Rheological, textural, and structural properties of heat-induced yam starch gels: Effects of high-amylose mung bean starch

Abstract

Yam starch (YS) has inherently weak gel strength because of its high amylopectin content, which critically restricts its advanced applications in the food industry. This study investigated the effect of incorporating high-amylose mung bean starch (MBS) (0–4%) as a natural modifier on YS gel properties. Results showed that MBS addition dose-dependently enhanced pasting temperature, viscosity, viscoelastic moduli, hardness, chewiness, and thermal stability of YS gels. Microstructurally, it led to a denser gel network with a smoother, more homogeneous nanoscale surface. X-ray diffraction (XRD) and Fourier transform infrared (FTIR) spectroscopy analyses indicated that MBS suppressed long-range crystallinity of YS gels but promoted short-range order and formed more junction zones within the amorphous network, resulting in an amorphously reinforced gel matrix. This reinforcement was ascribed to the competitive hydrogen bonding and cross-linking between MBS amylose and YS components. The work provides a green strategy for designing starch-based gels with tunable texture and improved stability for food applications.

Highlights

• •MBS addition enhanced the gelatinization temperature and viscosities of YS.
• •YS-MBS gels showed reinforced viscoelastic networks.
• •MBS addition dose-dependently improved hardness, cohesiveness and thermal stability of YS gels.
• •MBS reduced YS crystallinity by altered hydrogen bonding and molecular rearrangement.

1. Introduction

Starch, as a natural polysaccharide, is widely used in the food industry. The functional properties of starch in various products are influenced by the starch source, modification methods applied, and the amylose/amylopectin ratio (Navaf et al., 2024). During gelatinization, starch granules swell and leach amylose and amylopectin. Upon cooling, these chains reassociate through hydrogen bonding and hydrophobic interactions to form a three-dimensional viscoelastic gel network, which is a process referred to as retrogradation (Zheng & Zeng, 2025). The texture properties of starch gels directly affect their application properties. Single-network gels rely on reversible hydrogen bonds to form and maintain their network structure, which inherently leads to limitations such as low mechanical strength and poor thermal stability (Li, Wang, et al., 2024a). Although chemical crosslinking or esterification modification can improve performance, these methods face challenges such as complex processes and potential safety hazards (Wei, Wu, Ren, et al., 2024a). In contrast, starch blending modification has attracted much attention due to its green and safe nature. Regulating the synergistic effect of heterologous starch can optimize rheological properties, microstructure, and functional stability, offering a green approach for food texture design and shelf-life extension.

Starch-based gels, as critical texture modifiers and functional matrices in the food industry, have attracted sustained research interest in performance optimization (Wang et al., 2024). The diversity of starch sources leads to significant variations in crystalline types (A/B/C-type), molecular composition (amylose/amylopectin ratio), and pasting properties, which provides a theoretical basis for regulating gel functional characteristics through starch blending strategies (Xu & Kuang, 2024). Yam starch (YS), characterized by high amylopectin content, demonstrates exceptional viscosity and water retention capacity. However, its loosely branched structure results in low gel hardness and poor cohesiveness, limiting applications in structured food products (Chaniago et al., 2023). Mung bean starch (MBS), rich in amylose content, forms rigid networks through linear molecular chains and has been extensively employed to enhance gel-based food systems (Min, Yang, et al., 2025b). Therefore, it was chosen as a model high-amylose starch for this study due to its renowned gelling capacity, unique structural features, and relevance in fabricating structured foods.

Recent studies reveal that starch blending modification can improve the pasting characteristics and gel properties of the mixed gels (Wongkhueng et al., 2025). This approach leverages the inherent structural and functional complementarity between different starch sources to create composite systems with superior and tailored properties. The functional outcomes of such blends are predominantly governed by the specific molecular interactions between the constituent polymers, i.e., amylose and amylopectin (Yang et al., 2022). The branched amylopectin provides a viscous continuous phase, while the linear amylose reassociates upon cooling to reinforce the network; their synergy is key to gel texture and rheology (Wongphan et al., 2022). Although starch blending strategies based on the complementary effects of amylose and amylopectin have been studied, significant gaps remain in understanding the structure-forming mechanisms of such composite systems. The specific roles of crystalline reorganization versus amorphous network formation during the gel-strengthening process imparted by high-amylose starch are still inadequately elucidated, and fail to fully explain the complex performance variations.

Therefore, this study aims to use the YS-MBS composite system as a model to explore a phenomenon tentatively termed “amorphous reinforcement,” where gel strengthening occurs despite reduced long-range crystallinity. The contributions of the crystalline and amorphous phases will be quantitatively decoupled through the systematic correlation of intermolecular hydrogen bonding, crystalline structure, and microstructural network morphology with macroscopic rheological and textural properties. Texture profile analysis (TPA), rheometry, scanning electron microscopy (SEM), and rapid visco analysis (RVA) were used to evaluate gel hardness, springiness, adhesiveness, storage/loss moduli, microstructure, and pasting properties, respectively. This approach will provide a novel mechanistic framework and deliver actionable theoretical principles for the precise design of starch-based foods.

2. Materials and methods

2.1. Materials

Yam starch (with a purity ≥90%, amylose content of 18%, and a moisture content of 11%) was purchased from Baoji Liupanyun Biotechnology Co., Ltd. (Baoji, Shaanxi Province, China). Mung bean starch (with a purity ≥90%, amylose content of 40%, and a moisture content of 11%) was purchased from Hengshui Fuqiao Starch Co., Ltd. (Jingxian County, Hebei Province, China). All other reagents used were of analytical grade.

