Radiation — antennas, scattering, and radiation reaction

Intuition Beginner

A radio transmitter does not push current into empty space for nothing. The current sloshing up and down a metal wire accelerates the conduction electrons, and accelerating charges radiate electromagnetic waves — the same Larmor mechanism that makes a single shaken charge spit out light 10.07.01. A device engineered to turn a feed current into waves that fly away is called an antenna. The cleaner the conversion, the better the antenna.

The same physics runs in reverse. When an electromagnetic wave sweeps over a free electron, its electric field shakes the electron, and the shaking electron re-radiates. That re-radiation is called scattering. A single electron is a feeble scatterer: sunlight carries about 1360 watts per square metre, but one electron scatters only about $10^{-25}$ watts of it. Stack $10^{25}$ electrons into a cubic metre of air and the reradiated light paints the sky blue, because the scattering probability rises steeply at short (blue) wavelengths.

There is a third, unsettling consequence. If a charge radiates, it loses energy. That energy must come from the charge's own motion, so the radiation exerts a tiny recoil force back on its source — the radiation reaction. For an electron the recoil is almost always negligible, but it is never quite zero, and a careful treatment exposes a famous sickness: the classical equation that describes it predicts runaway speeds unless it is repaired.

The three threads share one root. Antennas, scattering, and radiation reaction are three faces of the fact that an accelerating charge couples to a $1/r$ field that carries energy to infinity 10.07.02. This unit assembles the three capstone applications that close out the radiation chapter.

Visual Beginner

System	Driven by	Key number
Half-wave dipole antenna	Feed current $I_0\cos \omega t$	$R_{\mathrm{rad}}\approx 73\Omega$
Free electron in sunlight	Incident wave, $I=1360$ W/m${}^2$	${\sigma }_T\approx 6.65\times 10^{-29}$ m${}^2$
Radiating charge	Its own emission	${\tau }_0\approx 6.3\times 10^{-24}$ s

The half-wave dipole is the workhorse of radio engineering because its radiation resistance sits next to the $75\Omega$ impedance of standard coaxial cable, so almost all the transmitter power leaves as waves instead of reflecting back into the amplifier. The scattering column records how small a target one electron presents. The reaction column records the absurdly short timescale on which a free electron would shed its own kinetic energy if nothing replenished it.

Worked example Beginner

Part A — a half-wave dipole. An FM transmitter feeds a half-wave dipole at $f=100$ MHz with a peak current $I_0=1.0$ A. The wavelength is $\lambda =c/f=3\times 10^8/10^8=3$ m, so the antenna is $1.5$ m long. The radiation resistance is $R_{\mathrm{rad}}=73\Omega$. Time-averaging the oscillation introduces a factor of one half:

$$⟨P⟩=\frac{1}{2}{I}_0^2{R}_{\mathrm{rad}}=\frac{1}{2}(1.0{)}^2(73)\approx 36.5\text{ W}.$$

About $37$ watts sail off as radio waves. The number $73\Omega$ is the price the feed current pays, measured in ohms, to fill space with radiation.

Part B — an electron in sunlight. The same sunlight delivers $I=1360$ W/m${}^2$. Each free electron scatters a slice of that flux equal to the Thomson cross-section ${\sigma }_T=6.65\times 10^{-29}$ m${}^2$:

$$P={\sigma }_{T}I=(6.65\times 10^{-29})(1360)\approx 9.0\times 10^{-26}\text{ W}.$$

A single electron reradiates about a tenth of a billionth of a billionth of a watt. The electron is a near-invisible scatterer, which is why clean air is transparent and only the immense column of air above us, holding roughly $10^{29}$ electrons per square metre, scatters enough blue light to colour the sky.

