The cause of bitter flavour development in toasted rolled oats (Avena sativa L.)
Hubbard Foods Limited of Auckland makes a variety of oats-based value-added products. In the preparation of a range of products at Hubbard Foods, technical staff has become aware of a bitterness problem that sporadically appears in toasted oats. Toasting involves dry heating to about 150°C resulting in the golden colour and flavour development necessary for range of products. Bitterness development has been described in the literature, but Hubbard staff is necessarily focussed on production issues, rather than on a sporadic problem seemingly outside the scope of production variables. The author of this thesis set out to identify the cause and suggest a remedy. Prior research with oats has shown that bitterness and associated off-flavours are linked to the accumulation of free fatty acids, their volatile oxidation products, and possibly amino acids and certain phenols. Oats are distinguished from related grains by their high relative fat content, about seven percent, and an associated very active lipase. The free fatty acids stem from the lipase activity that should be, but may not be, inactivated at source in Australia. This is achieved in the milling process by physical disruption and moist heating to a temperature at which the enzyme is denatured. However, residual lipase activity may adversely affect oats quality during time in storage and transit. A number of analytical methods for cereals were adapted to match the constraints of time and resources. These methods were for colour, moisture, peroxidase activity, p-anisidine, and fat and free fatty acids content, composition of fatty acids, total phenols, volatiles, and bitterness as perceived by an analytical sensory panel of four people. Determination of lipase activity is very expensive, so peroxidase activity is commercially used as an indicator. If the latter is inactive, the former will necessarily be also inactive. The designed methods were first applied to 17 oats lots passing through the Hubbard environment, where 14 were paired raw and toasted. The values of moisture, fat content, free fatty acids content and total phenols were within the normal limits expected for commercial lots of oats compared with the previous studies. Not much variation was observed among the 17 oats lots, with the exception of lot DWHE25. Lot DWHE25 was a faulty product, which had high moisture content, high free fatty acids content, and tasted very bitter. The results suggested that moisture content, free fatty acids and bitterness were usually correlated. In spite of the differences encountered and the clues provided by extremes, the data generated from Hubbard oats lots did not provide enough variation in quality to lead to a definitive chemical model of bitter flavour development. But perhaps crucially, it was found that most samples as received from Hubbard Foods were peroxidase-active which conflicted with the results reported on specification sheets prepared by the oats supplier. These specifications accompanied each lot delivered to Hubbard Foods. Therefore, the supplier’s method was examined and was found to be deficient in one critically important respect. Their method omitted the key reactant hydrogen peroxide. Therefore, it is possible that the lipase was active in many of the samples. Therefore, some experiments were conducted where raw oats, from Hubbard Foods and a supermarket, were treated with water additions and stored for a period to examine the effect of moisture content on the quality and flavour deterioration on subsequent storage. Water-treated oats were toasted to simulate a typical Hubbard process, yielding a total of 58 samples with carrying moisture contents. The data set was statistically analysed to identify the cause of bitterness and the means of its control. The free fatty acids content, volatile compounds particularly hexanal, and total phenols increased with moisture content and storage time. The correlations between chemical analysis and sensory test indicated that free fatty acids positively correlated with bitterness (r = 0.71), and hexanal was also positively correlated. Total phenols did not appear to correlate with bitterness. Oats lots with high peroxidase activity tended to have the poorest quality, strongly implicating residual lipase activity as the critical factor. There were no important interactions between water addition and toasting for most of the experiments. Therefore, it seems likely that the toasting procedures at Hubbard Foods are not responsible for bitterness formation. The cause(s) of bitterness is certainly at source, with a faulty peroxidase test strongly implicated.