Effect of tamarind seed polysaccharide on ice cream quality

Abstract

BACKGROUND

Ice recrystallization is a key factor limiting the storage stability of ice cream. Although tamarind seed polysaccharide (TSP) is recognized for its thickening and ice recrystallization inhibition (IRI) capabilities, its specific mechanism of action in frozen dairy systems remains underexplored. This study aimed to elucidate the stabilizing mechanism of TSP (0%, 0.2%, and 0.5% w/w) in ice cream formulations.

RESULTS

Rheological analysis revealed that TSP significantly modified the flow properties of the mixes. The consistency coefficient (K) increased from 0.4451 to 1.1593 Pa·sⁿ, and the flow behavior index (n) increased from 0.2876 to 0.6445, confirming the characteristic shear‐thinning behavior of the mixes. Oscillatory tests confirmed the formation of a weak gel‐like network (storage modulus (G') > loss modulus (G″)), imparting viscoelastic stability to the mix. At the microstructural level, TSP promoted the formation of a finer, more monodisperse emulsion and effectively inhibited Ostwald ripening, reducing the mean ice crystal area from 221.69 μm² to 135.78 μm². Furthermore, TSP retarded crystallization kinetics and reduced freezable water content, leading to a prolonged phase transition. These microstructural changes translated to enhanced cohesiveness (rising from 5.835 to 7.121) and substantially reduced hardness in the final product, although this was accompanied by an increase in the melting rate.

CONCLUSION

These findings demonstrate that TSP is a potent cryostabilizer that optimizes the microstructural integrity and textural quality of ice cream. This research provides a robust theoretical foundation for utilizing TSP as a functional ingredient in the frozen dessert industry. © 2026 Society of Chemical Industry.

INTRODUCTION

Ice cream is a widely consumed frozen dessert, prized for its smooth texture and unique structure. Based on hardness and production processes, it is generally classified into hard and soft ice cream. However, during processing, storage, transportation, and retail, ice cream inevitably experiences temperature fluctuations. These fluctuations induce ice recrystallization, a phenomenon that severely compromises both the texture and sensory quality of the product. Ice recrystallization is one of the primary factors leading to quality degradation in ice cream; therefore, effectively controlling and inhibiting ice crystal growth remains a significant challenge in the frozen food industry.^{1, 2, 3}

Current strategies to control ice recrystallization primarily focus on optimizing processing methods, storage conditions, and product formulations. Typically, hydrocolloids (stabilizers) are incorporated into the mix for their cryoprotective effects. These stabilizers hinder the recrystallization phenomenon by reducing the recrystallization rate and delaying ice crystal growth, thereby controlling ice crystal size. Furthermore, they enhance the viscosity of the ice cream mix, improve mouthfeel, and maintain the structural integrity of the final product. ⁴ The standard application concentration of these stabilizers in ice cream formulations generally ranges from 0.2% to 0.5%. ⁵

Most stabilizers are water‐soluble polysaccharides that form highly viscous solutions and contribute to the structural backbone of ice cream by interacting with proteins and salts. ⁶ Recently, it has been reported that unlike traditional polysaccharides (such as guar gum or locust bean gum), which primarily control ice crystal growth passively by increasing macroscopic viscosity, tamarind seed polysaccharide (TSP) exhibits superior ice recrystallization inhibition (IRI) capabilities. TSP's unique molecular structure and abundant hydroxyl groups form specific hydrogen bonds with water, constructing a restrictive three‐dimensional network. Furthermore, distinct from conventional gums, TSP can directly adsorb onto ice crystal surfaces to create steric hindrance and actively perturb the hydrogen‐bonded network of water during the nucleation process.^{7, 8, 9, 10}

Despite its broad application potential as a superior ice crystal stabilizer, the specific performance and efficacy of TSP in complex ice cream systems remain largely unexplored. Current research primarily focuses on macroscopic rheological properties and sensory evaluations, leaving a significant knowledge gap regarding the molecular‐level interactions between TSP and ice crystals, as well as the underlying mechanisms affecting ice cream quality.

To address this gap, this study systematically investigated the impact of TSP, utilized as a stabilizer, on the physicochemical properties of ice cream. The primary focus is to elucidate the interaction mechanisms between TSP and ice crystals, alongside its specific mechanism for inhibiting ice crystal growth. By evaluating rheological properties, ice crystal morphology, thermodynamic characteristics, and microstructural changes, this research intended to reveal the molecular mechanisms by which TSP regulates ice cream quality. Ultimately, this work will provide a robust theoretical foundation and technical support for the application of TSP in ice cream and other frozen food products.

