Compositional, morphological, and physicochemical properties of starches from red adzuki bean, chickpea, faba bean, and baiyue bean grown in China

Abstract Starches of four legume varieties grown in China were evaluated for composition, granule structure, turbidity, swelling power, solubility, and thermal and pasting properties. The similar granule shapes, surface fissures, polarized crosses, inner structure characteristics, and granule sizes could be observed among all legume varieties through a variety of microscopy techniques such as light microscopy, scanning electron microscopy, and confocal laser scanning microscopy. Amylose contents were in the range of 30.61%–33.55%. All of the starch varieties showed C‐type X‐ray pattern, but exhibited different relative crystallinity percentage. Significant differences were observed among starch varieties in swelling power, solubility, and light transmittance. Thermal analysis and pasting profile of legume starches showed that all the varieties’ differences are probably due to variation in amylose content. The thermal and pasting parameters of starches were evaluated using differential scanning calorimeter and Rapid Visco‐Analyser, respectively, and significant differences were observed in individual pasting and thermal parameters. The present study can be used for identifying differences between legume varieties for starch structural and physicochemical characteristics and could provide guidance to possible industries for their end use.

it is meaningful to study texture appearance, water-holding capacity, thermal stability, and enzyme digestibility of processed foodstuff (Hoover et al., 2010;Kaur et al., 2013). According to the existing reports, there are many different properties between legume starches and cereal starches, such as high gelation temperature, high resistant starch and amylose content, resistance to shear thinning, more viscosity and lower swelling and breakdown during heating, fast retrogradation, and high elasticity of gel (Maaran, Hoover, Donner, & Liu, 2014;Ratnayake, Hoover, & Warkentin, 2002;Singh, Nakaura, Inouchi, & Nishinari, 2007).
The native starch's internal architecture was made up of concentric alternating amorphous and semi-crystalline growth rings (Singh et al., 2007). The lamellae crystalline regions are formed by double helices of AP side chains. There are three polymorphic forms, namely A-type, B-type, and C-type. Most of the cereal starch has the A-type polymorphs, and B-type polymorphs can be easily found in tuber and high AM cereal starches. The C-type polymorph (a mixture of A-type and B-type polymorphs) is considered to be intermediate between A-type and B-type polymorphs, which are generally found in most legume starches (Hoover et al., 2010).
Physicochemical and structural properties of starches determine their applications in food and nonfood industries. In this study, four kinds of legumes including red adzuki bean (V. angularis L.; information for the applications of these starches in food and nonfood industries.

| Materials
The commercially mature red adzuki, faba bean, and chickpea seeds used in the study were purchased from a seed supplier at Mulei County in Xinjiang Province, China, while the baiyue bean seeds were obtained from Yushan County in Jiangxi Province, China.
8-amino-1, 3, 6-pyrenetrisulfonic acid (APTS) and sodium cyanoborohydride were purchased from Sigma Chemical (St. Louis, MO, USA). All other chemicals and reagents used in experiments were of analytical grade.

| Starch isolation
To isolate starches from four kinds of legumes, we adopt the method described by Singh, Kaur, Sandhu, and Guraya (2004), with proper modification. Briefly, firstly, seeds of legumes were washed with distilled water for several times until dust and other impurity substances were removed, and then, seeds were soaked in a solution which contains 0.16% sodium hydrogen sulfite for 24 hr at room temperature. After that, the steeped solution was drained off, and grains were ground in a laboratory blender. The obtained slurry passed through nylon cloth (100 mesh) successively, and the residue shall be washed by distilled water until no more starch was left. The material retained on the cloth all the time was discarded, and the filtrate slurry was allowed to stand for 1 hr. Afterward, the supernatant was removed by suction and the settled starch layer in distilled water was resuspended and the suspension was centrifuged at 2,200 g for 15 min, and then, the upper non-white layer was scraped off. Starch was then collected and dried in an oven at 40°C for 12 hr.
Then, the starches were smashed by universal mill grindings (100 mesh) and kept in a container at room temperature for further study.

| Chemical composition analysis of legume starch
The starch, protein, lipid, and ash content were performed in terms of the standard AACC methods 76-13, 46-30, 30-25, and 08-01, respectively (AACC, 2000). The amylose content was determined by using the commercially available AM/AP assay kit (catalog: JKY/K-AMYL, Megazyme International Ireland Ltd.). The results were reported on a dry basis.

| Light microscopy observation
A 1:1 glycerol solution (glycerol/H 2 O, V/V) was prepared, and starch granules were suspended in it. Then, an aliquot of starch slurry was transferred onto a glass slide, and a coverslip was placed on the top of the sample for microscopic examination. A polarizing light microscope (DMBA400; Motic China Group Co., Ltd, Guangzhou, China) with a × 40 objective was used to observe birefringence of the starch granules. The microscope was also used without the polarizing filter to observe the appearance of the sample.

