Reduction of droplet volume by controlling actuating waveforms in inkjet printing for micro- pattern formation This article has been downloaded from IOPscience. Please scroll down to see the full text article. 2009 J. Micromech. Microeng. 19 055010 (http://iopscience.iop.org/0960-1317/19/5/055010) Download details: IP Address: 131.187.94.93 The article was downloaded on 26/08/2013 at 15:17 Please note that terms and conditions apply. View the table of contents for this issue, or go to the journal homepage for more Home Search Collections Journals About Contact us My IOPscience IOP PUBLISHING JOURNAL OF MICROMECHANICS AND MICROENGINEERING J. Micromech. Microeng. 19 (2009) 055010 (8pp) doi:10.1088/0960-1317/19/5/055010 Reduction of droplet volume by controlling actuating waveforms in inkjet printing for micro-pattern formation H Y Gan1, Xuechuan Shan1,4, T Eriksson2,3, B K Lok1 and Y C Lam2 1 Singapore Institute of Manufacturing Technology (SIMTech), 71 Nanyang Drive, Singapore 638075 2 Mechanical & Aerospace Engineering, Nanyang Technological University, Singapore 639798 3 Department of Engineering Sciences, Uppsala University, Box 539, Uppsala, Sweden E-mail:
[email protected] Received 4 October 2008, in final form 17 December 2008 Published 28 April 2009 Online at stacks.iop.org/JMM/19/055010 Abstract Inkjet printing has proven to be a promising and flexible process methodology for low cost and drop-on-demand pattern formation in small-scale production of micro-electro-mechanical systems. To optimize the micro-patterns formed by inkjet printing, an accurate control of droplet volume is essential and critical. In this study, an inkjet system with a nozzle driven by a circular piezoelectric element was used to explore the impact of different waveforms on droplet volume. The investigation into this study included the impact of unipolar, bipolar, M-shaped and W-shaped waveforms as well as the effects of their amplitudes and pulse durations. The inkjetting behavior of Newtonian and non-Newtonian fluids under different actuating waveforms was studied in order to obtain a maximum reduction in ejected droplet sizes. An effective reduction of droplet volume in the range of 50–80% was demonstrated. The results of inkjetting PEDOT ink on a polished silicon surface showed that a 50% reduction in line width was achieved. (Some figures in this article are in colour only in the electronic version) 1. Introduction Inkjet printing is a non-contact, cost effective and direct additive technique for micro-pattern formation. Inkjet printing of metallic-, organic- and inorganic-based inks for direct writing has been reported, for instance, depositing electro- luminescent polymers for making light-emitting diodes and depositing metallic nano-particles for interconnect (de Gans et al 2004, Bidoki et al 2007, Lee et al 2008), dispensing solder balls on circuit boards (Orme and Smith 2000) or inkjetting organic droplets for biochips (Shena et al 1998). Piezoelectric driving is a popular actuating method for ejecting droplets from an inkjet nozzle. In this method, droplets are ejected from a squeezing mode, which is generated from a piezoelectric element. The actuating waveform of the piezoelectric element has a strong influence on the droplet ejection process, controlling both the size and velocity of 4 Author to whom any correspondence should be addressed. ejected droplets. Sakai (2000) applied a sequence of negative and positive pulses to modulate the actuating mode and obtained droplets with a diameter of about 70% of nozzle diameter, when a nozzle of 32 μm in diameter was used. Chen and Basaran (2002) investigated formation of small water– glycerin droplets and observed a tongue of liquid moving at a higher velocity than the nearby liquid by applying a successive negative and positive driving wave. In inkjet printing, research activities were largely dominated by improving processing speed and inkjetting quality. The processing speed can be improved by increasing the numbers of nozzles; and the inkjetting quality is closely related to the ejected droplets, including the droplet size, droplet consistency and satellite formation. The simplest way of decreasing the droplet size is to reduce the diameter of a nozzle orifice. In addition to its high manufacturing cost, a small orifice is more susceptible to clogging (Basaran 2002). Hence, it is advantageous to produce droplets that are smaller than the nozzle orifice. 