Quantification of underwater LIBS at varying ambient pressures towards deep-sea applications
Abstract
Advances in laser-induced breakdown spectroscopy have enabled rapid and simple analytical applications in the deep sea. However, high-pressure conditions pose an inevitable challenge: they significantly alter LIBS spectral features, with spectra of identical samples deviating substantially across varying pressures. This variability complicates subsequent spectral data analysis, especially quantitative analysis. To address this challenge, we develop a correction method rooted in functional data analysis (FDA), by making functional interpretations of the observed spectral variation patterns under varying ambient pressures. Then, the measured spectra at any pressure (0.1–50 MPa) can be easily converted into their mappings at standard pressure (0.1 MPa), which means aligning real-world measurements with the standardized pressure requirements of the established quantitative model. To assess the quantitative performance of underwater high-pressure LIBS combined with the proposed correction method, calibration curves of Mn, Sr and Li in aqueous solutions were then built through both univariate and multivariate calibration approaches. With the proposed method, the average RSDs of the predicted Mn, Sr and Li concentrations via the partial least squares regression (PLSR) algorithm for validation samples were reduced from 16.18% to 4.4%, and the AREs were also reduced from 29.47% to 5.91%. By bridging pressure-induced spectral discrepancies, this FDA-based correction method establishes a robust framework for reliable quantitative analysis in deep-sea LIBS applications, overcoming a critical bottleneck in underwater elemental detection.