NIR-Based Adulteration Screening of Rubus chingii Hu: A Two-Dimensional Correlation Spectroscopy-Guided Chemometric...

Area of Science:

• Food Science
• Analytical Chemistry
• Spectroscopy

Background:

• Economically motivated adulteration (EMA) of high-value geographical indication (GI) foods threatens consumer trust and safety.
• Rubus chingii Hu (RcH), a GI-designated fruit, is vulnerable to adulteration with cheaper alternatives.

Purpose of the Study:

• Develop a rapid, non-destructive method for identifying adulteration and predicting the ratio of GI RcH.
• Utilize near-infrared (NIR) spectroscopy and chemometrics for robust food authenticity analysis.

Main Methods:

• Near-infrared (NIR) spectroscopy combined with chemometric models (PLS-DA, SVM-DA, RF-DA, BPNN-DA) for adulteration identification.
• Two-dimensional correlation spectroscopy (2D-COS) to identify sensitive spectral regions (4200-6000 cm⁻¹).
• Wavenumber selection techniques (CARS, IWOA) and various preprocessing strategies (SNV, 1st D) were evaluated.

Main Results:

• The SNV + Smo-BPNN-DA model achieved 100% test accuracy for adulteration identification.
• The optimal regression model, SNV-BPNN-R, predicted adulteration ratios with R² = 96.96%.
• Wavenumber selection improved some regression models but showed a model-dependent effect.

Conclusions:

• A practical and interpretable NIR spectroscopy strategy was developed for rapid EMA screening of powder-based GI foods.
• The method supports on-site or online food authenticity control, enhancing food safety.
• Findings offer a transferable approach for detecting adulteration in valuable food ingredients.