Photomultiplier Tubes (PMTs) — Abridged Guide
Quick-reference guide to PMTs — photocathodes, gain, noise, timing, and selection. For full derivations and worked examples, see the Comprehensive Guide.
Comprehensive PMT Guide →
1.Introduction
PMTs are vacuum-based photodetectors that combine photoemission with secondary-electron multiplication to achieve internal gains of 10⁵–10⁸, enabling single-photon detection with sub-nanosecond timing.
PMTs remain the best choice when you need large detection area, ultra-low noise, and fast timing simultaneously. Solid-state alternatives (SiPMs) win on compactness, magnetic immunity, and NIR sensitivity.
2.Architecture and Types
All PMTs share the same core structure — window, photocathode, focusing electrode, dynode chain, and anode — enclosed in a vacuum envelope. Head-on (transmission) PMTs offer large uniform area; side-on (reflection) PMTs are compact for spectrophotometry.
For fastest timing, choose linear-focused dynodes or MCP-PMTs. For scintillation counting with large crystals, box-and-grid dynodes offer the best collection efficiency.
| Configuration | Best For | Active Area | Timing |
|---|---|---|---|
| Head-on | Scintillation, physics | 10–500+ mm | Moderate–fast |
| Side-on | Spectrophotometry | Small spot | Moderate |
| MCP-PMT | TCSPC, streak cameras | 10–50 mm | Ultra-fast (<100 ps) |
| Channel PMT | Confocal, low noise | Small | Moderate |
3.Photocathode Physics
QE–Responsivity Conversion
The photocathode material determines spectral range and QE. Bialkali (25–30% QE, 200–650 nm) is the workhorse. GaAsP (50% QE at 550 nm) is the premium choice for fluorescence. Multialkali extends to 850 nm at lower QE.
Always match the photocathode to your signal wavelength, not to the broadest available range. Bialkali has far lower dark current than multialkali — don't pay the noise penalty for red sensitivity you don't need.
| Cathode | Peak QE | Peak λ | Range | Dark Current |
|---|---|---|---|---|
| Bialkali | 25–30% | 370–400 nm | 200–650 nm | Low |
| Super bialkali | ~35% | 370–400 nm | 200–650 nm | Low |
| Multialkali | ~20% | 400–420 nm | 250–850 nm | Moderate |
| GaAsP | ~50% | 550 nm | 350–700 nm | Moderate–high |
| GaAs | ~12% @ 800 nm | Flat 350–850 | 300–930 nm | High |
4.Secondary Emission and Gain
PMT Gain
Gain depends exponentially on both the secondary emission ratio δ and the number of stages n. A 10-stage tube with δ = 5 gives G ≈ 10⁷. Gain scales steeply with supply voltage — 1% HV change ≈ 8% gain change for a 10-stage tube.
For photon counting, ensure first-dynode gain (δ₁) is as high as possible — this determines the quality of your pulse-height distribution and noise discrimination. GaP first dynodes (δ₁ > 20) give the best single-photon resolution.
5.Noise and SNR
Analog SNR (simplified)
NEP
PMT gain does not improve SNR — it only raises the signal above electronics noise. Once that threshold is crossed, SNR is determined solely by photocathode current, dark current, and bandwidth. The PMT's ENF of ~1.2–1.4 is far lower than APDs (>2), which is a key advantage.
Cooling a PMT to −20°C reduces dark current by 100–1000×. For the most demanding photon-counting applications, always use a cooled housing.
PMT Gain & Signal Calculator →6.Temporal Response
TTS (transit time spread) is the most important timing spec — it determines the PMT's intrinsic time resolution. Linear-focused: 0.2–1.5 ns. MCP-PMT: <50 ps. TTS limits TCSPC resolution and sets the instrument response function width.
Rise time and transit time are less important than TTS for most applications. A slow transit time just adds a fixed delay; TTS adds jitter that directly degrades measurement precision.
| Dynode Type | TTS (FWHM) | Rise Time | Best Application |
|---|---|---|---|
| Linear-focused | 0.2–1.5 ns | 1–3 ns | Photon counting, TCSPC |
| Circular cage | 1–2 ns | 2–4 ns | General purpose |
| Box-and-grid | 2–5 ns | 3–8 ns | Scintillation counting |
| MCP | <0.05 ns | <0.3 ns | Picosecond timing |
7.Operating Modes
Dead Time Correction
Photon Counting SNR
Analog mode measures average current (best above ~10⁶ photons/s). Photon counting mode counts individual pulses (best below ~10⁶ photons/s) and discriminates against sub-threshold noise. Photon counting SNR = √N_s when dark counts are negligible.
The resistive voltage divider bleeder current should be at least 10× the maximum expected anode signal current. For high-rate applications, add Zener diodes or use an active divider to prevent gain sag at the final stages.
8.Practical Considerations
Always use a µ-metal shield — even Earth's field degrades unshielded PMT performance. Keep average anode current below 1 µA for long life. Never expose a powered PMT to room light.
Allow 30 minutes of warm-up time after applying HV before taking quantitative measurements. Gain stabilizes as the tube reaches thermal equilibrium and charge redistribution settles.
9.PMTs vs. Solid-State
PMTs dominate for large area + low noise + fast timing. SiPMs dominate for compact + magnetic-immune + NIR. APDs fill the gap for moderate-gain, high-QE visible/NIR detection.
| Choose... | When you need... |
|---|---|
| PMT | Large area, lowest dark noise, UV/VUV, sub-ns timing |
| SiPM | Compact size, B-field immunity, NIR, low voltage |
| APD | Highest QE (visible–NIR), moderate gain, fast response |
10.Selection Workflow
Selection priority: spectral range → cathode material → window → active area → dynode type → gain → noise budget → housing/module → environment → scintillator match (if applicable).
For a first PMT purchase in a new lab application, a bialkali head-on tube in a current-output module with built-in HV supply is the most versatile starting point. Add cooling if dark current matters.
Continue Learning
The Comprehensive Guide includes 7 worked examples, 5 SVG diagrams, 3 data tables, and 10 references.