Silver nanoparticle piezoresistive sensors fabricated by roll-to-roll slot-die coating and laser direct writing Hyungman Lee,1,5,6 Dongjin Lee,2,6 Junghwan Hwang,1 Dongil Nam,2 Changwan Byeon,3 Seung Hwan Ko,4 and Seungsub Lee5,* 1
Korea Electronics Technology Institute, 25 Saenari-ro, Bundang-gu, Seongnam-si, Gyeonggi-do 463-816, South Korea School of Mechanical Engineering, Konkuk University, 120 Neungdong-ro, Gwangjin-gu, Seoul 143-701, South Korea 3 DCN, Co., #B-512, Migun Techno World II, 533-1, Yuseong-gu, Daejeon 305-500, South Korea 4 Department of Mechanical Engineering, Seoul National University, 1 Gwanak-ro, Gwanak-gu, Seoul 151-742, South Korea 5 School of Mechanical Engineering, Korea Advanced Institute Science and Technology, 291 Daehak-ro, Yuseong-gu, Daejeon 305-701, South Korea 6 Authors equally contributed to this work * [email protected] 2
Abstract: In this study, we propose new fusion technology to overcome the limitations of the current printing process for printed electronics. The combinative slot-die coating and laser direct writing on a roll-to-roll (R2R) basis was studied and applied to the fabrication of piezoresistive strain gauges. Piezoresistive sensors were fabricated for the first time by this developed fusion technology. A parametric study was performed for the R2R process, followed by a comparison of the fabricated functional devices with commercial products, which confirmed a 40% increase in the gauge factor of the fabricated sensors over the commercial product. The combinative manufacturing process for functional device fabrication will open a new chapter in the future of printed electronics due to its large area capacity at low cost. ©2014 Optical Society of America OCIS codes: (350.3390) Laser materials processing; (310.3840) Materials and process characterization; (310.1860) Deposition and fabrication; (220.4000) Microstructure fabrication.
References and links 1. 2. 3. 4. 5. 6. 7. 8. 9.
O. J. Romero, W. J. Suszynski, L. E. Scriven, and M. S. Carvalho, “Low-flow limit in slot coating of dilute solutions of high molecular weight polymer,” J. Non-Newton. Fluid Mech. 118, 137–156 (2004). B. G. Higgins and L. E. Scriven, “Capillary pressure and viscous pressure drop set bounds on coating bead operability,” Chem. Eng. Sci. 35(3), 673–682 (1980). F. C. Krebs, “Polymer solar cell modules prepared using roll-to-roll methods: Knife-over-edge coating, slot-die coating and screen printing,” Sol. Energy Mater. Sol. Cells 93(4), 465–475 (2009). F. C. Krebs, “All solution roll-to-roll processed polymer solar cells free from indium-tin-oxide and vacuum coating steps,” Org. Electron. 10(5), 761–768 (2009). F. C. Krebs, T. Tromholt, and M. Jørgensen, “Upscaling of polymer solar cell fabrication using full roll-to-roll processing,” Nanoscale 2(6), 873–886 (2010). J. J. Michels, S. H. P. M. de Winter, and L. H. G. Symonds, “Process optimization of gravure printed lightemitting polymer layers by a neural network approach,” Org. Electron. 10(8), 1495–1504 (2009). A. Sandström, H. F. Dam, F. C. Krebs, and L. Edman, “Ambient fabrication of flexible and large-area organic light-emitting devices using slot-die coating,” Nat Commun 3, 1002 (2012). J. Chang, C. Chi, J. Zhang, and J. Wu, “Controlled growth of large-area high-performance small-molecule organic single-crystalline transistors by slot-die coating using a mixed solvent system,” Adv. Mater. 25(44), 6442–6447 (2013). D. Angmo, M. Hösel, and F. C. Krebs, “All solution processing of ITO-free organic solar cell modules directly on barrier foil,” Sol. Energy Mater. Sol. Cells 107, 329–336 (2012).
10. J. Wang, R. C. Y. Auyeung, H. Kim, N. A. Charipar, and A. Piqué, “Three-dimensional printing of interconnects by laser direct-write of silver nanopastes,” Adv. Mater. 22(40), 4462–4466 (2010). 11. M. Aminuzzaman, A. Watanabe, and T. Miyashita, “Fabrication of conductive silver micropatterns on an organic–inorganic hybrid film by laser direct writing,” Thin Solid Films 517(20), 5935–5939 (2009). 12. K. Jain, M. Klosner, M. Zemel, and S. Raghunandan, “Flexible electronics and displays: high-resolution, roll-toroll, projection lithography and photoablation processing technologies for high-throughput production,” Proc. IEEE 93(8), 1500–1510 (2005). 13. H. W. Choi, D. F. Farson, J. Bovatsek, A. Arai, and D. Ashkenasi, “Direct-write patterning of indium-tin-oxide film by high pulse repetition frequency femtosecond laser ablation,” Appl. Opt. 46(23), 5792–5799 (2007). 14. O. Yavas and M. Takai, “Effect of substrate absorption on the efficiency of laser patterning of indium tin oxide thin films,” J. Appl. Phys. 85(8), 4207–4212 (1999). 15. O. Yavas and M. Takai, “High-speed maskless laser patterning of indium tin oxide thin films,” Appl. Phys. Lett. 73(18), 2558–2560 (1998). 16. M.-F. Chen, Y.-P. Chen, W.-T. Hsiao, and Z.-P. Gu, “Laser direct write patterning technique of indium tin oxide film,” Thin Solid Films 515(24), 8515–8518 (2007). 17. H. Shin, B. Sim, and M. Lee, “Laser-driven high-resolution patterning of indium tin oxide thin film for electronic device,” Opt. Lasers Eng. 48(7-8), 816–820 (2010). 18. C. Molpeceres, S. Lauzurica, J. L. Ocaña, J. J. Gandía, L. Urbina, and J. Cárabe, “Microprocessing of ITO and aSi thin films using ns laser sources,” J. Micromech. Microeng. 15(6), 1271–1278 (2005). 19. J. Chae, S. Appasamy, and K. Jain, “Patterning of indium tin oxide by projection photoablation and lift-off process for fabrication of flat-panel displays,” Appl. Phys. Lett. 90(26), 261102 (2007). 20. G. Račiukaitis, M. Brikas, M. Gedvilas, and T. Rakickas, “Patterning of indium–tin oxide on glass with picosecond lasers,” Appl. Surf. Sci. 253(15), 6570–6574 (2007). 21. M. S. Carvalho and H. S. Kheshgi, “Low-flow limit in slot coating: Theory and experiments,” AIChE J. 46(10), 1907–1917 (2000). 22. M. Cherrington, T. C. Claypole, D. Deganello, I. Mabbett, T. Watson, and D. Worsley, “Ultrafast near-infrared sintering of a slot-die coated nano-silver conducting ink,” J. Mater. Chem. 21(21), 7562–7564 (2011).
