d3 x = 0. Proceeding similarly with the x and y components, we arrive at Eq. (3.4.3). To establish Eq. (3.4.4) requires a bit more work, but the idea is the same. We consider the z component of this equation, in which we drop the primes on the integration variables. (At the same time we replace the vector r , which is constant with respect to the primed variables, by an arbitrary constant vector e.) We have e x jz d3 x = ex xjz d3 x + ey
V V

yjz d3 x + ez
V

zjz d3 x

for the left-hand side of the identity. Consider now (xz j ) = xz ( j ) + j (xz ) = xjz + zjx , which integrates to (xjz + zjx ) d3 x =
V S

xz j da = 0,

3.5

Problems

53

giving rise to the integral identity xjz d3 x = yjz d3 x = zjx d3 x.


V

Similarly, zjy d3 x.
V V

On the other hand, (z 2 j ) = z 2 ( j ) + j z 2 = 2zjz , and this yields zjz d3 x = 0.


V

Using these results, we go back to the left-hand side of Eq. (3.4.4) and write e x jz d3 x = 1 ex 2 1 + ey 2 1 = ex 2 xjz d3 x
3 V

zjx d3 x
V

yjz d x

zjy d3 x
V

1 (zjx xjz ) d3 x + ey 2 V

(yjz zjy ) d3 x.

Inside the rst integral we recognize the y component of the vector x j , while the second integral contains its x component. So e x jz d3 x = 1 ex (x j )y d3 x ey 2 V 1 = e (x j ) d3 x , 2 V z (x j )x d3 x

and we have arrived at the right-hand side of Eq. (3.4.4).

3.5

Problems

1. Show that for the current loop of Sec. 3.2, Eq. (3.4.5) reproduces Eq. (3.2.7). And show that Eq. (3.4.5) reduces to Eq. (3.3.4) for the rotating sphere of Sec. 3.3. 2. A uniformly charged sphere of radius R is rotating around the z axis with an angular velocity . Take the density of charge inside the sphere to be 3q/(4R3 ) (for r < R), where q is the spheres total charge. Calculate m, the spheres magnetic moment vector. 3. Calculate the vector potential A(x) inside the spinning sphere of the preceding problem. Express it in terms of the spheres magnetic moment vector m. 4. Prove that for a current loop of arbitrary shape that lies in a plane, Eq. (3.4.5) gives |m| = (current)(area enclosed by loop), and that the direction of m is normal to the plane.

54

Magnetostatics

Chapter 4 Electromagnetic waves in matter


4.1 Macroscopic form of Maxwells equations

[The material presented in this section is also covered in Sec. 6.6 of Jacksons text.] Our goal in this chapter is to describe how electromagnetic waves propagate in dielectric materials. Our rst task will be to reformulate Maxwells equations in a way that conveniently incorporates the electromagnetic response of the medium.

4.1.1 Microscopic and macroscopic quantities


Maxwells equations, as they were written down in Chapter 1, are the fundamental equations of classical electrodynamics; in principle they require no modication even in the presence of a nontrivial medium. But it is convenient to write them in a dierent form, in which the charges that are free to move within the medium are distinguished from the charges that are tied to the mediums molecules, and in which the microscopic uctuations of all the variables involved have been smoothed out. The resulting form of Maxwells equations involves only macroscopic densities of free charges and currents, and macroscopic electromagnetic elds. To distinguish them from macroscopic quantities, the fundamental, microscopic variables will be assigned dierent symbols in this chapter. The microscopic density of charge (including both free and bound charges) is denoted , while the microscopic density of current (including both free and bound currents) is denoted J . The microscopic electric eld is denoted e, and the microscopic magnetic eld is denoted b. These quantities are related by the microscopic Maxwell equations, e = 1 , 0 0, (4.1.1) (4.1.2) (4.1.3) (4.1.4)

b = b e+ = 0, t e = 0 J , b 0 0 t

which are the same as in Chapter 1. The macroscopic density of free charge will be denoted , while the macroscopic density of free current will be denoted j . These are obtained from the microscopic quantities by removing the contribution from the bound charges (those that are tied to the molecules and become part of the mediums electromagnetic response) and by macroscopic smoothing; the precise relationships will be given below. The macroscopic elds will be denoted E , D , 55

56

Electromagnetic waves in matter

B , and H . The macroscopic electric eld E is obtained from e by macroscopic smoothing, and D accounts for the electric polarizability of the medium; precise relationships will be given below. Similarly, the macroscopic magnetic eld B is obtained from b by macroscopic smoothing, and H accounts for the magnetization of the medium; precise relationships will also be given below.

4.1.2 Macroscopic smoothing


Microscopic quantities uctuate widely over atomic distances, and we would like to smooth out these uctuations to dene corresponding macroscopic quantities. For example, the charge density switches from zero to innity whenever a point charge is encountered, and we would like to replace this by a quantity that varies smoothly to reect only macroscopic changes of density. One simple-minded way of smoothing out a microscopic quantity f is to average it over a volume V (x) centered at some position x; this volume is imagined to be microscopically large (so that all microscopic uctuations will be averaged over) but macroscopically small (so that variations over relevant macroscopic scales are not accidentally discarded). The averaging operation is an integration of f over the volume V (x): f (x) = V 1 f (x ) d3 x , where the domain of integration consists of all points x that lie within V (x). If the volume is a sphere of radius R, then the condition is |x x | < R, and f (x) can be expressed as 1 V f (x )(R |x x |) d3 x ,

with a now innite domain of integration. In this averaging operation it is convenient to replace the sharply dened function V 1 (R |x x |) by a better-behaved function w(x x ) that smoothly goes from V 1 when x = x to zero when x is far away from x. Such a function, whose detailed specication is not required, will be called a smoothing function; a specic choice would be a Gaussian function of width R centered at x x = 0. The smoothing function is required to be normalized, in the sense that w(x x ) d3 x = 1. We dene the macroscopic average of a microscopic quantity f (t, x) to be F (t, x) f (t, x) f (t, x )w(x x ) d3 x , (4.1.5)

where w is a suitable smoothing function (assumed to be normalized) and the integration is over all space. The property that w does not have sharp boundaries implies that the operations of averaging and dierentiation commute. For example, a f (t, x) = a f (t, x) , (4.1.6)

and the same is true of time derivatives. This important result follows by straightforward manipulations. Consider, for example, f x w(x x ) d3 x x = f (t, x ) w(x x ) d3 x x f = w(x x ) d3 x x f . = x = f (t, x )

In the second line we use the fact that w depends on the dierence x x , so that a derivative with respect to x is minus a derivative with respect to x . In the third line

4.1

Macroscopic form of Maxwells equations

57

we integrate by parts, noting that the boundary terms at innity do not contribute because w vanishes there; the result is by denition the average of the microscopic quantity f /x. Following the usage introduced in Eq. (4.1.5), we dene the macroscopic electric eld to be E (t, x) = e(t, x) , (4.1.7) and the macroscopic magnetic eld to be B (t, x) = b(t, x) . (4.1.8)

Because dierentiation and averaging are commuting operations, the macroscopic form of Maxwells equations are E B B E+ t E B 0 0 t = = 1 , 0 0, (4.1.9) (4.1.10) (4.1.11) (4.1.12)

= 0, = 0 J .

The homogeneous equations (4.1.10) and (4.1.11) are the same as their microscopic version, but the inhomogeneous equations (4.1.9) and (4.1.12) involve a macroscopic averaging of the source terms.

4.1.3 Macroscopic averaging of the charge density


We distinguish between the charges that can move freely within the medium the free charges and the charges that are tied to the molecules the bound charges. We will consider the bound charges to be part of the medium and remove them from the right-hand side of Maxwells equations, which will then involve the free charges only. The inuence of the bound charges will thus be taken to the left-hand side of the equations; it will be described by new macroscopic quantities, P (electric polarizability of the medium) and M (magnetization of the medium). From these we will dene the new elds D 0 E + P and H B /0 M . In Fig. 4.1 we have a picture of the charge distribution in the medium. Let xA (t) be the position vector of the free charge qA labeled by A, and let vA (t) = dxA /dt be its velocity vector. Let xn be the position vector of the centre of mass of the molecule labeled by n; for simplicity we assume that the molecules do not move within the medium. Finally, let xB (t) be the position vector of the bound charge qB labeled by B within a given molecule, relative to the centre of mass of this molecule, and let vB (t) = dxB /dt be its velocity vector. The density of free charge is free (t, x) =
A

qA x xA (t) .

(4.1.13)

The density of bound charge within the molecule labeled by n is n (t, x) =


B

qB x xn xB (t) .
B qB

(4.1.14) = 0. The (4.1.15)

We assume that each molecule is electrically neutral, so that density of bound charge is then bound (t, x) =
n

n (t, x),

58

Electromagnetic waves in matter

xA xn xn

xB

Figure 4.1: The left panel shows the free charges (small solid disks) and molecules (large open circles) in the medium. The right panel shows the distribution of bound charges within a given molecule. and the total density of charge is = free + bound . We rst calculate the macroscopic average of the molecular charge density. We have n = =
B

n (t, x )w(x x ) d3 x qB (x xn xB )w(x x ) d3 x

=
B

qB w(x xn xB ).

The function w(x xn xB ) is peaked at x = xn + xB . Because the scale R over which w varies is very large compared with the intra-molecular displacement xB , we can approximate w by Taylor-expanding about x xn : 1 a b w(x xn xB ) = w(x xn ) xa B a w (x xn ) + xB xB a b w (x x ) + , 2 where to avoid confusing the notation we make the vectorial index on xB a superscript instead of a subscript. (Recalling that repeated indices are summed over, we have that xa B a stands for xB .) Substituting this into the previous expression yields n =
B

qB w

qB xa B a w +

1 2

b qB xa B xB a b w + ,

where w now stands for w(x xn ). We simplify this equation by recalling that each molecule is electrically neutral, so that the rst sum vanishes. In the second term we recognize the expression for the molecular dipole moment, pn = B qB xB , or pa n (t) =
B

qB xa B (t),

(4.1.16)

and in the third term we recognize the expression for the molecular quadrupole moment, b Qab qB xa (4.1.17) n (t) = 3 B (t)xB (t).
B

Notice that this actually diers from the denition made in Sec. 2.10: here our quadrupole moment is not dened as a tracefree tensor. Combining these results

4.1 gives

Macroscopic form of Maxwells equations

59

1 ab n (t, x) = pa n (t)a w (x xn ) + Qn (t)a b w (x xn ) + 6

(4.1.18)

for the molecular charge density. The macroscopic average of the density of bound charge is obtained by summing n over all molecules. This gives bound = pa n a w + pa nw
n

1 6

Qab n a b w + Qab n w + Qab n (x xn ) + ,

= a = a

1 + a b 6

pa n (x
n

1 xn ) + a b 6

because w w(x xn ) is the macroscopic average of (x xn ). The quantities Pa (t, x) and Qab (t, x) Qab n (t) (x xn ) macroscopic quadrupole density (4.1.20)
n

pa n (t) (x xn ) macroscopic polarization

(4.1.19)

have a straightforward physical interpretation. Consider, for example, Eq. (4.1.19). The quantity within the averaging brackets is n pa n (x xn ), the sum over all molecules of the product of each molecules dipole moment with a -function centered at the molecule. This is clearly the microscopic density of molecular dipole moments, and Pa is its macroscopic average. So the vector P (t, x), called the macroscopic polarization, is nothing but the macroscopic average of the dipole-moment density of the mediums molecules. Similarly, Qab (t, x) is the macroscopic average of the molecular quadrupole-moment density. In terms of these we have 1 bound (t, x) = a Pa (t, x) + a b Qab (t, x) + 6 for the macroscopic average of the density of bound charge. The average of the free-charge density is simply (t, x) free (t, x) qA (x xA ) , (4.1.22) (4.1.21)

and the macroscopic average of the total charge density is therefore 1 = P + a b Qab + . 6 Substituting this into Eq. (4.1.9) yields 1 a 0 Ea + Pa b Qab + 6 = . (4.1.23)

The vector within the large brackets is the macroscopic eld D (t, x), and we arrive at D = , (4.1.24)

60 with the denition

Electromagnetic waves in matter

1 (4.1.25) Da = 0 Ea + Pa b Qab + . 6 Notice that this macroscopic Maxwell equation involves only the density of free charges on the right-hand side, and that apart from a factor of 0 , the eective electric eld D is the electric eld E augmented by the electrical response of the medium, which is described by the molecular dipole-moment and quadrupolemoment densities. In practice the term involving Qab in Eq. (4.1.25) is almost always negligible, and we can write D = 0 E + P .

4.1.4 Macroscopic averaging of the current density


Following the denitions of Eqs. (4.1.13)(4.1.15) we introduce the current densities Jfree (t, x) Jn (t, x) Jbound (t, x) =
A

qa vA (t) x xA (t) , qB vB (t) x xn xB (t) , Jn (t, x),

(4.1.26) (4.1.27) (4.1.28)

=
B

=
n

and the total microscopic current density is J = Jfree + Jbound . The macroscopic average of the molecular current density is
a Jn

=
B

a vB w(x xn xB ) a q B vB w a b q B vB xB b w + ,

=
B

where w stands for w(x xn ) in the second line. In the rst term we recognize a the time derivative of the molecular dipole moment, dpa n /dt = (d/dt) B qB xB = a B qB vB . To put the second term in a recognizable form we dene the molecular magnetic moment by 1 qB (xB vB ) (4.1.29) mn (t) = 2
B

and we note that the vector w mn has components w mn


a

= = =

1 2 1 2

b a b qB xa B vB vB xB b w b a b qB xa B vB + vB xB b w a b q B vB xB b w

1 d ab b w Q 6 dt n
a b B qB xB xB

a b q B vB xB b w,

recalling the denition Qab n =3 Combining these results gives


a = Jn

for the molecular quadrupole moment.

1 pa w Qab b w + w mn t n 6 n

+ ,

(4.1.30)

and summing over n yields


a Jbound

1 pa n w b 6

Qab n w +

mn w
n

+ ,

4.1

Macroscopic form of Maxwells equations t 1 pa n (x xn ) b 6


a

61 Qab n (x xn )

+ Introducing M (t, x)

mn (x xn )

+ .

mn (t) (x xn ) macroscopic magnetization

(4.1.31)

as the macroscopic average of the magnetic-moment density of the mediums molecules, we nally arrive at the expression
a = Jbound

1 Pa Qab t 6

+ ( M )a +

(4.1.32)

for the macroscopic average of the density of bound current. The average of the free-current density is simply j (t, x) Jfree (t, x) qA vA (x xA ) , (4.1.33)

and the macroscopic average of the total current density is therefore J =j+ D 0 E + M + , t (4.1.34)

having used Eq. (4.1.25). Substituting Eq. (4.1.34) into Eq. (4.1.12) gives B /0 M D = j. t

The vector within brackets is the macroscopic eld H (t, x), and we arrive at H with the denition H = B /0 M . (4.1.36) Notice that this macroscopic Maxwell equation involves only the density of free currents on the right-hand side, and that apart from a factor of 0 , the eective magnetic eld H is the magnetic eld B diminished by the magnetic response of the medium, which is described by the molecular magnetic-moment density. D =j t (4.1.35)

4.1.5 Summary macroscopic Maxwell equations


The macroscopic Maxwell equations are D = = 0, (4.1.37) (4.1.38) (4.1.39) (4.1.40)

B B E+ t D H t

= 0, = j.

Here is the macroscopic free-charge density, and j is the macroscopic free-current density. The elds D and H are dened by D = 0 E + P (4.1.41)

62 and H= where P =
n

Electromagnetic waves in matter

1 B M, 0 pn (t) (x xn )

(4.1.42) (4.1.43)

is the macroscopic polarization (pn = B qB xB is the molecular dipole moment) and M= mn (t) (x xn ) (4.1.44)
n

the macroscopic magnetization (mn = moment).

1 2

B qB x B

vB is the molecular magnetic

4.2

Maxwells equations in the frequency domain

[The material presented in this section is also covered in Sec. 7.1 of Jacksons text.] We now begin our discussion of electromagnetic waves propagating in a dielectric medium. We suppose that the medium is not magnetized, so that we can set M = 0 is our equations. We further suppose that the waves are propagating in the absence of sources, so that we can also set = 0 and j = 0. Maxwells equations therefore reduce to D B B E+ t D B 0 t = = 0 0, (4.2.1) (4.2.2) (4.2.3) (4.2.4)

= 0, = 0,

where D = 0 E + P . We want to form plane-wave solutions to these equations, and examine how the waves interact with the medium. We shall assume that the medium is uniform, isotropic, and unbounded, and leave unexplored issues of reection and transmission at boundaries. To begin we will look for monochromatic waves oscillating with an angular frequency . We therefore set E (t, x) D (t, x) P (t, x) B (t, x) (x)eit , = E (x)eit , = D (x)eit , = P (x)eit , = B (4.2.5) (4.2.6) (4.2.7) (4.2.8)

with the understanding that only the real parts of these complex elds have physical meaning. The restriction to monochromatic waves is not too severe. Because the eld equations are linear (assuming, as we shall verify below, that the relation between D and E is itself linear), solutions with dierent frequencies can be added to form wave-packet solutions. For example, we can express the electric eld as 1 E (t, x) = 2

(, x)eit d, E

(4.2.9)

(, x); this is a Fourier representation of in terms of the frequency-domain eld E the electric eld. The inverse transformation is (, x) = 1 E 2

E (t, x)eit dt,

(4.2.10)

4.3

Dielectric constant

63

(, x), P (, x), B (, x) can be dened a similar and frequency-domain elds D way. The frequency-domain elds satisfy the equations D B i B E = = 0 0, (4.2.11) (4.2.12) (4.2.13) (4.2.14)

+ i0 D B and we have the relation

= 0, = 0,

= 0 E +P . D

(4.2.15)

These are the sourceless Maxwell equations in the frequency domain. They can only be solved once an explicit relationship (called a constitutive relation) is introduced and the electric eld E . The constitutive relation between the polarization P describes the mediums response to an applied electric eld, and its derivation is based on complicated molecular processes. While quantum mechanics is required for a proper treatment, in the next section we shall consider a simple classical model that nevertheless captures the essential features.

4.3

Dielectric constant

[The material presented in this section is also covered in Sec. 7.5 of Jacksons text.] We suppose that the medium is suciently dilute that the electric eld felt itself, the eld exerted by other by any given molecule is just the applied eld E molecules being negligible. We assume that in the absence of an applied eld, the molecular dipole moment is zero, at least on average. When an electric eld is applied, however, some of the molecular charges (the electrons) move away from their equilibrium positions, and a dipole moment develops. The displacement of charge qB from its equilibrium position is denoted B (t). This charge undergoes a motion that is governed by the applied electric eld (which tends to drive B away from zero) and the intra-molecular forces (which tend to drive the charge back to its equilibrium position). We model these forces as harmonic forces, all sharing the same natural frequency 0 . What we have, therefore, is a system of simple harmonic oscillators of natural frequency 0 that are driven at a frequency by an applied force. Since the motion of the charges will produce electromagnetic radiation, energy will gradually be removed from the oscillators (to be carried o by the radiation), and the oscillations will be damped; we incorporate this eect through a phenomenological damping parameter > 0. The equations of motion for the charge qB are therefore
2 B + B + 0 (, xn )eit , mB = qB E

(4.3.1)

where overdots indicate dierentiation with respect to t. Notice that the applied electric eld, which is macroscopic and varies slowly over intra-molecular distances, is evaluated at the molecules centre of mass. To solve the equations of motion B eit and we substitute this into in the steady-state regime we write B (t) = Eq. (4.3.1). This yields B ( ) = qB /mB 2 2 i E (, xn ) 0 (4.3.2)

for the frequency-domain displacement.

