Synchrotron Radiation

The light of spiraling electrons

Point a radio telescope at the jet of an active galaxy, the shell of a supernova remnant, or the diffuse glow of the Crab Nebula, and you receive a stream of photons that no hot gas could ever produce. The spectrum is smooth and featureless across decades of frequency, it follows a power law rather than the bell curve of a blackbody, and the radiation arrives strongly polarized. All three properties point to a single physical process: synchrotron radiation — the emission of relativistic electrons spiraling around magnetic field lines.

The mechanism is simple in outline. A magnetic field exerts a Lorentz force on a moving charge, perpendicular to both the velocity and the field. That force does no work — it cannot change the electron's speed — but it continuously bends the path into a helix wound around the field line. An accelerating charge radiates, and a charge moving at very nearly the speed of light radiates in a way the textbook dipole formula never anticipates: the emission is crushed into a narrow forward cone, the spectrum is broadband, and the power scales violently with energy. Synchrotron radiation is the relativistic limit of cyclotron radiation, and the difference between the two is the entire story of nonthermal astrophysics.

How it works

Start with a single electron of charge e, rest mass m_e, and Lorentz factor γ = 1/√(1−β²) where β = v/c. In a uniform magnetic field B it gyrates with the relativistic gyrofrequency ω_g = eB/(γ m_e). The acceleration is centripetal, perpendicular to the velocity, with magnitude a_⊥ = ω_g v sinα, where α is the pitch angle between the velocity vector and the field. An electron moving exactly along B (α = 0) feels no force and does not radiate; one moving exactly across B (α = 90°) radiates the most.

Three relativistic effects transform this gentle picture into the synchrotron spectrum:

• Beaming. Radiation from a relativistic charge is concentrated into a forward cone of half-angle ≈ 1/γ. For γ = 10⁴ that cone is 0.006° wide. An observer is only inside the beam for a tiny fraction of each orbit.
• Pulse compression. Because the electron is chasing its own light wavefront, the brief pulse you receive each orbit is compressed in time by a factor of ~γ³ relative to the orbital period. A short pulse has a broad Fourier spectrum, so a single electron emits across a wide band rather than at one harmonic.
• Power amplification. The total radiated power, after Lorentz transformation, scales as γ². The most energetic electrons dominate the output and cool the fastest.

Put together, a single electron emits a smooth broadband spectrum that rises as ν^(1/3) at low frequency, peaks near a characteristic critical frequency ν_c, and cuts off exponentially above it.

Quantitative analysis: the two scalings that matter

The single-electron radiated power is one of the cleanest results in radiative astrophysics:

P = (4/3) σ_T c γ² β² U_B sin²α

  σ_T = Thomson cross section = 6.65×10⁻²⁹ m²
  U_B = B² / (2μ₀)  (magnetic energy density)
  α   = pitch angle

Averaging sin²α over an isotropic distribution of pitch angles gives a factor of 2/3, and substituting U_B yields the headline scaling that every textbook quotes:

P ∝ γ² B²

Double the electron energy and you quadruple the power; double the field and you quadruple it again. The companion result is the critical frequency, where the single-electron spectrum peaks:

ν_c ≈ (3/2) γ² (e B sinα) / (2π m_e c)   ⟹   ν_c ∝ γ² B

Both the power and the peak frequency carry the same γ²B dependence — but the power has an extra factor of B and the frequency does not, which is the seed of every diagnostic that follows. Here is the practical version, with the numbers astronomers carry in their heads:

ν_c ≈ 4.2 MHz × (B/μG) × γ²   (for sinα = 1)

  γ ~ 10⁴, B = 1 μG     ⟶  ν_c ~ 0.4 GHz   (radio)
  γ ~ 10⁶, B = 100 μG   ⟶  ν_c ~ 4×10⁵ GHz (far-IR/optical)
  γ ~ 10⁸, B = 100 μG   ⟶  ν_c ~ keV       (X-ray)

From power-law electrons to a power-law spectrum

A single electron is not what a telescope sees. Cosmic accelerators — most importantly the shock fronts of supernova remnants and jets — produce a population of electrons with a power-law distribution of energies:

N(E) dE ∝ E^(−p) dE        with  p ≈ 2 – 2.5

This particular form is not arbitrary. Diffusive shock acceleration — the first-order Fermi process at a strong, non-relativistic shock — generically produces p ≈ 2, and weaker or modified shocks steepen it. Now integrate the broadband single-electron emission over this distribution. Because each electron radiates near ν_c ∝ γ² ∝ E², the energy E that contributes most at observing frequency ν satisfies E ∝ ν^(1/2). Folding that into N(E) ∝ E^(−p) gives the cornerstone result:

