Research highlights: microfluidically-fabricated materials

Abstract

Polymer particles with precise shapes or chemistries are finding unique uses in a variety of applications, including tissue engineering, drug delivery, barcoding, and diagnostic imaging. Microfluidic systems have been and are continuing to play a large role in enabling the precision synthesis of designer particles in a uniform manner. To expand the impact of these microfluidic-fabricated materials additional fundamental capabilities should still be developed. The capability to fabricate microparticles with complex three-dimensional shapes and increase the production rate of particles to an industrial scale will allow evaluation of shaped particles in a range of new applications to enhance biological, magnetic, optical, surface wetting, as well as other interfacial or mechanical properties of materials. Here we highlight work applying large collections of simple spherical microgels, with unique surface chemistry that allows in situ particle–particle annealing, to form microporous injectable scaffolds for accelerated tissue regeneration. We also report on two other techniques that are addressing the ability to create 3D-shaped microparticles by first sculpting a fluid precursor stream, and increasing the rate of production of particles using contact lithography to millions of particles per hour. The combination of these capabilities and the applications they will enable suggest a bright future for microfluidics in making the next materials.

Injectable microporous gels formed from microfluidically-generated building blocks

Injectable or flowable biomaterial scaffolds are uniquely suited to aid in tissue regeneration by molding to a wound site or allowing minimally-invasive application. Delivered as a liquid and polymerized in situ, previous injectable scaffolds possessed a fundamental trade-off between overall mechanical strength and porosity/degradability enabling tissue ingrowth. Highly-crosslinked scaffolds that can maintain a structural support often result in reduced cell migration and ingrowth. Generating micropores in a scaffold promotes cellular infiltration while separately modulating bulk material properties;¹ nonetheless these microporous scaffolds generated using leaching methods must be manufactured ex situ.²

Griffin et al. tackled these issues with a bottom-up approach: creating a scaffold from microfluidically-fabricated building blocks.³ Produced by a microfluidic water-in-oil emulsion method (Fig. 1a), uniform microsphere scaffold building blocks are polymerized, collected and brought into an aqueous solution where they are subsequently injected and annealed to one another enzymatically forming a Microporous Annealed Particle (MAP) gel (Fig. 1b and c). Micropores form as the network of void spaces between the covalently linked spherical gel particles. These building blocks are composed of a synthetic hydrogel mesh of multi-armed poly(ethylene) glycol-vinyl sulfone (PEG-VS) backbone decorated with cell-adhesive peptide (RGD), protease substrate crosslinkers, and two transglutaminase peptide substrates (K and Q). Via crosslinking of K and Q peptides by thrombin-activated Factor XIII, an enzyme responsible for blood clotting, these neighboring blocks dynamically form the MAP scaffold in situ with a seamless interface (Fig. 1d and e).

Fig. 1 Microfluidic generation of microsphere hydrogel building blocks for the creation of microporous annealed particle (MAP) scaffolds. (a) Scheme illustrating microgel formation using a microfluidic water-in-oil emulsion system. A pre-gel and crosslinker solution are segmented into monodisperse droplets followed by in-droplet mixing and crosslinking via Michael addition. (b) Microgels are purified into an aqueous solution and annealed using FXIIIa into a microporous scaffold, either in the presence of cells or as a pure scaffold. (c) Fluorescent images showing purified microgel building blocks (left) and a subsequent cell-laden MAP scaffold (right). (d) MAP scaffolds are mouldable to macroscale shapes, and can be injected to form complex shapes that are maintained after annealing. (e) This process can be performed in the presence of live cells. Reprinted with permission from Macmillan Publishers Ltd: Nature Materials. Griffin et al., Nat. Mater., 2015, 14, copyright 2015.

The chemical and physical properties of the scaffold can be tailored through microfluidic fabrication. The microporosity of the scaffold was modulated by the size of building blocks, which the authors precisely controlled with flow rate and geometry of the microfluidic device. The manipulation of storage moduli was achieved by varying PEG weight percentages and crosslinker stoichiometries, which were introduced into two separate inlet channels within the microfluidic device and only mixed once a droplet was formed. As a result, the moduli spans the stiffness regime necessary for mammalian soft tissue mimetics. In addition, the degradation of the scaffold was determined by the combination of microporosity and physical properties of the MAP gels.

The authors first demonstrated that cells could be seeded directly within the MAP gels prior to annealing, and following annealing extensive three-dimensional cellular networks rapidly formed for three human cell lines. They observed that cell networks increased in size and complexity through the entirety of the experiment and growth rate and cellular network formation greatly exceeded identical conditions with a non-porous gel. Furthermore, they were able to deliver the microgel building blocks directly to a wound site in murine skin by syringe injection, and found that the annealed MAP scaffold accelerated wound closure compared to control conditions or non-annealed scaffold by host-cell recruitment through microscale porosity. These results clearly support that the MAP scaffold prompts in vitro and in vivo cell spreading and migration as well as bulk tissue integration.

An important point is that imperfect self-assembly of the microgel building blocks leads to a robust formation of a porous scaffold, solving many issues with other bottom-up biomaterial approaches. Beyond wound healing, microfluidic-control over the building block generation provides a new bottom-up framework in tissue engineering scaffold fabrication; the self-assembled scaffold in situ combines the benefits of injectability, microporosity and modularity. Further increases in the production rate of microgel particles will be an important point to address in the future. Overall, this novel scaffold/gel should be able to improve tissue regeneration, organ-on-a-chip technologies, as well as stimulating clinical research and applications in wound healing.

Optofluidic fabrication of complex 3D particles using inertial flow sculpting

Particles with complex three-dimensional (3D) morphologies can offer unique functionalities, such as behavior under electrical stimulation, tighter packing density, and new optical characteristics, making them applicable to structural materials, photonics, and drug delivery. However, current manufacturing processes are limited in speed, resolution, and complexity of fabrication. Complex geometries can be produced using a wide range of materials with 3D printers, but at the cost of slow printing processes and low resolution. Injection molding techniques have similar advantages, but are limited to the shapes that can be produced with a mold.

In microfluidics, stop flow lithography (SFL) and optofluidic maskless lithography (OFML) can polymerize high-resolution 3D particles continuously by illuminating ultraviolet (UV) light on a static UV reactive fluid, but shapes have been mostly limited to 2D extrusions of the mask. Based on these principles, Paulsen, et al. have developed a new optofluidic fabrication method that relies on two sequential steps: (1) highly controllable inertial flow shaping in microfluidic channels⁴ and (2) UV photopolymerization of the shaped fluid stream.⁵

The optofluidic device uses two sheath fluid streams of poly(ethylene glycol) diacrylate (PEG-DA) from the side channels, sandwiching a photosensitive core fluid stream, PEG-DA with photoinitiator 2,2-dimethoxy-2-phenylacetophenone (DMPA). The first step requires permanent deformations in the central fluid stream, generated by strategically placed perturbation pillars that induce secondary flows (Fig. 2a). This requires the presence of slight inertial effects at the microscale where fluid fields irreversibly deform in a laminar flow regime (i.e. Reynolds number, Re = 10–100). This flow field deformation is enhanced with the addition of more pillars in the downstream direction. Numerical analysis shows that a series of side pillars can transform the side view velocity field of the center stream from a rectangle to an I shape at Re = 14.58, while little change is observed in Stokes flow (Re = 0.04). This flow shaping was demonstrated experimentally using co-flows of non-fluorescent sheath side streams with a Rhodamine B dyed core fluid stream, yielding significantly greater distribution of fluorescence intensity across the width of the stream for the inertial flow case. The second step of the fabrication process is similar to the aforementioned SFL technique. Downstream to the pillars, the shaped core fluid is exposed to UV light that is illuminated against a pre-designed photomask (Fig. 2a).

Fig. 2 Generation of complex 3D particle using optofluidics. (a) The flow field core photosensitive stream is inertially shaped with pillar obstacles, and is subsequently exposed to UV light against a patterned mask. (b) Keeping Reynolds number and photomask shape constant, the pillar arrangement can stretch or compress the cross-sectional area of the stream. (c) Unique pillar arrangements can yield predictable flow fields in fluid dynamic simulations which can then be applied with the photomask of choice to produce high-resolution intricate particles. All images were adapted from Paulsen et al. with permission.⁵

To start, the authors demonstrate the ability to shape fluid with central pillars vs. side pillars by masking with a simple slit-shaped photomask, exposing the shaped fluid after each pillar (Fig. 2b). In addition to pillar number and configuration, variation of Re was tested. Higher Re is characteristic of higher inertia, generally, leading to greater perturbations. Finally, as proof of concept, complex 3D particles were generated by combining flow stream sculpting and UV light shaped by a photomask (Fig. 2c).

In this work, the authors have demonstrated a high-throughput, inexpensive, continuous flow system for generating complex 3D particles in a systematic manner, overcoming the speed and resolution limitations of 3D printers, and the complexity limitation of other microfluidics techniques. One limitation of this work may be the non-intuitive translation of pillar arrangement to projected inertial flow field. Further, a better understanding of the space of shaped particles that can be fabricated is needed and how particle size can be scaled down.⁶ Nevertheless, the work should open up whole new classes of microparticles that should enable innovations across a variety of fields.

Scaling microparticle production with contact flow lithography

A challenge for microfluidic material fabrication, including the techniques discussed above, is scaling to higher production rates. Current technologies, for example, droplet-based microfluidic techniques and imprint lithography, face challenges to manufacture particles with non-spherical shapes in a high-throughput manner. Various approaches to flow-through photolithography have been developed to automate and scale up anisotropic microparticle synthesis. However, there is a bottleneck in particle synthesis rate because the illumination area with a lens-based setup is severely limited.

Le Goff et al. report a contact photolithography system to achieve an ultrahigh production rate of particle synthesis, about 10⁶ particles per hour.⁷ In the system, collimated illumination, generated by a high power UV light-emitting diode (LED), exposes eight channels (950 μm × 10 mm) simultaneously through a photomask in direct contact with the microfluidic device. The system produced ~6000 100 μm sized particles for each UV exposure (see Fig. 3a and b) with a cycle time of about 7.5 s between exposures. The authors show agreement of the qualitative shape and highly uniform size (CV = 3.3%) demonstrating the reproducibility of the approach (see Fig. 3c). As an add-on value of the instrument, the authors demonstrated the ability to pattern one- or two-layer hydrogel microstructures across 23 mm circular areas. In sum, particle synthesis using contact flow lithography enables many fundamental studies which require large numbers of particles and paves the way to create microstructures at an industrial scale, potentially benefiting the fields of rheology, drug formulation, biosensing, cell culture, and beyond.

Fig. 3 Contact flow lithography over large areas. (a) Schematic of the experimental setup for contact flow lithography. (b) Images of fabricated particles in eight channels after one UV exposure. (c) PEGDA hydrogel particles collected outside of the fabrication system. All images were adapted from Le Goff et al. with permission.⁷

References

1. M. Okamoto and B. John, Prog. Polym. Sci., 2013, 38, 1487–1503.
2. S. A. Bencherif, R. W. Sands, D. Bhatta, P. Arany, C. S. Verbeke, D. A. Edwards and D. J. Mooney, Proc. Natl. Acad. Sci. U. S. A., 2012, 109, 19590–19595.
3. D. R. Griffin, W. M. Weaver, P. O. Scumpia, D. Di Carlo and T. Segura, Nat. Mater., 2015, 14, 737–744.
4. H. Amini, E. Sollier, M. Masaeli, Y. Xie, B. Ganapathysubramanian, H. A. Stone and D. Di Carlo, Nat. Commun., 2013, 4, 1826.
5. K. S. Paulsen, D. Di Carlo and A. J. Chung, Nat. Commun., 2015, 6, 6976.
6. D. Stoecklein, C.-Y. Wu, K. Owsley, Y. Xie, D. Di Carlo and B. Ganapathysubramanian, Lab Chip, 2014, 14, 4197–4204.
7. G. C. Le Goff, J. Lee, A. Gupta, W. A. Hill and P. S. Doyle, Adv. Sci., 2015 DOI:10.1002/advs.201500149.

---