| Scanning electron microscopy (SEM) observation
Granular surface of legume starches was viewed with a scanning electron microscope (JSM-6360LV; JEOL, Japan) at an accelerating voltage of 20 kV. Starch samples which were dispersed evenly with a rubber suction bulb were mounted on an SEM stub with double-sided adhesive tape and coated with gold in an argon atmosphere.

| Confocal laser scanning microscopy (CLSM) observation
The microstructure of starch granules was viewed by confocal laser scanning microscopy. The observation of each starch sample was conducted according to the reported procedure. Starch granules (10 mg) were suspended in 15 μl of freshly made APTS, and 15 μl of 1 M sodium cyanoborohydride solution was added.
The reaction mixture was incubated at 30°C for 15-18 hr, and then, the granules were thoroughly washed five times with 1 ml of distilled water. The stained starch granules were then suspended in 20 μl of glycerol/water mixture (1:1, v/v) and then visualized using a Digital Eclipse C1plus CLSM system equipped with He/Ne and Ar lasers (Nikon Co., Ltd., Tokyo, Japan), and a stand for fixed fluorescent cell samples was used for the detection of the fluorescence signal from dye-stained starch granules (Chen et al., 2011;.

| X-ray diffraction analysis
X-ray powder diffraction (XRD) measurements were performed using an X-ray diffractometer (Rigaku D/max-2551/ PC; Rigaku Corporation, Tokyo, Japan) with the following operating conditions: radiation source, CuKa; angle of diffraction scanned from 5 to 60°; step size, 0.02; and step time, 2 s.
In order to avoid the negative effects of relative humidity on relative crystallinity, the starch samples were equilibrated in a 100% relative humidity chamber for 48 hr at room temperature in the first place.

| Determination of thermal properties
Gelatinization characteristics of starch thermal properties were measured by a differential scanning calorimeter (DSC, Q2000; TA Co., New Castle, USA). Starch (3 mg) was put into the aluminum DSC pan and distilled water (12 μl) was added with a microsyringe and then sealed hermetically, reweighed, and allowed it to stand overnight at room temperature in order to attain an even distribution of water before DSC analysis. Subsequently, the samples were heated from 30 to 120°C at 10°C/min. Meanwhile, an empty pan was used as reference for all measurements. During the scans, the space surrounding the sample chamber was flushed with dry nitrogen to avoid condensation.

| Determination of pasting properties
Pasting properties of starches were evaluated using a Rapid Visco-Analyser (RVA) model Master (Newport Scientific, Pty Ltd, Australia).
Firstly, starch (2.0 g, db) and deionized water (25.0 g) were put in the RVA canister, and the slurry was then manually homogenized using the plastic paddle to avoid lump formation before the RVA run. The starch slurry was heated from 50 to 95°C at the rate of 12°C/min, maintained at 95°C for 2.5 min, and then cooled to 50°C at the same rate. The speed of the paddle was 960 rpm for the first 10 s and then 160 rpm for the remainder of the experiment. Pasting temperature, peak viscosity (PV), breakdown viscosity, setback viscosity, and final viscosity were calculated.

| Statistical analysis
All the experiments were performed at least in triplicate and experimental data also were analyzed using analysis of variance (ANOVA) and expressed as mean value ± SD. Duncan's least significant test was used to compare means at the 5% significance level. All statistical computations and analyses were conducted using SPSS 16.0 to compare Simple Pearson's correlation and regression for Windows.

| Proximate composition of legume starches
The chemical composition of legume starches is presented in Table 1. Significant differences could be observed in the proximate constituents of different legume varieties. The starch contents indicate the purity of the isolated starches, were reasonably high (>96%), and ranged from 96.58% to 98.79%. Because of the presence of insoluble protein, fine fiber, and minerals within the starch granules, it is difficult to obtain pure starch from legumes, but it could be decreased after sedimentation and co-settled with the starch granules (Wang, Warkentin, Vandenberg, & Bing, 2014). The starch purity is generally influenced by the isolation method. The protein, fat, and ash contents were in the range of 0.23%-0.35%, 0.25%-0.42%, and 0.07%-0.14%, respectively, wherein baiyue bean showed significantly higher protein, fat, and ash contents value than rest of the legume values. The values of ash, fat, and protein content were in good agreement with the previous report on triangular pea, white pea, spotted colored pea and small white kidney bean, field pea, and lentil (Li, Xiao, Guo, et al., 2014;Wang et al., 2014). The low ash, fat, and protein contents in the isolated starches further indicated the high purity of the starches extracted from each sample.
The amylose content of different legume starches varied from 30.61% to 33.55%, and the highest amylose content was found in red adzuki starch, whereas the lowest amylose content was found in faba bean starch. These values were similar with the values previously reported for black gram (30.33%-40.64%; Wani, Sogi, & Gill, 2015), but relatively lower than those reported for chickpeas (46.5%) and pinto beans (52.4%; Yañez-Farias, Moreno-Valencia, Falcón-Villa, & Barrón-Hoyos, 1997), and field peas (42.9%-43.7%; Ratnayake, Hoover, Shahidi, Perera, & Jane, 2001). The different amylose content among different starch varieties could be attributed to different enzyme activity that involved in the biosynthesis of starch molecules within the starch granules (Krossmann & Lloyd, 2000), and the soil type and climatic conditions during growth and granule size distribution (Singh, Mc-Carthy, & Singh, 2006). The starch granules with high level of amylose always exhibited a slow hydrolysis rate and a rapid retrogradation rate than starch with low level of amylose.  The fissures structure could also be observed on mung bean starch under light microscope (Li, Shu, Zhang, & Shen, 2011). Cracks have been attributed to low granule integrity resulting from sub optimal packing of amylopectin double helices (Blennow et al., 2003). In addition, during processing, it is easy to facilitate rapid water or en-  Table 1.