0960-1317/09/055010+08$30.00 1 © 2009 IOP Publishing Ltd Printed in the UK J. Micromech. Microeng. 19 (2009) 055010 H Y Gan et al Figure 1. The inkjet printing system (Jetlab II from Microfab Technologies Inc.). Inkjet printing of polymeric inks has received immense attention recently due to the advancement of printed electronics. The typical inkjetting behavior of polymeric ink is that droplet remains connected to the nozzle tip via an elongating filament. This paper focuses on process study of reducing droplet volume of non-Newtonian fluids, i.e. polymeric inks, using a piezo-actuated inkjet nozzle. Poly(3,4-ethylenedioxythiophene) (PEDOT) was selected and employed as a polymeric ink due to its vast potential for applications in organic transistors, organic light-emitting diodes and solar cells. Another reason for selecting PEDOT was that volume reduction of the PEDOT droplet via waveform modification was highly expected in the piezo-actuated inkjet system. In this study, the effects of different waveforms, e.g. the unipolar waveform, bipolar waveform, M-shaped waveform and W-shaped waveform were investigated; the impacts of influential parameters were explored in order to observe inkjetting behavior and to obtain a desired reduction in the ejected droplet volume. 2. Experimental setup The process study on volume reduction of PEDOT droplet in inkjet printing was conducted by means of a commercialized inkjetting system (Jetlab II, Microfab Technologies Inc.), as shown in figure 1. It was a desk-top platform and inkjet printing was performed using a single-nozzle print-head actuated by a piezoelectric sleeve tube. The ink container was pressurized negatively by an external compressor to overcome any capillary dripping during an idle mode. The inkjet print- head consisted of a glass capillary bonded to a piezoelectric sleeve actuator, as shown in figure 2. The capillary tapered to Figure 2. X-ray image of the print-head. The image illustrates a glass nozzle tube with 50 μm inner diameter which is sleeved with a tube-type piezoelectric element. a fine orifice with a diameter of 50 μm, through which droplets were ejected when a suitable electrical pulse was applied to the piezoelectric actuator. The electrical pulse (i.e. the electrical voltage) caused a change in shape of the piezoelectric actuator that generated an acoustic pressure pulse in the capillary tube for droplet ejection. Therefore, the quality of ejected droplets was strongly governed by the precise shape of electrical pulse (hereafter referred to as waveform) used to drive the piezoelectric actuator. In this inkjetting system, a horizontal camera (i.e. camera 2 in figure 1) and a stroboscope LED were attached at both sides of the inkjet nozzle, so that images of droplet ejection could be recorded at different timings of droplet formation. In this study, at a constant jetting frequency of 550 Hz, different waveforms, e.g. unipolar, bipolar, M- shaped and W-shaped waveforms were designed in order to study the inkjetting behavior and to reduce droplet volume of a selected polymeric ink. 2 J. Micromech. Microeng. 19 (2009) 055010 H Y Gan et al Figure 3. The frequency sweep oscillatory test of PEDOT (in term of elastic modulus G′, viscous modulus G′′ and δ), indicating the behavior of a viscoelastic liquid. 3. Materials characterization A commercially available conducting polymer, Poly(3, 4-ethylenedioxythiophene) (PEDOT, Sigma Aldrich) was used in this study. PEDOT has excellent electric conductivity, chemical stability and optical transparency (Crispin et al 2003, Mabrook et al 2005); its applications have been expanded to different areas including conducting electrodes in FETs (Sirringhaus et al 2000) and LEDs (Kim et al 2002). PEDOT is a viscoelastic fluid which can behave both elastically and viscously upon different time scales of associated flow. The viscoelasticity of a polymer solution is generally attributed to the deformation of polymer chains and the consequent generation of unequal normal stresses. Molecules in a polymeric ink have a significant impact