Interaction of particles with matter

A particle physics detector never sees a particle — it sees the wake the particle leaves in matter. Only a handful of species live long enough and interact strongly enough to reach a sensor: the electron, proton and photon are stable, the neutrino is nearly invisible, and the muon, neutron, charged pions and kaons survive just long enough to travel a measurable $\sim\gamma v\tau$ before decaying. Everything we know about the subatomic world is reconstructed backwards, from the tracks, showers, glows and pulses these travellers deposit as they cross a gas, a crystal, a water tank or a wafer of silicon. Those deposits are a small family of electromagnetic and strong processes, each with its own scaling law. A charged particle drags on the atomic electrons and loses energy by ionisation, quantified by the Bethe–Bloch equation, whose $1/\beta^2$ velocity dependence makes slow particles the most destructive and fast ones nearly uniform. An electron, being light, instead radiates in the field of a nucleus — bremsstrahlung — once above a critical energy, and a high-energy photon converts to an $e^+e^-$ pair; together these seed the electromagnetic showers that fill a calorimeter, all measured in units of one radiation length. And a charge moving faster than light does in the medium sheds a Cherenkov cone whose opening angle fixes its velocity. This article grounds each of these four processes in one equation and shows how a detector reads them together to weigh, track and name a particle it can never touch.

Key takeaways

Charged particles lose energy by ionising atoms, described by the Bethe–Bloch equation, which falls as $1/\beta^2$ and depends on the medium only through $Z/A$ (hence, per g/cm², almost material-independent).
The loss bottoms out near $\beta\gamma\approx3$ – $4$ at about 2 MeV cm²/g — a minimum-ionising particle — then climbs slowly in the logarithmic relativistic rise.
Electrons above the critical energy $E_c\approx800/Z$ MeV lose energy mainly by bremsstrahlung; high-energy photons by $e^+e^-$ pair production. Both scale with the radiation length $X_0$ and seed electromagnetic showers.
Radiative loss is suppressed by $(m_e/m)^2$ , so muons scarcely radiate below $\sim100$ GeV and punch through metres of steel — the basis of muon identification.
Cherenkov radiation appears when $v>c/n$ , at angle $\cos\theta_c=1/(n\beta)$ , giving a velocity measurement and particle-identification thresholds.
Key idea: each particle type leaves a characteristic signature in matter — an ionisation trail, a burst of radiation, an electromagnetic shower, or a Cherenkov cone — and every detector is an engine for reading those signatures back into momenta and identities.

The core insight

Each particle type leaves a characteristic signature in matter — ionisation trails, radiated photons, electromagnetic showers, or Cherenkov cones — and detectors are engineered to read these signatures back into the momenta and identities of particles they can never touch directly.

Where this lives today

Every one of these processes is a working technology in a running experiment. The relativistic rise of $dE/dx$ is read out ionisation-by-ionisation in the time-projection chambers of ALICE and STAR, separating pions, kaons and protons track by track. Minimum-ionising particles set the energy scale of the silicon trackers at ATLAS and CMS, whose curvatures in a $2$ – $4$ T solenoid deliver momenta to better than a percent via $p_T=0.3\,BR$ . Radiation length is the design unit of every electromagnetic calorimeter, from CMS's lead-tungstate crystals ( $X_0=0.89$ cm) to lead-scintillator sandwiches that gauge a photon by the shower size of Worked Example 2. Cherenkov light is the whole detector at the largest scales: Super-Kamiokande reconstructs neutrinos from rings on a $50$ -kilotonne water tank, IceCube instruments a cubic kilometre of Antarctic ice, and ring-imaging (RICH) counters at LHCb turn the cone angle into the velocity that tags kaons in $B$ -meson decays. Even medicine runs on Bethe–Bloch: the $1/\beta^2$ divergence puts the Bragg peak of a proton beam precisely at a tumour's depth — the basis of proton and carbon-ion therapy. The physics is a century old; the instruments built on it define the experimental frontier.

The physical picture

Start with the question a detector actually asks: what did this thing do to the material it crossed? The answer sorts every particle into one of a few behaviours. A heavy charged particle — a proton, pion, kaon or muon — ploughs a nearly straight furrow, knocking electrons loose all along its path and losing energy smoothly by ionisation. A light charged particle — an electron — does that too, but above a critical energy it mostly radiates, spraying photons that convert to pairs that radiate again, a cascade that dies only when the pieces are too soft to multiply. A photon is invisible until it converts, then showers likewise. And any charge faster than light in the medium trails a coherent optical shock, a Cherenkov cone. The first figure lays these signatures side by side in one slab of matter.

