Fabricated Devices
The SEM images in Figs. 4(a ~ b) are taken from a fabricated microgripper, showing the uniform and smooth curling of all the four actuators due to residual stresses. In addition, four actuators form a spherical space after release, which is suitable for capturing and holding micro-objects. Devices with bimorph beams arranged at a 45-degree angle direction (as shown in Figs. 4(c ~ d)) and along the latitude direction (as shown in Figs. 4(e ~ f)) have also been fabricated. Notably, despite the three actuators having the same contour shape, different orientations of the bimorph beams result in significantly different deformation states. A reasonable explanation is that the orientation of the bimorph beams determines the direction of the residual stress acting on the actuator and control the actuators’ stiffness along the direction of the bimorph beams, thereby influencing the structure state after release. Figure S1 shows the images of micro-grippers arrays, which have good structural consistency for each unit. It is expected that the microgripper array will be useful for tasks requiring parallel control and manipulation of multiple targets simultaneously.
Actuation characteristics of the microgripper
After device fabrication, the microgripper is fixed on a PCB board to test its characteristics, where the electrical connection between the device and the PCB board is achieved by wire bonding. A DC power supply is used to provide the driving voltage and the resistance of each actuator is measured to be approximately 320 Ω. Figure 5 (a)(I) shows the profile of the bending angle of the microgripper at different applied voltages. Initially, the microgripper is in the closing state without power input. When the driving voltage is set from 1 V to 5 V, the bending angle of the actuator gets larger with the increase of the driving voltage in an approximately linear trend (between 1.5 V to 4.5 V). When the voltage reaches 5 V, the actuator reaches a fully expanded state and the opening angle of the tip exceeds 100 degrees. Figures 5(a)(Ⅱ~Ⅶ) show the deformation process of an actuator from 0 V to 5 V. Note that the actuator can precisely bend to and maintain a desired angle by finely modulating the driving voltage. As mentioned before, the four actuators can be controlled individually and independently to meet the specific capturing requirements. Figures S2(a ~ e) illustrate the results where the four actuators of the microgripper are driven in sequence at 5 V. Moreover, the simultaneous activation of all the four actuators is also feasible. Figures S2(f ~ j) depict the deformation procedure of a microgripper while the driving voltage on all the actuators ranges from 1 V to 5 V. It is estimated that the breakdown voltage of the actuator is around 8 V.
The mean operational temperature of one actuator is closely monitored utilizing the temperature coefficient of resistance (TCR). The detailed measurement procedure of TCR is elucidated in Figure S3. The TCR value of Pt is firstly quantified via a temperature-regulated chamber to be approximately 0.0026/℃. After that, by monitoring the alterations in the resistance of Pt at different driving voltages, the mean temperature of an actuator can be obtained. The correlation between the average working temperature and applied voltage is depicted in Fig. 5(b). Upon reaching a 5 V driving voltage, the average actuator temperature stays at approximately 220°C. The thermal distribution of the actuators is also characterized using an infrared thermal imaging system, as shown in the inset in Fig. 5(b). Notably, due to the absence of a thermal isolation structure between the actuator and the substrate, heat concentration is observed primarily at the actuator tips, while the root-part remains inadequately heated, thus exhibiting significant temperature gradients on the actuator. Furthermore, the power consumption of a single actuator is assessed through concurrent measurements with an ammeter and a voltmeter, and the results are presented in Fig. 5(c). Specifically, in our experimental setup, each actuator consumes approximately 50 mW when operating at the fully opening state, much smaller than previously reported work [33].
To achieve a fast capturing, the response speed of the actuator is a key parameter. In this work, both the thermal response time and mechanical response time are measured and quantified. Temperature changes of the actuator over time are monitored by an infrared thermal camera. In the tests, a continued square wave signal is applied on one of the actuators. The peak-to-peak voltage value is 5 V, and the frequency is set at 1 Hz. The temperature evolution over time is shown in Fig. 5(d). The enlarged views of the rise and fall periods are also provided. Consequently, the thermal response time is measured to be approximately 8 ms. Besides, the mechanical response time is assessed using a high-speed camera, capturing images at a rate of 2000 frames per second (f/s). Experimental results show that the mechanical response time is approximately 10 ms. Figure S4 displays six typical states during the opening process. Additionally, the detailed deformation process of the microgripper is shown in Supplementary Movie S1, where the playback speed is 1/30 of the actual speed.
Capturing Experiment
A specialized capturing experiment is conducted to assess the functionality of the microgripper. A schematic illustration of the setup and the test is shown in Fig. 6(a). Initially, the microgripper is attached onto a PCB board using UV glue, with four actuators electrically connected to the PCB board via gold wires. Subsequently, a PMMA microbead with a diameter of 500 µm is adhered to a plastic wire by electrostatic force. The mass of each microbead is measured to be 0.12 mg and the plastic wire is fixed to make the ball static. Then the microgripper is positioned on an XYZ stage with a movement accuracy at micron level.
