Polysaccharides reinforced sesame oil body interface as 3D printing inks for dysphagia food: focusing on structural characteristics and rheological properties

Abstract

Plant oil body-based emulsion gels as 3D printing inks with unique structures and mechanical properties hold great prospects in dysphagia-friendly foods. This study investigates the rheological properties and 3D printing characteristics of heated sesame oil body (HSOB)-based inks by blending three different anionic polysaccharides: κ-carrageenan (κ-GC), xanthan gum (XG), and gum arabic (GA). The influence of polysaccharide types on ink printing performance, rheological properties, microstructure, and moisture distribution was analyzed. The results indicate that κ-GC-HSOB ink exhibits excellent dimensional accuracy, self-supporting ability, smooth surface finish, and minimal defects. In contrast, XG-HSOB and GA-HSOB inks display rougher surface textures, structural collapse, and oil discharge. IDDSI experiments classified HSOB and GA-HSOB as Level 4 dysphagia foods, while κ-GC-HSOB and XG-HSOB were classified as Level 5. Overall, this study elucidates how different anionic polysaccharides enhance the 3D printing characteristics of HSOB-based inks, highlighting their potential for developing dysphagia-friendly food.

Graphical abstract

Highlights

• •Compared the microstructures of heated sesame oil body (HSOB) with different polysaccharides.
• •Incorporation of polysaccharides enhanced the viscosity and shear-thinning ability of inks.
• •κ-carrageenan (κ-GC)-HSOB ink has exceptional printing accuracy and self-supporting capabilities.
• •κ-GC-HSOB was formed by hydrophobic interactions, hydrogen bonding, and disulfide bonds.
• •κ-GC improved texture of HSOB, effectively modulating swallowing properties.

1. Introduction

In recent years, with the aging of the global population and the increasing incidence of dysphagia-related diseases, the demand for dysphagia-friendly foods has surged. Dysphagia, a swallowing disorder, can lead to malnutrition, aspiration pneumonia, and even death, making the development of safe, nutritious, and easy-to-swallow foods a critical research focus. Conventional dysphagia foods often suffer from monotonous textures, poor sensory properties, and limited nutritional diversity, failing to meet the diverse needs of patients (Xu et al., 2025). 3D food printing, as an emerging additive manufacturing technology, has revolutionized the food industry by enabling the customization of food shapes, textures, and nutritional compositions. Extrusion-based 3D food printing, in particular, has gained widespread attention due to its simplicity, versatility, and suitability for a wide range of food materials (Johannesson et al., 2023). However, the printability of food inks, which is closely related to their rheological properties, microstructure, and mechanical strength, remains a key challenge in 3D food printing (Yang, Su, et al., 2020). For dysphagia-friendly foods, the inks must not only exhibit excellent printability to form stable 3D structures but also meet specific texture requirements to ensure safe swallowing.

Emulsion gels, composed of oil-in-water or water-in-oil emulsions stabilized by gel networks, have emerged as promising 3D printing inks due to their tunable rheological properties, good mechanical strength, and ability to encapsulate bioactive compounds (Yan et al., 2024). Plant-derived oil bodies (OBs) have garnered considerable interest as sustainable and minimally processed alternatives to conventional emulsifiers and oil sources in food formulations (Yang, Cheng, et al., 2025). Sesame is the widely produced oilseed crop with the highest oil content (45.0% ∼ 65.0%) in Asia and Africa, and the global production of sesame seeds was about 67.0 lakh tons (Mahajan et al., 2025), which provides abundant materials for oil body preparation. Villa et al. (2025) reported that sesame seeds exhibit high nutritional value, as they are rich in unsaturated fatty acids and bioactive compounds, such as lignans, plant sterols, and tocopherols. These lipophilic components are localized within sesame oil bodies (SOB) (Chen et al., 2014), making SOB a promising source of essential nutrients for dysphagia people. As natural emulsion systems derived from sesame seeds, SOBs exhibited emulsion gel–like characteristics., which made them potential candidates for 3D printing inks (Nikiforidis et al., 2016; Yang et al., 2026). However, the inherent rheological properties and mechanical strength of OB alone may not be sufficient to meet the requirements of 3D printing, especially for the fabrication of complex structures (Shi et al., 2023). It was found that heat seeds can enhance the adsorption of exogenous proteins at the OB interface (Chen et al., 2023), and the increased interfacial protein further enhanced the interfacial strength of the OBs (Wang et al., 2025). Therefore, heat treatment (such as roast) is very likely to be beneficial for the 3D printing of SOB.

Polysaccharides are widely employed to modulate the texture and stability of food systems. Their ability to thicken, form gels, and interact with interfaces made them ideal candidates for structuring emulsion-based inks (Lin et al., 2025). The abundance, biocompatibility, non-toxicity, and affordability of κ-GC, XG, and GA made them stand out among other polymers as potent candidates for 3D printing applications (Sarab Badieh et al., 2025). Although κ-GC, XG, and GA are all anionic polysaccharides, their structures and roles in emulsion gel inks are different. Specifically, κ-GC is linear with approximately 20% sulfate groups and forms strong, brittle gels via helix aggregation (Li et al., 2024; Sarab Badieh et al., 2025). When interacting with OB, κ-GC may promote the formation of disulfide bonds, thereby strengthening the gel network structure. XG is a branched polysaccharide that provides high pseudoplastic viscosity and weak gel properties, primarily through molecular chain entanglement (Zheng et al., 2025). When interacting with OBs, XG may increase the gel's viscosity and high shear-thinning behavior, giving the OB ink both stability and easy flow during 3D printing. GA is a branched polysaccharide and an effective emulsifier, but it has limited gelling capacity (Bak, 2024). When interacting with OBs, GA may improve OB emulsion stability, which is essential for emulsion gel inks. Critically, the specific interactions between these polysaccharides and the modified interface of OBs, and how these interactions translate into the structural characteristics and rheological properties crucial for 3D printing performance, remain largely unexplored (Khubber et al., 2025). Moreover, the relationship between polysaccharide-induced structural changes in oil bodies from heated seeds (HSOB) and their suitability for 3D-printed dysphagia foods has not been systematically investigated. Understanding how polysaccharides influence the International Dysphagia Diet Standardization Initiative (IDDSI) classification of HSOB-based 3D-printed foods is crucial for their practical application.

