Small-Scale Machines Driven by External Power Sources

PROGRESS REPORT Microrobots www.advmat.de Small-Scale Machines Driven by External Power Sources Xiang-Zhong Chen, Bumjin Jang, Daniel Ahmed, Chengzhi Hu, Carmela De Marco, Marcus Hoop, Fajer Mushtaq, Bradley J. Nelson, and Salvador Pané* of these tiny devices. When the size of an object is reduced to the micro/nanometer realm, the ratio of inertial forces to viscous forces (characterized by the Reynolds number) becomes so small that inertial forces can be neglected.[9,10] That is, the small-scale object is in a highly “viscous” environment. In this case, power must be continuously provided to propel the smallscale robots. Because of the devices’ small size, traditional power sources such as capacitors and batteries are difficult to integrate into their design. Therefore, effectively driving small-scale robots and controlling their locomotion is challenging. Various propulsion techniques have been developed to power micro- and nanoswimmers. Propulsion strategies can be roughly classified into three categories: self-propulsion, external propulsion, and a hybrid strategy that involves both. The self-propulsion approach is usually based on local chemical reactions occurring on the small-scale machine’s surface, in which chemical energy is converted into kinetic energy. This energy transformation results in the locomotion of the micro- and nanoswimmers.[11–16] However, this strategy usually requires fuels such as hydrogen peroxide (H2O2) or hydrazine (N2H4), which are highly cytotoxic. Some alternatives to these fuels have been recently reported, for example, by Kastrup and co-workers, who show that microparticles made of calcium carbonate and tranexamic acid can swim through blood by carbon dioxide bubble propulsion.[17] While self-propelled devices can move very fast with velocities of up to 1.5 cm s−1, they lack directionality, which means that their direction and locomotion is not easily controlled. Therefore, the use of self-fueled micro- and nanomachines is limited to circumstances where precise positioning is not required. Alternatively, an external power source can be used to precisely control the locomotion of micro- and nanorobots. This approach is more desirable for biomedical applications in which precise positioning is essential. These external sources of energy can be provided by magnetic fields, light, acoustic waves, electric fields, thermal energy, or combinations of these.[13,18,19] The hybrid strategy usually consists of combining one method, either self-powered or externally powered, to provide force for propulsion and another method, based on the application of an external source of energy, to steer the devices. A very recent trend exploits the capabilities of motile microorganisms, which are combined with stimuli-responsive building blocks such as magnetic components so that they can be steered using external power sources.[20] The characteristics of all types of small-scale robots, including advantages and limitations, are summarized and listed in Table 1. It should be noted that there is not a small-scale robot Micro- and nanorobots have shown great potential for applications in various fields, including minimally invasive surgery, targeted therapy, cell manipulation, environmental monitoring, and water remediation. Recent progress in the design, fabrication, and operation of these miniaturized devices has greatly enhanced their versatility. In this report, the most recent progress on the manipulation of small-scale robots based on power sources, such as magnetic fields, light, acoustic waves, electric fields, thermal energy, or combinations of these, is surveyed. The design and propulsion mechanism of micro- and nanorobots are the focus of this article. Their fabrication and applications are also briefly discussed. 1. Introduction In 1959, in a talk given at an American Physical Society meeting, Richard Feynman mentioned that “it would be interesting in surgery if you could swallow the surgeon.” Seven years later, the classic movie “Fantastic Voyage” depicted a miniature submarine that was injected into a patient to clear a clot in a blood vessel to save his life. For the past half-century, researchers have made great progress toward this fantastic voyage. With the development of micro- and nanotechnology, researchers have been able to design and fabricate miniaturized mobile devices with sizes ranging from tens of nanometers to several hundreds of micrometers. These small-scale devices have shown great potential in minimally invasive surgery, targeted therapy, and cell manipulation.[1–4] Initial trials of these small-scale devices on animal models have recently been conducted.[5,6] In addition to applications in the biomedical arena, these devices also have potential applications in fields such as environmental remediation.[7,8] However, several challenges continue to impede the practical implementation of these devices including insufficient biocompatibility, imprecise operation, and a lack of efficient device tracking and monitoring capabilities in vivo. Over the last two decades, the community of small-scale robotics has focused its efforts on finding strategies for the locomotion Dr. X.-Z. Chen, B. Jang, Dr. D. Ahmed, Dr. C. Hu, Dr. C. De Marco, Dr. M. Hoop, F. Mushtaq, Prof. B. J. Nelson, Dr. S. Pané Multi-Scale Robotics Lab (MSRL) Institute of Robotics and Intelligent Systems (IRIS) ETH Zurich CH 8092, Zurich, Switzerland E-mail: [email protected] The ORCID identification number(s) for the author(s) of this article can be found under https://doi.org/10.1002/adma.201705061. DOI: 10.1002/adma.201705061 Adv. Mater. 2018, 1705061 1705061 (1 of 22) © 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.advancedsciencenews.com www.advmat.de design fitted for all kinds of applications. Micro- and nanoswimmers are currently developed according to targeted applications. In this Progress Report, we summarize the most recent research efforts realized in the area of externally powered micro- and nanorobots. For a summary of progress in the area of self-fueled small-scale swimmers, the readers are referred to previous publications.[4,7,11,13,15,16] While there are review papers summarizing the state of the art in this field extensively,[2,18,21–23] new ideas, new technologies, and new materials are constantly being integrated, resulting in rapid progress in this field in short periods of time. Therefore, in this Progress Report, we review not only the past classic groundbreaking work but also very recent developments in externally powered small-scale robots which have not been previously considered such as magnetoelectric micro- and nanomachines. Fabrication approaches and applications will also be briefly discussed. We end the report with conclusions and some prospective directions. 2. Microrobots and Micromachines Driven by External Power Sources 2.1. Magnetically Driven Microrobots and Micromachines Magnetic propulsion is promising for micro- and nanorobots locomotion because it is a noninvasive method that offers control and navigation.[1,19] When a magnetic object is placed in a magnetic field, it can be subjected to a magnetic force (field gradient) and a magnetic torque. The magnetic force F exerted on the magnetic object is F = ( m ⋅ ∇ )B (1) where B is the external magnetic field and m is the magnetic dipole moment of the object.[9,10] The magnetic dipole can either be induced (e.g., in a paramagnetic, diamagnetic, antiferromagnetic, or superparamagnetic material) or permanent (e.g., in a ferromagnetic or ferrimagnetic material). In a homogeneous magnetic field, the magnetic object will not experience gradient forces; thus, it cannot be moved in this way. However, if the dipole is not aligned with the direction of the applied magnetic field, a torque will be exerted on the magnetic dipole τ = m×B (2) causing the magnetic object to rotate until the direction of the dipole is aligned with the magnetic field. Therefore, to actuate a magnetic small-scale robot, either an inhomogeneous or a time-varying magnetic field must be applied. Figure 1 shows magnetic field-based driving modes used in the actuation of micro- and nanorobots.[10] Among these steering modalities, the field gradient driving mode is based on the use of inhomogeneous fields and the others, such as rotating and oscillating field driving modes are time-varying modalities. The magnetic field gradient driving mode is relatively straightforward and can be applied to actuate any magnetic object; therefore, we will not discuss this mode in this Progress Report. Here, we summarize the most recent progress on time-varying methods, which usually require the design of specific micro- and nanoarchitectures. Adv. Mater. 2018, 1705061 Xiangzhong Chen is currently a postdoctoral researcher at the Multi-Scale Robotics Lab (MSRL) of the Institute of Robotics and Intelligent Systems (IRIS) at ETH Zürich. He received his Ph.D. in 2013, majoring in polymer chemistry and physics from Nanjing University. In November 2013, he joined in MSRL. His Ph.D. thesis is about the application of ferroelectric polymers in the field of data storage, energy storage, and energy conversion. Now he is working on bridging magnetic and ferroelectric materials (ceramics, polymers, and composites) with robotic microdevices for biomedical applications such as cell stimulation and drug delivery. Bradley J. Nelson has been the Professor of Robotics and Intelligent Systems at ETH Zürich since 2002. Before moving to Europe, Prof. Nelson worked as an engineer at Honeywell and Motorola and served as a United States Peace Corps Volunteer in Botswana, Africa. He has also been a professor at the University of Minnesota and the University of Illinois at Chicago. He has over thirty years of experience in the field of robotics. He serves on the advisory boards of a number of academic departments and research institutes across North America, Europe, and Asia and is on the editorial boards of several academic journals. Salvador Pané is currently a senior research scientist with a permanent appointment at the Institute of Robotics and Intelligent Systems (IRIS) at ETH Zürich. He received his Ph.D. in chemistry (2008) from the Universitat de Barcelona in the field of the electrodeposition of magnetic materials. He became a postdoctoral researcher at IRIS in August 2008 and senior research scientist in 2012. Dr. Pané is currently working on bridging materials science, chemistry, and electrochemistry with small-scale robotics for various applications. 2.1.1. Rotating Fields Some of the most well-known small-scale robots driven by rotating magnetic fields are magnetic helical swimmers.[24,25] Inspired by bacteria that propel themselves by rotating their 1705061 (2 of 22) © 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.advancedsciencenews.com www.advmat.de Table 1. Key features, advantages, and limitations of small-scale robots classified according to the power source used for their propulsion. were covered by Ni and Ti to confer magnetism for actuation and biocompatibility, respectively. These helical micromachines have shown great promise Key features Advantages Limitations in drug delivery, gene delivery, and cell delivery, after loading their surface with payloads such Self-propulsion Use chemical reactions High swimming speed Continuous fuel to generate bubbles or when the fuel is adequate; supply is needed in the as liposomes.[26–30] By using stimuli-responsive chemical gradients for Low cost microenvironment for loading agents, the microrobots can perform smart propulsion powering; delivery upon pH or temperature changes. Their Most of the fuels are potential in vivo applications were also explored.[5] toxic; A swarm of ABFs, functionalized with a nearLack of directionality infrared fluorescence probe (NIR-797), were tracked External propulsion External power provided Position can be precisely Special manipulating in real time when they were actuated to navigate in by magnetic fields, light, controlled; equipment is usually the peritoneal cavity of a mouse model. This DLW acoustic fields, electric Good directionality required. method not only allows for batch fabrication but fields, etc., for actuation also provides flexibility in tuning the materials and Hybrid actuation Combination of different Relatively efficient in Live cells survive the shape of the structures.[20,27,31–36] ABFs made strategy actuation strategies, power output; only under certain of magnetic nanocomposites[34,37] (Figure 2B) and including self-powered Good control on conditions; helical-based micropumps and microsyringes[36] strategies, externallydirectionality; Precise control of the (Figure 2C) have greatly expanded the application powered strategies, Responsive to multiple live cells might be of these fascinating architectures. The latter microand motile live cells stimuli difficult robots comprise two nested components which (e.g., sperm cells) or microorganisms consist of a magnetically actuated main shaft with integrated parts embedded inside a passive cylinder. This sophisticated micromachinery can actively collect, encapsulate, transport, and controllably release micro- and flagella, a helical magnetic swimmer of comparable size to a nanoagents. The swimming performance of ABFs relies on the bacterium was first produced and reported by Nelson’s group at driving frequency and magnitude of magnetic fields, shape, ETH Zurich in 2007.[24] The helical microrobot had a magnetic magnetic properties, surface properties of structures, and the head and a helical semiconductor tail composed of a InGaAs/ viscosity and homogeneity of the liquid environment,[9,38–44] as GaAs bilayer thin film (Figure 2A(i)). It could rotate around its helical axis under a rotating magnetic field to move forward or has been well documented. Depending on the conditions, helbackward in the direction perpendicular to the plane of rotation ical swimmers can also perform 3D navigation in liquids. For (Figure 2A(ii)–(vii)). Because the microrobot mimics the shape detailed information, readers are referred to previously puband motion of a bacterium, it was named an artificial bacterial lished reviews.[9,10,40] flagellum (ABF). In 2012, the same group produced a new genNanoscale helical robots have also been manufactured. For eration of ABFs using direct laser writing (DLW).[26] Negative example, Fischer’s group fabricated nanoscale helical robots (length 1–2 µm, diameter 200–300 nm) using glancing-angle photoresists, such as SU-8 and IP-L, are directly polymerized deposition (GLAD).[42,43,45–47] These nanorobots were the smallest by two-photon lithography to form helical microarchitectures. The parameters of the helices, such as length, diameter, pitch, robots yet reported at the time of publication (Figure 2D).[46] and helicity angle, can be easily tuned. The written structures They are able to move in water at a speed of 40 µm s−1 under rotating magnetic fields. Other nanoscale helical robots have been fabricated through template-based electrodeposition. Park and co-workers fabricated Pd nanosprings by electrodepositing a Cu/Pd alloy in an anodic aluminum oxide (AAO) template followed by a de-alloying process.[48] Using this technique, Wang’s group produced nanoscale helical nanorobots in 2014 and studied their actuation behavior (Figure 2E).[49] The shape of the microrobots relied on the Cu/ Pd concentration ratio and the diameter of the AAO template. Using a rotating magnetic field of 150 Hz, velocities of up to 15 µm s−1 were achieved with 400 nm diameter nanohelices. In 2016, using the same fabrication approach, Hoop et al. fabricated antibacterial nanorobots by coating a layer of Ag onto such Figure 1. Magnetic actuation modes. a) Field vector rotated in a plane. b) Field rotated along structures.[50] These nanorobots can kill not the mantel of a cone. c) Oscillating “up-down” field in a plane. d) On-off field. e,f) Magnetic [10] only Gram-negative Escherichia coli (E. coli) field gradients. Adapted with permission. Copyright 2013, The Royal Society of Chemistry. Adv. Mater. 2018, 1705061 1705061 (3 of 22) © 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.advancedsciencenews.com www.advmat.de Figure 2. A) Artificial flagellum fabricated from a self-rolled semiconductor. Reproduced with permission.[39] Copyright 2009, American Institute of Physics. B) Array of helical microrobots fabricated from a nanocomposite of SU-8 and superparamagnetic nanoparticles by direct laser writing. The inset shows a magnified image of a composite microrobot. Reproduced with permission.[34] Copyright 2014, Wiley. C) Microcapsule and microsyringe fabricated by DLW. Reproduced with permission.[36] Copyright 2015, Wiley. D) An array of helical nanorobots and magnified image of a single helical nanorobot fabricated by GLAD. Reproduced with permission.[46] Copyright 2009, American Chemical Society. E) Cu helical nanorobot fabricated using electrodeposition and subsequent dealloying. Reproduced with permission.[49] Copyright 2014, Royal Chemical Society. F) Schematic of the metallization of a helical lipid structure into helical and tubular magnetic structures via electroless deposition of CoNiReP. Reproduced with permission.[51] Copyright 2012, Wiley. G) Magnetite ABF fabricated on a spirulina-based template. After ultrasound treatment, the ABF dissembled. Reproduced with permission.[53] Copyright 2015, Wiley. H) Soft microrobots fabricated with a bilayer hydrogel. The final shape of the helical structures can be controlled by aligned magnetic nanoparticles embedded in the hydrogel. Adapted with permission.[56] Copyright 2016, Nature Publishing Group. I) Flexible Au/ Ag/Ni metal nanowire motors driven by a rotating magnetic field. Reproduced with permission.[57] Copyright 2010, American Chemical Society. but also Gram-positive methicillin-resistant Staphylococcus aureus (S. aureus), both of which represent leading multidrugresistant bacterial pathogens. Helical small-scale structures are ubiquitous in nature, and many researchers have exploited them as templates or scaffolds for manufacturing micro- and nanoswimmers. For example, in 2012, Schuerle et al. developed a method to fabricate ABFs by coating self-assembled phospholipidic helices with a magnetic CoNiReP alloy via electroless deposition (Figure 2F).[51] In 2014, Wang’s group reported a biotemplate method to fabricate magnetic ABFs from helical plant vessels.[52] This approach consists of extracting the helical xylem vasculature of plants, which is subsequently coated with Ni and Ti. The structures are then carefully cut to be used as magnetic microrobots. In 2015, Zhang’s group reported another fabrication method of microrobots using spirulina as a template[53] (Figure 2G). The spirulina was coated with Fe3O4 via a sol-gel method, followed by high-temperature annealing to remove the template leaving Fe3O4 to crystallize. The hollow helical microrobots exhibit good biocompatibility and a high specific surface area and can be broken by ultrasonication, making them promising for targeted drug delivery applications. Adv. Mater. 2018, 1705061 Soft small-scale robots are attracting intensive attention, because their mechanical properties are much closer to those of biological systems than inorganic versions, and because they exhibit superior biocompatibility. Recently, Nelson’s group has developed soft microrobots made of stimuli-responsive hydrogel bilayers,[54–56] which consists of one thermally responsive layer made of poly(N-isopropylacrylamide) (PolyNIPAM) and another nonresponsive layer made of poly(ethylene glycol) diacrylate (PEGDA). Magnetic nanoparticles were blended with PolyNIPAM to endow the hydrogel with magnetism for actuation. The magnetic nanoparticles were also used for generating heat via magnetic fields or near-infrared irradiation to induce deformation of the temperature-sensitive PolyNIPAM layer. Another function of the magnetic nanoparticles was to control the rolling direction of the hydrogel and to program the magnetic shape anisotropy of the device (Figure 2H). These soft microrobots could not only be actuated by rotating magnetic fields but also adapt themselves to certain shapes, showing great potential in biomedical applications.[56] Several other interesting micro- and nanoarchitectures can be actuated using rotating magnetic fields. For example, Wang’s group reported micromotors (Figure 2I) made of segmented Au/Ag/Ni nanowires fabricated by electrodeposition.[57] Using 1705061 (4 of 22) © 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.advancedsciencenews.com www.advmat.de a selective chemical etchant, the Ag segment was thinned and subsequently transformed into a flexible hinge. Under applied rotating magnetic fields, the Ni tail rotated and the flexible Ag segment facilitated a cyclic mechanical deformation, which was consequently transformed into translational motion. Rotating fields can also be employed to actuate “surface walkers” or “surface rollers.” These microrobots are propelled on or near a surface. This propulsion mechanism works even without direct contact between the device and the surface. The rolling motion occurs in the same plane as the rotating magnetic field, unlike for helical propellers. When the roller is sufficiently heavy and is rolling on the surface, the friction between the surface and the roller provides the force to move the device forward. Such structures include but are not limited to microtubes and Janus particles (Figure 3A).[10,58–60] When the influence of gravity on the roller is negligible, the apparent viscosity increases towards the surface because of the presence of a wall or a surface; this drag imbalance causes forward motion (Figure 3B).[50,61–64] In this case, the rollers can not only walk on the surface but also climb a wall, as demonstrated by Zhang et al. and by Chen et al. using magnetic nanowires.[61,65] The vortex generated by the rolling of a surface walker can be used to trap and transport microobjects. Zhang’s group presented recently an upgraded version consisting of a dumbbell surface walker made by assembling a Ni nanowire to two polymer microbeads. By applying rotating magnetic fields, vortices with stronger forces for microobject transportation were demonstrated.[64] Previous examples dealing with nanowire-based surface walkers rotate Figure 3. A) (i) Schematic showing the manipulation of a Janus microsphere rolling on a surface. (ii) Time-lapse image of a moving Janus microsphere steered by a rotating magnetic field. A noncoated silica sphere is also shown for reference. Adapted with permission.[58] Copyright 2016, The Royal Society of Chemistry. B) (i) Schematic showing that velocities at the two ends of the nanowire differ because of inhomogeneous boundary conditions. Adapted with permission.[61] Copyright 2010, American Chemical Society. (ii) The nanowire is manipulated in the shape of a heart. Reproduced with permission.[65] Copyright 2017, Wiley. C) (i) An illustration of a rodbot. (ii–iv) A rodbot being used to transport a protein crystal onto a microgripper. Adapted with permission.[66] Copyright 2014, International Union of Crystallography. D) An SEM image (i) and illustration (ii) of an Au/Ni/Au segmented nanowire. (iii) How the nanowire exhibits kayak motion is shown. Adapted with permission.[68] Copyright 2017, The Royal Society of Chemistry. E) Geometry of colloidally assembled surface walkers. Adapted with permission.[69] Copyright 2010, National Academy of Sciences. F) (i) Schematic illustration showing how the colloidal wheel can be actuated. (ii) Translational movement of microwheel under rotating magnetic fields.[70] Adapted with permission.[70] Copyright 2016, Nature Publishing Group. G) Schematic of a doublet subjected to an external magnetic field H precessing around the y-axis. Adapted with permission.[72] Copyright 2008, American Chemical Society. Adv. Mater. 2018, 1705061 1705061 (5 of 22) © 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.advancedsciencenews.com www.advmat.de about their short axis because of the magnetic shape anisotropy. In principle, high-aspect-ratio shapes such as long rods and wires are preferentially magnetized through the long axis. There are other fabrication strategies that provide a design space to create a preferential magnetization to achieve rolling motion. For example, Tung et al.[66,67] fabricated a high-aspectratio cuboid, known as the Rodbot. It contains CoNi micropillars positioned parallel to each other and sufficiently separated to minimize magnetic dipole interactions between them. In this arrangement, the cuboid could be rotated about its long axis (Figure 3C) in order to roll on surfaces. Another interesting microrod called a magnetic microkayak was reported recently by Mair et al. (Figure 3D).[68] The microkayaks consisted of nanorods of two Au segments connected by a Ni segment. When magnetically actuated near a solid surface, these microrods exhibited a double-cone rotating pattern similar to the motion of a kayaker’s paddle, which induced translational motion. These microrods were actuated at high frequencies in the kilohertz range. In addition to individual robots, microassemblies were actuated through this method. Sing et al. reported that particle chains assembled by superparamagnetic colloids could be actuated by means of rotating magnetic fields (Figure 3E).[69] The translational motion speed depended on the frequency of the field. However, the chain broke at high frequencies. Tasci et al. fabricated microwheels through the assembly of superparamagnetic colloids (Figure 3F).[70] The microwheel was actuated by a rotating magnetic field and rolled on the surface of a substrate at high frequencies. The microwheels achieved velocities of up to 100 µm s−1 when the frequency was ≈500 rad s−1 (80 Hz). Conical fields can also be employed for the locomotion of magnetic micro- and nanoswimmers. A conical field is a variation of a rotating field and can be considered as a rotating field superimposed on a DC field perpendicular to the rotating plane. Tierno et al. demonstrated translational movement of DNAlinked anisotropic doublets consisting of two paramagnetic colloidal particles actuated as surface walkers (Figure 3G).[71,72] The frictional asymmetry caused by the surface was sufficient to provide the force for translational motion. Conical fields were also applied to drive core–shell magnetic nanowires to make them swim in 3D.[65] However, the underlying locomotion mechanisms are not yet fully understood. The authors assumed that 3D swimming behavior was probably associated with the asymmetric shape of the nanowires. Interestingly, Faivre’s group recently reported that a specific shape design might not be necessary for a magnetic propeller as long as the orientation of the magnetic moment with respect to the principal rotation axes of the object is not zero.[73–75] A geometrically achiral object can acquire apparent chirality because of its interaction with the external magnetic field. An optimized propeller results in propulsion speeds comparable to those of an optimally magnetized helix.[76] 2.1.2. Oscillating Fields Oscillating magnetic fields have also been widely used to propel magnetic micro- and nanoagents. The direction of the oscillating field is usually perpendicular to the traveling direction of the Adv. Mater. 2018, 1705061 swimmer. Small-scale devices propelled using this approach usually contain flexible joints or springs in their configuration. One of the first prototypes was demonstrated by Dreyfus et al., who fabricated a flexible artificial flagellum using DNA-linked chains of paramagnetic colloidal beads (Figure 4A).[77] The velocity and direction of motion could be controlled by the strength and frequency of the magnetic fields. Later, Jang et al. manufactured nanowire-based swimmers containing one or more magnetic Ni segments, and a flexible polypyrrole (PPy) tail connected through hinges made of poly(allylamine hydrochloride)/sodium poly(styrene sulfonate) (PAH/PSS) bilayer films (Figure 4B).[78] These nanoswimmers with one, two, or three links exhibited nonreciprocal motion when actuated by oscillating magnetic fields. Wang’s group also developed hinged nanoswimmers containing two Au segments at both ends and two Ni segments in the middle, with thinned Ag interconnectors in between (Figure 4C).[79] Under oscillating magnetic fields, the Ni segments rotated synchronously. Consequently, a traveling wave is generated through the Au tail causing the structure to propel. Misra and co-workers reported a sperm-like micromachine consisting of an ellipsoidal magnetic CoNi head and a flexible tail made of SU-8 (Figure 4D).[80] The microrobot was named MagnetoSperm and was fabricated using standard photolithography followed by e-beam evaporation of magnetic films. The oscillating fields can be used not only to actuate the microrobots but also to trigger functionality if specifically designed functional components are integrated. By integrating magnetoelectric materials, Pané and co-workers were able to produce microrobots that induced surface charge changes upon oscillating magnetic field stimulation.[58,65] The charge can be used for controlled drug release possibly due to repulsive electrostatic force. 2.1.3. ON-OFF Fields “ON-OFF” fields, or step fields, refer to periodic magnetic fields with an “ON” phase and an “OFF” phase. One of the first devices designed to propel using the “ON-OFF” approach was the MagMite, a spring-based microsystem, which was actuated by in-plane oscillating magnetic fields. A second generation of this type of micromachines were the PolyMites, which were produced via an easier, faster, and less expensive fabrication process (Figure 4E,F).[81] The basic actuation principle for these microrobots or mites is based on a “two-mass-spring oscillator system” and comprise three major components: a magnetic body that can stand on a surface, a magnetic hammer suspended above a surface, and a spring that connects the body and the hammer. When the magnetic field is ON, the two parts attract each other. When the magnetic field is OFF, the attraction considerably decreases and the restoring force pushes the two parts away from each other. When the frequency of the oscillating magnetic field is close to the resonant frequency of the device, the response of the body and the hammer to these forces can be magnified to generate a net displacement of the device. Using ON-OFF fields with specific angles were also employed to actuate microrobots resembling cilia (Figure 4G–I).[82] When actuated by the magnetic fields, the cilia moved nonreciprocally, which generated a net propulsive force. The motion 1705061 (6 of 22) © 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.advancedsciencenews.com www.advmat.de Figure 4. A) Schematic representation of a flexible magnetic filament. Magnetic particles are linked by double-stranded DNA. Adapted with permission.[77] Copyright 2005, Nature Publishing Group. B) Schematic and SEM images of one-, two-, and three-link swimmers. Adapted with permission.[78] Copyright 2015, American Chemical Society. C) Schematic (upper left) and SEM image (upper right) of a multilinked artificial nanofish. Scale bar: 800 nm. Lower panels: time-lapse microscope images showing magnetically controllable start (left), stop (middle), and restarted (right) movement of the nanofish. Scale bar: 5 µm. Reproduced with permission.[79] Copyright 2016, Wiley. D) Schematic and SEM image of a MagnetoSperm. Reproduced with permission.[80] Copyright 2014, American Institute of Physics. E) Schematic of how PolyMites function under an oscillating field. F) A PolyMite with polystyrene beads. Reproduced with permission[81] Copyright 2014, IEEE. G) SEM image of a Paramecium with cilia. Reproduced with permission.[167] Copyright 2011, Springer. H) SEM images of fabricated ciliary microrobots with and without a mask structure. I) Time-lapse images during translational motion of the ciliary microrobot. Adapted with permission.[82] Copyright 2016, Nature Publishing Group. was controlled by the shape parameters of the cilia, such as the cilium length and the cilium angle, and by the field strength and the field angle. These parameters were used to precisely control the position and orientation of the microrobot. 2.2. Acoustic Micro- and Nanoswimmers Recent studies have generated considerable interest in the field of propulsion that uses an acoustic field as a driving mechanism. Acoustics provide a new and attractive method for generating substantial propulsive forces, which could become important for applications in medicine and lab-onchip devices. However, before this technology can be used in vivo, a better understanding of the propulsion mechanism, its precise control, and its navigation in biologically suitable media is necessary. A unique advantage of acoustics over other field-driven systems is that this approach allows independent activation[83] and propulsion of each swimmer in a group, which can be a powerful tool in collaborative functions. Adv. Mater. 2018, 1705061 2.2.1. Acoustic Nanoswimmers Some early acoustic nanoswimmers exploited asymmetric shapes of rigid metallic nanorods.[84] The swimmers were levitated by introducing a standing wave acoustic field. The field was established between a piezo transducer at the bottom and a reflector at the top (Figure 5A). A series of nodes and antinodes developed across the transducer and the reflector. The nodes and antinodes were minimum and maximum pressure points, respectively, in the standing wave field. Nanoswimmers with density and compressibility values greater than those of the surrounding liquid were pushed toward the nodes. Wang and co-workers showed the axial motion of metallic nanorods at a velocity of ≈200 µm s−1 and noted that they demonstrated interesting motor–motor interactions.[84,85] Steering of the acoustic nanoswimmer was established by integrating Ni between the nanorods;[86,87] however, a large, external magnetic field of 40–50 mT was required to maneuver the swimmers. Recently, a helical, Ni-coated Pd structure was supplemented onto Au nanorods using a template-assisted fabrication method (Figure 5B) to demonstrate motion in acoustic and magnetic 1705061 (7 of 22) © 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.advancedsciencenews.com www.advmat.de Figure 5. A) Schematic of the experimental setup and of the acoustic propulsion of the nanorods. Adapted with permission.[84] Copyright 2012, American Chemical Society. B) Schematic of the magnetoacoustic hybrid nanoswimmer and fabrication method.[88] C) Image sequences demonstrate the motion of the nanoswimmer in acoustic and magnetic fields. B,C) Adapted with permission.[88] Copyright 2015, American Chemical Society. D) SEM image of the nanoswimmer composed of a flagellum-like polypyrrole (PPy) tail and an Au/Ni head. E) Oscillation of the nanoswimmer induces localized vortices or acoustic microstreaming in a liquid. F) Image sequences show the left-to-right translational motion of the nanoswimmer in water containing 0.5 µm diameter beads. D–F) Adapted with permission.[89] Copyright 2016, American Chemical Society. G) Schematic of a genesilencing strategy using acoustic nanoswimmers. H) Image sequence illustrates the intracellular rotary motion of a nanoswimmer. G,H) Adapted with permission.[92] Copyright 2016, American Chemical Society. I) Neutrophil rolls on vasculature before migrating into tissue. J) Magnetic particles aggregate due to dipole–dipole interaction in the presence of a rotating magnetic field. Scale bar, 25 µm. K) A rolling-type motion along the boundary is achieved by combining acoustic and magnetic fields. Image sequence demonstrates right to left transport of the assembled particles. Scale bar, 25 µm. I–K) Adapted with permission.[93] Copyright 2017, Nature Publishing Group. fields.[88] This swimmer takes advantage of the Au segment to move in an acoustic field and takes advantage of the helical structure to move in a rotating magnetic field. Figure 5C shows the controlled regulation of nanoswimmers using both fields. An advantage of the system is that it can propel nanoswimmers using an acoustic field and a magnetic field, and can steer nanoswimmers in a magnetic field. Although these swimmers show great promise, they may be difficult to use in in vivo environments because a predictable standing wave field can be difficult to establish in a human body. Recently, Ahmed et al. developed nanoswimmers that can be propelled in traveling acoustic waves.[89] These traveling waves are especially advantageous, because they interact directly with the swimmer’s body to generate propulsion, and because they are independent of the field developed in the chamber. Figure 5D shows a nanoswimmer comprising two parts: a flexible flagellum composed of polypyrrole, that is, the tail, and a metallic head. The nanoswimmers are fabricated using multistep electrodeposition techniques. In the presence of ultrasound, the flagellum-like tail of the nanoswimmer undergoes oscillations. The resulting oscillations induce a steady flow field in the surrounding liquid, which is characterized by a pair of counterrotating vortices, as shown in Figure 5E. The generation of Adv. Mater. 2018, 1705061 vortices around the swimmer is referred to as acoustic microstreaming. A perturbation expansion approach was used to model the flow field near the swimmer. The local acoustic microstreaming response of the fluid by the nanoswimmer can be characterized by a second-order system of equations (v2, p2, ρ2), which, in turn, is driven by the time-harmonic first-order fields (v1, p1, ρ1), where v, p, and ρ correspond to the fluid velocity, pressure, and density, respectively. A full description of the vortex flow profile of the swimmers was presented in other earlier publications.[89,90] The first- and second-order equations are successively solved using a finite element method (FEM) simulation using COMSOL Multiphysics to obtain the vibration velocity, v1, of the nanoswimmer and the corresponding streaming velocity v2 of the swimmer. The results showed that structural resonance of the swimmer can lead to maximal acoustic forces.[89] The force acting on the nanoswimmer is the surface integral of the stress σ2 due to microstreaming and the vibration, v1, acting on the surface of the nanoswimmer, out to a surface, ∂Ω1. Taking n to be outward normal pointing away from the swimmer, the force FA is given by[89] FA =  ∫ ∂Ω1 1705061 (8 of 22) σ 2 ⋅ ndA − ∫ ∂Ω1 ρ1 v 1 v 2 ⋅ ndA (3) © 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.advancedsciencenews.com www.advmat.de The angle bracket 〈·〉 represents time-averaging. The stress, σ2, is interlinked to v2. The first term in Equation (3) arises due to acoustic microstreaming, and the second term arises from the structural vibration of the nanoswimmer. The force that propels the nanoswimmer results from linear and nonlinear interactions of the swimmer in an acoustic field. Figure 5F shows the left-to-right translational motion of the nanoswimmer in water. Mallouk and co-workers demonstrated intracellular propulsion using acoustic nanoswimmers.[91] Swimmers were engulfed by HeLa cells via phagocytosis, which was achieved by controlling the incubation time of the swimmers. Researchers have also developed an intracellular gene-silencing strategy using acoustic nanoswimmers.[92] The nanoswimmers were wrapped with a rolling circle amplification (RCA) DNA strand, which assisted in gene silencing. The intracellular motion of the nanoswimmers led to accelerated siRNA delivery, which affected the fluorescence signal of the cells, as shown in Figure 5G. Figure 5H shows the rotary motion of the nanoswimmers inside a cell. Inspired by neutrophils rolling on endovascular walls before transmigrating to the disease site, a novel propulsive mechanism based on combined ultrasound and magnetic actuation modalities has been developed. Neutrophils undergo tethering, rolling, adhesion, and crawling through the endothelial cell walls before transmigrating into the tissue, as shown in Figure 5I.[93] The magnetic particles aggregate into a rolling sphere due to the dipole–dipole interaction in the presence of a rotating magnetic field (Figure 5J). The aggregate then migrates toward the wall due to the radiation force of an acoustic field. By combining magnetic and acoustic fields, a rolling-type motion along the boundaries is achieved, as shown in Figure 5K. Although acoustic-based nanoswimmers have been found to be useful in various bioapplications, additional research is needed to determine their propulsion mechanism and to assess their behavior in biologically relevant media. conducted experiments to measure the net propulsive flow generated by the oscillation of a bubble and accurately modeled the microstreaming flow near the swimmer at different heights. The flow near a bubble can reach 100 mm s−1. Figure 6C,D demonstrates the translation and rotational motion of the swimmer. Recently, Qiu et al. used resonant bubbles of various sizes to develop a bidirectional rotary motor,[98] as shown in Figure 6E. The resonance frequency of a spherical bubble is inversely proportional to its radius.[99] As a result, smaller bubbles have higher resonance frequencies. Their system contained thousands of microbubbles of two different sizes arranged in an array. The microbubbles were excited at different frequencies, which led to clockwise and counterclockwise motion, as shown in Figure 6F. Their study was the first to involve such a large number of microbubbles, and they achieved a significant force and torque generation of ≈0.45 mN and ≈0.5 mN mm, respectively. Resonant bubbles of unequal sizes have also been arranged within a microswimmer to achieve steerability in an acoustic field.[83,94] However, controlled motion was difficult to achieve because of the finite width of the overlapping resonances of the bubbles and coupling between bubbles.[83] Recently, Ahmed et al. developed a hybrid acoustomagnetic soft microrobot.[100] Their design consists of a microbubble at the center of the swimmer’s body and superparamagnetic particles aligned within a polymer matrix, as shown in Figure 6G. The microcavity supports a bubble trap, which enables propulsion in an external acoustic field, and the magnetic particles allow controlled motion in a magnetic field. The authors used this method to manipulate the swimmer along a path to write “ETH,” as shown in Figure 6H. This approach can be used to maneuver swimmers in a 3D environment. Although bubble-based acoustic swimmers show great potential for generating significant propulsive forces and can swim in a viscous medium, the air-filled bubbles remain stable for only a few hours. Next-generation swimmers will require a stable polymeric coating to maintain the stability of the bubbles. 2.2.2. Acoustic Microswimmers Another class of acoustic swimmers uses the resonance behavior of trapped air bubbles within parylene- or polyethylene glycolbased cavities in an acoustic field.[83,94] When a trapped bubble is exposed to an acoustic field with a wavelength much larger than the diameter of the bubble, the bubble begins to oscillate. As the frequency of the acoustic waves generated by a transducer approaches the resonance frequency of the trapped bubble, the oscillation amplitude of the bubble reaches its maximum value. The resulting oscillations induce counter-rotating vortices or microstreaming in the surrounding liquid, and these oscillations, along with microstreaming around the bubbles, generate a propulsive force.[95,96] Ahmed et al. demonstrated the propulsion of a polymer-based microswimmer with translation velocities as high as ≈8 mm s−1 (50 body lengths s−1) and 20 rotations s−1.[83] Figure 6A,B demonstrates translation and rotational motion of the swimmer. The bubble-based swimmers exhibited successful propulsion in glycerol and viscous hydrogel; however, the behavior of these swimmers in these liquids is not yet thoroughly understood. Bertin et al. demonstrated the propulsion of armored bubbles (Figure 6C) ranging from 10 to 20 µm in a traveling wave arrangement.[97] They Adv. Mater. 2018, 1705061 2.3. Light-Driven Small-Scale Machines Generating a highly localized magnetic, electric, and ultrasonic field at the sub-millimeter scale is challenging. When micro/ nanorobots are exposed to fields, collective motion occurs unless robots are individually designed. Inducing smart behavior, such as collaborative tasks by small groups or individual robots, is therefore difficult. In this respect, light can be a good alternative power source to drive robots because the size of the beam can be reduced to the sub-micrometer scale using optical devices such as objective lenses, filters, and mirrors. Additionally, selective wavelengths and multiple-beam radiation enable the generation of signals with better temporal and spatial resolution; such signals can be used for selective motion control of the robots. Together with the aforementioned merits of light, recent demand for the use of renewable energies has further motivated researchers toward light-driven micro/nanorobots. In the following sections, four different light-driven micro/ nanorobots are discussed. First, thermophoretic and photocatalytic propulsion mechanisms are discussed on the basis of the material properties of the robots. Further, the collective behavior 1705061 (9 of 22) © 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.advancedsciencenews.com www.advmat.de Figure 6. A,B) Image sequences demonstrating translation and rotational motion of acoustic microswimmers. Adapted with permission.[83] Copyright 2015, Nature Publishing Group. C) Microscopy of an armored microbubble and image sequence illustrating translational motion. D) Rotary motion of an armored acoustic swimmer. C,D) Adapted with permission.[97] Copyright 2016, American Physical Society. E) Schematic of a bidirectional rotary motor. F) Images show clockwise and counterclockwise motion in the presence of unequal frequencies of the acoustic field. E,F) Adapted with permission.[98] Copyright 2016, American Institute of Physics. G) Schematic of artificial acoustic-magnetic soft swimmers. H) The swimmer was manipulated to write “ETH” in water under acoustic and magnetic fields. Reproduced with permission.[100] Copyright 2017, Wiley. of the micro/nanorobots by diffusiophoretic mechanism will be introduced for robots that mimic behaviors observed in nature, for example, Dictyostelium, zebrafish, and Drosophila that secrete signaling chemicals to guide neighboring members.[101] Finally, soft microrobots composed of liquid crystalline elastomers are discussed. These robots crawl on the surface by contracting and relaxing their body under illumination by light of the appropriate wavelength. 2.3.1. Thermophoretically Driven Micro- and Nanorobots The first generation of light-driven microrobots was reported by Jiang et al., who fabricated microrobots by coating an Au thin Adv. Mater. 2018, 1705061 layer onto silica hemispheres (Figure 7a(i,ii)).[102] Under NIR illumination, Janus microstructures developed asymmetrical temperature distributions around their surfaces because of different NIR light absorption energies between Au and silica; this physical phenomenon is based on the surface plasmon resonance phenomenon. While absorption of NIR by silica is almost negligible, the Au component readily absorbs NIR radiation and undergoes a substantial temperature increase because of strong oscillations of its electrons. As a result, a thermal gradient forms around the robot resulting in its thermophoretic propulsion. The mechanism was confirmed by tracing fluorescent nanoparticles around a Janus microrobot tethered to a substrate under NIR illumination (Figure 7a(iii)). A decade later, Xuan et al. demonstrated that mesoporous Au/silica 1705061 (10 of 22) © 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.advancedsciencenews.com www.advmat.de Figure 7. Thermophoretically driven micro/nanorobots. a) Thermophoretically driven Janus Au/silica microrobots (3 µm diameter): (i) schemes of an Au/silica Janus microrobot (top image) placed in a fluidic chamber (bottom image). NIR was illuminated along the Z-axis from the bottom; (ii) optical image of the microrobots (scale bar: 1 µm). (iii) Trajectory of fluorescence tracers (40 nm of PS particles) around a tethered Janus microrobot under NIR illumination. The image was acquired by stacking images recorded for 5 s. Adapted with permission.[102] Copyright 2010, American Physical Society. b) Thermophoretically driven Janus Au/mesoporous silica nanorobots: (i) a simulated temperature distribution around a 50 nm Janus nanorobot under NIR illumination. (ii) Time-lapsed optical images of nanorobots with ON-OFF motion. Scale bar: 20 µm; (iii) corresponding time–speed plot of (ii). Adapted with permission.[103] Copyright 2016, American Chemical Society. Janus nanorobots could be driven by the same mechanism.[103] Similarly, a simulation showed an asymmetrical temperature distribution around the nanorobot under NIR exposure, verifying a substantial temperature gradient between the Au and silica faces. This result indirectly supported the thermophoretic mechanism suggested by Jiang et al. (Figure 7b(i)). The propulsion of the robot was stopped and reinitiated repeatedly by applying an ON-OFF signal of NIR illumination, as demonstrated in the optical image sequences in Figure 7b(ii) and in the corresponding time–speed plots in Figure 7b(iii). 2.3.2. Photocatalytically Driven Micro- and Nanorobots Photocatalytic micro- and nanorobots use light-induced catalytic activity for propulsion in a solution.[104] Under light illumination, the photocatalytic part of the microrobots absorbs light energy and creates electron–hole pairs. The holes and the electrons on the surface of the photocatalytic material react with species present in the surrounding solution generating byproducts. Photocatalytic robots adopt similar propulsion mechanisms as those observed in conventional catalytic microand nanorobots,[8,12,105] such as self-electrophoresis, self-diffusiophoresis, and bubble propulsion. Self-electrophoresis has mostly been observed for photocatalytic Janus-based micro- and nanorobots.[106,107] For example, Dong et al. presented microrobots consisting of photocatalytic TiO2 microspheres, half-coated with a thin Au layer (Figure 8a).[107] Under UV light illumination, holes and electrons were generated in the TiO2 hemisphere. Consequently, water was oxidized Adv. Mater. 2018, 1705061 at TiO2 producing an excess of protons, and electrons were transported from TiO2 to the Au hemispheres, which acted as electron sinks. The excess protons at the TiO2 surface migrated to the Au hemispheres through the robot surface because of proton deficiency on the surface of the Au hemisphere. This proton flow resulted in a slip velocity at the liquid–solid interface, thus causing the microparticle to propel. A recent study by Wu et al. demonstrated that the self-electrophoretic mechanism can be enhanced by adding photodegradable dyes to solution.[108] Self-diffusiophoresis has also been observed for Janus microspheres. For example, Zhang et al. demonstrated Janus microrobots composed of Au/WO3 on carbon microspheres, as shown in Figure 8b.[109] The authors observed that photocatalytically generated byproducts simply diffused away from the robot surface due to concentration gradients, resulting in a diffusiophoretic force. In contrast, bubble propulsion mostly occurs for microtubular and nanotubular TiO2 engines, as demonstrated by Mou[110] and by Enachi et al.,[111] respectively. Under light illumination, O2 bubbles were generated and coalesced to form bigger O2 bubbles in the inner space of the tube, as shown in Figure 8c. The bubbles were finally released at the opening tip, creating a thrust that drives the microengine. Although the propulsion mechanisms of photocatalytic micro/nanorobots are analogous to those of conventional catalytic robots, recent studies have revealed unique features in the motion of photocatalytic robots. First, photocatalytic robots can be propelled using pure water as a fuel.[106,107,109,122] Second, to a certain extent the photochemical reaction can be adjusted using light intensity, thereby providing the capability 1705061 (11 of 22) © 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.advancedsciencenews.com www.advmat.de Figure 8. Photocatalytically driven micro/nanorobots. Propulsion mechanisms: a) self-electrophoretic propulsion. Adapted with permission.[107] Copyright 2016, American Chemical Society. b) Self-diffusiophoretic propulsion. Adapted with permission.[109] Copyright 2017, American Chemical Society. c) Bubble propulsion. Adapted with permission.[110] Copyright 2015, Wiley. d,e) Light-guided trajectories of photocatalytic microrobots. (d) Adapted with permission.[113] Copyright 2016, Nature Publishing Group. (e) Adapted with permission.[114] Copyright 2017, Wiley. for prompt ON-OFF motion and further speed modulation of robots.[107,110,111,122] Such control is difficult in the case of conventional catalytic robots because of their perpetual (or autonomous) motion via uninterruptable reactions with the fuel solution unless their speed is modulated by external sources of energy such as temperature.[112] Another advantageous feature of photocatalytically driven micro- and nanoswimmers is their motion steerability. Using asymmetrical light exposure, Dai et al. and Chen et al.[113,114] maneuvered robots in desired Adv. Mater. 2018, 1705061 trajectories (Figure 8d,e), suggesting that such robots have potential applications as micromanipulation tools. Propulsion of photocatalytic motors has been mainly achieved with UV light illumination because of the limited absorbance spectrum of the employed photocatalytic materials.[106–111,113] The use of UV light is not only harmful for living organisms but also represents a very small portion (around 5%) of the solar spectrum, which consists primarily of visible light. Recent studies have focused on developing 1705061 (12 of 22) © 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.advancedsciencenews.com www.advmat.de new materials or expanding the absorption range of existing photocatalytic materials. The photocatalytic materials used to create visible-light-driven micromotors were BiOI,[115] Cu2O,[116] Fe2O3,[117–121] Recent study shows that a micromotor composed of black TiO2 (B-TiO2) micromotor can be propelled by the illumination of both UV and visible light.[122] This provides more flexibility regarding the selection of light when one explores a complex fluid environment that is sensitive to certain wavelengths. 2.3.3. Collective Behavior of Micro- and Nanorobots Collective behavior of micro- and nanorobots under light stimuli was first demonstrated by Ibele et al.,[123,124] who observed the schooling behavior of AgCl microparticles under UV illumination in water. The overall mechanism of this behavior was based on diffusiophoresis (Figure 9a(i)). Two different species (protons and chloride ions) were produced by the dissolution of AgCl under UV (Figure 9a(ii)). Consequently, an electric field toward the AgCl particles was generated because the diffusivity of protons is one order of magnitude greater than that of chloride ions. The electric field further developed two different flows: one is electrophoresis, which occurs far above the surface of the substrate, and the other is electroosmotic flow, which occurs via the double layer at the surface of the substrate (Figure 9a(iii)). Schooling/repulsive behavior of tracer particles with different zeta potentials was also studied; the zeta potential affected the magnitude of the osmotic flow (see the schooling behavior of silica particles near AgCl particles in Figure 9a(iv)). Figure 9. Collective behavior of micro and nanorobots under light illumination. a) Schooling behavior of AgCl microparticles and tracer particles. (i) Scheme of the diffusiophoresis mechanism around an AgCl particle in water under UV illumination. Small blue arrows in the bulk fluid indicate electrophoretic flows, and large blue arrows on the surface of the substrate indicate electroosmotic flows. (ii) Dissolution of Ag under UV illumination. (iii) Scheme of the schooling behavior of AgCl particles. (iv) Schooling behavior of tracer particles showing aggregation around the AgCl particles with time. Scale bar: 20 µm. (i,iii) Adapted with permission.[124] Copyright 2012, Springer. (ii,iv) Adapted with permission.[123] Copyright 2009, Wiley. b) Oscillatory behavior of AgCl particles in H2O2 under UV light. (i) Left panel: scheme of AgCl dissolution under UV irradiation. Green arrow is the generated electric field. Right panel: scheme of the re-formation of AgCl on the AgCl particle in H2O2; (ii) image sequences of oscillatory behavior of tracer particles near AgCl particles. Scale bar: 10 µm. Adapted with permission.[125] Copyright 2010, American Chemical Society. c) Transition between schooling and repulsive behavior of Ag3PO4 particles in the presence of NH3 and UV light. (i) An NOR gate: input parameters are NH3 and UV, and outputs are schooling or repulsive behavior. (ii) Repulsive behavior under UV. Scale bar: 20 µm. Adapted with permission.[126] Copyright 2013, American Chemical Society. d) Group behavior of azobenzene molecules. (i) Chemical structure of azobenzene molecules. Shift from trans to cis structure when UV light was switched to blue light; (ii) repulsive behavior of tracer particles by an osmotic flow; (iii) created patterns using collective behavior of tracer particles in response to light exposure. Adapted with permission.[128] Copyright 2016, Nature Publishing Group. Adv. Mater. 2018, 1705061 1705061 (13 of 22) © 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.advancedsciencenews.com www.advmat.de A year later, Ibele et al. further demonstrated the collective oscillatory behavior of AgCl particles in the presence of dilute hydrogen peroxide and UV light.[125] As previously described, AgCl particles were dissolved under UV illumination, and subsequently, an electric field was generated toward the AgCl particles. However, a reverse process occurred in the presence of H2O2, which led to the re-formation of an AgCl layer on the particle through a reaction involving Ag+ and H2O2. This reverse process developed an opposite electrophoretic force pointing from the particle to the diffused ions (Figure 9b(i)). The two countervailing forces caused the oscillatory stick-slip motion of AgCl particles against the underlying substrate. Similar motion occurred for trace particles when displaced around AgCl particles (see the sticking and release of tracer particles around the AgCl particles under UV in Figure 9b(ii)). Duan et al. further expanded knowledge of collective behavior by studying the transition between the schooling and exclusion behaviors of Ag3PO4 microparticles by adding NH3 in the solution and by irradiating the fluid with UV light.[126] The transition between schooling and repulsion due to the shift of the chemical equilibrium was demonstrated through modification of these two parameters (i.e., NH3 addition and UV illumination) (Figure 9c(ii)). Additionally, the concept of a NOR gate was demonstrated using this approach (Figure 9c(i)). Collective behavior also occurs with other photocatalytic particles and photosensitive molecules such as TiO2, as reported by Hong et al.,[127] Fe2O3, as reported by Palacci et al.,[120] and azobenzene, as reported by Feldmann et al.[128] In detail, TiO2 and Fe2O3 materials show advantages such as nondissolution of particles under light illumination and, hence, low toxicity in biological applications. Fe2O3 is additionally beneficial because of its magnetic functionality, which can be used as another input parameter to study collective behavior. A recent study on the photoisomerization of azobenzene molecules demonstrated the change from hydrophobicity to hydrophilicity by switching UV to blue light (Figure 9d(i)).[128] This switching behavior drove the collective behavior of tracer particles, as depicted in Figure 9d(ii)). Unique patterns of the tracer particles were achieved via selective exposure of the photosensitive molecule solution to light (Figure 9d(iii)). The collective behavior of tracer particles near tethered photoreactive patches of N-hydroxyphthalimide triflate (PAG-1) and p-Si/Pt was also studied by Yadav et al.[129] and by Esplandiu et al.,[130] respectively. 2.3.4. Light-Driven Soft Microrobots Among soft materials, liquid crystalline elastomers (LCE) have been widely used as motile bodies because of their large swelling/contraction properties with respect to its glass-transition temperature. The first generation of LCE-based microrobots was demonstrated by Zeng et al.,[131] who fabricated a flexible surface walker consisting of an LCE body with two pairs of IP-Dip polymer legs (Figure 10a(i)). The LCE structure was heated above its glass-transition temperature when illuminated with a focused green beam, because the dyes incorporated in the LCE matrixes absorb green light. Figure 10a(ii) shows around a 20% contraction of an LCE body under illumination Adv. Mater. 2018, 1705061 (Figure 10a(ii)). With repeated ON-OFF light signals, the structure showed a reversible contraction/swelling mode that enabled its locomotion along the surface of a glass substrate in air (Figure 10a(iii)). Using the same material property of LCE, Palagi et al. demonstrated an artificial ciliate.[132] Figure 10b(i) shows simulated contraction/swelling behavior of the LCE microcylinder when exposed/unexposed to green light. The author and co-workers used the beam scanning mode of a digital micromirror device (DMD) to propel the robot (Figure 10b(ii)). As the beam scanned along the body of the robot, the LCE microcylinder contracted locally its body where the beam interacted, consequently generating a traveling wave. This created a forward propulsion, similar to the swimming strategy of ciliates (Figure 10b(iii)). Crawling motion of microcaterpillars under green light was demonstrated by Rogóż et al.,[133] who fabricated microcaterpillars by curing LCEs and aligning them on a nonrubbed surface distinct from a rubbed surface; the LCEs aligned along the rubbing direction near the rubbed surface, whereas the rest showed a random orientation (Figure 10c(i)). The cured LCE body showed a sawtooth configuration above the glasstransition temperature (Figure 10c(ii)). By adopting the beam scanning mode used by Palagi et al., a crawling motion of the microcaterpillar was achieved (Figure 10c(iv)). 2.4. Electrically Driven Small-Scale Machines 2.4.1. Electroosmotic Propulsion of Diodes Self-propelled micro- and nanomotors are particularly suitable for in vivo biomedical applications. Electric fields, as an external energy source, can induce locomotion to miniaturized micro/nanomotors via various mechanisms. Electrokinetic phenomena are known to occur on particles exposed to an electric field and can be used to move objects. DC fields can be exploited to drive particles by electrophoretic motion on charged particles or molecules.[134,135] Dielectrophoretic forces can also be generated by nonuniform fields.[136,137] AC fields can also drive asymmetric particles by induced-charge electrophoresis and electroosmotic propulsion of diodes.[138,139] In the latter case, voltage can be induced as a result of rectification of an alternating electric field between the electrodes of each diode. The constant electric field between the electrodes leads to an electroosmotic flow, which propels the diodes or pumps the adjacent liquid. During this process, diodes suspended on the liquid surface are oriented in the direction of the field lines and move parallel to the electric field (Figure 11A). The semiconductors harvest electrical energy from external alternating electric fields and convert it into mechanical propulsion. The diodes can move with approximately constant velocity as high as millimeters per second until reaching the opposite electrode and can stop a few millimeters in front and above the electrode, where the intensity of the field decreases.[139] The combination of nanomaterials with electrode materials has opened new horizons. Electrochemical routes, particularly template-assisted electrodeposition, are very useful for preparing such nanostructured materials.[140] For example, segmented organic–inorganic polymer–metal nanowire diodes were 1705061 (14 of 22) © 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.advancedsciencenews.com www.advmat.de Figure 10. Light-driven liquid crystalline elastomer (LCE) microrobots. A) Light-driven surface microwalker; (i) SEM images of a top view (top panel) and side view (bottom panel) of a microwalker. (ii) Optical images of the surface walker with (left panel) and without (right panel) green-light illumination. Scale bar: 20 µm. (iii) Time-lapse optical images of a moving surface walker on a glass substrate in air under pulsed (chopped) laser illumination. Adapted with permission.[131] Copyright 2015, Wiley. B) Light-driven artificial ciliate; (i) simulated structural deformation of a cylindrical microrobot under light illumination. (ii) Scheme of optical manipulation. A DMD provides space and time variation for incoming light, resulting in a scanning mode for light along the robot. (iii) Forward locomotion of the robot under light that periodically scans along the body of the robot. Scale bar: 200 µm. Adapted with permission.[132] Copyright 2016, Nature Publishing Group. C) Light-driven artificial microcaterpillar (L: 14.8 mm, W: 3.8 mm, and t: 50 µm); (i) fabrication scheme of the artificial microcaterpillar: Selectively cured LC monomers under UV irradiation. (ii) Local contraction (T > Tp) and relaxation (T < Tp) of the structure under green light. Tp: phase-transition temperature of LC molecules. (iii) Time-lapse of optical images of the robot showing forward motion under the beam in scanning mode. Scale bar: 5 mm. Adapted with permission.[133] Copyright 2016, Wiley. synthesized by electrochemical deposition of Au into alumina templates, followed by the electrochemical polymerization of pyrrole (Ppy).[141] Polypyrrole-cadmium (PPy-Cd), Au-CdSe-Au, or CdSe-Au-CdSe nanowire hybrids with attractive diode properties were prepared in the same fashion.[138,142] These diode nanowires were propelled under a spatially uniform AC electrical field. Typical motion trajectories of template-grown PPy-Cd diode nanowires, monocomponent PPy, Cd and Au nanowires, and bisegmented PPy-Au nanowires when exposed to a spatially uniform AC square-wave electric field for 3 s are shown in Figure 11B. The PPy-Cd diode nanowire moved parallel to the field axis over a long trajectory. However, nanowires made of PPy, Cd, Au and bisegmented PPy-Au nanowires did not show any defined locomotion. This result indicated that the movement of nanowire-based diodes was caused by the local electroosmotic flux induced by the external field, and not as a Adv. Mater. 2018, 1705061 consequence of induced-charge electrophoresis, where the motion was perpendicular to the field direction. The electrically driven PPy-Cd diode nanowire exhibited a speed of ≈7 body lengths s−1, showing particular promise for use in a wide range of biomedical applications. 2.4.2. Bipolar Electrochemistry As an alternative approach for electric manipulation, numerous groups have shown that bipolar electrochemistry can be used to induce motion in conductive objects.[143,144] When a conducting particle is placed in an electric field, a maximum polarization voltage occurs between the two sides of the particle, oriented toward the electrodes. This voltage is dominated by the electric field strength and the characteristic 1705061 (15 of 22) © 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.advancedsciencenews.com www.advmat.de Figure 11. A) Schematic (upper panel) and overlay of a series of photographs (lower panel) showing the motion of diodes. Scale bar: 5 mm. Adapted with permission.[139] Copyright 2007, Nature Publishing Group. B) Schematic of the propulsion of PPy/Cd and CdSe-Au-CdSe nanowire diodes. The topleft inset shows the corresponding SEM images. Adapted with permission.[138] Copyright 2010, The Royal Chemical Society. C) a) Schematic of bipolar electrochemical water splitting on a conductive microbead. b) Optical microscopy image of a stainless steel bead experiencing an electric field. Scale bar: 250 µm. c) Motion of a glassy carbon microbead in a microchannel exposed to a 5.3 V mm−1 electric field in 7 × 10−3 M H2SO4. Scale bar: 100 µm. d) Scheme showing proton reduction and hydroquinone oxidation on a bipolar electrode sphere. e) Motion of a 1 mm diameter stainless steel bead exposed to a 1.3 V mm−1 electric field in 24 × 10−3 M HCl and 48 mm hydroquinone. Scale bar: 1 mm. f) Motion of a 275 mm diameter glassy carbon microbead in a microchannel with a 4.3 V mm−1 electric field in 7 mm HCl and 14 mm hydroquinone. Scale bar: 100 µm. Adapted with permission.[144] Copyright 2011, Nature Publishing Group. dimensions of the object. When the polarization voltage is sufficiently large, redox reactions can occur at opposite sides of the particle: oxidation reactions at the anodic pole and simultaneous reduction reactions at the cathodic pole. These reactions imply the evolution of gas bubbles (i.e., H2 or O2) on the microobject and can trigger its linear or rotational motion (Figure 11C).[144] Moreover, the moving objects can be fabricated from various conductive materials with different shapes and sizes, which makes this approach very versatile for different applications such as cargo transportation in microfluidic devices. However, because of the uncontrolled detachment of bubbles from the microobject and intrinsic variations in polarization intensity, this electrically induced motion is not completely regular.[145] Adv. Mater. 2018, 1705061 2.5. Thermally Driven Small-Scale Machines Heat-induced propulsion of micro- and nanorobots has been rarely reported, because heat is not easily transferred to microrobots. One way of using heat to actuate microrobots is to integrate an array of microheaters. This approach was realized with a MEMS microrobot containing shape memory alloy components. Upon heating, parts made of shape memory alloy deformed to actuate the microrobot.[146–148] However, electric wires were connected to the heaters to power them; thus, the microrobots had to be tethered, impeding further downscaling of such devices. Other researchers explored the possibility of driving microobjects, such as droplets or bubbles through direct temperature gradients on substrates.[149,150] For example, 1705061 (16 of 22) © 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.advancedsciencenews.com www.advmat.de Figure 12. A) Dodecane drop propelled on a program-heated surface. Reproduced with permission.[149] Copyright 2003, American Institute of Physics. B) SEM images of a thermally actuated three-leg microdevice, a schematic of an asymmetric driving and steering demonstration: with letters “N.” Adapted with permission.[151] Copyright 2006, American Institute of Physics. C) A bubble microrobot being used to move a glass bead around a feature on the substrate surface. Reproduced with permission.[152] Copyright 2011, American Institute of Physics. D) Schematic of the experiment: a coil is positioned under the sample consisting of magnetic Janus particles suspended in water. The motion is visualized using an optical microscope. E) Trajectories of a Janus motor (left) and a reference particle (right) moved in an AC magnetic field. Adapted with permission.[154] Copyright 2013, American Chemical Society. Darhuber et al. developed a substrate with four heaters aligned at the edges[149] (Figure 12A). Through coupling of the temperature field and surface tension, a droplet was precisely moved and positioned. However, all of these methods required direct contact between the microobject and the heat source. To wirelessly transfer heat onto a microrobot and power it, researchers have developed various strategies by combining specifically selected materials and powering methods. Light is usually used to induce heat to the body of microrobots to deform the device. A focused laser has been used to supply energy to a three-legged microrobot made of thin metal film bimorphs (400 nm Al and 200 nm Cr) to induce curvature of one leg, which leads to stepwise translational motion of the robot on a low-friction surface (Figure 12B). The speed and direction could be easily controlled by the laser beam.[151] Light-induced thermocapillary effects were introduced to drive the motion of a bubble (Figure 12C).[152] When the light was focused close to the bubble, the liquid heated up causing a flow toward the bubble and the development of two symmetric rotation flows on either sides of it. The temperature gradient resulted in a net movement of the gas bubble toward the warmest location. Therefore, the bubble can act as a microswimmer and can be used to manipulate and assemble micro-objects. Recent result showed that light-induced heat provided directional motion of a helical microrobot made of thermally responsive PNIPAm (poly(N-iso- Adv. Mater. 2018, 1705061 propylacrylamide)) microhydrogel containing Au nanorods.[153] Upon NIR irradiation, heat was generated on gold nanorods due to surface plasmon resonance, which caused the deformation of the hydrogel main body. The deformation of the helix caused a rotation around the principal axis of the helix, which in turn resulted in its translational motion. Magnetic fields are another effective way of wirelessly inducing heat. Magnetic hyperthermia was employed to induce thermophoresis to drive magnetic spherical Janus motors in liquid media (Figure 12D,E).[154] Full control over motion was achieved because of the specific properties of ultrathin 100-nmthick permalloy magnetic films, which exhibit a topologically stable magnetic vortex state in the cap structure of Janus motors. 2.6. Combined Strategy for Driving Small-Scale Machines Combinations of some of the methods to harvest energy from different power sources would strengthen the propulsive thrust of small-scale robots. Another advantage is the possibility of exploiting some energy sources to trigger the motion of micro- and nanorobots, while other stimuli could be used to control specific tasks, such as drug release. We can identify two different categories of micro- and nanomotors adopting combined strategies: (i) those 1705061 (17 of 22) © 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.advancedsciencenews.com www.advmat.de based on the combined use of microbes or cells and artificial compounds and (ii) those based on fully synthetic systems. The recent tendency for adopting microbes or cells in microand nanorobotic platforms is undoubtedly related to their intrinsic biocompatibility. Sitti’s group and Schmidt’s group have recently reviewed the use of bioengineered and biohybrid bacteria as systems for drug delivery.[155,156] Retaining their ability to swim driven by environmental gradients, bacteria can be integrated with other functional elements to be directed to the target location. For example, single E. coli bacteria have been trapped in Ppy microtubes grown electrochemically, internally modified with a bacteria-attractant layer of polydopamine (PDA), and electroplated with Ni for magnetic guidance.[157]E. coli bacteria have also been adopted to drive Janus metallic/PS microparticles with the anticancer agent doxorubicin (DOX) as cargo. The Fe cap provides directionality associated with the motion of the biohybrid beads under an applied magnetic field, suppressing random motion.[158] Martel’s group recently developed a strategy of delivering drug to tumors by covalently binding drug-containing nanoliposomes onto magnetotactic bacteria Magnetococcus marinus strain MC-1.[159] Each of these bacteria contains a chain of magnetic iron oxide nanocrystals. They tend to swim along local magnetic field lines. These bacteria prefer anoxic environments which are found widely spread in tumors. Making use of both magnetotaxis and anaerobicity of the bacteria, the researchers demonstrated successful targeted drug delivery, and showed that this combinatorial strategy significantly improved the therapeutic index of various nanocarriers in tumor hypoxic regions. Similarly, the magnetotactic bacteria Magnetosopirrillum gryphiswalense (MSR-1) were integrated with drug-loaded mesoporous silica microtubes to build hybrid microswimmers. These hybrid devices were capable of antibiotic delivery to an infectious biofilm.[160] Human cells have also been extensively investigated as drug carriers in functional robotic microsystems, especially red blood cells (RBCs) and sperm cells. Wang’s group used Fe3O4 NP-loaded RBCs as micromotors, combining ultrasound and magnetic fields for guidance.[161] Schmidt’s group recently reviewed how single spermatozoa can be surrounded by helical or tubular microstructures to obtain functional biohybrid microswimmers. These devices were termed spermbots, and they were mainly employed for in vitro fertilization applications. The sperm-trapping microhelices were magnetic to facilitate the steering of sperm cells with reduced mobility. The sperm-containing microtubes were made of thermoresponsive polymers in order to release spermatozoa when in contact with the oocyte.[162] Fully synthetic micro- and nanorobots combined self-fueled with external energy sources’ components. Unlike biohybrid versions, they can be employed in a broader spectrum of applications, for example, environmental purification of pollutant substances. For example, our group has recently shown that electrochemically fabricated coaxial TiO2-PtPd-Ni nanotubes (NTs) can be successfully used for cleaning water from dyes under natural sunlight and can be propelled using multiple locomotion strategies (magnetic and acoustic), along with autonomous actuation in the presence of H2O2.[62] The literature contains numerous examples of microand nanorobots, most of which are self-propelled in H2O2 solutions and magnetically steered.[163,164] Adv. Mater. 2018, 1705061 Adoption of toxic fuels such as H2O2 can be problematic in biomedical applications; thus, much recent research has opted for fuel-free micro/nanorobots.[18] For example, magnetoacoustic nanomotors composed of Au segment and a Pd-coated Ni nanospring segment was reported.[88] Nelson, Pané and co-workers have developed other strategies to trigger specific functions of micro/nanorobots using external stimuli in association with magnetic propulsion. For example, Huang et al. reported the fabrication of soft-micromachines made with a nonswelling layer of poly(ethylene glycol)diacrylate (PEGDA) selectively patterned onto a swelling thermoresponsive N-isopropylacrylamide (NIPAAm) hydrogel layer doped with Fe3O4 NPs that are magnetically aligned during polymerization.[56] Fusco et al. fabricated a steerable magnetic microrobot functionalized with electrodeposited chitosan, a pH-sensitive hydrogel, for the treatment of eye diseases.[165] The same authors developed a photolithographically structured hydrogel microrobot to load and release drugs in situ upon NIR-light stimulation.[54] Ahmed et al. demonstrated the fabrication of acoustomagnetic soft microswimmers. These bubble-based swimmers are fabricated using PEGDA, which is intrinsically hydrophilic. A perfluorochlorosilane coating was used to make the swimmer’s surface hydrophobic, which was necessary to trap air bubbles in the microcavities, and magnetic microparticles were added to magnetically steer them.[100] Wang and co-workers demonstrated the cargo capabilities of ultrasound-propelled Au nanowire motors encapsulated within a biocompatible pH-responsive coating loaded with caspase-3 (CASP-3). Upon entering a cell and with exposure to the higher intracellular pH, the polymer coating dissolved, releasing CASP-3 and inducing cellular apoptosis.[166] 3. Conclusions and Outlook Over the past decade substantial progress has been made in micro- and nanorobotics. We reviewed the recent developments of microrobots powered by external energy sources such as magnetic fields, light, acoustic waves, electric fields, and thermal energy. Table 2 compares micro- and nanorobots with different propulsive mechanisms with respect to some critical features such as speed and directionality. In comparison to early research in this area, more recent research has been heavily focused on new design concepts and fabrication techniques as well as the integration of additional features for their practical application. Despite the remarkable progress made in the field, many challenges remain in terms of design, fabrication, control, integration, and functionalization of small-scale robots. Due to the highly interdisciplinary nature of this field, close collaboration of researchers from various fields, including, but not limited to, robotics, material science, biology, and medicine will be necessary. In the future, micro- and nanorobots should be developed by taking the following aspects into consideration. 1) Design: Micro- and nanorobots at present, are predominantly designed by taking inspiration from nature and mimicking the motion of biological microorganisms. While doing so, these small-scale robots must be modified and optimized according to their potential applications. New designs with greater motion efficiency are also 1705061 (18 of 22) © 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.advancedsciencenews.com www.advmat.de Table 2. Comparison of different propulsive mechanisms in externally powered micromachines. Magnetic fields Speed Light Acoustic waves Electric fields Thermal ++ + +++ + + Directionality +++ ++ ++ + + Positioning precision +++ +++ + ++ + 3 3 2 2 2 Feasibility for biomedical application Dimensions of motion +++ +++ ++ + + Complexity of control system +++ ++ + + + expected. Apart from fabricating application oriented microrobots, researchers should focus on integrating them with complex functionalities which are compatible with the current fabrication processes. This approach would require that micro- and nanorobotics scientists consider the demands of different applications in the design stage and communicate with material scientists to realize the best approach for fabricating micro- and nanorobots. 2) Fabrication: As previously mentioned, the design of microand nanorobots must be tuned for their respective applications. Therefore, compatible fabrication methods must be carefully considered. For this case, customized fabrication methods are usually needed. Micro-3D rapid prototyping, such as direct laser writing techniques, are ideal customized fabrication methods, because they are very flexible for fabricating various geometries. However, only a few materials are currently available for 3D DLW. Furthermore, the incorporation of functional materials into micro- and nanorobots is highly limited and, hence, the development of new functional DLW materials is of utmost importance. Current micro3D rapid prototyping techniques are not typically suitable for mass production on a wafer scale, due to their intensive time demand. The development of cost-effective mass fabrication methods is also desirable. Under the current circumstances, we envisage a strategy to first use 3D DLW to explore the feasibility of the customized structure and optimize it, and then develop a mass production method to fabricate the optimized structures on a large scale. In the future, we also envision 3D rapid prototyping techniques that can be used to fabricate small-scale robots at the wafer scale with a high resolution. 3) Integration of functionalities: Another important aspect is to optimize the process to integrate functional materials with micro- and nanorobots for particular applications. Two strategies have been developed thus far. One strategy is to integrate stimuli-responsive materials into small-scale swimmers. In this case, functions can be triggered once the micro- and nanorobots are in the appropriate stimulating environment, for example, via pH and temperature changes. Stimuli-responsive materials endow micro- and nanorobots with intelligence, because they enable these devices to sense changes in their environment. However, off-target effects may occur because of unforeseen disturbances. The second strategy is to integrate certain functional materials that can be triggered only by external power sources on demand. This approach provides an alternative triggering of function with a high level of control and precision. To integrate functional materials, one should not only consider Adv. Mater. 2018, 1705061 process compatibility but also the requirements for certain applications, especially in the biomedical field. Biocompatibility, long-term stability, recyclability, and therapeutic efficacy must all be considered. 4) Control: Controlled locomotion and function triggering of micro- and nanorobots are not only determined by the robot itself but also by the control system. A control system with higher precision must still be developed. To this end, realtime tracking of micro- and nanorobots is indispensable.[2] In most cases, a single microrobot is not sufficient to fulfill a task. Typically, a swarm of microrobots must cooperate to complete the job. Therefore, developing an effective strategy for controlling a swarm of swimmers, and at the same time discriminative control over an individual or a subgroup of the swarm is needed. Currently, the swimming of most micro- and nanorobots is demonstrated in aqueous solution. The motion of micro- and nanomachines in viscous and nonuniform dispersants that mimic various body fluids will require careful investigation.[40] Although great progress has been made in this field over the last two decades, great challenges still persist. Researchers from different fields must collaborate in order to overcome these challenges and realize the use of these technologies toward practical applications outside the laboratory environment. Acknowledgements This work was partially financed by the European Research Council Starting Grant “Magnetoelectric Chemonanorobotics for Chemical and Biomedical Applications” (ELECTROCHEMBOTS) under Grant No. 336456. The authors also acknowledge the SBFI Cost Project No. C16.0061 under the COST Action MP1407. C.D.M. acknowledges the Marie Skłodowska-Curie Individual Fellowships (microMAGNETOFLUIDICS, No. 702128). Conflict of Interest The authors declare no conflict of interest. Keywords externally powered, fabrication, micromachines, microrobots, propulsion 1705061 (19 of 22) Received: September 5, 2017 Revised: November 3, 2017 Published online: © 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.advancedsciencenews.com www.advmat.de [1] B. J. Nelson, I. K. Kaliakatsos, J. J. Abbott, Annu. Rev. Biomed. Eng. 2010, 12, 55. [2] J. Li, B. Esteban-Fernández de Ávila, W. Gao, L. Zhang, J. Wang, Sci. Rob. 2017, 2, eaam6431. [3] M. Medina-Sanchez, O. G. Schmidt, Nature 2017, 545, 406. [4] M. Sitti, H. Ceylan, W. Hu, J. Giltinan, M. Turan, S. Yim, E. Diller, Proc. IEEE 2015, 103, 205. [5] A. Servant, F. Qiu, M. Mazza, K. Kostarelos, B. J. Nelson, Adv. Mater. 2015, 27, 2981. [6] W. Gao, R. Dong, S. Thamphiwatana, J. Li, W. Gao, L. Zhang, J. Wang, ACS Nano 2015, 9, 117. [7] J. G. S. Moo, M. Pumera, Chem. Eur. J. 2015, 21, 58. [8] W. Gao, J. Wang, ACS Nano 2014, 8, 3170. [9] J. J. Abbott, K. E. Peyer, M. C. Lagomarsino, L. Zhang, L. Dong, I. K. Kaliakatsos, B. J. Nelson, Int. J. Rob. Res. 2009, 28, 1434. [10] K. E. Peyer, L. Zhang, B. J. Nelson, Nanoscale 2013, 5, 1259. [11] J. Parmar, X. Ma, J. Katuri, J. Simmchen, M. M. Stanton, C. TrichetParedes, L. Soler, S. Sanchez, Sci. Technol. Adv. Mater. 2015, 16, 014802. [12] S. Sanchez, L. Soler, J. Katuri, Angew. Chem., Int. Ed. 2015, 54, 1414. [13] H. Wang, M. Pumera, Chem. Rev. 2015, 115, 8704. [14] W. Wang, W. Duan, S. Ahmed, A. Sen, T. E. Mallouk, Acc. Chem. Res. 2015, 48, 1938. [15] J. Katuri, X. Ma, M. M. Stanton, S. Sánchez, Acc. Chem. Res. 2017, 50, 2. [16] J. G. S. Moo, C. C. Mayorga-Martinez, H. Wang, B. Khezri, W. Z. Teo, M. Pumera, Adv. Funct. Mater. 2017, 27, 1604759. [17] J. R. Baylis, J. H. Yeon, M. H. Thomson, A. Kazerooni, X. Wang, A. E. St John, E. B. Lim, D. Chien, A. Lee, J. Q. Zhang, J. M. Piret, L. S. Machan, T. F. Burke, N. J. White, C. J. Kastrup, Sci. Adv. 2015, 1, e1500379. [18] T. Xu, W. Gao, L. P. Xu, X. Zhang, S. Wang, Adv. Mater. 2017, 29, 1603250. [19] X.-Z. Chen, M. Hoop, F. Mushtaq, E. Siringil, C. Hu, B. J. Nelson, S. Pané, Appl. Mater. Today 2017, 9, 37. [20] M. Medina-Sanchez, L. Schwarz, A. K. Meyer, F. Hebenstreit, O. G. Schmidt, Nano Lett. 2016, 16, 555. [21] S. Nain, N. N. Sharma, Front. Life Sci. 2014, 8, 2. [22] L. Zheng, L.-G. Chen, H.-B. Huang, X.-P. Li, L.-L. Zhang, Microsyst. Technol. 2016, 22, 2371. [23] H. Ceylan, J. Giltinan, K. Kozielski, M. Sitti, Lab Chip 2017, 17, 1705. [24] D. J. Bell, S. Leutenegger, K. M. Hammar, L. X. Dong, B. J. Nelson, in Proc. 2007 IEEE Int. Conf. Rob. Autom. IEEE, Rome, Italy 2007, p. 1128. [25] M. M. Hawkeye, M. J. Brett, J. Vac. Sci. Technol., A 2007, 25, 1317. [26] S. Tottori, L. Zhang, F. Qiu, K. K. Krawczyk, A. Franco-Obregon, B. J. Nelson, Adv. Mater. 2012, 24, 811. [27] S. Kim, F. Qiu, S. Kim, A. Ghanbari, C. Moon, L. Zhang, B. J. Nelson, H. Choi, Adv. Mater. 2013, 25, 5863. [28] F. Qiu, R. Mhanna, L. Zhang, Y. Ding, S. Fujita, B. J. Nelson, Sens. Actuators, B 2014, 196, 676. [29] F. Qiu, L. Zhang, K. E. Peyer, M. Casarosa, A. Franco-Obregón, H. Choi, B. J. Nelson, J. Mater. Chem. B 2014, 2, 357. [30] F. Qiu, S. Fujita, R. Mhanna, L. Zhang, B. R. Simona, B. J. Nelson, Adv. Funct. Mater. 2015, 25, 1666. [31] S. Tottori, B. J. Nelson, Biomicrofluidics 2013, 7, 061101. [32] T.-Y. Huang, F. Qiu, H.-W. Tung, X.-B. Chen, B. J. Nelson, M. S. Sakar, Appl. Phys. Lett. 2014, 105, 114102. [33] T.-Y. Huang, F. Qiu, H.-W. Tung, K. E. Peyer, N. Shamsudhin, J. Pokki, L. Zhang, X.-B. Chen, B. J. Nelson, M. S. Sakar, RSC Adv. 2014, 4, 26771. [34] C. Peters, O. Ergeneman, P. D. W. García, M. Müller, S. Pané, B. J. Nelson, C. Hierold, Adv. Funct. Mater. 2014, 24, 5269. Adv. Mater. 2018, 1705061 [35] M. A. Zeeshan, R. Grisch, E. Pellicer, K. M. Sivaraman, K. E. Peyer, J. Sort, B. Ozkale, M. S. Sakar, B. J. Nelson, S. Pane, Small 2014, 10, 1284. [36] T. Y. Huang, M. S. Sakar, A. Mao, A. J. Petruska, F. Qiu, X. B. Chen, S. Kennedy, D. Mooney, B. J. Nelson, Adv. Mater. 2015, 27, 6644. [37] C. Peters, M. Hoop, S. Pane, B. J. Nelson, C. Hierold, Adv. Mater. 2016, 28, 533. [38] K. E. Peyer, F. Qiu, L. Zhang, B. J. Nelson, in IEEE/RSJ Int. Conf. Intelligent Robots and Systems (IROS), IEEE, Vilamoura, Portugal 2012, p. 2553. [39] L. Zhang, J. J. Abbott, L. Dong, B. E. Kratochvil, D. Bell, B. J. Nelson, Appl. Phys. Lett. 2009, 94, 064107. [40] B. J. Nelson, K. E. Peyer, ACS Nano 2014, 8, 8718. [41] A. G. Mark, J. G. Gibbs, T. C. Lee, P. Fischer, Nat. Mater. 2013, 12, 802. [42] D. Schamel, A. G. Mark, J. G. Gibbs, C. Miksch, K. I. Morozov, A. M. Leshansky, P. Fischer, ACS Nano 2014, 8, 8794. [43] D. Walker, B. T. Kasdorf, H. H. Jeong, O. Lieleg, P. Fischer, Sci. Adv. 2015, 1, e1500501. [44] X. P. Wang, C. Hu, L. Schurz, C. De Macro, X.-Z. Chen, S. Pane, B. J. Nelson, ACS Nano, unpublished. [45] J. G. Gibbs, A. G. Mark, T. C. Lee, S. Eslami, D. Schamel, P. Fischer, Nanoscale 2014, 6, 9457. [46] A. Ghosh, P. Fischer, Nano Lett. 2009, 9, 2243. [47] D. Walker, M. Kubler, K. I. Morozov, P. Fischer, A. M. Leshansky, Nano Lett. 2015, 15, 4412. [48] L. Liu, S. H. Yoo, S. A. Lee, S. Park, Nano Lett. 2011, 11, 3979. [49] J. Li, S. Sattayasamitsathit, R. Dong, W. Gao, R. Tam, X. Feng, S. Ai, J. Wang, Nanoscale 2014, 6, 9415. [50] M. Hoop, Y. Shen, X. Z. Chen, F. Mushtaq, M. S. Sakar, A. Petruska, M. J. Loessner, B. J. Nelson, S. Pané, Adv. Funct. Mater. 2016, 26, 1063. [51] S. Schuerle, S. Pane, E. Pellicer, J. Sort, M. D. Baro, B. J. Nelson, Small 2012, 8, 1498. [52] W. Gao, X. Feng, A. Pei, C. R. Kane, R. Tam, C. Hennessy, J. Wang, Nano Lett. 2014, 14, 305. [53] X. Yan, Q. Zhou, J. Yu, T. Xu, Y. Deng, T. Tang, Q. Feng, L. Bian, Y. Zhang, A. Ferreira, L. Zhang, Adv. Funct. Mater. 2015, 25, 5333. [54] S. Fusco, M. S. Sakar, S. Kennedy, C. Peters, R. Bottani, F. Starsich, A. Mao, G. A. Sotiriou, S. Pane, S. E. Pratsinis, D. Mooney, B. J. Nelson, Adv. Mater. 2014, 26, 952. [55] S. Fusco, H. W. Huang, K. E. Peyer, C. Peters, M. Haberli, A. Ulbers, A. Spyrogianni, E. Pellicer, J. Sort, S. E. Pratsinis, B. J. Nelson, M. S. Sakar, S. Pane, ACS Appl. Mater. Interfaces 2015, 7, 6803. [56] H. W. Huang, M. S. Sakar, A. J. Petruska, S. Pane, B. J. Nelson, Nat. Commun. 2016, 7, 12263. [57] W. Gao, S. Sattayasamitsathit, K. M. Manesh, D. Weihs, J. Wang, J. Am. Chem. Soc. 2010, 132, 14403. [58] X. Z. Chen, N. Shamsudhin, M. Hoop, R. Pieters, E. Siringil, M. S. Sakar, B. J. Nelson, S. Pané, Mater. Horiz. 2016, 3, 113. [59] G. Chatzipirpiridis, E. Avilla, O. Ergeneman, B. J. Nelson, S. Pane, IEEE Trans. Mag. 2014, 50, 5400403. [60] J. Ali, U. K. Cheang, Y. G. Liu, H. Kim, L. Rogowski, S. Sheckman, P. Patel, W. Sun, M. J. Kim, AIP Adv. 2016, 6, 125205. [61] L. Zhang, T. Petit, Y. Lu, B. E. Kratochvil, K. E. Peyer, R. Pei, J. Lou, B. J. Nelson, ACS Nano 2010, 4, 6228. [62] F. Mushtaq, A. Asani, M. Hoop, X.-Z. Chen, D. Ahmed, B. J. Nelson, S. Pané, Adv. Funct. Mater. 2016, 26, 6995. [63] F. Mushtaq, M. Guerrero, M. S. Sakar, M. Hoop, A. Lindo, J. Sort, X.-Z. Chen, B. J. Nelson, E. Pellicer, S. Pané, J. Mater. Chem. A 2015, 3, 23670. 1705061 (20 of 22) © 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.advancedsciencenews.com www.advmat.de [64] Q. Zhou, T. Petit, H. Choi, B. J. Nelson, L. Zhang, Adv. Funct. Mater. 2016, 27, 1604571. [65] X.-Z. Chen, M. Hoop, N. Shamsudhin, T. Huang, B. Özkale, Q. Li, E. Siringil, F. Mushtaq, L. Di Tizio, B. J. Nelson, S. Pané, Adv. Mater. 2017, 29, 1605458. [66] H.-W. Tung, D. F. Sargent, B. J. Nelson, J. Appl. Crystallogr. 2014, 47, 692. [67] R. S. Pieters, H.-W. Tung, D. F. Sargent, B. J. Nelson, IFAC Proc. Vol. 2014, 47, 7480. [68] L. O. Mair, B. A. Evans, A. Nacev, P. Y. Stepanov, R. Hilaman, S. Chowdhury, S. Jafari, W. Wang, B. Shapiro, I. N. Weinberg, Nanoscale 2017, 9, 3375. [69] C. E. Sing, L. Schmid, M. F. Schneider, T. Franke, A. Alexander-Katz, Proc. Natl. Acad. Sci. USA 2010, 107, 535. [70] T. O. Tasci, P. S. Herson, K. B. Neeves, D. W. Marr, Nat. Commun. 2016, 7, 10225. [71] P. Tierno, R. Golestanian, I. Pagonabarraga, F. Sagues, Phys. Rev. Lett. 2008, 101, 218304. [72] P. Tierno, R. Golestanian, I. Pagonabarraga, F. Sagues, J. Phys. Chem. B 2008, 112, 16525. [73] P. J. Vach, N. Brun, M. Bennet, L. Bertinetti, M. Widdrat, J. Baumgartner, S. Klumpp, P. Fratzl, D. Faivre, Nano Lett. 2013, 13, 5373. [74] P. J. Vach, D. Faivre, Sci. Rep. 2015, 5, 9364. [75] P. J. Vach, P. Fratzl, S. Klumpp, D. Faivre, Nano Lett. 2015, 15, 7064. [76] K. I. Morozov, Y. Mirzae, O. Kenneth, A. M. Leshansky, Phys. Rev. Fluids 2017, 2, 044202. [77] R. Dreyfus, J. Baudry, M. L. Roper, M. Fermigier, H. A. Stone, J. Bibette, Nature 2005, 437, 862. [78] B. Jang, E. Gutman, N. Stucki, B. F. Seitz, P. D. Wendel-Garcia, T. Newton, J. Pokki, O. Ergeneman, S. Pane, Y. Or, B. J. Nelson, Nano Lett. 2015, 15, 4829. [79] T. Li, J. Li, H. Zhang, X. Chang, W. Song, Y. Hu, G. Shao, E. Sandraz, G. Zhang, L. Li, J. Wang, Small 2016, 12, 6098. [80] I. S. M. Khalil, H. C. Dijkslag, L. Abelmann, S. Misra, Appl. Phys. Lett. 2014, 104, 223701. [81] H.-W. Tung, M. Maffioli, D. R. Frutiger, K. M. Sivaraman, S. Pane, B. J. Nelson, IEEE Trans. Rob. 2014, 30, 26. [82] S. Kim, S. Lee, J. Lee, B. J. Nelson, L. Zhang, H. Choi, Sci. Rep. 2016, 6, 30713. [83] D. Ahmed, M. Lu, A. Nourhani, P. E. Lammert, Z. Stratton, H. S. Muddana, V. H. Crespi, T. J. Huang, Sci. Rep. 2015, 5, 9744. [84] W. Wang, L. A. Castro, M. Hoyos, T. E. Mallouk, ACS Nano 2012, 6, 6122. [85] S. Ahmed, D. T. Gentekos, C. A. Fink, T. E. Mallouk, ACS Nano 2014, 8, 11053. [86] S. Ahmed, W. Wang, L. O. Mair, R. D. Fraleigh, S. Li, L. A. Castro, M. Hoyos, T. J. Huang, T. E. Mallouk, Langmuir 2013, 29, 16113. [87] V. Garcia-Gradilla, J. Orozco, S. Sattayasamitsathit, F. Soto, F. Kuralay, A. Pourazary, A. Katzenberg, W. Gao, Y. Shen, J. Wang, ACS Nano 2013, 7, 9232. [88] J. Li, T. Li, T. Xu, M. Kiristi, W. Liu, Z. Wu, J. Wang, Nano Lett. 2015, 15, 4814. [89] D. Ahmed, T. Baasch, B. Jang, S. Pané, J. Dual, B. J. Nelson, Nano Lett. 2016, 16, 4968. [90] D. Ahmed, A. Ozcelik, N. Bojanala, N. Nama, A. Upadhyay, Y. Chen, W. Hanna-Rose, T. J. Huang, Nat. Commun. 2016, 7, 11085. [91] W. Wang, S. Li, L. Mair, S. Ahmed, T. J. Huang, T. E. Mallouk, Angew. Chem., Int. Ed. 2014, 53, 3201. [92] B. Esteban-Fernández De Ávila, C. Angell, F. Soto, M. A. Lopez-Ramirez, D. F. Báez, S. Xie, J. Wang, Y. Chen, ACS Nano 2016, 10, 4997. Adv. Mater. 2018, 1705061 [93] D. Ahmed, T. Baasch, N. Blondel, N. Laubli, J. Dual, B. J. Nelson, Nat. Commun. 2017, 8, 770. [94] J. Feng, J. Yuan, S. K. Cho, Lab Chip 2015, 15, 1554. [95] D. Ahmed, X. Mao, J. Shi, B. K. Juluri, T. J. Huang, Lab Chip 2009, 9, 2738. [96] D. Ahmed, X. Mao, B. K. Juluri, T. J. Huang, Microfluid. Nanofluid. 2009, 7, 727. [97] N. Bertin, T. A. Spelman, O. Stephan, L. Gredy, M. Bouriau, E. Lauga, P. Marmottant, Phys. Rev. Appl. 2015, 4, 064012. [98] T. Qiu, S. Palagi, A. G. Mark, K. Melde, F. Adams, P. Fischer, Appl. Phys. Lett. 2016, 109, 191602. [99] T. G. Leighton, The Acoustic Bubble, Academic, London 1994. [100] D. Ahmed, C. Dillinger, A. Hong, B. J. Nelson, Adv. Mater. Technol. 2017, 2, 1700050. [101] C. J. Weijer, J. Cell. Sci. 2009, 122, 3215. [102] H. R. Jiang, N. Yoshinaga, M. Sano, Phys. Rev. Lett. 2010, 105, 268302. [103] M. Xuan, Z. Wu, J. Shao, L. Dai, T. Si, Q. He, J. Am. Chem. Soc. 2016, 138, 6492. [104] M. Ni, M. K. H. Leung, D. Y. C. Leung, K. Sumathy, Renewable Sustainable Energy Rev. 2007, 11, 401. [105] M. Pumera, Nanoscale 2010, 2, 1643. [106] F. Mou, L. Kong, C. Chen, Z. Chen, L. Xu, J. Guan, Nanoscale 2016, 8, 4976. [107] R. Dong, Q. Zhang, W. Gao, A. Pei, B. Ren, ACS Nano 2016, 10, 839. [108] Y. Wu, R. Dong, Q. Zhang, B. Ren, Nano-Micro Lett. 2017, 9, 30. [109] Q. Zhang, R. Dong, Y. Wu, W. Gao, Z. He, B. Ren, ACS Appl. Mater. Interfaces 2017, 9, 4674. [110] F. Mou, Y. Li, C. Chen, W. Li, Y. Yin, H. Ma, J. Guan, Small 2015, 11, 2564. [111] M. Enachi, M. Guix, V. Postolache, V. Ciobanu, V. M. Fomin, O. G. Schmidt, I. Tiginyanu, Small 2016, 12, 5497. [112] A. A. Solovev, E. J. Smith, C. C. Bof’ Bufon, S. Sanchez, O. G. Schmidt, Angew. Chem., Int. Ed. 2011, 50, 10875. [113] B. Dai, J. Wang, Z. Xiong, X. Zhan, W. Dai, C. C. Li, S. P. Feng, J. Tang, Nat. Nanotechnol. 2016, 11, 1087. [114] C. Chen, F. Mou, L. Xu, S. Wang, J. Guan, Z. Feng, Q. Wang, L. Kong, W. Li, J. Wang, Q. Zhang, Adv. Mater. 2017, 29, 1603374. [115] R. Dong, Y. Hu, Y. Wu, W. Gao, B. Ren, Q. Wang, Y. Cai, J. Am. Chem. Soc. 2017, 139, 1722. [116] D. Zhou, Y. C. Li, P. Xu, N. S. McCool, L. Li, W. Wang, T. E. Mallouk, Nanoscale 2017, 9, 75. [117] H. Moyses, J. Palacci, S. Sacanna, D. G. Grier, Soft Matter 2016, 12, 6357. [118] J. Palacci, S. Sacanna, A. Abramian, J. Barral, K. Hanson, A. Y. Grosberg, D. J. Pine, P. M. Chaikin, Sci. Adv. 2015, 1, e1400214. [119] J. Palacci, S. Sacanna, S. H. Kim, G. R. Yi, D. J. Pine, P. M. Chaikin, Philos. Trans. R. Soc., A 2014, 372, 20130372. [120] J. Palacci, S. Sacanna, A. P. Steinberg, D. J. Pine, P. M. Chaikin, Science 2013, 339, 936. [121] J. Palacci, S. Sacanna, A. Vatchinsky, P. M. Chaikin, D. J. Pine, J. Am. Chem. Soc. 2013, 135, 15978. [122] B. Jang, A. Hong, H. E. Kang, C. Alcantara, S. Charreyron, F. Mushtaq, E. Pellicer, R. Buchel, J. Sort, S. S. Lee, B. J. Nelson, S. Pane, ACS Nano 2017, 11, 6146. [123] M. Ibele, T. E. Mallouk, A. Sen, Angew. Chem., Int. Ed. 2009, 48, 3308. [124] W. Duan, M. Ibele, R. Liu, A. Sen, Eur. Phys. J. E 2012, 35, 77. [125] M. E. Ibele, P. E. Lammert, V. H. Crespi, A. Sen, ACS Nano 2010, 4, 4845. [126] W. Duan, R. Liu, A. Sen, J. Am. Chem. Soc. 2013, 135, 1280. [127] Y. Hong, M. Diaz, U. M. Córdova-Figueroa, A. Sen, Adv. Funct. Mater. 2010, 20, 1568. [128] D. Feldmann, S. R. Maduar, M. Santer, N. Lomadze, O. I. Vinogradova, S. Santer, Sci. Rep. 2016, 6, 36443. 1705061 (21 of 22) © 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.advancedsciencenews.com www.advmat.de [129] V. Yadav, H. Zhang, R. Pavlick, A. Sen, J. Am. Chem. Soc. 2012, 134, 15688. [130] M. J. Esplandiu, A. A. Farniya, A. Bachtold, ACS Nano 2015, 9, 11234. [131] H. Zeng, P. Wasylczyk, C. Parmeggiani, D. Martella, M. Burresi, D. S. Wiersma, Adv. Mater. 2015, 27, 3883. [132] S. Palagi, A. G. Mark, S. Y. Reigh, K. Melde, T. Qiu, H. Zeng, C. Parmeggiani, D. Martella, A. Sanchez-Castillo, N. Kapernaum, F. Giesselmann, D. S. Wiersma, E. Lauga, P. Fischer, Nat. Mater. 2016, 15, 647. [133] M. Rogóż, H. Zeng, C. Xuan, D. S. Wiersma, P. Wasylczyk, Adv. Opt. Mater. 2016, 4, 1689. [134] Y. F. Lee, Y. F. Huang, S. C. Tsai, H. Y. Lai, E. Lee, Langmuir 2016, 32, 13106. [135] R. Liu, F. Wong, W. T. Duan, A. Sen, Adv. Mater. 2013, 25, 6997. [136] R. L. Urbano, A. M. Clyne, Lab Chip 2016, 16, 561. [137] K. J. Freedman, L. M. Otto, A. P. Ivanov, A. Barik, S. H. Oh, J. B. Edel, Nat. Commun. 2016, 7, 10217. [138] P. Calvo-Marzal, S. Sattayasamitsathit, S. Balasubramanian, J. R. Windmiller, C. Dao, J. Wang, Chem. Commun. 2010, 46, 1623. [139] S. T. Chang, V. N. Paunov, D. N. Petsev, O. D. Velev, Nat. Mater. 2007, 6, 235. [140] J. Wang, Faraday Discuss. 2013, 164, 9. [141] S. Park, S. W. Chung, C. A. Mirkin, J. Am. Chem. Soc. 2004, 126, 11772. [142] D. J. Pena, J. K. N. Mbindyo, A. J. Carado, T. E. Mallouk, C. D. Keating, B. Razavi, T. S. Mayer, J. Phys. Chem. B 2002, 106, 7458. [143] G. Loget, A. Kuhn, Lab Chip 2012, 12, 1967. [144] G. Loget, A. Kuhn, Nat. Commun. 2011, 2, 535. [145] M. Sentic, G. Loget, D. Manojlovic, A. Kuhn, N. Sojic, Angew. Chem., Int. Ed. 2012, 51, 11284. [146] E. Y. n. Erdem, C. Yu-Ming, M. Mohebbi, J. W. Suh, G. Kovacs, R. B. Darling, K. F. Böhringer, J. Microelectromech. Syst. 2010, 19, 433. [147] K. Saito, K. Iwata, Y. Ishihara, K. Sugita, M. Takato, F. Uchikoba, Micromachines 2016, 7, 58. [148] K. Saito, K. Sugita, Y. Ishihara, K. Iwata, Y. Asano, Y. Okane, S. Ono, S. Chiba, M. Takato, F. Uchikoba, Artif. Life Rob. 2016, 22, 118. [149] A. A. Darhuber, J. P. Valentino, J. M. Davis, S. M. Troian, S. Wagner, Appl. Phys. Lett. 2003, 82, 657. Adv. Mater. 2018, 1705061 [150] Z. J. Jiao, X. Y. Huang, N. T. Nguyen, J. Micromech. Microeng. 2008, 18, 045027. [151] O. J. Sul, M. R. Falvo, R. M. Taylor, S. Washburn, R. Superfine, Appl. Phys. Lett. 2006, 89, 203512. [152] W. Hu, K. S. Ishii, A. T. Ohta, Appl. Phys. Lett. 2011, 99, 094103. [153] A. Mourran, H. Zhang, R. Vinokur, M. Moller, Adv. Mater. 2017, 29, 1604825. [154] L. Baraban, R. Streubel, D. Makarov, L. Han, D. Karnaushenko, O. G. Schmidt, G. Cuniberti, ACS Nano 2013, 7, 1360. [155] Z. Hosseinidoust, B. Mostaghaci, O. Yasa, B. W. Park, A. V. Singh, M. Sitti, Adv. Drug Delivery Rev. 2016, 106, 27. [156] L. Schwarz, M. Medina-Sánchez, O. G. Schmidt, App. Phys. Rev. 2017, 4, 031301. [157] M. M. Stanton, B. W. Park, A. Miguel-Lopez, X. Ma, M. Sitti, S. Sanchez, Small 2017, 13, https://doi.org/10.1002/ smll.201603679. [158] M. M. Stanton, J. Simmchen, X. Ma, A. Miguel-López, S. Sánchez, Adv. Mater. Interfaces 2016, 3, 1500505. [159] O. Felfoul, M. Mohammadi, S. Taherkhani, D. de Lanauze, Y. Zhong Xu, D. Loghin, S. Essa, S. Jancik, D. Houle, M. Lafleur, L. Gaboury, M. Tabrizian, N. Kaou, M. Atkin, T. Vuong, G. Batist, N. Beauchemin, D. Radzioch, S. Martel, Nat. Nanotechnol. 2016, 11, 941. [160] M. M. Stanton, B. W. Park, D. Vilela, K. Bente, D. Faivre, M. Sitti, S. Sanchez, ACS Nano 2017, 11, 9968. [161] Z. Wu, T. Li, J. Li, W. Gao, T. Xu, C. Christianson, W. Gao, M. Galarnyk, Q. He, L. Zhang, J. Wang, ACS Nano 2014, 8, 12041. [162] V. Magdanz, M. Medina-Sanchez, L. Schwarz, H. Xu, J. Elgeti, O. G. Schmidt, Adv. Mater. 2017, 29, https://doi.org/10.1002/ adma.201606301. [163] A. X. Lu, Y. Liu, H. Oh, A. Gargava, E. Kendall, Z. Nie, D. L. DeVoe, S. R. Raghavan, ACS Appl. Mater. Interfaces 2016, 8, 15676. [164] A. K. Singh, K. K. Dey, A. Chattopadhyay, T. K. Mandal, D. Bandyopadhyay, Nanoscale 2014, 6, 1398. [165] S. Fusco, G. Chatzipirpiridis, K. M. Sivaraman, O. Ergeneman, B. J. Nelson, S. Pane, Adv. Healthcare Mater. 2013, 2, 1037. [166] B. Esteban-Fernandez de Avila, D. E. Ramirez-Herrera, S. Campuzano, P. Angsantikul, L. Zhang, J. Wang, ACS Nano 2017, 11, 5367. [167] A. Ghanbari, M. Bahrami, J. Intell. Rob. Syst. 2011, 63, 399. 1705061 (22 of 22) © 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim