Study of Micro-Droplet Behavior for a Piezoelectric Inkjet Printing Device Using a Single Pulse Voltage Pattern

Transcription

1 Materials Transactions, Vol. 45, No. 5 (24) pp to 181 #24 The Japan Institute of Metals EXPRESS REGULAR ARTICLE Study of Micro-Droplet Behavior for a Piezoelectric Inkjet Printing Device Using a Single Pulse Voltage Pattern Hsuan-Chung Wu 1, Tzu-Ray Shan 1, Weng-Sing Hwang 1; * and Huey-Jiuan Lin 2 1 Department of Materials Science and Engineering, National Cheng Kung University, Tainan, Taiwan, R. O. China 2 Department of Materials Science and Engineering, National United University, Miaoli, Taiwan, R. O. China The aim of this study is to investigate the formation and ejection behavior of droplets created by a squeeze mode piezoelectric inkjet printing device using a single pulse voltage pattern. The test liquids are de-ionized (DI) water and ethylene glycol. The experimental results and acoustic wave theory are used to discuss the effects of operating frequency, positive voltage keeping time and pulse voltage magnitude on the volume and velocity of the droplets. For this study, a squeeze mode piezoelectric printhead is employed. By coordinating an LED flash with droplet ejection, a CCD camera could be used to capture images of the droplets at different points in the formation and ejection process. These images were then used to estimate the volume and velocity of the droplets. The experimental results are consistent with the propagation theory of acoustic waves. The maximum allowable pulse frequency in DI water and ethylene glycol are 15 Hz and 14 Hz respectively. If the positive voltage keeping time equals the time required for the acoustic wave to propagate through the printhead, optimal ejection behavior is achieved. As the pulse voltage increases, both the velocity and volume of the droplet become larger. (Received February 27, 24; Accepted April 6, 24) Keywords: inkjet, piezoelectricity, squeeze-tube type, micro-droplet, acoustic wave theory 1. Introduction The development of manufacturing technologies has been driven by the need for automation, miniaturization, and the reduction of costs and environmental impact. To meet these needs, inkjet printing technology is an increasingly attractive alternative for the distribution and patterning of material in a wide variety of applications. Other than its application in computer printers, the prospect of adopting inkjet printing technology in various high-tech processes is very promising. Solder-Jet technology 1) has been applied in ball grid array (BGA) and flip-chip electronics packaging processes. More than 1,44 tin/lead (63/37) solder balls, each approximately 7 mm in diameter, were successfully deposited onto a test substrate at 22 C using Solder-Jet technique. Compared to conventional soldering processes, Solder Jet-based processes not only reduce the number of processing steps but are environmentally friendly, numeric processor driven, and highly flexible. Hence, they can increase productivity and lower costs. Inkjet printing technology also has great potential in flat panel display applications. 2) Organic light-emitting diodes (OLED) represent one of the most promising flat panel technologies. However, due to the restriction in the selection of materials, they are usually fabricated by thin film deposition and evaporation process with subsequent patterning through lithography. These processes and the required masks are very complicated and expensive. Tremendous benefits can be obtained if arrays of organic light emitting materials can be deposited directly by inkjet printing methods. The current practice for the creation of polymer light emitting diodes (PLED) is to deposit the color filter through spin coating. However, this process is restricted to the fabrication of a single color. With the application of inkjet printing technology, it is possible to fabricate the color filters by depositing tiny pixels of red, green and blue elements onto the substrate. Compared to the spin coating process, inkjet printing technology is a simpler, lower cost process that can not only achieve a higher resolution and material utility rate, but delivers large panel production capability as well. In addition to the benefits mentioned above, inkjet printing technology provides a unique micro-lithography process for the fabrication of micro-lens arrays as well as complex threedimensional structures. 3) A variety of actuation methods have been adopted to eject droplets. They include piezoelectric, thermal bubble, electrostatic and acoustic methods. 4) Among these methods, piezoelectric and thermal bubble devices are the most mature and popular systems for commercial inkjet printers. The ejection of droplets from the nozzle is induced either by the displacement of a piezoelectric diaphragm that is in contact with the fluid as shown in Fig. 1 or by the formation of a Liquid PZT Spacer Nozzle Plate Jet Nozzle Target Diaphragm *Corresponding author, wshwang@mail.ncku.edu.tw Fig. 1 A schematic diagram of a piezoelectric inkjet head.