MATERIALS AND METHODS

Experimental materials

TSP and monoglycerides were obtained from Yunnan Maoduoli Group Food Co., Ltd (Yuxi, China). Cold‐pressed coconut oil (Nanshi Yexiang, Hainan Hengcheng Xuguang Food Co., Ltd, Wenchang, China), sucrose (Yutang, Shanghai Bright Food (Group) Co., Ltd, Shanghai, China), and corn syrup (Zhanyi, Jiaxing Chenghao Jianong Food Co., Ltd, Jiaxing, China) were purchased via the online retail platform Taobao. Whole milk powder (Beiwei Muchang) was provided by Inner Mongolia Mengniu Dairy (Group) Co., Ltd (Hohhot, China). Phosphate‐buffered saline (PBS, 0.01 mol L⁻¹) and dimethyl sulfoxide (DMSO, AR, ≥ 99%) were purchased from Shanghai Yuanye Bio‐Technology Co., Ltd (Shanghai, China). Nile Red and fluorescein isothiocyanate (FITC) were acquired from Shanghai Aladdin Bio‐Chem Technology Co., Ltd (Shanghai, China). Deionized water was used throughout all experiments.

Preparation of ice cream

Three distinct ice cream formulations with varying TSP concentrations were prepared. The specific compositions for each TSP‐stabilized ice cream are summarized in Table 1.

Table 1.

Formulations of the ice cream mixes

Constituent (%)	0% TSP ice cream	0.2% TSP ice cream	0.5% TSP ice cream
Whole milk powder	12	12	12
Sucrose	12	12	12
Corn syrup	4	4	4
Coconut oil, cold‐pressed	4	4	4
Glycerol monostearate	0.2	0.2	0.2
Tamarind seed polysaccharide (TSP)	0	0.2	0.5
Water	67.8	67.6	67.3

Based on the experimental formulation, dry ingredients (stabilizers, monoglycerides, sucrose, and TSP) were weighed (PL2002; Mettler Toledo, Greifensee, Switzerland) and dry‐blended to ensure uniform dispersion. Syrup and water were added under constant agitation at 65 °C until dissolved, followed by the addition of coconut oil to form a homogeneous blend. The mix was then pasteurized at 70 °C for 10 min, ¹¹ cooled to ambient temperature, and dynamically frozen (SJ18EJ; Guangzhou Guangshen Electric Appliance Manufacturing Co., Ltd, Guangzhou, China). Finally, the semi‐frozen product was molded, hardened at −60 °C, and stored under these conditions prior to physicochemical analysis.

Rheological properties of ice cream mix

Steady‐state shear rheology

The apparent viscosity of the ice cream mixes was determined by adapting the method of Roy et al. ¹² with slight modifications. Rheological measurements were performed using a rheometer (Discovery HR‐3; TA Instruments Inc., Dallas, TX, USA) equipped with a 40 mm parallel plate geometry at a gap of 1000 μm. Samples were loaded onto the testing platform and allowed to equilibrate for 180 s at 0 °C. The apparent viscosity was then recorded over a shear rate range of 1 to 100 s⁻¹.

Amplitude sweep

Amplitude sweep was conducted according to the method of Xu et al. ¹³ with minor modifications. The measurements were performed at 0 °C with an equilibration period (soak time) of 30 s. The strain was varied from 0.01% to 100% at a fixed frequency of 1.0 Hz to determine the dynamic rheological properties of the samples.

Dynamic oscillatory shear rheology

The frequency sweep test was performed according to the method described by Roy et al. ¹² with slight modifications. The temperature was set to 0 °C with an equilibration time of 30 s. The storage modulus (G') and loss modulus (G") were measured within the linear viscoelastic region (LVR) over a frequency range of 0.1 to 10 Hz at a constant strain of 0.1%.

Oscillatory temperature ramp (OTR)

Oscillatory temperature ramp (OTR) was performed, adapting the protocol of Freire et al. ¹⁴ with minor adjustments. A constant strain of 0.1% was applied within the temperature range of 0 to 25 °C. The parallel plates were pre‐set to 0 °C. The samples were subjected to continuous oscillation at an angular frequency of 10 rad s⁻¹. Data for G', G", and the loss tangent (tanδ) were recorded every 4 s at a heating rate of 0.5 °C min⁻¹. To prevent moisture evaporation, a solvent trap was placed over the measuring geometry. The samples were first cooled from 25 °C to 0 °C, held at 0 °C for 10 min for thermal equilibration, and subsequently heated from 0 °C to 25 °C. Data analysis was performed using TRIOS software (TA Instruments).

Particle size analysis

The prepared ice cream mixes containing 0%, 0.2%, and 0.5% (w/w) TSP were diluted by factors of 100, 200, 300, 400, and 500, respectively. The average particle size (mean hydrodynamic diameter) of the diluted samples was determined using a Zetasizer Nano‐ZS90 (Malvern Instruments, Malvern, UK) at a detection temperature of 25 °C.