on the droplet in the inkjet process (Shore and Harrison 2005). The properties of an ejected droplet are strongly governed by the induced pressure wave within the nozzle chamber, physical dimension of the nozzle and the rheological properties of the ink such as viscosity, elasticity and surface tension. It is known that for most polymeric solutions, its viscosity changes depending on the shear rate. The apparent viscosity of a polymeric solution decreases with an increase of the shear rate. The shear viscosity of PEDOT ink was characterized using a standard rotational rheometer (Gemini HR Nano rheometer, Malvern Instruments Ltd). PEDOT exhibited shear-thinning behavior under a shear rate sweep and the measured zero-shear viscosity was 100 mPa s approximately by the cross-model regression fit. However, in extremely high shear flow (i.e. inkjet printing process), the polymer shear viscosity was masked by its solvent viscosity (Bird et al 1987, Gan and Lam 2008), which could not be quantified with existing facilities. An oscillatory frequency sweep test was also conducted to illustrate PEDOT viscoelastic behavior in terms of its elastic (G′) and viscous (G′′) modulus, as shown in figure 3. Prominent viscoelastic properties were demonstrated. With increase in oscillation frequency, the polymer structure of the temporary network- of-entanglements was showing more rigidity and therefore, the elastic behavior was showing increasing dominance by the increase of G′. Conversely, the importance of the viscous behavior was clearly reducing and thus G′′ was reducing. Moreover, the surface tension force is another important parameter. In the inkjet printing process, a pendant-like droplet hanging at the end of the nozzle tip will be detached and then drop down to substrate when the resultant of kinetic and gravitational forces exceeds its surface tension force. A low surface tension makes it easier for a stream of ink to break up into a series of droplets. An increase in surface tension requires an increase in the electrical driving voltage to generate a droplet. By rule of thumb in common practice, inks with a surface tension of 20–70 mN m−1 are favorable for inkjet printing. The interfacial tension of the PEDOT measured at room temperature was 31.5 mN m−1. Generally, for the inkjet printing process, Newtonian fluids will produce both a primary and satellite drops under specific inkjetting conditions. Conversely, polymeric fluids (i.e. non-Newtonian fluids) will suppress the satellite drop formation due to the inherent damping effects by its viscoelastic property (Shore and Harrison 2005). 4. Effects of actuating waveforms 4.1. Unipolar waveform A unipolar waveform, as shown in figure 4, was the simplest one for driving the piezoelectric actuator to generate a droplet. 3 J. Micromech. Microeng. 19 (2009) 055010 H Y Gan et al Figure 4. Profile of the unipolar waveform. tD stands for dwell time, while tR and tF represent the rise time and the fall time, respectively. The critical parameters of this waveform were its amplitude V and dwell time tD. In a practical inkjetting system, the waveform was actually in trapezoidal shape with a rise time (tR), a dwell time (tD) and a fall time (tF). The minimum achievable rise time tR and fall time tF were both 2 μs. The inkjet system had a fluid acoustic resonance due to the effect of compressibility. This indicated the existence of an optimal pulse width, which was defined as the one capable of achieving the highest drop velocity for a given pulse amplitude. For the nozzle used in this study, the typical dwell time tD was ranging from 20 to 30 μs. The rise time tR and fall time tF were referred to as the buffer time for a driving signal to reach its assigned amplitudes. Generally the rise time tR and fall time tF were kept longer for non-Newtonian fluids (i.e. polymeric fluids) than Newtonian fluids due to their viscoelastic effects. Figure 5(a) shows a curve of droplet volume versus dwell time and its corresponding regression curve fitting when PEDOT was used. The dwell time varied from tD = 16 μs to tD = 33 μs, and the droplet volume reached a maximum value at tD = 27 μs before decreasing. It was found that no ejection was observed when the dwell time was beyond the Figure 