Figure

Three signatures in one slab of matter — a heavy charged particle draws a straight ionisation trail governed by Bethe–Bloch, an electron or photon above the critical energy cascades into a shower that doubles every radiation length X₀, and a particle faster than c/n radiates a Cherenkov cone at cosθ = 1/(nβ), about 41° in water

The single most useful curve in the subject is the ionisation loss versus how fast the particle is going, and it is not monotonic. A slow particle lingers near each atom and loses energy furiously, so $dE/dx$ is huge at low velocity and falls as $1/\beta^2$ ; it reaches a broad minimum near $\beta\gamma\approx3$ – $4$ , the minimum-ionising particle every detector is calibrated against; then it creeps up in a logarithmic relativistic rise until the polarisation of the medium (the density effect) caps it into a plateau. The second figure is that curve, with the number a detector builder actually uses.

Figure

The Bethe–Bloch curve of mean ionisation loss versus βγ: the loss falls as 1/β² at low velocity, reaches a broad minimum of about 2 MeV cm²/g near βγ ≈ 3 (the minimum-ionising particle), then climbs through the logarithmic relativistic rise before the density effect flattens it into a plateau

The core equations

Ionisation is the workhorse, so its equation comes first. The Bethe–Bloch formula gives the mean rate at which a heavy charged particle loses energy to atomic electrons, per unit path (usually quoted per g/cm², i.e. divided by density). Read it as a constant, times the electron content $Z/A$ , times the velocity factor $1/\beta^2$ , times a slow logarithm.

-\left\langle\frac{dE}{dx}\right\rangle = K\,z^2\,\frac{Z}{A}\,\frac{1}{\beta^2}\left[\frac{1}{2}\ln\frac{2 m_e c^2\,\beta^2\gamma^2\,T_{\max}}{I^2} - \beta^2 - \frac{\delta(\beta\gamma)}{2}\right] \qquad (1)

where: $K=4\pi N_A r_e^2 m_e c^2 = 0.307\ \mathrm{MeV\,cm^2/mol}$ ; $z$ — charge of the incident particle in units of $e$ ; $Z,A$ — atomic number and mass of the medium; $\beta,\gamma$ — velocity and Lorentz factor of the particle; $m_ec^2=0.511$ MeV; $I$ — mean excitation energy of the medium ( $I\approx10\,Z$ eV); $T_{\max}$ — maximum energy transferable to one electron; $\delta(\beta\gamma)$ — the density-effect correction that flattens the relativistic rise. The whole material dependence is carried by $Z/A\approx0.5$ (nearly constant) and by $I$ inside a logarithm — which is why the loss per g/cm² barely changes from one solid to another. The kinematic ceiling $T_{\max}$ sets the top of the logarithm — the head-on-collision energy transfer — and for any projectile much heavier than the electron the recoil corrections drop out.

T_{\max}=\frac{2 m_e c^2\,\beta^2\gamma^2}{1+2\gamma\,m_e/M+(m_e/M)^2}\ \xrightarrow{\ M\gg m_e\ }\ 2 m_e c^2\,\beta^2\gamma^2 \qquad (2)

where: $M$ — mass of the incident particle. Substituting the heavy-particle limit into (1) collapses the bracket to the compact Thomson form $\ln(2m_ec^2\beta^2\gamma^2/I)-\beta^2$ : a single logarithm of $\beta^2\gamma^2$ over the excitation energy. The rise of $T_{\max}$ with $\gamma^2$ is one source of the relativistic rise; the other is the growth of the field's transverse reach. For a light particle the dominant loss is radiative. An electron in the field of a nucleus emits bremsstrahlung at a rate proportional to its energy, which integrates to exponential attenuation on the scale of the radiation length $X_0$ ; the energy where radiation overtakes ionisation is the critical energy $E_c$ .

