Fusion measurement guide
Fusion Diagnostics Methods Compared
On the nuclear side of a fusion experiment, diagnostics answer three questions: how many neutrons were produced, what charged particles came out, and what is in the fuel and exhaust gas. No single instrument answers all three. This guide compares the main methods for each: what it measures, where it is strong, where it fails, and when to use it.
It covers activation foils, proportional counters, scintillators, bubble detectors, CR-39, silicon detectors, magnetic spectrometers, residual gas analyzers, FT-ICR and magnetic-sector mass spectrometry, laser Raman spectroscopy, gas chromatography, ionization chambers, liquid scintillation counting and calorimetry.
By Noah D'Amico, PhD, BlankSlate Innovation · Last reviewed 18 November 2026
Starting point
What Does Each Fusion Reaction Give You to Measure?
The reaction sets the diagnostic. Neutrons leave the experiment and can be counted outside it. Charged products stop within microns to millimeters of material, so their detectors sit inside the chamber.
Deuterium-deuterium fusion, neutron branch
Deuterium-deuterium fusion, proton branch
Deuterium-tritium fusion
Deuterium-helium-3 fusion
+ protonn neutronEnergies are the kinetic energy each product carries away.
Group 1
How Are Fusion Neutrons Measured?
Neutron yield is the most direct evidence that fusion happened and how much. This is because other partial evidence, such as heat or accelerated ions, can be caused by chemical reactions or EM fields. On the other hand, a background neutron signal is easy to establish, and the creation of free neutrons necessitates nuclear reactions. The methods of neutron detection differ mainly in whether they give an integrated total, a rate or a time history, and in how well they survive the heat, X-ray flash and electromagnetic pulse of a pulsed experiment.
| Method | What it measures | Strengths | Limits | Typical use |
|---|---|---|---|---|
| Activation foils (indium, copper, zirconium) | Time-integrated neutron yield, from the gamma activity induced in the foil | Absolute yield from a simple setup. Immune to X-rays, gamma flash and EMP. Threshold reactions pick out DD or DT neutrons | No timing. The foil must be counted within a half-life or two, which ranges from minutes to days. Needs enough yield to activate | Shot yield on pulsed machines; cross-calibrating other detectors |
| Moderated proportional counters and fission chambers (He-3, BF3, U-235) | Neutron rate in real time | High efficiency and wide dynamic range. Good gamma rejection. Standard on steady and long-pulse sources | Moderation erases energy information. They saturate or pile up in short intense bursts. Helium-3 is scarce and costly | Yield-rate monitoring on tokamaks, neutron generators and long-pulse devices |
| Scintillators with time of flight | Neutron arrival time and energy spectrum | Nanosecond timing. The spectrum gives ion temperature and separates DD from DT neutrons. Some scintillators allow pulse-shape discrimination | They also respond to X-rays and gammas, so they need shielding and gating. Electronics are exposed to EMP. Need calibration | Burn history and ion temperature on pulsed and laser experiments |
| Bubble detectors | Time-integrated neutron dose or fluence, as visible bubbles | Passive, and read by eye or camera within minutes. Insensitive to gammas | Temperature-sensitive response. Limited dynamic range. Coarse energy information | Quick checks and surveys around small sources |
| CR-39 track detectors | Time-integrated fast-neutron fluence, through recoil-proton tracks | Passive, small and cheap, so many positions can be covered. Immune to X-rays, gammas and EMP. Permanent record | Low efficiency, roughly 1 in 10,000 to 1 in 1,000,000 neutrons. No timing. Etching and readout come after the shot | Yield and fluence mapping in harsh pulsed environments |
Group 2
How Are Charged Fusion Products Measured?
Protons, alphas and tritons carry the reaction's energy signature and, unlike neutrons, are detected with close to 100% efficiency by anything they hit. The difficulty is getting a detector close enough without it being blinded.
