Formulation Optimization of GG/SS/PVA/GEL Composite Hydrogels for Extrusion-Based Bioprinting Using Response Surface Methodology

Abstract

Extrusion-based bioprinting requires hydrogel materials with suitable rheological behavior, structural stability, and bio-related properties; however, the relationships among composition, network structure, and printing performance in multicomponent hydrogel systems remain insufficiently understood. In this study, a GG/SS/PVA/GEL composite hydrogel was developed and optimized using single-factor experiments combined with Box–Behnken response surface methodology. Rheological analysis, Fourier transform infrared spectroscopy, and scanning electron microscopy were further used to characterize the optimized system. The optimal formulation was identified as 0.14 g of GG, 0.60 g of SS, and 2.3 g of PVA. This formulation achieved a comprehensive score of 87 with a prediction error of less than 5%. The optimized hydrogel exhibited pronounced shear-thinning behavior, a printing fidelity of 98.6–101.4%, a maximum swelling ratio of approximately 403.5%, and an enzymatic degradation rate of 68.5%, together with a relatively uniform interconnected porous structure. These results indicate that the optimized composite hydrogel is a promising printable material candidate and provide a useful basis for formulation design in extrusion-based hydrogel systems.

1. Introduction

Extrusion-based bioprinting holds significant promise in potential applications, biomanufacturing, and personalized medicine, as it enables the precise fabrication of complex three-dimensional structures under mild conditions [1,2]. However, the advancement of this technology is constrained by the properties of printing materials, especially hydrogel inks. Such materials must exhibit sufficient flowability, maintain structural integrity after deposition, and possess suitable rheological and mechanical characteristics. Achieving a balance among these requirements is challenging, and the relationships between material composition, network structure, and printing performance in multicomponent hydrogels remain poorly understood [3].

Response Surface Methodology (RSM) is a statistical and mathematical tool for modeling and optimizing processes influenced by multiple variables. By quantifying the relationships between independent factors and response variables, RSM allows efficient determination of optimal conditions with fewer experiments. In hydrogel optimization, RSM provides a systematic method to correlate composition with rheological and mechanical properties, supporting the rational design of multi-component hydrogel systems for extrusion-based bioprinting.

Among biopolymers used in functional hydrogels, gellan gum (GG), silk sericin (SS), polyvinyl alcohol (PVA), and gelatin (GEL) have received significant attention for their unique properties and broad applications. GG, a microbial polysaccharide, demonstrates thermo-reversible gelation and structural stability, making it suitable as the primary network-forming component [4,5,6,7]. SS, a protein derived from Bombyx mori silk, exhibits high biocompatibility and adjustable biodegradability, and can regulate swelling and degradation in hydrogel systems [8,9,10]. GEL, derived from collagen, is widely employed in organohydrogels and can provide structural support while enabling functional responsiveness [11,12,13]. PVA, a synthetic polymer, is often incorporated to improve network rheology and mechanical strength. The combination of these four biopolymers leverages their complementary properties to enhance printability, mechanical performance, and swelling behavior, which are critical for extrusion-based bioprinting [14,15].

Despite these advances, major challenges persist. Most current studies rely on empirical adjustments targeting a single performance metric, without systematically considering the coupled effects of multiple functional properties [16,17]. The relationships among composition, network structure, rheological behavior, and overall printing performance in multicomponent hydrogels remain poorly understood [18]. Therefore, a multi-response optimization framework that quantitatively links formulation design with rheological, mechanical, and swelling properties is still lacking.

In this study, GG, SS, PVA, and GEL were strategically combined and optimized using RSM to exploit their complementary mechanical, swelling, and rheological properties. GG serves as the primary network former, SS regulates swelling and degradation, PVA enhances network rheology and mechanical strength, and GEL provides structural support. This systematic multi-response approach not only identifies the optimal formulation but also establishes clear correlations among material composition, network structure, and extrusion-based printing performance, highlighting the novelty of the work and offering guidance for rational design of multi-component hydrogel inks.

2. Materials and Methods

2.1. Materials and Reagents

GG (biotechnology grade, low-acyl type), SS (food grade, 99% purity), GEL (biotechnology grade), anhydrous calcium chloride, and phosphate-buffered saline (PBS; 0.2 M, pH 7.4, sterile and enzyme-free) were obtained from Macklin Biochemical Co., Ltd. (Shanghai, China). PVA (purity 98.0–99.0 mol%, Mw ≈ 190,000) was supplied by Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China).

