Distinguishing Ancient Mineral Pigments
A recent study demonstrates how Laser-Induced Breakdown Spectroscopy (LIBS) combined with machine learning can help distinguish two ancient mineral pigments – orpiment and realgar. These two pigments become extremely difficult to differentiate after aging because of their very similar appearance and identical elemental composition.
The LIBS measurements were performed using an ARYELLE series Echelle spectrometer from LTB Lasertechnik Berlin, covering the spectral range from 200 to 800 nm with ICCD detection. Based on 12,000 LIBS spectra, the researchers trained both CNN and LightGBM models, achieving remarkably robust classification results even across varying laser energies and simulated mural compositions.
What is especially fascinating is the practical validation: the method was applied to a real yellow mural fragment from Cave 196 of the Mogao Grottoes. Both machine learning models identified the pigment as orpiment, matching previous analytical findings.
It is exciting to see high-resolution LIBS contributing not only to materials analysis and industrial applications, but also to the preservation and scientific understanding of cultural heritage.
About LTB Lasertechnik Berlin
LTB Lasertechnik Berlin develops and produces laser processes, laser systems, and complete automation systems. These machines have proven themselves in a wide variety of locations: from clean rooms at pharmaceutical companies and research laboratories to production halls in the metalworking industry.
These products include:
For more information or a quotation for LTB products, contact IL Photonics.