Sorting zamak alloys via chemometric analysis of their LIBS spectra

Laser induced breakdown spectroscopy with chemometric methods has been employed for sorting zamak alloy (zinc based alloy) samples. Plasma plumes were created on the sample surface using 50 mJ Nd:YAG laser pulses at its fundamental wavelength (lambda = 1064 nm). Six types of zinc alloy samples with different elemental compositions have been used to perform the LIBS measurements under optimized experimental conditions. On the obtained LIBS data three different chemometric classification models were applied, namely K Nearest Neighbor (KNN), Support Vector Machine (SVM) and Artificial Neural Networks (ANN). The effect of the input database choice and its normalization on the classification model efficiency has been investigated and found to play a crucial role. It was found that KNN, SVM and ANNs permit an appropriate classification of different zinc alloys using LIBS spectra and the best results were obtained from the largest input database. The effect of data normalization has been discussed for each model. The obtained results demonstrate the feasibility of using chemometric methods associated with the LIBS technique as an industrial tool for in situ zamak sorting.