| Morphology characteristic of legume starches
The growth rings, hilum, cracks, and inner channels could be clearly identified among various starch granules. It could be seen that the morphological structure of inner channel and hilum differs significantly for the starches with different original source ( Figure 1a 5 -d 4 ), and more channels are observed in chickpea starch (Figure 1b 5 ). The amylopectin molecules are prominently located in the semi-crystalline growth rings, while amylose has been accepted to be mainly located within the amorphous zones of the two distinct granule domains (Pérez & Bertoft, 2010). However, the precise role of amylose and amylopectin on the structure of the crystalline cluster structures, as well as the mechanism of the growth ring formation, is still not fully understood. The starch granular pores and channel structure have been always being considered to facilitate the access of enzyme, water, and different solvent to the granule interior.

| X-ray diffraction analysis of legume starches
The crystal packing arrangements of four kind legume starches were investigated using an X-ray diffractometer, and the X-ray diffracto-  Table 2.
The positions of the peaks were found to be similar for the various starches. The chickpea starch exhibited the lowest relative crystallinity degree value among various starch granules, which might be due to the difference in the variety. In addition, there were many other factors such as crystal size, distribution of crystalline regions, and orientation of "crystals" within the semi-crystalline domains, and the extent of polymer interactions could result in different relative crystallinity in starch (Kaur et al., 2013).

| Swelling power and solubility of legume starches
The swelling power and solubility of the legume starches were investigated over the temperature range of 50-90°C, and the results are presented in Table 3 (Table 3), and the lower solubility was due to the integrity of the starch granules where restricted swelling and solubility were present. Meanwhile, the high solubility means that the hydrogen bonds can be easily weakened in the starch granules with high amylopectin content. For all legume starches, a rapid increment of solubility was shown after 70°C (Table 3), a pattern that is due to the higher disorganization of starch granules around the gelatinization temperature.
Swelling power is a measure of the ability of the starch to hydrate under specific conditions such as temperature and water content, and the greater swelling capacity of the starch granules is, the weaker of the binding forces (Hoover & Manuel, 1996). Baiyue bean starch exhibited higher swelling power than the other three kinds of starches at low temperature (50-60°C). Also, chickpea starch exhibited the highest swelling power at high temperatures (70-90°C) in this study. Again, a dramatic increase in the swelling power in all starch samples could be observed above 70°C. The rapid increase in swelling power at 70-90°C may be due to an increase in molecular mobility of the amorphous region, which causes unraveling, and melting of the double helices present within the amorphous and crystalline domains (Ratnayake et al., 2001). The different swelling power among various starch granules may be related on the amylose-lipid complexes, the intra-and inter molecular interactions, and the ratio and molecular weights of amylose and amylopectin (Hoover & Manuel, 1996;Tester & Morrison, 1990), and amylose always acts as an inhibitor of starch swelling.

| Light transmittance of legume starches
The light transmittance (turbidity, clarity, or turbidity) of the starch pastes of the starches from the different legume varieties for a period of 3 days is shown in Table.  The rapid decrease in clarity during the early storage may be due to the aggregation and crystallization of amylose, and then, the decrease rate in clarity during the rest of the storage could be attributed to the leached amylose and amylopectin chains that lead to the development of the functional zone which scatters a significant amount of light (Yu, Ma, Menager, & Sun, 2012). It has been reported that the difference in light transmittance among various starch granules is related to different granule remnants, granule swelling, leached amylose and amylopectin, amylose and amylopectin chain lengths, intraor intermolecular bonding, lipids, and substitution (Jacobson, Obanni, & BeMiller, 1997). As for legume starches, the legume starches contained varying amounts of phosphate monoester substituents might also contribute to variation in starch gel clarity (Jane, Kasemsuwan, & Chen, 1996).

