You have been redirected to your local version of the requested page
Semola and semolina quality control using NIR spectroscopy

AN-NIR-143

2026-09

Semola and semolina quality control using NIR spectroscopy

Nondestructive analysis of protein, moisture, gluten, and more in semolina-based products


Summary

Good pasta products start with good ingredients—including flour. Semola and semolina are special types of flour made from durum wheat, known for giving the resulting pasta a firm texture, bright yellow color, and superior cooking character. The difference is that semola is coarse (milled once) and semolina is exceptionally fine (milled twice). To make high-quality pasta, it is important to check key properties of these flours, like protein, moisture, ash content, gluten, and water absorption. This Application Note shows how monitoring these parameters with near-infrared spectroscopy (NIRS) can help pasta manufacturers quickly detect the best semola and semolina to keep their products consistent, tasty, and of high quality.


Experimental equipment

OMNIS NIR Analyzer Solid
Figure 1. OMNIS NIR Analyzer Solid.

Samples of semola and semolina flours were analyzed using an OMNIS NIR Analyzer Solid (Figure 1). Measurements were carried out in reflection mode while using a large sample cup. To enhance representativeness, the samples were rotated during the measurement, allowing the device to capture spectral data from multiple regions of each flour. OMNIS Software was used for all data acquisition and prediction model development.


Configuration


Result

The OMNIS NIR Analyzer was used to monitor the quality of semolina and semola flours, focusing on protein, moisture, ash, gluten content, and water absorption. Figure 2 shows an example of some semola spectra gathered by the OMNIS NIR Analyzer Solid.

For semolina, moisture showed the strongest performance (Figure 3), with low prediction errors and a high correlation between NIR predictions and laboratory reference values. Ash, protein, and gluten content (Figures 4–6) were also quantified with good accuracy.

For semola, protein content was predicted with excellent correlation to the reference data (Figure 7), while moisture, ash, gluten, and water absorption also showed reliable calibration performance (Table 1).

Near-infrared spectra of semola samples analyzed in reflection mode on an OMNIS NIR Analyzer Solid.
Figure 2. Near-infrared spectra of semola samples analyzed in reflection mode on an OMNIS NIR Analyzer Solid.

Result moisture content in semolina

Correlation diagram for the prediction of moisture in semolina measured with an OMNIS NIR Analyzer Solid.
Figure 3. Correlation diagram and the respective figures of merit for the prediction of moisture in semolina measured with an OMNIS NIR Analyzer Solid.
R2SEC (%)SECV (%)
0.9490.290.31

Result protein content in semolina

Correlation diagram for the prediction of protein in semolina measured with an OMNIS NIR Analyzer Solid.
Figure 4. Correlation diagram and the respective figures of merit for the prediction of protein in semolina measured with an OMNIS NIR Analyzer Solid.
R2SEC (%)SECV (%)
0.9070.600.63

Result ash content in semolina

Correlation diagram for the prediction of ash in semolina measured with an OMNIS NIR Analyzer Solid.
Figure 5. Correlation diagram and the respective figures of merit for the prediction of ash in semolina measured with an OMNIS NIR Analyzer Solid.
R2SEC (%)SECV (%)
0.7340.140.14

Result gluten content in semolina

Correlation diagram for the prediction of gluten in semolina measured with an OMNIS NIR Analyzer Solid.
Figure 6. Correlation diagram and the respective figures of merit for the prediction of gluten in semolina measured with an OMNIS NIR Analyzer Solid.
R2SEC (%)SECV (%)
0.5501.711.87

Result protein content in semola

Correlation diagram for the prediction of protein in semola measured with an OMNIS NIR Analyzer Solid.
Figure 7. Correlation diagram and the respective figures of merit for the prediction of protein in semola measured with an OMNIS NIR Analyzer Solid.
R2SEC (%)SECV (%)
0.9330.140.14

FOM from other semola prediction models

Table 1. Figures of merit for the prediction of moisture, ash, protein, gluten, and water absorption in semola.
Parameter (range)No. SpectraSEC (%)SECV (%)R2
Moisture (12.5–15.1%)1460.200.200.834
Ash (0.74–1.12%)990.040.040.787
Gluten (23–32%)430.670.810.815
Water absorption (53.7–58.8%)370.370.440.82

Conclusion

This Application Note demonstrates the feasibility of using NIR spectroscopy to determine multiple key parameters in semolina and semola flours. As shown in Table 2, various traditional analytical methods are typically used for these measurements. However, NIR spectroscopy has proven to be a faster, simpler, and highly accurate alternative, offering excellent precision while streamlining the analysis process.

 

Table 2. Overview of the ISO norms used as a reference method for the different parameters measured in semola and semolina flours.
ParameterMethod
ProteinISO 1871 – Kjeldahl method
MoistureISO 760 – Karl Fischer
AshISO 2171 – Determination of ash content 
GlutenISO 21415 – Methods for determining gluten 
Water Absorption ISO 5530-1 – Determination of water absorption using a farinograph
Contact

บริษัท เมทโธรห์ม สยาม จำกัด

979 111-115 อาคาร S.M. Tower ชั้น33, แขวงพญาไท, เขตพญาไท, ถ.พหลโยธิน
กรุงเทพมหานคร 10400

Contact