2.2. Instruments and Equipment

Extrusion-based bioprinter (EFL-BP-6601) was purchased from Suzhou Yong qin quan Intelligent Equipment Co., Ltd. (Suzhou, China). Vacuum drying oven (DZF-6020) was supplied by Changzhou Heng long Instrument Co., Ltd. (Changzhou, China). A high-precision analytical balance (LQ-C5001) was obtained from Ruian Le qi Trading Co., Ltd. (Ruian, China). Digital constant-temperature magnetic stirrer (85-2A) was purchased from Beijing Hongda Heng ye Technology Co., Ltd. (Beijing, China). Rotational rheometer (HR-1) was provided by Shanghai Di guan Industrial Co., Ltd. (Shanghai, China). Universal mechanical testing machine (CMT6104) was obtained from Shandong Drick Instruments Co., Ltd. (Jinan, China). Tabletop scanning electron microscope (TM4000) was supplied by Shenzhen Ke shi da Electronic Technology Co., Ltd. (Shenzhen, China).

2.3. Experimental Methods

2.3.1. Preparation of Hydrogel

The composite hydrogel was prepared through a sequential process involving thermal dissolution, stepwise blending, ionic pre-crosslinking, freeze–thaw treatment, extrusion-based printing, and secondary crosslinking. The contents of PVA, GG, and SS were treated as variable factors in subsequent single-factor experiments and RSM optimization, whereas the concentration of GEL was fixed at 6 wt%.

First, predetermined amounts of PVA and GG were added to deionized water and stirred at 80 °C for approximately 1.5 h until a homogeneous solution was obtained. The mixture was then cooled to approximately 50 °C, followed by the addition of GEL under continuous stirring to ensure complete dissolution. The SS solution was prepared separately by dissolving SS in deionized water at 40 °C for 1.5 h until a homogeneous solution was obtained. The system was further cooled to approximately 40 °C, after which the SS solution was gradually added under continuous stirring to ensure uniform mixing of all components.

After thorough mixing, CaCl₂ was added to achieve a final concentration of 1 wt%, thereby inducing ionic pre-crosslinking of GG and enhancing initial structural stability. The resulting mixture was kept in the beaker and subjected to freeze–thaw treatment. Specifically, the samples were frozen at −20 °C for 10 h and subsequently thawed at room temperature for 5 h. This cycle was repeated five times to promote the formation of a stable physical network. After freeze–thaw treatment, the hydrogel was transferred to the syringe barrel of an extrusion-based 3D printer and maintained at approximately 38 °C prior to printing. Printing was conducted under the conditions described in Section 2.3.2.

After printing, the scaffolds were fully immersed in a 2% (w/v) GA solution and crosslinked at 4 °C for 3 h. After crosslinking, the scaffolds were soaked in deionized water for 15 min and rinsed to remove residual GA, followed by immersion in a 0.1 mol/L glycine solution for 2 h at room temperature to block remaining aldehyde groups. The samples were subsequently rinsed three times with sterile PBS [19,20].

The formulations were expressed in terms of mass ratios. This choice reflects the use of biomacromolecular components such as gelatin and sericin, which possess heterogeneous molecular weight distributions. In addition, mass-based composition facilitates practical preparation and ensures reproducibility in extrusion-based bioprinting.

2.3.2. 3D Printing of Gel

The 3D printer used in this experiment was a pneumatic extrusion-based printer. The printing parameters were set as follows: nozzle travel speed of 10 mm/s, nozzle temperature of 38 °C, nozzle height of 0.5 mm above the platform, platform temperature of 0 °C, extrusion pressure of 20 kPa, filling density of 60%, and an extrusion nozzle outer diameter of 0.31 mm. Based on these parameters, square grid test models with a 1 mm spacing were printed. Three samples were printed for each material formulation. The printed samples were immediately subjected to performance evaluation.

2.3.3. Printing Fidelity Experiment

After scaffold printing was completed, a camera was immediately mounted on a tripod for image acquisition. The aperture, shutter speed, and ISO were uniformly set, and the shooting distance was maintained at 0.3 m to obtain clear, high-resolution images with uniform lighting and no noticeable distortion. The images were analyzed using ImageJ software version 1.52i. First, the scale was calibrated using reference objects around the sample to establish the relationship between pixel size and actual dimensions. The printed grid area was then measured.

In this study, printing fidelity is defined as the ratio of the actual printed area to the theoretical design area and is used to quantitatively characterize the geometric consistency between the printed macrostructure and the design model. This indicator reflects the combined effects of hydrogel ink spreading, structural collapse, and line diffusion on printing fidelity during extrusion-based bioprinting. Because hydrogel inks may spread after extrusion, the actual printed area can, under certain conditions, be slightly larger than the theoretical design area. Therefore, printing fidelity values may exceed 100%, which is consistent with the physical characteristics of extrusion-based hydrogel bioprinting.

2.3.4. Mechanical Properties Testing

Compression tests of cylindrical scaffolds (Φ10 × 10 mm) were performed using an Instron 5565 universal testing machine at a loading rate of 1 mm/min. The stress–strain curves were recorded, and the compressive modulus (E) and yield strength (σ_y) were calculated. Five replicates were tested for each group.

In addition, the compressive stress was calculated as the average stress within the initial strain region (0–15%) of the compression stress–strain curve. This parameter was used to characterize the initial resistance of the scaffold to deformation under compressive loading [21].

