Improving the Resolution of 3D-Printed Molds for Microfluidics by

Jan 23, 2017 - Breaking through technical barriers and cost reduction are critical issues for the development of microfluidic devices, and both rely g...
0 downloads 8 Views 3MB Size
Subscriber access provided by UNIV OF CALIFORNIA SAN DIEGO LIBRARIES

Technical Note

Improving the resolution of 3D-Printed molds for microfluidics by iterative casting-shrinkage cycles Miao Sun, Yanbo Xie, Jihong Zhu, Jun Li, and Jan Cornelis Titus Eijkel Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.6b05148 • Publication Date (Web): 23 Jan 2017 Downloaded from http://pubs.acs.org on January 28, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 7

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Improving the resolution of 3D-Printed molds for microfluidics by iterative casting-shrinkage cycles Miao Sun†, Yanbo Xie†*, Jihong Zhu‡, Jun Li†, Jan C.T. Eijkel§ †

MOE Key Laboratory of Space Applied Physics and Chemistry, School of Science, Northwestern Polytechnical University, China ‡

Engineering Simulation and Aerospace Computing, School of Mechanical Engineering, Northwestern Polytechnical University, Xi'an 710072, China § BIOS - Lab on a Chip group, MESA+ Institute for Nanotechnology and MIRA Institute for Biomedical Technology and Technical Medicine, University of Twente, the Netherlands ABSTRACT: Breaking through technical barriers and cost reduction are critical issues for the development of microfluidic devices, and both rely greatly on the innovation of fabrication techniques and use of new materials. The application of 3D printing definitely accelerated the prototyping of microfluidic chips by its versatility and functionality. However, the resolution of existing 3D printing techniques is still far below that of lithography, which makes it difficult to work on the scale of single cells and near impossible for single molecule work. In this paper, we present a facile way to increase the resolution of 3D printed microstructures to minimally 4µm by casting-shrinkage cycles of a polyurethane (PU) polymer. A water-PU liquid mixture poured on a 3D printed template quickly solidifies replicating the structures, which then isometrically shrink to half its size after solvent evaporation, downscaling the replicated structures. By repeating the casting-shrinkage cycles, we could downscale the (sub)millimeter structures of 3D printed structures on demand, until the working limit posed by the polymer properties, which we demonstrate by fabricating a micromixer. Moreover, we can even fabricate microfluidic chips from millimeter-scale manually assembled templates, fully independent of any micromachining facilities, significantly reducing the technical barriers and costs, thus opening up the microfluidics field to low-resource areas.

INTRODUCTION Lab on a Chip has become one of the most important technologies for the next generation research platforms in chemistry, biotechnology, fluids dynamics and many other areas.1 The emerging research fields with exciting possibilities stimulate the innovation of the chip and fabrication techniques, while new materials and chip fabrication methods also boost the development of the Lab on a chip technology now tracing back decades.2,3 On the basis of its easy manufacturing, high resolution obtainable, chemical resistance, excellent optical properties and low costs, PDMS (polydimethylsiloxane) casting from a manufactured master template is still the most popular materials for microfluidic chips.4-6 The most widely used process to fabricate the master for PDMS chips is photolithography of SU8 photoresist.7-9 However, emerging technologies for the template manufacturing like the shrinky-dink10,11 PM (plastic master) casting12 and 3D printing13-16 approaches, tremendously decrease the technical barrier of micromachining and costs when working with PDMS chips. Both methods however still have their own limitations. For instance, though using shrinky-dinks is an extremely easy method to fabricate the master for PDMS chips, thermo-responsive polymers (like polystyrene and polyolefins) only shrink once and aren’t able to downscale the crosssectional area of microstructures, since the materials printed/sputtered on (eg. ink) do not shrink as the thermoresponsive substrate.10,11,17 PM casting can precisely copy a