Check your understanding Beginner

Formal definition Intermediate+

Antennas

A linear antenna is a thin conducting wire of length $\ell$ carrying a specified current distribution $I(z,t)=\mathrm{Re}[\tilde{I}(z)e^{-i\omega t}]$ along its axis. In the radiation zone ($r\gg \lambda$) the vector potential reduces to

$$\mathbf{A}(\mathbf{r},t)=\frac{{\mu }_0{e}^{ikr}}{4\pi r}\hat{\mathbf{z}}{\int }_{-\ell /2}^{\ell /2}\tilde{I}(z^{\prime })e^{-ik{z}^{\prime }\cos \theta }d{z}^{\prime },$$

where $k=\omega /c$ and $\theta$ is the polar angle measured from the wire axis. The far-zone electric field is the transverse piece of $-{\partial }_t\mathbf{A}$:

$${\mathbf{E}}_{\mathrm{rad}}(r,\theta ,t)=\frac{ik{\mu }_0{e}^{i(kr-\omega t)}}{4\pi r}\sin \theta {\int }_{-\ell /2}^{\ell /2}\tilde{I}(z^{\prime })e^{-ik{z}^{\prime }\cos \theta }d{z}^{\prime }\hat{\mathbf{\theta }}.$$

The radiation resistance is the effective load presented to the feed, defined by equating radiated power to ohmic dissipation in a fictitious resistor:

$$⟨P⟩=\frac{1}{2}{R}_{\mathrm{rad}}|{\tilde{I}}_{\mathrm{feed}}{|}^2.$$

For a centre-fed half-wave dipole ($\ell =\lambda /2$) the current is $\tilde{I}(z)=I_0\cos (kz)$ (vanishing at the ends, as physically required), and evaluation of the integral gives $R_{\mathrm{rad}}=73.1\Omega$ with directivity $1.64$ ($2.15$ dBi).

An antenna array of $N$ identical elements separated by spacing $d$ and fed with progressive phase shift $\alpha$ has a radiation pattern that factorises:

$${\left(\frac{dP}{d\Omega }\right)}_{\mathrm{array}}={\left(\frac{dP}{d\Omega }\right)}_{\mathrm{element}}\times |A(\psi ){|}^2,A(\psi )=\frac{\sin (N\psi /2)}{N\sin (\psi /2)},\psi =kd\cos \theta +\alpha .$$

The second factor, the array factor, depends only on geometry and phasing — not on the element. Steering the beam amounts to tuning $\alpha$, which is how phased-array radars swing a beam electronically with no moving parts.

Scattering

The differential scattering cross-section is defined as the time-averaged reradiated power per unit solid angle, divided by the incident intensity:

$$\frac{d\sigma }{d\Omega }=\frac{⟨d{P}_{\mathrm{scat}}/d\Omega ⟩}{⟨I_{\mathrm{inc}}⟩}.$$

The total cross-section is its integral, $\sigma =\int (d\sigma /d\Omega )d\Omega$. For a free electron driven by a low-frequency wave, the result is Thomson scattering with

$$\frac{d{\sigma }_T}{d\Omega }=\frac{r_e^2}{2}(1+{\cos }^2\theta ),{\sigma }_T=\frac{8\pi }{3}{r}_e^2,$$

where $r_e=e^2/(4\pi {\epsilon }_0{m}_e{c}^2)\approx 2.818\times 10^{-15}$ m is the classical electron radius and $\theta$ is the scattering angle. For a bound, sub-wavelength scatterer of static polarisability ${\alpha }_0$, the result is Rayleigh scattering,

$${\sigma }_{\mathrm{Ray}}=\frac{k^4|{\alpha }_0{|}^2}{6\pi {\epsilon }_0^2}\propto {\omega }^4,$$

and the strong ${\omega }^4$ rise is what paints the sky blue. At photon energies approaching $m_e{c}^2$ the classical result is replaced by the quantum Klein–Nishina formula, which falls below ${\sigma }_T$ as energy grows.