2 Study of Micro-Droplet Behavior for a Piezoelectric Inkjet Printing Device Using a Single Pulse Voltage Pattern 1795 vapor bubble in the ink through the heating of a resistive film. Since piezoelectric inkjet printing systems have no need to vaporize the fluid, they can be used for the ejection and dispensing of polymers and liquid metals. The electronics manufacturing industry very much desires the capabilities offered by piezoelectric inkjet printing technology. As a framework for the experimental study of droplet ejection in inkjet printing, Bugdayci et al. 5) used the theories of stress and energy conservation to calculate the relationship between the voltage applied to the piezoelectric transducer (PZT) and the induced pressure for a squeeze type piezoelectric printhead. Bogy and Talke 6) calculated the variation of pressure at the nozzle inlet with time, which was caused by the contraction and expansion of the PZT for a similar device. In that study, experimental observations were compared with the theory using a voltage pulse too weak to eject a droplet from the nozzle. Shield 7) used both experimentation and twodimensional numerical simulation to study the ejection behavior of droplets of DI water and ethylene glycol. The simulation results were compared with the experimental results. However, the effects of the experimental conditions were insufficiently discussed. Chen 8) investigated pressure generation in the ink flow channel and ink droplet formation for a piezoelectric printhead. ANSYS and the characteristic method were used to solve the one-dimensional wave equation to obtain the transient pressure and velocity variations in the flow channel of the printhead. Print quality is known to be closely related to the final behavior of the ejected droplet impinging onto the substrate. This behavior is affected by the physical properties, volume, and velocity of the droplet. The aim of this study is to investigate the effects of droplet ejection behavior (including droplet volume and velocity) under different experimental conditions and to discuss these effects in terms of acoustic wave theory. 2. Experimental Method 2.1 Acoustic wave theory A squeeze mode piezoelectric printhead is employed in this study. As shown by the cross-sectional schematic diagram in Fig. 2, it is nearly cylindrical in shape. According to Bogy s 6) acoustic wave theory, the end at the nozzle end Open End Cavity Closed End Nozzle L Fluid Supply PZT Orifice Fig. 2 A schematic diagram of a squeeze mode piezoelectric inkjet printhead. Fig. 3 Voltage V t 1 t 2 t 3 t rise t dwell t fall Time Voltage pattern of a single pulse employed in this study. can be considered to be closed since the nozzle opening is a small fraction of the tube cross-section area. The supply end can be considered to be open since the inside diameter of the supply tube is considerably larger than the inside diameter of the squeeze tube. Therefore, a pressure wave retains its sign when it is reflected from the nozzle end (closed end) and it changes sign when it is reflected from the supply end (open end). A single pulse voltage pattern, imposed on the PZT as shown in Fig. 3, is employed in the study. t rise (¼ t 1 ) is the time required for the voltage to rise from zero to V. The polarization causes the PZT to move outward in the radial direction, which results in a pressure drop in the liquid contained in the cavity. The positive voltage V is maintained for the duration t dwell (¼ t 2 t 1 ). The PZT has no displacement during this time. However, the pressure generated from t ¼ t rise is propagated in the cavity. t fall (¼ t 3 t 2 ) is the time required for the voltage to drop from V to zero. The depolarization causes the PZT to move radially inward and results in an increase in pressure. After t 3, the PZT exerts no influence on the liquid in the cavity. The expansion and contraction of the inner radius of the PZT, as a result of the applied single pulse voltage, causes the propagation of a pressure wave within the printhead. The wave propagation schematic is shown in Fig. 4. The left end, near the fluid supply, is considered open and the right end, near the nozzle, is considered closed. The pressure values of the positive and negative acoustic waves are relative values to the atmospheric pressure. A negative acoustic wave is represented beneath the horizontal line, and a positive one is represented above the horizontal line. t opt, defined as the period of time required for an acoustic wave to propagate through a cavity of length L, is calculated according to the following equation. t opt ¼ L=v aco,liq ð1þ where v aco,liq is the velocity of acoustic wave propagation in the liquid. The velocities in DI water and ethylene glycol are v aco,wat ¼ 1435 m/s and v aco,eth ¼ 1658 m/s (6) respectively. Figure 4 shows the sequence of expected pressure pulse propagations and reflections. The polarization resulting from the increase in voltage from zero to V causes the PZT to move radially outward, resulting in a negative pressure in the liquid from t ¼ to t ¼ t 1 as shown in Fig. 4(a). The pressure wave is split in two and the split waves propagate in opposite directions (toward the open end and the closed end) with the

