From Sampling to Measurement: Ensuring Reliable TL Authentication

Thermoluminescence (TL) is a well-established experimental technique for authenticating ceramic and pottery artifacts. It provides absolute information on the last firing event recorded by mineral components. In culture heritage studies, where stylistic and typological analyses are increasingly challenged by sophisticated forgeries, TL serves as an independent tool. Despite its robustness, the reliability of TL authentication strongly depends on the experimental procedures used during sampling, preparation, measurement, and data analysis. This work presents a systematic investigation aimed at identifying and controlling the experimental parameters that govern TL signal intensity, stability, and reproducibility. The goal is to define a reference protocol for TL authentication. Several artifacts from different chronological periods were analyzed by introducing controlled variations in key procedural steps. These include alternative sampling strategies, such as standard extraction and drilling, exposure to ambient light during sampling under non-ideal conditions, chemical and physical sample preparation selecting different levels of treatment completeness, and TL measurement strategies based on regeneration and added dose protocols. The influence of acquisition parameters, including preheating and data treatment in dose-response curve fitting, was also evaluated. Results show that TL authentication outcomes are highly sensitive to procedural choices. Specific parameters critically affect authenticity classification and compatibility with the presumed age. Quantifying these effects allows the identification of best practices and supports the definition of optimized and standardized protocols. Careful procedural control is essential to achieve reliable, reproducible, and quantitatively robust discrimination between genuinely ancient ceramics and modern replicas. The comparison of different operational procedures shows that the methods used to sample the powder can significantly affect the final results. In particular, this work highlights that the use of the good light with drill procedure results in an underestimation of the equivalent dose (ED) and age by approximately 13% compared to sampling with the hammer procedure.