Lab on a Chip TECHNICAL INNOVATION Ultrasonic bonding method for heterogeneous microstructures using self-balancing jig† Cite this: Lab Chip, 2015, 15, 1412

Kyoung G. Lee,‡a Sujeong Shin,‡a Byeong Il Kim,b Nam Ho Bae,a Moon-Keun Lee,a Seok Jae Lee*a and Tae Jae Lee*a Perfect sealing of heterogeneous microstructures in plastic-based microfluidic devices is a significant and urgent challenge to be able to apply them in various microfluidic-based applications, including biosensing, biofiltering, chemical reactors and lab-on-a-chip. In this study we report a simple but practical and effective method to bond a microstructure-incorporated microfluidic device using an ultrasonic bonding method. The specially designed hemisphere-shaped jig, which is called a self-balancing jig, provides a free motion in all x, y, and z directions. These unique properties of the jig allow us to precisely adjust and bond the heterogeneous microstructures in the device. A conventional jig shows in solution leakages around the Received 16th December 2014, Accepted 22nd January 2015 DOI: 10.1039/c4lc01473a

heterogeneous microstructures while the self-balancing jig did not show any leakages in devices. Furthermore, the bonding performance was also confirmed by using a black ink and fluorescent dye solution. The micro-pillar arrays in the device also demonstrated its capability for selective filtering of microbeads. We believe that this technique would be a useful tool for producing microfluidic devices with heterogeneous

www.rsc.org/loc

microstructures.

1. Introduction Polymer-based microfluidic devices have brought significant attractions to research and industries due to their commercial advantages, including high-speed production, disposability, rapid processing, reliability, and high transparency.1–4 Recently, the polymer-based microfluidic devices face new challenges in terms of bonding microchannels and/or microstructures without any liquid leakage and deformation of structures.5–7 Typically, a microfluidic device should be strong enough to withstand high pressure, maintain transparency, not undergo any structural deformation, not have a gap at the interface between microstructures and substrate, and have liquid tightness. So far, many techniques have been proposed in the past such as thermal bonding and adhesive bonding.8,9 The low heat transfer coefficient of polymer makes it difficult to precisely control the glass transition temperature and this can lead to failure in bonding or damages to the devices. a

Department of Nano Bio Research, National Nanofab Center, 291 Daehak-ro, Yuseong-gu, Daejeon 305-806, Republic of Korea. E-mail: [email protected], [email protected] b T&S R&D Center, 13-27, 1gongdan-ro 6-gil, Gumi-si, Gyeongsangbuk-do, 730-906, Republic of Korea † Electronic supplementary information (ESI) available: Schematic drawing, design, and SEM images of the devices, photographic images and movie of the self-balancing jig are presented. See DOI: 10.1039/c4lc01473a ‡ These authors contributed equally to this work.

1412 | Lab Chip, 2015, 15, 1412–1416

Furthermore, an adhesive-assisted method gives better bonding performance but the potential solubility of the adhesive in organic solvents brings occasional detachment of the devices.10 To overcome such challenges, ultrasonic bonding has been considered as an alternative and effective way to seal a thermoplastic device without any adhesives.11,12 This method allows immediate melting and rapid cooling in milliseconds of specific plastic regions by applying ultrasonic energy. Because of such unique characteristics, numerous pioneers have successfully welded microfluidic channels and connectors from 500 μm to 1 mm so far.13,14 Although great advancement has been made in ultrasonic welding, the heterogeneous microstructure-incorporated microfluidic devices are still confronted with great technical challenges to secure microchannels and/or microstructures because of (1) random microscale thickness differences (Table S1†) due to shrinkage of polymers in the range of 0.5 to 2.5% (ref. 15 and 16) and (2) microscale imbalance of the devices. Furthermore, most of the ultrasonic bonding techniques have been used for flat and fixed metal jigs and would not be suitable for bonding complex devices. Therefore, these combinations of non-uniform microfluidic device and conventional metal jig openly bring miss-alignment and/or miss-leveling in all x, y, and z directions. Subsequently, bonding failure inevitably occurs due to unevenly distributed ultrasonic energy over the welding structures. These are major challenges in the

