Effects of heat–moisture, autoclaving, and microwave treatments on physicochemical properties of proso millet starch

Abstract Proso millet starch was modified by heat–moisture treatment (HMT), autoclaving treatment (AT), and microwave treatment (MT). The effects of these treatments on the starch physicochemical, structural, and molecular properties were investigated. The amylose and resistant starch contents were increased by AT and MT, but only slightly by HMT. HMT and AT significantly increased the water‐holding capacity, to 172.66% and 191.63%, respectively. X‐ray diffractometry showed that the relative crystallinity of the HMT sample decreased by 20.88%, and the crystalline peaks disappeared from the AT and MT sample patterns. The thermal treatments decreased the proso millet starch molecular weight to 1.769 × 106, 7.886 × 105, and 3.411 × 104 g/mol, respectively. The thermal enthalpy decreased significantly in HMT. Modification significantly changed the pasting profiles of the native proso millet starch, and the peak viscosity, setback, and breakdown values decreased. These results clarify the mechanism of starch changes caused by thermal treatment.

improved to extend the product shelf life (French, 1984). A typical process generally leads to viscosity reduction, structural loosing, and product deterioration under long-term storage conditions, especially in repeated freezing treatments.
Heat-moisture treatment (HMT) of starches refers to exposure of starch granules at a temperature above the glass transition temperature but below the gelatinization temperature for a certain time period and at restricted moisture content. HMT causes physical modifications of the starch granules with respect to size, shape, and birefringence (Hoover & Manuel, 1996).
Autoclaving treatment (AT) promotes hydration of the amorphous zone in starch granules under the action of a pressure field. The amorphous layer of the starch granule crystallization zone swells in water with increasing pressure, which causes rearrangement of the amylopectin double helices (Bravo, Siddhuraju, & Saura-Calixto, 1998). Microwave treatment (MT) affects starch through dielectric heating and electromagnetic polarization effects (Bilbao-Sáinz, Butler, Weaver, & Bent, 2007). Modification alters the starch physicochemical properties and can improve the functionality of native starch (Bemiller & Huber, 2015). Previously reported studies of proso millet starch have focused on crop breeding and genotype , or comparing the physicochemical properties of different varieties (Wang et al., 2018). Less work has been done on the effects of HMT, AT, and MT on the physicochemical properties of proso millet starch. The proximate composition of proso millet is similar to those of other common grains (e.g., rice, corn, and wheat) (Park et al., 2018;Silva et al., 2017;Zhao et al., 2018), and understanding the processing properties of proso millet starch would be conducive to its development and use.
The aim of this study was to explore the behaviors of native proso millet starch and starches modified by different thermal treatments that could improve the value of proso millet starch and proso millet in the food industry. This study will clarify the mechanisms involved in thermal treatments of starch and enable full use of proso millet resources.

| Materials
Proso millet starch with an amylose content of 13% was kindly supplied by the National Engineering Laboratory for Wheat and Corn Deep Processing, China. All chemicals were analytical grade and used as obtained, without further purification.

| Sample preparation
Proso millet starch (20 g, dry basis) and water (80 ml) were mixed well. The HMT sample was heated for 30 min at 100°C in a sealed glass bottle, the AT sample was heated for 30 min at 121°C (0.1 MPa), and the MT sample was heated at a microwave power of 500 W for 10 min. The samples were then freeze-dried, ground, and stored in plastic vials until further use.
A K-RSTAR kit (Megazyme International Ireland Ltd.) was used to determine the resistant starch content. The samples were hydrolyzed with α-amylase and amyloglucosidase (provided with the kit) for 16 hr at 37°C. In all the methods, free glucose formed by enzymatic hydrolysis was quantified colorimetrically with an oxidaseperoxidase glucose reagent.

| Water-holding capacity
The Water holding capacity (WHC) was determined by the method described by Yamazaki (1953). The sample (2 g, dry basis) was dissolved in distilled water (25 ml) by shock mixing, soaked at room temperature for 30 min to form a uniform paste, and centrifuged (2,000 g, 10 min). After separation of the supernatant, the water content of the sediment was determined directly by weighing.

