High-Throughput Calibration-Free Laser-Induced Breakdown Spectroscopy

Laser-induced breakdown spectroscopy (LIBS) offers a noninvasive, label-free technique for chemical analysis in challenging environments, including for deep-sea mineral resource evaluation and extra-terrestrial geology. We aim to improve the usefulness of LIBS spectral analysis in these applications. We propose an efficient, systematic procedure that uses calibration-free LIBS (CF-LIBS) to quantitatively estimate chemical compositions. This method combines baseline estimation and denoising using sparsity with the spectrum-adapted expectation-conditional maximization algorithm, enabling nonlinear background subtraction and high-throughput peak fitting. In addition, we introduce a Boltzmann plot regression based on Student's t-distribution that is robust against outliers. These techniques allow the chemical composition of metal and rock samples to be estimated using the CF-LIBS method, demonstrating its potential for use in comprehensive geological surveys in deep-sea and extra-terrestrial environments.