This journal is © The Royal Society of Chemistry 2015

Lab on a Chip

ultrasonic bonding of microfluidic devices, especially for the industrial scale bonding processes. Therefore, alternative methods still need to be developed for reducing bonding failure of the devices. Herein, we developed a self-balancing jig for ultrasonic bonding of a heterogeneous microstructure-based microfluidic device without using any external heat or adhesives. The heterogeneous micro- to macro-scaled structures incorporated microfluidic device was produced by microinjection molding of cyclo olefin copolymer (COC). This device is selected as a model to investigate the performance of the jig. The concave-shaped metal plates are composed of female and male parts. These were specially designed and manufactured for bonding of microstructures and microchannels. Additionally, fluorescent ink and microbeads were employed to confirm the performance of the newly developed jig for the fabrication of microfluidic device and its filtering performance was also investigated.

2. Experimental 2.1. Reagents and materials COC pellets (Grade-5013L, TOPAS Advanced Polymers, Germany), black ink (Waterman, France), Rhodamine 6G and polystyrene (PS) particles (7 μm in diameter) were purchased from Sigma-Aldrich and used without further purification.

Technical Innovation

ultrasonic bonding. One is a conventional jig which is purchased from Branson in Korea. The self-balancing jig was designed with a pair of male and female types of hemispherical knuckle-like structures and vacuum holes. 2.3. Ultrasonic bonding using conventional and self-balancing jig An ultrasonic bonder (Branson, 2000X-aef, USA) with a generator frequency of 20 kHz was employed to bond the microfluidic device. The device was loaded on the conventional jig and secured by a clamp. After moving the ultrasonic horn with a speed of 20 mm s−1, bonding pressure was applied to the substrates. Subsequently, an ultrasonic wave with amplitude of 90% was applied for 0.1 s to the device through the horn. After the ultrasonic energy was discharged, the bonding pressure was further applied for 10 s. Additional bonding process was performed using ultrasonic bonder and selfbalancing jig. The device was loaded on the upper jig and fixed with the clamp. Initially, the horn was slowly moved along the z-axis to balance the device, jig, and ultrasonic horn. After this, the jig was further vacuumed to secure the jig position using dry vacuum pump (Welch, model no. 2585-C-50, USA). Furthermore, the ultrasonic bonding and holding process was applied on the substrates with the same process conditions as for a conventional jig. Finally, both top and bottom microfluidic devices were perfectly bonded to each other.

2.2. Design and fabrication of microfluidic device A micro-filter device was designed and fabricated for testing ultrasonic bonding of a micro-sized energy director on a patterned substrate (60 × 60 mm2). The minimum gap size between two pillars was 2 μm, and the depth of the channel was 30 μm. All width and height of the energy director were 10 μm and it was fabricated around the device. All the detailed dimensions of the device are also shown in Fig. S1 and S2.† In order to produce the device, the Ni stamp was fabricated using a photolithography and a Ni electroplating method. In detail, the microstructures were initially constructed on a silicon wafer and then, Ni was additionally coated over the wafer through electroplating. After that, the Si wafer was completely removed and the Ni stamp was carefully cut in the desired dimensions and installed in the injection molding machine (ARBURG, Allrounder 270 C 400-100, Germany). COC pellets were placed in a dehumidifier at 70 °C to remove extra moisture. The temperature of the injection cylinder was set at 260 °C to melt and inject the COC into the mold under a speed of 150 mm s−1 and a pressure of 1500 bar. Subsequently, a packing pressure of injection molding ranging from 800 to 1500 bar was applied for 20 s to produce the patterned device. The mold was kept at 110 °C during the injection molding process. 2.3. Design and fabrication of jigs Two different types of jig assembly were designed and produced for comparison and investigate the performance of

