Ultrasonic-assisted binding of canistel (Lucuma nervosa A.DC) seed starch with quercetin

Highlights

• •Ultrasonic modified starch and quercetin combined was formed non-inclusion complex.
• •Addition of quercetin could protect the damage of starch granule size by ultrasonic.
• •Quercetin and starch binding through H-bonds and van der Waals interactions.
• •Ultrasonic modified and quercetin combined affected starch physicochemical properties.

Abstract

In order to provide a reference for improving the physicochemical properties of starch, the study of starch polyphenol complex interaction has aroused considerable interest. As a common method of starch modification, ultrasound can make starch granules have voids and cracks, and make starch and polyphenols combine more closely. In this research, canistel seed starch was modified by ultrasonic treatment alone or combined with quercetin. The molecular structure, particle characteristics and properties of starch were evaluated. With the increase of ultrasonic temperature, the particle size of the dextrinized starch granules increased, but the addition of quercetin could protect the destruction of starch granule size by ultrasonic; X-ray diffraction and infrared spectra indicated that quercetin was bound to the surface of canistel seed starch through hydrogen bonding, and the complex and the original starch had the same crystal structure and increased crystallinity; by molecular simulation, quercetin bound inside the starch molecular helix preserved the crystalline helical configuration of starch to some extent and inhibited the complete unhelicalization of starch molecules. Meanwhile, hydrogen bonding was the main driving force for the binding of starch molecules to quercetin, and van der Waals interactions also promoted the binding of both. In the physicochemical properties, as the temperature increased after the combination of ultrasonic modified starch combined with quercetin, the solubility, swelling force and apparent viscosity of the compound increased significantly, and it has higher stability and shear resistance.

1. Introduction

Canistel (Lucuma Nervosa A.DC) belongs to the family Sapotaceae. That is a tropical fruit and Its flesh color and texture are very similar to yolk, so it is so named [1]. It is widely distributed in South America (Belize, Guatemala, and Bahamas) and subtropical regions (Philippines and China) [2]. Canistel seeds contain a lot of starch (13.89%) (amylose content 33.65%), polysaccharide (8.04%), crude fat (6.27%), polyphenols (2.58%) and ash (2.22%) [1], [3]. Indeed, eating or using canistel pulp will be generated thousands of tons of seeds as waste. Besides, there are very few studies reported on canistel [4], [5], and even fewer studies focused on canistel seeds starch [6], [7], [8], [9], [10]. Canistel seeds have biological activities including anti-tumor property, prevention of heart disease, enhancement of immunity, alleviation of pain, anti-ulcer activity [6]. Canistel seeds have also been reported to have liver protection, anti-oxidation and treatment of inflammation with their rich polyphenols [11], [12].

The total phenol content of canistel seeds is 52.82 μmol GAE/g dry weight, and the total flavonoid content is 5.99 μmol Q/g dry weight. The main phenols are flavonoids (19 species), especially quercetin [9]. Quercetin is beneficial to human health, with antioxidant, anti-inflammatory, anti-viral, anti-cancer activities and alleviating functions of some cardiovascular diseases (heart disease, high blood pressure and high cholesterol) [13]. However, quercetin has little water solubility, poor chemical stability, and low bioavailability, which limits its application. Moreover, quercetin can change starch morphology, improve thermal stability and digestibility [14], [15]. Quercetin also interacts with buckwheat (rich in quercetin) starch during heating to affect its physicochemical properties [16].

Ultrasonic treatment of starch which is a thermodynamic method, may be an economical and feasible method for starch modification. Its main principle is to form cavitation in a short time and generate high heat and high pressure to destroy the surface of solid matrix, enhance mass transfer and accelerate diffusion [17]. Therefore, it can degrade starch polymers, releasing more amylose, and make it interact with other molecules under thermal conditions. At the same time, due to its advantages of high selectivity, less chemicals, and time optimization, ultrasonic treatment can be used as a “green” chemical treatment method. According to previous studies, ultrasonic treatment of starch (different temperature or different power) will affect the morphology and physicochemical properties of starch from different sources. In addition, it can also change the surface and the granule cavities of the starch [18].

Therefore, the objective of this study was to apply ultrasonic synergism to the preparation of the complex of canistel seed starch and quercetin. The physicochemical properties of the complex were characterized by laser diffraction particle size analysis, scanning electron microscopy (SEM), x-ray diffraction (XRD), Fourier transform infrared spectroscopy (FTIR), molecular simulation and other methods. At the same time, the solubility, swelling force and rheological properties of the complex were studied. The potential mechanism of synergism between ultrasonic-modified canistel seed starch and quercetin was studied, which provided a new method for developing functional starch materials.

