Detection of fraud in lime juice using pattern recognition techniques and FT-IR spectroscopy

Area of Science:

• Food Science
• Analytical Chemistry
• Spectroscopy

Background:

• Lime juice products are frequently subject to manufacturer fraud to reduce costs.
• Detecting adulteration in food products like lime juice is crucial for consumer safety and fair trade.

Purpose of the Study:

• To investigate and detect fraudulent practices in commercially distributed lime juice in Iran.
• To develop and validate a robust analytical method for lime juice authenticity assessment.

Main Methods:

• Collected 101 lime juice samples (natural and reconstructed) from Iranian markets.
• Utilized Fourier-transform infrared (FT-IR) spectroscopy for sample analysis.
• Applied chemometric algorithms including Variable Importance in Projection (VIP) and Partial Least Square-Discriminant Analysis (PLS-DA) for feature selection, and Counter Propagation Artificial Neural Networks (CPANN) for classification.

Main Results:

• Principal Component Analysis (PCA) showed limited initial sample clustering.
• VIP-PLS-DA identified key discriminatory wavenumbers for classifying lime juice origins.
• CPANN achieved high classification accuracy, reaching 0.96 for training and 0.87 for validation, successfully distinguishing between natural and reconstructed lime juice samples.

Conclusions:

• FT-IR spectroscopy combined with VIP-PLS-DA and CPANN offers a reliable and accurate method for detecting lime juice authenticity.
• This approach can effectively identify adulterated lime juice, safeguarding product integrity and consumer trust.