Fingerprinting black tea: When spectroscopy meets machine learning a novel workflow for geographical origin...

Area of Science:

• Food Science
• Analytical Chemistry
• Spectroscopy

Background:

• Food fraud, including tea authenticity issues, poses significant risks to global supply chains and consumer trust.
• Tea is a globally consumed beverage, making its authenticity and origin crucial for industry and consumers.
• Current methods for verifying tea origin lack the efficiency and speed needed for real-time supply chain monitoring.

Purpose of the Study:

• To develop an innovative workflow for discriminating black tea geographical indications (GI).
• To establish a rapid, low-cost, and user-friendly solution for authenticating black tea origin.
• To improve transparency and combat food fraud within the black tea supply chain.

Main Methods:

• Non-targeted spectroscopic fingerprinting using Fourier Transform Infrared (FTIR) and Near Infrared (NIR) spectroscopy.
• Analysis of 360 black tea samples from nine distinct GI regions worldwide.
• Application of machine learning algorithms, specifically k-nearest neighbours and support vector machine models, for data analysis.

Main Results:

• Achieved 100% accuracy in identifying black tea geographical origin using FTIR spectroscopy combined with machine learning.
• Demonstrated the effectiveness of spectroscopic fingerprinting for GI discrimination.
• Validated the potential for real-time, on-site determination of black tea origin.

Conclusions:

• The developed workflow offers a highly efficient and accurate solution for black tea geographical origin determination.
• This method provides a significant advancement in combating food fraud and ensuring supply chain integrity.
• The user-friendly and low-cost nature of this technique makes it suitable for widespread adoption in the tea industry.