5. Effects of dwell time in unipolar waveform on PEDOT with pulse amplitude U = 35 V, tR = tF = 3 μs. (a) Dwell time tD versus droplet volume; (b) dwell time tD versus droplet velocity. ‘×’ represents outlier data points. range shown in figure 5(a). In figure 5(a), outlier data points (indicated by ‘×’ marks) were excluded from the regression curve fitting. Smaller droplet volume was expected at shorter tD due to weaker reinforce pulse generated within the nozzle chamber (Bogy and Talke 1984). Figure 5(b) shows a curve of droplet velocity versus dwell time tD and the regression curve fitting. The droplet velocity increased gradually with the increase of tD up to tD = 29 μs, after which the velocity decreased rapidly and finally no ejection occurred as dwell time tD was beyond 33 μs. Hence, the optimal dwell time tD for generating a miniaturized droplet was from 27 μs to 30 μs. Figures 6(a) and (b) show the relationship of amplitude U of the unipolar waveform versus droplet volume and velocity, respectively. Both the droplet volume and velocity increased in linear proportion to the increase in voltage when the amplitude ranged from 28 V to 42 V. The smallest droplet volume generated was larger than 100 pL. The droplet volume increased about 50–60% when the amplitude of waveform increased from 30 to 40 V. Furthermore, the linearly proportionate curves in figure 6 were mainly attributed to the viscoelastic effect of PEDOT, which suppressed the formation of satellite droplets. However, due to the viscoelastic effect, the acoustic pulse induced by the driving waveform was largely damped off by the viscoelasticity of PEDOT. Thus, higher energy was needed to manipulate this non-Newtonian fluid, this in turn generated larger volume of ejected droplet. Table 1 presents the optimal operating conditions and droplet volumes of PEDOT ink and deionized (DI) water by means of a unipolar waveform. It was noted that optimal driving parameters for PEDOT were different from those for DI water. The driving waveform to obtain a minimum droplet volume of PEDOT was with higher amplitude. PEDOT required 50% higher voltage to be ejected while having a similar optimal value of tD and shorter rise time tR and fall time tF. The droplet volume of PEDOT was not as sensitive to the amplitude of driving waveform as DI water did. This mainly contributes to the viscoelastic properties of PEDOT; higher driving force and shorter time scale were needed to compensate the loss of momentum within this polymeric ink. 4 J. Micromech. Microeng. 19 (2009) 055010 H Y Gan et al Figure 6. Effects of driving voltage in the unipolar waveform on PEDOT with tD = 25 μs, tR = tF = 3 μs. (a) Driving voltage versus droplet volume; (b) driving voltage versus droplet velocity. Figure 7. Profile of the M-shaped waveform. tE represents interval time between the first and second pulses. Table 1. The smallest droplet volume achieved using the unipolar waveform. Droplet Fluids U (V) tR (μs) tD (μs) tF (μs) volume (pL) DI Water 20.0 5.0 25.0 5.0 60.5 PEDOT 30.0 2.0 25.0 2.0 92.3 4.2. M-shaped waveform The M-shaped waveform, which was composed of two unipolar waveforms as shown in figure 7, was designed with a purpose of minimizing satellite droplets particularly for PEDOT ink. The second unipolar wave with amplitude U2 and dwell time tD2 was introduced to expand the piezoelectric sleeve after ejecting the preceding droplet, so that high enough negative pressure was generated to ‘pull’ the filament tongue back into the nozzle. This second wave was carefully designed with low kinetic energy to avoid double ejection during the fall time period. The interval time tE between two waves, however, needed to be carefully adjusted. At tE = 10 μs the droplet had a long tail and created a cloud of Figure 8. The images of PEDOT droplets ejected using different M-shaped waveforms. Both the main droplets and the tails vary with the interval time tE (see figure 7) between two pulses. fine satellites, which led to a large droplet. The length of the tail and the droplet volume could be reduced by shortening the interval time tE. At tE = 2 μs, the breakup of droplet from the tail filament was taking place near to the nozzle tip, resulting in a small droplet. Figure 8 illustrates the images