This journal is © The Royal Society of Chemistry 2015

2.4. Fluid leakage and microbeads filtering tests Black ink was introduced to verify any fluid leakage of the bonded microfluidic devices using a conventional and a selfbalancing jig, respectively. Black ink and fluorescent dye (i.e. Rhodamine 6G) were injected into the device through the inlet and any solution leakages were investigated by both optical and fluorescent microscope (Eclipse Ti-S, Nikon, Japan). For the filtering test, PS micro particles (0.2 mL) were diluted with 1 mL of deionized water (DI water) and it was directly injected into the device. The filtered solution was collected from the outlet by using a syringe for further analysis. After filtering of the beads, 1 mL of DI water was further injected into the device for washing and it was finally dried. Both optical and SEM analysis were used to investigate the performance of the device.

3. Results and discussion Design and fabrication of microfluidic device The design and dimensions of the microfluidic device are shown in Fig. S1 and S2† and their bonding processes using conventional and self-balancing jigs are illustrated in Fig. 1. The white lines in the schematic diagram of the device indicates major functional structures in the microfluidic device for filtering microbeads, while yellow thin lines represent the welding lines for bonding the device (Fig. S2†). The plasticbased microfluidic devices were produced by combination of

Lab Chip, 2015, 15, 1412–1416 | 1413

Technical Innovation

Fig. 1 Schematic diagrams of ultrasonic bonding processes using (a) conventional jig and (b) self-balancing jig.

Ni stamp (Fig. S3a†) and microinjection molding techniques. The whole producing processes of plastic-based microfluidic chip and schematic illustration of microfluidic chip were previously reported by our group.17 The mirror image of the Ni stamp was transferred to the COC and all the structures were confirmed using scanning electron microscopy (SEM) under different magnifications as shown in Fig. S3c and S3d.† The micropillar arrays were located between an isosceles trapezoid and the major structures were properly replicated over COC surface. All of the thin welding lines also observed on the top of isosceles trapezoid structures (Fig. S3c†).

Self-balancing mechanism of jig and leakage test To investigate the influence of jigs on the devices, two devices were ultrasonically bonded using a conventional jig and a self-balancing jig at room temperature (Fig. 2). The top and bottom plates of the plastic device were placed on the flat metal holder and their position was secured using metal clamps on both the conventional and self-balancing jigs. Initially, an ultrasonic horn compressed the device then the ultrasound energies were transmitted to the top plate of the device in the form of sound. Ultrasonic induced intermolecular and interfacial friction of the device allows rapid melting of the polymeric energy director in a few seconds to bond with the bottom patterned plate.

Fig. 2 Picture of (a) conventional jig and (b) self-balancing jig loaded in ultrasonic bonder. (c) Photograph of disassembled self-balancing jig.

