Droplet Microfluidics for Fabrication of Non-Spherical Particles (ktian)
Written by Kevin Tian, AP 225, Fall 2011
--Ktian 17:04, 9 November 2011 (UTC)
Title: Droplet Microfluidics for Fabrication of Non-Spherical Particles
Authors: Ho Cheung Shum, Adam R. Abate, Daeyeon Lee, Andre´ R. Studart, Baoguo Wang, Chia-Hung Chen, Julian Thiele, Rhutesh K. Shah, Amber Krummel, David A. Weitz
Journal: Macromolecular Rapid Communications (2009), Volume: 31, Issue: 2 [Feature Article]
In general, all colloidal-scale particles assume a spherical shape naturally due to the dominance of surface tension during their creation. However although spherically shaped particles are far easier to generate, there are a number of potential applications for non-spherical shapes. For example it has been shown such shapes can have a denser packing compared to spherical packing. Anisotropic shapes also exhibit different hydrodynamic, electric and magnetic properties. In a more general setting, non-spherical shapes can also be used for structural reinforcement, through platelets and fibers. Since there is a need for these non-spherical shapes, a high throughput method of creating them is necessary.
This paper summarizes various efforts to utilize droplet-based microfluidics to create non-spherical droplets. Single and Multiple emulsion droplets are used as the basic platform to create these non-spherical droplets, which is achieved by a variety of spontaneous processing steps.
Single Emulsions - Droplets of one fluid (dispersed phase) are suspended in an immiscible fluid (continuous phase).
Double Emulsions - Droplets of one fluid are suspended inside of a drop of another fluid, which are suspended in a third outer-most continuous phase.
Multiple Emulsions - Droplets of a fluid suspended inside of a drop of another fluid, suspended inside of a drop of another fluid, suspended inside of a drop of another fluid...(ad infinitum).
As mentioned before, these emulsions are then used as templates for non-spherical particles. The flow of the processes is roughly as follows:
- Single Emulsions
- Arrested Coalescnece (of particle-stabilized droplets)
- Asymmtric solidification of droplets of polymer solutions
- Double Emulsions
- Polymerization in microfluidic flow
- Evaporation-driven Clustering [Clustering of Inner Droplets]
All above processes generally used PDMS microfluidic devices, unless aggressive chemicals were used. In which case 3D glass devices were used in fabrication procedures.
In general, one must note that the single emulsion techniques are far easier to implement, however are limited in the complexity of the shape they can create. Double emulsions (and beyond) are naturally more difficult to engineer but have far greater sophistication in terms of ability to engineer various shapes.
Arrested Coalescence [Single]
The idea behind arrested coalescence is to arrest the relaxation of a set of coalescing particles before the process is complete. This approach has been demonstrated through the use of Pickering emulsions, single emulsions containing droplets stabilized via interfacially adsorbed particles.
Though Pickering emulsions themselves are kinetically stable (for long but not indefinite periods of time) an incomplete coverage of the droplet surface by particles can be used to initiate coalescence. For particles that are irreversible adsorbed at the phase boundary between dispersed and continuous fluid phases, there is an associated increase of surface-to-volume ratio. This then has an accompanying increase of doplet surface coverage (by up to 26%) which can lead to the arresting of the coalescence process, due to two-dimensional jamming of the adsorbed particles. This effect is expected to occur with initial surface coverage ranging from 0.71 to 0.90.
In order to exploit this phenomenon to achieve well-defined structures the authors use monodisperse droplets produced in a glass microfluidic device as the base. These droplets are then coated with colloidal particles. The coverage can be tuned by altering the flow rates of the inner and outer fluids used in the droplet formation (but only to obtain partial coverage, as described above). It is noted that coalescence occurs within tenths of microseconds, however this only occurs when uncoated surfaces are exposed to each other. Thus droplet rotation/particle rearrangement at interfaces are necessary to initiate the coalescence. Though this naturally occurs after droplet contact, it takes several seconds after contact for this rearrangements to occur. This delay allows one to arbitrarily arrange droplet positions to create target geometries before coalescence begins.
One example is observed in Figure 3, where by using toluene as the inner phase, oil droplets were created that cream to the emulsion-air interface and self-assembled into hexagonal close packed arrays. Adjacent single droplets coalesce, leading to a non-spherically shaped particle. An interesting addition is that though jamming occurs at one site on the droplet surface, this does not jam the entire oil-water interface. This leads to one being able to create situations where one has multiple occurrences of arrested coalescence on a single droplet. Figure 4 illustrates the possibilities given the set of hexagonally arranged droplets in Figure 3, showing triangular arrangements as well as short chains of particles.
These structures can be dried and harvested without impacting the well-defined structures by polymerizing the inner dispersed phase. The authors note that the results are only the result of randomized particle rearrangement and arrested coalescence (which is somewhat disappointing, given that they emphasized how positioning the single droplets was the basis of this being used for construction of non-spherical structures). Though it certainly is a promising route of generating non-spherical structures/droplets/particles, the manipulation of these sets of particles into well defined templates and forcing arrested coalescence in a high throughput fashion is non-trivial. Whether this can lead to an effective technique remains to be seen.
Asymmetric Polymer Solidification [Single]
Polymer solidification is an alternative method of making non-spherical particles using the single emulsion precursor. The process involves dissolving a polymer in a volatile good solvent (whatever that means). This polymer solution is used as the dispersed phase of any emulsion, which can be evaporated to produce polymer particles. This evaporation process is typically isotropic, which would yield the spherical particles. Using microfluidic devices it is possible to cause anisotropic evaporation to occur.
First droplets of the polymer solution are prepared by a co-flow of some continuous phase (exact identity not mentioned). This yields the spherical polymer solution suspended as shown in Figure 5a. The flow is then directed into a cylindrical microchannel. As the flow is directed into the microchannel, a small amount of the solvent is dissolved into the continuous phase until it is saturated with the solvent (this is possible in spite of the immiscibility of the two fluids, though the amount is very limited). Since we have laminar flow (no turbulent mixing) we have differing velocities of the continuous phase along the axial direction of the channel. Since local stagnation of the continuous phase occurs right in front/back of the droplet, these regions remain saturated with solvent. On the other hand the outer regions has significantly less solvent due to the difference in flow speed.
The solidification of polymer thus initiates after polymer concentration hits a critical value (mostly at the circumference). We note that the remaining polymer in the droplet begins to deposit itself on the solidified surface, which decreases polymer concentration at the center of the droplet. Eventually the droplet is stretched and pinches out. This yields the torroidal shape shown in Figure 5b.
The strength of this approach is that it is applicable to a range of polymers (such as poly-ether sulfone (PES), polysulfon (PSF), etc.) assuming the appropriate solvent is used (such as N,N-dimethylformamide (DMF)). A surfactant is added over the course of droplet solidification in order to stabilize the drop (this surfactant must be immiscible with the solvent, yet allows solvent to diffuse into it). Functionalization of the torroidal particles is possible by incoporating components such as fluorescent dyes and magnetic nanoparticles into the solvent mixture, thus yielding various kinds of torroidal particles with several different applications (Figure 7).
It is believed that with careful tuning of the fluid flow (or controlling the environment), this concept can be applied to create more complex geometries with a very wide range of functions.
Polymerization in Microfluidic Flow [Double]
The basic method of creating a double emulsion is not necessarily more complicated however, as the basic schematic for a double emulsion microfluidic device shows in Figure 8. It simply involves an extra step over the single emulsion. however when the double emulsion drops are generated, the inner droplets do not typically remain in the center of the outer droplet due to viscous forces and density differences between the two fluids. Additionally if the channel height is insufficient to contain the diameter of the double emulsion drop, deformation of the drop occurs as they flow through the channel. Polymerizing the droplet while in the channel locks in the non-spherical shape, thus allowing for a relatively straightforward generation of a (basic) non-spherical shape.