2.2. Preparation of YS-MBS composite gels

YS and MBS powders were initially sieved through a 100-mesh sieve. Based on preliminary experiments, a YS concentration of 6% w/w was selected as the fixed matrix concentration because it forms a self-supporting gel with adequate mechanical integrity for further modification, while still allowing clear observation of reinforcement effects upon MBS addition. For pure YS gel preparation, 6.0 g of YS powder was suspended in deionized water to prepare a 6% w/w YS suspension. The suspension was magnetically stirred at room temperature (22 ± 1 °C) for 20 min, followed by heating in a water bath at 95 °C for 30 min and subsequently cooled at 4 °C overnight. For YS-MBS composite gels, MBS suspensions with four concentrations (1%, 2%, 3%, 4% w/w) were prepared by dispersing MBS powder in deionized water. The YS suspension (6% w/w) was then blended with the MBS suspensions (0%, 1%, 2%, 3%, or 4% w/w) to formulate the composite systems containing YS and MBS. After 20 min of magnetic stirring at room temperature, the mixed suspensions were heated in a 95 °C water bath for 30 min and subsequently cooled at 4 °C overnight to obtain the composite gels.

2.3. Pasting test

The pasting characteristics of YS-MBS composite gels were detected according to the method of Ji et al. (2021) using a rapid viscosity analyzer (RVA, Perten Instruments, Warriewood, Australia). YS-MBS suspensions with different MBS concentrations were prepared, stirred at room temperature for 20 min to ensure uniformity, and then subjected to RVA analysis. RVA parameters were set as follows: pre-shearing at 50 °C for 1 min, heating to 95 °C at a rate of 12 °C/min, holding at 95 °C for 2.5 min, cooling to 50 °C at a rate of 12 °C/min, holding at 50 °C for 2 min to complete the analysis, and the constant shear rate was set at 160 rpm.

2.4. Rheological properties

The rheological properties of gels were measured according to the method reported by Li et al. (2020). Gel samples of 4.0 g were placed between the upper and lower plates (diameter 40 mm, gap 1 mm) of an AR2000 rheometer (TA Instruments, New Castle, DE, USA) for oscillatory frequency sweep analysis. Each gel sample had a diameter of 20 mm and a thickness of 15 mm. The frequency range was 0.6283–62.83 rad/s, the strain amplitude was 1.0%, and the storage modulus (G') and loss modulus (G") of the samples were determined.

2.5. Texture profile analysis (TPA)

The cylindrical gel samples (20 mm in diameter and 15 mm in height) were analyzed using a TA.XT Plus texture analyzer (Stable Micro Systems Ltd., Godalming, UK). The test parameters were as follows: pre-test speed 1.0 mm/s, test speed 1.0 mm/s, post-test speed 5.0 mm/s, compression distance 3.0 mm, dwell time 5 s, trigger force 10 g, and a two-cycle compression mode was adopted.

2.6. Thermogravimetric analysis (TGA)

The thermal properties of the gel samples were investigated using a thermal analyzer (TA Instruments Inc., New Castle, DE, USA). Gel samples were freeze-dried for 48 h under identical vacuum conditions. The consistency of residual moisture across samples was verified by comparing the initial weight loss (100–150 °C) in the TGA curves, which was similar for all formulations. Approximately 10.0 mg of the sample powder was placed in an open alumina crucible. The analysis was performed under a continuous nitrogen purge (flow rate: 50 mL/min), with the sample being heated from 30 °C to 600 °C at a constant rate of 10 °C/min. The mass change (TGA) signal was recorded throughout the heating process.

2.7. Scanning electron microscopy (SEM)

The microstructure of the gels was observed by scanning electron microscopy following the method described by Min, Wang, et al. (2025)a. First, freeze-dried gel samples were fractured to obtain flat cross-sections. The fractured samples were then mounted on sample stubs using conductive adhesive and sputter-coated with gold. A scanning electron microscope (Q45, FEI, Hillsboro, OR, USA) was then used to observe and image the samples at an accelerating voltage of 20 kV.

2.8. Atomic force microscopy (AFM)

AFM observation was conducted following the method described by Min et al. (2024). The gel samples were lyophilized for 48 h, ground into fine powder, and then passed through a 100-mesh sieve. The powdered sample was dissolved and diluted with deionized water to a concentration of 10 μg/mL. Subsequently, a 5 μL aliquot of the sample solution was deposited onto a freshly cleaved mica substrate (approximately 15 mm in diameter) and vacuum-dried for 3 h. AFM imaging was performed using a Dimension Icon atomic force microscope (Bruker Co., Bremen, Germany) in tapping mode. A silicon cantilever with a spring constant of 0.4 N/m and a nominal resonance frequency of approximately 50 kHz was used. The scan rate was set to 1.0 Hz, and the scanning area was 5 μm × 5 μm with a resolution of 512 × 512 pixels. All images were processed using NanoScope Analysis software.

2.9. X-ray diffractometry (XRD)

The freeze-dried samples were analyzed using an X-ray diffractometer (Empyrean, PANalytical, Almelo, The Netherlands). The operating parameters were as follows: voltage 40 kV, current 40 mA, scanning range (2θ) from 5° to 40°, and scanning speed of 5°/min. The resulting diffractograms were analyzed using JADE 6.0 Software to calculate the relative crystallinity of each sample.

2.10. Fourier transform infrared (FTIR) spectroscopy

Gel samples were freeze-dried at −80 °C for 48 h. The dried samples were ground into a fine powder and stored in a desiccator. The powder samples were then mixed with potassium bromide (KBr) powder, ground thoroughly, pressed into pellets, and analyzed using a FTIR spectrometer (Vertex 70, Bruker, Germany). The FTIR spectra were recorded over the range of 400–4000 cm⁻¹ with a resolution of 4 cm⁻¹.

2.11. Contact angles measurements

The static contact angle of the samples was measured using an OCA20 optical contact angle analyzer (DataPhysics, Filderstadt, Germany) employing the sessile drop method. The gel samples were first lyophilized and compressed into smooth pellets. A 3 μL droplet of deionized water was automatically deposited onto the surface of the pellet at a constant rate of 1 μL/s. The measurement was performed at 22 ± 1 °C and a relative humidity of 50 ± 5%. The contact angle was recorded immediately after droplet stabilization (within 10 s). To ensure reproducibility, three independent measurements were taken at different surface locations for each sample, and the average value was calculated.

2.12. Statistical analysis

All experiments were performed in triplicate unless otherwise stated, and the results were expressed as mean ± standard deviation (SD). The data were analyzed using SPSS 22.0 software (IBM Corporation, New York, USA) via one-way analysis of variance (ANOVA) followed by Duncan's multiple range test. A P-value <0.05 was considered statistically significant.

