Open-source 3D printed air-jet for generating monodispersed alginate microhydrogels

Open-source designs represent an attractive and new tool for research as it provides both affordable and accessible options to the lab environment. In particular, with the advent of new and cheap additive manufacturing technologies, the open-source design of lab hardware enables others to perform research that would be difficult otherwise. This manuscript describes an air-jet system designed to be open-source and simple to produce with 3D printing. The fully 3D printed air-jet was designed for the generation of hydrogel microbeads of a controllable size. Alginate microbeads were used as a working model, given that it has many promising research applications due to their injectability and highly reproducible properties. A fit definitive design of experiments was performed to determine critical factors affecting diameter, index of dispersity, and circularity of microbeads from this air-jet design. By regulating alginate concentration, air pressure, pump speed, and needle diameter could achieve control over microbeads size from 200-800 µm with low variance. Furthermore, we also demonstrate the potential probiotic research applications of the open-source air-jet through the encapsulation of bacteria in alginate microbeads with controllable degradation. The results of this study exhibit an open-source platform for making microscale biomaterials with controllable properties that can be achieved through budget 3D printers.


44
Publishing replicable scientific work remains recognized as a critical endeavor that 45 researchers, funding agencies, and publishers should engage in cooperatively. Several efforts 46 have been made to encourage reproducible science [4][5][6], including free open-source hardware 47 (FOSH) and a growing number of online repositories, such as Github,Figshare,and 48 Protocols.io [7,8]. Specifically, FOSH cultivates participation in science by reducing supply 49 limitations, fostering new research opportunities, and facilitating translation of these tools for 50 educational purposes [8]. For example, open-source tools are currently available on the 51 National Institutes of Health (NIH) website for use in cell cultures, microfluidics, and drug 52 delivery systems [9]. Furthermore, the open-source license benefits researchers because 53 building equipment yields a deeper understanding and allows modifications of designs [10]. The 54 knowledge collected by users refines the open-source device to become robust while remaining 55 accessible to others. One appealing strategy for creating FOSH is through designs that utilize 56 3D printing. The accessibility of fabrication with 3D printing has furthered the inventiveness 57 within different communities, from laymen to aerospace engineers [10][11][12][13][14][15]. Moreover, the 58 reduced cost of 3D printers in the past couple of years is resulting in 3D printers becoming 59 available to the public at Makerspaces, public libraries, and universities [16]. 60 One potential tool that would benefit from becoming a FOSH is an air-based system to 61 produce microbeads. These microbead generators commonly employ polymer and crosslinker 62 solutions in the formation of microscale hydrogel [17][18][19][20]. Microbeads are currently utilized and 63 studied in different functions, including drug and cell delivery, cryopreservation, and scaffolds for 64 tissue engineering strategies [20][21][22]. In these biomedical science applications, alginate is a 65 particularly attractive polymer for microbeads, as it is biocompatible, undergoes rapid gelation 66 under gentle conditions, and enables controllable mesh size [23]. Alginate, a copolymer 67 comprised of (1,4)-linked β -D-mannuronate (M) and α -L-guluronate (G) residues, crosslinks in 68 the presence of divalent cations, such as Ca 2+ , resulting in the formation of a 3D polymeric 69 network characteristic of hydrogels [24][25][26]. Specifically, alginate microhydrogels have been 70 applied to guiding the morphogenesis of progenitor endothelial cells and control the delivery of 71 lentivectors and stem cells [27][28][29]. Alginate microbeads are commonly generated using droplet 72 microfluidics, coaxial airflow units, two-channel air-jackets, and high voltage [22,27,28,30]. 