--- title: Particle Detectors and Subsystems module: Accelerators and Detectors moduleNumber: 11 lessonNumber: 2 order: 1102 summary: > A detector reads a collision by the energy particles deposit as they cross matter. Charged particles ionize at the Bethe-Bloch rate, radiate in the field of nuclei above a critical energy, and emit Cherenkov light above a velocity threshold; electrons and photons build electromagnetic showers over a radiation length, and hadrons build wider showers over a nuclear interaction length. The onion of tracker, electromagnetic and hadronic calorimeters, and outer muon chambers turns these processes into momentum, energy, and identity, with neutrinos inferred from missing transverse momentum. topics: [Accelerators and Detectors] draft: false sources: - book: Perkins ref: "Ch. 4 (passage of particles through matter; detectors and their subsystems)" - book: Thomson ref: "Appendix (particle detectors; calorimetry and tracking)" - book: Tipler & Llewellyn ref: "§11-9 (interaction of radiation with matter)" - book: Griffiths ref: "Ch. 1 §1.11 (experimental methods and detectors)" --- A detector does not see a particle; it records the energy the particle sheds while crossing matter. Every measurement — momentum, energy, charge, identity — is reconstructed from these deposits. The design of a detector therefore follows from the physics of energy loss: which particles ionize, which radiate, which shower, and over what distances. This lesson develops the interaction mechanisms, then assembles them into the layered "onion" that a general-purpose collider detector is. Natural units where convenient; energies in GeV, lengths in the material-specific scales defined below. ## Ionization: the Bethe-Bloch rate A charged particle traversing matter loses energy chiefly by ionizing and exciting atomic electrons. The mean energy lost per unit path length is the **Bethe-Bloch** formula, $$ -\frac{\d E}{\d x} = K\,z^2\,\frac{Z}{A}\,\frac{1}{\beta^2} \left[\tfrac12\ln\!\frac{2m_e c^2\beta^2\gamma^2 T_{\max}}{I^2} - \beta^2 - \frac{\delta}{2}\right], $$ where $z$ is the projectile charge, $Z/A$ and $I$ are the target's atomic number over mass number and mean excitation energy, $\beta\gamma$ is the projectile velocity, and $K\approx0.307\,\text{MeV}\,\text{cm}^2/\text{mol}$. The shape carries the physics. - **The $1/\beta^2$ rise.** A slow particle spends longer near each atom and loses more; the rate falls steeply as the particle speeds up. - **The minimum.** Near $\beta\gamma\approx3$ the loss reaches a broad minimum of about $2\,\text{MeV}\,\text{cm}^2/\text{g}$ for most materials. A particle near this minimum is **minimum ionizing** (a "MIP"), the reference deposit for detector design. - **The relativistic rise.** The logarithm's $\gamma^2$ makes the loss climb slowly at high energy, until the **density effect** $\delta$ — polarization of the medium screening the distant field — flattens it into the Fermi plateau. The loss depends on velocity, not momentum, and at fixed momentum different masses have different $\beta$. That mass dependence is what turns $\d E/\d x$ into a particle identifier: a measured $\d E/\d x$ together with a measured momentum picks out the mass. $$ % caption: The Bethe-Bloch energy loss versus momentum for several particle species. % At a given momentum the lighter particle is faster, sits higher up the one-over-beta % squared curve, and loses less, so the bands separate and identify the mass. \begin{tikzpicture}[>=stealth, font=\footnotesize, scale=1.0] \definecolor{acc}{HTML}{4A6FA5} \draw[->, black] (0,0) -- (6.4,0) node[right, black] {momentum (log)}; \draw[->, black] (0,0) -- (0,4.0) node[above, black] {energy loss}; % each species: steep 1/beta^2 fall to a minimum, then slow rise. Shift right for % heavier mass (same beta needs more momentum). % electron / lightest (leftmost) \draw[very thick, domain=0.35:6.0, samples=90, variable=\x] plot ({\x}, {0.9 + 1.9/((\x-0.15)) + 0.10*ln(\x+1)}); % pion \draw[very thick, densely dotted, domain=0.95:6.0, samples=90, variable=\x] plot ({\x}, {0.9 + 1.9/((\x-0.75)) + 0.10*ln(\x+1)}); % kaon \draw[very thick, dashed, domain=1.7:6.0, samples=80, variable=\x] plot ({\x}, {0.9 + 1.9/((\x-1.5)) + 0.10*ln(\x+1)}); % proton (rightmost) \draw[very thick, domain=2.6:6.0, samples=70, variable=\x] plot ({\x}, {0.9 + 1.9/((\x-2.4)) + 0.10*ln(\x+1)}); \node[right] at (0.7,3.6) {electron}; \node[right] at (2.35,3.6) {pion}; \node[right] at (3.5,3.2) {kaon}; \node[right] at (4.7,2.7) {proton}; \end{tikzpicture} $$ ## Radiation, the radiation length, and electromagnetic showers Ionization