| Solubility and swelling power
The solubility and swelling power were determined by using a modified version of the leaching method reported by Adebowale, Afolabi, and Olu-Owolabi (2005). Starch (0.2 g) was dissolved in water (10 ml). The samples were subjected to intermittent shocks in a water bath for 30 min at 60, 70, 80, and 90°C. The treated samples were cooled to room temperature and then centrifuged at 2,000 g for 15 min. The supernatant was poured into an aluminum box and dried at 105°C to obtain the water-soluble starch. The precipitate was swelled starch.
where W ss is the weight of soluble starch (g), W sp is the weight of sediment paste (g), and W s is the weight of the sample (g). (1)

| HPSEC-MALLS-RI
The weight-average molecular weight (M w ), z-average radius of gyration (R z ), and polydispersity index (PDI) of the samples were determined by high-performance size-exclusion chromatography coupled with multiangle laser-light scattering and refractive index detection (HPSEC-MALLS-RI) under the conditions described by Zhang, Li, Chen, and Situ (2016) with some modifications. Starch (50 mg) was dispersed in dimethyl sulfoxide (DMSO; 2 ml). The suspension was heated in a boiling-water bath for 15 min with intermittent stirring.
Aqueous ethanol (95%, v/v, 6 ml) was added to the starch suspension to precipitate the starch. After standing for 15 min, the ethanolprecipitated starch was separated by centrifugation at 3,000 g for 10 min, the supernatant was discarded, and the tubes were drained on tissue paper for 15 min. The starch pellet was redissolved in DMSO (90%, v/v, 5 ml) with LiBr (50 mmol/L) and left at 60°C overnight. Before injection, the sample solution was filtered through a membrane filter (5.0 μm). The mobile phase was DMSO (HPLC grade, 90%, v/v) with LiBr (50 mmol/L) and was filtered through a 0.22-μm membrane filter and degassed by ultrasound before use; the injection volume was 100 μl. The flow rate was 0.5 ml/min, the column was maintained at 60°C, and the d n /d c value was 0.0740 ml/g. The data were analyzed by Astra software (Wyatt Technology).

| X-ray diffractometry
The crystal structures of the native and modified samples were Scanning was performed from 3° to 50° (2θ) with a step interval of 0.02° and a scanning rate of 10°/min. The relative crystallinity was calculated as the ratio of the crystalline peak area to the total diffraction area.

| Differential scanning calorimetry
All Differential scanning calorimetry (DSC) data were obtained with a Q2000 instrument (TA, USA). The sample processing method described by Hélène et al. (2007) was used. Briefly, samples were prepared in triplicate, by accurately weighing starch (2 mg) and dissolving it in deionized water (10 ml) in a pan, mixing, and holding for 12 hr to equilibrate. Samples were heated from 40 to 120°C at a rate of 10°C/min. An empty aluminum pan was used as the control.

| Pasting properties
The pasting properties of the samples were determined with a Rapid Visco Analyser (Model RVA; Newport Scientific) by using a standard starch profile. A mixture of the sample (3.5 g) in distilled water (25 ml) was stirred at 160 rpm. The samples were held at 50°C for 1 min and then heated to 95°C at 4°C/min and then held at 50°C for 5 min. The pasting temperature, peak viscosity (PV), breakdown (BD) value, final viscosity (FV), and setback (SB) value were recorded.

| Scanning electron microscopy
The sample particle microstructures were examined by Scanning electron microscopy (SEM) (PW-100-011; LASER Company); the method described by Kiseleva et al. (2005) was used. Freeze-dried samples were mounted on an aluminum stub by using double-sided sticky tape and coated with a thin film of gold. Images were recorded at an accelerating voltage of 20 kV.

| Statistical analysis
Sample analyses were performed in triplicate, and standard deviations were reported. A comparison of the means was ascertained by Tukey's test to be at a 5% level of significance by analysis of variance.

| Amylose, amylopectin, and resistant starch contents
The amylose, amylopectin, and resistant starch contents of proso in the starch by gelatinization. For MT, microwave gelatinization can destroy the starch crystallinity before particle expansion, and therefore, more of the molecular material was leached at higher temperatures (Palav & Seetharaman, 2006). Gelatinization under pressure may promote fracture of the amylopectin molecular structure or hydrolysis of long-chain amylose molecules to shortchain molecules.

| Water holding capacity
The WHCs of proso millet starch after HMT, AT, and MT increased significantly (p < .05) to 172. 66%, 191.63%, and 197.13%, respectively ( Table 2). The increase in the WHC was mainly the result of starch gelatinization caused by thermal treatment. In addition, this indicates that a mass of bound water was produced by covalent bonding between hydroxyl groups and the starch molecular chains. The amount of resistant starch can be increased by gelatinization, and the increase in the number of hydrophilic hydroxyl groups on the outside of the glucose unit enables moisture absorption. According to Pinnavaia and Pizzirani (1998), the WHC of starch has a significant correlation with its gelatinization degree. Singh and Adedeji (2017) reported a similar trend for HMT proso millet starch, and an increase in AT samples from different botanical sources was reported by Hoover (2010).

