Document Type

Dissertation - Open Access

Award Date

2020

Degree Name

Doctor of Philosophy (PhD)

Department / School

Agricultural and Biosystems Engineering

First Advisor

Kasiviswanathan Muthukumarappan

Keywords

Artificial Neural Network, Extrusion, Germination, Proso Millet, Quinoa, Response Surface Modelling

Abstract

According to Mordor Intelligence Research, the new compositional research and moisture content extrusion process are helping the growth of the plant protein market. The demand for plant proteins is growing at a fast rate, owing to change in lifestyle, lack of balanced dietary intake, and improved research and development in order to develop new kinds of plant-protein enriched products. It is necessary to identify the right plant protein sources and choose the right processing methodology to create highly digestible foods which can be consumed by infants and elderly as well. In addition, it is important to support local farmers and help them make value added products. The primary objective of this study was to investigate the feasibility of developing extruded foods from protein rich sprouted quinoa and proso millet flours with high protein and starch digestibility. The secondary objective was to understand the effect of extrusion processing conditions on the systems parameters and on the physical and physico-chemical properties of the extruded product. Quinoa and Proso millet were chosen as the grains of interest due to their high protein content as compared to wheat, rice, corn, amaranth and buckwheat. The study was broadly then divided in to three parts. In the first part, pre-treatment methods such as soaking and sprouting were analysed for effective reduction of saponins and phytic acid and increased starch and protein digestibility. The sprouting of quinoa increased the starch digestibility from 55.6% on Day 0 to 78.2% on Day 4. A similar increase in the protein digestibility was observed from Day 0 (42.2%) to Day 4 (75.5%). Sprouting also produced similar effects in proso millet where the starch digestibility increased from 51.7% (Day 0) to 77.1%. The breakdown of the phytates during sprouting of proso millet increased the protein digestibility from 46.4% on Day 0 to 76.8% on Day 4. Simultaneously the reduction of saponin content in quinoa and phytic acid content in both grains were observed. The saponin content in raw quinoa of 0.8g/100g was reduced to 0.1g/100g samples (Day 4) by germination. The phytic acid content in quinoa reduced from 1.1g/100g to 0.1g/100g (Day 4) and in proso millet reduced from 1.5g/100g to 0.2g/100g (Day 4). The color of the flour produced from the sprouted grains were significantly different from the unprocessed flour respectively. The L* value of sprouted quinoa flour was darker (L*=61.2) as compared to the control sample (L* = 82.6). Similarly, the sprouted proso millet flour was darker (L*=70.1) than the unprocessed proso millet flour (L*=84.2). In the second part, the extrusion process of sprouted quinoa was divided into three experiments. The first experiment is the single screw extrusion of sprouted quinoa. Using a response surface design to understand the influence of feed moisture content (15-25% w.b), die temperature (80-140°C), screw speed (90-220 rpm) and germination time (Days 0-4) on the physical and physico-chemical properties of sprouted quinoa extrudates was studied. The following responses were obtained: bulk density (116-154 kg/m3), hardness (1.05-1.8 N), water solubility index (11.5-16.5%), water absorption index (2.36-3.51), total color difference ΔE (14.8-21.7), expansion ratio (2.52-3.75), protein digestibility (80.5-86.5%) and starch digestibility (80.1-85.8%). The die temperature and germination time played a significant role in product responses. The second experiment was designed to create a puffed product with an inclusion of corn meal. Corn meal is the most common ingredient of expanded snacks in the food market. Because of its composition, ratio of vitreous to floury endosperm, and particle size, under optimal extruding conditions corn meal makes for a light, highly expanded, crunchy and soft product. Single-screw extrusion processing of sprouted quinoa-corn meal blend was studied using a response surface design to understand the influence of feed moisture content (15-25% w.b), die temperature (80-140°C), screw speed (90-220 rpm), corn meal ratio (0-30%) and quinoa time (Days 0-4) on the physical and physico-chemical properties of sprouted quinoa- corn meal blend extrudates. The following responses of the extrudates were measured and the values ranged from: bulk density (102-145 kg/m3), hardness (1.03-1.62 N), water solubility index (12.3-19.9%), water absorption index (2.44-3.79), total color difference ΔE (14.1-21.4), expansion ratio (2.75-3.97), protein digestibility (78.4-85.8%) and