Design and Fabrication of In-House Nozzle System to Extrude Multi-Hydrogels for 3D Bioprinting Process

The process of 3D bioprinting is widely acknowledged for creating intricate models tailored to individual patients, utilizing a variety of biomaterials that contain living cells. [1]. As a developing method for tissue engineering, this approach is gradually approaching the accurate replication of tissue-specific microarchitecture. Compared to inkjet and laser-based 3D bioprinting techniques, the extrusion-based method enables the placement of a range of biomaterials containing a greater proportion of cells. [2]. Due to their biocompatibility, low cytotoxicity, and high water content, natural hydrogels are promising options as bio-ink, which is a biomaterial that contains living cells (<90%) [3]. However, because of their low mechanical strength and crosslinking rate, only a few of them are typically utilized for bio-ink preparation [4]. In order to expedite the tissue regeneration process, it is necessary for different types of cells to interact successfully [5]. Thus, creating scaffolds that incorporate multiple materials encapsulating various types of cells has the potential to mimic the architecture of native tissue and take tissue regeneration efforts to the next level [6]. Various attempts have been made to construct scaffolds with multiple materials. For example, a multi-head bioprinting technique was utilized to fabricate scaffolds with chondrocytes and osteoblasts cells, as well as poly (ε-caprolactone) (PCL) and alginate [7]. Despite the multi-head system's capability to release multiple materials, it could only release a single type of material within a specific cross section. Consequently, it lacked the ability to achieve the controlled release of different materials within a single filament. In another effort, the ability to print hydrogel structures that are heterogeneous and multifunctional was demonstrated by adjusting process and material-related parameters such as chemical, electrical, mechanical, and biological properties [8]. However, the authors used an electrospinning system, and a similar effort was not addressed for the extrusion-based bioprinting technique. In order to include both suppleness and muscle growth on the one end, as well as stiffness and tendon development on the other end, a multi-head bioprinter was used to print polyurethane [9] with C2C12 cell and PCL with NIH/3T3 cell [10]. Other studies have documented the use of multiple print heads for constructing scaffolds made up of various materials [11–14]. Although the multi-head system had the ability to dispense multiple materials, it was limited to releasing only one type of material within a particular area. As a result, it was unable to achieve precise control over the release of different materials within a single filament. Nozzle and overall process designs for continuous fiber reinforcement for 3D-printed scaffold were reported [15–17]. However, the user-defined controlled distribution of matrix and reinforced continuous fiber was not identified in that research. A novel flow-focusing nozzle was developed to observe the material distribution controllability, numerical simulation, and finite element analysis having an intention to achieve better generation of tissue and organ structure. This work mainly focuses on the nozzle structure such as the relation between lead length, inner cavity diameter, and outer cavity diameter [18]. The printed structure with material distribution was not reported in this article. Moreover, this research did not include the conical nozzle structure in the simulation and experimental parts. In another work, even though the nozzle design for extruding plant-derived compounds was addressed, the simulation for identifying temperature and pressure distribution was not mentioned [19]. With the intention to achieve the ability to continuously mix non-Newtonian fluids at the microscale level, a new correlation between the size of mixers and their process parameters has been developed and tested. These correlations establish a foundation for designing active microfluidic mixers or print heads capable of effectively blending various substances by ensuring efficient homogenization and extruding them [20]. Even this research did not mention the systematic localization of two distinct materials at user-defined percentages without mixing, a systematic selection of extrusion without operating the rotating impeller can achieve that. Our proposed easy-to-switch material through nozzle connectors with a range of angles may provide a straightforward solution to this issue. For example, if a user seeks fabricate a scaffold to attain a distinct cross-sectional distribution of a filament wherein a certain material comprises a specific percentage while the other material remains absolutely separated, our novel system has the capacity to fulfill this requirement as shown in Fig 2(c). With proper process parameters change, the change of material distribution can be dynamic and filament with gradient can be achieved as shown in Figs. 15 and 17.

In another recently reported research, a rotational multimaterial 3D printing system was proposed that allowed precise manipulation of the local orientation of architecturally diverse filaments at a subvoxel level [21]. The proposed “shell–fan-core” geometry enabled controlled reinforcement of various materials to extrude as a filament. However, the research also did not discuss the distribution of materials within a given cross-sectional area. Considering that similar pressure and viscosity were applied, it is anticipated that the extruded area of reinforced materials through fans would be consistent, but their positioning would vary due to rotational extrusion. It was reported that tissue specificity and high biocompatibility can be achieved by controlled allocation of various hydrogels [22,23]. Our proposition can offer to achieve user-defined material distribution across a given region by controlling the switching time from one material to another with the appropriate selection of rheology and printing process parameters.

