Title

Author(s)

Citation

Issued Date

URL

Rights

Assessment and effects of starch properties in food materials

Wu, Kao; 吴考 Wu, K. [吴考]. (2015). Assessment and effects of starch properties in food materials. (Thesis). University of Hong Kong, Pokfulam, Hong Kong SAR. Retrieved from http://dx.doi.org/10.5353/th_b5699894. 2015

http://hdl.handle.net/10722/236338

The author retains all proprietary rights, (such as patent rights) and the right to use in future works.

ASSESSMENT AND EFFECTS OF STARCH PROPERTIES IN FOOD MATERIALS

Kao Wu

Ph.D. THESIS

THE UNIVERSITY OF HONG KONG November 2015

Abstract of thesis entitled

ASSESSMENT AND EFFECTS OF STARCH PROPERTIES IN FOOD MATERIALS

Submitted by Kao Wu For the Degree of Doctor of Philosophy at The University of Hong Kong in November 2015

Starch is an important food carbohydrate, and accounts for many of the properties of popular starchy foods such as noodles. Generally noodles can be divided into two types: wheat-based noodles (containing gluten) and starch-based noodles (gluten free). Physical properties of wheat noodles blended with buckwheat or millet flour, and mung bean starch noodles blended with other starches were investigated in this study. Commercial buckwheat and millet flour were added to wheat flour at different mixing ratios ranging from 0–100 %, and their pasting, thermal, gelling and rheological properties were investigated. Two endotherms were observed during DSC measurement of millet-wheat blends at high mixing ratio, as individual starch tended to gelatinize separately in excess water. Large and mostly non-additive changes were observed in pasting and rheological properties, indicating interactions existed between components. However, within a complex flour-water system, with many variables altered at the same time (protein, lipid, gluten content), it is difficult to accurately model these changes.

Mung bean is widely considered as the best material to make starch noodles. Potato, sweet potato, rice and sorghum were added to mung bean starch and their impact on pasting, thermal, swelling, rheological, and gelling properties were studied. Non-additive effects were found in most attributes, and were partially explained by amylose content and particle size distribution.

After studies on flour and starch blends, corresponding noodle qualities were investigated. Higher level of buckwheat and millet addition led to a weaker texture in wheat noodles but no big variation in cooking properties. A similar situation was found in most noodles made from starch blends. A 10 % addition of buckwheat and millet to wheat noodles can lower in vitro digestion, while addition of other starches all led to worse qualities.

A wire cutting and peeling test model was established, using seven kinds of noodle sheets for texture evaluation, but was not fully consistent with oral sensations on corresponding noodle strands, possibly due to impact of humidity variation on the sample surface and on the effect of saliva lubrication in the mouth. Further, an indentation test for elasticity evaluation was performed on nine types of noodle sheets, in combination with AFM imaging, pasting and texture study. Results indicated that relaxation ratio of noodle sheets may relate to final viscosity of flour and starches, but a larger sample size is needed for further confirmation.

Word count: 388

ASSESSMENT AND EFFECTS OF STARCH PROPERTIES IN FOOD MATERIALS

by

Kao Wu

B.Eng. (Huazhong Agricultural University, China)

A thesis submitted in partial fulfillment of the requirements for the Degree of Doctor of Philosophy at The University of Hong Kong

November 2015

DECLARATION

I declare that this thesis represents my own work, except where due acknowledgement is made, and that it has not been previously included in a thesis, dissertation or report submitted to this University or to any other institution for a degree, diploma or other qualifications.

Signed_____________________________ Kao Wu

i

ACKNOWLEDGEMENTS

This thesis describes my research work during the past four and half years. Time easily flows, however, those people who motivate me to the finish line will never vanish in my sea of memory. Here with fully gratitude, I would like to acknowledge deepest thanks to the following persons:

Prof. Harold Corke, my supervisor, for offering me an incredible precious chance to study in HKU, with wonderful exchange opportunities, continuous support and great freedom on research.

Prof. Anil Gunaratne and Dr. Lilia Collado, for equipping me with necessary research skills to start my expedition on the exciting science lands of starches.

Prof. Peter Lucas, for providing me a great opportunity to visit Kuwait University to explore the amazing science world of noodle texture under professional guidance. Appreciation of your full care, invaluable advice and unforgettable encouragements will last forever in my heart!

Dr. Fan Zhu, for providing me another precious exchange opportunity to visit Auckland University, with irreplaceable advice and excellent supervision on my research.

Dr. Jinsong Bao and Dr. Yizhong Cai, for kindly suggestions on my work. Dr. Man Xiao, for essential help in Wuhan.

ii

Special thanks to my uncle, Dr. Huaixiang Wu, for his invaluable suggestions and care all these years; and my cousin Mr. Zhenghui Wu, for vital assistance especially in my first year.

I would like to thank all lab colleagues, for the years we spend fighting together. Lab members in Hong Kong: Mingwei Fong; Joyce Wong; Vivian Cheung; Maggie Law; Ginny Ngai; Renyou Gan; Wilson Leung; Shuhong Dai; Darren Chan.

Sincere thanks to all friends that accompany with me overcoming various mountains in both real and mental world steps by steps. Friends in Hong Kong: Xinming Yang; Zehua Sun; Kehu Li; Yubin Yan; Aiqun Hu; Xuran Zuo; Jianbing Xu; Qinglong Wu; Hongxian Sun; Pan Liao; Yan Xue; Peihui Wang; Yan Wang; Yong Lai; Zhiyi Lv; Jiang Li; Jiaqu Yi; Jia Zhao; Hengzhi Liu; Gangyan Xu. Friends in Wuhan: Fan Ke; Mengfei Wang; Lu Wang; Yue Shen; Difei Huang; Li Wan; Xing Li; Yao Yao; Le Wang; Xiaoyang Zheng; Pei Jiang. Friends in Kuwait: Adam Casteren; Swapna; Yijia Xu; Jiuzhi Zhao; Chao Huang; Qi Zhao Friends in Auckland: Guantian Li; Zhi Yang; Zhao Li; Elisa Lam; Bee Ratchaneekorn; Paul; Xiufang Bi; Da Chen; Ling Zhang; Trang Xuan Ton Duong; Liz Xiong; Jin Yee Liew; Xin Yan Cheng; Rusiru Karunaratne; Rose Lu; Yuanbing Deng; Jeffry Sutanto; Elaine Huang.

Last but not least, I would like to thank my parents, for eternal loves and firm encouragements over thousand miles.

iii

ABBREVIATION LIST

PV

Peak viscosity (RVU)

G'

Storage modulus

HPV

Hot paste viscosity (RVU)

G''

Loss modulus

CPV

Cold paste viscosity (RVU)

tan δ

loss tangent

BD

Breakdown viscosity (RVU)

BF

Buckwheat flour

SB

Setback viscosity (RVU)

MF

Millet flour

SV

Swelling volume (mL/g)

WF

Wheat flour

To

Onset temperature (°C)

MB

Mung bean starch

Tp

Peak temperature (°C)

SP

Sweet potato starch

Tc

Conclusion temperature (°C)

PT

Potato starch

△H

Melting enthalpy (J/g)

S

Sorghum starch

HD

Hardness (g)

R

Rice starch

ADH

Adhesiveness (g.s)

COH

Cohesiveness

CHE

Chewiness

GUM

Gumminess

σ0

Yield stress

K

Consistency coefficient

n

Flow behavior

R2

Regression coefficient

iv

Contents DECLARATION................................................................................................................................i ACKNOWLEDGEMENTS ..............................................................................................................ii ABBREVIATION LIST...................................................................................................................iv Chapter 1. Introduction and literature review ...................................................................................1 1.1. Starch .................................................................................................................................1 1.1.1. Introduction.............................................................................................................1 1.1.2. Amylose and amylopectin .......................................................................................1 1.1.3. Swelling ..................................................................................................................2 1.1.4. Gelatinization ..........................................................................................................3 1.1.5. Retrogradation.........................................................................................................3 1.1.6. Gel texture...............................................................................................................4 1.1.7. Pasting.....................................................................................................................4 1.1.8. Rheology .................................................................................................................5 1.1.9. Digestibility.............................................................................................................6 1.1.10 Functionality improvement ....................................................................................7 1.2. Wheat-based noodle ...........................................................................................................9 1.2.1. Wheat noodle ..........................................................................................................9 1.2.2. Buckwheat noodle.................................................................................................10 1.2.3. Millet noodle ......................................................................................................... 11 1.3. Starch noodle....................................................................................................................12 1.3.1. Introduction...........................................................................................................12 v

1.3.2. Mung bean noodle.................................................................................................13 1.3.3. Sweet potato noodle ..............................................................................................14 1.3.4. Potato noodle.........................................................................................................15 1.3.5. Rice noodle ...........................................................................................................16 1.3.6. Sorghum noodle ....................................................................................................18 1.4. Glycemic index ................................................................................................................19 1.5. Texture evaluation. ...........................................................................................................20 1.6. Introduction and objectives of this study .........................................................................20 References...............................................................................................................................23 Chapter 2. Buckwheat and millet affect the thermal, rheological, and gelling properties of wheat flour.................................................................................................................................................38 Abstract ...................................................................................................................................38 2.1. Introduction......................................................................................................................39 2.2. Materials and methods .....................................................................................................40 2.2.1. Materials................................................................................................................40 2.2.2. Chemical composition...........................................................................................40 2.2.3. Particle size distribution........................................................................................41 2.2.4. Swelling volume....................................................................................................41 2.2.5. Pasting property ....................................................................................................42 2.2.6. Gel texture.............................................................................................................42 2.2.7. Thermal property...................................................................................................43 2.2.8. Steady shear analysis.............................................................................................43 vi

2.2.9. Dynamic oscillatory measurement ........................................................................44 2.2.10. Statistical analysis ...............................................................................................44 2.3. Results and discussion .....................................................................................................45 2.3.1. Chemical analysis..................................................................................................45 2.3.2. Particle size ...........................................................................................................45 2.3.3. Pasting properties..................................................................................................46 2.3.4. Swelling volume....................................................................................................48 2.3.5. Gel texture properties............................................................................................49 2.3.6. Thermal properties ................................................................................................49 2.3.7. Dynamic oscillatory properties .............................................................................50 2.3.8. Steady shear properties .........................................................................................53 2.4. Conclusion .......................................................................................................................54 References...............................................................................................................................65 Chapter 3. Buckwheat and millet flour affect the color, texture and digestibility of wheat-based noodles ............................................................................................................................................73 Abstract ...................................................................................................................................73 3.1. Introduction......................................................................................................................74 3.2. Materials and methods .....................................................................................................76 3.2.1. Materials................................................................................................................76 3.2.2. Noodle preparation................................................................................................76 3.2.3. Cooking properties ................................................................................................77 3.2.4. Color analysis........................................................................................................78 vii

3.2.5. Texture analysis of noodles ...................................................................................78 3.2.6. Total starch ............................................................................................................79 3.2.7. In vitro digestion ...................................................................................................79 3.2.8. Statistical analysis .................................................................................................81 3.3. Results and discussion .....................................................................................................81 3.3.1. Effect of BF/MF addition on textural and cooking properties ..............................81 3.3.2. Effect of BF/MF addition to color of cooked noodles ..........................................82 3.3.3. In vitro digestion ...................................................................................................83 3.4. Conclusion .......................................................................................................................84 References...............................................................................................................................89 Chapter 4. Thermal, rheological and gelling properties of mung bean starch blends with other starches............................................................................................................................................94 Abstract ...................................................................................................................................94 4.1. Introduction......................................................................................................................95 4.2. Materials and methods .....................................................................................................96 4.2.1. Materials................................................................................................................96 4.2.2. Particle size ...........................................................................................................97 4.2.3. Pasting property ....................................................................................................97 4.2.4. Gel texture.............................................................................................................97 4.2.5. Swelling volume....................................................................................................98 4.2.6. Thermal property...................................................................................................98 4.2.7. Steady shear property............................................................................................99 viii

4.2.8. Dynamic oscillatory measurement ........................................................................99 4.2.9. Statistical analysis ...............................................................................................100 4.3. Results and discussion ...................................................................................................100 4.3.1. Amylose content..................................................................................................100 4.3.2. Particle size .........................................................................................................101 4.3.3. Thermal properties ..............................................................................................101 4.3.4. Swelling volume..................................................................................................102 4.3.5. Pasting properties................................................................................................103 4.3.6. Gel texture properties..........................................................................................106 4.3.7. Dynamic oscillation properties............................................................................107 4.3.8. Steady shear properties ....................................................................................... 111 4.4. Conclusion ..................................................................................................................... 111 References.............................................................................................................................126 Chapter 5. Effect of starch blending on quality of mung bean starch noodles..............................132 Abstract .................................................................................................................................132 5.1. Introduction....................................................................................................................133 5.2. Materials and methods ...................................................................................................135 5.2.1. Materials..............................................................................................................135 5.2.2. Starch noodle preparation ...................................................................................135 5.2.3. Cooking properties ..............................................................................................135 5.2.4. Color analysis......................................................................................................136 5.2.5. Texture analysis of noodles .................................................................................136 ix

5.2.6. Total starch ..........................................................................................................137 5.2.7. In vitro digestion .................................................................................................138 5.2.8. Statistical analysis ...............................................................................................139 5.3. Results and discussion ...................................................................................................139 5.3.1. Textural, cooking and color properties................................................................139 5.3.2. In vitro digestion .................................................................................................142 5.3. Conclusion .....................................................................................................................144 References.............................................................................................................................149 Chapter 6. Adhesion, cohesion and friction estimated from combining cutting and peeling test results for thin noodle sheets* .......................................................................................................152 Abstract .................................................................................................................................152 6.1. Introduction....................................................................................................................153 6.2. Materials and methods ...................................................................................................155 6.2.1. Materials..............................................................................................................155 6.2.2. Sheet preparation.................................................................................................156 6.2.3. Mechanical test methodology..............................................................................157 6.2.4. Sensory analysis ..................................................................................................159 6.2.5. Data analysis .......................................................................................................159 6.3. Results and Discussion...................................................................................................160 6.4. Conclusion .....................................................................................................................166 References.............................................................................................................................174 Chapter 7. Indentation of thin noodle sheets.................................................................................180 x

Abstract .................................................................................................................................180 7.1. Introduction....................................................................................................................181 7.2. Materials and methods ...................................................................................................184 7.3. Results............................................................................................................................189 7.4. Discussion ......................................................................................................................191 7.5. Conclusion .....................................................................................................................193 References.............................................................................................................................201 Chapter 8. General conclusion and discussions ............................................................................206 Appendix: Combined references list.............................................................................................208

xi

Chapter 1. Introduction and literature review

1.1. Starch 1.1.1. Introduction

Produced by photosynthesis in green plants, starch is made up of a large number of glucose units bound by glycosidic bonds. It is the major component stored in cereal grains (e.g. rice and maize), and in root and tuber crops (e.g. potato and sweet potato), and is an important energy source in human diets. Not only consumed directly in foods, it also offers a wide range of desired functional properties to industry in particular related to its texturizing ability (Lehmann & Robin 2007). Most starch granules show a Maltese cross under polarized light due to their crystallinity (Gallant et al. 1997), and starch is usually divided into amylose and amylopectin based on molecular differences.

1.1.2. Amylose and amylopectin

Most starch consists of two alternating semi-crystalline and amorphous layers. Amylose is the essential component in the amorphous layer, and composed of essentially linear molecules of α(1→4)-linked D-glucopyranosyl units with average chain length 250-670, while amylopectin is highly branched macromolecule with mostly α-(1→4) linked glucose and about 5 % α-1, 6 branch linkages, mainly existing in the semi-crystalline layer (Blennow et al. 2000; Kong 2009). Chains of amylopectin can be divided into three types (Figure 1.1) (Gunja Smith et al. 1970): A-chain, defined as one connected to the remainder of the molecules only through its reducing chain end; B-chain, defined as one connected to the remainder of the molecules but carries other A and/or B chains at 1

one or more of its primary hydroxyl groups; and C-chain, which carries the reducing end of the whole amylopectin in relation to cluster structure. Amylopectin has a weight-average molecular weight of 107-109 Da, much higher than amylose (105-106 Da) (Blennow et al. 2000; Kong 2009). The degree of polymerization (DP) of amylopectin ranges from 9600 – 15900, much higher than amylose (324 – 4920) (Kong 2009). Based on starch granule crystalline properties determined by X-ray powder diffraction, starches can be divided into A-type, B-type and C-type: starches with amylopectins of relatively short average branch chain length (DP 23-29) display the A-type pattern, e.g. waxy maize, maize, rice, wheat, taro, tapioca and sweet potato; starches with amylopectins long branch chains (DP 30-44) exhibits the B-type pattern, such as potato, canna, high amylose maize; the C pattern is a mixture of both A and B types but occurs naturally like smooth-seeded pea starch and various bean starches (Cheetham & Tao 1998; Jane et al. 1997). Usually the differences in starch or flour functionalities can be explained by differences in amylose and amylopectin ratio and structure. But the minor components like protein and lipid in starch granules also have potential to modify starch properties (Tester et al. 2004), such as pasting profiles (Fitzgerald et al. 2003). Other minor components like phosphate, associated with potato starch in the amorphous region, can lead to great impact on rheological properties, resulting in clearer gels, high viscosity, and slow rate and extent of retrogradation (Blennow et al. 2000; Buléon et al. 1998; Hoover 2001).

1.1.3. Swelling

When starch is heated in excess water, the starch granule swells to greatly increased volume, representing one special characteristic of starches (Sasaki & Matsuki 1998), and causes viscisoty 2

increase. Swelling power and swelling volume are widely accepted as two methods for evaluating swelling property. Flour or starch swelling properties are considered to highly correlate with corresponding noodle eating and texture qualities (e.g. softness, elasticity, smoothness) (Crosbie 1991; Crosbie & Lambe 1993; Crosbie et al. 1992).

1.1.4. Gelatinization

Starch gelatinization was defined as the collapse or disruption of the starch granule manifested in irreversible changes in properties such as granular swelling, native crystallite melting, loss of birefringence and solubilization in the presence of water (Atwell et al. 1988). Usually differential scanning calorimetry (DSC) is used to analyze starch gelatinization properties in excess water and parameters such as onset (To), peak (Tp), conclusion (Tc) and melting enthalpy (△H) are recorded for evaluation.

1.1.5. Retrogradation

A gelatinized starch suspension will undergo molecular interactions (mainly hydrogen bonding between starch chains) during cooling and form an elastic gel, and this process is called retrogradation (Hoover 2001). This behavior is considered to have a major influence on texture, acceptability and digestibility of starch-based foods (Biliaderis 1991). It has been related with quality for food products such as in breakfast cereals and parboiled rice (Karim et al. 2000), and is also important for food preservation such as shelf life extension in bread (No et al. 2007). Various

3

means such as DSC, rheometry, NMR, FT-IR, X-ray diffraction, and Raman spectroscopy can be used for starch retrogradation analysis (Kong 2009).

1.1.6. Gel texture

During retrogradation, starch gel texture parameters like firmness and rigidity change greatly, and large deformation tests by Texture Profile Analysis (TPA) are widely used to evaluate these changes in actual food and model starch gel systems (Karim et al. 2000), and provide various data such as hardness, adhesiveness, cohesiveness, gumminess and chewiness for predictions of final texture acceptance by consumer.

1.1.7. Pasting

Pasting is defined to be a phenomenon following gelatinization in the dissolution of starch, which involves granular swelling, exudation of molecular components from the granules, and eventually total disruption of the granules (Atwell et al. 1988). Involving heating, holding and cooling stages in one test, Rapid Viscosity Analyzer (RVA) is widely used and preferred for pasting viscosity measurements due to the relatively small amount of sample required (typically 2-3.5 g). During heating in water, starch granule begins to swell and form a paste, exhibited as quick viscosity increasing, and further continuous heating will lead to disruption of starch granules, resulting in viscosity decrease. When the hot starch paste is subject to a cooling process, molecular reassociation or retrogradation happens and cause viscosity increase again. Peak viscosity (maximum viscosity

4

during heating), hot paste viscosity (or trough, holding strength, minimum viscosity after peak viscosity), breakdown viscosity (the difference between peak viscosity and hot paste viscosity), cold paste viscosity (or final viscosity, the viscosity at the end of the test), and setback viscosity (the difference between cold paste viscosity and hot paste viscosity) are commonly recorded. Based on viscosity pattern classification defined by Schoch & Maywald (1968), starch can be divided into four types. Type A starches show high peak and breakdown viscosity, and usually have high swelling property like potato, tapioca starch. Type B starches have moderate-swelling property and exhibite lower peak and breakdown viscosity, e.g. normal cereal starches. Type C starches are restrict swelling starch and usually are chemically cross-bonded products. Its showed no pasting peak but significant high viscosity. Type D starches have restricted swelling property, e.g. high amylose corn starches, and do not show significant viscosity at normal concentrations.

1.1.8. Rheology

Starch, a low-cost polysaccharide, is commonly used in food industry as an important functional ingredient, e.g. thickening liquid foods such as soups, sauces and custards (Evans & Haisman 1980). A dynamic rheometer allows continuous assessment of dynamic moduli under either temperature or frequency sweeping on starch suspensions (Singh et al. 2003), and can provide rheological properties of a starch-water systems simulating conditions in industry (Thebaudin et al. 1998). Starch gel is formed by continuous amylose network and dispersed swollen granule remnants, and therefore its viscoelastic behaviors are influence by the level and nature of leached material, the volume fraction and deformability of swollen granules and the interactions between these 5

components (Waterschoot et al. 2015). Storage modulus (G'), loss modulus (G'') and loss tangent (tan δ, G'' / G') are used to describe these rheological changes in starch paste and gels. G' is defined as the energy recovered per cycle of deformation and reflect the elastic nature of the material. G'' exhibits the energy dissipated as heat per cycle of deformation and describes the viscous character of the material. tan δ is calculated by G'' / G', and liquid-like material tends to have high value of tan δ, while solid-like material has low value of tan δ.

1.1.9. Digestibility The digestion process of starch involves amylolytic enzymes, mainly consisting of pancreatic αamylase and intestinal brush border glucoamylases (maltase-glucoamylase and sucrose-isomaltase) (Zhang et al. 2006). Starch digestibility has gained a lot of interest recently, due to the recognition of the importance of incomplete digestion and absorption of starch in the small intestine (Sajilata et al. 2006). From a nutritional viewpoint, starch can be classified intro three categories by Englyst test (Englyst et al. 1992) depending on digestion rate and extent: rapidly digested starch (RDS, starch fraction hydrolyzed to glucose within 20 min), slowly digested starch (SDS, starch fraction hydrolyzed to glucose between 20 min and 120 min) and resistant starch (RS, starch fraction not hydrolyzed within 120 min) (Dona et al. 2010). Resistant starch is widely accepted to be divided into four types (Nugent 2005): RS1 (physically protected starch), RS2 (ungelatinized resistant starch granules with B-type crystallinity and hydrolyzed slowly by α-amylases), RS3 (retrograded starch) and RS4 (chemically modified starch). Potential health benefits and functional properties of RS has attracted a lot of researchers, and healthy physiological effects in clinic colonic cancer prevention, 6

hypoglycemic effects, hypocholesterolemic effects, and fat accumulation inhibition (Fuentes Zaragoza et al. 2010). Besides, SDS also showed advantages over RDS due to its stabilizing effect on blood glucose level, and may benefit for diabetes management (Lehmann & Robin 2007).

1.1.10 Functionality improvement

Starch has advantages over other materials when used in industry as a renewable and biodegradable resource (Ellis et al. 1998). However, native starches are not optimal in many situations, thus chemical modifications (like cross-linking and/or acetylation) are often involved in order to improve its performance (Jacobs & Delcour 1998). Gunaratne & Corke (2007b) found acid treatment on wheat, potaot and maize starches could probably lead to cleavage of amylopectin branch points, and strenghthend the realignment and self-association of macromolecules in starch granlues, exhibiting as increase on gelatinization temperatures and melting enthalpy but decrease on peak viscosity. However, chemical modification is likely to be related with poteinal harm. Therefore another feasible method, food additive addition have been widely applied for property modification, e.g. xanthan gum, guar gum and cellulose gum to wheat starch (Christianson et al. 1981). Xanthan gum could reduce the tapicoa starch granule swell by totally wrapped them, and increas setback viscosity, while guar gum did not and led to oppsite results (Chaisawang & Suphantharika 2006). Besides them, more naturally annealing and heat-moisture treatments have been widely investigated for a physical modification of starch properties, and make hydrothermallly treated starch suitable for utilisation in the canned and frozen food industries (Zavareze & Dias 2011). It can modify raw sweet potato starch with Type A pasting profile to Type C, while led to no significant impact on gel texture 7

(Collado & Corke 1999). Collado et al. (2001) tried bihon-type noodle prepration by using heatmoisiture treatment sweet potato starch, but did not find significant quality improvments compared with raw sweet potato noodles. They claimed the possiblitie of blending other starches for better noodle properties, which is more econnomic and an alternative way for deirable functionalities, and commonly used in production of (extruded) snacks (Waterschoot et al. 2015).

Starch blends have been reported as another way to acquire new starch properties (Waterschoot et al. 2015), and even relates to a pudding patent (Kern & Stute 1994). Many research confirmes that individual components in starch blends tend to gelatinize separately (Waterschoot et al. 2015), e.g. canna and mung bean starch (Puncha Arnon et al. 2008), maize and yam starch (Karam et al. 2006). When their peak temperatures determined by DSC curves have a difference less than 6 °C, the two separate individual peaks tend to merge into one; while on the other side, two peaks occur (Waterschoot et al. 2015). However, their pasting behavior can have both addtive and non-addtive effects and cannot be easily predicted. This indicates interactions do exist among different components in some cirumstances. For example, Gunaratne & Corke (2007a) found significant nonadditive pasting behavior between native amaranth and heat-moisture-treatment potato starch, and indicated interactions may exist between amylose leaching from the two starches, and influenced the swelling and granular break down properties. Puncha Arnon et al. (2008) reported an additive behavior in peak viscosity in canna-potato mixtures. Waterschoot et al. (2015) concluded that in starch blends with larger granule difference, the small granules may fill the voids between large granule and therfore reduce the swelling power by creating a more compact system, accouting for the non-addtive behavior, and this was supporeted by corn, cassava and yam starch mixtures (Karam 8

et al. 2006) and potato and maize starch blends (Park et al. 2009). For starch mixtures with similar granule sizes, a certain level of amylose content may be in need to reach high viscosity than expected value, and a small ratio of one starch adding into another would likely to exhibit an none-addtive behavior (Waterschoot et al. 2015), e.g. potato and rice starch (Sandhu et al. 2010), potato and amaranth starch (Gunaratne & Corke 2007a).

1.2. Wheat-based noodle 1.2.1. Wheat noodle

As a representative starch food, noodles are one of the most distinctive foods in oriental cultures. They are very popular in Asia and thought to originate from China (Bharath Kumar & Prabhasankar 2014). They have great product diversity, potentially long shelf life, and varied cooking methods with desirable tastes. Noodles based on wheat are usually prepared with flour, water and salt. Wheat flour noodles can be classified into two categories based on salt difference: yellow alkaline noodles (alkaline salt) or white salt noodles (NaCl). The latter one is predominant in China (Fu 2008). Besides, various processing methods have created an abundant diversity of noodles such as fresh noodles, dried noodles, steamed noodles, boiled noodles, steamed and deep-fried noodles (Fu 2008). Almost 40 % of the wheat consumption in Asia is due to noodle production (Bharath Kumar & Prabhasankar 2014). Consisting of gliadins and glutenins, gluten protein comprises around 80 % of the total protein in wheat flour, responsible for wheat dough extensibility and strength (Kovacs et al. 2004), and both protein content and quality of protein can have significant relationships with processing and textural properties of white salted noodles (Park et al. 2003). Wheat flour with low 9

amylose content (＜ 23 %) tend to produce noodles with soft texture and high cohesiveness (Baik & Lee 2003). Pasting and swelling properties of wheat flour could explain most quality variations in white salted noodle made by 37 Australian flour samples, and were suggested as selection criteria for breeding programs (Yun & Moss 1996). Sweet potato, potato, or waxy corn starch additions can contribute to white salted noodle improvement (Chen et al. 2003b).

1.2.2. Buckwheat noodle

Buckwheat, a sharp, three-sided seed with dark brown color (Fu 2008), belongs to the family Polygonaceae (Fu 2008; Wei 1995). Originating from China, buckwheat is an important food plant mostly in Asia and Europe (Lin et al. 2005), and especially in developing countries lacking protein in the food (Lin et al. 2005). It is also a listed plant in traditional Chinese medicine (Lin et al. 2005), and can help maintain good blood circulation, and also works as an effective preventive measure against high blood pressure (Fu 2008; Lin et al. 2005). These benefits are mainly due to the valuable flavonoids like rutin, present in buckwheat but absent in cereals (Fu 2008; Krkošková & Mrazova 2005). Buckwheat groats contain a significant amount of resistant starch and thus may help prevent colon cancer (Skrabanja et al. 2004). It is a good dietary supplier for essential nutrients due to its high protein and mineral level (Ikeda & Asami 2000). Compared with other grain species, buckwheat is rich in minerals such as Fe, Ca, P, Cu, Zn, Mg, B, I, Ni, Co, Se (Lin et al. 2005).

Buckwheat noodle is a popular food in Japan, Korea and China (Fu 2008; Krkošková & Mrazova 2005; Li & Zhang 2001; Ma et al. 2013). It has a typical name ‘soba noodles’ in Japan. Buckwheat 10

noodles are often prepared without any additives (Ikeda & Asami 2000). However, as buckwheat does not contain any gluten, it cannot form a consistent dough alone, so buckwheat noodle production is usually made by blending buckwheat flour with appropriate percentage of wheat flour, which provides gluten as the binder. Sometimes it also involves partial gelatinization of buckwheat flour to increase its proportion by reducing reliance on gluten. Shibata et al. (2011) established a rapid method for buckwheat proportion prediction in buckwheat noodles based on fluorescence fingerprint. When mixed with water, the rutin in buckwheat flour is easily lost due to rutin-degrading enzymes, and hydrothermal treatments (steaming and autoclaving) had potential to prevent this reaction and needs further investigation on its effect on sensory property of final products before use in practice (Yoo et al. 2012). Generally the high dietary fiber content, nutrients and minerals in buckwheat make buckwheat noodle a healthy food (Ma et al. 2013).

1.2.3. Millet noodle

Millets are a group of small-seeded grasses with many varieties, and are popularly known as coarse cereals. Millet is the 6th cereal crop in world agricultural production, and has good resistance to drought, pest and disease, and a short growing season compared with other cereals (Saleh et al. 2013). In many African and Asian areas, millets serve as an important food component and are made into various food products like breads, porridges and snacks foods, while in North America and Europe, people are more attracted by their importance as an ingredient in multigrain and gluten-free cereal products (Chandrasekara & Shahidi 2011; Saleh et al. 2013). Nutritional value of foods is now a vital part of attracting consumers’ interest. Millets were found to have high nutritive value 11

and comparable to that of major cereals like wheat and rice (Parameswaran & Sadasivam 1994; Shinoj et al. 2006). Its grains are good sources of phytochemicals, micronutrients and also essential amino acids especially high in methionine (Saleh et al. 2013).

Millet noodle has a long history. Archaeological findings of northwest China in Lajia demonstrates that the conversion of millet flour into long, thin noodle strands can be dated back to 4000 years ago (Lu et al. 2005). However, due to the absence of gluten, flours of millets including foxtail and broomcorn millets can’t be made into noodles simply by mixing it with water (Ge et al. 2011). Thus it is often blended with wheat flours with appropriate mixing ratio to form noodles. Since people are becoming health conscious by having high fiber and low fat content in their diet, noodles with millets can provide such nutritional value and are thus receiving more attention (Vijayakumar et al. 2010). 30% blend of finger millet with refined wheat flour showed high nutritional value and hypoglycemic effects (Shukla & Srivastava 2014).

1.3. Starch noodle 1.3.1. Introduction

Asian noodles are not made exclusively of wheat or wheat-based formulations, many are made from isolated starches (Fu 2008). Starch noodles, as a traditional food, are widely welcomed in many Asian countries especially in China. In the absence of gluten, pre-gelatinized starch is often used as a binder mixed with ungelatinized starch to facilitate extrusion or sheeting to produce the noodle (Collado & Corke 1997). Thus processing steps for starch noodle preparation are rather similar for 12

all types. Starch noodles are often produced by blending dry starch with appropriate amount of water, with partial pre-gelatinized starch as a binder, to form a slurry or dough, then extruding it directly into boiling water to cook, cooling the produced noodles strands in cold water, transferring them to refrigerated or freezing environment, then drying them before packaging. Thus unlike wheat-based noodles, starch noodle qualities are highly related with corresponding starch functionalities. Starch noodles can offer customer certain nutritional benefits due to resistant starch formation as a result of retrogradation. Starches from potato and tapioca are widely used as a texture enhancing ingredient for noodles (Fu 2008).

1.3.2. Mung bean noodle

Mung bean has been mostly grown in Asian countries but is now also widely cultivated in Africa, South America, Australia, and the United States (Kim et al. 2007). Mung bean starch is widely known to be an excellent raw material for starch noodle preparation (Collado et al. 2001; Thao 2011). It has been processed into clear noodles with white color, smooth and pliable texture, and good cooking quality (Kasemsuwan et al. 1998; Zhu et al. 1990). Mung bean noodle has high tensile strength, a distinctive chewy elastic texture and low cooking loss even with prolonged cooking (Collado et al. 2001; Fu 2008). These textural characteristics are closely related with the unique starch property, and mung bean starch has very high amylose content and possesses restricted swelling and pasting qualities similar to chemically cross-linked starch as a Type C starch, based on viscosity pattern classification by Schoch & Maywald (1968) (Collado et al. 2001; Tan et al. 2006). The X-ray diffraction results indicated that crystallites within granules of mung bean starch were 13

more compactly packed than other legumes starches and mung bean starch had a C-type pattern (Hoover et al. 1997). Kim et al. (2007) reported amylose contents of five mung bean cultivars ranged from 31.7 to 33.8 % by iodine method. However, mung bean starch is expensive due to tedious processing methods and low production supply (Chen et al. 2003a; Kasemsuwan et al. 1998), and this limits its practical applications.