1. Introduction Conventional printing technologies are used for mass production of products, such as newspapers and labels. Printed electronics is a set of printing methods used to fabricate electrical devices on various substrates with the same benefits of conventional printing technologies. Printed electronics is one of the most important technologies in electronics and is used in applications including displays, photovoltaics, logic and memory devices, flexible printed circuit boards, RFID, and sensors. This new approach to the technology may allow for large-area, low-cost fabrication of devices that are foldable, rollable, transparent, stretchable, and wearable. Printing technologies seek to mass-produce fine patterned products at low cost. The roll-to-roll (R2R) process is an attractive solution to this need. The process is classified according to the patterning dimensions. An R2R slot-die achieves one-dimensional patterning, whereas gravure, gravure-offset, and flexo accomplish two-dimensional patterning. The patterning dimensions are very limited since the slot-die process is normally applied to parallel line patterns in the roll-feeding direction. The other methods listed can be applied to make two-dimensional patterns with a feature size of 30 μm. Slot-die coating is a general printing method and has been used for mass production. Coating phenomena have been mostly studied by mathematical analysis in the low-flow limit, which is defined by a mechanism that balances the viscous, capillary, and inertial forces in the fluid flow. Representative studies have focused on the modeling and verification of coating practices using viscocapillary modeling, coating bead and meniscus, and Couette and Poiseuille contributions [1, 2]. Recent applications of slot-die coating are mainly organic photovoltaics (OPV) and organic light-emitting diode (OLED) displays. The slot-die coating is advantageous for these applications because the solution-phase organic material can be coated directly into stripes in an atmospheric environment. The coating stripes become continuous once the R2R platform is incorporated. Previous studies were conducted for OPV fabrication that included slot-die coated organic multilayers such as P3HT:PCBM, PEDOT:PSS, and ZnO [3–5]. Krebs et al. used the slot-die coating for OPV in which PEDOT:PSS, the silver electrode, and other materials except the silver front electrode grid were coated using the R2R slot-die process resulting in multilayers of a silver
grid/PEDOT:PSS/P3HT-PCBM/ZnO/silver device [4]. Slot-die coating was also used for the fabrication of OLED devices [6, 7] and thin-film transistors (TFTs) [8]. Except for organic materials, there were examples of a silver (Ag) electrode layer coating instead of indium tin oxide (ITO) [9]. A semi-transparent Ag electrode was formulated with a sheet resistance of 5 Ω/ and transmittance of 30%. Laser direct-write (LDW) technology is a precision processing technology that forms fine patterning on a substrate without a mask using a noncontact method with a digital interface. Wang et al. [10] used a Ag nanopaste to study 3D LDW. Laser-induced forward transfer (LIFT) technique was applied to transfer the pattern to another substrate to form the ohmic contact interconnect with the Ag nanopaste. Laser-induced pyrolysis was also used to perform the LDW process [11]. The Ag electrode with a resolution of 5 to 75 μm was used, and the resistivity of the electrode was found to be 4.3 × 10−6 Ω·cm. Line patterns of 1 μm resolution were implemented with excimer laser projection lithography at 248 nm [12]. The lithography was extended to include the roll-to-roll process to fabricate patterns with a line width of 30 and 60 μm via holes. Many other studies on LDW for conductive ITO patterning on functional devices are available in the literature [13–20]. In this article, a new approach is proposed for printed electronics that combines roll-to-roll slot-die coating and LDW processes. The purpose of the method is to expand the onedimensional patterning in slot-die coating to two-dimensions, while fully exploiting the advantages of the digital process, no mask, noncontact, and low cost of LDW. First, a parametric study of the R2R slot-die coating was performed using silver nanoink, which exhibits large area capacity at low cost. The sheet resistance of Ag conductive thin films was evaluated. Optical and scanning electron microscopy was used for further characterization. 2D patterning in the LDW process was also optimized in terms of laser pulse energy, spot size, and scanning speed, which enabled controllability in the remaining Ag line width. Finally, the piezoresistive strain gauges were fabricated and evaluated to demonstrate the versatility of the proposed manufacturing scheme. 2. Experiment The overall manufacturing scheme is illustrated in Fig. 1, where the roll-to-roll process is used. First, the conductive silver nanoink is coated onto the flexible substrate using the slotdie method followed by a drying process using a near-infrared (NIR) lamp. Subsequently, two-dimensional micropatterning is performed on the conductive silver layer using laser direct writing. To the best of our knowledge, the combination of slot-die coating and LDW on the R2R basis for the fabrication of functional devices has not yet been reported. The R2R process has a maximum capacity of a 250 mm web width and a maximum speed of 5 m/min. The coating unit (DCN Co., South Korea) is composed of the slot-die module and drying section. The slot-die has a maximum ink pumping speed of 15 ml/min, and the NIR lamp has a maximum power of 2.4 kW. The nanosecond laser (F2w-40, YUCO Optics, USA) with pulse repetition rate of 100 kHz and wavelength of 532 nm is used in this study. The pulse width was greater than 10 ns. The conductive silver ink was purchased from Ink-Tek (TECPR-010, South Korea). The laser beam is steered using a galvano-scanner module and is focused on the coated silver conductive film. Two kinds of flexible substrates, polyimide (PI, SKC, South Korea) and polyethylene terephthalate (PET, SKC, South Korea), were used in this study.
Mirror 1
Mirror 2
Laser beam
Fig. 1. Overall manufacturing scheme used in this study.