1414 | Lab Chip, 2015, 15, 1412–1416

Lab on a Chip

Although, all the major microstructures were formed over the device, the milli- to micro-scaled thickness and surface roughness differences inevitably occurred due to the shrinkage of polymers after injection. The known thickness differences of thermoplastics are in the range of 0.2 to 2.5% from the original dimensions. In this case, each corner of the device was measured (Table S1†) and it revealed that the average thickness difference was about 2 μm (i.e. 1.68% compared to original thickness) due to plastic shrinkage. This may lead to micro-scaled gaps between the top and bottom plate and cause solution leakage around microchannels, especially when using a conventional jig. Furthermore, serious sonochemical vibration also can result in potential misalignment and partial bonding of the devices. In contrast, the self-balancing jig has a hemispherical shape and this benefits free movement along all x, y, z axes and smooth but accurate rotation is also possible (Fig. S4 and S5†). In this case, before sealing the device, an ultrasonic horn initially pressed the jig for alignment between the device and horn to minimize the bonding failure. This would provide a better performance especially for devices with heterogeneous microstructures. Furthermore, two clamps on the jig surface formed a square groove for holding the plastic to avoid the potential device having any misalignment. Additional vacuuming of the jig helps to prevent any potential distortion of the device to the external forces. The upper jig presented relatively free movement along with ultrasonic horn direction and diminished bonding failures. Instant ultrasonic oscillating pressure generates temporary friction energy which is highly localized to the top of the welding structures. This ultrasonic energy allows rapid melting and solidifying of the welding lines in a few seconds to enhance bonding ability. The performance of a conventional jig and a self-balancing jig were compared with each other using ultrasonically bonded devices (Fig. 3). As expected, the conventionally prepared device was partially bonded and most of the black ink leaked out from the channel (Fig. 3b). Furthermore, the ink spread over the micro-scaled gaps around the microstructures. This would be a natural phenomenon since the ultrasonic energy is localized and converted into vibrational energy to lead misalignment and partial bonding of the major channels and microstructures. For these reasons, the conventional jig would not be suitable for bonding such devices. In contrast, no indication of ink leaks were found from the device using a self-balancing jig and this result confirmed better performance and perfect sealing of the microchannel as shown in Fig. 3c. For further confirmation, the device was also filled with fluorescent dye solution. Practically, the dye solution penetrates relatively easily into any micro-scale cracks or voids while no fluorescent signals would be observed around perfectly bonded areas. Initially, the micropillar arrays were zoomed around the microstructures (Fig. 3d). No significant damage of the structures were observed even after injecting dyes. After switching fluorescent

This journal is © The Royal Society of Chemistry 2015

Lab on a Chip

Technical Innovation

Fig. 4 Photographic images of (a) the device and the microbeads solution before and after filtering and (b) packed beads using microstructures in the device. SEM images of (c) microbeads around the micropillars and (d) bonded microchannels of the highlighted area from (b).

4. Conclusion Fig. 3 Photographic images of (a) microfluidic device and highlighted areas from (a) in case of bonding with (b) conventional and (c) selfbalancing jig. (d) Optical image, (e) red fluorescence image, and (f) combined both optical and fluorescence image of the self-balanced bonding devices. (Scale bars in d–f are 50 μm).

mode, the strong red fluorescent color was only observed from voids in the device while micropillar arrays appeared in black color (Fig. 3e). The overlapped optical and fluorescent images even more clearly show the boundary of the bonded areas (Fig. 3f). These results provide strong evidence that self-balancing jigs give better performance than conventional jigs. For further investigation of the bonded device performance and its potential application, we selected PS microbead suspensions of 7 μm in diameter which are similar in size to white blood cells as an example model for white cell separation. After introduction of the white suspended solution to the device, the mixture solution of microbeads slowly flowed inside of the channels and filtered through the micropillar arrays as shown in Fig. 4a. The recovered solution became clear and no particles were observed in the discharged solution (Fig. 4a). The white precipitation (i.e. PS microbeads) was observed and evenly distributed around the circular micropillar zone as shown in Fig. 4b. The beads were densely packed around the structures and the highlighted area in Fig. 4b was further analysed using SEM imaging as shown in Fig. 4c. These results supported the proper replication of heterogeneous structures from the master mold and perfect bonding of the device (Fig. 4d). This microfluidic device not only validated enhancement of sealing channels and structures but also preserved effective filtering capability of the device without any collapse of the structures.

This journal is © The Royal Society of Chemistry 2015

An ultrasonic bonding method by employing a self-balancing jig for a polymeric-based microfluidic device was developed in this paper. The concave-shaped metal plate was firstly proposed and allowed even distribution of the ultrasonic energy over the devices for rapid but effective and perfect bonding of the device. The applied ultrasonic energy led to temporary melting of the welding lines and attached to the surface of the bottom plate for further solidification due to the rapid cooling. Moreover, the COC-based device was successfully bonded and was strong enough to secure the major structures. The heterogeneous microstructures in the device exhibited great mechanical stability and filtering capability to selectively separate microbeads from the mixture solution. These results proved that the combination of ultrasonic bonding and self-balancing jig could be an excellent way to bond macro- and micro-scaled heterogeneous structures and channels successfully for commercial microfluidic devices.