1.3.3. Sweet potato noodle

Sweet potato has long been one of the principal starch source in China dating from late 16th century (Tan et al. 2006; Zhu et al. 2011). Sweet potato production in China is 90 % of worldwide total and is mainly used for animal feed and human consumption (Zhu et al. 2011). It has great potential as a source of starch due to its various agronomic advantages over other root crops (Collado et al. 2001). Accounting for 50-80 % of root dry matter, starch in sweet potato, like other root crop starches such as cassava, potato, and arrowroot, is considered more free swelling, non-congealing and exhibits a Type A Brabender amylograph (moderate to high peak viscosity with a major breakdown and low cold paste viscosity with respect to the PV) (Collado et al. 2001), based on viscosity pattern classification (Schoch & Maywald 1968). According to X-ray results sweet potato starch had an Atype pattern but with fewer A-chains (Hanashiro et al. 1996; Jane et al. 1997), or intermediate between ‘A’ and ‘C’ (Moorthy 2002; Noda et al. 1996). Collado et al. (1999) reported a wide amylose range of isolated sweet potato starch from 14.0 to 29.7 % in 44 varieties. Sweet potato starch is polygonal or almost round in shape with a centric distinct hilum (Moorthy 2002; Zhu et al. 2011). Compared with mung bean starch noodles, qualities of sweet potato starch noodle are 14

considered not good with dull opaque appearance, moderately elasticity and high swelling and cooking loss (Tan et al. 2006). Chen et al. (2002) found that the color of sweet potato starch noodle (evaluated by sensory panel) had no correlations with starch color and paste clarity. Gel properties instead of pasting and gelatinization properties were considered more suitable for predicting sweet potato starch noodle qualities (Chen et al. 2002; Chen & Voragen 2003a). Compared with traditional sweet potato starch noodle, whose manufacturing is time-consuming with a freeze-thaw cycle, nonfrozen processing method is easier and acceptable, leading to softer and stickier noodle structure with limited impact on cooking properties (Lee et al. 2005).

1.3.4. Potato noodle

Potato has long been a principal food in many areas and recently China has pushed potato to be an alternative staple food after rice, wheat and corn. Differing from other starches, potato starch granule size is typically larger than others. The high phosphorus content leads to a relatively lower amylose content but very high viscosity and swelling power in potato starch (Zaidul et al. 2007). The majority of potato starch granules were flattened ellipsoid with few spherical (Mishra & Rai 2006). There were also significant differences in potato starch granule size and shape from small or oval to irregular or cuboidal as observed among five genotypes in Singh & Singh (2001) due to genetic variation. X-ray diffraction results suggested that crystalline structure of potato starch had a B-type pattern, which means more space available for water compared with A-type (Cheetham & Tao 1998; Hibi et al. 1993). Potato starch is a good raw material for starch noodle making, having neutral taste, much higher transparency and elasticity than cereal starches (Fu 2008). Potato starch are considered 15

more transparent than bean starch when made into noodles (Kim et al. 1996), and are reported to have potential to improve the quality of instant noodles (Noda et al. 2006). Adding rice starch to potato starch could lead to a lower gel hardness and cohesiveness, but resulting in higher cooking time and less cooking loss in corresponding starch noodles, and an equal proportion of mixed starches (1:1) showed the best sensory scores closing to commercial noodle sample (Sandhu et al. 2010). Singh et al. (2002) reported potato starch with large granule size and higher amylose content exhibited the best ability for noodle making. Substitution of up to 17 % with phosphorylated tapioca starch or up to 35 % with commercial phosphorylated tapioca starch could result in better potato starch noodle qualities, with higher strength of uncooked noodles, and lower stickiness and cooking loss (Muhammad et al. 1999).

1.3.5. Rice noodle

Not only consumed as cooked grains, rice can also be made into rice noodle as a traditional oriental food widely consumed in Southeast Asia (Bhattacharya et al. 1999). Rice protein lacks the functionalities of wheat gluten to avoid dough cracking in preparation for noodles. Thus for rice noodle production, usually part of the rice flour is pregelatinized to serve as a binder for the remaining flour. Excess pregelatinization of rice flour may not improve rice noodle quality, as increasing pregelatinized flour from 7 % to 10-15 % did not further improve rice noodle cooking qualities (Bhattacharya et al. 1999). At high pregelatinization ratio (15-30 %), Yalcin & Basman (2008) suggested rice noodle made by 25 % pregelatinized starch had the best cooking and sensory properties. It is reported that rice noodles with 7 % pregelatinized flour had much better quality than 16

100 % gelatinized raw flour (Resmini et al. 1979). Amylose content is generally recognized as one of the most important determinants for various rice products (Juliano 1998), and rice is usually divided into five groups based on amylose content: waxy (0-5 %), very low (5-12 %), low (12-20 %), intermediate (20-25 %) and high (25-33 %) (Biselli et al. 2014; Juliano 1992; Suwannaporn et al. 2007). Bao et al. (2006) reported a wide range of amylose contents in 577 rice accessions from 7.9 to 32.6 %. Amylose content, as well as other physicochemical properties such as swelling volume, RVA pasting and gel texture properties correlate highly with the cooking and textural properties of rice noodles, and are suggested to be used as indicators for rice noodle manufacturing (Bhattacharya & Corke 1999; Han & Koh 2011; Hormdok & Noomhorm 2007; Yalcin & Basman 2008), while rice protein had no impact on rice noodle quality (Han & Koh 2011). Traditionally rice noodles are made from long-grain rice with high or medium amylose content (＞22 % amylose) (Bhattacharya et al. 1999; Fu 2008), and high amylose content is preferable as it tends to give rice noodles with lower adhesiveness, leading to a desired clean and smooth texture (Han & Koh 2011). Special additives or treatments are widely investigated in rice starch aimed at improving noodle nutritional and eating qualities. Annealing and heat-moisture treatment on rice flour can result in good noodle textural qualities similar to commercial noodles, as a safe and completely natural product (Hormdok & Noomhorm 2007; Cham & Suwannaporn 2010). Natural fermentation of rice at 35 °C for 27 hours can lead to a better appearance and mouth feel when made into rice noodles, with limited effect on physicochemical properties of starches (Lu et al. 2003). Addition of transglutaminase improved dough machining ability, resulting in smoother noodle surface (Yalcin & Basman 2008), while Kim et al. (2014) demonstrated that further addition of rice protein isolated (RPI) to rice noodle with transglutaminase could build more extensive protein network capable for holding starch 17

components in the dough, and provide additional nutrition besides cooking and textural qualities. Unripe banana, edible canna, taro flours addition to rice noodle can lower digestibility by in vitro test (Srikaeo et al. 2011).

1.3.6. Sorghum noodle

Sorghum is the principal staple food of people in the semiarid regions of Africa and Asia (Beta & Corke 2001a). The traits of low-input cost and adaptability to a wide range of environments make sorghum a favorable potential candidate for various food and nonfood products (Zhu 2014). Amylose content of sorghum starch varies a lot among different genotypes, e.g. 24-33 % in 95 Zimbabwean sorghum landraces (Beta et al. 2001) and 16.1-55.8 % in 55 sorghum genotypes (Hill et al. 2012). Sorghum starch extracted from kernels exhibited an A-type crystallinity based on Xray diffraction (Shin et al. 2004; Zhu 2014). Sorghum starch exhibited a two-stage (namely, initial rapid, then restricted and rapid again) swelling and solubilization pattern in both normal and waxy genotypes, very different from other starches (Leach et al. 1959; Zhu 2014). The advantages of strong harsh environment tolerance, great competitiveness against weeds, and low-input cost of cultivation over other crops make sorghum a potential cheap supply of starches (Zhu 2014). Besides, sorghum has potential for starch noodle production due to a low cooking loss (Beta & Corke 2001b), and good control of sorghum grain and flour quality characteristics contributes to sorghum-based Chinese egg noodles with better physical attributes (Liu et al. 2012). Sorghum starch noodle elasticity varied great among different genotypes, and was highly related to starch pasting properties (hot paste viscosity and cold paste viscosity) (Beta & Corke 2001a). 18

1.4. Glycemic index

In addition to food evaluation by various nutrient contents, glycemic index (GI) concept was first introduced to classify foods based on their effect on postprandial glycaemia (Jenkins et al. 1981). GI is defined as the incremental blood glucose area following the test food, expressed as the percentage of the corresponding area following a carbohydrate equivalent load of a reference product (e.g. white bread) (Bharath Kumar & Prabhasankar 2014). GI ranges from less than 20 % to approximately 120 % using white bread as a reference (Bharath Kumar & Prabhasankar 2014). Food can be classified into high GI (GI ≥ 70) food, medium GI (56 – 69) food and low GI (≤ 55) food (Bharath Kumar & Prabhasankar 2014). GI has particular relevance to those chronic Western diseases associated with obesity and insulin resistance (Jenkins et al. 2002), and low GI diets are thought to contribute to antihyperglycemic medications of patients with type 2 diabetes by improving glycemic control (Jenkins et al. 2008). In addition, it may also relate to prevention of diet related cancer such as in colon, breast and prostate (Giovannucci 1999; Jenkins et al. 2002; McKeown Eyssen 1994). Avoiding high cost and time consuming in vivo tests for glycemic index determinations, various in vitro models are used simulating in vivo digestion and predict glycemic index effectively due to high correlations between them (Araya et al. 2002; Bharath Kumar & Prabhasankar 2014; Goni et al. 1997). Most of wheat products are known to have high glycemic index (Bharath Kumar & Prabhasankar 2014).

19

1.5. Texture evaluation.

To attract peoples’ eyes, nowadays various fascinating color and packaging are involved in abundant foods including noodles. Besides these surface decorations, the key points should be fantastic taste and texture feelings serving customers’ months. Texture feelings comes from composite sensory receptors lying in and around the mouth (Lucas 2004; Szczesniak 2002), and thus it is difficult for instrumental experiments to accurate reflect the true feelings, and sensory panel test involved of true human beings is always proved to be the most convincible way for consumer response predictions (Li et al. 2012). However, its great cost and time consuming limit its application mainly on testing small amount of food samples. Therefore cheap and easy-to-handle mechanical study remains as the major means used for texture evaluation. Besides the common two-cycle compressing program in TPA, wire cutting, peeling and indentation tests have also been previoulsy reported as three instrumental measuremnt on textures, e.g. wire cutting on cheese (Kamyab et al. 1998); peeling on flexible laminates (Kinloch et al. 1994); indentation on gels (Hu et al. 2010) and biological tissues (McKee et al. 2011), and can be used for evluation on cohesiveness fracture energy (toughness), adhesiveness fracture energy and elastic property (relaxation ratio), respectively. These attributes are important mechanical properties relating to eating characteristics (Briscoe et al. 1994; Gamonpilas et al. 2010; Gonzalez Gutierrez 2008; Kendall 2001).

1.6. Introduction and objectives of this study

In the past years our lab has completed various research on starch related fields: cereal grains like rice (Tan & Corke 2002); starches from abundant sources, e.g. sweet potato starch (Zhu et al. 2011), 20

sorghum starches (Beta & Corke 2001a), Amaranthus starch (Kong et al. 2009); noodles such as yellow alkaline and white salt noodles (Sui et al. 2006) and starch noodles (Collado & Corke 1997). As a potential economically and healthy way to achieve properties of chemically modified starch, blending of different starches has also been previously investigated in our lab, such as wheat and potato starch mixtures (Zhu & Corke 2011). However, these research achievements have been little extended to real food qualities such as in noodles. Mung bean starch is an excellent starch source but expensive. Few researchers investigate the possibilities on mung bean starch and noodle quality improvements or changes by other starch addition. Besides, understandings of impacts about different buckwheat and millet mixing ratio on wheat flour and noodle qualities are in need for their applications in foods. Noodle texture is an important property, however, mechanical methods like wire cutting, peeling and indentation tests are difficult to provide useful noodle texture data due to thin thickness and narrow width, and few researchers try the idea changing research objects from noodle strands to sheets to fit criteria for these tests. Therefore this study aims at further broadening above research area by assessing starch qualities in food products such as various noodles, and with the following objectives: (1) To investigate property changes of buckwheat and addition on wheat flour qualities, and possible explanations. (2) To further study the impact of granule size on the additive and non-additive behaviors of starch blends by addition of other starches into mung bean starch. (3) To check the impacts of flour and starch blends on qualities of corresponding noodles. (4) To verify the feasibility of texture evaluation methods on thin noodle sheets, including wire cutting, peeling and indentation tests. 21

Figure 1.1 Four graphs for amylopectin structure description in Buléon et al. (1998). (1): A-chain and B-chain. (2 & 3): A chain, B-chain, and C-chain (the one has reducing end) (4): S represents single chain (A-chain), L represent B-chain.

22

References

Araya, H., Contreras, P., Alvina, M., Vera, G. and Pak, N. (2002). A comparison between an in vitro method to determine carbohydrate digestion rate and the glycemic response in young men. European Journal of Clinical Nutrition, 56, 735-739. Atwell, W., Hood, L., Lineback, D., Varriano-Marston, E. and Zobel, H. (1988). The terminology and methodology associated with basic starch phenomena. Cereal Foods World, 33, 306311. Baik, B. K. and Lee, M. R. (2003). Effects of starch amylose content of wheat on textural properties of white salted noodles. Cereal Chemistry, 80, 304-309. Bao, J., Shen, S., Sun, M. and Corke, H. (2006). Analysis of genotypic diversity in the starch physicochemical properties of nonwaxy rice: apparent amylose content, pasting viscosity and gel texture. Starch‐Stärke, 58, 259-267. Beta, T. and Corke, H. (2001a). Genetic and environmental variation in sorghum starch properties. Journal of Cereal Science, 34, 261-268. Beta, T. and Corke, H. (2001b). Noodle quality as related to sorghum starch properties. Cereal Chemistry, 78, 417-420. Beta, T., Obilana, A. B. and Corke, H. (2001). Genetic diversity in properties of starch from Zimbabwean sorghum landraces. Cereal Chemistry, 78, 583-589. Bharath Kumar, S. and Prabhasankar, P. (2014). Low glycemic index ingredients and modified starches in wheat based food processing: A review. Trends in Food Science & Technology, 35, 32-41.

23

Bhattacharya, M., Zee, S. and Corke, H. (1999). Physicochemical properties related to quality of rice noodles. Cereal Chemistry, 76, 861-867. Biliaderis, C. G. (1991). The structure and interactions of starch with food constituents. Canadian Journal of Physiology and Pharmacology, 69, 60-78. Biselli, C., Cavalluzzo, D., Perrini, R., Gianinetti, A., Bagnaresi, P., Urso, S., Orasen, G., Desiderio, F., Lupotto, E. and Cattivelli, L. (2014). Improvement of marker-based predictability of apparent amylose content in japonica rice through GBSSI allele mining. Rice, 7, 1. Blennow, A., Bay Smidt, A. M., Olsen, C. E. and Møller, B. L. (2000). The distribution of covalently bound phosphate in the starch granule in relation to starch crystallinity. International Journal of Biological Macromolecules, 27, 211-218. Briscoe, B. J., Sebastian, K. S. and Adams, M. J. (1994). The effect of indenter geometry on the elastic response to indentation. Journal of Physics D: Applied Physics, 27, 1156.

Buléon, A., Colonna, P., Planchot, V. and Ball, S. (1998). Starch granules: structure and biosynthesis. International Journal of Biological Macromolecules, 23, 85-112. Chaisawang, M. and Suphantharika, M. (2006). Pasting and rheological properties of native and anionic tapioca starches as modified by guar gum and xanthan gum. Food Hydrocolloids, 20, 641-649. Cham, S. and Suwannaporn, P. (2010). Effect of hydrothermal treatment of rice flour on various rice noodles quality. Journal of Cereal Science, 51, 284-291. Chandrasekara, A. and Shahidi, F. (2011). Determination of antioxidant activity in free and hydrolyzed fractions of millet grains and characterization of their phenolic profiles by HPLC-DAD-ESI-MS. Journal of Functional Foods, 3, 144-158. 24

Cheetham, N. W. H. and Tao, L. (1998). Variation in crystalline type with amylose content in maize starch granules: an X-ray powder diffraction study. Carbohydrate Polymers, 36, 277-284. Chen, Z., Sagis, L., Legger, A., Linssen, J. P. H., Schols, H. A. and Voragen, A. G. J. (2002). Evaluation of starch noodles made from three typical Chinese sweet-potato starches. Journal of Food Science, 67, 3342-3347. Chen, Z., Schols, H. and Voragen, A. (2003a). Physicochemical properties of starches obtained from three varieties of Chinese sweet potatoes. Journal of Food Science, 68, 431-437. Chen, Z., Schols, H. and Voragen, A. (2003b). The use of potato and sweet potato starches affects white salted noodle quality. Journal of Food Science, 68, 2630-2637. Christianson, D., Hodge, J., Osborne, D. and Detroy, R. W. (1981). Gelatinization of wheat starch as modified by xanthan gum, guar gum, and cellulose gum. Cereal Chemistry, 58, 513-517. Collado, L. S. and Corke, H. (1997). Properties of starch noodles as affected by sweetpotato genotype. Cereal Chemistry, 74, 182-187. Collado, L. S. and Corke, H. (1999). Heat-moisture treatment effects on sweetpotato starches differing in amylose content. Food Chemistry, 65, 339-346. Collado, L. S., Mabesa, R. C. and Corke, H. (1999). Genetic variation in the physical properties of sweet potato starch. Journal of Agricultural and Food Chemistry, 47, 4195-4201. Collado, L. S., Mabesa, L. B., Oates, C. G. and Corke, H. (2001). Bihon-type noodles from heatmoisture-treated sweet potato starch. Journal of Food Science, 66, 604-609. Crosbie, G. B. (1991). The relationship between starch swelling properties, paste viscosity and boiled noodle quality in wheat flours. Journal of Cereal Science, 13, 145-150. Crosbie, G. B. and Lambe, W. J. (1993). The application of the flour swelling volume test for 25

potential noodle quality to wheat breeding lines affected by sprouting. Journal of Cereal Science, 18, 267-276. Crosbie, G. B., Lambe, W. J., Tsutsui, H. and Gilmour, R. F. (1992). Further evaluation of the flour swelling volume test for identifying wheats potentially suitable for Japanese noodles. Journal of Cereal Science, 15, 271-280. Dona, A. C., Pages, G., Gilbert, R. G. and Kuchel, P. W. (2010). Digestion of starch: in vivo and in vitro kinetic models used to characterise oligosaccharide or glucose release. Carbohydrate Polymers, 80, 599-617. Ellis, R. P., Cochrane, M. P., Dale, M. F. B., Duffus, C. M., Lynn, A., Morrison, I. M., Prentice, R. D. M., Swanston, J. S. and Tiller, S. A. (1998). Starch production and industrial use. Journal of the Science of Food and Agriculture, 77, 289-311. Englyst, H. N., Kingman, S. and Cummings, J. (1992). Classification and measurement of nutritionally important starch fractions. European Journal of Clinical Nutrition, 46, S3350. Evans, I. D. and Haisman, D. R. (1980). Rheology of gelatinised starch suspensions. Journal of Texture Studies, 10, 347-370. Fitzgerald, M., Martin, M., Ward, R., Park, W. and Shead, H. (2003). Viscosity of rice flour: a rheological and biological study. Journal of Agricultural and Food Chemistry, 51, 22952299. Fu, B. X. (2008). Asian noodles: history, classification, raw materials, and processing. Food Research International, 41, 888-902. Fuentes Zaragoza, E., Riquelme Navarrete, M. J., Sánchez Zapata, E. and Pérez Álvarez, J. A. 26

(2010). Resistant starch as functional ingredient: a review. Food Research International, 43, 931-942. Gallant, D. J., Bouchet, B. and Baldwin, P. M. (1997). Microscopy of starch: evidence of a new level of granule organization. Carbohydrate Polymers, 32, 177-191. Gamonpilas, C., Charalambides, M. N., Williams, J. G., Dooling, P. J. and Gibbon, S. R. (2010). On the study of wire cutting of starch gels. In 12th International Conference on Fracture.Coference. Curran Associates, Inc.

Ge, W. E. I., Liu, L. I., Chen, X. and Jin, Z. (2011). Can noodles be made from millet? An experimental investigation on noodles manufacture together with starch grain analysis. Archaeometry, 53, 194-204. Giovannucci, E. (1999). Insulin-like growth factor-I and binding protein-3 and risk of cancer. Hormone Research in Paediatrics, 51, 34-41. Goni, I., Garcia Alonso, A. and Saura Calixto, F. (1997). A starch hydrolysis procedure to estimate glycemic index. Nutrition Research, 17, 427-437. Gunaratne, A. and Corke, H. (2007a). Gelatinizing, pasting, and gelling properties of potato and amaranth starch mixtures. Cereal Chemistry, 84, 22-29. Gunaratne, A. and Corke, H. (2007b). Influence of prior acid treatment on acetylation of wheat, potato and maize starches. Food Chemistry, 105, 917-925. Gunja Smith, Z., Marshall, J., Mercier, C., Smith, E. and Whelan, W. (1970). A revision of the Meyer-Bernfeld model of glycogen and amylopectin. FEBS letters, 12, 101-104. Han, H. M., Cho, J. H. and Koh, B. K. (2011). Processing properties of Korean rice varieties in relation to rice noodle quality. Food Science and Biotechnology, 20, 1277-1282. 27

Hanashiro, I., Abe, J. I. and Hizukuri, S. (1996). A periodic distribution of the chain length of amylopectin as revealed by high-performance anion-exchange chromatography. Carbohydrate Research, 283, 151-159. Hibi, Y., Matsumoto, T. and Hagiwara, S. (1993). Effect of high pressure on the crystalline structure of various starch granules. Cereal Chemistry, 70, 671-671. Hill, H., Lee, L. S. and Henry, R. J. (2012). Variation in sorghum starch synthesis genes associated with differences in starch phenotype. Food Chemistry, 131, 175-183. Hoover, R. 2001. Composition, molecular structure, and physicochemical properties of tuber and root starches: a review. Carbohydrate Polymers, 45, 253-267. Hoover, R., Li, Y. X., Hynes, G. and Senanayake, N. (1997). Physicochemical characterization of mung bean starch. Food Hydrocolloids, 11, 401-408. Hormdok, R. and Noomhorm, A. (2007). Hydrothermal treatments of rice starch for improvement of rice noodle quality. LWT - Food Science and Technology, 40, 1723-1731. Hu, Y., Zhao, X., Vlassak, J. J. and Suo, Z. (2010). Using indentation to characterize the poroelasticity of gels. Applied Physics Letters, 96, 121904. Ikeda, K. and Asami, Y. (2000). Mechanical characteristics of buckwheat noodles. Fagopyrum, 17, 67-72. Jacobs, H. and Delcour, J. A. (1998). Hydrothermal modifications of granular starch, with retention of the granular structure:  a review. Journal of Agricultural and Food Chemistry, 46, 28952905. Jane, J. L., Wong, K. S. and McPherson, A. E. (1997). Branch-structure difference in starches of Aand B-type X-ray patterns revealed by their Naegeli dextrins. Carbohydrate Research, 300, 28

219-227. Jenkins, D. J., Kendall, C. W., Augustin, L. S., Franceschi, S., Hamidi, M., Marchie, A., Jenkins, A. L. and Axelsen, M. (2002). Glycemic index: overview of implications in health and disease. The American Journal of Clinical Nutrition, 76, 266S-273S. Jenkins, D. J., Kendall, C. W., McKeown Eyssen, G., Josse, R. G., Silverberg, J., Booth, G. L., Vidgen, E., Josse, A. R., Nguyen, T. H. and Corrigan, S. (2008). Effect of a low–glycemic index or a high–cereal fiber diet on type 2 diabetes: a randomized trial. The Journal of the American Medical Association, 300, 2742-2753. Jenkins, D. J., Wolever, T., Taylor, R. H., Barker, H., Fielden, H., Baldwin, J. M., Bowling, A. C., Newman, H. C., Jenkins, A. L. and Goff, D. V. (1981). Glycemic index of foods: a physiological basis for carbohydrate exchange. The American Journal of Clinical Nutrition, 34, 362-366. Juliano, B. (1992). Structure, chemistry, and function of the rice grain and its fractions. Cereal Foods World, 37, 772-774. Juliano, B. (1998). Varietal impact on rice quality. Cereal Foods World, 43, 207-222. Kamyab, I., Chakrabarti, S. and Williams, J. G. (1998). Cutting cheese with wire. Journal of Materials Science, 33, 2763-2770. Karam, L. B., Ferrero, C., Martino, M. N., Zaritzky, N. E. and Grossmann, M. V. E. (2006). Thermal, microstructural and textural characterisation of gelatinised corn, cassava and yam starch blends. International Journal of Food Science & Technology, 41, 805-812. Karim, A. A., Norziah, M. H. and Seow, C. C. (2000). Methods for the study of starch retrogradation. Food Chemistry, 71, 9-36. 29

Kasemsuwan, T., Bailey, T. and Jane, J. (1998). Preparation of clear noodles with mixtures of tapioca and high-amylose starches. Carbohydrate Polymers, 36, 301-312. Kendall, K. (2001). Molecular adhesion and its applications: the sticky universe. Kluwer Academic/Plenum Publishers: New York. Kern, H. and Stute, R. (1994). Starch mixtures as pudding starches. U.S. Patent (5,324,532).

Kim, S. H., Lee, B. H., Baik, M. Y., Joo, M. H. and Yoo, S. H. (2007). Chemical structure and physical properties of mung bean starches isolated from 5 domestic cultivars. Journal of Food Science, 72, C471-C477. Kim, Y., Kee, J. I., Lee, S. and Yoo, S. H. (2014). Quality improvement of rice noodle restructured with rice protein isolate and transglutaminase. Food Chemistry, 145, 409-416. Kim, Y. S., Wiesenborn, D. P., Lorenzen, J. H. and Berglund, P. (1996). Suitability of edible bean and potato starches for starch noodles. Cereal Chemistry, 73, 302-308. Kinloch, A. J., Lau, C. C. and Williams, J. G. (1994). The Peeling of Flexible Laminates. International Journal of Fracture, 66, 45-70. Kong, X. (2009). Molecular structure and functional properties of Amaranthus starch. (Ph.D. thesis). The University of Hong Kong: Hong Kong, China. pp 1-16. Kong, X., Bao, J. and Corke, H. (2009). Physical properties of Amaranthus starch. Food Chemistry, 113, 371-376. Kovacs, M. I. P., Fu, B. X., Woods, S. M. and Khan, K. (2004). Thermal stability of wheat gluten protein: its effect on dough properties and noodle texture. Journal of Cereal Science, 39, 919. Krkošková, B. and Mrazova, Z. (2005). Prophylactic components of buckwheat. Food Research 30

International, 38, 561-568. Leach, H. W., McCowen, L. D. and Schoch, T. J. (1959). Structure of the starch granule. I. Swelling and solubility patterns of various starches. Cereal Chemistry, 36, 534-544. Lee, S. Y., Woo, K. S., Lim, J. K., Kim, H. I. and Lim, S. T. (2005). Effect of processing variables on texture of sweet potato starch noodles prepared in a nonfreezing process. Cereal Chemistry, 82, 475-478. Lehmann, U. and Robin, F. (2007). Slowly digestible starch – its structure and health implications: a review. Trends in Food Science & Technology, 18, 346-355. Li, P. H., Huang, C. C., Yang, M. Y. and Wang, C. C. R. (2012). Textural and sensory properties of salted noodles containing purple yam flour. Food Research International, 47, 223-228. Li, S. and Zhang, Q. H. (2001). Advances in the development of functional foods from buckwheat. Critical Reviews in Food Science and Nutrition, 41, 451-464. Lin, R., Zhou, X., Ren, G., Bian, J. and Shan, F. (2005). Production and trading of buckwheat in China, nutrition and food. Food Science, 26, 259-263. Liu, L., Herald, T. J., Wang, D., Wilson, J. D., Bean, S. R. and Aramouni, F. M. (2012). Characterization of sorghum grain and evaluation of sorghum flour in a Chinese egg noodle system. Journal of Cereal Science, 55, 31-36. Lu, H., Yang, X., Ye, M., Liu, K. B., Xia, Z., Ren, X., Cai, L., Wu, N. and Liu, T. S. (2005). Culinary archaeology: millet noodles in late neolithic China. Nature, 437, 967-968. Lu, Z. H., Li, L. T., Cao, W., Li, Z. G. and Tatsumi, E. (2003). Influence of natural fermentation on physico‐chemical characteristics of rice noodles. International Journal of Food Science & Technology, 38, 505-510. 31

Lucas, P. W. (2004). Dental functional morphology: how teeth work. Cambridge University Press. Szczesniak, A. S. (2002). Texture is a sensory property. Food Quality and Preference, 13, 215-225. Ma, Y. J., Guo, X. D., Liu, H., Xu, B. N. and Wang, M. (2013). Cooking, textural, sensorial, and antioxidant properties of common and tartary buckwheat noodles. Food Science and Biotechnology, 22, 153-159. McKee, C. T., Last, J. A., Russell, P. and Murphy, C. J. (2011). Indentation versus tensile measurements of Young's modulus for soft biological tissues. Tissue Engineering Part B: Reviews, 17, 155-164. McKeown Eyssen, G. (1994). Epidemiology of colorectal cancer revisited: are serum triglycerides and/or plasma glucose associated with risk?. Cancer Epidemiology Biomarkers & Prevention, 3, 687-695. Mishra, S. and Rai, T. (2006). Morphology and functional properties of corn, potato and tapioca starches. Food Hydrocolloids, 20, 557-566. Moorthy, S. N. (2002). Physicochemical and functional properties of tropical tuber starches: a review. Starch‐Stärke, 54, 559-592. Muhammad, K., Kusnandar, F., Mat Hashim, D. and Rahman, R. A. (1999). Application of native and phosphorylated tapioca starches in potato starch noodle. International Journal of Food Science & Technology, 34, 275-280. No, H., Meyers, S., Prinyawiwatkul, W. and Xu, Z. (2007). Applications of chitosan for improvement of quality and shelf life of foods: a review. Journal of Food Science, 72, R87R100. Noda, T., Takahata, Y., Sato, T., Ikoma, H. and Mochida, H. (1996). Physicochemical properties of 32

starches from purple and orange fleshed sweet potato roots at two levels of fertilizer. Starch‐Stärke, 48, 395-399. Noda, T., Tsuda, S., Mori, M., Takigawa, S., Matsuura‐Endo, C., Kim, S. J., Hashimoto, N. and Yamauchi, H. (2006). Effect of potato starch properties on instant noodle quality in wheat flour and potato starch blends. Starch‐Stärke, 58, 18-24. Nugent, A. P. (2005). Health properties of resistant starch. Nutrition Bulletin, 30, 27-54. Parameswaran, K. P. and Sadasivam, S. (1994). Changes in the carbohydrates and nitrogenous components during germination of proso millet, Panicum miliaceum. Plant Foods for Human Nutrition, 45, 97-102. Park, C. S., Hong, B. H. and Baik, B. K. (2003). Protein quality of wheat desirable for making fresh white salted noodles and its influences on processing and texture of noodles. Cereal Chemistry, 80, 297-303. Park, E. Y., Kim, H. N., Kim, J. Y. and Lim, S. T. (2009). Pasting properties of potato starch and waxy maize starch mixtures. Starch‐Stärke, 61, 352-357. Puncha Arnon, S., Pathipanawat, W., Puttanlek, C., Rungsardthong, V. and Uttapap, D. (2008). Effects of relative granule size and gelatinization temperature on paste and gel properties of starch blends. Food Research International, 41, 552-561.

Resmini, P., Lenner, A., Debernardi, G. and Gerardo, D. (1979). Technology and ultrastructure of pasta (noodles) produced with rice flour. Centro Ricerche Soc. Braibanti: Milan, Italy. Sajilata, M. G., Singhal, R. S. and Kulkarni, P. R. (2006). Resistant starch - a review. Comprehensive Reviews in Food Science and Food Safety, 5, 1-17. Saleh, A. S. M., Zhang, Q., Chen, J. and Shen, Q. (2013). Millet grains: nutritional quality, 33

processing, and potential health benefits. Comprehensive Reviews in Food Science and Food Safety, 12, 281-295. Sandhu, K. S., Kaur, M. and Mukesh. (2010). Studies on noodle quality of potato and rice starches and their blends in relation to their physicochemical, pasting and gel textural properties. LWT - Food Science and Technology, 43, 1289-1293. Sasaki, T. and Matsuki, J. (1998). Effect of wheat starch structure on swelling power. Cereal Chemistry, 75, 525-529. Schoch, T. J. and Maywald, E. C. (1968). Preparation and properties of various legume starches. Cereal Chemistry, 45, 564-573. Shibata, M., Fujita, K., Sugiyama, J., Tsuta, M., Kokawa, M., Mori, Y. and Sakabe, H. (2011). Predicting the buckwheat flour ratio for commercial dried buckwheat noodles based on the fluorescence fingerprint. Bioscience, Biotechnology, and Biochemistry, 75, 1312-1316. Shin, S. I., Choi, H. J., Chung, K. M., Hamaker, B. R., Park, K. H. and Moon, T. W. (2004). Slowly digestible starch from debranched waxy sorghum starch: preparation and properties. Cereal Chemistry, 81, 404-408. Shinoj, S., Viswanathan, R., Sajeev, M. S. and Moorthy, S. N. (2006). Gelatinisation and rheological characteristics of minor millet flours. Biosystems Engineering, 95, 51-59. Shukla, K. and Srivastava, S. (2014). Evaluation of finger millet incorporated noodles for nutritive value and glycemic index. Journal of Food Science and Technology, 51, 527-534. Singh, J. and Singh, N. (2001). Studies on the morphological, thermal and rheological properties of starch separated from some Indian potato cultivars. Food Chemistry, 75, 67-77. Singh, N., Singh, J., Kaur, L., Sodhi, N. S. and Gill, B. S. (2003). Morphological, thermal and 34

rheological properties of starches from different botanical sources. Food Chemistry, 81, 219-231. Singh, N., Singh, J. and Singh Sodhi, N. (2002). Morphological, thermal, rheological and noodle‐ making properties of potato and corn starch. Journal of the Science of Food and Agriculture, 82, 1376-1383. Skrabanja, V., Kreft, I., Golob, T., Modic, M., Ikeda, S., Ikeda, K., Kreft, S., Bonafaccia, G., Knapp, M. and Kosmelj, K. (2004). Nutrient content in buckwheat milling fractions. Cereal Chemistry, 81, 172-176. Srikaeo, K., Mingyai, S. and Sopade, P. A. (2011). Physicochemical properties, resistant starch content and enzymatic digestibility of unripe banana, edible canna, taro flours and their rice noodle products. International Journal of Food Science & Technology, 46, 2111-2117. Sui, Z. Q., Lucas, P. W. and Corke, H. (2006). Optimal cooking time of noodles related to their notch sensitivity. Journal of Texture Studies, 37, 428-441. Suwannaporn, P., Pitiphunpong, S. and Champangern, S. (2007). Classification of rice amylose content by discriminant analysis of physicochemical properties. Starch‐Stärke, 59, 171177. Tan, H. Z., Gu, W. Y., Zhou, J. P., Wu, W. G. and Xie, Y. L. (2006). Comparative study on the starch noodle structure of sweet potato and mung bean. Journal of Food Science, 71, C447-C455. Tan, Y. and Corke, H. (2002). Factor analysis of physicochemical properties of 63 rice varieties. Journal of the Science of Food and Agriculture, 82, 745-752. Tester, R. F., Karkalas, J. and Qi, X. (2004). Starch—composition, fine structure and architecture. Journal of Cereal Science, 39, 151-165. 35

Thao, H. M. (2011). Physiochemical properties of sweet potato and mung bean starch and their blends for noodle production. Journal of Food Processing & Technology. 2, 105. Thebaudin, J. Y., Lefebvre, A. C. and Doublier, J. L. (1998). Rheology of starch pastes from starches of different origins: applications to starch-based sauces. LWT - Food Science and Technology, 31, 354-360. Vijayakumar, T. P., Mohankumar, J. B. and Srinivasan, T. (2010). Quality evaluation of noodles from millet flour blend incorporated composite flour. Journal of Scientific and Industrial Research, 69, 48-54. Waterschoot, J., Gomand, S. V., Fierens, E. and Delcour, J. A. (2015). Starch blends and their physicochemical properties. Starch‐Stärke, 67, 1-13. Wei, Y. M. (1995). Buckwheat production in China. Current Advances in Buckwheat Research, 2, 7-10. Yalcin, S. and Basman, A. (2008). Effects of gelatinisation level, gum and transglutaminase on the quality characteristics of rice noodle. International Journal of Food Science & Technology, 43, 1637-1644. Yun, S. H., Quail, K. and Moss, R. (1996). Physicochemical properties of Australian wheat flours for white salted noodles. Journal of Cereal Science, 23, 181-189. Yoo, J., Kim, Y., Yoo, S. H., Inglett, G. E. and Lee, S. (2012). Reduction of rutin loss in buckwheat noodles and their physicochemical characterisation. Food Chemistry, 132, 2107-2111. Zaidul, I. S. M., Yamauchi, H., Kim, S. J., Hashimoto, N. and Noda, T. (2007). RVA study of mixtures of wheat flour and potato starches with different phosphorus contents. Food Chemistry, 102, 1105-1111. 36

Zavareze, E. d. R. and Dias, A. R. G. (2011). Impact of heat-moisture treatment and annealing in starches: A review. Carbohydrate Polymers, 83, 317-328. Zhang, G. Y., Ao, Z. H. and Hamaker, B. R. (2006). Slow digestion property of native cereal starches. Biomacromolecules, 7, 3252-3258. Zhu, F. (2014). Structure, physicochemical properties, modifications, and uses of sorghum starch. Comprehensive Reviews in Food Science and Food Safety, 13, 597-610. Zhu, F. and Corke, H. (2011). Gelatinization, pasting, and gelling properties of sweetpotato and wheat starch blends. Cereal Chemistry, 88, 302-309. Zhu, F., Yang, X., Cai, Y. Z., Bertoft, E. and Corke, H. (2011). Physicochemical properties of sweetpotato starch. Starch‐Stärke, 63, 249-259. Zhu, J. H., Haase, N. U. and Kempf, W. (1990). Investigations on the laboratory scale separation of mung bean Starch. Starch - Stärke, 42, 1-4.

