Study On Determination Of Oleic And Linoleic Fatty Acids Contents In Peanut Seeds Varieties And Its Oil Product Using HSI And NIRS

The aim of this study was to determine the content of oleic fatty acid(OFA) and linoleic fatty acid(LFA)in peanut seeds varieties and peanut varieties’ oils product. The nutritional benefits of edible oils and its products depend much on the ‘fatty acid profiles’ and these vary greatly between oil types.Peanut oil is a good source of OFA and LFA, and is commonly used as an ‘economic’ olive oil. In recent years, cultivation and consumption of peanuts have increased widely particularly in the tropics and subtropics countries. However, in the world China is the leading country. Peanut has been very beneficial in different sectors including health, food and agriculture, industrial,environmental and economic sectors. Besides nutrition experts’ recommendations about plant based diet(such as peanut) as a healthy way to meet nutritional needs and lowering the risks of diseases such as coronary heart diseases(CHD) has a long standing interest. Peanuts were therefore chosen because they are nutrient dense foods and their consumption has been associated with reduced risk(potential prevention) of diseases such as heart diseases, diabetes and cancers.Peanuts help in reducing cholesterol levels by raising blood concentrations of high density lipoproteins while reducing low-density lipoproteins(bad cholesterol), weight management,lowering sugar and sucrose contents in the blood, and reduce blood pressure. The health benefits are partially attributable to the presence of high content of unsaturated fatty acids such oleic and linoleic acid in the peanut seed varieties.When it comes to analysis of fatty acids, a number of methods exist. These include Gas Chromatography(GC), Thin Layer Chromatography(TLC), High performance Liquid Chromatography(HPLC), Capillary Electrophoresis(CE) and Real-Time Polymerase Chain Reaction(RT-PCR). However, all these methods are slow, demanding, tiresome, time consuming,use a lot of sample and need sample pre-treatment. To overcome these problems, non-destructive methods such as Near Infra-Red spectroscopy(NIRS) and Hyperspectral Imaging(HSI) were developed. This study therefore focused on the use of spectroscopy(non-destructive techniques)for detecting OFA and LFA in peanut seeds’ varieties and its products. However, while the focus was on the non-destructive techniques, the study also used Gas Chromatography(GC). The GC method(wet chemistry) was used to compliment spectroscopy because it has been used since 1903 and has been an accurate and leading method in fatty acid analysis. In addition, the GC is used in this study because spectra data cannot stand alone; it needs chemical data for calibration.A sample of 96 seeds varieties and 83 oil products were used in the analysis. Oleic and linoleic fatty acids in the peanut seeds varieties were detected using Sisu CHEMA type hyperspectral imaging and DA 7200 array analyzer(NIRS). The detection of oleic and linoleic acid in the oil products was done using Micro NIR 1700 spectrometer. Spectra data were collected at the NIR range of 900-1700 nm for HSI and Micro NIR 1700 spectrometer, and 950-1650 nm for NIRS. Valuable information of collected spectra data was then extracted, summarized and calibrated with GC data using chemometric tools such as PCA and PLSR. Subsequently, outliers were detected and excluded and significant wavelengths selected. Further, prediction models for future detection of the contents of OFA and LFA were constructed. The calibrated and verified models(n >30)constructed for predicting fatty acids in all the three used techniques had a regression coefficient of R2 cv > 0.85 which indicates that the models are good and acceptable for future use. For example,the coefficient of determination(R2cv = 0.97) together with minor errors(SEP = 2.4 and RMSE =0.5) obtained by cross validated PLS models at 10 effective wavelengths out of 239 indicated that NIR spectral range(900.82-1647.7 nm) had a tremendous ability to predict OFA’ content using HSI.The study observed that nondestructive methods are truly real-time and a technological improvement in detecting food composition such as peanut oleic and linoleic fatty acids’ contents. This suggests that the continual use of these methods for food quality and safety monitoring and control systems is crucial to meet the ever-increasing interest from consumers concerning health-related properties of foods. Prominently, the use of appropriate and efficient spectroscopy techniques needs to be considered as they differ from one another. For example, the detection of fatty acids in seeds, HSI exhibited more information than the other two methods. The HSI was able to predict and map the composition of OFA and LFA which ranged from 18.8- 20.2 mg/100 g and 15-18 mg/100 g respectively, and using fewer number of test(prediction) set. This was so because of its ability to capture both spectral and spatial data. The HSI was unlike conventional NIRS which could not give the complete quantitative prediction of food composition(OFA and LFA); instead, it could only detect the content dimension but not the incomplete result of locating the analytes. Poor relationship on wavelengths selected for detection of one fatty acid using the three spectroscopy techniques was also observed.On the other hand, Micro NIR was useful in collection of spectra data in peanut oil as NIRS and HSI could not. The spectra data obtained was analyzed following similar steps used for the other spectra data obtained after scanning peanut seeds using HSI and NIRS. Though further sample’preparation on extraction of oils had to be done but still Micro NIR was good, fast, less tiresome,required little sample and simple to use just like NIRS and HSI.The study came up with six statistical models; three for detecting OFA and the other three for detecting LFA. The equations were developed based on the selected wavelengths and the corresponding coefficient value; and the offset of the model were also indicated. The study found that the non-destructive techniques were significant over the traditional methods and they could give future methods for predicting unknown composition/contents of peanut varieties. They were simple, fast,environmental friendly, using little sample without preparation and less tiresome. However, it was observed that they require computation information redundancy removal, relevant identification removal and modeling accuracy. The study’ findings suggest that these models are valid and verifiable. It is therefore recommended that the models be transferred to large scale laboratories for further validation and verification. Thereafter they can be used for future industrial application particularly on food quality and safety analysis as well as monitoring and control system to improve marketability of peanut oil and its products.