2. Materials and methods

2.1. Materials and sample preparation

Canistel seed starch was isolated according to the mothod reported by He et al. [4]. Fresh canistel seeds were put into an oven (DHG-9023A, Shanghai Yiheng Technology Co., Ltd, Shanghai, China) and dried at 45 °C before being hulled. After drying, the seeds were ground and dried with 80-mesh sieve. Then, the sample was added to n-hexane (1:3, m/v) and mixed and stirred for 24 h. After centrifugation, the precipitate was air-dried, and sodium hydroxide (0.04%, w/v) was added and stirred for 6 h. The mixture was centrifuged at 6000 rpm for 10 min, and the precipitate was collected and washed with water to a neutral pH. The precipitate was collected and the low molecular compounds were washed out with ethanol (70%). Finally, it is dried in an oven at 30 °C, and the ground powder was passed through an 80-mesh sieve for dry storage at 25 °C.

2.2. Preparation of ultrasonic modified starch-quercetin complex

Starch-quercetin complex was prepared according to the literature previously described by Gao et al. [19]. A 200.00 g (dry basis) of canistel seed starch was weighed to prepare a 10% (w/v) starch suspension by adding distilled water. The suspension was stirred at 70 °C for 20 min to obtain pre-gelatinized canistel seed starch (pre-CSS). The suspension was stirred, then put it into the ultrasonic processor (TL-1200Y, Jiangsu Tianlin Instrument Co. Ltd, Jiangsu, China). The ultrasonic power was set to 450 W for 20 min (after 5 s of ultrasonic treatment, stop for 5 s) and the temperature was controlled at 70 °C, 80 °C and 90 °C, then turned off the ultrasonic processor and kept stirring temperature for 1.5 h. The obtained samples were respectively named pre-CSS-U-70, pre-CSS-U-80, and pre-CSS-U-90. After preparing the 10% starch suspension, quercetin solution (10%, w/v) was added into the suspension and stirred. The suspension was stirred at 70 °C for 20 min to obtain pre-gelatinized starch-quercetin (pre-CSS-Q) [14]. The suspension was stirred, then put it into the ultrasonic processor. The ultrasonic power was set to 450 W for 20 min (after 5 s of ultrasonic treatment, stop for 5 s) and the temperature was controlled at 70 °C, 80 °C and 90 °C, then turned off the ultrasonic processor and kept stirring temperature for 1.5 h. The obtained samples were respectively named pre-CSS-U-Q-70, pre-CSS-U-Q-80, and pre-CSS-U-Q-90. The all samples were freeze-dried and passed through a 80-mesh sieve.

2.3. Particle size analysis

The size of the sample granules (expressed in % volume) was determined by the Mastersizer 3000 particle size analyzer (Malvern Instruments Ltd, Malvern, UK). The mothod was reported by He et al. [4].

2.4. Starch granules morphology analysis

The morphology of sample was observed on a scanning electron micrographs (SEM) (S4800, HITACHI, City, Japan) with a voltage of 5.0 kV.

2.5. X-ray diffraction (XRD) analysis

The Bruker AXS D8 Advance X-ray diffractometer (Bruker, Beijin, Germany) was used to obtain XRD pattern. The measured step of XRD was 0.02°, the scanning rate was 1°/min, and the scan range was 5 to 40° (2θ).

2.6. Fourier transform infrared spectral (FTIR) analysis

An FTIR spectrometer (Nicolet iS10 Nexus 670 FTIR, Nicolet Inc, City, USA) was used to obtain the sample’ spectrograms. Sample was mixed with KBr powder (1:100, w/w) and pressed into a tablet. The fourier infrared spectrometer settings: the scanning range was 4000 ∼ 400 cm⁻¹, the resolution was 4 cm⁻¹, and the scanning number was 32.