3 1796 H.-C. Wu, T.-R. Shan, W.-S. Hwang and H.-J. Lin Open End L Closed End (a) 1 5 Relative Pressure (Pa) (b) (c). 5t opt Time (t opt) -1 5 (d) 1.t opt Fig. 5 Pressure-time relation at nozzle inlet for a single electric pulse. (e) (f) (g) (h) (i) (j) 1.5t opt 2.5t opt 3.5t opt 4.5t opt 5.5t opt results in the separation of the ejected droplet from the nozzle. Note that the next large positive pressure wave arrives at the nozzle at 5:5t opt. If this positive pressure wave has enough energy, it will eject another droplet. The pressuretime relationship at the nozzle inlet is shown in Fig. 5. Both the energy dissipation due to the viscosity of the fluid and the reflection of the wave at the open end are sources of energy loss in the system, causing the amplitude of the pressure wave decrease over time. The higher viscosity of the liquid results in greater damping effect. The kinetic energy carried by the ejected droplet also causes the system to lose its energy, which hastens the return of the fluid in the nozzle to its undisturbed state. (k) 6.5t opt Fig. 4 Schematic illustrations of wave propagation and reflection under an initial pulse of pressure pulse with t dwell ¼ t opt in an open-closed squeeze tube. amplitude of half the initial pressure drop. The semi-ellipse shape is adopted to represent the pressure wave, as shown in Fig. 4(b). The pressure waves are reflected from the two ends as shown in Fig. 4(c). The dotted lines represent the pressure waves before reflection and the solid lines represent the pressure waves after reflection. The depolarization resulting from the voltage drop from V to zero causes the PZT to move radially inward, resulting in a positive pressure in the liquid from t ¼ t 2 to t ¼ t 3.Ift dwell ¼ t opt, it means the generation of the positive pressure wave coincides with the arrival of the two initial pressure waves after each has traveled a distance L as shown in Fig. 4(d). At this instant, the positive pressure wave is also split in two. The result is that the left propagating negative pressure wave is annihilated and the right propagating pressure wave is doubled, as shown in Fig. 4(e). By applying the voltage down-step at exactly t opt, the amplitude of the pressure wave is thus optimally enhanced. Fig. 4(f) shows the expected double-amplitude pressure wave arrives at the nozzle end. If the pressure is high enough, the liquid is ejected. In the absence of damping, if no other voltage steps are applied, the later sequence of pressure waves will occur as shown in parts g k of Fig. 4. At 3:5t opt, the negative pressure at the nozzle is maximized as shown in Fig. 4(h). This negative pressure wave produces the pulling force that 2.2 Experimental setup A schematic diagram of the experimental setup employed in this study is shown in Fig. 6. The drive electronics box includes a strobe driver and pulse controller. The strobe driver controls the delay time of the LED. The pulse controller controls the pulse voltage, pulse width and frequency output to the printhead. The jetting device consists of a printhead and PZT. The printhead is surrounded by metal for protection. The jetting device is connected to a fluid supply chamber and the drive electronics box. The schematic diagram and dimensions of the jetting device are shown in Fig. 2 and Fig. 7. A CCD camera is used to observe and record the evolution and movement of the droplet. The control system takes input data, which represents the operating conditions of the inkjet printing device, from the controllers and outputs the delay time to the strobe driver. A lens is used to project the image of the droplets onto a CCD. When a photon strikes a pixel of the CCD, a small electrical charge is generated and stored. Pixels Control System (PC) Fig. 6 CCD Camera Drive Electronics Box Fluid Supply Jetting Device LED A schematic diagram of the experimental setup.

