Bingjie
Wang
ab,
Pepijn
Prinsen
b,
Huizhi
Wang
c,
Zhishan
Bai
*a,
Hualin
Wang
a,
Rafael
Luque
b and
Jin
Xuan
*c
aState Environmental Protection Key Laboratory of Environmental Risk Assessment and Control on Chemical Process, School of Mechanical and Power Engineering, East China University of Science and Technology, Shanghai 200237, China. E-mail: baizs@ecust.edu.cn; wanghl@ecust.edu.cn
bDepartamento de Quimica Organica, Universidad de Cordoba, Campus de Rabanales, Edificio Marie Curie (C-3), Ctra Nnal IV-A, Km 396, Cordoba, E14014, Spain. E-mail: q62alsor@uco.es
cSchool of Engineering and Physical Sciences, Heriot-Watt University, Edinburgh, EH14 4AS, UK. E-mail: j.xuan@hw.ac.uk
First published on 20th January 2017
This article provides an up-to-date highly comprehensive overview (594 references) on the state of the art of the synthesis and design of macroporous materials using microfluidics and their applications in different fields.
Many applications require an open cellular structure with a specific average pore size and distribution.19 In particular, the need for hierarchical structures pleads for research in the field of micro-manufacturing. Hierarchical structures are those whose pores exist on different length scales from micro- to meso- to macropores regardless of their assembly being ordered or not. Obtaining hierarchical porosity is not a straightforward process; often it is found that the macroporous structure collapses during the creation of micropores or otherwise. Hierarchical porous systems have obvious advantages over monomodal ones for several reasons.30–46 In the first place, the hierarchy in porosity enables the selective entrance and exit of chemicals with different molecular dimensions while reducing the diffusion limitation effects.47 To understand this, it can be compared with the time needed for people to arrive at work by car; no matter the amount of available car parking areas (micropores), people will not arrive in time when the highways (macropores) or the exits to parking areas (mesopores) are not well designed. In this comparison, ‘people’ does not only refer to chemicals but also to energy such as electric currents30 or optical signals. A porous hierarchy facilitates the interaction with substrates of different sizes such as small molecules, biomacromolecules and even nanoparticles, and it allows better functionalization of the material itself.
Among others, suspension, precipitation, dispersion and seeded emulsion polymerization, membrane/microchannel emulsification and microfluidics are the main techniques for the fabrication of porous materials. Pore formation can be regarded as a process of phase separation, either in suspension or in swelling and polymerization. The pore size can be indirectly controlled by regulating several parameters, for instance, the effect of the ratio of a cross-linking monomer to porogen, the reaction temperature, the swelling ratio of the seed polymer, etc.48–53 However, when it comes to precise and strict control of the final morphology, the microfluidics technique is more attractive. Flow lithography, which combines microfluidics and photolithography, has emerged as a novel and efficient approach to fabricate complicated microparticles of equal size and specific shape in a continuous way.54
To date, several excellent reviews on the fabrication and applications of porous microparticles have been published already.47,55–75 The review by Okay55 in 2000 dealt with particle characteristics and fabrication methods of macroporous copolymers, but without an in-depth discussion of applications. Serra76 reviewed the use of microfluidic devices for the fabrication of microparticles, focusing on the final particle morphology as a function of the device geometry. In 2010, Kim71 reviewed the state-of-the-art fabrication of microparticles and microcolloids by microfluidic techniques, but with a focus on biology rather than on more widespread applications. In 2012, Gokmen58 published a comprehensive and complete review on porous polymer particles including their synthesis, characterization, functionalization and applications. However, the microfluidic method was introduced only briefly. Besides, it was not clear whether to compare the latter in terms of pore sizes. Testouri75 reviewed how millifluidic techniques can be used to generate porous solids with highly monodisperse and ordered pore structures and how to construct a versatile lab-on-a-chip. In summary, the existing reviews deal with different fabrication methods of porous materials in general. The reviews that describe the microfluidic method more in detail focus mainly on the fabrication of spherical morphologies and their deformations, such as hemispheres or rods. However, a comprehensive review pinpointing the fabrication of macroporous materials using microfluidic techniques is still lacking. This review aims to provide these insights, including their fabrication, functionalization and application. Moreover, not only randomly distributed macroporous materials are described, but also the inverse-opals and scaffolds with well-ordered uniform macropores. Moreover, attention is paid to materials with hierarchical porosity, as mentioned above. Ultimately, we outline the bottlenecks and future developments in this particular field.
![]() | ||
Fig. 2 Scheme of continuous flow photolithography. Reproduced with permission from ref. 54. Copyright 2006 Nature Publishing Group. |
Method (discovery) | Diameter of beads (μm) | Minimum CV of beads | Fabrication easiness/cost | Schematic illustration | Ref. |
---|---|---|---|---|---|
Suspension polymerization (1920s) | 5–2000 | Very high | Easy/cheap |
![]() |
55, 58, 59, 97 and 109–112 |
Dispersion polymerization (1970s) | 0.1–20 | 2–3% | Easy/cheap |
![]() |
55, 58, 59, 110, 111, 113 and 114 |
Seeded suspension polymerization (1980s) | 0.5–200 | 2–3% | Cheap/time consuming |
![]() |
55, 58, 97 and 110 |
Precipitation polymerization (1990s) | 0.1–8 | 2–3% | Easy/costly |
![]() |
55, 58, 110, 111, 115 and 116 |
Membrane/microchannel emulsification (1990s) | 10–1000 | 10% (membrane) or 2–3% (channel) | Difficult/costly |
![]() |
58, 117 and 118 |
Microfluidics (2000s) | 10–1000 | <1% | Costly device |
![]() |
18, 69, 70, 76, 119 and 120 |
The resins can be applied either as porous monoliths99,102,121–123 or as porous beads.124 Rigid monoliths perform better in high-throughput separation processes and in catalytic reactions.2,98–103 Unfortunately, they suffer from high initial flow resistance and alter ions in the breakthrough curve. In contrast, the size distribution of particulate macroporous resins obtained by milling macroscopic materials to form irregular particles is controlled by subsequent fractionations.125 Though easy to fabricate, these particles show poor packing behavior and may block the column due to the presence of tiny secondary particles.97 Alternatively, heterogeneous techniques such as suspension polymerization can be practical and useful for the synthesis of microspheres with diameters in the range of 0.1–1.5 mm and relatively broad size distributions.12,55
Suspension polymerization53,55,96 and seeded emulsion polymerization104,105 are among the most common preparation methods. In the former, drops of a mixture of a monomer, an initiator and an inert solvent (porogen) are dispersed in the immiscible continuous phase aided by vibration jetting, stirring or sonication. Often, the residual porogen is more miscible with the monomer rather than with the polymer phase. Residual porogens and monomers can be removed by solvent extraction or steam distillation, while enhancing the particle porosity. Generally, particles <200 μm are obtained with a broad range of sizes, depending on the precise control of the emulsification process and the hindrance of droplet coalescence in the polymerization stage. In contrast, seeded emulsion polymerization can render polymer particles <50 μm with more narrow size distributions. This occurs in 2 steps:104–107 (1) quick swelling of monodisperse latex emulsions (seeds) followed by polymerization to solid particles and (2) removal of immiscible solvent residues to form a porous structure. Using seed particles with a low cross-linking degree is time-consuming, especially when particles >100 μm are envisioned. For macroporous particles with diameter <10 μm, precipitation polymerization in organic solvents can be used. For instance, toluene and xylene were used as co-solvents in acetonitrile to fabricate monodisperse highly cross-linked divinylbenzene particles.108
Advantages of microfluidic devices include low reagent consumption, facile mass/heat transfer, high safety and a high surface-area-to-volume ratio, with various outstanding features superior to conventional approaches.126–131 Most importantly, the size distribution of the macroporous particles obtained by microfluidic synthesis can be controlled more precisely thanks to the flow-focusing mechanism of monomer emulsification. Besides, on-chip continuous polymerization prevents droplet coalescence. Particles synthesized by microfluidic synthesis have a finer and more ordered internal structure. However, the generation of particles with diameters in the range of 50–100 μm employing microfluidic methods may be more challenging; particles fabricated by suspension polymerization132 show specific surface areas around 400 m2 g−1 compared to only 30 m2 g−1 for particles with the same composition obtained by the microfluidic technology.133
Microfluidic-based porous particle fabrication methods | Merits | Demerits | Applicability |
---|---|---|---|
Polymerization with porogens |
Wide universality
Wide variety of porogens |
Need post-processing to obtain a porous structure
Pores’ random distribution Unmanageable pore size |
Radially “opened” macroporous particles with regular shape |
Polymerization without porogens | Without post-processing |
For specific phase component
Pores’ random distribution Unmanageable pore size |
Radially “skin” or “opened” macroporous particles with regular shape |
Sacrificial templates |
Excellent pore shape
Uniform pore size |
Complicated process | Ordered or radially hierarchical porous particles with regular or irregular shape |
Flow reaction | High porosity |
For specific phase component and reaction
Unstable pores’ formation |
Radially “skin” or “opened” macroporous particles with regular shape |
Self-assembling | High porosity |
Unmanageable particle shape
Pores’ random distribution Unmanageable pore size |
Radially “opened” macroporous particles with regular or irregular shape |
Flow photolithography |
Customized particle shape
Customized pore shape and distribution |
High cost of designed photomask
Low-yield |
Customized ordered or hierarchical porous particles with regular or irregular shape |
![]() | ||
Fig. 4 SEM images of (a and c) surface and (b and d) internal structure of “skin” macroporous poly(GMA–EGDMA) particles, synthesized in the presence of (a and b) DBP/DOP 1![]() ![]() |
The employment of porogens leads to porosity development,18,138 but it imposes an extra step for the removal of these agents. Watanabe et al.139 reported the controlled preparation of polyelectrolyte particles in the 10–100 μm range through selective solvent extraction in a cross-flow microfluidic device. Mixtures of poly(styrene sulfonate) and water (dispersed phase) were introduced into the T-junction to form plugs in hexadecane (continuous phase). Subsequently, the obtained plugs were introduced into methyl ethyl ketone which is miscible with hexadecane and water but not with poly(styrene sulfonate) in a flow-focusing microchannel (in situ precipitation) or in a prepared vessel (ex situ precipitation) (Fig. 5a and b). Smooth “skin” macroporous particles were obtained (Fig. 5c) with a gradient internal porous structure (Fig. 5d).
![]() | ||
Fig. 5 (left) Illustration of (a) in situ and (b) ex situ preparation via microfluidic emulsification, precipitation and solvent extraction. (right) SEM images of the (c) smooth “skin” surface and (d) internal structure. Reproduced with permission from ref. 139. Copyright 2014 American Chemical Society. |
Besides, Sim et al.140 fabricated microparticles whose top or side surface was not flat but curved, such as circular, square, triangular and stellate microdisks (Fig. 6a–d), by introducing a novel double-layered PDMS microfluidic device, combining conventional photolithography and multilayer soft-lithography (microfluidic molding). Two straight control channels, used for pneumatic actuation of the membrane to mold and release the particles, were situated on the top layer while a T-junction channel, used for suspension flow, was integrated in the bottom layer. Interestingly, by controlling the pressurization in the top channel, the membrane was either deformed to confine the suspension into the designed shape and then polymerized by UV-irradiation or recovered to the original shape by releasing the obtained particles. Finally, microparticles could exit through the bottom channel.
![]() | ||
Fig. 6 SEM images of (a) broken porous microparticles with a smooth surface and internal porous structure, (b) hemispherical microparticles, (c) triangular microparticles with an inclined side wall and (d) stellate microparticles with a curved surface. Scale bars are 1 μm (a) and 10 μm (b–d). Reproduced with permission from ref. 140. Copyright 2014 Wiley-VCH. |
The cycle of molding, polymerization and particle release could only be achieved by pneumatic force in the microfluidic device. By introducing a photo curable suspension of silica nanoparticles in the microfluidic molding process and subsequent selective wet-etching of the silica nanoparticles with a HF solution, a polymerized matrix with regular pores and structural colors was obtained.
Unlike the works of Pelzbauer et al.13 and Dubinsky et al.17 in which porogens were used to obtain closed macroporous particles, more recently, Udoh et al.141 employed a microfluidic solvent extraction approach to prepare “skin” polymer particles (Fig. 7). At the same time, they also deeply investigated the effect of polymer solution thermodynamics in the particle fabrication process by comparing two different non-solvents (ethyl acetate (EA) and methyl ethyl ketone (MEK)) in the extraction of a polymer/solvent (NaPPS/H2O) system. The outstanding control of particle size and closed porous structure verified that the solvent extraction method is more convenient than ones that need conventional porogens or a complex post-processing process.
![]() | ||
Fig. 7 SEM images of (A and E) external surface and (B–D and F–H) cross sections of “skin” polymer particles obtained by extraction in MEK and EA. (A, C, E and G) CNaPSS;t=0 = 3.5 wt%, (B and F) CNaPSS;t=0 = 1.0 wt%, (D and H) CNaPSS;t=0 = 10 wt%. Reproduced with permission from ref. 141. Copyright 2016 American Chemical Society. |
![]() | ||
Fig. 8 (a) Illustration of the fabrication steps of golf-ball-shaped microparticles. SEM images including close-up of microparticles with arrays of 235 nm silica nanoparticles (b) before and (c) after the wet etching process. Scale bars in (b) and (c) are 20 μm, while 1 μm in close-up images. Reproduced with permission from ref. 155. Copyright 2009 IOP Publishing Ltd and Deutsche Physikalische Gesellschaft. |
In the same year, Kim et al.158 adapted the technique for the controlled fabrication of hemispherical golf-ball-shaped structures or so-called photonic domes by using photocurable emulsion droplets as templates.156 The patterned polymeric domes (Fig. 9) were prepared in 3 steps: (1) fabrication of ethoxylated trimethylolpropane triacrylate (ETPTA) droplets containing a high load of silica nanoparticles, (2) UV-irradiation of droplets for 10 s inducing polymerization and (3) selective removal of silica nanoparticles on the surface by a HF solution.156,159 Kim et al.160 added a post-treatment of these golf-ball-shaped microparticles by reactive ion etching with sulfur hexafluoride resulting in microparticles composed of super hydrophobic and hydrophilic surfaces.
![]() | ||
Fig. 9 (a) Illustration of patterned polymeric domes, (b) SEM image of a single polymeric dome, including magnification of the porous surface morphology. Scale bar is 100 μm. Reproduced with permission from ref. 158. Copyright 2009 Wiley-VCH. |
Hwangbo et al.161 proposed a novel method to obtain microparticles with both a golf-ball-shaped surface and an internal porous structure using various polymers in a one-pot method (no post-treatment), similar to the breath figure method.148 In their fabrication process, an inert organic volatile phase change material (PCM) was employed, e.g. dichloromethane (DCM). Dimple patterns were created already in the first stage by fast DCM evaporation (Fig. 10a).
![]() | ||
Fig. 10 (a) Illustration of golf-ball-shaped microparticle fabrication. SEM images of the surface (up) and internal structure (down) using (b) poly(D,L-lactide-co-glycolide) (PLGA), (c) poly(methyl methacrylate) (PMMA), (d) polysulfone (PSf), (e) poly(D,L-lactide) (PDLA), (f) poly(L-lactide) (PLLA) and (g) poly(ε-caprolactone) (PCL). Scale bars are 50 μm (b, down), 100 μm (b, up and c) and 200 μm (d–g). Reproduced with permission from ref. 161. Copyright 2011 The Royal Society of Chemistry. |
Along with DCM evaporation, phase separation between the polymer and PCM led to PCM droplet formation inside the oil droplets. The particle solidification process was accompanied by volume shrinkage. Eventually, surface dimples formed on the PCM surface covering the oil phase droplet, while the PCM inside the oil phase droplet contributed to form a closed porous structure. Soft stirring of the aqueous medium before surface solidification generated smaller and more even dimples. Interestingly, the SEM images (Fig. 10b–g) showed that the golf-ball-shaped microparticles exhibited a dimple surface pattern with no pores using amorphous polymers (PSf and PDLA), while the ones obtained with semi-crystalline polymers (PLLA and PCL) contain surface pores.
Based on Hwangbo et al.'s work, Lee et al.162 could enhance the diameter of the surface dimples (thereby lowering the specific number of dimples) by increasing the PCM to PLGA ratio in the same tube-type microfluidic device (Fig. 11). A subsequent study found that because of this unique structure, active agents can be released faster from surface dimples than from interior dimples, which opened routes for drug release strategies. Compared to hollow and open porous microparticles, golf-ball-shaped ones can avoid interference from the external environment with the internal encapsulated material.
![]() | ||
Fig. 11 SEM images of golf-ball-shaped microparticles with PLGA![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
![]() | ||
Fig. 12 (a) Illustration of the cryo-polymerization process for the synthesis of macroporous microspheres. SEM images of (b) surface morphology and (c) internal structure. Scale bars are 500 μm (b) and 50 μm (c), respectively. Reproduced with permission from ref. 168. Copyright 2012 Elsevier B.V. |
Dubinsky et al.133 reported a semi-continuous photo-initiated microfluidic synthesis method of macroporous polymer microspheres in the size range of 10–100 μm with narrow size distribution. Monodisperse droplets were obtained with different compositions by employing different porogens: diethyl phthalate (DEP), diisobutyl phthalate (DBP), dioctyl phthalate (DOP) and diisodecyl phthalate (DDP), followed by UV-induced polymerization and washing with methanol and acetone for porogen removal (Fig. 13). The surface of the microspheres had a well-defined and interconnected porous structure, depending on the type of porogen used. The mean particle diameter increased in the order DEP < DBP < DOP < DDP, while the specific surface area decreased in the same order. Dubinsky et al.17 added variations to this preparation method by applying not only different porogens, but also their mixtures. Fig. 14 shows the surface morphology and the internal structure of macroporous poly(GMA–EGDMA) particles using a DBP:
DOP (3
:
1 v/v) mixture as a porogen. Interestingly, the interconnected morphology remained, but the surface morphology changed to porous or smooth. Additionally, they built a map (Fig. 15) demonstrating the relationship between the solubility parameter and the interfacial tension “map”, which predicts the surface morphology as a function of the monomer and porogen.
![]() | ||
Fig. 13 SEM images of surface morphology of macroporous microspheres prepared with (a) DEP, (b) DBP, (c) DOP and (d) DDP as porogens. Scale bar is 500 nm. Reproduced with permission from ref. 133. Copyright 2008 American Chemical Society. |
![]() | ||
Fig. 14 SEM images of the (a) internal structure and (b) surface morphology of macroporous microbeads, generated with mixtures of DBP![]() ![]() ![]() ![]() |
![]() | ||
Fig. 15 Map of interfacial tension (γ) and solubility parameter (δ) of porogens using a 2.5 wt% aqueous solution as the continuous phase and different porogen solvent compositions as the dispersed phase. Reproduced with permission from ref. 17. Copyright 2009 American Chemical Society. |
Kim et al.156 introduced a two-step method composed of photopolymerization and wet etching for the synthesis of porous photonic spheres in a co-flow capillary device (Fig. 16a). Interestingly, the pores packed on the outer layer showed well-ordered arrays, while near the core the structure became less ordered due to the high curvature. By tuning the diameters of the embedded silica particles from 145 to 152 and 190 nm, the emulsion droplets exhibited different colors from blue to green and red, opening potential application in optical research.
![]() | ||
Fig. 16 (a) Illustration of uniform droplets generated in the microfluidic device. SEM images of (b) whole photonic sphere embedded with 165 nm silica particles and (c) porous surface of the sphere. Reproduced with permission from ref. 156. Copyright 2008 Wiley-VCH. |
The introduction of porosity in microbeads requires porogens which are emulsified with other microbead components during the synthesis. To produce the pores effectively, porogens must be removed in a second leaching process, which requires harsh conditions. Therefore, the removal may be assisted by taking away the encapsulated active compounds also.173,174 This limitation can be overcome by using new miscible porogens or new synthesis strategies, like those that employ permanent geometric templates or scaffolds to avoid the leaching step.175–177 Duncanson et al.169 introduced a self-assembling perfluorinated–dendrimer–dye complex as a novel pore forming agent to generate monodisperse macroporous microspheres (Fig. 17) in a microfluidic flow-focusing device (FFD). After removing the organic solvent, residual water and gas were released from the microspheres under vacuum conditions followed by further drying and solidification of the droplets. The porogen not only acted as a pore formation agent, but also played a role as a carrier for additional active materials. It encapsulated efficiently active materials avoiding their disruptive removal. These microbeads could be applied in liquid–gas mixtures.
![]() | ||
Fig. 17 SEM micrographs of macroporous microspheres made by using the (a) [G 1.5]–dye complex and (b) [G 3.5]–dye complex. Reproduced with permission from ref. 169. Copyright 2012 The Royal Society of Chemistry. |
Jiang et al.178 created a new type of discrete immunebeads using microfluidic microspheres. Three interesting features are associated with their synthesis route: (1) SR454 serving as a trivinyl cross-linker increased the pore dimensions and therefore enhanced the transport and exchange of biomolecules, (2) decyl alcohol was adopted as a safer porogen rather than conventional phthalate plasticizers, without compromising the precise control of the final pore morphology and (3) the microfluidic apparatus was simple and cheap compared to conventional ones.77,179,180 The immune active ligands were first decorated on the porous surface followed by macropore formation (Fig. 18a). The proteins (anti-IgG) were anchored on the porous surface with 2 separate processing routes, using either glutaraldehyde (GLU) or N-gamma-maleimido-butyryloxy-succinimide (GMBS). In the former, the GMA surface epoxides were reacted first with ethylenediamine and then consented with GLU through aldehyde–amine interaction (Fig. 18b).181 The remaining free aldehyde group on GLU was then reacted with IgG primary amines to decorate the antibodies on the polymer surface.182 Some interference was observed due to background fluorescence and unsaturated imine formation. In the GMBS procedure, a thiol group was first grafted onto the monolith via epoxide–cysteamine reaction. Subsequently, GMBS was anchored on the newly formed thiol group, by reacting with the GMBS maleimide terminus. The auto fluorescence interference was repressed in this procedure. As a result, the microspheres grafted with bioactive proteins gradually evolved to discrete immune sensor elements.
![]() | ||
Fig. 18 Illustration of (a) macroporous microsphere fabrication and surface modification for ligand anchoring, (b) GMBS and GLU approaches for anchoring an antibody on the polymer surface and (c) optical image of the macroporous microspheres. SEM images of the (d) whole microsphere, (e) macroporous surface and (f) cleaved microspheres showing a uniform and cellular structure throughout the whole particle. Reproduced with permission from ref. 178. Copyright 2012 Elsevier Ltd. |
Emulsion templating is among the most convenient and commonly used macroporous microparticle fabrication methods. By fabricating monomer-comprised single emulsions as templates in a microfluidic device, with appropriate initiators and different types of porogens, microspheres can be controllably prepared with a monodisperse and highly interconnected structure. Zhang et al.18 proposed a novel and simple method to prepare monodisperse poly(HEMA–MMA) microbeads with macroporous structures, using oil-in-water (O/W) emulsions as templates with 2,2-dimethoxy-2-phenylacetophenone as a photo-initiator and poly(vinyl pyrrolidone) (PVP) as a porogen in a flow-focusing capillary microfluidic device.81,183 The microbeads showed slow biodegradability, increased chemical stability due to their polymeric 3D networks (formed after free radical co-polymerization of MMA with 2-hydroxyethyl methacrylate (HEMA))184 and high specific surface area. SEM images (Fig. 19) illustrate that upon altering the [PVP]/([HEMA] + [MMA]) mass ratio, the final morphology of microspheres varied significantly. Remarkably, the mechanical strength decreases with increasing contents of PVP.
![]() | ||
Fig. 19 SEM images of macroporous poly(HEMA–MMA) microspheres with PVP/(HEMA + MMA) wt% of (a) 3%, (b) 5% and (c) 7%. Scale bar is 20 μm. Reproduced with permission from ref. 18. Copyright 2009 Elsevier Inc. |
High-internal-phase emulsion (HIPE) is another emulsion templating method,185–189 where the volume fraction of the dispersed phase exceeds the 0.74 threshold value for monodisperse emulsions and can reach up to 0.99. PolyHIPE beads contain larger cavities within an average size range of 10–100 μm. These cavities are interconnected by a series of smaller pores that enable them to communicate with the adjacent ones. For their fabrication, first water, an initiator and a salt were mixed together for the formation of the interior phase. Then, this interior phase was mixed with the organic phase and an emulsifier to constitute a water-in-oil (W/O) HIPE. Finally, after polymerization of the continuous phase and removal of the aqueous droplets, porous particles with spherical geometry were obtained190–194 under controlled conditions. Gokmen et al.19,195,196 also fabricated poly(HIPE) microspheres using a microfluidic approach. Both dimension and morphology of the microspheres were precisely controlled. The obtained beads showed significantly smaller pore diameter, even down to 15 μm (Fig. 20). A post-modification was also demonstrated to modify the chemical functionality of the internal structure.
![]() | ||
Fig. 20 (a) Optical microscopy images of porous poly(HIPE) beads. SEM images of (b) whole microspheres, (c) surface of a microsphere and (d) inner structure of the microspheres. Reproduced with permission from ref. 19. Copyright 2009 American Chemical Society. |
Double emulsions such as water-in-oil-in-water (W–O–W) and oil-in-water-in-oil (O–W–O) are complex colloidal systems, where dispersed droplets contain smaller droplets. They are potential devices to encapsulate active substances, e.g. cosmetics,197 drugs,198–200 living cells,201 and food supplements.202,203 Stable double emulsions can be fabricated in fluidic devices based on microchannels, both in PDMS devices and in capillaries. Two-step droplet break-up at the T-junction204,205 and hydrodynamic flow focusing79,128,206,207 were the common setups used. The double emulsions often rendered excellent characteristics, such as uniformity in terms of diameter, shape, wall thickness and the number of encapsulated droplets. Moreover, simple fluidic devices constructed from glass capillary tubes, needles and PVC tubes were also developed for this purpose.196,208–210 Choi et al.175 introduced a fluidic device with one extra fluid channel in the above described system (3 flow channels totally), yielding uniform double emulsions (Fig. 21a). In this W–O–W system, both the W–O emulsion obtained in the first step and the W–O–W emulsion obtained after the second step were prepared in a dripping mode, leading to more uniform dimensions than those obtained in a jetting mode.211,212 The W–O–W emulsion droplets evolved gradually to large interconnected pores (Fig. 21b and c). This method was employed in numerous studies.192,213–215
![]() | ||
Fig. 21 (a) Illustration of the microfluidic apparatus for W–O–W emulsions. SEM images of (b) the porous surface including a close-up (scale bar is 5 μm) and (c) the cross-sections and the internal structure of PLGA microbeads. Reproduced with permission from ref. 175. Copyright 2009 Wiley-VCH. |
Jeong et al.170 created photo- and thermoresponsive macroporous core–shell spheres with reversible membrane permeability using a simple microfluidic device (Fig. 22). Highly uniform W–O–W double emulsion droplets with an aqueous core and interconnected porous shell were generated as templates, whose outer layer is then covered with an ethyl cellulose membrane with densely packed poly(N-isopropylacrylamide) (pNIPAAm) nanoparticles containing gold rods, which endowed the macroporous spheres with thermoresponsivity. The size of the pores was in the range of submicrometers to several micrometers. The ethyl cellulose membrane aided in the mechanical strength and was hardly affected by light, heat and chemicals. Interestingly, the permeability of the pNIPAAm nanoparticle embedded membrane was regulated by temperature, as a function of its phase transition behavior. When the temperature was raised beyond a certain value, the microspheres shrunk and start to allow much bigger molecules to pass through. This feature would not be achieved without the gold nanorods on the microspheres. By using near-infra-red (NIR) irradiation, the reversible membrane permeability could be remote controlled.
![]() | ||
Fig. 22 Illustration of (a) the fabrication and (b) the structure of macroporous core–shell pNIPAAm microspheres with the corresponding SEM images of (c and d) porous surface and (e and f) cross-sections. Reproduced with permission from ref. 170. Copyright 2013 American Chemical Society. |
Next to single and double emulsion templates, a three phase emulsion was used by Wan et al.134 to produce porous polyacrylamide microspheres. The emulsion consisted of gas, water and oil, applied in flow focusing or T-junction devices, FFD (Fig. 23a) and FFT (Fig. 23b), respectively. Individual gas bubbles were encapsulated by monodisperse photopolymerizable acrylamide (AAm) aqueous droplets (Fig. 23c). After UV-irradiation, these droplets gradually evolved to highly porous microspheres. The gas pressure played a major role in the droplet size in the FFT device, but it was only marginally affected in the DFF device.
![]() | ||
Fig. 23 Illustration of the (a) DFF and (b) FFT microchannel geometry of the microfluidic apparatus and (c) optical image of microspheres from three-phase emulsions (scale bar is 50 μm). Reproduced with permission from ref. 134. Copyright 2008 Wiley-VCH. |
The formation of macropores by reaction between reagents, e.g. thiol–ene and thiol–yne reactions,216,217 has become an attractive alternative because it omits tedious templates and porogen removal procedures. Prasath et al.164 presented a method to apply thiol–ene and thiol–yne reactions followed by UV-induced polymerization in a simple co-flow capillary device as a one-step preparation procedure of macroporous microspheres containing different functional groups with different morphologies (Fig. 24). Li et al.218 improved this strategy by two reactions occurring successively in a single droplet: an exothermic free radical polymerization of an acrylate monomer followed by an endothermic polycondensation with a urethane oligomer, activated by the heat generated from the first reaction. As a result, an interpenetrating polymer network is obtained within the macroporous structure.
![]() | ||
Fig. 24 SEM images of (a1–d1) whole microparticles and (a2–d2) surface morphology of macroporous microparticles generated from the “thiol–ene” formation of (a1 and a2) TT/(DAP + AA)α, (b1 and b2) TT/(DAP + AA)β, (c1 and c2) TT/PTEβ and (d1 and d2) (TT + MA)/TTTβ. SEM images of (e1 and f1) whole microparticles and (e2 and f2) relative surface morphology from the “thiol–yne” formation of (e1 and e2) TT/(OY + PA)β and (f1 and f2) TT/(OY + DPPD)β. Reproduced with permission from ref. 164. Copyright 2010 The Royal Society of Chemistry. |
As a variation of this method, Gong et al.163 prepared porous beads via simultaneous reactions within a single droplet in a microfluidic FFD. Monodisperse droplets, containing a H2O2 solution, were fabricated with several precursors: NOA 61 (a mixture of trimethylolpropane diallyl ether, isophorone diisocyanate ester, trimethylolpropane tris thiol and a benzophenone photo-initiator),219–222 EGDMA or tripropylene glycol diacrylate (TPGDA).127,223 The droplets were further encapsulated in the outer phase (either liquid paraffin for NOA 61 or silicon oil for EGDMA and TPGDA) and exposed to UV-irradiation until the desired number of cores was generated. The UV-light and/or heat induced the polymerization of the shell and the decomposition of H2O2, and the precursors were solidified while oxygen and water vapor (after further heating) were released through the shell creating a porous structure (Fig. 25). These porous microspheres had a pore dimension in the range of 1–100 μm and a maximum internal void volume fraction up to 70%, similar to the HIPE microbeads.
![]() | ||
Fig. 25 (a) Illustration of the structural evolution from emulsion to macroporous microspheres. SEM images of (b and c) polyTPGDA microspheres and (d) cross-section of polyNOA microspheres. Scale bar is 50 μm (c). Reproduced with permission from ref. 163. Copyright 2009 American Chemical Society. |
The microfluidic technology provides a method to precisely control the fabrication of porous beads. However, microfluidic devices are often limited in channel size and therefore complications may arise when materials are difficult to solidify, e.g. when they are light/heat sensitive or absorb light too strongly. Photopolymerization or photo cross-linking can help the precursors to react and solidify more rapidly. Superparamagnetic porous microbeads can be subjected to an external magnetic field to simplify the particle separation process.224–226 Moreover, particles can be selected on the basis of size and magnetic susceptibility in miniaturized on-chip systems.227 They are fabricated with porogens or double emulsions in microfluidic channels with good control over size and distribution.17–19,58,133,134,163,228
However, superparamagnetic iron oxide nanoparticles (SPIONs) absorb light strongly, thus being more difficult to photopolymerize. Paquet et al.167 addressed this limitation by preparing superparamagnetic microbeads based on the assembly of materials rather than any other chemical transformation such as polymerization or cross-linking, so making it suitable for fabrication with materials which are sensitive to light or heat. Different microbead morphologies resulted from alterations in the assembly processes, depending on the rate of solvent depletion from the droplet, the absorption of sodium dodecyl sulfate (SDS) at the droplet interface and the solubility of the SPIONs in the droplet phase. By adjusting the SDS concentration (Fig. 26a–d) and the polarity of the disperse phase (Fig. 26e–h), the porosity gradually evolved from non-porous to a dense and opened cellular structure. Wacker et al.166 combined evaporation-induced solidification and silica nanoparticle self-assembly to fabricate porous microspheres. Upon changing the nanoparticles’ dimension from 50 to 500 nm, the porous microspheres present different surface morphologies (Fig. 27). Their hypothesis states that if the pores were too small, evaporation of water to vapor affected the final morphology resulting in wrinkled surfaces. In contrast, when the pores were wide enough, the water vapor passed through the surface without any obstacle.
![]() | ||
Fig. 26 SEM images of the internal structure of the macroporous microbeads collected in (a) 0, (b) 25, (c) 50 and (d) 100 mM SDS solutions, and of the microbead interior using (e) chloroform, (f) THF![]() ![]() ![]() ![]() |
![]() | ||
Fig. 27 SEM images of porous spheres assembled by silica nanoparticles with increasing diameter (d/nm) and colloid weight concentrations (c). Scale bars are 10 μm. Reproduced with permission from ref. 166. Copyright 2011 American Chemical Society. |
![]() | ||
Fig. 28 (a) Optical microscopy of poly(HIPE) rods. SEM images of (b) partial rod and (c) porous surface of a rod. Reproduced with permission from ref. 19. Copyright 2009 American Chemical Society. |
Wacker et al.166 also obtained rod-like micro-objects, deformed from porous spheres (Fig. 29), in this case by evaporation-induced solidification and silica particle self-assembly. Increasing the oscillation time resulted in more elongated shapes and a too high oscillation frequency led to smaller clusters rather than porous rod-like structures.
![]() | ||
Fig. 29 SEM images of porous rod-like particles. Reproduced with permission from ref. 166. Copyright 2011 American Chemical Society. |
Important in view of possible applications, the morphology of microgels directly affected the response rate to external stimuli. For instance, conventional poly(N-isopropylacrylamide) PNIPAAm microgels usually showed a homogeneous but disconnected internal structure, which slowed down the response rate.230,231 Several attempts were made to improve the response rate of microgels, such as constructing a more heterogeneous internal structure,231–235 grafting functional groups onto the microgels236,237 or employing different porogens to improve the porosity,238–240 albeit with only moderate success. But, Chu and co-workers183 generated monodisperse pNIPAAm microgels with spherical voids via embedding solid polystyrene microbeads inside and then dissolving them with xylene, giving significantly improved response rates. Huang and his team241 also successfully fabricated pNIPAAm microgels with a good porous structure and response rate, using polyethylene glycols as porogens. Yue et al.242 improved their response rate as well, but by introducing nanogels into the microgels. Still, these fabrication methods are tedious.
pNIPAAm microcapsules can shrink upon exposure to external stimuli like heating and to certain chemicals, which finally ruptures the hydrogel surface membrane, while the oil cores inside the microcapsules squeeze out.243–246 Based on this, Mou et al.171 fabricated environmental stimuli-responsive microgels with a highly interconnected open cellular structure by combining homogeneous emulsification and polymerization by UV-irradiation (Fig. 30). The initial homogeneous O/W emulsions served as the inner phase and soybean oil with 5% w/v PGPR as the outer phase, which were successively injected into a co-flow capillary device to obtain monodisperse (O/W)/O emulsions. Upon exposure to UV-irradiation in an ice-water bath, followed by addition of isopropanol or heating, pNIPAAm microgels with open internal structures were obtained through volume shrinkage of the microgels. They compared them with microgels obtained under identical conditions but with conventional emulsions, where the aqueous phase replaced the O/W emulsion. The (O/W)/O directed microgels exhibited a much higher response to environmental stimuli, such as temperature, pH, etc.247–249 These results make them attractive for application in sensors, separations and actuators.250,251
![]() | ||
Fig. 30 (a) Illustration of porous pNIPAAM microgel formation. SEM images of pNIPAAm microgels obtained with (b1 and b2) W/O emulsions and (c and d) (O/W)/O emulsions, with molar ratios of MBAAm to NIPAM of (c1 and c2) 2% and (d1 and d2) 4%. Scale bars are 50 μm (b1, c1 and d1) and 20 μm (b2, c2 and d2). Reproduced with permission from ref. 171. Copyright 2014 American Chemical Society. |
Thin membranes can be fabricated in a microfluidic system, by cherry-picking the immiscible solutions and precisely tuning the viscosity to attain laminar flows, followed by polymerization at the liquid–liquid interface. However to date, thin membranes formed using conventional approaches have very small pore size and sometimes even lack pores, which limits their use in biological applications such as cell migration and molecule detection. Based on the study of surface tension and laminar flow behavior,252–254 an in situ fabrication method was reported by Kim and Beebe,255 relying on the interfacial reaction of a two-phase system to produce a porous nylon membrane with pore dimensions 0.1–1 μm (Fig. 31c–e). This approach made it possible to generate very thin porous membranes, firmly fixed inside the microchannel without any other bonding process. As shown in Fig. 31a, 2 partially separated microchannels existed in the fluidic apparatus.252–254 After washing one microchannel with hexadecane and the other one with a mixture of hexadecane and octadecyl trichlorosilane, the 2 microchannels turn hydrophilic and hydrophobic, respectively. Then an aqueous solution and organic solution, respectively, were introduced into the microchannels. 1,6-Diaminohexane in water and adipoyl chloride in toluene reacted at the interface resulting in a membrane. After 15 min, the aqueous and organic solutions are replaced with MeOH and toluene, respectively, followed by washing and drying. Most of the PEG (poly(ethylene glycol) was removed with the aqueous phase as it is highly soluble in water. Fig. 31b shows that different morphologies existed on both sides (asymmetric membrane). The organic side showed a ridge-and-valley structure (2–5 μm), while the aqueous side had a smooth skin surface (about 0.5 μm). Thus, the polymer transport occurred from the aqueous solution side to the organic solution side. This technique was also employed for the formation of chitosan and polycarbonate porous membranes.256,257
![]() | ||
Fig. 31 (a) Illustration of interfacial membrane fabrication and SEM images of (b) a porous membrane formed by interfacial polymerization (angled view) and membranes obtained with (c1 and c2) 0, (d1 and d2) 10 and (e1 and e2) 20 mg PEG in a 60% 1,6-diaminohexane aqueous solution. The magnifications are 10![]() ![]() |
Materials can also be decorated with a 3D porous “coat”, via photo-induced radical polymerization in a microfluidic device. Wang et al.172 fabricated a sensitive surface-enhanced Raman scattering (SERS) optrode from GMA, ETPTA and DMPA, densely decorated with silver nanoparticles, covering the tip of an optical fiber (Fig. 32). The porous layer thickness and the pore size increased with irradiation time, ranging from 4 to 149 μm and from 0.4 to 1.0 μm for 1 and 4 min polymerization time, respectively.
![]() | ||
Fig. 32 (a) SERS optrode-integrated microfluidic chip. SEM images of porous polymers covering the tip of optical fibers, prepared with UV-irradiation time (b) 1, (c) 2, (d) 3 and (e) 4 min. The insets are side views. Reproduced with permission from ref. 172. Copyright 2012 American Chemical Society. |
More recently, Liu et al.258 unprecedentedly reported a direct and versatile approach that combines a modified electrospinning setup based on a microfluidic nozzle with nonsolvent-induced phase separation to fabricate ultrafine fibers with interconnected macropores (Fig. 33). It exhibited higher specific surface area (48.66 ± 8.30 m2 g−1), larger pore size (116.73 nm) and pore volume (0.169 ± 0.007 cm3 g−1) in comparison to conventional electrospun porous fibers. In brief, two immiscible solvents, a polymer solution (20% polystyrene solutions with different THF/DMF ratios) and a mixing solvent (cyclohexane) were, respectively, pumped into the channel and well mixed before simultaneous electrospinning processes under selected operating conditions according to their previous studies.259,260 More importantly, various immiscible solvents/solutions can be electrospun by in situ mixing without solute precipitation. The oil adsorption results of macroporous fibers will be attractive to the investigation on and generation of heterogeneous fibers by in situ mixing electrospinning for potential applications.
![]() | ||
Fig. 33 (a) Schematic illustration of an in situ mixing electrospinning process for macroporous fiber fabrication. SEM images of the (b) surface and (c) cross section of macroporous fibers. Reproduced with permission from ref. 258. Copyright 2016 American Chemical Society. |
Chung et al.275 created porous scaffolds for tissue engineering via adjusting the flow rate of an alginate solution and the gas pressure to reach steady state monodisperse alginate droplets, which encapsulated nitrogen gas bubbles in a continuous mode generated in a microfluidic FFD. At a low liquid flow rate and high gas pressure, the microbubbles were less stable and more polydisperse. The microbubbles were closely packed in a container with a concentrated CaCl2 solution, where it gelated by vacuum degassing. The alginate scaffold exhibited a periodic interconnected porous structure (Fig. 34). Furthermore, the pore size could be adapted by varying the fluid velocity, gas pressure as well as the viscosity of the solutions. The pores were well-ordered near the surface, whereas deep inside the scaffold, the porous structure became less ordered and the size distribution of the pores became broader but still superior to scaffolds prepared by conventional methods. The feasibility of the microfluidic-generated scaffolds was further explored.28
![]() | ||
Fig. 34 (a) Co-flow microfluidic device for alginate microbubble generation, (b) monodisperse microbubbles, (c) honeycomb structure after gelation, (d and e) confocal microscope images of 3D ordered scaffold after vacuum degassing and (f) SEM image of a sponge-type alginate scaffold. Reproduced with permission from ref. 28. Copyright 2011 Elsevier Ltd. |
More recently, Wang et al.276 designed a novel microfluidic FFD device for the formation of highly organized honeycomb-like scaffolds. In their setup (Fig. 35a), the liquid phase and gas phase were introduced in the upper and lower inlets in the opposite direction. Gas-in-liquid microbubbles were generated at the liquid–gas interface. The scaffold showed a hexagonal close-packed porous texture, which facilitated cell attachment and physiology.275,277,278
![]() | ||
Fig. 35 (a) Counter-flow microfluidic FFC device, (b) confocal microscopy showing the layer assembly of microbubbles and (c) highly organized structure. Reproduced with permission from ref. 276. Copyright 2009 Wiley Periodicals, Inc. |
In the FFD setup of van der Net t al.,279 2 different solutions A and B were pumped separately at identical flow rates in a parallel mode to create the continuous phase in the microchannels (Fig. 36a).
![]() | ||
Fig. 36 (a) Parallel microfluidic FFD device and (b) unit cell of the fcc structure. SEM images of 3 ordered foam layers of the (c) (100) direction and (d) (111) direction of fcc packing in the swollen state. Scale bar is 1 mm. The inset images are taken with the foam in the dried state (scale bar is 500 μm). Reproduced with permission from ref. 279. Copyright 2009 Elsevier B.V. |
The gas phase is sheared by the continuous phase and deformed to form monodisperse microbubbles. The polymerization did not start until solutions A and B met. Although the obtained hexagonal structures in the foam (Fig. 36b) were compactly arranged, their packing directions vary. One is the (100) direction of the face-centered-cubic (fcc) packing (square arrangement of bubbles at the surface, Fig. 36c), whereas the other one is the (111) direction of fcc packing (triangle arrangement of bubbles at the surface, Fig. 36d). After polymerization, the hexagonally structured foam could be dried and re-swollen reversibly. However, incomplete polymerization was observed in some cases, as some bubbles maintain the initial spherical morphology.
Colosi et al.280 also obtained ordered and interconnected porous 3D scaffolds using a FFD microfluidic foaming technique, based on the work of Chung et al.275 But, here the alginate solution and nitrogen gas were replaced by an aqueous poly(vinyl alcohol) (PVA) solution and argon gas. Immediately after generation, the monodisperse microbubbles were cross-linked with glutaraldehyde (GLU) and freeze-dried by liquid nitrogen to obtain an irreversible scaffold structure. These were compared to the ones formed by a traditional gas foaming technique. The scaffolds generated by the microfluidic foaming technique presented a superior porous structure compared to the ones made by traditional gas foaming, based on the pore size and wall thickness distribution and high-resolution computed micro-tomography (μCT) and SEM images (Fig. 37). However, the microfluidic scaffolds showed rather scarce pore interconnectivity, moderated pore volume and a limited production rate.
![]() | ||
Fig. 37 μCT images of PVA scaffolds generated by (a) microfluidic foam and (b) traditional gas foam. SEM images of (c) microfluidic foam and (d) traditional gas foam. Reproduced with permission from ref. 280. Copyright 2012 American Chemical Society. |
3D scaffolds can also be prepared from HIPE monomers. However, polyHIPE scaffolds fabricated by conventional methods show rather poor reproducibility with uneven small pores, which may not satisfy the demand of tissue engineering. On the basis of Gokmen's study,19 Costantini and co-workers29 successfully prepared polyHIPE scaffolds from dextran-methacrylate (DEX-MA) using a microfluidic technique (Fig. 38a).281,282 In the first step, two immiscible phases added with DEX-MA or cyclohexane were introduced into a microfluidic FF chip to generate homogeneous O/W emulsions. In the second step, the O/W emulsions were cross-linked under UV-light (λ 365 nm) and then cured in an oven at 50 °C. In the third step, the product was placed in an organic polar solvent to extract the inner oil phase and create the macroporous structure. After washing with water and freeze-drying, the polyHIPEs presented a more interconnected and ordered architecture, compared to conventional HIPE templated scaffolds. By adjusting the flow rates of the 2 phases, distinct morphologies were obtained. The pore size of the scaffold increased upon using microchannels with larger dimensions (Fig. 38c–e). Three different lattice structures were present, including closed hexagonal packing together with [100] and [111] fcc packing (Fig. 38f–h). The concentration of the surfactant seemed to be vital to the final cellular structure. With higher surfactant concentrations, the closed porous structure evolved to a more open and interconnected one (Fig. 38i–k). However, the low production rate of the emulsions was still a limiting factor. To address this, parallelizing chips and junctions on a single chip were introduced.86,283,284
![]() | ||
Fig. 38 (a) Illustration of polyHIPE scaffold fabrication. SEM images of (b) conventional HIPE scaffolds and microfluidic HIPE scaffolds within (c) 1-fold, (d) 2-fold and (e) 5-fold volume of microfluidic FFD. Optical micrographs of microfluidic polyHIPEs with (f) hexagonal close packed, (g) fcc [100] and (h) fcc [111] crystal structures. SEM images of microfluidic HIPE scaffolds with (i) 1, (j) 7 and (k) 25% w/v Pluronic F68. Reproduced with permission from ref. 29. Copyright 2014 The Royal Society of Chemistry. |
Similar to 3D scaffolds, polymer foams are of particular interest for packing and insulation materials because of their low transport and thermal energy consumption. In 2016, Quell et al.285 used surfactant-stabilized HIPEs to fabricate two different kinds of highly ordered porous polymer foams, that is, closed-cell polyhedral foams (Fig. 39a) and open-cell spherical foams (Fig. 39c). For closed-cell polyhedral foams, monomers containing 1.28 mol% potassium persulfate (KPS) served as a dispersed phase. While for open-cell spherical foams, azobis(isobutyronitrile) (AIBN) replaced KPS with the same concentration and was added into the monomers. A mixture solution of styrene and divinylbenzene containing 10 wt% of the surfactant Hypermer 2296 was used as a continuous phase. Monodispersed emulsion templates were fabricated in the microfluidic chip, and then closely subsided in a continuous monomer matrix at a highly ordered level. Subsequently, emulsion templates were polymerized in an oil bath at 70 °C for 48 h. Eventually, the as-prepared polymer foams were purified by Soxhlet extraction with excess ethanol for 12 h before drying at room temperature. More importantly, they attributed different porous structures to the locus where the polymerization started. In the case of closed-cell polyhedral foams, initiation starts within the continuous monomer matrix and neighboring droplets are isolated by thin films. Otherwise, interconnected open-cell spherical foams were obtained.
![]() | ||
Fig. 39 (left) SEM images and (right) schemes of (a) highly ordered closed-cell polyhedral foam, (b) monodispersed emulsion templates in a continuous monomer matrix in a highly ordered level, and (c) open-cell spherical foam. Reproduced with permission from ref. 285. Copyright 2016 American Chemical Society. |
![]() | ||
Fig. 40 (a) Illustration of the generation of inverse-opal spheres. Reproduced with permission from ref. 288. Copyright 2014 American Chemical Society. SEM images of (b) whole inverse-opal spheres, (c) ordered porous surface and (d and e) ordered and interconnected internal cellular structure. Reproduced with permission from ref. 289. Copyright 2009 Wiley-VCH. |
Hexagonal close packed macropores were observed on the surface and in the interconnected internal structure. To obtain this morphology the volume fraction of polystyrene spheres to silica particles was important. More in particular, a volume ratio below 8 or above 10 gave a more disordered crystal structure and reduced the mechanical strength. A ratio of 9 led to greater surface area and more interaction sites, providing easier access for analyses to their recognition site. The inverse-opal photonic beads prepared by Zhao et al.291,292 were utilized for the label-free detection of biomolecules. This was demonstrated without applying time-consuming immune assay methods.
Zhao et al.293 also used silica colloidal crystal beads (SCCBs)290 as sacrificial templates and prepared macroporous hydrogel photonic spheres with a well-ordered inverse-opal structure. First, the SCCBs were fabricated in a home-made microfluidic device290 and then calcined to 800 °C for 3 h. The pre-gel solution (39% PEG, 1% initiator, 2% acrylic acid and 58% water) was filtered into the voids by capillary force. To ensure full void-filling, the SCCBs were pre-treated with piranha solution (30% H2O2 and 70% H2SO4) for 6 h, followed by washing and drying in nitrogen. The spheres got their opal architecture by polymerizing the pre-gel solution under UV-irradiation. After dissolving the template SCCBs in HF (1–2%), macroporous hydrogel photonic spheres were obtained. By increasing the PEG concentration, the mechanical strength was improved at the cost of structure order and a stable reflection peak position due to swelling. Fig. 41 shows the hexagonal close-packed SCCB surface and the well-ordered open porous and interconnected texture, before and after etching the templates. Conventional photonic band gap (PGB) crystals, which act selectively on photons like semiconductors do with electrons in electronic devices, can have opaline lattices which results in cooperative light scattering,294 often resulting in a very limited PBG and only a few narrow stop bands. Compared with conventional PGB crystals, macroporous hydrogel photonic spheres show angle-independent specific color with wider stop band gaps and are promising for advanced optical applications, such as reflection-based displays, barcodes and label-free sensors.
![]() | ||
Fig. 41 SEM images of a highly ordered surface of (a) template SCCB and (b) PEG inverse-opaline photonic sphere. Reproduced with permission from ref. 293. Copyright 2010 American Chemical Society. |
Wang et al.295 combined a microfluidic technique and a templating technique, which differed from the traditional top-down and bottom-up methods,288 to prepare multi-responsive poly(NIPAm-co-methacrylic acid (MAA)) photonic crystal microparticles (HPCMs) with an inverse-opal structure (Fig. 42). First, monodisperse emulsion droplets encapsulating silica nanoparticles with a diameter of around 220 nm were generated in a co-flow microfluidic device, followed by water evaporation. After washing with hexane to remove the oil phase and calcination, silica colloidal crystal microparticles (SCCMs) were obtained, which acted as templates for the formation of inverse-opal hydrogel microparticles. This was carried out in 3 steps: (1) immersion of the microparticles in a mixture of precursors and photo-initiators, (2) in situ UV-induced polymerization and (3) selective removal of templates by HF etching. Eventually, HPCMs with an inverse-opal structure were obtained. HPCMs bearing desired reactive functional groups (e.g. NIPAm as a temperature sensitive group and MAA as a pH sensitive group) showed rapid and sensitive responses to temperature and pH variations. Remarkably, the response rate was less than even 1 min. Other functional behaviors such as magnetic response could be introduced as well in HPCMs without affecting the temperature and pH responses too much.
![]() | ||
Fig. 42 Inverse-opal HPCM fabrication by combining a microfluidic technique I and a templating technique II. Reproduced with permission from ref. 295. Copyright 2013 American Chemical Society. |
More recently, Ye et al.296 developed novel microcapsules (Fig. 43a) consisting of densely packed opal photonic crystal (PhC) cores and responsive inverse-opal PhC hydrogel shells using similar SCCB templates.293
![]() | ||
Fig. 43 (a) Inverse-opaline hydrogel sphere formation. Reflection spectra and images (insets) of (b) the inverse-opal sphere, (c) SCCB and (d) PhC hydrogel beads. SEM images of highly ordered and interconnected porous structure of the (e) surface and (f) inner section. Reproduced with permission from ref. 296. Copyright 2014 Wiley-VCH. |
The microfluidically-generated SCCB templates had high monodispersity and brilliant structural colors due to the ordered arrangement of the spherical silica nanoparticles. After complete polymerization of the pre-gel solution, the hydrogel was disrupted by immersion in a buffer solution. Inverse-opal PhC hydrogel spheres were then fabricated by etching the SCCB templates inducing pore formation (±5 min). The etching conditions influenced the thickness of the inverse-opaline hydrogel shell, resulting in different relative refractive indices of the shell and core in the PhC microcapsule (Fig. 43b). Their reflection peaks differed only from those of SCCBs and PhC hydrogel spheres, exhibiting only one reflection peak (Fig. 43c and d, respectively). The PhC microcapsules displayed a highly ordered and interconnected macroporous structure (Fig. 43e and f), not only on the surface but also on the inside of the microcapsules, which is beneficial to diffusion imitated processes.
Inverse-opaline microsphere (IOM) crystals possess an isotropic band-gap property, differing from conventional 2D or 3D film-type colloidal ones,71,297,298 making them attractive for optical applications. Ionic liquids are salts consisting of organic anions and cations and are liquid below 100 °C. They have some advantages, such as negligible vapor pressure, incombustibility, reasonable chemical and thermal stability and high ionic conductivity.299 Poly-ionic liquid inverse-opaline microsphere (PIL-IOM) crystals showed unprecedented properties beyond those of their individual constituents. PIL-IOMs were fabricated by Cui et al.300 using a two-step procedure. Homogeneous silica nanoparticles were assembled into hexagonally packed crystalline lattices by droplet-based microfluidic synthesis,301,302 concomitantly with infiltration of imidazolium-based PIL monomers in the voids. The PIL-IOMs were isolated after UV-induced polymerization and removal of the silica nanoparticle templates with HF. The crystals exhibited brilliant color due to their uniformity in size and architecture (Fig. 44a). Further observation under SEM (Fig. 44b and c) illustrates the highly ordered interconnected macroporous structure, distributed over the whole particle surface. The relatively wide optical stop-band could be tuned from 400 to 700 nm by introducing different diameters of silica nanoparticles. PIL-IOMs could not only be used as stimuli-responsive photonic microgels, but also as multifunctional particles that mimic conventional molecules in terms of optical properties, molecular recognition, derivation and anisotropy.
![]() | ||
Fig. 44 (a) Optical image of the ordered macroporous PIL-IOMs. SEM images of the (b) surface and (c) cross-section. Reproduced with permission from ref. 300. Copyright 2014 Wiley-VCH. |
![]() | ||
Fig. 45 Schematic illustration of (a) monodispersed microbubble generation and (b) ordered porous film fabrication process. SEM images of the effect of the polymer concentration on the porous structure: (c) 0.1 wt%, (d) 0.3 wt%, and (e) 0.5 wt%. And SEM images of the effect of the surfactant (PEG-40S) concentration on the porous structure: (f) 0.25 wt%, (g) 0.5 wt%, and (h) 0.75 wt%. Reproduced with permission from ref. 304. Copyright 2016 American Chemical Society. |
![]() | ||
Fig. 46 (a) Fabrication and (b) SEM images of a linear gradient porous film. (c) Fabrication and (d) SEM images of a wedge-like gradient porous film. Scale bars are 20 μm (b) and 1 μm (d). Reproduced with permission from ref. 318. Copyright 2013 American Chemical Society. |
Methods to design meso and macroporous materials include phase separation phenomena,338 removal of templates such as bubbles, particles and droplets339–345 and conventional emulsification and foaming techniques, often leading to broad pore size distributions.340,346 Additional physical or chemical gelation reactions are required to fix the structure because the templating bubbles, particles and droplets tend to coalescence and show Ostwald ripening.210,342,347 In 2009, Lee and Weitz27 reported the fabrication of hierarchical porous inorganic microparticles via simple evaporation of the middle solvent layer. Moreover, droplet condensation on the surface of polymer solutions is regarded as an elegant physical method to generate pores of a few micrometers.348 Despite all these, this method was limited to obtain 2D structures with little flexibility to tune the pore size according to different requirements. Choi et al.175 employed the W–W/W–O emulsion as a modified microfluidic device (Fig. 47a) to prepare radially hierarchical porous microbeads. The dimension and morphology of the droplets were affected by the setup.211,212 In a dripping mode double emulsions were obtained. To form the hierarchical porous structure, they used fast evaporation of the organic solvent and emulsion templating. In contrast, slow evaporation of the solvent leads to microbeads with a dense structure. Interestingly, the final porous beads presented 2 different types of pore surface (Fig. 47b). The development of the small pores could be attributed to fast solvent evaporation, while the large pores were formed upon introducing a homogenized W/O emulsion.
![]() | ||
Fig. 47 (a) Scheme of a microfluidic apparatus for double W–W/O–W emulsions. SEM images of a (b) radially hierarchical porous surface and (c) cross-sections of PLGA microbeads. The scale bar of the inset in (b) is 5μm. Reproduced with permission from ref. 175. Copyright 2009 Wiley-VCH. |
Later, practical fabrication of bimodal macro/mesopore structures was introduced by Zhai et al.,349 based on a temperature-induced gelation (rather than a polymerization reaction) integrated with a microfluidic coaxial microdevice. Methylcellulose (MC, showing reversible-thermosensitive behavior) and PEG (featured with auxiliary functional groups) were used for the synthesis of radially hierarchical porous microspheres with diameter around 490 μm. First, monodispersed droplets were generated and treated in water baths at 90 °C (Fig. 48a). After complete gelation, the solid spheres were washed with acetone, dried at 80 °C and calcined at 550 °C. The spheres produced from sample 1 (containing 0.50 g MC) and sample 2 (0.25 g MC) showed distinct morphologies (Fig. 48b–g). Sample 1 spheres showed a densely porous surface and homogeneous macroporous interior structure (around 1 μm), whereas sample 2 spheres presented dense pores on the surface but smaller more homogeneous macropores (around 0.1 μm) in the interior. Mesopores also widely existed in the spheres with pore sizes of 13.8 nm in sample 1 and 20.5 nm in sample 2, as determined by nitrogen adsorption–desorption measurements.
![]() | ||
Fig. 48 (a) Schematic diagrams of the coaxial device. SEM images of (b and c) spheres, (d and f) porous surface and (e and g) internal structures of sample 1 and sample 2, respectively. Reproduced with permission from ref. 349. Copyright 2010 American Chemical Society. |
In addition, Kim et al.350 introduced a simple one-step double emulsion strategy to prepare multi-cored microcapsules with a porous surface in a dual-continuous coaxial microfluidic apparatus (Fig. 49a). Silica particles (900 nm), pretreated with dichlorodimethylsilane (DCDMS), were added to the ETPTA solution to obtain a silica-ETPTA suspension (middle phase). The emulsion droplets were generated in the capillary device and then polymerized by UV-irradiation for 1 s. By precisely adjusting the flow rates of the 3 phases, the number of cores could also be manipulated (Fig. 49b). The diameter of the silica nanoparticles embedded in the thin ETPTA membrane was slightly greater than the membrane thickness, resulting in partial exposure of the particles on both core and outer sides (Fig. 49c). Ultimately, the wet etching process with 5% HF solution dissolved the embedded silica nanoparticles creating a well-defined porous surface.
![]() | ||
Fig. 49 (a) Schematic illustration of multi-cored microcapsule fabrication. SEM images of a (b) microcapsule with 10 cores in gyro-elongated square dipyramid geometry and porous membrane, (c) surface and (d) cross-section. Reproduced with permission from ref. 350. Copyright 2011 Wiley-VCH. |
Uniform polymeric microparticles with controllable hierarchical interconnected porosity have also been reported using W/O/W emulsions351 (Fig. 50h). First, monodisperse W/O/W emulsions were generated in the glass-capillary microfluidic device. Herein, the partially water-soluble oil phase contained a surfactant (PGPR). Depending on the numbers of inner microdroplets (N, Fig. 50a), the inner structures were closely packed in distinct ways. Mass exchange between the inner and the outer aqueous phase occurred through the oil phase and was facilitated by the surfactant. The oil phase was partially miscible with the aqueous phases, resulting in the formation of fine aqueous nanodroplets (Fig. 50e and f). After UV-induced polymerization of the monodisperse W/O/W emulsions, hierarchical porous microparticles with well-defined micrometer- and nanometer-sized pores (Fig. 51a) left from the inner microdroplets and the mass transfer-induced aqueous nanodroplets, respectively (Fig. 50c, d and g). The combination of these two size-scaled highly interconnected pore structures was translated into advanced features, such as enhanced mass transfer, increased specific surface area and flexible morphology and functionality. The introduction of magnetic nanoparticles (Fig. 51b) showed synergy with the porous structure, useful for further application.
![]() | ||
Fig. 50 (a–g) Schematic illustration of the fabrication of microparticles with controllable interconnected hierarchical porosity using W/O/W emulsions and (h) corresponding SEM images (scale bar 200 μm). Reproduced with permission from ref. 351. Copyright 2015 American Chemical Society. |
![]() | ||
Fig. 51 SEM images of the cross-section of hierarchical microparticles obtained from a W/O/W emulsion with (a1) non-magnetic and (b1) magnetic nanoparticles, with the corresponding magnifications (a2 and a3) and (b2 and b3). Scale bars are 50 μm in (a1 and b1) and 20 μm in the rest. Reproduced with permission from ref. 351. Copyright 2015 American Chemical Society. |
Carroll et al.352 designed a surfactant micelle nano-emulsion mixture and combined it with tinplating silica microparticles to fabricate hierarchical porous spheres. Niño-emulsion droplets evolved into large pores with a diameter of tens of nm, while self-assembled surfactant micelle structures resulted in a subset of small pores with a diameter of several nanometers.353 First, the silica precursor droplets were formed in the microfluidic device with 2 different geometries, T-junction and flow focusing (Fig. 52a), and collected in a reservoir. The remaining water and the silicate reaction by product ethanol were eliminated via evaporation-induced solidification. Then, the solids were immersed in water or toluene to form the interconnected pores followed by washing and drying. Eventually, hierarchical porous spheres with honeycomb-like surface were obtained (Fig. 52b and c). Carroll also demonstrated that these particles can be used as replica electrocatalyst materials354,355 and high-surface-area scaffolds.356,357
![]() | ||
Fig. 52 (a) Illustration of microfluidic-derived hierarchical porous spheres obtained with a micellar nano-emulsion and silica microparticles. SEM images of (b) tha honeycomb-like porous surface and (c) corresponding magnification. Reproduced with permission from ref. 352. Copyright 2013 American Chemical Society. |
More recently, Fan et al.358 provided a simple and robust in situ microfluidical fabrication of hierarchical porous poly(ε-caprolactone)/silica (PCL/SiO2) hybrid microspheres. The superiority of this new strategy for radially hierarchical porous architecture lies in a synergistic effect between the sol–gel process (hydrolysis and condensation of a silane precursor) and solvent extraction in droplets. More importantly, this process not only avoided the complicated post-treatment by using porogens/sacrificed templates, but also ensures the uniform distribution of silica in the PCL matrix. In brief, the uniform droplets were firstly generated in the microchannel, and then collected in a certain concentration of PVA aqueous solution. After being solidified at room temperature for 1 day, the microspheres were centrifuged and washed with excess deionized water before the final freeze-dry process. It is found that ammonia, used as a catalyst for silane hydrolysis, played a decisive role in the formation of hierarchical porous structures. At the same time, via adjusting ammonia concentration, the porous structures and the surface morphology can be easily controlled (Fig. 53).
![]() | ||
Fig. 53 SEM images of PCL/SiO2 hybrid microspheres fabricated by adjusting the ammonia concentration in the continuous phase and/or collection solution. Reproduced with permission from ref. 358. Copyright 2016 American Chemical Society. |
Besides, Sommer et al.359 used monodisperse PCL particles obtained by microfluidics360 as sacrificed templates, combining with the 3D printing technique, to fabricate hierarchical silk porous materials with unprecedented structural control shown in Fig. 54. In detail, the microfluidically-synthesized PCL templates with different magnitudes in size were firstly added into silk fibroin-based ink to obtain the original materials by the 3D printing technique. Secondly, the removal of templates led to hierarchical porous silk-based materials. Finally, the PCL particles can be electrostatically coated with latex nanoparticles to build a porous architecture at three hierarchical levels. Due to moderate biodegradability of silk, the resulting structures turned from close to open porosity. Subsequently, these pore templating particles can be further modified with functional nanoparticles for various application fields from structural materials to thermal insulation. Meanwhile, Sommer and coworkers360 also introduced silk fibroin scaffolds with an inverse opal structure for tissue engineering application.
![]() | ||
Fig. 54 (a) Schematic illustration of the fabrication of 3D printed silk fibroin with a hierarchical porous structure. (b and c) SEM images of the resulting hierarchical porous silk fibroin structures. Reproduced with permission from ref. 359. Copyright 2016 American Chemical Society. |
![]() | ||
Fig. 55 (a) Illustration of the generation of uniform PCL microbeads. Reproduced with permission from ref. 210. Copyright 2009 Wiley-VCH. SEM images of (b) the freeze-dried ccp lattice, (b) top view of the inverse opal scaffold, (c) magnified view of the top surface and (d) side wall in a macropore with several smaller pores. Reproduced with permission from ref. 347. Copyright 2009 Wiley-VCH. |
Studart et al.367 described a versatile and simple approach to fabricate customized porous materials with up to 3 hierarchical levels. Surprisingly, the principle relied solely on the process of drying complex suspensions containing droplets, colloidal particles and surface active molecules dispersed in the absence of any type of gelation. Monodisperse polystyrene particles (400 nm) were prepared by surfactant-free emulsion polymerization. Homogeneous oil droplets were obtained in a dripping mode212 in 2 kinds of glass microcapillary devices: co-flow and flow-focusing geometry. The flow rates of dispersed and continuous phases were constant in all experiments. The resulting oil droplets were stabilized by adsorbing either in situ hydrophobized colloidal droplets (silica or polystyrene) or surface active molecules at the oil–water interface (Fig. 56).
![]() | ||
Fig. 56 (a and c) Schematic diagrams and (b and d) top views of surfactant-stabilized and particle-stabilized droplets, respectively. Hierarchical porous structures from (e) PVA-, (f) silica- and (g) CTAB-stabilized droplets. (h) Magnification of (e) illustrating the interstices between the macropores. (i and j) Magnified FESEM images of the region close to a macropore indicated in (g). Scale bars in the insets of (e) and (h) are 100 nm and 10 μm, respectively. Reproduced with permission from ref. 367. Copyright 2010 American Chemical Society. |
The stabilized droplets were polymerized and deposited on a glass substrate, followed by self-assembly (buoyancy effect) to highly ordered hexagonal arrays in the case of surfactant-stabilized droplets, whereas particle-stabilized droplets were able to form both compact ordered and random disordered arrays. For the particle-stabilized droplets, a textured coating on the surface of particles was observed (Fig. 56d). In contrast, only a thin air–water layer was present on the surface of the surfactant-stabilized droplets (Fig. 56b). Remarkably, the surfactant-stabilized droplets lead to an open ordered cellular structure (Fig. 56e), whereas the particle-stabilized droplets resulted in fully closed ordered arrays (Fig. 56f). Additionally, the nanoparticles compactly packed in the voids between the large pores (inset of Fig. 56e) were removed in the subsequent steps, resulting in a secondary pore hierarchy. The dried structures (Fig. 56i) presented ordered secondary pores in the vicinity of the large primary pores while more disordered pores were observed more distantly. The pore wall thickness could be adjusted by tuning the TEOS concentration in the aqueous feed (Fig. 56j), while the initial droplet diameter determined the ultimate aperture. Moreover, the type of droplet stabilizer (surfactants or particles) had a great influence on the interconnectivity between macropores (Table 3).
Morphology | Structure with closed ordered arrays | Structure with open ordered arrays | |
---|---|---|---|
Droplet type | Particle-stabilized | Surfactant-stabilized | |
PVA-stabilized | CTAB-stabilized | ||
Dispersed phase composition | Photosensitive oil mixture containing 1,6-hexanediol diacrylate + trimethylolpropane ethoxylate triacrylate (monomers), 2-hydroxy-2-methyl-L-phenyl-L-propanone (initiator) and toluene (diluent) | Octane | Octane |
Continuous phase composition | Aqueous suspension containing CTAB and silica particles | Aqueous suspension containing PVA and silica particles | Aqueous suspension containing CTAB, polystyrene particles and TEOS |
Average large pore size (μm) | ±140 | ±125 | ±140 |
Average small pore size (nm) | ±110 | ±110 | ±400 |
![]() | ||
Fig. 57 (a) W/O/W double emulsion template evolving to non-spherical colloidosome with multiple compartments. (b) Optical microscopy images of aqueous droplets encapsulated by W/O/W double emulsions. SEM images of colloidosomes with (c) 2, (d) 3, (e) 4, (f) 5 and (g) 6 compartments (N). Reproduced with permission from ref. 27. Copyright 2009 Wiley-VCH. |
Phase composition | Internal W/O drop number (n) | Compartment number (N) | Particle morphology | ||
---|---|---|---|---|---|
Outer | Middle | Inner | |||
0.2–2 wt% PVA aqueous solution | Toluene suspension containing 7.5 wt% hydrophobic silica nanoparticles | 0–2 wt% PVA aqueous solution | 2 | 2 | Peanut-shaped |
3 | 3 | Three links | |||
4 | 4 | Triangular dipyramid | |||
5 | 5 | Square pyramid | |||
6 | 6 | Pentagonal dipyramid |
Later, Fang et al.376 described a simple one-step method for the dropwise formation of doughnut-shaped silica microparticles using a novel microfluidic FFD377 device with a serpentine channel (Fig. 58a), which improved the control over shape and size. Along the serpentine microchannel the deformation became more pronounced and the droplets gradually shrunk during solidification from 75 μm spheres to 25μm doughnut-shapes (Fig. 58b and c), as a result of the buckling instability at the interface. A thin membrane (2–3 μm) connected the 2 (asymmetric) cavities (Fig. 58d).378 According to the drying hydrodynamics, either pancake-shaped or torus-shaped silica microparticles could also be created by adjusting the experimental conditions, such as the silica content in the dispersed phase (TEOS + triethyl amine), the flow rates of dispersed and continuous phases,297etc.
![]() | ||
Fig. 58 (a) PDMS device with serpentine channel, (b) optical microscope tilted side view of doughnut-shaped silica particles, (c) SEM image of whole particles and (d) top (*) or side (![]() |
In order to fabricate the non-spherical particles, novel microfluidic methods relying on photolithographic techniques were introduced recently.54,379 Arrays of mask-defined polymeric particles were formed under UV-light in the PDMS device in an automatized way. Fig. 59a shows a cross-section of the microfluidic device. Mask defined shapes were cross-linked in the centre of the device under a dynamic oxygen atmosphere, which was constantly diffusing through the PDMS wall and forming the inhibition layer above and beneath the particles. The thin oxide layer prevented clogging of particles allowing them to elute smoothly.380,381 One technique used a continuous flow of monomers passing through the microfluidic device producing arrays of arbitrary masked particles (continuous flow lithography, CFL). Another technique employed masks which were embedded inside the microfluidic device. Eventually, 4 types of flow lithography were developed (Table 5): CFL, stop flow lithography (SFL), stop flow interference lithography (SFIL) and lock and release lithography (LRL). An additional advantage over droplet-based methods is the possibility to fabricate particles with more than 3 different sections. Extending photolithography from photoresistant materials to other more bio-friendly materials could open the range of applications.
![]() | ||
Fig. 59 (a) Cross-section of a PDMS microfluidic flow lithography device. Reproduced with permission from ref. 380. Copyright 2008 American Chemical Society. (b) SFL device using a 3-way solenoid valve for rapid flow start and stop before and after polymerization. Reproduced with permission from ref. 379. Copyright 2007 The Royal Society of Chemistry. (c) SFIL. Reproduced with permission from ref. 382. Copyright 2007 Wiley-VCH. (d) LRL. Reproduced with permission from ref. 383. Copyright 2009 The Royal Society of Chemistry. |
Technique (year) | Characteristics |
---|---|
CFL (2006) |
Automatized fabrication of non-spherical particles with 2D-extruded shapes down to 3 μm
Low Pe flows and pattern blurring Structures with poor resolution |
SFL (2007) |
Automatized fabrication of heterogeneous biocompatible 2D-extruded particles
Improved resolution for more detailed structures (1–2 μm) Sharp interfaces between different sections of particles |
SFIL (2007) |
Predictable internal 3D structures in one shot
High surface-area-to-volume ratio Controllable 3D porosity |
LRL (2009) |
3D multifunctional particles with complex structures
Automatized fabrication of composite particles with multiple overlapping sections |
Most of the traditional approaches to make multifunctional encoded particles used for biomolecule analysis are complicated and time-consuming. Besides, they often generate only a limited number of particles. The processes for encoding, functionalizing or decoding active substrates (particles or surfaces) are expensive. Pregibon et al.384 presented a flexible and convenient CFL based approach, combining particle fabrication, encoding and probe incorporation into a single process. The obtained multifunctional particles possessed more than a million unique codes. The mix of 2 monomer flows (one loaded with an acrylate-modified probe and the other one with a fluorescent dye) constituted the laminar flow in a microfluidic channel and were then polymerized by UV-light (Fig. 60A). Particles with different regions were created in a single step: a graphically encoded region (fluorescent probe) and a probe-loaded region (modified probe). PEG, known as a bio-inert polymer, was used as the particle anchor point to “block” surfaces after the direct incorporation of the probes into the particles. PEG as a transparent material also allowed the fluorescent signal to pass through both sides of the particles. The resulting particles were extruded in 2D shapes (Fig. 60B–D) with macroporous “block” holes. Their chemical components were determined by the co-flowing monomer stream composition. Particles were designed to be “read” along 5 lanes along their length, with alignment indicators as the identification of the code position.
![]() | ||
Fig. 60 (A) Fabrication of dot-coded particles, (B) single-probe half-fluorescent particles, (C) particle subdomains for encoding and analyte detection and (D) differential interference contrast (DIC) image of particles (scale bar 100 μm). Reproduced with permission from ref. 384. Copyright 2007 Science. |
Later, Lee et al.165 studied the use of PEG as an effective porogen to create a macroporous hydrogel and a hydrogel slab within microfluidic channels. The inert PEG occupied the space within the polyacrylamide hydrogel without affecting the protein activity.385 In contrast to other common porogens, PEG is more hydrophilic and can be dissolved in water and therefore it was removed efficiently from the hydrogel.386 Pre-polymer solutions with 4% (w/v) polyacrylamide were exposed to 320–500 nm UV-light through a transparency mask, placed between the light source and microchannel. Upon addition of PEG porogens at concentrations higher than 5%, the polymerized hydrogel (Fig. 61a and b) became opaque and macroporous as a result of polymerization-induced phase separation. Higher loads of PEG porogens increased the porosity but decreased the structural strength. To increase the fabrication throughput, the flow rates of the different phases needed to be enhanced, but that leads almost inevitably to particle smearing and poor repeatability (unacceptable deformation) due to the limited UV-exposure time.379
![]() | ||
Fig. 61 (a) Optical image of 5% PEG700DA (Mw 700 Da) macroporous hydrogel with 20% PEG10000 porogen and (b) SEM image of 5% PEG700DA macroporous hydrogel slab with 20% PEG35000 porogen. Reproduced with permission from ref. 165. Copyright 2010 American Chemical Society. |
![]() | ||
Fig. 62 (a) Schematic illustration of SFL porous particle fabrication. SEM micrographs of (b) a certain number of polymerized porous particles, and (b) surface and (c) cross section of polymerized porous particles, respectively. Scale bars are 2 μm (b and c). Reproduced with permission from ref. 387. Copyright 2016 American Chemical Society. |
![]() | ||
Fig. 63 Concentric square masks, having differences between inner and outer sides of (a) 50, (b) 20 and (c) 10 μm, used for the SFL generation of macroporous square particles. (d, f and h) SEM images of SFL particles and (e, g and i) particles made with the same masks using CFL under the same conditions as SFL. The scale bars are 10 μm (d–i). Reproduced with permission from ref. 379. Copyright 2007 The Royal Society of Chemistry. |
![]() | ||
Fig. 64 SEM images of (a) glassy silica microgear sintered at 1150 °C for 10 h, including (b) high magnification and (c) porous silica microgear sintered at 1150 °C for 1 h, including (d) high magnification. Reproduced with permission from ref. 388. Copyright 2008 Wiley-VCH. |
More complex particles with multidimensional anisotropy could also be fabricated with advanced microfluidic designs.383,392–394 However, they sometimes lacked technological solutions to rapidly decode particles and quantify target binding in a high-throughput manner, suitable for clinical research applications.395–397 To expand the application of bar-coded hydrogel particles, Doyle and his group developed a versatile microfluidic flow scanner, which decoded the particles in an accurate way and up to 25 targets per s, with multiplex detection encoded. The device consisted of 4 major parts: (1) creation of bar-coded microparticles, (2) protein target labelling, (3) scanning and analyzing the particles and (4) automated target identification. The bar-coded hydrogel particles were produced using a well-designed SFL device (Fig. 65c) described by Dendukuri et al.,379 a flow control equipment used by Bong et al.398 and a scan system built on the microscope that was used for particle generation.399,400 Stop-flow cycles were adjusted by the pressure and synchronized with UV-light pulses to simultaneously synthesize, encode and functionalize. By high-resolution design of the particle layout, hosting different adjacent chemical functionalities, reliable code readouts could be achieved (Fig. 65a). Earlier, scans occurred on equally spaced probes,401 whereas Boyle used a 30 μm width blank region inserted in the particle to separate the code and the probe. Another blank region was incorporated at the end of the particle to ensure symmetrical exposure of the probe to the carrier (Fig. 65b). Other studies also explored bar-coded microparticles for protein detection and other applications.399,402–406
![]() | ||
Fig. 65 (a) Illustration of the fabrication of single-probed particles. (b) A bar-coded particle with code 20303, (c) microscope setup for synthesis and scanning: laser and PMT for scanning, UV for polymerization and CCD for synthesis and alignment and (d) blue colored feed to identify and adjust the monomer flow widths easily during the fabrication process: B: blank, P: probe and C: code. Reproduced with permission from ref. 381. Copyright 2011 Nature America. |
![]() | ||
Fig. 66 (a) Illustration of hydrogel sheet production, (b) Pacman-arena styled sheet, (c) gradient pore pattern and (d) cylindrical and conical pore formation. SEM images of the (e) cylindrical and (f) conical pores in the sheets and (g) multi-layered magnetic hydrogel sheets with square macropores and (h) corresponding magnification. Scale bars are 1 mm (b, c and g), 40 μm (e and f) and 100 μm (h). Reproduced with permission from ref. 409. Copyright 2014 American Chemical Society. |
![]() | ||
Fig. 67 SEM images of PEGDA structures with different morphologies and volume fractions, fabricated with the same phase mask and (a) 700 ms exposure time with 1.2 wt% hydroquinone (inhibitor), (b) 1000 ms exposure time and 0.3 wt% hydroquinone and (c) special flower shapes fabricated from ethanol suspension and 300 ms exposure time. Reproduced with permission from ref. 382. Copyright 2007 Wiley-VCH. |
Fig. 67a and b show the cross-sectional views of the 3D structures obtained at a short exposure time/high inhibitor load and long exposure time/low inhibitor load, respectively. The latter 3D structures contained smaller pores and had a higher volume fraction. Upon shrinking in ethanol, flower-shaped structures were formed (Fig. 67c).
The release time was also important. Particles could be released at a critical pressure as a function of the device deformation. While the low pressure was maintained, the monomer exchange continued and addition of new chemical groups was possible before the release was started. This way, composite particles with multiple chemical regions and complex porous structures were obtained (Fig. 68a). First, the multi-inlet was filled with one type of monomer and precisely locked. Subsequently, by tuning the pressures of the inlet streams (but keeping them below 10 psi), locked structures with chemistry 1 were covalently linked to chemistry 2 through mask overlapping, then polymerized under UV-light and released by high pressure (Fig. 68b–f). The mask was compatible with different chemistries to selectively polymerize through overlapping, resulting in different particle types. Fig. 68g shows a particle obtained by applying 3 chemistries. The lag times were critical, as they determined the fluidic exchange and mask alignment. However, the production throughput decreases when multiple chemistries are employed.
![]() | ||
Fig. 68 Composite particle (a) synthesis, (b) structure, (c) DIC image and (d) fluorescence microscopy image. Fluorescence microscopy images of (e and f) composite particles with an “autumn tree” pattern, with 6 thicker locks appearing as brighter regions and (g) composite particle with a “spring tree” pattern. Scale bars are 100 μm (c, d and f) and 50 μm (e and g). Reproduced with permission from ref. 383. Copyright 2009 The Royal Society of Chemistry. |
![]() | ||
Fig. 69 Common porous materials used as catalytic supports with increasing pore sizes and nanoreactors with gradually increased reaction spaces. |
Materials with hierarchical porous structures on the nano- and micrometer scales have also been explored to improve conventional micro-/meso-/macroporous materials as catalytic supports and nanoreactors.440–442 For this reason, the development of microfluidic techniques can play an important role in catalytic applications. Wacker et al.166 created magnetic macroporous spheres in a microfluidic device using silica particles and iron oxide nanoparticles and then functionalized them with horseradish peroxidase (HRP), frequently used as a signal amplifier in immune assays.443 This way the HRP functionalized particles could be recovered easily.444,445 Carroll et al.352 designed surfactant micelle/oil nanoemulsions and combined them with silica microtemplates to fabricate hierarchical porous spheres to improve the mass selective transfer in catalysis, but also for electrocatalysis materials.354,355 More recently, Kim et al.72 also demonstrated that the porous microparticles can be used as catalyst supports thanks to their intrinsic structural features. However, although microfluidically-generated porous materials possess so many distinct advantages, for future catalytic application, more efforts should be devoted to control their structures and morphologies. With the rapid progress of microfluidics in porous material fabrication, macroporous materials with higher specific surface area, larger pore size and pore volume, as well as better multi-functionalization superior to the current porous materials obtained by a conventional approach will certainly face a bright future in modern catalytic research and applications.
To date, only a few conventional approaches exist for the controlled creation of crystals from colloids. Crystal structures formed from colloids in a state of suspense often show fragile behavior. The most predominant approaches for this task are:156 (1) flow- or evaporation-induced colloidal packing492–495 and (2) non-close packing induced by electrostatic energy, followed by immobilization.496,497 CPCs with various morphologies, such as thin film or bulk, were fabricated. The Bragg diffraction that occurs upon irradiation of CPCs generates different structural colors when observed from different angles, useful in the construction of some optical materials and devices that require wider viewing angles.288 Other drawbacks such as a relatively slow response rate for certain stimuli and the lack of high throughput assays limit their application.498
Photonic crystal beads (PCBs) were created to address these drawbacks.297,498,499 Thanks to their structural symmetry, these spherical CPCs were rotation independent when their surface was irradiated at a fixed incident angle of the light beam.289,500,501 Crystallization to PCBs was induced by evaporation and polymerization of suspensions with droplet template colloidal nanoparticles.288 To increase the refractive index, the nanoparticles need to be removed selectively to generate inverse-opal structures with broader PBGs and reduced attenuation length. Characteristic to inverse-opaline PCBs, long-range ordering of pores was crucial to the optical performance, resulting in stable specific reflection peak positions.286,291 Under normal light incidence, the reflection (stop-band) wavelength λ of PCBs can be estimated using Bragg's equation:289,296,502
λ = 1.633 × d × naverage |
![]() | ||
Fig. 70 (a) Reflectance spectra and optical images of red, blue, and green PCBs composed of 190, 152, and 145 nm silica particles. (b) Normalized reflectance spectra and optical images of photonic balls before and after removing the 165 nm silica particles. Reproduced with permission from ref. 156. Copyright 2008 Wiley-VCH. (c) Reflection spectra and (d) 3D image (in water) of 7 inverse-opaline PCBs prepared with polystyrene spheres as sacrificial templates having, from left to right, 200, 215, 238, 262, 280, 298, and 315 nm sizes. Reproduced with permission from ref. 289. Copyright 2009 Wiley-VCH. (e) Reflection spectra of 5 kinds of inverse-opaline PCBs made with silica nanoparticles of, from top to down, 290, 260, 240, 220 and 200 nm sizes. Scale bars are 100 μm (a), 500 μm (d) and 200 μm (e). Reproduced with permission from ref. 296. Copyright 2014 Wiley-VCH. |
![]() | ||
Fig. 71 (a) Optical response and optical images (insets) of PIL-IOMs upon exposure to aqueous solution with 6 different counter ions. Reproduced with permission from ref. 300. Copyright 2014 Wiley-VCH. (b) Illustration of the microparticle structure and the payload release upon irradiation with a NIR laser and (c) normalized volume change of microparticles with (black squares) or without (red dots) gold nanorods according to a cycle laser. Reproduced with permission from ref. 170. Copyright 2014 American Chemical Society. |
![]() | ||
Fig. 72 (a) Reflection spectra and bright field microscopic images of 3 types of MIPBs. The cyan, green and red MIPBs were imprinted bovine haptoglobin (Hb), HRP and serum albumin (BSA), respectively. The dashed and solid lines are spectra of the MIPBs before and after multiplex detection. Reproduced with permission from ref. 291. Copyright 2009 Wiley-VCH. (b) Flow-through microfluidic device used to read reflection spectra of inverse-opaline PCBs. Reproduced with permission from ref. 289. Copyright 2009 Wiley-VCH. (c) Illustration of PCBs for label-free multiplex detection. Reproduced with permission from ref. 296. Copyright 2014 Wiley-VCH. |
Zhao and co-workers289 also developed a simple microfluidic device for in situ detection (Fig. 72b). When the PCBs passed the detection region, their reflection peaks could be detected using a microscope. More recently, by employing well-controlled wet etching processes, Ye et al.296 fabricated novel PCBs with close-packed opal cores as their encoding units offering stable diffraction peaks and responsive inverse-opal hydrogel shells. The latter acted as the sensing unit for target recognition, based on the shrinking behavior of the shells. These microparticles showed 2 reflection peaks, whose shifts could be used for the estimation of the target concentration (Fig. 72c).
Currently, the fabrication process of hierarchical porous particles with inverse opal structures is still complicated and apt to have deviation, which further affects their optical applications. One-step microfluidically hierarchical porous particle fabrication has a long way to go.
Recently, porous silicon (PSi) materials were developed as versatile carriers for drug delivery.523,525,534–538 The degradation rate of PSi materials to non-toxic silicic acid in vivo could be controlled by adjusting their porosity and surface chemistry.531,539–541 Furthermore, a wide variety of therapeutic and imaging agents have been successfully incorporated into PSi particles. However, the existence of free accessible pores imposed the risk of release of drugs from the particles during their transport.356,541,542 The pores could be sealed by using a matrix to ensure continuous release of the payloads for a prolonged time. In the future, customized ordered and/or hierarchical macroporous particles with regular or irregular shape can be easily obtained by advanced microfluidic technology according to the special requirements. It will accelerate the developments of microfluidically-synthesized macroporous particles in the wider biomedical field.
Removal of organic pollutants from surface or sub-surface aquifer water to reduce the ecological damage in the environment is a difficult task. The innovation of materials, especially porous particles, plays an important role in this area.563,573,574 To date, silica aerogels and activated carbons have been among the most common materials used for this task. However, with time these particles become unstable in such aqueous environments due to their inherent hydrophobic surface nature and eventually they need to be replaced.575–577 To achieve good water dispersion of hydrophobic particles, their surface morphology characteristics should be changed. Particles can absorb organic contaminants and remain well dispersed in aqueous environments by constructing a porous hydrophobic absorbing core and a hydrophilic surface. By combining mixing-induced phase separation and precipitation polymerization in a microfluidic device, particles with a hydrophobic porous core and a hydrophilic porous surface were designed to ensure excellent dispersion of particles in the aqueous phase with good adsorption capacity578 (Fig. 73a). Multilayered porous particles were obtained after removal of the porogen, assisted with a color change from dark to light (Fig. 73b), ultimately becoming transparent. The particles were effective for the uptake of organic oil directly from oil drops as for the uptake of organic molecules dissolved in an aqueous solution. The recyclability of these particles was also demonstrated.
![]() | ||
Fig. 73 (a) Schematic illustration of the porous particle fabrication. (b) The optical images describing the process absorption of oil by particles and consequent increase of inner layer brightness during 48 h time period. Reproduced with permission from ref. 578. Copyright 2013 Wiley-VCH. |
In addition to porous materials with only one pore dimension, recently Zhang et al.351 introduced highly interconnected hierarchical porous microparticles for the adsorption of oil droplets in water. Magnetic Fe3O4 nanoparticles (diameter 12 nm) were prepared and then dispersed by ultrasonic treatment in the middle phase to generate the magnetic hierarchical porous poly(MMA-co-EGDMA) microparticles. Similar to their previous work,533 first EGDMA dyed with red color was added into water (Fig. 74a) and then dispersed into microdroplets by shaking (Fig. 74b), serving as an oil contaminated sample. For the removal of the oil, the poly(MMA-co-EGDMA) porous microparticles were added and adequately mixed (Fig. 74c). Remarkably, the dyed oil droplets could be separated from the microparticles by directed guiding with a magnet (Fig. 74d). The microparticles were recovered by washing the absorbed oil with ethanol and recycled up to 20 times without structural deformation (Fig. 74e). The results were compared with those obtained with poly(MMA-co-EGDMA-co-GMA) to study the effect of nanopores and different hierarchical porous structures (Table 6).
![]() | ||
Fig. 74 Magnetic hierarchical porous poly(MMA-co-EGDMA) microparticles for magnetic-guided removal of dyed EGDMA microdroplets from water. Reproduced with permission from ref. 351. Copyright 2015 American Chemical Society. |
Structure type | Average pore size (μm) | Porosity (%) | Oil adsorption capacity (μL) |
---|---|---|---|
1 pore type (μm) | ±120 | 44 | <10 |
1 pore type (nm) | ±0.58 | 48 | ±20 |
2 pore types (μm) | — | — | ±30 |
3 pore types (μm) | — | — | ±40 |
The ever-increasing requirement for wastewater treatment and reuse gradually pushes conventional porous adsorbents to their limits. The solutions to the existing and future water challenges will depend on advanced adsorbents with unprecedented performances, such as higher specific surface area, distinguished hierarchical porous morphology, outstanding selective adsorption capacity, as well as less toxicity and acceptable mechanical properties. However, the current shortcomings of the microfluidic technique, for example, low yield and high cost, become another obstacle in the application of microfluidically-synthesized porous materials in water purification. Design and use of microfluidically-synthesized porous materials to solve these challenges become some of the most urgent problems in the future.
![]() | ||
Fig. 75 Encoded particles used for high throughput multiplex nucleotide diagnostics. Reproduced with permission from ref. 586. Copyright 2006 Macmillan Publishers. |
The advantage over other systems is that the probes were effectively printed directly into the encoded particles and only one fluorescent molecule was needed for both target identification and quantification. The technique succeeded in detecting DNA oligomer targets down to the picomolar level. Recently, other researchers designed special shaped particles obtained by CFL for the detection of some proteins. They were also used for the specific spatial sensing of ligands through development of polymeric substrates that contain a specific templating ligand: molecularly imprinted polymers emerging as a novel research area. With the help of microfluidic techniques the template was selectively removed from residual cavities and then the monodisperse MIP microbeads were used to detect the specific ligand molecule target,587 even in mixtures of various molecule analogs.588 With the rapid development and lasting improvement of encoded particles, we will witness the widespread use of microfluidically-synthesized encoded particles with faster response and improved accuracy in the future.
The fabrication of porous materials using microfluidic techniques addresses the drawbacks of traditional synthesis approaches. It provides accurate control of the particles' morphology in a highly reproductive manner. Microfluidic techniques also allow the space for further development with more complex configurations: from radially porous particles to ordered ones, from opened porous structures to closed ones, from regular porous spheres to irregular ones, from interconnected structures to multi-compartmentalized structures, from monomodal porous structures to hierarchical ones, etc. They have opened the way to the production of advanced materials such as inverse-opals, scaffolds and hierarchical porous materials, which would be difficult to generate otherwise. Flow lithography, as the result of simultaneous development of microfluidics and photolithography techniques, unlocks new possibilities for the synthesis of customized porous materials with a relatively wide range of morphologies and thus application fields. Microfluidic techniques allow particles to be fabricated with unprecedented control over their pore size with narrow distributions. They possess great potential for encapsulating a wide range of active materials, even living cells in biomedical applications. Nevertheless some drawbacks still exist, such as low throughput and tedious device preparation, limiting their wider range of applications. But this can be stated in many novel techniques and the research field is still developing. In this regard, multiple parallel microfluidic devices are currently considered as a practical but less efficient solution.593,594 Blocking phenomena occurring during fabrication still remain a challenge for microfluidic techniques. The existing methods for the minimization of blocking phenomena, such as surface modification, flow modulation and external force control, are rather sophisticated. A facile approach for minimizing this effect is still required.70 To date, one-step fabrication of macroporous particles has only rendered radially porous structures, whereas a combination of polymerization with sacrificial templates and wet etching processes results in various well ordered porous structures. Hierarchical porous structures are obtained by introducing sacrificial templating nanoparticles on different scales. However, modern microfluidics still looks for improvements. The main quest still remains how to simultaneously achieve accurate and independent control of the ordered hierarchical porous structures in one step.
In conclusion, future research will focus on the improvement of the device simplicity on one hand and on the particle morphology functional complexity on the other hand. This change undoubtedly conforms to the development of science from simple to complicated as well as from the shallower to the deeper. While initial inroads have been made, more work in the area of scale-up must still be done.
2D | Two-dimensional |
3D | Three-dimensional |
AAm | Acrylamide |
AIBN | Azobis(isobutyronitrile) |
AR | Aspect ratio |
BDK | 2,2-Dimethoxy-2-phenylacetophenone |
BMA | Butyl methacrylate |
BuAc | n-Butyl acetate |
ccp | Cubic close packed |
CFL | Continuous flow lithography |
CPC | Colloidal photonic crystal |
CTAB | Cetyltrimethylammonium |
DAP | Diallyl phthalate |
Darocur 1173 | 2-Hydroxy-2-methyl-1-phenyl-1-propanone |
DBP | Diisobutyl phthalate |
DCDMS | Dichlorodimethylsilane |
DCM | Dichloromethane |
DDP | Diisodecyl phthalate |
DDS | Drug-delivery system |
DEAP | 2,2-Diethoxyacetophenone |
DEP | Diethyl phthalate |
DMF | N,N-Dimethylformamide |
DMPA | 2,2-Dimethoxy-2-phenylacetophenone |
DOP | Dioctyl phthalate |
DPPD | 2,2-Di(pro-2-ynyl)-propane-1,3-diol |
DEX-MA | Dextran-methacrylate |
EA | Ethyl acetate |
EGDMA | Ethylene glycol dimethacrylate |
EPTBC | Ethyleneoxide propyleneoxide tri-block copolymer |
ETPTA | Ethoxylated trimethylolpropane triacrylate |
FFD | Flow-focusing device |
FFT | T-junction device |
FITC-dextran | Fluorescein isothiocyanate-dextran |
GLU | Glutaraldehyde |
GMA | Glycidyl methacrylate |
GMBS | N-Gamma-maleimido-butyryloxy-succinimide |
HEM | Hydroxyethyl methacrylate |
HF | Hydrofluoric acid |
HIPE | High-internal-phase emulsion |
HMPP | 2-Hydroxy-2-methyl-1-phenyl-1-propanone |
HPCM | Hydrogel photonic crystal microparticles |
HRP | Horseradish peroxidase |
KPS | Potassium persulfate |
LRL | Lock and release lithography |
MAA | Methacrylic acid |
MBAAm | N,N′-Methylene-bis-acrylamide |
MEK | Methyl ethyl ketone |
MIPB | Molecularly imprinted polymer bead |
MMA | Methyl methacrylate |
NIR | Near-infrared |
NOA 61 | Mixture of trimethylolpropane diallyl ether, isophorone diisocyanate ester, trimethylolpropane tristhiol and a benzophenone photo-initiator |
OY | 1,7-Octadiyne |
NIPAm | N-Isopropylacrylamide |
PBG | Photonic band gap |
PCB | Photonic crystal bead |
PCL | Poly(ε-caprolactone) |
PCM | Phase change material |
PDLA | Poly(D,L-lactide) |
PDMS | Poly(dimethyl siloxane) |
PEG | Poly(ethylene glycol) |
PEG-40S | Poly(ethylene glycol)-40-stearate |
PEGDA | Poly(ethylene glycol)diacrylate |
PEO | Poly(ethylene oxide) |
PGPR | Polyglycerol polyricinoleate |
PhC | Photonic crystal |
PIL | Polyionic liquid |
PIL-IOM | Polyionic liquid inverse opaline microspheres |
PLGA | Poly(D,L-lactide-co-glycolide) |
PLLA | Poly(L-lactide) |
Pluronic F108 | Ethylene oxide–propylene oxide–ethylene oxide tri-block copolymer |
PMMA | Poly(methyl methacrylate) |
pNIPAAm | Poly(N-isopropylacrylamide) |
PPO | Poly(propylene oxide) |
PSf | Polysulfone |
PTE | Pentaerythritol triallyl ether |
PVA | Poly(vinyl alcohol) |
SCCB | Silica colloidal crystal bead |
SDS | Sodium dodecyl sulfate |
SERS | Surface-enhanced Raman scattering |
SFIL | Stop flow interference lithography |
SFL | Stop flow lithography |
Span80 | Sorbitan mono-oleate |
SPIONs | Superparamagnetic iron oxide nanoparticles |
SPS | Sodium peroxodisulfate |
SR454 | Ethoxylated-trimethylolpropane-triacrylate |
TEMED | N,N,N′,N′-Tetramethylethylenediamine |
TEOS | Tetraethyl orthosilicate |
THF | Tetrahydrofuran |
TPGDA | Tripropylene glycol diacrylate |
TT | Pentaerythritol tetrakis-(3-mercaptopropionate) |
TTT | Triallyl-135 triazine trione |
Morphology | Sb/group ref. | Method | Phase composition |
Q
d![]() ![]() ![]() ![]() ![]() ![]() |
Porogen | Average particle diameter/μm | Average pore size/μm | Porosity | Specific surface area (SSA)/m2 g−1 | CV (%) | |||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Surface | Internal | Inner (dispersed) | Outer (continuous) | Surface | Internal | ||||||||
Smooth skin | Macropores | S. Dubinsky et al. | UV-induced polymerization with porogen | A mixture of GMA and porogen | Aqueous solution of PVA and surfactant or THF | 1![]() ![]() |
DBP/DOP 1![]() ![]() |
About 58a | 0 | — | — | — | — |
DOP (0.5 wt% Triton X-100) | About 63.5a | 0 | — | — | — | — | |||||||
Gradient porous structures | T. Watanabe et al. | Selective solvent extraction | Sodium poly(styrenesulfonate) | Hexadecane and Span80 | 1![]() ![]() ![]() ![]() |
— | About 79a | 0 | 10–100 | 10–50% | — | <5 | |
Macropores | J. Y. Sim et al. | Microfluidic molding | — | — | — | — | — | 0 | 0.166–0.2 | — | — | — | |
Micro- to macropores | C. E. Udoh et al. | Selective solvent extraction | Polymer solution | Hexadecane with Span80 | Q d = 10 μL min−1; Qc = 50–90 μL min−1 | — | 50–200 | 0 | 1–100 | — | — | — | |
Golf-ball-like dimples | Macropores | S.-H. Kim et al. | UV-induced polymerization with wet etching process | Silica–ETPTA and Darocur 1173 | Aqueous solution with Pluronic F108 | 1![]() ![]() |
— | About 110a | 0.14–0.21a | — | — | — | — |
About 197a | 0.59–1.18a | — | — | — | — | ||||||||
Macropores | K.-H. Hwangbo et al. | Selective solvent extraction | A mixture of DCM, polymer, and PCM | PVA aqueous solution | <1![]() ![]() |
— | 185–280 | 5–20a | 4.4–9.5a | — | — | — | |
Macropores | J. H. Lee et al. | Solvent evaporation and phase separation | A mixture of PLGA, PCM, dichloromethane | PVA aqueous solution | — | — | About 235 ± 5 | 4.2–33.2a | Smaller than surface dimples | — | — | — |
Morphology | Sb/group ref. | Method | Phase composition |
Q
d![]() ![]() ![]() ![]() ![]() ![]() |
Porogen | Average particle diameter/μm | Average pore size/μm | Porosity | Specific surface area (SSA)/m2 g−1 | CV (%) | ||
---|---|---|---|---|---|---|---|---|---|---|---|---|
Inner (dispersed) | Middle (dispersed) | Outer (continuous) | ||||||||||
Opened macroporous spheres | J. X. Yun et al. | Cryo-polymerization | A mixture of MBAAm, TEMED, AAm, APS | — | Ethyl enanthate and Span 80 | 1![]() ![]() ![]() ![]() |
— | 800–1500 | 3–50 | 86.2% | — | — |
S. Dubinsky et al. | UV-induced polymerization with porogen | A mixture of GMA, DMPA, EGDMA, porogen | — | PVA aqueous solution | 2![]() ![]() |
DEP | 75 | 0.033a | — | 28.7 | 1.0 | |
2![]() ![]() |
DBP | 74 | 0.107a | — | 13.9 | 2.9 | ||||||
1![]() ![]() |
DOP | 110 | 0.179a | — | 6.6 | 0.8 | ||||||
1![]() ![]() |
DDP | 131 | 0.250a | — | 3.4 | 2.7 | ||||||
S.-H. Kim et al. | UV-induced polymerization with wet etching process | Silica-ETPTA and Darocur 1173 | — | Aqueous solution with Pluronic F108, BASF, EPTBC | 1![]() ![]() |
— | About 247.4a | About 0.15a | — | — | — | |
W. J. Duncanson et al. | One step polymerization with porogen | DCM and PLA | — | PVA aqueous solution | — | Self-assembling perfluorinated-dendrimer–dye complex | About 45–60a | About 0.9–1.25a | — | — | <6 | |
K. Jiang et al. | UV-induced polymerization with porogen | A mixture of GMA, SR454, DA, DMPA | — | PVA aqueous solution with Triton X-100 | 1![]() ![]() |
DA | About 77 ± 2 | 1.6 ± 0.6 | 44–50% | — | — | |
H. Zhang et al. | Single emulsion template with porogen | A mixture of HEMA, MMA, BDK, PVP K30, PGPR, EGPDA | — | PVA aqueous solution with glycerol | 3![]() ![]() |
PVP K30 | 47.6–57.5 | About 1.1–2.5a | — | — | 3.5 | |
M. T. Gokmen et al. | UV-induced polymerization of HIPE emulsion template | A mixture of EGDMA, GMA, PEO–PPO–PEO 2800, CaCl2·2H2O (HIPE) | — | PVA aqueous solution | 1![]() ![]() |
— | About 400 | About 15 | 78–87% | 16 | — | |
J. B. Wacker et al. | Evaporation-induced consolidation and self-assembly | Silica colloid aqueous solution | — | Oleic acid | 1![]() ![]() ![]() ![]() |
— | About 13–130a | About 0.8a | — | — | — | |
J. Wan et al. | UV-induced polymerization of three phase emulsion template | Pure nitrogen gas | Deionized water with SDS, DEAP, MBAAm, acylamide | Mineral oil with DEAP |
Q
w![]() ![]() |
— | <210a | About 28a | — | — | — | |
R. A. Prasath et al. | UV-induced polymerization and flow reaction | Thiol–ene monomers with π-bond, DMPA, porogen | — | SDS aqueous solution | 1![]() ![]() ![]() ![]() |
DOP, Xy, BuAc, a mixture of Xy and BuAc | 240–325 | 0.05–3 | — | 0–4.9 | — | |
Thiol–yne monomers with π-bond, DMPA, porogen | — | 240–300 | 0.2–1 | — | 2.2–35.6 | — | ||||||
Macroporous core–shell spheres | W.-C. Jeong et al. | Double emulsion template | PVA aqueous solution with FITC–dextran | pNIPAAm particle-dispersed mixture of DCM, hexane and ethyl cellulose | PVA aqueous solution | 3![]() ![]() ![]() ![]() |
— | About 455a | Sub- to a few micro-meters | — | — | — |
X. Gong et al. | UV-induced polymerization and flow reaction | H2O2 solution with PVA | NOA 61 | Liquid paraffin | 3![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
— | About 50–200 | Several micro-meters to tens of micro-meters | 70% (max) | — | — | |
TPGDA | Silicon oil | |||||||||||
EGDMA | Silicon oil | |||||||||||
S.-W. Choi et al. | Double emulsion template and fast solvent evaporation | PVA aqueous solution | A mixture of PLGA and DCM | PVA aqueous solution | 3![]() ![]() ![]() ![]() |
— | About 709.1 | About 1.5 | — | — | 1.7 | |
Dense spheres with thin surface pores | C. Paquet et al. | Self-assembling of materials | Chloroform solvent with SPIONs and copolymer | — | PVA aqueous solution with glycerol | 1![]() ![]() |
— | About 113–131 | <1 | — | About 5.2 | 2–8 |
Opened macroporous spheres | Toluene solvent with SPIONs and copolymer | |||||||||||
Dense porous spheres | A mixture of THF and toluene with SPIONs and copolymer | |||||||||||
More dense channel-like structure | Hexane solvent with SPIONs and copolymer | |||||||||||
Macroporous rods | M. T. Gokmen et al. | UV-induced polymerization of HIPE emulsion template | A mixture of EGDMA, GMA, PEO–PPO–PEO 4400, CaCl2·2H2O (HIPE) | — | PVA aqueous solution | 4![]() ![]() |
— | About 200 | About 1.25a | — | — | — |
J. B. Wacker et al. | Evaporation-induced consolidation and self-assembly | Silica colloid aqueous solution | — | Oleic acid | 1![]() ![]() ![]() ![]() |
— | About 45—100a | About 0.8a | — | — | — | |
Sponge-like open-celled porous microgels (interconnected pores) | C.-L. Mou et al. | UV-induced polymerization and homogeneous emulsification | O/W emulsions with benzyl benzoate, soybean oil, PGPR, LR300, V-50, NIPAM, MBAAm, F-127, and glycerol | — | Soybean oil with PGPR | 1![]() ![]() ![]() ![]() |
Fine oil droplets (adding isopropanol or heat up) | — | About 8a | — | — | 1.6 |
Pinecone-like structure at one end and a tip-shaped tail structure at another end (disconnected pores) | Deionized water with NIPAM, MBAAm, V-50, F-127 and glycerol | (Adding isopropanol) | — | — | — | — | 1.8 |
Morphology | Sb/group ref. | Method | Phase composition |
Q
d![]() ![]() ![]() ![]() ![]() ![]() |
Porogen | Average swelling ratio | Average pore size/μm | Porosity | Specific surface area (SSA)/m2 g−1 | CV (%) | |
---|---|---|---|---|---|---|---|---|---|---|---|
Inner (dispersed) | Outer (continuous) | ||||||||||
Highly interconnected and uniform macroporous scaffold | K. Chung et al. | Single emulsion templates | Nitrogen gas | Alginate solution with Pluronic F127 | P = 6 psi; Qc = 18 mL h−1 | — | 91.83 ± 6.02% | About 250 | 86.90 ± 5.09% | — | — |
C.-C. Wang et al. | Single emulsion templates | Nitrogen gas | Alginate solution with acetic acid and Pluronic F127 | P = 5 psi; Qc = 12 mL h−1 | — | 2607.41 ± 340.10% | About 225a | 97.25 ± 0.84% | — | — | |
A. van der Net et al. | Microfluidic foaming technique | Air | Solution A (AAm aqueous solution with MBAAm, TEMED, Lutesol® AT18) + solution B (an aqueous solution with Lutesol® AT18 and SPS) | 1![]() ![]() |
— | — | About 207–353a | — | — | — | |
C. Colosi et al. | Microfluidic foaming technique | Argon gas | An aqueous solution of PVA and CTABr | P = 67 kPa, Qc = 0.135 mL h−1 | — | — | 140 ± 17 | 63% | — | — | |
M. Constantini et al. | Microfluidic polyHIPEs emulsions | A cyclohexane solution | An aqueous solution of DEX-MA with Irgacure 2959, APS, Pluronic F-68 | About 4![]() ![]() |
— | — | About 136–241 | About 80% | About 19–35 | — | |
Highly order closed-cell polyhedral foam and closed-cell polyhedral foam | A. Quell et al. | Microfluidic polyHIPE emulsions | PolyHIPE monomers with KPS or AIBN | A mixture solution of styrene and divinylbenzene containing the surfactant Hypermer 2296 | — | — | — | 59–71 (KPS) or 57–75 (AIBN) | — | — | — |
Morphology | Sb/group ref. | Method | Phase composition |
Q
d![]() ![]() ![]() ![]() ![]() ![]() |
Porogen | Average particle diameter/μm | Average pore size/μm | Porosity | Specific surface area (SSA)/m2 g−1 | CV (%) | |
---|---|---|---|---|---|---|---|---|---|---|---|
Inner (dispersed) | Outer (continuous) | ||||||||||
Highly ordered inverse opal structure | Y. Zhao et al. | Sacrificial template self-assembly, evaporation-induced consolidation and calcination | An aqueous suspension containing monodisperse polystyrene spheres and ultrafine silica particles | Silicon oil | — | — | About 270a | About 0.2a | — | — | — |
J. Wang et al. | Sacrificial template self-assembly, evaporation-induced consolidation and wet etching process | An aqueous suspension containing monodisperse silica NPs | Silicon oil | 1![]() ![]() |
— | — | About 0.22b | — | — | — | |
J. Cui et al. | Sacrificial template self-assembly, UV-induced polymerization and wet etching process | — | — | — | — | 0.3 | About 0.2a | — | — | — | |
Densely packed opal cores and inverse-opal shells | B. Ye et al. | Sacrificial template self-assembly, UV-induced polymerization and selected etching process | — | — | — | — | — | About 0.2a | — | — | — |
Highly ordered macroporous film | M. Elsayed et al. | Sacrificial template (efficient burst of templated bubbles) | Polymeric solution containing sodium alginate, PEG-40S and phospholipids | Gas | P = 95–165 kPa, Qc = 200 μL min−1 | — | — | 70 ± 30 μm to 122 ± 15 μm | — | — | — |
Morphology | Sb/group ref. | Method | Phase composition |
Q
d![]() ![]() ![]() ![]() ![]() ![]() |
Porogen | Average particle size/μm | Average pore size/μm | Porosity | Specific surface area (SSA)/m2 g−1 | CV (%) | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Inner (dispersed) | Middle (dispersed) | Outer (continuous) | Large | Small | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
a Measure (obtained by analyzing/measuring optical microscopy images/SEM). b Speculation. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Microspheres with hollow interior and radially hierarchical porous shell | S.-W. Choi et al. | Double emulsions template and fast solvent evaporation | PVA aqueous solution | A homogenized W/O emulsion containing PVA and PLGA | PVA aqueous solution | 0.7![]() ![]() ![]() ![]() |
— | About 446 | About 4 | <1 | — | — | — | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Microspheres with bimodal macropore–mesopore structure (radially hierarchical) | Z. Zhai et al. | Temperature-induced gelation | A silica sol aqueous solution containing MC, PEG, HCL and TEOS | — | A liquid paraffin solution containing Span 85 | 1![]() ![]() |
— | About 490 | About 0.1–1 | About 0.0138–0.0205 | — | About 309–390 | 2.1–2.5 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Multi-cored microcapsules with porous surface (radially hierarchical) | S.-H. Kim et al. | Double emulsion template, UV-induced polymerization and wet etching process | An aqueous solution of food coloring pigments or FITC-labeled dextran | A ETPTA suspension containing silica particles treated with DCDMS | An aqueous solution of ethyleneoxide–propylene oxide–ethylene oxide tri-block copolymer, BASF and Pluronic F2109 |
Q
M![]() ![]() |
— | About 200–300a | About 100b | About 0.7a | — | — | — | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Radially hierarchical interconnected porous spheres | N. J. Carroll et al. | Microfluidic templating and evaporation-induced consolidation | An aqueous solution with CTAB, TEOS, and HCl | Pure hexadecane oil | Hexadecane solution with ABIL EM 90 |
Q
M![]() ![]() ![]() ![]() |
— | 7.7 ± 1 | Tens of nanometers | Single nanometer | — | About 650–1250 | — | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Radially hierarchical porous microspheres | Y. Fan et al. | The synergistic effect between the sol–gel process and solvent extraction in droplets | Dichloromethane solution with PCL and TEOS | — | Mixture of PVA and ammonia solution | — | — | About 65 | 7.5a | — | — | 333.698 | — | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Chitosan inverse opal scaffold with ordered hierarchical pores | S.-W. Choi et al. | Sacrificial template, heat treatment and sedimentation–evaporation method | A PCL solution in MC | — | PVA aqueous solution | — | — | 4.5 mm in diameter and 1.5 mm in thickness | About 110a | About 5a | — | — | <5 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Uniform microparticles with controllable highly interconnected hierarchical porous structures | M.-J. Zhang et al. | Double emulsion template and mass transfer | An aqueous solution containing glycerin and Pluronic F127 | MMA containing EGDMA, PGPR, and HMPP | An aqueous solution containing glycerin and Pluronic F127 | — | — | About 217a | About 130–167a | About 2a | 0–50% | — | — | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Hierarchical porous silk-based materials | M. R. Sommer et al. | Sacrificial template and 3D printing technology | DCM solution with PCL | — | PVA aqueous solution | Q d = 1–5 mL h−1; Qc = 5–30 mL h−1 | — | — | <335b | — | — | — | — |
This journal is © The Royal Society of Chemistry 2017 |