2.7. Molecular dynamics simulation

The initial structure of Quercetin was downloaded from Pubchem with the CID 5280343, optimized using the Avogadro program and the ORCA program, using water as the SMD solvent, the topology was created using the GAFF force field-based anchamber tool; and its RESP charge was fitted using Multiwfn [20], [21]. The canistel seed starch was created using the LEap program of the AMBER suite for the V-helical starch structure, which contains 26 glucose units, where the dihedral angles φ(O5-C1-O4′-C4′) and ψ (C1-O4′-C4′–C3′) were set to 107.5° ± 4.5° and 110.8° ± 5.5°, respectively [5], [6]. The topological parameters of this starch were given by the GLYCAM_06j-1 force field. In order to better investigate the interaction between quercetin and starch, three starch-containing simulations were constructed in this study: a starch alone system (Starch), a complex system with quercetin bound inside the starch helix (intra), and a complex system with quercetin away from starch (inter). All MD simulations were performed using the GROMACS 19.5 package for each system, then the trajectories were analyzed using the GROMACS 22.3 package, and visual analysis was performed using VMD 1.9.3 software.

2.8. Swelling power (SP) and solubility (SOL)

The SP and SOL of the sample were based on the previously described method with a slight modification [4]. The temperature of measuring SP and SOL is 55 °C, 65 °C, 75 °C and 85 °C respectively. The value of the SOL was the ratio of the weight of soluble sample to the weight of the sample. The value of SP was the ratio of the weight of water in the swollen granules.

2.9. Differential thermogravimetric (DTG)

The 5.0 ± 0.5 mg sample was put into an aluminum crucible, and TGA and DTG curves were obtained by using a thermogravimetric differential thermal analyzer (TG-DTA8122, Rigaku Group Inc, City, USA). Under the dynamic atmosphere of dry nitrogen, the flow rate was 20 mL/min, and the heating rate was 10 °C/min in the temperature range of 25.0 ∼ 600.0 °C.

2.10. Rheological properties

The rheological properties of the samples were measured at 25 °C using a strain-controlled rheometer equipped (Haake MARS 40, Thermo Fisher Scientific, City, Germany) with parallel plates with diameter of 50 mm (plate spacing of 1 mm). The sample was prepared into 6% (w/v) suspension, stirred in boiling water bath for 30 min and stirred continuously to make it fully gelatinized, cooled to room temperature, and balanced at room temperature for 1 h, then loaded on the rheometer plate (diameter 40 mm, gap 1 000 m) at 25 °C. By performing a dynamic sweep test in the linear viscoelastic region (LVR), the angular frequency range is 0.01 ∼ 15 Hz. The viscosity curve was drawn, shear rate is 0.01 ∼ 300 s⁻¹. Before each experiment, the edge of the sample is covered with a thin layer of silicone oil to prevent water evaporation. The data were modeled by Herschel-Bulkley equation.

$$\delta ={\delta }_0+K{\gamma }^n$$

δ, shear stress (Pa); γ, shear rate (S^-1); K, consistency coefficient (Pa·sⁿ); n, flow behavior index; δ₀, yield stress (Pa).

2.11. Statistical analysis

All experimental data were statistically analyzed and expressed as mean ± standard deviation (SD). The figures were obtained using the Origin 2017 software. The data was statistically analyzed by SPSS 19.0 using one-way analysis of variance test (ANOVA) and Duncan's multiple range test (Duncan’s multiple range test), and the significance level was set at P less than 0.05.

3. Results and discussion

3.1. Particle size analysis

The particle size and distribution of all samples are shown in the Table 1 and Fig. 1, and the difference of particle size of all samples were statistically significant. The pre-gelatinized fruit seed starch (pre-CSS) had a bimodal distribution, while the sonicated samples had a unimodal distribution. Ultrasonic treatment at 80 °C had the largest particle size with an average diameter of 153.88 μm. The results showed that the starch granules became larger after ultrasonic treatment. In the case of large-size granules, the volume percentage increases significantly under the influence of ultrasonic treatment at different temperatures. This might be the possible reason that the compactly bound lamellae structures (crystalline and amorphous lamellar structures) became loose as the ultrasound treatment made the starch particles larger [22]. Another reason may be that the canistel starch undergoes aging, causing larger granular substances in the starch to increase the particle size of the sample. At the same time, ultrasonic treatment at different temperatures could cause the change of starch particle size, thus changing the physicochemical properties of starch [23], and ultrasonic treatment at 80 °C had the greatest effect.

Table 1.

The size and distribution of pregelatinized canistel seed starch and canistel seed starch-quercetin complexes.

