Nanoparticles enhanced laser-induced breakdown spectroscopy for characterisation and discrimination of gemstones

• Main Authors: Al-Juboori, Haider, Haider, Zhaib, Hang, Wei, Roslan, Sufi
• Format: Conference or Workshop Item
• Language: English
• Published: 2024
• Subjects: QD Chemistry
• Online Access: http://eprints.uthm.edu.my/12006/1/Nanoparticles%20enhanced%20laser-induced%20breakdown%20spectroscopy%20for%20characterisation%20and%20discrimination%20of%20gemstones.pdf; http://eprints.uthm.edu.my/12006/; https://doi.org/10.1051/epjconf/202430500015

Laser-induced breakdown spectroscopy (LIBS) is an excellent technique for rapid on-site investigations that attracts interest from diverse research areas. Gemmology is no exception. The application of LIBS for gemstone characterisation is limited due to ineffective ablation and crack formation, even more so with cost-effective, non-gated LIBS systems. Nanoparticle-enhanced LIBS (NELIBS) has proven to improve the effectiveness of LIBS by minimizing sample damage and enhancing the spectral features. Therefore, this study is dedicated to exploring the advantages of NELIBS, for characterising Sapphire and Opal and discrimination based on spectral differences. Our objective is to explore enhancing spectral features and performing discriminant analysis using the PLS-DA algorithm. Nanoparticles (NPs) were deposited in two layers by sequentially drying two drops (2µL) of a colloidal solution of 20nm gold nanoparticles (AuNPs) on the sample surface. Targeted areas were shot with 3 pulses of Nd:YAG laser (~350mJ, 10ns, 1064nm, 1Hz) for collecting NELIBS spectra with the OceanOptics HR4000 spectrometer. The procedure was repeated without NPs for comparative analysis with conventional LIBS (CLIBS). Results have shown a significant enhancement in spectral features, i.e., the emergence of several new spectral lines of major gemstone elements in the UV-Vis regions of the NELIBS spectra, while the CLIBS spectra were devoid of any meaningful spectral information. The PLS-DA model was trained and validated using a 4- fold cross-validation approach. The model discriminated gemstones with 99.48% accuracy at the 4th fold and exhibited a mean cross-validation accuracy of 98.97%. This preliminary investigation demonstrates the effectiveness of NELIBS for characterization and the potential for onsite identification of gemstones.