Therefore, this study aims to develop HSOB-based inks reinforced with κ-GC, XG, and GA, and systematically investigate the effects of these polysaccharides on the structural characteristics (including microstructure, molecular interactions, and protein secondary structures) and rheological properties of the inks. The printability of the inks, as evaluated by parameters such as hardness, self-supporting capability, and printing deviation, will be correlated with their structural and rheological properties. Additionally, the IDDSI classification of the 3D-printed products will be determined to assess their suitability for dysphagia patients. The findings of this study are expected to provide valuable insights into the development of polysaccharide-reinforced emulsion gel inks for 3D printing of dysphagia-friendly foods, and contribute to the advancement of 3D food printing technology in the field of medical nutrition.

2. Materials and methods

2.1. Materials and reagents

Sesame seeds were provided by Shandong Shilixiang Sesame Products Co., LTD. The κ-carrageenan (κ-GC, CAS: 1114-20-8), xanthan gum (XG, CAS: 11138-66-2), and gum arabic (GA, CAS: 9000-01-5) were obtained from Shanghai Yuanye Bio-Technology Co., Ltd. All experimental reagents utilized in this study were of analytical grade and procured from Tianjin Kaitong Chemical Reagent Co., Ltd.

2.2. Extraction of SOB and HSOB

Sesame seeds were roasted in an oven (SM2-522H, Sinmag Machinery (Wuxi) Co., Ltd., China) under different temperatures and times. For the temperature treatment, seeds were roasted at 50, 70, 90, 110 °C and 130 °C for 20 min. For the time treatment, seeds were roasted at 110 °C for 10, 15, 20 and 25 min.

The roasted and unroasted sesame seeds were mixed with deionized water in a weight-to-weight ratio of 1:7 for 10 h at 4 °C. Next, the soaking seeds were mixed with fresh deionized water (seeds: water = 1:7), and each mixture was ground using a high-speed blender (JYL-C16D/C16V, Joyoung Co., Ltd., China) at 22,000 rpm for 3 min. After grinding, filter the mixture through four layers of gauze to obtain sesame slurry. The slurry was centrifuged at 7000 ×g for 30 min at 4 °C, and the creams were collected, dispersed into deionized water, and centrifuged at the same conditions. The floating fraction was collected as oil bodies (OBs). Oil bodies from unroasted sesame seeds were defined as SOB (control), whereas those from roasted sesame seeds were defined as HSOB.

2.3. Preparation of HSOB 3D printing inks with different polysaccharides

Add κ-GC, XG, and GA to HSOB separately, maintaining a polysaccharide to oil body mass ratio of 0.5%. The κ-GC concentration was selected based on the results shown in Figs. S1 ∼ S4, and the addition of 0.5% κ-GC to HSOB resulted in better physicochemical properties. The concentrations of XG and GA were chosen according to Li et al. (2025) and Sukhotu et al. (2016), respectively. HSOB complexes with the different polysaccharides were prepared, and designated as HSOB, κ-GC-HSOB, XG-HSOB, and GA-HSOB. Samples were stored in a refrigerator at 4 °C for subsequent use.

2.4. Determination of physical and chemical indicators

2.4.1. Analysis of SOB and HSOB composition

SOB and HSOB primarily consist of water, proteins, and lipids. Assess the moisture content, total lipid levels, and protein amounts through methods such as direct drying, Soxhlet extraction, and Kjeldahl nitrogen analysis (Shen et al., 2023).

2.4.2. SDS-page

1 g of HSOB was thoroughly mixed with 3 mL of deionized water. Then, 0.5 mL of this sample was mixed with 0.5 mL of sample loading buffer (containing 2% (v/v) β-mercaptoethanol and bromophenol blue). The mixture was heated at 100 °C for 5 min, followed by centrifugation at 4000 rpm for 5 min. A 10 μL aliquot of the supernatant was loaded into the gel sample well. The stacking gel and separating gel concentrations were 4% and 16%, respectively. Electrophoresis was performed at 40 mV through the stacking gel. When the samples entered the separating gel, the voltage was switched to 100 mV. The gel was fixed in fixing solution for 1 h, stained with Coomassie Brilliant Blue staining solution for 1 h, and finally destained using glacial acetic acid. Imaging was performed using a ChemiDoc gel documentation system (Bio-Rad company, USA).

2.4.3. Average particle size and particle size distribution

Mix 1 g sample with 10 mL of deionized water and stir evenly using a magnetic stirrer. The volume average particle size (D_{4, 3}) is measured with a particle size analyzer (Beijing Haixinrui Technology Co., Ltd., China) at 25 °C.

2.4.4. Zeta potential

1 g of the sample was mixed with 10 mL of deionized water and stirred uniformly using a magnetic stirrer. A 20 μL aliquot of the emulsion to be tested was mixed evenly with 8 mL of fresh deionized water and measured using a Nano-ZS Laser Particle Size Analyzer (Malvern Instruments Ltd., London, UK) at 25 °C.

2.4.5. Optical microstructure

Mix 1 g of the sample with 10 mL of deionized water, ensuring an even mixture using a magnetic stirrer. Place a drop of the stirred lotion sample in the center of a glass slide, cover it with a cover slip, and capture the microstructure image of the sample using an inverted fluorescent microscope (BX-51, TKO Optical Instruments Co., Ltd., Japan) equipped with a 10× eyepiece and a 40× objective lens at a temperature of 25 °C.