4 Study of Micro-Droplet Behavior for a Piezoelectric Inkjet Printing Device Using a Single Pulse Voltage Pattern cm 711µ m (A) 89µ m Flying Distance (B).32cm 4µm Diameter (C) Fig cm 711µ m Geometry and dimensions of the inkjet printhead device. Table 1 Surface tension coefficient, density, viscosity and acoustic velocity of DI water and ethylene glycol. D v aco (N/m) (g/cm 3 ) (Ns/m) (m/s) DI water Ethylene glycol that are more intensely illuminated receive more photons and thus develop larger charges. In this way the image projected onto the CCD can be captured. The delay time of the LED flash can be controlled to capture the stages of droplet formation and flight. The images are displayed on a monitor and stored on the control system. Two types of liquid, DI water and ethylene glycol, are employed in this study. DI water is a high surface tension, low viscosity liquid. Ethylene glycol is a high surface tension, high viscosity liquid. The surface tension coefficients, densities, viscosities and acoustic velocities of these liquids are shown in Table 1. Note that the viscosity of ethylene glycol is significantly higher than that of DI water. 2.3 Setup of experimental condition At the beginning of the experiment, the input voltage pattern and LED delay time are input to the controller. Though a single pulse or double pulse voltage pattern can be selected, only the single pulse pattern, as shown in Fig. 3, is employed in this study. After determining the pulse type, the voltage magnitude (V), t rise, t dwell, t fall and pulse frequency are all entered into the controller. An insufficient t rise or t fall will not allow the controller to output the full desired voltage and thus the PZT will not reach the desired expansion. Therefore, the voltage pulse output by the pulse controller to the printhead must meet with the requirements given below. V V 15 and 15 ð2þ t rise t fall That is to say, the slope of voltage variation can not be more than 15 volt/ms. Fig. 8 Measurements of diameter and flying distance of the liquid droplet from an experimental observation. The experimental variables in this study are the operating frequency, the positive voltage keeping time and the voltage magnitude. The operating frequency is defined as the number of voltage pulses per second. Since t rise and t fall are governed by the target voltage (according to equation 2), t dwell is the only variable for the pulse time. The voltage magnitude must be high enough to cause droplet ejection, but not so high that the ejection becomes chaotic and difficult to observe. In order to represent the experimental conditions conveniently, the notation of t rise t dwell t fall voltage frequency is used. That is to say, voltage rise time(ms) positive voltage keeping time(ms) voltage fall time(ms) voltage(v) operating frequency(hz). 2.4 Measurement of droplet volume and flight distance Figure 8 illustrates how the droplet diameter and flight distance are determined from the experimental observations. The width of the glass tube underneath the nozzle (A) is 711 mm. The flight distance of the droplet can be calculated according to the relation; ðb=aþ711 mm. In order to measure the droplet volume, an image in which the droplet is spherical is selected. The diameter can be calculated using the equation ðc=aþ711 mm, and the liquid droplet volume can be calculated using ð4r 3 Þ=3, with r being the radius of the liquid droplet. The droplet velocity is determined by measuring the flight distances B at time t and B at time t, and is calculated by the following equation. 3. Results and Discussion v ¼ B B t t 3.1 Operating frequency effects Table 2 shows the experimental conditions used to discuss the effects of operating frequency on droplet ejection behavior for DI water and ethylene glycol. The DI water operating conditions are frequency. As mentioned in Section 2.3, these parameters represent the experimental conditions of t rise ¼ 1:5 ms, t dwell ¼ 15 ms, t fall ¼ 1:5 ms and voltage ¼ 17 V with various frequencies(5, 1, 15, 2, 25, 3, 35 and 4 Hz). The other ð3þ

