Feasibility study of detecting some milk adulterations using a LED-based Vis-SWNIR photoacoustic spectroscopy system

Area of Science:

• Analytical Chemistry
• Food Science
• Spectroscopy

Background:

• Milk adulteration poses a significant threat to public health and the food industry.
• Traditional methods for detecting milk adulterants are often time-consuming and labor-intensive.

Purpose of the Study:

• To evaluate a developed photoacoustic spectroscopy (PAS) system for detecting common adulterants in cow's milk.
• To assess the system's effectiveness across a range of adulterants including formalin, urea, and detergent powder.

Main Methods:

• Utilized a photoacoustic spectroscopy system with visible to short-wave near infrared (Vis-SWNIR) light sources (395-940 nm).
• Applied principal component analysis (PCA) for data visualization and differentiation of adulterated samples.
• Employed artificial neural networks (ANN) for classification of adulteration type and level.

Main Results:

• Principal Component Analysis (PCA) effectively differentiated between various milk adulterations.
• Artificial Neural Networks (ANN) achieved high classification accuracy: 97.6% for adulteration type and nearly 100% for adulteration level.
• The Vis-SWNIR PAS system proved to be a reliable tool for detecting multiple milk adulterants.

Conclusions:

• The Vis-SWNIR photoacoustic spectroscopy system is a potent instrument for detecting cow's milk adulterations.
• Future research should involve diverse milk sources to generalize findings.
• Expanding light sources to long-wave NIR could enable application to other liquid food quality assessments.