dominates for heavy charged particles at ordinary energies, but a light charged particle in the strong field near a nucleus loses energy a second way: it **radiates** a photon (bremsstrahlung). The radiative loss grows linearly with energy and inversely with the squared mass, $$ -\frac{\d E}{\d x}\bigg|_{\text{rad}} \;\propto\; \frac{E}{m^2}, $$ so it matters overwhelmingly for electrons and is suppressed by $(m_\mu/m_e)^2\approx4\times10^4$ for muons. The energy at which radiative loss equals ionization loss is the **critical energy** $E_c$, a few MeV to tens of MeV depending on material. Above $E_c$ an electron radiates faster than it ionizes. The characteristic length is the **radiation length** $X_0$: the distance over which an electron's energy falls to $1/e$ of its value by radiation, and also (up to a factor $\tfrac79$) the mean free path of a high-energy photon for pair production. Two processes feed each other above $E_c$: - a high-energy electron radiates a photon, - a high-energy photon converts to an $e^+e^-$ pair. Each step roughly halves the energy per particle and doubles the particle count, so a single incident electron or photon triggers a cascade — an **electromagnetic shower** — that multiplies until the typical particle energy drops below $E_c$, after which ionization absorbs the rest. The shower's longitudinal depth scales as $X_0$ and grows only logarithmically with the incident energy, $t_{\max}\sim\ln(E/E_c)$, so a calorimeter a couple of dozen radiation lengths deep contains showers over a wide energy range in a compact volume. The transverse spread is set by the **Molière radius** $R_M$, within which about 90% of the energy is deposited. $$ % caption: Electromagnetic shower development. A single electron or photon initiates a % cascade: bremsstrahlung and pair production alternate, roughly doubling the particle % count and halving the energy each radiation length, until the particles fall below % the critical energy and stop multiplying. \begin{tikzpicture}[>=stealth, font=\footnotesize, scale=1.0] \definecolor{acc}{HTML}{4A6FA5} % incoming \draw[acc, very thick, ->] (-6.6,0.0) -- (-5.6,0.0); % generation 1 split \draw[thick] (-5.6,0.0) -- (-4.3,0.7); \draw[thick] (-5.6,0.0) -- (-4.3,-0.7); \fill (-5.6,0.0) circle (1.6pt); % generation 2 split \draw[thick] (-4.3,0.7) -- (-3.0,1.15); \draw[thick] (-4.3,0.7) -- (-3.0,0.35); \draw[thick] (-4.3,-0.7) -- (-3.0,-0.35); \draw[thick] (-4.3,-0.7) -- (-3.0,-1.15); \fill (-4.3,0.7) circle (1.4pt); \fill (-4.3,-0.7) circle (1.4pt); % generation 3 split (fan) \foreach \yy in {1.15,0.35,-0.35,-1.15}{ \draw[thin] (-3.0,\yy) -- (-1.7,\yy+0.22); \draw[thin] (-3.0,\yy) -- (-1.7,\yy-0.22); \fill (-3.0,\yy) circle (1.2pt); } \node[acc, left] at (-6.6,0.0) {electron}; % depth axis \draw[->, black] (-6.4,-1.8) -- (-1.4,-1.8) node[right, black] {depth in $X_0$}; \node[black] at (-3.9,1.75) {shower maximum}; \end{tikzpicture} $$ ## Hadronic showers and the nuclear interaction length A hadron entering dense matter loses energy by ionization like any charged particle, but it also undergoes strong inelastic collisions with nuclei, producing secondary pions, protons, and neutrons that in turn collide. The resulting **hadronic shower** is governed not by $X_0$ but by the **nuclear interaction length** $\lambda_I$, the mean free path for an inelastic nuclear collision. In most materials $\lambda_I$ is much longer than $X_0$ (in iron, $\lambda_I\approx17\,\text{cm}$ against $X_0\approx1.8\,\text{cm}$), so hadronic showers start later, penetrate deeper, and spread wider than electromagnetic ones. A hadron calorimeter must therefore be many $\lambda_I$ thick and sits outside the electromagnetic calorimeter. Hadronic showers are also intrinsically noisier to measure. A variable fraction of the energy goes into breaking up nuclei, into slow neutrons, and into neutrinos and muons from pion decay that escape, and neutral pions decay to photons and feed an electromagnetic subshower. This lost and fluctuating "invisible" energy makes hadronic energy resolution worse than electromagnetic resolution and is the reason jets are measured less precisely than electrons and photons. ## Cherenkov radiation and velocity thresholds A charged particle moving through a medium faster than the local speed of light in that medium, $v>c/n$, emits a coherent shock of light — **Cherenkov radiation** — at a fixed angle to its track, $$ \cos\theta_c = \frac{1}{n\beta}. $$ Light appears only above the **threshold** $\beta>1/n$, and the emission angle grows with velocity up to a maximum at $\beta\to1$. Both features are exploited for