Radiation reaction

The Abraham–Lorentz force is the non-relativistic self-force that accounts for the energy carried away by a charge's own radiation:

$${\mathbf{F}}_{\mathrm{rad}}=\frac{q^2}{6\pi {\epsilon }_0{c}^3}\ddot{\mathbf{v}}=m{\tau }_0\ddot{\mathbf{v}},{\tau }_0=\frac{q^2}{6\pi {\epsilon }_{0}m{c}^3}=\frac{2r_e}{3c}.$$

For the electron ${\tau }_0\approx 6.26\times 10^{-24}$ s. The equation of motion is third-order in time,

$$m\dot{\mathbf{v}}={\mathbf{F}}_{\mathrm{ext}}+m{\tau }_0\ddot{\mathbf{v}},$$

which is the source of the runaway and pre-acceleration pathologies. The covariant generalisation is the Lorentz–Abraham–Dirac (LAD) equation; its perturbative reduction to a second-order equation is the Landau–Lifshitz equation, the form actually used in accelerator and plasma codes.

Key derivation Intermediate+

Theorem (Thomson cross-section from Larmor). A free charge $q$ of mass $m$ illuminated by a linearly polarised plane wave of peak field $E_0$ reradiates time-averaged power

$$⟨P_{\mathrm{scat}}⟩=\frac{q^4{E}_0^2}{12\pi {\epsilon }_0{m}^2{c}^3},$$

so that, dividing by the incident intensity $⟨I⟩=c{\epsilon }_0{E}_0^2/2$, the total cross-section is

$${\sigma }_T=\frac{8\pi }{3}(\frac{q^2}{4\pi {\epsilon }_{0}m{c}^2}{)}^2=\frac{8\pi }{3}{r}_e^2.$$

Proof. Take the incident wave to drive the charge along $\hat{\mathbf{x}}$ with $E_x(t)=E_0\cos \omega t$. Below any resonance the magnetic force is negligible and the equation of motion reduces to $m{\dot{v}}_x=q{E}_0\cos \omega t$, giving acceleration $a(t)=(q{E}_0/m)\cos \omega t$. The instantaneous Larmor power 10.07.01 is

$$P(t)=\frac{q^2{a}^2}{6\pi {\epsilon }_0{c}^3}=\frac{q^4{E}_0^2}{6\pi {\epsilon }_0{m}^2{c}^3}{\cos }^2\omega t.$$

Time-averaging $⟨{\cos }^2\omega t⟩=1/2$ yields the stated $⟨P_{\mathrm{scat}}⟩$. The reradiation is exactly the dipole pattern of 10.07.02 with $p_0=q{E}_0/m{\omega }^2$: the driven charge is an oscillating dipole, and the angular distribution follows from the dipole ${\sin }^2\Theta$ law in the polarisation frame.

The incident intensity is the time-averaged Poynting flux,

$$⟨I⟩=\frac{1}{2}c{\epsilon }_0{E}_0^2.$$

The cross-section is the ratio of scattered power to incident flux:

$${\sigma }_T=\frac{⟨P_{\mathrm{scat}}⟩}{⟨I⟩}=\frac{q^4{E}_0^2}{12\pi {\epsilon }_0{m}^2{c}^3}\cdot \frac{2}{c{\epsilon }_0{E}_0^2}=\frac{q^4}{6\pi {\epsilon }_0^2{m}^2{c}^4}.$$

Collecting $q^4/(6\pi {\epsilon }_0^2{m}^2{c}^4)$ as $(8\pi /3)[q^2/(4\pi {\epsilon }_{0}m{c}^2){]}^2=(8\pi /3)r_e^2$ gives the stated result. For an electron the numerical value is ${\sigma }_T\approx 6.65\times 10^{-29}$ m${}^2$, independent of frequency precisely because the $E_0^2$ and $\omega$ dependencies cancel between the driven acceleration and the Larmor power. $\square$

Bridge. This derivation builds toward the antenna and scattering machinery developed below, where the same retarded-potential integral appears again in the half-wave dipole and in the array factor. The foundational reason a thin wire radiates at all is that the accelerating conduction electrons along it produce the $1/r$ fields of Larmor; this is exactly the dipole radiation of 10.07.02 specialised to a driven, spatially extended current. Putting these together, the retarded-potential picture generalises the single-charge Larmor formula 10.07.01 to extended sources, and the bridge is the radiation resistance $R_{\mathrm{rad}}=73\Omega$ that converts a feed current into watts of outgoing wave — the antenna is engineering, but the physics is Larmor all the way down.

