Multiway Calibration Strategies in Laser-Induced Breakdown Spectroscopy: A Proposal

Area of Science:

• Analytical Chemistry
• Spectroscopy
• Chemometrics

Background:

• Investigated novel multiway algorithms for laser-induced breakdown spectroscopy (LIBS) data modeling.
• Utilized parallel factor analysis (PARAFAC) and unfolded-partial least-squares with residual bilinearization (U-PLS/RBL) for multidimensional data analysis.
• Compared multiway approaches with univariate calibration for elemental determination.

Discussion:

• Evaluated LIBS data modeling for printed circuit board (PCB) waste (Al, Cu, Fe) using depth as a dimension.
• Assessed LIBS data modeling for human mineral supplements (Ca, K, Mg) using time-delay as a dimension.
• Examined the impact of calibration data prepared from salts on model performance.

Key Insights:

• PARAFAC demonstrated satisfactory results for PCB waste analysis, offering low prediction errors and good accuracy.
• U-PLS/RBL exhibited superior performance in the quantitative analysis of mineral supplement samples.
• Multiway algorithms provide effective solutions for complex LIBS data, outperforming univariate methods in specific applications.

Outlook:

• Further research into advanced multiway algorithms can enhance LIBS capabilities for diverse sample matrices.
• Optimization of data acquisition strategies (e.g., depth profiling, time-resolved measurements) can improve analytical accuracy.
• Expanding the application of these chemometric methods to other waste streams and biological samples is recommended.