Thermoluminescence of undoped Na2Gd2B2O7: An energetically heterogeneous trap manifold and evidence for non-monotonic post-irradiation storage behaviour.

The thermoluminescence (TL) properties and trapping behavior of undoped Na2Gd2B2O7 were systematically investigated. X-ray diffraction and Rietveld refinement confirmed the monoclinic phase, while optimization of detection and thermal-treatment conditions enabled selective examination of the dominant high-temperature TL band near 350 °C. The integrated TL response was approximately linear over 1.4-8.6 Gy and became progressively sublinear up to 1001 Gy; the composite maximum shifted from approximately 353 to 332 °C. Ten successive irradiation-preheat-readout cycles on a single aliquot demonstrated good reusability, with coefficients of variation of 0.80% for peak intensity, 1.03% for integrated area (230-400 °C), and 0.17% for peak-maximum temperature. Partial-heating Tm-Tstop measurements resolved thermally distinct regions near 140 and 350 °C and revealed unresolved deeper-side structure. Initial-rise analysis using a 5-15% rising-flank interval showed that low post-switch slopes remain overlap-dominated; from Tstop ≈ 280 °C, the apparent activation energy increased from ∼1.38 eV to predominantly 1.77-1.90 eV over 310-385 °C, supporting an energetically heterogeneous high-temperature manifold. Whole-curve screening retained a five-component GOK model as a conservative shape representation, while the preheated band was described separately by four effective GOK components; neither set is interpreted as a unique trap count, and several fitted kinetic parameters, including the effective pre-exponential parameter of the dominant full-curve contribution, are not independently interpretable as microscopic constants. Post-irradiation storage measurements, based on one measurement per storage time, showed a pronounced non-monotonic response: the integrated 230-430 °C signal reached ∼444% and ∼728% of its prompt value after 3 and 6 h, respectively, before returning close to the prompt level at 12 h and then increasing modestly again to ∼111% at 1 d and ∼124% at 10 d. A stricter 280-400 °C control integration still yielded a 7.15-fold enhancement at 6 h and accounted for 89.4% of the net 0-6 h increase, showing that the anomaly is not confined to low-temperature signal or terminal >400 °C background. The enhancement is therefore a direct experimental observation, whereas its microscopic origin and component-wise partition remain unresolved. Independent replicate storage measurements and wavelength-resolved TL spectroscopy are required to establish repeatability and identify the emitting center.