Exercises Intermediate+

Lean formalization Intermediate+

lean_status: none. Mathlib has the wave equation and a developing theory of Maxwell's equations, but it does not yet formalise the retarded radiation integrals over extended current distributions needed for antenna theory, the asymptotic far-zone energy flux that defines a scattering cross-section, the Thomson or Rayleigh formulae, or the Abraham–Lorentz–Dirac self-force and its Landau–Lifshitz reduction. A faithful formalisation would also require a measure-theoretic account of radiated power at null infinity, which is not currently in Mathlib.

Advanced results Master

The half-wave dipole in detail

For a centre-fed dipole of length $\ell =\lambda /2$ the current distribution that satisfies the boundary condition $I(\pm \lambda /4)=0$ is $\tilde{I}(z)=I_0\cos (kz)$. The far-field radiation integral evaluates to

$${\mathbf{E}}_{\mathrm{rad}}=\frac{i{\eta }_0{I}_0{e}^{i(kr-\omega t)}}{2\pi r}\frac{\cos [(\pi /2)\cos \theta ]}{\sin \theta }\hat{\mathbf{\theta }},$$

giving the angular pattern $|\cos [(\pi /2)\cos \theta ]{|}^2/{\sin }^2\theta$. This is the half-wave analogue of the short-dipole ${\sin }^2\theta$ pattern of 10.07.02: both vanish on axis and peak broadside, but the half-wave version is slightly narrower because the contributions from different points along the wire add with phase that reinforces the broadside direction. Integrating the Poynting vector over the sphere and dividing by $\frac{1}{2}{I}_0^2$ yields the radiation resistance

$$R_{\mathrm{rad}}=\frac{{\eta }_0}{2\pi }{\int }_0^{\pi }\frac{{\cos }^2[(\pi /2)\cos \theta ]}{\sin \theta }d\theta \approx 73.1\Omega ,$$

a result first obtained by Carter in 1932 and now tabulated in every antenna handbook. The proximity of $73\Omega$ to the $75\Omega$ standard for coaxial cable is a happy accident of nature, not a convention: it is why centre-fed dipoles are the natural match to standard transmission lines.

Multipole scattering and the optical theorem

A compact scatterer of size $a\ll \lambda$ has its scattering amplitude decomposed into the same electric-dipole, magnetic-dipole, and electric-quadrupole channels developed in 10.07.03. Each multipole $E1$, $M1$, $E2$ contributes a partial cross-section suppressed by $(ka{)}^2$ relative to its predecessor, so the leading channel dominates unless forbidden by symmetry — the selection-rule logic of 10.07.03 applied to absorption and scattering. The forward scattering amplitude is tied to the total cross-section by the optical theorem,

$${\sigma }_{\mathrm{tot}}=\frac{4\pi }{k}\mathrm{Im}f(\theta =0),$$

which expresses the physically deep fact that the wave removed from the incident beam (extinction) interferes destructively with the forward-going scattered wave. Rayleigh's ${\omega }^4$ law is the leading $E1$ term for a dielectric sphere, and the anomalous scattering resonances that appear near molecular absorption lines are the classical shadows of quantum transitions.