Ice crystal simulation

To simulate the growth of ice crystals during temperature fluctuations, sugar solutions with corresponding TSP concentrations were prepared (excluding whole milk powder, coconut oil, and monoglycerides from the base ice cream formulation). The microstructure of the samples was observed using an Olympus BX51 microscope (Olympus, Tokyo, Japan) equipped with a Linkam BCS196 cryo‐stage (Linkam Scientific Instruments, Redhill, UK). A 1.5 μL sample, diluted 50‐fold with PBS buffer, was placed between two circular glass coverslips and pressed gently to ensure uniform spreading. Subsequently, the temperature was first decreased to −20 °C at a rate of 30 °C min⁻¹ and held for 3 min, then increased to −8 °C at a rate of 10 °C min⁻¹. and held for 40 min. ¹⁵ Once the temperature reached −8 °C, ice crystal images were captured every 10 min at 500× magnification using a color CCD camera (Q Imaging Retiga, Hangzhou Hikvision Digital Technology Co., Ltd, Hangzhou, China). The resulting images were analyzed using Image J software.

Texture properties of ice cream mixes

Textural attributes of the ice cream mixes were assessed using a TA‐XT Plus Texture Analyzer (Stable Micro Systems Ltd., Godalming, UK) equipped with a 10 kg load cell. Back extrusion tests were conducted at ambient temperature employing an A/BE probe assembly (35 mm diameter disc). Operational parameters were defined as follows: pre‐test speed, 1.0 mm s⁻¹; test speed, 2.0 mm s⁻¹; and post‐test speed, 10.0 mm s⁻¹. A penetration depth of 10 mm and a trigger force of 5 g were applied. Data acquisition and analysis were performed using the instrument's proprietary software to extract parameters from the force–time curves, specifically cohesiveness and index of viscosity.

Multiple fluorescence staining and microstructure analysis

Stock solutions of Nile Red and FITC were prepared at a concentration of 1 mg mL⁻¹ by dissolving the fluorophores in DMSO and PBS, respectively. The two dye solutions were then thoroughly mixed to obtain a composite fluorescent staining solution for subsequent microscopic analysis. Subsequently, 20 μL of the staining solution was added to 1 mL of the ice cream emulsion and incubated in the dark at ambient temperature for 1 h to facilitate dye penetration and labeling. Microstructural analysis was performed using a Leica DFC7000T fluorescence microscope (Leica Microsystems GmbH, Wetzlar, Germany).

Quality analysis of ice cream

Hardness

The hardness of the ice cream was determined at room temperature using a texture analyzer equipped with a 2 mm diameter stainless steel cylindrical probe. The probe was programmed to penetrate the sample to a depth of 5 mm. The pre‐test, test, and post‐test speeds were set at 1.0, 1.0, and 3.0 mm s⁻¹, respectively, with a trigger force of 5 g. All experimental data were recorded and processed for further analysis.

Melting rate

The melting rate of the ice cream was determined by placing a sample (20.0 ± 1.0 g) on an eight‐mesh sieve, with a pre‐weighed beaker positioned underneath to collect the melted portion. The test was conducted in a constant temperature incubator (TH2‐D; Suzhou Peiying Experimental Equipment Co., Ltd, Suzhou, China). The measurements were performed at three different temperature and time combinations: 25 °C for 45 min, 30 °C for 35 min, and 35 °C for 25 min. The initial weight of the sample (m), the weight of the empty beaker (m ₁), and the weight of the beaker containing the melted ice cream (m ₂) were recorded. The melting rate was calculated using Eqn (1):

$$\text{Melting rate}\left(\%\right)=\frac{m_2-m_1}{m}\times 100\left(1\right)$$

Freezing curve

The freezing profiles of the ice cream mixes with different TSP concentrations were determined using a digital thermometer (YZ‐200E; China Yangzi Group Co., Ltd, Chuzhou, Anhui, China). The temperature probe was inserted into the geometric center of each sample. Temperature data were recorded at 10‐min intervals for a total duration of 6 h. After the freezing process was complete, the core temperature data were retrieved to plot the freezing curves.

Statistical analysis

All experiments were performed at least in triplicate. Data processing and graphical illustrations were conducted using Microsoft Excel and Origin software. Statistical significance was evaluated by one‐way analysis of variance (ANOVA) using SPSS (Chicago, IL, USA). Results were expressed as means ± standard deviation, and differences were considered statistically significant at P < 0.05.

RESULTS AND DISCUSSION

Rheological properties

Steady‐state shear rheology

The shear‐dependent flow behavior of ice cream mixes containing varying concentrations of TSP (0%, 0.2%, and 0.5%, w/w) is shown in Fig. 1(a). All samples exhibited a pronounced decrease in apparent viscosity with increasing shear rate, followed by a gradual plateauing effect. This shear‐thinning behavior is characteristic of pseudoplastic (non‐Newtonian) fluids. ¹⁶ Such behavior is attributed to the progressive structural breakdown of the system under shear stress and the subsequent molecular rearrangement. ¹⁷

Figure 1.