3. Results and discussion

3.1. Pasting properties

The formation of the YS-MBS composite gel system is based on the swelling and pasting of both MBS and YS, as well as the gelation process of amylose, which endows the composite gels with distinct pasting properties. The pasting characteristics of YS-MBS composite gels with varying MBS concentrations were systematically analyzed using RVA, with corresponding RVA profiles and viscosity parameters presented in Fig. 1A and Table 1. Compared with the control gel (6% YS), the composite gel containing 4% MBS exhibited a significant increase in pasting temperature from 62.2 °C to 74.3 °C, indicating improved thermal stability. The increase in pasting temperature is attributed to the competitive hydration between MBS and YS and the weaker water-binding capacity of amylose-rich MBS, which delays water penetration into starch granules (Waterschoot et al., 2015). Additionally, amylose leached from MBS may coat granule surfaces, further restricting water ingress and raising the gelatinization threshold (Tongbram et al., 2026).

Fig. 1.

The RVA profiles of yam starch (YS) gel and the YS-mungbean starch (MBS) composite gels.

Table 1.

Viscosity values of the yam starch (YS) gel and the YS-mungbean starch (MBS) composite gels. Different letters (a-e) in the same column indicate significant differences (P < 0.05).

Sample	Pasting temperature (°C)	Peak viscosity (cP)	Trough viscosity (cP)	Final viscosity (cP)	Breakdown (cP)	Setback (cP)
6% YS	62.2 ± 1.9b	1293 ± 56e	833 ± 15e	1282 ± 59e	460 ± 31d	449 ± 15e
+1% MBS	69.3 ± 2.5a	1701 ± 51d	886 ± 20d	1591 ± 68d	815 ± 21b	705 ± 26d
+2% MBS	70.6 ± 2.8a	1947 ± 62c	970 ± 28c	2065 ± 43c	977 ± 19a	1095 ± 18b
+3% MBS	72.8 ± 3.1a	2193 ± 69b	1618 ± 37b	2471 ± 91b	575 ± 17c	853 ± 30c
+4% MBS	74.3 ± 5.0a	2632 ± 75a	1849 ± 51a	3405 ± 96a	783 ± 20b	1556 ± 37a

The peak viscosity increased 2-fold (P < 0.05), reflecting improved water absorption capacity and granule swelling potential. This phenomenon arises from the formation of ordered double-helical structures by the abundant amylose in MBS during gelatinization and cooling, which aggregate into crystalline junction zones. These structures synergize with the viscous continuous phase formed by YS amylopectin to collectively build a robust gel matrix (Wang, 2024). The trough viscosity increased 2.2-fold (P < 0.05), suggesting reduced starch degradation under prolonged heating conditions. This could be attributed to MBS-based crystalline cross-links that stabilize the amylopectin matrix of YS, thereby mitigating chain depolymerization (Wang et al., 2006). Notably, the final viscosity demonstrated a 2.7-fold enhancement (P < 0.05), highlighting stronger retrogradation effects.

The breakdown and setback viscosities provide deeper insights into the thermal stability and retrogradation tendency of the starch gels, respectively. Breakdown viscosity showed a non-monotonic variation, increasing initially from 460 cP (6% YS) to a maximum of 977 cP at 2% MBS and then decreasing to 783 cP at 4% MBS. This non-monotonic change reflects a transition from YS-dominated to MBS-dominated network structure. At low MBS concentrations (≤2%), the system lacks a continuous reinforcing network, and the swollen granules are not effectively stabilized, making them susceptible to shear-induced disruption. As a result, breakdown viscosity is higher. When the MBS concentration reaches 3% or higher, the leached amylose chains form a sufficient three-dimensional network (reaching a percolation threshold). This network physically entraps and stabilizes the swollen granules, allowing the whole structure to collectively resist shear forces, thereby reducing breakdown viscosity (Umnajkitikorn et al., 2024). This interpretation is also supported by the rheological data (Fig. 2A, B), which show a marked increase in storage modulus G′ at MBS ≥3%, indicating the formation of a stronger elastic network.

Fig. 2.

The storage modulus (G′) (A), loss modulus (G′′) (B), TGA and DTG curves (C), and degradation temperature and weight loss rate (D) of yam starch (YS) gel and the YS-mungbean starch (MBS) composite gels. Different letters (a-c) indicate significant differences among the groups (P < 0.05).

Conversely, setback viscosity exhibited a consistent and significant increase with MBS addition, rising from 449 cP for the control gel to 1556 cP for the composite with 4% MBS (P < 0.05). This parameter reflects the extent of starch chain reassociation and the formation of crystalline structures (recrystallization) during cooling. This 3.5-fold enhancement provides direct rheological evidence for the accelerated and intensified retrogradation process driven by the linear amylose chains from MBS. Amylose, due to its linear structure, leaches out during gelatinization and rapidly realigns upon cooling to form a three-dimensional network, which is primarily responsible for the initial gel setting and the substantial increase in final viscosity and setback (Wang, 2024). This strong short-term retrogradation contributes to immediate gel setting and may be associated with improved freeze-thaw stability (Xie et al., 2022). The same robust amylose network could potentially also improve long-term stability, although this remains to be directly tested in future studies.

The pasting properties of starch blends are highly governed by the molecular features of the additive starch. As reported by Wongkhueng et al. (2025), highly cross-linked tapioca starch increased pasting temperature, final viscosity, and setback when mixed with rice flour, whereas acetylated starch hardly affected pasting temperature but significantly lowered final viscosity. This contrast indicates that pasting behavior changes are governed by the additive's specific interactions, not merely starch addition. In the YS-MBS system, pasting temperature, final viscosity, and setback viscosity increase. These effects align with network-promoting components (e.g., MBS amylose or HCL cross-links), not with hydrophilicity-altering modifications (e.g., acetylation). Thus, MBS acts by providing linear amylose that reassociates during cooling to form a reinforcing network, which is a different mechanism from most chemically modified starches.

