One cheap glass, four kinds of radiation: a scintillator that senses X-rays, gamma rays, neutrons and particles

A single glass that answers to almost every kind of radiation could let hospitals, cargo scanners and particle-physics labs share one low-cost detector material.

Scintillators convert invisible radiation into light, and they sit inside the machines that image the body, inspect baggage, and clock the debris of particle collisions. Crystals such as BGO do this job well, but they must be grown slowly and at high cost. Plastics are cheap yet low in density and quick to degrade. Glass has the right economics — cheap, moldable, and producible in large sheets — but for nearly seven decades it has failed to combine high density with decent light output.

Now a team led by Zhehao Hua, Sen Qian and Haizheng Tao reports a cerium-doped, gadolinium-rich fluoro-oxide glass (GFO:Ce) that does both. The material reaches a density of about 6.0 g/cm3 — high enough to stop gamma rays efficiently — while still emitting useful light. It is melted under a reducing atmosphere and can be cast into plates, blocks or thin screens; the researchers made 5-mm-thick imaging screens, 5-mm cubes for gamma spectroscopy and 4 × 4 × 1 cm blocks for beam tests.

The glass resolves 29.8 line pairs per millimeter in X-ray imaging, among the highest values reported for a glass scintillator, enough to pick out the spring inside a capsule and the copper traces at the edge of a chip. Under a caesium-137 source it delivers a light yield of 2,076 ± 51 photons per MeV with an energy resolution of 22.4 ± 0.5% at 662 keV, and an average decay time of 447.2 ns. Its response stays linear with dose down to a detection limit of 16.58 microgray per second, and the emission survives 20 on-off irradiation cycles. At 493 K it still gives 52.2% of its room-temperature photoluminescence.

The same glass detects neutrons, with distinct response peaks near 80 and 180 keV and a resolution of 42.6 ± 1.7% at the upper peak. In high-energy beams it registers minimum-ionizing particles at 99.2 ± 0.6 photoelectrons for muons and 112.8 ± 2.2 for pions, above the 50-photoelectron benchmark set for hadron calorimeters.

The performance traces back to local structure. Gadolinium-oxygen-fluorine polyhedra tend to cluster within the disordered network and pass energy to cerium centres with 85.5% efficiency, giving a photoluminescence quantum yield of 80.40%.

The limits are equally clear. Energy migration efficiency is only 11–15%, and Auger-type quenching still trims the light yield; X-ray excited luminescence reaches 62.1% of BGO's. The authors present the glass as a platform for industrial and medical radiography and for calorimetry at future particle colliders, with optimization still to come.