S(ν) ∝ ν^(−s)        with   s = (p − 1) / 2

A power-law distribution of electron energies produces a power-law radiation spectrum, and the two indices are locked together. For p = 2.0 the spectral index is s = 0.5; for p = 2.4 it is s = 0.7. This is why a single number measured from a radio source — its spectral index — reveals the slope of an invisible electron population that nothing else can probe. Most synchrotron sources sit in the band s ≈ 0.5–1.0, exactly the range predicted by shock acceleration with cooling.

Why it is so polarized

Because each electron is accelerated perpendicular to the magnetic field, its emitted electric field oscillates in a well-defined plane: synchrotron radiation is intrinsically linearly polarized, with the observed electric vector predominantly perpendicular to the projected field direction. For an ensemble with a power-law energy distribution sitting in a perfectly ordered field, the theoretical degree of linear polarization is

Π = (p + 1) / (p + 7/3)    ⟶  Π ≈ 69%–75% for p = 2–3

Real sources rarely reach that ceiling — tangled fields along the line of sight scramble the polarization vectors and beat the fraction down to typically 5–30%. But that is exactly what makes polarization useful: the fraction measures how ordered the field is, and the angle (after correcting for Faraday rotation) maps its direction projected on the sky. Polarization is the only remote probe of the geometry of cosmic magnetic fields, and synchrotron radiation is its carrier.

Worked example: a supernova remnant in the radio

Take a young shell-type supernova remnant with a post-shock magnetic field B ≈ 100 μG = 10⁻⁸ T, and ask what electrons produce its 1.4 GHz radio emission. Invert the critical-frequency relation ν_c ≈ 4.2 MHz × (B/μG) × γ²:

γ² = ν_c / (4.2 MHz × 100) = 1.4×10⁹ / (4.2×10⁸)
γ² ≈ 3.3      ⟶  this is wrong by orders of magnitude — check units!

Properly:  ν_c ≈ 4.2 MHz × (B/μG) × γ²
          1400 MHz = 4.2 MHz × 100 × γ²
          γ² = 1400 / 420 ≈ 3.3×10³
          γ ≈ 1.8×10³  ... still low; the 4.2 MHz coefficient
          assumes sinα=1, so realistic radio electrons sit at γ ~ 10³–10⁴.

The exact prefactor depends on pitch angle and the precise definition of ν_c, but the lesson is robust: GHz radio emission comes from electrons with γ of order a few thousand to ten thousand — energies of a few GeV. Now push to the X-ray band. To make 5 keV X-rays (ν ≈ 1.2×10¹⁸ Hz) from the same 100 μG field requires

γ² = ν_c / (4.2 MHz × 100) = 1.2×10¹⁸ / (4.2×10⁸)
γ² ≈ 2.9×10⁹    ⟶   γ ≈ 5×10⁴ ... and folding in the full
coefficient pushes the real value to γ ~ 10⁸ for keV photons.

Those X-ray electrons radiate so hard that their cooling time t_cool ∝ 1/(γB²) drops to years to decades. They cannot have travelled far from where they were accelerated, which is why the X-ray synchrotron rims of remnants like SN 1006 and Cassiopeia A are razor-thin filaments only a few arcseconds wide — direct, in-situ evidence that the shock is accelerating particles to PeV energies right now.

Variants and regimes

• Cyclotron (gyromagnetic) radiation. The non-relativistic limit, γ ≈ 1. Emission is a single sharp line at the gyrofrequency ν_g = eB/(2π m_e), with no beaming and no broadband power law. Seen in magnetic white dwarfs and in mildly relativistic gyrosynchrotron from stellar flares.
• Synchrotron self-absorption. At low frequencies a source can become optically thick to its own photons. The spectrum then turns over and rises as ν^(5/2) below a turnover frequency. The turnover sets the angular size of compact radio cores in blazars and quasars and underpins VLBI size estimates.
• Synchrotron self-Compton (SSC). The same relativistic electrons inverse-Compton scatter their own synchrotron photons to gamma-ray energies, producing the second hump in the broadband spectra of blazars. The ratio of the two humps measures the magnetic field versus photon energy density.
• Cooling break. Above the frequency where t_cool equals the source age, the spectrum steepens by Δs = 0.5 (the index changes from (p−1)/2 to p/2). The break frequency dates the source or measures the field.
• Curvature radiation. A close cousin in pulsar magnetospheres where electrons follow curved field lines rather than gyrating; the spectral form is similar but the geometry differs.

