Milk powder analysis by near-infrared spectroscopy

Significance of the Topic

Quality control of milk powder is essential to ensure compliance with regulatory standards, maintain nutritional integrity, and extend shelf life, especially for sensitive applications like infant formula.

Study Objectives and Overview

This application note aimed to develop and validate near-infrared spectroscopy (NIRS) models to quantify moisture, protein, fat, and lactose in milk powder. A diverse set of over 600 samples was analyzed to capture global variability and support real-time quality monitoring.

Methodology

• Sample Analysis: Milk powder samples measured in diffuse reflectance mode with rotating cups to average heterogeneity.
• Spectrometer: OMNIS NIR Analyzer; spectra collected without sample preparation in less than 10 seconds.
• Reference Methods: AOAC 927.05 for moisture, 939.02 for protein, 932.06 for fat; phenol–sulfuric acid assay for lactose.
• Model Development: Calibration and external validation sets; chemometric regression models relating NIR spectra to reference data.

Instrumentation Used

• OMNIS NIR Analyzer Liquid/Solid (diffuse reflectance)
• OMNIS sample cups and large holder for solids
• OMNIS Software for model creation and prediction

Key Results and Discussion

• High correlation between predicted and reference values across all parameters (R² values typically >0.95).
• Low prediction errors demonstrated model robustness for moisture, protein, fat, and lactose.
• External validation confirmed predictive performance with minimal bias despite sample diversity.

Benefits and Practical Applications

• Rapid, reagent-free analysis enables inline or at-line monitoring on production lines.
• Minimal sample preparation accelerates throughput and reduces waste.
• Integration into quality workflows supports immediate corrective actions and consistent product integrity.

Future Trends and Potential Applications

Near-infrared spectroscopy is expected to advance with improved chemometric algorithms, enhanced detector technologies, and wider integration into automated production systems. Potential applications include real-time monitoring of other dairy ingredients, multicomponent process control, and deployment in remote or resource-limited environments.

Conclusion

The study demonstrates that NIRS, using the OMNIS NIR Analyzer and validated chemometric models, provides a reliable, fast, and non-destructive method for quantifying critical quality parameters in milk powder. Its implementation in production environments can streamline quality assurance and optimize process control.

Reference

• AOAC Official Method 927.05: Moisture in Food
• AOAC Official Method 939.02: Protein (Kjeldahl method)
• AOAC Official Method 932.06: Fat (Soxhlet method)
• Phenol–sulfuric acid method for lactose determination