Our team recently created a nozzle system that enables the continuous deposition of multiple filaments through a single outlet by means of an asymmetric Y-connector [14]. The purpose of this system was not to achieve a uniform mixing of two materials or produce a filament evenly shared by both materials. Instead, it focused on dynamically switching the deposition of materials from one type to another, enabling users to achieve their desired material distribution within any specific region. That device was fixed at a specific switching angle [24], was made out of plastic, and was suitable for one-time use. In this article, we present an upgraded version of the nozzle system, where we explored four different angles such as 30 deg, 45 deg, 60 deg, and 90 deg (vertical and tilted) between the two materials. It was reported that if the flow is expanded initially before entering into another pipe with a bigger diameter compared to the nozzle diameter, the backflow can be controlled [25]. Changing the angles helps us figure out how that affects the control of backflow and minimizes the overall material switching time in the nozzle.

To manufacture the new asymmetric Y-connector nozzle, we chose stainless steel as the material. This updated version of the nozzle system is compatible with a 3D bioprinting system, using luer lock connectors, which are commonly used in printing systems, along with plastic tips, 3 ml syringes, and check valves. Our proposed nozzle can be easily customized and made to work with existing resources such as the 3D bioprinter, biomaterial, and pressure source. We opted for an asymmetric Y-connector configuration, which is commonly used in other designs made of polycarbonate, to ensure better alignment with the bioprinters currently in use. We used 304 and 316L stainless steel to manufacture our proposed connector, which offers the advantage of sanitization and reuse. We measured the material switching time and compared it for various angles to study their effects. To test the nozzle system, we used our previously developed hybrid hydrogel [26]. Our in-house fabricated nozzle connectors are easy to clean, sterile, and reusable, allowing smooth material transition and flow.

The rest of the article is organized as follows: Sec. 2 describes the materials and methods including features of flow simulations, preparation for 3D printing, design, and manufacturing of the proposed nozzle system. The result of flow simulation and extrusion of different materials, distribution of materials, and total switching time are presented in Sec. 3. Section 4 includes the discussion. Finally, Sec. 5 wraps up, highlighting the findings and future work of the article.

Our purpose in conducting this simulation study was not to optimize the nozzle structure, rather understanding the effects of various connection angles of our proposed nozzle connectors. A non-Newtonian material model was used in solidworks, defined by our viscosity data collected from a rheometer. For flow simulation, the viscosities of two materials published earlier such as 8% alginate (A₈) and 2% alginate–6% carboxymethyl cellulose (CMC; A₂C₆) were used in this article [27]. The viscosity of A₈ and A₂C₆ at 1 s⁻¹ shear rate are 11231 mPa.s and 26559 mPa.s, respectively. The velocity, vorticity, shear rate, and shear stress distributions were simulated for three different applied pressures such as 182 kPa (26 psi), 202 kPa (29 psi), and 222 kPa (32 psi). The print pressures were selected based on our earlier published works and other articles [28–31]. The boundary conditions and the number of iterations (13–85) needed to reach the convergence of the selected goal are presented in Table 1. X_min, X_max, Y_min, Y_max, Z_min, and Z_max are the size of the computational domain of the flow simulation we conducted. “Inlet Volume Flow” and “Pressure Openings-Environmental Pressure” were selected as inlet and outlet boundary conditions, respectively. We selected “Normal to Face” with uniform flow as a flow parameter for the whole model. The simulation package we used did not allow us to select the specific error tolerance number rather just allowed us to select either of the three available options such as minimum, maximum, and average to reach the goal. We ran the simulation for all three options and got the convergence with the number of iterations mentioned in Table 1. The detailed engineering drawing of all nozzle connectors is shown in Fig. 1.