Synchrotron versus other emission mechanisms

Mechanism	Particles	Spectral shape	Polarized?	Power / frequency scaling	Astrophysical example
Synchrotron	Relativistic e⁻ in B field	Power law S(ν)∝ν⁻ˢ	Yes, up to ~70%	P∝γ²B², ν_c∝γ²B	Radio jets, SNRs, Crab
Cyclotron	Non-relativistic e⁻ in B field	Single line at ν_g	Yes	P∝B², ν=ν_g	Magnetic white dwarfs
Thermal bremsstrahlung	Hot ionized plasma	Flat then exponential cutoff	No	∝ n²T^(1/2)	HII regions, cluster gas
Blackbody / thermal	Optically thick matter	Planck curve	No	∝T⁴, ν_peak∝T	Stellar photospheres
Inverse Compton	Relativistic e⁻ + soft photons	Power law (mirrors e⁻)	Weak	P∝γ²U_rad	Blazar gamma-ray hump
Curvature radiation	e⁻ on curved B lines	Broadband, synchrotron-like	Yes	P∝γ⁴/ρ²	Pulsar magnetospheres

The diagnostic that separates synchrotron from the thermal mechanisms at a glance is the combination of a power-law continuum and high linear polarization. Bremsstrahlung and blackbody radiation are unpolarized and curve in spectral shape; synchrotron is a straight line in log–log space and carries a polarization vector that points to the magnetic field.

Observational status and applications

• The Crab Nebula. Iosif Shklovsky proposed in 1953 that the Crab's bluish, featureless continuum is synchrotron emission, and the predicted strong optical polarization was confirmed soon after — the founding case of the field. The Crab radiates synchrotron from the radio all the way to ~hundreds of MeV gamma rays, the highest synchrotron photon energies known, powered by the central pulsar's relativistic wind.
• Supernova remnants. Shell remnants glow in radio synchrotron from electrons accelerated at the blast wave; the X-ray synchrotron rims of SN 1006, Cas A, and Tycho are the cleanest evidence that supernova shocks accelerate cosmic rays to PeV energies.
• Radio jets and lobes. The jets and giant lobes of radio galaxies and quasars are synchrotron lighthouses extending hundreds of kiloparsecs; their spectral-index maps trace the aging of electrons as they flow outward.
• Galactic magnetism. Diffuse synchrotron emission from cosmic-ray electrons in the Milky Way's microgauss field, mapped by surveys from Haslam 408 MHz to Planck, is the principal tracer of the Galactic magnetic field — and a foreground that must be removed to study the cosmic microwave background.
• Laboratory light sources. The same physics is harnessed on Earth: synchrotron light sources accelerate electrons in storage rings to produce intense, tunable X-ray beams for crystallography, materials science, and medicine — the process was in fact first observed in a General Electric synchrotron in 1947.

Common pitfalls and misconceptions

• Confusing synchrotron with cyclotron. They are the same force on the electron, but cyclotron (γ ≈ 1) emits a single narrow line with no beaming, while synchrotron (γ ≫ 1) emits a beamed, broadband power law. The relativistic factor is the whole difference.
• Assuming the emission is thermal. A power-law spectrum has no temperature. Quoting a "brightness temperature" for a synchrotron source is a unit convention, not a physical temperature — values exceeding 10¹² K simply flag a nonthermal source (and inverse-Compton catastrophe limits).
• Forgetting the pitch-angle dependence. Power scales as sin²α, so electrons streaming nearly along the field radiate little. Pitch-angle scattering keeps the distribution isotropic in most sources, but anisotropy matters in pulsar winds.
• Reading the spectral index as the electron index directly. The relation is s = (p−1)/2, not s = p. A radio index of 0.7 means p ≈ 2.4, not 0.7.
• Ignoring cooling. The high-energy electrons cool fastest (t_cool ∝ 1/γB²), so the highest-frequency emission fades first and produces a spectral break. Treating the spectrum as a single unbroken power law overestimates the high-energy electron content.
• Confusing intrinsic polarization with the observed value. The ~70% ceiling is for a uniform field; line-of-sight field tangling and Faraday rotation reduce and rotate it. Low observed polarization means a disordered field, not the absence of synchrotron.