3D Printing Colloidal Gels: Navigating the Printability Barrier

Overview of Printability of Biomaterials

Additive manufacturing, or 3D printing, is a pivotal technology applied across domains from health care to electronics. 3D printing builds objects in a controlled, layer-by-layer fashion, minimizing waste and shortening fabrication time as opposed to subtractive and formative manufacturing. This technique enables the creation of intricate structures from polymers, ceramics, metals, and even biological materials. Common techniques include extrusion, inkjet, stereolithography, and laser-assisted printing, each offering distinct advantages in terms of resolutions and material compatibilities. ¹ Among these, extrusion-based 3D printing is the most prevalent due to its adaptability to a wide range of materials, especially for tissue engineering applications.^2,3 Ideal materials for extrusion-based 3D printing are those that flow under high shear or exhibit shear-thinning behavior and possess sufficient viscosity to retain the printed shape postprinting. ⁴ The convergence of 3D printing and tissue engineering has led to the emergence of bioprinting, a subfield focused on developing materials that meet these viscoelastic requirements while incorporating bioactive components essential for tissue engineering applications.

Hydrogels are hydrophilic polymer networks capable of absorbing large volumes of water that are well-suited for bioprinting because of their tissue-like swelling behavior. ⁴ While their biocompatibility has been well characterized in molded or injectable forms, their inherently low shape fidelity has hindered broader 3D printing applications. To overcome this, researchers have modified hydrogels to enable crosslinking, thereby bolstering shape stability postprinting. In our previous work, we demonstrated that naturally derived nanoparticle colloidal hydrogels are a promising bioprinting ink, highlighting the potential of colloidal hydrogel platforms to enhance tissue engineering applications.^5,6 Unlike continuous polymer networks, colloidal gels form through the self-assembly of particles in suspension, driven by weak attractive forces, such as Van der Waals forces and electrostatic interactions. ⁷ The resulting network exhibits shear-thinning flow with self-healing, viscoelastic properties that facilitate extrusion printing and can serve as quantitative predictors of printability.^8,9 With this introduction of colloidal gels as well as other novel hydrogel inks as bioprinting inks, there lacks the literature background necessary to determine optimal printing parameters.

Design of experiments (DOE), including full factorial designs, has been used to identify the primary factors affecting printability and to guide the development of related material systems. ¹⁰ However, DOE methods require large datasets, consuming significant time and resources. To overcome these challenges, data-driven machine learning models have been employed to optimize printability metrics on materials like gelatin methacryloyl (GelMA), ¹¹ CELLINK start gel, ¹² and granular hydrogels, ¹³ examining variables including composition, temperature, extrusion pressure, and print speed. However, an underlying assumption of these approaches is that a fixed parameter set will yield reproducible results. Here, a colloidal gel was selected as a novel ink platform to show that dynamic variables, such as elapsed time and batch-to-batch variability, also influence print fidelity and should be considered in other hydrogel printing optimization strategies. This work shows how print fidelity can be impacted by time, represented by scaffold count, and batch-to-batch variability, reflected in the range of printing parameters and material response across different ink batches.

Exploring the Printability of Colloidal Gels

Representative Effects of Colloidal Gel Printability

Necessary changes to printing parameters throughout the printing process

The BioAssembly Bot 400 allows changes to the printing parameters, namely pressure and speed, while printing, which take effect on the next layer. This allows an experienced user to respond to changes in shape fidelity while printing by either adjusting the pressure, speed, or pausing the print to clear a clogged needle. These changes in shape fidelity and the response by the user were documented within annotated videos of various ink batches. Supplementary Video S1 shows the annotated video displaying these changes for one ink. These changes to the print parameters and shape fidelity from Supplementary Video S1 were summarized within Figure 1A. From this ink batch, it is noted that as time or the number of printed scaffolds increased, an increase in printing pressure is needed to maintain the same shape fidelity. Time is represented by the scaffold count based on how long it takes to print six scaffolds or a full PD. The time scale for printing a full PD is on average 25 min per PD, varying based on the print speed used. All scaffolds from one ink batch were printed within the same day, and each petri was printed back-to-back, with 15 min of rest in between. With this set-up, the scaffold count is continuous over time and can serve as a representation of time. Figure 1B further illustrates how scaffolds with different printing parameters can result in the same printing outcome and how the same printing parameters can result in different shape fidelity.

OPEN IN VIEWER

FIG. 1.

User adjustments to printing parameters (speed and pressure) during the 3D printing process to maintain the fiber diameter for one ink. (A) An overview of the changes to the pressure and speed necessary to print 30, 10 × 10 mm scaffolds with a colloidal ink and their resulting average fiber diameter (mm). The average fiber diameter is denoted by color and the size of the symbol. (B) Representative images of scaffolds throughout the printing session to demonstrate how the quality of the scaffolds can change with or without changes to the printing parameters. S# represents the printed scaffold number from Figure 1A. The scale bar shown is 2 mm and applies to all images.

