Machine learning coupled laser-induced breakdown self-reversal isotopic spectrometry for determining lithium isotope...

Area of Science:

• Analytical Chemistry
• Spectroscopy
• Isotope Ratio Measurement

Background:

• Isotopic analysis is crucial for various scientific fields.
• Accurate quantification of lithium isotopes (6Li and 7Li) presents analytical challenges.
• Existing methods may lack precision or require complex sample preparation.

Purpose of the Study:

• To implement and evaluate Laser-Induced Breakdown Self-Reversal Isotopic Spectrometry (LIBRIS) for lithium isotope quantification.
• To establish a correlation between spectral self-reversal shift and 6Li atom percent.
• To develop a machine learning model for accurate determination of isotopic concentrations.

Main Methods:

• Utilized LIBRIS to measure the self-reversal shift of the Li 670.8 nm peak.
• Prepared LiOH·H2O samples with varying 6Li atom percentages (3-95%).
• Trained supervised machine learning regression models, including a stacked ensemble, on spectral data.

Main Results:

• Observed a spectral self-reversal peak center wavelength shift of 13.813±1.21 pm across the tested isotopic range.
• The stacked ensemble model achieved a Root Mean Square Error (RMSE) of 5.66 atom %.
• Established a detection limit of 18.8 atom % for 6Li concentration.

Conclusions:

• LIBRIS is a viable technique for non-destructively quantifying lithium isotopic composition.
• Machine learning, particularly stacked ensembles, significantly enhances the accuracy of LIBRIS-based isotopic analysis.
• The developed method offers a sensitive and precise approach for isotopic concentration determination.