64

Electromagnetic waves in matter

The molecules dipole moment is pn = B qB (xB + B ), where xB is the equilibrium position of charge qB . Because the dipole moment vanishes at equilibrium, this reduces to pn = B qB B . Substituting Eq. (4.3.2) then gives p n =
B 2 0 2 qB /mB (, xn ). E 2 i

2 Since all the contributing charges qB are electrons, we can write B qB /mB = 2 Ze /m, where Z is the number of electrons per molecule, e the electronic charge, and m the electrons mass. We therefore arrive at

p n ( ) =

Ze2 1 (, xn ) E 2 m 0 2 i

(4.3.3)

for the molecular dipole moment in the frequency domain. The polarization vector is then P =
n 2

p n (x xn ) (, xn ) (x xn ) E (x xn ) .

= =

1 Ze 2 2 i m 0

Ze2 1 (, x) E 2 m 0 2 i

The quantity within the averaging brackets is the microscopic molecular density. Its macroscopic average is therefore the mediums density N (the number of molecules per macroscopic unit volume), which we take to be uniform. We have arrived at
2 1 (, x) = Ze N (, x) P E 2 m 0 2 i

(4.3.4)

for the frequency-domain polarization. We recall that Z is the number of electrons per molecule, that N is the molecular number density, and that e and m are the electrons charge and mass, respectively. We notice that the relationship between and E is linear, as was anticipated in Sec. 4.2. P 0 E +P and E also is linear. We express it as The relationship between D
2 1 = 0 1 + Ze N , D E 2 0 m 0 2 i

or as (, x) = ( )E (, x), D (4.3.5) having introduced the (complex, and frequency-dependent) dielectric constant
2 p ( ) =1+ 2 0 0 2 i

(4.3.6)

and the mediums plasma frequency p = Ze2 N 0 m


1/ 2

(4.3.7)

Equation (4.3.6) is the product of a simplistic classical model. But a proper quantum-mechanical treatment would produce a similar result, except for the fact

4.4

Propagation of plane, monochromatic waves

65

0.5

1.5 w

2.5

Figure 4.2: Real and imaginary parts of the dielectric constant, as functions of /0 . To t both curves on the same graph we removed 1 from Re(/0 ). The real part of the dielectric constant goes through 1 at resonance. The imaginary part of the dielectric constant achieves its maximum at resonance. that the electrons actually oscillate with a discrete spectrum of natural frequencies j and damping coecients j . A more realistic expression for the dielectric constant is then ( ) fj 2 = 1 + p 2 2 i 0 j j j where fj is the fraction of electrons that share the same natural frequency j . In the rest of this chapter we will continue to deal with the simple expression of Eq. (4.3.6). It is good to keep in mind, however, that the model is a bit crude. In Fig. 4.2 we plot the real and imaginary parts of the dielectric constant. You will notice that for most frequencies, when is not close to 0 , Re(/0 ) increases with increasing , and Im(/0 ) is very small; this typical behaviour is associated with normal dispersion. When is very close to 0 , however, the electrons are driven at almost their natural frequency, and resonance occurs. Near resonance, then, Re(/0 ) decreases with increasing , and Im(/0 ) is no longer small; this behaviour is associated with anomalous dispersion. In realistic models the number of resonances is larger than one: there is one resonance for each natural frequency j . At low frequencies the dielectric constant behaves as /0 1 + (p /0 )2 , and is therefore a constant slightly larger than unity. At high frequencies it behaves as /0 1 (p / )2 , and therefore approaches unity from below.

4.4

Propagation of plane, monochromatic waves

[The material presented in this section is also covered in Sec. 7.1 of Jacksons text.] After making the substitution of Eq. (4.3.5), the frequency-domain Maxwell equations (4.2.11)(4.2.14) become E = 0 (4.4.1)

66

Electromagnetic waves in matter B i B E = 0, (4.4.2) (4.4.3) (4.4.4)

+ i( )0 E B

= 0, = 0.

eit We recall that the time-domain elds are reconstructed, for example, as E = E if the wave is monochromatic, or as in Eq. (4.2.9) if the wave possesses a spectrum of frequencies. and B satIt is easy to show that as a consequence of Eqs. (4.4.1)(4.4.4), both E isfy a frequency-domain wave equation. For example, taking the curl of Eq. (4.4.3) gives ) i B = 0; ( E

) = substituting Eq. (4.4.4) and making use of the vectorial identity ( E 2 ( E ) E puts this in the form (, x) = 0, 2 + k 2 E where k2 2 ( ) . c2 0 (4.4.5)

(4.4.6)

We have introduced a complex wave number k ( ), and we recall that c (0 0 )1/2 is the speed of light in vacuum. Similar manipulations return (, x) = 0, 2 + k 2 B (4.4.7)

the statement that the magnetic eld also satises a wave equation. To see how monochromatic electromagnetic waves propagate in a dielectric . Then suppose that or B , and call it medium, consider any component of E = (z ). The the wave is a plane wave that propagates in the z direction, so that wave equation simplies to d2 (, z ) = 0. + k2 dz 2 (4.4.8)

To be concrete we suppose that the second term within the large brackets is small. Taking the square root gives k= or k = kR + ikI , where kR =
2 2 2 ) p (0 1+ 2 2 )2 + 2 2 c 2 (0 2 p 1+ 2 2 i c 2 0

= eikz . To help A right-moving wave is described by the particular solution identify the physical properties of this wave we substitute Eq. (4.3.6) into Eq. (4.4.6) and get 2 p 2 k2 = 2 1 + 2 . (4.4.9) c 0 2 i

(4.4.10)

is the real part of the complex wave number, while kI =


2 c 2 2 c 2 (0 )2 + 2 2

4.4

Propagation of plane, monochromatic waves

67

is its imaginary part. The right-moving wave is therefore described by (, z )eit = ekI z ei(tkR z) . (t, z ) (4.4.11)

From this expression we recognize that the wave is indeed traveling in the positive z direction, with a speed c vp = (4.4.12) kR ( ) n( ) known as the phase velocity; n ckR / is the mediums index of refraction. From our previous expression for kR we obtain n=1+
2 2 )2 + 2 2 2 (0 2 2 2 ) p (0

2 and we see that n 1 + 1 2 (p /0 ) > 1 for low frequencies, so that vp < c. On the 1 2 other hand, n 1 2 (p / ) < 1 for high frequencies, so that vp > c. The phase velocity of the wave can therefore exceed the speed of light in vacuum! This strange fact does not violate any cherished relativistic notion, because a monochromatic wave does not carry information. To form a signal one must modulate the wave and superpose solutions with dierent frequencies. As we shall see, in such situations the wave always travels with a speed that does not exceed c. Another feature of Eq. (4.4.10) is that the wave is an exponentially decreasing function of z provided that kI is positive, which is guaranteed to be true if > 0. A positive damping constant means that some of the waves energy is dissipated by the oscillating electrons (which emit their own electromagnetic radiation), and the wave must therefore decrease in amplitude as it travels through the medium. It is interesting to note that in the case of lasers and masers, the electronic oscillations actually reinforce the wave and lead to an exponential amplication instead of an attenuation; in such media the damping constant is negative. Having constructed solutions to Eqs. (4.4.5) and (4.4.7), we should examine what constraints the full set of equations (4.4.1)(4.4.4) place on these solutions. We write (, z ) = e E E0 ( )eik()z (4.4.13)

and (, z ) = B bB0 ( )eik()z , (4.4.14) where e and b are real unit vectors, and where E0 and B0 are complex amplitudes. Substituting these expressions into Eqs. (4.4.1) and (4.4.2) informs us that the unit vectors are both orthogonal to z , e z = bz = 0; (4.4.15)

the elds are therefore transverse to the direction in which the wave propagates. Equation (4.4.3) then produces B0 b = (k/ )E0 (z e ), which implies B0 ( ) = and b=z e . (4.4.17) These equations state that the magnetic eld is orthogonal to the electric eld, and that its amplitude is linked to that of the electric eld. Finally, Eq. (4.4.4) gives us B0 (z b) = (k/ )E0 e , which merely conrms Eqs. (4.4.16) and (4.4.17). k ( ) E0 ( ) (4.4.16)

68

Electromagnetic waves in matter

4.5

Propagation of wave packets

[The material presented in this section is also covered in Secs. 7.8 and 7.9 of Jacksons text.]

4.5.1 Description of wave packets


In this section we shall assume, for simplicity, that ( ) is real: we work away from resonances and ignore dissipative eects. We let k ( ) be the positive solution to Eq. (4.4.6). A wave packet is obtained by superposing solutions to Eq. (4.4.8) with dierent frequencies . Including both left- and right-moving waves, we write 1 (t, z ) = 2

a+ ( )eik()z + a ( )eik()z eit d,

(4.5.1)

where a ( ) are complex amplitudes that determine the shape of the wave packet. Alternatively, and more conveniently for our purposes, we can instead superpose solutions with dierent wave numbers k . Dening (k ) to be the positive solution to Eq. (4.4.6), we write 1 (t, z ) = 2

A+ (k )ei(k)t + A (k )ei(k)t eikz dk,

(4.5.2)

and it is not too dicult to show that the decomposition of Eq. (4.5.2) is equivalent to that of Eq. (4.5.1). The complex amplitudes A (k ) are determined by the initial conditions we wish to impose on (t, z ) and its time derivative. It follows from Eq. (4.5.2) that
1 A+ (k ) + A (k ) eikz dk, (0, z ) = 2 and Fouriers inversion theorem gives

1 A+ (k ) + A (k ) = 2 (0, z ) = 1 2 and inversion gives


(0, z )eikz dz.

On the other hand, dierentiating Eq. (4.5.2) with respect to t produces i (k ) A+ (k ) A (k ) eikz dk,

1 i (k ) A+ (k ) A (k ) = 2 Solving for A (k ), we nally obtain A (k ) = 1 1 2 2


(0, z )eikz dz.

(0, z )

1 (0, z ) eikz dz. i (k )

(4.5.3)

Equation (4.5.3) then simplies to

The procedure to follow to construct a wave packet is therefore to choose an ini (0, z ), then calculate A (k ) using tial conguration by specifying (0, z ) and Eq. (4.5.3), and nally, obtain (t, z ) by evaluating the integral of Eq. (4.5.2). In the following discussion we will choose the initial conguration to be timesymmetric, in the sense that (0, z ) 0. (4.5.4) A(k ) A (k ) = 1 1 2 2

(0, z )eikz dz,

(4.5.5)

and we have an equal superposition of left- and right-moving waves.

4.5

Propagation of wave packets

69

4.5.2 Propagation without dispersion


Let us rst see what the formalism of the preceding subsection gives us when there is no dispersion, that is, when n c|k |/ is a constant independent of k . We then have (k ) = vp |k |, (4.5.6)

where vp = c/n is the waves phase velocity. For time-symmetric initial data, Eq. (4.5.2) reduces to 1 (t, z ) = 2

A(k ) ei|k|vp t + ei|k|vp t eikz dk.

The factor within the square brackets is 2 cos(|k |vp t) = 2 cos(kvp t), and we can therefore replace the previous expression with 1 (t, z ) = 2 But from Eq. (4.5.5) we have A(k ) = and substituting this gives (t, z ) = = or 1 2 1 2

A(k ) eik(z+vp t) + eik(zvp t) dk.

1 1 2 2

(0, z )eikz dz ,

dz (0, z )

1 2

eik(z+vp tz ) + eik(zvp tz ) dk

dz (0, z ) (z + vp t z ) + (z vp t z ) ,

1 1 (0, z + vp t) + (0, z vp t). (4.5.7) 2 2 The rst term on the right-hand side is (half) the initial wave packet, (0, z ), translated in the z direction by vp t; it represents that part of the wave packet that travels undisturbed in the negative z direction with a speed vp . The second term, on the other hand, is the remaining half of the wave packet, which travels toward the positive z direction with the same speed vp . The main features of wave propagation without dispersion are that the wave packet travels with the phase velocity vp = /k and that its shape stays unchanged during propagation. (t, z ) =

4.5.3 Propagation with dispersion


Let us now return to the general case of propagation with dispersion. Here is a nonlinear function of k , and for time-symmetric initial data we have 1 (t, z ) = 2

A(k ) ei(k)t + ei(k)t eikz dk.

To be concrete, suppose that A(k ) is sharply peaked around k = k0 , with k0 some arbitrary wave number. [A proper treatment would have A(k ) be a symmetric function peaked about both k = k0 and k = k0 . Unlike what we shall nd below, this would produce a real wave function (t, z ). It is not dicult to make this change, but we shall not do so here.] Suppose further that (k ) varies slowly near k = k0 , so that it can be approximated by
(k ) 0 + 0 (k k0 ) = (0 0 k0 ) + 0 k,

(4.5.8)

70

Electromagnetic waves in matter

1.04

1.03

1.02

1.01

0.5

1.5 w

2.5

Figure 4.3: Group velocity vg in a dielectric medium, in units of c, as a function of /0 .


(d/dk )(k0 ). Then the wave packet can be expressed where 0 (k0 ) and 0 as

1 2

A(k ) eik(z+0 t) ei(0 0 k0 )t + eik(z0 t) ei(0 0 k0 )t dk A(k )eik(z+0 t) dk



1 = ei(0 0 k0 )t 2

+e or where

i(0 0 k 0 )t

1 2

A(k )eik(z0 t) dk, (4.5.9)

(t, z ) = ei(0 0 k0 )t (0, z + vg t) + ei(0 0 k0 )t (0, z vg t), vg =

d (k0 ) (4.5.10) dk is known as the group velocity. From Eq. (4.5.9) we learn that in the presence of dispersion, the wave packet travels with the group velocity vg instead of the phase velocity vp = k0 /0 . And the phase factors indicate that the shape of the wave suers a distortion during propagation. These are the main features of wave propagation in a dispersive medium. If the dispersion relation is expressed as k = n( )/c, in terms of the index of refraction n( ), then dk/d = c1 [n + (dn/d )] and 1 vg = . c n + (dn/d ) For the dielectric medium described in Sec. 4.3, n2 = 1 +
2 (0 2 2 (0 2 ) p , 2 )2 + 2 2

(4.5.11)

and the group velocity is plotted in Fig. 4.3. You will notice that vg is smaller than c for most frequencies, but that it seems to exceed c near resonance. This

4.5

Propagation of wave packets

71

phenomenon is entirely articial: when is close to 0 it is no longer true that is a slowly varying function of k , and the approximations involved in dening the group velocity are no longer valid. Furthermore, the imaginary part of , together with the dissipative eects it represents, cannot be neglected near resonance. The bottom line is that the very notion of group velocity is no longer meaningful near a resonance.

4.5.4 Gaussian wave packet


As a concrete example of wave propagation with dispersion we consider an initial conguration given by 2 2 (0, z ) = ez /(2L ) cos(k0 z ); (4.5.12) this is an oscillating wave (with wave number k0 ) modulated by a Gaussian envelope of width L. Once more we take the initial data to be time-symmetric, and we (0, z ) = 0. In order to be able to carry out the calculations we shall take impose the dispersion relation to be 1 (k ) = 1 + 2 k 2 , 2 (4.5.13)

where is a minimum frequency and an arbitrary length scale. This dispersion relation is not particularly realistic, but its quadratic form will allow us to evaluate the integrals that will appear in the course of our analysis. These will have the form of Gaussian integrals, and we quote the standard result

eaz

+bz +c

dz =

b2 exp +c . a 4a

(4.5.14)

Here the constants a, b, and c can be complex, but Re(a) must be positive for the integral to converge. We rst calculate the Fourier amplitudes associated with Eq. (4.5.12). Equation (4.5.5) gives A(k ) = = 1 1 2 2 1 1 4 2

(0, z )eikz dz

ez

/(2L2 )

ei(kk0 )z + ei(k+k0 )z dz.

For the rst integral we have a = 1/(2L2 ), b = i(k k0 ), c = 0, and b2 /(4a) = 1 (k k0 )2 L2 ; for the second integral we simply make the substitution k0 k0 2 in this result. Equation (4.5.14) then gives A(k ) =
2 2 1 L 1 (kk0 )2 L2 e 2 + e 2 (k+k0 ) L . 4

(4.5.15)

We see that A(k ) is a symmetric function of k , and that it is peaked at both k = k0 and k = k0 . The width of the distribution around both peaks is equal to 1/L. The wave packet at times t = 0 is given by Eq. (4.5.2), 1 = 2 1 A(k )ei[kz(k)t] dk + 2

A(k )ei[kz+(k)t] dk,

which we may also express as 1 = 2 1 A(k )ei[kz(k)t] dk + 2


A(k )ei[kz(k)t] dk,

72

Electromagnetic waves in matter

0.5 initial wave 0 -0.5

-40 0.6

-20

0 z

20

40

no dispersion dispersion

0.4

propagated wave

0.2

-0.2

-0.4 -40 -20 0 z 20 40

Figure 4.4: Wave propagation with a high-frequency dispersion relation (k ) = 2 + c2 k 2 )1/2 , calculated numerically with the help of a Fast Fourier Transform (p routine. The upper panel shows the initial wave prole, described by Eq. (4.5.12). The lower panel shows two versions of the propagated wave. The rst (lower curve) is obtained without dispersion by setting p = 0; this wave propagates with speed c without altering its shape. The second (higher curve the wave function is translated upward for ease of visualization) is obtained with dispersion; this wave spreads out as it travels with speed vg < c. Notice the drop in amplitude as the initial wave packet splits into an equal superposition of left- and right-moving waves.

4.6

Problems

73

having changed the variable of integration to k in the second integral. Because A(k ) = A(k ) and (k ) = (k ), we have = =
1 A(k )ei[kz(k)t] dk + complex conjugate, 2 2 Re A(k )ei[kz(k)t] dk. 2

Substituting Eq. (4.5.15) then gives L = Re 2 2


e 2 (kk0 )

L2 i[kz (k)t]

dk + (k0 k0 ) .

Taking into account Eq. (4.5.13), we see that the argument of the exponential function is a second-order polynomial in k , which we write as ak 2 + bk + c, with 1 2 2 (L2 + it2 ), b = k0 L2 + iz , and c = (it + 1 a= 2 2 k0 L ). With this we have b2 (z ik0 L2 )2 = 2 , 4a 2L (1 + itf 2 ) where f /L, and Eq. (4.5.14) yields =
1 2 2 1 (z ik0 L2 )2 eit 2 k0 L + (k0 k0 ) . Re (1 + itf 2 )1/2 exp 2 2 2L (1 + itf 2 )

It is then a matter of straightforward algebra to bring this to its nal form, (t, z ) = (z vg t)2 1 ei[k0 z(k0 )t] Re (1 + itf 2 )1/2 exp 2 2 2L (1 + itf 2 ) + (k0 k0 ) , where f = /L and d (k0 ) (4.5.17) dk is the group velocity at k = k0 for the dispersion relation of Eq. (4.5.13). Equation (4.5.16) informs us that the wave packet retains its Gaussian shape as it propagates through the medium, but that its width increases with time: vg = 2 k0 = L(t) = L 1 + (f 2 t)2 . (4.5.18) (4.5.16)

The wave therefore spreads out, and this is a typical feature of propagation with dispersion. In addition, we see that while the oscillations travel at the phase velocity vp = (k0 )/k0 , the centre of the Gaussian peak travels at the group velocity vg ; this result was expected after the discussion of the preceding subsection. A more realistic example of wave dispersion (calculated numerically) is displayed in Fig. 4.4. Notice that the features discussed in the preceding paragraphs are present in this example as well: the wave packet spreads out as it propagates with a group velocity vg that is smaller than c.

4.6

Problems

1. (This problem is adapted from Jacksons problem 7.19.) A wave packet in one dimension has an initial shape given by (0, z ) = f (z )eik0 z , where f (z ) is the modulation envelope; the wave packet is assumed to be time-symmetric. For each of the forms f (z ) given below, calculate the wave-number spectrum |A(k )|2 of the wave packet, sketch | (0, z )|2 and |A(k )|2 , calculate the root-mean-square deviations z and k from the means (dened in terms of | (0, z )|2 and |A(k )|2 , respectively), and test the inequality z k 1 2.

