FTIR spectroscopy identification of fraud in coffee powder: a study on preprocessing methods

Area of Science:

• Analytical Chemistry
• Food Science
• Spectroscopy

Background:

• Coffee adulteration is a significant economic and quality concern.
• Detecting adulterants requires sensitive and reliable analytical methods.
• Fourier Transform Infrared (FTIR) spectroscopy offers a non-destructive approach for chemical analysis.

Purpose of the Study:

• To develop and validate an FTIR-based method for detecting and quantifying adulteration in ground coffee.
• To evaluate the efficacy of various spectral preprocessing techniques and regression models.
• To identify key spectral markers indicative of common coffee adulterants.

Main Methods:

• Fourier Transform Infrared (FTIR) spectroscopy was employed to analyze coffee samples adulterated with barley, chickpea, and date pit powders at varying concentrations (10-100%).
• Eight spectral preprocessing techniques (e.g., SNV, MSC, Savitzky-Golay) were applied to enhance spectral features and reduce noise.
• Machine learning regression models (Bagging, Random Forest, Gradient Boosting, XGBoost) were trained and optimized for predictive analysis.

Main Results:

• The FTIR-chemometrics pipeline demonstrated high accuracy in detecting adulteration.
• Bagging Regressor combined with SNV or Savitzky-Golay preprocessing achieved the best predictive performance.
• Accurate quantification of adulterants was achieved, with R² values up to 0.91 and low RMSE values (10.85-13.68).

Conclusions:

• FTIR spectroscopy coupled with advanced chemometrics is a powerful tool for identifying and quantifying coffee adulteration.
• The developed method is robust, non-destructive, and capable of detecting adulteration at levels as low as 10%.
• This approach provides a viable solution for real-time food fraud detection in the coffee industry.