microstructure and create a new master thus greatly decreasing the costs of fabricating PDMS chip templates, but still relies on the photolithography fabrication process for an initial template. The emerging technology of 3D printing, is considered as the booster of lab on a chip technology in its third decade.13,15,16,18-20 However, the resolution of current 3D printers still is above fifty micrometers,21,22 which makes them unsuitable for the research of single cell or single molecules, even when not considering the chemical instability of some printed materials.23 In this paper, we describe a method to optimize the use of 3D printed microfluidic chip templates, by greatly increasing the resolution to 4µm.24 Repeating a shrink process for many cycles enables to downscale a millimeter-scale structures to the microscale. Finally, we demonstrate the fabrication of a microfluidic mixer from an originally 3D printed master. Furthermore, we present a chip that is fabricated by manually assembling a master without using any specific micromachining facilities, thus greatly decreasing the technical barriers and costs of collecting microfluidic fabrication-related facilities in the laboratory. We use polyurethane polymer (commercially available as Smooth-On “hydroshrinkTM”) as the material of a soft lithography template. Adding polyurethane to water (1:3 by weight) and pouring the mixture on a 3D printed template, the mixed liquid turns into solid within 1-2 min replicating the structure

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 7

Figure 1. (a) The schematic picture of the fabrication process “shrink cycles”; (b) The setup for casting the liquid PU-water mixtures; (c) the fresh PU polymer with enlarged view; (d) the cured PU polymer with shrunk micro-structures; of the template. As the solvents are extracted from the solids, the structure gradually shrinks to “Shrinky-dink” process that relies on the shrinkage of the substrate and not of the materials (eg. ink) that are patterned on it. The shrinky-dink process causes the channel cross-sectional area to remain constant when ink is printed on top of the substrate or to become even larger in the case a channel is engraved.10 In contrast, our method can isometrically shrink the micro-patterns, enabling the renewed use of the shrunken structure as a new template in a “casting-shrinkage cycle” (Figure 1a), downscaling the structures for two or even many times on demand. By using these shrink cycles and a 3D printed template, our method greatly helps to increase the resolution of microstructures while still keeping a low cost level.

EXPERIMENTS The work flow of fabricating the PU master and PDMS chips is shown in the figure 1a. The original master was printed or manually assembled at (sub)millimeter scale (Figure 1a, ①). Subsequently the mixed solution of polyurethane and water was casted (Figure 1a, ②). The liquids rapidly become solid within 1-2 minutes after adding water to the PU, creating the (sub)millimeter-scale structures. To prevent gas bubbles (formed during the mixing of PU and water) sticking on the structures, we built a vacuum chamber to extract the gas and let the liquids fully fill the embedded master (setup can be seen in Figure 1b). After the liquids solidified, we took the cured PU out (Figure 1a, ③) and placed it in the oven (60 °C) to speed up the evaporation process. After all the volatile solvents had evaporated from the solids (4-6 hrs. depending on the thickness of PU), the substrate color turned from white (fresh PU, Figure 1c) to clear yellow (cured PU, Figure 1d) while the size in three dimensions had shrunken to about half the original size, including the replicated structures (Figure 1a, ④). By pouring liquid PDMS on this new shrunken master (Figure 1a, ⑤) and hardening in the oven, PDMS chips with half the size of the original structures can be fabricated (Figure 1a, ⑥ ). Now there are two options: either the fabricated PDMS chip are bonded to achieve downscaled PDMS microfluidic chips (Figure 1a, ⑦), or the PDMS chips are used as a new PU master for the next shrink cycle (Figure 1a, ⑧), to fabricate the next generation of microfluidic chips now a quarter of the original structure size. Shrink cycles can be repeated

until the structures are downscaled to the target value. Another short path is available to achieve the cycles (④→②, green dashed line), where the fresh PU can replicate the (hydrophobized) cured PU and skip the process of PDMS chip casting, however creating opposite structures to the last shrink cycle. It needs mentioning that both the original (3D printed) master and PU templates can be reused without damaging any of the structures after peeling off the PDMS. We found that the shrinking ratio was only determined by the mass ratio of water and PU, while physical conditions like pressure, humidity and temperatures play an important role in the shrinkage speed. For example, the shrinking process can last 14 days at room temperature in the open air. Higher temperature, vacuum environment and low humidity definitely increase the speed of volatile solvent evaporation from the solid, which is the reason of placing fresh PU in the oven in our experiments. However, a too fast evaporation process can damage fine structures, for instance evaporating at 70°C in our experiments proved likely to fail prototyping the microstructures.