74 a) f (z ) = e|z|/2 ; b) f (z ) = e
2 2

Electromagnetic waves in matter

z /4

c) f (z ) = 1 |z | for |z | < 1 and f (z ) = 0 for |z | > 1; d) f (z ) = 1 for |z | < 1 and f (z ) = 0 for |z | > 1. 2. We have seen that the relationship between the frequency-domain elds D is and E (, x) = ( )E (, x), D where
2 p ( ) =1+ 2 . 0 0 2 i

Prove that as a consequence of this, the relationship between the time-domain elds D and E is D (t, x) = 0 E (t, x) +
2 p 0

e /2 sin( ) E (t , x) d ,

2 1 2 . This reveals that D at time t depends on the electric where 0 4 eld at all previous times: the relationship is nonlocal in time. The exponential factor, however, indicates that delays longer than 1/ do not matter very much: only the electric elds recent history is important.

Chapter 5 Electromagnetic radiation from slowly moving sources


5.1 Equations of electrodynamics

In this chapter and the next we will explore some of the mechanisms by which electromagnetic radiation is produced. While Chapter 6 will be concerned with relativistic situations, here we shall restrict our attention to charge and current distributions whose internal velocities are small compared with the speed of light. We recall the main equations of electrodynamics, as we have formulated them in Chapter 1. We have wave equations for the scalar and vector potentials, (t, x) = and A(t, x) = 0 j (t, x), where 1 2 + 2 c2 t2 (5.1.2) 1 (t, x) 0 (5.1.1)

(5.1.3)

is the wave operator and c (0 0 )1/2 is the speed of light. For Eqs. (5.1.1) and (5.1.2) to be equivalent to Maxwells equations, the potentials must satisfy the Lorenz gauge condition, 1 A+ 2 = 0. (5.1.4) c t Maxwells equations imply that the charge and current densities satisfy the continuity equation, + j = 0, (5.1.5) t which states that charge is locally conserved. In the absence of boundaries, the wave equations have solutions (t, x) = and A(t, x) = 1 40 0 4 (t |x x |/c, x ) 3 d x |x x | j (t |x x |/c, x ) 3 d x; |x x | (5.1.6)

(5.1.7)

these are retarded solutions, which incorporate the physically appropriate causal relationship between sources and elds. 75

76

Electromagnetic radiation from slowly moving sources From the potentials we obtain the elds via the relations E= A t (5.1.8)

and B = A. (5.1.9) And nally, we recall from Sec. 1.4 that the elds carry energy. The electromagnetic eld energy density is given by (t, x) = 1 1 0 E 2 + B2 2 20 (5.1.10)

and the energy ux vector (Poynting vector) is S (t, x) = 1 E B. 0 (5.1.11)

5.2

Plane waves

The simplest situation featuring the production of electromagnetic waves involves monochromatic plane waves. These are generated by a planar distribution of charges, all oscillating in the plane with the same frequency . We take the charge distribution to be uniform in the x-y plane, so that the charge density is given by (t, x) = (z ), (5.2.1) where is a constant surface density. Notice that the charge density is actually time independent. The resulting scalar potential will also not depend on time, and the electromagnetic waves will be described entirely by the vector potential. We take each charge to be oscillating in the x direction with a velocity v (t, x) = v0 x cos t. (5.2.2)

Notice that the velocity vector is actually independent of position in the plane, and that x is actually an arbitrary direction within the plane; we could pick another direction by rotating the x-y coordinates around the z axis. The current density is j (t, x) = v0 (z )x cos t, (5.2.3)

and we will calculate the associated vector potential without approximations. In principle we might proceed by substituting Eq. (5.2.3) into Eq. (5.1.7), but it is easier to start instead with the equivalent form A(t, x) = 0 4 j (t , x ) (t t |x x |/c) 3 dt d x , |x x |

in which we substitute Eq. (5.2.3) expressed in terms of primed variables. We have A(t, x) = where (t t , x) = (z ) 0 v0 x 4 dt (t t , x) cos t , (t t |x x |/c) 3 d x |x x |

is an integral that we must now evaluate.

5.2

Plane waves Integration over dz is immediate: = (t t (x x )2 + (y y )2 + z 2 /c) dx dy ,

77

(x x )2 + (y y )2 + z 2

or =

(t t

x2 + y 2 + z 2 /c)

x2 + y 2 + z 2

dx dy

after translating the origin of the (x , y ) coordinates to the point (x, y ); this is allowed, because the domain of integration is the whole x -y plane. To integrate this we make the change of variables x = cos , y = sin . This gives, after writing dx dy = dd and integrating over d, (t t 2 + z 2 /c) d. = 2 2 + z 2 0 The easiest way to integrate over the -function is to make its argument be the integration variable. So let s = 2 + z 2 /c t, where t t t . Because is nonnegative, we have that s is bounded below by |z |/c t. We also have d d , = 2 ds = 2 2 c (s + t) c +z and the integral becomes

= 2c
|z |/ct

(s) ds,

or (t t , x) = 2c(t t |z |/c), where (s) is the step function, which is unity when s > 0 and zero otherwise. Substituting this result into our previous expression for the vector potential, we obtain A(t, x) = = 0 (v0 )(2c)x 4 0 2v0 c x 4

t|z |/c

dt (t |z |/c t ) cos t cos t dt .

The dt integral evaluates to 1 sin (t |z |/c), apart from an ambiguous constant associated with the integrals innite lower bound. Since this constant will never appear in expressions for the electric and magnetic elds, we simply set it to zero. Our nal expression for the vector potential is therefore A(t, z ) = where E0 sin (t |z |/c)x , (5.2.4)

1 0 v0 c (5.2.5) 2 has the dimensions of an electric eld. For z > 0 we have that sin (t |z |/c) = sin (t z/c) = sin k (z ct), where k = /c is the wave number, and we see that the wave travels in the positive z direction with speed c. For z < 0 we have instead sin (t |z |/c) = sin (t + z/c) = sin k (z + ct), and the wave now travels in the negative z direction. Thus, Eq. (5.2.4) describes a plane wave that propagates away from z = 0, the place where it is generated; the wave has a frequency , a wave number k = /c, and it travels with the speed of light. E0 =

78

Electromagnetic radiation from slowly moving sources

The time-dependent part of the electric eld is obtained from Eqs. (5.1.8) and (5.2.4): E (t, z ) = E0 cos (t |z |/c)x . (5.2.6) We see that the electric eld behaves as a plane wave that travels away from z = 0 with the speed of light, and that it points in the x direction, which is perpendicular to the direction of propagation. The magnetic eld, on the other hand, is obtained by taking the curl of Eq. (5.2.4). This involves dierentiating |z | with respect to z , which is +1 for z > 0 and 1 if z < 0. Denoting this function by (z ) (z ) (z ), we have E0 (z ) cos (t |z |/c)y . (5.2.7) B (t, z ) = c The magnetic eld also is a plane wave that travels away from z = 0 at the speed of light, and it points in the y direction, which is perpendicular to both the direction of propagation and the direction of the electric eld. Notice that E and B are both proportional to cos (t |z |/c): the elds are in phase. Notice also that |B | is smaller than |E | by a factor 1/c. To nish o our discussion of plane electromagnetic waves, we calculate the elds energy density and ux vector. From Eqs. (5.1.10), (5.2.6), and (5.6.7) we obtain 2 (t, z ) = 0 E0 cos2 (t |z |/c), (5.2.8) and from Eq. (5.1.11) we get S (t, z ) = c(z )(t, z )z . (5.2.9)

This shows that the electromagnetic eld energy travels with the wave, in the positive or negative z direction, away from the source, and with the speed of light. Let us summarize the main points: The plane wave travels in the (positive or negative) z direction with speed c. The electric and magnetic elds are perpendicular to z and to each other. The elds are in phase. Their magnitudes are related by |B |/|E | = 1/c. The electromagnetic eld energy travels with the wave. These properties are generic: they are shared by all electromagnetic waves. The only caveat is that in general, the direction of propagation is not uniform: the constant vector z goes into a vector n that may change from point to point.

5.3

Spherical waves the oscillating dipole

[The material presented in this section is also covered in Secs. 9.1 and 9.2 of Jacksons text.]

5.3.1 Plane versus spherical waves


Any function of the form (t, z ) = f (t z/c), (5.3.1) where f ( ) is an arbitrary function of = t z/c, is a solution to the wave equation = 0. If = t z/c, the plane wave travels in the positive z direction with speed c; if instead = t + z/c, the wave travels in the negative z direction. To see that Eq. (5.3.1) does indeed form a valid solution to the wave equation, we note that

5.3

Spherical waves the oscillating dipole

79

/t = f ( ) while /z = f ( )/c, where a prime indicates dierentiation with respect to ; substitution into the wave operator yields = c2 f + c2 f = 0, as required. Plane waves are not a very good approximation to realistic waves, which originate in a localized region of space and propagate outward with a decreasing amplitude. A better approximation that incorporates these features is the spherical wave 1 (5.3.2) (t, r) = f (t r/c), r where f (u) is an arbitrary function of u = t r/c. This is also a solution to = 0. Here the wavefronts are spherical instead of planar, and the wave propagates outward in the radial direction; its amplitude decays as 1/r, giving rise to an intensity that goes as 1/r2 . That Eq. (5.3.2) forms a valid solution to the wave equation follows by direct substitution: In the absence of an angular dependence the wave operator reduces to 1 2 1 = 2 2 + 2 r2 . c t r r r , while we have r2 /r = Dierentiating with respect to t yields 2 /t2 = r1 f rf /c f and 1 1 1 1 2 r = f + 2 rf + f = 2 rf , r r c c c c where an overdot indicates dierentiation with respect to u. All this gives = 0, as required. Note that a similar calculation would reveal that = r1 f (t + r/c) also is a solution to the wave equation; this wave travels inward instead of outward.

5.3.2 Potentials of an oscillating dipole


The simplest electromagnetic spherical wave is produced by an oscillating dipole located at the origin of the coordinate system. Here and in the following subsections we will calculate the potentials and elds produced by this dipole, and examine the nature of the corresponding electromagnetic radiation. In Sec. 2.3 we saw that the charge density of a point dipole could be expressed as (x) = p (x), where p was a constant dipole moment vector. Here we introduce a time dependence by making p oscillate with angular frequency : p p cos t. The charge density of an oscillating dipole is then (t, x) = p (x) cos t. (5.3.3)

Notice that in this equation, p is still a constant vector, which we align with the z axis: p = pz . A time-varying charge density must be associated with a current density j , because j = = p (x) sin t = p (x) sin t . t

The simplest solution to this equation is j (t, x) = p (x) sin t, (5.3.4)

and this is the current density of an oscillating dipole. We rst compute the vector potential produced by the current density of Eq. (5.3.4). Once more it is convenient to start with the expression A(t, x) = 0 4 j (t , x ) (t t |x x |/c) 3 dt d x . |x x |

80

Electromagnetic radiation from slowly moving sources

After substituting Eq. (5.3.5) we obtain A(t, x) 0 p 4 0 = p 4 = (t t |x x |/c) 3 d x |x x | (t t r/c) dt sin t , r dt sin t (x )

sin (t r/c) 0 p . (5.3.5) 4 r This has the structure of a spherical wave, with its expected dependence on u = t r/c and its amplitude decaying as 1/r. We may now compute the scalar potential. This calculation is a bit more laborious, and we start with A(t, x) = (t, x) = 1 40 (t , x ) (t t |x x |/c) 3 dt d x . |x x |

or

After substituting Eq. (5.3.3) we get (t, x) = where (t t , x) = (x ) 1 p 40 dt (t t , x) cos t , (t t |x x |/c) 3 d x |x x |

is a vectorial integral that we must now evaluate. Integration by parts allows us to write = (x )

(t t |x x |/c) 3 d x. |x x |

Because the dependence on x is contained entirely in |x x |, and because |x x | = (x x )/|x x |, evaluating the gradient gives = (t t |x x |/c) 1 (t t |x x |/c) x x 3 + d x |x x |2 c |x x | |x x | (t t r/c) 1 (t t r/c) x + , = r2 c r r (x ) (t t , x) = r (t t r/c) 1 (t t r/c) , + r2 c r

or

where r = x/r. Substituting this result into our previous expression for the scalar potential, we obtain (t, x) = 1 pr 40 dt (t t r/c) 1 (t t r/c) cos t , + r2 c r

and integrating against the -functions gives (t, x) = cos (t r/c) sin (t r/c) 1 . pr 40 r2 c r

One must be careful with the second term: a minus sign comes from the integration by part needed to convert (t t r/c) into an actual -function, but this minus

5.3

Spherical waves the oscillating dipole

81

sign is canceled out because the prime on the -function indicates dierentiation with respect to t t r/c instead of t . We express our nal result as = near + wave , where near (t, x) = and wave (t, x) = 1 pr cos(u) 2 40 r 1 sin(u) pr , 40 c r (5.3.6)

(5.3.7)

(5.3.8)

and where u = t r/c is retarded time. The scalar potential contains two terms. Of these, only wave has the mathematical structure of a spherical wave. The other term, near , has the structure of a static dipole eld, 1 pr , dipole (x) = 2 40 r to which a time dependence cos(u) is attached. Unlike wave , this term does not represent electromagnetic radiation. Instead, near describes the eld of a point dipole, and this eld follows the dipoles time dependence with the usual delay: the cos t dependence of the dipole is replaced by cos (t r/c). As the name indicates, near dominates when r is small, while wave dominates when r is large. The scalar potential becomes purely radiative when r is suciently large that near can be neglected; the region of space for which this is true is called the wave zone. On the other hand, the region of space for which r is suciently small that wave can be neglected in front of near is called the near zone. In the near zone we must have 1/r2 /(cr), or r 1 c = = k 2 (near zone), (5.3.9)

where k is the wave number and the wavelength. In the near zone, u = t r/c t and the delay contained in the retarded time is hardly noticeable the near-zone eld responds virtually instantaneously to changes in the source. In the wave zone we have instead 1/r2 /(cr), or r c 1 = = k 2 (wave zone). (5.3.10)

Here the time delay is important, and the scalar potential behaves as a true spherical wave. In the region of space where r is comparable to c/ , between the near zone and the wave zone (a region that is sometimes called the induction, or intermediate, zone), near and wave are comparable, and the scalar potential behaves neither as an instantaneously-changing dipole eld nor as a spherical wave.

5.3.3 Wave-zone elds of an oscillating dipole


Let us calculate the electric and magnetic elds in the wave zone. We use the expressions sin(u) 0 (5.3.11) A(t, x) = p 4 r and 1 sin(u) (t, x) = pr , (5.3.12) 40 c r where u = t r/c is retarded time. Notice that Eq. (5.3.11) is exact, while Eq. (5.3.12) is valid in the wave zone only.

82

Electromagnetic radiation from slowly moving sources To do this calculation we will need to compute the gradient of (t, r) sin(u) . r

We have = (/r)r = (/r)r and sin(u) cos(u) = , r r2 c r because u/r = 1/c. But since we are working in the wave zone, the rst term can be neglected and we obtain the useful approximation sin(u) cos(u) = r . r c r (5.3.13)

We might also need the gradient of p r . To see what this is, consider its x component: p x x px x = 2 + , x p r r r r p (p r )r . (5.3.14) r The fact that this is of order 1/r means that the gradient of p r will be negligible in our forthcoming calculations. The electric eld is related to the potentials by Eq. (5.1.8), (p r ) = E= A . t so that

Substituting Eqs. (5.3.11) and (5.3.12), we obtain E= 1 sin(u) sin(u) 0 2 cos(u) . p + (p r ) + (p r ) 4 r 40 c r r

Using Eq. (5.3.14) shows that the rst term within the square brackets is of order 1/r2 and can be neglected in the wave zone, while using Eq. (5.3.13) gives E= cos(u) 1 2 0 2 cos(u) (p r ) p r . 2 4 r 40 c r

Collecting the terms and recalling that 0 0 = 1/c2 brings this to its nal form Ewave (t, x) = 0 2 cos(u) p (p r )r . 4 r (5.3.15)

This has the structure of a spherical wave, and we notice that Ewave is transverse to the direction in which the wave propagates: Ewave r = 0. (5.3.16)

If, for concreteness, we let p point in the z direction, then p = pz , r = sin cos x + sin sin y + cos z , p r = p cos , and p (p r )r = p sin (cos cos x + , where is a unit vector that points in the direction cos sin y sin z ) = p sin of increasing . This yields Ewave (t, x) = for the electric eld. cos(u) 0 2 p sin 4 r (5.3.17)

5.3

Spherical waves the oscillating dipole

83

The magnetic eld is obtained by taking the curl of the vector potential. We have sin(u) 0 p , B (t, x) = 4 r and Eq. (5.3.13) gives us Bwave (t, x) = 0 2 cos(u) (r p) . 4 c r (5.3.18)

The magnetic eld also is a spherical wave, and like the electric eld it is transverse to the direction of propagation: Bwave r = 0. (5.3.19)

Furthermore, we see from Eqs. (5.3.15) and (5.3.18) that the elds are mutually orthogonal, because [p (p r )r ] (r p) = p (r p) (p r )r (r p) = 0, since the vector r p is perpendicular to both p and r . So the elds satisfy Ewave Bwave = 0 (5.3.20)

in the wave zone. For p = pz we have r p = p sin ( sin x + cos y ) = , and Eq. (5.3.18) becomes p sin Bwave (t, x) = 0 2 cos(u) . p sin 4 c r (5.3.21)

5.3.4 Poynting vector of an oscillating dipole


1 We now calculate S = 0 E B , the energy ux vector associated with the electromagnetic radiation produced by the oscillating dipole. It is easy to check that in the wave zone, the electric and magnetic elds are related by

Ewave = cBwave r .

(5.3.22)

This follows because the magnetic eld is proportional to r p, and (r p) r = (r r )p (r p)r = p (p r )r , which gives the direction of the electric eld. Using once more the same vectorial identity, the Poynting vector is S= c c Br B = (B B )r (B r )B . 0 0

The second term vanishes because B is orthogonal to r , and we arrive at Swave (t, x) = c 2 . Bwave (t, x) r 0 (5.3.23)

This shows that the energy is transported in the radial direction, the same direction in which the wave propagates. For p = pz we can substitute Eq. (5.3.21) into Eq. (5.3.23) and obtain S = = c 0 0 4
2

4 p2 sin2 cos2 (u) r c2 r2

0 p2 4 sin2 cos2 (u) r . 16 2 c r2

The energy ux oscillates in time, and after averaging over a complete wave cycle we get 0 p2 4 sin2 Swave = r . (5.3.24) 32 2 c r2

84

Electromagnetic radiation from slowly moving sources

From the angular dependence of this expression we see that most of the energy is emitted near the equatorial plane ( = 2 ), and that none of the energy propagates along the z axis ( = 0 or = ). The average of any function f of retarded time u T over a complete wave cycle is dened by f (u) T 1 0 f (u) du, where T = 2/ is the oscillation period. Using this rule we get cos2 (u) = 1 2 , and this gives Eq. (5.3.24). The total power radiated (energy per unit time) is obtained by integrating S (energy per unit time per unit area) over any closed surface S . The averaged power is then P = S da. (5.3.25)
S

It is simplest to take S to be a sphere of constant r. Then da = r r2 sin dd, and Eq. (5.3.25) becomes 0 p2 4 P = sin3 d. 2 32 2 c 0 The integral evaluates to 4/3, and we obtain P = 0 2 4 p . 12c (5.3.26)

5.3.5 Summary oscillating dipole


We have seen that in the wave zone (r c/ ), the elds of an oscillating dipole are given by cos (t r/c) 0 E (t, r, , ) = 2 p sin 4 r and 0 2 cos (t r/c) B (t, r, , ) = p sin . 4 c r These have the structure of a spherical wave, which travels outward at the speed of light. They are produced by a charge distribution whose dipole moment vector is given by p(t) = p cos(t)z . The eld amplitudes are related by |B |/|E | = 1/c, and both elds are in phase. The electric eld is directed along = cos cos x + cos sin y sin z while the magnetic eld is directed along = sin x + cos y . The elds are mutually orthogonal, and they are both transverse to r = sin cos x + sin sin y + cos z , the direction in which the wave is traveling. This is also the direction in which the eld energy is transported. The angular prole of the energy-ux (Poynting) vector is described by sin2 , so that most of the energy is emitted in the directions orthogonal to the dipole.

