Lab on a Chip HIGHLIGHT
Cite this: Lab Chip, 2014, 14, 1491
View Article Online View Journal | View Issue
Research highlights: printing the future of microfabrication
Published on 27 March 2014. Downloaded on 22/10/2014 19:23:15.
Peter Tseng, Coleman Murray, Donghyuk Kim and Dino Di Carlo In this issue we highlight emerging microfabrication approaches suitable for microfluidic systems with a focus on “additive manufacturing” processes (i.e. printing). In parallel with the now-wider availability of low cost consumer-grade 3D printers (as evidenced by at least three brands of 3D printers for sale in a recent visit to an electronics store in Akihabara, Tokyo), commercial-grade 3D printers are ramping to higher and higher resolution with new capabilities, such as printing of multiple materials of different transparency, and with different mechanical and electrical properties. We highlight new work showing that 3D printing (stereolithography approaches in particular) has now risen as a viable technology to print whole microfluidic devices. Printing on 2D surfaces such as paper is an everyday experience, and has been used widely in analytical chemistry for printing conductive materials on paper strips for glucose and
DOI: 10.1039/c4lc90023e
other electrochemical sensors. We highlight recent work using electrodes printed on paper for digital microfluidic droplet actuation. Finally, we highlight recent work in which printing of membrane-bound droplets that interconnect through bilayer membranes may open up an entirely new approach to
www.rsc.org/loc
microfluidic manufacturing of soft devices that mimic physiological systems.
Mail-order microfluidics While the vast majority of microfluidic systems have been prototyped using polydimethylsiloxane (PDMS) soft lithography, transitioning from prototype to product remains a challenge due to the fact that soft lithography requires a significant capital investment to automate and scale up. Hand-made PDMS devices can scale to volumes of hundreds, however reproducibility and polymer flexibility often lead to commercialization challenges.1 Additionally high volume production methods, such as injection molding, are extremely expensive for early stage companies that only need hundreds of reproducible parts for clinical trials or proof-of-concept studies. Separately, PDMS-based molding off of silicon wafers has limited most microfluidic designs to simple multilayer 3D shapes. To address this niche, Folch and colleagues2 demonstrated that commercial stereolithography (i.e. a 3D printing methodology) can serve as a medium volume production and prototype friendly microfluidic fabrication technique. Although 3D printed materials have recently been used as molds for soft lithography,3 here the authors are one of the first to show that stereolithography can generate 3D microfluidic devices directly2,4 with relative ease compared to the
Department of Bioengineering, California NanoSystems Institute, Jonsson Comprehensive Cancer Center, University of California Los Angeles, 420 Westwood Plaza, 5121 Engineering V, Box 951600, Los Angeles, California 90095, USA. E-mail: [email protected]
This journal is © The Royal Society of Chemistry 2014
quasi 3D designs obtainable using soft lithography. Both 2D and 3D microfluidic devices were drafted in Autodesk Inventor and sent to FineLine Prototyping Inc., which produced the devices using a 3D Systems Viper SL stereolithography system and sent the completed devices back by mail. The devices were fabricated with DSM Somos® WaterShed XC 11122 resin due to its transparency and biocompatibility. Folch et al. discuss the limiting factors of this approach, including drainage of uncured polymer, optical clarity, and the SL system's z-height resolution. In analyzing the fabrication resolution for stereolithography, the authors report that though most SL systems have a resolution of approximately 100 μm, the limiting factor for microfluidic devices is the effective drainage of uncured liquid resin. The “hydrodynamic” resolution is limited to microchannels that are 500 μm × 500 μm or 635 μm × 635 μm depending on device complexity. To further quantify the resolution, the authors printed microchannels with a 500 μm height and widths ranging from 50 to 1000 μm. Channels with 50 to 200 μm widths could not be effectively cleared of uncured resin, while a 300 μm channel width geometry was successfully cleared but had significant necking. Channels with widths ranging from 400 μm to 1000 μm were successfully cleared and demonstrated a deviation from the desired geometry of approximately 5%. An observed fabrication byproduct was surface roughness due to laser over-curing. The surface roughness was determined to be approximately ±2.54 μm, which constitutes
Lab Chip, 2014, 14, 1491–1495 | 1491
View Article Online
Published on 27 March 2014. Downloaded on 22/10/2014 19:23:15.
