In recent years, there has been a surge in methods to synthesize geometrically and chemically complex microparticles. Analogous to atoms, the concept of a “periodic table” of particles has emerged and continues to be expanded upon. Complementing the natural intellectual curiosity that drives the creation of increasingly intricate particles is the pull from applications that take advantage of such high-value materials. Complex particles are now being used in fields ranging from diagnostics and catalysis, to self-assembly and rheology, where material composition and microstructure are closely linked with particle function. This is especially true of polymer hydrogels, which offer an attractive and broad class of base materials for synthesis. Lithography affords the ability to engineer particle properties a priori and leads to the production of homogenous ensembles of particles. This review summarizes recent advances in synthesizing hydrogel microparticles using lithographic processes and highlights a number of emerging applications. We discuss advantages and limitations of current strategies, and conclude with an outlook on future trends in the field.
►Lithography yields microgels with sophisticated morphology and functionality. ► These processes provide new motifs for design of particles and suspensions. ► Process scalability has led to new avenues of research and applications.
Keywords: Hydrogels; Microgels; Synthesis; Lithography; Microfluidics; Assembly; Delivery
- Lithographic patterning of hydrogel colloids
- 2.1. Materials for cross-linked, functionalized, and composite hydrogels
- 2.2. Lithographic synthesis processes for hydrogel colloids
- 2.2.4. Particle recovery
- 2.2.5. Process throughput
- 2.2. Lithographic synthesis processes for hydrogel colloids
- Properties of colloidal hydrogels
- 3.1. Particle architecture: size, shape, and chemical anisotropy
- 3.2. Particle microstructure: cross-linked networks and porosity
- 3.3. Colloidal behavior and stability
- 3.2. Particle microstructure: cross-linked networks and porosity
- 4.1. Particle assembly
- 4.2. Micromechanical systems
- 4.3. Targeted delivery and encapsulation of biologicals
- 4.4. Molecular separations and multiplexed detection
- 4.5. Economic considerations
- 4.2. Micromechanical systems
Cross-linked hydrophilic polymer gels, or “hydrogels”, have become an important class of materials for applications in nanotechnology, biotechnology, and medicine due to their unique material properties . They can be prepared from a wide variety of natural and synthetic precursors ranging from commodity to designer chemicals, and as a result, can be readily commercialized . Hydrogels are highly swelling in water and combine the ability to transport molecular and nano-scale species throughout the material while maintaining solid-like mechanical properties . This feature has led to the widespread use of hydrogels as a scaffolding material in biomedical applications including drug delivery, tissue engineering, and wound healing , ,  and . Their microstructure and interfacial properties can also be rendered responsive to various stimuli through chemical and physical cues, as well as applied fields, resulting in “smart” materials which can respond to their local environment.
Hydrogels are typically prepared and processed as bulk materials such as monolithic structures or supported films. However, emerging applications require miniaturization and tailoring of the hydrogel architecture at increasingly small length scales for delivery and transport purposes in microscopic environments. This has spurred the development of various processes for the synthesis of colloidal and nanoparticle hydrogels, or “microgels”. The majority of these techniques are based on traditional batch polymerization methods such as dispersion and emulsion polymerization . More recent techniques based on microfluidic methodologies provide a higher degree of control over the size distribution of the microgels . In either case, droplets of a polymerizable phase are generated within an inert carrier phase and stabilized by a surfactant . The droplets are subsequently polymerized, after which they must be transferred from the carrier fluid to an appropriate solvent depending on further processing or end-use application, eventually yielding a suspension of colloidal hydrogel particles.
Microgels serve as model “soft colloids”, as they are easily stabilized and their swelling and mechanical properties can be tuned using various physicochemical stimuli . They have become an important class of materials for various aspects of fundamental colloid science, including colloidal interactions, phase behavior, and rheology , , , ,  and . Furthermore, microgels are being developed for a number of potential applications in nanomaterial synthesis , optics and photonics ,  and , and medicine . Although utilized for a wide range of fundamental and applied studies, the traditional “bottom-up” microgel synthesis techniques mentioned above have generally been limited to the production of spherical or spheroidal particles with uniform microstructure and surface chemistry, as well as moderate polydispersity. As a result, these techniques are ill-suited for emerging applications that demand high degrees of both uniformity and complexity and that call for the use of chemically and morphologically anisotropic particle architectures .
These technological requirements have driven the development of advanced methodologies for the synthesis of “designer” colloidal hydrogels with a number of desired process and particle characteristics. The ideal process would be able to produce highly uniform particles with a well-defined shape or chemical anisotropy, with primary particle sizes spanning the colloidal domain. Furthermore, it would be versatile enough to accommodate different hydrogel materials, including changes to the fundamental polymer chemistry, chemical functionalization, and encapsulation of functional materials. Finally, the ideal process would be scalable to quantities of commercial interest, while remaining cost-effective by efficiently consuming reagents.
Lithographic processing methods, in which hydrogel colloids are templated in a highly controlled and reproducible manner, have emerged as an attractive “top-down” alternative to traditional colloidal hydrogel synthesis techniques because they meet many of these requirements . Existing capabilities in microfabrication methods such as imprint lithography, photolithography, and microfluidics have been combined to create designer colloids with tailored architectures and customizable internal microstructures that cannot be achieved with bottom-up processes. Lithographic synthesis has thereby significantly expanded the design space of colloidal hydrogels, enabling enhanced applications in self-assembled and stimuli-responsive materials, micromechanical systems, pharmaceuticals, and medical diagnostics. In this review, we highlight the progress to date in the synthesis and application of complex hydrogel colloids by lithographic techniques. We begin by describing the materials and methods commonly used in lithography of colloidal hydrogels, noting advantages and limitations of current strategies. We then examine the most crucial design properties of colloidal hydrogels, and how these can be controlled using the various forms of lithography-based synthesis. Finally, we outline a number of applications enabled by complex structured hydrogel particles, and conclude with an outlook on future trends and opportunities.
2. Lithographic patterning of hydrogel colloids
2.1. Materials for cross-linked, functionalized, and composite hydrogels
In principle, lithographically-patterned hydrogel colloids can be prepared by any of the routes typically used to produce a cross-linked polymer network, including chemical cross-linking, physical association, and molecular self-assembly . However, the most prevalent method of hydrogel formation in lithographic processing is through covalent cross-linking via polymerization reaction. Nearly all studies to date involve photoinitiated free radical polymerization due to controlled initiation and relatively fast propagation kinetics compared to other types of polymerization. The vast majority of these methods involve the processing of a hydrogel precursor which consists of, at minimum, a photoinitiator, a monomer or reactive polymer, and a cross-linking agent.
Hydrogel precursors used in lithographic processing are typically based on acrylic-functionalized species, including acrylates and methacrylates. The most commonly used precursors are based on poly(ethylene) glycols (PEGs) , ,  and , as they are relatively inexpensive, available in a wide variety of molecular weights and derivative chemistries, biocompatible, and exhibit negligible cytotoxicity . Other biocompatible polymers, such as polylactic acid (PLA) and polyglycolic acid (PGA), both of which exhibit tunable biodegradation , have also been utilized. These systems form a valuable complement of materials for use in a wide variety of applications in biotechnology and high-performance materials. Moreover, co-polymerization of multiple different hydrogel precursors, either by the use of co-polymers in the precursor itself or by random co-polymerization during lithographic processing, provides an additional degree of flexibility in the synthesis of hydrogel colloids, and has been used to create particles with highly tunable equilibrium and dynamic properties, including swelling , biodegradation , and self-assembly .
For many applications, it is desirable to conjugate the hydrogel material with chemical functionality in order to carry out specific reactions within the hydrogel or at the particle surface after synthesis. In situ conjugation with acrylic-functional species can be used for covalent incorporation of functional species directly into the hydrogel matrix. For example, homogeneous or spatially-selective fluorescent labeling provides a particularly useful method for characterization of hydrogel colloids with complex internal structure and chemistry. Several methods have been developed to either covalently incorporate or otherwise encapsulate fluorescent species during lithographic processing, including molecular dyes such as rhodamine , fluorescein  and , and FITC-dextran , as well as fluorescent beads  and quantum dots .
Because lithographic production of hydrogel colloids typically occurs as a single-step synthesis, molecular and colloidal objects can be easily encapsulated into the resulting hydrogel particles, provided that these species are miscible and/or readily suspended in the hydrogel precursor. This can be accomplished by simply adding the material to be encapsulated into the precursor fluid, and provides a facile method to impart functionality and novel properties to colloidal hydrogels. For example, encapsulation of inorganic nanoparticles and colloids has previously allowed for the facile production of colloidal nanocomposites , while encapsulation of electrically or magnetically active species such as magnetic beads  and  has been used to create field-responsive colloids.
The encapsulation of biological and bioactive materials within colloidal hydrogels is of particular interest in biotechnology applications  and . Lithography-based processing techniques provide significant advantages for this purpose, since the bioactive material only contacts the hydrogel precursor (and in some cases, the lithographic template). This is in stark contrast to traditional emulsion or microfluidic droplet-based methods for encapsulation in hydrogel colloids, where the process must be carefully designed in order to ensure that the bioactive material is compatible and non-interacting with process additives such as stabilizers, carrier fluids, and microfluidic materials . Examples of bioactive materials that have been incorporated into colloidal hydrogels by lithographic processing include cells , ,  and , enzymes , proteins  and , and small molecule therapeutics .
2.2. Lithographic synthesis processes for hydrogel colloids
At its essence, lithographic synthesis of hydrogel colloids involves the transfer of a pre-designed template pattern, containing the various geometrical features of the particles to be synthesized, to the hydrogel precursor, followed by (or concomitant with) polymerization and/or cross-linking. The variety of emerging lithographic techniques for the synthesis of complex hydrogel colloids fall into three distinct categories: imprint lithography, photolithography, and flow lithography. Despite differences in format and processing details, these techniques generally share a common workflow methodology, which is illustrated in Fig. 1.
|Full-size image (102K) |
High-quality image (934K)
Schematic workflow diagram for lithographic synthesis of hydrogel colloids, illustrated here for imprint lithography (left), photolithography (center), and flow lithography (right). (I) The patterned template is prepared. (II) The fluid reservoir is filled with a hydrogel pre-cursor fluid(s). (III) Hydrogel colloids are synthesized by simultaneous pattern transfer and cross-linking reaction. (IV) Particles are recovered from the fluid reservoir.
2.2.1. Template design and production
First, the lithographic template is generated, which contains feature patterns that determine the size and shape of the particles to be synthesized (I). In imprint lithography, the lithographic template is a patterned mold, typically polymeric in nature, with negative features corresponding to the particles to be synthesized and which is prepared from a master template . Imprint lithography of hydrogel colloids has developed as an extension of standard soft lithography techniques originally used for the production of patterned polymeric substrates such as microfluidic devices . The first process to demonstrate imprint lithography of individual hydrogel colloids was the particle replication in non-wetting templates (PRINT) method developed by Desimone and co-workers . Similar processes and variations have been presented in the literature  and . In recent years, these processes have enjoyed significant enhancements in performance due to advances in microfabrication technology. For example, improved resists and write methods for master templates have enabled the generation of two- and three-dimensional patterns with feature sizes below 100 nm . Furthermore, the ability to form template patterns from any free-standing object has led to the interesting ability to template various nanoscale objects and thus replicate them in the form of hydrogel colloids . Unfortunately, the complexity of the hydrogel colloids generated by imprint lithography is constrained by practical limitations in the production of soft templates. Bending and buckling of the lithographic template, for instance, limit the transfer of patterns with high-aspect ratios or internal features .
In photolithography and flow lithography, the template is light (or other radiation) incident on the material to be synthesized, which has been patterned through various optical elements. As a result, the minimum feature size for photolithography is limited by the wavelength of the light used in the diffraction-limiting case. Typically, the light source used is the same as that used to initiate photopolymerization, such that pattern transfer and hydrogel synthesis occur simultaneously. Common processes for photolithographic hydrogel synthesis use transparency masks to pattern the incident light. In most cases, this approach limits the morphology of the structures formed to two-dimensional extruded shapes, where the thickness of the particle is determined by the height of the fluid that is exposed to the patterned light. However, more recent studies have demonstrated the ability to form nearly arbitrary three-dimensional hydrogel structures through photolithography, by three-dimensional projection of patterned light with multi-photon  and , interference pattern , or holographic  and  sources. Alternatively, three-dimensional features can be achieved by pre-forming the hydrogel precursor into a three-dimensionally patterned or flexible mold  and , essentially combining the capabilities of photolithography with imprint lithography. Photolithographic techniques feature several attractive characteristics not offered by imprint lithography. For example, the design and preparation of light-patterning elements such as transparency masks are significantly less resource-intensive than microfabrication. In addition, recent variations have used digital micromirror devices (DMDs) in place of transparencies, resulting in the ability to dynamically change the template pattern during processing, as well as create highly complex and interconnected shapes ,  and  not achievable using imprint lithography.
2.2.2. Loading of hydrogel pre-cursor
Once the lithographic template is prepared, the hydrogel pre-cursor is introduced into a reservoir that is aligned coincident with the lithographic template (II). For imprint lithography, the fluid reservoir is often the negative pattern features of the template itself, which is drop-filled either manually or automatically into the polymer mold . Unfortunately, this approach often results in poor transfer of the template pattern to the hydrogel particles due to a non-uniform filling. To prevent this, other methods use a fluid film that is cast onto a planar substrate . The fluid film is then sandwiched between the planar substrate and soft template, the latter of which is filled by capillary forces  and . A major advance introduced by the PRINT technique is the utilization of perfluoropolyether (PFPE) soft templates and substrates, which are highly non-wetting to both polar and non-polar organic solvents. This allows for the filling of multiple immiscible fluid layers within the template, thereby producing chemically anisotropic particles with simple shapes .
The simplest photolithographic methods involve pattern transfer into a stationary fluid film or reservoir ,  and  prepared in a manner similar to that used in imprint lithography. Recently, a number of so-called “flow lithography” techniques have been introduced, in which photolithography is performed in situ within a microfluidic environment . Initial iterations of flow lithography involved the continuous flow of the hydrogel precursor within a microfluidic channel, which was exposed at regular intervals to ultraviolet (UV) light patterned by a photomask to produce discrete particles . Because of the interplay between polymerization and flow, this resulted in poor transfer of pattern features to the synthesized particles. This shortcoming led to the development of stop-flow lithography (SFL), in which the flow of precursor is stopped prior to a patterned exposure, resulting in more precise pattern transfer for feature sizes approaching 1 μm  and .
Flow lithography involves the introduction of easily manipulated laminar microflows into photolithographic processing. As such, flow lithography techniques such as stop-flow lithography (SFL)  and optofluidic maskless lithography  have led to a number of new motifs of hydrogel colloid synthesis previously unavailable using droplet-based microfluidic techniques . For instance, the ability to create co-flowing layered streams allows for the facile production of hydrogels with multiple polymer chemistries on the same particle of arbitrary one-dimensional configuration . Unlike imprint lithography, the different polymer chemistries need not be immiscible, providing for nearly arbitrary multi-functionality. Furthermore, if the co-flowing precursor streams are immiscible, the interfacial tension between the fluids can be used to induce three-dimensional curvature to the particle shape . Initial studies were limited to co-flowing streams in the plane of the pattern transfer, lowering throughput and limiting chemical anisotropy to “striped” particles . However, by stacking the different chemical layers vertically with respect to the incident irradiation, these limitations can be avoided, allowing for synthesis of particles with more complex designs of chemical “patchiness” . As mentioned previously, the geometry of the microchannel in which the flow lithography is performed can also be adjusted in order to impart three-dimensional features to the hydrogel particles  and . Combined with the ability to alternate the flow of different precursor streams, this design strategy can be leveraged to impart highly sophisticated chemical patterns to a single hydrogel particle .
2.2.3. Polymerization and cross-linking
After aligning the fluid with the template features, the polymerization reaction is initiated (III), resulting in the formation of cross-linked particle structures within the fluid. In the common case of photopolymerization, this is done by exposure of the fluid to a light source for a pre-defined period of time dictated by the polymerization kinetics. This requires the fluid, fluid reservoir, and/or template to be transparent to the required wavelength(s) of radiation. Typically, the fidelity with which the patterned template features are transferred to the synthesized polymer structure depends critically on several factors relating to the polymerization kinetics and mass transport within the precursor fluid. In photolithography processes, the rate of diffusion of reacting species relative to the rates of propagation and termination defines the lower limit of the feature resolution . Furthermore, the presence and transport of polymerization-quenching species such as oxygen will provide further limitations on the fidelity of pattern transfer, although such inhibition provides a particularly important role in flow lithography techniques, as it prevents the adhesion of particles to the microfluidic device walls .
2.2.4. Particle recovery
After the polymerization has completed (or is quenched), the particles are then recovered from the fluid reservoir (IV). This final step is often difficult in practice, as the newly synthesized particles can adhere to the template and/or fluid reservoir depending on the surface chemistry and wetting properties of the materials used, as well as the way in which the fluid is filled. In the case of imprint lithography, there is often a residual polymerized film which connects adjacent particles that must be eliminated before isolated particles can be recovered . First, the particles and so-called “flash layer” are isolated from the patterned mold, typically by mechanical delamination. The flash layer is then eliminated by various physical or chemical means such as wet or dry etching . In the case of the PRINT method and its variations, careful design of the mold chemistry results in a fluid seal which prevents the formation of a flash layer . Even so, difficulties arise in removing newly formed particles from their mold features. More recent iterations of the PRINT method harvest particles from the mold by adhering a polymer that is wetting to the hydrogel particles but not to the PFPE mold, which is then used to mechanically delaminate the particles from the mold. Subsequent agitation of the adhesive allows for the recovery of the particles . In other cases, the mold material was cleverly chosen such that it can be dissolved or swollen in order to remove the synthesized hydrogel colloids .
Since photolithographic techniques do not require the contact of the precursor fluid with the patterned photomask, particle recovery is much simpler than for imprint lithography. For stationary photolithography, the newly formed hydrogel particles can simply be washed from the film or fluid reservoir, provided that the precursor is non-wetting and the hydrogel does not adhere to the substrate. This can be ensured through a proper choice of substrate chemistry . In the case of flow lithography techniques, the presence of flow is advantageous in that the newly formed hydrogel particles can be freely advected away from the region of synthesis. Other specific features of conventional microfluidic fabrication and operation can also aid in particle recovery. As mentioned previously, in SFL, oxygen-induced inhibition of the polymerization reaction is heightened near the walls for oxygen-permeable materials such as poly(dimethylsiloxane) (PDMS) , thereby preventing newly formed hydrogel particles from adhering to the synthesis channel walls. In the case of hydrodynamic focusing lithography (HFL), the ability to co-flow non-polymerizable fluid layers between the particles and the channel walls enables the use of oxygen-impermeable microfluidic materials of construction such as glass . For cases in which the microfluidic synthesis channel has three-dimensional features that are polymerized into or around, the particles can either be held in place in order to perform additional synthesis steps, or can be released upon the application of a pressure drop capable of deforming the microchannel  and .
Polymeric materials synthesized by lithography-based processes typically require some form of purification after recovery in order to remove unreacted material or other impurities left over from the synthesis process. For hydrogel colloids, this is typically accomplished through a solvent exchange in which the particles receive several “washes” against a desired final suspending medium. Isolation of the particles during these wash steps can be aided by centrifugation, filtration, or magnetic separation . However, since hydrogel materials are soft by nature, the conditions under which purification is performed must be carefully chosen to ensure that the hydrogel colloids are not adversely affected by extreme shear or compression. In the case of microfluidic-based flow lithography techniques, the particles can be transported and manipulated downstream under gentle laminar flow conditions, allowing for facile additional processing in situ after synthesis, such as purification , functionalization , or assembly .
2.2.5. Process throughput
The utility and eventual commercialization of particle synthesis technologies depend critically on process throughput and yield, as well as operating cost. Since traditional soft lithography techniques are now fairly routine, both imprint lithography and microfluidic-based flow lithography of hydrogel colloids present a minimal expense relative to material costs. However, because of limitations on contemporary microfabrication, both techniques have suffered from fairly low throughput. Imprint lithography suffers due to limitations on the size of molds that can be prepared from typical soft lithography methods. For example, the first demonstrations of the PRINT method reported throughputs on the order of several milligrams per hour of material . Significant progress has been made recently in scaling up the process, which can now be performed via roll-to-roll processing with throughputs on the order of ten grams of material per hour . In the case of photolithography, the primary limiting factor is the available field of irradiation. Since most photolithography-based processes to produce colloidal hydrogels are based on optical projection, throughput is limited to the field of view of the objective used