5.4

Electric dipole radiation

[The material presented in this section is also covered in Sec. 9.2 of Jacksons text.] We now consider a much more general situation, in which electromagnetic radiation is produced by an arbitrary distribution of charges and currents, with an arbitrary time dependence (not necessarily oscillating with a single frequency ). Our only restrictions are that

5.4

Electric dipole radiation the source is conned to a bounded region V of space; the charges are moving slowly.

85

These conditions will allow us to formulate useful approximations for the behaviour of the electric and magnetic elds. Although the present context is very general, we will see that the results to be derived in this section are very close to those obtained in the preceding section.

5.4.1 Slow-motion approximation; near and wave zones


To make the slow-motion approximation precise, and to dene near and wave zones in the general case, we introduce the following scaling quantities: rc tc vc c c characteristic length scale of the charge and current distribution, characteristic time scale over which the distribution changes, rc characteristic velocity of the source, tc 2 characteristic frequency of the source, tc 2c ctc characteristic wavelength of the radiation. c

The characteristic length scale is dened such that if (t, x ) and j (t, x ) are the charge and current densities, respectively, then |x | is at most of order rc throughout the source. This implies that the distribution of charge and current is localized 3 within a region whose volume is of the order of rc . The characteristic time scale is dened such that /t is of order /tc throughout the source. If, for example, oscillates with a frequency , then tc 1/ and c . The slow-motion approximation means that vc = rc /tc is much smaller than the speed of light: vc c. (5.4.1) This condition gives us rc = vc tc ctc = c , or rc c . (5.4.2)

The source is therefore conned to a region that is much smaller than a typical wavelength of the radiation. We can now dene the near and wave zones: near zone: wave zone: |x x | c c = , 2 c c c = . |x x | 2 c (5.4.3) (5.4.4)

Because rc c , the wave zone can also be dened by the condition r c /(2 ). We will compute the potentials and elds in these zones, using Eqs. (5.4.1)(5.4.4) to simplify our expressions.

5.4.2 Scalar potential


We begin by calculating the scalar potential, (t, x) = 1 40 (t |x x |/c, x ) 3 d x, |x x |

86

Electromagnetic radiation from slowly moving sources

where V is the region of space occupied by the source. In the near zone we can treat |x x |/c as a small quantity and Taylor-expand the charge density about the (t)2 + we have current time t. Because (t ) = (t) (t) + 1 2 1 (t |x x |/c) = (t) (t)|x x | + , c where an overdot indicates dierentiation with respect to t. Relative to the rst term, the second term is of order |x x |/(ctc ) = |x x |/c , and by virtue of Eq. (5.4.3), this is small in the near zone. The third term would be smaller still, and we neglect it. We then have (t, x) = = 1 40 1 40 (t, x ) 3 d x |x x | (t, x ) 3 d x |x x | 1 (t, x ) d3 x + c 1 d (t, x ) d3 x + . c dt V

The second term vanishes, because it involves the time derivative of the total charge (t, x ) d3 x , which is conserved. We have obtained near (t, x) = 1 40 |x x |2 (t, x ) 1+O |x x | 2 c d3 x . (5.4.5)

We see that the near-zone potential takes its usual static expression, except for the small correction of order |x x |2 /2 c and the fact that charge density depends on time. The time delay between the source and the potential has disappeared, and what we have is a potential that adjusts instantaneously to the changes within the distribution. The electric eld it produces is then a time-changing electrostatic eld. This near-zone eld does not behave as radiation. To witness radiative eects we must go to the wave zone. Here |x x | is large and we can no longer Taylor-expand the density as we did previously. Instead we must introduce another approximation technique. We use the fact that in the wave zone, r is much larger than r , so that |x x | = = (x x ) (x x ) r2 2x x + r2
1/ 2

2 2 = r 1 2r x /r + O(rc /r ) 2 2 = r 1r x /r + O(rc /r ) 2 = rr x + O(rc /r).

This gives (t |x x |/c) where is retarded time and is of order Let us now Taylor-expand the charge density about the retarded time u instead of the current time t. We have 1 (t |x x |/c) = (u) + (u) r x + + , c where an overdot now indicates dierentiation with respect to u. We see that relative to r x , the term is of order rc /r 1 and can be neglected. Now relative
2 rc /r.

= (t r/c + r x /c + /c) = (u + r x /c + /c), u = t r/c (5.4.6)

5.4

Electric dipole radiation

87

to (u), the time-derivative term is of order rc /(ctc ) = vc /c, and this is small by virtue of Eq. (5.4.1). We therefore have (t |x x |/c) = (u) + in the wave zone. Inside the integral for we also have 1 1 1 2 2 x /r + O(rc /r ) = 1 + O(rc /r) , = 1+r |x x | r r and it is sucient to keep the leading term only. The wave-zone potential is then (t, x) = = 1 1 40 r 1 1 40 r (u) +
V

r x 2 2 (u) + O vc /c c

(5.4.7)

r x (u) + d3 x c r (u, x )x d3 x + . (u, x ) d3 x + c u V

In the rst integral we recognize Q, the total charge of the distribution; this is actually independent of u by virtue of charge conservation. In the second integral we recognize p(u), the dipole moment vector of the charge distribution; this does depend on retarded time u. Our nal expression for the potential is therefore wave (t, x) = where Q=
V

1 r p (u) 1 Q + + , 40 r 40 cr (u, x ) d3 x

(5.4.8)

(5.4.9)

is the total charge and p(u) =


V

(u, x )x d3 x

(5.4.10)

the dipole moment. The rst term on the right-hand side of Eq. (5.4.8) is the static, monopole potential associated with the total charge Q. This term does not depend on time and is not associated with the propagation of radiation; we shall simply omit it in later calculations. The second term, on the other hand, is radiative: it depends on retarded time u = t r/c and decays as 1/r. We see that the radiative part of the scalar potential is produced by a time-changing dipole moment of the charge distribution; it is nonzero whenever dp/dt is nonzero. You should check that wave (t, x) 1 r p (u) 40 cr

reduces to Eq. (5.3.8) when the dipole moment is given by p(t) = p0 cos(t), where p0 is a constant vector.

5.4.3 Vector potential


The exact expression for the vector potential is A(t, x) = 0 4 j (t |x x |/c, x ) 3 d x. |x x | |x x | c

In the near zone we approximate the current density as j (t |x x |/c) = j (t) 1 + O

88 and we obtain

Electromagnetic radiation from slowly moving sources

Anear (t, x) =

0 4

|x x | j (t, x ) 1+O |x x | c

d3 x .

(5.4.11)

We see that here also, the vector potential takes its static form, except for the correction of order |x x |/c and the fact that the current density depends on time. The potential responds virtually instantaneously to changes in the distribution, and there are no radiative eects in the near zone. In the wave zone we have instead j (t |x x |/c) = j u + r x /c = j (u) 1 + O(vc /c) and Awave (t, x) = 0 1 4 r j (u, x ) 1 + O(vc /c) d3 x .
V

In static situations, the volume integral of j vanishes recall Eq. (3.4.3). But here the current density depends on time, and as we shall prove presently, we have instead j (u, x ) d3 x = p (u). (5.4.12)
V

The vector potential is therefore Awave (t, x) = 0 p (u) + 4 r (5.4.13)

in the wave zone. This has the structure of a spherical wave, and we see that the radiative part of the vector potential is produced by a time-changing dipole moment. You may verify that Eq. (5.4.13) reduces to reduces to Eq. (5.3.5) when the dipole moment is given by p(t) = p0 cos(t), where p0 is a constant vector. To establish the z component of Eq. (5.4.12) we start with the identity (z j ) = z ( j ) + j (z ), in which we substitute the statement of charge conservation, j = /t. After integration over the source we have
V

jz d3 x

d dt

z d3 x =
V V

(z j ) d3 x =

z j da,

where S is the two-dimensional surface bounding V . Because no current is crossing this surface, the right-hand side vanishes and we have jz d3 x =
V

d dt

z d3 x =
V

dpz . dt

This is the same thing as the z component of Eq. (5.4.12), except for the fact that and j are expressed in terms of the variables (t, x) instead of (u, x ). The other components of the equation are established with similar manipulations.

5.4.4 Wave-zone elds


The potentials wave (t, x) = and p (u) 1 r 1 + O(vc /c) 40 cr (5.4.14)

0 p (u) 1 + O(vc /c) , (5.4.15) 4 r where u = t r/c and r = x/r, are generated by time variations of the dipole moment vector p of the charge and current distribution. They therefore give rise Awave (t, x) =

5.4

Electric dipole radiation

89

to electric-dipole radiation, the leading-order contribution, in our slow-motion approximation, to the radiation emitted by an arbitrary source. We now compute the electric and magnetic elds in this approximation. To get the electric eld we keep only those terms that decay as 1/r, and neglect terms that decay faster. For example, when computing the gradient of the scalar potential we can neglect r1 = r /r2 ; we only need (r p ), which we calculate as x (r p ) y z x p x + p y + p z r r r 1 1 1 x y z 1 = p x x r + p y x r + p z x r + O r c r c r c r 1 x 1 , = r p +O c r r = x 1 1 . r p r + O c r

so that r p = (5.4.16)

To obtain this result we made use of the fact that p depends on x through u = t r/c, so that acting with amounts to taking a derivative with respect to u and multiplying by c1 r = c1 r . The electric eld is then E = A t 0 p 1 (r p )r = + , 4 r 40 c2 r

or Ewave (t, x) = p (u) r (u) r p (u) r 0 r 0 p = . 4 r 4 r (5.4.17)

Notice that the wave-zone electric eld behaves as a spherical wave, and that it is transverse to r , the direction in which the wave propagates. To get the magnetic eld we need to compute p (u). To see what this is we examine its x component: p
x

= y p z z p y 1 y 1 z z + p y = p c r c r 1 p x, = r c (5.4.18)

so that

1 p (u) = r p (u). c

The magnetic eld is then Bwave (t, x) = p (u) 0 r . 4 cr (5.4.19)

Notice that the wave-zone magnetic eld behaves as a spherical wave, and that it is orthogonal to both r and the electric eld. Notice nally that the elds are in phase they both depend on p (u) and that their magnitudes are related by |B |/|E | = 1/c. The properties listed in Sec. 5.3.5 are therefore shared by all electric-dipole radiation elds.

90

Electromagnetic radiation from slowly moving sources

5.4.5 Energy radiated


1 in the wave zone The Poynting vector is S = 0 E B , and since E = cB r as follows from Eqs. (5.4.17) and (5.4.19) we have S = (c/0 )(B r ) B = (c/0 )[|B |2 r (B r )B ]. The last term vanishes by virtue of Eq. (5.4.19), and we obtain c 2 , (5.4.20) Bwave r Swave = 0

the same result as in Eq. (5.3.23). The fact that the Poynting vector is directed along r shows that the electromagnetic eld energy travels along with the wave. The energy crossing a sphere of radius r per unit time is given by P = S da, where da = r r2 d and d = sin dd. Substituting Eqs. (5.4.19) and (5.4.20) yields 0 2 P = r p (u) d. (4 )2 c To evaluate the integral we use the trick of momentarily aligning the z axis with the instantaneous direction of p (u) we must do this for each particular value of u. Then r p (u) = |p (u)| sin and P = 0 p (u) (4 )2 c
2

sin2 d =

0 p (u) (4 )2 c

2 0

d
0

sin3 d.

The angular integration gives (2 )(4/3) and we arrive at P = 0 2 p (u) . 6c (5.4.21)

This is the total power radiated by a slowly-moving distribution of charge and current. The powers angular distribution is described by dP 0 2 = p (u) sin2 (u), d (4 )2 c where (u) is the angle between the vectors r and p (u). (5.4.22)

5.4.6 Summary electric dipole radiation


To leading order in a slow-motion approximation, the wave-zone elds of an arbitrary distribution of charges and currents are given by Bwave (t, x) = and Ewave = cBwave r . The elds are generated by time variations of the sources dipole moment vector, p(t) =
V

0 r p (t r/c) 1 + O(vc /c) 4 cr

(t, x)x d3 x.

The elds behave as spherical waves, they are each transverse to r , and they are mutually orthogonal. The eld energy ux vector is Swave = c 2 , Bwave r 0

and the total power radiated is given by P = S da = 0 2 p (u) . 6c

5.5
15

Centre-fed linear antenna


15

91

10

10

-5

-5

-10

-10

-15 0 2 4 6 8 oscillating dipole: t = 0 10 12 14

-15 0 2 4 6 8 oscillating dipole: t = 1 10 12 14

15

15

10

10

-5

-5

-10

-10

-15 0 2 4 6 8 oscillating dipole: t = 2 10 12 14

-15 0 2 4 6 8 oscillating dipole: t = 3 10 12 14

Figure 5.1: Electric eld lines of an oscillating dipole, p(t) = p0 cos(t) z , at selected moments of time, for = 1. Witness the transition between near-zone behaviour and wave-zone behaviour which occurs near r = = 2c/ 1. The curves were obtained by following the method detailed in Sec. 2.11. The powers angular prole is proportional to sin2 (u), where (u) is the angle between the vectors r and p (u). The electric eld lines of an oscillating dipole, p(t) = p0 cos(t) z , are represented in Fig. 5.1. The eld lines of a rotating dipole, p(t) = p0 [cos(t) x + sin(t) y ], are represented in Fig. 5.2.

5.5

Centre-fed linear antenna

[The material presented in this section is also covered in Sec. 9.4 of Jacksons text.] As an example of a radiating system we consider a thin wire of total length 2 which is fed an oscillating current through a small gap at its midpoint. The wire runs along the z axis, from z = to z = , and the gap is located at z = 0 (see Fig. 5.3.) For such antennas, the current typically oscillates both in time and in space, and it is usually represented by j (t, x) = I sin k ( |z |) (x) (y )z cos t, where k = /c. (5.5.2) The current is an even function of z (it is the same in both arms of the antenna) and its goes to zero at both ends (at z = ). The current at the gap (at z = 0) is I sin k, and I is the currents peak value. We want to calculate the total power radiated by this antenna, using the electricdipole approximation. To be consistent we must be sure that vc c for this (5.5.1)

92

Electromagnetic radiation from slowly moving sources

-2

-2

-4

-4

-6

-6

-8 -5 -4 -3 -2 -1 0 1 rotating dipole: t = 0 2 3 4 5

-8 -6 -4 -2 0 rotating dipole: t = 1 2 4 6

-2

-4

-6

-8 -8 -6 -4 -2 0 rotating dipole: t = 2 2 4 6 8

Figure 5.2: Electric eld lines of a rotating dipole, p(t) = p0 [cos(t) x + sin(t) y ], at selected moments of time, for = 1. Witness the transition between near-zone behaviour and wave-zone behaviour which occurs near r = = 2c/ 1.

z=l

antenna coaxial feed

z = l
Figure 5.3: Centre-fed linear antenna. The oscillating current is provided by a coaxial feed.

5.5

Centre-fed linear antenna

93

distribution of current, or equivalently, that rc c 2/k refer back to Eq. (5.4.2). In other words, we must demand that k 1, which means that k |z | is small throughout the antenna. We can therefore approximate sin[k ( |z |)] by k ( |z |), and Eq. (5.5.1) becomes j (t, x) = I0 1 |z |/ (x) (y )z cos t, where I0 = Ik (5.5.4) is the value of the current at the gap. In this approximation the current no longer oscillates in space: it simply goes from its peak value I0 at the gap to zero at the two ends of the wire. In Sec. 9.4 of his book, Jackson gives an exact treatment of the antenna of Eq. (5.5.1). To compute the power radiated by our simplied antenna we rst need to calculate p (t), the second time derivative of the dipole moment vector. For this it is ecient to turn to Eq. (5.4.12), p (t) = j (t, x) d3 x, (5.5.3)

in which we substitute Eq. (5.5.3). We have

p (t) = I0 z cos t

1 |z |/ dz,

and evaluating the integral gives p (t) = I0 z cos t. Taking an additional derivative and making the substitution t u = t r/c yields p (u) = I0 z sin(u), or after using Eq. (5.5.2). Since the vector p (u) is always aligned with the z axis, we have that (u) = , where is the angle between the vectors r and z . According to Eq. (5.4.22), therefore, the angular distribution of the power is given by dP d = = 0 2 p (u) sin2 2 (4 ) c 0 (I0 c)2 (k)2 sin2 (u) sin2 . (4 )2 c dP d 0 c I0 k 32 2 p (u) = (I0 c)(k)z sin(u) (5.5.5)

After averaging over a complete wave cycle, this reduces to =


2

sin2 .

(5.5.6)

To obtain the total power radiated we must integrate over the angles. Using sin2 d = 8/3, we arrive at our nal result 0 c 2 I0 k . (5.5.7) P = 12 For a xed frequency , the power increases like the square of the feed current I0 . For a xed current, the power increases like the square of the frequency, so long as the condition k 1 is satised. From Eq. (5.5.6) we learn that most of the energy is radiated in the directions perpendicular to the antenna; none of the energy propagates along the axis.

94

Electromagnetic radiation from slowly moving sources

5.6

Classical atom

The classical picture of a hydrogen atom contains a proton at the centre and an electron on a circular orbit around it. Supposing for simplicity that the proton is at rest, the systems dipole moment is p(t) = ea(t) where a(t) is the electrons position vector. Since the electron is accelerated by the Lorentz force F exerted by the proton, we have p (t) = ea = (e/m)F . Because this is nonzero, the atom emits electromagnetic radiation. The atom therefore loses energy to the radiation, and as a result, the orbital radius a gradually decreases; it eventually goes to zero. As is well known, the classical model leads to an unstable atom. In this section we calculate the lifetime of the classical atom. The Coulomb force exerted by the proton is F = where a = a(t)/a. This gives p (t) = 1 e e2 a (t) 40 m a2 (5.6.1) 1 e2 a , 40 a2

for the second derivative of the dipole moment vector. According to the electricdipole approximation, the total energy radiated by the atom per unit time is given by Eq. (5.4.21), 0 2 p (u) . P = 6c Substituting Eq. (5.6.1) then gives P = e6 0 . 6c (40 )2 m2 ca4 (5.6.2)

Energy conservation dictates that the energy carried away by the radiation must come at the expense of the atoms orbital energy; we therefore have dEorb = P. dt The orbital energy of a classical atom is Eorb = 1 e2 . 40 2a (5.6.4) (5.6.3)

Combining the last three equations, it is easy to show that the orbital radius must decrease, at a rate given by da 1 e2 4 . = dt 3 (40 )2 m2 c3 a2 (5.6.5)

To obtain the lifetime we integrate Eq. (5.6.5), starting with the initial condition that at t = 0, the orbital radius is equal to Bohrs radius a0 = 40 2 . me2

(In this step we make a leap and momentarily leave the classical realm, since the Bohr radius is very much a consequence of quantum mechanics. But we may still think of a0 as giving an appropriate initial condition, as it sets the atomic

5.7

Magnetic-dipole and electric-quadrupole radiation


0

95 dt =

length scale.) Integration of Eq. (5.6.5) is elementary, and we obtain 0 (dt/da) da as a0 (40 )5 c3 6 . = 4me10 To put this in friendlier terms we introduce and C 2 2.43 1012 m electrons Compton wavelength. mc 1 C 1.6 1011 s. 85 c e2 1 ne structure constant (40 ) c 137

(5.6.6)

(5.6.7)

The lifetime can then be expressed as = (5.6.8)

This result shows that under classical laws, the hydrogen atom would be extremely short lived. Equation (5.6.8), by the way, gives an order-of-magnitude estimate of the time required by an atom to make a transition from an excited state to the ground state. The results of this section, especially Eq. (5.6.5), leave us with a curious paradox. Ignoring the fact that the classical laws of electrodynamics do not apply to a real hydrogen atom, we seem to nd an inconsistency in our description of the classical atom. On the one hand, the fact that the electron feels the Coulomb eld of the proton guarantees that the electrons motion will be circular and that its orbital radius will not change with time. (The most general orbit allowed by the Coulomb force is actually an ellipse, but the point remains that the size of this ellipse cannot change with time.) On the other hand, we know that the electron is accelerated, that it emits electromagnetic waves, that the radiation carries energy away from the system, and that the orbital radius cannot possibly stay constant. The paradox is therefore this: The electrons equations of motion do not account for the decrease in orbital radius; they unambiguously predict that the radius should be constant. Is classical electrodynamics internally inconsistent? Later on, in Chapter 6, we will show that the electrons radiation reaction is actually contained in Maxwells theory. The eect is there, but it is well hidden, and it is a subtle matter to reveal it.