A three-axis multi-head (three-extruder) Bio-X (CELLINK, Boston, MA) 3D bioprinter was used to extrude hydrogels. A 3 ml disposable barrel reservoir (EFD, Nordson, Westlake, OH) was used to store prepared hydrogels and extruded pneumatically. We used a plunger that extruded the material in the printer from the barrel reservoir. Material transition time, i.e., the time required to change the material flow from one type (100% M₁ from the vertical nozzle) to another type (100% M₂, from tilted nozzle) in the coaxial nozzle systems was determined as shown in Fig. 2(c). Material from one nozzle was fully extruded into the empty nozzle connector until it had reached the tip. The material from the other nozzle was then extruded continuously until it was visibly extruded at the nozzle tip. The time required for the two angles (axial and tilted nozzle) was recorded and analyzed to see how the angle variation affects the material transition. Our previously developed hybrid hydrogel (4% alginate and 4% CMC; A₄C₄) [26] was used as a test material to flow through the nozzle system. The viscosity of A₄C₄ is 22209 mPa.s at 1 s⁻¹ shear rate. An extrusion pressure of 110 kPa was applied. A₄C₄ demonstrated better cell viability (86% after 23 incubation days) compared to pure 4% alginate (80%) in our earlier research [26]. In addition to that, we explored the cell viability of various cell lines such as BxPC3, prostate stem cancer cell, and human embryonic kidney cells encapsulating into A₄C₄. It resulted in more than 89% cell viability after 15 incubation days [32]. Those promising results motivated us to select this composition as a working material to demonstrate this proposed nozzle system. In the future, we plan to encapsulate various cell lines into A₄C₄ and extrude simultaneously using this nozzle system.

To fabricate the scaffolds extruding through nozzle connectors having 30 deg, 45 deg, 60 deg, and 90 deg angles, the prepared A₄C₄ hybrid hydrogel was stored in two disposal syringes and extruded pneumatically following a layer-upon-layer fashion through a nozzle having 410 µm diameter on a stationary build plane. Various printing parameters such as nozzle diameter, air pressure, nozzle speed, and print distance (i.e., the perpendicular distance between the nozzle tip and print bed) can control the deposition rate of the material [33].

A computer-aided design (CAD) software, rhino 6.0,³ was used to design and define the vectorized toolpath of a scaffold. slicer,⁴ a G-code generator software, is used to generate a Bio-X compatible file including the toolpath coordinates and all process parameters to fabricate the scaffold. Two syringes of A₄C₄ were prepared and dyed with either red or blue food coloring. They were loaded onto the printer, and the 45 deg nozzle was attached to both syringes with check valves between the nozzle and syringes. A 0.41 mm plastic tapered syringe tip was used for printing. Each syringe was pressurized initially to fill in the empty space of the nozzle and come out of the plastic tip. The model used to print was a prismatic box of dimensions 20 mm × 20 mm × 1 mm, the layer height was set to 0.3 mm, and an infill percent was 11%. The print pressure was set to 130 kPa, and the print speed was set to 7.0 mm/s. The print distance used in fabricating the scaffold is 0.405 mm. The layer height, applied pressure, and print speed were optimized in our earlier works [26,28]. The print was recorded to examine the color-changing and mixing behavior. The overall scaffold fabrication process is schematically shown in Fig. 2.

After modeling all nozzle connectors using solidworks 3d modeling software, we printed them using a Raised 3D printer (Irvine, CA) as shown in Fig. 3. The outer diameter (OD) and inner diameter (ID) of the nozzle connector were 5.35 mm and 6.35 mm, respectively. Nozzle connectors were fabricated using stainless steel because the material being used will potentially be composed of live cell cultures. The intent is to sterilize the fittings between tests, so the fittings can be reused.

303 stainless steel Leur fittings and 304 Stainless steel tubing were purchased from McMaster-Carr McMaster-Carr (Elmhurst, IL). The fittings were modified in-house with the Bridgeport machine to fit onto the 304 stainless steel tubing. The previous plastic prototype fitting had an OD of 6.25 mm and an ID of 3.86 mm. The tubing for four connectors had an OD of 6.3 mm and an ID of 4.3 mm. Tubing was initially cut with a small horizontal bandsaw (General International: Model BS5205, Whitehouse, OH) to a length of 20.3 mm. The ends were then milled perpendicular on a Bridgeport Milling Machine (Atlanta, GA) fitted with a quick-release C5 Collet fixture to hold the tubing. A device with a three-jaw chuck was mounted on a separate Bridgeport machine to set different angles. The device was used to mill one half the desired angle on each tube. The angle was on one end of the tube and went to half of the diameter of the tube. There was a pair of tubes for each part. The exception was the 90 deg fitting. That cut was not made at the end of the tube for ease of manufacturing. After the cut was made, the three-jaw chuck would be rotated 180 deg, and a relief cut using a 3.96 mm diameter cutter was made on one part of the pair to provide clearance for the flow.