The parametric study in the R2R slot-die coating was performed considering the ink flow rate, web speed, and NIR power. The coating was performed at web speeds of 20, 30, and 40 mm/s. The ink flow rate was between 2 and 7 ml/min with a step of 1 ml/min. The low flow rate limit, beyond which a stable liquid film cannot be obtained, is dependent on the ink properties and web speed and can be explained theoretically with the viscocapillary model [21]. The power of the NIR lamp used was between 30 and 90% of the maximum power. The sheet resistance of the conductive Ag film was measured using a four-point probe meter (Advanced Instrument Technology Co., South Korea). The average sheet resistance was determined from five consecutive points along the centerline of the substrate. In addition, the power and scanning speed of the laser beam were altered to produce the optimal shape of micropatterns in the LDW process. The size of the laser focus was 7 μm, based on the specification of an f-theta lens (JENOPTIK, f = 102 mm). The line width was optimized by changing the laser power, focus position, and writing speed. The laser power was changed by adjusting the current input to the laser diode and was measured at the LDW sample position with a power meter (GENTEC, Duo, UP19K and UP50N). The focus position was adjusted upward to 0, 20, and 40 μm from the focal point. The writing speed was varied between 100 and 1,000 mm/s with a step size of 100 mm/s. The samples were analyzed with scanning electron microscopy (SEM) and optical microscopy. Finally, the functional device of the piezoresistive sensor was fabricated using optimized process conditions and the combination of the R2R slot-die coating and LDW processes. The performance of the piezoresistive sensors was characterized statically and dynamically in a custom-made testing setup with a cantilever beam made of polymethyl methacrylate (PMMA) and compared with a commercial product (Vishay Co., South Korea). 3. Results and discussion The reason for using the NIR lamp has been reported in a previous study by Cherringon et al. [22]. The sheet resistance of the conductive silver film is dependent on the type and duration of the drying process. NIR, IR, and a convective oven were used, and the results showed that drying by NIR lamp was effective at obtaining low sheet resistance with a short process time. Fixing the web speed at 33.3 mm/s, the coating gap and ink flow rate were changed to obtain the stable silver conductive film on the substrate. We obtained a stable silver film at a coating gap of 250 μm and ink flow rate of 2 ml/min. After drying using NIR lamp, the sheet resistance was measured for different NIR lamp powers. The results are presented in Fig. 2(a). The sheet resistance decreases with an increase in lamp power. The sheet resistance was 2.97 Ω/ at 0.72 kW lamp power and 0.72 Ω/ at 2.16 kW. A higher curing temperature, which is achieved at higher NIR lamp powers, may induce lower sheet resistance. The inset shows an image of the substrate with the conductive Ag film, which appears reflective, as expected.
(a)
(b)
Fig. 2. Sheet resistance of conductive Ag film with respect to NIR lamp power (a) and ink flow rate (b). The inset in (a) shows the image of the Ag film on a PI substrate. The web speed was reduced to 20 mm/s (1.2 m/min) and 30 mm/s (1.8 m/min) on the PET substrate.
To observe the effect of the ink flow rate, various experiments were performed, and results are shown in Fig. 2(b). The samples with the stable conductive film were used for the measurement of sheet resistance. Two types of substrates were used to determine the effect of the substrate material. The power of the NIR lamp power was 90% of maximum, and the gap was set to 100 μm. First, the coating was deposited on a PI substrate at a fixed web speed of 40 mm/s (2.4 m/min). As the ink flow rate increases, the sheet resistance decreases, presumably due to the increased thickness of the Ag film. Specifically, sheet resistances of 426 and 200 mΩ/ were observed for ink flow rates of 4 and 7 ml/min, respectively. The stable Ag film was not formed on the PI substrate at a flow rate less than 3 ml/min. For a PET substrate, however, we found that a minimum ink pumping speed of 8 ml/min was required when the web speed was 40 mm/s. The sheet resistance of the Ag film in this case was 375 mΩ/. We also confirmed that the thickness of the Ag film is smaller on the PET substrate than on the PI substrate. This result is attributed to the difference in surface tension due to the wetting properties of the ink. As Fig. 2(b) shows, lower resistance was observed for lower web speeds at the same ink flow rate, presumably due to increased film thickness. To observe surface morphology and film thickness, the Ag conductive film was cut with a focused ion beam (FIB) as shown in Fig. 3. A top-view image of the Ag conductive film, showing silver nanoparticles, is illustrated in Fig. 3(a). FIB was used to make a 10 μm cut as shown in Fig. 3(b). The thickness of the Ag film was measured using cross-sectional SEM images in Figs. 3(c) and 3(d). A Pt layer is observed on top of the Ag film. The image in Fig. 3(c) is of the Ag film fabricated at a web speed of 20 mm/s and an ink flow rate of 2 ml/min, whereas the image in Fig. 3(d) is of the Ag film fabricated at a web speed of 30 mm/sec and an ink flow rate of 7 ml/min. The film thickness was 67 and 252 nm in Figs. 3(c) and 3(d), respectively. From this thickness information, the resistivity of the Ag film fabricated using the R2R slot-die coating was determined to be 8.7 and 10 μΩ·cm, respectively.
(a)
(b)
(c)
(d)
Fig. 3. Scanning electron microscopy images of top view of Ag film (a), cutting using focused ion beam (b), cross-sectional images of Ag film fabricated at web speed of 20 mm/s and ink flow rate of 2 ml/min (c) and at web speed of 30 mm/s and ink flow rate of 7 ml/min (d).
The LDW process is a fabrication step that removes a small portion of the conductive Ag film after annealing by exposing it to the laser. A pulse laser is required to create the pattern, and the results vary with experimental parameters such as repetition rate, pulse width, spot size, and the power of the laser pulse. As the laser beam moves, which is controlled by the operation of scanning mirrors, an ablated line is formed on the conductive Ag film that has been fabricated from the slot-die coating process. Once the ablated line is defined, the remaining Ag film forms the pattern as shown in Fig. 4(a). The pulse energy was adjusted such that the ablated line width was 10 μm. The laser spot size was 7 μm, and the pitch was 40 μm. The detailed SEM image confirms that the Ag film has been removed completely in the ablated region. By forming a series of ablated lines, the remaining Ag line width was characterized with respect to the laser pulse energy as shown in Fig. 4(b) for a laser spot size of 20 μm and a pitch of 50 μm. As the laser pulse energy increases, the ablated line width increases, leading to a decrease in the line width of the remaining Ag.