Highlight
approximately 0.5% of the channel topography for a 500 μm high channel. To reduce imaging challenges due to surface roughness, the authors utilized Wacker AP 150 silicone oil with a refractive index of 1.510 to match the chips refractive index of 1.512. By coating the outer surfaces of the devices during imaging, the authors were able to reduce the roughness prominence on the outer surfaces and enhance the visibility of the internal geometry. Additionally the authors changed the channel build orientation such that the overcured features occurred along the channel's walls as opposed to its roof. However, flipping the build axis resulted in a stair-stepping effect due to the fact that the region of photopolymer cured by the laser beam has a Gaussian or parabolic profile. To assess the feasibility of imaging cells within SL produced microchannels, CHO-K1 cells stained with Calcein AM and seeded into the flat-built three-inlet device were compared to stained CHO-K1 cells on a glass slide as a control. Phase contrast imaging at 20× magnification demonstrated relatively crisp images with slightly increased background due to resin surface roughness. Fluorescence imaging at the same magnification yielded similar results with some autofluorescence. In a cost analysis of soft lithography fabrication, the authors concluded that a single PDMS chip costs approximately $215 due to the manual process steps required. The authors asserted that stereolithography fabrication would cost significantly less because SL systems are highly automated. Furthermore, devices can either be purchased via mail order SL companies or SL systems can be purchased at prices ranging from ~$2000 to $13 000 and brought in house. Developments to improve 3D printing resolution5 and further reduce costs are expected to expand the use of such a 3D printing approach to a wider range of microfluidic applications
Lab on a Chip
that depend on smaller channel dimensions. Folch predicts the transition will occur within the next 5 years: “Once the remaining challenges are overcome, I believe that the field of microfluidics will have largely abandoned PDMS molding by 2018”. The authors also demonstrate a quite exciting aspect of the work in which complex 3D structures can be fabricated in an integrated fashion, including devices with fluid paths that are folded to reduce footprint and seamlessly integrated world to chip fluid connections such as Luer couplings (Fig. 1). This new capability should allow the imagination of the designer to roam with many new innovative and useful design strategies expected compared to the usual 2D layer-bylayer thinking tied to traditional lithography.
Active, paper-based, digital microfluidics Paper-based microdevices have recently seen significant interest due the ubiquity of their base material–pressed cellulose fibers, their ease of processing (via straightforward writing and printing), and their low-cost, portability, and disposability.6 It is believed that these characteristics will facilitate their use in resource-limited settings by providing end-users with inexpensive and straight-forward devices to operate. Conventional paper-based microfluidic chips allow controlled reactions between reagent and sample streams by generating regions of hydrophilicity and hydrophobicity via a number of approaches (wax patterning, flexographic printing of styrene, or chemical modification of cellulose). Transport can then naturally occur via capillary and wetting forces in the hydrophilic regions. A large number of components have been developed in conjunction with these devices, including
Fig. 1 Photographs of (a) a coil-shaped microchannel device and (b) the interfacing of tubing with integrated Luer connectors. Micrographs demonstrating (c) how void areas can be removed from the footprint of SL devices, (d) how closed chambers can be optically accessed from precisely-defined sidewalls of a device, (e) an inlet distributor with a folded geometry, and (f) a Luer inlet oriented parallel to the plane of the chip. Adapted from Folch et al.
1492 | Lab Chip, 2014, 14, 1491–1495
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 27 March 2014. Downloaded on 22/10/2014 19:23:15.