starch digestibility (77.5-85.6%). The addition of corn meal improved the color difference and expansion ratio of the extrudates. The third experiment employed a twin-screw extruder to understand the influence of feed moisture content (15-25% w.b), die temperature (80-140°C), screw speed (90-220 rpm) and germination time (Days 0-4) on the physical and physico-chemical properties of sprouted quinoa flour extrudates. The bulk density (132-175 kg/m3), hardness (1.56-2.14 N), water solubility index (14.4-18.5%), water absorption index (2.93-3.41), ΔE (16.7- 20.8), expansion ratio (2.28-2.83), protein digestibility (78.4-84.1%) and starch digestibility (77.1-83.4%) were measured. All independent parameters had statistically significant effects on all the extrudate characteristics. Twin screw extrusion did not improve the extrudate characteristics significantly. Poor expansion ratio was observed as compared to single-screw extrusion. Similar experiments were carried out for sprouted proso millet flour. Single-screw extrusion processing of sprouted proso millet flour was studied using a Box Behnken response surface design to understand the influence of feed moisture content (15-25% w.b), die temperature (80-140°C), screw speed (90-220 rpm) and germination time (Days 0-4) on the physical and physico-chemical properties of proso millet extrudates. The following responses of the extrudates were measured and the values ranged from: bulk density (101-137 kg/m3), hardness (1.01-1.3N), water solubility index (14.8-18.7%), water absorption index (4.0-4.41), ΔE (13.0-17.1), expansion ratio (3.28-3.75), protein digestibility (78.2-86.6%) and starch digestibility (80.9-87.7%). Both feed moisture content and extruder die temperature had statistically significant effects on all the extrudate characteristics. Extruder screw speed had minimal effect on the extrudate properties. It was followed by a single-screw extrusion processing of sprouted proso millet-corn meal blend with feed moisture content (15-25% w.b.), die temperature (80-140°C), screw speed (90-220 rpm), corn meal ratio (0-30%) and proso millet germination time (Days 0- 4) on the physical and physico-chemical properties of sprouted proso millet-corn meal blend extrudates. The following responses of the extrudates were measured and the values ranged from: bulk density (100-143 kg/m3), hardness (0.84-1.59 N), water solubility index (12.5-20.1%), water absorption index (2.55-3.90), total color difference ΔE (13.07-20.37), expansion ratio (2.74-3.96), protein digestibility (78.5-85.9%) and starch digestibility (77.7-85.8%). The addition of 30% corn meal significantly increased the expansion ratio from 3.29-3.96. Last study was to investigate the effects of twin-screw extrusion processing of sprouted proso millet with feed moisture content (15-25% w.b), die temperature (80-140°C), screw speed (90-220 rpm) and germination time (Days 0-4) on the physical and physicochemical properties of sprouted proso millet extrudates. The responses from the experiment include: bulk density (132-175 kg/m3), hardness (1.56-2.14 N), water solubility index (14.4-18.5%), water absorption index (2.93-3.41), ΔE (16.7-20.8), expansion ratio (2.28-2.83), protein digestibility (78.4-84.1%) and starch digestibility (77.1-83.4%). The expansion ratio was significantly poor when compared to the single screw extrusion of sprouted proso millet. In the third and last part, a comparative study on the efficacy of Response Surface Methodology (RSM) and Artificial Neural Network (ANN) in modelling of single and twin-screw extrusions were conducted. We optimized the neural network topology to predict the error and regression coefficient and root mean square error using artificial neural network and compared the results with response surface methodology for single screw (sprouted quinoa, sprouted quinoa-corn meal blend, sprouted proso millet, sprouted proso millet-corn meal blend) and twin screw (sprouted quinoa and sprouted proso millet) extrusion processes. Irrespective of the ingredient composition and blend for all extrusion processes, ANN predictions have regression coefficients greater than 0.9 and RSM predictions are greater than 0.8. Similarly, the root mean square error values were low in all ANN predictions are compared to RSM. Based on error analysis results, the prediction capability of ANN model is found to be the best of all the prediction models investigated, irrespective of food composition and extrusion processes.

Library of Congress Subject Headings

Extrusion process.
Quinoa -- Composition.
Quinoa -- Processing.
Broomcorn millet -- Composition.
Broomcorn millet -- Processing.
Plant proteins as food.

Format

application/pdf

Number of Pages

548

Publisher

South Dakota State University

Download

Share

COinS
 

Rights Statement

In Copyright