5 1798 H.-C. Wu, T.-R. Shan, W.-S. Hwang and H.-J. Lin Table 2 Parameters used to test frequency. Liquid DI water Ethylene glycol Experimental Conditions Control variables t rise t dwell t fall voltage frequency Frequency (Hz) frequency 5, 1, 15, 2, 25, 3, 35, 4 1, 15, 2, 3, frequency 4, 8, 12, 14, 16 Table 3 Parameters used to test t dwell. Experimental conditions Control variables Liquid t rise t dwell t fall voltage frequency pulse width-t dwell (ms) 8, 1, 12, 14, DI water 1.5 t dwell , 18, 2, 22, 24 Ethylene 7, 8, 9, 1, 11, 12, 13, 3 t dwell glycol 14, 15, 16, 17, 18, 19, 2 (a) DI water with t rise =1.5, t dwell =15, t fall =1.5, voltage=17 conditions, with frequency varying from 5Hz to 4,Hz. Table 4 Parameters used to test voltage. Liquid Experimental conditions Control variables t rise t dwell t fall Voltage Frequency voltage (V) DI water voltage 15 16, 17, 18 Ethylene glycol voltage 15 37, 39, 41 experimental conditions given in Tables 2, 3 and 4 are represented in the same manner. t ¼ is defined as the time of pulse initiation. Figs. 9(a) and (b) show the images taken at t ¼ 15 ms of DI water and ethylene glycol droplets formed at various pulse frequencies. The droplet morphologies for DI water are nearly identical at frequencies within 15 Hz, as shown in Fig. 9(a). The droplet morphologies for ethylene glycol are also very similar at frequencies within 14 Hz, as shown in Fig. 9(b). An excessive pulse frequency results in chaotic droplet ejection. This is because the previous pressure wave does not have time to decay fully and, therefore, interacts with the pressure wave produced by the next voltage pulse. In contrast, lower frequencies result in a lower droplet ejection rate, causing decreased production efficiency. In order to optimize efficiency, the operating frequency should be as high as possible without causing chaotic droplet ejection. Therefore, the optimum frequencies for DI water and ethylene glycol are 15 Hz and 14 Hz, respectively. As mentioned above, the decay time of the pressure wave determines the optimal operating frequency of the inkjet system. The decay time associated with DI water is approximately 667 ms under ,5 conditions. The decay time of the pressure wave in ethylene glycol is approximately 83 ms under conditions. The decay time in ethylene glycol is shorter than that in DI water because the viscosity of ethylene glycol is substantially higher resulting in greater damping. Thus, a higher pulse frequency can be used without the problems mentioned earlier. The delay time can be calculated using the equation below. (b) Ethylene glycol with t rise =3, t dwell =13, t fall =3, voltage=4 conditions, with frequency varying from 1,Hz to 16,Hz. Fig. 9 Experimental observations of droplet with various frequencies at 15 ms. (a) DI water, (b) ethylene glycol. 1 t decay ¼ frequency opt 3.2 Pulse time effects Table 3 shows the experimental conditions used to discuss the effects of positive voltage keeping time (t dwell ) on droplet ejection behavior for DI water and ethylene glycol. Figure 1(a) shows the images taken at t ¼ 15 ms of DI water droplets formed using various values for t dwell. The droplet morphologies for DI water are approximately identical, but the velocities are different for t dwell equal to 14, 16 and 18 ms under the conditions 1.5 t dwell Only one low velocity droplet is ejected when t dwell equals 12 or 2 ms. There is no droplet ejection if t dwell is above 2 ms or below 12 ms. Figure 1(b) shows the images taken at t ¼ 2 ms of ethylene glycol droplets formed using various t dwell values. The droplet morphologies for ethylene glycol are also similar, but the velocities are different for t dwell times between 9 and 16 ms under the conditions 3 t dwell A droplet is not ejected if t dwell is above 16 ms or below 9 ms. The relationships between average droplet velocity and positive voltage keeping time are shown in Fig. 11. The velocity variations with respect to t dwell for DI water and ethylene glycol follow the same trend from to 15 ms. ð4þ