of ejected PEDOT droplets under different interval time tE. It was seen that the optimal interval time tE was ranging from tE = 4 μs to 1 μs. Further refinement of the interval time in order to realize the smallest droplet was necessary but not conducted in this study due to the hardware limitation. It could be concluded, however, that tE = 2 μs was the acceptable optimal value. Based on the feasible working condition, the first square wave was chosen as a normal unipolar waveform except that the pulse amplitude U1 was set slightly higher for higher velocity to compensate the damping effects induced by the second square wave. The actual U1 was in the range of 29–35 V for PEDOT, and U2 was in the range of 35–45 V. 5 J. Micromech. Microeng. 19 (2009) 055010 H Y Gan et al Table 2. The smallest droplet volume achieved using the M-shaped waveform. Fluids U1 (V) U2 (V) tR1 (μs) tD1 (μs) tF1 (μs) tE (μs) tR2 (μs) tD2 (μs) tF2 (μs) Droplet volume (pL) DI water 18.0 27.0 5.0 20.0 2.0 5.0 1.0 3.0 1.0 20.4 PEDOT 37.0 45.0 5.0 22.0 2.0 4.0 1.0 4.0 1.0 25∼41 Figure 9. Profile of the bipolar waveform. Figure 10. Profile of the W-shaped waveform. Table 2 indicates the cases of PEDOT and DI water when the M-shaped waveform was used. It was found that the M-shaped waveform was very effective for droplet volume reduction to PEDOT ink as well as DI water. The droplets of PEDOT and DI water were reduced to 27–44% and 33%, respectively, in droplet volume compared to the case of the unipolar waveform in table 1. It was also found that dwell time tD2 of the second square wave was a critical parameter to minimize the satellite phenomenon; the range of tD2 was from 3 to 5 μs. 4.3. Bipolar waveform Figure 9 depicts a so-called bipolar driving waveform, which consisted of a succession of two square-wave pulses, the first Table 3. The smallest droplet volume achieved using the bipolar waveform. Droplet U1 U2 tR1 tD tF tZ tR2 volume Fluids (V) (V) (μs) (μs) (μs) (μs) (μs) (pL) DI water 10.0 −26.0 1.0 11.0 2 0.5 0.0 15.2 PEDOT 25.0 −40.0 0.2 4.0 0.5 1.0 0.0 52.4 was positive and the second negative. This waveform allowed a high enough voltage difference without pushing the pulse amplitude too high. The negative amplitude was introduced as a solution to eliminate satellite droplets. The satellite droplets resulted from a combination of the ink properties and the driving waveform, one possible reason was the remaining fluid momentum after the main drop formation. Thus, this negative wave generated a ‘suction’ effect immediately after ejection, slightly different from the M-shaped waveform mechanism, which pulled the extra volume of the ink back into the nozzle cavity that would otherwise become tiny satellite droplets. It was seen from table 3 that the droplet volumes of both PEDOT and DI water were obviously reduced by using the bipolar waveform. However, the droplet volume of DI water was reduced more significantly than PEDOT because the viscoelasticity of PEDOT made it more difficult to pull the ink back effectively. 4.4. W-shaped waveform A new driving waveform, which was shown in figure 10 and referred to as the W-shaped waveform, was introduced. The W-shaped waveform was composed of a succession of three waves with different peak amplitudes. The first wave was negative, the second positive and the third negative again. The first negative wave was introduced for the purpose of eliminating any residual acoustic wave from the last ejection cycle. The subsequent positive and negative waves, the combination of which was a modified bipolar wave (see figure 9), were used for ejecting the ink droplet. Droplet volumes were obviously reduced compared to the previous three waveforms. As illustrated in table 4, the volume of ejected PEDOT droplets varied from 14 to 25 pL; and the droplet volume of 11.8 pL for DI water was produced corresponding to a droplet radius of 14 μm when a nozzle with 50 μm diameter was used. The potential differences between U2 and U3 of this waveform were 115 V for PEDOT and 34 V for DI water, respectively. Figure 11 illustrates the real time inkjetting of PEDOT ink. One should highlight here the drawback of utilizing the W-shaped waveform for PEDOT inkjet printing. The 6 J. Micromech. Microeng. 