2.4.6. Cryo-SEM

The microstructure of different polysaccharides-HSOB emulsions was observed using a Cryo-electron scanning (Regulus 8100, Hitachi, Tokyo, Japan) equipped with a low-temperature preparation and transfer system (PP3010T, Quorum, UK). In brief, approximately 2.0 μL of the sample was frozen in liquid nitrogen, then transferred to the sample preparation chamber, where it was sublimated for 8 min at −80 °C and sputtered for 60 s at 10 mA, before being transferred to the electron microscope chamber for observation at 3.0 kV. (Yu et al., 2024).

2.5. Characterization of rheological properties

2.5.1. Apparent viscosity

Select the rotor model PP50 and determine the mode of speed control, targeting a shear rate range of 0.1–100 s⁻¹. Data should be collected using logarithmic sampling, resulting in a total of 19 data points. The consistency coefficient (K, Pa•sⁿ) and the flow behavior index (n) was calculated according to Liu, Wu, et al. (2024).

2.5.2. Frequency sweeps

Select the rotor of the PP50 model and choose the oscillation mode as frequency scanning for the measurement. Set the number of data points to 25, with the sampling time determined by the equipment. Maintain a constant shear strain of 1%, while allowing the frequency to vary logarithmically from 0.1 to 100 Hz, ensuring a data point density of 8. Additionally, tan δ is calculated based on the storage modulus (G') and the energy dissipation modulus (G") (Wang et al., 2025).

2.5.3. Amplitude sweep

The measurement mode is configured to oscillation mode amplitude scanning, utilizing a data point count of five and a sampling time determined by the device. The scanning frequency is maintained at a constant value of 1 Hz, while the shear strain varies logarithmically from 0.1% to 100%. The intersection point between the storage modulus (G') and the loss modulus (G") indicates the yield stress value (Rahman & Farooq, 2025).

2.5.4. Three-stage thixotropy test (3ITT)

The 3ITT method comprises three distinct sections of shear rate testing. The first section applies a constant shear rate of 1 s⁻¹ for 120 s, simulating the state of inks prior to the 3D printing process (Koo et al., 2026). The second section employs a constant shear rate of 100 s⁻¹ for 180 s, representing the conditions of inks during the 3D printing process. Finally, the third section replicates the conditions of the first section to simulate the state of inks following the completion of 3D printing (Liu et al., 2025).

2.6. Measurement of 3D printing performance of inks

Load the sample into the 3D printer bin and initiate printing using an extrusion-type 3D printer at room temperature. The parameters for printing are set as follows: printing temperature at 25 °C, printing speed at 70 mm/s, nozzle diameter at 0.84 mm, and material extrusion speed at 100%. The first printed shape is a triangular prism with a side length of 40 mm and a height of 10 mm. After printing, allow the sample to rest at room temperature for 1 h before capturing a photograph for documentation. The second printed shape is cylindrical, with a diameter of 30 mm and a height of 15 mm. Measure the side length and height of the samples immediately after printing and after 1 h of resting. Subsequently, calculate the accuracy and stability of the 3D printed samples using Eqs. (1), (2), respectively (Cheng et al., 2025).

$$3\mathrm{D}\text{printing accuracy}\left(\%\right)=\left(1-\frac{\left|{\mathrm{A}}_{\mathrm{c}}-{\mathrm{A}}_{\mathrm{a}}\right|}{{\mathrm{A}}_{\mathrm{c}}}\right)\times 100$$ (1)

In the formula: Ac is the printer setting data; Aa is the actual measurement data.

$$3\mathrm{D}\text{printing stability}\left(\%\right)=\left(1-\frac{\left|{\mathrm{A}}_{\mathrm{b}}-{\mathrm{A}}_{\mathrm{a}}\right|}{{\mathrm{A}}_{\mathrm{a}}}\right)\times 100$$ (2)

In the formula, A_a represents the length, width, and height data of the sample after printing is completed; A_b represents the length, width, and height data of the sample after being left at room temperature for 1 h.

2.7. Texture profile analysis (TPA)

A texture test was conducted on the 3D printed sample using the TA/0.5 gel probe. The pre-test speed was set at 0.5 mm/s, which was maintained throughout the test, including the post-test phase. The trigger point for the test was established at 3.0 gf, with a 3.0-s interval between consecutive tests. The target displacement was defined as 7.0 mm, while the trigger force was set to 5.0 gf, and a return distance of 10 mm was utilized (Farooq et al., 2025).

2.8. Surface-enhanced Raman spectroscopy

Surface-Enhanced Raman Spectroscopy (SERS) analyzes molecular stretching and bending vibration transitions through inelastic scattering, thereby providing insights into molecular properties and structures (Yang, Cheng, et al., 2025). Raman spectra were collected from samples using the Renishaw inVia Qontor confocal Raman microscopy system (Renishaw plc, Gloucestershire, UK). In this study, a 532 nm laser and a 5× objective lens were employed to capture spectra over the range of 400–4000 cm⁻¹.

2.9. Chemical forces

Portions of 0.5 g inks were separately mixed with 5 mL of five different solutions: Solution A, B, C, D, and E. Solution A was 0.05 mol/L NaCl, Solution B was 0.6 mol/L NaCl, Solution C contained 0.6 mol/L NaCl and 1.5 mol/L urea, Solution D contained 0.6 mol/L NaCl and 8 mol/L urea, and Solution E contained 0.6 mol/L NaCl, 8 mol/L urea, and 0.5 mol/L DTT (dithiothreitol). The mixtures were stirred for 1 h, then incubated at 4 °C for 1 h, followed by centrifugation at 8000 rpm for 20 min. The protein content in the supernatant was determined using the Coomassie Brilliant Blue method. The differences in protein concentration between Solution B and Solution A, between Solution C and Solution B, between Solution D and Solution C, and between Solution E and Solution D represent electrostatic interactions, hydrogen bonding interactions, hydrophobic interactions, and covalent interactions, respectively.