Effect of tamarind seed polysaccharide (TSP) concentration on the rheological properties of ice cream mixes: (a) apparent viscosity versus shear rate; (b) storage modulus (G') and loss modulus (G") versus shear stress; (c) moduli (G', G") versus angular frequency. The temperature dependence of the viscoelastic properties is shown as moduli versus temperature during cooling (d) and heating (f), and loss tangent (tanδ) versus temperature during cooling (e) and heating (g).

Furthermore, the apparent viscosity followed an upward trend as the TSP concentration increased from 0% to 0.5% (w/w), indicating that TSP incorporation enhanced the structural integrity of the colloidal network. This likely stems from increased molecular density, which facilitates intermolecular entanglement and macromolecular chain associations at higher concentrations. ¹⁶ In the low‐shear regime, the system maintains a highly associated state, resulting in elevated viscosity. However, as the shear rate increases, the TSP macromolecular chains undergo hydrodynamic orientation along the flow field. This alignment reduces steric hindrance and frictional resistance between polymer chains, thereby manifesting as the decline in the viscosity curve. ¹⁸ From a sensory perspective, this shear‐thinning property is desirable, as it contributes to a smoother and more refined mouthfeel in the final product.

The apparent viscosity data of the ice cream mixes were fitted to the Power Law model (τ = Kγ ⁿ ), where the consistency coefficient (K) is related to the fluid viscosity, with a higher K value indicating greater sample viscosity, and n represents the flow behavior index. The resulting parameters are summarized in Table 2.

Table 2.

Fitting parameters of the Power Law model for the steady‐state flow curves of ice cream mixes and ice crystal growth characteristics in sugar solutions with different tamarind seed polysaccharide (TSP) concentrations

Parameter	TSP concentration
	0%	0.2%	0.5%
The following data were measured on the ice cream mixes
K (Pa·sn)	0.4451 ± 0.0987b	0.4568 ± 0.0568b	1.1593 ± 0.0410a
n	0.2876 ± 0.0827c	0.4801 ± 0.0500b	0.6445 ± 0.0054a
R 2	0.9474 ± 0.0358b	0.9600 ± 0.0236ab	0.9691 ± 0.0317a
The following data were measured on the sucrose solutions
Area (μm2)	221.69 ± 54.79a	180.71 ± 73.22b	135.78 ± 66.71c
Maximum Feret (μm)	18.22 ± 2.36a	16.06 ± 3.60b	15.86 ± 3.59c
Aspect ratio	1.15 ± 0.11a	1.17 ± 0.16a	1.15 ± 0.14a

Note: Different lowercase letters indicate a significant difference (P < 0.05) among ice cream mixes and sugar solutions with varying TSP contents. K, consistency coefficient; n, flow behavior index; R ², coefficient of determination.

As the concentration increased, the K value of the ice cream mixes gradually increased, while n showed a rising trend. This indicates that higher TSP levels lead to greater viscosity and distinct pseudoplasticity. Meanwhile, the significant increase in the n value from 0.2876 to 0.6445 suggested that the shear‐thinning behavior of the mixes tends to be more moderate, reflecting a weakening of the pseudoplastic character. The marked elevation in K values directly confirmed that TSP, acting as a hydrocolloid, effectively constructs and reinforces the three‐dimensional network structure of the mix through molecular chain entanglement and hydration. Furthermore, the increase in the n value revealed the structural characteristics of this network. ¹⁹ The network formed by higher TSP concentrations possesses greater elasticity and resistance to shear‐induced disruption. Consequently, the relative decrease in viscosity within the shear field was smaller, manifesting as a milder shear‐thinning behavior.

Amplitude sweep

The G' and G" characterize sample viscoelasticity, reflecting elastic energy storage and viscous dissipation, respectively. ²⁰ A predominantly elastic, gel‐like state is indicated when G' > G″, whereas G" > G' reflects viscous, fluid‐like behavior.