-\left(\frac{dE}{dx}\right)_{\text{rad}}=\frac{E}{X_0}\ \Rightarrow\ E(x)=E_0\,e^{-x/X_0},\qquad E_c:\ \left(\frac{dE}{dx}\right)_{\text{ion}}=\left(\frac{dE}{dx}\right)_{\text{rad}} \qquad (3)

where: $X_0$ — radiation length, the depth over which an electron's energy falls by $1/e$ (lead $0.56$ cm, water $36$ cm); $E_c$ — critical energy, $\approx800/Z$ MeV, where ionisation and radiative losses are equal; $E_0$ — initial energy. A high-energy photon pair-produces on the same scale, with mean free path $\tfrac{9}{7}X_0$ , so $X_0$ is the universal ruler of electromagnetic showers. Because the radiative rate scales as $1/M^2$ , replacing the electron with a muon multiplies $E_c$ by $(m_\mu/m_e)^2\approx4\times10^4$ . A charged particle faster than the local speed of light radiates coherently in the optical, and the geometry of that shock front fixes both a threshold and an angle that measures velocity directly.

\cos\theta_c=\frac{1}{n\beta},\qquad \frac{dN}{dx}=2\pi\alpha\,z^2\sin^2\theta_c\left(\frac{1}{\lambda_1}-\frac{1}{\lambda_2}\right) \qquad (4)

where: $n$ — refractive index of the medium; $\theta_c$ — half-angle of the Cherenkov cone measured from the track; $\alpha\approx1/137$ ; $\lambda_{1,2}$ — the wavelength window of the detector. Radiation exists only above threshold $\beta>1/n$ (where the cosine would exceed one); as $\beta\to1$ the angle saturates at $\cos\theta_c=1/n$ ( $41°$ in water). The photon yield of a few hundred per cm is modest but directional. Finally, the ionisation trail is read out in a magnetic field, where the curvature of the helix converts a geometric radius into a momentum — the tracker's central measurement.

p_T\,[\mathrm{GeV}/c]=0.3\,B\,[\mathrm{T}]\,R\,[\mathrm{m}] \qquad (5)

where: $p_T$ — momentum component transverse to the field; $B$ — magnetic flux density; $R$ — radius of curvature of the track. The full momentum follows from the dip angle $\lambda$ of the helix as $p\cos\lambda=0.3\,BR$ . A $1$ GeV track in a $2$ T solenoid curls with $R\approx1.7$ m — large but precisely measurable, which is why trackers are big.

Derivation AThe stopping power from a classical collision picture

The goal is to show where the two defining features of Bethe–Bloch — the $1/\beta^2$ velocity law and the logarithm of $\beta^2\gamma^2$ — come from, using only the impulse a passing charge delivers to one atomic electron and a sum over the electrons in a slab. This is Bohr's 1913 classical argument; Bethe's quantum treatment changes only the contents of the logarithm, not its structure, so the classical route already exposes every scaling the full formula carries.

Step 1. Consider the incident charge $ze$ flying past one electron at impact parameter $b$ with speed $v$ . The transverse electric impulse delivers a momentum kick $\Delta p\approx\dfrac{2ze^2}{4\pi\varepsilon_0\,b\,v}$ , so the electron picks up energy $\Delta E(b)=\dfrac{\Delta p^2}{2m_e}=\dfrac{2z^2e^4}{(4\pi\varepsilon_0)^2 m_e v^2\,b^2}$ . Why: the fast charge acts for a time $\sim b/v$ , so a slower particle lingers longer and transfers more — this single factor is the origin of the $1/v^2$ that makes slow particles ionise the most.

Step 2. Count the electrons at impact parameter between $b$ and $b+db$ in a thickness $dx$ : they lie in a cylindrical shell of volume $2\pi b\,db\,dx$ , with electron density $n_e=N_A Z\rho/A$ , so the energy lost to that shell is $\Delta E(b)\times n_e\,2\pi b\,db\,dx$ . Why: turning one collision into energy per unit length means integrating over every electron the particle sweeps past, and the geometry supplies one power of $b$ that will tame the $1/b^2$ .

Step 3. Integrate over impact parameters: $-\dfrac{dE}{dx}=\dfrac{4\pi z^2 e^4 n_e}{(4\pi\varepsilon_0)^2 m_e v^2}\displaystyle\int_{b_{\min}}^{b_{\max}}\dfrac{db}{b}=\dfrac{4\pi z^2 e^4 n_e}{(4\pi\varepsilon_0)^2 m_e v^2}\ln\dfrac{b_{\max}}{b_{\min}}$ . The $1/b^2$ times $b\,db$ leaves $db/b$ , whose integral is a logarithm — the second signature feature. Why: the logarithm, not a power law, is what makes the loss only weakly dependent on the details; almost every material ends up within a factor of two.