19 (2009) 055010 H Y Gan et al Table 4. The smallest droplet volume achieved using the W-shaped waveform. Fluids U1 (V) U2 (V) U3 (V) tF1 (μs) tD1 (μs) tR1 (μs) tD2 (μs) tF2 (μs) tD3 (μs) Droplet volume (pL) DI water −10.0 10.0 −24.0 1.0 6.0 0.5 7.0 0.5 3.0 11.8 PEDOT −20.0 65.0 −50.0 1.0 2.0 0.2 3.0 0.5 1.0 14–25 Figure 11. PEDOT droplets ejected with the W-shaped waveform, which were captured at 10 μs intervals. parameters in the W-shaped waveform must be carefully fine- tuned otherwise instable ejection or no-ejection would occur. A slight increase in amplitude U could increase the inkjetting instability. For example, a slight adjustment of U3 with 0.1 V resulted in total instable jetting. 5. Discussions The unipolar waveform was the simplest one, which could lead to satellite formation and result in asymmetric droplet formation of inks with even moderate viscosity. This satellite droplet, a tiny droplet formed just after the main one, could cause the droplet out-of-print and affect resolution and quality of inkjetting. Therefore, it is obvious that reducing or eliminating the satellite droplet would effectively stabilize the ejected droplet volume. For Newtonian and low viscosity fluids, the induced acoustic pressure could have enough residual kinetic energy to eject a small volume of fluid that forms satellites after the formation of the main drop. This small droplet will be pulled by surface drag force and then join the main droplet. For Newtonian fluids, this satellite problem can be overcome by reducing the amplitude voltage of the driving waveform. As illustrated in table 3, a bipolar waveform was effective to minimize the droplet volume of Newtonian fluids like DI water. Hence, the bipolar waveform is sufficient to minimize the droplet volume of Newtonian fluids. For non-Newtonian fluids like polymeric inks, a relatively long and thin tail will be generated behind the main droplet due to its viscoelasticity. This tail will break down into a small cloud of fine satellites, which trail the main drop. The M-shaped waveform was then introduced in this study to minimize satellite droplets especially for PEDOT tailed droplets. The second square wave was introduced to generate a negative pressure pulse to pull the filament tongue back into the nozzle. The fine-tuning window of this second wave was very narrow which could not destruct the main droplet formation or have enough energy to eject another droplet. Figure 12. Comparison of individual droplets of PEDOT ejected on a silicon substrate. (a) By unipolar waveform, (b) by bipolar waveform, (c) by M-shaped waveform and (d) by W-shaped waveform. On the other hand, the W-shaped waveform was designed to enhance the satellite suppression. The idea of this waveform was to remove any former residual acoustic wave from the last ejection cycle, and subsequently a modified bipolar wave was adopted to cut off the filament tongue near the tip of the nozzle, this led to minimized satellite phenomenon. 6. Inkjet printing results The PEDOT droplets ejected on a silicon wafer by using four different waveforms were illustrated in figure 12. The optimal operating conditions for each waveform to obtain minimal volume of PEDOT droplets, as shown in tables 1–4, were employed. The dot diameters were decreased from 150 μm (by unipolar waveform) to less than 65 μm (by W-shaped waveform) when a 50 μm diameter nozzle was used. The operation of line formation by means of inkjetting a series of dots was demonstrated in figure 13. The dots were ejected with different pitches to study the merging effect caused by ink spreading. For a large dot pitch, a series of dots could not be merged to form a continuous line. However, if the dot pitch was too small, the dots would be overlapped and then created a larger line width than expected. Generally, a small ejected droplet generated a small dot on a substrate, which led to a fine line under a certain pitch. Therefore, as illustrated 7 J. Micromech. Microeng. 19 (2009) 055010 H Y Gan et al Figure 13. Comparison of line width of PEDOT inkjetted under different dot pitches using different waveforms on a silicon substrate. (a) By unipolar waveform, (b) by bipolar waveform, (c) by M-shaped waveform and (d) by W-shaped waveform. in figure 13, the fine line inkjet printing was achieved via the W-shaped waveform in PEDOT application. 