The strain sweep profiles of the ice cream mixes are illustrated in Fig. 1(b). With strain increased, both G' and G" exhibited an initial upward trend followed by a decline; notably, G" consistently remaining exceeded G', classifying these mixes into the strong strain overshoot category. ²¹ All samples exhibited a constant modulus in the low‐strain region (< 1%), defined as the LVR (Fig. 1(b)). Beyond this region, both G' and G" decreased sharply with increasing strain, demonstrating typical strain‐thinning behavior. The LVR values for all samples ranged from 0.01% to 1%. Within the LVR, G' was higher than G", suggesting that the mix system constitutes a weak gel network dominated by elasticity. As the strain continued to increase, the values of G' and G" decreased until a crossover point (G' = G”) occurred, which can be interpreted as the transition from elastic to viscous behavior. ²²

Oscillatory shear rheology

In the low‐frequency region, G" > G', indicating a predominantly viscous response. At approximately 1 rad s⁻¹, a crossover point (G' = G″) marked a viscoelastic transition from liquid‐like behavior to an elasticity‐dominated weak gel state (Fig. 1(c)). ²³ Beyond this critical frequency, both moduli exhibited a frequency‐dependent increase, with G' consistently exceeding G". This elastic dominance is a hallmark of typical weak gel systems. TSP incorporation significantly elevated both moduli, as this functional hydrocolloid forms a three‐dimensional network through chain entanglements and hydrogen bonding, augmenting system stability. Higher concentrations promoted intermolecular associations, increasing junction zones and structural complexity. Conversely, lower concentrations favored intramolecular interactions, contributing minimally to the intermolecular network.^{24, 25} Furthermore, the widening divergence between G' and G" at elevated concentrations underscored a more robust intermolecular network, imparting a more pronounced solid‐like structural character to the mixes.

Oscillatory temperature ramp (OTR) properties

The cooling curve (25 to 0 °C) is illustrated in Fig. 1(d), as the temperature decreased, G" exhibited an initial increase followed by a gradual decline, which was accompanied by a continuous rise in viscosity. The initial enhancement in G' reflected the strengthening of intermolecular interactions and the elevation of baseline viscosity, eventually reaching a maximum that corresponded to the development of an elastic network. However, further cooling appeared to impair the established network structure.

During heating (0 to 25 °C, Fig. 1(f)), G' initially increased before declining across all formulations. This suggests that thermal energy facilitates polysaccharide segment mobility during early heating, rapidly reorganizing and reinforcing the three‐dimensional network. ²⁶ Throughout the heating and cooling cycles, G' remained consistently higher than G", indicating that the system predominantly exhibited solid‐like behavior. As the TSP concentration increased from 0% to 0.5%, both G' and G" increased, confirming the reinforcing effect of TSP on the network structure. The tanδ, defined as the ratio of G" to G', characterizing viscoelasticity, decreased with falling temperatures and higher TSP content (Fig. 1(e),(g)), indicating diminished fluidity; its value remained < 1, denoting elastic dominance. ²⁷

These observations demonstrated that TSP forms a temperature‐dependent three‐dimensional network, stabilized by molecular entanglements and intermolecular interactions. At lower temperature, the attenuation of molecular thermal motion reinforces the network structure, resulting in a progressive increase in moduli. The peak modulus likely represents the optimal stability of the polysaccharide network. Beyond this critical temperature, increased thermal energy disrupts the cohesive forces maintaining the network integrity, leading to a subsequent decline in both G' and G". ²⁸ Furthermore, the continuous increase in tanδ during heating signified a transition from elasticity‐dominated to viscosity‐dominated behavior, whereas the cooling process exhibited a reversible transition toward elastic dominance.

Particle size

Ice cream mix particle size critically influences final textural quality. The effect of varying TSP concentrations on the particle size distributions of the mixes is illustrated in Fig. 2(a). A distinct inverse relationship was evident at constant dilution factors, where particle size declined with increasing TSP content. At lower dilutions, elevated system concentration and turbidity likely caused multiple light scattering interference, leading to data instability. However, as optical interference was attenuated at higher dilution levels, the impact of TSP became increasingly evident. Specifically, increasing TSP content reduced the average particle size from 405.78 nm to 316.36 nm at a 300‐fold dilution and from 374.74 nm to 270.20 nm at a 400‐fold dilution, with the reduction being most significant at a 500‐fold dilution. ²⁹

Figure 2.

Combined particle size distribution and freezing curve of the ice cream mix. (a) Particle size diagram of ice cream mixes with different concentrations of tamarind seed polysaccharide (TSP); (b) freezing curves of ice cream mixes with different TSP concentrations. Different lowercase letters indicate significant differences among ice cream mixes with different TSP contents (P < 0.05).