37

Chapter 2. Buckwheat and millet affect the thermal, rheological, and gelling properties of wheat flour

Abstract

Buckwheat (BF) and millet (MF) are recommended as healthy foods due to their unique chemical composition and health benefits. This study investigated the thermal and rheological properties of buckwheat‒wheat and millet‒wheat flour blends at various ratios (0:100‒100:0). Increasing BF or MF concentration led to higher hot paste viscosity, cold paste viscosity, and setback of pasting, gel adhesiveness, storage modulus (G') and loss modulus (G'') of dynamic oscillatory rheology, and yield stress (σ0) of flow curve of wheat flour (WF). BF and MF addition decreased peak viscosity and breakdown of pasting, gel hardness, swelling volume, consistency coefficient (K) of flow curve of WF. Thermal properties of the blends were additive. Non-additive effects were observed for some property changes in the mixtures, and indicated interactions between flour components. This study may provide a basis for using BF and MF in formulating WF based products.

38

2.1. Introduction

Buckwheat is widely recognized as a type of functional food ingredient in countries like China, Japan, and Poland (Dietrych-Szostak & Oleszek 1999), and has been gaining popularity due to its potential health effects (Torbica et al. 2012). Rutin, a flavonoid with antioxidative, antiinflammatory and anticarcinogenic effects, exists in buckwheat at a relatively high level (QuettierDeleu et al. 2000; Sun & Ho 2005). Tomotake et al. (2006) showed that buckwheat flour (BF) reduced the cholesterol and body fat in rats and mice. Skrabanja et al. (2001) showed that buckwheat products could reduce the postprandial glucose and insulin response in vitro and in vivo. The phenolic composition, antioxidant capacity, and gluten free character make buckwheat suitable for a range of nutritional applications (Inglett et al. 2009).

Millet is also considered another healthy food with a range of nutrients such as dietary fiber, calcium, iron, and phenolic compounds (Devi et al. 2014; Shukla & Srivastava 2014). It has been used in producing snacks (Shinoj et al. 2006), biscuit (Saha et al. 2011), and noodles (Shukla & Srivastava 2014), and showed nutritional benefits. For example, wheat noodle with 30 % millet addition showed lower glycemic index than pure wheat noodles (Shukla & Srivastava 2014).

Thus, blending buckwheat or millet flour with wheat flour can be expected to enhance the nutritional properties of the resulting products. However, due to the lack of gluten-type protein in buckwheat and millet, the quality of the final products may be detrimentally affected. For example, when the proportion of buckwheat was over 60 %, the mixture could hardly form any consistent dough (Wei

39

et al. 1995). When the proportion of millet was over 50 %, the quality of noodles dropped dramatically (Saleh et al. 2013). Previous studies showed that information on flour and dough properties can provide a great insight into the quality of final products, and the pattern is product type dependent (Fajardo & Ross 2015; Marti et al. 2015). There is a lack of information on the impact of buckwheat and millet flour addition on diverse physicochemical properties of wheat flour. This seriously hinders the further utilization of these functional grains for diverse food applications. The objective of this study was to study the thermal and rheological properties of buckwheat-wheat flour and millet-wheat flour mixtures. This study may provide a basis to diversify the uses of buckwheat and millet for new product development.

2.2. Materials and methods 2.2.1. Materials

Commercial supplies of buckwheat (Buckwheat flour, Doves Farm, UK), foxtail millet (Organic Millet, Dafengshou, China) and wheat flour (Lion Plain Flour, Anchor foods, Australia), coded BF, MF, WF, were bought from local supermarkets in Hong Kong. BF and MF were mixed WF with proportion 5 %, 10 %, 20 %, 40 %, 60 % for experiments, coded from B-5, M-5 to B-60, M-60, respectively.

2.2.2. Chemical composition

Moisture content was determined by using a moisture analyzer (HB43-S, Mettler Toledo, Switzerland) with a standard method ‘Flour, white’. Ash content was determined by placing 3 g 40

sample in a 550 °C oven until constant weight, followed AACC (2000) methods 08-01. Protein content was determined by improved Kjeldahl method as described in AACC (2000) methods 4610. Lipid content were involved with ethyl ether extraction following acid hydrolysis of the sample with HCl, according to AACC (2000) methods 30-10. Amylose content was determined by iodine method described in Chrastil (1987). Total starch content was measured by total starch assay kit (KTSTA, Megazyme, Co. Wicklow, Ireland).

2.2.3. Particle size distribution

Particle size distribution of flours was measured by Mastersizer 2000 (Malvern Instruments, Malvern, UK). Samples were mixed with Milli-Q® water into small volume dispersion unit before being dispersed at a stirrer speed of 3,000 rpm.

2.2.4. Swelling volume

Swelling volume was determined as described by Crosbie & Lambe (1993). Flour (0.40 g, db) was weighed into 125 mm × 16 mm Pyrex tubes with 12.5 mL 0.05 M AgNO3 solution. The tubes were equilibrated at 25 °C for 5 min, transferred to 92.5 °C water bath for 30 min with occasional mixing. Sample was cooled in ice water for 1 min, placed in 25 °C water bath for 5 min and centrifuged at 1000×g for 15 min. The height of the gel was measured and converted to volume of gel per unit of dry weight of sample.

41

2.2.5. Pasting property

Pasting properties of flour were determined using a Rapid Visco-Analyzer (RVA) (Newport Scientific, Warriewood, Australia). Flour (3.5 g, 14 % moisture basis) samples were mixed with 0.05 M AgNO3 solution (25.0 g) in a canister and loaded using STD1 heating and cooling profile. The temperature was at 50 oC for 1 min, then increased at 12 oC/min to 95 oC, held at 95 oC for 2.5 min, reduced at 12 oC/min to 50 oC, and held for 2 min. Peak viscosity (PV), hot paste viscosity (HPV), cold paste viscosity (CPV), breakdown (BD = PV – HPV), and setback (SB = CPV – HPV) were recorded.

2.2.6. Gel texture

Texture profile analysis was determined on the flour gel made in the RVA testing using a TA-XT2 Texture Analyzer (Stable Micro Systems, Godalming, England). After RVA testing, the paddle was removed and the starch paste in the canister was covered by Parafilm and stored at 4 oC for 24 h. The gel was compressed at a speed of 1.0 mm/s to a distance of 10 mm with a 7 mm cylindrical probe with a flat end. Texture parameters of hardness (HD, g), adhesiveness (ADH, g·s), and cohesiveness (COH) were derived from the instrument software. HD is defined as the maximum force to fracture the gel. ADH is the area under the curve as the probe moves back to initial location. COH was defined as the ratio of the positive force areas during the second compression to that of the first compression.

42

2.2.7. Thermal property

Gelatinization parameters were measured using TA 2920 Modulated DSC Thermal Analyzer different scanning calorimeter (DSC) equipped with a thermal analysis data station (TA Instruments, Newcastle, DE) following the method according to Gunaratne et al. (2011). Flour (2.0 mg, db) was weighed onto the aluminum pan, and distilled water (6 μL) was added with a microsyringe. Pans were sealed, and allowed to stand for 1 h at room temperature. The scanning temperature range and heating rate were 30 ‒ 120 °C and 10 °C/min, respectively, using an empty pan as reference. Onset (To), peak (Tp), and completion (Tc) temperatures, and enthalpy change (ΔH) of gelatinization were recorded by a Universal Analysis Program, Version 2.5H (TA instruments, Newcastle, DE).

2.2.8. Steady shear analysis

Steady shear properties were measured by a rheometer (Physica MCR 301, Anton Paar, Austria) at 25 °C according to the method described by Kong et al. (2010). Starch pastes with 5 % (dry weight basis) of total solids were prepared by heating the starch dispersions at 90 °C for 30 min before transferring onto the plate. Sample edges were covered with an oil layer to prevent evaporation. Sample was equilibrated at 25 °C for 2 min. The pastes were sheared from 0.1 to 1000 s-1 (upward), then reduced from 1000 to 0.1 s-1 (downward) to describe the flow behavior, and the data were fitted to the power law (Eq. 1) and Herschel–Bulkley (Eq. 2) models: σ = Kγn

(1)

σ = σ0 + Kγn

(2)

where, σ is the shear stress (Pa), γ the shear rate (s-1), K the consistency coefficient (Pa Sn), n the 43

flow behavior index (dimensionless), and σ0 is the yield stress (Pa).

2.2.9. Dynamic oscillatory measurement

Dynamic oscillatory measurements were determined mainly following the procedures described previously (Kong et al. 2010; Zhu & Wang 2012). Starch suspensions made of 15 % solids (dry basis) were loaded onto the plate. Sample edge was covered with a thin layer of oil to minimize evaporation. Small-deformation experiments on shear were performed using the same instrument as in the previous section, which was equipped with parallel plate geometry of 25 mm in diameter. The gap size, strain and frequency were 1000 mm, 2 % (within the linear viscoelastic region) and 1 Hz, respectively. Starch suspensions were heated from 40 to 90 oC at a scan rate of 1 °C/min and then cooled from 90 to 25 °C at the same rate. Storage modulus (G', solid component of the network), loss modulus (G'', liquid component), and loss tangent (tanδ = G'' /G'), were recorded as a function of temperature. Following the temperature sweep, materials were held for 5 min at 25 °C to achieve equilibrium before subjecting to a frequency sweep over the range of 0.1 to 25 Hz to monitor the viscoelastic properties.

2.2.10. Statistical analysis

All tests were run at least in triplicate. SPSS (version 19, Endicott, NY) was used for analysis of variance (ANOVA) by Tukey's multiple range test at p < 0.05, and Pearson correlation analysis for determining the relationship between different variables at p < 0.05 and p < 0.01. OriginPro (version

44

8.5.0, OriginLab, Northampton, MA) was used for figure drawing and model fitting.

2.3. Results and discussion 2.3.1. Chemical analysis

Moisture, protein, lipid, ash, amylose and total starch contents of WF, BF, and MF were in Table 2.1. The moisture contents were in the range of 10.8‒15.7 %. WF showed the lowest lipid (1.0 %) and ash content (0.64 %). BF showed the highest protein (11.3 %) and ash content (1.73 %). MF had the lowest protein content (9.7 %) and highest lipid content (3.4 %). Composition of BF was comparable to that reported by Bonafaccia et al. (2003) and Krkošková & Mrázová (2005). Data of MF were comparable to yellow foxtail millet flour reported by Kamara et al. (2009), with little lower protein and higher ash content. This difference may be due to the difference in the variety and processing conditions. WF (24.7 %) and BF (25.6 %) had similar amylose content, while MF (22.9 %) was little lower. MF showed the highest starch content (82.8 %), while BF had the lowest (74.9 %).

2.3.2. Particle size

Particle size distribution image was shown in Fig 2.1. Flour characteristics and behavior are known to be affected by particle size distribution (Torbica et al. 2012; Wronkowska & Haros 2014). Smaller particle size granules need more protein to bind each other and thus acted as a major factor for texture properties like stretchability and foldability on tortilla (Wang & Flores 2000). Higher amount 45

of small flour particles contributes to a dough with lower extensibility, fluidability, and higher water absorption, while narrower particle size distribution lead to higher water absorption capacity and enzymatic susceptibility (Gull et al. 2015; Torbica et al. 2012). All sample shared similar distribution range (2‒158 µm, data not shown) in this study, but WF showed the highest percentage for smaller granules, while BF and MF showed little higher percentage for larger granules.

2.3.3. Pasting properties

Pure WF, BF, and MF had very different pasting properties (Table 2.2). PV indicates the waterbinding capacity of the flour mixtures or maximum swelling of granules, and often correlates with final product quality, providing indications for a likely viscous load encountered by a mixing cooker (Ragaee & Abdel Aal 2006). After PV, further starch granule disruption in the flour mixture was reached during the holding period of RVA test, resulting in amylose leaching out and alignment. The holding strength (HPV) during the period is usually accompanied by a breakdown (BD). CPV indicates the ability to form a viscous paste or gel after cooking and cooling. Setback (SB) as a result of viscosity change during starch re-association in the cooling process, reflects degree of retrogradation of starch and often correlates with texture properties. Increasing amylose content decreases the melting temperature by disrupting crystallinity in the granule structure, and amylose content were significant negative correlated with PV (Blazek & Copeland 2008). Besides, Chen et al. (2010) found gluten addition to wheat starch could significantly decreased PV. However in our study, MF had the lowest amylose content and less gluten than WF but still lowest PV. Possible explanation for this contradictory result was that smaller size granules had greater impact beyond 46

the higher amylose, as Peterson & Fulcher (2001) indicated that larger granules tended to have lower PV (Ragaee & Abdel Aal 2006). In this study MF had higher larger granules distribution than BF and WF. Leaching amylose alignment happens during heat holding period and BD reflects heat stability of the mixtures. Pearson correlation analysis for all mixtures showed that amylose content played a major role in HPV, and was positively related with HPV (0.956, p ＜ 0.01) and similar strong correlation was reported in Bao et al. (2006); while total starch played a major role in BD (0.734, p ＜ 0.01). Sasaki et al. (2000) reported that low amylose content would lead to lower CPV as the increase in viscosity during cooling (SB) is induced by leached out amylose rearranging and forming a thin amylose gel layer, and was supported Juhasz & Salgo (2008) and Yilmaz et al. (2015). That may be the reason MF showed higher CPV than BF. However, that could not explain why WF had higher amylose content but lower CPV than MF. This may possibly be accredited to the gluten presence. Amylose chain entanglement may be hindered by gluten fibrils embedded between the starch granules, hindering and weakening the formation of the starch network (Champenois et al. 1998), and this was supported by Chen et al. (2010), who found gluten addition lead to CPV decrease in wheat starches. To check the mixing ratio impact on mixture properties, a quadratic equation rather than linear model was used to describe relationships between parameters of starch mixtures (Zhu & Corke 2011): Parameter = a*Ratio2 + b*Ratio + c, where a, b, c were constants and depended on specific parameters (e.g. PV). The higher a value was, the better polynomial model could be used (higher R2 compared with linear model). For BF/WF mixtures the equation for PV was PV = 26.98*Ratio2 - 23.27*Ratio + 246.59 (R2 = 0.8298), while for MF/WF mixtures it would be PV = 47.50* Ratio2-7.54*Ratio + 242.49 (R2 = 0.9915). For a linear model BF/WF mixtures showed a low R2 (0.8386), while for MF/WF mixtures R2 was 0.9213. Thus compared with linear model, PV 47

of MF/WF mixtures showed higher non-additive effects, while BF/WF mixture was more complicated. We demonstrated that interactions may exist between different components in the mixtures. However, besides factors above, pasting behavior of flour is also influenced by other factors such as protein content (Singh & Singh 2010) and lipid content (Copeland et al. 2009; Fitzgerald et al. 2003), thus it may not be easily described by one simple equation.

2.3.4. Swelling volume

Swelling volume (SV) has been used as a way to predict noodle eating quality, and often showed correlations with starch paste peak viscosity (Crosbie 1991). SV of all samples ranged from 9.3 (BF) to 15.8 (M-10) mL/g (Table 2.2). Zhao et al. (1998) reported that SV were correlated with PV (p ＜ 0.01) with PV from around 150 to 400 RVU. Similarly in our study addition of BF/MF to WF both lowered down SV and PV value in each case, and SV and PV were correlated positively in each case (BF/WF mixture: 0.908, p＜0.01; MF/WF mixture: 0.930, p ＜ 0.01). However, when all mixtures were subjected to a combined analysis, SV and PV were not correlated (p ＞ 0.05). Possible explanation is that PV ranged in a relatively narrow area from 188 to 246 RVU in our study. Besides, as swelling test mainly refer to starch, starch concentration should also have an impact on SV and we found total starch content highly correlated with SV (0.791, p ＜ 0.01). Relatively good linearity effects were observed in BF/WF mixtures (R2 = 0.93), but not present in MF/WF mixtures (R2 = 0.70). A quadratic model was used to describe the flour mixtures, and SV could be better modeled by SV = 3.20*Ratio2 – 8.96*Ratio + 15.13 (R2 = 0.95, BF/WF mixtures) and SV = 2.24*Ratio2 + 0.58*Ratio + 15.49 (R2 = 0.80, MF/WF mixtures). Thus it is not clear whether 48

different components (e.g. starches) in BF and WF have interactions with each other or not. It is difficult to fit SV data of MF/WF mixtures using both equations, possibly due to their similar data (13.8 & 15.8 g/ml).

2.3.5. Gel texture properties

Gel texture data (HD, ADH, and COH) were in Table 2.3. WF showed the highest HD and lowest ADH. COH of all samples were not significantly different except for WF. Addition of BF and MF lowered HD, and increased ADF of WF. Gel is formed due to starch retrogradation, and differences in starch type (e.g. amylose content) may be one reason for the variations. Amylose content in wheat flour positively correlated with hardness in cooked noodles (Baik & Lee 2003). Besides Ikeda et al. (1999) found protein content correlated negatively with HD and positively with ADH in buckwheat dough. Yilmaz et al. (2015) reported wheat flour had the highest increasing effect on HD when mixed with buckwheat and rice flour as gluten-free food had a softer structure. They also indicated protein content in each flour type may take responsibility for their effects on textural parameters. Both linear and quadratic model could not describe the flour mixtures for HD and ADH well (R2 ＜ 0.81 for all). The narrow HD range (24.6-34.3 g), and varying attributes (starch concentration, gluten, amylose, etc.) at the same time may lower down the feasibility to fit the pattern well.

2.3.6. Thermal properties To, Tp, Tc and ΔH of thermal properties monitored by DSC were in Table 2.4. To ranged from 58.1 °C

49

(WF) to 66.5 oC (MF). Pure BF data were in conformity with other studies. To of five buckwheat starches were around 60 oC (Qian & Kuhn 1999). Zhou et al. (2009) reported thermal properties of one BF with To of 61.8 oC, Tp of 68.5 oC, Tc of 76.9 oC. Millet starch with high degree of crystallinity and high transition temperatures are more resistant to heating (Shinoj et al. 2006). Higher mixing amount of BF and MF could increase heat stability of the mixtures by increasing gelatinization temperature.

Individual components tend to gelatinize independently in excess water (＞ 65 % water content) (Waterschoot et al. 2015). To have two separate endotherms, the difference between Tp of both individual starches should at least exceed 6 oC. Otherwise, the gelatinization endotherms overlap and merge into one (Waterschoot et al. 2015). In addition, when the proportion of one component is much lower, corresponding peak may not be obvious enough. This could possibly explain that one endotherm was observed for all BF/WF mixtures, while two for MF/WF mixture at higher mixing ratio (M-40 and M-60). Difference in Tp between BF and WF was 7.6 oC, much lower than 11.7 oC between MF and WF.

2.3.7. Dynamic oscillatory properties

Dynamic rheology with small deformation avoids gel network disruption (Liu et al. 2006). Storage modulus (G') is a measure of the energy stored subsequently released per cycle of deformation per unit volume, and is the property that relates to the molecular events of elastic nature. Loss modulus (G'') refers to energy dissipated or lost as heat per cycle of deformation per unit volume and is the 50

property that relates to the molecular events of viscous nature (Gunasekaran & Ak 2000). G' was higher than G'', indicating all sample pastes had a tendency to behave like weak gel (Shinoj et al. 2006). During controlled heating, G' and G'' of all samples started to increase starting around 60 °C, reached a peak (TG'max) then fell (Figure 2.2 & Table 2.5). The initial increase in G' and G'' could be attributed to leaching amylose chains and swollen starch granules to fill the entire available volume of the system (Eliasson 1986; Kong et al. 2010; Singh et al. 2003). Further heating leads to melting of the crystalline region in the swollen starch granule and indicates gel structure is destroyed (Eliasson 1986; Singh et al. 2003; Tsai et al. 1997), as a result of decrease in G'. MF mixtures reached peak earlier than BF mixtures both in G' and G'', and possibly due to its gluten presence, and higher larger granule amount. Singh & Kaur (2004) found larger granule size fractions lead to lower TG'max. Champenois et al. (1998) suggested gluten presence could delay or partly reduce the contact between granules, which meant the amylose had not leached out of the granules at the temperature corresponding to early stage of swelling. However, this is contradict to the higher TG'max in BF than in WF. In this case higher minor flour components (lipids and solubles, etc) may show their impact as it possibly lead to a loose bounding in protein-carbohydrate networks formation, and thus network tended to be disrupted earlier at high temperature due to unfavorable proteincarbohydrate interactions or competition for water between gluten and starch (Addo et al. 2001). High amylose content showed higher moduli (Kong et al. 2010), and that possibly why pure BF had higher G' and G'' than MF. WF had the lowest G' and G'' perhaps due to the gluten fibrils embedded between the starch granules, as they can hinder the formation of the starch network that normally occurs either by granule-granule interactions (early stage) or by amylose chain entanglements (later stage) especially when starch concentration was low, thus weakened the starch network 51

(Champenois et al. 1998). This was supported by Chen et al. (2010), who showed decreased gluten addition lead to increased G' and G'' in wheat starch. Tan δ is the ratio of the energy lost to the energy stored for each cycle. It denotes relative effects of viscous and elastic components in a viscoelastic behavior (Gunasekaran & Ak 2000) and Tan δG'max differed little for all samples ranging from 0.096 (M-40) to 0.109 (B-5). A well fitted quadratic model was used to describe G'max in both mixtures, and the equation for BF/WF mixtures was G'max = 2369*Ratio2 + 2136*Ratio + 1707, R2 = 0.9924, and for MF/WF mixtures was G'max = 2346*Ratio2 - 678*Ratio + 1710, R2 = 0.9789, indicating none-additive effects in both mixtures, which meant interactions happened between the components.

During controlled cooling, re-association of the starch molecules happened, and G' and G'' increased for all samples (Figure 2.3), and Tan δ25°C differed little from 0.071 (MF) to 0.129 (BF) (Table 2.5). The low tan δ in all samples indicated that the gel possessed a strong gel-like character (Kong et al. 2009). Similarly like heating process, addition of BF and MF lead to increment G' and G'' (Figure 2.3 & Table 2.5). A quadratic model was used to describe G'25°C, and fitted equations were G'max = 4362*Ratio2 + 2238*Ratio + 2008, R2 = 0.9852 (BF/WF) and G'max = 3129*Ratio2 + 1025*Ratio + 2197, R2 = 0.9753 (MF/WF), again supporting non-additive behavior happened in both mixtures, and meant interactions may exist between the mixtures. As the dispersion was cooled, the formation of a gel structure lead to increment in G' and G''. Champenois et al. (1998) suggested gluten played as the default zones in the among the starch network and weakened the gel.

Frequency sweeps have been used to further provide insights into the structure of biomaterials (Rao 52

2003) (Figure 4 & Table 2.5). G'25Hz was correlated with SB (0.928, p＜0.01), indicating this additive effect of BF and MF on G' and G'' while the former may be attributed to differences in retrogradation differences of the mixtures. The quadratic model could describe G'25Hz for both mixtures: G'25Hz = 6248*Ratio2 + 1713*Ratio + 2589, R2 = 0.9791 (BF/WF); G'25Hz = 4157*Ratio2 + 423*Ratio + 2811, R2 = 0.9708 (MF/WF) and also showed non-additive effects for the mixtures.

Rheological results were correlated with pasting results, e.g. G'max and PV (-0.747, p＜0.01), G'25°C with FV (0.904, p＜0.01), △G' with SB (0.834, p＜0.01). Eliasson (1986) list three major factors that might influence the rheological behavior of a gel: starch granule (dispersed phrase), amylose/amylopectin matrix (continuous phrase) and interactions between the components. We found non-additive behavior in all heating, cooling and frequency sweeping periods, and that was possibly due to effects of gluten presence. However, with other varying components (e.g. lipid), the mechanism of their interacting impact remains unclear and needs further investigation.

2.3.8. Steady shear properties

The flow properties (Figure 2.5 & Table 2.6) indicate WF had higher shear stress than BF/MF. Gelatinized starch suspensions are usually represented by the power law or Herschel-Bulkley model in the range (1-1500 s-1) (Lagarrigue & Alvarez 2001). Both power law and Herschel-Bulkley equations could fit well the curves. The latter was then employed to describe flow property due to a little higher R2 in most cases. All the pastes showed pseudoplastic, shear-thinning behavior (n＜ 1). Yield stress (σ0) demonstrates the minimum stress required to initiate flow (Achayuthakan et al. 53

2006; Kong et al. 2010; Viturawong et al. 2008). BF/MF addition leads to increment of σ0 in most cases, indicating more cross-linked or interactive structure in the flour mixtures had to be broken down before flow can occur (Achayuthakan et al. 2006; Kong et al. 2010; Viturawong et al. 2008). Flow behavior (n) had no big variations, suggesting the mixtures had no significant differences in pseudoplastic behavior. The differences in K may be accredited to the differences in starch type and starch concentration (Lagarrigue & Alvarez 2001).

2.4. Conclusion

Addition of BF and MF greatly affected the pasting, swelling, gelling and rheology of WF. Large differences were shown in pasting and rheological behavior. Increasing BF and MF led to increase of FV, SB, G', G'' and decrease of PV, BD. Blending WF with BF/MF weakened the gel. Differences of BF/WF and MF/WF mixtures were mainly on HPV and thermal properties. Individual component in the mixture tended to gelatinize separately. Large difference in Tp between WF and MF led to two endotherm peaks at higher proportion of MF. A quadratic model was used to describe pasting, gel texture and rheology properties and most indicated non-additive effects. Interactions between individual components may exist. However, with different starches, protein and lipid changing at the same time in mixtures, it is difficult to accurately explain the trend for property changes. Further studies are needed to confirm their relationship with mechanical data.

54

Table 2.1. Chemical composition of wheat (WF), buckwheat (BF), and millet (MF) flours Sample

Moisture

Protein

Lipid

Ash

Amylose content

Total starch content

(%)

(%)

(%)

(%)

(g/100g, db)

(g/100g, db)

WF

15.7±0.1

10.5±0.3ab

1.0±0.0a

0.64±0.1a

24.7±0.9a

79.5±1.9ab

BF

14.8±0.1

11.3±0.5b

2.8±0.0b

1.73±0.2b

25.6±0.2a

74.9±0.5a

MF

10.8±0.1

9.7±0.4a

3.4±0.0c

1.06±0.1c

22.9±0.2b

82.8±0.5b

WF: wheat flour; BF: buckwheat flour; MF: millet flour. All data are expressed as mean ± standard deviation of mean; same superscripts indicate data that do not differ at significance level p < 0.05.

55

Table 2.2. Pasting and swelling properties of BF/WF and MF/WF mixtures Sample

PV

HPV

BD

CPV

SB

SV

WF

246±0.7de

141±4.4abc

100±1.9c

224±3.5a

83±0.9a

15.7±0.4d

B-5

252±1.4e

143±6.8abc

109±5.5c

228±3.2a

85±3.6a

14.3±0.1c

B-10

247±1.7de

139±3.8ab

108±2.1c

228±2.6a

89±1.2a

13.8±0.1c

B-20

240±3.1cd

133±2.2a

107±0.8c

226±1.4a

92±0.9ab

13.8±0.1c

B-40

223±2.5b

150±2.6bc

73±0.1b

236±1.0a

86±1.6a

11.8±0.5b

B-60

231±4.1bc

155±1.5c

76±2.7b

256±5.5b

100±4.1b

11.2±0.1b

BF

195±2.8a

176±1.3d

20±1.5a

349±0.1b

174±1.2c

9.3±0.1a

WF

246±0.7a

141±4.4d

100±1.9ab

224±3.5a

83±0.9a

15.7±0.4a

M-5

241±5.1a

142±0.7d

99±4.4ab

226±3.1ab

84±2.4a

15.1±0.1a

M-10

243±0.5a

139±4.2d

104±4.7ab

229±4.3ab

90±0.1ab

15.8±0.3a

M-20

241±1.1a

133±0.2cd

108±0.9b

228±2.7ab

94±2.9b

15.5±0.7a

M-40

232±0.6b

123±0.5bc

109±1.2b

230±0.7ab

106±1.2c

15.2±0.1a

M-60

219±1.5c

113±2.2b

105±3.7b

236±1.2b

123±0.4d

15.2±0.3a

MF

188±0.9d

97±2.8a

91±3.7a

251±1.2c

154±0.5e

13.8±0.3b

WF: wheat; BF: buckwheat; MF: millet; B-5 to B-60 and M-5 to M-60 refers to BF/MF proportion (5-60 %) in the mixture. PV = peak viscosity (RVU); HPV = hot pates viscosity (RVU); BD = breakdown viscosity (RVU); CPV = cold paste viscosity (RVU); SB = setback (RVU); SV = swelling volume (mL/g) All data are expressed as mean ± standard deviation of mean; same superscripts indicate data that do not differ at significance level p < 0.05.

56

Table 2.3. Gel texture of BF/WF and MF/WF mixtures Sample

HD

ADH

COH

Sample

HD

ADH

COH

WF

34.3±3.5a

-27.6±2.2d

0.517±0.01a

WF

34.3±3.5c

-27.6±2.2a

0.517±0.01a

B-5

33.8±2.5a

-43.9±9.4c

0.505±0.01ab

M-5

34.1±1.3c

-40.0±8.6a

0.513±0.01a

B-10

27.6±0.6b

-53.2±3.4abc

0.480±0.01b

M-10

27.6±1.3ab

-40.4±9.1a

0.504±0.01a

B-20

33.9±0.9a

-51.6±2.5bc

0.495±0.01ab

M-20

28.1±1.5ab

-59.3±5.1b

0.487±0.01a

B-40

28.7±1.1b

-66.8±4.6a

0.495±0.01ab

M-40

25.1±0.9a

-57.0±6.0b

0.506±0.03a

B-60

28.3±1.0b

-62.9±6.8ab

0.487±0.01ab

M-60

24.9±0.5a

-63.9±5.8b

0.497±0.02a

BF

24.6±1.2b

-57.3±8.0abc

0.503±0.02ab

MF

31.5±2.7bc

-63.3±3.6b

0.493±0.01a

WF: wheat; BF: buckwheat; MF: millet; B-5 to B-60 and M-5 to M-60 refers to BF/MF proportion (5-60 %) in the mixture. HD = hardness (g), ADH = adhesiveness (g.s), COH= cohesiveness. All data are expressed as mean ± standard deviation of mean; same superscripts indicate data that do not differ at significance level p < 0.05.

57

Table 2.4. Thermal properties of buckwheat-wheat and millet-wheat mixtures To

Tp

Tc

△H

WF

58.1±0.4a

63.3±0.2a

74.6±0.4ab

6.2±0.8a

B-5

58.6±0.4a

63.6±0.3a

73.5±1.4a

4.9±0.6a

B-10

58.6±0.4a

63.6±0.3a

72.0±2.2a

4.5±1.0a

B-20

58.8±0.5a

63.7±0.6ab

74.6±2.5ab

4.8±1.5a

B-40

59.4±0.3a

64.7±0.5b

77.9±1.8bc

5.1±0.8a

B-60

59.4±0.7a

69.9±0.9c

79.2±1.5c

6.0±1.3a

BF

64.6±1.8b

70.7±0.3c

78.9±1.2c

5.8±0.7a

WF

58.1±0.4a

63.3±0.2a

74.6±0.4c

6.2±0.8d

M-5

58.3±0.6a

63.5±0.1a

73.6±0.6bc

5.6±1.4d

M-10

58.7±0.3ab

63.0±0.3a

68.5±0.9a

2.9±0.6bc

M-20

58.7±0.5ab

63.4±0.3a

70.4±3.7ab

3.4±1.3c

Peak 1

58.9±0.8ab

63.4±0.4a

69.0±0.6a

2.0±0.4abc

Peak 2

71.5±0.4d

75.0±0.2c

83.9±1.4d

0.4±0.2a

Peak 1

59.6±0.5b

63.5±0.5a

69.5±0.5a

1.1±0.1ab

Peak 2

70.1±0.4e

75.0±0.6c

82.8±2.1d

1.2±0.4ab

66.5±0.3c

74.0±3.0b

83.7±0.7d

5.5±0.5d

Sample

M-40

M-60 MF

WF: wheat; BF: buckwheat; MF: millet; B-5 to B-60 and M-5 to M-60 refers to BF/MF proportion (5-60 %) in the mixture. To: onset temperature (°C); Tp: peak temperature (°C); Tc: conclusion temperature (°C); △H: gelatinization enthalpy (J/g). All data are expressed as mean ± standard deviation of mean; same superscripts indicate data that do not differ at significance level p < 0.05.