| Method | What it measures | Strengths | Limits | Typical use |
|---|---|---|---|---|
| CR-39 track detectors | Count, position and, with calibration, energy of protons, alphas and tritons | Every track is one particle. Insensitive to X-rays, gammas and EMP. Works inside the chamber with no cables | Time-integrated. Tracks overlap at high fluence. Needs etching, and energy needs calibration or filters | Charged-particle yield and spectra on pulsed, laser and beam-target experiments |
| Silicon charged-particle detectors (surface-barrier, PIPS) | Energy of each particle in real time | Good energy resolution, tens of keV. Immediate electronic readout | Sensitive to light, X-rays and EMP. Radiation damage, and pile-up in intense bursts. Need bias, feedthroughs and electronics | Beam-target work and other steady, low-rate sources |
| Magnetic and Thomson-parabola spectrometers | Spectra resolved by momentum or charge-to-mass ratio | Separate particle species and give the best energy resolution | Bulky, with a small solid angle and careful alignment. Still need a detector at the focal plane, often CR-39 or image plates | Precision spectra on large laser and pulsed-power facilities |
Group 3
How Are Fusion Fuel, Exhaust Gas and Tritium Measured?
The gas side has its own problem: the species that matter share the same nominal mass. Helium-4, D2 and HT all sit at mass 4. Helium-3, HD and tritium all sit at mass 3. Methods are separated by whether they can tell these apart, and by whether they work inline.
| Method | What it measures | Strengths | Limits | Typical use |
|---|---|---|---|---|
| Quadrupole residual gas analyzer (RGA) | Partial pressures by nominal mass, in real time | Inexpensive, fast and already fitted to most vacuum systems | Unit mass resolution: cannot separate He-4 from D2 at mass 4, or He-3 from HD at mass 3. Workarounds such as threshold ionization cost sensitivity | Vacuum health, leak checking, coarse gas composition |
| FT-ICR mass spectrometry | Exact-mass spectrum of the gas, resolving species at the same nominal mass | Resolving power above 10,000 at mass 3 and 4 separates He-3 from HD and He-4 from D2 directly. One spectrum covers light and heavy species | Samples gas from a line or volume, not the plasma itself. Needs a magnet and ultra-high vacuum. Costs more than an RGA | Fuel and exhaust isotope accounting, helium ash, D-He3 work |
| Magnetic-sector mass spectrometry | High-precision isotope ratios | The reference method for isotope-ratio precision | Large, lab-bound and expensive, usually with gas purification ahead of it | Off-line reference analysis of gas samples |
| Laser Raman spectroscopy | Concentrations of the six hydrogen isotopologues: H2, HD, D2, HT, DT, T2 | Non-contact and inline, so the gas is not consumed. Distinguishes all six isotopologues at once | Blind to helium, which has no Raman signal. Sensitivity falls at low pressure. Optical windows must tolerate tritium | Composition monitoring in tritium loops and fuel-cycle plants |
| Gas chromatography | Separated hydrogen isotopologues and helium | Quantitative, and long established in tritium plants | Slow, minutes per sample. Needs cryogenic columns. Batch sampling | Process control and accountancy samples |
| Ionization chambers | Total tritium activity in a gas stream, in real time | Simple and robust, with a wide range | No isotopic or chemical detail. Surface contamination causes memory effects. Response depends on the carrier gas | Process and room-air tritium monitoring |
| Liquid scintillation counting | Tritium activity in liquid samples | The most sensitive routine tritium measurement | Off-line. Gas and surface samples must first be converted or washed into liquid. Quench corrections and mixed waste | Environmental, wipe and bubbler samples; low-level accountancy |
| Calorimetry | Tritium inventory, from decay heat of about 0.324 W per gram of tritium | Non-destructive and independent of chemical form. Accurate for gram-scale inventories in storage beds and containers | Slow: hours to days to reach thermal equilibrium. Insensitive to small amounts. Says nothing about composition | Inventory of stored tritium |
The mass differences are small. D2 and helium-4 differ by 0.0256 mass units, which needs a resolving power of about 160 to separate. HD and helium-3 differ by 0.0059, which needs about 510. HT and D2 differ by 0.0043, which needs about 930. In practice more is needed when one species is far more abundant than its neighbor.