2.3.5. Swelling Performance Experiment

Cylindrical hydrogel samples with relatively uniform sizes were prepared, with an average dimension of approximately 1 mm × 2 mm (diameter × height). The samples were dried in an oven until the weight difference between two consecutive measurements was less than 0.005 g, indicating that a constant weight had been reached. The weight at this stage was recorded as W_d. The xerogel samples were then immersed in 150 mL of deionized water, and the temperature was maintained at 37 °C to simulate physiological conditions. The temperature was controlled using a thermostatic water bath with a fluctuation range not exceeding ±2 °C. After each measurement, the samples were gently blotted with filter paper to remove excess surface liquid, ensuring that no visible droplets remained before weighing. The sample weight was measured using an electronic balance until the difference between successive measurements was within 0.05 g. The final weight was recorded as W_t [22].

$$\mathit{Swelling}\mathit{Ratio}={\displaystyle \frac{W_t-W_d}{W_d}}\times 100\%$$

(1)

Deionized water was selected to evaluate the intrinsic swelling behavior of the hydrogel without interference from external ions, which could affect osmotic pressure and network interactions.

2.3.6. Enzymatic Degradation Experiment

The enzymatic biodegradation experiment was conducted by first measuring the initial weight of the hydrogel (W¹), followed by immersion in 0.0006% (w/v) type I collagenase solution in a 24-well plate and incubation at 37 °C for 14 days. The enzyme solution was then removed, and the hydrogel was washed with distilled water to eliminate residual salts within the porous structure. The hydrogel was subsequently weighed to obtain the final weight (W²) [23,24].

$$\mathrm{Enzymatic}\mathrm{Degradation}\mathrm{Rate}={\displaystyle \frac{\left(W^1-W^2\right)}{W^1}}\times 100$$

(2)

2.3.7. Single-Factor Experiment

Single-factor experiments were conducted to investigate the effects of GG, SS, PVA, and GEL on the properties of the hydrogel system.

For the GG experiment, different amounts of GG (0.04, 0.06, 0.08, 0.10, 0.12, and 0.14 g) were added into 50 mL beakers, while the amounts of SS (0.60 g), PVA (2.3 g), GEL (1.2 g), and deionized water (15.8 mL) were kept constant.

For the SS experiment, SS was added at 0.40, 0.50, 0.60, 0.70, 0.80, and 0.90 g, while GG (0.10 g), PVA (2.3 g), GEL (1.2 g), and deionized water (15.8 mL) were fixed.

For the PVA experiment, PVA was varied at 2.0, 2.1, 2.2, 2.3, 2.4, and 2.5 g, while GG (0.10 g), SS (0.60 g), GEL (1.2 g), and deionized water (15.8 mL) remained unchanged.

In all experiments, the mixtures were stirred in a water bath at 60 °C until homogeneous and then transferred to a 3D printer for extrusion-based printing at a constant temperature of 38 °C. The printed samples were subsequently subjected to compression testing and other performance evaluations.

By comprehensively evaluating the print fidelity, mechanical properties, swelling behavior, and enzymatic degradation characteristics, data support is provided for further optimizing the optimal concentrations of GG, SS, and PVA.

2.3.8. Response Surface Experiment

Based on the results of the single-factor experiments, a Box–Behnken design (BBD) was employed for experimental design, data analysis, and optimization. In previous studies on composite hydrogels, such as the work by Labus et al. [25], GEL was used as a fixed matrix component at 10% (w/v), whereas the PVA content was treated as a variable parameter to evaluate its influence on material performance. This approach demonstrates that using GEL as a constant base component is a reasonable and widely adopted experimental strategy. Similarly, in the present study, PVA was selected as the variable parameter, while GEL was maintained at a fixed proportion. Response surface methodology (RSM) was applied to optimize the formulation of the remaining major components. The addition amounts of GG (A), SS (B), and PVA (C) were selected as the primary experimental factors, each evaluated at three levels [26], as shown in

Table 1. Response Surface Design Levels.

Level	A: Gellan Gum Addition	B: Silk Sericin Addition	C: Polyvinyl Alcohol Addition
−1	0.06	0.40	2.1
0	0.10	0.50	2.3
1	0.14	0.60	2.5

. The overall performance of the printed scaffolds was used as the response variable (

Table 2. Comprehensive Performance Evaluation Criteria of the Scaffold.