58

Table 2.5. Dynamic rheological properties of buckwheat-wheat and millet-wheat mixtures Sample

TG'max (°C)

G'max (Pa)

tan δG'max

G'90°C (Pa)

tan δG'90°C

G'25°C (Pa)

tan δ25°C

△G' (Pa) a

G'25Hz(Pa) a

tan δ25Hz

WF B-5 B-10 B-20 B-40 B-60 BF

a

86.9 86.8a 86.7a 86.3a 87.0a 88.0b 89.5c

a

1700 1857a 1800a 2433b 2817b 3877c 6213d

bc

0.107 0.109c 0.104ab 0.103a 0.103ab 0.100a 0.100a

a

1428 1563a 1517a 2080b 2533c 3747d 6207e

de

0.119 0.121e 0.116cd 0.114c 0.107b 0.102a 0.100a

a

1908 2103a 2100a 3100b 3627b 4653c 8683d

a

0.098 0.090a 0.092a 0.086a 0.094a 0.101a 0.129a

480 540a 583a 1020a 1093a 907a 2477b

2480 2607a 2580a 3820b 4397b 5427c 10667d

0.170a 0.167a 0.177a 0.148a 0.155a 0.163a 0.179a

WF M-5 M-10 M-20 M-40 M-60 MF

86.9f 85.7ef 85.5e 83.6d 80.4c 78.4b 76.7a

1700a 1780a 1547a 1720a 1750a 2203b 3367c

0.107c 0.104bc 0.104bc 0.100ab 0.096a 0.097a 0.102bc

1428c 1410c 1203ab 1208ab 1090a 1313bc 2087a

0.119f 0.113e 0.114e 0.105d 0.094c 0.085b 0.073a

1908a 2473b 2297b 2793c 3033c 3807d 6400e

0.098a 0.079b 0.078b 0.074b 0.071b 0.071b 0.071b

480a 1063b 1093b 1585c 1943d 2493e 4313f

2480a 3147b 2810ab 3363c 3487c 4490d 7430e

0.170d 0.143c 0.148cd 0.128bc 0.126abc 0.116ab 0.104a

WF: wheat; BF: buckwheat; MF: millet; B-5 to B-60 and M-5 to M-60 refers to BF/MF proportion (5-60 %) in the mixture. TG'max: value of temperature when highest G' is reached during heating; G'max: highest G' value during heating; tan δG'max: value of tan δ at TG'max; G'90°C: value of G' at 90 °C; tan δG'90°C: value of tan δ at 90 °C; G'25°C: value of G' at 25 °C; tan δ25°C: value of tan δ at 25 °C; △G': difference between G'25°C and G'90°C; G'25Hz: value of G' at 25 Hz; Tan δ25Hz: value of tan δ at 25 Hz. All data are mean value; values with the same superscript are not significantly different (p ＜ 0.05).

59

Table 2.6. Power law and Herschel-Bulkley parameters of buckwheat-wheat and millet-wheat mixtures Upward Power Law

Sample

Downward Herschel-Bulkley

Power Law

Herschel-Bulkley

K(Pa Sn)

n(-)

R2

σ0（Pa）

K(Pa Sn)

n(-)

R2

K(Pa Sn)

n(-)

R2

σ0（Pa）

K(Pa Sn)

n(-)

R2

WF B-5 B-10 B-20 B-40 B-60 BF

7.61 6.01 5.66 4.80 4.24 3.59 4.09

0.44 0.46 0.47 0.48 0.48 0.48 0.39

0.9891 0.9903 0.9910 0.9955 0.9905 0.9789 0.9773

1.63 0.78 1.46 1.64 3.93 5.30 3.80

6.85 5.67 5.05 4.16 2.76 1.66 2.04

0.45 0.47 0.48 0.50 0.54 0.58 0.49

0.9888 0.9899 0.9908 0.9958 0.9943 0.9901 0.9851

2.07 1.79 1.66 1.72 2.22 2.39 2.38

0.60 0.61 0.61 0.61 0.55 0.52 0.47

0.9948 0.9950 0.9946 0.9957 0.9970 0.9950 0.9963

3.38 2.98 2.95 2.68 2.40 2.83 1.82

1.35 1.19 1.07 1.16 1.57 1.52 1.65

0.66 0.67 0.68 0.66 0.60 0.59 0.52

0.9981 0.9981 0.9980 0.9986 0.9995 0.9993 0.9992

WF M-5 M-10 M-20 M-40 M-60 MF

7.61 6.20 5.29 6.25 6.06 5.67 4.38

0.44 0.45 0.46 0.43 0.42 0.40 0.42

0.9891 0.9936 0.9927 0.9964 0.9974 0.9942 0.9882

1.63 1.76 1.91 2.13 3.02 4.16 4.61

6.85 5.43 4.49 5.23 4.55 3.51 2.20

0.45 0.47 0.48 0.46 0.46 0.47 0.52

0.9888 0.9937 0.9930 0.9971 0.9993 0.9992 0.9977

2.07 1.88 1.73 1.77 1.65 1.35 1.29

0.60 0.60 0.60 0.60 0.60 0.61 0.59

0.9948 0.9940 0.9952 0.9945 0.9936 0.9934 0.9922

3.38 3.34 2.74 2.88 2.98 2.69 2.65

1.35 1.18 1.14 1.14 1.01 0.81 0.72

0.66 0.67 0.66 0.66 0.67 0.68 0.67

0.9981 0.9978 0.9985 0.9980 0.9981 0.9988 0.9985

WF: wheat; BF: buckwheat; MF: millet; B-5 to B-60 and M-5 to M-60 refers to BF/MF proportion (5-60 %) in the mixture. Upward: 0-1000 s-1; downward, 1000-0 s-1. σ0: yield stress; K: consistency coefficient; n: flow behavior; R2: regression coefficient

60

Figure 2.1. Particle size distribution range of wheat, buckwheat and millet flour. Figure 2.2. Changes in storage (G') and loss (G'') modulus of 15 % flour suspensions during heating (heating rate: 1 °C/min, strain: 2 %, frequency: 1 Hz). (A & C) buckwheat (BF), wheat (WF) and their mixtures. (B & D) millet (MF), wheat (WF) and their mixtures.

61

Figure 2.3. Changes in storage (G') and loss (G'') modulus of 15 % flour suspensions during cooling (cooling rate: 1 °C/min, strain: 2 %, frequency: 1 Hz). (A & C) buckwheat (BF), wheat (WF) and their mixtures. (B & D) millet (MF), wheat (WF) and their mixtures.

62

Figure 2.4. Changes in storage (G') and loss (G'') modulus of 15 % flour suspensions during a frequency sweep (strain: 2 %, temperature: 25 °C). (A & C) buckwheat (BF), wheat (WF) and their mixtures. (B & D) millet (MF), wheat (WF) and their mixtures.

63

Figure 2.5. Flow curves of flour pastes (5 % concentration). (A & C) upward (0-1000 s-1) and downward (1000-0 s-1) for buckwheat (BF), wheat (WF) flour and their mixtures. (B & D) upward and downward for millet (MF), wheat (WF) flour and their mixture pastes.

64

References

AACC International. Approved Methods of Analysis, 10th Ed. Method 08-01. Approved April 13, 1961, renewed November 3, 1999; Method 30-10. Approved April 13, 1961, renewed November 3, 1999; Method 46-10. Approved April 13, 1961, renewed November 3, 1999. AACC International, St. Paul, MN, U.S.A. Achayuthakan, P., Suphantharika, M. and Rao, M. A. (2006). Yield stress components of waxy corn starch-xanthan mixtures: Effect of xanthan concentration and different starches. Carbohydrate Polymers, 65, 469-478. Addo, K., Xiong, Y. L. and Blanchard, S. P. (2001). Thermal and dynamic rheological properties of wheat flour fractions. Food Research International, 34, 329-335. Baik, B. K. and Lee, M. R. (2003). Effects of starch amylose content of wheat on textural properties of white salted noodles. Cereal Chemistry, 80, 304-309. Bao, J., Shen, S., Sun, M. and Corke, H. (2006). Analysis of genotypic diversity in the starch physicochemical properties of nonwaxy rice: apparent amylose content, pasting viscosity and gel texture. Starch‐Stärke, 58, 259-267. Blazek, J. and Copeland, L. (2008). Pasting and swelling properties of wheat flour and starch in relation to amylose content. Carbohydrate Polymers, 71, 380-387. Bonafaccia, G., Marocchini, M. and Kreft, I. (2003). Composition and technological properties of the flour and bran from common and tartary buckwheat. Food Chemistry, 80, 9-15. Champenois, Y., Rao, M. A. and Walker, L. P. (1998). Influence of gluten on the viscoelastic properties of starch pastes and gels. Journal of the Science of Food and Agriculture, 78,

65

119-126. Chen, J., Deng, Z., Wu, P., Tian, J. and Xie, Q. (2010). Effect of gluten on pasting properties of wheat starch. Agricultural Sciences in China, 9, 1836-1844. Chrastil, J. (1987). Improved colorimetric determination of amylose in starches or flours. Carbohydrate Research, 159, 154-158. Collado, L. S., Mabesa, L. B., Oates, C. G. and Corke, H. (2001). Bihon-type noodles from heatmoisture-treated sweet potato starch. Journal of Food Science, 66, 604-609. Copeland, L., Blazek, J., Salman, H. and Tang, M. C. (2009). Form and functionality of starch. Food Hydrocolloids, 23, 1527-1534. Crosbie, G. B. (1991). The relationship between starch swelling properties, paste viscosity and boiled noodle quality in wheat flours. Journal of Cereal Science, 13, 145-150. Crosbie, G. B. and Lambe, W. J. (1993). The application of the flour swelling volume test for potential noodle quality to wheat breeding lines affected by sprouting. Journal of Cereal Science, 18, 267-276. Devi, P. B., Vijayabharathi, R., Sathyabama, S., Malleshi, N. G. and Priyadarisini, V. B. (2014). Health benefits of finger millet (Eleusine coracana L.) polyphenols and dietary fiber: a review. Journal of Food Science and Technology, 51, 1021-1040. Dietrych Szostak, D. and Oleszek, W. (1999). Effect of processing on the flavonoid content in buckwheat (Fagopyrum esculentum Möench) grain. Journal of Agricultural and Food Chemistry, 47, 4384-4387. Eliasson, A. C. (1986). Viscoelastic behaviour during the gelatinization of starch I. comparison of wheat, maize, potato and waxy - barley starches. Journal of Texture Studies, 17, 253-265.

66

Fajardo, C. A. and Ross, A. S. (2015). Exploring relationships between pancake quality and grain and flour functionality in soft wheats. Cereal Chemistry. (Accepted). Fitzgerald, M., Martin, M., Ward, R., Park, W. and Shead, H. (2003). Viscosity of rice flour: a rheological and biological study. Journal of Agricultural and Food Chemistry, 51, 22952299. Gull, A., Prasad, K. and Kumar, P. (2015). Effect of millet flours and carrot pomace on cooking qualities, color and texture of developed pasta. LWT - Food Science and Technology, 63, 470-474. Gunaratne, A., Bentota, A., Cai, Y. Z., Collado, L. and Corke, H. (2011). Functional, digestibility, and antioxidant properties of brown and polished rice flour from traditional and new‐ improved varieties grown in Sri Lanka. Starch‐Stärke, 63, 485-492. Gunasekaran, S. and Ak, M. M. (2000). Dynamic oscillatory shear testing of foods - selected applications. Trends in Food Science & Technology, 11, 115-127. Ikeda, K., Fujiwara, J., Asami, Y., Arai, R., Bonafaccia, G., Kreft, I. and Yasumoto, K. (1999). Relationship of protein to the textural characteristics of buckwheat products: analysis with various buckwheat flour fractions. Fagopyrum, 16, 79-83. Inglett, G. E., Xu, J., Stevenson, D. G. and Chen, D. (2009). Rheological and pasting properties of buckwheat (Fagopyrum esculentum Möench) flours with and without jet-cooking. Cereal Chemistry, 86, 1-6. Juhasz, R. and Salgo, A. (2008). Pasting behavior of amylose, amylopectin and their mixtures as determined by RVA curves and first derivatives. Starch‐Stärke, 60, 70-78. Kamara, M., Huiming, Z., Kexue, Z., Amadou, I. and Tarawalie, F. (2009). Comparative study of

67

chemical composition and physicochemical properties of two varieties of defatted foxtail millet flour grown in China. American Journal of Food Technology, 4, 255-267. Kong, X., Kasapis, S., Bao, J. and Corke, H. (2009). Effect of gamma irradiation on the thermal and rheological properties of grain amaranth starch. Radiation Physics and Chemistry, 78, 954960. Kong, X., Kasapis, S., Bertoft, E. and Corke, H. (2010). Rheological properties of starches from grain amaranth and their relationship to starch structure. Starch - Stärke, 62, 302-308. Krkošková, B. and Mrázová, Z. (2005). Prophylactic components of buckwheat. Food Research International, 38, 561-568. Lagarrigue, S. and Alvarez, G. (2001). The rheology of starch dispersions at high temperatures and high shear rates: a review. Journal of Food Engineering, 50, 189-202. Liu, H., Eskin, N. M. and Cui, S. W. (2006). Effects of yellow mustard mucilage on functional and rheological properties of buckwheat and pea starches. Food Chemistry, 95, 83-93. Marti, A., Qiu, X., Schoenfuss, T. C. and Seetharaman, K. (2015). Characteristics of perennial wheatgrass (Thinopyrum intermedium) and refined wheat flour blends: the impact on rheological properties. Cereal Chemistry. (Accepted). Peterson, D. G. and Fulcher, R. G. (2001). Variation in Minnesota HRS wheats: starch granule size distribution. Food Research International, 34, 357-363. Qian, J. Y. and Kuhn, M. (1999). Evaluation on gelatinization of buckwheat starch: a comparative study of Brabender viscoamylography, rapid visco-analysis, and differential scanning calorimetry. European Food Research and Technology, 209, 277-280. Quettier Deleu, C., Gressier, B., Vasseur, J., Dine, T., Brunet, C., Luyckx, M., Cazin, M., Cazin, J.

68

C., Bailleul, F. and Trotin, F. (2000). Phenolic compounds and antioxidant activities of buckwheat

(Fagopyrum

esculentum

Moench)

hulls

and

flour.

Journal

of

Ethnopharmacology, 72, 35-42. Ragaee, S. and Abdel Aal, E. S. M. (2006). Pasting properties of starch and protein in selected cereals and quality of their food products. Food Chemistry, 95, 9-18. Rao, M. (2003). Phase transitions, food texture and structure. In: B. M. McKenna, (Ed.), Texture in food. Woodhead Publishing: Cambridge, U.K.. pp. 36-62. Saha, S., Gupta, A., Singh, S., Bharti, N., Singh, K., Mahajan, V. and Gupta, H. (2011). Compositional and varietal influence of finger millet flour on rheological properties of dough and quality of biscuit. LWT - Food Science and Technology, 44, 616-621. Saleh, A. S. M., Zhang, Q., Chen, J. and Shen, Q. (2013). Millet grains: nutritional quality, processing, and potential health benefits. Comprehensive Reviews in Food Science and Food Safety, 12, 281-295. Sasaki, T., Yasui, T. and Matsuki, J. (2000). Effect of amylose content on gelatinization, retrogradation, and pasting properties of starches from waxy and nonwaxy wheat and their F1 seeds. Cereal Chemistry, 77, 58-63. Shinoj, S., Viswanathan, R., Sajeev, M. S. and Moorthy, S. N. (2006). Gelatinisation and rheological characteristics of minor millet flours. Biosystems Engineering, 95, 51-59. Shukla, K. and Srivastava, S. (2014). Evaluation of finger millet incorporated noodles for nutritive value and glycemic index. Journal of Food Science and Technology, 51, 527-534. Singh, N. and Kaur, L. (2004). Morphological, thermal, rheological and retrogradation properties of potato starch fractions varying in granule size. Journal of the Science of Food and

69

Agriculture, 84, 1241-1252. Singh, S. and Singh, N. (2010). Effect of debranning on the physico-chemical, cooking, pasting and textural properties of common and durum wheat varieties. Food Research International, 43, 2277-2283. Singh, N., Singh, J., Kaur, L., Sodhi, N. S. and Gill, B. S. (2003). Morphological, thermal and rheological properties of starches from different botanical sources. Food Chemistry, 81, 219-231. Skrabanja, V., Liljeberg Elmståhl, H. G., Kreft, I. and Björck, I. M. (2001). Nutritional properties of starch in buckwheat products: studies in vitro and in vivo. Journal of Agricultural and Food Chemistry, 49, 490-496. Sun, T. and Ho, C. T. (2005). Antioxidant activities of buckwheat extracts. Food Chemistry, 90, 743-749. Tomotake, H., Yamamoto, N., Yanaka, N., Ohinata, H., Yamazaki, R., Kayashita, J. and Kato, N. (2006). High protein buckwheat flour suppresses hypercholesterolemia in rats and gallstone formation in mice by hypercholesterolemic diet and body fat in rats because of its low protein digestibility. Nutrition, 22, 166-173. Torbica, A., Hadnađev, M. and Dapčević Hadnađev, T. (2012). Rice and buckwheat flour characterisation and its relation to cookie quality. Food Research International, 48, 277283. Tsai, M. L., Li, C. F. and Lii, C. Y. (1997). Effects of granular structures on the pasting behaviors of starches 1. Cereal Chemistry, 74, 750-757. Viturawong, Y., Achayuthakan, P. and Suphantharika, M. (2008). Gelatinization and rheological

70

properties of rice starch/xanthan mixtures: Effects of molecular weight of xanthan and different salts. Food Chemistry, 111, 106-114. Wang, L. and Flores, R. A. (2000). Effects of flour particle size on the textural properties of flour tortillas. Journal of Cereal Science, 31, 263-272. Waterschoot, J., Gomand, S. V., Fierens, E. and Delcour, J. A. (2015). Starch blends and their physicochemical properties. Starch‐Stärke, 67, 1-13. Wei, Y. M., Zhang, G. Q. and Li, Z. X. (1995). Study on nutritive and physico-chemical properties of buckwheat flour. Food / Nahrung, 39, 48-54. Wronkowska, M. and Haros, M. (2014). Wet-milling of buckwheat with hull and dehulled – The properties of the obtained starch fraction. Journal of Cereal Science, 60, 477-483. Yilmaz, M. T., Yildiz, Ö., Yurt, B., Toker, O. S., Karaman, S. and Baştürk, A. (2015). A mixture design study to determine interaction effects of wheat, buckwheat, and rice flours in an aqueous model system. LWT - Food Science and Technology, 61, 583-589. Zhao, X., Batey, I., Sharp, P., Crosbie, G., Barclay, I., Wilson, R., Morell, M. and Appels, R. (1998). A single genetic locus associated with starch granule properties and noodle quality in wheat. Journal of Cereal Science, 27, 7-13. Zhou, Y. G., Li, D., Wang, L. J., Li, Y., Yang, B. N., Bhandari, B., Chen, X. D. and Mao, Z. H. (2009). Effect of water content on thermal behaviors of common buckwheat flour and starch. Journal of Food Engineering, 93, 242-248. Zhu, F. and Corke, H. (2011). Gelatinization, pasting, and gelling properties of sweetpotato and wheat starch blends. Cereal Chemistry, 88, 302-309. Zhu, F. and Wang, Y. J. (2012). Rheological and thermal properties of rice starch and rutin mixtures.

71

Food Research International, 49, 757-762.

72

Chapter 3. Buckwheat and millet flour affect the color, texture and digestibility of wheat-based noodles

Abstract

Wheat flour is widely used as a raw material for Asian noodle production, but its nutritional content may be limited. Thus two grains recognized for health properties, buckwheat and millet, were added to wheat flour for making Asian salt noodles to improve the nutritional composition. The effects on noodle qualities were studied, and results showed that noodle texture became weaker with increasing levels of buckwheat and millet, but changes in cooking properties were limited. In vitro digestion results showed that less than 10 % buckwheat and millet flour addition could lead to little lower or similar estimated glycemic index. However, higher than 20 % mixing ratio would increase estimated glycemic index.

73

3.1. Introduction

Asian noodles are one of the most popular foods consisting mainly of wheat flour. To improve its quality and nutritional basis, in recent years researchers have tried adding various ingredients to wheat flour, such as sweet potato flour (Collado & Corke 1996; Collins & Pangloli 1997), soy flour (Collins & Pangloli 1997; Shogren et al. 2006), banana flour (Ovando Martinez et al. 2009; Saifullah et al. 2009), buckwheat (Kreft & Skrabanja 2002) and millets (Vijayakumar et al. 2010). As two recognized healthful crops, buckwheat and millet have long been used to make noodles combined with wheat flour, resulting in improvements in nutritional diversity. Buckwheat noodles are popular in parts of China and in Japan (Ikeda & Asami 2000; Kreft & Skrabanja 2002), while millet noodles may even date back 4000 years (Lu et al. 2005). Recently public interest in healthy foods has increased. Buckwheat can be a valuable constituent of the diet, providing dietary carbohydrates in different forms (Kreft & Skrabanja 2002). Its beneficial health qualities for improvement on prevention of chronic diseases are gaining increasing recognition from food scientists (Li & Zhang 2001; Torbica et al. 2012). Spaghetti with appropriate buckwheat flour addition (15-20 %) demonstrated little change in sensory properties (Chillo et al. 2008). For millet, a 20-30 % combination with wheat noodles can be more nutritious and beneficial for health due to lower glycemic response (Shukla & Srivastava 2014; Vijayakumar et al. 2010).

To classify foods digestibility by their postprandial blood glucose response, Jenkins et al. (1981) introduced the concept of glycemic index (GI). GI is defined as the postprandial incremental glycemic area after a test meal, expressed as the percentage of the corresponding are after an equicarbohydrate portion of a reference food (glucose or white bread) (Goni et al. 1997). However, 74

postprandial glucose studies have to recruit subjects, involve training, and are expensive and inconvenient. Thus in vitro digestion method are more practical and feasible for researchers. Whether in vitro data could reflect accurately in vivo postprandial glucose rate needs to be investigated, as real digestion process are more complex. Established by Goni et al. (1997), a simple and easy in vitro method was well correlated with in vivo response for the same food, and has been wide accepted by many studies, e.g. in rice (Frei et al. 2003; Hu et al. 2004), starches (Chung et al. 2008), noodles (Ge et al. 2014), and breads (Zabidi & Aziz 2009).

On the other hand, noodle appearance is always a key quality determinant in white salted noodles based on wheat flour (Wang et al. 2004). Good brightness is preferred in all types of noodles while the Asian consumer likes a bright creamy appearance (Wang et al. 2004). Wang et al. (2004) summarized several factors that could have an impact on noodle color and discoloration: intrinsic flour color, ash content, flour extraction rate, flour particle size, sprout damage, protein content and enzyme activity. Besides, a noodle with good cooking and textural properties can better meet customer’s expectations and criteria. Wu et al. (2015) reported a similar sensory scores on textural attributes about buckwheat and millet noodles. Zhu et al. (2010) evaluated cooking properties together with firmness and tensile force on Asian salted noodles mixed with natural pigments. Multiple blends containing oat, rye, buckwheat and common wheat flour could have higher nutritional value, better sensory attributes and lower starch hydrolysis rate (Angioloni & Collar 2011).

Previously in Chapter 2 we have studied impact of buckwheat and millet addition to WF on flour

75

properties. However, limited research has been done on wheat noodle combinations with buckwheat and millet, especially the latter. Due to lack of gluten, pure buckwheat and millet cannot form a consistent dough. Buckwheat proportion should be less than 60 % to avoid dough cracking (Wei et al. 1995), and for millet, it was suggested to be less than 50 % (Saleh et al. 2013). The objective of this study was to investigate the impact of buckwheat and millet addition to wheat noodles on color, cooking, texture and digestion properties, at different mixing ratios (60 % or less for buckwheat and 40 % or less for millet).

3.2. Materials and methods 3.2.1. Materials

Materials were prepared as the same described in Chapter 2. Commercial supplies of buckwheat (Buckwheat flour, Doves farm, UK), foxtail millet (Organic Millet, Dafengshou, China) and wheat flour (Lion Plain Flour, Anchor Foods, Australia), coded BF, MF, WF, were bought from local supermarkets in Hong Kong. Buckwheat was mixed with wheat flour with proportions 5 %, 10 %, 20 %, 40 %, 60 %, and for millet the same values up to 40 %, coded from B-5 to B-60, and M-5 to M-40, respectively.

3.2.2. Noodle preparation

Noodle making followed methods described by Zhu et al. (2010) with some modifications. Wheat flour or flour mixtures (30.0 g, db) were mixed with 12.5 g of 2 % (w/w) NaCl solution to make

76

salted noodles. The mixture was manually kneaded for 5 min in a plastic bowl to obtain a crumbly dough ball. After resting for 5 min in the bowl, the crumbly dough was then sheeted through the first roll gap 15 times and folded in half each time on a domestic-type electric pasta machine (Atlas Electric model 150, Marcato, Campodarsego, Italy). Then the dough sheet was further passed through the next four successively reduced gaps 8 times each without intermediate resting. Finally, the resultant thin dough sheet was cut through a slotted roll to produce noodles of around 2 mm width. The raw noodles were placed on plastic bags to dry at 25 °C for around 8 h in an oven with the ventilation setting for fresh air at six (maximum) (Model 600, Memmert, Schwabach, Germany). Then dried raw noodles were stored in transparent plastic bags at room temperature (23-26 °C) until use within a week.

3.2.3. Cooking properties

Cooking yield and cooking loss were determined following Zhu et al. (2010) with little modifications. Dried raw noodles (3-4 g) were cooked in boiling water (200 mL) in a covered beaker for 5 min until the white core disappeared. They were cooled in water (20 °C) for 10 s before drying on a paper towel for 15 min until further texture and color testing. For cooking yield and loss analysis, cooked noodle strands were placed on a paper towel and softly blotted for about 1 min to remove the excess surface moisture on the noodle surface. Cooking yield (CY) was calculated as: CY = Mco/Md Where Mco = weight of cooked noodles (g); Md = weight of dried raw noodles (g). After cooking, the remaining water in the beaker was further boiled for evaporation until less than

77

50 mL was left. Then remaining water in the beaker was transferred to a petri dish by washing with 50 mL distilled water, and then further dried in an oven at 100 °C until constant weight (Mc, g). Cooking loss % (CL) was calculated as: CL % = (Mc-Mb)*100 / (Md-Mm) Where Mb = weight of beaker (g); Mm = weight of moisture in dried raw noodles (g). All analysis was conducted at least in triplicate.

3.2.4. Color analysis

The Hunter color parameters L, a, and b (CIE 1976) were measured by a colorimeter (Chroma Meter CR-301, Minolta Co., Osaka, Japan) on surface of dried noodles, and on cooked noodles around 20 min after cooking. At least three replications were performed and L, a, b value were recorded.

3.2.5. Texture analysis of noodles

Tensile and compression tests were performed on a TA-XT2 Texture Analyzer (Stable Micro Systems, Godalming, England) according to the method described in Zhu et al. (2010) with modifications. Noodles were cooked in separate batches for 5 min and tested separately between 15 and 20 min after cooking to avoid rapid textural changes of cooked noodles right after cooking. For tensile strength analysis, the instrument was equipped with spaghetti/noodle tensile grips (A/SPR). The noodle strand was wound two, three or four times around parallel friction roller of the grip to anchor the samples and avoid slippage. The distance between the parallel rollers was 4 cm. The mode was measure force in tension. Pre-test and test speeds were 3.0 mm/s, post-test speed was 5.0 78

mm/s. Distance was 100 mm. The trigger type was auto with a trigger force of 5.0 g. The data acquisition rate was 200 pps. The maximum force required to break the strand was termed tensile force (g). At least 10 individual strands were tested for each group. The five minimum values were discarded. For compression analysis, the probe was a cylindrical probe with a diameter of 35 mm (P/35). The mode was measuring force in compression with single cycle. The pre-test, test, and posttest speeds were 2.0 mm/s. The strain was 75 %. The trigger type was auto with a trigger force of 5.0 g. The data acquisition rate was 200 pps. Two noodle strands were placed close and parallel on the platform and tested close together at a time. The maximum force (g) was noted as firmness. At least six tests per group were conducted.

3.2.6. Total starch

Total starch content was measured by total starch assay kit (K-TSTA, Megazyme, Co. Wicklow, Ireland) as described in Chapter 2.

3.2.7. In vitro digestion

In vitro digestion followed Goni et al. (1997) with modifications. Approximately 50 mg (db) noodle strands were cooked with 5 mL distilled water in boiled water bath for 5 min. Subsequently after cooking, samples were homogenized for 2 min using an Ultra Turrax homogenizer T25 (Janke and Kunkel, Ika Labortechnik, München, Germany) and 10 mL of HCl–KCl buffer pH 1.5 were added. Then, 0.2 mL of a solution containing 40 mg of pepsin from porcine gastric mucosa (P7125, Sigma) in 10 mL HCl–KCl buffer, pH 1.5, were added to each sample, followed by 60 min of incubation in 79

a shaking water bath at 40 °C . The volume was raised to 25 mL with Tris–Maleate buffer (pH 6.9) and 5 mL of Tris–Maleate buffer containing 2.6 IU of α-amylase from porcine pancreas (ref. A3176, Sigma) were then added to each sample. Immediately the sample was transferred to a shaking water bath at 37 °C to start starch hydrolysis. Aliquots (0.5 mL) were taken every 30 min from 0 to 180 min. Each aliquot was immediately put in a boiling water bath for 10 min to inactivate αAmylase, and kept at 4 °C condition until the end of incubation time. Then 1.5 mL of 0.4 M sodium– acetate buffer, pH=4.75, and 30 μL of amyloglucosidase from Aspergillus niger (ref. 102 857, Roche) were added to aliquots. Digested starch were then hydrolyzed into glucose by incubating samples at 60 °C for 45 min in a shaking water bath. Finally, glucose concentration was measured using the D-glucose assay kit (K-GLUC, Megazyme, Co. Wicklow, Ireland). The rate of starch digestion was expressed as a percentage of the total starch hydrolyzed at different times (30, 60, 90, 120, 150 and 180 min).

To calculated digestion kinetics and expected glycemic index (eGI), a non-linear model established by Goni et al. (1997) was applied as follows: 𝐶 = 𝐶∞ (1 − 𝑒 −𝑘𝑡 ), where C is the percentage of starch hydrolyzed at time t, C∞ is the maximum concentration of hydrolyzed starch, k is the kinetic constant and t is chosen time (min). The hydrolysis index (HI) was calculated by dividing the area under the hydrolysis curve of each starch sample by the corresponding area obtained from the reference sample (white bread). The eGI was calculated using the equation described by Goni et al. (1997): eGI = 39.71 + 0.549×HI

80

3.2.8. Statistical analysis

SPSS (version 19, Endicott, NY) was used for analysis of variance (ANOVA) by Tukey's multiple range test at p < 0.05. OriginPro, version 8.5.0 (OriginLab, Northampton, MA) was used for parameter estimating and figure drawing.

3.3. Results and discussion 3.3.1. Effect of BF/MF addition on textural and cooking properties

Cooking and textural properties of all samples were shown in Table 3.1. Cooking yield varied little ranging from 2.63 (WF/M-40) to 2.92 (M-20), and cooking loss from 6.37 % (B-10) to 8.53 % (M40). Similarly Baik & Lee (2003) reported cooking loss ranging from 7.7 to 9.2 % in wheat noodles. Cooking loss is considered as a measure of resistance of the noodles to disintegration upon prolonged cooking, and low cooking loss is preferred for high quality noodles (Tan et al. 2009). A lower degree of protein network may account for the high cooking loss as a result of less ability to hold flour components together (Baik & Lee 2003). As a special protein in wheat, gluten works as the most important protein to bind other components together to form a dough. 60 % BF addition to WF, which means 60 % reduction of gluten, could just meet the minimum requirements for dough formation to make noodles, while preliminary testing indicated that 60 % MF addition to WF could not form a dough. Thus protein in MF may be weaker than BF, explaining a significant high cooking loss in M-40 in contrast to other samples.

Deformation and tension properties of cooked noodles were analyzed by compression and tensile 81

tests. Tensile testing provides indications of how the samples hold together during cooking by evaluating the elasticity and breaking strength, and reflects cooking tolerance and cooking quality of the noodles (Bhattacharya et al. 1999). Noodle firmness ranged from 3630 (B-60) to 4686 (B-10) g, and tensile force from 11.7 (M-40) to 18.2 (B-5) g. Both BF and MF addition led to softer wheat noodles, while at higher than 20 % replacement proportion, more significant differences were found. Clearly higher amounts of BF and MF addition tended to contribute a lesser strength to hold all components together as a result of gluten reduction. Each type of noodle has its own optimum protein range, and protein content correlates positively with noodle firmness, while sometimes negatively with elasticity (Fu 2008). Bhattacharya et al. (1999) and Zhu et al. (2010) showed that noodle cooking and textural properties could be predicted by flour pasting, swelling and textural properties. However, in our study, we found no correlations among them. Possibly because the limited sample size (n = 10) and narrow data variation range make it difficult to identify any relationships.

3.3.2. Effect of BF/MF addition to color of cooked noodles

BF addition to WF contributed to a lower L, higher a, and similar b, while MF addition lead to lower L, similar a, but higher b (Table 3.2). These can be attributed to natural differences in flour colors. Besides, fragments of bran and enzymatic browning effects during dough processing are also thought to be two major sources to influence noodle color, (Oh et al. 1985). Salt inside noodles could lower enzymatic browning as a result of oxidative activity reduction (Fu 2008; Oh et al. 1985), while protein content had greater impact on noodle sheet color than flour color, and brightness of

82

noodle sheets is strongly affected by water absorption (Wang et al. 2004). After cooking, a noticeable reduction of L and a occurred.