Lab on a Chip
fluidic modulators (switches, valves, timers, and meters), paper batteries, and displays.6 These have been used in the generation of a large variety of paper-based bio-assays and chemical sensors. Even with these respective technological advances, paper chips still possess a number of limitations, including the required pre-definition of fluidic pathways, and the pre-implantation of chemicals and probes. In an attempt to leverage the cost and other advantages of paper chips in combination with the active control enabled by digital microfluidics,7 Shin and colleagues used inkjet printing to pattern electrodes onto cellulose and allow electric field manipulation on paper-based chips.8 Such patterning of electrode materials onto paper is similar to previous approaches used in mass-producing paper-based glucose test strips. However, the authors subsequently used this capability to perform dynamic electro-wetting manipulation of droplets via an external power-source, which they term as an active paper open chip (APOC) (Fig. 2). This integration allows the addition of active transportation to traditionally passive paper-based devices. Electro-wetting operates when the application of an electric field to a fluid changes its surface tension via the translocation of charge. By controlling the locality of this phenomenon via electrodes, droplets can be controllably steered. In this paper, the authors printed these electrodes by formulating a conductive carbon nanotube (1 μm length) ink, to be used with an inkjet printer, which they subsequently printed onto photo paper. These electrodes were passivated by deposition of a thin film layer of parylene-C (1 μm), and through the spinning of Teflon AF polymer (200 nm),
Highlight
yielding a smooth, hydrophobic surface. Lastly, a thin film layer of silicone oil was spin-coated onto the paper chip to lubricate droplet motion and adjust the surface tension. Droplets could subsequently be manipulated via application of field in an open air configuration. A number of droplet manipulations were demonstrated, including the continuous manipulation of droplets in a circle of electrodes, and the merging and splitting of droplets on electrode rails. These were on electrodes with a pitch of 1.5 to 2 mm, and spacings of around 0.5 mm. Due to the limitation of inkjet printing resolution, the authors attempted a number of configurations and studied their effect on droplet motion, including varying the excitation voltage, whether it was AC or DC, and the surface of the substrate (Teflon, or with oil). They found an AC voltage of 70 Vrms was required to optimally overcome the kinetics of the system, and with this they could manipulate droplets of above 300 μL. In a final demonstration, the authors created a prototype of a functional digital paper chip, by introducing intersecting lanes carrying dye. Droplets of different colored dyes could be merged and subsequently mixed by controlled voltages, undergo multiple steps of merging and manipulation, and finally be analyzed at a final location. In summary, Shin and colleagues demonstrated a potential method of activating paper-based microchips for digital microfluidic applications via the deposition of carbon nanotube inks, which were capable of precisely controlling the trajectories of droplets along the chip surface. It remains an open question of how transitioning to such a lower cost base material will lead to new digital microfluidic applications.
Fig. 2 a) Design of active, paper-based microfluidic chips. b) and c) Printed paper-electrodes and their manipulation of droplets. d) Compatibility of the technique with existing sheets of photo paper. Adapted from Shin et al.
This journal is © The Royal Society of Chemistry 2014
Lab Chip, 2014, 14, 1491–1495 | 1493
View Article Online
Highlight
Published on 27 March 2014. Downloaded on 22/10/2014 19:23:15.
Printed droplet networks As discussed above, a variety of fabrication methods are being explored to achieve enclosed or open microfluidic manipulation of fluids at various scales, and Villar et al. recently presented an innovative fabrication method that may open up a new genre of fluid manipulation designs.9 The human body is a complex network of micro-/macro-scale fluidic systems in which the individual components are cells that, when connected, can communicate fluidically through membrane junctions and protein pores. Microfabrication approaches have enabled controlled placement of compartments separated by bilayer membranes, however, such systems have not yet realized large assemblies and complex geometries that could be of use for complex assays or mimicking physiological connections.10,11 Villar et al. addressed this challenge by inventing a software-controlled printing system that creates lipid-coated aqueous droplet networks9 (Fig. 3). Their approach was to build up layers of homo- or heterologous aqueous droplets
Lab on a Chip
by ejecting droplets into a bath of lipid-containing oil. The position of the oil bath was controlled by a software-driven, motorized micromanipulator such that a horizontal crosssection of homogeneous/heterogeneous aqueous droplets from two separate droplet ejectors was layered over previously built layers in an automated manner. Their home-built printing system was capable of printing at a rate of ~1 droplet per second, and the resulting printed network consisted of aqueous compartments within single lipid bilayers, which were stable for weeks. With a stable, self-supportive network of droplets, Villar et al. first tested its capability of realizing electrical connections through the network. They used staphylococcal α-hemolysin (αHL) to define an electrical pathway within the droplet network. First they printed a pathway consisting of αHL-containing droplets (αHL+) within droplets without αHL (αHL−). Large drops of αHL pore-containing buffer were used to make the electrical contact between the entire droplet network and an electrode pair (Ag/AgCl electrode). When the circuit was connected through an αHL+ pathway at 50 mV in
Fig. 3 (A) Illustration of a printed droplet network. (B) Three orthogonal views of a single network printed according to a design with a scale bar of 1 mm. (C) Micrographs of a 3-D droplet network printed in bulk aqueous solution with a scale bar of 400 μm. (D) Schematic of a printed network with an αHL+ pathway. Green indicates drops with αHL. (E) Image of a printed network–electrode pair connection through an αHL+ pathway. Scale bar of 500 μm. (F) Amperogram showing the stepwise current increase from the configuration shown in (E) at 50 mV. (G) Images of a flower-shaped droplet network folding into a hollow sphere structure by osmosis. Scale bar of 200 μm. (H) Folding simulation of the processes illustrated in (G). Images from ref. 9 with permission.