These results substantiate that TSP incorporation significantly reduces particle size in the ice cream mix, primarily driven by its stabilizing properties. As a high‐molecular‐weight polysaccharide, TSP adsorbs onto the oil–water interface, forming a protective steric barrier around fat globules. This barrier effectively inhibited droplet flocculation and coalescence, ensuring a narrower and more uniform particle size distribution. ³⁰ Drawing upon Stokes' Law, the sedimentation or creaming velocity of dispersed droplets is proportional to the square of their diameter. Therefore, reducing particle size is pivotal for enhancing the long‐term stability of the emulsion. A finer particle size not only mitigates gravitational separation by lowering the creaming rate but also indicates a robust emulsified state with high interfacial stability. Ultimately, this homogeneous microstructure serves as the structural basis for the smooth and consistent macroscopic texture observed in the final ice cream product.^{31, 32}

Freezing curve

The effect of varying TSP concentrations on the freezing curves of the ice cream formulations is illustrated in Fig. 2(b). The thermal history is delineated into three characteristic regimes: pre‐cooling, phase transition, and sub‐cooling. ³³ A defining feature of the phase transition regime is the thermal recoil (recalescence), driven by the exothermal release of latent heat during crystallization. This stage typically encompasses the zone of maximum ice crystal formation (−5 to 0 °C), where more than 80% of the water fraction freezes. ³⁴

Stage I (pre‐cooling)

All samples exhibited rapid and overlapping temperature drops. This implies TSP minimally alters specific heat capacity and thermal conductivity. Convective and conductive transfer dominate heat removal. ³⁵

Stage II (phase transition plateau)

This regime is critical for structure formation. The control sample entered the plateau at about 50 min, whereas the 0.2% and 0.5% (w/w) TSP samples showed a significant thermal lag, delaying the onset to about 65 and 80 min, respectively. This delay evidences the ability of TSP to inhibit nucleation kinetics. Moreover, the control curve displayed a steeper slope, reflecting rapid crystal growth, while TSP inclusion resulted in a broader and flatter plateau. This is attributed to the hydrocolloid nature of TSP: its hydrophilic network entraps free water via hydrogen bonds, increasing the energy barrier for heterogeneous nucleation. Consequently, water molecules are immobilized, leading to a controlled, uniform crystallization process under a lower driving force. ³⁶

Stage III (sub‐cooling)

The elevated final temperatures observed in TSP‐fortified samples suggest a reduction in the fraction of freezable water. By sequestering water molecules within its macromolecular structure, TSP converts free water into bound water in the unfrozen serum phase. This mechanism not only stabilizes the microstructure but also modifies the path towards the glass transition temperature (T _g), potentially enhancing storage stability.

Ice crystals

Given that ice crystal morphology directly dictates sensory attributes, the maintenance of small crystals is paramount for imparting a smooth, dense, and high‐quality texture to the final product.

In 0% (w/w) TSP, rapid ice crystal propagation formed a coarse and highly branched network indicative of Ostwald ripening, which compromises textural smoothness. Conversely, 0.2% TSP significantly attenuated crystal growth, fostering numerous fine crystals and inhibiting coalescence. Optimal inhibition was observed at 0.5% (w/w) TSP (Fig. 3).

Figure 3.

Micrographs of ice crystals in sugar solutions with different tamarind seed polysaccharide (TSP) concentrations.

The temporal stability of the microstructure further underscored the concentration‐dependent efficacy of TSP. With increasing storage time, the ice crystal network in 0% (w/w) TSP coarsened drastically. Initially exhibiting a relatively uniform state, the system rapidly developed into coarse dendrites that dominated the field of view, clearly demonstrating the Ostwald ripening process. In contrast, the microstructure of the 0.2% TSP sample remained relatively stable; although slight network development occurred between 10 and 40 min, a fine and uniform morphology was maintained. The 0.5% TSP sample exhibited minimal morphological changes throughout the observation period. Quantitative analysis at the final observation point highlighted this inhibitory effect: compared to the control, the 0.5% (w/w) TSP sample showed a significant reduction in crystal dimensions. Specifically, as the TSP concentration increased to 0.5% (w/w), the average ice crystal area decreased from 221.69 to 135.78 μm², while the maximum Feret diameter dropped from 18.22 to 15.86 μm (Table 2). These results confirm that the addition of TSP effectively retards ice recrystallization. ⁸

Mechanistically, TSP provides superior cryoprotection through an active multi‐scale interaction with the freezing front, distinguishing it from traditional stabilizers like guar gum and carboxymethyl cellulose (CMC). While the latter primarily retard ice crystal growth passively by increasing viscosity, TSP's branched galactoxyloglucan backbone actively adsorbs onto the ice–liquid interface via dense hydrogen bonding.^{37, 38} This creates a non‐freezing hydration layer that serves as a potent thermodynamic barrier, significantly increasing the activation energy for water migration and arresting Ostwald ripening. Moreover, the TSP‐mediated three‐dimensional network imposes a robust physical confinement effect in the cryo‐concentrated phase, restricting water mobility more effectively than thermal‐dependent gums like locust bean gum. These attributes highlight TSP as a high‐efficiency, multi‐functional alternative to conventional viscosity‐based stabilizers.^{8, 10, 39}

Texture profile analysis (TPA) of ice cream mix

The textural characteristics of the ice cream mixes containing different TSP concentrations are detailed in Table 3. Analysis revealed no significant variance (P > 0.05) in liquidity or firmness.