Indicator (Weight)	Evaluation Criteria	Score (Points)
Printing fidelity (40%)	Shape intact, smooth surface, neat edges. A 98.6% ≤ Printing fidelity ≤ 101.4%	31~40
	Slight deformation, smooth and even surface. A 96.3% ≤ Printing fidelity < 98.6% or 101.4% < Printing fidelity ≤ 103%	11~30
	Incomplete shape, rough and uneven surface. Printing fidelity < 96.3% or Printing fidelity > 103%	0~10
Mechanical Properties (20%)	High load-bearing capacity with strong resistance to deformation. Compressive stress ≥ 0.175 Mpa	11~20
	Balanced support and flexibility with overall moderate performance. A 0.115 Mpa ≤ compressive stress < 0.175 Mpa	6~10
	Soft texture with low resistance to deformation. Compressive stress < 0.115 Mpa	0~5
Bio-related Performance (40%)	Excellent swelling and degradation performance. Swelling ratio ≥ 480%, enzymatic degradation rate ≥ 72.3%	31~40
	Moderate swelling and degradation performance. A 400% ≤ swelling ratio < 480% or 65.4% ≤ enzymatic degradation rate < 72.3%	11~30
	Poor swelling and degradation performance. Swelling ratio < 400%, enzymatic degradation rate < 65.4%	0~10

). Response surface analysis was then conducted to determine the optimal ratio of the three components.

The weights assigned to the evaluation indicators were determined based on the practical performance requirements of extrusion-based bioprinting and previous studies on bioink optimization [27,28,29]. In this scheme, print fidelity and bio-related performance were each assigned weights of 40%, whereas mechanical properties were assigned a weight of 20%. These weights reflect the relative influence of each property on scaffold functionality, as supported by prior studies showing that print fidelity critically affects structural accuracy and cell distribution, whereas bio-related properties, such as swelling behavior and enzymatic degradation, determine material adaptability and long-term performance [29]. Mechanical properties, although essential for scaffold stability, typically have a secondary impact compared with print fidelity and bio-related performance under standard extrusion-based conditions [27].

The threshold values for each evaluation range in Table 2 were defined based on the statistical distribution of single-factor experimental results. These values were further classified according to the comprehensive requirements of shape fidelity, structural stability, and biologically relevant performance in extrusion-based bioprinting, thereby establishing a robust multi-performance evaluation system.

Multi-criteria evaluation and optimization methods based on RSM have been widely applied to hydrogel formulations and related material systems. Through experimental designs such as the Box–Behnken design, statistical models can describe the relationships between independent variables and multiple response variables, enabling the prediction of optimal conditions. For example, Leanpolchareanchai [30] applied RSM to optimize multiple performance indicators in hydrogel microneedle systems, demonstrating the effectiveness and reliability of this approach for multifactorial, multi-response material optimization.

2.3.9. Microstructural Characterization

After low-temperature curing, the rectangular hydrogel samples were immersed in a 2% (w/v) GA solution for crosslinking for 2 h. The samples were then pre-frozen in liquid nitrogen and freeze-dried using a freeze dryer (ALPHA 2–4 LD plus, Martin Christ GmbH, Osterode, Germany) at −80 °C for 24 h. The freeze-dried samples were fractured in liquid nitrogen, sputter-coated with gold, and then observed using a field-emission scanning electron microscope (SEM, TM4000, Hitachi Ltd., Tokyo, Japan) at an accelerating voltage of 15 kV.

2.3.10. Chemical Structure Characterization

Fourier transform infrared (FTIR) spectra were recorded using an FTIR spectrometer (Nicolet iS10, Thermo Fisher Scientific Inc., Waltham, MA, USA). The spectra were collected in the range of 500–4000 cm⁻¹ with a spectral resolution better than 0.4 cm⁻¹.

2.3.11. Rheological Property Testing

The sample’s viscosity as a function of shear rate was measured using a rheometer at 38 °C, corresponding to the printing temperature. Oscillatory frequency sweep tests (0.1–100 rad/s) and temperature-dependent oscillatory frequency tests (10–50 °C) were also performed. The storage modulus (G′) and loss modulus (G″) were recorded to evaluate the material’s shear-thinning behavior and structural recovery.

2.3.12. Scaffold Printability Evaluation

To assess the printability and shape fidelity of the composite hydrogel scaffolds, a regular grid model was selected as the test object (Figure 1). This model was chosen because its well-defined struts and pore units enable evaluation of extrusion continuity, filament formation, and layer stacking stability. In addition, regular grid structures are among the most commonly used models in extrusion-based printing, facilitating comparison of different material systems in terms of shape fidelity, pore structure controllability, and multilayer formation capability. These characteristics make the grid model a useful tool for assessing the potential of the composite hydrogel for scaffold construction [31].

2.4. Data Analysis

Data processing and graphing were conducted using Origin 2021, printing data were extracted using ImageJ software version 1.52i, and response surface analyses and graphical representations were performed with Design-Expert software version 13.0. All experiments were performed at least in triplicate unless otherwise stated, and the results are presented as mean ± standard deviation. For mechanical testing, five replicates were used for each group.

3. Results

3.1. Single-Factor Experiment Results

3.1.1. GG Addition Amount

Figure 2 illustrates the effect of GG addition on the gel’s printing fidelity and compressive stress. As GG content increases, printing fidelity first rises and then decreases, indicating an optimal network density that balances chain segment mobility and gel rigidity. At low GG content, the gel network lacks sufficient structural support, leading to partial collapse and reduced printing fidelity. Conversely, excessive GG content results in a denser network, restricting molecular chain mobility, reducing flowability, and compromising printing stability. Under the conditions of this study, adding 0.10 g of GG achieves an optimal balance between the material’s flow behavior and structural recovery, resulting in maximal printing fidelity.