5.7

Magnetic-dipole and electric-quadrupole radiation

[The material presented in this section is also covered in Sec. 9.3 of Jacksons text.] Electric-dipole radiation corresponds to the leading-order approximation of the electromagnetic eld in an expansion in powers of vc /c, where vc is a typical internal velocity of the source. In some cases, however, the dipole moment p either vanishes or does not depend on time, and the leading term is actually zero. In such cases, or when higher accuracy is required, we need to compute the next term in the expansion. This is our task in this section. We will see that at the next-to-leading-order, the wave-zone elds depend on the dipole moment vector p, the magnetic moment vector m, and the electric dipole moment tensor Qab . We recall the denitions of these objects: p(t) =
V

(t, x )x d3 x ,

(5.7.1)

96

Electromagnetic radiation from slowly moving sources 1 2


V

m(t) Qab (t)

= =

x j (t, x ) d3 x ,

(5.7.2) (5.7.3)

2 3 (t, x ) 3x a xb r ab d x ,

where V is the volume that contains the charge and current distribution; this volume is bounded by the surface S . In the foregoing manipulations we will introduce the vector Q(t, x) dened by Qa (t, r ) = Qab (t)rb , (5.7.4)

where rb xb /r are the components of the unit vector r = x/r, and in which summation over the repeated index b is understood.

5.7.1 Wave-zone elds


To make our calculations more ecient we rst derive a few useful results that follow from the fact that in this section, we place ourselves exclusively in the wave zone. So throughout this discussion we shall assume that r c c = . 2 c (5.7.5)

Notice that this condition is independent of any assumption that concerns vc : there is a wave zone whether or not the source is moving slowly. But if vc /c 1, then we also have that rc c /(2 ) r, as was shown in Sec. 5.4.1. In this subsection we rely on the wave-zone condition of Eq. (5.7.5), but we do not rely on the slow-motion assumption; this is will be incorporated at a later stage. We can anticipate that the wave-zone potentials will have the form of a spherical wave. For example, we shall verify that the vector potential is given by A(t, x) = 0 w(u) 4 r (5.7.6)

in the wave zone, where u = t r/c and w is a vector that will be determined. We will not need an expression for the scalar potential; as we shall see, in the wave zone E can be obtained directly from B , which is determined by the vector potential. Given the form of the vector potential, its curl is A= w(t r/c) 0 w(t r/c) 0 = (r) . 4 r 4 r r

The gradient of r is the vector r , and the derivative of w(u)/r with respect to r is c1 w (u)/r, up to terms of order 1/r2 that can be neglected in the wave zone. We therefore have w (u) 0 r B (t, x) = (5.7.7) 4 cr for the wave-zone magnetic eld. Later we will nd that w is not just a function of retarded time u, but that it depends also on r : w = w(u, r ). This does not aect our result for the magnetic eld, nor any other result derived in this subsection. To see this, let us calculate more carefully the x derivative of the vector w: x w = w w x u + x r . u r

In the rst term we substitute x u = c1 x r = c1 rx and we recover our previous expression. We then recognize that x r is of order 1/r and that the

5.7

Magnetic-dipole and electric-quadrupole radiation

97

second term can be neglected in the wave zone. We conclude that the dependence of w on r is invisible to so long as all computations are carried out in the wave zone. To get the electric eld we go back to one of Maxwells equations, B = 0 j + 1 E . c2 t

Because the current density vanishes outside the volume V , we have that j = 0 in the wave zone, and 0 c r w (u) E . = c2 B = t 4 r By the same argument as the one just given, we can treat r as a constant vector when evaluating the curl, and we can also pull the 1/r factor in front of the derivative operator. This yields and we obtain r w (u) 0 r E = . t 4 r Integrating this with respect to t gives E (t, x) = w (u) r 0 r = cB (t, x) r 4 r (5.7.8) 1 r w r w (u) r w (u) = (r) = r , r r r c r

for the wave-zone electric eld. Using Eqs. (5.7.7) and (5.7.8) it is easy to show that the wave-zone Poynting vector is given by S (t, x) = w (u) 0 r c |B |2 r = 2 0 16 c r2
2

r .

(5.7.9)

Evaluating this on a sphere of constant r gives dP = S da = or 0 dP 2 = r w (u) , d 16 2 c (5.7.10) 0 2 r w (u) d, 16 2 c

where d = sin dd is an element of solid angle. Equation (5.7.10) gives the radiations angular distribution, and here it becomes important to specify the dependence of w on the radial vector r . Notice that the elds and the radiations angular prole depend only on r w , and that they are insensitive to an eventual component of w along r . We will use this observation later, and discard from w any longitudinal component. From these results we infer, once more, that irrespective of the exact expression for the vector w(u, r ), the elds are both transverse (orthogonal to r ) and mutually orthogonal, that their amplitudes dier by a factor of c, that they are in phase, and that the eld energy travels in the radial direction. To be more concrete we shall have to calculate the vector w.

98

Electromagnetic radiation from slowly moving sources

5.7.2 Charge-conservation identities


Before we get to this we need to establish the useful identities j (t, x ) d3 x = p (t)
V

(5.7.11)

and
V

1 d 1 (t, r ) + r j (t, x )(r x ) d3 x = m(t) r + Q 6 6 dt

(t, x )r2 d3 x , (5.7.12)


V

where p(t) is the electric dipole moment, m(t) the magnetic dipole moment, and Q(t, r ) the vector related to the electric quadrupole moment which we introduced in Eq. (5.7.3). Equation (5.7.11) was already derived near the end of Sec. 5.4.3; refer back to Eq. (5.4.12). Equation (5.7.12) is a generalization of Eq. (3.4.4), which covers static situations (and which was derived in Sec. 3.4), and we derive it using similar methods. Consider the z component of Eq. (5.7.12). We write the left-hand side as Iz = e jz (t, x)x d3 x,
V

in which we have changed the integration variables from x to x and replaced the constant vector r by the arbitrary constant vector e. More explicitly, Iz is given by Iz = ex
V

xjz d3 x + ey
V

yjz d3 x + ez
V

zjz d3 x.

Consider now the relation = xz j + xj z + z j x + xjz + zjx . = xz t Integrating both sides over V and using Gauss theorem (together with the statement that j da = 0 on the bounding surface S ) gives
V

(xz j )

xjz d3 x = yjz d3 x =

zjx d3 x +
V

d dt

xz d3 x.
V

We obtain zjy d3 x +
V V

d dt

yz d3 x
V

in a similar way. Consider now (z 2 j ) = z 2 j + 2z j z = z 2 integrating this over V leads to zjz d3 x =


V

+ 2zjz ; t

1 d 2 dt

z 2 d3 x.
V

Let us now substitute these results into our previous expression for Iz . We have Iz 1 d ex xz d3 x xjz d3 x zjx d3 x + 2 dt V V V 1 d 1 d yz d3 x + ez yjz d3 x + ey zjy d3 x + 2 dt V 2 dt V V 1 1 = ex (x j )y d3 x + ey (x j )x d3 x 2 2 V V 1 d 3 ex xz d x + ey + yz d3 x + ez z 2 d3 x . 2 dt V V V =

z 2 d3 x
V

5.7

Magnetic-dipole and electric-quadrupole radiation

99

In this last expression, the rst two terms are equal to ex my + ey mx = (m e)z . The bracketed terms are related to Qz : Keeping in mind that we use e as a substitute for r , we have Qz (t, e) = Qzx ex + Qzy ey + Qzz ez = ex
V

(3xz ) d3 x + ey
V

(3yz ) d3 x + ez
V

(3z 2 r2 ) d3 x

3[ ] ez

r2 d3 x,
V

where [ ] stands for the bracketed terms in our previous expression for Iz . Using these new results we nally obtain Iz = (m e)z + 1 d Qz + ez 6 dt r2 d3 x ,
V

which is the same statement (after the substitutions x x and e r ) as Eq. (5.7.12).

5.7.3 Vector potential in the wave zone


We are now ready to calculate the vector w. We begin with the exact expression for the vector potential, A(t, x) = 0 4 j (t |x x |/c, x ) 3 d x. |x x |

In this we substitute the approximations 1 1 |x x | r and t |x x |/c u + r x /c, where u = t r/c is retarded time, which are appropriate in the wave zone. This gives 0 1 A(t, x) = j u+r x /c, x d3 x (5.7.13) 4 r V for vector potential. Notice that this expression, though restricted to the wave zone, is otherwise general: it is not yet limited by a slow-motion assumption. We now incorporate the slow-motion approximation by Taylor-expanding the current density: j 1 x ) (u) + . j (u + r x /c) = j (u) + (r c u Relative to the leading term, the second term is smaller by a factor of order c1 rc /tc vc /c. This term was neglected in our discussion of electric-dipole radiation, but we shall now keep it. We will, however, neglect all remaining terms, which contribute at order (vc /c)2 . Substituting the expansion for j into Eq. (5.7.13) gives A(t, x) = 0 1 4 r j (u, x ) d3 x +
V

1 c u

j (u, x )(r x ) d3 x + .

100

Electromagnetic radiation from slowly moving sources

According to Eq. (5.7.11), the rst integral is equal to p (u). And according to Eq. (5.7.12), the second integral is equal to 1 d 1 (u, r ) + r m(u) r + Q 6 6 du (u, x )r2 d3 x .
V

The quantity within the square brackets denes the vector w, and we have w=p (u) + 1 1 d2 1 m (u) r + Q (u, r ) + r c 6 6c du2 (u, x )r2 d3 x .
V

The last term is proportional to r . But we have seen in Sec. 5.7.1 that the electric and magnetic elds, as well as the radiated power, depend only on r w . This means that any component of w along r will not contribute to the wave-zone elds. We can therefore discard this last term in our expression for w. We have obtained w(u, r ) = p (u) + 1 1 m (u) r + Q (u, r ) , c 6 (5.7.14)

and in terms of this the wave-zone vector potential is given by A(t, x) = 0 w(u, r ) , 4 r (5.7.15)

an expression that was already anticipated in Eq. (5.7.3). The rst term on the right-hand side of Eq. (5.7.14) gives rise to electric-dipole radiation; this is the leading term in the expansion of the elds in powers of vc /c. The second and third terms give rise to magnetic-dipole radiation and electric-quadrupole radiation, respectively; these contribute at order vc /c beyond the leading term. To calculate the radiated power we shall need r w =r p (u) + 1 1 m (u) r m (u) r + r Q(3) (u, r ), c 6c (5.7.16)

where Q(3) 3 Q/u3 and we have used the vectorial identity r (m r ) = m (r m )r .

5.7.4 Radiated power (magnetic-dipole radiation)


Let us calculate the power radiated by a distribution of charges and currents for which p (u) = Q(3) (u, r ) = 0. (5.7.17) Such a source will emit magnetic-dipole radiation. Under the conditions of Eq. (5.7.17), Eq. (5.7.16) becomes r w = so that r w
2

1 m r m )r c

1 |m |2 (r m )2 , c2 which we rewrite in component form as = r w


2

1 m am b ab ra rb , c2

where ra = xa /r are the components of the vector r , and where summation over both repeated indices is understood. Substituting this into Eq. (5.7.10) gives 0 dP m am b ab ra rb , = d 16 2 c3

5.7

Magnetic-dipole and electric-quadrupole radiation

101

and integrating over the angles yields P = where 0 m am b ab ra rb , 4c3

1 1 ra rb d = ab . 4 3 These angular integrations are worked out in Sec. 5.7.7. Our nal result for the radiated power is therefore ra rb Pmagnetic-dipole = 0 2 m (u) . 6c3 0 2 p (u) . 6c (5.7.18)

We might compare this result with Eq. (5.4.21), Pelectric-dipole =

3 In orders of magnitude we have that the electric dipole moment is p (rc )(rc ) 4 4 3 4 2 4 rc /tc . On the other hand, m (jrc )(rc ) jrc vc rc , and rc and that p 4 2 m vc rc /tc . The ratio of powers is then

Pmagnetic-dipole Pelectric-dipole

m cp

vc c

So for slowing-moving sources, the power emitted in magnetic-dipole radiation is typically smaller than the power emitted in electric-dipole radiation by a factor of order (vc /c)2 1.

5.7.5 Radiated power (electric-quadrupole radiation)


Let us now calculate the power radiated by a distribution of charges and currents for which p (u) = m (u) = 0. (5.7.19) Such a source will emit electric-quadrupole radiation. Under the conditions of Eq. (5.7.19), Eq. (5.7.16) becomes r w = so that r w
2

1 r Q(3) , 6c

1 r Q(3) r Q(3) . 36c2 The inner product can be expressed as = (r r ) Q(3) Q(3) r Q(3) r Q(3) = Q(3)
2

r Q(3) ,

and we have r w
2

1 (3) Q(3) Q ab ra rb . 36c2 a b

But the vector Q(3) depends on r , because by the denition of Eq. (5.7.4), Qa = Qac rc . Writing also Qb = Qbd rd , we arrive at r w
2

1 (3) Q(3) Q ab rc rd ra rb rc rd . 36c2 ac bd

Substituting this into Eq. (5.7.10) gives 1 0 dP (3) Q(3) Q = ab rc rd ra rb rc rd , 2 d 16 c 36c2 ac bd

102

Electromagnetic radiation from slowly moving sources

and integrating over the angles yields P = The angular integrals rc rd and ra rb rc rd 1 4 ra rb rc rd d = 1 ab cd + ac bd + ad bc 15 1 4 rc rd d = 1 cd 3 0 1 (3) Q(3) Q ab rc rd ra rb rc rd . 4c 36c2 ac bd

are evaluated in Sec. 5.7.7. Taking these results into account, we obtain Q(3) ac Qbd ab rc rd ra rb rc rd 1 1 (3) (3) (3) Qac Qbd ab cd Q(3) Q ab cd + ac bd + ad bc = 3 15 ac bd 1 (3) (3) 1 1 (3) (3) 1 (3) = Qac Qac Q(3) Qaa Qbb Q(3) Q(3) ac Qac 3 15 15 15 ac ca 2 1 (3) Q(3) = ac Qac 3 15 1 (3) (3) = Q Q , 5 ac ac
(3)

where we have used the property that Qab is a symmetric, tracefree tensor (so that Qca = Qac and Qaa = 0). Our nal result for the radiated power is therefore Pelectric-quadrupole = 0 (3) (3) Q (u)Qab (u), 720c3 ab (5.7.20)

where summation over the two repeated indices is understood. In orders of mag(3) 2 3 5 5 3 nitude, Qab (rc )(rc ) rc , and Qab rc /tc p rc /tc p vc . We therefore have Pelectric-quadrupole Pelectric-dipole vc c
2

for slowing-moving distributions, the power emitted in electric-quadrupole radiation is smaller than the power emitted in electric-dipole radiation by a factor of order (vc /c)2 1. We see also that electric-quadrupole radiation is of the same order of magnitude as magnetic-dipole radiation.

5.7.6 Total radiated power


In situations in which all three types of radiation contribute, the total power is the sum of individual contributions, P = 0 p (u) 6c
2

1 m (u) c2

1 (3) (3) 4 4 Q (u)Qab (u) + O vc /c 120c2 ab

(5.7.21)

There is no interference between the dierent terms in Eq. (5.7.16), because the angular proles associated with each term are orthogonal.

5.8

Pulsar spin-down

103

5.7.7 Angular integrations


In this subsection we establish the formulae ra ra rb ra rb rc ra rb rc rd 0, 1 ab , = 3 = 0, 1 = ab cd + ac bd + ad bc , 15 = (5.7.22) (5.7.23) (5.7.24) (5.7.25)

where ra = xa /r are the components of the vector r , and (4 )1 ( ) d is the average of the quantity ( ) over a sphere of constant r. These relations can be derived by brute-force computation, by simply substituting r = (sin cos , sin sin , cos ) and integrating over d = sin dd. But there is an easier, prettier way that relies on arguments of symmetry. Take the zero result for ra . If these integrals were not zero, they would have to be equal to some vector a that does not depend on the angles and (because they have been integrated over) nor on r (because r appears nowhere on the left-hand side of the equation). But there is no such vector available, because there is no preferred direction in a homogeneous three-dimensional space. The right-hand side must therefore be zero. To arrive at the result for ra rb we observe that the right-hand side must be a constant tensor that is symmetric in the pair of indices a and b. The only such tensor available is ab , and we must therefore have ra rb = ab , for some constant . To determine we simply take the trace of the left-hand side, ab ra rb = ab ra rb = 1 = 1, and the trace of the right-hand side, ab ab = 3. Equating the results produces = 1/3, and this agrees with the result of Eq. (5.7.23). The zero result for ra rb rc follows from the fact that we do not have a vector that could be combined with ab to form a symmetric, three-index tensor. To derive the result for ra rb rc rd we note that the right-hand side must be a four-index tensor that is symmetric in all pairs of indices. We must form this tensor with two copies of ab , and it is easy to see that it must be equal to (ab cd + ac bd + ad bc ). To determine we evaluate the double trace of the left-hand side, which gives 1, and the double trace of the right-hand side, which gives ab cd ab cd + ac bd + ad bc = (3 3 + 3 + 3) = 15 . Equating the results produces = 1/15, in agreement with Eq. (5.7.24).