Klein–Nishina and the quantum boundary

The Thomson formula holds for $\hslash \omega \ll m_e{c}^2$. When the photon energy becomes comparable to the electron rest energy, the scattering becomes inelastic (Compton scattering), the scattered photon is red-shifted by the recoil, and the total cross-section is given by the Klein–Nishina formula:

$${\sigma }_{\mathrm{KN}}=\frac{3{\sigma }_T}{4}\left[\frac{1+x}{x^3}\right(\frac{2x(1+x)}{1+2x}-\ln (1+2x))+\frac{\ln (1+2x)}{2x}-\frac{1+3x}{(1+2x{)}^2}],$$

with $x=\hslash \omega /(m_e{c}^2)$. In the low-energy limit this reduces to ${\sigma }_T$ term by term. At high energy ${\sigma }_{\mathrm{KN}}\sim (3{\sigma }_T/8x)(\ln 2x+\frac{1}{2})$, falling roughly as $1/\omega$: gamma rays above a few MeV interact only weakly with electrons, which is why high-energy gamma astronomy can see photons that have crossed most of the universe. In astrophysical sources such as blazars, relativistic electrons up-scatter low-energy synchrotron photons to gamma energies by inverse Compton scattering 10.07.05, and the same Klein–Nishina decline ultimately caps the gamma-ray spectrum.

The Lorentz–Abraham–Dirac equation and its pathologies

The covariant form of the radiation-reaction equation, derived by Dirac in 1938 by careful regularisation of the self-field, is

$$m\frac{d{U}^{\mu }}{d\tau }=F_{\mathrm{ext}}^{\mu }+\frac{q^2}{6\pi {\epsilon }_0{c}^3}(\frac{d^2{U}^{\mu }}{d{\tau }^2}+\frac{U^{\mu }}{c^2}\frac{d{U}_{\nu }}{d\tau }\frac{d{U}^{\nu }}{d\tau }).$$

The second term inside the bracket keeps $U_{\mu }{U}^{\mu }=c^2$ conserved under the recoil. The equation is third-order in proper time and inherits the runaway solutions of the non-relativistic Abraham–Lorentz form. Eliminating the runaway by imposing the future boundary condition $d{U}^{\mu }/d\tau \to 0$ as $\tau \to +\infty$ shifts the pathology to pre-acceleration: the charge begins to move a time $\sim {\tau }_0$ before the external force arrives. For the electron ${\tau }_0\approx 6.3\times 10^{-24}$ s, far below any directly resolvable timescale, so the pre-acceleration is operationally harmless — but it signals that classical electrodynamics with point charges is not a closed theory.

The Landau–Lifshitz equation

The standard resolution treats the self-force perturbatively. Substituting the lowest-order equation $md{U}^{\mu }/d\tau =F_{\mathrm{ext}}^{\mu }$ back into the LAD reaction term reduces the dynamics to a regular second-order equation,

$$m\frac{d{U}^{\mu }}{d\tau }=F_{\mathrm{ext}}^{\mu }+\frac{q^2}{6\pi {\epsilon }_0{c}^3}[\frac{d{F}_{\mathrm{ext}}^{\mu }}{d\tau }+\frac{U^{\mu }{F}_{\mathrm{ext}}^{\nu }{F}_{\mathrm{ext},\nu }}{m{c}^2}],$$

with no runaways and no pre-acceleration. This is the Landau–Lifshitz equation, used universally in accelerator tracking codes and in strong-field plasma simulations. Its regime of validity is ${\tau }_0\omega \ll 1$ and $\chi =\gamma |\mathbf{E}|/E_{\mathrm{crit}}\ll 1$, where $E_{\mathrm{crit}}=m^2{c}^3/(q\hslash )\approx 1.3\times 10^{18}$ V/m is the Schwinger field. Only in the magnetospheres of pulsars ($B\sim 10^8$ T) or in the focus of petawatt lasers does $\chi$ approach unity and the classical reaction give way to stochastic quantum radiation reaction, an active frontier of ultra-intense laser physics.