3.3.3. In vitro digestion

Noodle hydrolysis rates of different noodle mixtures are shown in Fig 3.1. Hydrolysis rate of all samples increased steeply within the 30 min, then rose slowly and finally remained almost the same level. This could be attributed to all sample contains higher levels of rapidly digestible starch (Ge et al. 2014). Starch hydrolysis parameters of all samples were calculated (Table 3.3). Kinetic constant (k) reflects the rate of hydrolysis, indicates release speed of glucose from test materials (Chung et al. 2008). Ranging from relatively low 0.043 (B-10 & M-5) to 0.087 (M-20), a low value of k indicated slow release of glucose from starch. Ge et al. (2014) reported a low k value (0.070.08) for six starch noodles. Hydrolysis index (HI) is common to be used in comparison of starch digestibility in different food materials. HI ranged from 84.3 (M-10) to 98.2 (B-60) and were used for eGI calculation. eGI of all samples were generally high, ranging from 103 (B-10 & M-10) to 113 (B-60), indicating noodles were high GI foods (≥70) (Bharath Kumar & Prabhasankar 2014). Pure WF noodle (WF) had an eGI of 103. We noticed that high amount (over 20 %) of BF and MF addition to WF noodles tended to increase eGI (B-60, 113; M-40, 111), while 10 % BF & MF addition both could lower eGI but only by 3. Both raw BF and MF were thought to have high resistant starch. However, it was highly unstable when made into food products. Resistant starch content could be decreased from hulled to dehulled (Chen et al. 2010), and food processing leads to higher starch digestibility (Berti et al. 2004). Thus their impact on noodle digestibility might be

83

limited, as one possible explanation for high estimated glycemic index. Skrabanja et al. (2001) found resistant starch decreased with high BF ratio in dough for bread making, and suggested the smaller size of BF starch granules with higher surface area had higher water binding capabilities, and thus might be more easily to completely gelatinize and be sensitive to amylolytic enzymes. However, that is not a good explanation for our results as BF and MF both had larger granule size than WF. Thus this may simply relate to higher resistant starch content in BF and MF, leading to a lower eGI. However, when their proportion increases to a certain amount (20 %), the large reduced gluten can only provide limited protective addition not only for BF/MF starch granules but also WF starch granules, making starch granules easier to gelatinize and increase eGI. Due to higher starch-protein interaction, gluten presence in white flour may account for the lower glycemic response compared with gluten-free foods (Jenkins et al. 1987). This may be supported by the tension and compression results, as high addition of BF/MF lead to a looser noodle texture. Moreover, we observed high proportion of BF and MF addition leading to noodles strands being easier to break, which may also mean more surface area for granule swelling.

3.4. Conclusion

Increasing addition of BF and MF to WF lead to greater decrease in lightness and tension strength, but did not significantly influence cooking yield and cooking loss. Noodle textures may be important on in vitro digestibility, and thus higher than 20 % proportion of BF/MF blended with WF led to higher eGI. However, small amount of BF and MF (less than 10%) can lead to lower eGI, and may explained by more resistant starch content involved while noodle texture were maintained.

84

Table 3.1. Cooking and texture properties of noodle samples

Cooking property

Noodle texture

Sample Cooking yield

Cooking loss (%)

Tensile (g)

Firmness (g)

WF

2.63a

7.12ab

17.4±0.5d

4487cd

B-5

2.77ab

6.75a

18.2±0.5d

4326cd

B-10

2.75ab

6.37a

16.6±0.5cd

4686d

B-20

2.77ab

6.82a

16.6±0.3cd

3900ab

B-40

2.73ab

7.58ab

13.3±2.2ab

3824a

B-60

2.64a

7.89ab

12.5±2.1ab

3630a

WF

2.63a

7.12ab

17.4±0.5d

4487cd

M-5

2.62a

7.13a

16.2±0.4cd

4232bc

M-10

2.66a

7.85ab

14.6±0.6bc

4306cd

M-20

2.92b

7.66ab

13.2±0.4ab

3714a

M-40

2.63a

8.53b

11.7±0.7a

3634a

WF: wheat noodle; B-5 to B-60 and M-5 to M-40 refers to buckwheat (BF) / millet (MF) proportion (5-60 %, 5-40 %) in the noodle mixtures. All data are mean value, and tensile force data is expressed as mean ±standard deviation of mean; same superscripts indicate data that do not differ at significance level p < 0.05.

85

Table 3.2. Color analysis of dried and cooked noodles Dry noodle color

Cooked noodle color

Sample L

a

b

L

a

b

WF

80.4f

1.4a

13.5ab

74.6ef

-0.2f

10.9ab

B-5

77.7cde

1.5a

14.3bc

74.2de

-1.9c

12.7cd

B-10

77.6cde

1.5a

14.2bc

73.7d

-1.2e

12.0bc

B-20

75.2ab

2.1b

14.0ab

72.7c

0.1fg

11.7bc

B-40

75.4abc

2.5c

14.3bc

70.8b

0.2g

10.9ab

B-60

73.8a

2.8d

12.9a

68.5a

-1.5d

10.3a

WF

80.4f

1.4a

13.5ab

74.6ef

-0.2f

10.9ab

M-5

79.8ef

1.2a

15.5cd

75.2f

-2.8a

13.8de

M-10

77.2bcd

1.5a

16.8e

74.4def

-1.6d

13.6de

M-20

80.2f

1.3a

16.1de

74.4def

-1.2e

14.3e

M-40

78.6def

1.4a

17.1e

74.7ef

-2.4b

17.4f

WF: wheat noodle; B-5 to B-60 and M-5 to M-40 refers to buckwheat (BF) / millet (MF) proportion (5-60 %, 5-40 %) in the noodle mixtures. Values with same superscript are not significantly different (p ＜ 0.05).

86

Table 3.3. Model parameters, hydrolysis index and estimated glycemic index of cooked noodles. Sample

Total starch (%)

C∞

k

Calculated HI

eGI

WF

79.5

91.4±0.1b

0.053a

121±1.0b

106±0.5b

B-5

79.3

91.3±0.3b

0.047a

120±0.8b

105±0.4b

B-10

79.0

88.9±0.6b

0.043a

115±0.6a

103±0.3a

B-20

78.6

95.0±0.1c

0.049a

125±0.2c

108±0.1c

B-40

77.7

97.0±0.4de

0.064b

132±0.3d

112±0.2d

B-60

76.7

98.2±0.0e

0.064b

133±0.1d

113±0.0d

WF

79.5

91.4±0.1b

0.053a

121±1.0b

106±0.5b

M-5

79.7

91.6±0.3b

0.043a

119±0.5b

105±0.3b

M-10

79.8

84.3±0.0a

0.070bc

115±0.3a

103±0.2a

M-20

80.2

95.4±0.2cd

0.087d

133±0.5d

113±0.3d

M-40

80.8

94.8±1.0c

0.077c

131±1.4d

111±0.8d

WF: wheat noodle; B-5 to B-60 and M-5 to M-40 refers to buckwheat (BF) / millet (MF) proportion (5-60 %, 5-40 %) in the noodle mixtures. HI: hydrolysis index; eGI: estimated glycemic index. Data of C∞, k, Calculated HI and eGI are expressed as mean ± standard deviation of mean; same superscripts indicate data that do not differ at significance level p < 0.05.

87

Fig 3.1. In vitro starch hydrolysis rate of noodles made by buckwheat-wheat and millet-wheat flour blends. (A) buckwheat flour (BF) portion ranges from 0 to 60 %. (B) millet flour (MF) portion ranges from 0 to 40 %.

88

References

Angioloni, A. and Collar, C. (2011). Nutritional and functional added value of oat, Kamut®, spelt, rye and buckwheat versus common wheat in breadmaking. Journal of the Science of Food and Agriculture, 91, 1283-1292. Baik, B. K. and Lee, M. R. (2003). Effects of starch amylose content of wheat on textural properties of white salted noodles. Cereal Chemistry, 80, 304-309. Berti, C., Riso, P., Monti, L. D. and Porrini, M. (2004). In vitro starch digestibility and in vivo glucose response of gluten–free foods and their gluten counterparts. European Journal of Nutrition, 43, 198-204. Bharath Kumar, S. and Prabhasankar, P. (2014). Low glycemic index ingredients and modified starches in wheat based food processing: A review. Trends in Food Science & Technology, 35, 32-41. Bhattacharya, M., Zee, S. Y. and Corke, H. (1999). Physicochemical properties related to quality of rice noodles. Cereal Chemistry, 76, 861-867. Chen, L., Liu, R., Qin, C., Meng, Y., Zhang, J., Wang, Y. and Xu, G. (2010). Sources and intake of resistant starch in the Chinese diet. Asia Pacific Journal of Clinical Nutrition, 19, 274-282. Chillo, S., Laverse, J., Falcone, P. M., Protopapa, A. and Del Nobile, M. A. (2008). Influence of the addition of buckwheat flour and durum wheat bran on spaghetti quality. Journal of Cereal Science, 47, 144-152. Chung, H. J., Liu, Q., Donner, E., Hoover, R., Warkentin, T. D. and Vandenberg, B. (2008). Composition, molecular structure, properties, and in vitro digestibility of starches from newly released Canadian pulse cultivars. Cereal Chemistry, 85, 471-479. 89

Collado, L. and Corke, H. (1996). Use of wheat-sweet potato composite flours in yellow-alkaline and white-salted noodles. Cereal Chemistry, 73, 439-444. Collins, J. and Pangloli, P. (1997). Chemical, physical and sensory attributes of noodles with added sweetpotato and soy flour. Journal of Food Science, 62, 622-625. Frei, M., Siddhuraju, P. and Becker, K. (2003). Studies on the in vitro starch digestibility and the glycemic index of six different indigenous rice cultivars from the Philippines. Food Chemistry, 83, 395-402. Fu, B. X. (2008). Asian noodles: history, classification, raw materials, and processing. Food Research International, 41, 888-902. Ge, P., Fan, D., Ding, M., Wang, D. and Zhou, C. (2014). Characterization and nutritional quality evaluation of several starch noodles. Starch - Stärke, 66, 880-886. Goni, I., Garcia Alonso, A. and Saura Calixto, F. (1997). A starch hydrolysis procedure to estimate glycemic index. Nutrition Research, 17, 427-437. Hu, P. S., Zhao, H. J., Duan, Z. Y., Zhang, L. L. and Wu, D. X. (2004). Starch digestibility and the estimated glycemic score of different types of rice differing in amylose contents. Journal of Cereal Science, 40, 231-237. Ikeda, K. and Asami, Y. (2000). Mechanical characteristics of buckwheat noodles. Fagopyrum, 17, 67-72. Jenkins, D., Thorne, M. J., Wolever, T., Jenkins, A. L., Rao, A. V. and Thompson, L. U. (1987). The effect of starch-protein interaction in wheat on the glycemic response and rate of in vitro digestion. The American Journal of Clinical Nutrition, 45, 946-951. Jenkins, D., Wolever, T., Taylor, R. H., Barker, H., Fielden, H., Baldwin, J. M., Bowling, A. C.,

90

Newman, H. C., Jenkins, A. L. and Goff, D. V. (1981). Glycemic index of foods: a physiological basis for carbohydrate exchange. The American Journal of Clinical Nutrition, 34, 362-366. Kreft, I. and Skrabanja, V. (2002). Nutritional properties of starch in buckwheat noodles. Journal of Nutritional Science and Vitaminology, 48, 47-50. Li, S. and Zhang, Q. H. (2001). Advances in the development of functional foods from buckwheat. Critical Reviews in Food Science and Nutrition, 41, 451-464. Lu, H., Yang, X., Ye, M., Liu, K. B., Xia, Z., Ren, X., Cai, L., Wu, N. and Liu, T. S. (2005). Culinary archaeology: millet noodles in late neolithic China. Nature, 437, 967-968. Oh, N., Seib, P., Ward, A. and Deyoe, C. (1985). Noodles IV. Influence of flour protein, extraction rate, particle size, and starch damage on the quality characteristics of dry noodles. Cereal Chemistry, 62, 441-446. Ovando Martinez, M., Sáyago Ayerdi, S., Agama Acevedo, E., Goñi, I. and Bello Pérez, L. A. (2009). Unripe banana flour as an ingredient to increase the undigestible carbohydrates of pasta. Food Chemistry, 113, 121-126. Saifullah, R., Abbas, F., Yeoh, S. and Azhar, M. (2009). Utilization of green banana flour as a functional ingredient in yellow noodle. International Food Research Journal, 16, 373-379. Saleh, A. S. M., Zhang, Q., Chen, J. and Shen, Q. (2013). Millet grains: nutritional quality, processing, and potential health benefits. Comprehensive Reviews in Food Science and Food Safety, 12, 281-295. Shogren, R., Hareland, G. and Wu, Y. (2006). Sensory evaluation and composition of spaghetti fortified with soy flour. Journal of Food Science, 71, S428-S432.

91

Shukla, K. and Srivastava, S. (2014). Evaluation of finger millet incorporated noodles for nutritive value and glycemic index. Journal of Food Science and Technology, 51, 527-534. Skrabanja, V., Liljeberg Elmståhl, H. G., Kreft, I. and Björck, I. M. (2001). Nutritional properties of starch in buckwheat products: studies in vitro and in vivo. Journal of Agricultural and Food Chemistry, 49, 490-496. Tan, H. Z., Li, Z. G. and Tan, B. (2009). Starch noodles: history, classification, materials, processing, structure, nutrition, quality evaluating and improving. Food Research International, 42, 551-576. Torbica, A., Hadnađev, M. and Dapčević Hadnađev, T. (2012). Rice and buckwheat flour characterisation and its relation to cookie quality. Food Research International, 48, 277283. Vijayakumar, T. P., Mohankumar, J. B. and Srinivasan, T. (2010). Quality evaluation of noodles from millet flour blend incorporated composite flour. Journal of Scientific and Industrial Research, 69, 48-54. Wang, C., Kovacs, M. I., Fowler, D. and Holley, R. (2004). Effects of protein content and composition on white noodle making quality: color 1. Cereal Chemistry, 81, 777-784. Wei, Y. M., Zhang, G. Q. and Li, Z. X. (1995). Study on nutritive and physico-chemical properties of buckwheat flour. Food / Nahrung, 39, 48-54. Wu, K., Gunaratne, A., Collado, L. S., Corke, H. and Lucas, P. W. (2015). Adhesion, cohesion, and friction estimated from combining cutting and peeling test results for thin noodle sheets. Journal of Food Science, 80, E370-E376. Zabidi, M. A. and Aziz, N. A. A. (2009). In vitro starch hydrolysis and estimated glycaemic index

92

of bread substituted with different percentage of chempedak (Artocarpus integer) seed flour. Food Chemistry, 117, 64-68. Zhu, F., Cai, Y. Z. and Corke, H. (2010). Evaluation of Asian salted noodles in the presence of Amaranthus betacyanin pigments. Food Chemistry, 118, 663-669.

93

Chapter 4. Thermal, rheological and gelling properties of mung bean starch blends with other starches

Abstract

Blending diverse starches is an alternative and easy way to improve starch quality rather than chemical modification. Potato (PT), sweet potato (SP), rice (R) and sorghum (S) starch were added to mung bean (MB) starch with mixing ratio from 0 % to 100 %, and their gelling, thermal and rheological properties were studied and compared. Thermal results showed endotherms of mixed components merged into one and tended to gelatinize separately. Swelling volume (SV) of the mixtures were in the middle of two pure starches. However, great variation was found in blended starch pasting, gel texture and rheological properties, suggesting interactions occurred during heating and cooling. Granule size distribution and molecular weight might play an important role for the non-additive changes in addition to amylose content. However, amylose structure, together with amylopectin may also have some influence and should be further studied for a deeper interpretation.

94

4.1. Introduction

Not only considered as an important carbohydrate in human diet, starch also has wide applications in food industry. However, native starches are not optimal in many situations and chemical modifications, and are often involved in order to improve its performance (Jacobs & Delcour 1998). However, natural food components are desirable in food industries and market nowadays and therefore new ways to improve the properties of native starches avoiding chemical modifications are of interest to food producers (Obanni & Bemiller 1997; Ortega Ojeda & Eliasson 2001). Food additives are commonly used as one way to optimize the process operation and achieve desirable properties (Gunaratne & Corke 2007). However, for developing countries, both chemically modified starches and food additive are sometimes not economical or available (Zhu & Corke 2011). In such conditions another way is by mixing different starches together, which may have more possibilities, and has already been studied by many researchers, such as for sweet potato and wheat starch (Zhu & Corke 2011), and potato and Amaranthus starches (Gunaratne & Corke 2007). It has even been related with a pudding patent (Kern & Stute 1994), and suggested as a possibility to reduce retrogradation (Obanni & Bemiller 1997). A mixture with less swelling and less retrogradation during storage than normal rice starch might be used as a retrogradation retarding agent in the processed rice products (Yao et al. 2003). Puncha Arnon et al. (2008) studied properties of potato starch blends with mung bean starch and canna starch blends with rice starch.

Rapid Visco Analyzer (RVA) is an instrument commonly used to evaluate starch pasting properties, and is often combined with a gel texture study by a Texture Profile Analyzer (TPA), while a Differential Scanning Colorimeter (DSC) thermal is used for determining gelatinization properties. 95

From the food standpoint, rheology has been defined as the study of the deformation and flow of the raw materials, processing intermediates and the final food products and a knowledge of the rheological properties of foods is of great significance in processing, quality control, sensory evaluation and structural studies (Chun & Yoo 2004; Sopade & Kassum 1992). All these instruments are useful to evaluate starch quality for diverse food applications.

Mung bean starch as a valuable source of starch, is considered to be the best raw material to make transparent noodles (Li et al. 2011a), and mung bean noodle is consequently regarded as the best among starch noodles (Ge et al. 2014; Li et al. 2008; Tan et al. 2009). However, its relatively high price, low production supply and tedious processing methods limit its applications in food industry (Kasemsuwan et al. 1998). Thus in this study, we tried combinations of mung bean starch mixture with other common crop starches (potato, sweet potato, sorghum and rice) and investigated its impact on pasting, gel texture, thermal, swelling and rheological properties, and may provide suggestions for formulations of starch mixtures with required functionalities.

4.2. Materials and methods 4.2.1. Materials

Mung bean (Split Green Beans, Yu Pin King, China), rice (Thai Premium Rice, Select, Thailand), sweet potato (Orange Sweet Potato, China), potato (Local Potato, Hong Kong) and sorghum (Organic Sorghum Grain, Seesang, China) were bought in local supermarket and starch were extracted using methods of Bao et al. (2005), Collado et al. (1999), Beta & Corke (2001). The starch

96

were ground to pass 212 μm sieve after drying in a 35 °C oven, and were coded MB, R, SP, PT, S, respectively. Amylose contents were determined by iodine method of Chrastil (1987).

4.2.2. Particle size

Particle size distribution of flours was measured by Mastersizer 2000 (Malvern Instruments, Malvern, UK). Samples were mixed with Milli-Q® water into the small volume dispersion unit before being dispersed at a stirrer speed of 3,000 rpm.

4.2.3. Pasting property

Mixed starch (2.0 g, 14 % moisture basis) was mixed with distilled water (25.0 g) in a canister and loaded to Rapid Visco Analyzer (RVA) model 3D (Newport Scientific, Warriewood, Australia) using STD 1 heating and cooling profile. The temperature was at 50 oC for 1 min, then increased at 12 oC/min to 95 oC, held at 95 oC for 2.5 min, reduced at 12 oC/min to 50 oC, and held for 2 min. Peak viscosity (PV), hot paste viscosity (HPV), cold paste viscosity (CPV), breakdown (BD = PV – HPV), and setback (SB = CPV – HPV) were recorded.

4.2.4. Gel texture

Texture profile analysis was determined on the flour gel made in the RVA testing using a TA-XT2 Texture Analyzer (Stable Micro Systems, Godalming, England). After RVA testing, the paddle was removed and the starch paste in the canister was covered by Parafilm and stored at 4 oC for 24 h.

97

The gel was compressed at a speed of 1.0 mm/s to a distance of 10 mm with a 7 mm cylindrical probe with a flat end. Texture parameters of hardness (HD, g), adhesiveness (ADH, g·s), and cohesiveness (COH) were derived from the instrument software. HD is defined as the maximum force to fracture the gel. ADH is the area under the curve as the probe moves back to initial location. COH was defined as the ratio of the positive force areas during the second compression to that of the first compression.

4.2.5. Swelling volume

Swelling volume (Crosbie 1991) was determined following the method of Collado & Corke (1999). Starch samples (0.35 g, db) were weighed into 125×16 mm Pyrex tubes with 12.5 mL distilled water. The tubes were equilibrated at 25 °C for 5 min, and transferred to a 92.5 °C water bath with occasional mixing. Samples were cooled in ice water for 1 min, placed in a 25 °C bath for 5 min and centrifuged at 1000×g for 15 min. The height of the gel was measured and converted to volume of gel per unit dry weight of the sample.

4.2.6. Thermal property

Gelatinization parameters were measured using a TA 2920 Modulated DSC Thermal Analyzer differential scanning calorimeter (DSC) equipped with a thermal analysis data station (TA Instruments, Newcastle, DE). Mixed starch (2 mg, db) was weighed onto the aluminium DSC pan and distilled water (6 µL) was added with a microsyringe. Pans were sealed and allowed to stand for 2 h at room temperature. The scanning temperature range and heating rate were 30 °C to 120 °C 98

and 10 °C min-1, respectively, using an empty pan as reference. Onset (To), peak (Tp), and completion (Tc) temperatures, and enthalpy change (ΔH) of gelatinization were recorded by a Universal Analysis Program, Version 2.5H (TA instruments, Newcastle, DE).

4.2.7. Steady shear property

Steady shear properties were measured by a rheometer (Physica MCR 301, Anton Paar, Austria) at 25 °C according to the method described by Kong et al. (2010). Starch pastes with 5 % (dry weight basis) of total solids were prepared by heating the starch dispersions at 90 °C for 30 min before transferring onto the plate. Sample edges were covered with an oil layer to prevent evaporation. Sample was equilibrated at 25 °C for 2 min. The pastes were sheared from 0.1 to 1000 s-1 (upward), then reduced from 1000 to 0.1 s-1 (downward) to describe the flow behavior, and the data were fitted to the power law (Eq. 1) models: σ = Kγn

(1)

where, σ is the shear stress (Pa), γ the shear rate (s-1), K the consistency coefficient (Pa Sn), n the flow behavior index (dimensionless).

4.2.8. Dynamic oscillatory measurement

Dynamic oscillatory measurements were determined follow the procedures described by Kong et al. (2010) with minor modifications. Starch suspensions made of 20 % solids on a dry weight basis were loading on the bottom platen of the rheometer. Small-deformation experiments on shear were performed using the same instrument as in the previous section, which was equipped with a parallel 99

plate measuring geometry of 25 mm diameter. The gap size, strain and frequency were set at 1000 mm, 2 % (within the linear viscoelastic region) and 1 Hz, respectively. Starch suspensions were heated from 40 to 90 °C at a scan rate of 1 °C/min and then cooled from 90 to 25 °C at the same rate, all along the sample edges being covered with a thin layer of soya oil to minimize evaporation. The parameters of storage modulus (G'; solid component of the network), loss modulus (G''; liquid component) and their ratio, loss tangent (tan δ = G'' / G' ), were recorded as a function of temperature. Following the temperature sweeps, materials were held undisturbed for 5 min at 25 °C to achieve a state of equilibrium and then subjected to a frequency sweep at the same temperature over the range of 0.1 to 25 Hz to once more monitor the aforementioned viscoelastic functions.

4.2.9. Statistical analysis

All tests were run at least in triplicate. SPSS (version 19, Endicott, NY) was used for analysis of variance (ANOVA) by Tukey's multiple range test at p < 0.05, and Pearson correlation analysis for determining the relationship between different variables at p＜0.05 and p＜0.01. OriginPro (version 8.5.0, OriginLab, Northampton, MA) was used for figure drawing and model fitting.

4.3. Results and discussion 4.3.1. Amylose content

Amylose content of five pure starches were shown in Table 4.1. Mung bean (MB), sorghum (S) and sweet potato (SP) starch all showed similar high amylose content (35.0-36.6 %), while rice (R)

100

starch was little lower (30.5 %) and potato (PT) was the lowest (19.8 %).

4.3.2. Particle size

Starch granular properties and characteristics were the major factors for the starch rheological behavior, followed by amylose (Lii et al. 1996). Granule size is important for digestion, pasting (Mahasukhonthachat et al. 2010), and rheological properties (Rao & Tattiyakul 1999). Granule sizes of R, SP, S, MB and PT varied between 2-23, 3-35, 7-46, 8-46, and 12–79 μm respectively, and could be roughly separated into three categories: small, medium and large size (Fig 4.1 and Table 4.1).

4.3.3. Thermal properties

Tp of five pure starches were 67.5 (PT), 69.2 (MB), 69.6 (R), 71.2 (S), 76.2 (SP) °C (Table 4.2). Gelatinization range of pure MB starch was similar to MB starches extracted from a cultivar (Jilv 9239-8) in Li et al. (2011b), and also very close to values reported in Puncha-arnon et al. (2008). Gelatinization behavior seemed to have no relationship with granule size for pure starches (Li & Yeh 2001). When mixed in excess water ( ＞ 65 % water content), individual starch tends to gelatinize independently, especially when large difference in gelatinization temperature exists (at least 6 °C) between components, thus avoiding overlap (Waterschoot et al. 2015). This was supported by studies observing two endotherms in starch blends, e.g. canna and potato (Punchaarnon et al. 2008), and cassava and yam (Karam et al. 2006). Overlapped endotherms were also observed in many studies such as PT and maize starch mixture (Obanni & Bemiller 1997), and canna 101

and maize (Puncha-arnon et al. 2008). The differences in Tp among P, R, S and MB starch here were relatively small (1.7, 0.4 and 3.0 °C, respectively), thus gelatinization peaks of individuals merged into one. Only the Tp difference between SP and MB was high at 7.0 °C (over 6 °C), however, it was still hard to separate distinguish two peaks in the endotherm. The reason may be that both starches had large gelatinization ranges (MB, 15.5 °C from 61.9 to 77.4 °C; SP, 14.6 °C from 69.2 °C to 83.8 °C) and shared a large overlapped area from 69.2 °C to 77.4 °C, covering more than 50 % for each endotherm, thus making it difficult to separate them well. Generally △H of mixtures were close to summation of each components and showed a relatively good linearity (R2 = 0.9633 in P/MB mixtures; R2 =0.9294 in SP/MB mixtures; R2 = 0.9635 in S/MB mixtures) in this study, and this was supported by Ortega Ojeda & Eliasson (2001) who claimed that gelatinization occurs independently based on calculations of gelatinization enthalpies. Similar results happened in canna and maize (Puncha-arnon et al. 2008). Only R/MB mixture had different behavior, and all △H showed no big difference. This non-additive behavior was possibly due to the small difference (1.7 J/g) between pure R and MB starch, making it difficult to separate each mixture clearly. Obanni & Bemiller (1997) reported possible interactions between potato (large granule) and rice (small granule) starch due to its unusual △H. However, we did not find any clear evidence in this study for molecular interactions in large and small granule (P/MB blends), medium and medium granule (SP/MB, S/MB blends), or medium and small granule (R/MB blends).

4.3.4. Swelling volume

Swelling volume (SV) is a rapid method with small sample size to predict noodle eating quality, and

102

reflects the water-holding ability of starch (Bao et al. 2002; Crosbie 1991). During gelatinization, starch granules swell to several times their initial volume (Singh et al. 2007). Swelling volume of starch is affected by amylose content and the structure of amylopectin (Huang et al. 2007). Amylose content correlated negatively with SV (Bao et al. 2004), while higher proportions of long chains of amylopectin or more strongly associated amylopectin chains within crystalline regions could contribute to high swelling properties (Hoover et al. 1997; Sasaki & Matsuki 1998). SV of five pure starches were 22.5 (R), 25.2 (MB), 32.8 (S), 36.0 (SP), 62.5 (PT) mL/g (Table 4.3). PT starch had much higher SV value than others and reached the maximum value for this test (62.5 mL/g) together with P-80, while rice showed the lowest value (22.5 mL/g). Increasing PT mixing ratio lead to increasing SV of P/MB mixtures. The similar SV value of R and MB made their mixtures show very little difference in SV. SV data of SP/MB and S/MB data were in the middle of the two individual starches.

4.3.5. Pasting properties

Pasting properties are dependent on the rigidity of starch granules, which in turn affect the granule swelling potential and the amount of amylose leaching out in the solution (Morris 1990). Pasting data of all starches or mixtures are shown (Table 4.3). The increase in viscosity during heating of starch-water system is mainly due to swollen granules, and the decrease during further continuous heating at high temperature (95 °C) is caused by weakening, disintegration and disruption of the gelatinized granules, while increase in cooling related to the retrogradation, or aggregation of the amylose fraction (Hagenimana & Ding 2005; Thiewes & Steeneken 1997). Wide ranges were

103

recognized in all parameters for five pure starches, with PV ranging from 69 (R) to 227 (PT) RVU, HPV from 48 (S) to 184 (PT) RVU, BD from 6 (R) to 64 (S) RVU, CPV from 105 (R) to 259 (PT) RVU, SB from 43 (R) to 79 (MB) RVU. Significantly different from other four starches, potato (PT) starch had very high viscosity, and this was related to its phosphorus. Phosphorus is always covalently bound to the amylopectin molecules and leads to potato starch with a relatively lower amylose content but high viscosity (Zaidul et al. 2007), and higher phosphor leads to higher peak viscosity (PV) (Sandhu et al. 2010; Zaidul et al. 2007). Increasing PT ratio leads to increasing PV, HPV and CPV values for PT/MB mixtures, but PV of the mixtures were lower than calculated summation of individual starches (except PT-80), while both HPV and CPV showed higher value. This non additive behavior indicated interactions existed among the large granule (PT) and medium granule (MB). Waterschoot et al. (2015) pointed out that smaller granules (MB) will fill the voids between the larger granules (PT), producing a packed system and reduce the swelling power, while swelling power was significantly positively correlated with PV (Crosbie 1991). In addition pure MB starch had an earlier To and lower △H than PT starch and will swell earlier surrounding big PT starch granules, thus possibly limiting PT starch swelling and giving more heat resistance to PT starch granule. This swelling inhibition behavior could also contribute to less BD in the blends as granule rigidity was better maintained, which leads to non-additive effects on HPV (Park et al. 2009; Waterschoot et al. 2015). Most blends showed higher CPV than summation of individual components except P-80. A similar phenomenon was observed in PT and waxy maize starch mixtures (Park et al. 2009). However, as PT had lower amylose content than MB, this non-additive behavior results in CPV perhaps more affected by increasing phosphor content in PT starch, and entanglements of the leached starch molecules and leached molecules with gelatinized starch

104

granules (Zhang et al. 2011).

As small granule representative, pure rice (R) starch showed the lowest pasting values (PV, BD, CPV and SB). PV of R/MB mixtures were all lower value than the sum of the two individual starches, indicating interactions between them. This again may be attributed to the voids between MB granules filling with small R starch granules and thus creating a compact system reducing the swelling power and PV, similar to the P/MB mixture as described above. However, differing from PT/MB mixtures, R/MB mixtures all showed lower HPV than the sum of individual starches, and the HPV of the mixtures tended to shift to pure R starch especially at high ratio (R-60 and R-80). This is consistent with results for canna and R starch mixtures in Puncha-arnon et al. (2008). R had lower amylose content than MB and had no phosphorus but much smaller granule size than MB. It should swell earlier than MB due to low △H. Moreover its low SV and low viscosity nature may give limited protection for MB granule rigidity and thus lead to different results. CPV of R/MB blends showed non-additive effects and were in the middle of two single starches but lower than the sum of individual components, indicating R addition significantly weakened the gel network. A relatively low value of CPV could usually be explained by a reduction in swelling power and thus carbohydrate leaching of one starch by the other (Puncha-arnon et al. 2008; Waterschoot et al. 2015). Moreover, R had lower amylose content than MB, and lower amylose content starch exhibited lower tendency for starch paste retrogradation (Kaur et al. 2008; Kong et al. 2010). Puncha-arnon et al. (2008) reported pieces of canna starch granules were covered by swollen rice starch granules, limiting their high ability to retrograde, leading to a lower CPV in blends.

105

SP and S had close or similar particle distribution range with MB and also no significant difference in amylose content, but not the same trend in pasting properties. PV of MB, SP and S were very close, and most of their mixtures did not differ significant from each other. They both showed good linearity or additive effects for HPV (R2 = 0.9933 in S/MB, all values of SP/MB were similar). The only difference occurred for CPV, as SP/MB mixture gradually shifted from the CPV of MB to that of SP, while S/MB mixtures had a sudden decrease (S-20) and all showed lower values than the sum of individual components, indicating retrogrdation was partly restricted. MB should swell much earlier than SP due to a lower Tp and △H, while S shall swell little later than MB due to larger △ H in a narrow gelatinization range. This difference may lead to difference in retrogradation during cooling. Besides, their difference in granules may also have some impact as Mua & Jackson (1997) reported molecular weight and structure of amylose and amylopectin contributed greatly to rheological changes during pasting. However, further studies are in still need to clearly explain this phenomenon.

4.3.6. Gel texture properties

Gel texture data for all starches and mixtures are shown in Table 4.4. MB produced a strong gel textures and had greatly different properties to the others as it had very high hardness (HD, 151 g), low adhesiveness (ADH, -4 g·s), high gumminess (GUM, 64) and chewiness (CHE, 100). Li et al. (2011b) also reported a significant higher HD and lower ADH in MB starch gel compared with SP gel. In contrast to MB, R showed a very low HD (4 g), low GUM (2), low CHE (2), but little higher COH than all others. Molecular weight of amylose fractions, and branching ratio and molecular

106

weight of amylopectin fractions could lead to gel texture differences (Mua & Jackson 1997; Mua & Jackson 1998). Addition of other starches significantly affected the gel texture by lowering HD, GUM, CHE, and increasing ADH, while COH for all samples did not differ much except in R/MB mixtures. The greatest reduction in HD (40 - 70 g) all occurred at 20 % addition of other starches to MB starch. This is in agreement with Waterschoot et al. (2015) that the largest reduction of firmness occurred at 10-25 % of one starch (lower firmness) added into the other (higher firmness), supported by studies such as on PT and R blends (Sandhu et al. 2010), and canna and R blends (Puncha-arnon et al. 2008). This non-additive reduction in swelling power and carbohydrate leaching of one starch (MB) due to other starches’ presences may well explain this phenomenon (Waterschoot et al. 2015). Puncha-arnon et al. (2008) suggested granule size played an important role and the greater differences in granule size, the larger the HD reduction. This was partial supported by our results e.g. R (smallest) starch granule had the greatest reduction effect. Furthermore Yao et al. (2003) introduced a concept ‘distribution of regional moisture content’ about the effect of uneven distribution or heterogeneity of the moisture content in individual swollen granules to explain nonadditive behavior of blended starch gel nature. This may explain the non-additive effects of blended gel texture between starch mixtures within two medium granule size (SP/MB, S/MB). However, Puncha-arnon et al. (2008) also indicated that confirmation of different interactions between leached amylose and amylopectin from each starch component should be essential for a full explanation.