Increasing GG content also enhances compressive stress and modulus until an optimal point is reached. Beyond this point, excessive GG causes rigidity that hinders smooth extrusion. This effect is attributed to the coil–helix transition of GG molecules in water, followed by their assembly into double-helix structures via intermolecular hydrogen bonding, forming a cation-induced physically crosslinked network [32]. The trends in compressive modulus and yield strength mirrored those of compressive stress, indicating that compressive stress effectively represents the overall mechanical load-bearing capacity.

Therefore, compressive stress was chosen as the representative mechanical parameter for the subsequent comprehensive performance evaluation.

Figure 3 illustrates the effect of GG content on the swelling ratio and enzymatic degradation rate of the gel system. As GG content increased, the swelling ratio first increased, reaching a maximum of 403.5% at 0.10 g, and then decreased. This trend is attributed to the ability of optimal GG content to modify the gel network, increasing pore number and size, thus facilitating water penetration into the gel matrix. Excessive GG results in an overly crosslinked, dense network that restricts water diffusion, thus reducing the swelling ratio.

Similarly, the enzymatic degradation rate was significantly affected by GG content, following a similar increase–decrease trend with a maximum at 0.10 g. Optimal GG content improves the gel network, increasing the contact area between enzymes and substrates, thus promoting enzymatic degradation. At GG content, insufficient network formation limits enzyme penetration, leading to a lower degradation rate. In contrast, excessive GG causes over-crosslinking, hindering enzyme diffusion into the gel interior. These observations align with previous studies showing that both insufficient network formation and high crosslinking density limit enzyme access and reduce enzymatic degradation in polysaccharide-based hydrogels [33,34,35].

3.1.2. SS Content

SS is a natural protein with good biocompatibility. In composite hydrogel systems, SS mainly contributes to network formation through intermolecular interactions. Figure 4 illustrates the effects of varying SS content on the printing fidelity and compressive stress of the gel system. As SS content increased, printing fidelity remained largely unchanged, suggesting that within the studied range, SS has a limited direct influence on material flow during extrusion. This observation indicates that the macroscopic stability of extrusion-based shaping is primarily governed by the physical network formed by GG and PVA, while SS serves mainly as a supplementary regulatory role [36].

The effect of SS content on compressive stress first increased and then decreased. At an optimal concentration, SS molecular chains form hydrogen bonds with GG and PVA, creating a synergistic interaction that enhances network integrity, improves stress transfer, and increases gel compressive stress. However, excessive SS may introduce additional protein molecules, causing local network heterogeneity, reducing effective crosslinking density, weakening load-bearing capacity, and ultimately decreasing compressive stress [37].

Figure 5 illustrates the effects of varying SS content on hydrogel properties relevant to biological applications, including swelling behavior and enzymatic degradation. With increasing SS content, both the swelling ratio and the enzymatic degradation rate initially increased and then decreased. Incorporation of an optimal amount of SS renders the gel network more open and organized, creating additional pathways for water and enzymes to penetrate, thereby enhancing swelling and promoting enzymatic degradation. At low SS content, the network structure is insufficiently developed and lacks structural integrity, limiting enzyme penetration and resulting in a lower degradation rate. Conversely, excessive SS produces a denser network that restricts water diffusion and enzyme penetration, leading to reduced swelling and enzymatic degradation. These results are consistent with previous studies demonstrating that polymer network density and connectivity strongly influence enzymatic degradation in multi-component hydrogel systems [33,34].

3.1.3. PVA Content

PVA, a hydrophilic polymer, plays a key role in regulating the structural stability and rheological properties of composite hydrogels. Figure 6 illustrates the effects of varying PVA content on the printing fidelity and compressive stress of the gel system. As PVA content increased, printing fidelity gradually improved and tended to exceed the theoretical value at higher concentrations. This trend aligns with previous reports showing that moderate PVA incorporation enhances filament continuity and shear-thinning behavior, improving spreading during extrusion, whereas excessive PVA may cause chain entanglement and non-uniform flow [38,39].

Regarding compressive stress, the gel system first increased and then decreased as PVA content rose. At moderate PVA content, hydrogen bonding among PVA chains and between PVA, GG, and SS strengthens the network, enhancing load-bearing capacity and compressive stress [39]. At excessive PVA content, although chain entanglement increases, uneven polymer distribution induces stress concentration, impairing the network’s ability to dissipate stress. Consequently, compressive stress decreases, and the gel becomes more brittle [38]. Therefore, optimal PVA content improves network integrity, maximizing compressive stress and enhancing compression resistance.

Figure 7 illustrates the effects of varying PVA content on hydrogel properties relevant to biological applications, including swelling behavior and enzymatic degradation. As PVA content increased, the gel’s swelling ratio gradually decreased and stabilized at higher concentrations. This behavior is attributed to increased molecular entanglement and crosslinking of PVA chains, which restrict water diffusion into the hydrogel network [40]. Similarly, the enzymatic degradation rate decreased with increasing PVA content, showing a more pronounced decline at higher concentrations. Excessive PVA likely hinders enzyme penetration into the network and limits interaction with the substrate, thereby suppressing degradation. Similar effects of polymer network density on enzyme diffusion and activity have been observed in multi-component hydrogels, where denser or more entangled networks restricted enzyme mobility and slowed degradation [33,40].

3.2. Response Surface Experimental Results and Analysis

3.2.1. Optimization Results and Analysis of Response Variables

An optimization experiment was conducted using Design-Expert software version 13.0 based on the Box–Behnken design. The addition levels of GG, SS, and PVA were selected as response surface factors, while printing fidelity, mechanical performance, and bioapplication-related properties served as evaluation indicators. Seventeen experimental runs were conducted in total. A regression model was established to describe the relationship among the three variables: GG content (A), SS content (B), and PVA content (C). The resulting regression model was expressed as: y = 80.2 − 0.375A + 1.5B + 0.125C − 1.75AB − 1.5AC + 3.65A² + 0.4B² − 2.35C², with R² = 0.9014 (

Table 3. Optimization Experimental Results and Analysis.

Run No	A(g)	B(g)	C(g)	Comprehensive Performance Evaluation of the Scaffold (Score)
1	0.06	0.40	2.3	77
2	0.14	0.40	2.3	85
3	0.06	0.60	2.3	78
4	0.14	0.60	2.3	87
5	0.06	0.50	2.1	75
6	0.14	0.50	2.1	82
7	0.06	0.50	2.5	79
8	0.14	0.50	2.5	86
9	0.10	0.40	2.1	79
10	0.10	0.60	2.1	86
11	0.10	0.40	2.5	79
12	0.10	0.60	2.5	80
13	0.10	0.50	2.3	83
14	0.10	0.50	2.3	79
15	0.10	0.50	2.3	81
16	0.10	0.50	2.3	82
17	0.10	0.50	2.3	80

).

Analysis of variance (ANOVA) and significance testing of the regression model coefficients were performed. The results revealed an F value of 1.83 and a p value of 0.0005 (p < 0.001), indicating that the regression model is highly significant. The lack-of-fit term exhibited an F value of 4.14 (>0.01) and a p value of 0.1018 (>0.05), which was not statistically significant, indicating that the model fit the experimental data well. Further evaluation of the quadratic regression equation indicated an R² value of 0.9014, showing that the model accounts for 90.14% of the variation in response values, demonstrating its practical relevance. Therefore, the model demonstrates good fit and high reliability, and the regression equation can be used to predict the optimal formulation of the GG–SS–PVA 3D printing gel (

Table 4. Regression Model and Analysis of Variance Table.

Source of Variance	Sum of Squares (SS)	Degrees of Freedom (DF)	Mean Square (MS)	F-Value	p-Value	Statistical Significance
Model	123.45	9	13.72	1.83	0.0005	Statistical significance
Amount of A-type GG Added	1.13	1	1.13	0.1499	0.0002
Amount of B-type SS Protein Added	18	1	18	2.4	0.1654
Amount of C-type PVA Added	0.125	1	0.125	0.0167	0.0014
AB	12.25	1	12.25	1.63	0.3422
AC	9	1	9	1.2	0.0001
BC	6.25	1	6.25	0.8325	0.2919
A2	56.09	1	56.09	7.47	0.0002
B2	0.6737	1	0.6737	0.0897	0.1732
C2	23.25	1	23.25	3.1	0.0068
Residual	52.55	7	7.51
Lack of Fit Error	39.75	3	13.25	4.14	0.1018	Not statistically significant
Pure Error	12.8	4	3.2
Total	176	16

).

3.2.2. Analysis of the Interaction Effects Between Two Factors

The significance test for the regression model coefficients revealed p-values for the linear terms as follows: PA = 0.0002, PB = 0.1654, and PC = 0.0014, PA, PB, and PC represent the effects of factors A (GG), B (SS), and C (PVA). Since PA and PC are both less than 0.01, GG content (factor A) and PVA content (factor C) significantly influence the comprehensive performance evaluation of the scaffold. Among the quadratic terms, A² and C² also showed highly significant effects, and the interaction term AC was significant, indicating that the combined effect of GG and PVA contents plays a crucial role in scaffold performance.

As shown in Figure 8, the contour lines are denser along the GG content axis (factor A) than along the SS content axis (factor B), indicating that GG content exerts a greater influence on the scaffold’s overall performance than SS content. Additionally, the response surface exhibits a clear curved shape, suggesting an interaction between the two factors. As GG content increases, the scaffold’s performance score initially rises and then levels off, while increasing SS content at a constant GG level also improves performance. These observations align with previous studies using response surface methodology to analyze factor contributions and interactions in multi-component hydrogel systems.

As shown in Figure 9, the response surface and contour plots reveal a strong interaction between GG content (A) and PVA content (C). The curved, twisted shape of the response surface indicates that these two factors jointly influence scaffold performance. The contour lines are denser along the GG axis than along the PVA axis, suggesting that GG has a stronger effect. When both factors vary simultaneously, the scaffold evaluation score first increases and then decreases, forming a peak that corresponds to an optimal combination. This is consistent with RSM theory, where quadratic and interaction terms can produce non-linear response surfaces and predict optimal factor combinations.

Figure 10 shows that the response value changes nonlinearly with variations in SS (B) and PVA (C) content. Within a certain range, increasing both factors initially improves scaffold performance. However, once the combination exceeds a critical threshold, the response value decreases. The contour lines are denser along the PVA axis than along the SS axis, indicating that PVA has a greater influence than SS. These trends align with RSM analyses that quantify both main effects and factor interactions in hydrogel optimization.

Based on the regression equation analysis, the influence of each factor on the scaffold’s overall performance is ranked as follows [41]: GG content (A) > PVA content (C) > SS content (B). This ranking is supported by the response surface and contour analyses described above and aligns with RSM principles for multi-factor optimization.

The printing parameters used in the validation tests were consistent with those in the single-factor experimental stage: nozzle travel speed of 10 mm/s, extrusion pressure of 20 kPa, nozzle temperature of 38 °C, and platform temperature of 0 °C. Using Design-Expert software version 13.0, the predicted optimal formulation for the GG-SS-PVA 3D printing gel was A-0.14 g, B-0.60 g, C-2.3 g, with a predicted value of 87. The hydrogel was prepared using the optimal process and evaluated comprehensively. The absolute deviation between the predicted and experimental values was less than 5%, indicating that the binomial equation model provides good predictive performance and can serve as a reference for optimizing GG-SS-PVA 3D printing gels (

Table 5. Validation of the optimized formulation (n = 3).

Run No	A/g	B/g	C/g	Predicted Value	Experimental Value	Deviation (%)
1	0.14	0.60	2.3	87	87	0
2	0.14	0.60	2.3	87	86	−1.1
3	0.14	0.60	2.3	87	89	+2.3

Subsequent validation experiments were performed based on the optimized sample (GG-0.14 g, SS-0.60 g, PVA-2.3 g).

).

3.2.3. Validation Experiment

FTIR Chemical Characterization

The FTIR spectra of GG, SS, GEL, PVA, and the composite hydrogel (GG/SS/GEL/PVA) are presented in Figure 11. Analysis of the spectra of the starting materials enables identification of the characteristic functional groups present in the composite hydrogel.

A broad absorption band in the range of 3200–3400 cm⁻¹ is generally attributed to the stretching vibrations of hydroxyl (–OH) and amide (–NH) groups. This band primarily arises from hydroxyl groups in PVA and amide groups in GEL and sericin. Hydroxyl groups within the polysaccharide structure of GG may also contribute to this absorption [42,43,44,45].

The absorption peak at approximately 2900 cm⁻¹ is commonly attributed to C–H stretching vibrations, primarily originating from –CH₂ groups in PVA chains and C–H bonds in the backbones of proteins and polysaccharides [44,45].

For the protein components (GEL and SS), characteristic amide bands appear at approximately 1650 and 1540 cm⁻¹, corresponding to amide I (C=O stretching) and amide II (N–H bending and C–N stretching), respectively. These bands are characteristic spectral features of proteins and are widely used to characterize protein structures [43,44].

The absorption band at approximately 1400 cm⁻¹ is generally attributed to the symmetric stretching of carboxylate groups (–COO⁻), primarily originating from glucuronic acid residues in the GG structure. These carboxyl groups enable GG to interact with multivalent cations and form ionically crosslinked gel networks [42].

An absorption band at 1080–1100 cm⁻¹ is commonly associated with C–O and C–O–C stretching vibrations, attributed to C–O bonds in PVA and glycosidic linkages in the polysaccharide backbone of GG [42,45].

Compared with the spectra of the individual components, the composite hydrogel exhibits a broadened band in the 3200–3400 cm⁻¹ region, indicating the formation of strong hydrogen-bonding interactions among hydroxyl and amide groups. Changes in the carboxyl-related spectral region indicate that GG carboxyl groups may participate in Ca²⁺-mediated ionic crosslinking. Moreover, variations in PVA-related bands may be associated with the formation of crystalline domains during the freeze–thaw process, which serve as physical crosslinking points within the hydrogel network [42,45].

Overall, the FTIR spectra of the composite hydrogel retain the characteristic absorption peaks of GG, SS, GEL, and PVA. Changes in peak shape and position indicate significant intermolecular interactions among the components. Together with Ca²⁺-induced ionic crosslinking of GG and freeze–thaw-induced crystallization of PVA, these interactions contribute to the formation of a physically crosslinked, multicomponent hydrogel network [42,45].

