Chiswick School students along with Orbyts Fellow Sergio López Martínez from UCL Medical Physics & Biomedical Engineering embarked on a project to evaluate the capabilities of X-ray Fluorescence (XRF) Spectroscopy for material identification. Through the use of a Metrix Micro-CT machine and an external Amptek X-123 detector, they’ve attempted to identify characteristic X-rays emitted by materials and compared them against known purity samples to determine their identity.

The students demonstrated that XRF is a valuable tool for material identification, especially when combined with Micro-CT to reveal composition. A key finding was the successful identification of silver in a ring sample placed in the Micro-CT machine. While the ring exhibited a lower count than a 100% pure silver reference, this observation highlighted XRF's ability not only to identify materials but also to determine their concentration.

However, the research also unveiled limitations of the technique, particularly when analysing complex objects like a RAM card. The XRF analysis of the RAM card presented no clear material identification, showing significant alignment with a background spectrum obtained with no sample present, suggesting a potential misidentification. Furthermore, other peaks did not clearly align with known elements typically found in RAM cards, and many lower peaks corresponded to rare-earth metals or elements unlikely to be present.

This indicated that while XRF excels at detecting clearly dominant materials, its effectiveness diminishes for identifying elements in intricate devices or those present in low concentrations. The presence of background noise also posed challenges, particularly for identifying lower energy peaks. Despite these challenges, the students developed a Python program to recalibrate spectra and omit background noise, improving data analysis. This study underscores XRF's potential in material identification for high concentrations, while also highlighting areas for further development to enhance its utility for complex or trace-level analyses!