3.2. Rheological properties

Storage modulus (G') and loss modulus (G") are critical parameters for evaluating the viscoelastic properties of starch gels. The storage modulus represents the elastic component of the gel, reflecting its ability to store energy under deformation, while the loss modulus corresponds to the viscous component, indicating energy dissipation during flow (Wang et al., 2024). Both G' and G" of YS-MBS composite gels increased progressively with rising frequency under dynamic oscillatory conditions, as shown in Fig. 2A and B. This trend demonstrates enhanced elastic and viscous responses at higher frequencies, as accelerated deformation rates reduce the time available for starch molecular relaxation (Mojarrad & Rafe, 2017). The intensified molecular interactions under high-frequency stress amplify interchain friction, thereby elevating overall viscoelasticity.

At equivalent frequencies, incremental additions of MBS systematically enhanced both moduli. Compared to the control gel (6% YS), the composite gel (6% YS + 4% MBS) exhibited a 2.2-fold increase in G' and a 2.7-fold increase in G". These pronounced enhancements suggest that MBS-derived amylose likely forms helical junctions that reinforce the branched amylopectin network of YS (Wang, 2024). The molecular synergy between MBS amylose and YS amylopectin strengthens elastic resilience and structural coherence. Specifically, MBS amylose may act as a crosslinking agent, generating a denser and more ordered network that restricts chain mobility and amplifies intermolecular friction, a mechanism analogous to filler-induced matrix constraint in polymer composites (Aberle et al., 1994).

The observed linear increase in G' with MBS concentration (1% to 4%) highlights additive interactions consistent with starch gelation theory, where amylose-driven network reinforcement dominates over amylopectin's viscous contributions (Gu et al., 2023). The elevated viscosity under identical shear rates correlates with amylose entanglement structures that enhance molecular friction (Gu et al., 2024). Notably, the high G' values make YS-MBS gels ideal candidates for structured foods requiring chewiness and elasticity, such as plant-based meat products (Baig et al., 2024). These rheological modifications critically influence gel fluidity and stability. In practical food processing, understanding such viscoelasticity modulation provides essential guidance for starch selection, process optimization, and quality control in product development.

3.3. Textural characteristics

Textural properties are critical for evaluating the mechanical behavior of food gels, including hardness, gumminess, adhesiveness, cohesiveness, and chewiness (Tanikawa & Watanabe, 2024). The introduction of MBS significantly altered the textural properties of YS-MBS composite gels, as shown in Table 2. Increasing MBS content from 0% to 4% not only reconfigured the compositional matrix but also markedly enhanced gel hardness from 118.5 g to 208.7 g (P < 0.05). During gelation, the high amylose content in MBS facilitated the formation of a denser network. The strong intermolecular interactions between amylose chains stabilize the internal structure and enhances deformation resistance, which is consistent with previous research on the relationship between amylose content and gel hardness (Ribeiro et al., 2024). This structural reinforcement was further evidenced in gumminess, which increased by 2.4-fold, indicating suppressed molecular chain slippage due to crosslinked networks.

Table 2.

The textual properties of the yam starch (YS) gel and the YS-mungbean starch (MBS) composite gels. Means with different letters (a-e) differ significantly (P < 0.05) among the data in the same column.

Sample	Hardness (g)	Gumminess (g)	Adhesivebess (g·s)	Cohesiveness	Chewiness (g)
6% YS	118.5 ± 9.3b	38.8 ± 2.6e	−3.43 ± 0.21e	0.16 ± 0.02e	1.94 ± 0.16e
+1% MBS	137.3 ± 11.6b	48.7 ± 2.9d	−4.12 ± 0.16d	0.56 ± 0.03d	2.92 ± 0.19d
+2% MBS	145.5 ± 14.3b	56.6 ± 3.2c	−7.35 ± 0.25c	0.64 ± 0.02c	3.96 ± 0.20c
+3% MBS	176.5 ± 14.1ab	80.8 ± 5.1b	−13.39 ± 0.33b	0.71 ± 0.03b	8.08 ± 0.23b
+4% MBS	208.7 ± 16.2a	94.9 ± 6.6a	−27.73 ± 0.52a	0.84 ± 0.05a	18.03 ± 0.46a

Notably, the absolute values of adhesiveness, which were inherently negative in texture analysis due to the measurement principle during probe retraction, increased 8.1-fold (P < 0.05). This phenomenon is attributed to the entanglement density induced by MBS: the topological interactions between long amylose chains and short amylopectin branches enhance the viscous resistance at gel-contact interfaces (Lu et al., 2025). Cohesiveness increased 5.3-fold (P < 0.05), which signifies a more homogeneous internal structure and aligns with the synergistic increase of G' and G" observed in rheological analysis (Fig. 2A and B). Furthermore, the chewiness of the composite gels exhibited a dose-dependent enhancement with the incorporation of MBS (Table 2). The chewiness value for the pure 6% YS gel was recorded at 1.94 g. Upon the addition of 4% MBS, this value increased significantly to 18.03 g, representing an 8.3-fold increase (P < 0.05). This profound augmentation aligns seamlessly with the observed trends in hardness and cohesiveness. It confirms that the reinforcement of the gel network is not merely superficial, but translates into significant improvements in bulk structural integrity and resistance to deformation during mastication (Min et al., 2022). The escalation in chewiness, particularly the sharp increase between 3% and 4% MBS, can be attributed to the synergistic reinforcement mechanisms discussed previously. The linear amylose chains from MBS permeate and crosslink the amylopectin-dominated network of YS, creating a denser and more continuous three-dimensional matrix. This inherently requires greater energy input for mechanical breakdown, directly manifesting as elevated chewiness.

