Optimizing the quantitative analysis of solid biomass fuel properties using laser induced breakdown spectroscopy...

Area of Science:

• Analytical Chemistry
• Spectroscopy
• Biomass Science

Background:

• Solid biomass fuel property analysis is crucial due to feedstock variability.
• Laser-Induced Breakdown Spectroscopy (LIBS) offers rapid elemental analysis.
• Existing linear models like Partial Least Squares (PLS) struggle with non-linear LIBS spectral data.

Purpose of the Study:

• To apply Kernel Partial Least Squares (KPLS) for analyzing solid biomass fuel properties.
• To compare KPLS performance against traditional PLS models.
• To investigate the impact of data normalization methods on KPLS accuracy.

Main Methods:

• Utilized LIBS to acquire spectral data from solid biomass fuels.
• Implemented KPLS, a non-linear chemometric method, for quantitative analysis.
• Evaluated KPLS with and without three normalization techniques: C internal standardization, total intensity standardization, and standard normal variate.

Main Results:

• KPLS demonstrated improved quantitative analysis compared to PLS.
• C internal standardization with KPLS yielded the best results for volatile matter and ash content (RMSEP 1.365% and 0.290%).
• KPLS without normalization provided optimal results for gross calorific value and fixed carbon content (RMSEP 0.198 MJ kg⁻¹ and 0.746%).

Conclusions:

• KPLS is a superior method for non-linear LIBS spectral data analysis in biomass fuel characterization.
• The choice of normalization method significantly impacts KPLS model performance depending on the target fuel property.
• This study advances rapid and accurate biomass fuel analysis for efficient utilization.