Enhancing beer authentication, quality, and control assessment using non-invasive spectroscopy through bottle and...

Area of Science:

• Food Science and Technology
• Analytical Chemistry
• Spectroscopy

Background:

• Rising economic impact of fraud in alcoholic beverages due to counterfeiting and adulteration.
• Need for non-destructive analytical methods for quality control and authentication in the beverage industry.

Purpose of the Study:

• To develop a method using near-infrared (NIR) spectroscopy and machine learning (ML) for beer authentication and quality assessment through the bottle.
• To predict fermentation type, sensory descriptors, and volatile aromatic compounds in beer.
• To provide a tool for brewers and retailers to ensure quality and prevent fraud.

Main Methods:

• Utilized near-infrared (NIR) spectroscopy (1596-2396 nm) for non-destructive analysis of 25 commercial beer samples.
• Employed quantitative descriptive analysis with trained panelists and gas chromatography-mass spectroscopy (GC-MS) for ground-truth data.
• Developed artificial neural network (ANN) models using NIR absorbance values to predict beer characteristics.

Main Results:

• High accuracy achieved in predicting fermentation type (99%), sensory descriptors (R=0.92), and volatile compounds (R=0.94).
• Models demonstrated robust performance when deployed on new samples (95% for fermentation type, R=0.83 for sensory descriptors).
• The method allows for through-the-bottle analysis, preserving sample integrity.

Conclusions:

• Near-infrared spectroscopy coupled with ML modeling offers a rapid, accurate, and non-destructive approach for beer authentication and quality assessment.
• This technology can help combat fraud, ensure provenance, and monitor quality during transport and storage.
• Potential for expansion to assess additional quality traits like physicochemical parameters and origin through further model training.