The textural enhancement observed in the YS-MBS system aligns with the general principle that high-amylose starch reinforces gel networks, yet it also reveals system-specific characteristics. For instance, in binary composite gels of wheat flour and high-amylose corn starch (Hylon VII), Mojarrad and Rafe (2017) similarly reported increased gel firmness (hardness) with the addition of high-amylose starch, which was attributed to amylose retrogradation. However, a key divergence was noted. Hardness increased, but springiness decreased, and no significant changes in adhesiveness or cohesiveness were observed. This suggests a simple reinforcement without major alterations to the network's interactive or cohesive nature. In contrast, the YS-MBS composites show a more comprehensive enhancement. The hardness increased significantly, while cohesiveness and chewiness rose dose-dependently (5.3-fold and 8.3-fold, respectively) (P < 0.05). This indicates that MBS not only strengthens the YS gel but also improves its internal structural homogeneity and resistance to permanent fracture. The difference likely stems from the distinct base materials and interaction mechanisms. The linear amylose from MBS may integrate more effectively with the amylopectin-dominated YS matrix, forming a denser yet more elastic cross-linked network, rather than merely retrogressing as a separate reinforcing phase. This comparison underscores that the YS-MBS synergy achieves a superior texture profile. It simultaneously enhances strength, cohesiveness, and chewing resistance, which is particularly desirable for structured food applications such as plant-based meat analogs. In summary, MBS incorporation systematically modulates YS gel textural properties by governing starch molecular interactions and network architecture. This enables the targeted design of a food matrix structure.

3.4. Thermal properties

Thermal stability is a critical functional attribute of starch-based gels, determining their performance during thermal processing and long-term storage (Wang et al., 2025). The thermal degradation behavior of YS-MBS composite gels was characterized by TGA and derivative thermogravimetry (DTG), as shown in Fig. 2C and D. All samples exhibited an obvious thermal degradation profile, consistent with the typical pattern of starch-based biomaterials. During the initial stage (100–250 °C), the gel samples exhibited minimal weight loss, which is primarily attributed to the evaporation of bound water and the rupture of weak surface hydrogen bonds (Min et al., 2024). The gel samples showed a weight-loss between approximately 250 °C and 400 °C, corresponding to the thermal decomposition and carbonization of starch molecules (Cai et al., 2026). A clear trend was observed within this major decomposition region, and the residual weight at any given temperature increased with higher levels of MBS. For instance, at 400 °C, the residual weight followed the order: +4% MBS > +3% MBS > +2% MBS > +1% MBS > 6% YS, indicating enhanced thermal stability. The quantitative trends in degradation temperature and weight loss rate further clarify the impact of MBS. The degradation temperature increased from approximately 318 °C (0% MBS) to about 333 °C (4% MBS). The values plateaued within the 1%–3% MBS range before showing a significant rise at 4% MBS. This suggests that MBS incorporation progressively enhanced the thermal stability of the composite gels, with the most pronounced improvement observed for the 4% MBS formulation. The total weight loss rate decreased from about 74% (0% MBS) to approximately 67% (4% MBS), showing a consistent downward trend with increasing MBS concentration. The reduction in weight loss confirms that gels with MBS underwent less severe thermal degradation, as the densified network suppressed polymer chain scission (Wang et al., 2025).

The enhancement in thermal stability upon MBS incorporation can be attributed to synergistic molecular interactions between MBS amylose and YS amylopectin. The high-amylose MBS can form double-helical cross-links and intermolecular hydrogen bonds with YS amylopectin, constructing a denser, more cohesive three-dimensional network. This network restricts molecular mobility, requiring a higher temperature to be disrupted, which leads to increased degradation temperatures and reduced weight loss rates. This mechanism aligns with the concept of “amorphous reinforcement,” whereby a dense, short-range ordered network forms at the expense of long-range crystallinity. As subsequently demonstrated by XRD and FTIR results, this network yields a gel matrix with superior resistance to both mechanical stress and thermal degradation (Min et al., 2024). Consequently, the enhanced thermal properties of the YS-MBS composites correlate directly with their previously demonstrated superior pasting, rheological, and textural properties. The dose-dependent improvement in thermal stability suggests that YS-MBS composite gels are promising candidates for thermally processed foods, such as high-temperature sterilized sauces or plant-based meat products. Among the formulations, the +4% MBS recipe, with the highest degradation temperature and the lowest weight loss rate, exhibits optimal heat resistance, ensuring structural integrity during high-temperature processing and storage.

3.5. Microstructure

3.5.1. SEM

The microstructure of the gels was analyzed by SEM, as shown in Fig. 3. The control gel (6% YS) exhibited a relatively homogeneous, continuous network structure with no discernible starch granule residues, confirming complete pasting of YS. In contrast, the YS-MBS composite gels displayed progressive microstructural refinement with increasing MBS incorporation. As the MBS ratio rose, the pore size within the gel network gradually decreased, accompanied by more compact and intricately distributed architectures (Fig. 3). At higher MBS concentrations (3% or 4%), the network density increased substantially, demonstrating a direct correlation between structural compactness and enhanced textural properties (Mahajan et al., 2024). This structural evolution aligns with macroscopic textural characteristics. The densified network at 4% MBS gives significant improvements in hardness (208.7 g) and cohesiveness (0.84), corroborating the textural results and providing microstructural evidence for biomimetic starch-based foods (Guo et al., 2024). By controlling the addition of MBS (3% to 4%), a gel matrix with high elasticity and muscle-like fiber orientation can be directionally prepared to meet the demand for texture simulation in plant-based meat products (Dinali et al., 2024).

Fig. 3.

Scanning electron micrographs, image segmentation, and pore size statistics of the yam starch (YS) gel and the YS-mungbean starch (MBS) composite gels. 500 × .

3.5.2. AFM

The nanoscale surface topography and molecular arrangement of the gels were probed using AFM, providing direct evidence for the structural refinement induced by MBS. Representative two-dimensional (2D) and three-dimensional (3D) AFM images are presented in Fig. 4. The surface of the pure 6% YS gel exhibited considerable heterogeneity. This was evidenced by high color contrast in the 2D image and prominent peaks and valleys in the 3D rendering, indicating a relatively loose and uneven molecular packing. In contrast, the incorporation of MBS led to a progressive transformation of the surface morphology. With increasing MBS concentration (1% to 4%), the color contrast in the 2D images diminished, and the 3D topographies showed markedly reduced height fluctuations. The surface became notably smoother and more homogeneous, suggesting that MBS molecules effectively filled voids within the YS network and promoted a denser aggregation. This visual assessment was quantified by analyzing the molecular chain height profiles and surface roughness parameters (Fig. 4). The height profile of the pure YS gel showed the largest fluctuations, with peak heights reaching approximately 12 nm, indicative of loose and disordered chain arrangements. The addition of MBS progressively suppressed these fluctuations. For the composite with 4% MBS, the peak height was reduced to about 3 nm, demonstrating that MBS facilitated a more compact and regular alignment of starch molecular chains, thereby minimizing excessive chain protrusion (Tian et al., 2025). The trend toward structural compaction and homogenization was further confirmed by the significant reduction in surface roughness parameters. Both the root-mean-square roughness (Rq) and the arithmetic average roughness (Ra) decreased linearly with higher MBS content. Specifically, Rq declined from 2.07 nm (6% YS) to 1.3 nm (6% YS + 4% MBS), while Ra decreased from about 1.04 nm to 0.506 nm. This quantifiable smoothening at the nanoscale is a critical structural signature of a more integrated and continuous gel network (Min et al., 2024).