(a)
Remained Ag pattern
Remained Ag film
Ablated region
(b)
Remained Ag Line Width (μm)
Ablated line
Laser Pulse Energy (μJ) Fig. 4. Silver line patterns fabricated by laser direct writing process: (a) SEM images with pitch of 40 μm, (b) variation in remaining Ag line width as a function of laser pulse energy for pitch of 50 μm.
The repetition rate and scanning speed influence the quality of the lines as shown in Fig. 5(a). The combination of a repetition rate of 100 kHz and a scanning speed of 1000 mm/s corresponds to 10 μm/pulse. This means that the laser beam moves by 10 μm during one pulse of the laser. Figure 5(a) confirms that the distance between consecutive convex parts is in good agreement with this value. Therefore, the maximum scanning speed of the laser beam
is limited depending on the repetition rate. Furthermore, the laser pulse energy affects the LDW patterned line quality as shown in Fig. 5(b). The repetition rate was fixed at 100 kHz. At low energy of 0.2 μJ, a maximum speed of 450 mm/s was observed to form stable line patterns. At 550 mm/s, a series of dot patterns were observed and Ag lines were not formed. At 0.4 and 0.6 μJ, fine patterns without breakage were obtained throughout the tested scanning speeds. The ablated line width varies from 9.5 to 10.5 μm when the scanning speed is between 350 and 550 mm/s. It is clear that higher pulse energy increases the maximum scanning speed by joining the consecutive dots resulting from removal of more Ag. It is confirmed that LDW can be optimized by adjusting the laser pulse energy and scanning speed to laser specifications. (b)
(a)
Scan Speed (mm/sec) 350
400
450
500
550
Pulse Energy (μJ)
0.2
0.4
0.6
Fig. 5. Characteristics of line patterns using pulsed laser: (a) ablated line fabricated at scan speed of 1,000 mm/s, (b) silver line pattern formation at various pulse energies and scanning speeds.
Instead of using the typical strain gauge fabrication process that includes copper etching, the combination of slot-die coating and LDW processes on the R2R basis was used in this paper. Figure 6 shows the scheme of fabricated strain gauges and the test setup. The metal line resistor in Fig. 6(a) was designed to be 10 mm by 5 mm in size with bonding pads at both ends. The Ag line widths were 54, 120, 180, and 260 μm, and the length was varied from 2.5 to 3.5 mm to produce the appropriate initial resistance. To evaluate the fabricated strain sensors, a cantilever beam was chosen as shown in Fig. 6(b). Using the Euler beam bending model, tension stress was applied to the top surface of the beam. The beam was made of a polymethyl methacrylate plate with dimensions 150 mm x 50 mm x 2.9 mm. The fabricated strain gauge was attached onto the PMMA plate, and a commercial strain gauge (Vishay Co., GF: 2) was used at the same x-position for comparison as shown in Fig. 6(c).
(a)
Strain gauge
(b)
(c)
Fig. 6. Silver nanoparticle piezoresistive strain gauge fabricated by combination of R2R slotdie and LDW: (a) CAD design of single strain gauge, (b) cantilever beam test setup, and (c) photograph of strain gauges attached to cantilever.
The fabricated piezoresistive sensors were evaluated using various methods. First, the time response was characterized for a static load of 108 g as shown in Fig. 7(a). The sensor device had an initial resistance of 120 Ω which is the same as the commercial sensor. Within a few seconds, steady state was attained. The strain was calculated assuming that the fabricated Ag film sensor had the same gauge factor as the commercial product. This means that the fabricated Ag film sensors had a larger change in resistance since the real strain was assumed to be constant. The change in resistance is determined from the following equation: ΔR = GF × ε + αΔT , R
where R is the resistance, GF is the gauge factor, ε is the strain, α is the coefficient of thermal expansion, and ΔT is the temperature change. The change in resistance was about 40% higher in the fabricated sensors than in the commercial products. Additionally, cyclic loading was applied at the tip of cantilever using an oval-shaped ring. Since the rotation speed of the ring was 200 rpm, the period of both sensors was observed as 0.15 s. Similar to the static case, a change in resistance was observed for the fabricated sensor that was almost 40% higher than in the commercial product at the maximum strain. Therefore, the gauge factor for the fabricated Ag film strain sensors is 40% higher than the tested commercial product. Finally, the strain gauges demonstrated good linearity over a range of static loads regardless of the initial resistance as shown in Fig. 7(c). The strain on top of the PMMA plate is proportional to the stress and has a linear relationship with the applied load, F, at the tip. In this experiment, both strain gauges are assumed to have a gauge factor of 2. Consequently, the strain value was measured to be larger for the fabricated Ag film strain gauge than the commercial product, and the fabricated sensor had a larger difference in the resistance at the same strain than the commercial product. The slope shown in Fig. 7(c) for the Ag film sensor is 40% larger than the commercial product. This yields a gauge factor of 2.8 for the fabricated Ag film piezoresistive sensors.
(a)
(b)
(c)
Fig. 7. Performance evaluation of fabricated piezoresistive sensors assuming GF = 2: time response for static load (a) and cyclic load (b), and strain versus static load in strain gauges (c).
4. Conclusion In this study, the combination of a roll-to-roll slot-die coating and laser direct-writing process was proposed to fabricate uniform Ag films over a large area with low sheet resistance. The
piezoresistive sensors, i.e., strain gauges, were fabricated and their performance was evaluated. The R2R slot-die coating was studied in terms of the web speed, ink flow rate, and NIR lamp power. The fabricated Ag film had a sheet resistance of 200 mΩ/ that resulted in a resistivity eight to ten times higher than bulk silver. LDW was successfully applied on the R2R basis, leading to powerful 2D micropatterning. The LDW process was optimized using scan speed and laser power. The design flexibility can be accommodated in a CAD program to operate the laser beam. The laser pulse energy and scanning speed were also optimized. Piezoresistive strain gauge sensors were fabricated for the first time by the combination of slot-die and LDW on an R2R basis. The comparison to commercial products shows a 40% increase in the gauge factor. The combinative manufacturing process for functional device fabrication will open a new chapter in the future of printed electronics. Acknowledgment This work was supported by the Korea Institute of Energy Technology Evaluation and Planning (KETEP) grant funded by the Korea government Ministry of Knowledge Economy (grant number 2011T100100479) and the technology development program (grant number SL122685) funded by the Small and Medium Business Administration. This research was also supported by Leading Foreign Research Institute Recruitment Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT and Future Planning (MSIP) (grant number 2010-00525) and the Global Leading Technology Program of the Office of Strategic R&D Planning (OSP) by the Ministry of Commerce, Industry and Energy, South Korea (grant number 10042421).