Decision table
Which Method for Which Question?
| Question | First choice | Add when needed |
|---|---|---|
| Did fusion happen, and how much? | Activation foils or CR-39 for a time-integrated yield | A moderated counter for a real-time rate |
| When did it happen within the shot? | A scintillator with time of flight | Shielding and gating against the X-ray flash |
| What energies did the products have? | CR-39 with filters, or a silicon detector on steady sources | A magnetic spectrometer for precision spectra |
| Will electronics survive near the experiment? | Passive detectors: CR-39, activation foils, bubble detectors | Shielded scintillators at a distance |
| Is that mass-4 signal helium or deuterium? | FT-ICR mass spectrometry | A threshold-ionization RGA as a lower-cost partial answer |
| What is the isotopologue mix in my fuel loop? | Laser Raman inline, or gas chromatography | FT-ICR when helium must be measured too |
| How much tritium do I hold? | Calorimetry for storage; ionization chambers for process gas | Liquid scintillation for low-level and environmental samples |
Most experiments end up with a pair on the nuclear side: a passive, time-integrated detector that cannot be fooled by the electromagnetic environment, and an electronic detector that supplies timing. On the gas side, an RGA watches the vacuum while a higher-resolution method does the isotope accounting.
Available here
Two of These Methods Are Made in Lubbock, Texas
BlankSlate Innovation manufactures CR-39 track detectors for fusion diagnostics ($2/cm2, from 3 business days) and is building a tabletop FT-ICR mass spectrometer for fusion fuel and exhaust. We also supply metal deuterides for targets.
Because we sell two of the methods on this page, we have tried to state their limits as plainly as everyone else's.
CR-39 product overview · FT-ICR vs RGA for fusion · Neutron detection methods compared · Talk to our team
Questions people ask
Fusion Diagnostics FAQ
What are the main methods of fusion diagnostics?
On the nuclear side, fusion diagnostics fall into three groups. Neutron diagnostics include activation foils, moderated proportional counters, scintillators with time of flight, bubble detectors and CR-39. Charged-particle diagnostics include CR-39, silicon detectors and magnetic spectrometers. Fuel and exhaust diagnostics include residual gas analyzers, FT-ICR and magnetic-sector mass spectrometry, laser Raman spectroscopy, gas chromatography, ionization chambers, liquid scintillation counting and calorimetry.
How are fusion neutrons measured?
Time-integrated neutron yield is usually measured with activation foils or CR-39 track detectors, which are passive and immune to X-rays and electromagnetic pulse. Real-time rates come from moderated helium-3, boron trifluoride or fission counters. Timing and the neutron energy spectrum come from fast scintillators using time of flight, which also gives the ion temperature.
Can a residual gas analyzer tell helium-4 from deuterium?
Not directly. Helium-4 and D2 both appear at mass 4, and a standard quadrupole residual gas analyzer has roughly unit mass resolution, so the two peaks overlap. Separating them by mass takes a resolving power of a few hundred or more. Threshold ionization is a partial workaround. High-resolution instruments such as FT-ICR or magnetic-sector mass spectrometers resolve the two species directly.
How is tritium inventory measured?
It depends on the form. Stored tritium is measured by calorimetry, using its decay heat of about 0.324 watts per gram. Tritium in process gas is monitored with ionization chambers, and its isotopic composition with laser Raman spectroscopy, gas chromatography or mass spectrometry. Low-level and environmental samples are measured by liquid scintillation counting.
Why is CR-39 used in fusion experiments?
CR-39 is a passive plastic track detector that records individual protons, alpha particles and tritons, and fast neutrons through recoil protons. It needs no power or cables and does not respond to X-rays, gamma rays or electromagnetic pulse, so it works where electronic detectors fail. Its limits are that it is time-integrated and must be etched and read after the shot.
Sources
References and Further Reading
- I. H. Hutchinson, Principles of Plasma Diagnostics, 2nd ed., Cambridge University Press, 2002.
- G. F. Knoll, Radiation Detection and Measurement, 4th ed., Wiley, 2010.
- F. H. Séguin et al., "Spectrometry of charged particles from inertial-confinement-fusion plasmas," Review of Scientific Instruments 74, 975 (2003). doi.org/10.1063/1.1518141
- J. A. Frenje et al., "Absolute measurements of neutron yields from DD and DT implosions at the OMEGA laser facility using CR-39 track detectors," Review of Scientific Instruments 73, 2597 (2002). doi.org/10.1063/1.1487889
- M. Schlösser et al., "Accurate calibration of the laser Raman system for the Karlsruhe Tritium Neutrino Experiment," Journal of Molecular Structure 1044, 61 (2013). doi.org/10.1016/j.molstruc.2012.11.022
- N. D'Amico et al., "AI-Based Large-Area Nuclear Particle Track Analysis System," Nuclear Engineering and Technology 57, 103738 (2025). doi.org/10.1016/j.net.2025.103738