Fig. 4.

The atomic force micrographsm, molecular chain height profiles, and surface roughness parameters of the yam starch (YS) gel and the YS-mungbean starch (MBS) composite gels.

The AFM findings are consistent with and provide a nanoscale foundation for the macroscopic properties reported earlier. The formation of a denser, smoother network, as revealed by AFM, directly correlates with the enhanced gel strength (higher storage modulus G' and hardness), improved thermal stability (elevated decomposition temperature), and superior textural uniformity. This structural evolution can be attributed to the role of MBS-derived linear amylose in forming extensive short-range ordered junction zones (Tian et al., 2025). These junctions act as cross-links that tighten the overall matrix, suppress large-scale structural irregularities, and embody the “amorphous reinforcement” mechanism at the supramolecular level. Thus, AFM analysis conclusively demonstrates that MBS modulates the architecture of YS gels from the nanoscale upward, which is fundamental to tailoring their functional performance.

3.6. XRD analysis

XRD analysis revealed a dose-dependent reduction in the relative crystallinity of the composite starches with increasing MBS incorporation, as shown in Fig. 5A. The control group (6% YS) had a relative crystallinity of 18.61%, while the 4% MBS composite gel showed 12.25% (a 34% decrease). This indicates that MBS incorporation significantly inhibited YS recrystallization during gel formation in a dose-dependent manner. The reduction in overall crystallinity reveals a nuanced impact of MBS on the retrogradation of the composite system. While the linear amylose from MBS undergoes rapid recrystallization (contributing to short-term network formation) (Kou et al., 2022), its integration into the gel network appears to inhibit the long-range ordered recrystallization of the YS amylopectin fraction. We propose that the MBS amylose forms a dense, cross-linked network that physically restricts the mobility and reorganization of YS amylopectin chains, thereby suppressing their tendency to form large, stable crystalline domains over time. This mechanism aligns with the “amorphous reinforcement” concept, where gel strength may be enhanced primarily by an amorphous network even when long-range crystallinity does not increase (or even decreases). Hydrogen bonding, a critical force governing starch molecular alignment (Hao et al., 2025), explains this phenomenon. In pure YS, a tightly coordinated hydrogen bond network facilitates ordered packing of amylopectin into high-crystallinity domains. The introduction of MBS leads to a significant decrease in crystallinity, possibly due to the formation of new hydroxyl-mediated hydrogen bonds between the amylose of MBS and YS molecules, which disrupts the original crystal nucleation sites (Fan et al., 2019). On the other hand, during gel formation and cooling, amylose-rich MBS chains compete with amylopectin-dominated YS chains for molecular recombination. This competitive recombination hinders the orderly stacking of YS chains, thereby significantly reducing crystallinity (Lu et al., 2024).

Fig. 5.

X-ray diffraction patterns (A), FTIR spectra (B), and contact angles (C) of the yam starch (YS) gel and the YS-mungbean starch (MBS) composite gel with varying MBS content.

The XRD patterns of YS-MBS composite gels with different MBS contents exhibited distinct key peak positions (Fig. 5A). The main characteristic peak of YS appeared at a diffraction angle (2θ) of approximately 17°. This characteristic peak represented a significant feature in the YS crystal structure and was a manifestation of its ordered crystal structure (Zou et al., 2022). With increasing MBS addition, the peak intensity near 17° gradually decreased, indicating disruption of the B-type crystalline domains in YS. The enhanced scattering intensity at 5.6° was associated with amorphous regions or weakly ordered domains. This reflected an increase in amorphous content (Sanchez et al., 2024). This elevated amorphous fraction correlated with improved molecular chain flexibility and reversible deformation capacity, ultimately enhancing gel elasticity (Wei, Wu, Zhang, et al., 2024b).

The observed structural modulation, in which MBS incorporation reduced long-range crystallinity while enhancing gel properties, contrasted with scenarios where high-amylose starch acts primarily as an inactive filler. Apostolidis et al. (2021) reported that high-amylose maize starch can remain ungelatinized in binary systems with native starch under moderate heating (e.g., 95 °C), retaining its B-type crystalline structure and producing an intense XRD peak near 5°. In such cases, the high-amylose component contributed to overall crystallinity but might not integrate molecularly into the gel network, leading to distinct structure-property relationships. In the YS-MBS system, no intense residual granular peak was observed. Instead, the B-type peak of YS at ∼17° diminished progressively with increasing MBS, and the scattering intensity at ∼5.6° increased. These observations were consistent with substantial gelatinization of MBS under our conditions (95 °C, 30 min). The gelatinization temperature of MBS is approximately 70 °C (Puncha-arnon et al., 2008). However, high-amylose starches may not fully gelatinize at 95 °C without pressure assistance (Tian et al., 2022). Consequently, although the current XRD data suggest that a significant proportion of MBS experienced gelatinization, enabling molecular interaction with YS components, the precise degree of gelatinization was not directly measured in this research and thus remained a limitation. Notably, the observed dose-dependent decline in YS crystallinity does not rely on full gelatinization of MBS. Even partial gelatinization can lead to structural rearrangement of YS, and the gradual attenuation of the YS B-type peak is compatible with the proposed amorphous reinforcement effect.

