Plasma Physics
Magnetic Mirror: How a Field Pinch Reflects Charged Particles
A proton spiraling into a region where the magnetic field triples in strength will slow its forward motion to zero and get flung back the way it came — without ever touching a wall, without losing a joule of energy. This is the magnetic mirror effect: a place where converging field lines act like a reflecting surface for charged particles. It is not a material mirror but a geometric one, built entirely out of the shape of the field.
The magnetic mirror is the reflection of a gyrating charged particle by a region of increasing magnetic field strength. It works because a particle's magnetic moment μ = mv⊥²/2B is very nearly conserved as it drifts along field lines. As B rises toward a "pinch," the perpendicular energy μB must rise too, and since total kinetic energy is fixed, the parallel energy drains away until the particle stops and reverses. Two such pinches facing each other form a magnetic bottle that traps plasma — the principle behind Earth's Van Allen belts and an entire family of fusion machines.
- TypeAdiabatic reflection of charged particles
- RegimeMagnetized collisionless plasma, slowly varying B
- Key equationμ = mv⊥²/2B = const (adiabatic invariant)
- Loss conesin²θ_lc = B_min/B_max = 1/R_m
- First analyzedFermi (1949 cosmic rays); mirror machines 1950s
- Observed inVan Allen belts, aurorae, mirror & tandem-mirror fusion devices
Interactive visualization
Press play, or step through manually. The visualization is yours to drive — try it before reading on.
Watch the 60-second explainer
A condensed visual walkthrough — narrated, captioned, under a minute.
The Physical Setup: A Bottle Made of Field Lines
Picture two coaxial coils carrying current in the same direction, separated by a gap. Midway between them the magnetic field is weakest; near each coil the field lines squeeze together and B is strongest. This bulging, barrel-shaped field is a magnetic bottle, and each high-field end is a magnetic mirror.
A charged particle in a magnetic field does not travel in a straight line — the Lorentz force F = qv × B makes it spiral, executing tight circular gyration around a field line while streaming along it. So its velocity splits into two parts:
- v⊥ — the perpendicular (gyration) component, setting the radius of the helix;
- v∥ — the parallel component, carrying it along the axis toward a mirror.
The pitch angle θ = arctan(v⊥/v∥) measures how steeply the helix climbs. A particle moving nearly along the axis (small θ) barely feels the pinch and escapes; one gyrating fast across the field (large θ) is turned around. The mirror does not confine everything — and that leak is the loss cone.
The Mechanism: The Magnetic Moment as an Adiabatic Invariant
The reflection follows from one deep fact: the orbital magnetic moment
μ = m·v⊥² / (2B)
is an adiabatic invariant — nearly constant so long as B changes little over one gyro-orbit and one gyro-period (the slowly-varying regime). Here m is mass, v⊥ perpendicular speed, B field strength.
Now conserve energy. The magnetic force does no work, so total kinetic energy E = ½m·v∥² + ½m·v⊥² is fixed, and the perpendicular part equals μB. Substitute:
½m·v∥² = E − μB
As the particle moves into stronger field, B rises, μB rises, and v∥² must fall. When B reaches B_turn = E/μ, we get v∥ = 0: the particle halts along the axis, all its energy now in gyration, and the field gradient (the −μ∇B force, the grad-B force) pushes it back. It has been reflected without any collision. The whole argument rests on μ-conservation; if the field varies too sharply, μ breaks and particles leak through.
Key Quantities: The Loss Cone and a Worked Example
Define the mirror ratio R_m = B_max/B_min. A particle reflects only if it turns around before reaching B_max. Combining μ- and energy-conservation between the midplane (B_min, pitch angle θ) and the throat gives the confinement condition:
sin²θ ≥ B_min/B_max = 1/R_m
Particles with pitch angle smaller than the loss-cone angle θ_lc = arcsin(1/√R_m) escape. Notice it depends only on the field ratio, not on particle mass, charge, or energy — a 10 keV electron and a 100 keV proton share the same loss cone.
Worked example. Take R_m = 4 (B goes from 1 T at center to 4 T at the throat). Then sin²θ_lc = 0.25, so θ_lc = 30°. Any particle whose velocity lies within 30° of the axis at the midplane is lost; the rest bounce. For an isotropic plasma the trapped fraction is √(1 − 1/R_m) = √0.75 ≈ 0.87 — 87% confined, 13% gone almost instantly. Pushing to R_m = 10 shrinks θ_lc to 18.4° and traps 95%.
How It's Observed, Measured, and Applied
The magnetic mirror is not a theorist's toy — it is directly measured in nature and engineered in the lab.
- Van Allen radiation belts. Earth's dipole field is a giant natural bottle: field lines are weak at the equator and pinch toward the magnetic poles (R_m in the thousands). Trapped protons and electrons bounce pole-to-pole in ~0.1–1 s while also drifting in longitude, forming the belts James Van Allen discovered with Explorer 1 in 1958.
- Aurorae. Particles scattered into the loss cone by waves precipitate into the upper atmosphere and light it up — the loss cone made visible.