RESULTS AND DISCUSSION 3D printed microfluidic chip As already mentioned in the introduction, the emerging 3D printing technology enables rapid prototyping of new designs for liquid handling devices and device integration that can usher in a new era for the “chip”.14 However, the resolution of current low cost printers is still far from the working limit of soft lithography. Taking advantage of the shrink cycles, we downscaled microstructures printed by a 3D printer to increase the resolution of microfluidic channels. We 3D printed a typical microfluidic mixer with a channel width of 609µm and depth of 439µm as an original master (Figure 2a, fabricated by stereolithography, ZRapid Tech SLA600 with ZR580 resin), a microfluidic mixer structure widely used for example in medicine production.11,25 After three shrink cycles, the microfluidic chip was shrunk from wafer-size to dime-size, with a channel width of 68.3µm and depth of 43.6µm for the 3rd generation (Figure 2b, c). After bonding to a glass slide (use 2nd generation chip as a demo), we injected red and blue ink for demonstration of liquid mixing.26 Solution streams with different mixing ratios of red and blue ink as judged from the color

2

ACS Paragon Plus Environment

Page 3 of 7

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

were collected at the 5 outlets of the microfluidic chip (Figure 2d). Facility-independent microfluidic chip fabrication

By using more than one shrink cycle, we expected to be able to fabricate microfluidic chips from the templates with millimeter-scale structures constructed with extremely simple means. As

Figure 2. (a) 3D printed mixer template and three generations PDMS chips before bonding; (b) close-up of the microfluidic channels (filled by blue ink) in PDMS chips under the microscope; (c) cross-sectional view of the microfluidic channel in the three generations, with a scale bar of 100µm; (d) red and blue inks injected from two inlets of the microfluidic mixer, with a gradual mixture ratio in the five outlets. an example, we manually assembled plastic sticks (about 1mm square cross-section and 4-6cm long) to form an H-filter on a flat glass plate as an original master (Figure 3a). The H-filter is a basic microfluidic structure used for cell separations and

Figure 3. (a) original H-type and flow-focusing (d) structures with plastic sticks as millimeter-scale master; (b) parallel flow generated in the chip; (c) microscopic snapshot showing parallel flow on the microscale; (e) droplets generated in a 2nd generation PDMS chip; (f) and (g) droplets generated at different flow-rate ratios (water to oil are 3:5 in Fig 4f and 1:27 in Fig 4g) in a 3rd generation PDMS chip. Table 1. The dimensions of channels by manually assembled templates

Original

1st generation

2nd generation

3rd generation

Width (µm)

887

480

242

124

Height (µm)

911

511

256

122

chemical synthesis.27-32 After three shrink cycles, the structure was reproduced on the corresponding generation of PDMS, with dimensions shown in table 1. We used the 2nd and 3rd generation PDMS chips with square channels of respectively 250 and 120 µm width and height to perform fluidic tests after bonding on a glass slide. With the diluted blue and red inks injected by 50mbar pressure (operated by a Fluigent MFCS pump) into the two inlets of the microfluidic channel, as can be seen (Figure 3b, c), parallel flow was successfully generated in the H-filter microfluidic chip. The microdroplet platform is now one of the most important platforms in microfluidics with applications like DNA sequencing33, Point of Care34, digital PCR35, chemical synthesis36,37 and even display38. Similar as described above, we assembled plastic sticks into a flow focusing structure as original template of a droplet generator, with two phase inlets (water and oil) as shown in the diagram (Figure 3d). After two or three shrink cycles, a PDMS microfluidic droplet generator with square microchannels with respectively height and width of 250µm (Figure 3e) and 123µm (Figure 3f, g) were fabricated. After bonding to a glass slide, we treated the surface by FDTS (Perfluorodecyltrichlorosilane), to create a hydrophobic monolayer on the glass slide for water droplet generation dispersed in the oil phase. An aqueous red ink solution was used as the dispersed phase and mineral oil (transparent) as the continuous phase with addition of surfactant SPAN80 (5wt%) in the oil phase to aid emulsification. With applied pressure at both inlets, red ink droplets were generated with excellent mono-dispersity (Figure 3f and g), and the droplet size can be adjusted by the ratio of flow rate in two phases. The quality of the microfluidic chip will depend on the initial master, since the PU can precisely replicate the structures including the

3

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

flaws existing in the master. It is important to notice that by the method of assembling plastic sticks on the millimeter scale width/depth and shrinking to the microscale, microfluidic structures can be replicated simply with an oven, greatly de-

Page 4 of 7

creasing the cost of PDMS chip fabrication and opening up the field of microfluidic chip manufacturing to low-resource countries.

