Lighting up FLASH and beyond: TLD response across four orders of magnitude in proton dose rate

Thermoluminescent dosimeters (TLDs) are widely used for proton therapy dosimetry and remote audits. Yet their response under pulsed delivery and the ultrahigh dose rates relevant to proton FLASH remains incompletely characterized. In proton beams, apparent dose-rate effects can be masked by changes in linear energy transfer (LET), so we designed measurements to separate true dose-rate dependence from LET-driven changes in TLD sensitivity.

To characterize TLD-100 (LiF:Mg,Ti) detector response to synchrotron-based proton beams across in-pulse dose rates spanning approximately 0.1 to 2000 Gy/s.

TLD-100 powder capsules were irradiated in the entrance region of 87.2 MeV protons using two Hitachi ProBeat synchrotron beamlines: a clinical spot-scanning beamline (0.3 Gy/s in-pulse) and an experimental beamline (approximately 0.1-2000 Gy/s). Three dose levels (3, 9, and 18 Gy) were evaluated across three experimental sessions. Expected doses to TLDs were calculated using benchmarked Geant4 Monte Carlo models. In-pulse dose rates were determined from dose per pulse divided by pulse duration. Pulse duration was measured using an oscilloscope for all dose rates. For dose rates ≤375 Gy/s, dose per pulse was measured using an Advanced Markus ion chamber (P_ion = 1.003, ≥99% saturation). For dose rates exceeding the chamber's range, dose per pulse was calculated using Geant4. TLD response was defined as measured-to-expected dose ratio. Positioning at the entrance region isolated dose-rate effects from LET-dependent variations.

TLD response demonstrated systematic in-pulse dose-rate dependence with three distinct regions. At in-pulse dose rates below 7 Gy/s, TLDs exhibited 1%-7% over-response attributed to efficient charge carrier trapping. Near-unity response (0%-5%) was observed between 7-300 Gy/s, representing balanced trapping and recombination. Above 300 Gy/s, TLDs showed 1%-7% under-response due to enhanced electron-hole recombination. Two-way ANOVA confirmed in-pulse dose rate as the dominant factor (82.9% of variance, p < 0.0001), while dose had minimal effect (2.3% of variance). One-way ANOVA with Šídák's test showed no significant difference among dose levels (p > 0.05), confirming dose independence across 3-18 Gy.

TLD-100 response to synchrotron-based protons is dose-independent across 3-18 Gy but follows a characteristic in-pulse dose-rate profile governed by charge carrier dynamics: over-response between 0.1-7 Gy/s, nearly unity between 7-300 Gy/s, and under-response between 300-2000 Gy/s.