2.10. LF-NMR

The prepared 3D printing ink was weighed (2 g) and covered with a polyethylene film to prevent moisture loss. Subsequently, the gel was placed into a 15 mm diameter glass tube and analyzed using a MicroMR pulsed NMR analyzer (Niumag Co., Shanghai, China) for low-field NMR (LF-NMR) analysis, with a focus on the T₂ relaxation time. The Carre-Purcelle-Meiboom-Gill (CPMG) radio frequency pulse sequence was employed. The T₂ relaxation time curve was obtained using multi-exponential inversion software (Xu et al., 2025).

2.11. FTIR

FTIR spectroscopy was employed to examine alterations in the secondary structure of HSOB-based inks across various treatment groups. Lyophilized proteins were thoroughly mixed with KBr in a 1:100 (w/w) ratio, then ground and compressed for the analysis. FTIR spectra (IRTracer-100, Shimadzu Corporation, Japan) were collected at wavenumbers ranging from 4000 to 400 cm⁻¹, with a resolution set at 1 cm⁻¹.

2.12. IDDSI

The IDDSI Framework (https://iddsi.org/Framework) offers globally standardized terminology and test methods for describing foods or liquids suitable for dysphagia diets, dividing both beverages and food products into eight distinct categories (0–7). IDDSI has been widely applied to the evaluation of 3D printed gels for dysphagia diets (Hou et al., 2023; Yang et al., 2026). In the fork pressure assessment, a printed sample was pressed with a thumb until the fork deformed the material, applying a force nearly equivalent to 17 kPa. This pressure approximates that exerted by the tongue when swallowing. Additionally, all ink formulations were subjected to a spoon tilt test to evaluate their cohesion and adhesion characteristics.

2.13. Statistical analysis

The data were examined with the use of IBM SPSS version 22.0 statistical software (SPSS Inc., Chicago, USA). Differences deemed significant between the groups (p < 0.05) were identified through analysis of variance (ANOVA) along with Duncan's multiple range test.

3. Results and discussions

3.1. Basic components and rheological properties of SOB and HSOB

During extraction, OBs undergo fragmentation, leading to the release and dissolution of exogenous proteins in the aqueous phase. At this stage, the exogenous proteins can bind to OBs through various interactions, which subsequently thickness the interfacial layer (Yang et al., 2026). Fig. 1a illustrates the SDS-PAGE analysis of SOB and HSOB. With heating temperature and time increase, a greater quantity of exogenous proteins were adsorbed onto the HSOB interface, particularly the 2S, 7S and 11S proteins. The observation that endogenous proteins, such as Oleosin-H and Oleosin-L, remain largely unchanged further substantiates the authenticity of exogenous protein adsorption within the external interfacial layer.

Fig. 1.

Components and rheological properties of SOB and HSOB. (a) SDS-PAGE, (b, c) basic composition at different heating temperatures and times, (d, e) apparent viscosity, (f, g) storage modulus G′ of LVR. Letters a ∼ e indicate significant differences between groups (p < 0.05).

Fig. 1b and c quantitatively demonstrate that the protein content increases with both temperature and time, high heating temperature (130 °C) decreased the lipid content to below 80%. Furthermore, a significant increase in particle size was observed at both 130 °C for 20 min and 110 °C for 30 min (Fig. S5), indicating decreased colloidal stability. HSOB extracted from seeds heated at 110 °C for 25 min, contained higher lipid content (91.95 ± 1.64%) and higher protein content (4.18 ± 0.28%). Research indicated that higher concentrations of interfacial proteins can enhance texture and improve 3D printing performance (Ghazani et al., 2024). As shown in Fig. 1d and e, heating at 110 °C for 25 min resulted in HSOB exhibiting the highest consistency coefficient (K) and the lowest flow behavior index (n). A higher K indicates greater viscosity at low shear rates, preventing oil bodies from dripping during 3D printing, while a lower n reflects stronger shear-thinning behavior that enables smooth extrusion under high shear rates and reduces the risk of nozzle clogging (Liu, Chen, et al., 2024; Yang et al., 2026). The strain sweep results (Fig. S5) showed all samples had similar the linear viscoelastic region (LVR), and 110 °C for 25 min made HSOB had the highest storage modulus (G' in Fig. 1g), indicating the formation of a stable elastic network structure. Therefore, heating at 110 °C for 25 min was selected for HSOB preparation. Under this condition, the storage modulus and viscosity of HSOB exhibit favorable levels, which are beneficial for the preparation of gels with polysaccharides.