	D [4,3] (μm)	Cumulative percentage(%) diameter (μm)
		d (0.1)	d (0.5)	d (0.9)
pre-CSS	69.07 ± 3.63e	7.17 ± 0.07 g	38.36 ± 1.17e	178.04 ± 11.36e
pre-CSS-Q	46.82 ± 1.24 g	11.13 ± 0.03e	29.27 ± 0.17f	111.09 ± 3.93 g
pre-CSS-U-70	105.25 ± 0.74c	14.08 ± 0.03c	84.24 ± 0.58c	230.60 ± 1.95c
pre-CSS-U-80	153.88 ± 4.88a	23.44 ± 0.45a	133.59 ± 4.13a	311.91 ± 10.20a
pre-CSS-U-90	134.59 ± 0.75b	18.60 ± 0.05b	113.95 ± 0.79b	281.48 ± 1.57b
pre-CSS-U-Q-70	53.60 ± 1.29f	10.05 ± 0.03f	23.30 ± 0.21 g	154.62 ± 4.01f
pre-CSS-U-Q-80	48.83 ± 1.57 fg	10.95 ± 0.09e	27.96 ± 0.65f	121.07 ± 4.02 g
pre-CSS-U-Q-90	80.28 ± 7.50d	12.32 ± 0.54d	48.30 ± 4.21d	199.11 ± 19.58d

D [4,3] (volume-weighted mean diameter) indicates the average particle size; d(0.1), d(0.5) and d(0.9) indicates that the particles with a particle size smaller than the diameter account for 10%, 50% and 90%.

Fig. 1.

Particle size distribution of pregelatinized canistel seed starch and canistel seed starch-quercetin complexes.

The particle size of the existed quercetin fiber strips and starch granules (46.82 μm) was smaller than that of the directly pretreated starch (69.07 μm). Under the same conditions, the particle size of starch-quercetin complex was smaller than that of starch, which may be due to the restriction of starch aggregation by the addition of quercetin [24]. Among the canistel seed starch-quercetin complex compounded by ultrasound at different temperatures, pre-CSS-U-Q-80 had the smallest particle size (48.83 μm), and pre-CSS-U-Q-90 had the largest particle size (80.28 μm). The same was true in the cumulative distribution percentage of starch volume. This indicates that addition of quercetin would reduce the size of the starch granules of the composite, which may be due to the combination of quercetin and starch to make the overall structure more compact. Meanwhile, starch granule size difference can be explained by the following structural characterization.

3.2. Scanning electron microscope

After ultrasonic treatment, the surface of canistel seed starch granules became rougher and more dispersed, while the deformation was serious and the flakes were obvious. At the same time, this phenomenon became more obvious with the increase of temperature (Fig. 2). It was attributed to that the energy of ultrasound produces high-frequency vibration, which made starch produce friction, destroyed the granule morphology of starch, and caused the change of starch characteristic morphology [25]. On the other hand, starch and water gelatinize at higher temperature could absorb water to change its morphology and destroy the integrity of starch [26].

Fig. 2.

SEM pictures of pregelatinized canistel seed starch and canistel seed starch-quercetin complexes. (A)pre-CSS; (B)pre-CSS-Q; (C)pre-CSS-U-70; (D)pre-CSS-U-80; (E)pre-CSS-U-90; (F)pre-CSS-U-Q-70; (G)pre-CSS-U-Q-80; (H)pre-CSS-U-Q-90.

The natural canistel seed starch forms a bell or sphere with smooth surface. The pre-CSS-Q showed an irregular shape. Quercetin adhered to the surface of starch granules and mixed. The surface of starch granules became rough and partially gelatinized. These results indicated that quercetin complex can change the shape of starch granules, which was consistent with previous studies [14]. When the complex of starch and quercetin was treated by ultrasound, the damage of starch granules was more obvious with the increase of ultrasound temperature. Compared with canistel seed starch sonicated at 70 °C, ultrasonic modified canistel seed starch and quercetin combined at 70 °C has more integrity, which indicated that quercetin could protect the damage of starch granules by ultrasonic. When canistel seed starch and quercetin were compounded at 80 °C, starch granules became smoother and tighter. Quercetin can penetrate starch through starch channels, adhere to the surface of starch, and make starch granules smoother [27]. At the same time, starch in a high moisture environment at a certain temperature may cause changes in the surface structure of starch particles, making the surface of starch smoother [28]. When quercetin was added by ultrasound at 90 °C, the starch was destroyed completely, but a network surface structure was formed on the surface. It can be considered that quercetin leached polymers, especially amylose molecules, through hydrogen bonding in high temperature ultrasound, and the complex increased with the increase of amylose synthesis [29].

3.3. Recrystallization

The original starch and the pretreated starch (pre-CSS) had the same A-type crystal structure, and the main diffraction peaks were 15.1°, 17.0°, 18.0° and 23.0° (2θ). The results showed that the crystal type of canistel seed starch was not changed after direct pretreatment (Fig. 3A). To further study the crystalline structure and crystallinity of all simple, certain XRD patterns were observed.