Parts were hand assembled and welded with a Coherent Rofin StarWelder (Baasel Lasertech, Gilching, Germany). Single laser pulses were applied with a foot pedal, while the parts were held under a microscope. Each pulse was applied for 5 ms with an average power of 2.3 kW and 0.3 mm diameter. The pulse shape is divided into five segments with power settings of 80%, 100%, 90%, 75%, and 50%. Pulse overlap was between 50% and 75%. All external seams were welded. Some stainless steel 304 filler wire was added to fill any voids. The 304 stainless in the tubing was easier to weld because of the lower sulfur content. Parts were cleaned with a small wire brush and returned to their respective baggies. The overall nozzle manufacturing steps are shown in Fig. 4.

Based on the objective function, the simulation runs with a certain number of iterations (13–85) to reach convergence. This number of iterations depends on the angle of nozzle connectors and applied pressure. As a representative of all simulations, Figs. (5)–(8) show the distribution of vorticity, shear rate, shear stress, and velocity for 182 kPa for 30 deg, 45 deg, 60 deg, and 90 deg, respectively. Analysis of the shear stress distribution indicates that the shear stress values for 45 deg and 90 deg are almost equivalent, while 30 deg and 60 deg exhibit 43% and 2.2% higher shear stress, respectively, in comparison to 45 deg and 90 deg.

Simulations of vorticity, which is the curl of the velocity field, indicate that vorticity originated at the inlets and the intersection. It shows that the fluid would briefly enter all angled connector pieces before flowing back down. The simulations also did not show backflow traveling up the straight part of the nozzles. However, the backflow increases with increasing the inclined angle from 30 deg to 90 deg. This increase can be attributed to the pressure changes caused by the reduction of the tilted path (30 deg) to a straight path (90 deg). This scenario results in the release of applied pressure towards a straight path (90 deg) generating the most amount of shear stress at the nozzle tip of the 30 deg nozzle connector compared to other connectors. This result also supports the earlier statement: if the flow is expanded initially before entering into another pipe with a bigger diameter compared to the nozzle diameter, the backflow can be controlled [25]. Since it was reported in our earlier published work that shear stress directly controls the extrusion velocity $(τ∞γ˙n∞vn)$ [32], 30 deg nozzle connector showed higher velocity distribution into the nozzle compared to other connectors as shown in Figs. 5(d)–8(d). Every nozzle tested had some degree of backflow, so none of the designs were able to negate that on their own, but the problem was easily fixed by using check valves.

The shear rate distribution for all angles at 182 kPa applied pressure was analyzed as shown in Fig. 9(a). We observed the shear rate created by 45 deg and 90 deg was almost similar at nozzle tip, which matches with the shear stress distribution. Flow simulations for 30 deg nozzle connector were conducted for three different applied pressures such as 182, 202, and 222 kPa from nozzle tip to the endpoint of the arrangement. From the shear rate distribution for each nozzle connector, it is clear that higher applied pressure showed larger shear rate at the tip. Figure 9(b) shows overall shear strain distribution for three applied pressures of 182, 202, and 222 kPa from the nozzle tip to 73.47 mm. The simulation result shows 50% and 20% higher shear rate for 22% and 10% increment of applied pressure compared to the applied pressure of 182 kPa. Similar characteristics were resulted in for 45 deg, and 90 deg nozzle connectors.

Following the methods described in Sec. 2.3, we fabricated a total of four nozzle connectors having 30 deg, 45 deg, 60 deg, and 90 deg angles as shown in Fig. 10(b).

All connectors were used to determine the material transition time from one material (M₁) to another (M₂) as shown in Fig. 11. Material from one nozzle was fully extruded into the empty nozzle connector until it reached the tip. The time required for this operation is defined as “axial delay” time. The material from the other nozzle extruded continuously until it visibly extruded at the nozzle tip. The time needed for this operation is defined as “tilted delay” time. The summation of “axial delay” and “tilted delay” is termed as the “total time.” A nozzle connector having 30 deg showed the highest axial time, where 90 deg showed the lowest. A possible reason is the intersection lengths of axial and tilted connectors of nozzles having angles of 30 deg, 45 deg, and 60 deg are 100%, 41%, and 15.5% larger, respectively, compared to the nozzle having angle of 90 deg as shown in Fig. 12. Even if we used a check valve to reduce the backflow of material during extrusion through the axial connector, a higher intersection length may allow material entry to a tilted connector.

In case of a tilted connector, a nozzle having 45 deg and 90 deg showed the lowest titled delay time where 60 deg showed the highest. For both 45 deg and 90 deg, material M₁ showed minimal entry to the tilted nozzle during axial flow resulting in quickest flow of material M₂. Therefore, a nozzle connector having angles 45 deg and 90 deg showed the lowest total time to shift from material M₁ to material M₂ as shown in Fig. 13.

To analyze the distribution of material during flow through the nozzle connector closely, we used the connector without a plastic tip. A set of filaments were fabricated using all nozzle connectors having angles of 30 deg, 45 deg, 60 deg, and 90 deg. As an example, a filament fabrication process using 30 deg nozzle connector is shown in Fig. 14(a). The material distribution throughout the filament was analyzed following a technique shown in Figs. 14(b) and 14(c). The fabricated filament was crosslinked with CaCl₂ for 5–7 min and sliced to get the cross sections at different locations of filament to analyze the material distribution as shown in Fig. 14(b). Even though the fabricated filament was close to circular in shape after crosslinking, during slicing, the filament failed to maintain a similar shape. Cross sections of filaments fabricated by all nozzle connectors are shown in Fig. 16 where all of them showed material distribution at various levels. rhino and imagej software were used to process the images of those cross sections to analyze the material distribution. From the cross-sectional view of each filament (Fig. 15) and material distribution calculation (Fig. 16), we observed that filament extruded through the nozzle connectors having angles of 30 deg and 45 deg showed smooth material transition. One possible explanation could be that the applied pressure does not release toward the tilted paths (e.g., 30 deg and 45 deg), resulting in the nozzle tip experiencing the highest shear stress compared to other connectors.

Finally, we fabricated scaffolds with 30 deg, 45 deg, and 90 deg nozzle connectors. Even though we observed a successful material transition for all nozzle connectors as shown in Fig. 12, we did not consider 60 deg nozzle connector to fabricate scaffolds to expedite the demonstration. Two syringes of A₄C₄ were prepared and dyed with either red or blue food coloring. They were loaded onto the printer, and the 30 deg, 45 deg, and 90 deg nozzle connectors were attached to both syringes with check valves between the nozzle and syringes. First, one type (M₁) material was extruded for the first layer followed by the other (M₂) material. From the flow diagram of Figs. 17(a) and 17(b), the shifting of material from one type (M₁) to other (M₂) was clearly visible for a nozzle connector having 45 deg angle at different layer heights. This phenomenon can help control the required material mix at various layer heights based on applications.

The material switches were also observed for nozzle connectors having 30 deg and 90 deg angle as shown in Figs. 18(a) and 18(b). The nozzle connector with a 30 deg angle displayed a consistent flow of material and maintained the desired shape fidelity, while the nozzle connector with a 90 deg angle encountered difficulties in achieving this. The controllability to allocate materials region-wise of our proposed nozzle connectors can help achieve tissue specificity and high biocompatibility [22,23], which can lead toward successful tissue regeneration in the long run.

This article represents an improved version of our earlier system where we explored four different angles such as 30 deg, 45 deg, 60 deg, and 90 deg using stainless steel as the material for the nozzle. To minimize the number of experiments due to limited resources, we selectively chose a wide range of gap (e.g., mostly 15 deg and 30 deg) between two nozzle connectors. In the future, we intend to employ a systematic simulation approach to determine an optimal nozzle connector angle that minimizes even more the switching time between the extrusion of different material types. We continued with the asymmetric Y-connector configuration, which is frequently used in other designs, to ensure compatibility with existing bioprinters. In our recently published work, we printed 30 deg, 45 deg, 60 deg, and 90 deg nozzle connectors and successfully extruded the material through them. However, we observed that higher applied pressure was required to extrude materials with similar viscosities through 3D-printed nozzle connectors compared to the materials extruded through the nozzle connectors used in our earlier proposition [14]. The staircase effect of 3D-printed nozzle connectors may add internal roughness and eventually require higher applied pressure to extrude through [35]. However, we did not observe a significant difference in the required applied pressure to extrude material between our previous polymeric nozzle connector and the proposed stainless steel connector.

The updated system was successfully used with a 3D bioprinter to extrude and switch materials. The use of 304 and 316L stainless steel will allow users to sanitize and reuse as required. Even the simulation result for vorticity showed a chance of backflow at various levels for different nozzle connectors, using check valve during the experiment nullified this phenomenon and resulted in smooth material flow and switch for all nozzle connectors. The measurement and comparison of the material switching time for different angles indicates a nozzle connector having 45 deg worked the best. The simulation outcomes indicated that the distributions of shear rate and shear stress were comparable for the 45 deg and 90 deg nozzle connectors. For any specific nozzle connector (such as 30 deg), applied pressure controlled the shear rate. Figure 16 displays cross-sectional images of filaments produced by all nozzle connectors, demonstrating material distribution at different levels in all of them. This observation suggests that by selecting appropriate process parameters (nozzle connectors, applied pressure, print speed, print distance), the desired material mix specified by the user can be achieved. This is one of the areas for future research in this field. By examining the cross-sectional views of each filament (Fig. 15) and calculating the material distribution (Fig. 16), we noticed that the filaments extruded through the nozzle connectors with 30 deg and 45 deg angles exhibited a seamless transition of material.

The nozzle connector with an angle of 30 deg had the highest axial time, while that with an angle of 90 deg had the lowest. We hypothesized that the reason is that the intersection lengths of the axial and tilted connectors in the 30 deg, 45 deg, and 60 deg nozzles are 100%, 41%, and 15.5%, respectively, larger than that of the 90 deg nozzles. Although a check valve was used to reduce the backflow during axial flow, a higher intersection length could allow material to enter the tilted connector. As for the tilted connector, the nozzle with angles of 45 deg and 90 deg had the shortest tilted delay time, while that with an angle of 60 deg had the longest. For both angles of 45 deg and 90 deg, material M₁ had minimal entry into the tilted nozzle during the axial flow, resulting in the quickest flow of material M₂. Therefore, the nozzle connectors with angles of 45 deg and 90 deg had the shortest total time to switch from material M₁ to material M₂. However, it is clear that even 90 deg had the shortest total time to switch, it struggled to display a consistent flow of material and maintained the desired shape fidelity. One plausible explanation for this observation is that the applied pressure got released toward comparatively the straight paths (e.g., 60 deg and 90 deg), causing the nozzle tip to face challenges in achieving higher shear stress, shear rate, and consequently extrusion viscosity compared to 30 deg and 45 deg connectors. This observation suggests that by selecting appropriate process parameters (nozzle connectors, applied pressure, print speed, print distance), the desired material mix specified by the user can be achieved. This is one of the areas for future research in this field.

As an extension of our previous work, we considered 30 deg, 45 deg, 60 deg, and 90 deg angles (vertical and tilted) between the two materials and chose stainless steel as a material to fabricate nozzle connectors. We determined and compared the overall material switching time to analyze the effects of those various angles. Our previously developed hybrid hydrogel (4% alginate and 4% CMC) was used as a test material to flow through the nozzle system. We observed closely the material distribution into the filament during the extrusion through a nozzle connector and a nozzle itself. The main outcomes for this research are as follows: (i) all connectors were able to switch material from one to another, (ii) flow simulation resulted in backflow increases with increasing the angle from 30 deg to 90 deg, (iii) the nozzle with angles of 45 deg and 90 deg had the shortest tilted delay time, while that with an angle of 60⁰ had the longest, (iv) the nozzle connectors with angles of 45 deg and 90 deg had the shortest total time to switch from one material to another, (v) we observed that filament extruded through nozzle connectors having angles of 30 deg and 45 deg showed smooth material transition, and (v) the measurement and comparison of the material switching time for different angles indicate that a nozzle connector having 45 deg worked the best. We identified the material transition time during extrusion through the nozzle connecting it to all nozzle connectors. In the future, we will also identify the effect of material viscosity (using material composition other than A₄C₄) on the material transition for all connectors and nozzle connected to the connector. Finally, our long-term goal is to use those nozzle connectors to extrude multiple materials encapsulating living cells.

• National Science Foundation (Grant No. OIA-1355466).

• Department of Transportation US-DOT No. 693JK31850009CAAP.

• National Institute of Health under COBRE: CDTSPC (Grant No. P20GM109024).

There are no conflicts of interest.

The datasets generated and supporting the findings of this article are obtainable from the corresponding author upon reasonable request.