Table 3.

Textural parameters of ice cream mixes and final ice cream products with different tamarind seed polysaccharide (TSP) concentrations

Parameter	TSP concentration
	0%	0.2%	0.5%
The following data were measured on the ice cream mixes
Liquidity (g s−1)	3.225 ± 0.335a	3.611 ± 0.374a	3.540 ± 0.179a
Firmness (g)	10.656 ± 1.552a	10.704 ± 1.196a	11.103 ± 1.195a
Consistency (g·s)	40.370 ± 0.219b	40.627 ± 0.395ab	41.421 ± 0.805a
Cohesiveness (g)	5.835 ± 0.127a	6.124 ± 0.171b	7.121 ± 0.056c
Viscosity index (g·s)	3.830 ± 0.185a	4.224 ± 0.191b	6.195 ± 0.251c
The following data was measured on the final ice cream products
Hardness (g)	8600.22 ± 345.79a	2713.79 ± 26.36c	3912.42 ± 220.76b

Note: Different lowercase letters indicate significant differences among ice cream mixes and final ice cream products with different TSP contents (P < 0.05).

However, viscosity index, quantified here as the work of adhesion, demonstrated a significant TSP‐dependent increase, elevating from 3.830 g·s to 6.195 g·s. This phenomenon was attributed to the establishment of a three‐dimensional network by high‐molecular‐weight polysaccharides in the continuous phase. As a branched polysaccharide rich in glucose, xylose, and galactose, TSP extends fully in solution, engaging in extensive interactions such as hydrogen bonding, Van der Waals forces, and steric entanglements. These interactions foster a continuous weak gel network that heightens internal frictional resistance, macroscopically translating to increased viscosity. Given its shear‐thinning nature, TSP is effectively positioned as a functional thickener for stabilizing frozen desserts.^{30, 40}

Cohesiveness, which reflects the internal bond strength preserving structural integrity, showed a similar upward trend. ⁴¹ The 0.5% (w/w) TSP formulation achieved a cohesiveness value of 7.121, significantly surpassing the control group (5.835). This indicates that TSP facilitates the formation of a more robust and compact gel network. ⁴² The enhanced intermolecular forces, driven by hydrogen bonding and chain entanglement, enable the system to withstand structural rupture under shear or compressive stress.

As a key textural property reflecting structural changes, cohesiveness represents the work required to overcome the internal binding forces of the network. ⁴³ The implications of elevated cohesiveness are two‐fold. Microscopically, a cohesive network acts to immobilize ice crystals, retarding their growth and coarsening mechanisms, a prerequisite for a smooth and non‐grainy mouthfeel. Macroscopically, it imparts superior shape retention and resistance to collapse during storage and scooping. Furthermore, this textural reinforcement aligns with the rheological findings, specifically the increase in G', providing comprehensive evidence of TSP's role in enhancing the elastic and solid‐like behavior of the ice cream mix.

Microstructure observation via multiple fluorescence staining

Multiple fluorescence staining was employed to visualize the microstructural architecture of the ice cream mixes, as depicted in Fig. 4. The protein matrix, labeled with FITC, appears as a yellow‐green continuous phase, whereas the dispersed oil droplets, stained with Nile Red, are visualized as orange‐red spheres. The sample without TSP exhibited considerable heterogeneity in oil droplet size. In contrast, TSP incorporation induced a transition towards a finer and more monodisperse distribution. This morphological change corroborates the particle size reduction observed in earlier analyses. The stabilizing mechanism is two‐fold. First, TSP molecules adsorb at the interface or structure the continuous phase to create a steric barrier. This barrier mitigates excessive flocculation, thereby regulating partial fat coalescence to an appropriate degree necessary for structural stability and textural optimization. ⁴⁴ Second, the rheological modification plays a critical role. The addition of TSP significantly increases the system's viscosity. This elevated viscosity increases the shear stress applied to fat globules during homogenization, thereby promoting more efficient droplet disruption and minimizing particle size. Consequently, the formation of these finer droplets serves as a foundational factor in enhancing the long‐term stability of the mix.^{45, 46}

Figure 4.

Fluorescence micrographs of ice cream mixes with different tamarind seed polysaccharide (TSP) concentrations.

