[Improving component analysis ability of the complex mixed solutions by multi-dimensional diffuse transmittance...

Area of Science:

• Analytical Chemistry
• Spectroscopy
• Biomedical Optics

Context:

• Near-infrared (NIR) spectroscopy is crucial for analyzing complex biological and chemical mixtures.
• Traditional methods struggle with highly scattering or absorbing samples.
• Developing advanced spectroscopic techniques is essential for accurate component analysis.

Purpose:

• To investigate the efficacy of multi-point diffuse transmittance spectroscopy for analyzing complex solutions.
• To assess the impact of multiple spectral data points on component concentration prediction accuracy.
• To determine the relationship between spectral signal-to-noise ratio and sample properties.

Summary:

• A novel device combining a xenon light source, scanning stage, and spectrometer was used to collect diffuse transmittance spectra at 20 points (0-5 mm) from complex mixtures (Intralipid-20%, India ink, C6H12O6).
• Partial least squares regression was employed for modeling and prediction of component concentrations.
• Results indicated improved modeling and prediction accuracy for Intralipid and India ink with increased data points, while C6H12O6 accuracy did not improve, suggesting enhanced signal-to-noise for strong absorbers.

Impact:

• Multi-point spectral collection enhances signal-to-noise ratio for strongly scattering and absorbing substances.
• System accuracy improvements are needed for precise analysis of weakly absorbing substances.
• This method offers potential for improved quantitative analysis in complex fluid systems.