An optimized spectral reconstruction method for shift excitation Raman differential spectroscopy.

Area of Science:

• Analytical Chemistry
• Spectroscopy
• Materials Science

Background:

• Raman spectroscopy is powerful but susceptible to noise and strong fluorescence interference.
• Fluorescence background can obscure weak Raman signals, hindering qualitative and quantitative analysis.
• Shifted Excitation Raman Differential Spectroscopy (SERDS) is an effective technique for fluorescence removal.

Purpose of the Study:

• To develop and validate a Tikhonov regularized least squares (TRLS) reconstruction method for SERDS.
• To improve the accuracy and reliability of Raman spectral data by mitigating fluorescence interference.
• To optimize the TRLS method using artificial and real-world datasets.

Main Methods:

• Development of a Tikhonov regularized least squares (TRLS) reconstruction algorithm.
• Verification and optimization using four groups of artificial datasets with varying characteristics.
• Performance evaluation on real Raman spectral datasets using RMSE, R, and RPD metrics.

Main Results:

• The TRLS method effectively mitigates oscillations compared to direct unconstrained least squares (DULS).
• Optimizing the TRLS parameter α reduced the relative standard deviation (RSD) in reconstructed datasets.
• Quantitative analysis showed enhanced prediction accuracy and practicality with the TRLS method.

Conclusions:

• The Tikhonov regularized least squares (TRLS) based reconstruction method significantly improves differential Raman spectra.
• The TRLS method offers enhanced stability and reliability for analyzing complex Raman spectral data.
• This approach is practical for accurate qualitative and quantitative analysis in Raman spectroscopy.