Korea Electronics Technology Institute, 25 Saenari-ro, Bundang-gu, Seongnam-si, Gyeonggi-do 463-816, South Korea School of Mechanical Engineering, Konkuk University, 120 Neungdong-ro, Gwangjin-gu, Seoul 143-701, South Korea 3 DCN, Co., #B-512, Migun Techno World II, 533-1, Yuseong-gu, Daejeon 305-500, South Korea 4 Department of Mechanical Engineering, Seoul National University, 1 Gwanak-ro, Gwanak-gu, Seoul 151-742, South Korea 5 School of Mechanical Engineering, Korea Advanced Institute Science and Technology, 291 Daehak-ro, Yuseong-gu, Daejeon 305-701, South Korea 6 Authors equally contributed to this work * [email protected] 2
Abstract: In this study, we propose new fusion technology to overcome the limitations of the current printing process for printed electronics. The combinative slot-die coating and laser direct writing on a roll-to-roll (R2R) basis was studied and applied to the fabrication of piezoresistive strain gauges. Piezoresistive sensors were fabricated for the first time by this developed fusion technology. A parametric study was performed for the R2R process, followed by a comparison of the fabricated functional devices with commercial products, which confirmed a 40% increase in the gauge factor of the fabricated sensors over the commercial product. The combinative manufacturing process for functional device fabrication will open a new chapter in the future of printed electronics due to its large area capacity at low cost. ©2014 Optical Society of America OCIS codes: (350.3390) Laser materials processing; (310.3840) Materials and process characterization; (310.1860) Deposition and fabrication; (220.4000) Microstructure fabrication.
References and links 1. 2. 3. 4. 5. 6. 7. 8. 9.
O. J. Romero, W. J. Suszynski, L. E. Scriven, and M. S. Carvalho, “Low-flow limit in slot coating of dilute solutions of high molecular weight polymer,” J. Non-Newton. Fluid Mech. 118, 137–156 (2004). B. G. Higgins and L. E. Scriven, “Capillary pressure and viscous pressure drop set bounds on coating bead operability,” Chem. Eng. Sci. 35(3), 673–682 (1980). F. C. Krebs, “Polymer solar cell modules prepared using roll-to-roll methods: Knife-over-edge coating, slot-die coating and screen printing,” Sol. Energy Mater. Sol. Cells 93(4), 465–475 (2009). F. C. Krebs, “All solution roll-to-roll processed polymer solar cells free from indium-tin-oxide and vacuum coating steps,” Org. Electron. 10(5), 761–768 (2009). F. C. Krebs, T. Tromholt, and M. Jørgensen, “Upscaling of polymer solar cell fabrication using full roll-to-roll processing,” Nanoscale 2(6), 873–886 (2010). J. J. Michels, S. H. P. M. de Winter, and L. H. G. Symonds, “Process optimization of gravure printed lightemitting polymer layers by a neural network approach,” Org. Electron. 10(8), 1495–1504 (2009). A. Sandström, H. F. Dam, F. C. Krebs, and L. Edman, “Ambient fabrication of flexible and large-area organic light-emitting devices using slot-die coating,” Nat Commun 3, 1002 (2012). J. Chang, C. Chi, J. Zhang, and J. Wu, “Controlled growth of large-area high-performance small-molecule organic single-crystalline transistors by slot-die coating using a mixed solvent system,” Adv. Mater. 25(44), 6442–6447 (2013). D. Angmo, M. Hösel, and F. C. Krebs, “All solution processing of ITO-free organic solar cell modules directly on barrier foil,” Sol. Energy Mater. Sol. Cells 107, 329–336 (2012).
10. J. Wang, R. C. Y. Auyeung, H. Kim, N. A. Charipar, and A. Piqué, “Three-dimensional printing of interconnects by laser direct-write of silver nanopastes,” Adv. Mater. 22(40), 4462–4466 (2010). 11. M. Aminuzzaman, A. Watanabe, and T. Miyashita, “Fabrication of conductive silver micropatterns on an organic–inorganic hybrid film by laser direct writing,” Thin Solid Films 517(20), 5935–5939 (2009). 12. K. Jain, M. Klosner, M. Zemel, and S. Raghunandan, “Flexible electronics and displays: high-resolution, roll-toroll, projection lithography and photoablation processing technologies for high-throughput production,” Proc. IEEE 93(8), 1500–1510 (2005). 13. H. W. Choi, D. F. Farson, J. Bovatsek, A. Arai, and D. Ashkenasi, “Direct-write patterning of indium-tin-oxide film by high pulse repetition frequency femtosecond laser ablation,” Appl. Opt. 46(23), 5792–5799 (2007). 14. O. Yavas and M. Takai, “Effect of substrate absorption on the efficiency of laser patterning of indium tin oxide thin films,” J. Appl. Phys. 85(8), 4207–4212 (1999). 15. O. Yavas and M. Takai, “High-speed maskless laser patterning of indium tin oxide thin films,” Appl. Phys. Lett. 73(18), 2558–2560 (1998). 16. M.-F. Chen, Y.-P. Chen, W.-T. Hsiao, and Z.-P. Gu, “Laser direct write patterning technique of indium tin oxide film,” Thin Solid Films 515(24), 8515–8518 (2007). 17. H. Shin, B. Sim, and M. Lee, “Laser-driven high-resolution patterning of indium tin oxide thin film for electronic device,” Opt. Lasers Eng. 48(7-8), 816–820 (2010). 18. C. Molpeceres, S. Lauzurica, J. L. Ocaña, J. J. Gandía, L. Urbina, and J. Cárabe, “Microprocessing of ITO and aSi thin films using ns laser sources,” J. Micromech. Microeng. 15(6), 1271–1278 (2005). 19. J. Chae, S. Appasamy, and K. Jain, “Patterning of indium tin oxide by projection photoablation and lift-off process for fabrication of flat-panel displays,” Appl. Phys. Lett. 90(26), 261102 (2007). 20. G. Račiukaitis, M. Brikas, M. Gedvilas, and T. Rakickas, “Patterning of indium–tin oxide on glass with picosecond lasers,” Appl. Surf. Sci. 253(15), 6570–6574 (2007). 21. M. S. Carvalho and H. S. Kheshgi, “Low-flow limit in slot coating: Theory and experiments,” AIChE J. 46(10), 1907–1917 (2000). 22. M. Cherrington, T. C. Claypole, D. Deganello, I. Mabbett, T. Watson, and D. Worsley, “Ultrafast near-infrared sintering of a slot-die coated nano-silver conducting ink,” J. Mater. Chem. 21(21), 7562–7564 (2011).