To demonstrate the utility of the approach the authors formed microparticles with uniformly anisotropic structures. The particles consisted of a hydrogel shell with a enclosed magnetic core. This is made by making an oil-in-water-in-oil double emulsion, that is locked in via photo-polymerization. The photo-polymerization is achieved by making the shells with a acrylamide monomer solution. By adjusting flow rates during formation, one can control size and internal structure (by drop size and encapsulating fewer or more particles). These particles are then solidified, dried and redispersed in water . Though the aprticles swell somewhat, the essential anisotropic structure remains intact. This entire example is illustrated in Figure 9.
In general though it appears that polymerization during flow of the particles in varying microfluidic channels can create a small variety of shapes and sizes. However the greatest advantage is the there is a great amount of variation in the materials that can be selected for use. In addition the double emulsion allows for various "cores" of different functionality to be encapsulated in the structure.
Clustering of Inner Droplets [Double]
Another method described in the paper involves the packing of multiple inner droplets inside a small volume of the outer shell phase. If one selects solutions that have the property that the interface between inner and outer droplet/fluid phases is less deformable than the interface between the outer droplet and the continuous phase, then one obtains deformed droplets; i.e. non-spherical double emulsions. This approach is illustrated with double emulsions with multiple inner droplets to form non-spherical colloidalsomes with multiple compartments.
Colloidalsomes are defined as capsules whose shell consists entirely of densely packed colloidal particles. The typical fabrication method involves particle-covered water-in-oil emulsion droplets that are then transferred into an aqueous phase. By using different particles to form the shells, one can tune colloidosomes to different applications to suit the permeability, biocompatibility, and functionality required. Shape has been the greatest challenge in so far as colloidosome fabrication.
Figure 12 illustrates the process of generating the double emulsion with multiple inner droplets. The number of inner droplets per outer droplet can be tuned by varying the relative flow rates of the three components. The multi-compartment colloidosomes are generated by simply removing the outer oil phase from the double emulsion. The reasoning is that due to the Laplace pressure across water/oil interfaces being equal, dependent on surface curvature, the inner droplets maintain their spherical shape during solvent removal. The result is shown in Figure 14. This technique relies on the jamming of the particles in order to solidify the precursor emulsion drops.
The technique is not limited to colloidosome fabrication, since the only requirements pertain to relative ease of deformation of the inner droplets compared to the encapsulating outer droplet and the stability of the double emulsion droplets. This allows for the inner, less deformable "droplets" to actually be solid particles as well, which can be used to encapsulate solids in a non-spherical double emulsion (figure 15).
The essential advantage of this technique over polymerization is that this non-spherical shape can be maintained in the absence of a flow or confining walls (since the non-spherical shape is the equilibrium state). Thus the formation of the double-emulsion and solidification need not necessarily be done in the same step. The need for a polymerization, which is certainly at times a non-trivial implementation, is also absent.
The general theme of the paper has been the various ways that droplet microfluidics can be used to fabricate non-spherical particles using single and multiple emulsions. Complicated geometries and particle structures are possible (in principle) and some basic forays into this realm have been explored. However as the authors aptly note, the major challenge is the mass production of these particles. Though many techniques are waiting to be discovered, the up-scaling of these technologies determines whether the commercial application of non-spherical particles is possible.
Another note is whether microfluidics is capable of fabricating particles in the sub-micron range (as all techniques previously discussed are in the micron range). Though the techniques are theoretically still applicable for nanometer sized particles, considerations that were simple before begin to become significant practical concerns. This includes fluid pumping pressures, materials selection, among others. However in general one can confidently say that this area holds much promise. The utilization of very simple colloidal systems to create micron-sized particles with reasonable throughput (if one parallelizes the microfluidic devices).