Development of a novel NIR spectroscopy-based chemometric model for sourdough bread authentication

Area of Science:

• Food Science
• Analytical Chemistry
• Spectroscopy

Background:

• Growing global demand for sourdough necessitates reliable authentication methods.
• Traditional biochemical analyses for bread authentication are time-consuming and labor-intensive.
• Near-infrared (NIR) spectroscopy presents a rapid, non-destructive alternative for food analysis.

Purpose of the Study:

• To develop and validate a NIR spectroscopy-based method for distinguishing sourdough bread from bread leavened with baker's yeast.
• To assess the efficacy of multivariate analysis techniques for spectral data interpretation.
• To provide a rapid and accurate authentication tool for the food industry.

Main Methods:

• Production of various bread types with different leavening agents (baker's yeast, sourdoughs) and fermentation parameters.
• Acquisition of near-infrared (NIR) spectra from bread samples.
• Application of spectral preprocessing and multivariate analysis, including Partial Least Squares Discriminant Analysis (PLS-DA), for classification.

Main Results:

• A PLS-DA model with eight latent variables achieved 100% sensitivity and 89% specificity in classifying bread types.
• Key spectral regions linked to sourdough metabolites were identified as crucial for differentiation.
• Permutation testing confirmed the model's robustness and reliability.

Conclusions:

• NIR spectroscopy is suitable for the rapid, non-invasive authentication of sourdough bread.
• The developed method addresses the need for accurate food traceability and transparent labeling.
• This approach offers a practical solution for identifying sourdough products in the market.