Taken together, the consequent decrease in overall crystallinity underscores that the reinforcement stems not from the persistence of crystalline fillers, but from the formation of a new, molecularly integrated amorphous network. This comparison is consistent with the “amorphous reinforcement” scenario we describe, highlighting the role of sufficient gelatinization to enable molecular-level synergy. The results demonstrate that MBS effectively modulates YS recrystallization through molecular-level interactions, establishing a theoretical framework for engineering functional starch-based materials with tailored structural properties.

3.7. FTIR analysis

FTIR spectroscopy is an effective technique for analyzing the structural characteristics of macromolecular substances. By detecting stretching vibration variations in chemical bonds between functional groups, it provides infrared absorption spectra that enable comprehensive characterization of molecular architecture (Min et al., 2022). As shown in Fig. 5B, the FTIR spectra of YS-MBS composite gels exhibited absorption peaks largely consistent with those of the control gel (6% YS), indicating no formation of new functional groups upon MBS incorporation.

All gel samples displayed a prominent broad peak in the 3200–3600 cm⁻¹ region, characteristic of O—H stretching vibrations associated with hydrogen bonding and bound water molecules (Johnson et al., 2023). With increasing MBS, this peak shifted systematically from 3395 cm⁻¹ (6% YS) to 3464 cm⁻¹ (6% YS + 4% MBS) (Fig. 5B). The blue shift indicates weakening or reconfiguration of the hydrogen-bonding network, i.e., the O—H covalent bond shortens/strengthens due to weaker or longer hydrogen bonds (Shen et al., 2025). Thus, MBS disrupts the original ordered H-bond network of pure YS, creating a more heterogeneous environment. This reconfiguration (not strengthening) is consistent with the decreased long-range crystallinity (XRD) and increased short-range order (FTIR 1045/1022). Accordingly, the “amorphous reinforcement” arises from a redistributed network of short-range ordered junction zones held together by weaker but more numerous hydrogen bonds, rather than from stronger individual H-bonds. The enhanced absorption at 2920 cm⁻¹ likely reflects molecular chain ordering (e.g., double helices or microcrystal formation), a process typically associated with hydrophobic interaction-driven structural reorganization (Cai et al., 2021).

In the 1000–1150 cm⁻¹ region, absorption peaks corresponding to C–O–C and C–O–H stretching vibrations (glycosidic bonds in starch) exhibited positional shifts upon MBS addition. This indicates that MBS modifies the local environment or bonding characteristics of glycosidic linkages in YS, thereby influencing starch chain conformation (Hao et al., 2023). The intensity ratio of the bands at 1045 cm⁻¹ and 1022 cm⁻¹ is a sensitive indicator of the relative crystallinity and non-crystallinity of starch (Gu et al., 2023). As shown in Fig. 5B, the ratio value of 1045/1022 for the pure 6% YS gel was 1.013. Contrary to the conventional expectation that this ratio would decrease with a loss of order, it exhibited a consistent, slight increase with MBS addition, reaching 1.027 for the composite with 4% MBS. This increase suggests an enhancement in short-range order within the composite gels (Kou et al., 2022). We interpret this as evidence for the formation of numerous localized and ordered structures, such as double-helical junction zones or tightly bound cross-links, between the linear amylose of MBS and the amylopectin of YS. These junction zones are physical crosslinking points that contribute to the amorphous reinforcement. Importantly, this gain in short-range order occurs simultaneously with a significant loss in long-range crystalline order, as unambiguously demonstrated by the decrease in XRD relative crystallinity (Fig. 5A). The apparent dichotomy can be interpreted in light of what we term “amorphous reinforcement”. MBS amylose disrupts the long-range crystalline architecture of YS while constructing a new, pervasive network. This network is amorphous at the long-range scale but densely populated with short-range ordered cross-links at the molecular level. These ordered junctions act as robust reinforcement points within the amorphous continuum, ultimately responsible for the enhanced viscoelastic and textural properties (Asiamah et al., 2026). MBS exerts a dual impact on the retrogradation behavior of YS. It facilitates the short-term network formation of amylose while inhibiting the long-term recrystallization of amylopectin. These effects structurally correspond to the increase in short-range ordered junctions and the decrease in long-range crystalline order, respectively. This reconfiguration of molecular order is identified as the fundamental reason why the “amorphously reinforced” gel network we revealed possesses excellent immediate strength. These FTIR findings elucidate molecular-level structural modifications induced by MBS in YS gels, offering critical theoretical foundations for advancing starch hybrid systems in food science and material engineering applications.

3.8. Contact angles

Surface wettability, which influences critical properties such as water adhesion, coating potential, and component compatibility (Zhang et al., 2012), was assessed by static contact angle measurements. As summarized in Fig. 5C, the incorporation of MBS induced a systematic alteration in the surface characteristics of the composite gels. The pure YS gel exhibited a contact angle of 65.7°, indicating a relatively hydrophilic surface. In stark contrast, the addition of MBS drastically increased the surface hydrophilicity. The contact angle sharply decreased to 37.8° for the gel containing 1% MBS and continued to decline progressively with higher MBS loadings, reaching 27.5° for the composite gel with 4% MBS.

This monotonic decrease in contact angle with increasing MBS content indicates a significant enhancement in surface hydrophilicity. The phenomenon can be partly attributed to the higher amylose content of MBS, which introduces additional hydroxyl groups that increase surface polarity (Białopiotrowicz, 2003; Chaniago et al., 2023). However, given that both YS and MBS are starch-based polymers rich in hydroxyl groups, the dramatic reduction from 65.7° to 27.5° cannot be explained solely by changes in hydroxyl density. Surface topography also contributes to wettability. According to the Wenzel model, for an intrinsically hydrophilic surface (θ < 90°), an increase in roughness should decrease the apparent contact angle (more hydrophilic), whereas a decrease in roughness should increase the contact angle (less hydrophilic) (Drelich, 2019). The AFM results showed that MBS incorporation progressively reduces surface roughness (Fig. 4). For the intrinsically hydrophilic YS gel (contact angle 65.7°), a reduction in roughness would be expected to increase the contact angle according to the Wenzel model. However, the experimentally observed contact angle decreases significantly to 27.5°. This seemingly contradictory observation reveals that enhanced surface polarity, arising from greater exposure of hydroxyl groups in the dense short-range ordered network, dominates over the physical topographical effect. Thus, surface chemical composition prevailed over physical roughness as the primary determinant of the YS-MBS system wettability. In guar gum adsorption-modified systems (Pawliszak et al., 2024), combined AFM topography and contact angle analyses also demonstrated that chemical modification-induced changes in surface polarity often dominate wettability, which was consistent with our observations. Enhanced surface wettability may improve water-holding capacity, reduce syneresis, and benefit adhesion to hydrophilic substrates or starch-based film development (Zhang et al., 2012).

