Rancidification of gluten-free whole flours

Millet and rice are gluten-free cereals, which are mostly consumed in Africa and Asia. The global availability of the whole flour of these two grains is however limited, because of rapid rancidity processes, which occur due to the large endosperm. Consequently, the whole flour only has a rather short shelf-life. That is why, this research project aims at investigating the rancidity processes on a molecular level using various spectroscopic and wet-chemical analyses. More specifically, it will be investigated if and how strongly substances which inhibit or decelerate oxidation processes are able to counter rancidity processes. Furthermore, water content, fat content and water activities of whole flour samples and the influence of these parameters on the deterioration will also be studied. Lastly, differences regarding rancidity processes in various millet and rice cultivars will also be analyzed, in order to identify which cultivars would be more preferable for storage as whole flours. In the research project wet- chemical will be applied to investigate specific molecules or molecular classes and spectroscopic methods will be used to obtain a more holistic impression of the flour samples. Since a vast amount of data will be recorded, mathematical algorithms are needed for data analysis. In doing so, explorative methods will be applied to find pattern in the data sets. This will then allow to interpret which spectral regions are related to rancidity processes and how these regions are related to wet-chemical data. Consequently, these mathematical algorithms help to extract relevant information and interpret the data. The project will lead to a better understanding of deterioration processes in whole flours of millet and rice. A focus will be put on studying the interactions between molecular classes during rancidification and how strongly they influence the process. The results can then serve as a basis for the development of strategies to prolong shelf-life of whole flours.

Millet and rice are widely used gluten-free grains in Asian and African diets. Whole flours of these grains are prone to deterioration because of a higher fat content due to bigger endosperms compared to e.g. wheat. In this research project, the deterioration process was followed by recording various nutritional parameters as well as markers of rancidity. Differences in deterioration process of millet and rice as well as between various cultivars of these grains were successfully investigated. The results of this research indicate that deterioration processes vary for different molecular classes. While fat content decreases continuously and fat acidity increases continuously, the peroxide value shows a cyclic behavior. This indicates that only after a certain threshold is reached, the degradation of peroxides is kick-started within the flours. In addition, the research also found that the deterioration does not have a linear correlation to temperature, in fact deterioration processes are highly accelerated above 42 C. Usually water content and water activity is also a strong predictor for deterioration processes, however experiments within this research project found that for flours the influence is negligent, as water content and water activity are both fairly low and stayed approximately the same for the duration of the experiments. Near-infrared spectroscopy (NIRS) analysis revealed that changes in deterioration processes varied more between millet cultivars than rice cultivars. In general however, the spectral changes wich do occur are similar between millet and rice whole flours. Furthermore, tracking changes with spectral analysis also showed that C=O vibrations change very early in the deterioration process, indicating that fats are the first molecular class to deteriorate. C-H & O-H vibrations on the other hand, change later and not at the same time as C=O. This means that oxidation of fatty acids to peroxides only occurs after the saponification of fats. This is also supported by wet-chemical analysis. In addition to new findings regarding the deterioration of gluten-free whole flours, this research also lead to a new algorithm for automated spectral pre-processing and quantitative as well as qualitative analysis. The algorithms permutes through various spectral pre-processing methods, optimizes parameters for regression models and saves all results in a comprehensive list. This way, the often tedious task of testing various pre-treatments and various parameters for regression is automated and results can be reviewed immediately. The algorithm also proves that while certain spectral pre-treatments are necessary, others do not improve the quality of the analysis and should therefore not be applied in order to better preserve the spectral information. In summary, this research was able to uncover new insights for deterioration processes for gluten-free grains and generate and implement a novel algorithm for improved spectral data evaluation and interpretation.