1494 | Lab Chip, 2014, 14, 1491–1495
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 27 March 2014. Downloaded on 22/10/2014 19:23:15.
Lab on a Chip
1 M KCl, a stepwise ionic current increase was measured. When the circuit was connected through an αHL− pathway, on the other hand, only a capacitive transient ionic current was measured. This confirmed that the observed stepwise ionic current increase was due to αHL pores interfacing between the electrode buffer drop and printed droplet network. They also performed a computational simulation for the electrical connection through αHL+/αHL− droplet pathways and the calculated current from this model matched with the experimentally obtained current increase, confirming the desired electrical connection through the printed droplet network. Lastly, Villar et al. tested whether their printed droplet network, with defined electrical connections through macroscopic assembly of droplets, could allow fluid flow. They verified that water permeates through the bilayers at the interface of droplets, and thus confirmed that an osmolarity gradient among the droplets could be used to program a macroscopic shape change of the entire droplet network. They next built a model to predict the final, folded geometry from the initial geometry of the printed network. With the model, they designed droplet networks folding as desired, and successfully achieved various network geometries from circular (simple and easily printable) to hollow spheres (difficult to directly print). Their system successfully enables designs of 3-D geometries of soft materials that contain microcompartment structures of aqueous fluid bound by lipid bilayers. The bilayer is the means by which the macroscopic droplet network communicates to other droplets via membrane proteins or osmotic flow of water. More sophisticated individual/collective signal (fluid, chemicals, and others) transduction mechanisms may be achieved by utilizing additional membrane proteins or stimulus-responsive molecules, to replicate
This journal is © The Royal Society of Chemistry 2014
Highlight
physiological fluidic networks, or perform complex assays. Further development in "fixing" such soft systems once configured will be important to transition to robust devices that can be widely used in assays, suggesting that the first uses will be in scientific investigations of membrane networks.
References 1 E. Sollier, C. Murray, P. Maoddi and D. Di Carlo, Lab Chip, 2011, 11, 3752–3765. 2 A. K. Au, W. Lee and A. Folch, Lab Chip, 2014, 14, 1294–1301. 3 P. H. King, G. Jones, H. Morgan, M. R. R. de Planque and K.-P. Zauner, Lab Chip, 2014, 14, 722–729. 4 A. I. Shallan, P. Smejkal, M. Corban, R. M. Guijt and M. C. Breadmore, Anal. Chem., 2014, 86, 3124–3130. 5 M. Hatzenbichler, M. Geppert, R. Seemann and J. Stampfl, Proc. SPIE, 2013, 8618, 86180A. 6 A. K. Yetisen, M. S. Akram and C. R. Lowe, Lab Chip, 2013, 13, 2210. 7 K. Choi, A. H. C. Ng, R. Fobel and A. R. Wheeler, Annu. Rev. Anal. Chem., 2012, 5, 413–440. 8 H. Ko, J. Lee, Y. Kim, B. Lee, C.-H. Jung, J.-H. Choi, O.-S. Kwon and K. Shin, Adv. Mater., 2014, DOI: 10.1002/ adma.201305014. 9 G. Villar, A. D. Graham and H. Bayley, Science, 2013, 340, 48–52. 10 C. E. Stanley, K. S. Elvira, X. Z. Niu, A. D. Gee, O. Ces, J. B. Edel and A. J. deMello, Chem. Commun., 2010, 46, 1620–1622. 11 S. Thutupalli, S. Herminghaus and R. Seemann, Soft Matter, 2011, 7, 1312.