4.3.7. Dynamic oscillation properties

Dynamic rheometer provides structural information by continuous assessment of dynamic moduli

107

during temperature and frequency sweep testing of the starch suspensions in a non-destructive way (Singh et al. 2003). The storage modulus (G') is measure of the energy stored in the material and recovered from it per cycle, relating to molecular events of elastic nature; while the loss modulus (G'') is a measure of the energy dissipated or lost per cycle of sinusoidal deformation, relating to molecular events of viscous nature; and the tangent (tan δ = G'' / G') denotes relative effects of viscous and elastic components in viscoelastic behavior (Gunasekaran & Ak 2000; Singh et al. 2003). Temperature sweeping of dynamic oscillation results are shown in Fig 4.2 & 4.3 and Table 4.5. During controlled heating, G' and G'' of all starches almost showed almost no change and remained horizontal before 60 °C. With amylose molecules dissolved from swollen starch particles, the starch suspension transformed into a ‘sol’ as a result of slight increase in G' (Hsu et al. 2000), before a steep increase in the range 65 to 70 °C to a maximum (G'max). Hagenimana et al. (2005) conclude three factors accounted for the initial increase to G'max: a close-packed network created by progressive swelling of starch granules, solubilized amylose molecules realized during heating and increase in gel volume. Low molecular weight amylopectin could also interact with amylose matrix to strengthen the continuous phase or network contributing to G' increment (Hsu et al. 2000). TG'max did not differ much and ranged from 72.0 (PT) to 77.4 (R), while G'max had great variation ranging from 4303 (PT) to 12200 (MB). This difference should be credited to the difference in granule structure (Hagenimana et al. 2005). The starch components, and form and size distribution of granules will affect the packing characteristics (Mendez Montealvo et al. 2008). Non additivity in G' and G'' in the mixtures were clearly shown for G'max especially for the great reduction at 20 % mixing ratio of other starches to MB starch. G' of PT/MB and SP/MB mixtures were gradually falling from MB to PT or SP. In contrast, we observed certain ratio in R/MB (R-60, R-80) can lead

108

to a lower G'max even than pure R and S starches, while S-40 to pure S are all similar. Like the pasting process, earlier swelling granules may fill in the void and limit other granules from swelling. However, granule weight might play an important role here as R granules were significantly smaller and possibly lighter than others while PT granule were larger and heavier. This might lead to R tending to appear in the upper layer of the dispersion system while PT granules settle at the bottom in a long time test. So the properties might shifted more to MB for PT/MB mixtures and to R for R/MB mixtures. SP, S and MB had similar amylose content and granule distribution range. However, gelatinization temperature of SP were higher than MB, while MB and S were very close. Waterschoot et al. (2015) stated that competition for water between starches in a blend can lead to a disproportional distribution of water between the starches and to incomplete starch swelling. Thus the competition for water of S and MB granules to swell might be particularly strong in S/MB mixtures and may have greater impact on rheological properties. G' and G'' decreased in further heating are due to extended rupture and disintegration of starch granules, melting of crystallites, and weakening of inter-chain interactions and the rigidity of the starch network due to increased molecular mobility (Kong et al. 2010; Mendez Montealvo et al. 2008). Similar to G'max, R and S addition had larger impact on G'90°C. Tan δ90 °C of pure R (0.084) and S (0.080) were lowest when compared with PT (0.161), SP (0.176) and MB (0.117), indicating pure R and S were less viscous at high temperature. This is in agreement with pasting data as pure R and S had lowest HPV.

During controlled cooling, intermolecular association in polymer-rich regions leads to a quick gel formation and amylose aggregation (Mendez Montealvo et al. 2008), and G' of all starch blends all increased, reflecting retrogradation. A big tan δ decrease was observed in all starches or mixtures,

109

indicating this gel formation. Non-additive effects occurred as G'25 °C of all blends were lower than the sum of individual components. Leached amylose in the gel largely determines G' as it forms a three-dimensional gel network (Waterschoot et al. 2015). Pure MB showed highest G', while R showed lowest. Amylose contributes to G' increment while amylopectin hinders formation of a continuous amylose network and thus decrease G' (Waterschoot et al. 2015). The greater reduction in the R/MB mixture might be due to a lower amylose content and higher amylopectin correspondingly brought to the system. Besides, uneven distribution of reassociating molecules might also exist as in heating process. R granules were much smaller and probably lighter than MB, thus R molecules may appear more on the upper part of the gel during amylose aggregation in cooling, as a result mixture properties would more likely shift to R. Similarly PT/MB mixtures had lower influence as MB might be in the upper layer, besides the high swelling volume of PT starch. Moreover, competition for water between starches can lead to a disproportional distribution of water between the starches and to incomplete starch swelling (Waterschoot et al. 2015), and may lead to different results in SP/MB and S/MB mixtures.

Dynamic frequency sweep test is widely used to gain further insights on the structure of biomacromaterials (Rao 2003). Here testing was performed over a range of 0.1-25 Hz to evaluate viscoelastic nature of gel formed after heating and cooling (Fig 4.4 & 4.5). Similar results occurred in all samples and all mixture data were between the two individual starches. In all cases, G' did not change and G'' increased rapidly. However, added starch significantly lowered down both G' and G'' with different extent. R showed greatest non additive impact and might because R granules were more likely in the upper part of the R/MB gel mixture, and representing more R gel properties.

110

4.3.8. Steady shear properties

Steady shear rheological properties decide the flow property of the material, and are critical to understanding some important engineering processes (Gunasekaran & Ak 2000). The flow curves of starch mixtures exhibited pseudoplastic, shear-thinning behavior (Figure 4.3 & 4.4), which means a curvature downwards on the shear rate axis (Thebaudin et al. 1998). All flow curve of mixtures were mostly between those of individual pure starches (Fig 4.6 & 4.7). Experimental data of flow behavior for both upward and downward curves were fitted well with power law (Table 4.6) with high R2. The value of K, as a measure of viscosity, is very dependent on type of starch, concentration and temperature, while n, a measure of the non-Newtonian behavior, is independent of the three factors (Dzuy Nguyen et al. 1998). During upward scanning, consistency indices (K) greatly decreased, and flow behavior (n) was increased when adding other starches to MB, showing the mixture became less pseudoplastic. During downward sweeping, K again decreased while n increased when adding other starches for most cases. Clearly addition of other starches could significantly change the MB’s flow behavior and may be due to influences on volume occupied by starch gel particles and gel particle size distribution as listed in Wong & Lelievre (1982).

4.4. Conclusion

Though starches tended to gelatinize individually when mixed with other starch, interactions may still occur and big changes were shown in pasting, gel texture, and rheological property changes. Granule size distribution, and amylose content could possibly explain impact of PT (big granules) 111

and R (small granules) addition to MB (medium granules). However, that is not adequate to interpret of property changes in SP/MB and S/MB mixtures as they shared similar amylose content and granule size distribution range. Further studies on granule structure, such as molecular weight, chain length, branching ratio of amylose and amylopectin should be studied and discussed to more fully explain their non-additive behaviors.

112

Table 4.1. Amylose content and particle size distribution of pure starches

Amylose content (%)

Particle size distribution range (μm)

Granule size

PT

19.8±1.0b

12-79

Large

MB

35.0±1.1a

8-46

Medium

S

36.6±0.5a

7-46

Medium

SP

35.0±2.3a

3-35

Medium

R

30.5±0.8c

2-23

Small

Sample

PT = potato; MB = mung bean; S = sorghum; SP = sweet potato; R = rice. Data of amylose content are expressed as mean ± standard deviation of mean; same superscripts indicate data that do not differ at significance level p < 0.05.

113

Table 4.2. Thermal properties of starches and their blends Sample

To

Tp a

△H

Tc a

a

MB PT-20 PT-40 PT-60 PT-80 PT

61.9±0.2 63.2±0.1b 63.8±0.2c 63.5±0.2bc 63.7±0.2bc 63.9±0.2c

69.2±0.5 67.9±0.3b 67.8±0.1b 67.6±0.4b 67.4±0.3b 67.5±0.4b

77.4±0.6 75.2±0.3a 75.0±0.9a 75.8±0.2a 76.1±1.0a 77.3±2.9a

6.0±0.6a 6.7±0.2a 7.9±0.9ab 10.2±0.7bc 12.6±1.0cd 13.8±2.2d

MB SP-20 SP-40 SP-60 SP-80 SP

61.9±0.2a 61.4±1.2a 61.7±0.3a 61.4±0.3a 65.2±0.4b 69.2±1.0c

69.2±0.5a 69.7±1.0ab 70.5±0.4b 75.5±0.5c 75.8±0.4c 76.2±0.4c

77.4±0.6a 79.6±1.4b 81.6±0.9c 83.0±0.6cd 83.5±0.9cd 83.8±0.5d

6.0±0.6a 7.7±0.4a 7.8±0.6a 10.3±0.9b 10.5±0.4b 11.7±1.8b

MB R-20 R-40 R-60 R-80 R

61.9±0.2a 63.0±0.4b 63.6±0.4bc 64.0±0.4cd 63.9±0.7c 64.8±0.4d

69.2±0.5a 69.2±0.2a 69.4±0.2a 69.6±0.1a 69.5±0.4a 69.6±0.2a

77.4±0.6a 74.9±0.6b 74.5±0.5b 74.7±0.5b 74.6±0.4b 74.1±0.4b

6.0±0.6a 4.4±0.6b 4.1±0.8b 4.6±0.7b 4.6±0.5b 4.3±0.8b

MB S-20 S-40 S-60 S-80 S

61.9±0.2a 62.5±0.2a 64.0±1.0b 64.3±0.3bc 65.6±0.3c 65.6±0.3c

69.2±0.5a 70.0±0.2ab 70.5±0.5bc 70.6±0.3bc 71.1±0.5c 71.2±0.2c

77.4±0.6a 78.0±0.5ab 79.4±1.8abc 79.2±0.4abc 81.3±0.6c 80.3±1.1bc

6.0±0.6a 7.4±1.2ab 7.8±1.9ab 8.3±0.3ab 9.7±1.1b 10.6±1.5b

PT-20 to PT-80 refers to potato starch proportion (%) in the blends, same for other blends (SP/MB, R/MB and S/MB). To: onset temperature (°C); Tp: peak temperature (°C); Tc: conclusion temperature (°C); △H: gelatinization enthalpy (J/g). All data are expressed as mean ± standard deviation of mean; same superscripts indicate data that do not differ at significance level p < 0.05.

114

Table 4.3. Pasting and swelling properties of starches and their blends Sample

PV

HPV

BD

CPV

SB

SV

MB PT-20 PT-40 PT-60 PT-80 PT

114±1.9a 129±1.4b 147±1.1c 181±0.9d 223±1.5e 227±3.9e

89±1.4a 110±0.2b 137±3.1c 170±3.9d 181±1.3d 184±7.1d

25±0.5b 18±1.6ab 10±1.9a 11±4.9a 42±0.2c 43±3.2c

167±2.0a 200±1.1b 219±3.6c 228±0.4d 223±1.3cd 259±0.5e

79±0.6cd 90±1.2d 81±0.5cd 59±3.5b 43±0.0a 76±7.5c

25.2±1.2a 31.3±0.8b 43.3±1.9c 57.0±0.5d 62.5±0.0e 62.5±0.0e

MB SP-20 SP-40 SP-60 SP-80 SP

114±1.9a 115±2.5ab 115±0.3a 113±0.4a 120±0.6b 126±0.2c

89±1.4a 91±3.3a 92±0.9a 89±0.5a 93±1.5a 94±0.4a

25±0.5bc 24±0.8ab 23±0.6a 24±0.1ab 27±0.8c 32±0.6d

167±2.0a 164±5.4a 158±1.7ab 148±0.6bc 148±1.4d 141±0.5d

79±0.6a 74±2.1b 66±0.8c 59±0.2d 55±0.1e 47±0.2f

25.2±1.2a 25.7±0.3a 27.5±0.0b 31.0±0.5c 33.3±0.6d 36.0±0.5e

MB R-20 R-40 R-60 R-80 R

114±1.9a 97±0.4b 87±1.1c 80±0.1d 74±0.7e 69±0.3f

89±1.4a 80±0.6b 70±0.6c 65±0.1d 63±1.4d 62±0.4d

25±0.5a 17±0.2b 17±0.5bc 15±0.2c 12±0.7d 6±0.6e

167±2.0a 137±0.6b 118±1.5c 113±0.4d 109±0.2de 105±1.1e

79±0.6d 57±0.1c 48±0.9b 48±0.4b 46±1.2ab 43±1.5a

25.2±1.2a 25.0±0.0a 26.0±0.5a 26.3±0.8a 25.5±1.3a 22.5±0.9b

MB S-20 S-40 S-60 S-80 S

114±1.9b 107±1.1a 106±0.3a 110±2.9ab 110±0.1ab 112±0.9ab

89±1.4a 80±2.2b 70±1.2c 62±0.3d 55±0.2e 48±0.1f

25±0.5a 28±1.1a 36±0.9b 48±2.7c 55±0.4d 64±1.1f

167±2.0a 136±1.9b 120±1.1c 114±1.5d 109±0.5d 109±1.3d

79±0.6e 57±0.4c 49±0.2a 52±1.2ab 54±0.3bc 61±1.4d

25.2±1.2a 25.8±0.3ab 27.3±0.3b 29.2±0.3c 30.5±0.5c 32.8±0.8d

PT-20 to PT-80 refers to potato starch proportion (%) in the blends, same for other blends (SP/MB, R/MB and S/MB). PV = peak viscosity (RVU); HPV = hot pates viscosity (RVU); BD = breakdown viscosity (RVU); CPV = cold paste viscosity (RVU); SB = setback (RVU); SV = swelling volume (mL/g). All data are expressed as mean ± standard deviation of mean; same superscripts indicate data that do not differ at significance level p < 0.05.

115

Table 4.4. Gel texture properties of starches and their blends Sample

HD

ADH

COH

GUM

CHE

MB PT-20 PT-40 PT-60 PT-80 PT

151±9a 93±5b 79±3c 63±0d 44±1e 28±3f

-4±4a -10±7a -13±3a -17±12a -17±5a -14±4a

0.427±0.1a 0.439±0.1a 0.459±0.0a 0.478±0.0a 0.512±0.0a 0.509±0.0a

64±11d 41±5c 36±4c 30±2bc 23±2ab 14±1a

100±36c 57±10b 52±3b 42±5ab 23±2ab 14±1a

MB SP-20 SP-40 SP-60 SP-80 SP

151±9d 112±5c 72±6b 55±2a 36±1a 28±3a

-4±4a -29±41ab -52±14a -39±14ab -33±14ab -19±5ab

0.427±0.1a 0.466±0.0a 0.440±0.1a 0.443±0.0a 0.476±0.0a 0.465±0.0a

64±11d 52±3c 31±5b 24±2a 17±1a 13±1a

100±36b 69±10b 32±3a 23±2a 16±1a 13±1a

MB R-20 R-40 R-60 R-80 R

151±9d 81±4c 31±2b 12±5a 4±0a 4±1a

-4±4a -58±39a -38±18ab -31±4ab -21±7ab -19±8ab

0.427±0.1a 0.432±0.0a 0.533±0.0b 0.562±0.0b 0.576±0.0b 0.532±0.0b

64±11d 35±1c 17±1b 7±3ab 2±0a 2±1a

100±36a 33±2b 16±1b 6±3b 2±0b 2±1b

MB S-20 S-40 S-60 S-80 S

151±9d 92±15c 60±3b 46±4ab 34±2a 32±5a

-4±4b -37±15a -38±7a -46±9a -39±11a -33±7a

0.427±0.1a 0.443±0.0a 0.450±0.0a 0.463±0.1a 0.434±0.0a 0.448±0.0a

64±11d 40±6c 27±1b 21±1ab 15±2a 14±2a

100±36b 40±6a 26±2a 20±2a 14±2a 13±2a

PT-20 to PT-80 refers to potato starch proportion (%) in the blends, same for other blends (SP/MB, R/MB and S/MB). HD = hardness; ADH = adhesiveness; COH = cohesiveness; GUM = gumminess; CHE = chewiness All data are expressed as mean ± standard deviation of mean; same superscripts indicate data that do not differ at significance level p < 0.05.

116

Table 4.5. Dynamic rheological properties of starches and their blends Sample

TG'max (°C)

G'max (Pa)

tan δG'max

G'90°C (Pa)

tan δ90 °C

G'25°C (Pa)

tan δ25°C

△G' (Pa)

G'25Hz (Pa)

tan δ25Hz

MB PT-20 PT-40 PT-60 PT-80 PT

72.9b 72.7ab 72.4ab 72.4ab 72.4a 72.0a

12200e 8695d 7280c 6153b 5227ab 4303a

0.211a 0.194a 0.186a 0.182a 0.183a 0.198a

3743a 3275b 3227b 3170b 3113b 3000b

0.117a 0.134ab 0.144bc 0.155bc 0.161c 0.161c

13067a 11300b 10333c 9180d 8143e 7243f

0.034a 0.035ab 0.035ab 0.038b 0.042c 0.048d

9323a 8025b 7107c 6010d 5030e 4243f

14400d 12050c 11700c 10433b 9577ab 8880a

0.063a 0.067ab 0.067ab 0.075bc 0.081c 0.094d

MB SP-20 SP-40 SP-60 SP-80 SP

72.9a 73.4a 74.4b 75.1bc 75.2cd 74.7d

12200e 7683d 5233c 3940b 3447ab 2755a

0.211ab 0.191b 0.211ab 0.227b 0.231b 0.226b

3743e 2860d 2310c 1953b 1840ab 1570a

0.117a 0.132b 0.148c 0.162d 0.167de 0.176e

13067a 10107b 8130c 6267d 5307e 4023f

0.034a 0.034a 0.035a 0.039b 0.044c 0.049d

9323a 7247b 5820c 4313d 3467e 2453f

14400a 10933b 9100c 7137b 6183e 4738f

0.063a 0.067ab 0.074b 0.086c 0.098d 0.109e

MB R-20 R-40 R-60 R-80 R

72.9a 72.7a 72.9a 75.6b 76.4c 77.4d

12200d 8093c 6880b 5747a 6153ab 6510ab

0.211c 0.123b 0.103ab 0.105ab 0.094a 0.083a

3743b 2943a 2840a 2743a 2910a 3587b

0.117a 0.125a 0.124a 0.128a 0.129a 0.084b

13067a 7573b 5333c 4273d 3743d 3477d

0.034a 0.041b 0.047c 0.048c 0.049c 0.056d

9323e 4630d 2493c 1530b 833b -110a

14400d 8283c 5727b 4717ab 4190a 3980a

0.063a 0.091b 0.116c 0.121cd 0.129de 0.137e

MB S-20 S-40 S-60 S-80 S

72.9a 73.2ab 73.7bc 73.9c 74.3cd 74.6d

12200c 8823b 7287a 7680ab 8203ab 8297ab

0.211c 0.139b 0.130ab 0.126ab 0.112a 0.113a

3743a 2770b 2360b 2423b 2618b 2597b

0.117c 0.127c 0.120c 0.111bc 0.095ab 0.080a

13067d 10967c 10030bc 9300ab 9170ab 8240a

0.034c 0.034c 0.032bc 0.032bc 0.031ab 0.028a

9323e 8197d 7670cd 6877bc 6553ab 5643a

14400d 11633c 10600bc 9820ab 9608ab 8780a

0.063a 0.061a 0.066a 0.066a 0.066a 0.066a

PT-20 to PT-80 refers to potato starch proportion (%) in the blends, same for other blends (SP/MB, R/MB and S/MB). TG'max: value of temperature when highest G' is reached during heating; G'max: highest G' value during heating; tan δG'max: value of tan δ at TG'max; G'90°C: value of G' at 90 °C; tan δG'90°C: value of tan δ at 90 °C; G'25°C: value of G' at 25 °C; tan δ25°C: value of tan δ at 25 °C; △G': difference between G'25°C and G'90°C; G'25Hz: value of G' at 25 Hz; Tan δ25Hz: value of tan δ at 25 Hz. All data are mean value; same superscripts indicate data that do not differ at significance level p < 0.05.

117

Table 4.6. Fitted coefficients of power law for 5 % starch pastes Sample

Upward

Downward K(Pa Sn)

n(-)

R2

0.9894 0.9983 0.9995 0.9995 0.9973 0.9952

28.90 29.19 27.12 22.74 22.98 22.69

0.39 0.40 0.42 0.44 0.46 0.48

0.9834 0.9915 0.9935 0.9957 0.9970 0.9974

0.19 0.30 0.32 0.33 0.34 0.37

0.9894 0.9942 0.9968 0.9970 0.9991 0.9993

28.90 24.66 19.46 15.47 12.31 8.09

0.39 0.41 0.44 0.46 0.47 0.49

0.9834 0.9931 0.9958 0.9963 0.9967 0.9969

129.63 99.16 61.82 40.39 32.24 28.09

0.19 0.22 0.25 0.28 0.28 0.24

0.9894 0.9606 0.9960 0.9823 0.9847 0.9900

28.90 23.08 13.22 7.56 5.70 6.33

0.39 0.41 0.46 0.51 0.52 0.45

0.9834 0.9890 0.9866 0.9857 0.9838 0.9880

129.63 63.10 64.45 75.15 66.95 69.68

0.19 0.29 0.27 0.21 0.20 0.19

0.9894 0.9949 0.9887 0.9805 0.9802 0.9729

28.90 31.59 25.39 14.59 11.30 10.40

0.39 0.38 0.39 0.43 0.44 0.44

0.9834 0.9903 0.9886 0.9790 0.9739 0.9721

K(Pa Sn)

n(-)

R

MB PT-20 PT-40 PT-60 PT-80 PT

129.63 77.87 86.46 82.69 62.05 61.58

0.19 0.27 0.26 0.27 0.33 0.35

MB SP-20 SP-40 SP-60 SP-80 SP

129.63 58.24 49.76 41.83 35.19 22.10

MB R-20 R-40 R-60 R-80 R MB S-20 S-40 S-60 S-80 S

2

PT-20 to PT-80 refers to potato starch proportion (%) in the blends, same for other blends (SP/MB, R/MB and S/MB). Upward: 0-1000 s-1; downward, 1000-0 s-1. K: consistency coefficient; n: flow behavior; R2: regression coefficient.

118

Figure 4.1. Particle size distribution range of mung bean (MB), sweet potato (SP), potato (PT), rice (R) and sorghum (S) starch.

119

Figure 4.2. Changes in storage modulus (G') of 20 % starch suspensions during heating and cooling (heating and cooling rate: 1 °C/min, strain: 2 %, frequency: 1 Hz). (A) mung bean (MB), potato (PT) starch and their blends. (B) mung bean (MB), sweet potato (SP) starch and their blends. (C) mung bean (MB), rice (R) starch and their blends. (D) mung bean (MB), sorghum (S) starch and their blends.

120

Figure 4.3. Changes in loss modulus (G'') of 20 % starch suspensions during heating and cooling (heating and cooling rate: 1 °C/min, strain: 2 %, frequency: 1 Hz). (A) mung bean (MB), potato (PT) starch and their blends. (B) mung bean (MB), sweet potato (SP) starch and their blends. (C) mung bean (MB), rice (R) starch and their blends. (D) mung bean (MB), sorghum (S) starch and their blends.

121

Figure 4.4. Changes in storage modulus (G') of 20 % starch suspensions during a frequency sweep (strain: 2 %, temperature: 25 °C). (A) mung bean (MB), potato (PT) starch and their blends. (B) mung bean (MB), sweet potato (SP) starch and their blends. (C) mung bean (MB), rice (R) starch and their blends. (D) mung bean (MB), sorghum (S) starch and their blends.

122

Figure 4.5. Changes in loss modulus (G'') of 20 % starch suspensions during a frequency sweep (strain: 2 %, temperature: 25 °C). (A) mung bean (MB), potato (PT) starch and their blends. (B) mung bean (MB), sweet potato (SP) starch and their blends. (C) mung bean (MB), rice (R) starch and their blends. (D) mung bean (MB), sorghum (S) starch and their blends.

123

Figure 4.6. Upward (0-1000 s-1) flow curves of starch pastes (5 % concentration). (A & C) upward and downward for pure starches and their blend pastes. (A) mung bean (MB), potato (PT) starch and their blends. (B) mung bean (MB), sweet potato (SP) starch and their blends. (C) mung bean (MB), rice (R) starch and their blends. (D) mung bean (MB), sorghum (S) starch and their blends.

124

Figure 4.7. Downward (1000-0 s-1) flow curves of starch pastes (5 % concentration). (A & C) upward and downward for pure starches and their blend pastes. (A) mung bean (MB), potato (PT) starch and their blends. (B) mung bean (MB), sweet potato (SP) starch and their blends. (C) mung bean (MB), rice (R) starch and their blends. (D) mung bean (MB), sorghum (S) starch and their blends.

125

References

Bao, J., Ao, Z. and Jane, J. L. (2005). Characterization of physical properties of flour and starch obtained from gamma-irradiated white rice. Starch - Stärke, 57, 480-487. Bao, J. S., Sun, M. and Corke, H. (2002). Analysis of the genetic behavior of some starch properties in indica rice (Oryza sativa L.): thermal properties, gel texture, swelling volume. Theoretical and Applied Genetics, 104, 408-413. Bao, J. S., Sun, M., Zhu, L. H. and Corke, H. (2004). Analysis of quantitative trait loci for some starch properties of rice (Oryza sativa L.): thermal properties, gel texture and swelling volume. Journal of Cereal Science, 39, 379-385. Beta, T. and Corke, H. (2001). Noodle quality as related to sorghum starch properties. Cereal Chemistry, 78, 417-420. Chrastil, J. (1987). Improved colorimetric determination of amylose in starches or flours. Carbohydrate Research, 159, 154-158. Chun, S. Y. and Yoo, B. (2004). Rheological behavior of cooked rice flour dispersions in steady and dynamic shear. Journal of Food Engineering, 65, 363-370. Collado, L. S. and Corke, H. (1999). Heat-moisture treatment effects on sweetpotato starches differing in amylose content. Food Chemistry, 65, 339-346. Collado, L. S., Mabesa, R. C. and Corke, H. (1999). Genetic variation in the physical properties of sweet potato starch. Journal of Agricultural and Food Chemistry, 47, 4195-4201. Crosbie, G. B. (1991). The relationship between starch swelling properties, paste viscosity and boiled noodle quality in wheat flours. Journal of Cereal Science, 13, 145-150. Dzuy Nguyen, Q., Jensen, C. T. B. and Kristensen, P. G. (1998). Experimental and modelling studies 126

of the flow properties of maize and waxy maize starch pastes. Chemical Engineering Journal, 70, 165-171. Ge, P., Fan, D., Ding, M., Wang, D. and Zhou, C. (2014). Characterization and nutritional quality evaluation of several starch noodles. Starch - Stärke, 66, 880-886. Gunaratne, A. and Corke, H. (2007). Gelatinizing, pasting, and gelling properties of potato and amaranth starch mixtures. Cereal chemistry, 84, 22-29. Gunasekaran, S. and Ak, M. M. (2000). Dynamic oscillatory shear testing of foods - selected applications. Trends in Food Science & Technology, 11, 115-127. Hagenimana, A. and Ding, X. (2005). A comparative study on pasting and hydration properties of native rice starches and their mixtures. Cereal Chemistry, 82, 70-76. Hagenimana, A., Pu, P. and Ding, X. (2005). Study on thermal and rheological properties of native rice starches and their corresponding mixtures. Food Research International, 38, 257-266. Hoover, R., Li, Y. X., Hynes, G. and Senanayake, N. (1997). Physicochemical characterization of mung bean starch. Food Hydrocolloids, 11, 401-408. Hsu, S., Lu, S. and Huang, C. (2000). Viscoelastic changes of rice starch suspensions during gelatinization. Journal of Food Science, 65, 215-220. Huang, J., Schols, H. A., van Soest, J. J. G., Jin, Z., Sulmann, E. and Voragen, A. G. J. (2007). Physicochemical properties and amylopectin chain profiles of cowpea, chickpea and yellow pea starches. Food Chemistry, 101, 1338-1345. Jacobs, H. and Delcour, J. A. (1998). Hydrothermal modifications of granular Starch, with retention of the granular structure:  a review. Journal of Agricultural and Food Chemistry, 46, 28952905.

127

Karam, L. B., Ferrero, C., Martino, M. N., Zaritzky, N. E. and Grossmann, M. V. E. (2006). Thermal, microstructural and textural characterisation of gelatinised corn, cassava and yam starch blends. International Journal of Food Science & Technology, 41, 805-812. Kasemsuwan, T., Bailey, T. and Jane, J. (1998). Preparation of clear noodles with mixtures of tapioca and high-amylose starches. Carbohydrate Polymers, 36, 301-312. Kaur, L., Singh, J., Singh, H. and McCarthy, O. J. (2008). Starch–cassia gum interactions: A microstructure – Rheology study. Food Chemistry, 111, 1-10. Kern, H. and Stute, R. (1994). Starch mixtures as pudding starches. U.S. Patent (5,324,532). Kong, X., Kasapis, S., Bertoft, E. and Corke, H. (2010). Rheological properties of starches from grain amaranth and their relationship to starch structure. Starch - Stärke, 62, 302-308. Li, J. Y. and Yeh, A. I. (2001). Relationships between thermal, rheological characteristics and swelling power for various starches. Journal of Food Engineering, 50, 141-148. Li, S., Ward, R. and Gao, Q. (2011a). Effect of heat-moisture treatment on the formation and physicochemical properties of resistant starch from mung bean (Phaseolus radiatus) starch. Food Hydrocolloids, 25, 1702-1709. Li, W., Shu, C., Zhang, P. and Shen, Q. (2011b). Properties of starch separated from ten mung bean varieties and seeds processing characteristics. Food and Bioprocess Technology, 4, 814821. Li, Z. G., Liu, W. J., Shen, Q., Zheng, W. and Tan, B. (2008). Properties and qualities of vermicelli made from sour liquid processing and centrifugation starch. Journal of Food Engineering, 86, 162-166. Lii, C. Y., Tsai, M. L. and Tseng, K. H. (1996). Effect of amylose content on the rheological property

128

of rice starch. Cereal Chemistry, 73, 415-420. Mahasukhonthachat, K., Sopade, P. A. and Gidley, M. J. (2010). Kinetics of starch digestion in sorghum as affected by particle size. Journal of Food Engineering, 96, 18-28. Mendez Montealvo, G., Garcia Suarez, F. J., Paredes Lopez, O. and Bello Perez, L. A. (2008). Effect of nixtamalization on morphological and rheological characteristics of maize starch. Journal of Cereal Science, 48, 420-425. Morris, V. (1990). Starch gelation and retrogradation. Trends in Food Science & Technology, 1, 26. Mua, J. and Jackson, D. (1997). Relationships between functional attributes and molecular structures of amylose and amylopectin fractions from corn starch. Journal of Agricultural and Food Chemistry, 45, 3848-3854. Mua, J. P. and Jackson, D. S. (1998). Retrogradation and gel textural attributes of corn starch amylose and amylopectin fractions. Journal of Cereal Science, 27, 157-166. Obanni, M. and Bemiller, J. N. (1997). Properties of some starch blends. Cereal Chemistry, 74, 431436. Ortega Ojeda, F. E. and Eliasson, A. C. (2001). Gelatinisation and retrogradation behaviour of some starch mixtures. Starch‐Stärke, 53, 520-529. Park, E. Y., Kim, H. N., Kim, J. Y. and Lim, S. T. (2009). Pasting properties of potato starch and waxy maize starch mixtures. Starch‐Stärke, 61, 352-357. Puncha-arnon, S., Pathipanawat, W., Puttanlek, C., Rungsardthong, V. and Uttapap, D. (2008). Effects of relative granule size and gelatinization temperature on paste and gel properties of starch blends. Food Research International, 41, 552-561.

129

Rao, M. (2003). Phase transitions, food texture and structure. Texture in Food, 1, 36-62. Rao, M. A. and Tattiyakul, J. (1999). Granule size and rheological behavior of heated tapioca starch dispersions. Carbohydrate Polymers, 38, 123-132. Sandhu, K. S., Kaur, M. and Mukesh. (2010). Studies on noodle quality of potato and rice starches and their blends in relation to their physicochemical, pasting and gel textural properties. LWT - Food Science and Technology, 43, 1289-1293. Sasaki, T. and Matsuki, J. (1998). Effect of wheat starch structure on swelling power. Cereal Chemistry, 75, 525-529. Singh, J., Kaur, L. and McCarthy, O. J. (2007). Factors influencing the physico-chemical, morphological, thermal and rheological properties of some chemically modified starches for food applications - A review. Food Hydrocolloids, 21, 1-22. Singh, N., Singh, J., Kaur, L., Sodhi, N. S. and Gill, B. S. (2003). Morphological, thermal and rheological properties of starches from different botanical sources. Food Chemistry, 81, 219-231. Sopade, P. A. and Kassum, A. L. (1992). Rheological characterization of akamu, a semi-liquid food made from maize, millet and sorghum. Journal of Cereal Science, 15, 193-202. Tan, H. Z., Li, Z. G. and Tan, B. (2009). Starch noodles: History, classification, materials, processing, structure, nutrition, quality evaluating and improving. Food Research International, 42, 551-576. Thebaudin, J. Y., Lefebvre, A. C. and Doublier, J. L. (1998). Rheology of starch pastes from starches of different origins: applications to starch-based sauces. LWT - Food Science and Technology, 31, 354-360.

130

Thiewes, H. J. and Steeneken, P. A. (1997). Comparison of the Brabender Viskograph and the Rapid Visco Analyser. 1. Statistical evaluation of the pasting profile. Starch‐Stärke, 49, 85-92. Waterschoot, J., Gomand, S. V., Fierens, E. and Delcour, J. A. (2015). Starch blends and their physicochemical properties. Starch‐Stärke, 67, 1-13. Wong, R. and Lelievre, J. (1982). Rheological characteristics of wheat starch pastes measured under steady shear conditions. Journal of Applied Polymer Science, 27, 1433-1440. Yao, Y., Zhang, J. and Ding, X. (2003). Retrogradation of starch mixtures containing rice starch. Journal of Food Science, 68, 260-265. Zaidul, I. S. M., Yamauchi, H., Kim, S. J., Hashimoto, N. and Noda, T. (2007). RVA study of mixtures of wheat flour and potato starches with different phosphorus contents. Food Chemistry, 102, 1105-1111. Zhang, Y., Gu, Z., Hong, Y., Li, Z. and Cheng, L. (2011). Pasting and rheologic properties of potato starch and maize starch mixtures. Starch‐Stärke, 63, 11-16. Zhu, F. and Corke, H. (2011). Gelatinization, pasting, and gelling properties of sweetpotato and wheat starch blends. Cereal Chemistry, 88, 302-309.