1. Introduction Conventional printing technologies are used for mass production of products, such as newspapers and labels. Printed electronics is a set of printing methods used to fabricate electrical devices on various substrates with the same benefits of conventional printing technologies. Printed electronics is one of the most important technologies in electronics and is used in applications including displays, photovoltaics, logic and memory devices, flexible printed circuit boards, RFID, and sensors. This new approach to the technology may allow for large-area, low-cost fabrication of devices that are foldable, rollable, transparent, stretchable, and wearable. Printing technologies seek to mass-produce fine patterned products at low cost. The roll-to-roll (R2R) process is an attractive solution to this need. The process is classified according to the patterning dimensions. An R2R slot-die achieves one-dimensional patterning, whereas gravure, gravure-offset, and flexo accomplish two-dimensional patterning. The patterning dimensions are very limited since the slot-die process is normally applied to parallel line patterns in the roll-feeding direction. The other methods listed can be applied to make two-dimensional patterns with a feature size of 30 μm. Slot-die coating is a general printing method and has been used for mass production. Coating phenomena have been mostly studied by mathematical analysis in the low-flow limit, which is defined by a mechanism that balances the viscous, capillary, and inertial forces in the fluid flow. Representative studies have focused on the modeling and verification of coating practices using viscocapillary modeling, coating bead and meniscus, and Couette and Poiseuille contributions [1, 2]. Recent applications of slot-die coating are mainly organic photovoltaics (OPV) and organic light-emitting diode (OLED) displays. The slot-die coating is advantageous for these applications because the solution-phase organic material can be coated directly into stripes in an atmospheric environment. The coating stripes become continuous once the R2R platform is incorporated. Previous studies were conducted for OPV fabrication that included slot-die coated organic multilayers such as P3HT:PCBM, PEDOT:PSS, and ZnO [3–5]. Krebs et al. used the slot-die coating for OPV in which PEDOT:PSS, the silver electrode, and other materials except the silver front electrode grid were coated using the R2R slot-die process resulting in multilayers of a silver
grid/PEDOT:PSS/P3HT-PCBM/ZnO/silver device [4]. Slot-die coating was also used for the fabrication of OLED devices [6, 7] and thin-film transistors (TFTs) [8]. Except for organic materials, there were examples of a silver (Ag) electrode layer coating instead of indium tin oxide (ITO) [9]. A semi-transparent Ag electrode was formulated with a sheet resistance of 5 Ω/ and transmittance of 30%. Laser direct-write (LDW) technology is a precision processing technology that forms fine patterning on a substrate without a mask using a noncontact method with a digital interface. Wang et al. [10] used a Ag nanopaste to study 3D LDW. Laser-induced forward transfer (LIFT) technique was applied to transfer the pattern to another substrate to form the ohmic contact interconnect with the Ag nanopaste. Laser-induced pyrolysis was also used to perform the LDW process [11]. The Ag electrode with a resolution of 5 to 75 μm was used, and the resistivity of the electrode was found to be 4.3 × 10−6 Ω·cm. Line patterns of 1 μm resolution were implemented with excimer laser projection lithography at 248 nm [12]. The lithography was extended to include the roll-to-roll process to fabricate patterns with a line width of 30 and 60 μm via holes. Many other studies on LDW for conductive ITO patterning on functional devices are available in the literature [13–20]. In this article, a new approach is proposed for printed electronics that combines roll-to-roll slot-die coating and LDW processes. The purpose of the method is to expand the onedimensional patterning in slot-die coating to two-dimensions, while fully exploiting the advantages of the digital process, no mask, noncontact, and low cost of LDW. First, a parametric study of the R2R slot-die coating was performed using silver nanoink, which exhibits large area capacity at low cost. The sheet resistance of Ag conductive thin films was evaluated. Optical and scanning electron microscopy was used for further characterization. 2D patterning in the LDW process was also optimized in terms of laser pulse energy, spot size, and scanning speed, which enabled controllability in the remaining Ag line width. Finally, the piezoresistive strain gauges were fabricated and evaluated to demonstrate the versatility of the proposed manufacturing scheme. 2. Experiment The overall manufacturing scheme is illustrated in Fig. 1, where the roll-to-roll process is used. First, the conductive silver nanoink is coated onto the flexible substrate using the slotdie method followed by a drying process using a near-infrared (NIR) lamp. Subsequently, two-dimensional micropatterning is performed on the conductive silver layer using laser direct writing. To the best of our knowledge, the combination of slot-die coating and LDW on the R2R basis for the fabrication of functional devices has not yet been reported. The R2R process has a maximum capacity of a 250 mm web width and a maximum speed of 5 m/min. The coating unit (DCN Co., South Korea) is composed of the slot-die module and drying section. The slot-die has a maximum ink pumping speed of 15 ml/min, and the NIR lamp has a maximum power of 2.4 kW. The nanosecond laser (F2w-40, YUCO Optics, USA) with pulse repetition rate of 100 kHz and wavelength of 532 nm is used in this study. The pulse width was greater than 10 ns. The conductive silver ink was purchased from Ink-Tek (TECPR-010, South Korea). The laser beam is steered using a galvano-scanner module and is focused on the coated silver conductive film. Two kinds of flexible substrates, polyimide (PI, SKC, South Korea) and polyethylene terephthalate (PET, SKC, South Korea), were used in this study.
Mirror 1
Mirror 2
Laser beam
Fig. 1. Overall manufacturing scheme used in this study.