Fig. 3.

The structure of pregelatinized canistel seed starch and canistel seed.

Starch gelatinization resulted in no obvious characteristic peaks of canistel seed in the treated samples, but the diffraction pattern of quercetin showed that it had crystal properties, with several obvious peaks at 10.68°, 13.06°, 16.80°, 17.76°, 21.68°, 25.12°, and 27.00°. After interacting with quercetin, there were still some crystal peaks of quercetin in the starch. Compared with the canistel seed starch, there were new diffraction peaks (10.68°, 12.42°, and 21.68°), but weaker than quercetin, indicating that these changes are migrating from quercetin to starch granules, but the intermolecular or intramolecular hydrogen bond changes are not enough to destroy the crystal structure [30]. The canistel seed starch-quercetin complex displayed a special peak at 27.38°of 2θ, and its peak intensity decreased with increasing temperature. Perhaps this peak was where quercetin and starch were bonded. The results showed that quercetin and starch molecules could interact. The canistel seed starch-quercetin complex treated with ultrasonic treatment at various temperatures did not have a peak of V-shaped crystal structure in the diffractogram, indicating that it was not a V-shaped complex. This finding is consistent with previous studies that no V-type complex formed between buckwheat starch and quercetin [31]. This may be because the inner diameter of starch amylose is smaller than quercetin, which limits quercetin to enter the starch double helix, thus quercetin is compounded with the side chain in the starch helical structure, or quercetin lacks hydrophobicity [32]. It can be considered that the molecular interaction between canistel seed starch and quercetin is mainly hydrogen bond type. The composite sample of starch and quercetin after ultrasonic treatment at 80 °C had the highest crystallinity (26%) (Table 2). It has been reported that the complexation between starch and quercetin through hydrophobic interaction was an endothermic process [33]. The higher temperature is conducive to the formation of the complex, but higher temperature will destroy the structure between starch chains and become an obstacle to the formation of the complex [34].

Table 2.

Short range order characteristics and crystallinity of pregelatinized canistel seed starch and canistel seed starch-quercetin complexes.

	1047/1022	1022/995	Relative crystallinity (%)
pre-CSS	1.35 ± 0.03b	0.80 ± 0.06d	18.81 ± 0.31f
pre-CSS-Q	1.29 ± 0.12bc	0.61 ± 0.09g	21.32 ± 0.32d
pre-CSS-U-70	1.21 ± 0.08c	0.74 ± 0.11e	20.20 ± 1.01e
pre-CSS-U-80	1.04 ± 0.06e	0.94 ± 0.10a	22.38 ± 0.67c
pre-CSS-U-90	1.09 ± 0.34e	0.90 ± 0.03b	23.55 ± 0.75b
pre-CSS-U-Q-70	1.34 ± 0.20b	0.71 ± 0.22f	23.12 ± 0.62bc
pre-CSS-U-Q-80	1.12 ± 0.04d	0.86 ± 0.02c	26.08 ± 0.17a
pre-CSS-U-Q-90	1.48 ± 0.11a	0.47 ± 0.09h	25.29 ± 0.31a

3.4. Conformational dynamics and IGM analysis

Constructing starch models from starch crystal structures and exploring the molecular interaction mechanism between starch and quercetin based on molecular dynamics simulations has become a mainstream technique complementary to experiments [35], [36], and the results are shown in Fig. 4, Fig. 5 and Fig. S1-2. Fig. S1 indicates the conformational change pattern of starch in 200 ns simulations, and all three groups of simulated systems reached equilibrium in 200 ns simulations, where The quercetin binding inside the starch had the most significant conformational change on the starch, with the average RMSD remaining at 2.0 nm, followed by the quercetin binding outside the starch with a slightly higher RMSD of about 1.3 nm than the starch alone group (RMSD of about 1.2 nm), indicating that the presence of quercetin exacerbated the conformational change of the starch (Fig. S1A). The changes in SASA of starch in the three simulated systems tended to be consistent, and the binding of quercetin inside the starch resulted in the lowest SASA of starch molecules and a significant decrease of its Rg compared with the blank group, which may indicate that quercetin bound inside the starch helix may induce the starch to refold or reheal during the movement to reduce the surface area in contact with water, and intensify the structural denseness of starch molecules (Fig. S1BC). Fig. 2 is a further observation of the effect of quercetin on the distribution of the dominant conformation of starch, where the starch molecules in the blank group existed mainly in three dominant conformations, with the A conformation showing almost complete unhelicalization (Fig. 4A). Compared with the blank group, quercetin binding inside the starch resulted in the starch molecules existing mainly in the double-stranded helical conformation (A conformation), followed by the refolded conformation (BCD conformation). (BCD conformation), suggesting that quercetin bound inside the helix of the starch molecule preserves the crystalline helical conformation of starch to some extent (Fig. 4B). In addition, quercetin located in solution spontaneously approaches the starch molecule and binds to the surface of the starch molecule with a high degree of helicity at its binding site, which is consistent with the experimental X-ray and IR analysis results. Moreover, this was confirmed by observing the distribution of φ(O5-C1-O4′-C4′) and ψ(C1-O4′-C4′–C3′), compared to the blank group (Fig. S2A), and the presence of quercetin inhibited the complete unhelixation of starch molecules (Fig. S2 B).

Fig. 4.

Effect of quercetin on the distribution of mainstream starch conformation. (A) a starch alone system (Starch), (B) a complex system with quercetin bound inside the starch helix (intra), and (C) a complex system with quercetin away from starch (inter).

Fig. 5.

Conformational IGM visualization in two simulated systems: quercetin-starch (intra) and quercetin-starch (inter).

Fig. 5 visualize the most dominant conformations in the two simulated systems of quercetin-starch (intra) and quercetin-starch (inter), respectively, and it is evident that a clear hydrogen bonding interaction (blue) is formed between the quercetin and starch molecules, which is consistent with the results analyzed by IR spectroscopy, and secondly, van der Waals interactions are also observed (green) and were involved in the binding of both. In terms of atomic contribution, the interaction between the hydroxyl group on the sugar chain and the oxygen atom on the quercetin A ring in the starch molecule in the quercetin-starch (intra) system can account for more than 25% of the total interactions (Fig. 5A). In the quercetin-starch (inter) system, in addition to the two pairs of molecular interactions mentioned above, the carbon atoms on the A ring of quercetin and the oxygen atoms on the B ring can form van der Waals interactions and hydrogen bonds with the hydroxyl groups on the sugar chain in the starch molecule, accounting for about 9.78% and 3.44%, respectively (Fig. 5B). Previously, [36] studied starch complexes wrapped with C10 aromatic molecules and found that hydrogen bonding could drive the binding of C10 aromatic molecules to starch [36]. In conclusion, we found that hydrogen bonding is the main driving force for the binding of starch molecules to quercetin, and van der Waals interactions also promote the binding of both.

3.5. Fourier transform infrared spectral

Infrared analysis can know more about the effect of quercetin on the structure of canistel starch. A strong band was shown at 3420 cm⁻¹ in Fig. 3B, which was attributed to the absorption the stretching vibration of the intermolecular or intramolecular hydroxyl. Compared with the pretreated starch, the tensile strength of ultrasonic treated starch at 3420 cm⁻¹ became wider and the peak intensity became larger, which indicated that ultrasonic treatment could increase the number of hydroxyl groups. The peak intensity ratio of 1045/1022 cm⁻¹ and 1022/995 cm⁻¹ in Table 2 reflected the degree of order and crystallinity of starch molecules. Under 80 °C ultrasonic treatment, the pretreatment had the minimum 1045/1022 cm⁻¹ ratio and the maximum 1022/995 cm⁻¹ ratio, indicating that the degree of order was the lowest, but there was the greatest degree of crystallinity.

The infrared spectrum had a broad absorption peak at 3420 cm⁻¹, indicating that the hydroxyl groups (OH) of starch and quercetin were stretched [37]. The characteristic carbonyl absorption band shifts from 1668 cm⁻¹ to 1665 cm⁻¹, indicating that quercetin may interact with starch molecules [15]. Compared with pre-CSS, the pre-CSS-Q and pre-CSS-U-Q did not form a new characteristic peak, but the absorption peak of the complex at 3420 cm⁻¹ became wider and stronger with the increase of temperature, indicating that it did not form a new covalent bond with quercetin, but the hydrogen bond was stronger [38]. Therefore, quercetin interacted with starch through non-covalent bonds during the gelatinization process. The pre-CSS had a lower ratio of 1045/1022 cm⁻¹ and a higher ratio of 1022/995 cm⁻¹ compared with the pre-CSS-Q, indicating that the addition of quercetin would reduce the order of starch and increase crystallinity.