73 To generate alginate microbeads, we designed a novel 3D printed air-jet system that 74 generates control droplets of solution. An air-jet pushes air away from a source, making it 75 similar to air bifurcation or electrostatic bead generators [22,30,31]. Our air-jet uniformly 76 extrudes air across a needle attached to a syringe. A syringe pump supplied an alginate 77 polymer solution at the needle tip, which formed droplets that fall into a calcium bath and 78 generate hydrogel beads. Our objective for the device design was to permit modularity and 79 sterility while having a low manufacturing cost by being compatible with budget 3D printers. We 80 hypothesized controlling airflow, pump speed, and needle size for the 3D printed air-jet can 81 create well-defined alginate microbeads. The device was characterized via a design of 82 experiments (DOE) analysis using a fit definitive screening to define factors that altered 83 resulting alginate bead characteristics. We expect this system to be reproducible between 84 experiments and between laboratories due to well-defined set-up parameters that are typically 85 less abundant in other publications using air-based droplet generating methods [31][32][33]. 86 87

89
The air-jet was designed using Fusion 360® software (Autodesk, Inc) and prepared for 90 3D printing using Cura software (Ultimaker), an open-source slicer program. The device was 91 split into three components, including the air-jet, the air intake, and supports. The air intake was 92 printed with the air inlet parallel to the plane of the printing bed to prevent the nozzle from 93 fracturing into the air tubing along with the printed layers. The air-jet was oriented with the outlet 94 normal to the plane of the printing bed, such that the outlet for droplets was printed last, to 95 prevent the need for support material. All components were printed on an MP Select Mini V2 96 (Monoprice, Inc) with poly (lactic acid) (Hatchbox®) at a layer thickness of 0.0875 mm and a 97 printing speed of 50 mm/s. The extrusion temperature was 190 °C, and the heat bed 98 temperature was 60 °C. 99 100 Generation of alginate microbeads 101 LF 10/60 alginate polymer (~120-150 kDa) with higher G-block content (>60% as 102 specified by the manufacturer) obtained from Novamatrix (FMC) was used to generate the 103 microbeads. Alginate solutions were prepared by dissolving alginate polymer in either deionized 104 water (diH 2 O) or phosphate buffer solution with added magnesium and calcium (PBS ++ ; Life 105 Technologies), as previously described [34]. The open-source air-jet and a syringe pump 106 (Braintree Scientific) were run in parallel in a vertical position above a calcium bath. Syringes 107 (Becton Dickinson) with varying needle sizes (PrecisionGlide TM ; Becton Dickinson) were 108 positioned within the syringe pump and air-jet. The distance from the needle to the calcium bath 109 was constant at 15 cm. The air-jet was then centered over the needle to generate a uniform air 110 flow across the needle, with the tip of the needle slightly protruding from the outlet of air-jet. 111 Nitrogen gas was then run through the air-jet before starting the syringe pump to prevent 112 alginate from getting stuck in the air-jet chamber. After activating the syringe pump, the direction 113 of droplets was checked using a flat surface. Adjustments were made in response to poor 114 alignment of the needle through the center of the air-jet to ensure alginate droplets fell directly 115 downwards. Finally, 5 mL of 100 mM calcium chloride (CaCl 2 ) (Sigma) was used as the calcium 116 bath for all experiments and placed underneath the air-jet. Alginate microbeads were generated 117 for one minute with this setup before being analyzed directly from the wells using ImageJ 118 software (NIH). 119 120 Definitive Screen Design to characterize nitrogen pressure, 121 needle gauge, and pump speed effect on microbead 122 formation 123 A DOE approach with a definitive screening design (DSD; JMP software) was utilized to 124 determine the impact of several factors on the properties of microbead generated with this air-jet 125 design. A fit definitive screening test was used to ascertain active main effects and second-126 order effects of buffer type, needle diameter, air pressure, alginate concentration, and pump 127 speed on bead diameter, the index of dispersion of bead diameters, and circularity. The effect of 128 polymer solution viscosity was investigated using alginate dissolved at different concentrations 129 (1, 2, and 3% (w/v)) and in different solvents (PBS ++ and diH 2 O). The various needle gauges of 130 27, 21, and 18 were tested, which had approximate diameters of 0.210, 0.524, and 0.840 mm, 131 respectively. Nitrogen pressure was altered between 200 to 800 kPa using a pressure gauge 132 attached to a compressed nitrogen gas tank. Syringe pump speeds were also varied between 133 100, 250, and 400 µL/min. Index of dispersion is the variance normalized to the mean and was 134 used to analyze bead uniformity. The circularity is defined below in the following equation: 135 where the value of bead circularity varies between 0 to 1, with 1 being a perfect circle. 137 Microbeads were analyzed with ImageJ, and significance was determined via the fit definitive 138 screening analysis. DSD utilized a minimum of 18 groups, where a group is a set of differing 139 factors, with n = 8 samples per group. The residual was calculated for each examined factor by 140 subtracting the mean of all measured values of a specific factor from the average of a specific 141 condition of the factor. If conditions of a factor are determined to be significantly different, trends 142 were described with predictive plots generated using least-square regression or polynomial 143 least squares.  solution, containing alginate lyase, and microbeads were generated with the air-jet as described 168 above. These alginate microbeads were left to gel for 15 minutes and subsequently washed in 169 DI water. Next, approximately 100 µL of beads were transferred to 12-well plates, topped in 4 170 mL of EGM-2MV media (Lonza), and incubated at 37 °C. At 1 hour and 1, 3, 5, 7, 14, and 21 171 days microbeads were imaged, and the size of the beads was analyzed with ImageJ. The initial 172 distribution of microbead sizes at 1 hour, and the change in microbead size over 21 days was 173 reported (n = 50). 174

175
Comparisons were assessed by Student's unpaired t-tests. Differences between 176 conditions were considered significant if P < 0.05. All analyses were performed using GraphPad 177 Prism software (GraphPad Software) and JMP software. 178 179

181
The air-jet system was engineered for the rapid generation of alginate microbeads with 182 consistent and controllable size. The design of the air-jet was optimized for a Fused Deposition 183 Modeling (FDM) 3D printer, which enabled the internal features to be printed without support 184 material. The distance between the inner cylinder that holds the needle and the outer wall was 1 185 mm to prevent fusing between features while printing ( Fig 1A). Additionally, air flows around this 186 gap (teal), which has a cross-sectional area of approximately 33 mm 2 . This area expands to a 187 maximum of 103.9 mm 2 before being constricted to an area of 7.1 mm 2 at the exit hole (green). 188 The small needle channel (orange) has a cross-sectional area of approximately 1.77 mm 2 , 189 which is further reduced with the inserted needle. This region reduces air from traversing up the 190 needle inlet and increases air velocity at the tip of the needle. Including an angled inlet (red), the 191 overall profile of the air-jet is compact to allow needles to fit through the air-jet (Fig 1B). 192 Furthermore, the construction permits the incorporation of a needle and syringe, either through 193 direct attachment to the syringe pump or separately through adjustable stands (Fig 1C-D).  Figure). 206 From this definitive screening design, prediction trends were calculated for factors exhibiting a 207 significant first-order or second-order relationship with microbead size, variance, or circularity. 208 The equation for these predictive trends of continuous factors are presented in each subfigure 209 (Fig 2A-C). Alginate concentration was determined to have a second-order effect on bead 210 diameter, with higher alginate concentrations resulting in increased bead diameter (Fig 2A), 211 decreased bead variance (Fig 2B), and increased circularity (Fig 2C). Similar trends were also 212 observed with a decrease in air pressure. Needle diameter had a small effect on bead diameter, 213 with a smaller needle diameter creating smaller beads. Additionally, the buffers used to dissolve 214 the alginate polymer were found to influence the index of dispersity, with PBS ++ usually leading 215 to a larger distribution of bead sizes. Representative images of the beads under similar 216 conditions are shown. Beads with lower alginate concentrations were smaller and less uniformly 217 round (Fig 2D). Using a faster pump speed, from 250 to 400 µL/min, led to more uniformly round 218 beads that appear slightly larger (Fig 2E). Beads formed with alginate dissolved in diH 2 O appear 219 more uniformly round compared to alginate dissolved in a phosphate buffer (Fig 2F), while 220 beads were smaller using a smaller needle size (Fig 2G). controlling bead sizes. A 27G (0.21mm) needle was used to make beads with 2% (w/v) alginate. 240 The addition of airflow had a drastic effect on the bead size (Fig 3A). From the initial addition of 241 airflow, the bead size decreased with a one-phase decay (R 2 of 0.9925). The equation for the 242 nonlinear fit is 243 where air pressure is in kPa. For the range where airflow was used, the decrease in bead size 245 with increasing air pressure is approximately linear (R 2 = 0.8165), described by the following 246 equation: 247 . 248 Increasing nitrogen gas pressure was found to both decrease bead size and increase the index 249 of dispersion (Fig 3A-B). While no airflow creates beads with the largest index of dispersion, the 250 variance in bead diameters increased as air pressure is increased beyond 400 kPa (Fig 3B). A 251 rise in air pressure beyond 400 kPa also led to a reduction in bead circularity and greater 252 deviation in circularity (Fig 3C). Representative images of these beads show drastic changes in 253 bead diameters with the addition of air and loss of circularity at higher air pressures (Fig 3D). properties. Alginate microbeads were created with various pump speeds using a 2% w/v 271 alginate polymer solution and 400 kPa air pressure. The bead diameters were determined to 272 linearly decrease with pump speed, R 2 = 0.8427 (Fig 4A). The bead diameter increased by 273 approximately 40 µm per 100 µL/min change in pump speed in the range tested regardless of 274 whether a surfactant was used or not. There is no significant difference in the index of 275 dispersion of bead diameters with the addition of surfactant, nor for bead circularity (Fig 4B-C). 276 Together, these data demonstrate how bead diameter formation is dependent on factors that 277 occur at the air-jet, while circularity of alginate microbeads is adjusted through factors relating to 278 the calcium bath. Alginate microbeads generated from the air-jet was interrogated for potential biomedical 295 applications. Bacteria were successfully encapsulated in alginate microbeads of varying 296 diameters by changing the air pressure ( Fig 5A). Furthermore, these bacteria were viable and 297 could proliferate within the alginate microbeads (Fig 5B). The air-jet could also generate 298 degradable alginate microbeads. Alginate microbeads containing various alginate lyase 299 concentrations were found to initially have similar distributions of bead diameters (Fig 5C). The 300 alginate microbeads containing alginate lyase decrease in size over 21 days, with the largest 301 change within the first few days (Fig 5D). An approximate 27.3% reduction in microbead 302 diameter was observed with 50 mU/mL of the enzyme. 303

316
The designed open-source air-jet reliably generated microbeads without the need for 317 complex assembly or expensive lab equipment. To our knowledge, this is the first fully 3D 318 printable air-jet system. Importantly, this device controls needle placement and airflow, making 319 set-up easy and reproducible compared to other air-jet systems [18,36]. Furthermore, the 320 geometries of internal features take advantage of FDM 3D printers to generate the air-jet as one 321 piece benefiting the accessibility and affordability of the design. Overall, the described air-jet 322 system offers control over microbeads with defined set-up parameters that can be created with 323

budget 3D printers. 324
The open-source air-jet described here provides an advantage in ensure reproducibility 325 and reliability by allowing defined set-up parameters. Specifically, the separation from the 326 needle and the airflow, comparable to other air-jets that shear droplets off the needle tip; 327 however, the fixed needle position ensures central alignment with the airflow [20,33,37]. 328 Consequently, the described air-jet requires consideration of needle size is important to prevent 329 the backflow of solution or air. In contrast, electrostatic bead generators do not need to consider 330 needle size or backflow of components, but the air-jet is still advantageous for not requiring an 331 electrical power source [18,38]. To further prevent possible backflow or entrapment of solution, 332 smaller diameter holes at the needle inlet were designed. Additionally, the air was flowing 333 through the device before running the syringe pump with the needle placed approximately 0.5 334 mm from the outlet. Together, these considerations result in monodispersed alginate 335 microbeads in minutes, even with highly viscous solutions, for what would take hours with 336 microfluidics [28,39]. 337 The fit definitive screening test determined polymer concentration, air pressure, needle 338 diameter, and buffer type as factors influencing microbead formation. Advantageously, the DOE 339 methodology dramatically reduced the total number of groups, from 162 to 18, where a group 340 represents a specific combination of different factor values that could manipulate microbead 341 properties. The results suggest that a higher alginate concentration, larger needle diameter, or 342 lower air pressure leads to larger circular beads, an outcome consistent with alginate 343 microbeads formed via other air-bifurcation or air-jet systems [17,18,37,39]. Whereas lower 344 concentrations of alginate and higher airflow speeds could generate smaller beads, the 345 prediction trends suggest this could lead to problematic increase variance and decrease 346 circularity. Although the circularity was improved with higher alginate concentrations, higher 347 polymer densities can be disadvantageous for cell applications where cells are required to 348 migrate through the hydrogel scaffold [37,39]. Therefore, alternative methods, such as air 349 pressure and needle size, should be utilized with the optimal polymer concentrations to achieve 350 circular hydrogel microbeads of the desired diameter. Taken together, alginate microbead 351 properties can be controlled by adjusting multiple parameters involved with this air-jet design. 352 Pump speed and air pressure were interrogated as methods to regulate microbead size, 353 circularity, and uniformity. While air pressure could control the size of beads from 2 mm to 200 354 µm, adjusting pump speed changed bead sizes from 350 to 425 µm, suggesting air pressure 355 provides a more dynamic range of bead sizes. However, both methods generate beads in the 356 range of 200-800 microns in diameter with a coefficient of variation of less than 10% that is 357 often desired for encapsulation of cell applications [40]. Although microbeads smaller than 200 358 microns can be made with higher air pressures, tear-shaped droplets will form without sufficient 359 distance and time for the droplet to become round [41]. Additionally, the surface tension of the 360 bath also contributes to deviations in circularity. Consistent with previous reports, the addition of 361 a surfactant to the calcium bath did improve the circularity of beads in this system [39]. Overall,362 the designed open-source system generates microbeads with desirable characteristics, with 363 similar considerations as other air-jet systems being required to improve bead uniformity. 364 The encapsulation of bacteria and the enzyme alginate lyase highlight the distinct 365 benefits of the described air-jet system for encapsulating cargo that can degrade the hydrogel 366 matrix. Here, we load microbeads with bacteria, and their growth over 24 hours results in 367 pockets of bacteria within the hydrogel matrix. The encapsulation of bacteria is promising for 368 numerous biomedical applications, including probiotic delivery, especially for anaerobic bacteria 369 sensitive to atmospheric oxygen concentrations [21,[42][43][44][45]. By using the air-jet system with 370 nitrogen gas in a nitrogen-enriched chamber, microbeads could be rapidly generated with 371 anaerobic bacteria while displacing oxygen [21]. Furthermore, the use of enzymes to degrade 372 alginate has been previously used to control the delivery of endothelial progenitor cells and 373 adeno-associated vectors [46,47]. The rapid degradation of microbeads observed here 374 suggests heterogenous microbeads could result from a slower output rate, such as with 375 microfluidic methods [28]. Herein, we demonstrate the specialized utility of this open-source air-376 jet system that can rapidly generate monodispersed microbeads for encapsulation applications. 377 In conclusion, we present an open-source air-jet with simple features that can be printed 378 on a small budget 3D printer. The system's modular design allows easy cleaning, 379 interchangeability to customized set-ups, and rapid replacement. We have demonstrated that 380 alginate microbeads can be generated using this open-source air jet system with reasonable 381 microbead diameters and variance. We anticipate the system will enable other research 382 laboratories, as well as other fields, to easily generate polymeric microbeads.

388
Authors have no conflict of interest or competing interest to declare. 389