particle identification: at a fixed momentum, a lighter (faster) particle may exceed threshold while a heavier one does not, and the Cherenkov angle measures $\beta$ directly. Combined with a momentum measurement, either observation yields the mass. Detectors built on this include threshold counters (light or no light) and ring-imaging Cherenkov detectors that reconstruct the cone angle from the radius of the light ring. $$ % caption: Cherenkov emission. A charge crossing a medium faster than light does in % that medium radiates a coherent wavefront on a cone whose half-angle satisfies cosine % theta equals one over n beta; the light appears only above the velocity threshold % beta greater than one over n. \begin{tikzpicture}[>=stealth, font=\footnotesize, scale=1.0] \definecolor{acc}{HTML}{4A6FA5} % track \draw[very thick, ->] (-3.4,0.0) -- (3.2,0.0) node[right] {track}; \fill (0.0,0.0) circle (1.8pt); % wavefront cone (two lines making the Mach-like cone behind the particle) \draw[acc, thick] (0.0,0.0) -- (-2.2,1.5); \draw[acc, thick] (0.0,0.0) -- (-2.2,-1.5); % emitted wavefronts as short parallel segments \draw[black] (-0.7,0.0) -- (-1.2,0.55); \draw[black] (-1.4,0.0) -- (-1.9,0.55); \draw[black] (-0.7,0.0) -- (-1.2,-0.55); \draw[black] (-1.4,0.0) -- (-1.9,-0.55); % angle marker \draw[black] (0.9,0.0) arc (180:146:0.9); \node[black] at (0.62,0.5) {angle}; \node[above] at (-2.2,1.5) {wavefront}; \end{tikzpicture} $$ ## The onion: a general-purpose collider detector A detector at a collider surrounds the interaction point with concentric layers, ordered so that each particle type is measured before it is stopped, and only the particles meant to reach a given layer do. From the beam pipe outward: - **Tracker (inner).** Thin, low-mass layers of silicon or gas in a strong solenoidal magnetic field. Charged particles leave hits that reconstruct their curved trajectories while depositing little energy, so the particle survives to the outer layers. The curvature gives momentum; the direction of the bend gives the sign of the charge; the ionization density can give $\d E/\d x$. - **Electromagnetic calorimeter (ECAL).** A dense absorber, tens of $X_0$ deep, in which electrons and photons deposit their full energy as electromagnetic showers. Hadrons pass through mostly intact because $\lambda_I\gg X_0$. - **Hadronic calorimeter (HCAL).** A thicker, coarser absorber, several $\lambda_I$ deep, in which hadrons shower and deposit their energy. Together the two calorimeters absorb everything except muons and neutrinos. - **Muon chambers (outer).** Tracking layers beyond the calorimeters. Only muons and neutrinos reach here; muons ionize as MIPs and register tracks, neutrinos do not. The layers cooperate. Momentum comes from the inner tracker; energy comes from the calorimeters; the combination identifies the particle by which layers it touched. $$ % caption: Cross-section of a general-purpose collider detector. Concentric layers, % from the beam pipe outward: tracker in a magnetic field, electromagnetic % calorimeter, hadronic calorimeter, and muon chambers. Each particle is measured % before it is absorbed, and only muons and neutrinos reach the outermost layer. \begin{tikzpicture}[>=stealth, font=\footnotesize, scale=1.0] \definecolor{acc}{HTML}{4A6FA5} % concentric arcs (quarter section to keep labels clear) \fill (0,0) circle (1.2pt); \draw[thick] (0,0) circle (0.9); \draw[thick] (0,0) circle (1.7); \draw[thick] (0,0) circle (2.5); \draw[thick] (0,0) circle (3.2); \node at (0.0,0.45) {tracker}; \node at (0.0,1.30) {ECAL}; \node at (0.0,2.10) {HCAL}; \node at (0.0,2.85) {muon}; % a few outgoing tracks along the right \draw[very thick, ->] (0,0) -- (3.6,0.4); \draw[very thick, ->] (0,0) -- (2.9,-1.7); \node[right] at (3.6,0.4) {muon}; \node[right] at (2.9,-1.7) {hadron}; \end{tikzpicture} $$ ## Momentum from curvature In the tracker's solenoidal field $B$ a charged particle follows a helix whose transverse projection is a circle of radius $r$. The transverse momentum follows from the same relation that governs the accelerator's magnets, $$ p_T\,[\text{GeV}] = 0.3\,B\,[\text{T}]\;r\,[\text{m}], $$ so measuring the sagitta of the arc over the tracker's radial span measures $p_T$, and the direction of curvature fixes the charge sign. The **momentum resolution** degrades with momentum: a stiffer track is straighter, its sagitta smaller, and the fractional error grows linearly, $$ \frac{\sigma_{p_T}}{p_T} \;\propto\; p_T, $$ so at high momentum the tracker's precision runs out. Calorimeters have the opposite trend — their fractional energy resolution improves