Step 4. Fix the cutoffs physically. The closest approach $b_{\min}$ is set by the largest allowed momentum transfer ( $b_{\min}\sim\hbar/\gamma m_e v$ ), and the largest $b_{\max}$ by the adiabatic limit — beyond $\sim\gamma v\hbar/I$ the atom responds too slowly to absorb energy. Their ratio gives $\ln(b_{\max}/b_{\min})\sim\ln(\gamma^2 m_e v^2/I)$ , and with $e^2/4\pi\varepsilon_0=r_e m_e c^2$ the prefactor becomes $4\pi N_A r_e^2 m_e c^2\,(Z/A)(1/\beta^2)$ . Why: the two cutoffs are where the $\gamma^2$ and the excitation energy $I$ enter, completing the structure $K(Z/A)\beta^{-2}\ln(2m_ec^2\beta^2\gamma^2/I)$ that Bethe's quantum calculation reproduces with the right numerical factors. *Verified: prefactor $K=4\pi N_A r_e^2 m_e c^2=0.307$ MeV cm²/mol; loss $\propto z^2(Z/A)\beta^{-2}$ with a logarithm rising as $\gamma^2$ .*

Go deeper · Derivation B · 2 minCritical energy and the radiation length

Derivation BCritical energy and the radiation length

The aim is to explain why light particles behave so differently from heavy ones by comparing the two loss mechanisms on one axis: ionisation, nearly constant with energy, versus radiation, which grows in proportion to it. Where the curves cross defines the critical energy $E_c$ ; the constant setting the radiative slope is the radiation length $X_0$ . Together they predict exponential attenuation and one universal length scale shared by electrons and photons — the fact that makes calorimeter design possible.

Step 1. Write the two losses. Ionisation is $(dE/dx)_{\text{ion}}\approx$ constant (the broad Bethe–Bloch minimum, a few MeV per g/cm² with only a logarithmic energy dependence). Radiation in the nuclear field grows linearly, $(dE/dx)_{\text{rad}}=E/X_0$ , because each bremsstrahlung photon carries off a fixed fraction of the electron's energy. Why: one loss is roughly flat and the other rises without limit, so at high enough energy radiation must win — the only question is where the crossover sits.

Step 2. Define the critical energy by equality: $E_c$ is the energy at which $(dE/dx)_{\text{ion}}=(dE/dx)_{\text{rad}}=E_c/X_0$ . Empirically $E_c\approx800/Z$ MeV, so it is $\sim10$ MeV in lead and $\sim100$ MeV in air. Above $E_c$ an electron radiates faster than it ionises. Why: $E_c$ is the single number that says whether a given electron will make a shower or just leave a track — it is the pivot of the whole electromagnetic-shower story.

Step 3. Integrate the radiative loss alone: $dE/E=-dx/X_0$ gives $E(x)=E_0\,e^{-x/X_0}$ . So $X_0$ is by definition the depth over which the mean electron energy falls by a factor $e$ . Why: exponential attenuation with a material-specific but energy-independent length is what lets us quote shower depths in dimensionless units of $X_0$ and compare lead to water directly.

Step 4. Tie in the photon. A high-energy photon converts to $e^+e^-$ with mean free path $\tfrac{9}{7}X_0$ — the same scale, since both processes are the same $Z^2$ nuclear-field diagram read in two directions. An electron radiates a photon, the photon makes a pair, each member radiates again: the particle count roughly doubles every $X_0$ until the average energy drops to $E_c$ and multiplication stops. Why: sharing one length scale makes the shower a self-similar cascade whose depth $\sim\ln(E_0/E_c)$ and size $\sim E_0/E_c$ are predictable — the basis of every sampling calorimeter. *Verified: $E_c\approx800/Z$ MeV; $E(x)=E_0e^{-x/X_0}$ ; pair-production mean free path $\tfrac{9}{7}X_0$ ; radiative rate $\propto1/M^2$ .*

Go deeper · Derivation C · 2 minThe Cherenkov angle from a shock-wave construction

Derivation CThe Cherenkov angle from a shock-wave construction

The goal is to derive the Cherenkov relation $\cos\theta_c=1/(n\beta)$ purely from wavefront geometry — the Huygens construction that also gives the Mach cone of a supersonic aircraft. Cherenkov light is not emitted specially by any one atom; it is the coherent envelope of ordinary spherical wavelets, which add constructively only when the source outruns them. That one condition produces both the threshold and the angle a detector turns into a velocity.