Lab Chip, 2014, 14, 1491–1495 | 1495
Cite this: Lab Chip, 2014, 14, 1491
View Article Online View Journal | View Issue
Research highlights: printing the future of microfabrication
Published on 27 March 2014. Downloaded on 22/10/2014 19:23:15.
Peter Tseng, Coleman Murray, Donghyuk Kim and Dino Di Carlo In this issue we highlight emerging microfabrication approaches suitable for microfluidic systems with a focus on “additive manufacturing” processes (i.e. printing). In parallel with the now-wider availability of low cost consumer-grade 3D printers (as evidenced by at least three brands of 3D printers for sale in a recent visit to an electronics store in Akihabara, Tokyo), commercial-grade 3D printers are ramping to higher and higher resolution with new capabilities, such as printing of multiple materials of different transparency, and with different mechanical and electrical properties. We highlight new work showing that 3D printing (stereolithography approaches in particular) has now risen as a viable technology to print whole microfluidic devices. Printing on 2D surfaces such as paper is an everyday experience, and has been used widely in analytical chemistry for printing conductive materials on paper strips for glucose and
DOI: 10.1039/c4lc90023e
other electrochemical sensors. We highlight recent work using electrodes printed on paper for digital microfluidic droplet actuation. Finally, we highlight recent work in which printing of membrane-bound droplets that interconnect through bilayer membranes may open up an entirely new approach to
www.rsc.org/loc
microfluidic manufacturing of soft devices that mimic physiological systems.
Mail-order microfluidics While the vast majority of microfluidic systems have been prototyped using polydimethylsiloxane (PDMS) soft lithography, transitioning from prototype to product remains a challenge due to the fact that soft lithography requires a significant capital investment to automate and scale up. Hand-made PDMS devices can scale to volumes of hundreds, however reproducibility and polymer flexibility often lead to commercialization challenges.1 Additionally high volume production methods, such as injection molding, are extremely expensive for early stage companies that only need hundreds of reproducible parts for clinical trials or proof-of-concept studies. Separately, PDMS-based molding off of silicon wafers has limited most microfluidic designs to simple multilayer 3D shapes. To address this niche, Folch and colleagues2 demonstrated that commercial stereolithography (i.e. a 3D printing methodology) can serve as a medium volume production and prototype friendly microfluidic fabrication technique. Although 3D printed materials have recently been used as molds for soft lithography,3 here the authors are one of the first to show that stereolithography can generate 3D microfluidic devices directly2,4 with relative ease compared to the
Department of Bioengineering, California NanoSystems Institute, Jonsson Comprehensive Cancer Center, University of California Los Angeles, 420 Westwood Plaza, 5121 Engineering V, Box 951600, Los Angeles, California 90095, USA. E-mail: [email protected]
This journal is © The Royal Society of Chemistry 2014
quasi 3D designs obtainable using soft lithography. Both 2D and 3D microfluidic devices were drafted in Autodesk Inventor and sent to FineLine Prototyping Inc., which produced the devices using a 3D Systems Viper SL stereolithography system and sent the completed devices back by mail. The devices were fabricated with DSM Somos® WaterShed XC 11122 resin due to its transparency and biocompatibility. Folch et al. discuss the limiting factors of this approach, including drainage of uncured polymer, optical clarity, and the SL system's z-height resolution. In analyzing the fabrication resolution for stereolithography, the authors report that though most SL systems have a resolution of approximately 100 μm, the limiting factor for microfluidic devices is the effective drainage of uncured liquid resin. The “hydrodynamic” resolution is limited to microchannels that are 500 μm × 500 μm or 635 μm × 635 μm depending on device complexity. To further quantify the resolution, the authors printed microchannels with a 500 μm height and widths ranging from 50 to 1000 μm. Channels with 50 to 200 μm widths could not be effectively cleared of uncured resin, while a 300 μm channel width geometry was successfully cleared but had significant necking. Channels with widths ranging from 400 μm to 1000 μm were successfully cleared and demonstrated a deviation from the desired geometry of approximately 5%. An observed fabrication byproduct was surface roughness due to laser over-curing. The surface roughness was determined to be approximately ±2.54 μm, which constitutes
Lab Chip, 2014, 14, 1491–1495 | 1491
View Article Online
Published on 27 March 2014. Downloaded on 22/10/2014 19:23:15.