Figures 6(b ~ g) show the detailed manipulation process of the microgripper. First, the platform moved upward to bring the device close to the microbead (Fig. 6(b)). Then, DC voltages are applied to the four actuators and gradually increased to ensure the opening is large enough to wrap the microbead, and then precisely control the position of the platform within the plane so that the microgripper can accurately touch and capture the microbead (Figs. 6(c ~ d)). Next, the platform is further lifted precisely to achieve an accurate contact between the device and the microbead (Fig. 6(e)). The applied voltages are then gradually diminished until zero, closing the actuators accordingly under the influence of residual stresses (Fig. 6(f)). Finally, the platform moves downward. As depicted in Fig. 6(g), the device successfully captures the microbead and takes it from the plastic holder. Figures 6(h ~ i) show the SEM images of the microgripper after capturing. Despite that only the tips of the actuators contact with the surface, the captured microbead is effectively grasped by the microgripper. In addition, we conducted more experiments aiming to capture multiple targets using a single microgripper, and the results are presented in Figure S5.
Reliability test
Vibration and impact tests are conducted to evaluate the grasping strength and reliabliy of the microgripper. These tests are very important for the microgripper when it comes to practical use. Figures 7(a ~ b) illustrate the initial capturing scenarios of two kinds of PMMA microbeads with diameters of 500 µm and 400 µm, respectively. In these cases, two different wrapping modes including half wrapping (defined as Type-A) and full wrapping (defined as Type-B) are observed. Then, the ability of the devices to resist physical vibration is evaluated by a vibration test system (SignalCalc 901 DP), as shown in Fig. 7(c). The vibration of the platform is triggered by a sinusoidal signal with varying frequency and amplitude. Following the JESD22-B103B standard30 [34], the vibration frequency is set to sweep from 20 to 2000 Hz and then back to 20 Hz, which is increased on a logarithmic scale with the duration set to 12 min. For each frequency sweep, the acceleration amplitude is incrementally raised by 1 g until the captured-ball bounces off from the microgripper. Throughout the entire vibration test, the status of the microgripper can be directly observed by naked eyes. The microgripper is affixed to the platform using a double-sided adhesive tape, where the platform has only one freedom of movement in the Z direction as depicted in Fig. 7(d). Figure 7(e) displays the results of the vibration tests. For Type A, the maximum endurable acceleration is approximately 5 g, whereas it is about 35 g for the Tpye-B. A detailed failure process of Type A is shown in supplementary Movie S2. The condition of the microgripper is monitored both before and after the vibration test, and the results are depicted in Figure S6. Despite the ball detaching from the microgripper due to vibration, the overall shape and functionality of the device remain unaffected. That is, under the influence of residual stress, the actuators are capable of recovering to initial closing state. Furthermore, upon the reapplication of the excitation signal to the microgripper, the actuators can resume normal operations.
Figures 7(f ~ g) shows the setup of the impact test. Two types of the microgripper are respectively affixed to the surface of the lifting platform using a double-sided tape. A precise control over the impact experiment's height is achieved utilizing a control system. The lifting platform is equipped with an acceleration sensor, enabling real-time monitoring of the impact acceleration during descent and collision via a computer interface. The equipment is capable of generating a maximum acceleration of 1600g, corresponding to a drop height of 1.3 meters on the lifting platform. In each impact test, the drop height is incremented by 5 cm until the microbead falls from the microgripper. The test results are shown in Fig. 7(h). For Type A, the maximum endurable acceleration is approximately 600g, while for Type B, the ball remains securely grasped by the microgripper even under an impact acceleration of 1600g. We also evaluated the structural change of the microgripper after the impact tests, and the results are shown in Figure S7. Figures S7(a ~ c) illustrate the morphology of the device for Type A after the tests, where the microgripper does not revert to its initial closing state. A reasonable assumption would be: upon experiencing an instantaneous impact, the PMMA microbead generates a strong collision with the actuators. Given that the entire microbead is constrained solely by the tips of the actuators, external impact readily breaks the constraint. Additionally, the impact of the PMMA microbead induces a certain degree of irreversible deformation of the actuators. In contrast, when the PMMA microbead is fully wrapped by the microgripper, our study indicates that the influence of the microbead on the microgripper are small and with little effect on the residual stress of the actuators. Consequently, the microgripper remains a closing state, merely indistinguishable from its initial state upon release. Subsequent to the impact, the captured microbead slightly shifts its position within the microgripper, as illustrated in Figures S7(d ~ f). These results demonstrate the superior performance of the microgripper to withstand strong vibration and physical impact.