| Thermal properties of legume starches
The gelatinization thermograms of starches isolated from different legume varieties are shown in Figure 3, and the thermal transition temperatures onset transition temperature (T o ), peak temperature (T p ), conclusion transition temperature (T c ), gelatinization temperature range (ΔT r ) and gelatinization enthalpy (ΔH) on a dry starch basis are presented in Table 5. The DSC endothermic peak is linked to starch gelatinization which can reflect the loss of ordered structures (double helices/crystallites) of starch granules (Cooke & Gidley, 1992). The gelatinization transition parameters and gelatinization enthalpy of starches from various legume varieties differed significantly as shown in Table 5. Gelatinization temperature is a reflection of the perfectness of starch crystallites (Tester & Morrison, 1990).
The  Values within the same column for the same wheat genotype followed by a common letter are not significantly different (p < 0.05). The data are presented as mean ± SD (triplicate analysis). A granule during gelatinization (Cooke & Gidley, 1992). The lower ΔH indicates a lower degree of organization in or a lower stability of the crystals (Chiotelli & Meste, 2002). Red adzuki bean and faba bean starch had broader gelatinization temperature ranges and higher gelatinization enthalpy than other starch samples (Table 3). These results are in agreement with the observation that red adzuki bean and faba bean starch had a higher relative crystallinity degree (Table 2) and required more energy for gelatinization.

| Pasting properties
Pasting viscosity profiles of starches isolated from various legumes are illustrated in Figure 4, and the related parameters are summarized in Table 6. The overall shape of the RVA pasting curves was similar for the four kinds of legume starches (Figure 4). However, there were significant differences (p < 0.05) in PV, trough viscosity (TV), breakdown (BD), final viscosity (FV), setback (SB), peaking time (PT), and pasting temperature (GT) between the starches (Table 6). Changes in viscosity during heating process give an indication of the starch stability, and the changes occurring during cooling show the consistency of the product when consumed (Kaur, Sandhu, & Lim, 2010). Peak viscosity of rice starch samples ranged from 2657 to 3911 cP, and the highest PV was shown by red adzuki bean starch, whereas the lowest PV was showed by baiyue bean starch. Peak viscosity refers to the maximum viscosity attained by gelatinized starch during heating in water; in the meantime, it can also indicate the water-binding capacity of the starch granule (Shimelis, Meaza, & Rakshit, 2006). The difference in PV among various legume starches could be due to variation in the water absorption rate and starch swelling properties during heating.
Trough viscosity was the highest for red adzuki bean starch and the lowest for baiyue bean starch, and BD viscosity was the highest for baiyue bean starch and the lowest for chickpea starch. TV has been reported to be influenced by the extent of amylose leaching, amylose-lipid complex formation, friction between swollen granules, granule swelling, and competition between leached amylose and remaining granules for free water (Olkku & Rha, 1978). Breakdown is an indication of the degree of its organization within the starch granule and a tendency to lose viscosity upon holding and shearing. It has been reported that the higher the breakdown viscosity value is, the lower the ability of the starch granules to withstand heating and shear stress during cooking (Adebowale, Olu-Owolabi, Olawunmi, & Lawal, 2005). Thus, the low breakdown value of chickpea starch indicates its restricted starch granule swelling and slow tendency to lose viscosity, which might be able to withstand more heating and shear stress than other starch varieties that possess higher breakdown values.
The FV value varied from 2573 cP to 4764 cP, and the highest FV was recorded for chickpea starch and the lowest for baiyue bean starch. FV can show the ability of starch to form a viscous paste, which is largely influenced by the retrogradation tendency of the soluble amylose during cooling stage (Olkku & Rha, 1978

| CON CLUS ION
Structural and physicochemical properties of starches from red adzuki bean, chickpea, faba bean, and baiyue bean starch were characterized, and a high variability among various legume starches properties was found. There were minor differences in morphology characteristic properties among various legume starches. The X-ray diffraction patterns of all legume starches showed C-type pattern with relative crystallinity degree varied among the varieties. The swelling power, solubility, light transmittance, and gelatinization parameter values of legume starches varied among the varieties.
Pasting properties significantly varied and are useful for assessing the cooking quality of legume starches. Differences in physicochemical properties were influenced by the differences in morphology, composition, and structure of various legume starches. The findings would be useful in selecting legume variety or legume starch for specific food and ingredient applications, and also for further investigations on fine structure of starch granules and modifications targeting food ingredient applications.

ACK N OWLED G M ENTS
This study was financially supported by the Fundamental Research Funds for the Natural Science Foundation of China (31301557) and the Central University Basic Scientific Research Funds (Northwest A&F University, 2014YB019).

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.