3.6. Swelling power (SP) and solubility (SOL)

The solubility and swelling power of all samples gradually increased with increasing temperature (Table 3). Compared with pre-CSS-U, pre-CSS had the lowest solubility and swelling power at the same temperature. The increase of solubility can be attributed to the change in the interaction between starch and water [39]. Ultrasound can destroy the surface of starch granules (starch granule crushing and mechanical damage), making starch more disorderly, and promoting the interaction of water molecules with the free hydroxyl groups of amylose and amylopectin through hydrogen bonds, resulting in the increase of starch solubility and swelling power [40], [41]. The solubility and swelling power of all pre-CSS-U-Q were significantly increased. With the increase of ultrasonic temperature, the solubility and swelling power of the pre-CSS-U-Q increase, but it had the highest solubility and swelling power at 80 °C. This can be attributed to the higher disorder of the pre-CSS-U-Q-80, making the solubility and swelling power of the complex starch the most obvious increase, which was the same as the previous X-ray diffraction results.

Table 3.

Solubility and swelling power of pregelatinized canistel seed starch and canistel seed starch-quercetin complexes.

		55 °C	65 °C	75 °C	85 °C
SOL(g/g)	pre-CSS	2.16 ± 0.16d	3.09 ± 0.80f	8.54 ± 0.82bcd	9.64 ± 0.71c
	pre-CSS-Q	5.00 ± 0.04 cd	3.57 ± 0.85e	3.71 ± 0.25d	8.03 ± 0.49c
	pre-CSS-U-70	10.34 ± 0.72b	10.79 ± 0.62c	20.10 ± 8.01a	16.15 ± 0.50b
	pre-CSS-U-80	11.27 ± 0.02b	13.43 ± 0.06b	18.38 ± 0.26a	21.57 ± 0.40a
	pre-CSS-U-90	14.60 ± 0.36a	14.59 ± 0.35ab	16.65 ± 0.07a	23.72 ± 4.86a
	pre-CSS-U-Q-70	6.62 ± 0.06c	5.81 ± 0.08d	6.89 ± 0.34 cd	9.13 ± 0.01c
	pre-CSS-U-Q-80	15.79 ± 2.98a	15.71 ± 0.87a	15.20 ± 0.28ab	16.48 ± 0.53b
	pre-CSS-U-Q-90	13.75 ± 0.38a	14.32 ± 0.33b	13.78 ± 1.26abc	19.41 ± 1.29ab
SP(%)	pre-CSS	2.21 ± 0.17d	3.19 ± 0.85f	9.35 ± 0.98cde	10.67 ± 0.87c
	pre-CSS-Q	5.27 ± 0.05cd	3.70 ± 0.91f	3.85 ± 0.27e	8.73 ± 0.58c
	pre-CSS-U-70	11.54 ± 0.90b	12.10 ± 0.78d	25.78 ± 12.61a	19.27 ± 0.70b
	pre-CSS-U-80	12.71 ± 0.02b	15.51 ± 0.08c	22.52 ± 0.38a	27.50 ± 0.66a
	pre-CSS-U-90	17.10 ± 0.49a	17.09 ± 0.48b	19.98 ± 0.11ab	31.36 ± 8.37a
	pre-CSS-U-Q-70	7.09 ± 0.07b	6.17 ± 0.10e	7.40 ± 0.39de	10.05 ± 0.01c
	pre-CSS-U-Q-80	18.82 ± 4.21a	18.64 ± 1.23a	17.92 ± 0.38abc	19.73 ± 0.76b
	pre-CSS-U-Q-90	15.94 ± 0.51a	16.71 ± 0.45bc	16 ± 1.70abc	24.10 ± 1.99ab

3.7. Thermogravimetry analysis

Fig. 3C and D showed the thermogravimetric curves and thermogravimetric differential curves of all samples. It can be seen from the TGA curve that the starch sample had three successive quality loss steps. In the first stage, the mass loss reduction corresponded to 30-150 °C, which was caused by the evaporation loss of water molecules in the starch granules. In the second stage, the mass loss reduction stage corresponded to 150-378 °C, and the mass loss ranged from 62.04% to 75.54%. The main reason was that the pyrolysis and dissipation of starch leaded to an obvious linear decline; while the quality loss in the third stage was slower than that in the second stage, which may be due to the decomposition of the decomposed canistel seed starch and it took a certain time to reach a constant weight [42]. The maximum weight loss rates of pre-CSS and pre-CSS-Q were 317 °C and 310 °C, respectively, which indicated that the thermal stability of composite quercetin decreased. When the temperature reached 400 °C, the remaining rate of quercetin was about 69.12%, while the remaining rate of starch added quercetin was less than 40%, indicating that a complex was formed. DTG curve showed that the maximum weight loss rate of the pretreated canistel seed starch was higher than that of the complex, which indicated that the weight of the complex decreased at the same temperature and showed better stability [15]. Compared with pre-CSS-U-Q-70 and pre-CSS-U-Q-90, pre-CSS-U-Q-80 had smaller maximum weight loss rate, indicating that the stability of the composite was related to ultrasonic treatment at different temperatures, and it had the lowest stability at 80 °C.

