magnetic proton recoil is a neutron spectrometry technique that converts neutrons into recoil protons in a thin plastic foil and momentum-analyses those protons with a magnet, turning the energy spectrum of a neutral particle into a position measurement.
Neutrons carry no charge and cannot be steered, so the technique first converts them into something that can be. A collimated neutron beam strikes a thin polyethylene foil, where a small fraction scatter elastically off the hydrogen nuclei in it. An aperture selects only the protons knocked forward along the beam — the direction in which the scattering cross-section is largest and in which the proton inherits nearly the neutron's full energy. Those protons enter a magnetic system that disperses them by momentum and focuses them onto a curved focal plane, where an array of scintillators called the hodoscope records where each one landed. Position on that plane is proton momentum, and proton momentum is neutron energy.
What separates the method from time-of-flight spectrometry, its main rival, is calibration. Elastic neutron-proton scattering is among the best-characterized cross-sections in nuclear physics and a magnet's dispersion follows from its geometry, so the instrument's response function can be constructed from first principles rather than fitted against a source. That makes it an absolute measurement of the 📝neutron flux spectrum, and so of 📝fusion power, that leans on no external standard — bought at the cost of a large magnet, a long uninterrupted sightline to the 📝plasma, and heavy 📝shielding. Developed at Uppsala University and installed on 📝JET in 1996, it measured the neutron spectra of the 1997 DTE1 📝deuterium-📝tritium campaign; the upgraded MPRu instrument followed in 2005 and has run in the campaigns since.
📝SPARC's version, the core spectrometer of its planned 📝neutron diagnostics set, is being designed by the 📝MIT Plasma Science and Fusion Center with 📝Commonwealth Fusion Systems and two Milan groups. Two 2024 papers set out the pieces: an ion optical design that reaches a modeled energy resolution better than one percent once nonlinear image aberrations are corrected, and a hodoscope design settling on EJ276D scintillator rods 0.7 cm wide and 13 cm long after bench tests with radioactive sources weighed silicon photomultipliers against photomultiplier tubes. Those bench tests are genuine measurements — of detectors, not of a plasma. No SPARC neutron spectrum exists.