131

Chapter 5. Effect of starch blending on quality of mung bean starch noodles

Abstract

Potato (PT), sweet potato (SP), rice (R) and sorghum (S) starches were added to mung bean (MB) starch to produce starch noodles and their impact on noodle qualities of cooking yield, cooking loss, color, and in vitro digestibility were determined. Only PT starch addition brought noticeable changes in cooking yield, while R addition led to more cooking loss. Though starch noodles are generally high glycemic index foods, starch hydrolysis rate of pure MB starch noodles were the lowest and most starch noodles produced by starch blends were intermediate between pure MB and the corresponding added starch for this trait. Correlation analysis showed that noodle qualities could be predicted from starch qualities. SP starch may be the best among the four for addition to MB starch for noodle manufacture.

132

5.1. Introduction

Starch-based foods such as rice, steamed bread and noodles are daily staple foods in Oriental cultures. Asian noodles are not made exclusively of wheat, many are made from rice, buckwheat, or starches derived from mung bean and potato (Fu 2008). Starch noodles are usually prepared from isolated or purified starches from various plant sources such as mung bean, potato, or sweet potato. Starch noodles have a long history and have been a Chinese favorite food for more than 1400 years (Tan et al. 2009). Unlike wheat-based noodles, partial pre-gelatinized starch is used as the binder to form a consistent dough with the other remaining starch and water, due to absence of gluten. Thus starch itself plays an important role in both starch noodle processing and final starch noodle quality (Chen et al. 2002). Starch noodles showed less roughness and higher smoothness than wheat-based flour noodles according to a sensory test (Wu et al. 2015). Mung bean starch has high amylose content and behaves like chemically cross-linked starch that exhibiting restricted swelling and solubilization and these properties contribute to a starch gel which is both resilient and transparent (Fu 2008). Thus, traditionally mung bean starch is considered to be the most suitable raw material for starch noodle making as it gives a product with good nutritional quality, desired appearance and texture. mung bean starch noodle is, consequently, regarded as the best of all kinds of starch noodles (Ge et al. 2014; Li et al. 2008; Tan et al. 2009). This favored quality profile is thought to result from the high amylose content and restrict granule swelling (Kim et al. 1996). Besides a short cooking requirement, mung bean noodles have a low loss of solids on prolonged cooking and a unique translucent, chewy and elastic texture (Fu 2008; Kasemsuwan et al. 1998). These textural characteristics are also due to the distinguished properties of the starch, which has great hot-paste stability (Fu 2008). However, clear noodles made from mung bean starch are expensive compared 133

with other types of noodles as a result of mung bean starch’s low production supply and tedious processing methods (Kasemsuwan et al. 1998). Studies on noodles based on sweet potato starch are of interest to many developing and developed countries because it plays a vital role in food production, such as in substitution for expensive mung bean starch (Tan et al. 2009; Zhu & Wang 2014).

Blending different starches together has been proved as an efficient way to change starch quality such as reduce retrogradation (Obanni & Bemiller 1997). Researchers have put a lot of effort on properties of starch blends, like canna and mung bean (Puncha-arnon et al. 2008), or sweet potato and wheat (Zhu & Corke 2011). It is possible to improve noodle qualities through starch blends. Noodle products made from a mixture of tapioca and high-amylose starch had been preferred in sensory testing over native mung bean noodles (Kasemsuwan et al. 1998). Blending pigeon pea starch with rice starch could significantly improve starch noodle qualities on transparency, slipperiness and reduce cooking loss (Yadav et al. 2011). However, still limited research has studied applications of starch blends in noodle production. Besides texture and color, digestibility should also be studied as it relates more to human health. Previously we studied the effect of four common starches (potato, sweet potato, rice and sorghum) addition to mung bean starch on starch qualities. Thus the objective of this study is to further elucidate their impact on the starch noodle quality parameters of texture, color, and digestibility. This may help in starch noodle improvement.

134

5.2. Materials and methods 5.2.1. Materials

Mung bean (MB), potato (PT), sweet potato (SP), rice (R) and sorghum (S) starch were prepared as described in Chapter 4, coded MB, PT, SP, R and S, respectively. Corresponding amylose content were 35.0 % (MB), 19.8 % (PT), 35.0 % (SP), 30.5 % (R) and 36.6 % (S).

5.2.2. Starch noodle preparation

Starch noodles were prepared as previously described (Collado et al. 2001) with modifications. Starch dough was prepared by partial gelatinization of 6 % (w/w) of the total starch to be used which serves as binder. Gelatinization was done by boiling starch and water (1:7 w/v) for 5 min. Gelatinized starch was then mixed with the remaining 94 % ungelatinized starch to form a dough at 40-55 % moisture. The dough was kneaded by hand for 15 min or until a uniform consistency was achieved. The starch dough was extruded using the fabricated extruder into boiling water for 2 to 3 min (Lii and Chang 1981), transferred to cold water, drained (3×), and dried at 35 °C in a convection drier, cooled to RT, and sealed in polyethylene bags until used for analysis and sensory evaluation. The final moisture of dried starch noodles ranged from 8 to 10 %.

5.2.3. Cooking properties

Cooking yield and cooking loss were determined mainly following Zhu et al. (2010). Dried raw noodles (3-4 g) were cooked in boiling water (200 mL) in a beaker with lid for 5 min until the white

135

core disappeared. They were cooled in cool water (20 °C) for 10 s before dried on a paper towel for 15 min until further texture and color testing. For cooking yield and loss analysis, cooked noodle strands on the paper towel were lightly blotted for about 1 min to remove the excess surface moisture on the noodles surface. Cooking yield (CY) was calculated as: CY = Mco/Md Where Mco = weight of cooked noodles (g); Md = weight of dried raw noodles (g). After cooking, the remaining water in the beaker was further boiled for evaporation until less than 50 mL was left. Then remaining liquid in the water was transferred to a petri dish by washing with 50 ml distilled water, and then further dried in an oven at 100 °C until constant weight (Mc, g). Cooking loss % (CL) was calculated as: CL % = (Mc-Mb)*100 / (Md-Mm) Where Mb = weight of beaker (g); Mm = weight of moisture in dried raw noodles (g). All analysis was conducted at least in triplicate.

5.2.4. Color analysis

The Hunter color parameters L, a, and b (CIE 1976) were measured by a colorimeter (Chroma Meter CR-300, Minolta, Osaka, Japan) on surface of dried noodles, and cooked noodles at around 20 min after cooking. At least three replications were performed and L, a, b value were recorded.

5.2.5. Texture analysis of noodles

Tensile and compression tests were used in this study according to the method described in Zhu et 136

al. (2010) with modifications by using a TA-XT2 Texture Analyzer (Stable Micro Systems, Godalming, England). Noodles were cooked in separate batches for 5 min, and were tested between 15 and 20 min after cooking to avoid the rapid textural changes of cooked noodles right after cooking. For tensile strength analysis, the instrument was equipped with spaghetti/noodle tensile grips (A/SPR). The noodle strand was wound two, three or four times around the parallel friction roller of the grip to anchor the samples and avoid slippage. The distance between the parallel rollers was 4 cm. The mode was measure force in tension. Pre-test and test speeds were 3.0 mm/s, post-test speed was 5.0 mm/s. Distance was 100 mm. The trigger type was auto with a trigger force of 5.0 g. The data acquisition rate was 200 pps. The maximum force required to break the strand was termed tensile force (TN, g). At least 10 individual strands were tested for each group. For compression analysis, the probe was cylinderical with a diameter of 35 mm (P/35). The mode was measuring force in compression with single cycle. The pre-test, test, and post-test speeds were 2.0 mm/s. The strain was 75 %. The trigger type was auto with a trigger force of 5.0 g. The data acquisition rate was 200 pps. Two noodle strands were put parallel on the platform and tested close together at a time. The maximum force (g) was recorded as firmness. At least six tests within each group were conducted.

5.2.6. Total starch

Total starch content of flours was determined by total starch assay kit (K-TSTA, Megazyme, Co. Wicklow, Ireland).

137

5.2.7. In vitro digestion

In vitro digestion mainly followed Goni et al. (1997) with modifications. Approximately 50 mg (db) noodle strands were cooked with 5 mL distilled water in boiled water bath for 5 min. Subsequently after cooking, samples were homogenized for 2 min using an Ultra-Turrax homogenizer T25 (Janke and Kunkel, Ika Labortechnik, München, Germany) and 10 mL of HCl–KCl buffer pH 1.5 were added. Then, 0.2 mL of a solution containing 40 mg of pepsin from porcine gastrine mucosa (P7125, Sigma) in 10 mL HCl–KCl buffer, pH 1.5, was added to each sample, followed by 60 min of incubation in a shaking water bath at 40 °C. The volume was raised to 25 mL with Tris–Maleate buffer (pH 6.9) and 5 ml of Tris–Maleate buffer containing 2.6 IU of α-amylase from porcine pancreas (ref. A-3176, Sigma) was then added to each sample. Immediately the sample was transferred to a shaking water bath at 37 °C to start starch hydrolysis. Aliquots (0.5 mL) were taken every 30 min from 0 to 180 min. Each aliquot was immediately put in a boiling water bath for 10 min to inactivate α-amylase, and kept in 4 °C condition until the end of incubation time. Then 1.5 mL of 0.4 M sodium–acetate buffer, pH = 4.75, and 30 μL of amyloglucosidase from Aspergillus niger (ref. 102 857, Roche) were added to aliquots. Digested starch was then hydrolyzed into glucose by incubating samples at 60 °C for 45 min in a shaking water bath. Finally, glucose concentration was measured using the D-glucose assay kit (K-GLUC, Megazyme, Co. Wicklow, Ireland). The rate of starch digestion was expressed as a percentage of the total starch hydrolyzed at different times (30, 60, 90, 120, 150 and 180 min).

To calculated digestion kinetics and expected glycemic index (eGI), a non-linear model established by Goni et al. (1997) was applied as follows: 138

𝐶 = 𝐶∞ (1 − 𝑒 −𝑘𝑡 ), where C is the percentage of starch hydrolyzed at time t, C∞ is the maximum concentration of hydrolyzed starch, k is the kinetic constant and t is chosen time (min). The hydrolysis index (HI) was calculated by dividing the area under the hydrolysis curve of each starch sample by the corresponding area obtained from the reference sample (white bread). The eGI was calculated using the equation described by Goni et al. (1997): eGI = 39.71 + 0.549×HI

5.2.8. Statistical analysis

SPSS (version 19, Endicott, NY) was used for analysis of variance (ANOVA) by Tukey's multiple range test at p < 0.05. OriginPro, version 8.5.0 (OriginLab, Northampton, MA) was used for parameter estimating and figure drawing.

5.3. Results and discussion 5.3.1. Textural, cooking and color properties.

Cooking and textural properties of all starch noodles are shown in Table 5.1. Among five pure starch noodle samples, cooking yield ranged from 1.45 (S) to 2.53 (PT) while cooking loss varied from 0.41 % (MB) to 1.41 % (R). Cooking yield indicates noodle water absorption capabilities. Only PT starch addition to MB had significantly increased cooking yield. High cooking loss is not preferable as it usually comes with high solubility of starch, resulting in turbid cooking water, weak cooking tolerance, and sticky mouthfeel (Bhattacharya et al. 1999), and low cooking loss represented a better

139

performance to maintain their integrity during cooking (Zhu et al. 2010). Beta & Corke (2001) reported a cooking loss ranged from 1.8 to 3.1 % in sorghum noodles. Collado & Corke (1997) list the cooking loss range (1.1 to 1.9 %) in SP starch noodles. Chen et al. (2003) demonstrated a very high cooking loss (8.8 %) in PT starch noodle than SP (2.1 and 5.2 %). All showed higher than ours, possibly due to the differences in origin of materials and processing methods. The high cooking yield in PT starch noodle was possibly due to its high swelling properties (Kim et al. 1996).Lii & Chang (1981) claimed that starch with high amylose content were related to better starch noodle production, while Chen et al. (2003) found smaller granules tended to have lower cooking loss. This may related to the relatively high cooking loss (1.02 %) of pure PT starch noodle as PT starch had a lowest amylose content (19.8 %) and largest granule size among the four. However, other starch all had a high amylose content (all over 30 %) but MB starch noodles showed predominate advantages over them, and the R had the smallest starch granule size but showed highest cooking loss (1.41 %) when made into noodles. Starch origin and chemical structure can also have an impact on starch noodle quality (Chen et al. 2003; Tan et al. 2006). Therefore, differences in starch structure may play an important role among the five pure starch noodles. Generally cooking properties of starch blends were in the middle of the two individual starches. As cooking loss of SP was the similar to pure MB, SP may be the best choice among the four for addition in terms of this trait.

Tensile and compression data varied a lot among the pure starch noodle samples. MB were significantly had higher tensile force (TN) (102 g) and firmness (FN) (9630 g) force than others pure starch noodles (33-44 g), and much higher than wheat noodles (TN, 17 g; FN, 4487 g) reported in Chapter 3. This is similar to some other reports. Elasticity (same as TN) ranged from 17.3 to

140

44.5 g in ten types of S noodles in Beta & Corke (2001). Hormdok & Noomhorm (2007) reported a low TN (20.6 g) in R starch noodles. Addition of other starches decreased TN and FN with increasing mixing ratio except SP-20 and S-20. Bhattacharya et al. (1999) reported a tensile range of 1.7 to 20.2 g for rice noodles. Cooked mung bean starch noodle had much higher FN than other sweet potato starch noodles by a cutting test (Chen et al. 2002).

All cooked noodle color parameters showed little variations. L, a, b ranged from 65.1 (P-80) to 75.2 (R-20), -1.9 (S-20/S-40) to 1.5 (SP-40), -0.2 (S) to 4.1 (SP) (Table 5.1). Those small differences shall be accredited to processing differences of starch isolation such as purities. MB had the highest total starch content and showed the greatest value of lightness, while the others were little lower. SP showed noticeably higher b value than others, indicating existence of some pigment or color components, which came from few impurities left after starch extraction and purification.

Bhattacharya et al. (1999) and Zhu et al. (2010) demonstrated that flour properties could be used for noodle quality predictions. Thus in this study noodle cooking and texture properties were compared with swelling, pasting and gel textural data of corresponding starch and starch blends with correlation analysis results (Table 5.3). Unlike wheat based noodles discussed previously in Chapter 3, where no correlation was observed, starch noodle textural properties correlated better with starch properties and better in textural properties rather than pasting. For example, cooking yield correlated positively with swelling volume (SV) (0.915, p ＜ 0.01), cooking loss positively with cohesiveness (COH) (0.742, p ＜ 0.01), while TN (0.896, p ＜ 0.01) and FN (0.808, p ＜ 0.01) both correlated significantly with gel hardness (HD). Therefore HD and SV might be the most

141

important attributes in pasting and textural properties as they correlated with more attributes in noodle qualities. The high gel hardness could predict low swelling index, and good quality of starch noodles (Chen et al. 2002). This is in agreement with Chen et al. (2002) who found starch gel properties was more suitable for predicting starch noodle qualities, e.g. cooking loss significantly related with COH of gel.

5.3.2. In vitro digestion

Starch hydrolysis rate of all noodle samples are shown in Fig 5.1. All pure starch noodles had a rapid increase of starch hydrolysis rate at 30 min, then rose slowly or kept almost horizontal. Significantly different from the other four, MB starch noodle showed the lowest starch hydrolysis at all time period, indicating a lower postprandial glucose response. The mean value of final starch hydrolysis rate for all starch noodles were: MB (72.7 %), PT (90.7 %), SP (92.9 %), S (96.8 %), R (96.7 %). Starch hydrolysis rate of all starch noodles made by starch mixtures were mostly additive of the corresponding two pure starch noodles.

To better compare hydrolysis results of all starch noodle, a non-linear model described by Goni et al. (1997) was used and corresponding hydrolysis parameters and estimated glycemic index (eGI) were then calculated (Table 5.2). C∞ ranged from 71.3 % (MB) to 97.3 % (S), showing MB starch noodle was much harder to digest while S was easier. All other starch exhibited higher C∞ than MB starch noodle and C∞ of blends were in the middle between individual starches. Kinetic constant (k) was used to compare the release speed of glucose from test materials, reflecting hydrolysis rate

142

(Chung et al. 2008). k showed some variations among the samples and ranged from 0.025 (P-20) to 0.097 (SP). Among pure starch noodles, MB (0.031) showed the lowest k while SP (0.097) the highest. Ge et al. (2014) reported a low k value for six starch noodles including mung bean (0.08), sweet potato (0.08) and potato (0.07) starch noodles. However, it was much lower in our study. HI value were calculated and used to for estimation of glycemic index based on WB value. For pure samples, only MB showed an eGI lower than 100, while the other four were P (102.5), S (109.8), SP (110.4), R (112.8), showing some difference from values reported by Ge et al. (2014), and this may be due to variations in materials and processing methods. eGI of starch noodles made by starch blends were mostly in the middle of values of pure starch noodles. Thus all addition of other starches lead to increase in eGI. Foods could be classified into low GI foods (≤ 55) such as parboiled rice and beans, medium GI foods (56-69) like brown rice, oatmeal, and high GI foods (≥ 70) such as white bread, short grain rice (Bharath Kumar & Prabhasankar 2014). In Chapter 3 we demonstrated that wheat based noodles were high GI foods. Clearly, both wheat and starch noodles may all lead to high postprandial glucose response and would generally not be suitable for diabetes patients. Frei et al. (2003) reported retrograded cooked rice grain had significantly lower eGI and higher amylose content meant higher retrogradation effects, but without recooking. Lii & Chang (1981) claimed that starch with high amylose content were related to better starch noodle production. This may related to the weak properties (high cooking loss) of pure PT starch noodle with a lowest amylose content (19.8 %) in PT starch among the four. However, other starch all had a high amylose content (all over 30 %) but MB starch noodles showed predominate advantages over them. Tan et al. (2006) pointed that besides amylose content, the chemical structure of amylose and amylopectin strongly affecting noodle digestibility, and highly branched amylopectin and low molecular weight of the

143

constituent fractions led to high digestibility of starch noodles. Compared with SP starch, the portion of long chains in amylopectin in MB starch was greater than SP starch, avoiding enzyme penetration by creating a dense pack of starch chains (Tan et al. 2006). Correlation analysis between eGI and starch properties (Table 5.3) showed that eGI correlated significantly with textural properties instead of amylose content. Thus from a macroscopic view, a starch noodle with a harder texture perhaps may refer to that dense packing of starch chains, difficult for enzyme hydrolysis, and lead to a lower digestion rate. Among the four starches, R might be the worst choice because it lead to highest increase of eGI at 20 % ratio.

5.3. Conclusion

Mung bean was the most favored material for starch noodle production, not only in terms of noodle quality, but also for lowest digestibility. Gel texture properties may be better for prediction of starch noodle cooking and textural qualities than pasting properties. Higher gel hardness or TN and FN may be associated with a lower in vitro digestion or lower estimated glycemic index. Other starches all lead to worse qualities of mung bean starch noodle, and SP may be the best starch to mung bean starch noodles, as it caused the least change in qualities.

144

Table 5.1. Cooking, texture and color of starch noodles. PT-20 to PT-80 refers to potato starch proportion (%) in the blends, same for other blends (SP/MB, R/MB and S/MB). Sample

Cooking property

Noodle texture

Color

Cooking yield

Cooking loss (%)

Tensile force (g)

Firmness (g)

L

a

b

MB PT-20 PT-40 PT-60 PT-80 PT

1.54a

0.41a

102e

9630a

72.4a

-0.6b

1.70b 1.92c 2.06cd 2.12d 2.53e

0.84bc 0.68ab 1.08c 0.80bc 1.02c

96de 85cd 69b 71bc 44a

9447a 8889a 8453a 6751b 5671b

71.2a 68.8b 67.4b 65.1a 68.3b

-0.2c -0.2c -0.2c 0.0c -0.9a

0.7ab -0.1a 0.1a -0.1a 0.5ab 1.2b

MB SP-20 SP-40 SP-60 SP-80 SP

1.54a 1.64a 1.65a 1.57a 1.57a 1.81a

0.41a 0.62a 0.55a 0.51a 0.47a 0.55a

102c 119d 97c 68b 50a 37a

9630d 10990c 8318b 7793b 7205b 4296a

72.4d 68.9c 68.2bc 68.8c 66.6a 67.0ab

-0.6a 1.4d 1.5d 0.1b 1.2c 0.1b

0.7a 0.8a 1.3b 2.9c 3.1c 4.1d

MB R-20 R-40 R-60 R-80 R

1.54a 1.56a 1.51a 1.46a 1.44a 1.48a

0.41a 0.74ab 1.01ab 0.99ab 1.98c 1.41bc

102d 71c 51b 48b 43ab 33a

9630e 8491d 7596c 6486b 5928b 4754a

72.4c 75.2d 70.5e 68.8b 67.7a 68.1ab

-0.6b -1.7a -1.7a -0.5c -0.4d 0.1e

0.7bc 1.3d 0.8c 0.3a 0.5abc 0.4ab

MB S-20 S-40 S-60 S-80 S

1.54a 1.54a 1.49a 1.39a 1.44a 1.45a

0.41a 0.67ab 0.74ab 0.82ab 0.90b 0.91b

102d 86c 85c 64b 49a 40a

9630cd 9939d 8689bc 9820cd 7752ab 6888a

72.4c 70.3ab 70.1a 70.6ab 70.7ab 71.3b

-0.6c -1.9a -1.9a -1.6b -1.7b -0.1d

0.7b 1.4c 1.4c 0.6b 0.4b -0.2a

All data are mean value; values with the same superscript are not significantly different (p ＜ 0.05).

145

Table 5.2. In vitro digestion and estimated glycemic index Sample

Total starch (%)

C∞

MB PT-20 PT-40 PT-60 PT-80 PT

99.9 99.3 98.6 97.9 97.3 96.6

71.3±0.4a 81.3±0.6b 89.6±0.2c 89.5±0.2c 83.5±0.7d 87.2±0.2e

MB SP-20 SP-40 SP-60 SP-80 SP

99.9 99.6 99.3 99.0 98.7 98.4

MB R-20 R-40 R-60 R-80 R MB S-20 S-40 S-60 S-80 S

k

Calculated HI

eGI

0.031ab 0.025a 0.040bc 0.046bc 0.050c 0.048c

87.1±0.7a 94.1±0.3b 114.8±1.0de 116.9±0.4e 110.2±0.5c 114.3±0.3d

87.5±0.4a 91.4±0.1b 102.7±0.5de 103.9±0.2e 100.2±0.3c 102.5±0.2d

71.3±0.4a 78.2±0.3b 88.0±0.3c 86.5±0.3c 89.9±0.4d 92.0±0.8e

0.031a 0.033a 0.042ab 0.050b 0.080c 0.097d

87.1±0.7a 96.5±0.1b 113.5±0.6c 114.1±1.0c 124.3±1.2d 128.8±1.1e

87.5±0.4a 92.7±0.0b 102.0±0.3c 102.3±0.5c 107.9±0.6d 110.4±0.6e

99.9 99.2 98.5 97.8 97.1 96.4

71.3±0.4a 90.2±0.5b 93.8±0.3c 96.5±0.1d 94.0±0.0c 96.9±0.0d

0.031a 0.036ab 0.047ab 0.049b 0.063c 0.075c

87.1±0.7a 113.5±0.3b 122.9±0.0c 127.0±0.8d 127.1±0.0d 133.2±0.5e

87.5±0.4a 102.0±0.2b 107.2±0.0c 109.4±0.4d 109.5±0.0d 112.8±0.3e

99.9 99.3 98.7 98.1 97.4 96.8

71.3±0.4a 80.0±0.5b 84.1±0.5c 85.9±0.2d 95.6±0.3e 97.3±0.3f

0.031a 0.033a 0.047bc 0.036ab 0.041abc 0.048c

87.1±0.7a 98.5±0.9b 110.1±0.8c 107.9±0.7c 122.8±1.2d 127.7±0.6e

87.5±0.4a 93.8±0.5b 100.2±0.4c 99.0±0.4c 107.1±0.7d 109.8±0.3e

PT-20 to PT-80 refers to potato starch proportion (%) in the blends, same for other blends (SP/MB, R/MB and S/MB). HI: hydrolysis index; eGI: estimated glycemic index. Data of C∞, k, Calculated HI and eGI are expressed as mean ± standard deviation of mean; same superscripts indicate data that do not differ at significance level p < 0.05.

146

Table 5.3. Correlation analysis between cooking properties, eGI and pasting, gel texture properties. Pasting properties BD

CPV

Gel texture properties

PV

HPV

SB

SV

HD

ADH

COH

GUM

CY

0.921**

0.949**

0.020

0.915**

0.318

0.910**

0.018

0.483*

0.085

0.058

0.085

CL

-0.193

-0.110

-0.231

-0.201

-0.363

0.002

-0.610**

0.176

.0742**

-0.611**

-0.511*

TN

0.097

0.170

-0.181

0.324

0.611**

-0.122

0.896**

0.010

-0.537*

0.908**

0.852**

FN

-0.048

-0.044

-0.011

0.112

0.520*

-0.227

0.808**

-0.204

-0.574**

0.819**

0.738**

eGI

-0.218

-0.225

-0.005

-0.379

-0.645**

0.026

-0.901**

-0.168

0.537*

-0.906**

-0.864**

CY = cooking yield; CL = cooking loss; TN = tension; FN = firmness; eGI = estimated glycemic index. *significant at p < 0.05，**significant at p < 0.01.

147

CHE

Figure 5.1 In vitro starch hydrolysis rate of noodles made by pure starch and starch blends (A) potato - mung bean (PT - MB) blends. (B) sweet potato - mung bean (SP - MB) blends. (C) rice mung bean (R - MB) blends. (D) sorghum - mung bean (S - MB) blends.

148

References

Beta, T. and Corke, H. (2001). Noodle quality as related to sorghum starch properties. Cereal Chemistry, 78, 417-420. Bharath Kumar, S. and Prabhasankar, P. (2014). Low glycemic index ingredients and modified starches in wheat based food processing: A review. Trends in Food Science & Technology, 35, 32-41. Bhattacharya, M., Zee, S. Y. and Corke, H. (1999). Physicochemical properties related to quality of rice noodles. Cereal Chemistry, 76, 861-867. Chen, Z., Sagis, L., Legger, A., Linssen, J. P. H., Schols, H. A. and Voragen, A. G. J. (2002). Evaluation of starch noodles made from three typical Chinese sweet-potato starches. Journal of Food Science, 67, 3342-3347. Chen, Z., Schols, H. A. and Voragen, A. G. J. (2003). Starch granule size strongly determines starch noodle processing and noodle quality. Journal of Food Science, 68, 1584-1589. Chung, H. J., Liu, Q., Donner, E., Hoover, R., Warkentin, T. D. and Vandenberg, B. (2008). Composition, molecular structure, properties, and in vitro digestibility of starches from newly released Canadian pulse cultivars. Cereal Chemistry, 85, 471-479. Collado, L. S. and Corke, H. (1997). Properties of starch noodles as affected by sweetpotato genotype. Cereal Chemistry, 74, 182-187. Collado, L. S., Mabesa, L. B., Oates, C. G. and Corke, H. (2001). Bihon-type noodles from heatmoisture-treated sweet potato starch. Journal of Food Science, 66, 604-609. Frei, M., Siddhuraju, P. and Becker, K. (2003). Studies on the in vitro starch digestibility and the

149

glycemic index of six different indigenous rice cultivars from the Philippines. Food Chemistry, 83, 395-402. Fu, B. X. (2008). Asian noodles: history, classification, raw materials, and processing. Food Research International, 41, 888-902. Ge, P., Fan, D., Ding, M., Wang, D. and Zhou, C. (2014). Characterization and nutritional quality evaluation of several starch noodles. Starch - Stärke, 66, 880-886. Goni, I., Garcia Alonso, A. and Saura Calixto, F. (1997). A starch hydrolysis procedure to estimate glycemic index. Nutrition Research, 17, 427-437. Kasemsuwan, T., Bailey, T. and Jane, J. (1998). Preparation of clear noodles with mixtures of tapioca and high-amylose starches. Carbohydrate Polymers, 36, 301-312. Kim, Y. S., Wiesenborn, D. P., Lorenzen, J. H. and Berglund, P. (1996). Suitability of edible bean and potato starches for starch noodles. Cereal Chemistry, 73, 302-308. Li, Z. G., Liu, W. J., Shen, Q., Zheng, W. and Tan, B. (2008). Properties and qualities of vermicelli made from sour liquid processing and centrifugation starch. Journal of Food Engineering, 86, 162-166. Lii, C. Y. and Chang, S. M. (1981). Characterization of red bean (Phaseolus radiatus var. Aurea) starch and its noodle quality. Journal of Food Science, 46, 78-81. Obanni, M. and Bemiller, J. N. (1997). Properties of some starch blends. Cereal Chemistry, 74, 431436. Puncha-arnon, S., Pathipanawat, W., Puttanlek, C., Rungsardthong, V. and Uttapap, D. (2008). Effects of relative granule size and gelatinization temperature on paste and gel properties of starch blends. Food Research International, 41, 552-561.

150

Tan, H. Z., Gu, W. Y., Zhou, J. P., Wu, W. G. and Xie, Y. L. (2006). Comparative study on the starch noodle structure of sweet potato and mung bean. Journal of Food Science, 71, C447-C455. Tan, H. Z., Li, Z. G. and Tan, B. (2009). Starch noodles: history, classification, materials, processing, structure, nutrition, quality evaluating and improving. Food Research International, 42, 551-576. Wu, K., Gunaratne, A., Collado, L. S., Corke, H. and Lucas, P. W. (2015). Adhesion, cohesion, and friction estimated from combining cutting and peeling test results for thin noodle sheets. Journal of Food Science, 80, E370-E376. Yadav, B. S., Yadav, R. B. and Kumar, M. (2011). Suitability of pigeon pea and rice starches and their blends for noodle making. LWT - Food Science and Technology, 44, 1415-1421 Zhu, F., Cai, Y. Z. and Corke, H. (2010). Evaluation of Asian salted noodles in the presence of Amaranthus betacyanin pigments. Food Chemistry, 118, 663-669. Zhu, F. and Corke, H. (2011). Gelatinization, pasting, and gelling properties of sweetpotato and wheat starch blends. Cereal Chemistry, 88, 302-309. Zhu, F. and Wang, S. N. (2014). Physicochemical properties, molecular structure, and uses of sweetpotato starch. Trends in Food Science & Technology, 36, 68-78.

151

Chapter 6. Adhesion, cohesion and friction estimated from combining cutting and peeling test results for thin noodle sheets*

Abstract

The aim of this study was to estimate the adhesive and cohesive fracture energies, and frictional characteristics of seven types of cooked starch and flour sheets and combine these into a model framework for textural analysis. Cutting tests with wires of diameter 0.30-0.89 mm were performed with and without lubrication. Plots of the work done, normalized to the area cut by the wire, showed that this to be linearly related to wire diameter irrespective of lubrication. The oil had little impact on the intercept of these plots, giving cohesive fracture energy (Gc) ranges for these foods between 6.8 - 32.5 J m-2. However, lubrication had a strong influence on the slope of the plots. From a comparison of the slopes for lubricated vs. unlubricated tests, the kinetic coefficient of friction 𝜇𝑘 could be calculated. Values for 𝜇𝑘 between 0.007 - 0.521 for different foods were obtained. Peeling tests were performed by lifting sheets vertically away from a fresh mica surface. The adhesive fracture energy Ga, varied from 2.5 – 4.8 J m-2. The results can be modeled by plotting the ratio of cohesive to adhesive fracture energy against the coefficient of friction. Thresholds in both axes suggest a physical basis for distinguishing textural perceptions. However, sensory testing with 12 subjects using the seven food types could not establish whether this framework, however wellestablished physically, would apply to oral sensations. A much larger test would be required. *This chapter has been published in Journal of Food Science, 80 (2), E370-E376. 152

6.1. Introduction

Foods are usually multi-component, multi-phase systems containing complex mixtures of water, polysaccharides, proteins, lipids and numerous minor constituents (Copeland et al. 2009). Starch, an important macro-constituent in many foods, represents a major energy resource, providing 5070 % of the energy in the average human diet (Copeland et al. 2009). Noodles and pasta represent a diverse group of starch-containing products, where the consumer has a very exacting expectation for texture, often connected to longstanding cultural traditions.

The ease of fracture of these foods is one important mechanical property that determines eating characteristics, and this takes place during mastication when the structure of the food is broken down and flavor and aroma are released (Gamonpilas et al. 2010; Goh et al. 2005; Kamyab et al. 1998). Fracture behavior is also an important characteristic to assess during food production, an understanding of this being a necessary first step towards optimizing food processes and designing new consumer products (Gamonpilas et al. 2010; Goh et al. 2005; Luyten & Van Vliet 1995).

The cutting of food materials with wires is common practice in food processing. From such tests, it is claimed that firmness differences can be evaluated, e.g. between sweet potato and mung bean noodle (Chen et al. 2002). Additionally, the test has been developed as a promising method for measuring the material properties of soft foods. The procedure is to push a taut horizontal wire of known diameter into a regularly-shaped flat-surfaced block of food material of known breadth (Gamonpilas et al. 2010; Kamyab et al. 1998). A complete wire cutting curve (Figure 6.1 A)

153

contains three stages. There is a pre-crack indentation phase up to a fracture point. This is followed by a rapid fracture phase. Finally, the fracture stabilizes, when each further movement of the wire produces an equal increment of growth in crack length (Czerner & Martucci 2013). An equation relating the steady cutting force to both wire diameter and friction was developed by Kamyab et al. (1998): 𝐹 𝑏

= 𝐺𝑐 + (1 + μ𝑘 )𝜎𝑦 𝑑

[1]

where F is the cutting force, d is wire diameter, b the sample width, 𝜎𝑦 is a characteristic stress, 𝜇𝑘 the kinetic coefficient of friction and Gc is the cohesive fracture energy (toughness). By performing tests with various wire diameters, F/b can plotted against d. Provided that the plot is linear, then the value of F/b extrapolated to the y-intercept, i.e. to a fictive zero-diameter wire d = 0, gives an estimate of Gc (Figure 6.1 B). However, this theory applies strictly only to the steadystate phase. Unfortunately, thin sheets, i.e., those 1; ‘rough’: Spearman’s R = +0.643; p > 0.1; ‘slippery’: Spearman’s R = -0.464; p > 0.1), although the direction (positive/negative) of these correlation 161

coefficients was as predicted. However, none of the sensory attributes was significantly correlated with physical measures of adhesion, cohesion, or their ratio, or with the coefficient of friction (p > 0.1). Yet all this is based on only seven foods: a test with a larger number of foods would be required to fully evaluate the statistical associations.

