Non-traditional grains in low and high moisture extrusion applications –residence time, physico-chemical properties and resistant starch

Abstract

Sorghum, millets and teff are important staple crops worldwide, especially in semi-arid regions of Africa and India because of their drought tolerance. However, lack of research and other limitations have restricted their usage in food products. This study is focused on extrusion processing of low and high tannin sorghum varieties, millet and teff for high and low moisture applications, and to investigate process characteristics such as residence time distribution and specific mechanical energy, physico-chemical properties of resultant food products (such as pre-cooked pasta and expanded snacks) and their resistance starch content. Results from preliminary lab scale extrusion, including optimization of starch type and level for pre-cooked pasta and in-barrel moisture for expanded snacks, were used to design pilot-scale studies on a twin-screw extruder. In the first pilot-scale experiment, decorticated white sorghum blends prepared with addition of mono-glycerides (0.5%, 1% and1.5%) and salt (1%) were processed at three different in-barrel moisture contents 40%, 44% and 48% (wet basis) for processing of precooked pasta. The optimum formulation containing 1% mono-glycerides and process conditions corresponding to 48% in-barrel moisture were also used to develop precooked teff and millet pasta. The non-traditional grain based pastas were investigated for cooking quality, thermal characteristics using differential scanning calorimeter, pasting properties using rapid visco analyzer and texture profile analysis. In general, increasing in-barrel moisture led to reduction in solid losses (ranging from 4.0-8.2% for all treatments), indicating improvement in cooking quality. On the other hand, increase in mono-glycerides concertation led to higher cooking losses, and also affected pasting and textural properties significantly. Sorghum-based precooked pasta was of best quality while millet pasta was poorest in cooking quality, visual and textural attributes. Cooking loss for control pre-cooked pasta produced in this experiment using semolina was 4.5%, and commercial semolina pasta was 3.2%. Residence time distribution in pilot-scale twin screw extruder, during high moisture process conditions used for pre-cooked pasta, was also investigated at three different in-barrel moistures (40%, 44% and 48%) and monoglycerides/ lipid (0.5, 1% and 1.5%) concentrations. Increase in in-barrel moisture significantly decreased mean residence time. For example, mean residence time was 4.47 min at 40% moisture, 3.89 min at 44% and 3.74 min at 48%. On the contrary, residence time significantly increased with lipid level. For example, mean residence time was 3.87 min at 0.5% concentration of mono-glycerides, 4.48 min at 1% and 4.70 min 1.5%. In the second experiment focusing on low moisture applications, pilot-scale twin screw extrusion was used to process decorticated white sorghum and high tannin sumac sorghum for expanded snacks. The addition of sumac bran decreased the specific mechanical energy input (366-578 kJ/kg) and expansion ratio (6.4-7.9), and resulted in higher piece density of extrudates. Use of sumac bran and sumac flour led to increase in resistant starch content, although it was less than 1% for all treatments. Therefore, extrusion with ingredients having high tannin content does not provide value, despite tannins being associated with resistant starch at least in raw materials.