NIR vs chimie humide : des méthodes complémentaires pour évaluer la qualité du tourteau de soja

Understanding trypsin inhibitor levels in soybean meal is essential for optimizing protein utilization and supporting animal performance. Two primary analytical approaches are used to measure trypsin inhibitor activity: traditional wet chemistry and near-infrared spectroscopy, commonly referred to as NIR. While both methods aim to quantify the same component, they differ significantly in how results are generated, interpreted, and applied in practical settings. 

Wet chemistry methods, such as the AOCS Ba 12a-2020 procedure, are considered the reference standard for measuring trypsin inhibitor (TI) activity. This approach relies on direct chemical reactions and laboratory procedures to quantify TI levels in a sample. Because it measures the analyte directly, wet chemistry provides reliable and well-established results. However, this accuracy comes with tradeoffs. The process is time-intensive, requires specialized laboratory equipment, and depends on trained personnel. As a result, wet chemistry is often used for calibration, validation, and situations where precise measurement is critical. 

Near-infrared spectroscopy (NIR) offers a different approach. Rather than measuring trypsin inhibitor activity directly, NIR uses light in the near-infrared region of the electromagnetic spectrum, typically between 780 and 2,500 nanometers (nm), to analyze the physical and chemical characteristics of a sample. The NIR method relies on predictive modeling. Samples are first analyzed using wet chemistry, and those results are then correlated with their corresponding NIR spectra. Over time, this process builds a robust equation that can estimate trypsin inhibitor levels based on spectral data alone. 

The key advantage of NIR is speed and efficiency. Once a calibration model is established, NIR can deliver rapid results with minimal sample preparation. This makes it a practical tool for routine analysis, quality control, and real-time decision-making in feed production environments. However, because NIR is an indirect method, its accuracy depends on the strength and relevance of the calibration model. It does not replace wet chemistry but instead builds upon it. 

An important consideration when comparing these methods is variability. All analytical techniques have an inherent margin of error. For wet chemistry, variability is often expressed as relative standard deviation. In the case of TI analysis, this variability has been calculated at approximately plus or minus 13.5% for typical soybean meal ranges. NIR, on the other hand, has its own margin of error, expressed as a standard deviation of approximately +/- 1.27 mg/g. 

These differences highlight a critical point. Results from wet chemistry and NIR should not be expected to match exactly. Instead, they should be evaluated within their respective margins of variability. When the result ranges from both methods overlap, the values are considered aligned and within acceptable analytical limits. If they do not overlap, the sample may be an outlier and could require further investigation to refine the predictive model. 

In practice, wet chemistry and NIR are complementary tools. Wet chemistry provides the scientific foundation and validation needed to ensure accuracy. NIR extends that foundation by enabling faster, scalable analysis that supports timely decisions. Together, they offer a balanced approach to managing variability in soybean meal quality and maintaining confidence in nutritional assessments. 

Paula Fisher

Como Gerente Sênior de Serviços Técnicos, a Sra. Fisher gerencia a equipe que fornece análises de ingredientes, materiais e produtos para os grupos de Inovação, R&D e Atendimento ao Cliente da NOVUS.

Read more by Paula Fisher