PHYS370 Advanced Electromagnetism

Part 7: Electromagnetic Potentials

Electromagnetic Potentials

In this section, we consider:

The electric scalar and magnetic vector potentials. The wave equations for the electromagnetic potentials.

Advanced Electromagnetism

Part 7: Electromagnetic Potentials

Potentials

A potential is a function whose derivative gives a eld. Fields are associated with forces; potentials are associated with energy. The magnetic vector potential A is dened so that the magnetic eld B is given by: B= A (1)

The electric scalar potential is dened so that the electric eld E is given by: A t Note that in general, the scalar and vector potentials are functions of position and time. E = (2)

Advanced Electromagnetism

Part 7: Electromagnetic Potentials

Electrostatic Potential

The electric eld in the presence of a static charge distribution (r) is found from Coulombs law: 1 (r )(r r ) dV (3) 40 |r r |3 where the integral extends over all space. Note that the prime on the coordinates indicates that the coordinate is associated with the charge. E(r) =
Advanced Electromagnetism

Part 7: Electromagnetic Potentials

Electrostatic Potential

Calculating the potential is simpler than calculating the eld directly; and one can then use E = to nd the electric eld.
Advanced Electromagnetism

In terms of the scalar potential, for a static charge distribution, we have: (r ) 1 dV (4) (r) = 40 |r r |

Part 7: Electromagnetic Potentials

Electrostatic Potential Since we have from Maxwells equations: D = where D = E it follows that in an homgeneous, isotropic medium: E = = and so: 2 (r) = (r) Equation (8) is called Poissons equation. (6) (5)

(7)

(8)

Equation (4) is the solution to Poissons equation, expressed as an integral.

Advanced Electromagnetism

Part 7: Electromagnetic Potentials

Electrostatic Potential

The behaviour of a charged particle in an electric eld is determined by the eld E, rather than by the potential. Since E = for an electrostatic eld, we can add any function with vanishing gradient to the potential , and obtain the same physics. In other words, the behaviour of any electrostatic system is the same under the transformation: (r) (r) + 0 where 0 is a constant (independent of position). The freedom that we have in choosing the potential is called gauge invariance. This allows us to choose arbitrarily the point at which (r) = 0. (9)

Advanced Electromagnetism

Part 7: Electromagnetic Potentials

Electrostatic Potential

Note that if we write the solution to Poissons equation (4):

(r) =

1 40

(r ) dV |r r |

(10)

then implicitly (assuming that all charges are within a nite distance from the origin), we make the gauge choice:

(r) 0

as

|r|

(11)

Advanced Electromagnetism

Part 7: Electromagnetic Potentials

Magnetic Vector Potential

In the presence of sources for the magnetic eld (i.e. a current distribution), the magnetic eld B can be found from the Biot-Savart law: B(r) = 0 4 J(r ) (r r ) dV |r r |3 (12)

where the integral extends over all space.

Generally, the Biot-Savart law is dicult to apply. It is often easier to rst calculate the magnetic vector potential; but rst, we need to derive the dierential equation for the vector potential.

Advanced Electromagnetism

Part 7: Electromagnetic Potentials

Magnetic Vector Potential

In a static case (constant elds, charges and currents), the magnetic eld is related to the current density by: B = J Substituting B = A, and using the vector identity: we nd:
2A

(13)

A A = J

2A

(14)

(15)

This looks like a complicated equation; but there is a way to simplify it...

Advanced Electromagnetism

Part 7: Electromagnetic Potentials

Magnetic Vector Potential

Suppose that: A=f where f is some function of position. Let us dene a new vector potential A : A =A+ Since: 0 0 (18) 0 (17) (16)

for any function 0, the new vector potential A gives exactly the same magnetic eld as the old vector potential A.