3.8. Rheological properties

The flow curves of all samples exhibited that the shear stress increased with the increase of the shear rate, that is, the starch paste and the compound paste showed shear thinning properties (Fig. 6A and B). Compared with the direct pretreatment starch, the apparent viscosity increased after adding quercetin. This might be due to the formation of a strong and continuous network, resulting in an increase in the consistency of the starch-quercetin complex [43] (Fig. 6C and D). The decrease in the apparent viscosity of the canistel seed starch after ultrasonic treatment at different temperatures could be attributed to the rupture of amylose chains and the debranching of amylopectin molecules caused by the action of ultrasound in the C-O-C bond [44].

Fig. 6.

Static rheological curve of pregelatinized canistel seed starch and canistel seed starch-quercetin complexes.

Since all samples had shear stress when the shear rate was zero, the Herschel-Bulkley rheological model was used to fit the steady-state rheological characteristic curves of the samples. The fitting results were shown in the Table 4. The Herschel-Bulkley rheological model had a high degree of fit of R² (0.990–0.996), indicating that the model fitted well. The pre-CSS-U-Q were significantly higher than the directly pretreated canistel seed starch, indicating that the ultrasonic composite starch had stronger structural strength and formed a strong and continuous network, resulting in the increase of the composite consistency, and more shear required to initiate the flow of the starch paste. The flow behavior index (n) of the composite paste represented the degree of shear thinning behavior, and the range was 0.341 to 0.534. The starch n value of ultrasonic compound quercetin decreased, which was caused by the change of starch network during the shearing process. The ultrasonic composite starch at 80 °C also had the lowest yield stress and viscosity coefficient, indicating that the ultrasonic composite at 80 °C had the least effect on the structural strength of starch. Studies have shown that the high swelling force was believed to be related to the high viscosity and less pseudoplasticity of the starch paste [45], which was consistent with the previous swelling force and stable shear analysis results. Stable shear analysis indicated that starch had higher stability and shear resistance when compounded with canistel seed starch and quercetin.

Table 4.

The Herschel-Bulkley parameters of pregelatinized canistel seed starch and canistel seed starch-quercetin complexes.

	δ0 yield stress (Pa)	K consistency coefficient (Pa·sn)	n flow index	R2 coefficient of determination.
pre-CSS	0.440	0.185	0.735	0.990
pre-CSS-U-70	1.193	1.023	0.411	0.996
pre-CSS-U-80	0.266	0.31	0.527	0.994
pre-CSS-U-90	0.192	0.396	0.420	0.995
pre-CSS-U-Q-70	2.860	4.094	0.341	0.997
pre-CSS-U-Q-80	0.471	0.407	0.534	0.989
pre-CSS-U-Q-90	0.547	0.985	0.488	0.998

4. Conclusion

The present study showed that the ultrasonic modified canistel seed of starch combined with quercetin, which leads to the enlargement of starch granules. Canistel seed starch and quercetin compounded, quercetin adhered to the surface of starch granules and mixed, the surface of starch granules became smooth, and quercetin could protect the damage of starch granules by ultrasound. At the same time, the ultrasound-assisted starch complex did not form new covalent bonds and did not destroy the original crystalline type of starch, but the hydrogen bonds were stronger, indicating that the complex was compounded in the form of non-inclusion compounds. These structural changes lead to the increase of exothermic stability of starch complex, which has high stability and shear resistance. Overall, the results of these studies provide theoretical technical support and basic data for the processing and utilization of canistel seed starch.

CRediT authorship contribution statement

Rui He: Conceptualization, Methodology, Writing – original draft. Yong-gui Pan: Software, Data curation. Wen-Ting Shang: Visualization, Software. Geng Zhong: Investigation. Wu-Yang Huang: Supervision. Dong Xiang: Validation. Fei Pan: Software, Data curation. Wei-min Zhang: Writing – review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article: