2.1 Design of multifunctional MDTs
Our transport platform is simply fabricated by closely sandwiching ferrofluid droplets between a hybrid wetting surface (upper plate) and a superhydrophobic surface (lower plate) under an external magnetic field (Fig. 1a). Here, aluminum-based hydrophilic/hydrophobic substrates are prepared by laser ablation and low surface energy modification (Fig. S1, S2, Supplementary Information). The magnetic field can guide the movement of the ferrofluid droplets, which allows the movement of the upper plate, thus enabling contactless transportation control of the objects. For the bottom system, the shape and the movement of ferrofluid droplet can be easily controlled by an external magnetic field on a superhydrophobic surface (Fig. 1b). For the top hybrid wetted surface, the ferrofluid droplets can be firmly trapped by the superhydrophilic patterned regions due the strong adhesion between the patterned regions and such droplets, whereas the opposite is true for the superhydrophobic surfaces, allowing the droplets to easily roll off the surfaces (Fig. 1c). By aligning the patterned region of the upper plate with the ferrofluid droplets and gently placing it down, the upper plate self-assembles with the ferrofluid droplets due to its self-weight and the attraction of the superhydrophilic patterned regions to the ferrofluid droplets (Fig. 1d). Interestingly, if cleverly design the array and structure of such self-assembled structures (FDSS) based on individual ferrofluidic droplets, the unique capabilities of MDTs consisting of four FDSSs for cargo transportation and energy conversion can be realized. The MDTs can not only deliver cargo on plane, but also transfer on narrow and long tubular structures, with high transportation velocity and accuracy (Fig. 1e). Furthermore, the MDTs can realize energy conversion from mechanical energy to electricity based on Faraday’s law of electromagnetic induction, and the output electrical signals in turn can be used to monitor its movement (Fig. 1f).
2.2 Conceptualization of FDSS and its mechanical stability and wettability
Superhydrophobicity is common in nature, and the superelasticity and low sliding resistance usually exhibited by droplets on their surfaces. Through clever surface design, the kinetic energy of the droplets can be converted into other forms of motion and energy output30,31. For example, two superhydrophobic structures can be used to drive the movement of a droplet through compression and release (model 1, 2, Fig. S3a, Supplementary Information). Here, two fluids are used for comparison, a water droplet and a ferrofluid droplet. However, it is not possible to achieve directional transport through droplet manipulation. Accordingly, structures with hybrid wettable surfaces can be manipulated in a restricted domain for droplets (Model 3, 4, Fig. S3a, Supplementary Information), which is expected to enable droplet motion control. Compared to Model 3, Model 4 is prone to failure when an external force is applied, in which case the minimum gap (hmin) between the upper and lower plates after compression is maintained at 0.2 mm. Besides, structures (Model 3) that are superhydrophobic on one side and hybrid wettable surfaces on the other are more resistant to compression than dual superhydrophobic surfaces (Model 1 & 2) and dual hybrid wettable surfaces (Model 4), as shown in Fig. S3b, Video S1, Supplementary Information. Thus, Model 3 is expected to realize the restricted domain control of droplets for transport.
Compared to water droplets, ferrofluid droplets are more controllable by external magnetic field, especially on super-sliding surfaces32. When the ferrofluid droplets is subjected to an external magnetic field, it is automatically shaped and its stiffness is increased. Therefore, based on model 3, we propose a hybrid model (model 5, the aforementioned FDSS) with capillary adhesion and magnetic actuation that can easily achieve rapid recovery after extrusion (Fig. S3a, Supplementary Information). Compared to Model 3, FDSS has greater stress resistance and recovery. When both models are subjected to external pressure and compressed to their limit positions (hmin <0.1 mm), the adhesive forces are unable to maintain the confinement of the droplets, causing the droplets in model 3 to be extruded (Fig. 2a), while the droplets in FDSS can be easily recovered (Fig. 2b). As for FDSS, the vibration between the maximum spreading diameter (d) of the ferrofluid droplet on the lower plate and the distance (h) between the top and lower plates is plotted in Fig, 2c, which shows that the compression and recovery curves basically overlap, indicating that the FDSS has a good deformation recovery capacity. The maximum pressure that can be withstood by the FDSS is ~ 248 mN, which is ~ 10.8 times that of the model when h is varied by 400 µm·s− 1 to 0.2 mm and then recovered (Fig. 2d). Furthermore, our hybrid adhesive is reusable and highly durable. As evidenced by Fig. 2e, the weight retention of the ferrofluid droplet is sustained over 97.5% after 60 cycles.
The robustness of FDSS depends not only on the guidance and reinforcement of ferrofluid droplets by the magnetic field, but also on the adhesion of aqueous ferrofluid droplets by different wettability substrates. Here, the hybrid wetting and superhydrophobic surfaces are fabricated by laser ablation and low surface energy modification. After laser processing, the aluminum surfaces are covered with micro-nanoclusters (Fig. S2, Supplementary Information). After low surface energy modification, the static contact angle (CA) of the ferrofluid droplets on superhydrophobic surfaces is ~ 154°, while after secondary laser patterning, the CA of the ferrofluid droplet on the patterned region is ~ 24° (Fig. 2f). The hydrophilic pattern has a strong capillary adhesion that firmly trapped the ferrofluid droplets and prevents them from falling off the inclined surfaces, whereas in superhydrophobic region, the ferrofluid droplets roll off easily. Furthermore, ferrofluid droplets can be shaped and stiffened by external magnetic field to overcome the instability of the liquid. Therefore, the combination of hybrid wettability and magnetic field guidance would be a good candidate for robust cargo transportation, in which ferrofluid droplets can not only be shaped and slide rapidly over the superhydrophobic surfaces when actuated by an external magnetic field, but also can drive the upper plate for stable transport.
2.3 Mechanical performance of MDTs
To achieve robust cargo transportation, we propose a synergistic transport system (MDTs) consisting of four sets of FDSS, as shown in Fig. 3a. Here, the surface magnetic strength of the magnet used is ~ 302.6 mT (Fig. 3b). Load stability is important for microrobot to carry cargo. As schemed in Fig. 3c, we design two main aspects for probing the stability of MDTs, including the ferrofluid droplet volume and the structure of the system. The vibration curve of the droplet volume on the pressure resistance is shown in Fig. 3d. The system’s tolerance does not increase with increasing droplet volume, but reaches a maximum of ~ 930 mN at 18 µL and then decreases with increasing volume (18 to 36 µL). Besides, in the absence of an external force, the almost linear increase in pristine spread diameter (L) and pristine ferrofluid height (H) with increasing volume and the decreasing rate of change in between top and lower plates (Fig. 3e), which are mainly attributed to the fact that the further away from the magnetic field, the weaker the magnetic field's ability to bind the droplet, leading to a decrease in the droplet's stiffness. Here, the spacing (D) of adjacent superhydrophilic patterns, the superhydrophilic pattern diameter (d) and the minimum gap after applying pressure are 13 mm, 4 mm and 0.3 mm, respectively.
Furthermore, to demonstrate the mechanical performance of the system, we systematically investigate the dependence of the pressure tolerance on the MDTs structure, and summarize the results in Fig. 3f-h. Here, the structure of the robot is quantified by the spacing (D), the diameter (d) and the distance (h), and the droplet volume used and the minimum gap after applying pressure are 18 µL and 0.3 mm, respectively. Firstly, the system with a d of 4 mm can support a ~ 920 mN downforce when D ≤ 13 mm, while the ferrofluid may spill from the gap, due to the limited area of the upper substrate (Fig. 3f). Secondly, in general, increasing d contributes to the increase the load carrying capacity of the robot, while spill of the ferrofluid droplets occurs when enlarging d by more than 6 mm, as shown in Fig. 3g. Finally, when increasing h to 0.3 mm, the load capacity of the robot is maximized, but decrease as continues to increase (h > 0.3 mm), as shown in Fig. 3h. Besides, the MDT has excellent mechanical stability under long-term compression, as shown in Video S2, Supplementary Information.
Dynamic stability is also important for transporters. When the MDTs are impacted by glass beads (weight: 0.65g) falling from a height of 30 cm on the surface, they can be stabilized and restored to their original state without loss (Fig. S4, Supplementary Information). More importantly, regardless of whether the MDTs are loaded or unloaded, after being subjected to a certain range of lateral forces, they can automatically recover due to the action of capillary adhesion forces, as shown in Fig. S5, Video S2, Supplementary Information, which further prove the stability of the MDTs.
2.4 Applications
Static and dynamic stability of MDTs demonstrates its potential for cargo transportation in practice. For cargo transportation, the lower plate is designed to be large enough to allow the MDTs to move easily over the surface. The load supported by the upper plate can be moved along a predefined trajectory by programmable movement of the bottom magnet array, as shown in Fig. 4a. Here, the MDTs are named robot. To demonstrate the robot’s transport behaviors, we used two important evaluation metrics, load weight (W) and transport velocity (V), to systematically investigate their effects on the robot's transport performance, and summarize the results in Fig. 4b. The thickness of the superhydrophobic lower plate is 0.3 mm and the upper plate weight is ~ 0.435 g.
In general, stable cargo transportation is achieved with moderate W and V; stable transport also occurs when reducing W and/or V, while increasing W and/or V so that the robot will be in a transition state between stabilization and failure (i.e., unstable transport, Fig. S6, Supporting Information); the robot will fail to transport when continuing to W and/or V. For example, the robot is capable of transporting a solid cargo of W ≤ 8 g on the surface at a speed of V ≤ 80 mm/s, while a unstable transport state will occur if the load enlarges to 9 ~ 12 g. When increasing W to 15 g or V to 130 mm/s, the cargo, along with the upper plate, will slide off the surfaces of the ferrofluid droplets due to the fact that the friction driving the transport of the cargo will be greater than the adhesion of the superhydrophilic region to the ferrofluid droplets (Fig. S6, Supporting Information). Besides, the robot can transport the largest loads (W = 15.5 g) when the velocity V ≤ 40 mm/s, which is ~ 35.6 times the mass of the upper plate used to carry the load, indicating excellent cargo transportation capabilities. Physically, cargo transportation behavior of the robot is directly related to solid-liquid contact interface angles. When the robot is loaded, the load angle (𝛼1 and 𝛼2) hardly changes with the change of load mass or transportation state due to the solid-liquid interface property 33,34. During robot loading and transportation, the advancing angle (𝜃2) decreases as the transportation speed (V) increases, and the receding angle (𝜃1) remains constant, which is attributed to the shaping and strengthening of the magnetic droplets by the magnetic field, as shown in Fig. S7, Supporting Information.
Moreover, we demonstrate the dynamic programmability and applicability of the robot for transportation in planar/curved environments, as most of the reported droplet-based manipulation systems for cargo transportation become invalid due to load carrying capacity, uneven operating surfaces, or limited space. Figure 4c displays the programmable cargo transportation control for robot on flat surfaces. Firstly, the robot can be simply assembled in the self-assembly area when subjected to the influence of the magnetic field on the ferrofluid droplets and the adhesion of the superhydrophilic region to the ferrofluid droplets. Secondly, the robot can be dragged and moved to the loading zone for cargo loading by using a magnetic field to orient the ferrofluid droplets. In addition, the magnetic field can be directed to control the robot, including migration, rotation, and movement on an inclined plane, as shown in Fig. S8, S9, Supporting Information. After loading, the cargo is transported along a preset programmed path to the unloading area and complete rotation, commutation and unloading. Finally, the robot returns to its original position after a single transport, the process that takes only 27 s (Video S3, Supplementary Information). Furthermore, the robot can also be applied in limited spaces, such as narrow tubular structures and boxes, which is important in in the field of contactless handling and transportation. By manipulating the movement of the magnets, the robot can easily move and robustly transport cargo inside the PEF tube without mass loss (Fig. 4d, S9, Video S3, Supporting Information). Here, the lower plate is designed as a curved FEP tube with a thickness of 0.8 mm, and the superhydrophilic pattern of the upper plate and the ferrofluid droplets can be designed to match very well for adhesion on the curved surface. Besides, inverted MDTs are also suitable for underwater transportation. The robot can not only directionally and stably tow underwater objects, but also perform programmable rapid movement of cargo (Fig. S10, Video S4, Supporting Information).
Accurate motion monitoring and energy conversion of kinetic energy are notable in contact-free manipulation and transportation, especially in confined unsupervised spaces, which is widely used in microchannel delivery, chemical reactions, and biotechnology. For illustration, we design a unique platform for energy conversion and velocity motion state detection based on Faraday's law of electromagnetic induction35, as shown in Fig. 5a. The lower plate of the platform consists of a double layer, and multiple array coils are placed in the middle of the double layer substrate with a coil spacing of 25 mm. Here, the coils used have a diameter of 20 mm, a thickness of 0.2 mm, an internal resistance of 34 Ω, and a coil number of 400 turns, and the overall thickness of the lower plate is 0.8 mm. As the robot moves rapidly across the platform, the magnetic flux through the coil changes rapidly (Fig. 5b), resulting in an open-circuit voltage (VOC) that can be recorded by an oscilloscope, thus realizing the conversion of kinetic energy to electrical energy for driving external electronic devices such as LED arrays “DUTs” (Fig. 5c, Video S5, Supplementary Information). The external circuit used is shown in Fig. S11, Supplementary Information. Besides, when the robot moves at different speeds and generate different voltages. In general, the faster the robot is transported, the higher the VOC is (Fig. 5d). For example, when robot transports at a speed of 2.8 cm/s, the resulting peak VOC is 24.6 mV, while at a speed of 34.2 cm/s, the VOC is 94.8 mV.
Furthermore, the VOC generated by the different coils can be used to monitor the robot’s motion, such as the instantaneous speed and motion direction. By programmable manipulation of the ferrofluid droplets using a magnetic field, we perform sequential kinematic state monitoring of the robot. As shown in Fig. 5e, the VOC generated by the coil at positions 1, 2, and 5 is ~ 25.0 mV, at positions 6, 9, 8 is 30.2 mV, and at positions 7 and 4 is 99.6 mV, which indicates that the robot move along positions 1, 2, and 5 to position 6 at a uniform speed of V1 = 3.0 cm/s, and then move along positions 9 and 8 to position 7 at a uniform speed of V2 = 12.0 cm/s, finally, move to position 4 at a uniform speed of V3 = 34.5 cm/s. Therefore, our platform provides a versatile pathway for automated control and monitoring of contactless droplet manipulation and transportation. In practice, the durability of the system must be given priority. To verify its durability, the robot is performed by continuously moving back and force linearly with a speed of V = 14.3 cm/s for 40 s (Fig. 5f). The robot can output a sufficiently stable electric response over long periods after the move, and no significant mechanical damage or performance degradation is observed.
Besides, the platform also can be used as a self-power device for wireless signal transmission. The robot can output a stable electrical response activated by a magnetic field, indicating that the electrical output performance of the robot can be evaluated by connecting different external loads. As shown in Fig. 5g, different external resistors are connected in series with the robot, and their peak voltage and power are tested. The output voltages of the robot increased with load resistance, while the corresponding power increases and then decreases, reaching a peak power of ~ 48 µW at R = 34 Ω, which indicates that the system can be used to supply power wirelessly based on the previous study36–38.