Highlight
approximately 0.5% of the channel topography for a 500 μm high channel. To reduce imaging challenges due to surface roughness, the authors utilized Wacker AP 150 silicone oil with a refractive index of 1.510 to match the chips refractive index of 1.512. By coating the outer surfaces of the devices during imaging, the authors were able to reduce the roughness prominence on the outer surfaces and enhance the visibility of the internal geometry. Additionally the authors changed the channel build orientation such that the overcured features occurred along the channel's walls as opposed to its roof. However, flipping the build axis resulted in a stair-stepping effect due to the fact that the region of photopolymer cured by the laser beam has a Gaussian or parabolic profile. To assess the feasibility of imaging cells within SL produced microchannels, CHO-K1 cells stained with Calcein AM and seeded into the flat-built three-inlet device were compared to stained CHO-K1 cells on a glass slide as a control. Phase contrast imaging at 20× magnification demonstrated relatively crisp images with slightly increased background due to resin surface roughness. Fluorescence imaging at the same magnification yielded similar results with some autofluorescence. In a cost analysis of soft lithography fabrication, the authors concluded that a single PDMS chip costs approximately $215 due to the manual process steps required. The authors asserted that stereolithography fabrication would cost significantly less because SL systems are highly automated. Furthermore, devices can either be purchased via mail order SL companies or SL systems can be purchased at prices ranging from ~$2000 to $13 000 and brought in house. Developments to improve 3D printing resolution5 and further reduce costs are expected to expand the use of such a 3D printing approach to a wider range of microfluidic applications
Lab on a Chip
that depend on smaller channel dimensions. Folch predicts the transition will occur within the next 5 years: “Once the remaining challenges are overcome, I believe that the field of microfluidics will have largely abandoned PDMS molding by 2018”. The authors also demonstrate a quite exciting aspect of the work in which complex 3D structures can be fabricated in an integrated fashion, including devices with fluid paths that are folded to reduce footprint and seamlessly integrated world to chip fluid connections such as Luer couplings (Fig. 1). This new capability should allow the imagination of the designer to roam with many new innovative and useful design strategies expected compared to the usual 2D layer-bylayer thinking tied to traditional lithography.
Active, paper-based, digital microfluidics Paper-based microdevices have recently seen significant interest due the ubiquity of their base material–pressed cellulose fibers, their ease of processing (via straightforward writing and printing), and their low-cost, portability, and disposability.6 It is believed that these characteristics will facilitate their use in resource-limited settings by providing end-users with inexpensive and straight-forward devices to operate. Conventional paper-based microfluidic chips allow controlled reactions between reagent and sample streams by generating regions of hydrophilicity and hydrophobicity via a number of approaches (wax patterning, flexographic printing of styrene, or chemical modification of cellulose). Transport can then naturally occur via capillary and wetting forces in the hydrophilic regions. A large number of components have been developed in conjunction with these devices, including
Fig. 1 Photographs of (a) a coil-shaped microchannel device and (b) the interfacing of tubing with integrated Luer connectors. Micrographs demonstrating (c) how void areas can be removed from the footprint of SL devices, (d) how closed chambers can be optically accessed from precisely-defined sidewalls of a device, (e) an inlet distributor with a folded geometry, and (f) a Luer inlet oriented parallel to the plane of the chip. Adapted from Folch et al.
1492 | Lab Chip, 2014, 14, 1491–1495
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 27 March 2014. Downloaded on 22/10/2014 19:23:15.