Step 1. As the charged particle crosses the medium it briefly polarises the atoms along its path; each disturbed atom relaxes and emits a spherical wavelet that travels at the phase speed of light in the medium, $c/n$ . Why: the medium, not the particle, is the radiator — this is why Cherenkov emission depends on the refractive index and vanishes in vacuum where there is nothing to polarise.

Step 2. In a time $t$ the particle advances a distance $\beta c\,t$ along the track, while a wavelet emitted at the start has expanded to radius $(c/n)t$ . If $\beta c>c/n$ the particle outruns its own wavelets, and their envelope is a cone — exactly as a supersonic source builds a Mach cone. Why: the inequality $\beta>1/n$ is the threshold; below it the wavelets always stay ahead of the particle, no coherent front forms, and there is no radiation.

Step 3. The cone's surface is tangent to every wavelet. In the right triangle formed by the particle's path (hypotenuse $\beta c\,t$ ) and the wavelet radius (opposite side $(c/n)t$ ), the half-angle $\theta_c$ between the track and the wavefront-normal satisfies $\cos\theta_c=\dfrac{(c/n)t}{\beta c\,t}=\dfrac{1}{n\beta}$ . Why: this is the whole result — a purely geometric ratio of two distances covered in the same time, with no dynamics, which is why it measures $\beta$ so cleanly.

Step 4. Read off the limits. At threshold $\beta=1/n$ , $\cos\theta_c=1$ and the cone collapses to the forward direction; as $\beta\to1$ the angle saturates at $\cos\theta_c=1/n$ , the maximum ( $41°$ in water). Measuring $\theta_c$ therefore pins $\beta$ , and combined with momentum from curvature it yields the mass — particle identification. Why: because $\theta_c$ maps one-to-one onto $\beta$ between threshold and saturation, a ring-imaging detector reads velocity straight off the radius of the light cone. *Verified: $\cos\theta_c=1/(n\beta)$ ; threshold $\beta=1/n$ ; water $n=1.33\Rightarrow\theta_c^{\max}=41°$ , $\gamma_{\text{thr}}=1.52$ .*

Limiting cases

Case	Result	Meaning
Low velocity, $\beta\gamma\ll1$	$dE/dx\propto1/\beta^2$	loss diverges as the particle slows — the Bragg peak at the end of a range
Minimum ionising, $\beta\gamma\approx3$ – $4$	$dE/dx\approx2$ MeV cm²/g	broad minimum, nearly the same per g/cm² for all solids — the detector's calibration point
Relativistic rise / plateau	$dE/dx$ grows as $\ln(\beta\gamma)^2$ , then flat	rise from the growing transverse field; the density effect $\delta$ caps it
Electron above $E_c$	radiation dominates, $E=E_0e^{-x/X_0}$	seeds electromagnetic showers; below $E_c$ it just ionises
Heavy particle (muon)	radiative loss $\propto1/M^2$ suppressed	ionisation dominates to $\sim100$ GeV; muons penetrate metres of steel
Photon, $E\gtrsim10$ MeV	pair production dominates	mean free path $\tfrac{9}{7}X_0$ ; below, Compton then photoelectric
Charged, $v>c/n$	Cherenkov cone at $\cos\theta_c=1/(n\beta)$	threshold $\beta=1/n$ ; a direct velocity measurement

Common trapdE/dx does not depend strongly on the material

The intuitive guess is that a dense, high- $Z$ material stops a particle far more effectively per centimetre — true, but only because of density. The physically meaningful quantity is the mass stopping power, $dE/dx$ divided by density (MeV cm²/g), and there the material dependence is astonishingly weak: it scales as $Z/A$ , near $0.5$ for everything from carbon to lead, plus a slow logarithm through $I$ . So the minimum ionisation is about $2$ MeV cm²/g in almost any solid, and the difference between two detectors is mostly density, not chemistry. A second, opposite error is to blur ionisation and radiation together. They obey different laws: ionisation is roughly flat in energy and mass-independent, while bremsstrahlung grows linearly with energy and is suppressed by $(m_e/m)^2$ — which is why an electron and a muon of the same energy behave completely differently, and why you must never quote one particle's radiative loss for another's.

Worked examples

1 · Cherenkov light in a water tank (fully worked)

A water Cherenkov detector has $n=1.33$ . Find the velocity threshold, the momentum threshold for a muon ( $m_\mu=105.7$ MeV), the maximum cone angle, and the number of visible photons ( $300$ – $600$ nm) emitted per cm by a relativistic track.

Step 1. Threshold velocity is where the cone collapses, $\cos\theta_c=1$ , i.e. $\beta=1/n=1/1.33=\mathbf{0.752}$ . The corresponding Lorentz factor is $\gamma_{\text{thr}}=1/\sqrt{1-\beta^2}=1/\sqrt{1-0.565}=\mathbf{1.52}$ . Only particles above this $\beta$ radiate at all.

Step 2. Momentum threshold for a muon: $p=\beta\gamma\,m_\mu c=(0.752)(1.52)(105.7)=1.14\times105.7=\mathbf{121\ MeV}/c$ . A muon slower than this leaves no Cherenkov ring — a built-in trigger threshold.

Step 3. Maximum cone angle at $\beta\to1$ : $\cos\theta_c^{\max}=1/n=0.752$ , so $\theta_c^{\max}=\arccos(0.752)=\mathbf{41°}$ . A relativistic track paints a ring whose radius encodes this angle; a slower track gives a smaller ring — that is how the tank measures velocity.

Step 4. Photon yield from Frank–Tamm, $\dfrac{dN}{dx}=2\pi\alpha z^2\sin^2\theta_c\left(\dfrac1{\lambda_1}-\dfrac1{\lambda_2}\right)$ . With $\sin^2\theta_c=1-1/n^2=1-0.565=0.435$ , $z=1$ , and $1/\lambda_1-1/\lambda_2=1/(300\,\text{nm})-1/(600\,\text{nm})=1.67\times10^4\ \mathrm{cm^{-1}}$ : $\dfrac{dN}{dx}=2\pi\cdot\dfrac{1}{137}\cdot0.435\cdot1.67\times10^4\approx\mathbf{330\ photons/cm}$ . A relativistic muon radiates a 41° cone of ≈ 330 visible photons per cm, once its momentum exceeds 121 MeV/c.

2 · An electromagnetic shower in lead

A $10$ GeV electron enters lead ( $X_0=0.56$ cm, $E_c\approx7.4$ MeV). Use the Heitler toy model — the number of particles doubles every radiation length and each carries half the energy — to estimate the depth of shower maximum, the number of particles there, and the total lead thickness to contain it.

Step 1. After $t$ radiation lengths there are $N(t)=2^t$ particles, each with energy $E_0/2^t$ . Multiplication stops when that per-particle energy reaches the critical energy, $E_0/2^{t_{\max}}=E_c$ , so $t_{\max}=\log_2(E_0/E_c)=\log_2(10000/7.4)=\log_2(1350)=\mathbf{10.4}$ radiation lengths.

Step 2. Convert to a real depth: $x_{\max}=t_{\max}X_0=10.4\times0.56\ \text{cm}=\mathbf{5.8\ cm}$ of lead. The whole $10$ GeV cascade peaks less than three inches into the metal — the reason calorimeters are compact.

Step 3. Particle count at maximum: $N_{\max}=2^{t_{\max}}=E_0/E_c=10000/7.4=\mathbf{1350}$ particles, all near the critical energy. The measured signal (ionisation or Cherenkov light) is proportional to the total track length $\propto E_0/E_c$ , which is why a calorimeter's response scales linearly with the incident energy.

Step 4. Containment needs several more radiation lengths past the peak, since the tail falls slowly; empirically $\sim(t_{\max}+4)$ to hold $95\%$ , so about $14\,X_0\approx\mathbf{8\ cm}$ of lead. A 10 GeV shower peaks at ≈ 5.8 cm (10 X₀) with ≈ 1350 particles and is contained in ≈ 8 cm of lead.

Physics of the portrait

The equation that governs how every charged particle reveals itself carries one name above all others, and its author is the same man who worked out how the electrons and the radiating electrons behave — Hans Bethe supplied both the ionisation formula and, with Heitler, the bremsstrahlung and pair-production theory that seeds every shower.