3.9. Schematic mechanism

Based on the aforementioned intermolecular interactions and the evolution of gel properties, a possible mechanism for the formation of YS-MBS composite gels is proposed hereof, as illustrated in Fig. 6. This model posits a structural evolution pathway encompassing competitive interaction at the molecular level, consequent reorganization of the supramolecular architecture, and eventual microstructural densification. The process initiates during the heating and pasting of the composite system. The granular structures of both YS and MBS are disrupted, releasing primarily the highly branched amylopectin from YS and the linear amylose from MBS as molecular chains into an aqueous, entangled state. This disentanglement creates the prerequisite for subsequent intermolecular interactions (Kong et al., 2024).

Fig. 6.

The schematic diagram of yam starch (YS) gel and the YS-mungbean starch (MBS) composite gels.

During the critical cooling stage, the system undergoes a competitive molecular rearrangement. The leached linear amylose chains from MBS compete with YS molecules (both amylose and amylopectin side chains) for available hydrogen bonding sites. This competition and the subsequent formation of new intermolecular bonds are directly substantiated by the FTIR results, which showed a progressive blue shift in the O—H stretching vibration, indicating a significant reconfiguration of the hydrogen-bonding network. The intervention of MBS amylose fundamentally redirects the system's self-assembly pathway. Rather than allowing the YS amylopectin to reorganize into its native, long-range B-type crystalline order, the MBS amylose acts as an effective cross-linking agent (Cheng et al., 2024). It preferentially forms numerous junction zones characterized by short-range order (as opposed to long-range crystalline order) with itself and with linear segments of YS components. This molecular rearrangement leads to a structural transition at the supramolecular level (Li, Zhang, et al., 2024b). The system evolves from a crystalline-dominated state to one characterized by a highly interconnected amorphous cross-linked network. This core observation, which we refer to as “amorphous reinforcement,” is consistent with the XRD data (dose-dependent decrease in crystallinity) and FTIR results (concurrent increase in short-range order).

The extensive cross-linking drives the final microstructural densification and network maturation. The nascent, coarse gel network assembles into a much denser, finer, and more homogeneous three-dimensional matrix. Direct evidence for this stage is provided by SEM micrographs, which show a progressive reduction in pore size and increased continuity with higher MBS content. Crucially, AFM provides nanoscale validation, quantifying a significant decrease in surface roughness (Rq, Ra) and molecular chain height, confirming the formation of a smoother, more compact gel surface as a result of this densification. This coherent multi-scale structural evolution is the direct origin of all observed macroscopic property enhancements. The densified, reinforced amorphous network translates into superior mechanical strength (increased hardness, chewiness from TPA) and viscoelasticity (elevated storage modulus G‘from rheology). The tight, cross-linked matrix enhances thermal stability, as evidenced by the elevated decomposition temperature and slower mass loss rate in TGA. The homogeneous and finely structured network with abundant polar junctions drastically improves surface hydrophilicity, explaining the sharp decrease in water contact angle upon MBS addition.

In particular, the term “amorphous reinforcement” describes the observation that in YS-MBS gels, increased MBS concentration reduces long-range crystallinity (XRD) but increases short-range order (FTIR) and improves gel strength. This contrasts with the conventional expectation that reinforcement requires long-range crystals. The underlying molecular interactions (hydrogen bonding, retrogradation) are not new. The term simply highlights a structure where short-range ordered junctions in an amorphous matrix, rather than crystalline lamellae, provide reinforcement, a scenario that may occur when branched amylopectin limits crystal growth. This is a complementary pathway, not a replacement of classical theory, and requires validation in other starch systems. The proposed mechanism coherently integrates all experimental observations. It depicts how MBS reconfigures the structural hierarchy of YS gels by molecular competition and cross-linking, suppressing long-range crystallinity while promoting a pervasive, short-range ordered amorphous network. This “amorphous reinforcement” perspective helps bridge molecular interactions with macroscopic functionality and may inform the design of starch-based composite materials with tailored properties.

4. Conclusions

This study describes an observation in YS-MBS composite gels that we tentatively term “amorphous reinforcement”, where gel strengthening occurs despite decreased long-range crystallinity, likely due to the formation of a network of short-range ordered junction zones. This structural reconfiguration simultaneously enhances hardness, cohesiveness, chewiness, viscoelasticity, and thermal stability. The multi-scale synergy originates from competitive hydrogen bonding and molecular rearrangement between MBS amylose and YS amylopectin, leading to a denser, smoother, and more homogeneous gel network. These findings expand the current understanding of how high-amylose starches modulate gel properties in binary systems, although the generality of this behavior requires further investigation. This work focuses on a model binary starch gel system (YS-MBS). The observed “amorphous reinforcement” offers a green strategy for texture modulation without chemical modification, but direct translation to complex food products (e.g., plant-based meat analogs) would require integration with proteins, extrusion processing, and sensory evaluation. Future studies should explore other high-amylose starches and real food matrices to validate the practical relevance of this mechanism. Within the defined scope of this model system, the study provides a useful basis for the rational design of functional starch materials for food, pharmaceutical, and biodegradable packaging applications.

CRediT authorship contribution statement

Cong Min: Writing – original draft, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Miao He: Formal analysis, Data curation. Rongxiang Wang: Resources, Conceptualization. Cheng He: Resources, Methodology, Conceptualization. Miaomiao Liu: Methodology, Formal analysis. Yanlong Li: Validation, Software. Li Feng: Software, Methodology. Yungang Cao: Writing – review & editing, Supervision.

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.

Data availability

Data will be made available on request.