3.2. Visual appearance and structural analysis of HSOB-based inks

Fig. 2 illustrated the network structures formed by the crosslinking of different polysaccharides with HSOB, where HSOB droplets accumulated and interfacial proteins cross-link, resulting in a network structure. The κ-GC-HSOB exhibited changes in the interface and network entanglement cross-linking, potentially resulting in superior mechanical strength. This was confirmed by the rheological results shown in Fig. 4e–h, where κ-GC-HSOB displayed the highest storage modulus (G′) within the linear viscoelastic region (LVR) as well as enhanced thixotropic recovery. The κ-GC-HSOB, due to its high water absorption capacity, allows the molecules to extend in water and form a dense three-dimensional network structure, which is closely associated with the interactions. As shown in Fig. 6d, addition of GC markedly increased hydrogen bonding, hydrophobic interactions, and disulfide bond. The κ-GC enhanced molecular interactions with HSOB and increased network gel strength. XG-HSOB showed HSOB filling within the XG gel network, creating a uniformly dense network structure, although the oil droplets increase in size (Fig. 2). This structure endowed XG-HSOB with moderate gel strength because of its intermediate G′ in the LVR and thixotropic recovery (Fig. 4e–h). In addition, GA-HSOB demonstrated partial emulsification phenomena, where GA disrupts the HSOB interface, leading to droplet aggregation (Dickinson, 2013). The network structure of the XG-HSOB based emulsion gel becomes highly interconnected and denser through hydrophobic interaction and ion bonds. Zhu et al. (2025) reported noncovalent interactions were considered to be the main driving forces between GA, XG, and peanut OBs. The Schiff effect occurs between GA and the HSOB interface for cross-linking, but due to the accumulation of HSOB, the entire system collapses after adding GA, possibly due to phase separation caused by excessive GA addition (Farooq et al., 2024). This GA-induced HSOB collapse may explain the poor gel structure, and GA-HSOB exhibits reduced thixotropic recovery (Fig. 4e–h). All samples exhibited characteristics similar to high internal phase emulsions, remaining stable when inverted, making them suitable for the development of 3D printed foods.

Fig. 2.

Cryo-SEM and structural models of different polysaccharide-HSOB inks. The scale is 10 μm.

Fig. 4.

Rheological properties of different HSOB-based inks. (a) Frequency sweeps, (b) Tan δ, (c, d) viscosity behavior, (e) Strain sweeps, (f) LVR (G′), (g) 3-ITT, (h) Recovery rate. Letters a ∼ d indicated significant differences between groups (p < 0.05).

Fig. 6.

Microscopic indicators characterize intermolecular interactions. (a) LF-NMR, (b) Moisture type, (c) Raman spectra, (d) Chemical forces, (e) FTIR, (f) Protein secondary structure. Letters a ∼ d indicate significant differences between groups (p < 0.05).

3.3. Effects of the physicochemical properties of HSOB-based inks

Fig. 3 illustrated the changes in microstructure, zeta potential, particle size, and particle size distribution of HSOB following the addition of various polysaccharides (0.5%, w/w). Fig. 3a illustrated that the addition of κ-GC and XG led to a shift in the size distribution of the HSOB toward smaller particles, whereas the addition of GA had no significant effect. The κ-GC-HSOB displayed the smallest particle size and contained no significant large droplets (Fig. 3b). In contrast, other samples exhibited noticeably larger droplets. D_4,3, which is based on the volume mean diameter, represents the most characteristic average droplet diameter. Particle sizes ranged from 4.07 μm to 7.77 μm among the samples, following the order: GA-HSOB > HSOB > XG-HSOB > κ-GC-HSOB (Fig. 3c). In addition, Fig. S1 showed that both a low concentration (0.2%) and a high concentration (1.1%) of κ-GC resulted in larger particle sizes and broader size distribution of HSOB.

Fig. 3.

Physical and chemical indicators of different polysaccharides-HSOB inks. (a) Size distribution, (b) Optical microstructure (the scale is 20 μm.), (c) Average particle size, (d) Zeta potential. Letters a ∼ c indicate significant differences between groups (p < 0.05).

As depicted in Fig. 3d, HSOB-polysaccharide ink exhibited distinct zeta potential characteristics. Although GA, XG, and κ-GC are all anionic polysaccharides, they altered the zeta potential of the HSOB-based ink to different extents. Specifically, κ-GC and XG significantly decreased the zeta potential, which ranged from −32.53 mV to −34.23 mV. Notably, the zeta potential of GA-HSOB showed no statistically significant difference from that of HSOB. This variation is predominantly governed by the negative charge density of the polysaccharides, as reported previously (Wang et al., 2024). 0.2% κ-GC induced only a limited increase in the negative charge of HSOB, whereas concentrations above 0.5% resulted in HSOB zeta potentials below −30 mV, indicating the strong electrostatic repulsion (Fig. S1). The changes in zeta potential after polysaccharide addition are presumably associated with the specific intermolecular interactions (e.g., electrostatic interactions, hydrogen bonding) between HSOB and polysaccharides, which promote the formation of the HSOB-polysaccharide composite structures (Miao et al., 2025).

3.4. Rheological analysis of HSOB-based inks with different polysaccharides

The rheological properties of HSOB-based inks were studied. As shown in Fig. 4a and b, for all inks, G′ was always greater than G″, indicating the elastic behavior of the inks. The G′ and G″ of HSOB-based inks exhibited minimal frequency dependence, suggesting that the gel structure was insensitive to frequency fluctuations (Yang, Feng, et al., 2020). The viscosity curves of HSOB-based inks (Fig. 4c), demonstrated that viscosity decreased gradually with increasing shear rate, indicating typical shear thinning behavior (Sweeney et al., 2017). The addition of polysaccharides significantly enhanced the apparent viscosity of HSOB. Moreover, higher apparent viscosity contributed to a more stable system, which was beneficial for achieving good printability in 3D printing (Zambrano & Vilgis, 2023). As shown in Fig. 4d, the addition of polysaccharides to HSOB inks led to a significant increase in the K value and a decrease in the n value, which might be attributed to the formation of disulfide bonds and hydrophobic interactions between the polysaccharides and HSOB, resulting in higher viscosity for the HSOB inks (Fig. 6d).

As illustrated in Fig. 4e, when the strain exceeds 10%, G′ began to decrease due to the deformation of the gel network. Meanwhile, G″ showed a slight increase before it starts to decrease with increasing strain, indicating the presence of “Type III nonlinear behavior” in HSOB-based inks (Hou et al., 2025). This rheological behavior may facilitate 3D printing, as the material tends to soften under relatively high stress during extrusion and solidifies once the stress is removed. With the addition of κ-GC, XG, and GA, the G′ and G″ values of the gels gradually increased, indicating an improvement in viscoelasticity and gel strength (Colodel et al., 2019). Specifically, the addition of κ-GC, XG, and GA significantly enhanced their G' of the gels (Fig. 4f), and κ-GC-HSOB had the highest G′, indicating the strongest gel structure. This might be attributed to the increased hydrogen bond and hydrophobic interaction between polysaccharides and proteins (Zhu et al., 2024).

The three-interval thixotropy test (3-ITT) was employed to simulate the various shear rates that the inks underwent before, during, and after 3D printing, in order to evaluate its 3D printing performance. After the HSOB-based 3D printing inks undergo shear thinning and pass through the extrusion nozzle, the shear force dissipates. At this point, it is crucial for the viscosity of the slurry to rapidly return to its pre-extrusion state. Otherwise, the low-viscosity slurry may flow and diffuse, leading to decreased product stability (Choi et al., 2025). As shown in Fig. 4g and h, the shear recovery characteristics of various HSOB-based inks were measured. Initially, the material was subjected to low shear rates of 1 s⁻¹ to simulate the condition of material accumulation within the barrel. Subsequently, the slurry's behavior was simulated under high shear rates of 100 s⁻¹, mimicking the state as it passed through the extrusion head. Finally, the material's condition post-extrusion was simulated again at a low shear rate of 1 s⁻¹ (Huang et al., 2021). The addition of κ-GC demonstrated significantly higher viscosity under high-speed shear, which aided in maintaining the shape of printed products at room temperature. Furthermore, κ-GC-HSOB had the highest recovery rate (73%), while HSOB exhibited a recovery rate of only 43.5%. This improvement indicated that polysaccharide-enhanced inks demonstrated better thixotropic recovery, thereby enhancing printability (Gao et al., 2025). As shown in Fig. S2, although increasing κ-GC concentration led to higher viscosity and G′ values, the recovery rates of samples containing 0.2–0.8% κ-GC were similar and higher than that of the 1.1% κ-GC sample, suggesting that excessive κ-GC may hinder structural recovery.

3.5. 3D printing performance of HSOB-based inks with different polysaccharides

Triangular and cylindrical shapes were printed to evaluate the 3D printing performance of HSOB-based inks. Fig. 5 demonstrated that the HSOB exhibited a relatively favorable 3D printing effect. However, the printing accuracy of the oil matrix with the addition of GA and XG has markedly diminished. Furthermore, visual inspection revealed that the sesame oil matrix containing GA experiences oil leakage, suggesting that the incorporation of GA not only fails to enhance the stability of the sesame oil matrix but also exacerbates its instability. In contrast, κ-CG-HSOB displayed excellent printing stability. Notably, the κ-CG-HSOB samples exhibited the smoothest and clearest surface texture, closely matching the expected geometric structures. In addition, we also evaluated the oxidation stability of four samples (Fig. S6). As the storage time increased, the POV values of all samples slightly increased. Limited increase in POV is likely due to the strong antioxidant properties of lignans and tocopherol (fat-soluble components) present in the sesame oil (Senouwa et al., 2023). Interesting, κ-CG-HSOB appeared to reduce oxidation, which was attributed to the synergistic effect of polysaccharides that form a three-dimensional network, physically hindering oxygen from entering the oil phase and effectively slowing down the oxidation process. Fig. 5e–h illustrated that κ-CG-HSOB exhibited the highest printing accuracy. Over time, the stability of all samples decreased, as evidenced by the increase in width and height deviations after one hour of storage post-printing. Among these, the κ-CG-HSOB sample showed the smallest deviation, corresponding to its highest G' in the LVR and recovery rate in the 3-ITT test (Fig. 4). Although GA-HSOB exhibited moderate recovery behavior and G′, its printing precision and structural stability were relatively worse than HSOB. This was mainly attributed to GA-induced droplet coalescence in HSOB (Fig. 2), which was likely to be further exacerbated during the 3D printing process, ultimately leading to oil exudation on the surface of the printed constructs. In addition, Fig. S3 showed that κ-CG at concentrations of 0.5–0.8% resulted in the highest printing accuracy and stability of HSOB. Overall, the κ-CG-HSOB sample demonstrated the best printing performance, with exceptional accuracy and stability.

Fig. 5.

3D printing performance of different HSOB-based inks. (a) and (b) are the triangles printed for 0 min and 60 min, respectively. (c) and (d) are cylinders printed for 0 min and cylinders printed for 60 min. (e), (f), (g), and (h) are the printing accuracy of the 3D printed. Letters a ∼ c indicate significant differences between groups (p < 0.05).

3.6. TPA of 3D printed HSO-based inks with different polysaccharides

The rheological and microstructural characteristics of HSOB ink are related to its structural properties, which directly affect its application in food processing. Table 1 illustrated the structural characteristics of the inks with different formulations. It was observed that the addition of an appropriate amount of polysaccharides enhanced the hardness, Springiness and cohesiveness of the HSOB-based inks. Polysaccharides serve as the filling component of the OB gel network, highlighting the importance of self-supporting 3D printed foods. Among these, the κ-GC-HSOB-based ink exhibited the highest hardness, followed by the XG-HSOB and GA-HSOB-based inks (Cheng et al., 2026). In terms of elasticity, the inks containing κ-GC, XG, and GA outperformed the control group, indicating an enhanced ability to recover quickly from deformation. The elasticity of the κ-GC-HSOB-ink was greater than that of the other groups. This may be attributed to its flexible gel network structure, which resists external deformation, resulting in greater elastic potential energy during the extrusion process. Furthermore, the addition of 0.2% κ-GC only slightly increased the hardness of the HSOB-based ink. Increasing the concentration to 0.8–1.1% significantly enhanced hardness (Table S1). In contrast, 0.5% κ-GC resulted in moderate hardness, which is more suitable for the development of dysphagia-friendly foods. The cohesiveness of the HSOB-based inks with polysaccharide addition was significantly greater than that of the HSOB-based inks, reflecting an enhanced binding capacity of the particles within the gel. The introduction of polysaccharides stabilized and continuously reinforced the oil body network structure, thereby enhancing its cohesiveness. Moreover, these changes in texture may alter the IDDSI rating of the printed gel. The increase in cohesiveness allows the food to form clumps after chewing, thereby preventing dispersion and choking hazards.

Table 1.

Total texture analysis of different HSOB-based inks.

	Hardness (gf)	Springiness	Adhesiveness (gf)	Chewiness (gf)	Cohesiveness	Resilience
HSOB	8.55 ± 0.41b	0.59 ± 0.08b	−58.15 ± 2.95a	5.85 ± 0.74c	0.58 ± 0.04a	0.50 ± 0.04b
κ-GC-HSOB	13.14 ± 0.7a	0.87 ± 0.04a	−74.00 ± 2.53b	8.51 ± 0.22a	0.69 ± 0.12a	0.60 ± 0.01ab
XG-HSOB	12.66 ± 0.11a	0.79 ± 0.01a	−67.72 ± 2.72b	6.85 ± 0.36bc	0.70 ± 0.04a	0.73 ± 0.03a
GA-HSOB	11.87 ± 1.16a	0.74 ± 0.07ab	−65.26 ± 4.31ab	7.65 ± 0.38ab	0.61 ± 0.02a	0.65 ± 0.01ab

Letters a ∼ c indicate significant differences between groups (p < 0.05).

3.7. Microscopic indicators characterize intermolecular interactions

The water distribution in HSOB-based inks was measured using LF-NMR. As shown in Fig. 6a, the relaxation times of all samples were divided into three peaks, corresponding to three states of water: bound water (T₂₁, 1–100 ms), macromolecular surface water (T₂₂, 100–400 ms), and water fixed in the gel network (T₂₃, 600–1000 ms) (Liu et al., 2025). Compared to the control group, the T₂ peak of the HSOB-based inks shifted to the left, indicating that the addition of polysaccharides reduced the mobility of water and enhanced the stability of the ink. Notably, the κ-GC-HSOB-based inks exhibited the shortest relaxation time, indicating a robust network structure and strong water-binding capacity. The κ-GC can interact with the hydrophilic groups of proteins in solution and water, thereby reducing the mobility of water. In contrast, the GA-HSOB-based inks showed lower water retention capacity, attributed to its lower viscosity and higher fluidity, making it suitable for rapid printing but prone to bottom collapse. On the other hand, the XG-HSOB ink exhibited longer T₂ and poorer stability, which may be due to the extensive branching of XG that could lead to steric hindrance, obstructing HSOB crosslinking and potentially forming a loose network structure that affected network construction, resulting in higher water mobility and bottom collapse. Overall, the κ-GC-HSOB inks demonstrated relatively moderate water retention capacity, providing better structural stability and moderate fluidity, making it easier to control during the 3D printing process.

Surface-Enhanced Raman Spectroscopy (SERS) was conducted to investigate the interactions involved in printing inks. The addition of polysaccharides did not produce new peaks, indicating that no new covalent interactions were formed between the polysaccharides and HSOB. Raman bands at 1600–1700 cm⁻¹ (amide I band region) are sensitive to protein secondary structure, with α-helix at 1645–1657 cm⁻¹, β-sheet at 1665 and 1680 cm⁻¹, and random coil around 1660 cm⁻¹. As shown in Fig. 6c, the principal peak in the amide I band region is associated with the α-helix, and GA-HSOB exhibited the lower peak intensity, indicating decreasing α-helix content, consistent with the results in Fig. 6f. The addition of polysaccharides resulted in a decreased peak intensity at 1021, 1260, and 1300 cm⁻¹, which are associated with phenylalanine, CH₂, and = CH, which might be attributed to hydrophobic interactions among proteins, lipids, and polysaccharides (Huang et al., 2023). Polysaccharide addition enhanced protein–water interactions, leading to the exposure of hydrophobic groups in HSOB. This interaction enhanced hydrophobic interactions, promoted the unfolding of a tighter network structure at the interface, and induced structural changes in the proteins (Wang et al., 2025). The 540 cm⁻¹ band in Raman spectroscopy is attributed to the stretching vibration of S—S bonds, confirming the formation of disulfide bonds. XG-HSOB and κ-GC-HSOB exhibited stronger peaks, indicating a higher content of disulfide bonds in these inks. In addition, κ-GC-HSOB contained the highest content of disulfide bonds (Fig. 6d), which reinforced the network structure of the ink. This explains its highest storage modulus (G', Fig. 4e and f) and superior 3D printing performance (Fig. 5). Notably, the observed decrease in ion bond following the addition of polysaccharides can be attributed to the anionic nature of all three polysaccharides. Given that the HSOB interface is negatively charged, the formation of ion bonds with these polysaccharides becomes challenging. Therefore, the increase in hydrophobic and disulfide bond likely constitutes the primary factor to affect the texture and rheological properties of the polysaccharide-HSOB inks.

The interaction mechanism between HSOB and three anionic polysaccharides was studied using FTIR (Fig. 6e). Compared to the HSOB, spectral analysis indicated that no new characteristic peaks were observed, suggesting that no new functional groups were induced between HSOB and the three anionic polysaccharides. The bands at 2923 cm⁻¹, 1745 cm⁻¹ and 1160 cm⁻¹ correspond to the stretching vibrations of -CH₂, C O, and C—O groups in the oil, respectively (Cheng et al., 2026; Zhao et al., 2026). The intensities of these bands were lowest in κ-GC-HSOB than other inks, suggesting more effective encapsulation of oil in the former, which is consistent with the observed microstructure (Fig. 2). Liu et al. (2026) found that the C O peaks of soybean oil decreased with increasing concentrations of cyclic dextrin, which was associated with enhanced encapsulation of oil droplets. In addition, the broad peak observed in the 3200–3600 cm⁻¹ region corresponds to O—H stretching vibrations from polysaccharide (Zheng et al., 2025), and high peak intensity indicates strong hydrogen bonds. The κ-GC-HSOB exhibited the strongest peak intensity, followed by HSOB, while GA-HSOB showed the weakest signal, indicating that κ-GC enhanced hydrogen bonding whereas GA reduced it. This result is consistent with the findings presented in Fig. 6d. Therefore, κ-GC more effectively encapsulated HSOB and enhanced hydrogen bonding within the HSOB-based ink, which may be one of the key factors contributing to the strengthened network structure and improved 3D printing fidelity.

The amide I band associated with C O stretching vibrations is highly sensitive to the secondary structure of proteins. Therefore, FTIR analysis was conducted in the range of 1600–1700 cm⁻¹ to characterize the secondary structure of proteins, as shown in Fig. 6f. This range includes intramolecular β-sheets (1612–1642 cm⁻¹), intermolecular β-sheets (1615–1625 cm⁻¹), random coils (1640–1650 cm⁻¹), α-helices (1650–1660 cm⁻¹), β-turns (1661–1690 cm⁻¹), and intermolecular antiparallel β-sheets (Yang, Deng, et al., 2025). The addition of different anionic polysaccharides led to varying changes in the secondary structure conformation of HSOB interface proteins. After the addition of κ-GC, the β-sheets and β-turns of HSOB increased, while the addition of XG resulted in an increase in the random coils of HSOB. The increase of β structure changes the overall conformation of the surface protein from “flexible” to “rigid”, making the gel structure more stable (Yang, Deng, et al., 2025). This is consistent with the observation that κ-GC-HSOB, which exhibited the highest β-sheet content and lowest random coils, also showed a higher storage modulus (G′), greater hardness and superior 3D printing performance. Conversely, after adding GA, the secondary structure of HSOB remained largely consistent. Changes in the secondary structure of proteins affected the molecular interactions within the system, thereby influencing the texture, rheology, and 3D printing characteristics of the inks (Liu et al., 2025).

3.8. IDDSI

The spoon tilt and fork pressure tests were conducted to identify Level 4 and Level 5 dysphagia-friendly foods within the IDDSI framework. Fig. 7 illustrated the results of the IDDSI tests on food items 3D printed with HSOB ink containing various polysaccharides. In the fork pressure test, all samples easily passed through the fork gaps. In the spoon tilt test, HSOB and GA-HSOB smoothly slid off the spoon. This indicated compliance with the Level 4 dysphagia-friendly food standard, suitable for individuals with significantly reduced tongue control. The entire spoon samples of κ-GC-HSOB and XG-HSOB slid off completely with a little residue. This indicated compliance with the Level 5 dysphagia-friendly food standard, allowing for movement of this texture with less chewing or tongue force. However, HSOB retained the most residues on the spoon than other samples, which was relative to high adhesiveness (Table 1). Table 1 also showed that HSOB had the lowest hardness (8.55 gf) and springiness (0.59), potentially leading to an overly soft, mushy mouthfeel and a higher aspiration risk, as the material can easily fragment into small particles during mastication (Hou et al., 2023). XG-HSOB and κ-GC-HSOB had high hardness and low adhesiveness, which balanced swallowing safety and mouthfeel. Similar textural properties and IDSSI results of sesame OB extracted at pH 6.5 were reported by Yang et al. (2026). In summary, different polysaccharides can effectively regulate the IDDSI levels of HSOB-based ink, which is crucial for developing HSOB products for dysphagia management (Yang, Cheng, et al., 2025).

Fig. 7.

The IDDSI tests of HSOB inks with different polysaccharides.

4. Conclusions

This study investigated HSOB-based inks containing polysaccharides (κ-GC, XG, and GA) for extrusion-based 3D printing of dysphagia-friendly foods. The results indicated that polysaccharide type was a key factor to affect the rheological properties, microstructure, and interactions of the inks, thereby changing the 3D printability. The κ-GC-HSOB ink exhibited superior gel strength, high hardness, good self-supporting capability, and minimal deformation during printing. This was attributed to κ-GC promoting the formation of additional disulfide bonds, hydrophobic interactions, and more β-sheet, as well as facilitating the cross-linking network formation by adsorbing onto the HSOBs. GA-induced HSOB collapse resulted in the poor gel structure and 3D printability. IDDSI experiments found that HSOB and GA-HSOB are classified as Level 4 dysphagia foods, while κ-GC-HSOB and XG-HSOB are classified as Level 5 dysphagia foods. This study provides a strategy to enhance the 3D printing performance of oil body-based foods for dysphagia, providing a theoretical basis for developing functional colloidal systems in special dietary products.

CRediT authorship contribution statement

Yanan Yang: Writing – original draft, Visualization, Methodology, Investigation, Formal analysis, Data curation. Ruizhi Yang: Writing – review & editing, Writing – original draft, Visualization, Formal analysis, Data curation. Wentao Wang: Resources, Formal analysis. Yuan Fang: Formal analysis. Luping Zhao: Writing – review & editing, Validation, Supervision, Project administration, Funding acquisition, Conceptualization.

Declaration of competing interest

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

Appendix A. Supplementary data

Supplementary material

Data availability

Data will be made available on request.