Lab on a Chip
fluidic modulators (switches, valves, timers, and meters), paper batteries, and displays.6 These have been used in the generation of a large variety of paper-based bio-assays and chemical sensors. Even with these respective technological advances, paper chips still possess a number of limitations, including the required pre-definition of fluidic pathways, and the pre-implantation of chemicals and probes. In an attempt to leverage the cost and other advantages of paper chips in combination with the active control enabled by digital microfluidics,7 Shin and colleagues used inkjet printing to pattern electrodes onto cellulose and allow electric field manipulation on paper-based chips.8 Such patterning of electrode materials onto paper is similar to previous approaches used in mass-producing paper-based glucose test strips. However, the authors subsequently used this capability to perform dynamic electro-wetting manipulation of droplets via an external power-source, which they term as an active paper open chip (APOC) (Fig. 2). This integration allows the addition of active transportation to traditionally passive paper-based devices. Electro-wetting operates when the application of an electric field to a fluid changes its surface tension via the translocation of charge. By controlling the locality of this phenomenon via electrodes, droplets can be controllably steered. In this paper, the authors printed these electrodes by formulating a conductive carbon nanotube (1 μm length) ink, to be used with an inkjet printer, which they subsequently printed onto photo paper. These electrodes were passivated by deposition of a thin film layer of parylene-C (1 μm), and through the spinning of Teflon AF polymer (200 nm),
Highlight
yielding a smooth, hydrophobic surface. Lastly, a thin film layer of silicone oil was spin-coated onto the paper chip to lubricate droplet motion and adjust the surface tension. Droplets could subsequently be manipulated via application of field in an open air configuration. A number of droplet manipulations were demonstrated, including the continuous manipulation of droplets in a circle of electrodes, and the merging and splitting of droplets on electrode rails. These were on electrodes with a pitch of 1.5 to 2 mm, and spacings of around 0.5 mm. Due to the limitation of inkjet printing resolution, the authors attempted a number of configurations and studied their effect on droplet motion, including varying the excitation voltage, whether it was AC or DC, and the surface of the substrate (Teflon, or with oil). They found an AC voltage of 70 Vrms was required to optimally overcome the kinetics of the system, and with this they could manipulate droplets of above 300 μL. In a final demonstration, the authors created a prototype of a functional digital paper chip, by introducing intersecting lanes carrying dye. Droplets of different colored dyes could be merged and subsequently mixed by controlled voltages, undergo multiple steps of merging and manipulation, and finally be analyzed at a final location. In summary, Shin and colleagues demonstrated a potential method of activating paper-based microchips for digital microfluidic applications via the deposition of carbon nanotube inks, which were capable of precisely controlling the trajectories of droplets along the chip surface. It remains an open question of how transitioning to such a lower cost base material will lead to new digital microfluidic applications.
Fig. 2 a) Design of active, paper-based microfluidic chips. b) and c) Printed paper-electrodes and their manipulation of droplets. d) Compatibility of the technique with existing sheets of photo paper. Adapted from Shin et al.
This journal is © The Royal Society of Chemistry 2014
Lab Chip, 2014, 14, 1491–1495 | 1493
View Article Online
Highlight
Published on 27 March 2014. Downloaded on 22/10/2014 19:23:15.
Printed droplet networks As discussed above, a variety of fabrication methods are being explored to achieve enclosed or open microfluidic manipulation of fluids at various scales, and Villar et al. recently presented an innovative fabrication method that may open up a new genre of fluid manipulation designs.9 The human body is a complex network of micro-/macro-scale fluidic systems in which the individual components are cells that, when connected, can communicate fluidically through membrane junctions and protein pores. Microfabrication approaches have enabled controlled placement of compartments separated by bilayer membranes, however, such systems have not yet realized large assemblies and complex geometries that could be of use for complex assays or mimicking physiological connections.10,11 Villar et al. addressed this challenge by inventing a software-controlled printing system that creates lipid-coated aqueous droplet networks9 (Fig. 3). Their approach was to build up layers of homo- or heterologous aqueous droplets
Lab on a Chip
by ejecting droplets into a bath of lipid-containing oil. The position of the oil bath was controlled by a software-driven, motorized micromanipulator such that a horizontal crosssection of homogeneous/heterogeneous aqueous droplets from two separate droplet ejectors was layered over previously built layers in an automated manner. Their home-built printing system was capable of printing at a rate of ~1 droplet per second, and the resulting printed network consisted of aqueous compartments within single lipid bilayers, which were stable for weeks. With a stable, self-supportive network of droplets, Villar et al. first tested its capability of realizing electrical connections through the network. They used staphylococcal α-hemolysin (αHL) to define an electrical pathway within the droplet network. First they printed a pathway consisting of αHL-containing droplets (αHL+) within droplets without αHL (αHL−). Large drops of αHL pore-containing buffer were used to make the electrical contact between the entire droplet network and an electrode pair (Ag/AgCl electrode). When the circuit was connected through an αHL+ pathway at 50 mV in
Fig. 3 (A) Illustration of a printed droplet network. (B) Three orthogonal views of a single network printed according to a design with a scale bar of 1 mm. (C) Micrographs of a 3-D droplet network printed in bulk aqueous solution with a scale bar of 400 μm. (D) Schematic of a printed network with an αHL+ pathway. Green indicates drops with αHL. (E) Image of a printed network–electrode pair connection through an αHL+ pathway. Scale bar of 500 μm. (F) Amperogram showing the stepwise current increase from the configuration shown in (E) at 50 mV. (G) Images of a flower-shaped droplet network folding into a hollow sphere structure by osmosis. Scale bar of 200 μm. (H) Folding simulation of the processes illustrated in (G). Images from ref. 9 with permission.