Figure 4. SEM images of the templates and PDMS chips. A PDMS micropillar array fabricated by standard SU8 lithography was used as the original template (a) The first generation of cured PU (b) and PDMS (c) copies with microstructures downscaled to half the size. The second generation of PU master (d) and PDMS (e) further downscaling the microstructures to one fourth the original size. The scale bar of figure (a)-(e) is 100µm; (f) Minimum PU microstrips with 4µm width.

Discussion To characterize the working limits of the method, we used an array of square PDMS micropillars fabricated from a SU8 photoresist master for the PU shrink replication cycles. Characterization was done by a stylus profiler (DekTak XT) and Scanning Electronic Microscope. The original dimensions of the square pillars are 64µm in width and 51µm in height. After PU casting and one shrink cycle, a PU master was obtained that had an array of micro pits 37.8µm in width and 26.5µm in depth (see table 2). According to the number of shrink cycles we termed this the 1st generation PU master. In a PDMS chip from the 1st generation PU master, the dimension of pillar structures was 36.7µm in width and 25.0µm in height, with an average shrinkage of 52.3% from the original to the 1st generation PDMS. After a second shrink cycle, the micro pits became 21.2µm in width, 7.60µm in distance and 14.1µm in height, showing an average 52.4% shrinkage compared to the 1st generation and 27.4% compared to the original PDMS structures. The distance between the micro pillars presented the workable lower limit of this method as the distance of 7.60µm in the second generation is close to the general working limit for structures on PDMS chips, as structures begin to Table 2. The dimensions of the different generation structures. Structures

Original

1st PU

1st PDMS

2nd PU

2nd PDMS

Pillar (r)(µm)

65.8

37.8

36.7

21.2

19.5

Distance (d) (µm)

33.5

16.7

17.6

7.60

10.0

Height (h) (µm)

51.3

26.5

25.0

14.1

12.4

Pillar

0.574

0.558

0.322

0.296

Distance

0.499

0.525

0.227

0.299

height

0.516

0.487

0.275

0.241

Shrink factor

deform and stick, presumably by capillary forces at this height to width ratio. However, when we decrease the height to width ratio, it proved possible to fabricate 4µm wide microstrips (Figure 4f). Further downscaling was impossible since the sidewalls of the structures easily stuck together. In addition, as can be seen from the enlarged view of the 2nd generation PU master (Figure 4f), wrinkles appear on the PU surface possibly caused by the crystallization of the polymers.17,39,40 The wrinkles are around 2µm in width and 300nm in height, and will possibly constrain the fabrication of nanostructures. As we have discussed above, there are some more benefits of using the casting-shrinkage cycles to fabricate microfluidic chips: Lower equipment needs - compared to the soft lithography by SU8 templates, it only needs an oven which greatly decreases the costs of facilities in the lab; Low costs - the costs of chemicals (13 dollars per pound for more than 20 casting shrinkage cycles) made the chip fabrication much cheaper than using photolithography, without even considering the absence of the need for a cleanroom environment; Time saving – since the time needed includes time for the fabrication of photomasks and SU8 templates, using 3D printed templates saves the time of collecting the PDMS templates (6-8hrs for each generation including PDMS and PU templates); also, simple structures can be manually prepared. The method has some limitations as well. Care must be taken to make the entire initial mold to scale, as all dimensions equally shrink. The high viscosity and short phase transition time of mixed liquids of water and PU make it difficult to print. Besides, like most other soft lithography, the casting-shrinkage cycles cannot be used for downsizing 3D structures since it is impossible to peel them off from the templates.