Synthesis. The three subjects of this unit are unified by the exchange of energy and momentum between a charge distribution and the electromagnetic field it radiates or absorbs. The foundational reason antennas, scattering, and radiation reaction belong in one chapter is that each is a different facet of the Larmor mechanism 10.07.01: an antenna is a designed current distribution radiating coherently, scattering is the radiation re-emitted by a charge driven by an incoming wave, and radiation reaction is the self-force that very mechanism exerts on its own source. This is exactly the content of energy-momentum conservation for the coupled particle-field system. The central insight is that the same transverse $1/r$ dipole field of 10.07.02 appears in all three, and only the prescription of the source changes. Putting these together, the radiation resistance generalises Ohm's law to free space, the Thomson cross-section generalises the dipole formula to a driven oscillator, and the Landau–Lifshitz force generalises Newton's second law to include the recoil of emission. The bridge is the energy balance $⟨I⟩{\sigma }_T=⟨P_{\mathrm{scat}}⟩$, which ties scattering to radiation reaction through the very Larmor power from which this whole chapter began.

Full proof set Master

Proposition (Thomson differential cross-section). The differential cross-section for Thomson scattering of an unpolarised plane wave by a free charge is

$$\frac{d{\sigma }_T}{d\Omega }=\frac{r_e^2}{2}(1+{\cos }^2\theta ),$$

and its integral over solid angle is ${\sigma }_T=(8\pi /3)r_e^2$.

Proof. Take the incident wave to propagate along $\hat{\mathbf{z}}$ and let its electric field point along $\hat{\mathbf{x}}$ with peak amplitude $E_0$. The charge accelerates along $\hat{\mathbf{x}}$ with $a(t)=(q{E}_0/m)\cos \omega t$, and reradiates with the dipole pattern of 10.07.02. For an observation direction $\hat{\mathbf{n}}$ making polar angle ${\Theta }_x$ with the acceleration axis $\hat{\mathbf{x}}$, the time-averaged radiated power per unit solid angle is

$$⟨\frac{dP}{d\Omega }⟩=\frac{q^2{a}_0^2}{32{\pi }^2{\epsilon }_0{c}^3}{\sin }^2{\Theta }_x=\frac{q^4{E}_0^2}{32{\pi }^2{\epsilon }_0{m}^2{c}^3}{\sin }^2{\Theta }_x,$$

with $a_0=q{E}_0/m$. Dividing by the incident intensity $⟨I⟩=c{\epsilon }_0{E}_0^2/2$ gives the polarised differential cross-section

$$(\frac{d\sigma }{d\Omega }{)}_{\hat{\mathbf{x}}}=\frac{q^4}{16{\pi }^2{\epsilon }_0^2{m}^2{c}^4}{\sin }^2{\Theta }_x=r_e^2{\sin }^2{\Theta }_x.$$

For an unpolarised beam we average over two orthogonal polarisations, $\hat{\mathbf{x}}$ and $\hat{\mathbf{y}}$. If the scattering direction $\hat{\mathbf{n}}$ makes angle $\theta$ with the incident propagation axis $\hat{\mathbf{z}}$, then $\cos {\Theta }_x=\hat{\mathbf{n}}\cdot \hat{\mathbf{x}}=n_x$ and $\cos {\Theta }_y=n_y$, with $n_x^2+n_y^2+n_z^2=1$ and $n_z=\cos \theta$. Averaging,

$$⟨{\sin }^2{\Theta }_x+{\sin }^2{\Theta }_y⟩=2-(n_x^2+n_y^2)=2-(1-n_z^2)=1+{\cos }^2\theta .$$

So

$$\frac{d{\sigma }_T}{d\Omega }=\frac{r_e^2}{2}(1+{\cos }^2\theta ).$$

Integrating,

$${\sigma }_T=\frac{r_e^2}{2}{\int }_0^{2\pi }d\varphi {\int }_0^{\pi }(1+{\cos }^2\theta )\sin \theta d\theta =\frac{r_e^2}{2}\cdot 2\pi \cdot \left[\right(-\cos \theta )-\frac{{\cos }^3\theta }{3}{]}_0^{\pi }.$$