1494 | Lab Chip, 2014, 14, 1491–1495
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 27 March 2014. Downloaded on 22/10/2014 19:23:15.
Lab on a Chip
1 M KCl, a stepwise ionic current increase was measured. When the circuit was connected through an αHL− pathway, on the other hand, only a capacitive transient ionic current was measured. This confirmed that the observed stepwise ionic current increase was due to αHL pores interfacing between the electrode buffer drop and printed droplet network. They also performed a computational simulation for the electrical connection through αHL+/αHL− droplet pathways and the calculated current from this model matched with the experimentally obtained current increase, confirming the desired electrical connection through the printed droplet network. Lastly, Villar et al. tested whether their printed droplet network, with defined electrical connections through macroscopic assembly of droplets, could allow fluid flow. They verified that water permeates through the bilayers at the interface of droplets, and thus confirmed that an osmolarity gradient among the droplets could be used to program a macroscopic shape change of the entire droplet network. They next built a model to predict the final, folded geometry from the initial geometry of the printed network. With the model, they designed droplet networks folding as desired, and successfully achieved various network geometries from circular (simple and easily printable) to hollow spheres (difficult to directly print). Their system successfully enables designs of 3-D geometries of soft materials that contain microcompartment structures of aqueous fluid bound by lipid bilayers. The bilayer is the means by which the macroscopic droplet network communicates to other droplets via membrane proteins or osmotic flow of water. More sophisticated individual/collective signal (fluid, chemicals, and others) transduction mechanisms may be achieved by utilizing additional membrane proteins or stimulus-responsive molecules, to replicate
This journal is © The Royal Society of Chemistry 2014
Highlight
physiological fluidic networks, or perform complex assays. Further development in "fixing" such soft systems once configured will be important to transition to robust devices that can be widely used in assays, suggesting that the first uses will be in scientific investigations of membrane networks.
References 1 E. Sollier, C. Murray, P. Maoddi and D. Di Carlo, Lab Chip, 2011, 11, 3752–3765. 2 A. K. Au, W. Lee and A. Folch, Lab Chip, 2014, 14, 1294–1301. 3 P. H. King, G. Jones, H. Morgan, M. R. R. de Planque and K.-P. Zauner, Lab Chip, 2014, 14, 722–729. 4 A. I. Shallan, P. Smejkal, M. Corban, R. M. Guijt and M. C. Breadmore, Anal. Chem., 2014, 86, 3124–3130. 5 M. Hatzenbichler, M. Geppert, R. Seemann and J. Stampfl, Proc. SPIE, 2013, 8618, 86180A. 6 A. K. Yetisen, M. S. Akram and C. R. Lowe, Lab Chip, 2013, 13, 2210. 7 K. Choi, A. H. C. Ng, R. Fobel and A. R. Wheeler, Annu. Rev. Anal. Chem., 2012, 5, 413–440. 8 H. Ko, J. Lee, Y. Kim, B. Lee, C.-H. Jung, J.-H. Choi, O.-S. Kwon and K. Shin, Adv. Mater., 2014, DOI: 10.1002/ adma.201305014. 9 G. Villar, A. D. Graham and H. Bayley, Science, 2013, 340, 48–52. 10 C. E. Stanley, K. S. Elvira, X. Z. Niu, A. D. Gee, O. Ces, J. B. Edel and A. J. deMello, Chem. Commun., 2010, 46, 1620–1622. 11 S. Thutupalli, S. Herminghaus and R. Seemann, Soft Matter, 2011, 7, 1312.
Lab Chip, 2014, 14, 1491–1495 | 1495