4

ACS Paragon Plus Environment

Page 5 of 7

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

CONCLUSION In conclusion, we describe a method to greatly increase the resolution of 3D printed microfluidic templates to a few µm, by using casting-shrinkage cycles of a mixture of polyurethane polymer and water. After three cycles, 3D printed (sub)millimeter scale structures can be downscaled to the microscale less than 100µm. In this paper we also successfully demonstrated microfluidic structure manufacturing from manually assembled millimeter-sized plastic sticks as the original masters to form H-filter and flow focusing structures, consequently fabricating microfluidic chips for parallel flow and droplet generation, without using any specific micromachining facilities. We demonstrated this technique is a strong complement to the flexible 3D printing technology for microfluidics by enabling downsizing, and furthermore the manufacturing of microscale structures without 3D printing from manually assembled millimeter-scale structures.

AUTHOR INFORMATION Corresponding Author *E-mail: [email protected]

Author Contributions M.S. and Y.X. performed experiments and analyzed results. M.S., Y.X. J.E wrote the manuscript; Jh.Z. provided the manufacturing of 3D printed templates; J.L. gave valuable suggestions.

Notes The authors declare no competing financial interest.

ACKNOWLEDGMENT Financial support from NSFC grant (NO. 21602176) and Fundamental Research Funds for the Central Universities No. 3102015ZY059) are gratefully acknowledged.

REFERENCES (1) Escandón, J.; Santiago, F.; Bautista, O.; Méndez, F. International Journal of Thermal Sciences 2014, 86, 246-257. (2) Castillo-León, J. Microfluidics and Lab-on-a-Chip Devices: History and Challenges, 2014, p 1-15. (3) Brittain, S.; Paul, K.; Zhao, X. M.; Whitesides, G. Physics World 1998, 11, 31-36. (4) Martinez, A. W.; Phillips, S. T.; Wiley, B. J.; Gupta, M.; Whitesides, G. M. Lab on A Chip 2009, 8, 2146-2150. (5) Bruzewicz, D. A.; Reches, M.; Whitesides, G. M.; Chem., A. Analytical Chemistry 2008, 80, 3387-3392. (6) Ren, K.; Zhou, J.; Wu, H. Accounts of Chemical Research 2013, 46, 2396-2406. (7) Jo, B. H.; Lerberghe, L. M. V.; Motsegood, K. M.; Beebe, D. J. Journal of Microelectromechanical Systems 2000, 9, 76-81. (8) Xia, Y.; Whitesides, G. M. Angewandte Chemie International Edition 1998, 37, 550-575. (9) Xia, Y.; Mcclelland, J. J.; Gupta, R.; Qin, D.; Zhao, X. M.; Sohn, L. L.; Celotta, R. J.; Whitesides, G. M. Advanced Materials 1997, 9, 147-149. (10) Chen, C. S.; Breslauer, D. N.; Luna, J. I.; Grimes, A.; Chin, W. C.; Lee, L. P.; Khine, M. Lab on A Chip 2008, 8, 622-624. (11) Grimes, A.; Breslauer, D. N.; Long, M.; Pegan, J.; Lee, L. P.; Khine, M. Lab on A Chip 2008, 8, 170-172. (12) Desai, S. P.; Freeman, D. M.; Voldman, J. Lab on A Chip 2009, 9, 1631-1637.