Rheological Property Testing

Rheological analysis revealed that the composite system exhibited pronounced shear-thinning behavior, facilitating smooth extrusion and good flowability. At low shear rates, viscosity was relatively high, whereas at higher shear rates, it decreased rapidly, approaching near-zero values, thus enabling precise macroscopic scaffold construction (Figure 12).

Oscillatory frequency scans showed that at the printing temperature of 38 °C, the storage modulus (G′) was significantly higher than the loss modulus (G′′) across the entire frequency range, with minimal frequency dependence, indicating the formation of a stable solid-like network (Figure 13).

Temperature-sweep rheological tests showed that within the range of 10–50 °C, the storage modulus (G′) remained consistently higher than the loss modulus (G″), indicating that the system maintained a stable solid-like structure without a noticeable gel–sol transition. These results suggest that the hydrogel retains its state within this temperature range, making it suitable for subsequent printing and processing (Figure 14).

Scanning Electron Microscopy (SEM) Analysis

SEM was used to characterize the microstructure of the composite hydrogel. As shown in Figure 15, the optimized sample exhibits a well-defined, three-dimensional porous network with a sponge-like morphology. The pores are interconnected, ranging in size from several tens to several hundred micrometers, and the pore walls are relatively rough and uniformly distributed throughout the structure. The formation of this porous structure is mainly attributed to the ionically crosslinked network formed by GG in the presence of Ca²⁺, along with hydrogen bonding interactions among PVA, SS, and GEL. This interconnected porous architecture may facilitate water transport and mass diffusion, and has been reported to enhance the penetration and transport of molecules in hydrogel systems [46]. Overall, the composite hydrogel exhibits a well-defined, interconnected three-dimensional porous network, consistent with observations in other polysaccharide- and protein-based composite hydrogels.

Scaffold Printing and Secondary Crosslinking Evaluation

The final printed structures are shown in Figure 16. The optimized composite hydrogel was smoothly extruded under the defined printing parameters. The single-layer grids were well-defined, with clear filaments and intact structures. In three- and five-layer prints, the scaffolds maintained structural stability, strong interlayer adhesion, and uniform pore distribution, meeting the requirements for multilayer printing [27]. Overall, the material demonstrates excellent printability, high shape fidelity, and robust structural stability, indicating its potential for fabricating complex three-dimensional hydrogel scaffolds.

After secondary crosslinking with GA, the GG–SS–PVA–GEL 3D-printed hydrogel scaffold exhibited a yellow coloration (Figure 17A, secondary crosslinked product; Figure 17B, stretching behavior; Figure 17C, shape recovery). Mechanical testing demonstrated that the scaffold maintained structural integrity after repeated stretching and exhibited good shape recovery, with no significant permanent deformation, indicating excellent elastic resilience. These observations are consistent with previous reports showing that glutaraldehyde crosslinking enhances the mechanical stability of hydrogels while preserving elasticity [47].

4. Discussion

Previous studies, such as Hu et al., 2025 [28] and Torres-Ayala et al., 2025 [48], primarily focused on single-property optimization. Specifically, Hu et al. optimized the rheological and printability properties of alginate–xanthan gum composite hydrogels, whereas Torres-Ayala et al. evaluated self-supporting capacity and shape fidelity. These studies made significant progress in extrusion-based bioprinting.

In this study, single-factor experiments combined with the Box–Behnken response surface methodology were employed to optimize multiple properties of GG/SS/PVA/GEL composite hydrogels, including print fidelity, mechanical properties, and bio-related performance. The results showed that the optimized hydrogel exhibited excellent performance across all evaluated indicators. Subsequent validation experiments further confirmed the accuracy of the predictions, providing a reliable reference for composite hydrogel design. This multi-property optimization approach provides a valuable strategy for developing high-performance composite hydrogels for extrusion-based bioprinting applications.

5. Conclusions

The optimal composition of GG/SS/PVA/GEL composite hydrogels was predicted using single-factor experiments combined with the Box–Behnken response surface methodology and subsequently validated experimentally. The results indicated that the optimal formulation consisted of 0.14 g of GG, 0.60 g of SS, and 2.3 g of PVA. Validation experiments showed a prediction error of less than 5%, confirming the reliability of the model. The optimized hydrogels exhibited excellent print fidelity, mechanical properties, and bio-related characteristics, enabling high-precision fabrication of multilayered structures with well-defined pore architecture and robust elastic recovery.

Overall, this composite system shows potential for applications in tissue engineering scaffolds, drug delivery carriers, and customized three-dimensional bioprinting. However, this study has not evaluated cytocompatibility or performed in vitro functional assessments. Future work could focus on validating cytocompatibility and extending the optimization approach to other multicomponent hydrogel systems, with the aim of developing high-performance, biocompatible, and scalable hydrogels for extrusion-based bioprinting applications.