In the present work, femtosecond Laser Induced Breakdown Spectroscopy, fs-LIBS, is applied for the first time for the analysis of the inorganic content and the identification of the animal origin of milk samples (i.e., cow, goat, and sheep). The fs-LIBS spectra were analyzed by different machine learning algorithms (i.e., multivariate statistical analysis), and their performance was assessed. The obtained classification accuracies of the milk samples based on the animal origin were found to attain values up to 95 %, depending on the algorithm used. For comparison purposes, LIBS experiments were also carried out employing nanosecond laser pulses, ns-LIBS. The obtained results from both the fs- and ns-LIBS experiments were found to be in excellent agreement between them. The present findings, using fs laser excitation, demonstrate and extend the capabilities of LIBS technique, for the identification of the animal origin of milk samples, thus making fs- and/or ns-LIBS technique a powerful and useful tool for milk authentication, quality, and safety control related research.
Read full abstract