- Fusion mirror machines. Devices from the 1950s onward — Q-machines, 2XIIB, and the giant MFTF-B and TMX at Livermore — used mirrors to hold fusion-grade plasma. The tandem mirror (Dimov, Fowler & Logan, 1976) adds high-field end plugs to electrostatically stopper the loss cone.
Confinement is quantified by measuring the pitch-angle distribution and watching for the characteristic empty wedge of velocity space — the loss cone — in particle detectors.
Comparison to Related Regimes and Cousins
The mirror effect sits inside a family of magnetic-confinement ideas, and it helps to see the boundaries.
- Mirror vs. tokamak. A mirror is an open (linear) system — field lines exit the ends, so there is always a loss cone. A tokamak is closed (toroidal); field lines never leave, eliminating end losses but at the cost of complex geometry. Ironically, tokamaks contain their own internal mirrors: the field is stronger on the inboard side, trapping "banana-orbit" particles.
- Mirror vs. Fermi acceleration. A moving mirror does work. Enrico Fermi (1949) showed that particles reflecting off converging moving magnetic clouds gain energy on each head-on bounce — Fermi acceleration, a leading mechanism for cosmic rays.
- Adiabatic vs. non-adiabatic. The whole picture holds only while μ is invariant. When the gyroradius approaches the field's scale length, or at a sharp field null, μ is violated and particles cross the loss cone freely — the basis of chaotic scattering in the magnetotail.
Significance, Open Questions, and Famous Cases
The magnetic mirror explains why space near Earth is dangerous (trapped MeV protons), why the aurorae glow, and why one whole branch of fusion research exists. Its beauty is economy: a single conserved quantity, μ, predicts reflection, the loss cone, and the trapped fraction all at once.
The unfinished business is the loss cone itself. Because it removes particles moving along the field, a mirror plasma has an anisotropic velocity distribution, which drives instabilities — the loss-cone and DCLC (drift-cyclotron loss-cone) modes — that scatter particles into the cone faster than pure collisions would. Taming these was the central struggle of 20th-century mirror fusion.
- Historic case. The U.S. mirror program peaked with MFTF-B (completed 1986, ~$372 M) — then cancelled the day after it was finished as funding shifted to tokamaks.
- Revival. Since ~2020, startups and labs (WHAM at Wisconsin, Novatron, Gauss Fusion concepts) have revisited high-field mirrors with modern high-temperature superconducting magnets, betting that R_m of tens with sheared rotation can finally quiet the loss cone.
| Mirror ratio R_m | Loss-cone angle θ_lc = arcsin(1/√R_m) | Trapped fraction √(1−1/R_m) | Example / regime |
|---|---|---|---|
| 2 | 45.0° | 0.71 | Modest lab coil pair |
| 4 | 30.0° | 0.87 | Typical mirror machine |
| 10 | 18.4° | 0.95 | High-field pinch |
| 100 | 5.7° | 0.995 | Strong tandem-mirror plug |
| ~1000+ | ≈1.8° | ≈0.9995 | Earth's field, equator→pole |
Frequently asked questions
Why does a charged particle bounce off a magnetic mirror if the field does no work?
The magnetic force never changes a particle's total kinetic energy, but it can redistribute it between parallel and perpendicular motion. Because the magnetic moment μ = mv⊥²/2B stays constant, entering stronger field B forces perpendicular energy μB to grow, so parallel energy must shrink. When parallel velocity hits zero the particle is turned around — reflected without any energy gain or loss.
What is the loss cone in a magnetic mirror?
The loss cone is the range of velocity directions, within a half-angle θ_lc = arcsin(1/√R_m) of the field axis, for which a particle's pitch angle is too small to reflect. Such particles reach the high-field throat before turning around and escape out the end. It appears as an empty cone-shaped void in the plasma's velocity distribution.
What is the mirror ratio and why does it matter?
The mirror ratio R_m = B_max/B_min is the ratio of the strongest field (at the throat) to the weakest (at the center). A larger R_m gives a narrower loss cone and traps a larger fraction √(1 − 1/R_m) of particles. It is the single most important design number for any mirror confinement device.
How do magnetic mirrors relate to the Van Allen radiation belts?
Earth's dipole field is naturally weak at the equator and pinches toward the poles, acting as a magnetic bottle with an enormous mirror ratio. Trapped protons and electrons bounce back and forth between mirror points near the north and south poles in a fraction of a second, forming the Van Allen belts discovered by Explorer 1 in 1958.
Does the loss cone depend on a particle's energy or charge?
No. The confinement condition sin²θ ≥ 1/R_m depends only on the pitch angle and the field ratio, not on mass, charge, or speed. A slow electron and a fast proton at the same pitch angle are either both trapped or both lost. This mass- and energy-independence is one of the most striking features of the mirror effect.
Why did magnetic mirror fusion reactors fall out of favor?
The loss cone is unavoidable in an open-ended mirror, and the resulting anisotropic velocity distribution drives loss-cone and drift-cyclotron instabilities that scatter particles out faster than collisions alone. After large machines like MFTF-B were built and cancelled in the 1980s, funding shifted to closed tokamaks — though high-field superconducting mirrors are now being revisited.