Quality characteristics of ice cream

Hardness

Hardness serves as a pivotal textural parameter, directly reflecting the microstructural architecture, particularly the dimensions and spatial distribution of ice crystals. ⁴⁷ Quantified as the peak deformation force, it indicates the frozen matrix's rigidity. As detailed in Table 3, TSP significantly reduced hardness, exhibiting an inverse correlation with increasing concentration. This stems from TSP's cryostabilizing effect, which retards recrystallization and minimizes crystal size. Interestingly, hardness slightly increased in the 0.5% (w/w) TSP sample as compared to the 0.2% counterpart. This deviation may be explained by the increased number of densely fine crystals, which, despite their small size, reinforce the structural network. Fundamentally, larger crystals form a rigid and continuous ice skeleton leading to a harder texture, whereas finer crystals yield a softer, smoother product. ⁴⁸

The rheological properties of the unfrozen serum phase also play a governing role. Elevated viscosity and restricted molecular mobility impede ice crystal propagation. Existing literature confirms that hardness scales positively with ice crystal size and total ice phase volume.^{49, 50} In conclusion, TSP addition promotes the development of a cohesive and robust hydrocolloid network. This structured matrix effectively disrupts the continuity of the ice phase and reduces the overall ice fraction, thereby resulting in the observed decrease in macroscopic hardness relative to the control.

Melting rate

The melting rates of the ice cream mixes formulated with different TSP concentrations are detailed in Table 4. Analysis revealed that melting behavior is significantly modulated by both TSP dosage and temperature. The melting rate showed a distinct dose‐dependent upward trend: at 25 °C and 30 °C, the values for the 0.5% (w/w) TSP samples were 76.43% and 84.92%, respectively, both significantly higher than those of the 0% (w/w) TSP samples (50.93% and 66.94%). These data demonstrate that TSP incorporation reduces meltdown resistance, leading to faster structural collapse at ambient temperatures.

Table 4.

Melting rate of ice cream with different concentrations of tamarind seed polysaccharide (TSP)

TSP concentration (%)	Temperature (°C)	Melting rate (%)
0	25	50.93 ± 1.52Cc
	30	66.94 ± 6.38Ab
	35	60.41 ± 9.15ABa
0.2	25	63.62 ± 3.88Bb
	30	76.94 ± 3.86Aa
	35	60.54 ± 3.37Ba
0.5	25	76.43 ± 3.58Ba
	30	84.92 ± 2.30Aa
	35	63.87 ± 1.26Ca

Note: Different lowercase letters indicate significant differences among ice creams with various TSP levels at the same temperature (P < 0.05); different uppercase letters denote significant differences among different temperatures for the same TSP level (P < 0.05).

Mechanistically, this acceleration is attributed to the freezing point depression induced by TSP. By increasing the soluble solids content, TSP elevated the volume fraction of the unfrozen water phase, leaving a smaller mass of ice to resist thermal degradation at a specific temperature. However, at 35 °C, the absence of significant differences between groups (all rates > 60%) suggests that the high ambient temperature's thermodynamic driving force overshadowed the specific effects of TSP concentration. Thus, while TSP served as an excellent cryostabilizer for texture, its application involved a functional trade‐off regarding meltdown resistance, aligning with its role in accelerating melting while inhibiting recrystallization. ⁹

The melting behavior of ice cream is governed primarily by fat destabilization and the microstructural distribution of ice crystals. ⁵¹ Crucially, the particle size distribution of ice crystals serves as a determining factor. It is well‐established that prolonged freezing durations introduce frictional heat, which accelerates Ostwald ripening – a recrystallization phenomenon characterized by the sacrificial melting of smaller crystals to fuel the growth of larger ones. ⁵² Thermodynamically, melting is driven by heat transfer from the surrounding warm air to the ice cream interface, inducing an endothermic phase transition. Melting initiates at the exterior surface, establishing a temperature gradient. The generated meltwater dilutes the unfrozen serum phase, reducing its viscosity and facilitating gravitational drainage. As sufficient thermal energy penetrates the core, the structural integrity of the internal ice network is compromised, resulting in the complete collapse and flow of the fluid matrix.

CONCLUSIONS

This study elucidated the functional role of TSP at varying concentrations (0%, 0.2%, and 0.5%, w/w) in modulating the physicochemical and microstructural properties of ice cream. Rheological characterization revealed that TSP significantly reinforced the network structure, evidenced by increased viscosity, cohesiveness, and G' and G", thereby imparting a robust gel‐like behavior to the mix. Microstructurally, TSP incorporation not only reduced the mean particle size of the mix but also significantly decreased ice crystal dimensions and product hardness in the final dessert, contributing to a smoother texture. Notably, while the melting rate increased, the structural stability was maintained. Freezing curve analysis further demonstrated a distinct extension of the phase transition plateau in TSP‐fortified samples. Mechanistically, this is ascribed to the hydrocolloid's ability to immobilize free water via hydrogen bonding, which thermodynamically inhibits ice nucleation and recrystallization. In conclusion, TSP serves as a potent cryostabilizer that optimizes the microstructure and quality attributes of ice cream. This research provides a valuable theoretical and empirical basis for the utilization of TSP as a novel stabilizer in the frozen dessert industry.

CONFLICT OF INTEREST

The authors declare no conflict of interest.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.