5.8

Pulsar spin-down

As an application of magnetic-dipole radiation we consider the oblique-rotator model of a pulsar, a rotating neutron star that emits pulses of electromagnetic radiation at regular intervals. The most famous pulsar is located in the Crab nebula, and the pulsar is believed to be the source of energy for this supernova remnant. The Crab pulsar is observed to spin down: the frequency of the pulses decreases with time. If we associate the pulse frequency with the neutron stars spin frequency,

104

Electromagnetic radiation from slowly moving sources

z m R

Figure 5.4: The neutron star rotates with angular velocity around the z axis. The magnetic moment m makes an angle of with respect to this axis, and it rotates along with the star. then we can interpret this phenomenon in terms of the pulsar losing rotational energy, which is given by 1 (5.8.1) Erot = I 2 , 2 where I is the stars moment of inertia and its angular velocity. If P denotes the pulses period (which is not to be confused with the power radiated by electromagnetic waves), then = 2/P . If the neutron star is modeled as a solid sphere 2 of uniform mass density, then I = 2 5 M R , where M is the pulsars mass and R its radius. The loss of rotational energy translates into a decrease of , or an increase of P : rot = I = (2 )2 I P . (5.8.2) E P3 4 1013 s/s and P 0.03 s. For the Crab we have the observational values P If we take the neutron star to have a mass of 1.4 solar masses and a radius of 12 km, then the rate of loss of rotational energy is rot 7 1031 J/s. E This is comparable to what is required to power the Crab nebula. The energetics of the Crab can therefore be explained by the pulsar losing its rotational energy. What is responsible for this loss? We can imagine that the neutron star is carrying a magnetic eld, and that the spinning motion of this eld produces magneticdipole radiation. The energy carried o by the radiation will then come at the expense of the stars rotational energy. For this model to work we need the elds orientation to dier from the stars rotational axis otherwise we would have a stationary situation analogous to the one analyzed in Sec. 3.3. This is the obliquerotator model depicted in Fig. 5.4. We imagine that the neutron star maintains a magnetic dipole moment m that is oriented at an angle with respect to the rotation axis. This vector rotates along with the star, also with an angular velocity . If the rotation axis is aligned with the z direction, we can write m(t) = m0 sin cos t x + sin sin t y + cos z , (5.8.3)

5.9

Problems

105

where m0 is the magnitude of the pulsars magnetic moment. This can be related to the magnitude of the magnetic eld on the surface of the neutron star. Because the star is located well within the near zone, we can use Eq. (3.4.2) to describe the surface eld: )r m 0 3(m r , B= 3 4 R where R is the stars radius. The eld is maximum at the magnetic pole, where r is aligned with m. We therefore have Bmax = 0 2m0 , 4 R3 (5.8.4)

a relationship between m0 , the stars radius, and the maximum surface magnetic eld. The time-changing magnetic moment of Eq. (5.8.3) produces magnetic-dipole radiation. This radiation takes energy away from the star, at a rate given by Eq. (5.7.18), 2 rad = 0 m (u) , E 3 6c where u = t r/c is retarded time. Substituting Eq. (5.8.3) gives rad = 0 m0 2 sin 2 , E 6c3 or rad = E 2 Bmax 2 R3 sin 30 c3
2

(5.8.5)

(5.8.6)

after involving Eq. (5.8.4). rad and |E rot | 7 1031 J/s, we need a To produce an equality between E magnetic eld such that Bmax sin 5 108 T. This is very large, but not unreasonable. A main-sequence star typically supports a magnetic eld of 0.1 T. When the star collapses to form a neutron star, the magnetic eld is frozen in and the magnetic ux 4R2 B is conserved. Because R decreases by a factor of 105 during the collapse, B increases by a factor of 1010 , and elds of the order of 108 T can easily be achieved for neutron stars.

5.9

Problems

1. A point electric dipole is rotating around the z axis at an angular velocity , so that p(t) = p0 cos(t) x + sin(t) y , where p0 is a constant. a) Calculate the wave-zone electric and magnetic elds of this rotating dipole. and . Express the elds in terms of the unit vectors b) Calculate dP/d , the angular prole of the radiated power, averaged over a complete wave cycle. c) Calculate the total radiated power. How does it compare with the power radiated by an oscillating dipole? 2. Two equal charges q are moving with constant angular velocity on a circle of radius b. The charges are situated at diametrically opposite points on the circle. Calculate, to leading order in a slow-motion approximation, the total power radiated by this distribution of charges.

106

Electromagnetic radiation from slowly moving sources

3. The current density inside a centre-fed, linear antenna is given by j (t, x) = I cos(t) sin(k k |z |) (x) (y ) z , where I is the currents peak amplitude, the oscillation frequency, the antennas half-length, and k = /c. In this problem we do not impose the slow-motion approximation, so that k is not assumed to be small. a) Show that in the wave zone (kr 1), the vector potential can be expressed as 0 w(u, ) , A(t, x) = 4 r and nd an expression for the vector w(u, ). Here, u = tr/c is retarded time and is the angle between x and the z axis. b) Calculate dP/d , the angular prole of the radiated power, averaged over a complete wave cycle. For the two special cases k = /2 (halfwave antenna) and k = (full-wave antenna), provide parametric plots of this angular prole (like those displayed in Jacksons Figure 9.7). Compare these proles with the sin2 dependence obtained for electric-dipole radiation. c) From the general expression for dP/d derived in part b), obtain the short-antenna limit k 1. Make sure that your result matches the one derived in the text. 4. An oscillating current I (t) = I0 et sin(t) ows in a circular loop of radius a placed in the equatorial plane of the coordinate system (r, , ). The current is started at t = 0, and it decays exponentially due to dissipation within the loop; we assume that , so that the time scale for dissipation is much longer than the oscillation period. The current density is given by I (t) . (r a) ( /2) a Calculate P , the total power radiated by the loop, averaged over an oscillation period. Calculate also the total amount of energy radiated by the loop. j (t, x) = 5. A nearly spherical surface described by r = R(t, ) a(1 + cos cos t), where a, , and are constants, supports a nearly uniform distribution of charge. It is assumed that 1, and the charge density is given by q r R(t, ) , (t, x) = 4a2 where q is the total charge. Calculate P , the electromagnetic power emitted by this distribution of charge, averaged over a complete wave cycle. You may work to leading order in the slow-motion approximation. You may also work to leading order in . 6. A ring of radius b is placed in the x-y plane, and the origin of the coordinate system is placed at the centre of the ring. Initially, the ring is at rest and it supports a linear charge density proportional to sin , where is the angle from the x-axis. The ring is then made to rotate with a uniform angular velocity , so that the linear charge density becomes proportional to sin( t). The rings volume charge density is described by (t, r, , ) = (r b) ( 2 ) sin( t). b

5.9

Problems

107

(a) Calculate, to leading order in a slow-motion approximation, the total power radiated by the ring. At what angle is most of the radiation emitted? (b) Find an exact expression for (t, 0), the scalar potential at the centre of the ring. (c) Find an exact expression for A(t, 0), the vector potential at the centre of the ring. 7. Three charges are located along the z axis: a charge +2q stays at the origin z = 0, a charge q is at z = +a cos t, and the nal charge q is at z = a cos t; a and are constants. It is assumed that a 2c and that the charges simply go through one another when they meet at z = 0. For this situation calculate dP/d , the angular prole of the radiated power, averaged over a complete wave cycle. (Express this in terms of the angle from the z axis.) Calculate also P , the total radiated power. What is the frequency of the electromagnetic waves emitted by this system of charges? 8. A current I is suddenly established in an innite wire that extends along the z axis. The current density is given by j (t, x) = I(t) (x) (y )z , where (t) is the step function. Because j = 0, it is consistent to set (t, x) = 0. Calculate, without approximations, the vector potential A(t, x) associated with this distribution of current. From your answer you should be able to verify that at late times, the electric and magnetic elds that are produced by the distribution of current have the components Ez 0 2I , 4 t Bx 0 2Iy , 4 x2 + y 2 By 0 2Ix . 4 x2 + y 2

The electric eld eventually disappears, and the magnetic eld settles down to the value it would have for an unchanging current.

108

Electromagnetic radiation from slowly moving sources

Chapter 6 Electrodynamics of point charges


6.1 Lorentz transformations

[The material presented in this section is also covered in Secs. 11.3, 11.9, and 11.10 of Jacksons text.] In this nal chapter we will look at the electrodynamics of charge and current distributions for which the characteristic speed vc is close to the speed of light. To keep the discussion concrete we will restrict it to the electromagnetic eld produced by a single point particle in arbitrary (relativistic) motion. To set the stage we begin with a review of the Lorentz transformations of special relativity. We consider a reference frame S that moves with constant speed v relative to another frame S ; we take the motion to take place in the x direction, so that the frames velocity vector is v = v x . The coordinates (t , x , y , z ) of an event seen in S are related to the coordinates (t, x, y, z ) of the same event seen in S by the well-known relations x = (x vt), where = y = y, z = z, t = t v x , c2 (6.1.1)

1 1 v 2 /c2

(6.1.2)

These are the Lorentz transformations between the two reference frames. The reversed transformation is obtained by reversing the sign of v : x = (x + vt ), y = y , z = z, t = t + v x . c2 (6.1.3)

We now would like to derive how the electric and magnetic elds transform under such a change of reference frame. We note rst that Lorentz transformations alter the dierential operators. For example, applying the chain rule on an arbitrary function gives t x y z = + + + , t t t x t y t z t or = +v t t x 109

110

Electrodynamics of point charges

after using Eq. (6.1.3). Doing a similar calculation for all other partial derivatives, we obtain . = +v t t x (6.1.4) It follows from this that the wave operator is a Lorentz invariant:

v , = + x x c2 t

= , y y

= , z z

1 2 2 2 2 + + + c2 t2 x2 y 2 z 2 1 +v +v = 2 2 c t x t x 2 2 v v + + + 2 + 2 + 2 x c t x c t y 2 z 2 2 2 2 2 1 = 2 2 + + + c t x2 y 2 z 2 = . =

In other words, = . (6.1.5) To gure out how the electric and magnetic elds transform under a change of reference frame, we rst work out how and j should transform. This will tell us how the potentials and A transform, and from this we will be able to get to the elds. It is sucient to consider a simple special case: a point charge q moving in the x direction with a speed v (the same speed that enters in the Lorentz transformation). As seen from S we have = q (x vt) (y ) (z ), jx = qv (x vt) (y ) (z ), and jy = jz = 0. As seen from S , however, the particle is not moving, and = q (x ) (y ) (z ), while jx = jy = jz = 0. We can expect that the Lorentz transformation will involve a linear mixture of and jx , and that the other components of the current density will be left alone. Rewriting in terms of the unprimed coordinates gives = q [ (x + vt)] (y ) (z ) q (x vt) (y ) (z ). =

This expression for involves a factor of 1/ instead of the expected , and jx has not yet made an appearance. But we can write it in the equivalent form = q (1 v 2 /c2 ) (x vt) (y ) (z ) v = q (x vt) (y ) (z ) 2 qv (x vt) (y ) (z ) c v = 2 jx , c

and this is the expected expression of a Lorentz transformation. This shows that = transforms as t, and we further expect that jx will transform as x, that is, jx (jx v). This is conrmed by the fact that the right-hand side is zero for the situation considered here. We conclude that under a change of reference frame, the charge and current densities transform as
jx = (jx v), jy = jy , jz = jz ,

v jx . c2

(6.1.6)

6.1

Lorentz transformations

111

While the derivation of these transformation rules was restricted to a very simple situation, it is possible to show that they are in fact completely general. To see how the potentials transform we recall the wave equations (/c2 ) = /(0 c2 ) = 0 and A = 0 j from Sec. 5.1. Because is an invariant, these equations tell us that /c2 must transform as , and that A must transform as j . Thus, A x = Ax v , c2 A y = Ay , A z = Ax , = ( vAx ) (6.1.7)

are the Lorentz transformations for the scalar and vector potentials. These are designed to leave the form of the wave equations unchanged under a change of reference frame. You should check that the Lorenz gauge condition, A+ 1 = 0, c2 t

also takes the same form in S . All the equations satised by the potentials are therefore preserved under a Lorentz transformation. This implies that Maxwells equations themselves are invariant under a change of reference frame. We can now gure out how the elds transform. We simply write down how the elds in S are related to the potentials, E = A , t B = A ,

and we substitute Eqs. (6.1.4) and (6.1.7). For example,


Ex

+v t x Ax = t x = Ex . = 2

Ax

v v 2 + c2 x c2 t

vAx

Or
Ey

Ay vAx +v t x y Ay Ax Ay v = + t y x y

= Ey vBz . Another example is


Bx =

A A Az Ay y z = = Bx . y z y z

Yet another is
By

A A z x z x v v Ax 2 Az + = z c x c2 t Ax v Az Az = + 2 z x c t z v = By + 2 Ez . c =

112

Electrodynamics of point charges

The complete summary of these results is that the electric eld transforms as
Ex = Ex , Ey = Ey vBz , Ez = Ez + vBy

(6.1.8)

under a change of reference frame, while the magnetic eld transforms as


Bx = Bx , By = By +

v Ez , c2

Bz = Bz

v Ey . c2

(6.1.9)

Two observers in relative motion will therefore disagree as to the value of the electric and magnetic elds. There are interesting special cases to the foregoing results. Suppose rst that B = 0 in some reference frame S . In another reference frame S we have Ex = Ex , 2 2 Ey = Ey , Ez = Ez , Bx = 0, By = (v/c )Ez = (v/c )Ez = (v E )y /c2 , and Bz = (v/c2 )Ey = (v/c2 )Ey = (v E )z /c2 . In other words, B=0 B = 1 v E. c2 (6.1.10)

A pure electric eld in S is therefore perceived as a mixture of electric and magnetic elds in S . Suppose next that E = 0 in S . Then Bx = Bx , By = By , Bz = Bz , Ex = 0, Ey = vBz = vBz = (v B )y , and Ez = vBy = vBy = (v B )z . In other words, E = 0 E = v B. (6.1.11) A pure magnetic eld in S is therefore perceived as a mixture of electric and magnetic elds in S . We conclude that E and B are observer-dependent elds; only the electromagnetic eld as a whole has an observer-independent existence.

6.2

Fields of a uniformly moving charge

[The material presented in this section is also covered in Sec. 11.10 of Jacksons text.] We can use the Lorentz transformations of Eqs. (6.1.8) and (6.1.9) to calculate the electric and magnetic elds of a point charge q that moves with constant speed v in the x direction. The charges position vector is r (t) = (vt, b, 0), (6.2.1)

where b is the charges displacement above the x axis. We want E and B as measured at O, the origin of the reference frame S . We shall rst do the calculation in S , which moves along with the particle. As seen in the reference frame S , the charges position vector is r = (0, b, 0), but the observer at O has a position rO = (vt , 0, 0), because relative to S , O moves in the negative x direction with speed v . In S there is only an electric eld E , and its value at O is given by the standard expression
q r rO |3 . 40 |r rO We have r rO = (vt , b, 0), |r rO | = b2 + v 2 t2 , and the primed components of the electric eld are

E =

Ex =

q vt , 40 (b2 + v 2 t2 )3/2

Ey =

b q , 2 40 (b + v 2 t2 )3/2

Ez = 0. (6.2.2)

As measured in S , the electric eld lines are distributed isotropically around r , the charges position.

6.2

Fields of a uniformly moving charge

113

t 3 2 1 1 2 3

0.5

1.5

Figure 6.1: Electric eld of a uniformly moving charge, as measured by an observer at the origin of a xed reference frame. The x component of the eld goes through zero at t = 0, while the y component is always negative and achieves its maximum (in magnitude) at that time. We now transform the eld to the frame of reference S . The coordinates of O are x = y = z = 0, and the time coordinates are therefore related by t = t. The unprimed elds are obtained by inverting Eqs. (6.1.8) and (6.1.9), which is done by reversing the sign of v . Because B = 0 we have Ex = Ex , Ey = Ey , E z = Ez = 0, 2 2 Bx = Bx = 0, By = (v/c )Ez = 0, and Bz = (v/c )Ey . We therefore obtain Ex = and q vt , 40 (b2 + 2 v 2 t2 )3/2 Bx = 0, Ey = By = 0, q b , 40 (b2 + 2 v 2 t2 )3/2 Bz = Ez = 0 (6.2.3) v Ey (6.2.4) c2 for the elds measured at O, in the S frame. Plots of Ex and Ey as functions of time are shown in Fig. 6.1. The elds can also be expressed in terms of the charges position vector, r (t) = (vt, b, 0). For this purpose we let be the angle between the vectors r and v . This is dened by v r (t) , (6.2.5) cos = vr(t) where r |r | = b2 + v 2 t2 . We have v r = v 2 t, so cos = vt/ b2 + v 2 t2 , which implies sin = b/ b2 + v 2 t2 . We also have b2 + 2 v 2 t 2 = 2 (b2 / 2 + v 2 t2 ) v2 = 2 b2 + v 2 t 2 2 b2 c b2 v2 = 2 (b2 + v 2 t2 ) 1 2 2 c b + v 2 t2 v2 = 2 r2 1 2 sin2 , c

114

Electrodynamics of point charges

and this can be substituted into Eqs. (6.2.3). The end result is E= 1 q 40 2 1 (v/c)2 sin2 1 vE c2
3/ 2

r r3

(6.2.6)

for the electric eld at O, and B= (6.2.7)

for the magnetic eld. We see from Eq. (6.2.6) that E is always directed along r (t), a vector that points from the charge to O. The electric eld therefore points away from the charge, and this means that the electric eld lines originate at the position of the moving charge. The eld lines, however, are not distributed isotropically around the charge; this is revealed by the presence of sin2 in the expression for the electric eld. When = 0, that is, when r is directed along v , we have that [1 (v/c)2 sin2 ]3/2 = 1, and the electric eld is reduced by a relativistic factor of 2 < 1. But when instead = /2, so that r is orthogonal to v , then [1 (v/c)2 sin2 ]3/2 = 3 > 1, and the eld is amplied by a factor > 1. The electric eld, as seen in S , is strongest in the directions transverse to the motion, and weakest in the longitudinal direction. The results derived in this section are at once remarkable and unremarkable. It is unremarkable that the eld lines should emanate from the position of the charge we have the same picture in the charges rest frame. And it is unremarkable that the eld lines should be distorted by a Lorentz contraction in the direction of the motion. But it is remarkable that the eld registered at O does not involve a time delay associated with the propagation of a signal from charge to observer. Allowing for such a delay would suggest that the eld at time t should point not toward the charge at its current position r (t), but instead toward an earlier position r (t ). In this picture, a signal originates at the charge when it was at the earlier position, and it propagates to O in a time t t t determined by the condition that the signal travels at the speed of light; since the distance traveled is |r (t )|, we must have t t = |r (t )|/c, which in principle can be solved for t . During the time in which the signal propagates, the particle moves to its current position, but the eld registered at O should not yet be aware of this; it should know only of the earlier position, and E should therefore point in the direction of r (t ). But this is not what we have found: the eld does point in the direction of r (t), the current position of the charge. Our picture, according to which information should be delayed, is curiously wrong. We will return to this issue later on, and investigate it more fully.

6.3

Fields of an arbitrarily moving charge

[The material presented in this section is also covered in Secs. 6.5 and 14.1 of Jacksons text.] We now allow the point charge q to move on an arbitrary trajectory r (t). We let v (t) = dr /dt be its velocity vector, and a(t) = dv /dt its acceleration vector. The charge density is (t, x) = q x r (t) , (6.3.1) and the current density is j (t, x) = q v (t) x r (t) . (6.3.2)

To obtain the elds produced by this particle we shall rst calculate the potentials.

