Application of Near-Infrared Filters in Measuring the Content of Oleic and Linoleic Acid in Peanut Oil

When determining the content of oleic acid and linoleic acid in different varieties of peanuts and peanut oil, hyperspectral imaging technology and near-infrared spectroscopy can be used. Using narrowband filters, the spectrum emitted by LED or halogen lamps is divided into ten narrow spectral regions, and these spectra are then used to image the peanuts. The imaging results at different wavelengths are analyzed using a mathematical model for regression. The wavelength range for near-infrared filters generally spans from 900nm to 1700nm, with a bandwidth of 10~15nm.

The nutritional value of edible oils largely depends on the content of fatty acids, which varies significantly among different plant varieties. Peanut oil is an excellent source of oleic acid and linoleic acid, often referred to as the "economical" olive oil. In recent years, peanuts have been widely cultivated in most tropical and subtropical countries, with China being the largest producer. Peanuts are critically important in various fields, including health, food, agriculture, industry, environment, and economics. The consumption of peanuts is directly associated with reducing the risk of coronary heart disease. The nutritional value of peanuts is mainly attributed to their high content of unsaturated fatty acids, such as oleic acid (ω9) and linoleic acid (ω6). The presence of unsaturated fatty acids can increase high-density lipoprotein levels in the blood, reduce low-density lipoprotein (poor cholesterol) levels, and thus help prevent diseases (such as heart disease, diabetes, and cancer), regulate weight, and lower blood sugar and blood pressure. This study uses non-destructive spectroscopic techniques to measure the content of oleic and linoleic acids in peanuts. Traditional gas chromatography (GC) is also used to provide chemical values for model development. Gas chromatography (a wet chemical method) can provide accurate reference values, but it is slow, time-consuming, complicated, and requires a large number of samples. The calibration set spectral data for standard fatty acids is obtained using non-destructive analysis methods. 96 varieties of peanut kernels and 83 varieties of peanut oil were analyzed experimentally. Spectral data for peanut kernels were obtained using a hyperspectral imaging system (Sisu CHEMA) and near-infrared spectroscopy equipment (DA 7200), and spectral data for peanut oil were obtained using Micro NIR 1700. Outliers were removed and significant wavelengths selected, using chemometric methods such as PCA and PLS to extract useful spectral information and build models. Both the calibration model and the prediction model had good regression coefficients, indicating successful results. For example, a PLS model established from 10 effective wavelengths within the 900nm-1700nm near-infrared spectral range had a regression coefficient of 0.97, with errors of 2.4 and 0.5 respectively, showing great potential for predicting oleic acid content. The study demonstrates that spectroscopic detection techniques can measure components in food (such as oleic and linoleic acids) in real-time, enabling continuous monitoring of food quality and safety, and establishing control systems to meet the growing consumer concern for food health and quality. Appropriate and efficient spectroscopic techniques must consider differences among various technologies. For example, in detecting peanut kernels, hyperspectral imaging provides more information than NIR methods. Because hyperspectral imaging can obtain both spectral and spatial data, it can predict the content of oleic and linoleic acids using a small amount of testing samples, with oleic and linoleic acid contents ranging from 18.820.2 mg/100 g and 1518 mg/100 g respectively. Traditional near-infrared spectroscopy (NIRS) does not provide spatial information on food components (fatty acids such as oleic and linoleic acids), while hyperspectral imaging can detect the three-dimensional information of components, thus providing comprehensive results. Additionally, the study results show that the three spectroscopic techniques have poor correlation in detecting the same fatty acid content at optimal wavelengths. Meanwhile, Micro NIR can be used to collect spectral data for peanut oil, while DA7200 NIR and HSI equipment currently lack the necessary accessories for oil sample testing. Spectral data collected for peanut oil using Micro NIR can be analyzed in a similar manner to the processing of peanut kernel data by HSI and NIRS, further establishing prediction models. Besides extracting oils, the modeling performance using Micro NIR is comparable to that of NIRS and HSI. This study has established six data models using the three aforementioned devices: three for detecting oleic acid and three for detecting linoleic acid. Mathematical models were established based on optimal wavelengths and corresponding regression coefficients, and model biases were predicted. The models established in this study need further verification and confirmation in multiple large laboratories to determine their suitability for future industrialized food testing and control. This study has achieved significant breakthroughs compared to traditional methods, providing a rapid, non-destructive method to predict unknown peanut samples.

Product Recommendation:

BP900nm - 10/15

BP1700nm - 10/15