6 Study of Micro-Droplet Behavior for a Piezoelectric Inkjet Printing Device Using a Single Pulse Voltage Pattern 1799 (a) Open End L Closed End (b) (a) DI water with t rise =1.5, t fall =1.5, voltage=18, frequency=1,5 conditions, with t dwell varying from 8µs to 24µ s. (c) (d) (e) 1. 5t opt 1.25t opt (f) 1.5t opt (g) 2.5t opt (b) Ethylene glycol with t rise =3, t fall =3, voltage=45, frequency=1,5 conditions, with t dwell varying from 7µs to 2µ s. Fig. 1 Experimental observations of droplet with various t dewll. (a) DI water, (b) ethylene glycol (h) (i) (j) 3.5t opt 4.5t opt 5.5t opt 4 35 Glycol DI Water (k) 6.5t opt Velocity (cm/s) Dwell time (us) Fig. 11 Average droplet velocity in the first 15 ms for DI water and ethylene glycol under various dwell times. The droplet velocity increases with t dwell to a maximum value and then decreases as t dwell increases further. The maximum droplet velocity for DI water occurs at t dwell ¼ 14 ms and that for ethylene glycol occurs at t dwell ¼ 12 ms. As mentioned in Section 2.1, the pressure distribution within the printhead is greatly affected by the chosen t dwell under optimal conditions. The maximum positive pressure at the nozzle occurs at 1:5t opt. It is this maximum pressure that determines the droplet velocity. If t dwell is selected randomly it is very unlikely that this maximum positive pressure will be achieved. The pressure wave distribution within the printhead is more chaotic, leading to decreased droplet ejection velocity. The pressure wave distribution for t dwell ¼ 1:25t opt is shown in Fig. 12. The positive pressure resulting from the PZT s inward motion propagates in opposite directions. This Fig. 12 Schematic illustrations of wave propagation and reflection under an initial pulse of pressure pulse with t dwell ¼ 1:5t opt in an open-closed squeeze tube. positive pressure wave is at a distance of L/4 from the initial pressure created by the PZT s move outward motion. Thus, the pressure waves will not constructively interfere, as shown in Figs. 12(d) (e). Afterward, the pressure waves retain their sign if they reflect from the closed end and change sign if they reflect from the open end as shown in Figs. 12(f) (k). Comparing the positive pressure at the nozzle in Fig. 12(f) at t ¼ 1:5t opt with that in Fig. 4(f), the positive pressure magnitude is reduced by half, and thus the ejected droplet velocity is lower. The t opt of DI water and ethylene glycol can be calculated from the printhead length L (2.235 cm) shown in Fig. 7 and eq. (1). The t opt of DI water is t opt,wat ¼ 2:235(cm)=1435(m/s) 15:57 ms. The t opt of ethylene glycol is t opt,eth ¼ 2:235(cm)=1658(m/s) 13:48 ms. Either DI water or ethylene glycol, the theoretical calculation of t opt differs only slightly from the experimentally determined optimal t dwell by approximately 1.5 ms. This is because the voltage rising time must be considered in the experiment. Therefore, setting the positive voltage keeping time equal to t opt, creates the optimum condition and maximizes droplet velocity. 3.3 Voltage effects Table 4 shows the experimental conditions used to discuss the effects of voltage magnitude on droplet ejection behavior

7 18 H.-C. Wu, T.-R. Shan, W.-S. Hwang and H.-J. Lin for DI water and ethylene glycol. The voltage magnitudes used to investigate DI water were 16, 17 and 18 volts at a pulse frequency of 15 Hz. Ethylene glycol was tested at 37, 39 and 41 volts at a pulse frequency of 15 Hz. As mentioned in Section 3.1, the droplet morphologies for ethylene glycol are approximately identical at frequencies under 14 Hz. Thus, the 15 Hz frequency was adopted for ethylene glycol as well as for DI water. Figure 13 shows the droplet volume variation with respect to voltage for DI water. The three droplet diameters and volumes are taken from images of the spherical droplet at t ¼ 2 ms. The droplet diameters are 71.63, and mm respectively. The droplet volumes are , and pl respectively. As the voltage increases, the droplet diameter and volume both become larger. Figure 14 shows the variations of main droplet velocity as measured at the front of the droplet with respect to time. As the voltage increases, so does the droplet velocity. The maximum velocity in all cases occurs at 35 ms. Figure 5 shows that the pressure at the nozzle starts to become negative at 2:5t opt. Also from the experimental results of Section 3.2, the t opt,wat of DI water is approximately 14 ms. Thus, the value of 