6.3

Fields of an arbitrarily moving charge

115

6.3.1 Potentials
To calculate the potentials associated with the densities of Eqs. (6.3.1) and (6.3.2) we employ a technique that was rst put to use in Secs. 5.2 and 5.3. For example, we express the vector potential as A(t, x) = 0 4 j (t , x ) (t t |x x |/c) 3 d x dt , |x x |

in terms of the retarded Greens function obtained in Sec. 1.9. Substituting Eq. (6.3.2) yields A(t, x) = q0 4 v (t ) x r (t ) (t t |x x |/c) 3 d x dt , |x x |

and integrating over d3 x returns A(t, x) = q0 4 v (t ) (t t |x r (t )|/c) dt . |x r (t )| (6.3.3)

The same manipulations produce (t, x) = q 40 (t t |x r (t )|/c) dt |x r (t )|


(t ) = R(t ) . R R(t )

(6.3.4)

for the scalar potential. At this stage it is convenient to introduce the quantities R(t ) = x r (t ), R(t ) = |x r (t )|, (6.3.5)

In terms of these the argument of the -function is t t R(t )/c. To evaluate the integrals of Eqs. (6.3.3) and (6.3.4) we let s t + R(t )/c t be the new variable of integration. We have ds/dt = 1 + c1 dR/dt , and it is easy v . With this we obtain ds/dt = , where to verify that dR/dt = R v /R = R (t )/c. 1 v (t ) R The integral for the vector potential is then A = = = Similarly, = q 1 40 R .
s=0

(6.3.6)

q0 v (s) dt ds 4 R ds q0 v (s) ds 4 R q0 v . 4 R s=0

The condition s = 0 gives R(t )/c = t t , or |x r (t )| = c(t t ), (6.3.7)

and this implicit equation can in principle be solved for t . The meaning of Eq. (6.3.7) is clear: The signal received at position x and time t depends on the state of motion of the particle at an earlier time t ; the delay (t t ) is set by the time required by light to travel the distance R(t ). We shall call t the retarded time corresponding

116

Electrodynamics of point charges

to the spacetime point (t, x), and Eq. (6.3.7) will be referred to as the retarded condition. We shall write the potentials as (t, x) = and A(t, x) = 1 q 40 R q0 v 4 R (6.3.8)
ret

,
ret

(6.3.9)

indicating that the quantities within the square brackets must be evaluated at the time t determined by the retarded condition of Eq. (6.3.7). These expressions are known as the Li enard-Wiechert potentials. Apart from the presence of in the denominator, the potentials take their familiar form from time-independent situations, and they implement the notion that the potentials here and now should depend on the retarded position and velocity of the particle. Given that the retardation eect is present in the Li enard-Wiechert potentials, it is all the more curious that the eect was not seen in the preceding section. (Stay tuned.) The factor of 1/ accounts for the fact that (t |x x |/c, x ) d3 x is not equal to the charge q . While we do have (t, x ) d3 x = q , in the previous expression we sample the charge density at dierent times while performing the integration, and the result is not q . Instead, (t |x x |/c, x ) d3 x = = q = q (t , x ) (t t |x x |/c) d3 x dt (t t |x r (t )|/c) dt ,
ret

/c. and this result claries the origin of = 1 v R

6.3.2 Fields
Having obtained the potentials, we can calculate the elds by straightforward differentiation, A E= , B = A. t The dependence of A on t comes from its explicit dependence on t and the retarded condition |x r (t )| = c(t t ). Its dependence on x comes from the explicit dependence of R and on position, but there is also an implicit dependence contributed by the retarded condition. This mixture of explicit and implicit dependence complicates the task of taking derivatives. But we can simplify things by going back to earlier forms of Eqs. (6.3.3) and (6.3.4), (t, x) = and A(t, x) = q 40 1 (s) dt R(t ) v (t ) (s) dt , R(t ) (6.3.10)

q 1 40 c2

(6.3.11)

6.3

Fields of an arbitrarily moving charge

117

where s = t + R(t )/c t. Their advantage is that t inside the integrand is no longer restricted, so that the dependence of the potentials on t and x is fully explicit. It will then be straightforward (if tedious) to dierentiate rst, and integrate next (thereby setting t by the retarded condition). Before we proceed we recall the denitions of (t )/c = ds/dt . Eqs. (6.3.5) and (6.3.6): R(t ) = x r (t ) and = 1 v (t ) R We begin with the scalar potential. We have = q 40 1 1 (s) + (s)s dt , R R

and since s depends on x through R, we can replace s by R/c. This gives = = = q 40 q 40 q 40 R 1 (s) + (s) dt R cR R 1 1 (s) + (s) ds R cR d R 1 1 (s) ds, R ds cR

where we have integrated the second term by parts. The derivative with respect to s can now be expressed in terms of d/dt , and we nally obtain = q 40 1 1 1 d R c dt R R (s) ds.

Leaving this result alone for the time being, we move on and calculate A/t. We have A t v s q 1 (s) dt 2 40 c R t v q 1 (s) ds = 40 c2 R d v q 1 (s) ds = 40 c2 ds R q 1 v 1 d (s) ds. = 40 c2 dt R =

In the rst step we used the fact that the only t dependence is contained in s. In the second step we substituted s/t = 1. In the third step we integrated by parts. And in the fourth step we replaced the s-derivative by a t -derivative. Collecting these results gives us an expression for the electric eld: E = = q 40 q 40 1 1 1 1 d + R c dt 1 d 1 + R c dt 1 d R v 2 R c dt R R v /c (s) ds. R (s) ds

This can be integrated at once, and we obtain E= q 1 1 1 d + 40 R c dt R v /c R .


ret

, This expression badly needs to be simplied. It is easy to establish that R = R 1 2 so that R = R/R , and we already have obtained dR/dt = v R in the preceding subsection. We now set to work on d dt R v /c R = 1 dR a R dt c v /c d R R , 2 R2 dt

118

Electrodynamics of point charges

/dt = d(R/R)/dt = where a = dv /dt is the particles acceleration. We have dR 2 v /R + (R/R )(v R) = [v (v R)R]/R, and also d(R)/dt = d(R v (a R v v )/c = v R a R/c + v 2 /c. Plugging these R/c)/dt = v R results back into our previous expression gives 1 d c dt R v /c R 1 /c)R a v /c (v R R2 Rc2 v /c R /c + v 2 /c2 + R v /c a R v R 2 2 R 2 R 2 c2 1 + R v /c 1 v 2 /c2 = 2 v /c (1 )R R 2 R 2 1 v /c)(a R ) a + 2 2 (R Rc v /c v /c v /c R R R R = 1 v 2 /c2 2+ 2 2 2 R R R R2 1 v /c)(a R ) a . + 2 2 (R Rc =

The rst and fourth terms cancel out, and the expression within the square brackets [(R v /c) a] = (R a)(R v /c) R (R v /c)a = (R a)(R v /c) a. is R This gives 1 d c dt R v /c R = (R v /c) a v /c R R R + + . R2 2 2 R 2 2 c2 R

Substituting all this into our previous result for the electric eld returns E (t, x) = (R v /c) a v /c R R q + 2 3 2 40 R 3 c2 R ,
ret

(6.3.12)

our nal expression for the electric eld of a moving point charge. The electric eld is naturally decomposed into a velocity eld that decays as 1/R2 and an acceleration eld that decays as 1/R. The velocity eld represents that part of the electric eld that stays bound to the particle as it moves; it is the particles generalized Coulomb eld. The acceleration eld, on the other hand, represents electromagnetic radiation that propagates independently of the charge. To obtain the magnetic eld we take the curl of Eq. (6.3.11), which gives A = = = = so that B= q 1 40 c2 q 1 40 c2 q 1 40 c2 q 1 40 c2 s v 1 v (s) + (s) dt R R R v 1 1 v (s) + (s) ds R cR d R v 1 1 v (s) ds R ds cR 1 d R v 1 1 v (s) ds, R c dt R R v R .
ret

1 q 1 1 1 d v 40 c2 R c dt

v )/R2 , We now work on simplifying this expression. We have (1/R) v = (R and d(R v /R)/dt can be expressed as v /c) v d (R d = dt R dt v /c R R v+ v /c R a, R

6.3

Fields of an arbitrarily moving charge

119

which allows us to use results that were generated during the calculation of the electric eld. We have v /c) v 1 d (R c dt R = v /c (R a)(R v /c) a R R v + 2 2 2 + 2 R R 2 Rc2 v /c R + a cR v a)R v av v R (R R + 2 2 2+ = 2 2 R R Rc2 Rc2 a av R + + cR Rc2 v a)R v R v a R (R R = + 2 2 2+ + . 2 2 2 R R Rc cR

Collecting these results, we arrive at B= or B (t, x) = v a)R v /c + R a (R q 1 R + 40 c2 2 3 R2 c3 R .


ret

v a)R v R a R (R q 1 40 c2 2 3 R 2 3 Rc2 c2 R

,
ret

(6.3.13)

A little more algebra reveals that this can also be expressed as B= 1 R c


ret

E.

(6.3.14)

The magnetic eld also is decomposed into a velocity eld that stays bound to the particle, and an acceleration eld that represents electromagnetic radiation. Our general results for the electric and magnetic elds allow us to resolve the issue that was rst brought up near the end of Sec. 6.2: where does the electric eld point? Does it point toward r (t ), the retarded position of the particle, as it was argued it might do? Or does it point toward r (t), its current position, as we actually found in Sec. 6.2? The answer is that in general, the electric eld follows neither option. Consider our expression (6.3.12) for the bound electric eld. (We exclude the radiative eld from this discussion, as we expect it to behave largely independently of the charge.) It shows that the eld points in the direction opposite to the vector R + Rv /c evaluated at the retarded time t . If the observation point is at the origin of the coordinates (x = 0), then R = r and the bound eld points in the direction opposite to r (t ) + (t)v (t ), where t is determined by the retarded condition t t t = r(t )/c. This result is interesting: the eld points not in the direction opposite to the charges retarded position r (t ), nor in the direction opposite to the charges current position r (t), but in the direction opposite to the charges anticipated position ra r (t ) + (t)v (t ). This is where the charge would be after a time t if it were moving with a constant velocity v (t ); the time interval corresponds to the time required for a light signal to travel from the charges retarded position to the observation point. (The situation is described in Fig. 6.2.) So the eld brings information to the

120

Electrodynamics of point charges

retarded position v (t) r(t)

anticipated position ra

current position

Figure 6.2: Retarded position, current position, and anticipated position of an accelerated charge. observer that does indeed originate from the retarded position, but the eld orients itself as though it were anticipating the motion of the particle during the time the information propagates from the charge to the observer. Because the extrapolation is based only on partial information (the values of r and v at the retarded time, and nothing more), it is not fully accurate, and the eld misses the actual position of the particle. Notice that uniform motion is a special case: when the velocity is constant, the charges anticipated position ra coincides with its current position r (t), and the eld miraculously orients itself in this direction. This explains the result found in Sec. 6.2: the information needed time to reach the observer, but it was sucient to correctly anticipate the actual motion of the charge.

6.3.3 Uniform motion


Let us verify that the general expression of Eq. (6.3.12) reduces to Eq. (6.2.3) in the case of uniform motion in the x direction. We set r (t) = (vt, b, 0), r (t ) = (vt , b, 0) r , and v (t) = v (t ) = v x . We once more take the observation point to be at x = 0. Because a = 0, Eq. (6.3.12) reduces to q r q r + r v /c + v /c E= = , 2 3 2 40 r 40 2 3 r3 or q (r + r v /c) . (6.3.15) E= 40 (r )3 We recall that = 1 + v r /c. The retarded condition is r = and it gives, after squaring, (1 v 2 /c2 )t2 2tt + t2 b2 /c2 = 0. This is a quadratic equation for t , which we do not need to solve. (6.3.16) b2 + (vt )2 = c(t t ),

6.3

Fields of an arbitrarily moving charge

121

The numerator of Eq. (6.3.15) involves the vector r + r v /c. Its x component is vt + r v/c = vt + c(t t )v/c = vt, and this vector therefore coincides with r . In the denominator we have (r )2 raised to the power 3/2. This is (r )2 = 2 (r + v r /c)2 = 2 (r + v 2 t /c)2 = 2 c2 (t t + v 2 t /c2 )2 = 2 c2 (t t / 2 )2

= c2 ( 2 t2 2tt + t2 / 2 ). We now use Eq. (6.3.16) to eliminate t and obtain (r )2 = c2 b2 /c2 + 2 t2 (1 1/ 2 ) = b2 + 2 v 2 t 2 .

Substituting these results into Eq. (6.3.15) gives E= r (t) q , 40 (b2 + 2 v 2 t2 )3/2 (6.3.17)

the same statement as Eq. (6.2.3).

6.3.4 Summary
The scalar and vector potentials of a point charge in arbitrary motion described by the position vector r (t) are given by the Li enard-Wiechert expressions (t, x) = and A(t, x) = q0 v 4 R ,
ret

q 1 40 R

(6.3.18)
ret

(6.3.19)

= R/R, where R(t ) = x r (t ), R = |R|, R (t )/c, = 1 v (t ) R and where t is determined by the retarded condition |x r (t )| = c(t t ). The electric eld is E (t, x) = (R v /c) a v /c R q R + 2 3 2 40 R 3 c2 R 1 R c ,
ret

(6.3.20)

(6.3.21)

(6.3.22)

and the magnetic eld is B=


ret

E.

(6.3.23)

Both elds are naturally decomposed into velocity (bound) elds that decay as 1/R2 , and acceleration (radiative) elds that decay as 1/R.

122

Electrodynamics of point charges

S retarded position

R(t) current position


Figure 6.3: Sphere S centered at the retarded position r (t ) of the accelerated charge. The sphere has a constant radius R(t ) |x r (t )|.

6.4

Radiation from an accelerated charge

[The material presented in this section is also covered in Secs. 14.2, 14.3, and 14.4 of Jacksons text.] At large distances from the particle in the wave zone we can neglect the velocity elds and approximate the electric eld by E (t, x) = and the magnetic eld by B= 1 R c
ret

(R v /c) a R q 40 3 c2 R

,
ret

(6.4.1)

E.

(6.4.2)

= R/R, = 1 v (t ) R (t )/c, and We recall that R = x r (t ), R = |R|, R that t is determined by the retarded condition t t = R(t )/c. In this section we use these expressions to calculate how much power is radiated by the accelerated charge. We also determine the angular distribution of the radiated power.

6.4.1 Angular prole of radiated power


Because our expressions for the elds refer directly to the charges retarded position, it is a relatively simple task to calculate the power that crosses a sphere S centered at this position (see Fig. 6.3); it would be much more complicated to adopt any other reference point. The sphere S has a constant radius R(t ) and its element of (t ) d , where d is an element of solid angle on the surface area is da = R2 (t )R sphere. The energy crossing an element of S per unit time t is, by denition, S (t ) da , where S = E B /0 is the Poynting vector. But because all quantities are expressed in terms of retarded time t instead of t, it is convenient to convert this per-unitt-time ux to a per-unit-t -time measure. We have that S (t ) da dt dt

is the energy crossing an element of S per unit t -time, and the conversation factor (t )/c . The displayed quantity can is dt/dt = 1 + c1 dR(t )/dt = 1 v (t ) R

6.4

Radiation from an accelerated charge

123

(t )R2 (t ) d , which is a denition for dP = therefore be expressed as S (t ) R (dP /d ) d . We have arrived at dP d = R2 S R = (6.4.3)

energy crossing an element of S per unit t -time and unit solid angle.

Because it is understood that all quantities refer to the retarded time t , we no longer have a need for the redundant notation [ ]ret . Integrating Eq. (6.4.3) over d gives the total power crossing S . Integrating this over dt (which involves displacing S in response to the particles motion during integration) returns the total energy radiated by the charge. The Poynting vector is given by S= 1 1 E ) = 1 |E |2 R , EB = E (R 0 0 c 0 c

. This gives because E is orthogonal to R [(R v /c) a] 2 R dP q2 1 1 2 2 = R | E | = , d 0 c 0 c (40 )2 5 c4 or [(R v /c) a] 2 0 q 2 R dP = . d (4 )2 c 5

(6.4.4)

The general expression for the angular prole is complicated, and it depends on the individual directions of v , a, and R. To gain insight into this it will help to look at specic simple situations; we shall do so in the following subsections. To ease the notation we shall now agree to drop the primes on dP /d .

6.4.2 Slow motion: Larmor formula


Our rst special case will be the limit v c. In this slow-motion approximation in Eq. (6.4.4), and also set 1. We then have we can neglect v /c in front of R dP d 0 q 2 a) 2 R (R (4 )2 c 0 q 2 )R 2 a (a R (4 )2 c 0 q 2 )2 . |a|2 (a R (4 )2 c

, then |a|2 (a R )2 = |a|2 (1 If denotes the angle between the vectors a and R 2 2 2 cos ) = |a| sin , and we nally obtain 0 dP q |a| d (4 )2 c
2

sin2 .

(6.4.5)

This expression could have been derived on the basis of the electric-dipole approximation of Sec. 5.4. According to the results of that section, in particular Eq. (5.4.22), the angular prole of radiation produced by a slowly-moving source is given by 0 dP = |p |2 sin2 , d (4 )2 c

124

Electrodynamics of point charges

where p is the dipole moment of the charge distribution, and the angle between the vectors r x/|x| and p . Here p = q r (t), so that p = q a. And because the plays the same role as r sphere S is centered on the particle, the vector R , so that . The electric-dipole approximation therefore produces the same result as in Eq. (6.4.5). Integration of dP/d over all angles yields P 0 2 q |a| . 6c (6.4.6)

This result is known as Larmors formula.

6.4.3 Linear motion


In this subsection we no longer restrict the size of v/c, but we assume that v and a are pointing in the same direction. This condition might hold at all times, but it is sucient to assume that it holds only momentarily. At this moment the particle is and the common moving on a straight line. We now let be the angle between R direction of v and a. We have that v a = 0, so Eq. (6.4.4) becomes (R a) 2 0 q 2 R dP = . d (4 )2 c 5 (R a)|2 = |a|2 sin2 , and we recall that = 1v R /c = We already know that |R 1 v cos /c. Substituting these results gives 0 dP = q |a| d (4 )2 c
2

sin2 . (1 v cos /c)5

(6.4.7)

For v/c 1 this reproduces the sin2 behaviour encountered in the preceding subsection. For v/c approaching unity, however, the angular prole is strongly modied it tips forward in the direction of v ( = 0), and the power radiated at small angles is very large (see Fig. 6.4). For such values of v/c, the radiation is beamed in the forward direction. The angular function is f= sin2 . (1 v cos /c)5 (6.4.8)

Dierentiating with respect to cos and putting the result to zero indicates that the maximum of the distribution is at the value of that solves (3v/c)2 + 2 5v/c = 0. So most of the radiation is emitted at angles close to max , which is determined by 1 + 15(v/c)2 1 . (6.4.9) cos max = 3(v/c)
5 3 (v/c) 75 For v/c 1, this expression is well approximated by cos max 2 8 (v/c) + , and max /2: most of the radiation is emitted in the directions transverse to v . This is the usual nonrelativistic situation. For v/c close to unity, however, so 43 4 2 that 1, Eq. (6.4.9) is well approximated by cos max 1 1 512 + , 8 and max is very small: most of the radiation is emitted in the forward direction. 1 2 max + , and this gives us the For such small angles we have cos max = 1 2 approximate relation 1 , 1. (6.4.10) max 2

6.4

Radiation from an accelerated charge

125

0.5

0.4 0.2

0.2

0.4

0.5

4000 2000 0 2000 4000

5000

10000

15000

20000

25000

Figure 6.4: The rst plot shows the angular prole of radiation produced by a nonrelativistic particle in linear motion; the radiation propagates mostly in the directions transverse to the charges velocity (the horizontal direction in the plot). The second plot shows the angular prole of radiation produced by a relativistic particle in linear motion; the radiation is beamed in the forward direction with an angular width of order 1/(2 ). Notice the dierent scales in the two diagrams.

126

Electrodynamics of point charges

This is the statement that when the charge moves with a highly relativistic speed, the radiation is strongly beamed in the direction of the velocity vector. The total power radiated is obtained by integrating Eq. (6.4.7) with respect to d = sin dd. We have P = 0 2 q |a| 2 2 (4 ) c
0

sin3 d , (1 v cos /c)5

6 and the integral evaluates to 4 3 . The nal answer is therefore

P =

0 2 q |a| 6 . 6c

(6.4.11)

This generalizes the Larmor formula of Eq. (6.4.6). It is instructive to express this in terms of the force acting on the particle. This force is F = dp/dt, where p = m v is the particles relativistic momentum. For linear motion in the x direction we have px = mv/(1 v 2 /c2 )1/2 and (mv )(v/c2 )a ma dpx ma + = . = 2 2 1 / 2 dt (1 v /c ) (1 v 2 /c2 )3/2 (1 v 2 /c2 )3/2 This result can be expressed as F = m 3 a, and Eq. (6.4.11) becomes P = 0 q |F | 6c m
2

(6.4.12)

This is the total power radiated by a charge in instantaneous linear motion.