Acknowledgements This work was supported by the Pioneer Research Center Program (2014M3C1A3051460) through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (MSIP) of Korea (2014M3C1A3051476); the Public Welfare & Safety research program through the NRF funded by the MSIP of Korea (NRF-2013M3A2A1073991); the ICT R&D program of MSIP/IITP (10041870, Development of Growth Management Technology for Integrating Complex Raising Building Forms of Fish and Shellfish); and the NRF grant funded by MSIP (no. 2014R1A5A2010008).

Notes and references 1 K. Liu and Z. H. Fan, Analyst, 2011, 136, 1288–1297. 2 B.-H. Kwon, K. G. Lee, T. J. Park, H. Kim, T. J. Lee, S. J. Lee and D. Y. Jeon, Small, 2012, 8, 3257–3262.

Lab Chip, 2015, 15, 1412–1416 | 1415

Technical Innovation

3 L. Gervais, N. De Rooij and E. Delamarche, Adv. Mater., 2011, 23, H151–H176. 4 E. Roy, J.–C. Galas and T. Veres, Lab Chip, 2011, 11, 3193–3196. 5 Z. Zhang, X. Wang, Y. Luo, S. He and L. Wang, Talanta, 2010, 81, 1331–1338. 6 Q. Chen, L. Zhang and G. Chen, RSC Adv., 2014, 4, 56440–56444. 7 P. N. Nge, C. I. Rogers and A. T. Woolly, Chem. Rev., 2013, 113, 2550–2583. 8 J. Liu, S. Yang, C. S. Lee and D. L. DeVoe, Electrophoresis, 2008, 29, 2241–2250. 9 C.-W. Tasa and D. L. DeVoe, Microfluid. Nanofluid., 2009, 6, 1–16. 10 Y.-J. Pan and R.-J. Yang, J. Micromech. Microeng., 2006, 16, 2666–2672.

1416 | Lab Chip, 2015, 15, 1412–1416

Lab on a Chip

11 Y. Luo, Z. Zhang, X. Wang and Y. Zheng, Microelectron. Eng., 2010, 87, 2429–2436. 12 Z. Zhang, Y. Luo, X. Wang, S. He, F. Meng and L. Wang, Microsyst. Technol., 2010, 16, 2043–2048. 13 J. S. Kuo and D. T. Chiu, Lab Chip, 2011, 16, 2656–2665. 14 M. Bu, I. R. Perch-Nielsen, K. S. Sørensen, J. Skov, Y. Sun, D. D. Bang, M. E. Pedersen, M. F. Hansen and A. Wolff, J. Micromech. Microeng., 2013, 23, 074002. 15 K. G. Lee, B. G. Choi, B. I. Kim, T. Shyu, M. S. Oh, S. G. Im, S.-J. Chang, T. J. Lee, N. A. Kotov and S. J. Lee, Adv. Mater., 2014, 26, 6119–6124. 16 M. Heckele and W. K. Schomburg, J. Micromech. Microeng., 2004, 14, R1–R14. 17 B. I. Kim, K. G. Lee, T. J. Lee, B. G. Choi, J. Y. Park, C. Y. Jung, C.-S. Lee and S. J. Lee, Lab Chip, 2013, 13, 4321–4325.

This journal is © The Royal Society of Chemistry 2015

Copyright of Lab on a Chip - Miniaturisation for Chemistry & Biology is the property of Royal Society of Chemistry and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.

Ultrasonic bonding method for heterogeneous microstructures using self-balancing jig.

Perfect sealing of heterogeneous microstructures in plastic-based microfluidic devices is a significant and urgent challenge to be able to apply them ...
2MB Sizes 0 Downloads 7 Views