Advanced Electromagnetism

10

Part 7: Electromagnetic Potentials

Magnetic Vector Potential However, if we choose 0 such that:

2 0 = f 2

(19)

then: A = A+
0=0

(20)

In other words, given a vector potential, we can always choose to work with another vector potential that gives the same eld as the original one, but that has zero divergence. Assuming that we make such a choice, then equation (15) for the vector potential becomes:
2 A = J

(21)

This is again Poissons equation or rather, three Poisson equations, one for each component of the vectors involved.
Advanced Electromagnetism

11

Part 7: Electromagnetic Potentials

Magnetic Vector Potential

Since we already know the solution to Poissons equation for the scalar potential, we can immediately write down the solution to Poissons equation for the vector potential: A(r) = 0 4 J(r ) dV |r r | (22)

This integral is generally easier to perform than the one involved in the Biot-Savart law. Once we have obtained the vector potential, we can derive the magnetic eld from B = A.

Advanced Electromagnetism

12

Part 7: Electromagnetic Potentials

Gauge Transformations of the Magnetic Vector Potential Notice that to derive equation (22), we made use of a gauge transformation of the vector potential: the magnetic eld is unchanged if we make the transformation: AA+ for any scalar function 0. In particular, we made a gauge transformation so that: A=0 (24) 0 (23)

The gauge condition (24) is known as the Coulomb gauge, and is implicit in equation (22). The Coulomb gauge is a convenient choice for static systems; but as we shall see later, there is a better choice for dynamic systems.
Advanced Electromagnetism

13

Part 7: Electromagnetic Potentials

Summary (so far) For time-independent elds, we can perform calculations more simply using the electric scalar and magnetic vector potentials. The potentials obey Poissons equation in the presence of sources: (25) The physics is invariant under gauge transformations of the scalar and vector potentials: + 0, AA+ 0 (26)
2 (r) = (r) , 2 A(r) = J(r)

where 0 is a constant, and 0 is any scalar eld. One possible choice of gauge is such that: (|r| = ) = 0, 1 40 (r ) dV , |r r |
14

A=0 0 4 J(r ) dV |r r |

(27)

The potentials can be calculated directly from the sources: (r) = A(r) = (28)

Advanced Electromagnetism

Part 7: Electromagnetic Potentials

Wave Equations with Sources We are interested in the case of electromagnetic waves produced by time-dependent sources, i.e. charge and current distributions that vary with time. Note that we can (somewhat articially) divide electric currents into two sorts:

External currents that cause electromagnetic waves. Induced currents caused by electromagnetic waves. We will consider only the case where the conductivity of the medium is zero; then we can neglect induced currents, and include only external sources of power.
Advanced Electromagnetism

15

Part 7: Electromagnetic Potentials

Wave Equations with Sources

As usual, we start from Maxwells equations: D = E = B Taking the curl of E = ( E we nd: E)

2 E =

B =0 H =J +D H = J E (29)

We nd that the wave equation for the electric eld, in the presence of sources, is:
2 2 E E = J + t2 t

(/)

(30)

Advanced Electromagnetism

16

Part 7: Electromagnetic Potentials

Wave Equations with Sources Following a similar procedure, starting by taking the curl of H, we obtain the wave equation for the magnetic eld H:
2 2 H H = t2

(31)

Note that as the sources and J approach zero, we obtain the usual source-free wave equations. Also note that the wave equation for the magnetic eld has only a source term dependent on the electric current J, whereas the wave equation for the electric eld includes a source term for the electric charge . This is an consequence of the fact that there are no magnetic monopoles.
Advanced Electromagnetism

17

Part 7: Electromagnetic Potentials

Wave Equations with Sources Equations (30) and (31) are best solved using the electric scalar and magnetic vector potentials, and with an appropriate gauge choice. For static electromagnetic elds, we have already seen the Coulomb gauge: A=0 AA+ 0 (32)

Recall that by making an appropriate gauge transformation: (33)

we can x A to be anything we like, while leaving A unchanged. For time-dependent elds, we make the choice of gauge: =0 (34) A + t The condition (34) is called the Lorenz gauge. Working in the Lorenz gauge simplies the solution of the wave equations.
Advanced Electromagnetism

18

Part 7: Electromagnetic Potentials

Wave Equations for the Potentials Let us substitute: B= A, E = A (35)

into Maxwells equation: B = J + E We obtain: ( A)

2 A = J

(36)

(37)

Imposing the Lorenz gauge (34): A + = 0 we obtain:

2 2 A A = J t2

(38)

(39)

This is the wave equation for the vector potential A. Note that in the static case, it reduces to equation (21).
Advanced Electromagnetism

19

Part 7: Electromagnetic Potentials

Wave Equations for the Potentials Now we substitute: E = A into Maxwells equation: E = to obtain: 2 Imposing the Lorenz gauge (34): A + = 0 we nd:
2 2 = t2

(40)

(41)

A=

(42)

(43)

(44)

Equation (44) is the wave equation for the electric scalar potential, with sources. In the static case, it reduces to Poissons equation (8).
Advanced Electromagnetism

20

Part 7: Electromagnetic Potentials

Wave Equations for the Potentials We have derived, using Maxwells equations and the Lorenz gauge condition, the wave equations (39) and (44) for the magnetic vector and electric scalar potentials:
2 2 t2 2 2 t2

A = J =

(45) (46)

Let us take the divergence of equation (45), plus multiplied by the time derivative of equation (46):
2 2 t2

A +

= t

J +

(47)

But from the Lorenz gauge condition (34), the left hand side of equation (47) must be zero. Hence: =0 (48) t Equation (48) is the continuity equation, that expresses the local conservation of electric charge. J +
Advanced Electromagnetism

21

Part 7: Electromagnetic Potentials

Wave Equations for the Potentials

In free space, = 0 and = 0. Recall that: c= 1 00 (49)

where c is the speed of light in a vacuum. We dene the dAlembertian operator : =

2 2 1 c2 t2

(50)

In terms of the dAlembertian operator, the wave equations for the potentials in free space can be written: A = 0J = 0 (51) (52)

Advanced Electromagnetism

22

Part 7: Electromagnetic Potentials

Solution to the Wave Equation with Sources The nal step is to write down the solution to the wave equations (39) and (44) for the vector and scalar potentials in the presence of sources. The equations have essentially the same form, so let us consider just the equation for the scalar potential:
2 2 = t2

(53)

In the absence of any charge, we know there are solutions in which changes in propagate through space at speed v = 1/ . In the presence of a static charge, we know that a solution can be written: (r ) 1 (54) dV (r) = 4 |r r |
Advanced Electromagnetism

23

Part 7: Electromagnetic Potentials

Solution to the Wave Equation with Sources Putting aside conservation of charge for a moment, imagine a system in which a point charge q appears at some point r in space at time t , and then disappears a moment later. An observer measuring the potential at point r and time t will nd a potential: q (55) = 4|r r | if: t=t + |r r | v (56)

This suggests the solution to the wave equation with sources: (r, t) = 1 4 (r , t ) dV |r r | (57)

where t and t are related by (56).

Advanced Electromagnetism

24

Part 7: Electromagnetic Potentials

Solution to the Wave Equation with Sources

Equation (57) is in fact a correct solution to the wave equation with sources. The integral is very similar to the static case; but the nite speed of propagation of signals through space is taken into account by evaluating the charge density at the source at an earlier time than the observation time.
Advanced Electromagnetism

25

Part 7: Electromagnetic Potentials

Solution to the Wave Equation with Sources

Having obtained the solution to the wave equation for the scalar potential, we can immediately write down the solution to the wave equation for the vector potential: A(r, t) = where, as before: t=t + 4 J(r , t ) dV |r r | |r r | v (58)

(59)

In the next part of the course, we shall apply these equations to nd the electric and magnetic elds generated by an oscillating dipole. In other words, we shall investigate the generation of electromagnetic waves.

Advanced Electromagnetism

26

Part 7: Electromagnetic Potentials

Electromagnetic Potentials: Summary You should be able to: Write expressions for the electric and magnetic elds in terms of the scalar and vector potentials. Explain that under a gauge transformation of the scalar and vector potentials, the electric and magnetic elds remain unchanged. Starting from Maxwells equations and the expressions for the elds in terms of the potentials, derive wave equations for the potentials in the Lorenz gauge. Write integral expressions for the scalar and vector potentials in terms of the source charges and currents, for both static and dynamic systems.
Advanced Electromagnetism

27

Part 7: Electromagnetic Potentials