[Quickly determination of titanium dioxide content in juice based on Vis/NIR spectroscopy technique]

Area of Science:

• Analytical Chemistry
• Food Science
• Spectroscopy

Background:

• Titanium dioxide is a common additive in food products.
• Accurate detection of titanium dioxide content is crucial for food safety and quality control.
• Existing detection methods can be time-consuming or destructive.

Purpose of the Study:

• To develop and validate a rapid, non-destructive method for quantifying titanium dioxide in juice.
• To compare the performance of chemometric models like Partial Least Squares (PLS) and Principal Component-Artificial Neural Network (PC-ANN).
• To identify optimal spectral preprocessing techniques for accurate analysis.

Main Methods:

• Vis/NIR spectroscopy was employed to collect spectral data from juice samples.
• Chemometric techniques, including PLS and PC-ANN, were utilized for data analysis.
• Various spectral preprocessing methods (SNV, MSC, etc.) were evaluated to optimize model performance.

Main Results:

• Multivariate Scatter Correction (MSC) proved to be the most effective preprocessing method.
• The PLS model achieved a correlation coefficient (R2) of 0.9008 and RMSEP of 0.05.
• The PC-ANN model demonstrated strong performance with R2 of 0.8684 and RMSEP of 0.04.

Conclusions:

• Vis/NIR spectroscopy combined with chemometrics offers a feasible and accurate approach for non-destructive titanium dioxide detection in juice.
• The developed method provides a reliable foundation for future forecasting models of titanium dioxide content.
• This technique supports rapid quality assessment and adulteration detection in the food industry.