The parametric study in the R2R slot-die coating was performed considering the ink flow rate, web speed, and NIR power. The coating was performed at web speeds of 20, 30, and 40 mm/s. The ink flow rate was between 2 and 7 ml/min with a step of 1 ml/min. The low flow rate limit, beyond which a stable liquid film cannot be obtained, is dependent on the ink properties and web speed and can be explained theoretically with the viscocapillary model [21]. The power of the NIR lamp used was between 30 and 90% of the maximum power. The sheet resistance of the conductive Ag film was measured using a four-point probe meter (Advanced Instrument Technology Co., South Korea). The average sheet resistance was determined from five consecutive points along the centerline of the substrate. In addition, the power and scanning speed of the laser beam were altered to produce the optimal shape of micropatterns in the LDW process. The size of the laser focus was 7 μm, based on the specification of an f-theta lens (JENOPTIK, f = 102 mm). The line width was optimized by changing the laser power, focus position, and writing speed. The laser power was changed by adjusting the current input to the laser diode and was measured at the LDW sample position with a power meter (GENTEC, Duo, UP19K and UP50N). The focus position was adjusted upward to 0, 20, and 40 μm from the focal point. The writing speed was varied between 100 and 1,000 mm/s with a step size of 100 mm/s. The samples were analyzed with scanning electron microscopy (SEM) and optical microscopy. Finally, the functional device of the piezoresistive sensor was fabricated using optimized process conditions and the combination of the R2R slot-die coating and LDW processes. The performance of the piezoresistive sensors was characterized statically and dynamically in a custom-made testing setup with a cantilever beam made of polymethyl methacrylate (PMMA) and compared with a commercial product (Vishay Co., South Korea). 3. Results and discussion The reason for using the NIR lamp has been reported in a previous study by Cherringon et al. [22]. The sheet resistance of the conductive silver film is dependent on the type and duration of the drying process. NIR, IR, and a convective oven were used, and the results showed that drying by NIR lamp was effective at obtaining low sheet resistance with a short process time. Fixing the web speed at 33.3 mm/s, the coating gap and ink flow rate were changed to obtain the stable silver conductive film on the substrate. We obtained a stable silver film at a coating gap of 250 μm and ink flow rate of 2 ml/min. After drying using NIR lamp, the sheet resistance was measured for different NIR lamp powers. The results are presented in Fig. 2(a). The sheet resistance decreases with an increase in lamp power. The sheet resistance was 2.97 Ω/ at 0.72 kW lamp power and 0.72 Ω/ at 2.16 kW. A higher curing temperature, which is achieved at higher NIR lamp powers, may induce lower sheet resistance. The inset shows an image of the substrate with the conductive Ag film, which appears reflective, as expected.
(a)
(b)
Fig. 2. Sheet resistance of conductive Ag film with respect to NIR lamp power (a) and ink flow rate (b). The inset in (a) shows the image of the Ag film on a PI substrate. The web speed was reduced to 20 mm/s (1.2 m/min) and 30 mm/s (1.8 m/min) on the PET substrate.
To observe the effect of the ink flow rate, various experiments were performed, and results are shown in Fig. 2(b). The samples with the stable conductive film were used for the measurement of sheet resistance. Two types of substrates were used to determine the effect of the substrate material. The power of the NIR lamp power was 90% of maximum, and the gap was set to 100 μm. First, the coating was deposited on a PI substrate at a fixed web speed of 40 mm/s (2.4 m/min). As the ink flow rate increases, the sheet resistance decreases, presumably due to the increased thickness of the Ag film. Specifically, sheet resistances of 426 and 200 mΩ/ were observed for ink flow rates of 4 and 7 ml/min, respectively. The stable Ag film was not formed on the PI substrate at a flow rate less than 3 ml/min. For a PET substrate, however, we found that a minimum ink pumping speed of 8 ml/min was required when the web speed was 40 mm/s. The sheet resistance of the Ag film in this case was 375 mΩ/. We also confirmed that the thickness of the Ag film is smaller on the PET substrate than on the PI substrate. This result is attributed to the difference in surface tension due to the wetting properties of the ink. As Fig. 2(b) shows, lower resistance was observed for lower web speeds at the same ink flow rate, presumably due to increased film thickness. To observe surface morphology and film thickness, the Ag conductive film was cut with a focused ion beam (FIB) as shown in Fig. 3. A top-view image of the Ag conductive film, showing silver nanoparticles, is illustrated in Fig. 3(a). FIB was used to make a 10 μm cut as shown in Fig. 3(b). The thickness of the Ag film was measured using cross-sectional SEM images in Figs. 3(c) and 3(d). A Pt layer is observed on top of the Ag film. The image in Fig. 3(c) is of the Ag film fabricated at a web speed of 20 mm/s and an ink flow rate of 2 ml/min, whereas the image in Fig. 3(d) is of the Ag film fabricated at a web speed of 30 mm/sec and an ink flow rate of 7 ml/min. The film thickness was 67 and 252 nm in Figs. 3(c) and 3(d), respectively. From this thickness information, the resistivity of the Ag film fabricated using the R2R slot-die coating was determined to be 8.7 and 10 μΩ·cm, respectively.
(a)
(b)
(c)
(d)
Fig. 3. Scanning electron microscopy images of top view of Ag film (a), cutting using focused ion beam (b), cross-sectional images of Ag film fabricated at web speed of 20 mm/s and ink flow rate of 2 ml/min (c) and at web speed of 30 mm/s and ink flow rate of 7 ml/min (d).
The LDW process is a fabrication step that removes a small portion of the conductive Ag film after annealing by exposing it to the laser. A pulse laser is required to create the pattern, and the results vary with experimental parameters such as repetition rate, pulse width, spot size, and the power of the laser pulse. As the laser beam moves, which is controlled by the operation of scanning mirrors, an ablated line is formed on the conductive Ag film that has been fabricated from the slot-die coating process. Once the ablated line is defined, the remaining Ag film forms the pattern as shown in Fig. 4(a). The pulse energy was adjusted such that the ablated line width was 10 μm. The laser spot size was 7 μm, and the pitch was 40 μm. The detailed SEM image confirms that the Ag film has been removed completely in the ablated region. By forming a series of ablated lines, the remaining Ag line width was characterized with respect to the laser pulse energy as shown in Fig. 4(b) for a laser spot size of 20 μm and a pitch of 50 μm. As the laser pulse energy increases, the ablated line width increases, leading to a decrease in the line width of the remaining Ag.
(a)
Remained Ag pattern
Remained Ag film
Ablated region
(b)
Remained Ag Line Width (μm)
Ablated line
Laser Pulse Energy (μJ) Fig. 4. Silver line patterns fabricated by laser direct writing process: (a) SEM images with pitch of 40 μm, (b) variation in remaining Ag line width as a function of laser pulse energy for pitch of 50 μm.
The repetition rate and scanning speed influence the quality of the lines as shown in Fig. 5(a). The combination of a repetition rate of 100 kHz and a scanning speed of 1000 mm/s corresponds to 10 μm/pulse. This means that the laser beam moves by 10 μm during one pulse of the laser. Figure 5(a) confirms that the distance between consecutive convex parts is in good agreement with this value. Therefore, the maximum scanning speed of the laser beam
is limited depending on the repetition rate. Furthermore, the laser pulse energy affects the LDW patterned line quality as shown in Fig. 5(b). The repetition rate was fixed at 100 kHz. At low energy of 0.2 μJ, a maximum speed of 450 mm/s was observed to form stable line patterns. At 550 mm/s, a series of dot patterns were observed and Ag lines were not formed. At 0.4 and 0.6 μJ, fine patterns without breakage were obtained throughout the tested scanning speeds. The ablated line width varies from 9.5 to 10.5 μm when the scanning speed is between 350 and 550 mm/s. It is clear that higher pulse energy increases the maximum scanning speed by joining the consecutive dots resulting from removal of more Ag. It is confirmed that LDW can be optimized by adjusting the laser pulse energy and scanning speed to laser specifications. (b)
(a)
Scan Speed (mm/sec) 350
400
450
500
550
Pulse Energy (μJ)
0.2
0.4
0.6
Fig. 5. Characteristics of line patterns using pulsed laser: (a) ablated line fabricated at scan speed of 1,000 mm/s, (b) silver line pattern formation at various pulse energies and scanning speeds.
Instead of using the typical strain gauge fabrication process that includes copper etching, the combination of slot-die coating and LDW processes on the R2R basis was used in this paper. Figure 6 shows the scheme of fabricated strain gauges and the test setup. The metal line resistor in Fig. 6(a) was designed to be 10 mm by 5 mm in size with bonding pads at both ends. The Ag line widths were 54, 120, 180, and 260 μm, and the length was varied from 2.5 to 3.5 mm to produce the appropriate initial resistance. To evaluate the fabricated strain sensors, a cantilever beam was chosen as shown in Fig. 6(b). Using the Euler beam bending model, tension stress was applied to the top surface of the beam. The beam was made of a polymethyl methacrylate plate with dimensions 150 mm x 50 mm x 2.9 mm. The fabricated strain gauge was attached onto the PMMA plate, and a commercial strain gauge (Vishay Co., GF: 2) was used at the same x-position for comparison as shown in Fig. 6(c).
(a)
Strain gauge
(b)
(c)
Fig. 6. Silver nanoparticle piezoresistive strain gauge fabricated by combination of R2R slotdie and LDW: (a) CAD design of single strain gauge, (b) cantilever beam test setup, and (c) photograph of strain gauges attached to cantilever.
The fabricated piezoresistive sensors were evaluated using various methods. First, the time response was characterized for a static load of 108 g as shown in Fig. 7(a). The sensor device had an initial resistance of 120 Ω which is the same as the commercial sensor. Within a few seconds, steady state was attained. The strain was calculated assuming that the fabricated Ag film sensor had the same gauge factor as the commercial product. This means that the fabricated Ag film sensors had a larger change in resistance since the real strain was assumed to be constant. The change in resistance is determined from the following equation: ΔR = GF × ε + αΔT , R
where R is the resistance, GF is the gauge factor, ε is the strain, α is the coefficient of thermal expansion, and ΔT is the temperature change. The change in resistance was about 40% higher in the fabricated sensors than in the commercial products. Additionally, cyclic loading was applied at the tip of cantilever using an oval-shaped ring. Since the rotation speed of the ring was 200 rpm, the period of both sensors was observed as 0.15 s. Similar to the static case, a change in resistance was observed for the fabricated sensor that was almost 40% higher than in the commercial product at the maximum strain. Therefore, the gauge factor for the fabricated Ag film strain sensors is 40% higher than the tested commercial product. Finally, the strain gauges demonstrated good linearity over a range of static loads regardless of the initial resistance as shown in Fig. 7(c). The strain on top of the PMMA plate is proportional to the stress and has a linear relationship with the applied load, F, at the tip. In this experiment, both strain gauges are assumed to have a gauge factor of 2. Consequently, the strain value was measured to be larger for the fabricated Ag film strain gauge than the commercial product, and the fabricated sensor had a larger difference in the resistance at the same strain than the commercial product. The slope shown in Fig. 7(c) for the Ag film sensor is 40% larger than the commercial product. This yields a gauge factor of 2.8 for the fabricated Ag film piezoresistive sensors.
(a)
(b)
(c)
Fig. 7. Performance evaluation of fabricated piezoresistive sensors assuming GF = 2: time response for static load (a) and cyclic load (b), and strain versus static load in strain gauges (c).
4. Conclusion In this study, the combination of a roll-to-roll slot-die coating and laser direct-writing process was proposed to fabricate uniform Ag films over a large area with low sheet resistance. The
piezoresistive sensors, i.e., strain gauges, were fabricated and their performance was evaluated. The R2R slot-die coating was studied in terms of the web speed, ink flow rate, and NIR lamp power. The fabricated Ag film had a sheet resistance of 200 mΩ/ that resulted in a resistivity eight to ten times higher than bulk silver. LDW was successfully applied on the R2R basis, leading to powerful 2D micropatterning. The LDW process was optimized using scan speed and laser power. The design flexibility can be accommodated in a CAD program to operate the laser beam. The laser pulse energy and scanning speed were also optimized. Piezoresistive strain gauge sensors were fabricated for the first time by the combination of slot-die and LDW on an R2R basis. The comparison to commercial products shows a 40% increase in the gauge factor. The combinative manufacturing process for functional device fabrication will open a new chapter in the future of printed electronics. Acknowledgment This work was supported by the Korea Institute of Energy Technology Evaluation and Planning (KETEP) grant funded by the Korea government Ministry of Knowledge Economy (grant number 2011T100100479) and the technology development program (grant number SL122685) funded by the Small and Medium Business Administration. This research was also supported by Leading Foreign Research Institute Recruitment Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT and Future Planning (MSIP) (grant number 2010-00525) and the Global Leading Technology Program of the Office of Strategic R&D Planning (OSP) by the Ministry of Commerce, Industry and Energy, South Korea (grant number 10042421).