The bracket evaluates to $[1+\frac{1}{3}]-[-1-\frac{1}{3}]=8/3$, giving

$${\sigma }_T=\frac{r_e^2}{2}\cdot 2\pi \cdot \frac{8}{3}=\frac{8\pi }{3}{r}_e^2,$$

completing the proof. $\square$

Corollary (forward–backward symmetry). Because $d{\sigma }_T/d\Omega$ depends on $\theta$ only through ${\cos }^2\theta$, Thomson scattering is symmetric between $\theta$ and $\pi -\theta$: a free electron scatters as much light backwards as forwards. This symmetry is broken in the quantum (Klein–Nishina) regime, where recoil preferentially suppresses backward scattering at high photon energy.

Connections Master

• Larmor formula 10.07.01 is the single result from which all three threads of this unit descend. The antenna radiation integral is the Larmor field summed coherently over an extended current; the Thomson cross-section is Larmor power divided by incident intensity; the Abraham–Lorentz force is the Larmor energy loss written as a recoil on the source. Whenever this unit computes a radiated or scattered power, the Larmor prefactor $q^2{a}^2/(6\pi {\epsilon }_0{c}^3)$ is doing the work.

• Electric dipole radiation 10.07.02 supplies the angular pattern and the retarded-potential machinery. The half-wave dipole is an extended oscillating dipole whose far field is the integral of the elementary dipole field of 10.07.02 over the sinusoidal current; Rayleigh scattering is dipole radiation from the polarisability-induced dipole $p={\alpha }_{0}E$; and the ${\sin }^2\Theta$ law is what fixes the $(1+{\cos }^2\theta )$ shape of the Thomson differential cross-section.

• Multipole radiation 10.07.03 generalises dipole radiation to higher channels and governs the corrections to both antenna patterns and scattering amplitudes. The $E2$ and $M1$ terms suppressed by $(ka{)}^2$ describe the deviations of a finite antenna from the ideal dipole, and the multipole decomposition of a compact scatterer reduces Rayleigh and anomalous scattering to the same selection rules that control forbidden spectral lines.

• Synchrotron radiation 10.07.05 is the relativistic, circular-motion limit of the Larmor mechanism, and its complement is inverse Compton scattering, in which the relativistic electrons of a synchrotron source up-scatter ambient low-energy photons to gamma energies via the Thomson/Klein–Nishina cross-sections of this unit. The same ${\gamma }^4$ beaming that makes synchrotron sources bright also makes inverse Compton the dominant gamma-ray channel in blazars and pulsar wind nebulae.

Historical & philosophical context Master

The three threads of this unit were woven between 1897 and 1962, and each carries a cautionary tale about how far a classical theory can be pushed.

Joseph Larmor published the power formula $P=2q^2{a}^2/(3c^3)$ (Gaussian units) in 1897 ^{[Larmor 1897]}, derived from the Hertzian dipole. Within a decade it was clear that the same result described both emission (antennas) and re-emission (scattering). Guglielmo Marconi's transatlantic transmission of 1901 used a crude antenna whose radiation resistance was a few ohms at most; the modern theory of the half-wave dipole and its $73\Omega$ radiation resistance was worked out by Carter (1932) once the sinusoidal current distribution was properly imposed, turning antenna design from empiricism into calculation.

Joseph John Thomson measured the scattering of X-rays by electrons in 1906 ^{[Thomson 1906]} and showed that the cross-section, of order $10^{-29}$ m${}^2$, implied a particle of radius $\sim 10^{-15}$ m — the classical electron radius $r_e$. The Thomson cross-section was the first quantitative bridge between electrodynamics and the newly discovered electron, and its independence of frequency was a clean classical prediction that quantum mechanics would later refine via the Klein–Nishina formula (1929) once Compton's 1923 measurement of the inelastic photon recoil showed the classical picture failing above $m_e{c}^2$. Lord Rayleigh's earlier ${\omega }^4$ scattering law (1871) ^{[Rayleigh 1871]} fell into place as the dipole limit of the same theory, explaining the blue sky two decades before the electron was discovered.