Material responses to the 3D printing process

This constant adjustment to the printing parameters has not been previously reported and raised concerns about changes to the material’s printability caused by the printing process. To assess the material’s response to the printing process, images of the ink cartridge were collected, and the ink’s viscoelastic properties were measured before and after printing. Cartridge images revealed a consistent decrease in piston height between printing each PD, indicating uniform extrusion as seen in Figure 2. Notably, the pneumatic print head compresses the ink during printing but may cause the piston to retract above the ink level once printing the PD is completed. The ink appeared visually homogenous throughout within images in Figure 2 and the video seen in Supplementary Video S2, with no evidence of phase separation between the nanoparticles and the suspension medium. Despite this, rheological measurements showed noticeable changes in material properties. As shown in Figure 3A, viscosity increased after printing across multiple ink batches. Additionally, comparison of the change of the storage and loss moduli during the printing revealed batch-dependent variation in differences seen in Figure 3B. While some batches showed minimal difference, others exhibited a significant increase in storage modulus postprinting, suggesting material properties are impacted during the printing process.

OPEN IN VIEWER

FIG. 2.

Illustrating the change in the colloidal gel ink within the print cartridge after 3D printing 6 scaffolds within 1 petri dish (PD) with a pneumatic extrusion-based system. Line-up of the ink cartridge, showing the change in ink and piston position after printing 30 scaffolds and waiting 15 min in between each PD.

OPEN IN VIEWER

FIG. 3.

Rheological evaluation of the viscoelastic properties of the GNP colloidal gel inks before and after printing. (A) Flow ramp showing the change in viscosity under high shear rate. (B) Difference of storage and loss moduli before and after printing. Error bars show means ± s.d. for n = 4.

Discussion/Perspective

Extrusion-based 3D printing enables the fabrication of complex structures using diverse materials and has become a powerful tool in tissue engineering. Colloidal gels, formed by the self-assembly of nanoparticles, exhibit promising viscoelastic properties, including shear-thinning and self-healing, making them strong candidates for 3D printing. Before printing, it is required to tune the printing parameters of any material to achieve optimal printability by using DOE or machine learning methods. This study explored how the printing process itself influences the properties of colloidal gels and, ultimately, the printability of the gels that is not considered during the initial tuning studies. Maintaining shape fidelity requires continual adjustments to the pressure and speed during the printing process (Fig. 1A) for several separate batches of colloidal gel inks (Supplementary Figs. S3, S4, S5 and S6). In addition, there is notable variation in the range of pressures needed to extrude a uniform fiber between ink batches. For instance, two ink batches printed well at a pressure of 4 psi, while another achieved similar fidelity at only 1.5 psi. Such variability between inks limits the direct transferability of optimized parameters between batches and undermines the assumption commonly made in DOE and machine learning approaches that a single parameter set yields a consistent print outcome. This inconsistency is often simplified as a “printing range” in literature, which highlights a broader challenge across printing colloidal gel platforms.

To investigate the source of this variability, we measured the change in the viscoelastic properties of the colloidal gel inks before and after printing. Results showed a general increase in storage modulus postprinting, indicating a rise in material stiffness. This shift likely explains the need for increased pressure during printing. We hypothesize that this change arises from a pressure-induced phase separation within the GNP colloidal gel ink, leading to a GNP concentration gradient in the ink cartridge. Phase separation during printing may cause a nonuniform distribution of water leaving the syringe, resulting in progressively less water remaining in the syringe. This process creates a nanoparticle gradient, with a higher particle concentration accumulating toward the end of printing potentially causing aggregation and precipitation of GNPs. Such precipitation is expected during 3D printing, given that hydrated gelatin has a higher density (1.2 g/mL) than water, which explains why water exits the syringe more rapidly. The resulting increase in GNP concentration is expected to raise the viscosity and storage modulus of the ink, as previous work has shown both properties to be dependent on GNP concentration. ⁵ This subtle change in material properties would be visually undetectable to users (Fig. 2) but results in fluctuations in fiber diameter, print uniformity, and storage modulus.

These findings emphasize the importance of documenting real-time parameter adjustments and printing conditions in bioprinting studies. Variations in print performance should not be dismissed as noise but recognized as fundamental behavior of the material and printing system. This work aims to encourage the gathering of diverse data, including dynamic factors and captured data, to further improve the predictive strength of optimization processes. Capturing supporting data such as photographs and videos during printing (Supplementary Videos S1 and S2) can enrich datasets, enabling more accurate models that link material properties and printing parameters with outcome print quality. Thus, a deeper understanding of colloidal biomaterial behavior in extrusion-based 3D printing is essential for developing adaptive, intelligent biofabrication strategies.