| Solubility and swelling power
The

| Molecular weight distribution
The M w of the heat-treated samples were determined by HPSEC-MALLS-RI; the results are shown in Table 3. The M w of the proso millet starch used in this work was 8.694 × 10 6 g/mol, which is lower than previously reported values. This can be attributed to differences botanical origins of the starch, processing methods, and starch structures (Yoo & Jane, 2002;Zhang et al., 2014). In this study, thermal treatment of starch led to decreases in both the M w and R z values. HMT, AT, and MT decreased the M w values, especially MT. The M w values were 1.769 × 10 6 g/mol after HMT, 7.886 × 10 5 g/mol after AT, and 3.411 × 10 4 g/mol after MT. These changes may be related to changes in the amylose content. Previous research showed that M w decreases with increasing amylose content (Aberle, Burchard, Vorwerg, & Radosta, 1994;Fishman, Rodriguez, & Chau, 1996), and this is also consistent with the amylose content results. The results show that thermal treatment decreased the M w of proso millet starch.
The z-average radius of gyration (R z ) refers to the theoretical probability of finding a molecule at a given distance from the center. The branch length and chain-branching mode affect the R z value (Jackson, 2010). The R z value is therefore affected by the botanical origins of the starch. The high R z of the proso millet used in this work can be attributed to high amylopectin content. The R z value was decreased by thermal treatment; this result is consistent with previous research (Yang et al., 2017;Zeng, Ma, Kong, Gao, & Yu, 2015). In addition, the decreased R z indicates a reduction in the branching degree.
The PDI is the ratio of the weight-average molecular weight to the number-average molecular weight; it is related to the molecular mass and polydispersity of the starch (Shin et al., 2009 branched chains in the starch system. The results show that thermal treatment alters the structure and properties of proso millet starch.

| X-ray diffraction
The XRD patterns and peak intensities of the samples are shown in Figure 2. The native proso millet starch and HMT sample gave an Starch recrystallization is a complicated process, which involves conformational changes, chain alignment, crystal packing, and phase propagation (Sun, Gong, Li, & Xiong, 2014). Thermal treatment may have destroyed the starch crystalline structures. During thermal treatment, starch crystallites may have been disrupted or their orientation may have changed, as a result of partial or complete gelatinization and movement of double helices (Gunaratne & Hoover, 2002).

| Differential scanning calorimetry
Thermograms of the samples are shown in Figure 3. It has been reported that the enthalpy values of gelatinized starch indicate melting of crystallites that were formed during gelatinization by association between adjacent double helices, and the endotherm peak is attributed to the melting of gelatinized amylopectins other than amylose (Krueger, Knutson, Inglett, & Walker, 2010).

| Pasting properties
The pasting properties of the samples are summarized in Table 4.
All the modified starches showed significant changes (p < .05) in their pasting properties compared with those of native proso mil-   (Doutch et al., 2012;Hoover, 2010).

| Scanning electron microscopy
Scanning electron microscopy was used to investigate and confirm changes in the surface morphologies of the starch gels ( Figure 4).
Changes in the morphologies of starch granules are related to interactions between crystalline and amorphous regions in the starch (Wasserman et al., 2004). The microstructure of the gel obtained by HMT was a honeycomb with a thick stromal wall and pores of nonuniform size. The AT gel had more surface folds, and its stromal wall was thicker, its network structure looser, and pores larger than those of the HMT gel. The stromal wall of the MT gel was thin, and it had a uniform pore size and a neat regular network structure.
The main reason for the differences among the gel microstructures is that different expansion and rupture spaces were provided for the starch granules during different heating processes. HMT, AT, and MT evidently affected the form and degree of agglomeration of granules. This is reasonable because of the partial gelatinization caused by moisture and the different thermal energies during HMT, AT, and MT. HMT leads to inconsistent swelling of the granules and the appearance of concavities on the surfaces; AT, which adds a pressure field to HMT, further increases the degree of gelatinization. In a microwave field, the effects of dielectric heating and electromagnetic polarization give rapid and regular starch gelatinization. The surface morphologies of the samples varied depending on the treatment.
This suggests that HMT, AT, and MT destroyed the surface structures of the proso millet starch granules to different extents; this is in agreement with previously reported results (Zavareze, Storck, Castro, Schirmer, & Dias, 2010;Zhong et al., 2017).

| CON CLUS IONS
This study explored the effects of HMT, AT, and MT on the physico-

CO N FLI C T O F I NTE R E S T
The authors declare that they do not have any conflict of interest.

E TH I C A L A PPROVA L
The study did not involve any human or animal testing.

I N FO R M E D CO N S E NT
Written informed consent was obtained from all study participants.