Our physical experimentation here was in vitro. Information on the behavior of food in the mouth has advanced considerably in the last 15 years with the advent of ‘in-mouth’ measurement (Davidson et al. 1998). Intraoral observations of interactions between food and oral surfaces have shown the importance of friction and, to a limited extent, adhesion for understanding the perception of stickiness, creaminess, roughness, smoothness, slipperiness, greasiness etc. (de Wijk et al. 2011; de Wijk & Prinz 2005; 2007; Dresselhuis et al. 2008; Dresselhuis et al. 2007; Prinz et al. 2006; Rossetti et al. 2009; van Aken et al. 2007). Yet much of this work is based just on measurement of friction. When adhesion and cohesion are brought into the picture, together as competing forces, then methods that have their roots in ‘push-pull-off’ texture profile analysis still tend to be employed (e.g. Dunnewind et al. (2004)). In other words, up to now, there has been no simple set of tests, inmouth or otherwise, that could estimate all properties involved and bring cohesion, adhesion and friction together.

Figure 6.4 indicates how we believe that a fuller picture may be achieved just using peeling and cutting tests. The logarithmic scales used in Figure 6.4 match the standard presumption for the sensory perception of physical parameters. On the vertical axis of the graph, the ratio of adhesive fracture energy (from peeling) to cohesive fracture energy (wire cutting) describes the competing 162

attractive forces that encourage the formation of surfaces, either within the food particle via internal fracture or else that favor adherence to the external oral surface. We term this ratio ‘adhesion/cohesion’ following Kendall (1975b). On the horizontal axis, the coefficient of friction provides information on ‘sticking’ versus ‘slipping’. We have placed thresholds on these axes based on the theoretical analysis of interfaces in materials Kendall (1975b). For a food that has an elastic modulus identical to that of oral mucosa, then for an adhesion/cohesion ratio < 0.106, food will slip cleanly against the oral surface, while above this ratio, particles will stick Kendall (1975b). An elastic mismatch (Dunders 1969) between food and mucosa can raise this threshold ratio to 0.2 or greater (Evans et al. 1999; Evans et al. 1990; Kendall 1975b) and the value of Poisson’s ratio also comes into play since this can be very high in many animal soft tissues (Lucas 2004). Although the elastic moduli of food gels and oral mucosa are probably not very different, we indicate the possibility of such a mismatch by placing a ‘band’ for the threshold instead of a line (Figure 6.4). There is no such objective threshold for friction on the horizontal axis of Figure 6.4. However, one could be envisaged by assuming that sensory detection involves a disturbance of the normal level of intraoral friction when food is absent. Human sensory systems often act like this to detect disturbances to the equilibrium state of the body very sensitively. [For example, the changes in facial skin color during emotional states and in body temperature when there is a fever are both very finely perceived (Changizi 2009)]. Both involve tiny changes in physical scales even though subjects are incapable of registering much larger changes in external environmental conditions accurately when asked to do so.] In vitro experiments of whole saliva against a substrate resembling oral mucosa give a ‘clean mouth’ frictional equilibrium as 𝜇𝑘 ≈ 0.02 (Bongaerts et al. 2007). Intraoral friction may be actually considerably higher than this simply due to the roughness of the tongue alone (de 163

Wijk & Prinz 2007; Dresselhuis et al. 2008), so again we broaden the threshold to a band to indicate this uncertainty. Application of these ‘threshold bands’ to Figure 6.4 then produces four distinct physical ‘domains’ of food behavior.

As far as textural analysis is concerned, the following model of food behavior is purely a hypothesis at this point. On Figure 6.4, we have tentatively labeled the four domains as ‘sticky’, ‘smooth, ‘slippery’ and ‘rough’. These predictions follow logic and are not driven by the results of our sensory analysis (which does not conform to it in any exact fashion). When adhesion is high (relative to cohesion) and friction is high too, then food is hypothesized to feel sticky. The converse, when adhesion and friction are both low, will result in slippery foods. A truly slippery surface texture is one with very low adhesion or friction. Such a food would be very difficult to control in the mouth. It would resemble the ‘oyster’ in the model of Hutchings & Lillford (1988) and could be swallowed immediately. If adhesion is relatively high but friction is low, then a food could feel smooth. In such circumstances, pieces may break off the food mass easily and so coat the oral surface - a factor implicated in creaminess (de Wijk & Prinz 2005; 2007) - but if so, these pieces will be easy to clear afterwards with the tongue because friction is low. Lastly, if adhesion is relatively low but friction is high, then the food is likely to feel rough: any particles that adhere to the mouth and then fracture will be difficult to clear, and so they will roughen the oral surface. Were such a hypothesis to be capable of describing intraoral sensations, then the intraoral evaluation of sticky vs. slippery or rough vs. smooth would integrate information from both axes of Figure 6.4, thus suggesting substantial sensory integration.

164

Quantitative values for the behavior of the noodle sheets placed on Figure 6.4 are taken directly from Table 6.2. In the sensory analysis, sweet potato was evaluated as sticky, as was wheat. This approximates to their positions in the domains on Figure 6.4. Also, semolina and millet were judged to be rough, again partially in accord with Figure 6.4. However, the positions of potato, buckwheat and mung bean in the figure are clearly at variance with the sensory data. There is no doubt that sensory analysis must be preferred to such physical modeling until the latter passes many such tests. Sensory analysis allows the overall characteristics of foods to be evaluated in an integrated manner (Li et al. 2012). Individual physical measurements cannot be related to this unless valid models can be constructed. While in-mouth measurement has improved our understanding of the mechanisms behind these sensory assessments, we believe that physical modeling, such as in Figure 6.4, is bound to be required to advance understanding further.

Our modeling here had two aims. One was an objective characterization of the physical behavior of foods suitable for fundamental food research and quality control. As far as sensory correlation is concerned, our experiments could be improved in respect of lubrication. Noodles need to have the correct humidity, yet be free of a surface coating of free water (Ross 2006). Attempts were made to achieve this here, but it was difficult to maintain. More important is the influence of saliva in the mouth. For foods containing cooked starch, this serves not just as a masticatory lubricant facilitating the sliding of tooth surfaces during chewing, but as rapid digester (Janssen et al. 2009). Some starches (e.g. potato) are digested by salivary amylase much more rapidly than others (Janssen et al. 2009). Saliva could be introduced as the lubricant in wire-cutting tests to compare to mineral oil. Ideally, the wire could be coated either with a ceramic to mimic tooth enamel (if friction were to be 165

judged between the teeth) or with a low-modulus polymer so as to emulate oral mucosa. Peeling tests could be performed on saliva-covered substrates (Bongaerts et al. 2007; Dresselhuis et al. 2007). The cohesive fracture energy would only be affected though if saliva were given time to penetrate the interior of the noodle.

6.4. Conclusion

The experiments describe a novel methodology for obtaining the surface properties of thin noodle sheets against standardized in vitro substrates (steel and mica) relating to cohesion, adhesion and friction. These methods could be applied to many kinds of semisolid foods and recommended for food manufacturing situations. A new model for understanding textural attributes that includes stickiness and smoothness is suggested. Further research including sensory analysis of a large number of foods is needed to test the hypothesis.

166

Table 6.1. Results for cutting where W/bt (in J m-2) was regressed against d (in mm). Sample

Treatment

Intercept

Slope

R2

Buckwheat

Unlubricated

16.56 ±1.92

26.3±3.3

0.825

Lubricated

15.41 ±1.79

25.9±2.9

0.856

Unlubricated

26.59 ±1.54

51.8±2.6

0.963

Lubricated

24.65 ±1.79*

43.0±2.9*

0.922

Unlubricated

18.93 ±3.32

66.1±5.2

0.909

Lubricated

12.39 ±1.89*

65.1±3.1

0.957

Unlubricated

27.56 ±1.64

27.8±2.8

0.808

Lubricated

27.16 ±1.58

26.4±2.7

0.842

Unlubricated

11.14 ±1.74

21.5±2.7

0.774

Lubricated

12.20 ±2.00*

16.5±3.2

0.630

Unlubricated

30.79 ±4.24

69.0±6.8

0.859

Lubricated

32.54 ±3.36

60.6±5.3

0.902

Unlubricated

6.14 ±1.64

24.5±2.5

0.859

Lubricated

6.80 ±0.78*

16.2±1.2*

0.886

Wheat

Mung bean

Millet

Potato

Semolina

Sweet potato

Intercept and slope are expressed as mean ±standard error of mean; *means lubricated samples are significantly different (p < 0.05) from unlubricated ones; regression models are all significant at p < 0.001.

167

Table 6.2. Properties obtained from cutting and peeling tests. Cohesive fracture

Adhesive fracture

Adhesive/cohesive

Kinetic Coefficient of

energy (J m-2)

energy (J m-2)

fracture energy

Friction

Mung bean

12.39 ±1.89

4.6 ±0.43ab

0.37

0.023

Potato

12.20 ±2.00

2.6 ±0.25c

0.21

0.247

Sweet potato

6.80 ±0.78

3.4 ±0.21bc

0.50

0.521

Buckwheat

15.41 ±1.79

3.2±0.17c

0.21

0.007

Millet

27.16 ±1.58

2.5 ±0.33c

0.09

0.068

Wheat

24.65 ±1.79

4.8 ±0.33a

0.19

0.206

Semolina

32.54 ±3.36

3.3 ±0.32c

0.10

0.125

Sample

Cohesive and adhesive fracture energy data are expressed as mean ±standard error of mean; values with the same superscript are not significantly different (p ＜ 0.05).

168

Table 6.3. Scores on the Likert scale for the sensory attributes of different noodle types. Sample

Stickiness

Slipperiness

Roughness

Smoothness

Potato

1.8±0.6a

4.6±1.5ad

2.0±1.0a

4.8±1.3a

Mung bean

1.6±0.5a

4.7±1.5ac

2.2±1.0a

4.8±1.5a

Sweet potato

4.0±1.0bc

3.4±1.4abcd

2.4±1.0a

4.0±1.2a

Millet

3.6±1.4bc

1.8±0.8b

4.8±0.8bc

1.7±0.5bc

Semolina

2.6±1.2c

2.0±1.0b

5.1±1.1bc

1.6±0.7bc

Buckwheat

4.9±1.2b

2.2±1.5bc

5.6±1.2c

1.4±0.7c

Wheat

4.1±1.4bc

2.5±0.8bd

3.6±1.5ab

2.8±1.3ab

Likert scale ranges from 1 point (‘extremely not’ sticky, slippery, rough or smooth as relevant) up to 7 point (‘extremely’) at the other endpoint. All data are mean ± standard deviation; same superscripts indicate values that do not differ from each other at significance level p ＜ 0.007 (i.e. 0.05/7).

169

Figure 6.1. The behavior of semisolid foods in wire cutting experiments. A thick block shows three phases of behavior as the wire sinks into it (A), culminating in a relatively steady cutting force. The force per unit width of the block in the steady phase, which is the same as the work done to cut unit area of material, is linearly related to wire diameter (B). The inset in B shows the crack path being tracked by a digital microscope equipped with a ring of LED’s for illumination. With thin noodle sheets (C), no steady state is seen, but there is also no precipitate fast fracture, the force falling to zero as the wire nears the base of the sheet. The work done per unit area of fractured noodle can be plotted against wire diameter, as in B. The effect of lubrication on these plots (D) allows the kinetic coefficient of friction to be estimated. 170

Figure 6.2. The geometry of peeling tests on noodle sheets (F, dragging force; θ, dragging angle; b, sample width).

171

Figure 6.3. Regressions of the work done in the cutting test, divided by cut area, against wire diameter for flour-based (A) and starch-based (B) noodles. All plots were effectively linear. Lubrication reduced the slopes as predicted, with less effect on the intercepts (except for mung bean).

172

Figure 6.4. Postulated model of noodle surface texture based on properties given in Table 6.2 (see text).

173

References

Adhikari, B., Howes, T., Bhandari, B. R. and Truong, V. (2001). Stickiness in foods: a review of mechanisms and test methods. International Journal of Food Properties, 4, 1-33. Atkins, A. G. and Mai, Y. (1985). Elastic and plastic fracture: metals, polymers, ceramics, composites, biological materials. Ellis Horwood: Chichester. Baik, B., Czuchajowska, Z. and Pomeranz, Y. (1994). Role and contribution of starch and protein contents and quality to texture profile analysis of oriental noodles. Cereal Chemistry, 71, 315-320. Beta, T. and Corke, H. (2001). Noodle quality as related to sorghum starch properties. Cereal Chemistry, 78, 417-420. Bongaerts, J., Rossetti, D. and Stokes, J. (2007). The lubricating properties of human whole saliva. Tribology Letters, 27, 277-287. Bouzakis, K. D., Tsouknidas, A. and Maliaris, G. (2010). A wire cutting test to check the temperature distribution in inductively heated cylindrical aluminium billets at semi-solid material state. Journal of Materials Processing Technology, 210, 166-173. Changizi, M. A. (2009). The vision revolution. Ben Bella: Dallas. Chen, Z., Sagis, L., Legger, A., Linssen, J. P. H., Schols, H. A. and Voragen, A. G. J. (2002). Evaluation of starch noodles made from three typical Chinese sweet-potato starches. Journal of Food Science, 67, 3342-3347. Collado, L. S., Mabesa, L. B., Oates, C. G. and Corke, H. (2001). Bihon-type noodles from heatmoisture-treated sweet potato starch. Journal of Food Science, 66, 604-609. Collado, L. S., Mabesa, R. C. and Corke, H. (1999). Genetic variation in the physical properties of 174

sweet potato starch. Journal of Agricultural and Food Chemistry, 47, 4195-4201. Copeland, L., Blazek, J., Salman, H. and Tang, M. C. (2009). Form and functionality of starch. Food Hydrocolloids, 23, 1527-1534. Czerner, M. and Martucci, J. (2013). Mechanical and fracture Behavior of gelatin gels M. In 13th International Conference on Fracture. Beijing, China. Curran Associates, Inc. Davidson, J. M., Linforth, R. S. T. and Taylor, A. J. (1998). In-mouth measurement of pH and conductivity during eating. Journal of Agricultural and Food Chemistry, 46, 5210-5214. de Wijk, R. A., Janssen, A. M. and Prinz, J. F. (2011). Oral movements and the perception of semisolid foods. Physiology & Behavior, 104, 423-428. de Wijk, R. A. and Prinz, J. F. (2005). The role of friction in perceived oral texture. Food Quality and Preference, 16, 121-129. de Wijk, R. A. and Prinz, J. F. (2007). Fatty versus creamy sensations for custard desserts, white sauces, and mayonnaises. Food Quality and Preference, 18, 641-650. Dresselhuis, D., De Hoog, E., Stuart, M. C. and Van Aken, G. (2008). Application of oral tissue in tribological measurements in an emulsion perception context. Food Hydrocolloids, 22, 323-335. Dresselhuis, D., Klok, H. J., Stuart, M. C., de Vries, R., van Aken, G. and de Hoog, E. A. (2007). Tribology of o/w emulsions under mouth-like conditions: determinants of friction. Food Biophysics, 2, 158-171. Dunders, J. (1969). Edge-bonded dissimilar orthogonal elastic wedges under normal and shear loading. Journal of Applied Mechanics, 36, 650-652. Dunnewind, B., Janssen, A. M., Van Vliet, T. and Weenen, H. (2004). Relative importance of 175

cohesion and adhesion for sensory stickiness of semisolid foods. Journal of Texture Studies, 35, 603-620. Evans, A., Hutchinson, J. and Wei, Y. (1999). Interface adhesion: effects of plasticity and segregation. Acta Materialia, 47, 4093-4113. Evans, A., Rühle, M., Dalgleish, B. and Charalambides, P. (1990). The fracture energy of bimaterial interfaces. Metallurgical Transactions A, 21, 2419-2429. Gamonpilas, C., Charalambides, M. N. and Williams, J. G. (2009). Determination of large deformation and fracture behaviour of starch gels from conventional and wire cutting experiments. Journal of Materials Science, 44, 4976-4986. Gamonpilas, C., Charalambides, M. N., Williams, J. G., Dooling, P. J. and Gibbon, S. R. (2010). On the study of wire cutting of starch gels. In 12th International Conference on Fracture. Ottawa, Canada. Curran Associates, Inc. Goh, S. M., Charalambides, M. N. and Williams, J. G. (2005). On the mechanics of wire cutting of cheese. Engineering Fracture Mechanics, 72, 931-946. Hutchings, J. B. and Lillford, P. (1988). The perception of food texture - the philosophy of the breakdown path. Journal of Texture Studies, 19, 103-115. Janssen, A. M., Van de Pijpekamp, A. M. and Labiausse, D. (2009). Differential saliva-induced breakdown of starch filled protein gels in relation to sensory perception. Food hydrocolloids, 23, 795-805. Kamyab, I., Chakrabarti, S. and Williams, J. G. (1998). Cutting cheese with wire. Journal of Materials Science, 33, 2763-2770. Kendall, K. (1975a). Thin-film peeling-the elastic term. Journal of Physics D: Applied Physics, 8, 176

1449. Kendall, K. (1975b). Transition between cohesive and interfacial failure in a laminate. Proceedings of the Royal Society of London Series A, 344, 287-302. Kendall, K. (2001). Molecular adhesion and its applications: the sticky universe. Kluwer Academic/Plenum Publishers: New York. Kilcast, D. and Roberts, C. (1998). Perception and measurement of stickiness in sugar-rich foods. Journal of Texture Studies, 29, 81-100. Kinloch, A. J., Lau, C. C. and Williams, J. G. (1994). The peeling of flexible laminates. International Journal of Fracture, 66, 45-70. Kinloch, A. J. and Williams, J. G. (2002). The mechanics of peeling tests. In: A. V. Pocius and D. A. Dillard, (Eds.). The Mechanics of Adhesion. Elsevier: Amsterdam. pp. 273-301. Li, P. H., Huang, C. C., Yang, M. Y. and Wang, C. C. R. (2012). Textural and sensory properties of salted noodles containing purple yam flour. Food Research International, 47, 223-228. Liu, W. J. and Shen, Q. (2007). Studies on the physicochemical properties of mung bean starch from sour liquid processing and centrifugation. Journal of Food Engineering, 79, 358-363. Lucas, P. W. (2004). Dental functional morphology: how teeth work. Cambridge University Press: Cambridge. pp. 112-115. Luyten, H. and Van Vliet, T. (1995). Fracture properties of starch gels and their rate dependency. Journal of Texture Studies, 26, 281-298. Michalski, M. C., Desobry, S. and Hardy, J. (1997). Food materials adhesion: A review. Critical Reviews in Food Science and Nutrition, 37, 591-619. Peleg, M. (1976). Texture profile analysis parameters obtained by an instron universal testing 177

machine. Journal of Food Science, 41, 721-722. Pesika, N. S., Tian, Y., Zhao, B., Rosenberg, K., Zeng, H., McGuiggan, P., Autumn, K. and Israelachvili, J. N. (2007). Peel-zone model of tape peeling based on the gecko adhesive system. The Journal of Adhesion, 83, 383-401. Pons, M. and Fiszman, S. M. (1996). Instrumental texture profile analysis with particular reference to gelled systems. Journal of Texture Studies, 27, 597-624. Prinz, J. F., Huntjens, L. and de Wijk, R. A. (2006). Instrumental and sensory quantification of oral coatings retained after swallowing semi-solid foods. Archives of Oral Biology, 51, 10711079. Rosinski, M. J., Barmore, C. R., Dick, R. L. and Acton, J. C. (1989). Film sealant and vacuum effects on two measures of adhesion at the sealant-meat interface in a cook-in packaging system for processed meat. Journal of Food Science, 54, 863-866. Ross, A. S. (2006). Instrumental measurement of physical properties of cooked Asian wheat flour noodles. Cereal Chemistry, 83, 42-51. Rossetti, D., Bongaerts, J. H. H., Wantling, E., Stokes, J. R. and Williamson, A. M. (2009). Astringency of tea catechins: more than an oral lubrication tactile percept. Food Hydrocolloids, 23, 1984-1992. Szczesniak, A. S., Brandt, M. A. and Friedman, H. H. (1963). Development of standard rating scales for mechanical parameters of texture and correlation between the objective and the sensory methods of texture evaluation. Journal of Food Science, 28, 397-403. van Aken, G. A., Vingerhoeds, M. H. and de Hoog, E. H. A. (2007). Food colloids under oral conditions. Current Opinion in Colloid & Interface Science, 12, 251-262. 178

Wei, Y. and Hutchinson, J. W. (1998). Interface strength, work of adhesion and plasticity in the peel test. In W. G. Knauss and R. A. Schapery (Eds.), Recent Advances in Fracture Mechanics. Springer Netherlands. pp 315-333.

179

Chapter 7. Indentation of thin noodle sheets

Abstract

The deformation characteristics of thin noodle sheets made from either starch or flour were examined with spherical indenters. A load relaxation method was employed that partitioned elastic and viscous behaviors. Indentations were restricted to below 10 % of the thickness of a sheet, with all measurements made after loading specimens to ~6-8 % nominal strain. The effect of three indenters of different tip radii were examined. Estimates of instantaneous and infinite moduli for 3.175 mm and 1.65 mm tips were similar, but the latter was less variable. Estimates were most variable for the smallest (0.72 mm) indenter, with values often higher than for the larger probes. Flour-based noodle sheets were less elastic than starch-based ones with indications that the degree of elasticity was inversely related to final viscosity, as measured during gel formation in a Rapid Visco Analyzer (RVA). To examine if differing results with probe sizes reflected characteristic microstructural dimensions, and thus the relative chance of contacting different structures, nonindented and post-indented specimens of wet sweet potato noodle sheets were imaged by AFM. Preliminary indications were that contacts with 0.72 mm probes might be more likely to involve swollen (disrupted) starch granules rather than the surrounding matrix. However, most of the deformation probably takes place within the matrix. Thus, the size effect might be explained by characteristic granule size.

180

7.1. Introduction

Nowadays, abundant noodle and pasta products are consumed as staple foods in most parts of the world. However, ‘noodle’ is a generic term that masks substantial variation in many physicochemical attributes, such as base ingredients, preparation methods, size, color and texture. There are two major types of noodle: wheat flour-based noodles (e.g. white salted or alkaline noodles) that contain gluten as a binder, versus starch/gluten-free noodles (e.g. mung bean or sweet potato noodles), which use gelatinized starch to bind the dough. Cultural preferences play a large role in patterns of consumption in different geographical locations. For example, alkaline noodles made with hexaploid wheat flour are consumed in large amounts in southeast Asia, southern China and Japan (Ross et al. 1997), but are absent from local cuisines elsewhere. Starch noodles, also named glass or clear noodles, are popular in most parts of Asia (Chen et al. 2002; Collado & Corke 1997; Kim et al. 1996; Lee et al. 2005; Tan et al. 2009; Yuan et al. 2008), while durum flour-based noodles (in the form of pasta) predominate in Europe. Egg-pasta is popular in Italy and appreciation is growing in Europe and United States (Materazzi et al. 2008). However, most noodles made in Asian countries do not contain eggs (Oh et al. 1983).

These noodle types differ in appearance, making it easy for a consumer to differentiate them, but it is in their texture where key characteristics for acceptability to consumers are located (Ross 2006; Szczesniak 2002). Texture is a perception about the physical mouthfeel of foods formed from a composite of inputs from sensory receptors lying in and around the mouth (Lucas 2004; Szczesniak 2002), and the eating quality of cooked noodles depends largely on their firmness, resilience, and surface characteristics (Oh et al. 1985). The elasticity of noodles, in terms of shape recovery after 181

deformation, is known to be an important attribute (Ross et al. 1997). It has proved difficult to model textural evaluation from instrumental testing and sensory panel testing remains the most reliable way of predicting consumer response (Li et al. 2012). However, this is time-consuming and costly, and becomes impractical when large numbers of samples are to be evaluated. Thus, a variety of instrumental methods have been developed to estimate the texture of certain foods, e.g. noodles (Edwards et al. 1993). These potentially provide researchers with a convenient and cost-effective way to track changes in the mechanical and physical properties of foods that could relate to the perception of texture (Ross 2006). Increasingly, physiological techniques are closing the gap between in vitro instrumentation and sensory perception by attempting to measure the mechanical behavior of foods in the mouth, particularly surface textural properties (Davidson et al. 1998; de Wijk & Prinz 2005; Prinz et al. 2006; van Aken et al. 2007). Nevertheless, the understanding of such behavior in fundamental terms still demands the accurate measurement of food properties.

Indentation is one of the oldest types of mechanical test. It is convenient and inexpensive without need to shape specimens into standard sizes and shapes, as required for compression and tension (Briscoe et al. 1994; Gonzalez Gutierrez 2008; Huang et al. 2002; Ma et al. 2003). Originally, as a load-controlled technique without measurement of displacement, indentation was limited to establishing the hardness of stiff solids that approximate to ideal elastic-plastic behavior, i.e. a linear elastic response up to a yield point, followed by plastic flow at a constant stress (Tabor, 1951). Indenters were generally sharp and designed so to produce indentations of geometrically-similar shape whatever their depth. Hardness was defined as that load which produces a given area of indentation, measured in the plane of the surface. Revived in the early 1990’s, indentation is now 182

showing promise for providing essential information on the deformation characteristics of many different types of solids. Generally, it is the depth of indentation that is now controlled while the load is allowed to vary. Hardness is still estimated similarly to load-controlled techniques, but the elastic modulus is also estimated from rebound of material against the probe from the deepest part of the indentation as it is withdrawn (Oliver & Pharr 1992). Extension of the method to polymers to include viscous flow (Briscoe & Sebastian 1996; Briscoe et al. 1994; Goh & Scanlon 2007; Oyen & Cook 2003) has broadened its utility and a growing literature includes data on foods and biological soft tissues (Liu et al. 2003), cartilage (Simha et al. 2007), gels (Hu et al. 2010), biological tissues (McKee et al. 2011).

Our aims have been to institute mechanical tests of noodle quality that would be easy to make accurately in industrial application, and which are grounded in recent theoretical developments. A study of the surface properties of noodle sheets was published recently (Wu et al. 2015). In this study, we employed a load-relaxation method with spherical indentation established by Chua & Oyen (2009) to test the internal mechanical properties of five starch-based (mung bean, rice, potato, sweet potato, sorghum) and four flour-based noodle sheets (wheat, buckwheat, millet, and semolina). These sheets were only 3-5 mm thick. This restricted the depth of indentation to much less than 10 % of sheet thickness or else the influence of the specimen substrate (in this study, a glass slide) would begin to influence the results (Galli & Oyen 2008). Indenter size also plays a role limiting the depth of indentation (Galli & Oyen 2008) and we therefore made tests with indenters of three differing radii to ascertain the effect of this. However, miniaturization creates other potential issues because smaller indenters may only contact certain aspects of the microstructure of the noodle sheets rather 183

than produce a ‘structural average’ as a macro-indentation might. To examine this, we made a preliminary investigation of non-indented and post-indented surfaces of one noodle sheet type to examine the mechanism of deformation.

7.2. Materials and methods

The food materials were wheat (Red Bicycle, Hong Kong), millet, buckwheat, and semolina flours, plus mung bean, sorghum, rice, sweet potato and potato starches, all purchased in local supermarkets in Hong Kong. Starches were extracted following the corresponding methods of Collado et al. (1999), Bao et al. (2005), Beta & Corke (2001) and Liu & Shen (2007) with minor modifications. Each starch was ground to pass through a 212 μm aperture sieve after drying in a 35°C oven. Starch sheets were then prepared following Beta & Corke (2001) and Collado et al. (2001), with minor changes. Starch dough was initiated with partial gelatinization of a subset 10 % (w/w) of the total starch used, which served as a binder for the whole mass. Gelatinization was achieved by boiling this starch subset in water (1:7 w/v) until fully gelatinized. This was then mixed with the remaining ungelatinized starch to form a dough of 45 % moisture content. The whole mass was then kneaded by hand for 15 min or until a uniform consistency was achieved. The starch dough was passed through a pasta making machine (KitchenAid Artisan Series KSM150PSES, Whirlpool Corporation, Michigan, USA) at setting 1, then cut into wide strands (3 mm thick) before being dropped into boiling water until thoroughly cooked. After this, the noodle sheets were transferred to cold water for 10 min, re-cooked in boiling water for 10 min, and then finally cooled to room temperature for 1 h in polyethylene bags to prevent moisture loss before testing. For flour sheets, 65 % flour and

184

35 % distilled water were mixed to form a consistent piece of dough. The dough was then passed through the pasta making machine, cut into wide strands (4-5 mm thick), cooked in boiling water until thoroughly cooked, then cooled to RT for 1 h in polyethylene bags to prevent moisture loss. Buckwheat and millet sheets were produced by mixing 40 % buckwheat/millet flour with 60 % wheat flour.

Flour and starch pasting properties were determined by using a Rapid Visco Analyzer (RVA, Newport Scientific, Warriewood, Australia) mainly following methods described by Ohm et al. (2006) and Fonseca et al. (2015). RVA is a commonly-used instrument relating functionality to structural properties by measuring viscous properties of starch and flours (Singh et al. 2010). The RVA method records several parameters. Peak viscosity is the maximum viscosity achieved during the heating process, trough viscosity is minimum paste viscosity achieved during holding period, while the final viscosity is that measured at the end of cooling process. The difference between the peak and trough is termed the ‘breakdown’ viscosity, while that between the breakdown and the final viscosity is the setback. The peak viscosity indicates the water-binding capacity of the flour mixtures, while the final viscosity gives the ability of starch to form a viscous paste after cooking and cooling (Singh et al. 2010). For these measurements, flour samples (3.5 g, 14 % moisture basis) was mixed with 25 g 0.5 mM AgNO3 in an RVA canister; while starch sample (2.0 g, 14 % moisture basis) was mixed with 25 g distilled water. Then an ‘STD 1’ heating and cooling profile was loaded to make measurements as follows: RVA temperature was set at 50 °C for 1 min, then increased at 12 °C /min to 95 °C. It was then held at 95 °C for 2.5 min, reduced at 12 °C /min to 50 °C, and finally held there for 2 min. Total time was 13 min. 185

Following this, gel texture analysis was performed on the gels made by RVA using a TA-XT2 Texture Analyzer (Stable Micro Systems, Godalming, Surrey, England). After RVA, the paddle was removed and the gel in the canister covered by Parafilm and stored at 4 °C for 24 h. The gel was compressed at a speed of 1.0 mm/s to a distance of 10 mm with a 7 mm flat-ended cylindrical probe in a standard two-cycle program. The test was repeated twice. The following parameters were calculated automatically by Texture Profile Analysis (TPA). Gel hardness is defined in TPA as the maximum force (in grams) to fracture the gel. Adhesiveness (in grams-seconds) is the area under the force-time curve as the probe moves back to its initial location. Finally, cohesiveness is the ratio of the area of the positive part of the force-time plot during the second compression relative to that during the first compression.

Indentation tests followed the ‘bulk indentation’ procedure of Chua & Oyen (2009). These were run on a small mechanical tester equipped with a 50 N miniature load cell (Transducer Techniques, Temecula, CA), monitoring displacement with an LVDT (Measurement Specialties DC-EC 2000, Hampton, U.S.). Signals were amplified (giving load accuracy ± 0.003 N; displacement ± 2 µm), converted digitally (14-bit ADC), and displayed in real-time using Labview 2012 software (National Instruments, Texas, U.S.). In the test, a noodle sheet was loaded slowly and evenly by moving one of three blunt probes, each 0.72, 1.65 or 3.175 mm in radius, down onto the specimen for 10 s (Figure 7.1). All displacements were kept to below 10 % of the sheet thickness to avoid substrate ℎ

effects (Galli & Oyen 2008). The strain 𝜖 in the noodles was calculated as 0.2√(𝑅) (Johnson, 1985) where h is the depth of indentation produced by a spherical probe of radius R. A 3-mm thick 186

noodle sheet indented by a probe with R = 3.175 mm reaches 𝜖 ~ 0.06 at 0.3 mm of displacement. However, under an indenter of 0.72 mm radius pushed to the same depth, 𝜖 ~ 0. 13. Since cooked noodles often have nonlinear strain-stress curves (Sui et al. 2006), then in order to compare results with probes of different radii, they should be loaded to similar strains. Accordingly, the displacements were reduced for smaller probes to produce 𝜖 = 0.06-0.08, the actual displacement involved being dependent not only on probe radius, but also on noodle thickness.

After this loading period, known as a ‘load ramp’, the probe was stopped and the displacement held constant (inset in Figure 7.1C) while the load allowed to decay for a further 90 seconds (Figure 7.1C). At any given time point t, the decay in the force F was then assumed to behave according to a series F(t) = Bo + B1exp(-t/τ1) + B2exp(-t/τ2) + … Bnexp(-t/τn)

(1)

where Bn is a fitting constant and τn, its time constant (Chua & Oyen 2009). A parallel series to equation (1) can be written for the time-dependent shear modulus G(t) G(t) = Co + C1exp(-t/τ1) + C2exp(-t/τ2) + … Cnexp(-t/τn)

(2)

where Cn is termed an amplitude coefficient (Chua & Oyen 2009). Equations (1) and (2) can be related by 𝐶0 =

𝐵0 8𝑅0.5 ) 3

ℎ1.5 (

(3)

and 𝐶𝑘 =

𝐵𝑘 8𝑅0.5 1.5 ℎ ( )𝑅𝐶𝐹𝑘 3

(4)

where RCFk is an adjustment factor to account for the length of time needed to ramp the specimen up to its maximum load (Chua & Oyen 2009; Oyen et al. 2008). A convenient assumption, entirely 187

appropriate for gels such as noodles, is to take their Poisson ratio as 0.5 (Sui et al. 2006). Then, the instantaneous modulus is given by E0 = 1.5(C0 + C1 + … Cn), with the infinite modulus E∞ = 1.5C0 (Chua & Oyen 2009). A curve fitting procedure, described by Altenbach (2011), utilizing the Levenberg-Marquardt algorithm, was then applied to the force-time decay curves, defining the best fit to the data as that which minimized the correlation coefficients between each of the amplitude and time constants. Two-term exponential decay functions fitted the data well (Figure 7.1). The average relaxation ratio 𝐸∞ /𝐸0 describes the fraction of behavior of a noodle that is elasticallybased (Chua & Oyen 2009), i.e. its effective ability to recover its shape after deformation.

Finally, to examine the potential interaction between indenter probe and noodle microstructure, preliminary investigations on the deformation of non-indented and post-indented (1.65 mm probe) surfaces of sweet potato noodle sheets were made with an atomic force microscope (Agilent 5500, Keysight, Santa Rosa, CA). Specimens were placed in a humidity chamber, as described for starch granules investigated by Park et al. (2011). Stiff sharp cantilevers (SSS-NCL, Nanoworld, Neuchâtel, Switzerland) having a force constant averaging 48 N m-1, resonant frequency 190 kHz and a tip sharpness