6.4.4 Circular motion


We now take a and v to be orthogonal vectors. This condition might hold at all times, but it is sucient to assume that it holds only momentarily. At this moment the particles motion is circular. For concreteness we let a be directed along the x axis, and v points in the z direction. We let the angles (, ) give the direction of relative to this set of axes. We therefore have a = ax the vector R , v = v z , and = sin cos x R + sin sin y + cos z . = v cos , a R = a sin cos , = 1 v cos /c, and According to this, v R (R v /c) a R = = a)(R v /c) R (R v /c)a (R a)(R v /c) a. (R

The squared norm of this vector is [(R v /c) a] R


2

a)2 (R v /c) (R v /c) (R a)(R v /c) a + 2 a2 2(R a)2 (1 2R v /c + v 2 /c2 2) + 2 a2 = (R a)2 (1 v 2 /c2 ) + 2 a2 = (R a (R )2 = 2 a2 1 . 2 2 =

Substituting this result into Eq. (6.4.4) gives a (R )2 0 q 2 a2 dP 1 , = 2 3 2 d (4 ) c 2

6.4

Radiation from an accelerated charge

127

400 0 400 2000 4000 6000 8000

Figure 6.5: Angular prole of radiation produced by a relativistic charge in circular motion. The radiation is beamed in the (horizontal) direction of the velocity vector. dP q 2 |a|2 sin2 cos2 0 1 . (6.4.13) = d (4 )2 c (1 v cos /c)3 2 (1 v cos /c)2 This angular prole is rather complex, and quantitatively dierent from the one encountered in the preceding subsection; but it still features a strong beaming in the forward direction (direction of v ) when v is close to the speed of light. This is illustrated in Fig. 6.5. The total power radiated is obtained by integrating dP/d over the angles and . The result is 0 2 q |a| 4 . (6.4.14) P = 6c This is a dierent generalization of the Larmor formula of Eq. (6.4.6). We may also express this in terms of the force F acting on the particle. In this case the force changes only the direction of the vector v , but not its magnitude. So F = dp/dt = d(m v )/dt = m a, and Eq. (6.4.14) becomes P = 0 q |F | 6c m
2

or

2.

(6.4.15)

This is the total power radiated by a charge in instantaneous circular motion. You should compare this with Eq. (6.4.12) and notice the extra factor of 2 .

6.4.5 Synchrotron radiation


We continue our discussion of circular motion and provide additional details. We now assume that the charge moves at all times on a circle of radius at a constant angular velocity . Its position vector is r (t ) = cos(t )x + sin(t )y , its velocity vector is v (t ) = v sin(t )x + cos(t )y , (6.4.17) (6.4.16)

128 where v = , and its acceleration vector is

Electrodynamics of point charges

a(t ) = v cos(t )x + sin(t )y .

(6.4.18)

We wish to calculate the electric eld that would be measured by a detector situated far away, in the wave zone. We place this detector on the y axis, at position x = ry , and we assume that r . The wave-zone electric eld is given by Eq. (6.4.1), which we rst evaluate as a function of retarded time t . The relation between t and true time t is given by Eq. (6.4.20). It is useful to recall that t is time measured by a clock that moves with the charged particle, while t is time measured by a clock attached to the detector. As we shall see, these measurements of time dier substantially when v is comparable to c, and this leads to an interesting modulation of the electromagnetic wave. We use the condition r to calculate E approximately. We shall need some level of precision in the calculation of t(t ), because it is this function which determines the phasing of the electric eld. A much cruder approximation will suce when calculating the elds amplitude. Working to rst order in /r, it is easy to see that the length of the vector R x r (t ) is given by R(t ) = r sin t . Equation (6.3.21) then gives t = t + r/c (/c) sin t , or (t r/c) = t v sin t . c (6.4.19)

This gives the relationship between t r/c, the retarded time measured at the detector, and t , the time measured at the particle. From this it follows that = dt v = 1 cos t . dt c (6.4.20)

We see that will deviate strongly from 1 when v is close to the speed of light and when | cos t | is close to unity. To perform the remaining calculations we allow ourselves to neglect all terms =y linear in /r. We take R , v /c = (v/c) sin(t )x R 1 (v/c) cos(t ) y , (R v /c) a = v (v/c cos t )x and R . Substituting all this into Eq. (6.4.1) gives E = E x , with E= q v v/c cos t . 40 c2 r [1 (v/c) cos t ]3 (6.4.21)

The electric eld can be expressed in terms of detector time t by inverting Eq. (6.4.19). Its behaviour is displayed in Fig. 6.6 for selected values of = [1 (v/c)2 ]1/2 . The plots reveal that for moderately larger than unity, the wave deviates strongly from a simple sinusoid. The phase modulations which originate from the discrepancy between t and t produce a wave that is characterized by two distinct timescales: a long one associated with the rotational frequency , and a short one associated with the pulse duration. The frequency spectrum of the radiation is presented in Fig. 6.7. These plots show that in the relativistic regime, the radiation is carried by a large number of harmonics of the fundamental frequency ; as increases the dominant harmonic is shifted to higher frequencies. To understand these features it is helpful to consider the extreme relativistic regime, in which 1. In this regime the basic mechanism behind the narrowing of the pulses is very easy to identify: When v/c 1 the denominator of Eq. (6.4.21)

6.4

Radiation from an accelerated charge

129

1 field kappa

2.5 field kappa 2

0.5 1.5

0.5 -0.5 0

-1 -0.5

0.5 1 1.5 synchrotron radiation: gamma = 1.0, v/c = 0.

2.5

-0.5 -0.5

0.5 1 1.5 synchrotron radiation: gamma = 1.2, v/c = 0.55

2.5

3 field kappa 2.5

3.5 field kappa 3

2.5 2 2 1.5 1.5 1 1 0.5 0.5 0

-0.5 -0.5

0.5 1 1.5 synchrotron radiation: gamma = 1.4, v/c = 0.70

2.5

-0.5 -0.5

0.5 1 1.5 synchrotron radiation: gamma = 1.6, v/c = 0.78

2.5

3.5 field kappa 3

3.5 field kappa 3

2.5

2.5

1.5

1.5

0.5

0.5

-0.5 -0.5

0.5 1 1.5 synchrotron radiation: gamma = 1.8, v/c = 0.83

2.5

-0.5 -0.5

0.5 1 1.5 synchrotron radiation: gamma = 2.0, v/c = 0.87

2.5

Figure 6.6: Electric eld produced by a charged particle in circular motion, as a function of detector time t, for selected values of the relativistic factor. The plots also show = dt/dt , whose deviation from 1 measures the magnitude of the relativistic time-dilation eect. The relativistic modulation of the wave pulse for large values of v/c is clearly apparent. In particular, the plots reveal that the radiation is characterized by two distinct time scales: a long one associated with the rotational frequency , and a short one associated with the pulse duration.

130

Electrodynamics of point charges

10 9 8 7

4.5

3.5

3 6 2.5 5 2 4 1.5 3 2 1 0 0 1 2 3 synchrotron radiation: gamma = 1.0, v/c = 0, wc = 1. 4 5 1

0.5

0 0 2 4 6 8 10 12 synchrotron radiation: gamma = 1.2, v/c = 0.55, wc = 1.7 14

2.5

1.8

1.6 2

1.4

1.2 1.5 1

0.8 1 0.6

0.5

0.4

0.2

0 0 5 10 15 20 25 30 synchrotron radiation: gamma = 1.4, v/c = 0.70, wc = 2.7 1.2

0 0 5 10 15 20 25 30 35 40 45 50 synchrotron radiation: gamma = 1.6, v/c = 0.78, wc = 4.1 0.9

0.8 1 0.7

0.8

0.6

0.5 0.6 0.4

0.4

0.3

0.2 0.2 0.1

0 0 10 20 30 40 50 60 70 80 synchrotron radiation: gamma = 1.8, v/c = 0.83, wc = 5.8

0 0 20 40 60 80 100 120 synchrotron radiation: gamma = 2.0, v/c = 0.87, wc = 8.0

Figure 6.7: Fourier transform of the electric eld produced by a charged particle in circular motion, for selected values of the relativistic factor. The plots reveal that the number of relevant harmonics of the fundamental frequency increases with . In addition, the dominant harmonic is shifted to higher frequencies.

6.4

Radiation from an accelerated charge

131

approaches zero whenever cos t = 1, which happens whenever t is a multiple of 2 . Most of the radiation is therefore emitted during a very short portion of the particles orbit, just as the particle crosses the x axis in the positive sense. This radiation is strongly beamed in the forward direction, and it travels along the y axis toward the detector. During the remaining portion of the orbit little radiation is emitted, and the detector must wait for the completion of a full orbit before receiving the next pulse. 2 + O( 4 ) and we To express these ideas mathematically we write v/c = 1 1 2 1 2 approximate sin t t and cos t 1 2 (t ) . From Eq. (6.4.20) we obtain 1 1 + ( t )2 , 2 2

which shows that is small (of order 2 1) for a duration of order ( )1 in terms of particle time t . We also obtain v/c cos t and Eq. (6.4.21) reduces to E q v 4 1 ( t )2 . 4 40 c2 r [1 + ( t )2 ]3 (6.4.22) 1 1 ( t )2 , 2 2

Again, this shows that in terms of particle time t , the radiation is on for a duration of order ( )1 . To see how long the pulse lasts in terms of detector time t, we approximate 1 2 t , which gives Eq. (6.4.19) as (t r/c) 2 t 2 3 (t r/c). (6.4.23)

This relation reveals that the pulse duration, as measured by a clock attached to the detector (and as displayed in Fig. 6.6) is given by tpulse 1 3 . (6.4.24)

It is reduced by a factor 3 1 relative to the orbital timescale associated with the angular velocity . As a consequence, the characteristic frequency c of the radiation is shifted relative to the fundamental frequency . It is given by c 3 , and the displacement of the spectrum toward higher frequencies when increases is seen very clearly in Fig. 6.7.

6.4.6 Extremely relativistic motion


To conclude this section we note that at any point on the arbitrary trajectory of a point charge, the applied force can be decomposed into a component that is parallel to the velocity vector v , and a component F that is perpendicular to it (F v = 0). When the motion is extremely relativistic, 1 and the total power emitted by the particle is well approximated by Eq. (6.4.15), P = 0 q |F | 6c m
2

2.

The contribution from the parallel force is smaller by a factor of 1/ 2 1, and it can be neglected. In this limit, therefore, the parallel force plays no role in the energetics of the radiation.

132

Electrodynamics of point charges

6.5

Radiation reaction

[The material presented in this section is also covered in Secs. 16.1, 16.2, and 16.3 of Jacksons text.] In Sec. 5.6 we saw that a classical electron orbiting a classical proton emits electromagnetic radiation. As a result, the classical hydrogen atom loses energy at a rate determined by Larmors formula, 0 dE 2 = P = q |a| . dt 6c This loss of energy translates into a decrease in the electrons orbital radius. The orbit gradually shrinks, and the classical atom is unstable to the emission of electromagnetic radiation. It is natural to assume that the electron is subjected only to the electric eld produced by the proton. This picture, however, is incomplete: So long as the electron feels only the protons Coulomb eld, its motion must stay circular (or more generally, elliptical). The electrons inspiraling motion cannot take place in this picture, in contradiction with our previous description. We are forced to conclude that the protons electric eld cannot be the only eld acting on the electron. The electron must also be subjected to its own electric eld. It is the electrons self-eld which provides the force that drives the electrons inspiraling motion. This conclusion is dicult to understand, because the electrons own eld diverges at the electrons position. How can it then produce a (presumably nite) radiation-reaction force that drives the inspiral? The answer is that the divergent part of the electrons self-eld exerts no force it just travels along with the electron. The part of the self-eld that is responsible for the radiation reaction is nite. To gure out what this part is, let us ask what would happen if instead of imposing outgoing-wave boundary conditions on our spherical-wave solutions to Maxwells equations, we were to impose incoming-wave boundary conditions. In this case we would have incoming waves bringing energy to the system instead of outgoing waves taking energy away. The atoms energy would then increase, and the electron would move out instead of in. We would, in fact, have a mirror image (more precisely, a time reversal) of the correct picture, and the radiation-reaction force would be acting in the opposite direction (pushing the orbit out instead of in). We can gure out which part of the electrons eld is producing the radiationreaction force by examining what happens when we switch from outgoing-wave boundary conditions to incoming-wave boundary conditions. We will see that while part of the eld is not aected by this switch, the remaining part changes sign. And we will see that while the unaected part is innite at the electrons position, the remaining part is nite. It is this part of the eld that is responsible for the radiation reaction. Let us examine the vector potential rst. The physical solution to Maxwells equations is the retarded solution, A (t, x) = 0 4 j (t |x x |/c, x ) 3 d x, |x x |

where j is the electrons current density. This describes waves that are properly outgoing in the wave zone. These waves remove energy from the system, and the radiation reaction drives the electron inward. An unphysical solution to Maxwells equations is the advanced solution, A+ (t, x) = 0 4 j (t + |x x |/c, x ) 3 d x, |x x |

6.5

Radiation reaction

133

and this describes waves that are incoming in the wave zone. These waves bring energy to the system, and the radiation reaction drives the electron outward. To isolate the part of the (retarded) vector potential that is responsible for the radiation reaction, we must examine both solutions simultaneously, and determine what changes when we choose one sign over the other. We want to evaluate A near the electron, in the near zone. We can therefore take |x x | to be small, and Taylor-expand the current density: j (t |x x |/c) = j (t) |x x | j (t) + . c t

We see that the rst term is the same irrespective of the choice of boundary conditions (outgoing waves versus incoming waves); it cannot, therefore, contribute to a radiation-reaction force. The second term, however, changes sign under a change of boundary conditions, and the force it contributes will also change sign; this must be the vector potentials contribution to the reaction-reaction force. Because the retarded solution provides the correct sign, we shall write Arr (t, x) = 0 d 4c dt j (t, x ) d3 x + , (6.5.1)

Relative to the leading term, the second term is of order (rc /tc )/c = vc /c, and we shall take this to be small we are working under a slow-motion approximation. We have 0 j (t, x ) 3 0 d A = j (t, x ) d3 x + . d x 4 |x x | 4c dt

and we will see that the elds this potential generates produce the radiation-reaction force that acts on the electron. For the scalar potential we have the two solutions (t, x) = 1 40 (t |x x |/c, x ) 3 d x, |x x |

and we Taylor-expand the charge density: (t |x x |/c) = (t) |x x | |x x |2 2 (t) + (t) c t 2c2 t2 |x x |3 3 (t) + . 6c3 t3

This gives = 1 (t, x ) 3 1 d (t, x ) d3 x d x 40 |x x | c dt 1 2 (t, x )|x x | d3 x + 2 2 2c t 1 3 (t, x )|x x |2 d3 x + . 3 3 6c t

The rst and third terms are insensitive to boundary conditions; they have nothing to do with radiation reaction. The second and fourth terms, on the other hand, change sign under a switch of boundary conditions, and they are associated with radiation reaction. The second term, however, vanishes by virtue of charge conservation, and we conclude that the radiation-reaction part of the scalar potential is given by (recall that c2 = 0 0 ) rr (t, x) = 0 1 3 4c 6 t3 (t, x )|x x |2 d3 x + . (6.5.2)

134

Electrodynamics of point charges

Up to now the charge and current densities have been left unspecied, and our considerations have been general: Eqs. (6.5.1) and (6.5.2) give the radiation-reaction potentials for an arbitrary distribution of change and current. These expressions will apply, in particular, to a point charge q if we set (t, x ) = q (x r (t)) and j (t, x ) = q v (t) (x r (t)), where v = dr /dt. Substituting this into our expressions for the potentials returns rr (t, x) = and 0 q 3 x r (t) 4c 6 t3
2

0 d q v (t). 4c dt We have |x r |2 = (x r ) (x r ), and dierentiating once with respect to t gives 2(x r ) v . Dierentiating once more gives 2(x r ) a +2v v , and dierentiating a third time yields 2(x r ) a + 2v a + 4v a = 2(x r ) a + 6v a. We nally arrive at 1 0 q (x r ) a va (6.5.3) rr (t, x) = 4c 3 and 0 q a. (6.5.4) Arr (t, x) = 4c Here we have dened a = dv /dt and a = da/dt. These are the radiation-reaction potentials acting on a point charge q moving on a trajectory r (t). The radiation-reaction part of the electric eld is given by Arr (t, x) = Err = or Because Arr Arr 0 1 rr = q a a , t 4c 3 Err = (6.5.5)

0 qa . 6c does not actually depend on x, we also have Brr = 0.

(6.5.6)

The radiation-reaction force is therefore Frr = q Err , or 0 2 . (6.5.7) q a Frr = 6c This force acts on the point charge, and it originates from the charges own electric eld! It is easy to verify that the radiation-reaction force does negative work on the particle and takes energy away from it. As the charge moves with velocity v the force does work at a rate = Frr v = 0 q 2 a W v. 6c We write d a v = (a v ) |a|2 , dt over a time interval T . This gives and we average W W = = 1 T
T 0

dt W 1 T d 1 (a v ) dt T 0 dt T T 1 (a v ) |a|2 . T 0
T 0

0 2 q 6c 0 2 q 6c

|a|2 dt

6.6

Problems

135

Supposing now that the motion is periodic with period T (as it would be in the case of atomic motion), or else that it begins and ends with a vanishing acceleration, we may discard the boundary terms and obtain = 0 q 2 |a|2 = P , W 6c (6.5.8)

where P is the power radiated by the point charge, as given by the Larmor formula of Eq. (6.4.6). The work done by the radiation-reaction force therefore matches the energy taken away by electromagnetic radiation, and we have energy conservation. (It is noteworthy that this statement holds only on the average.) We have a complete resolution of the paradox stated at the beginning of this section. The electrons own eld, or at least the nite part that is sensitive to the choice of boundary conditions, acts on the electron and prevents it from keeping its circular motion. (Indeed, it is easy to check that for a circular orbit, the radiationreaction force of Eq. (6.5.7) points in the direction of v and drives the orbit to a smaller radius.) Furthermore, the work done by this radiation-reaction force matches the amount of energy taken away by the radiation, as it well should if energy is to be conserved in the process.

6.6

Problems

1. A point charge q moves on a circle of xed radius a with an angular frequency (t) that changes slowly with time. (The time scale over which changes is much longer than the rotational period.) We take a to be smaller than c, but we do not assume that it is much smaller than c; the charge has a Lorentz -factor that can be large compared with unity. The charge emits electromagnetic radiation. It therefore loses energy and this causes the angular frequency to change with time. You are asked to derive an equation for d/dt and to express your result in terms of , a, and other relevant constants. (You may need to know that the charges relativistic energy is E = mc2 , where m is the charges rest mass.) 2. (This problem is adapted from Jacksons problem 14.12.) A charge q moves in simple harmonic motion along the z axis. Its position is given as a function of t by z (t ) = z0 cos(t ), where z0 and are constants. It is assumed that z0 /c is fairly close to unity, so that the motion is relativistic. a) Show that the time-averaged angular prole of the radiation is described by 2 2 dP 0 2 2 1 + (/2) cos sin2 , = ( qz ) 0 d 32 2 c (1 2 cos2 )7/2 where is the angle from the z axis. The time average is dened by T ( ) T 1 0 ( ) dt , where T 2/ is the period. b) Produce a parameteric plot of the angular function for = 0.95. Comment on its appearance in view of the beaming eect discussed in Sec. 6.3.3. 3. A charge q moves on a trajectory described by z (t ) = b2 + (ct )2 , where b is a constant. The particle comes in from innity, it turns around at z = b when t = 0, and it eventually returns to innity. Calculate P , the total power emitted by the particle. Show in particular that this does not depend on time t .

136

Electrodynamics of point charges

4. (This problem is adapted from Jacksons problem 14.7.) A nonrelativistic particle of charge q , mass m, and initial speed v is incident on a xed charge Q with an impact parameter b that is large enough to ensure that the particles deection in the course of the collision is very small. Calculate the total energy radiated by the particle. Because the deection angle is so small, it is sucient to assume that the particles trajectory is a straight line and that its velocity vector is constant. But this approximation must be introduced only after you have calculated the particles acceleration.

You might also like