Radiation reaction has always been the uncomfortable part of the story. Max Abraham (1904–1905) ^{[Abraham 1905]} and Hendrik Lorentz independently derived the self-force ${\mathbf{F}}_{\mathrm{rad}}\propto \ddot{\mathbf{v}}$ by computing the field of an extended charge and shrinking it to a point, exposing immediately the third-order time derivative and its runaway solutions. Paul Dirac's 1938 covariant reformulation ^{[Dirac 1938]} made the equation Lorentz-invariant but did not remove the runaway; the future boundary condition that does remove it merely relocates the trouble to pre-acceleration. The pragmatic resolution came from Lev Landau and Evgeny Lifshitz in their 1941 Classical Theory of Fields ^{[Landau Lifshitz 1941]}, where the radiation reaction is treated as a perturbation and reduced to a well-posed second-order equation. Fritz Rohrlich's Classical Charged Particles (1965, 3rd ed. 2007) ^{[Rohrlich 2007]} remains the canonical defence of the consistency of classical point-charge electrodynamics, while the modern frontier — stochastic quantum radiation reaction at $\chi \sim 1$ — is now probed directly in petawatt-laser experiments.

The philosophical lesson is that a theory can be extraordinarily accurate and still be open at the centre: the Larmor mechanism predicts the radiation perfectly but cannot, within classical point-particle electrodynamics, consistently describe the recoil that this same radiation produces. The antenna, the scatterer, and the self-force together mark the boundary where classical electrodynamics hands the baton to quantum field theory.

Bibliography Master

@article{larmor1897,
  author  = {Larmor, Joseph},
  title   = {On the theory of the magnetic influence on light; and on the rotation of the plane of polarisation of a beam of light by a magnetic field},
  journal = {Philosophical Magazine},
  series  = {5},
  volume  = {44},
  pages   = {503--512},
  year    = {1897}
}

@article{rayleigh1871,
  author  = {Strutt, John William (Lord Rayleigh)},
  title   = {On the light from the sky, its polarisation and colour},
  journal = {Philosophical Magazine},
  series  = {4},
  volume  = {41},
  pages   = {107--120 and 274--279},
  year    = {1871}
}

@article{abraham1905,
  author  = {Abraham, Max},
  title   = {Theorie der Elektrizit\"at, vol.~II: Elektromagnetische Theorie der Strahlung},
  publisher = {Teubner (Leipzig)},
  year    = {1905}
}

@article{thomson1906,
  author  = {Thomson, Joseph John},
  title   = {Conduction of Electricity through Gases},
  edition = {2nd},
  publisher = {Cambridge University Press},
  year    = {1906},
  note    = {Chapter on scattering of {R\"ontgen} rays; introduces the classical electron radius}
}

@article{dirac1938,
  author  = {Dirac, Paul A. M.},
  title   = {Classical theory of radiating electrons},
  journal = {Proceedings of the Royal Society of London A},
  volume  = {167},
  pages   = {148--169},
  year    = {1938}
}

@book{landauLifshitz1941,
  author    = {Landau, Lev D. and Lifshitz, Evgeny M.},
  title     = {The Classical Theory of Fields},
  edition   = {4th},
  publisher = {Butterworth-Heinemann},
  year      = {1975},
  note      = {Original Russian edition 1941; \S\S76--78 on radiation damping}
}

@book{jackson1999,
  author    = {Jackson, John David},
  title     = {Classical Electrodynamics},
  edition   = {3rd},
  publisher = {Wiley},
  year      = {1999}
}

@book{griffiths2017,
  author    = {Griffiths, David J.},
  title     = {Introduction to Electrodynamics},
  edition   = {4th},
  publisher = {Cambridge University Press},
  year      = {2017}
}

@book{rohrlich2007,
  author    = {Rohrlich, Fritz},
  title     = {Classical Charged Particles},
  edition   = {3rd},
  publisher = {World Scientific},
  year      = {2007}
}