with energy, $\sigma_E/E\propto 1/\sqrt E$, because a higher-energy shower contains more sampled particles and the relative statistical fluctuation shrinks. The two measurements are complementary: the tracker wins at low momentum, the calorimeter at high energy, and their crossover guides how a detector weights each for a given particle. $$ % caption: Complementary resolutions. Tracker momentum resolution degrades with % momentum (fractional error rises linearly), while calorimeter energy resolution % improves as one over the square root of energy; the crossover sets which subsystem % measures a given particle best. \begin{tikzpicture}[>=stealth, font=\footnotesize, scale=1.0] \definecolor{acc}{HTML}{4A6FA5} \draw[->, black] (0,0) -- (6.4,0) node[right, black] {energy or momentum}; \draw[->, black] (0,0) -- (0,4.0) node[above, black] {fractional error}; % tracker: rising linearly \draw[very thick, domain=0.2:6.0, samples=2, variable=\x] plot ({\x}, {0.55*\x}); % calorimeter: 1/sqrt(E) falling \draw[very thick, dashed, domain=0.15:6.0, samples=80, variable=\x] plot ({\x}, {1.5/sqrt(\x)}); \node[right] at (5.2,3.35) {tracker}; \node[right] at (4.7,0.85) {calorimeter}; \end{tikzpicture} $$ ## Particle identification and missing energy Putting the layers together, each particle type leaves a distinct **signature** across the detector, and the pattern identifies it. | Particle | Tracker | ECAL | HCAL | Muon | | --- | --- | --- | --- | --- | | Electron | curved track | full shower | none | none | | Photon | no track | full shower | none | none | | Charged hadron | curved track | small deposit | full shower | none | | Neutral hadron | no track | small deposit | full shower | none | | Muon | curved track | MIP deposit | MIP deposit | track | | Neutrino | none | none | none | none | An electron and a photon both shower in the ECAL; the presence or absence of a matching track distinguishes them. A charged and a neutral hadron both shower in the HCAL; again the track separates them. A muon is the only particle that leaves a track before and after the calorimeters. A neutrino leaves nothing at all. Because a neutrino is invisible, it is inferred from what is missing. In a collider the initial transverse momentum is zero, so the vector sum of all measured transverse momenta must also vanish; any imbalance is **missing transverse momentum**, the signature of one or more neutrinos (or of any other non-interacting particle, which is why missing momentum is also the collider handle on invisible new physics). The longitudinal balance cannot be used at a hadron collider because the colliding partons carry unknown longitudinal momentum fractions, so only the transverse component is conserved and measurable. Missing transverse momentum reconstructing the neutrino is how leptonic $W$ decays and many searches are found. ## Summary A detector measures the energy particles deposit while crossing matter. Charged particles ionize at the Bethe-Bloch rate, which falls as $1/\beta^2$ to a minimum-ionizing plateau and, through its velocity dependence, identifies mass at fixed momentum. Light charged particles and photons radiate and pair-produce above the critical energy, cascading into electromagnetic showers over a radiation length $X_0$; hadrons shower over the longer nuclear interaction length $\lambda_I$ with worse resolution from invisible energy. Cherenkov light above $\beta>1/n$ measures velocity and thresholds on mass. The onion — tracker in a magnetic field, ECAL, HCAL, muon chambers — measures each particle before absorbing it: momentum from track curvature ($p_T=0.3Br$, resolution degrading as $p_T$), energy from calorimeters (resolution improving as $1/\sqrt E$), identity from the pattern of layers touched, and neutrinos from missing transverse momentum. Turning these reconstructed events into a discovery is the subject of the [next lesson](/particle-physics/experiment/how-discoveries-are-made). [^detmatter]: The Bethe-Bloch formula, the minimum-ionizing rate, radiation and critical energy, the radiation length and electromagnetic showers, the nuclear interaction length and hadronic showers, and Cherenkov radiation are treated in Perkins, Ch. 4, and Tipler & Llewellyn, §11-9. Detector architecture — tracking in a solenoidal field, electromagnetic and hadronic calorimetry, muon systems, and momentum-from-curvature $p_T=0.3Br$ — follows Thomson, Appendix, and Griffiths, §1.11. Material constants ($X_0$, $\lambda_I$, $\d E/\d x$, mean excitation energies) and the passage-of-particles review are tabulated by the Particle Data Group, [pdg.lbl.gov](https://pdg.lbl.gov).