(13) Au, A. K.; Bhattacharjee, N.; Horowitz, L. F.; Chang, T. C.; Folch, A. Lab on A Chip 2015, 15, 1934-1941. (14) Au, A. K.; Huynh, W.; Horowitz, L. F.; Folch, A. Angewandte Chemie International Edition 2016, 55, 3862-3881. (15) Erkal, J. L.; Selimovic, A.; Gross, B. C.; Lockwood, S. Y.; Walton, E. L.; Mcnamara, S.; Martin, R. S.; Spence, D. M. Lab on A Chip 2014, 14, 2023-2032. (16) Ho, C. M.; Ng, S. H.; Li, K. H.; Yoon, Y. J. Lab on A Chip 2015, 15, 3627-3637. (17) Nguyen, D.; Taylor, D.; Qian, K.; Norouzi, N.; Rasmussen, J.; Botzet, S.; Lehmann, M.; Halverson, K.; Khine, M. Lab on A Chip 2010, 10, 1623-1626. (18) Symes, M. D.; Kitson, P. J.; Yan, J.; Richmond, C. J.; Cooper, G. J.; Bowman, R. W.; Vilbrandt, T.; Cronin, L. Nature Chemistry 2012, 4, 349-354. (19) Comina, G.; Suska, A.; Filippini, D. Lab on A Chip 2014, 14, 2978-2982. (20) Lee, A. Lab on A Chip 2013, 13, 1660-1661. (21) Comina, G.; Suska, A.; Filippini, D. Lab on A Chip 2014, 14, Accepted Manuscript. (22) Bhattacharjee, N.; Urrios, A.; Kang, S.; Folch, A. Lab on A Chip 2016, 16, 1720-1742. (23) Capel, A. J.; Edmondson, S.; Christie, S. D.; Goodridge, R. D.; Bibb, R. J.; Thurstans, M. Lab on A Chip 2013, 13, 45834590. (24) Qin, D.; Xia, Y.; Whitesides, G. M. Nat. Protocols 2010, 5, 491-502. (25) Escobedo, C.; Chou, Y.-W.; Rahman, M.; Duan, X.; Gordon, R.; Sinton, D.; Brolo, A. G.; Ferreira, J. Analyst 2013, 138, 14501458. (26) Liu, A.-l.; He, F.-y.; Wang, K.; Zhou, T.; Lu, Y.; Xia, X.-h. Lab on a Chip 2005, 5, 974-978. (27) Xia, N.; Hunt, T. P.; Mayers, B. T.; Alsberg, E.; Whitesides, G. M.; Westervelt, R. M.; Ingber, D. E. Biomedical Microdevices 2006, 8, 299. (28) Atencia, J.; Beebe, D. J. Nature 2005, 437, 648-655. (29) Squires, T. M.; Quake, S. R. Reviews of Modern Physics 2005, 77, 977-1026. (30) Zhang, L.; Feng, Q.; Wang, J.; Sun, J.; Shi, X.; Jiang, X. Angewandte Chemie International Edition 2015, 54, 3952-3956. (31) Zhang, L.; Feng, Q.; Wang, J.; Zhang, S.; Ding, B.; Wei, Y.; Dong, M.; Ryu, J. Y.; Yoon, T. Y.; Shi, X. Acs Nano 2015, 9, 9912. (32) Yu, Y.; Wen, H.; Ma, J.; Lykkemark, S.; Xu, H.; Qin, J. Advanced Materials 2015, 26, 2494-2499. (33) Liu, P.; Li, X.; Greenspoon, S. A.; Scherer, J. R.; Mathies, R. A. Lab on A Chip 2011, 11, 1041-1048. (34) Wen, H.; Yu, Y.; Zhu, G.; Jiang, L.; Qin, J. Lab on A Chip 2015, 15, 1905-1911. (35) Men, Y.; Fu, Y.; Chen, Z.; Sims, P. A.; Greenleaf, W. J.; Huang, Y.; Chem., A. Analytical Chemistry 2012, 84, 4262-4266. (36) Teh, S.-Y.; Lin, R.; Hung, L.-H.; Lee, A. P. Lab on a Chip 2008, 8, 198-220. (37) Sui, X.; Shui, L.; Cui, J.; Xie, Y.; Song, J.; Van, d. B. A.; Hempenius, M. A.; Vancso, G. J. Chemical Communications 2014, 50, 3058-3060. (38) Tao, H.; Jin, M.; Eijkel, J. C. T.; Zhou, G.; Shui, L. Biomicrofluidics 2016, 10, 011908. (39) Fu, C.-C.; Grimes, A.; Long, M.; Ferri, C. G. L.; Rich, B. D.; Ghosh, S.; Ghosh, S.; Lee, L. P.; Gopinathan, A.; Khine, M. Advanced Materials 2009, 21, 4472-4476. (40) Yoo, P. J.; Suh, K. Y.; Park, S. Y.; Lee, H. H. Advanced materials 2002, 14, 1383-1387.

5

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 7

6

ACS Paragon Plus Environment

Page 7 of 7

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Figure 1. (a) The schematic picture of the fabrication process “shrink cycles”; (b) The setup for casting the liquid PU-water mixtures; (c) the fresh PU polymer with enlarged view; (d) the cured PU polymer with shrunk micro-structures. 169x54mm (300 x 300 DPI)

ACS Paragon Plus Environment