A layered lead-halide perovskite scintillator distinguished fast-neutron events from a gamma-ray background in a laboratory test, researchers at MIT and Seoul National University report in a Sept. 23 arXiv preprint. A 5-millimeter sample separated signals from a deuterium-deuterium neutron source and accompanying gamma rays. The result suggests a material for future fusion-neutron diagnostics, but the paper does not report a reactor test.
The sample separated neutron and gamma-ray events
Fusion experiments need neutron measurements to describe plasma conditions, but gamma rays can also trigger a detector. The team tested a cylindrical scintillator with a 490-square-millimeter area and 5-millimeter thickness in a mixed radiation field. Its deuterium-deuterium generator produced 2.45-MeV neutrons; interactions in surrounding materials created the accompanying gamma-ray background.
The researchers used pulse-shape discrimination, comparing the light collected over an early 22-nanosecond gate with a longer 160-nanosecond gate. At the acquisition-specific integrated-charge threshold of 2,515, the neutron and gamma-ray populations reached a separation figure of merit of 3.51. That measure compares the distance between the two pulse populations with their combined spread. Raising the threshold improved separation by excluding dimmer events, so the score does not show how many neutrons the sample detects overall.
In a separate test with cesium-137 gamma rays, the sample showed a 662-kiloelectronvolt photopeak with 5.4% energy resolution. This describes the spread of measured gamma-ray energies, not neutron-detection efficiency.
Lower-bandgap domains make emitted light less likely to be reabsorbed in thicker samples
Making a scintillator thicker gives incoming neutrons more material to interact with, but it also gives emitted light a longer path through the crystal, where it can be reabsorbed. The team could vary internal composition separately from sample dimensions, then build the material into samples millimeters thick and centimeters wide.
The base material, PEA2PbBr4, is a layered lead-bromide perovskite. The team added 3 mol% formamidinium lead bromide and 5 mol% extra lead bromide, producing a small population of domains with narrower bandgaps inside a wider-bandgap matrix. Excitation funnels into the lower-energy domains, which emit light that the surrounding matrix is less likely to absorb. The authors call this a multi-quantum-well structure; its measured photoluminescence lifetime was 6.6 nanoseconds under 405-nanometer optical excitation.
The preprint does not establish detector efficiency
The radiation measurements used 5-millimeter samples. A separate 25-millimeter-wide, 3.5-millimeter-thick disk demonstrates a form the researchers can fabricate, but it was not the geometry used for the reported neutron test.
A calculation for a 5-millimeter sample estimated that 10.1% of 2-MeV neutrons would interact with it and that the average energy deposited would be 3.6% of the incident neutron energy. Those are model results at 2 MeV, not measured efficiency; the direct discrimination test used 2.45-MeV neutrons. The paper reports a material-level laboratory result rather than a packaged sensor tested around a fusion device. Its disclosure also identifies Shaun Tan, Zachary Benesh and Moungi Bawendi as inventors on a provisional patent related to the work.
Sources
Tan et al., “Fast-Neutron Scintillation with Multi-Quantum-Well 2D Perovskites” (arXiv:2609.28468v1, submitted Sept. 23, 2026) · Full paper (PDF)
Illustrative lead image: Savannah River Site, bismuth germanate scintillator crystals, a U.S. Department of Energy work in the public domain, via Wikimedia Commons. The crystals are not from this perovskite study; the image is unchanged.
.jpg){kind=link}