2:5t opt,wat is approximately equal to 35 ms. After 35 ms, the negative pressure starts to act on the droplet, which has still not separated from the orifice, and results in a decrease in droplet velocity. From this, the consistency between the experimental results and the theory can be seen. The droplet separates from the nozzle at nearly 65 ms, after which the droplet is only affected by gravity and the velocity slowly increases. Figure 15 shows the droplet volume variation with respect to voltage for ethylene glycol. The three droplet diameters and volumes are taken from images of the spherical droplet at t ¼ 2 ms. The droplet diameters are 57.95, 6.58 and mm respectively. The droplet volumes are 11.89, and pl respectively. As the voltage increases, the droplet diameter and volume both grow. Figure 16 shows the main droplet velocity variations with respect to time. As the voltage increases, so does the droplet velocity. The maximum velocity in all cases occurs at 3 ms. From the experimental results in Section 3.2, the t opt,ehy of ethylene 135 Volume (pl) Volume (pl) Volts (V) Fig. 13 Volume of main droplet for DI water at 2 ms under various voltages Voltage (v) Fig. 15 Volume of main droplet for ethylene glycol at 2 ms under various voltages V 17 V 16 V V 39 V 37 V Velocity (m/s) Velocity (m/s) Time (us) Fig. 14 Velocity of main droplet as functions of time for DI water under various voltages Time (us) Fig. 16 Velocity of main droplet as functions of time for ethylene glycol under various voltages.

8 Study of Micro-Droplet Behavior for a Piezoelectric Inkjet Printing Device Using a Single Pulse Voltage Pattern 181 glycol is approximately 12 ms. Thus, the value of 2:5t opt,wat is approximately equal to 3 ms. After 3 ms, the negative pressure starts to act on the droplet, which has still not separated from the orifice, and results in a decrease in droplet velocity. Again, the consistency between the experimental results and the theory can be seen. The droplet separates from the nozzle at nearly 12 ms. Afterward, the droplet velocity oscillates within small range. The velocity variation of ethylene glycol is smaller than that of DI water. This is due to the fact that ethylene glycol has greater viscosity. The selection of voltage magnitude should take into consideration the trade-offs between droplet velocity and volume. Increased voltage results in greater velocity, but also in larger droplet size, effectively reducing the print resolution. Increased resolution can be gained by using lower voltages at the expense of droplet velocity. The importance of these variables will depend greatly on the specific application. 4. Summary In this study, the effects of operating frequency, positive voltage keeping time and voltage magnitude, on droplet ejection behavior of single pulse voltage pattern microdroplet inkjet devices were investigated. Two types of liquid, DI water and ethylene glycol, were used in the experiments. The conclusions taken from this experiment are as follows. (1) The droplet ejection behaviors can be successfully explained through the propagation theory of acoustic waves. (2) The maximum operating frequencies for DI water and ethylene glycol are 15Hz and 14Hz, respectively. (3) When the positive voltage keeping time is equal to the time required for the pressure wave to propagate through the printhead, the velocity of the ejected droplet is maximized. (4) As the voltage increases, the velocity and volume of the liquid droplet become larger. Acknowledgements This work has been supported by the National Science Council in Taiwan (NSC E-6-55), for which the authors are grateful. REFERENCES 1) Q. Liu: J. Mater. Proc. Tech. 115 (21) ) T. R. Hebner, C. C. Wu, D. Marcy, M. H. Lu, J. C. Strum: Appl. Phys. Lett. 72 (1998) ) K. Yamaguchi, K. Sakai, T. Yamanaka, T. Hirayama: Precision Eng. 24 (2) ) L. P. Hue: J. Imaging Sci. and Tech. 42 (1998) ) N. Bugdayci, D. B. Bogy, F. E. Talk: IBM J. Res. Dev. 27 (1983) ) D. B. Bogy, F. E. Talk: IBM J. Res. Dev. 28 (1984) ) T. W. Shield, D. B. Bogy, F. E. Talke: IBM J. Res. Dev. 31 (1987) ) P. H. Chen, H. Y. Peng, H. Y. Liu, S. L. Chang, T. I. Wu and C. H. Cheng: Int. J. Mech. Sci. 41 (1999)