Origins · 1 minHans Bethe and how this was discovered

Hans Bethe

1906–2005 · the quantum theory of energy loss and radiation in matter

Hans Bethe was arguably the most complete theoretical physicist of the twentieth century, equally at home in nuclear structure, quantum electrodynamics, solid-state theory and astrophysics — and the detection of particles rests on two of his results. In 1930, building on Niels Bohr's 1913 classical estimate, Bethe applied the new quantum mechanics to the energy a fast charge loses to atomic electrons, deriving the stopping-power formula (soon refined by Felix Bloch) that still bears his name and still calibrates every tracking detector. Four years later, with Walter Heitler, he computed the cross-sections for bremsstrahlung and electron–positron pair production in the field of a nucleus — the Bethe–Heitler theory — which fixed the radiation length and made the electromagnetic shower a calculable object. Bethe went on to explain the carbon–nitrogen cycle powering the stars (his 1967 Nobel Prize), led the theoretical division at Los Alamos, and performed the first renormalised calculation of the Lamb shift in 1947, opening modern QED. That the same person gave us both how a particle ionises and how it radiates is fitting: the two processes, and their competition at the critical energy, are the whole physics of a particle passing through matter, and Bethe wrote the theory of each.

History

1913

— Bohr derives the classical stopping-power formula from impulse and impact parameter

1930

— Bethe gives the quantum energy-loss formula, soon refined by Bloch

1934

— Bethe and Heitler compute bremsstrahlung and pair production in the nuclear field

1934

— Cherenkov observes the faint blue glow later named for him

1937

— Frank and Tamm explain Cherenkov light as a coherent shock front, giving cosθ = 1/(nβ)

1940

— Fermi adds the density-effect correction that flattens the relativistic rise

Going further

Texts. Thomson, Modern Particle Physics, §1.2 (interactions of particles with matter) · Griffiths, Introduction to Elementary Particles, 2nd ed., §1.5 and Ch. 6 · Halzen & Martin, Quarks and Leptons, §6 (QED processes underlying bremsstrahlung and pair production) · Peskin & Schroeder, An Introduction to Quantum Field Theory, §5.5 (bremsstrahlung) and §6.1 · Schwartz, Quantum Field Theory and the Standard Model, Ch. 13 · Particle Data Group, Review of Particle Physics, "Passage of particles through matter" (the definitive tables) · Leo, Techniques for Nuclear and Particle Physics Experiments, Ch. 2.

Provenance. The Bethe–Bloch prefactor $K=4\pi N_A r_e^2 m_e c^2=0.307$ MeV cm²/mol and the classical impact-parameter derivation follow Bohr and Bethe; the radiation length and pair-production mean free path $\tfrac{9}{7}X_0$ are the Bethe–Heitler results, developed in detail by the bremsstrahlung and electromagnetic-showers articles and recalled here by name. A verifier should confirm: (1) $K(Z/A)\beta^{-2}$ scaling with $Z/A\approx0.5$ ; (2) the minimum ionisation $\approx2$ MeV cm²/g; (3) $E_c\approx800/Z$ MeV and $E(x)=E_0e^{-x/X_0}$ ; (4) Cherenkov threshold $\beta=1/n$ , angle $41°$ in water, yield $\approx330$ photons/cm ( $300$ – $600$ nm); (5) the shower estimate $t_{\max}=\log_2(E_0/E_c)\approx10.4$ , $N_{\max}\approx1350$ for $10$ GeV in lead; (6) $p_T=0.3\,BR$ . Material constants are standard PDG values (lead $X_0=0.56$ cm, water $X_0=36$ cm, $n_{\text{water}}=1.33$ ).

Prerequisites. Special-relativity kinematics · Natural units · The particle zoo. Leads to. Ionisation energy loss · Bremsstrahlung · Electromagnetic showers · Cherenkov radiation · Tracking detectors.

Ask the tutor: Ask me to walk through the Bethe–Bloch curve point by point, or to work out the Cherenkov angle, shower depth or momentum resolution for any particle, medium and detector you name.

Why does the ionisation loss rise again at high energy — shouldn't fast particles interact less?

