Determination of olive oil quality parameters and adulteration with NIR spectroscopy

Importance of the Topic

Virgin olive oil is among the most prized vegetable oils and is highly vulnerable to adulteration and quality degradation. Rapid and reliable analytical methods are essential for producers, regulators and quality control laboratories to ensure authenticity, safety and compliance with international standards. Near-infrared spectroscopy (NIRS) offers an efficient alternative to traditional wet-chemical techniques by minimizing chemical waste, lowering operational costs and delivering immediate results.

Goals and Study Overview

The primary objective of this study was to develop and validate NIR-based prediction models for key olive oil quality parameters and to demonstrate the feasibility of NIRS for routine quality control and adulteration screening. A diverse set of 137 commercial and experimental olive oil samples was analyzed, covering a broad range of physicochemical properties. Model performance was assessed by comparing NIR predictions with standard reference methods.

Methodology and Instrumentation

Samples were analyzed on the OMNIS NIR Analyzer Liquid in transmission mode across the 1000–2250 nm spectral range. Eight-millimeter disposable glass vials were maintained at 40 °C using an integrated temperature sensor to ensure consistent measurement conditions. The OMNIS Software platform handled data acquisition, spectral preprocessing and multivariate model development. A calibration set (75 %) and validation set (25 %) were defined to evaluate predictive accuracy.

Main Results and Discussion

Quantitative models were built for iodine value, free fatty acids (FFA), refractive index, K232, peroxide value (PV), induction time and major fatty acids (C16:0, C18:0, C18:1, C18:2, C18:3). Key performance indicators included standard error of calibration (SEC), standard error of cross-validation (SECV), standard error of prediction (SEP) and the coefficient of determination (R²CV). High correlation was achieved for:

• Iodine value (R²CV = 0.974, SEP ≈ 0.38 mg/100 g)
• Oleic acid C18:1 (R²CV = 0.980, SEP ≈ 0.75 %)
• Linoleic acid C18:2 (R²CV = 0.985, SEP ≈ 0.43 %)
• Refractive index (R²CV = 0.998, SEP ≈ 0.00012)
• Induction time (R²CV = 0.908, SEP ≈ 0.34 h)

Moderate correlations were observed for stearic acid C18:0 (R²CV = 0.778), FFA (R²CV = 0.746), peroxide value (R²CV = 0.719) and α-linolenic acid C18:3 (R²CV = 0.633). Model residuals and prediction errors confirm that NIRS can reliably replace or complement reference methods in routine analyses.

Benefits and Practical Applications

Compared with conventional techniques, NIRS offers substantial advantages:

• Measurement time below 10 seconds per sample versus 5 minutes to 15 hours for traditional methods
• No chemical reagents or hazardous solvents required
• Reduced sample preparation and waste generation
• Lower operational costs and higher laboratory throughput
• Capability to determine multiple parameters simultaneously from a single scan

The approach is well suited for quality control in olive oil production, import/export inspection and regulatory compliance checks.

Future Trends and Possibilities

Emerging developments in NIR technology and chemometrics will further enhance its applicability:

• Expansion of calibration databases to include sterol profiles, moisture and minor bioactive compounds
• Integration with automation platforms for inline and at-line monitoring in processing plants
• Implementation of portable or handheld NIR devices for field and spot testing
• Application of machine learning algorithms to improve model robustness and adaptability to new olive varieties and geographic origins

Conclusion

This study demonstrates that near-infrared spectroscopy using the OMNIS NIR Analyzer Liquid enables rapid, accurate and solvent-free determination of critical olive oil quality parameters. The validated models exhibit strong agreement with reference methods, offering significant time and cost savings while maintaining high analytical reliability.

References

No specific literature references were cited in the original text.