7. Conclusions Volume reduction of droplet in inkjet printing was studied using different actuating waveforms. The conclusions from the experimental results were summarized as follows. (1) The inkjetting behavior using the simplest unipolar waveform showed that droplet volume varied obviously with dwell time and applied amplitude. (2) The M-shaped waveform, consisting of two unipolar waveforms, showed a significant reduction in droplet volume both for PEDOT and DI water with acceptable stability and consistency of droplet ejection. (3) The bipolar waveform demonstrated obvious reduction in droplet volume; the effect of droplet volume reduction was more significant while inkjetting Newtonian fluids like DI water. (4) The process window of the W-shaped waveform was relatively narrow because the ink performance was very sensitive to minute changes of operating parameters. However, the W-shaped waveform could reduce droplet volume in an order of 80% compared with the unipolar waveform, leading to a 50% reduction in inkjetted line width. (5) For inks with Newtonian or near Newtonian properties, the bipolar waveform was recommended due to its relative simplicity and inkjet quality. For non-Newtonian inks such as PEDOT, the M- or W- shaped waveforms with optimal tuned parameters were recommended due to the significant reduction in droplet volume. Acknowledgments The authors would like to express heartfelt gratitude to Mr Y N Liang and Dr C W Lu for their cooperation and to the Singapore Institute of Manufacturing Technology (SIMTech) for funding and support. References Basaran O A 2002 Small-scale free surface flows with breakup: drop formation and emerging applications AIChE J. 48 1842–48 Bidoki S M, Lewis D M, Clark M, Vakorov A, Millner P A and McGorman D 2007 Ink-jet fabrication of electronic components J. Micromech. Microeng. 17 967–74 Bird R B, Armstrong R C and Hassager O 1987 Dynamics of Polymeric Liquids 1 (New York: Wiley) Bogy D B and Talke F E 1984 Experimental and theoretical study of wave propagation phenomena in drop-on-demand ink jet devices IBM J. Res. Dev. 28 314–21 Chen A U and Basaran O A 2002 A new method for significantly reducing drop radius without reducing nozzle radius in drop-on-demand drop production Phys. Fluids 14 L1–4 Crispin X, Marciniak S, Osikowicz W, Zotti G, Denier van der Gon W, Louwet F, Fahlman M, Groenendaal L, De Schryver F and Salaneck W R 2003 Conductivity, morphology, interfacial chemistry and Stability of PEDOT J. Polym. Sci. B 41 2561–83 de Gans B J, Duineveld P C and Schubert U S 2004 Inkjet printing of polymers: state of the art and future developments Adv. Mater. 16 203–13 Gan H Y and Lam Y C 2008 Viscoelasticity Encyclopedia of Microfluidics and NanoFluidics (Berlin: Springer) pp 2147–55 Kim W H, Makinen A J, Nikolov N, Shashidhar R, Kim H and Kafafi Z H 2002 Molecular organic light-emitting diodes using highly conductive polymers as anodes Appl. Phys. Lett. 80 3844–46 Lee S-H, Shin K-Y, Hwang J Y, Kang K T and Kang H S 2008 Silver inkjet printing with control of surface energy and substrate temperature J. Micromech. Microeng. 18 075014 Mabrook M F, Pearson C and Petty M C 2005 An inkjet-printed chemical fuse Appl. Phys. Lett. 86 013507 Orme M and Smith R F 2000 Enhanced aluminum properties by means of precise droplet deposition ASME, J. Manuf. Sci. Eng. 122 484–93 Sakai S 2000 Dynamics of piezoelectric inkjets printing systems Proc. IS&T NIP 16 15–20 Shena M, Heller R A, Theriault T P, Konrad K, Lachenmeier E and Davis R W 1998 Microarrays: biotechnology’s discovery platform for functional genomics Trends Biotechnol. 16 301–6 Shore H J and Harrison G M 2005 The effect of added polymers on the formation of drops ejected from a nozzle Phys. Fluids 17 033104 Sirringhaus H, Kawase T, Friend R H, Shimoda T, Inbasekaran M, Wu W and Woo E P 2000 High-resolution inkjet printing of all-polymer transistor circuits Science 290 2123–6 8 1. Introduction 2. Experimental setup 3. Materials characterization 4. Effects of actuating waveforms 4.1. Unipolar waveform 4.2. M-shaped waveform 4.3. Bipolar waveform 4.4. W-shaped waveform 5. Discussions 6. Inkjet printing results 7. Conclusions Acknowledgments References