Two effects turn the curve back up past the minimum. First, as $\gamma$ grows the particle's electric field is Lorentz-contracted into a thin transverse pancake that reaches electrons further from the track, so more atoms contribute — the maximum energy transfer $T_{\max}$ climbs as $\gamma^2$ , and that $\gamma^2$ sits inside the Bethe–Bloch logarithm. This is the relativistic rise, a slow logarithmic climb of perhaps $10\%$ per decade. Second, the rise does not continue forever: at high enough energy the medium polarises in response to the passing field and screens the distant electrons — the density effect first computed by Fermi, which caps the loss into the Fermi plateau. The full shape is a steep $1/\beta^2$ fall, a broad minimum, a gentle relativistic rise, then a flat plateau.

Is the energy loss really the same in every material?

Per gram per square centimetre, very nearly. The Bethe–Bloch loss scales as $Z/A$ , the electrons per nucleon, which is close to $0.5$ for everything heavier than helium ( $0.5$ for carbon, $0.46$ for iron, $0.40$ for lead); the only other material dependence is a logarithm of the excitation energy $I$ . So a minimum-ionising particle deposits roughly $1.5$ – $2$ MeV per g/cm² whether it crosses plastic, water, silicon or lead. What does change enormously is the loss per centimetre, since that is the mass loss times the density. The rule of thumb: measure thickness in g/cm² and the physics is almost universal; measure it in cm and density does all the work.

What exactly is the radiation length, and why is the same length used for electrons and photons?

The radiation length $X_0$ is the mean thickness over which a high-energy electron loses all but $1/e$ of its energy to bremsstrahlung — from $-dE/dx=E/X_0$ you get $E=E_0e^{-x/X_0}$ . It is a material property: $0.56$ cm in lead, $\sim36$ cm in water. A photon shares the same scale because bremsstrahlung (electron radiates a photon) and pair production (photon makes an $e^+e^-$ pair) are the same $Z^2$ nuclear-field diagram read in two directions; the photon's conversion mean free path works out to $\tfrac{9}{7}X_0$ . So an electron and a photon of the same energy build cascades on one common ruler — the coincidence that makes an electromagnetic shower simple enough to design a calorimeter around.

Why don't muons shower like electrons if both are charged leptons?

Because the muon is $\sim207$ times heavier, and radiative loss scales as $1/m^2$ . Bremsstrahlung is emission during acceleration, $a=F/m$ , so a particle $207$ times heavier accelerates $207$ times less and radiates $207^2\approx4\times10^4$ times more weakly. The critical energy where radiation overtakes ionisation is thus $\sim10$ MeV for electrons in lead but of order $1$ TeV for muons. Below that a muon loses energy only by slow ionisation, so a $10$ GeV muon sheds barely $1$ GeV crossing a metre of iron and emerges intact. That is why the outermost layers of a collider detector, beyond the calorimeters that stop everything else, are the muon chambers: the muon is the one charged particle that reliably gets there.

How does a detector turn Cherenkov light into a particle identity?

The cone angle measures velocity, and velocity plus momentum gives mass. A ring-imaging (RICH) detector lets a track radiate in a gas or liquid radiator, then focuses the cone with a mirror onto a photodetector plane, where it appears as a ring whose radius fixes $\theta_c$ and hence $\beta$ through $\cos\theta_c=1/(n\beta)$ . The tracker separately gives the momentum $p$ from curvature, and since $p=\gamma m\beta c$ with $\beta$ now known, the mass $m$ follows. Two particles of equal momentum but different mass — a pion and a kaon — travel at slightly different $\beta$ and make rings of slightly different radius, which is how LHCb tags kaons. Threshold counters use the binary fact that only particles above $\beta=1/n$ radiate, so a well-chosen index makes pions glow while heavier kaons of the same momentum stay dark.

Why do slow particles deposit the most energy, and how does cancer therapy exploit it?

The $1/\beta^2$ factor means a particle loses energy fastest when it is moving slowest, because it lingers near each atom and delivers a bigger impulse. As a charged particle slows to a stop it therefore dumps a sharply increasing dose right at the end of its path — the Bragg peak. A photon beam, by contrast, attenuates roughly exponentially and deposits most of its dose near the entrance. Proton and carbon-ion therapy exploit the difference: by tuning the beam energy, clinicians place the Bragg peak exactly at a tumour's depth, delivering a concentrated dose there while sparing the tissue in front of and — crucially — behind it, since the particle stops. The same $1/\beta^2$ that calibrates a tracker is thus a tool for depositing radiation with millimetre precision inside a patient.